Modeling late-onset Alzheimer's disease neuropathology via direct neuronal reprogramming.

Late-onset Alzheimer's disease (LOAD) is the most common form of Alzheimer's disease (AD). However, modeling sporadic LOAD that endogenously captures hallmark neuronal pathologies such as amyloid-ß (Aß) deposition, tau tangles, and neuronal loss remains an unmet need. We demonstrate that neurons generated by microRNA (miRNA)-based direct reprogramming of fibroblasts from individuals affected by autosomal dominant AD (ADAD) and LOAD in a three-dimensional environment effectively recapitulate key neuropathological features of AD. Reprogrammed LOAD neurons exhibit Aß-dependent neurodegeneration, and treatment with ß- or γ-secretase inhibitors before (but not subsequent to) Aß deposit formation mitigated neuronal death. Moreover inhibiting age-associated retrotransposable elements in LOAD neurons reduced both Aß deposition and neurodegeneration. Our study underscores the efficacy of modeling late-onset neuropathology of LOAD through high-efficiency miRNA-based neuronal reprogramming.

Statistical considerations when estimating time-saving treatment effects in Alzheimer's disease clinical trials.

INTRODUCTION: Estimating treatment effects as time savings in disease progression may be more easily interpretable than assessing the absolute difference or a percentage reduction. In this study, we investigate the statistical considerations of the existing method for estimating time savings and propose alternative complementary methods. METHODS: We propose five alternative methods to estimate the time savings from different perspectives. These methods are applied to simulated clinical trial data that mimic or modify the Clinical Dementia Rating Sum of Boxes progression trajectories observed in the Clarity AD lecanemab trial. RESULTS: Our study demonstrates that the proposed methods can generate more precise estimates by considering two crucial factors: (1) the absolute difference between treatment arms, and (2) the observed progression rate in the treatment arm. DISCUSSION: Quantifying treatment effects as time savings in disease progression offers distinct advantages. To provide comprehensive estimations, it is important to use various methods. HIGHLIGHTS: We explore the statistical considerations of the current method for estimating time savings. We proposed alternative methods that provide time savings estimations based on the observed absolute differences. By using various methods, a more comprehensive estimation of time savings can be achieved.

γ-Secretase activity, clinical features, and biomarkers of autosomal dominant Alzheimer's disease: cross-sectional and longitudinal analysis of the Dominantly Inherited Alzheimer Network observational study (DIAN-OBS).

BACKGROUND: Genetic variants that cause autosomal dominant Alzheimer's disease are highly penetrant but vary substantially regarding age at symptom onset (AAO), rates of cognitive decline, and biomarker changes. Most pathogenic variants that cause autosomal dominant Alzheimer's disease are in presenilin 1 (PSEN1), which encodes the catalytic core of γ-secretase, an enzyme complex that is crucial in production of amyloid ß. We aimed to investigate whether the heterogeneity in AAO and biomarker trajectories in carriers of PSEN1 pathogenic variants could be predicted on the basis of the effects of individual PSEN1 variants on γ-secretase activity and amyloid ß production. METHODS: For this cross-sectional and longitudinal analysis, we used data from participants enrolled in the Dominantly Inherited Alzheimer Network observational study (DIAN-OBS) via the DIAN-OBS data freeze version 15 (data collected between Feb 29, 2008, and June 30, 2020). The data freeze included data from 20 study sites in research institutions, universities, hospitals, and clinics across Europe, North and South America, Asia, and Oceania. We included individuals with PSEN1 pathogenic variants for whom relevant genetic, clinical, imaging, and CSF data were available. PSEN1 pathogenic variants were characterised via genetically modified PSEN1 and PSEN2 double-knockout human embryonic kidney 293T cells and immunoassays for Aß37, Aß38, Aß40, Aß42, and Aß43. A summary measure of γ-secretase activity (γ-secretase composite [GSC]) was calculated for each variant and compared with clinical history-derived AAO using correlation analyses. We used linear mixed-effect models to assess associations between GSC scores and multimodal-biomarker and clinical data from DIAN-OBS. We used separate models to assess associations with Clinical Dementia Rating Sum of Boxes (CDR-SB), Mini-Mental State Examination (MMSE), and Wechsler Memory Scale-Revised (WMS-R) Logical Memory Delayed Recall, [11C]Pittsburgh compound B (PiB)-PET and brain glucose metabolism using [18F] fluorodeoxyglucose (FDG)-PET, CSF Aß42-to-Aß40 ratio (Aß42/40), CSF log10 (phosphorylated tau 181), CSF log10 (phosphorylated tau 217), and MRI-based hippocampal volume. FINDINGS: Data were included from 190 people carrying PSEN1 pathogenic variants, among whom median age was 39·0 years (IQR 32·0 to 48·0) and AAO was 44·5 years (40·6 to 51·4). 109 (57%) of 190 carriers were female and 81 (43%) were male. Lower GSC values (ie, lower γ-secretase activity than wild-type PSEN1) were associated with earlier AAO (r=0·58; p<0·0001). GSC was associated with MMSE (ß=0·08, SE 0·03; p=0·0043), CDR-SB (-0·05, 0·02; p=0·0027), and WMS-R Logical Memory Delayed Recall scores (0·09, 0·02; p=0·0006). Lower GSC values were associated with faster increase in PiB-PET signal (p=0·0054), more rapid decreases in hippocampal volume (4·19, 0·77; p<0·0001), MMSE (0·02, 0·01; p=0·0020), and WMS-R Logical Memory Delayed Recall (0·004, 0·001; p=0·0003). INTERPRETATION: Our findings suggest that clinical heterogeneity in people with autosomal dominant Alzheimer's disease can be at least partly explained by different effects of PSEN1 variants on γ-secretase activity and amyloid ß production. They support targeting γ-secretase as a therapeutic approach and suggest that cell-based models could be used to improve prediction of symptom onset. FUNDING: US National Institute on Aging, Alzheimer's Association, German Center for Neurodegenerative Diseases, Raul Carrea Institute for Neurological Research, Japan Agency for Medical Research and Development, Korea Health Industry Development Institute, South Korean Ministry of Health and Welfare, South Korean Ministry of Science and ICT, and Spanish Institute of Health Carlos III.

