Cerebrospinal fluid-based metabolomics to characterize different types of brain tumors.

ABSTRACT

BACKGROUND: Brain tumors cause significant morbidity and mortality due to rapid progression and high recurrence risks. Reliable biomarkers to improve diagnosis thereof are desirable. OBJECTIVE: This work aimed to identify panels of biomarkers for diagnostic purposes using cerebrospinal fluid (CSF)-based metabolomics. METHODS: A cohort of 163 histologically-proven patients with brain disorders was involved. Comprehensive CSF-based metabolomics was achieved by liquid chromatography-quadrupole time-of-flight spectrometric (LC-Q/TOF-MS) and multivariate statistical analyses. The diagnostic performance of the metabolic markers was evaluated using receiver operating characteristic curves. RESULTS: A total of 508 ion features were detected by the LC-Q/TOF-MS analysis, of which 27 metabolites were selected as diagnostic markers to discriminate different brain tumor types. The area under the curve (AUC) was 0.91 for lung adenocarcinoma patients with brain metastases (MBT) vs. lung adenocarcinoma patients without brain metastases (NMBT), 0.83 for primary central nervous system lymphoma (PCNSL) vs. secondary central nervous system involvement of systemic lymphoma (SCNSL), 0.77 for PCNSL vs. MBT, 0.87 for SCNSL vs. MBT, 0.86 for MBT vs. nontumorous brain diseases (NT), and 0.80 for PCNSL vs. NT. Perturbed metabolic pathways between the comparisons related mainly to amino acids and citrate metabolism. CONCLUSIONS: CSF-based metabolomics to a large extent reliably identifies significant metabolic differences between different brain tumors and shows great potential for diagnosis of brain tumors.