King's College London

TY - JOUR

T1 - Blood metabolite markers of neocortical amyloid-β burden

T2 - discovery and enrichment using candidate proteins

AU - Voyle, N

AU - Kim, M

AU - Proitsi, P

AU - Ashton, N J

AU - Baird, A L

AU - Bazenet, C

AU - Hye, A

AU - Westwood, S

AU - Chung, R

AU - Ward, M

AU - Rabinovici, G D

AU - Lovestone, S

AU - Breen, G

AU - Legido-Quigley, C

AU - Dobson, R J B

AU - Kiddle, S J

PY - 2017/2

Y1 - 2017/2

N2 - We believe this is the first study to investigate associations between blood metabolites and neocortical amyloid burden (NAB) in the search for a blood-based biomarker for Alzheimer's disease (AD). Further, we present the first multi-modal analysis of blood markers in this field. We used blood plasma samples from 91 subjects enrolled in the University of California, San Francisco Alzheimer's Disease Research Centre. Non-targeted metabolomic analysis was used to look for associations with NAB using both single and multiple metabolic feature models. Five metabolic features identified subjects with high NAB, with 72% accuracy. We were able to putatively identify four metabolites from this panel and improve the model further by adding fibrinogen gamma chain protein measures (accuracy=79%). One of the five metabolic features was studied in the Alzheimer's Disease Neuroimaging Initiative cohort, but results were inconclusive. If replicated in larger, independent studies, these metabolic features and proteins could form the basis of a blood test with potential for enrichment of amyloid pathology in anti-amyloid trials.

AB - We believe this is the first study to investigate associations between blood metabolites and neocortical amyloid burden (NAB) in the search for a blood-based biomarker for Alzheimer's disease (AD). Further, we present the first multi-modal analysis of blood markers in this field. We used blood plasma samples from 91 subjects enrolled in the University of California, San Francisco Alzheimer's Disease Research Centre. Non-targeted metabolomic analysis was used to look for associations with NAB using both single and multiple metabolic feature models. Five metabolic features identified subjects with high NAB, with 72% accuracy. We were able to putatively identify four metabolites from this panel and improve the model further by adding fibrinogen gamma chain protein measures (accuracy=79%). One of the five metabolic features was studied in the Alzheimer's Disease Neuroimaging Initiative cohort, but results were inconclusive. If replicated in larger, independent studies, these metabolic features and proteins could form the basis of a blood test with potential for enrichment of amyloid pathology in anti-amyloid trials.

UR - http://www.scopus.com/inward/record.url?scp=85016160687&partnerID=8YFLogxK

U2 - 10.1038/tp.2015.205

DO - 10.1038/tp.2015.205

M3 - Article

C2 - 26812040

VL - 6

JO - Translational psychiatry

JF - Translational psychiatry

SN - 2158-3188

M1 - e719

ER -