Blood levels of histone-complexed DNA fragments are associated with coagulopathy, inflammation and endothelial damage early after trauma.

BACKGROUND: Tissue injury increases blood levels of extracellular histones and nucleic acids, and these may influence hemostasis, promote inflammation and damage the endothelium. Trauma-induced coagulopathy (TIC) may result from an endogenous response to the injury that involves the neurohumoral, inflammatory and hemostatic systems. AIMS: To study the contribution of extracellular nucleic constituents to TIC, inflammation and endothelial damage. SETTING AND
DESIGN: Prospective observational study.
MATERIALS AND METHODS: We investigated histone-complexed DNA fragments (hcDNA) along with biomarkers of coagulopathy, inflammation and endothelial damage in plasma from 80 trauma patients admitted directly to the Trauma Centre from the scene of the accident. Blood was sampled a median of 68 min (IQR 48-88) post injury. Trauma patients with hcDNA levels >median or ≤median were compared.
RESULTS: Trauma patients with high plasma hcDNA had higher Injury Severity Score (ISS) and level of sympathoadrenal activation (higher adrenaline and noradrenaline) and a higher proportion of prolonged activated partial thromboplastin time (APTT) and higher D-dimer, tissue-type plasminogen activator (tPA), Annexin V and soluble CD40 ligand (sCD40L) concurrent with lower plasminogen activator inhibitor (PAI)-1) and prothrombin fragment (PF) 1 + 2 (all P < 0.05), all indicative of impaired thrombin generation, hyperfibrinolysis and platelet activation. Furthermore, patients with high hcDNA had enhanced inflammation and endothelial damage evidenced by higher plasma levels of terminal complement complex (sC5b-9), IL-6, syndecan-1, thrombomodulin and tissue factor pathway inhibitor (all P < 0.05).
CONCLUSIONS: Excessive release of extracellular histones and nucleic acids seems to contribute to the hypocoagulability, inflammation and endothelial damage observed early after trauma.

INTRODUCTION

Tissue injury accompanying massive trauma results in immediate increases in circulating levels of damage-associated molecular patterns (DAMPs) like high-mobility group box1 (HMGB1) and nucleic acids including histone-complexed DNA (nucleosomes),[123456] and endogenous molecules that signal tissue and cell damage.[67] Trauma remains a major cause of death and disability worldwide[8] and since hemorrhage accounts for most early trauma deaths,[9] much attention has been given to the early trauma-induced coagulopathy (TIC).[10] Though the exact pathophysiologic mechanism(s) contributing to TIC remain elusive, there is emerging consensus that this coagulopathy reflects an endogenous response to the injury involving both the neurohumoral, inflammatory, complement and hemostatic systems.[1011121314]

Extracellular RNA (and to a lesser degree DNA) augments (auto) activation of proteases of the contact phase pathway of blood coagulation[15] and extracellular histones (particularly histones complexed with DNA) promote thrombin generation through platelet-dependent mechanisms,[16] the latter in part by reducing thrombomodulin (TM)-mediated protein C activation.[17]

Though many trauma patients with extensive tissue injuries display some degree of TIC,[1011121314] the association between DAMP release and TIC has not previously been studied. The aim of the present study was to investigate the association between circulating DAMPs, in the form of histone-complexed DNA fragments (hcDNA), and coagulopathy, inflammation and endothelial damage in trauma patients upon hospital admission.

MATERIALS AND METHODS

Study design and patients

Prospective observational cohort study of trauma patients admitted directly to a Level I equivalent Trauma Centre (TC) at a tertiary hospital between March 2010 and November 2010. The study is part of an ongoing larger multicentre study, Activation of Coagulation and Inflammation after Trauma 3 (ACIT3),[18] approved by the Regional Ethics Committee and the National Data Protection Agency and conducted in accordance with the 2nd Declaration of Helsinki. Written informed consent was obtained from the patients or next of kin.

Inclusion criteria: Adult trauma patients (≥18 years of age) meeting criteria for trauma team activation and insertion of an arterial cannula. Exclusion criteria as previously described in other publications based on this cohort of patients.[192021222324] Data on demography, clinical and biochemical parameters, investigations, management and 30-day mortality were recorded and Injury Severity Score (ISS) scores were obtained from the Trauma Audit and Research Network database or scored locally.

Here we report on findings related to a cohort of 80 patients recruited to the ACIT3 study who had extensive blood samples and analysis performed.

Blood sampling and ELISA measurements

Blood was sampled for standard arterial blood gas (ABG, Radiometer ABL 725/735), routine biochemistry (analyzed in a DS/EN ISO 15189 standardized laboratory (Sysmex XE-2100 (hemoglobin, platelets, leukocytes), ACL TOP (APTT, INR, AT, fibrinogen)) and research analyses (rapidly processed plasma/serum, stored at −80°C). Biomarkers were measured by commercially available immunoassays in uniplicate in plasma/serum: EDTA plasma: Histone-complexed DNA fragments (hcDNA, Cell Death Detection ELISAPLUS, Roche, Hvidovre, Denmark); adrenaline and noradrenaline (2-CAT ELISA, Labor Diagnostica, Nordhorn, Germany); Annexin V (ADI, Stamford, CT, US); soluble thrombomodulin (sTM, Nordic Biosite, Copenhagen, Denmark); D-dimer (ADI) and soluble CD40 ligand (sCD40L, RandD Systems Europe). Citrate plasma: Protein C (PC, Helena Laboratories, Beaumont, TX, US); activated protein C (APC, USCNLIFE, Ltd. Wuhan, China); tissue-type plasminogen activator (tPA, ADI); plasminogen activator inhibitor-1 (PAI-1, Assaypro, Sct. Charles, MO, US); prothrombinfragment 1 and 2 (PF1.2, USCNLIFE); thrombin/antithrombin complex (TAT, USCNLIFE); tissue factor pathway inhibitor (TFPI, ADI); factor XIII (FXIII, Assaypro); terminal complement complex (sC5b-9, MicroVue sC5b-9 plus EIA Kit, Quidel Corp., San Diego, CA, US) and interleukin-6 (IL-6, Quantikine HS, RandD Systems Europe). Serum: Syndecan-1 (Diaclone SAS, Besancon, France).

Statistics

Statistical analysis was performed using SAS 9.1 (SAS Institute Inc., Cary, NC, US). Data from patients stratified according to hcDNA level (>median vs. ≤median) were compared by Wilcoxon Rank Sum tests and Chi-square/Fisher's exact tests, as appropriate. Correlations were investigated by Spearman correlations and presented by rho and P values. Data are presented as medians with inter-quartile ranges (IQR). The number of patients included in the present study was not based on power calculations but on the number of samples available for analysis at one ELISA kit since this was a hypothesis-generating study where we performed a large number of biomarker analyses (19 biomarkers*80 samples = 1,520 analyses). P < 0.05 were considered significant.

RESULTS

Patients

The present study included 80 trauma patients (median age 46 years (IQR 33-64) and ISS 17 (IQR 10-28), 91% with blunt trauma and 31% had severe traumatic brain injury (sTBI, Abbreviated Injury Score head >3)). Blood samples were drawn a median of 68 min (IQR 48-88) after the injury. Twelve patients (15%) had increased APTT and/or INR, 14% received massive transfusion (>10 RBC the initial 24 h) and overall 30-day mortality was 18%.

Clinical presentation and biomarker profile in patients with high vs. low hcDNA

When stratifying patients according to median level of hcDNA (above vs. below median corresponding to high and low hcDNA groups), patients with high hcDNA had significantly higher ISS and plasma catecholamines and also a higher proportion of prolonged APTT and higher D-dimer, tPA, Annexin V and sCD40L concurrent with lower PAI-1 and PF1.2 indicating impaired thrombin generation, hyperfibrinolysis and platelet activation. Furthermore, patients with high hcDNA had enhanced inflammation and endothelial damage evidenced by higher plasma levels of sC5b-9, IL-6, syndecan-1, sTM and TFPI, the latter normally embedded in the glycocalyx. No significant differences in demography, platelet count, massive transfusion or 30-day mortality were found according to the hcDNA groups [Table 1].

Table 1

Demography, injury severity, mortality, sympathoadrenal activation, biochemistry, hemostasis and biomarkers of coagulopathy in 80 trauma patients stratified according to histone-complexed DNA fragments (hcDNA) into high hcDNA>median and low hcDNA-median groups

The circulating hcDNA levels correlated with ISS (rho = 0.53, P < 0.001), adrenaline (rho = 0.58, P < 0.001), noradrenaline (rho = 0.37, P = 0.002), APTT (rho = 0.25, P = 0.025), D-dimer (rho = 0.55, P < 0.001), tPA (rho = 0.30, P = 0.008), sCD40L (rho = 0.43, P < 0.001), Annexin V (rho = 0.41, P < 0.001), FXIII (rho= -0.26, P = 0.019), IL-6 (rho = 0.34, P = 0.002), syndecan-1 (rho = 0.49, P < 0.001) and sTM (rho = 0.61, P < 0.001) i.e., markers of injury severity, sympathoadrenal activation, coagulation factor consumption, hyperfibrinolysis, platelet activation, inflammation and endothelial damage. Furthermore, hcDNA correlated with SBE (rho = 0.25, P = 0.035), WBC (rho = 0.35, P = 0.002), glucose (rho = 0.32, P = 0.006) and RBC transfusion the first 6 (rho = 0.29, P = 0.009) and 24 (rho = 0.32, P = 0.004) hours.

