Prehospital immune responses and development of multiple organ dysfunction syndrome following traumatic injury: A prospective cohort study

Abstract

Background: Almost all studies that have investigated the immune response to trauma have analysed blood samples acquired post-hospital admission. Thus, we know little of the immune status of patients in the immediate postinjury phase and how this might influence patient outcomes. The objective of this study was therefore to comprehensively assess the ultra-early, within 1-hour, immune response to trauma and perform an exploratory analysis of its relationship with the development of multiple organ dysfunction syndrome (MODS).

Methods and findings: The immune and inflammatory response to trauma was analysed in 89 adult trauma patients (mean age 41 years, range 18-90 years, 75 males) with a mean injury severity score (ISS) of 24 (range 9-66), from whom blood samples were acquired within 1 hour of injury (mean time to sample 42 minutes, range 17-60 minutes). Within minutes of trauma, a comprehensive leukocytosis, elevated serum pro- and anti-inflammatory cytokines, and evidence of innate cell activation that included neutrophil extracellular trap generation and elevated surface expression of toll-like receptor 2 and CD11b on monocytes and neutrophils, respectively, were observed. Features consistent with immune compromise were also detected, notably elevated numbers of immune suppressive CD16BRIGHT CD62LDIM neutrophils (82.07 x 106/l ± 18.94 control versus 1,092 x 106/l ± 165 trauma, p < 0.0005) and CD14+HLA-DRlow/- monocytes (34.96 x 106/l ± 4.48 control versus 95.72 x 106/l ± 8.0 trauma, p < 0.05) and reduced leukocyte cytokine secretion in response to lipopolysaccharide stimulation. Exploratory analysis via binary logistic regression found a potential association between absolute natural killer T (NKT) cell numbers and the subsequent development of MODS. Study limitations include the relatively small sample size and the absence of data relating to adaptive immune cell function.

Conclusions: Our study highlighted the dynamic and complex nature of the immune response to trauma, with immune alterations consistent with both activation and suppression evident within 1 hour of injury. The relationship of these changes, especially in NKT cell numbers, to patient outcomes such as MODS warrants further investigation.

Fig 1. Flow diagram showing recruitment and analysis of study subjects.

(A) A total of 12 patients were lost between the T≤1-hour and T = 4–12-hour time points as a result of steroid treatment (n = 4), refusal of sampling (n = 6), hospital discharge (n = 1), and mortality (n = 1). (B) Patients lost as a result of mortality (n = 2), difficulty in bleeding (n = 3), refusal of sampling (n = 5), hospital discharge (n = 2), and steroid treatment (n = 1). Three patients who refused blood sampling at the T = 4–12-hour time point provided samples at the T = 48–72-hour time point. Thus, a total of 10 patients were lost between the T = 4–12-hour and T = 48–72-hour time points. Insufficient sample volume and equipment breakdown account for the differences in patient numbers between each parameter analysed. CF-DNA, cell-free DNA; fMLF, formyl-methionine-leucine-phenylalanine; LPS, lipopolysaccharide; mtDNA, mitochondrial DNA; nDNA, nuclear DNA; PMA, phorbol 12-myristate 13-acetate; ROS, reactive oxygen species; TOI, time of injury.

Fig 2. Neutrophil ROS production postinjury.

(A–B) Neutrophil ROS generation in response to PMA stimulation was assessed across time post-trauma. Data are presented as (A) percentage of neutrophils that produced ROS and (B) oxidative capacity. The number of samples analysed is indicated below each time point. The horizontal line for HC data depicts the median value. *p < 0.05, ***p < 0.0005 versus HCs. HC, healthy control; MFI, mean fluorescence intensity; PMA, phorbol 12-myristate 13-acetate; ROS, reactive oxygen species.

Fig 3. Evidence of NET formation during the ultra-early immune response to sterile traumatic injury.

(A–B) Plasma concentrations (ng/ml) of nuclear (A) and mitochondrial (B) DNA across time post-trauma. The number of samples analysed is indicated below each time point. The horizontal line for HC data depicts the median value. *p < 0.05, ***p < 0.0005 versus HCs. (C–D) Western blots showing the levels of citrullinated histone H3 in plasma samples obtained from 9 trauma patients within 1 hour of injury (C) and in 3 trauma patients across the 3 postinjury time points (D). +ve CNT, positive control; HC, healthy control; mtDNA, mitochondrial DNA; nDNA, nuclear DNA; NET, neutrophil extracellular trap.

Fig 4. Traumatic injury results in immediate alterations in the composition of the circulating neutrophil pool.

(A) Flow cytometry plots depicting the percentage of CD16^BRIGHT CD62L^DIM neutrophils (lower right quadrant) in blood samples from a single healthy control (left panel) and a trauma patient (right panel). (B–C) Prospective analysis of the absolute number (B) and frequency (C) of circulating CD16^BRIGHT CD62L^DIM neutrophils post-trauma. The number of samples analysed is indicated below each time point. ***p < 0.0005 versus HCs. The horizontal line for HC data depicts the median value. HC, healthy control.

Fig 5. Traumatic injury results in elevated percentages and absolute numbers of circulating CD14⁺HLA-DR^-/low immunosuppressive monocytes.

(A) Representative flow cytometry plots depicting the percentage of CD14⁺HLA-DR^-/low monocytes (upper left quadrant) in a single HC and a trauma patient across time. (B–C) Prospective assessment of the percentage (B) and absolute number (C) of CD14⁺HLA-DR^-/low monocytes post-trauma. The number of samples analysed is indicated below each time point. The horizontal line for HC data depicts the median value. *p < 0.05, ***p < 0.0005 versus HC. HC, healthy control.

Fig 6. Serum cytokine and chemokine concentrations post-trauma.

Serum concentrations of IL-6 (A), IL-8 (B), G-CSF (C), IL-1Ra (D), TNF-α (E), and IL-10 (F) across time post-trauma. The number of patient and HC samples analysed is indicated below each time point. The horizontal line for HC data depicts the median value. **p < 0.005, ***p < 0.0005 versus HCs. G-CSF, granulocyte-colony stimulating factor; HC, healthy control; IL, interleukin; IL-1Ra, interleukin-1 receptor antagonist; TNF- α, tumour necrosis factor-alpha.

Fig 7. Serum cortisol concentrations post-trauma.

The number of patient and HC samples analysed is indicated below each time point. The horizontal line for HC data depicts the median value. **p < 0.005, ***p < 0.0005 versus HCs. HC, healthy control.