Screening and diagnostic efficacy evaluation of metabolomics biomarkers in patients with active tuberculosis / 中华检验医学杂志

ABSTRACT

Objective:Based on UPLC-Q-TOF/MS, to screen a panel of plasma metabolite biomarkers for TB diagnosis and evaluate its diagnostic efficacy.Methods:102 active TB patients [49 males, 53 females, age 40.0(24.0, 48.5) years] from Shanghai Pulmonary Hospital and Shanghai East Hospital from January 2017 to January 2018, 100 TB-IGRA positive patients [55 males, 45 females, age 44.0(37.0, 52.0) years] and 96 healthy controls [55 males, 41 females, age 43.0(32.2, 52.8) years] from Shanghai East Hospital were randomly enrolled. UPLC-Q-TOF/MS technology was used to detect small molecule metabolites in plasma. Combined with multivariate statistical methods VIP and univariate statistic analysis Student's t-test, the main differential metabolites in the plasma of patients with active tuberculosis were filtered. The ROC curve was analyzed for the differential metabolites, and the AUC value, specificity, and sensitivity for diagnosis were used to screen metabolic biomarkers with diagnostic potential. Results:All the samples examined resulted in 10 266 variables, and 1 153 substances were identified by qualitative retrieval through the human metabolome database. After pairwise comparison of samples from the three groups, differential metabolites that simultaneously satisfied VIP > 1 and P<0.05 were plotted into a Venn diagram, and the resulting intersection set contained 38 major differential metabolites. The ROC curve analysis of 38 major metabolites showed that the area under the curve of lactic acid, dopamine, 9-pentadecenoic acid, and 12,13-dihydroxy octadecadienoic acid in the diagnosis of active tuberculosis were 0.92, 0.98, 0.94, and 0.94, respectively, the specificity was both more than 90% and the sensitivity was both more than 80%. The specificity and sensitivity of four metabolites in the combined diagnosis of active tuberculosis were both 94%. Conclusion:Lactic acid, dopamine, 9-pentadecenoic acid, and 12, 13-dihydroxy octadecadienoic acid can be used as potential metabolic biomarkers for tuberculosis diagnosis.