GC-MS metabolomics reveals potential biomarkers for infection-specific subtypes in sepsis.

OBJECTIVES: Sepsis remains one of the most severe complications of infection. Specific and easily measurable biomarkers could help us predict the type of infection in sepsis. This study aimed to explore biomarkers for early identification of sepsis with different infection. METHODS: Serum was obtained from 37 patients with sepsis and 8 healthy individuals. Metabolites were screened using gas chromatography-mass spectrometry (GC-MS), and then potential biomarkers were screened. Receiver operating characteristic curve (ROC) analysis was used to evaluate potential biomarkers. RESULTS: A total of 24 differential metabolites were found between sepsis subgroup with pulmonary infection and healthy controls, and they were enriched with 8 metabolic pathways. Similarly, 27 differential metabolites were identified in abdominal infection cases, and 64 in urinary tract infection (UTI) cases, which were associated with 23 and 26 metabolic pathways, respectively. Hydroxylamine and N-acetylputrescine may be used as candidate biomarkers to distinguish pulmonary infection and abdominal infection in sepsis, respectively, while glucose, phosphoenolpyruvate and d-ribose were good discriminators in UTI. CONCLUSIONS: Our GC-MS-based metabolomic study identified potential biomarkers for sepsis across various infection types. These biomarkers show promise for the rapid and precise diagnosis of sepsis onset. Our GC-MS-based metabolomic study identified candidate metabolites associated with different infection sources in sepsis. Given the limited sample size and the correspondingly limited statistical power, these findings are exploratory and require prospective validation in an adequately powered, independent cohort before any diagnostic application can be considered.