Baseline levels and longitudinal changes in plasma Aß42/40 among Black and white individuals.

Blood-based biomarkers of Alzheimer disease (AD) may facilitate testing of historically under-represented groups. The Study of Race to Understand Alzheimer Biomarkers (SORTOUT-AB) is a multi-center longitudinal study to compare AD biomarkers in participants who identify their race as either Black or white. Plasma samples from 324 Black and 1,547 white participants underwent analysis with C2N Diagnostics' PrecivityAD test for Aß42 and Aß40. Compared to white individuals, Black individuals had higher average plasma Aß42/40 levels at baseline, consistent with a lower average level of amyloid pathology. Interestingly, this difference resulted from lower average levels of plasma Aß40 in Black participants. Despite the differences, Black and white individuals had similar longitudinal rates of change in Aß42/40, consistent with a similar rate of amyloid accumulation. Our results agree with multiple recent studies demonstrating a lower prevalence of amyloid pathology in Black individuals, and additionally suggest that amyloid accumulates consistently across both groups.

Unexpected Low Rate of Amyloid-ß Pathology in Multiple Sclerosis Patients.

The life expectancy of people with multiple sclerosis (MS) has increased, yet we have noted that development of a typical Alzheimer disease dementia syndrome is uncommon. We hypothesized that Alzheimer disease pathology is uncommon in MS patients. In 100 MS patients, the rate of amyloid-ß plasma biomarker positivity was approximately half the rate in 300 non-MS controls matched on age, sex, apolipoprotein E proteotype, and cognitive status. Interestingly, most MS patients who did have amyloid-ß pathology had features atypical for MS at diagnosis. These results support that MS is associated with reduced Alzheimer disease risk, and suggest new avenues of research. ANN NEUROL 2024.

Prediction of future cognitive decline among cognitively unimpaired individuals using measures of soluble phosphorylated tau or tau tangle pathology.

Plasma p-tau217 and Tau-PET are strong prognostic biomarkers in Alzheimer's disease (AD), but their relative performance in predicting future cognitive decline among cognitively unimpaired (CU) individuals is unclear. In this head-to-head comparison study including 9 cohorts and 1534 individuals, we found that plasma p-tau217 and medial temporal lobe Tau-PET signal showed similar associations with cognitive decline on a global cognitive composite test (R2 PET=0.32 vs R2 PLASMA=0.32, pdifference=0.812) and with progression to mild cognitive impairment (Hazard ratio[HR]PET=1.56[1.43-1.70] vs HRPLASMA=1.63[1.50-1.77], pdifference=0.627). Combined plasma and PET models were superior to the single biomarker models (R2=0.36, p<0.01). Furthermore, sequential selection using plasma p-tau217 and then Tau-PET reduced the number of participants required for a clinical trial by 94%, compared to a 75% reduction when using plasma p-tau217 alone. We conclude that plasma p-tau217 and Tau-PET showed similar performance for predicting future cognitive decline in CU individuals, and their sequential use (i.e., plasma p-tau217 followed by Tau-PET in a subset with high plasma p-tau217) is useful for screening in clinical trials in preclinical AD.

Benchmarking of a multi-biomarker low-volume panel for Alzheimer's Disease and related dementia research.

Alzheimer's Disease (AD) biomarker measurement is key to aid in the diagnosis and prognosis of the disease. In the research setting, participant recruitment and retention and optimization of sample use, is one of the main challenges that observational studies face. Thus, obtaining accurate established biomarker measurements for stratification and maximizing use of the precious samples is key. Accurate technologies are currently available for established biomarkers, mainly immunoassays and immunoprecipitation liquid chromatography-mass spectrometry (IP-MS), and some of them are already being used in clinical settings. Although some immunoassays- and IP-MS based platforms provide multiplexing for several different coding proteins there is not a current platform that can measure all the stablished and emerging biomarkers in one run. The NUcleic acid Linked Immuno-Sandwich Assay (NULISA™) is a mid-throughput platform with antibody-based measurements with a sequencing output that requires 15µL of sample volume to measure more than 100 analytes, including those typically assayed for AD. Here we benchmarked and compared the AD-relevant biomarkers including in the NULISA against validated assays, in both CSF and plasma. Overall, we have found that CSF measures of Aß42/40, NfL, GFAP, and p-tau217 are highly correlated and have similar predictive performance when measured by immunoassay, mass-spectrometry or NULISA. In plasma, p-tau217 shows a performance similar to that reported with other technologies when predicting amyloidosis. Other established and exploratory biomarkers (total tau, p-tau181, NRGN, YKL40, sTREM2, VILIP1 among other) show a wide range of correlation values depending on the fluid and the platform. Our results indicate that the multiplexed immunoassay platform produces reliable results for established biomarkers in CSF that are useful in research settings, with the advantage of measuring additional novel biomarkers using minimal sample volume.

