Targeted metabolomics for serum amino acids and acylcarnitines in patients with lung cancer.

Abstract

Lung cancer is one of the most prevalent types of cancer, but accurate diagnosis remains a challenge. The aim of the present study was to create a model using amino acids and acylcarnitines for lung cancer screening. Serum samples were obtained from two groups of patients with lung cancer recruited in 2015 (including 40 patients and 100 matched controls) and 2017 (including 17 patients and 30 matched controls). Using a metabolomics method, 21 metabolites (13 types of amino acids and 8 types of acylcarnitines) were measured using liquid chromatography-tandem mass spectrometry. Data (from the 2015 and 2017 data sets) were analysed using a Mann-Whitney U test, Student's t-test, Welch's F test, receiver-operator characteristic curve or logistic regression in order to investigate the potential biomarkers. Six metabolites (glycine, valine, methionine, citrulline, arginine and C16-carnitine) were indicated to be involved in distinguishing patients with lung cancer from healthy controls. The six discriminating metabolites from the 2017 data set were further analysed using Partial least squares-discriminant analysis (PLS-DA). The PLS-DA model was verified using Spearman's correlation analysis and receiver operating characteristic curve analysis. These results demonstrated that the PLS-DA model using the six metabolites (glycine, valine, methionine, citrulline, arginine and C16-carnitine) had a strong ability to identify lung cancer. Therefore, the PLS-DA model using glycine, valine, methionine, citrulline, arginine and C16-carnitine may become a novel screening tool in patients with lung cancer.