Metabolomic in severe traumatic brain injury: exploring primary, secondary injuries, diagnosis, and severity

Abstract

Background: Traumatic brain injury (TBI) is a major public health concern worldwide, contributing to high rates of injury-related death and disability. Severe traumatic brain injury (sTBI), although it accounts for only 10% of all TBI cases, results in a mortality rate of 30-40% and a significant burden of disability in those that survive. This study explored the potential of metabolomics in the diagnosis of sTBI and explored the potential of metabolomics to examine probable primary and secondary brain injury in sTBI.

Methods: Serum samples from 59 adult patients with sTBI and 35 age- and sex-matched orthopedic injury controls were subjected to quantitative metabolomics, including proton nuclear magnetic resonance (¹H-NMR) and direct infusion/liquid chromatography-tandem mass spectrometry (DI/LC-MS/MS), to identify and quantify metabolites on days 1 and 4 post-injury. In addition, we used advanced analytical methods to discover metabo-patterns associated with sTBI diagnosis and those related to probable primary and secondary brain injury.

Results: Our results showed different serum metabolic profiles between sTBI and orthopedic injury (OI) controls, with significant changes in measured metabolites on day 1 and day 4 post-brain injury. The number of altered metabolites and the extent of their change were more pronounced on day 4 as compared to day 1 post-injury, suggesting an evolution of mechanisms from primary to secondary brain injury. Data showed high sensitivity and specificity in separating sTBI from OI controls for diagnosis. Energy-related metabolites such as glucose, pyruvate, lactate, mannose, and polyamine metabolism metabolites (spermine and putrescine), as well as increased acylcarnitines and sphingomyelins, occurred mainly on day 1 post-injury. Metabolites of neurotransmission, catecholamine, and excitotoxicity mechanisms such as glutamate, phenylalanine, tyrosine, and branched-chain amino acids (BCAAs) increased to a greater degree on day 4. Further, there was an association of multiple metabolites, including acylcarnitines (ACs), lysophosphatidylcholines (LysoPCs), glutamate, and phenylalanine, with injury severity at day 4, while lactate, glucose, and pyruvate correlated with injury severity on day 1.

Conclusion: The results demonstrate that serum metabolomics has diagnostic potential for sTBI and may reflect molecular mechanisms of primary and secondary brain injuries when comparing metabolite profiles between day 1 and day 4 post-injury. These early changes in serum metabolites may provide insight into molecular pathways or mechanisms of primary injury and ongoing secondary injuries, revealing potential therapeutic targets for sTBI. This work also highlights the need for further research and validation of sTBI metabolite biomarkers in a larger cohort.

Keywords: Metabolites; Metabolomics; Primary and secondary injury; sTBI.

Fig. 1

OPLS-DA Plots to discriminate the OI control and sTBI patients on day 1 and day 4. OPLS-DA analysis to establish prediction models comparing sTBI samples on days 1 and 4 vs OI controls based on the most differentiating metabolites (VIP > 1.0) obtained by DI/LC–MS/MS and ¹H-NMR. (A) and (C) sTBI samples on day 1, (B) and (D) sTBI samples on day 4. A higher R2 (R²Y) reflects a better model and a higher Q2 (Q²Y) reflects a higher predictability of the models

Fig. 2

Heatmaps analysis to show the differences in metabolites of OI control and sTBI patients on day 1 and day 4. Clustered heatmaps show the alterations of top significant metabolites (VIP > 1) between sTBI samples on days 1 and 4 post-sTBI compared to OI controls. (A) DI/LC–MS/MS, (B) ¹H-NMR. The relative quantification scale is on the top right varying from − 6 to + 6

Fig. 3

Boxplots of ANOVA analysis of OI control and sTBI patients on day 1 and day 4. Measured specific metabolite levels revealing metabolic phenotype patterns (A) Metabolites increase from OI control to sTBI on day 1 and increase to day 4 post-injury. (B) Metabolites decrease from OI control to sTBI on day 1 and decrease to day 4 post-injury. (LC/DI-MS/MS dataset)

Fig. 4

Boxplots of ANOVA analysis of OI control and sTBI patients on day 1 and day 4. Measured specific metabolite levels revealing metabolic phenotype patterns (A) Metabolites increased from OI control to sTBI on day 4 but were higher on day 1 post-injury (i.e. the metabolites were elevated on day 1 but were less elevated on day 4 post-injury). (B) Metabolites decrease from OI control to sTBI on day 4 and then on day 1. (i.e. metabolite concentrations went down on days 1 and 4 but were lower on day 1 than day 4 posy-injury). (LC/DI-MS/MS dataset)

Fig. 5

Diagram of metabolite changes and related pathways and mechanisms on day 1 and day 4. A summary of the serum metabolite alteration on days 1 and 4 post-injury and the potentially disrupted pathophysiological mechanisms of brain injury (boxes) based on the most differentiating metabolites obtained by the multivariate (VIP) and univariate (p value) analysis. The red box shows increased metabolic mechanisms of injury, and the blue box decreases metabolic mechanisms of injury related to potential pathophysiological pathways involved in brain injury. Arrows show increased (up) associated metabolites and decreased (down) associated metabolites with the related known pathophysiologic mechanisms of injury