Biomarker response to injury in patients with high vs. low hcDNA

To study discrepancies in the hemostatic response to injury in relation to hcDNA, we investigated correlations between ISS and biomarkers in patients stratified according to median hcDNA. Only in patients with low hcDNA did ISS correlate with TAT whereas ISS and TAT did not correlate in high hcDNA patients [Figure 1a]. Furthermore, only in patients with high hcDNA did ISS correlate with APTT and INR, global markers of coagulopathy, whereas ISS and APTT/INR did not correlate in low hcDNA patients [Figure 1b and c]. Other biomarkers displaying a discrepant response to injury were noradrenaline (positively correlated with ISS only in low hcDNA patients (rho = 0.49, P = 0.002) vs. no correlation) and Annexin V (positively correlated with ISS only in high hcDNA patients (rho = 0.47, P = 0.002) vs. no correlation).

Figure 1

Correlations between injury severity score and thrombin-antithrombin complex, activated partial thromboplastin time and international normalized ratio on admission in 80 trauma patients stratified according to histone-complexed DNA (hcDNA); high hcDNA (>median) vs. low hcDNA (≤median). Rho and P values are shown for correlations between ISS and the investigated variables in patients with high hcDNA (black circles, filled lines) or low hcDNA (white circles, dashed lines): (a) ISS vs. TAT (ng/ml), (b) ISS vs. APTT (sec) and (c) ISS vs. INR (ratio)

DISCUSSION

The main findings in the present study were that patients with high circulating levels of histone-complexed DNA fragments displayed evidence of coagulopathy with prolonged APTT, hyperfibrinolysis, platelet activation and reduced thrombin generation along with enhanced sympathoadrenal activation, inflammation and endothelial damage. Furthermore, only in patients with high circulating hcDNA did higher injury severity correlate with increases in the plasma based coagulation tests, INR and APTT.

The finding that the plasma level of hcDNA in trauma patients was associated with both clinical injury severity and biomarkers indicative of endothelial damage, inflammation and coagulopathy may not be surprising given that hcDNA is released from dying cells and hence expected to be considerably higher in the most severely injured patients. However, the recent finding that nucleic acids are potent activators of the contact phase pathway of blood coagulation[15] and of platelets[16] and that they influence the natural anticoagulant Protein C pathway[17] makes it tempting to speculate that high circulating hcDNA levels may directly contribute to coagulopathy in trauma patients. Despite the finding here of an association between high hcDNA and platelet activation (sCD40L, Annexin V), in alignment with previous findings,[16] thrombin generation as evaluated by PF1.2 was reduced [Table 1]. This finding is notable since it indicates that thrombin generation was inhibited by the natural anticoagulants or impaired due to platelet exhaustion.[2025] The levels of circulating natural anticoagulant factors was however comparable in patients with high vs. low hcDNA [Table 1] and hcDNA did not correlate with PC or APC.

Recently, Xu et al., (2009) reported that extracellular histones induce endothelial damage, organ failure and death in sepsis and that in vivo administration of histones induces endothelial damage, hemorrhage and thrombosis in mice.[26] Historically, extracellular histones and nucleic constituents have been proposed to exert anticoagulant effects[2728] because in vitro addition of these prolongs plasma-based clotting times.[172728] This is in accordance with our in vivo findings that patients with high plasma hcDNA displayed enhanced endothelial glycocalyx (syndecan-1) and cell (sTM) damage concurrent with reduced thrombin generation and increased APTT. Thus, whether extracellular histones and nucleic acids are deemed pro- or anticoagulant probably depends on the context in which their effects on hemostasis are evaluated i.e., plasma vs. whole blood vs. in vivo in a systems’ biology perspective that includes the endothelium.[12]

From a systems’ biology perspective, taking into account the possible opposite effects of extracellular nucleic constituents on the endothelial (solid) and blood (fluid) phase of the vascular system, we hypothesize that the hypocoagulability induced by these in plasma[172728] and in circulating blood from trauma patients (as observed here) may reflect an evolutionary adapted response that counterbalances the injury/shock/catecholamine-driven[110111314] and histone/nucleic acid-driven[26] endothelial activation and damage.[12] We infer that the overall aim of this response is to maintain blood flow in order to keep the progressively more procoagulant microvasculature open so tissue perfusion is prioritized above hemostasis.[12]

The results presented here are subject to the limitations inherent to observational studies and, thereby, do not allow independent evaluation of the cause-and-effect relationship suggested. Furthermore, the low number of subjects included in the present study increases the risk of introducing a Type II error e.g., the comparable blood transfusion requirement and 30-day mortality in patients with high vs. low hcDNA levels though patients with high hcDNA displayed a clinical and biomarker profile indicating worse outcome, emphasizing that the reported findings should be confirmed in a larger cohort of patients. The low number of patients also precludes us from identifying possible gender differences in hcDNA fragments secondary to injury. Finally, we did not analyze later blood samples and therefore cannot reveal any potential late downstream influence of high hcDNA levels on organ function, etc.

In conclusion, we found that excessive release of extracellular histones and nucleic acids seems to promote coagulopathy, inflammation and endothelial damage following trauma.