Cerebrospinal Fluid Metabolomic Profiling in Tuberculous and Viral Meningitis: Screening Potential Markers for Differential Diagnosis

Zihui Li,Boping Du,Jing Li,Jinli Zhang,Xiaojing Zheng,Hongyan Jia,Aiying Xing,Qi Sun,Fei Liu,Zongde Zhang
DOI: https://doi.org/10.1016/j.cca.2017.01.002
IF: 6.314
2017-01-01
Clinica Chimica Acta
Abstract:Background: Tuberculous meningitis (TBM) is the most severe and frequent form of central nervous system tuberculosis. The current lack of efficient diagnostic tests makes it difficult to differentiate TBM from other common types of meningitis, especially viral meningitis (VM). Metabolomics is an important tool to identify disease specific biomarkers. However, little metabolomic information is available on adult TBM.Methods: We used H-1 nuclear magnetic resonance-based metabolomics to investigate the metabolic features of the CSF from 18 TBM and 20 VM patients. Principal component analysis and orthogonal signal correction partial least squares-discriminant analysis (OSC-PLS-DA) were applied to analyze profiling data. Metabolites were identified using the Human Metabolome Database and pathway analysis was performed with MetaboAnalyst 3.0.Results: The OSC-PLS-DA model could distinguish TBM from VM with high reliability. A total of 25 key metabolites that contributed to their discrimination were identified, including some, such as betaine and cyclohexane, rarely reported before in TBM. Pathway analysis indicated that amino acid and energy metabolism was significantly different in the CSF of TBM compared with VM.Conclusions: Twenty-five key metabolites identified in our study may be potential biomarkers for TBM differential diagnosis and are worthy of further investigation. (C) 2017 Elsevier B.V. All rights reserved.
What problem does this paper attempt to address?