SEARCH

SEARCH BY CITATION

Keywords:

  • coagulopathy;
  • endothelial glycocalyx;
  • platelets;
  • sCD40L;
  • sympathoadrenal activation;
  • trauma

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References

Summary.  Background: Severe injury activates the sympathoadrenal, hemostatic and inflammatory systems, but a maladapted response may contribute to a poor outcome. Soluble CD40L is a platelet-derived mediator that links inflammation, hemostasis and vascular dysfunction. Objectives: To investigate the association between the sCD40L level and tissue injury, shock, coagulopathy and mortality in trauma patients. Methods: A prospective, observational study of 80 trauma patients admitted to a Level I Trauma Center. Data on demography, biochemistry, Injury Severity Score (ISS) and 30-day mortality were recorded and admission plasma/serum analyzed for sCD40L and biomarkers reflecting sympathoadrenal activation (adrenaline, noradrenaline), tissue/endothelial cell/glycocalyx damage (histone-complexed DNA fragments [hcDNA], Annexin V, thrombomodulin and syndecan-1), coagulation activation/inhibition (PF1.2, TAT-complex, antithrombin, protein C, activated protein C, sEPCR, TFPI, von Willebrand factor [VWF], fibrinogen and factor [F] XIII), fibrinolysis (D-dimer, tissue plasminogen activator [tPA] and plasminogen activator inhibitor-1 [PAI-1]) and inflammation (interleukin-6 [IL-6] and sC5b-9). We compared patients stratified by median sCD40L level and investigated predictive values of sCD40L for mortality. Results: High circulating sCD40L was associated with enhanced tissue and endothelial damage (ISS, hcDNA, Annexin V, syndecan-1 and sTM), shock (pH, standard base excess), sympathoadrenal activation (adrenaline) and coagulopathy evidenced by reduced thrombin generation (PF1.2), hyperfibrinolysis (D-dimer), increased activated partial thromboplastin time (APTT) and inflammation (IL-6) (all P < 0.05). A higher ISS (P = 0.017), adrenaline (P = 0.049) and platelet count (P = 0.012) and lower pH (P = 0.002) were associated with higher sCD40L by multivariate linear regression analysis. High circulating sCD40L (odds ratio [OR] 1.84 [95% CI 1.05–3.23], P = 0.034), high age (P = 0.002) and low Glasgow Coma Score (GCS) pre-hospital (P = 0.002) were independent predictors of increased mortality. Conclusions: High early sCD40L levels in trauma patients reflect tissue injury, shock, coagulopathy and sympathoadrenal activation and predict mortality. As sCD40L has pro-inflammatory activity and activates the endothelium, sCD40L may be involved in trauma-induced endothelial damage and coagulopathy.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References

Trauma is a major cause of death and disability worldwide [1] and although hemorrhage accounts for most early trauma-related deaths [2,3], mortality in patients who survive their initial injuries is often as a result of complications, such as infections and (multiple) organ failure, driven by an overwhelming and/or dysregulated inflammatory response [4–6]. Given the interactions at multiple levels between inflammation, coagulation and sympathoadrenal activation [7–10], an early maladapted response to injury in any of these systems may contribute to the development of late post-injury complications and mortality [11–14].

Recently, a great deal attention has been given to the soluble CD40 ligand (sCD40L), a platelet-derived mediator with proinflammatory, procoagulant and immunomodulatory function, which serves as a link between inflammation, hemostasis and vascular dysfunction [15–20]. CD40L is an activation antigen on platelets that is expressed within seconds after activation [15,21]. It is a glycoprotein belonging to the tumor necrosis factor (TNF) family with its receptor, CD40, being a member of the TNF receptor superfamily [17]. Interaction between platelet CD40L and endothelial CD40 elicits an inflammatory reaction of the endothelium [15,19] whereas CD40L/CD40 (cis) interaction on activated platelets [21] or soluble metalloproteinases released from activated platelets [17] cause cleavage of platelet CD40L with release of sCD40L [17,21]. Thus, most (95%) of the circulating sCD40L is of platelet origin [15,18]. The released sCD40L exerts procoagulant activity by being a GPIIb/IIIa ligand and platelet agonist that, through GPIIb/IIIa-mediated platelet activation, enhances microparticle formation, promotes fibrinogen binding to GPIIb/IIIa and stabilizes platelet aggregation and thrombus formation [16,22]. In addition, sCD40L exerts pro-inflammatory activity, including platelet–leukocyte conjugation, through its interaction with CD40 expressed on a broad range of immune cells [17–19,23,24]. Thus, through downstream effects on platelets, endothelial and immune cells, sCD40L may promote inflammation, endothelial cell dysfunction and vascular destabilization [17–19]. This notion is supported by the finding that circulating levels of sCD40L are associated with a poor outcome in diseases with combined inflammatory and vascular pathology [17–19,23,24], including sepsis [25]. No previous studies have, to our knowledge, investigated the level of sCD40L early after trauma.