Genetic and multi-omic resources for Alzheimer disease and related dementia from the Knight Alzheimer Disease Research Center.

The Knight-Alzheimer Disease Research Center (Knight-ADRC) at Washington University in St. Louis has pioneered and led worldwide seminal studies that have expanded our clinical, social, pathological, and molecular understanding of Alzheimer Disease. Over more than 40 years, research volunteers have been recruited to participate in cognitive, neuropsychologic, imaging, fluid biomarkers, genomic and multi-omic studies. Tissue and longitudinal data collected to foster, facilitate, and support research on dementia and aging. The Genetics and high throughput -omics core (GHTO) have collected of more than 26,000 biological samples from 6,625 Knight-ADRC participants. Samples available include longitudinal DNA, RNA, non-fasted plasma, cerebrospinal fluid pellets, and peripheral blood mononuclear cells. The GHTO has performed deep molecular profiling (genomic, transcriptomic, epigenomic, proteomic, and metabolomic) from large number of brain (n = 2,117), CSF (n = 2,012) and blood/plasma (n = 8,265) samples with the goal of identifying novel risk and protective variants, identify novel molecular biomarkers and causal and druggable targets. Overall, the resources available at GHTO support the increase of our understanding of Alzheimer Disease.

A Comprehensive Head-to-Head Comparison of Key Plasma Phosphorylated Tau 217 Biomarker Tests.

Plasma phosphorylated-tau 217 (p-tau217) is currently the most promising biomarkers for reliable detection of Alzheimer's disease (AD) pathology. Various p-tau217 assays have been developed, but their relative performance is unclear. We compared key plasma p-tau217 tests using cross-sectional and longitudinal measures of amyloid-ß (Aß)-PET, tau-PET, and cognition as outcomes, and benchmarked them against cerebrospinal fluid (CSF) biomarker tests. Samples from 998 individuals (mean[range] age 68.5[20.0-92.5], 53% female) from the Swedish BioFINDER-2 cohort were analyzed. Plasma p-tau217 was measured with mass spectrometry (MS) assays (the ratio between phosphorylated and non-phosphorylated [%p-tau217WashU]and ptau217WashU) as well as with immunoassays (p-tau217Lilly, p-tau217Janssen, p-tau217ALZpath). CSF biomarkers included p-tau217Lilly, and the FDA-approved p-tau181/Aß42Elecsys and p-tau181Elecsys. All plasma p-tau217 tests exhibited high ability to detect abnormal Aß-PET (AUC range: 0.91-0.96) and tau-PET (AUC range: 0.94-0.97). Plasma %p-tau217WashU had the highest performance, with significantly higher AUCs than all the immunoassays (P diff<0.007). For detecting Aß-PET status, %p-tau217WashU had an accuracy of 0.93 (immunoassays: 0.83-0.88), sensitivity of 91% (immunoassays: 84-87%), and a specificity of 94% (immunoassays: 85-89%). Among immunoassays, p-tau217Lilly and plasma p-tau217ALZpath had higher AUCs than plasma p-tau217Janssen for Aß-PET status (P diff<0.006), and p-tau217Lilly outperformed plasma p-tau217ALZpath for tau-PET status (P diff=0.025). Plasma %p-tau217WashU exhibited higher associations with all PET load outcomes compared to immunoassays; baseline Aß-PET load (R2: 0.72; immunoassays: 0.47-0.58; Pdiff<0.001), baseline tau-PET load (R2: 0.51; immunoassays: 0.38-0.45; Pdiff<0.001), longitudinal Aß-PET load (R2: 0.53; immunoassays: 0.31-0.38; Pdiff<0.001) and longitudinal tau-PET load (R2: 0.50; immunoassays: 0.35-0.43; Pdiff<0.014). Among immunoassays, plasma p-tau217Lilly was more strongly associated with Aß-PET load than plasma p-tau217Janssen (P diff<0.020) and with tau-PET load than both plasma p-tau217Janssen and plasma p-tau217ALZpath (all P diff<0.010). Plasma %p-tau217 also correlated more strongly with baseline cognition (Mini-Mental State Examination[MMSE]) than all immunoassays (R2 %p-tau217WashU: 0.33; immunoassays: 0.27-0.30; P diff<0.024). The main results were replicated in an external cohort from Washington University in St Louis (n =219). Finally, p-tau217Nulisa showed similar performance to other immunoassays in subsets of both cohorts. In summary, both MS- and immunoassay-based p-tau217 tests generally perform well in identifying Aß-PET, tau-PET, and cognitive abnormalities, but %p-tau217WashU performed significantly better than all the examined immunoassays. Plasma %p-tau217 may be considered as a stand-alone confirmatory test for AD pathology, while some immunoassays might be better suited as triage tests where positive results are confirmed with a second test.

