NMR metabolomic profiles associated with long-term risk of prostate cancer

Abstract

Introduction Prostate cancer is a multifactorial disease whose aetiology is still not fully understood. Metabolomics, by measuring several hundred metabolites simultaneously, could enhance knowledge on the metabolic changes involved and the potential impact of external factors. Objectives The aim of the present study was to investigate whether pre-diagnostic plasma metabolomic profiles were associated with the risk of developing a prostate cancer within the following decade. Methods A prospective nested case-control study was set up among the 5141 men participant of the SU.VI.MAX cohort, including 171 prostate cancer cases, diagnosed between 1994 and 2007, and 171 matched controls. Nuclear magnetic resonance (NMR) metabolomic profiles were established from baseline plasma samples using NOESY1D and CPMG sequences. Multivariable conditional logistic regression models were computed for each individual NMR signal and for metabolomic patterns derived using principal component analysis. Results Men with higher fasting plasma levels of valine (odds ratio (OR) = 1.37 [1.07–1.76], p = .01), glutamine (OR = 1.30 [1.00–1.70], p = .047), creatine (OR = 1.37 [1.04–1.80], p = .02), albumin lysyl (OR = 1.48 [1.12–1.95], p = .006 and OR = 1.51 [1.13–2.02], p = .005), tyrosine (OR = 1.40 [1.06–1.85], p = .02), phenylalanine (OR = 1.39 [1.08–1.79], p = .01), histidine (OR = 1.46 [1.12–1.88], p = .004), 3-methylhistidine (OR = 1.37 [1.05–1.80], p = .02) and lower plasma level of urea (OR = .70 [.54–.92], p = .009) had a higher risk of developing a prostate cancer during the 13 years of follow-up. Conclusions This exploratory study highlighted associations between baseline plasma metabolomic profiles and long-term risk of developing prostate cancer. If replicated in independent cohort studies, such signatures may improve the identification of men at risk for prostate cancer well before diagnosis and the understanding of this disease.