Dataset Information

Multi-cohort and longitudinal Bayesian clustering study of stage and subtype in Alzheimer's disease.

ABSTRACT: Understanding Alzheimer's disease (AD) heterogeneity is important for understanding the underlying pathophysiological mechanisms of AD. However, AD atrophy subtypes may reflect different disease stages or biologically distinct subtypes. Here we use longitudinal magnetic resonance imaging data (891 participants with AD dementia, 305 healthy control participants) from four international cohorts, and longitudinal clustering to estimate differential atrophy trajectories from the age of clinical disease onset. Our findings (in amyloid-β positive AD patients) show five distinct longitudinal patterns of atrophy with different demographical and cognitive characteristics. Some previously reported atrophy subtypes may reflect disease stages rather than distinct subtypes. The heterogeneity in atrophy rates and cognitive decline within the five longitudinal atrophy patterns, potentially expresses a complex combination of protective/risk factors and concomitant non-AD pathologies. By alternating between the cross-sectional and longitudinal understanding of AD subtypes these analyses may allow better understanding of disease heterogeneity.

SUBMITTER: Poulakis K

PROVIDER: S-EPMC9355993 | biostudies-literature | 2022 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Multi-cohort and longitudinal Bayesian clustering study of stage and subtype in Alzheimer's disease.

Poulakis Konstantinos K Pereira Joana B JB Muehlboeck J-Sebastian JS Wahlund Lars-Olof LO Smedby Örjan Ö Volpe Giovanni G Masters Colin L CL Ames David D Niimi Yoshiki Y Iwatsubo Takeshi T Ferreira Daniel D Westman Eric E

Nature communications 20220805 1

Understanding Alzheimer's disease (AD) heterogeneity is important for understanding the underlying pathophysiological mechanisms of AD. However, AD atrophy subtypes may reflect different disease stages or biologically distinct subtypes. Here we use longitudinal magnetic resonance imaging data (891 participants with AD dementia, 305 healthy control participants) from four international cohorts, and longitudinal clustering to estimate differential atrophy trajectories from the age of clinical dise ...[more]

PMID: 35931678

Similar Datasets

Project description:IntroductionWith the pressing need to develop treatments that slow or stop the progression of Alzheimer's disease, new tools are needed to reduce clinical trial duration and validate new targets for human therapeutics. Such tools could be derived from neurophysiological measurements of disease.Methods and analysisThe New Therapeutics in Alzheimer's Disease study (NTAD) aims to identify a biomarker set from magneto/electroencephalography that is sensitive to disease and progression over 1 year. The study will recruit 100 people with amyloid-positive mild cognitive impairment or early-stage Alzheimer's disease and 30 healthy controls aged between 50 and 85 years. Measurements of the clinical, cognitive and imaging data (magnetoencephalography, electroencephalography and MRI) of all participants will be taken at baseline. These measurements will be repeated after approximately 1 year on participants with Alzheimer's disease or mild cognitive impairment, and clinical and cognitive assessment of these participants will be repeated again after approximately 2 years. To assess reliability of magneto/electroencephalographic changes, a subset of 30 participants with mild cognitive impairment or early-stage Alzheimer's disease will also undergo repeat magneto/electroencephalography 2 weeks after baseline. Baseline and longitudinal changes in neurophysiology are the primary analyses of interest. Additional outputs will include atrophy and cognitive change and estimated numbers needed to treat each arm of simulated clinical trials of a future disease-modifying therapy.Ethics and data statementThe study has received a favourable opinion from the East of England Cambridge Central Research Ethics Committee (REC reference 18/EE/0042). Results will be disseminated through internal reports, peer-reviewed scientific journals, conference presentations, website publication, submission to regulatory authorities and other publications. Data will be made available via the Dementias Platform UK Data Portal on completion of initial analyses by the NTAD study group.

