TY - JOUR
T1 - Multivariate analyses of peripheral blood leukocyte transcripts distinguish Alzheimer's, Parkinson's, control, and those at risk for developing Alzheimer's
AU - Delvaux, Elaine
AU - Mastroeni, Diego
AU - Nolz, Jennifer
AU - Chow, Nienwen
AU - Sabbagh, Marwan
AU - Caselli, Richard J.
AU - Reiman, Eric M.
AU - Marshall, Frederick J.
AU - Coleman, Paul
N1 - Publisher Copyright:
© 2017 Elsevier Inc.
PY - 2017/10
Y1 - 2017/10
N2 - The need for a reliable, simple, and inexpensive blood test for Alzheimer's disease (AD) suitable for use in a primary care setting is widely recognized. This has led to a large number of publications describing blood tests for AD, which have, for the most part, not been replicable. We have chosen to examine transcripts expressed by the cellular, leukocyte compartment of blood. We have used hypothesis-based cDNA arrays and quantitative PCR to quantify the expression of selected sets of genes followed by multivariate analyses in multiple independent samples. Rather than a single study with no replicates, we chose an experimental design in which there were multiple replicates using different platforms and different sample populations. We have divided 177 blood samples and 27 brain samples into multiple replicates to demonstrate the ability to distinguish early clinical AD (Clinical Dementia Rating scale 0.5), Parkinson's disease (PD), and cognitively unimpaired APOE4 homozygotes, as well as to determine persons at risk for future cognitive impairment with significant accuracy. We assess our methods in a training/test set and also show that the variables we use distinguish AD, PD, and control brain. Importantly, we describe the variability of the weights assigned to individual transcripts in multivariate analyses in repeated studies and suggest that the variability we describe may be the cause of inability to repeat many earlier studies. Our data constitute a proof of principle that multivariate analysis of the transcriptome related to cell stress and inflammation of peripheral blood leukocytes has significant potential as a minimally invasive and inexpensive diagnostic tool for diagnosis and early detection of risk for AD.
AB - The need for a reliable, simple, and inexpensive blood test for Alzheimer's disease (AD) suitable for use in a primary care setting is widely recognized. This has led to a large number of publications describing blood tests for AD, which have, for the most part, not been replicable. We have chosen to examine transcripts expressed by the cellular, leukocyte compartment of blood. We have used hypothesis-based cDNA arrays and quantitative PCR to quantify the expression of selected sets of genes followed by multivariate analyses in multiple independent samples. Rather than a single study with no replicates, we chose an experimental design in which there were multiple replicates using different platforms and different sample populations. We have divided 177 blood samples and 27 brain samples into multiple replicates to demonstrate the ability to distinguish early clinical AD (Clinical Dementia Rating scale 0.5), Parkinson's disease (PD), and cognitively unimpaired APOE4 homozygotes, as well as to determine persons at risk for future cognitive impairment with significant accuracy. We assess our methods in a training/test set and also show that the variables we use distinguish AD, PD, and control brain. Importantly, we describe the variability of the weights assigned to individual transcripts in multivariate analyses in repeated studies and suggest that the variability we describe may be the cause of inability to repeat many earlier studies. Our data constitute a proof of principle that multivariate analysis of the transcriptome related to cell stress and inflammation of peripheral blood leukocytes has significant potential as a minimally invasive and inexpensive diagnostic tool for diagnosis and early detection of risk for AD.
KW - Alzheimer's disease
KW - Blood test
KW - Prodromal detection
KW - RNA
UR - http://www.scopus.com/inward/record.url?scp=85023755441&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85023755441&partnerID=8YFLogxK
U2 - 10.1016/j.neurobiolaging.2017.05.012
DO - 10.1016/j.neurobiolaging.2017.05.012
M3 - Article
C2 - 28716532
AN - SCOPUS:85023755441
SN - 0197-4580
VL - 58
SP - 225
EP - 237
JO - Neurobiology of Aging
JF - Neurobiology of Aging
ER -