Main author: Sigurður Trausti Karvelsson
Institution or Company: University of Iceland
Co-Authors, Institution or Company:
Freyr Jóhannsson, University of Iceland. Óttar Rolfsson, University of Iceland.
Introduction: Trauma is a disparate group of conditions united by the alterations to endothelial permeability and changes in the blood-clotting pathway. Recently, research has been focused on the metabolic characterization of trauma. In this study, we collected blood samples from 99 patients with a wide range of traumatic injuries (moderate, serious, and severe), examined their plasma metabolomic composition, and assessed whether it associated with the varying severity of the trauma.
Methods: Using five different LC-MS methods, we measured over 5000 metabolic features, both polar and non-polar, in 99 trauma patient plasma samples. Following data pre-processing, the metabolic features were binned into 46 co-abundance groups, or modules, based on the correlation of feature intensities. These modules were subsequently used to cluster the trauma patients. Discriminating modules for the groups were identified using random forest, and the most representative metabolic features within each module were obtained and tested as biomarkers using logistic regression models and receiver-operating characteristic (ROC) analysis.
Results: We could classify the trauma patients into three groups of different trauma severity based on the metabolic modules. A module that was functionally annotated as methionine and cysteine metabolism was most discriminating for the group of patients with severe trauma. The performance of taurine, a metabolite within that pathway, was superior to current measurements of trauma severity in stratifying trauma patients.
Conclusions: The plasma metabolomic composition of trauma patients is directly associated with the severity of the traumatic injury and can be used for prognostic evaluation of the patients.