Plasma Phosphorylated Tau 217 and Aß42/40 to Predict Early Brain Aß Accumulation in People Without Cognitive Impairment.

Importance: Phase 3 trials of successful antiamyloid therapies in Alzheimer disease (AD) have demonstrated improved clinical efficacy in people with less severe disease. Plasma biomarkers will be essential for efficient screening of participants in future primary prevention clinical trials testing antiamyloid therapies in cognitively unimpaired (CU) individuals with initially low brain ß-amyloid (Aß) levels who are at high risk of accumulating Aß. Objective: To investigate if combining plasma biomarkers could be useful in predicting subsequent development of Aß pathology in CU individuals with subthreshold brain Aß levels (defined as Aß levels <40 Centiloids) at baseline. Design, Setting, and Participants: This was a longitudinal study including Swedish BioFINDER-2 (enrollment 2017-2022) and replication in 2 independent cohorts, the Knight Alzheimer Disease Research Center (Knight ADRC; enrollment 1988 and 2019) and Swedish BioFINDER-1 (enrollment 2009-2015). Included for analysis was a convenience sample of CU individuals with baseline plasma phosphorylated tau 217 (p-tau217) and Aß42/40 assessments and Aß assessments with positron emission tomography (Aß-PET) or cerebrospinal fluid (CSF) Aß42/40. Data were analyzed between April 2023 and May 2024. Exposures: Baseline plasma levels of Aß42/40, p-tau217, the ratio of p-tau217 to nonphosphorylated tau (%p-tau217), p-tau231, and glial fibrillary acidic protein (GFAP). Main Outcomes and Measures: Cross-sectional and longitudinal PET and CSF measures of brain Aß pathology. Results: This study included 495 (BioFINDER-2), 283 (Knight ADRC), and 205 (BioFINDER-1) CU participants. In BioFINDER-2, the mean (SD) age was 65.7 (14.4) with 261 females (52.7%). When detecting abnormal CSF Aß-status, a combination of plasma %p-tau217 and Aß42/40 showed better performance (area under the curve = 0.949; 95% CI, 0.929-0.970; P <.02) than individual biomarkers. In CU participants with subthreshold baseline Aß-PET, baseline plasma %p-tau217 and Aß42/40 levels were significantly associated with baseline Aß-PET (n = 384) and increases in Aß-PET over time (n = 224). Associations of plasma %p-tau217 and Aß42/40 and their interaction with baseline Aß-PET (%p-tau217: ß = 2.77; 95% CI, 1.84-3.70; Aß42/40: ß = -1.64; 95% CI, -2.53 to -0.75; %p-tau217 × Aß42/40: ß = -2.14; 95% CI, -2.79 to -1.49; P < .001) and longitudinal Aß-PET (%p-tau217: ß = 0.67; 95% CI, 0.48-0.87; Aß42/40: ß = -0.33; 95% CI, -0.51 to -0.15; %p-tau217 × Aß42/40: ß = -0.31; 95% CI, -0.44 to -0.18; P < .001) were also significant in the models combining the 2 baseline biomarkers as predictors. Similarly, baseline plasma p-tau217 and Aß42/40 were independently associated with longitudinal Aß-PET in Knight ADRC (%p-tau217: ß = 0.71; 95% CI, 0.26-1.16; P = .002; Aß42/40: ß = -0.74; 95% CI, -1.26 to -0.22; P = .006) and longitudinal CSF Aß42/40 in BioFINDER-1 (p-tau217: ß = -0.0003; 95% CI, -0.0004 to -0.0001; P = .01; Aß42/40: ß = 0.0004; 95% CI, 0.0002-0.0006; P < .001) in CU participants with subthreshold Aß levels at baseline. Plasma p-tau231 and GFAP did not provide any clear independent value. Conclusions and Relevance: Results of this cohort study suggest that combining plasma p-tau217and Aß42/40 levels could be useful for predicting development of Aß pathology in people with early stages of subthreshold Aß accumulation. These biomarkers might thus facilitate screening of participants for future primary prevention trials.

A genetic and proteomic comparison of key AD biomarkers across tissues.