The aim of the present study was to investigate the early association between the sCD40L level and markers of tissue and endothelial damage, shock, sympathoadrenal activation, coagulopathy, inflammation and mortality in trauma patients. We hypothesized that high circulating sCD40L post-injury would predict a poor outcome through potentially damaging effects on the endothelium and hence a potentially enhanced trauma-induced hit on the vascular system [26].

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References

Study design

A prospective, observational cohort study of trauma patients admitted directly to a Level I Trauma Center (TC) at a tertiary hospital (Rigshospitalet, Copenhagen, Denmark, covering 2.5 million inhabitants) between March 2010 and November 2010. The present study is part of an ongoing larger multicenter study, Activation of Coagulation and Inflammation after Trauma 3 (ACIT3), approved by the Regional Ethics Committee (H-4-2009-139), the Danish Data Protection Agency and conducted in accordance with the 2nd Declaration of Helsinki. Written informed consent was obtained from the patients or their next of kin. In the present study, we report on the findings related to a cohort of 80 patients recruited to the ACIT3 study, from whom extended blood sampling was performed.

Patient selection

ACIT3 study inclusions: adult trauma patients (≥ 18 years) who met the criteria for full trauma team activation and had an arterial cannula inserted. The latter was chosen since only patients with expected severe injuries have an arterial cannula placed immediately upon TC admission. Exclusion criteria were arrival in the TC > 2 h after injury; > 2000 mL of intravenous fluids administered before hospital arrival; and transfer from another hospital or burns > 5% total body surface area. Patients were retrospectively excluded if they were taking anticoagulant/antiplatelet medications (except aspirin), had moderate or severe liver disease or had known bleeding diathesis.

The 80 included patients were selected from the first 100 patients recruited to the ACIT3 study with complete data. We intended to include 80 patients because we measured an extensive number of biomarkers by ELISA, with each ELISA kit providing analysis of 80 samples. We aimed at including the most severely injured and/or coagulopathic patients and selected the 80 patients according to: outcome (mortality or intensive care unit [ICU] admission post trauma; yes), transfusion of red blood cells within 6 h (yes), revised trauma score (RTS) (< 5.00, this score was used because we did not have access to the Injury Severity Score [ISS] before or later on) or coagulopathy (activated partial thromboplastin time [APTT] ≥ 35 s, International Normalized Ratio [INR] ≥ 1.2, Ly30 > 1%/Cl30 < 95%; yes). This yielded 70 severely injured/coagulopathic patients, and an additional 10 patients (age 48 years interquartile range [IQR 43–52], 60% males) were selected blinded from the remaining 30 patients to match their age and gender (see Table 1 for details on demography, injury severity etc.). The 20 patients not included in the present study, had, compared with the included patients, comparable age and gender (41 years [IQR 33–53], 40% males) and APTT (26 [IQR 23–27], NS) but had, as expected, lower ISS (4 [IQR 2–10], P < 0.001), mortality (0%, P = 0.037) and INR (1.1 [IQR 1.0–1.1], P = 0.007).

Table 1.   Demography, injury severity, biochemistry, hemostasis, transfusion requirements and mortality in 80 trauma patients admitted directly to a Level I Trauma Center at a tertiary hospital (Rigshospitalet, Copenhagen, Denmark) and included as part of a prospective Multicenter study, Activation of Coagulation and Inflammation after Trauma 3 (ACIT3)
  Patients
  1. Data are presented as medians (IQR) or n (%). ISS, injury severity score; sTBI, severe Traumatic Brain Injury, Abbreviated Injury Score head > 3; PH, pre-hospital at the site of injury; GCS, Glascow Coma Score scale; RBC, red blood cells; MT, massive transfusion > 10 RBC the initial 24 h; APTT, activated partial thromboplastin time; INR, international normalized ratio.

