Related Experiment Video
Updated: Sep 27, 2026
Identification and Quantification of Deranged Metabolites in Critically Ill Patients Using NMR-Based Metabolomics
Published on: November 29, 2024
Metabolomics Profiles and Prognostic Biomarkers in Sepsis Survivors and Nonsurvivors
Jing Xu1, Zhou Huang2, He Li2
1Department of Emergency, The First Affiliated Hospital, Guangxi Medical University, Nanning, Guangxi, China; Guangxi University Key Laboratory of Emergency Medicine, The First Affiliated Hospital, Guangxi Medical University, Nanning, Guangxi, China; Hunan Provincial Key Laboratory of Emergency and Critical Care Metabonomics, Department of Emergency Medicine, Clinical Research Center For Emergency and Critical Care In Hunan Province, Hunan Provincial Institute of Emergency Medicine, Hunan Provincial People's Hospital, The First Affiliated Hospital of Hunan Normal University, Changsha, Hunan, China.
Introduction:
Sepsis imposes an overwhelming global critical care burden, contributing to an estimated 49 million annual deaths worldwide. Although multiple sepsis-related metabolic profiles have been characterized in prior research, no metabolomic biomarkers have yet entered routine clinical prognostic practice. This study aimed to identify actionable serum metabolite biomarkers to stratify 28-d mortality risk among adult sepsis patients.
Methods:
We performed high-resolution untargeted serum metabolomics to profile molecular signatures across three cohorts: 60 sepsis nonsurvivors, 60 sepsis survivors, and 20 age- and sex-matched healthy controls (HCs). Supervised partial least squares discriminant analysis implemented via MetaboAnalyst 5.0 was used to screen differential circulating metabolites. Receiver operating characteristic curve analysis quantified the predictive performance of candidate metabolites for sepsis mortality. The correlations between differential metabolite concentrations and clinical laboratory/severity indices were analyzed using R software.
Results:
Partial least squares discriminant analysis models demonstrated clear metabolic separation among HCs, sepsis survivors, and nonsurvivors. Pairwise comparison identified 1092 differential metabolites between survivors and HCs, 757 metabolites between survivors and nonsurvivors, and 1230 metabolites between nonsurvivors and HCs. A core set of 200 mortality-linked metabolites was enriched in five canonical Kyoto Encyclopedia of Genes and Genomes pathways: steroid hormone biosynthesis, lysine degradation, starch and sucrose metabolism, riboflavin metabolism, and pantothenate/coenzyme A biosynthesis. Five metabolites exhibited superior prognostic power relative to conventional clinical markers: tsangane L-3-glucoside, 5-L-glutamyl-L-alanine, 15-methyl-15S-PGE2, allysine, and 4-nitrophenol. These analytes displayed strong linear correlations with validated disease severity and organ dysfunction biomarkers, supporting their utility as mortality prognostic indicators.
Conclusion:
Distinct serum metabolomic landscapes differentiate healthy individuals, sepsis survivors, and fatal sepsis patients. Tsangane L-3-glucoside, 5-L-glutamyl-L-alanine, 15-methyl-15S-PGE2, allysine, and 4-nitrophenol represent promising preliminary candidate biomarkers to early discriminate sepsis patients at high risk of 28-d mortality.