INTRODUCTION: Plasma has been proposed as an alternative to cerebrospinal fluid (CSF) for measuring Alzheimer's disease (AD) biomarkers, but no studies have analyzed in detail which biofluid is more informative for genetics studies of AD. METHOD: Eleven proteins associated with AD (α-synuclein, apolipoprotein E [apoE], CLU, GFAP, GRN, NfL, NRGN, SNAP-25, TREM2, VILIP-1, YKL-40) were assessed in plasma (n = 2317) and CSF (n = 3107). Both plasma and CSF genome-wide association study (GWAS) analyses were performed for each protein, followed by functional annotation. Additional characterization for each biomarker included calculation of correlations and predictive power. RESULTS: Eighteen plasma protein quantitative train loci (pQTLs) associated with 10 proteins and 16 CSF pQTLs associated with 9 proteins were identified. Plasma and CSF shared some genetic loci, but protein levels between tissues correlated weakly. CSF protein levels better associated with AD compared to plasma. DISCUSSION: The present results indicate that CSF is more informative than plasma for genetic studies in AD. HIGHLIGHTS: The identification of novel protein quantitative trait loci (pQTLs) in both plasma and cerebrospinal fluid (CSF). Plasma and CSF levels of neurodegeneration-related proteins correlated weakly. CSF is more informative than plasma for genetic studies of Alzheimer's disease (AD). Neurofilament light (NfL), triggering receptor expressed on myeloid cells 2 (TREM2), and chitinase-3-like protein 1 (YKL-40) tend to show relatively strong inter-tissue associations. A novel signal in the apolipoprotein E (APOE) region was identified, which is an eQTL for APOC1.

A multicenter, randomized, double-blind, placebo-controlled ascending dose study to evaluate the safety, tolerability, pharmacokinetics (PK) and pharmacodynamic (PD) effects of Posiphen in subjects with early Alzheimer's Disease.

BACKGROUND: Amyloid beta protein (Aß) is a treatment target in Alzheimer's Disease (AD). Lowering production of its parent protein, APP, has benefits in preclinical models. Posiphen, an orally administered small molecule, binds to an iron-responsive element in APP mRNA and decreases translation of APP and Aß. To augment human data for Posiphen, we evaluated safety, tolerability and pharmacokinetic and pharmacodynamic (PD) effects on Aß metabolism using Stable Isotope Labeling Kinetic (SILK) analysis. METHODS: Double-blind phase 1b randomized ascending dose clinical trial, at five sites, under an IRB-approved protocol. Participants with mild cognitive impairment or mild AD (Early AD) confirmed by low CSF Aß42/40 were randomized (within each dose arm) to Posiphen or placebo. Pretreatment assessment included lumbar puncture for CSF. Participants took Posiphen or placebo for 21-23 days, then underwent CSF catheter placement, intravenous infusion of 13C6-leucine, and CSF sampling for 36 h. Safety and tolerability were assessed through participant reports, EKG and laboratory tests. CSF SILK analysis measured Aß40, 38 and 42 with immunoprecipitation-mass spectrometry. Baseline and day 21 CSF APP, Aß and other biomarkers were measured with immunoassays. The Mini-Mental State Exam and ADAS-cog12 were given at baseline and day 21. RESULTS: From June 2017 to December 2021, 19 participants were enrolled, randomized within dose cohorts (5 active: 3 placebo) of 60 mg once/day and 60 mg twice/day; 1 participant was enrolled and completed 60 mg three times/day. 10 active drug and 5 placebo participants completed all study procedures. Posiphen was safe and well-tolerated. 8 participants had headaches related to CSF catheterization; 5 needed blood patches. Prespecified SILK analyses of Fractional Synthesis Rate (FSR) for CSF Aß40 showed no significant overall or dose-dependent effects of Posiphen vs. placebo. Comprehensive multiparameter modeling of APP kinetics supported dose-dependent lowering of APP production by Posiphen. Cognitive measures and CSF biomarkers did not change significantly from baseline to 21 days in Posiphen vs. placebo groups. CONCLUSIONS: Posiphen was safe and well-tolerated in Early AD. A multicenter SILK study was feasible. Findings are limited by small sample size but provide additional supportive safety and PK data. Comprehensive modeling of biomarker dynamics using SILK data may reveal subtle drug effects. TRIAL REGISTRATION: NCT02925650 on clinicaltrials.gov (registered on 10-24-2016).

Predicting continuous amyloid PET values with CSF tau phosphorylation occupancies.

INTRODUCTION: Cerebrospinal fluid (CSF) tau phosphorylation at multiple sites is associated with cortical amyloid and other pathologic changes in Alzheimer's disease. These relationships can be non-linear. We used an artificial neural network to assess the ability of 10 different CSF tau phosphorylation sites to predict continuous amyloid positron emission tomography (PET) values. METHODS: CSF tau phosphorylation occupancies at 10 sites (including pT181/T181, pT217/T217, pT231/T231 and pT205/T205) were measured by mass spectrometry in 346 individuals (57 cognitively impaired, 289 cognitively unimpaired). We generated synthetic amyloid PET scans using biomarkers and evaluated their performance. RESULTS: Concentration of CSF pT217/T217 had low predictive error (average error: 13%), but also a low predictive range (ceiling 63 Centiloids). CSF pT231/T231 has slightly higher error (average error: 19%) but predicted through a greater range (87 Centiloids). DISCUSSION: Tradeoffs exist in biomarker selection. Some phosphorylation sites offer greater concordance with amyloid PET at lower levels, while others perform better over a greater range. HIGHLIGHTS: Novel pTau isoforms can predict cortical amyloid burden. pT217/T217 accurately predicts cortical amyloid burden in low-amyloid individuals. Traditional CSF biomarkers correspond with higher levels of amyloid.

AmyloidPETNet: Classification of Amyloid Positivity in Brain PET Imaging Using End-to-End Deep Learning.

Background Visual assessment of amyloid PET scans relies on the availability of radiologist expertise, whereas quantification of amyloid burden typically involves MRI for processing and analysis, which can be computationally expensive. Purpose To develop a deep learning model to classify minimally processed brain PET scans as amyloid positive or negative, evaluate its performance on independent data sets and different tracers, and compare it with human visual reads. Materials and Methods This retrospective study used 8476 PET scans (6722 patients) obtained from late 2004 to early 2023 that were analyzed across five different data sets. A deep learning model, AmyloidPETNet, was trained on 1538 scans from 766 patients, validated on 205 scans from 95 patients, and internally tested on 184 scans from 95 patients in the Alzheimer's Disease Neuroimaging Initiative (ADNI) fluorine 18 (18F) florbetapir (FBP) data set. It was tested on ADNI scans using different tracers and scans from independent data sets. Scan amyloid positivity was based on mean cortical standardized uptake value ratio cutoffs. To compare with model performance, each scan from both the Centiloid Project and a subset of the Anti-Amyloid Treatment in Asymptomatic Alzheimer's Disease (A4) study were visually interpreted with a confidence level (low, intermediate, high) of amyloid positivity/negativity. The area under the receiver operating characteristic curve (AUC) and other performance metrics were calculated, and Cohen κ was used to measure physician-model agreement. Results The model achieved an AUC of 0.97 (95% CI: 0.95, 0.99) on test ADNI 18F-FBP scans, which generalized well to 18F-FBP scans from the Open Access Series of Imaging Studies (AUC, 0.95; 95% CI: 0.93, 0.97) and the A4 study (AUC, 0.98; 95% CI: 0.98, 0.98). Model performance was high when applied to data sets with different tracers (AUC ≥ 0.97). Other performance metrics provided converging evidence. Physician-model agreement ranged from fair (Cohen κ = 0.39; 95% CI: 0.16, 0.60) on a sample of mostly equivocal cases from the A4 study to almost perfect (Cohen κ = 0.93; 95% CI: 0.86, 1.0) on the Centiloid Project. Conclusion The developed model was capable of automatically and accurately classifying brain PET scans as amyloid positive or negative without relying on experienced readers or requiring structural MRI. Clinical trial registration no. NCT00106899 © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Bryan and Forghani in this issue.

An interpretable machine learning-based cerebrospinal fluid proteomics clock for predicting age reveals novel insights into brain aging.

Machine learning can be used to create "biologic clocks" that predict age. However, organs, tissues, and biofluids may age at different rates from the organism as a whole. We sought to understand how cerebrospinal fluid (CSF) changes with age to inform the development of brain aging-related disease mechanisms and identify potential anti-aging therapeutic targets. Several epigenetic clocks exist based on plasma and neuronal tissues; however, plasma may not reflect brain aging specifically and tissue-based clocks require samples that are difficult to obtain from living participants. To address these problems, we developed a machine learning clock that uses CSF proteomics to predict the chronological age of individuals with a 0.79 Pearson correlation and mean estimated error (MAE) of 4.30 years in our validation cohort. Additionally, we analyzed proteins highly weighted by the algorithm to gain insights into changes in CSF and uncover novel insights into brain aging. We also demonstrate a novel method to create a minimal protein clock that uses just 109 protein features from the original clock to achieve a similar accuracy (0.75 correlation, MAE 5.41). Finally, we demonstrate that our clock identifies novel proteins that are highly predictive of age in interactions with other proteins, but do not directly correlate with chronological age themselves. In conclusion, we propose that our CSF protein aging clock can identify novel proteins that influence the rate of aging of the central nervous system (CNS), in a manner that would not be identifiable by examining their individual relationships with age.

Molecular neuroimaging in dominantly inherited versus sporadic early-onset Alzheimer's disease.

Approximately 5% of Alzheimer's disease patients develop symptoms before age 65 (early-onset Alzheimer's disease), with either sporadic (sporadic early-onset Alzheimer's disease) or dominantly inherited (dominantly inherited Alzheimer's disease) presentations. Both sporadic early-onset Alzheimer's disease and dominantly inherited Alzheimer's disease are characterized by brain amyloid-ß accumulation, tau tangles, hypometabolism and neurodegeneration, but differences in topography and magnitude of these pathological changes are not fully elucidated. In this study, we directly compared patterns of amyloid-ß plaque deposition and glucose hypometabolism in sporadic early-onset Alzheimer's disease and dominantly inherited Alzheimer's disease individuals. Our analysis included 134 symptomatic sporadic early-onset Alzheimer's disease amyloid-Positron Emission Tomography (PET)-positive cases from the University of California, San Francisco, Alzheimer's Disease Research Center (mean ± SD age 59.7 ± 5.6 years), 89 symptomatic dominantly inherited Alzheimer's disease cases (age 45.8 ± 9.3 years) and 102 cognitively unimpaired non-mutation carriers from the Dominantly Inherited Alzheimer Network study (age 44.9 ± 9.2). Each group underwent clinical and cognitive examinations, 11C-labelled Pittsburgh Compound B-PET and structural MRI. 18F-Fluorodeoxyglucose-PET was also available for most participants. Positron Emission Tomography scans from both studies were uniformly processed to obtain a standardized uptake value ratio (PIB50-70 cerebellar grey reference and FDG30-60 pons reference) images. Statistical analyses included pairwise global and voxelwise group comparisons and group-independent component analyses. Analyses were performed also adjusting for covariates including age, sex, Mini-Mental State Examination, apolipoprotein ε4 status and average composite cortical of standardized uptake value ratio. Compared with dominantly inherited Alzheimer's disease, sporadic early-onset Alzheimer's disease participants were older at age of onset (mean ± SD, 54.8 ± 8.2 versus 41.9 ± 8.2, Cohen's d = 1.91), with more years of education (16.4 ± 2.8 versus 13.5 ± 3.2, d = 1) and more likely to be apolipoprotein ε4 carriers (54.6% ε4 versus 28.1%, Cramer's V = 0.26), but similar Mini-Mental State Examination (20.6 ± 6.1 versus 21.2 ± 7.4, d = 0.08). Sporadic early-onset Alzheimer's disease had higher global cortical Pittsburgh Compound B-PET binding (mean ± SD standardized uptake value ratio, 1.92 ± 0.29 versus 1.58 ± 0.44, d = 0.96) and greater global cortical 18F-fluorodeoxyglucose-PET hypometabolism (mean ± SD standardized uptake value ratio, 1.32 ± 0.1 versus 1.39 ± 0.19, d = 0.48) compared with dominantly inherited Alzheimer's disease. Fully adjusted comparisons demonstrated relatively higher Pittsburgh Compound B-PET standardized uptake value ratio in the medial occipital, thalami, basal ganglia and medial/dorsal frontal regions in dominantly inherited Alzheimer's disease versus sporadic early-onset Alzheimer's disease. Sporadic early-onset Alzheimer's disease showed relatively greater 18F-fluorodeoxyglucose-PET hypometabolism in Alzheimer's disease signature temporoparietal regions and caudate nuclei, whereas dominantly inherited Alzheimer's disease showed relatively greater hypometabolism in frontal white matter and pericentral regions. Independent component analyses largely replicated these findings by highlighting common and unique Pittsburgh Compound B-PET and 18F-fluorodeoxyglucose-PET binding patterns. In summary, our findings suggest both common and distinct patterns of amyloid and glucose hypometabolism in sporadic and dominantly inherited early-onset Alzheimer's disease.

Novel non-linear models for clinical trial analysis with longitudinal data: A tutorial using SAS for both frequentist and Bayesian methods.

Longitudinal data from clinical trials are commonly analyzed using mixed models for repeated measures (MMRM) when the time variable is categorical or linear mixed-effects models (ie, random effects model) when the time variable is continuous. In these models, statistical inference is typically based on the absolute difference in the adjusted mean change (for categorical time) or the rate of change (for continuous time). Previously, we proposed a novel approach: modeling the percentage reduction in disease progression associated with the treatment relative to the placebo decline using proportional models. This concept of proportionality provides an innovative and flexible method for simultaneously modeling different cohorts, multivariate endpoints, and jointly modeling continuous and survival endpoints. Through simulated data, we demonstrate the implementation of these models using SAS procedures in both frequentist and Bayesian approaches. Additionally, we introduce a novel method for implementing MMRM models (ie, analysis of response profile) using the nlmixed procedure.

CSF neurofilament light chain profiling and quantitation in neurological diseases.

Neurofilament light chain is an established marker of neuroaxonal injury that is elevated in CSF and blood across various neurological diseases. It is increasingly used in clinical practice to aid diagnosis and monitor progression and as an outcome measure to assess safety and efficacy of disease-modifying therapies across the clinical translational neuroscience field. Quantitative methods for neurofilament light chain in human biofluids have relied on immunoassays, which have limited capacity to describe the structure of the protein in CSF and how this might vary in different neurodegenerative diseases. In this study, we characterized and quantified neurofilament light chain species in CSF across neurodegenerative and neuroinflammatory diseases and healthy controls using targeted mass spectrometry. We show that the quantitative immunoprecipitation-tandem mass spectrometry method developed in this study strongly correlates to single-molecule array measurements in CSF across the broad spectrum of neurodegenerative diseases and was replicable across mass spectrometry methods and centres. In summary, we have created an accurate and cost-effective assay for measuring a key biomarker in translational neuroscience research and clinical practice, which can be easily multiplexed and translated into clinical laboratories for the screening and monitoring of neurodegenerative disease or acute brain injury.

Comparison of tau spread in people with Down syndrome versus autosomal-dominant Alzheimer's disease: a cross-sectional study.