N 80
AgeYears46 (33–64)
Genderm%68% (54)
Blunt trauma% (n)91% (73)
ISSScore17 (10–28)
sTBI% (n)31% (22)
GCS pre-hospitalScore13 (6–15)
pH 7.34 (7.29–7.39)
SBEmmol L−1−2.0 (−4.0–0.0)
Lactatemmol L−11.7 (1.2–2.7)
SatO2 pre-hospital%98 (93–100)
Shockindex pre-hospitalHR/SBP0.62 (0.50–0.75)
Hemoglobinmmol L−18.4 (7.3–9)
Platelet count109 L−1208 (173–253)
APTT > 35 s%8% (6)
INR > 1.2%13% (10)
Saline pre-hospitalmL350 (0–1000)
MT (> 10 RBC in 24 h)% (n)14% (11)
Mortality% (n)18% (14)

Data on demography, clinical and biochemical parameters, investigations, management and 30-day mortality were recorded and ISS scores were obtained from the Trauma Audit & Research Network (TARN) database.

No patients received tranexamic acid, adrenaline or noradrenaline prior to blood sampling.

Blood sampling

Blood was sampled immediately upon arrival for standard arterial blood gas (ABG, Radiometer ABL 725/735, Copenhagen, Denmark), routine biochemistry and research analyzes (citrate, heparin, EDTA plasma and serum). Routine biochemistry samples were analyzed in a DS/EN ISO 15189 standardized laboratory by a Sysmex XE-2100 (hemoglobin, platelets and leukocytes) and ACL TOP (APTT, INR, antithrombin [AT] and fibrinogen). Plasma samples were ice-cooled immediately whereas serum samples were kept at room temperature for 1 h before centrifugation (one [serum] or two [plasma] times 1800 g at 5 °C for 10 min) and storage at −80 °C.

ELISA measurements

Soluble biomarkers of sympathoadrenal activation, tissue, endothelial cell and glycocalyx damage, coagulation activation/inhibition, fibrinolysis, platelet activation and inflammation were measured in uniplicate using commercially available immunoassays according to the manufacturer’s instructions. In each patient, all 21 biomarkers were measured corresponding to a total of 21 × 80 = 1680 measurements, with only three missing measurements. The biomarkers were analyzed in EDTA/citrate plasma or serum as follows: EDTA plasma: adrenaline and noradrenaline (2-CAT ELISA; Labor Diagnostica Nord GmbH & Co. KG, Nordhorn, Germany; lower limit of detection [LLD]) 11 pg mL−1 (adrenaline, normal reference < 100 pg mL−1) and 44 pg mL−1 (noradrenaline, normal reference < 600 pg mL−1), respectively. Histone-complexed DNA fragments (hcDNA, Cell Death Detection ELISAPLUS; Roche, Hvidovre, Denmark; LLD not stated, relative quantification); Annexin V (American Diagnostica Inc., Stamford, CT, USA; LLD not stated, normal reference < 10 ng mL−1); soluble thrombomodulin (sTM) (Nordic Biosite, Copenhagen, Denmark; LLD 0.38 ng mL−1); D-dimer (ADI; LLD 2–4 ng mL−1); and sCD40L (R&D Systems Europe, Abingdon, UK; LLD 4.2 pg mL−1). Citrate plasma: protein C (PC, Helena Laboratories, Beaumont, TX, USA; LLD 5% of reference plasma); activated protein C (APC, USCNLIFE; LLD 4.2 pg mL−1); soluble endothelial protein C receptor (sEPCR; R&D Systems Europe; LLD 0.064 ng mL−1); protein S (PS, ADI; LLD not stated, quantified relative to provided reference plasma); tissue-type plasminogen activator (tPA, ADI, detects sc-tPA, tc-tPA and tPA/PAI-1 complexes; LLD 1 ng mL−1); plasminogen activator inhibitor-1 (PAI-1, Assaypro, Sct. Charles, MO, USA; LLD 0.2 ng mL−1); prothrombinfragment 1 and 2 (PF1.2, USCNLIFE; LLD 0.043 nmol L−1); (TAT, USCNLIFE; LLD 0.215 ng mL−1); tissue factor pathway inhibitor (TFPI, ADI, detects intact TFPI, truncated TFPI, TF/FVIIa/TFPI complexes; LLD 0.18 ng mL−1); von Willebrand factor antigen (VWF, Helena Laboratories, LLD 5% of reference plasma); factor (F) XIII (Assaypro; LLD 50 pg mL−1); terminal complement complex (sC5b-9, MicroVue sC5b-9 plus EIA Kit; Quidel Corp., San Diego, CA, USA; LLD 3.7 ng mL−1); and interleukin-6 (IL-6, Quantikine HS; R&D Systems Europe; LLD 0.039 pg mL−1). Serum: Syndecan-1 (Diaclone SAS, Besancon, France; LLD 2.56 ng mL−1).

