Brain regional volume analysis to differentiate Alzheimer’s disease (AD) and vascular dementia (VD) from healthy control (HC): Machine learning approach

In Alzheimer’s disease (AD), the contribution of pathophysiological mechanisms other than amyloidosis and tauopathy is now widely recognized, although not clearly quantifiable by means of fluid biomarkers. We aimed to identify quantifiable protein biomarkers reflecting neuroinflammation in AD using multiplex proximity extension assay (PEA) testing. Cerebrospinal fluid (CSF) samples from patients with mild cognitive impairment due to AD (AD-MCI) and from controls, i.e., patients with other neurological diseases (OND), were analyzed with the Olink Inflammation PEA biomarker panel. A machine-learning approach was then used to identify biomarkers discriminating AD-MCI (n: 34) from OND (n: 25). On univariate analysis, SIRT2, HGF, MMP-10, and CXCL5 showed high discriminatory performance (AUC 0.809, p = 5.2 × 10−4, AUC 0.802, p = 6.4 × 10−4, AUC 0.793, p = 3.2 × 10−3, AUC 0.761, p = 2.3 × 10−3, respectively), with higher CSF levels in AD-MCI patients as compared to controls. These same proteins were the best contributors to the penalized logistic regression model discriminating AD-MCI from controls (AUC of the model 0.906, p = 2.97 × 10−7). The biological processes regulated by these proteins include astrocyte and microglia activation, amyloid, and tau misfolding modulation, and blood-brain barrier dysfunction. Using a high-throughput multiplex CSF analysis coupled with a machine-learning statistical approach, we identified novel biomarkers reflecting neuroinflammation in AD. Studies confirming these results by means of different assays are needed to validate PEA as a multiplex technique for CSF analysis and biomarker discovery in the field of neurological diseases.

Download Full-text

Predicting future regional tau accumulation in asymptomatic and early Alzheimer’s disease

10.1101/2020.08.15.252601 ◽

2020 ◽

Author(s):

Joseph Giorgio ◽

William J Jagust ◽

Suzanne Baker ◽

Susan M. Landau ◽

Peter Tino ◽

...

Keyword(s):

Machine Learning ◽

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Clinical Trial ◽

Trial Design ◽

Clinical Trial Design ◽

Learning Approach ◽

Multimodal Data ◽

Machine Learning Approach ◽

Β Amyloid

AbstractThe earliest stages of Alzheimer’s disease (AD) involve interactions between multiple pathophysiological processes. Although these processes are well studied, we still lack robust tools to predict individualised trajectories of disease progression. Here, we employ a robust and interpretable machine learning approach to combine multimodal biological data and predict future tau accumulation, translating predictive information from deep phenotyping cohorts at early stages of AD to cognitively normal individuals. In particular, we use machine learning to quantify interactions between key pathological markers (β-amyloid, medial temporal atrophy, tau and APOE 4) at early and asymptomatic stages of AD. We next derive a predictive index that stratifies individuals based on future pathological tau accumulation, highlighting two critical features for optimal clinical trial design. First, future tau accumulation provides a better outcome measure compared to changes in cognition. Second, stratification based on multimodal data compared to β-amyloid alone reduces the sample size required to detect a clinically meaningful change in tau accumulation. Further, we extend our machine learning approach to derive individualised trajectories of future pathological tau accumulation in early AD patients and accurately predict regional future rate of tau accumulation in an independent sample of cognitively unimpaired individuals. Our results propose a robust approach for fine scale stratification and prognostication with translation impact for clinical trial design at asymptomatic and early stages of AD.One Sentence SummaryOur machine learning approach combines baseline multimodal data to make individualised predictions of future pathological tau accumulation at prodromal and asymptomatic stages of Alzheimer’s disease with high accuracy and regional specificity.

Download Full-text

A blood-based signature of cerebral spinal fluid Aβ1–42 status

10.1101/190207 ◽

2017 ◽

Author(s):

Benjamin Goudey ◽

Bowen J Fung ◽

Christine Schieber ◽

Noel G Faux ◽

◽

...

Keyword(s):

Machine Learning ◽

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Minimally Invasive ◽

Cerebral Spinal Fluid ◽

Spinal Fluid ◽

Carrier Status ◽

Learning Approach ◽

Risk Indicator ◽

Machine Learning Approach

ABSTRACTIt is increasingly recognized that Alzheimer’s disease (AD) exists before dementia is present and that shifts in amyloid beta occur long before clinical symptoms can be detected. Early detection of these molecular changes is a key aspect for the success of interventions aimed at slowing down rates of cognitive decline. Recent evidence indicates that of the two established methods for measuring amyloid, a decrease in cerebral spinal fluid (CSF) amyloid β1−42 (Aβ1−42) may be an earlier indicator of Alzheimer’s disease risk than measures of amyloid obtained from Positron Emission Topography (PET). However, CSF collection is highly invasive and expensive. In contrast, blood collection is routinely performed, minimally invasive and cheap. In this work, we develop a blood-based signature that can provide a cheap and minimally invasive estimation of an individual’s CSF amyloid status using a machine learning approach. We show that a Random Forest model derived from plasma analytes can accurately predict subjects as having abnormal (low) CSF Aβ1−42 levels indicative of AD risk (0.84 AUC, 0.78 sensitivity, and 0.73 specificity). Refinement of the modeling indicates that only APOEε4 carrier status and four analytes are required to achieve a high level of accuracy. Furthermore, we show across an independent validation cohort that individuals with predicted abnormal CSF Aβ1−42 levels transitioned to an AD diagnosis over 120 months significantly faster than those predicted with normal CSF Aβ1−42 levels and that the resulting model also performs reasonably across PET Aβ1−42 status.This is the first study to show that a machine learning approach, using plasma protein levels, age and APOEε4 carrier status, is able to predict CSF Aβ1-42 status, the earliest risk indicator for AD, with high accuracy.

Download Full-text