BACKGROUND: In people with genetic forms of Alzheimer's disease, such as in Down syndrome and autosomal-dominant Alzheimer's disease, pathological changes specific to Alzheimer's disease (ie, accumulation of amyloid and tau) occur in the brain at a young age, when comorbidities related to ageing are not present. Studies including these cohorts could, therefore, improve our understanding of the early pathogenesis of Alzheimer's disease and be useful when designing preventive interventions targeted at disease pathology or when planning clinical trials. We compared the magnitude, spatial extent, and temporal ordering of tau spread in people with Down syndrome and autosomal-dominant Alzheimer's disease. METHODS: In this cross-sectional observational study, we included participants (aged ≥25 years) from two cohort studies. First, we collected data from the Dominantly Inherited Alzheimer's Network studies (DIAN-OBS and DIAN-TU), which include carriers of autosomal-dominant Alzheimer's disease genetic mutations and non-carrier familial controls recruited in Australia, Europe, and the USA between 2008 and 2022. Second, we collected data from the Alzheimer Biomarkers Consortium-Down Syndrome study, which includes people with Down syndrome and sibling controls recruited from the UK and USA between 2015 and 2021. Controls from the two studies were combined into a single group of familial controls. All participants had completed structural MRI and tau PET (18F-flortaucipir) imaging. We applied Gaussian mixture modelling to identify regions of high tau PET burden and regions with the earliest changes in tau binding for each cohort separately. We estimated regional tau PET burden as a function of cortical amyloid burden for both cohorts. Finally, we compared the temporal pattern of tau PET burden relative to that of amyloid. FINDINGS: We included 137 people with Down syndrome (mean age 38·5 years [SD 8·2], 74 [54%] male, and 63 [46%] female), 49 individuals with autosomal-dominant Alzheimer's disease (mean age 43·9 years [11·2], 22 [45%] male, and 27 [55%] female), and 85 familial controls, pooled from across both studies (mean age 41·5 years [12·1], 28 [33%] male, and 57 [67%] female), who satisfied the PET quality-control procedure for tau-PET imaging processing. 134 (98%) people with Down syndrome, 44 (90%) with autosomal-dominant Alzheimer's disease, and 77 (91%) controls also completed an amyloid PET scan within 3 years of tau PET imaging. Spatially, tau PET burden was observed most frequently in subcortical and medial temporal regions in people with Down syndrome, and within the medial temporal lobe in people with autosomal-dominant Alzheimer's disease. Across the brain, people with Down syndrome had greater concentrations of tau for a given level of amyloid compared with people with autosomal-dominant Alzheimer's disease. Temporally, increases in tau were more strongly associated with increases in amyloid for people with Down syndrome compared with autosomal-dominant Alzheimer's disease. INTERPRETATION: Although the general progression of amyloid followed by tau is similar for people Down syndrome and people with autosomal-dominant Alzheimer's disease, we found subtle differences in the spatial distribution, timing, and magnitude of the tau burden between these two cohorts. These differences might have important implications; differences in the temporal pattern of tau accumulation might influence the timing of drug administration in clinical trials, whereas differences in the spatial pattern and magnitude of tau burden might affect disease progression. FUNDING: None.

A multicenter, randomized, double-blind, placebo-controlled ascending dose study to evaluate the safety, tolerability, pharmacokinetics (PK) and pharmacodynamic (PD) effects of Posiphen in subjects with Early Alzheimer's Disease.

Background: Amyloid beta protein (Aß) is a treatment target in Alzheimer's Disease (AD). Lowering production of its parent protein, APP, has benefits in preclinical models. Posiphen binds to an iron-responsive element in APP mRNA and decreases translation of APP and Aß. To augment human data for Posiphen, we evaluated safety, tolerability and pharmacokinetic and pharmacodynamic (PD) effects on Aß metabolism using Stable Isotope Labeling Kinetic (SILK) analysis. Methods: Double-blind phase 1b randomized ascending dose clinical trial, at five sites, under an IRB-approved protocol. Participants with mild cognitive impairment or mild AD (Early AD) with positive CSF biomarkers were randomized (within each dose arm) to Posiphen or placebo. Pretreatment assessment included lumbar puncture for CSF. Participants took Posiphen or placebo for 21-23 days, then underwent CSF catheter placement, intravenous infusion of 13C6-leucine, and CSF sampling for 36 hours. Safety and tolerability were assessed through participant reports, EKG and laboratory tests. CSF SILK analysis measured Aß40, 38 and 42 with immunoprecipitation-mass spectrometry. Baseline and day 21 CSF APP, Aß and other biomarkers were measured with immunoassays. The Mini-Mental State Exam and ADAS-cog12 were given at baseline and day 21. Results: From June 2017 to December 2021, 19 participants were enrolled, in dose cohorts (6 active: 2 placebo) of 60 mg once/day and 60 mg twice/day; 1 participant was enrolled and completed 60 mg three times/day. 10 active drug and 5 placebo participants completed all study procedures. Posiphen was safe and well-tolerated. 8 participants had headaches related to CSF catheterization; 5 needed blood patches. Prespecified SILK analyses of Fractional Synthesis Rate (FSR) for CSF Aß40 showed no significant overall or dose-dependent effects of Posiphen vs. placebo. Comprehensive multiparameter modeling of APP kinetics supported dose-dependent lowering of APP production by Posiphen. Cognitive measures and CSF biomarkers did not change significantly from baseline to 21 days in Posiphen vs placebo groups. Conclusions: Posiphen was safe and well-tolerated in Early AD. A multicenter SILK study was feasible. Findings are limited by small sample size but provide additional supportive safety and PK data. Comprehensive modeling of biomarker dynamics using SILK data may reveal subtle drug effects. Trial registration: NCT02925650 on clinicaltrials.gov.