Statistics

Statistical analysis was performed using SAS 9.1 (SAS Institute Inc., Cary, NC, USA). Data from patients stratified according to sCD40L level (> median vs. ≤ median) were compared using Wilcoxon’s Rank Sum tests and χ2/Fisher’s exact tests, as appropriate. Correlations were investigated using Spearman’s correlations and presented by rho and P-values. The contribution of platelets, tissue injury, shock, endothelial damage and inflammation (variables expected to contribute to sCD40L) to the variation in circulating sCD40L was investigated using univariate and multivariate linear regression analysis, presented by regression coefficients β with (standard errors), t- and P-values and R2 for the multivariate model. Platelet count was included in the multivariate model in spite of it being only borderline significant in the univariate model (P = 0.074) because 95% of circulating sCD40L is reported to be of platelet origin [15,18].

The predictive value of sCD40L for 30-day mortality was investigated using univariate logistic regression analysis and by backwards multivariate logistic regression analysis after adjusting for variables with known prognostic value for trauma mortality [27–29]: age, ISS, pre-hospital GCS, pH and APTT. Furthermore, all studied biomarkers were investigated for their univariate predictive value for 30-day mortality using logistic regression, and significant univariate variables were confronted with the multivariate model (including only the significant values to avoid over fitting). Results are presented by odds ratio (OR) (95% CI), χ2 and P-values. Data are presented as medians with IQR. P-values < 0.05 were considered significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References

Patients

The present study included 80 trauma patients with ISS in the entire range (median 17 [IQR 10–28]; ISS > 26 n = 23, 15–26 n = 26 and < 15 n = 30), the majority suffering from blunt trauma and approximately one-third had severe traumatic brain injury (sTBI, Abbreviated Injury Score head > 3) (Table 1). Most patients (96%) were referred by mobile emergency care units (MECU) staffed with anesthetists (28% by helicopter) and blood samples were drawn a median of 68 min (IQR 48–88) after injury. Twelve patients (15%) had increased APTT and/or INR, 14% received a massive transfusion (> 10 RBC the initial 24 h) and overall 30-day mortality was 18% (Table 1).

Biomarker profile in patients with high vs. low sCD40L

When comparing patients stratified according to median sCD40L, patients with a high sCD40L demonstrated a higher degree of ISS and tissue injury (hcDNA and Annexin V), shock (lower pH, SBE) and endothelial cell, including glycocalyx [30], damage (higher sTM, syndecan-1) as well as increased sympathoadrenal activation (higher adrenaline and glucose) and blood cell mobilization (higher white blood cell [WBC], neutrophils, monocytes and platelets) (Table 2). Furthermore, patients with a high sCD40L displayed evidence of coagulopathy, hyperfibrinolysis and inflammation with higher median APTT, lower PF1.2 and higher D-dimer and IL-6 (Table 2). Patients with a high sCD40L also tended to have a higher 30-day mortality (P = 0.078, Table 2).

Table 2.   Demography, injury severity, transfusion requirements, mortality, sympathoadrenal activation, biochemistry, hemostasis and biomarkers of coagulopathy in 80 trauma patients stratified according to sCD40L (high sCD40L > median, low sCD40L ≤ median)
  High sCD40LLow sCD40LP-values
  1. Data are presented as medians (IQR) or n (%), with P-values shown for variables with P < 0.200, and in bold for P < 0.050. Groups were compared using Wilcoxon’s Rank Sum tests or χ2/Fisher’s exact tests, as appropriate. ISS, injury severity score; sTBI, severe Traumatic Brain Injury, Abbreviated Injury Score head > 3; GCS, Glascow Coma Score scale; RBC, red blood cells; MT, > 10 RBC the initial 24 h; APTT, activated partial thromboplastin time; INR, international normalized ratio. Biomarker abbreviations, see Materials and Methods section, ELISA.