Project description:New research criteria for preclinical Alzheimer's disease have been proposed, which include stages for cognitively normal individuals with abnormal amyloid markers (stage 1), abnormal amyloid and neuronal injury markers (stage 2), or abnormal amyloid and neuronal injury markers and subtle cognitive changes (stage 3). We aimed to investigate the prevalence and long-term outcome of preclinical Alzheimer's disease according to these criteria.Participants were cognitively normal (clinical dementia rating [CDR]=0) community-dwelling volunteers aged at least 65 years who were enrolled between 1998 and 2011 at the Washington University School of Medicine (MO, USA). CSF amyloid-?1-42 and tau concentrations and a memory composite score were used to classify participants as normal (both markers normal), preclinical Alzheimer's disease stage 1-3, or suspected non-Alzheimer pathophysiology (SNAP, abnormal injury marker without abnormal amyloid marker). The primary outcome was the proportion of participants in each preclinical AD stage. Secondary outcomes included progression to CDR at least 0·5, symptomatic Alzheimer's disease (score of at least 0·5 for memory and at least one other domain and cognitive impairments deemed to be due to Alzheimer's disease), and mortality. We undertook survival analyses using subdistribution and standard Cox hazards models and linear mixed models.Of 311 participants, 129 (41%) were classed as normal, 47 (15%) as stage 1, 36 (12%) as stage 2, 13 (4%) as stage 3, 72 (23%) as SNAP, and 14 (5%) remained unclassified. The 5-year progression rate to CDR at least 0·5, symptomatic Alzheimer's disease was 2% for participants classed as normal, 11% for stage 1, 26% for stage 2, 56% for stage 3, and 5% for SNAP. Compared with individuals classed as normal, participants with preclinical Alzheimer's disease had an increased risk of death after adjusting for covariates (hazard ratio 6·2, 95% CI 1·1-35·0; p=0·040).Preclinical Alzheimer's disease is common in cognitively normal elderly people and is associated with future cognitive decline and mortality. Thus, preclinical Alzheimer's disease could be an important target for therapeutic intervention.National Institute of Aging of the National Institutes of Health (P01-AG003991, P50-AG05681, P01-AG02676), Internationale Stichting Alzheimer Onderzoek, the Center for Translational Molecular Medicine project LeARN, the EU/EFPIA Innovative Medicines Initiative Joint Undertaking, and the Charles and Joanne Knight Alzheimer Research Initiative.

Project description:IntroductionAnticipating the diagnosis of Alzheimer's disease (AD) at an early asymptomatic at-risk stage, where therapeutics can more effectively delay conscious cognitive decline, is currently among the biggest challenges in the field. Herein, we aimed to compare the capacity of the Memory Binding Test (MBT) with the official diagnostic tool, the Free and Cued Selective Reminding Test (FCSRT), to anticipate AD diagnosis at an early preclinical stage based on the associative memory component of MBT (binding), suggested as more sensitive to the emergence of subtle episodic memory (EM) deficits (AD hallmark).MethodsWe assessed the tests performance longitudinally (over 5 years) in 263 cognitively-normal elderly individuals at risk of AD (>6 months of subjective memory complaints) using linear mixed-effect models controlled for age, sex, and education. We stratified participants in 2 models: amyloid-β (Aβ)/neurodegeneration (N) model, assessing Aβ burden and neurodegeneration effect [3 groups: controls (Aβ-/N-); stable/N- (Aβ+); stable/N+ (Aβ+)]; and the stable/progressors model, assessing progression to prodromal-AD effect [2 groups: stable (Aβ+); progressors (Aβ+)], based on 15 subjects who progressed to AD during follow-up (excluded once diagnosed).ResultsAβ burden was associated with significantly less MBT-intrusions, while Aβ burden and neurodegeneration together, with the most. Progression status had a strong negative effect on both tests performance. When compared with the FCSRT, the MBT seems to anticipate diagnosis based on a worst performance in a higher number of scores (including binding) in at least a year.DiscussionAnticipation of diagnosis to an asymptomatic at-risk stage, while participants remain cognitively-normal according to FCSRT cut-offs and unaware of objective EM deficits, has the potential to delay the onset of AD-linked cognitive decline by applying promising therapeutics before decline becomes too advanced.

Project description:Recent research on multi-view clustering algorithms for complex disease subtyping often overlooks aspects like clustering stability and critical assessment of prognostic relevance. Furthermore, current frameworks do not allow for a comparison between data-driven and pathway-driven clustering, highlighting a significant gap in the methodology. We present the COPS R-package, tailored for robust evaluation of single and multi-omics clustering results. COPS features advanced methods, including similarity networks, kernel-based approaches, dimensionality reduction, and pathway knowledge integration. Some of these methods are not accessible through R, and some correspond to new approaches proposed with COPS. Our framework was rigorously applied to multi-omics data across seven cancer types, including breast, prostate, and lung, utilizing mRNA, CNV, miRNA, and DNA methylation data. Unlike previous studies, our approach contrasts data- and knowledge-driven multi-view clustering methods and incorporates cross-fold validation for robustness. Clustering outcomes were assessed using the ARI score, survival analysis via Cox regression models including relevant covariates, and the stability of the results. While survival analysis and gold-standard agreement are standard metrics, they vary considerably across methods and datasets. Therefore, it is essential to assess multi-view clustering methods using multiple criteria, from cluster stability to prognostic relevance, and to provide ways of comparing these metrics simultaneously to select the optimal approach for disease subtype discovery in novel datasets. Emphasizing multi-objective evaluation, we applied the Pareto efficiency concept to gauge the equilibrium of evaluation metrics in each cancer case-study. Affinity Network Fusion, Integrative Non-negative Matrix Factorization, and Multiple Kernel K-Means with linear or Pathway Induced Kernels were the most stable and effective in discerning groups with significantly different survival outcomes in several case studies.

Dataset Information

Multi-cohort and longitudinal Bayesian clustering study of stage and subtype in Alzheimer's disease.

Publications

Multi-cohort and longitudinal Bayesian clustering study of stage and subtype in Alzheimer's disease.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets