Li H, Lin Y, Zheng S, Yu T, Xie Y, Yin Z. Untargeted metabolomics analysis of cerebrospinal fluid in patients with leptomeningeal metastases from non-small cell lung cancer.
Biotechnol Genet Eng Rev 2024;
40:815-832. [PMID:
36942709 DOI:
10.1080/02648725.2023.2191069]
[Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/01/2023] [Accepted: 03/06/2023] [Indexed: 03/23/2023]
Abstract
OBJECTIVE
To explore and analyze the diagnostic value of metabolic markers in cerebrospinal fluid (CSF) in leptomeningeal metastases (LM) of non-small cell lung cancer (NSCLC).
METHODS
Forty-six CSF samples from patients with NSCLC-LM were collected. Another 48 CSF samples from patients with nonmalignant neurological diseases were selected as control group. Metabolomic analysis of CSF was performed by high-performance liquid chromatography-mass spectrometry. Principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) were applied for modeling. A multi-criteria evaluation system (variable importance value >1, multiple of change >2 and P < 0.05 for univariate analysis) was used to find differential metabolites between two groups. The subject working characteristic curves and pathway enrichment analysis were used to screen metabolites and pathways associated with NSCLC-LM.
RESULTS
The PCA model and OPLS-DA model showed good overall data quality. Thirty endogenous differential metabolites were screened, and six potential biomarkers were further identified, including tyrosine (t = 3.37, P = 0.024, AUC = 0.967), phenylalanine (t = 3.98, P < 0.001, AUC = 0.992), pyruvate (t = 4.48, P < 0.001, AUC = 0.976), tryptophan (t = -2.5, P = 0.014, AUC = 0.935), adenosine monophosphate (t = -6.13, P < 0.001, AUC = 0.932) and glucose (t = -4.00, P < 0.001, AUC = 0.993). Thirty differential metabolites screened were subjected to metabolic pathway enrichment analysis and matched to 20 relevant metabolic pathways, of which the four most likely to cause metabolite changes were as follows: glycolysis and sugar metabolism synthesis, pyruvate metabolism, phenylalanine metabolism, and phenylalanine, tyrosine and tryptophan biosynthesis.
CONCLUSIONS
Untargeted metabolomics can effectively screen for CSF metabolites specific to NSCLC-LM patients, and six potential metabolites and their metabolic pathways might be involved in the pathogenesis of NSCLC-LM.
Collapse