Demography
 N 4040 
 AgeYears42 (30–67)48 (39–63)ns
 Genderm%68% (27)68% (27)ns
 Blunt trauma% (n)25% (10)13% (5)ns
 ISSScore25 (17–32)13 (8–24)< 0.001
 sTBI% (n)26% (10)38% (12)ns
 GCS (pre-hospital)Score12 (5–15)13 (7–15)ns
 SatO2 (pre-hospital)%98 (92–100)99 (95–100)ns
 Shock index (pre-hospital)HR/SBP0.63 (0.49–0.77)0.61 (0.53–0.72)ns
 MT (> 10 RBC in 24 h)% (n)15% (6)13% (5)ns
 Mortality% (n)25% (10)10% (4)0.078
Sympathoadrenal activation, biochemistry and hemostasis
 Adrenalinepg mL−1965 (262–1496)191 (85–361)0.001
 Noradrenalinepg mL−1716 (310–1360)354 (207–1259)0.183
 pH 7.31 (7.25–7.37)7.36 (7.34–7.41)0.001
 Lactatemmol L−12.0 (1.2–2.9)1.6 (1.0–2.6)ns
 Glucosemmol L−19.1 (7.8–11.8)7.1 (6.3–8.8)< 0.001
 SBEmmol L−1−2.95 (−5.03–−0.93)−1.70 (−2.98–0.60)0.017
 Hemoglobinmmol L−18.5 (7.1–9.0)8.4 (7.7–9.1)ns
 Platelet count109 L−1232.5 (187.75–273.5)187.5 (168–218.25)0.002
 WBC109 L−116.4 (12.1–21.9)9.8 (8.1–13.2)< 0.001
 Fibrinogeng L−12.3 (1.8–2.7)2.5 (2.2–2.9)0.070
 FXIIIμg mL−130 (21–37)29 (24–37)ns
 VWF%192 (114–219)204 (145–230)0.137
 APTTs26 (24–29)25 (23–26)0.013
 INRRatio1.1 (1.1–1.2)1.1 (1.1–1.1)ns
Tissue, endothelial cell and glycocalyx damage
 Histone complexed DNA%10.6 (4.3–23.5)2.2 (0.0–9.7)< 0.001
 Annexin Vng mL−137.2 (23.6–47.0)23.2 (18.5–29.0)< 0.001
 sTMng mL−12.8 (1.2–4.1)1.1 (0.8–2.7)0.001
 Syndecan-1ng mL−142.9 (29.5–79.1)27.5 (14.7–35.0)< 0.001
Platelet activation and thrombin generation
 sCD40Lpg mL−1418 (336–557)226 (169–261)NA
 PF1.2nmol L−13.4 (1.0–17.3)9.3 (3.6–18.8)0.014
 TATng mL−136.5 (30.5–41.4)36.8 (30.9–43.4)ns
Natural anticoagulation
 AT103 U L−10.91 (0.82–1.01)0.93 (0.82–1.03)ns
 PC%102 (88–125)110 (98–124)ns
 APCng mL−19.7 (8.0–12.5)10.0 (7.7–12.0)ns
 sEPCRng mL−1210 (171–268)252 (170–432)ns
 PS%65 (59–70)65 (61–72)ns
 TFPIng mL−164 (46–80)58 (46–77)ns
Fibrinolysis
 D-dimerng mL−1173 (164–176)134 (72–170)< 0.001
 tPAng mL−17.2 (4.7–13.1)5.7 (3.6–11.3)ns
 PAIng mL−124 (11–41)22 (16–39)ns
Inflammation
 sC5b-9ng mL−11059 (904–1257)1025 (884–1189)ns
 IL-6pg mL−1104 (68–122)30 (10–88)< 0.001

Correlations between sCD40L, injury severity, blood cells and biomarkers

In alignment with the differences between patients with high vs. low sCD40L, sCD40L correlated moderately with the degree of tissue injury (ISS [Fig. 1A], hcDNA [rho = 0.43, P < 0.001], Annexin V [Fig. 1D]), shock (pH [Fig. 1C], SBE [rho = −0.27, P = 0.020]) and endothelial and glycocalyx damage (sTM [rho = 0.39, P < 0.001], syndecan-1 [Fig. 1E]), sympathoadrenal activation (adrenaline [Fig. 1B] and noradrenaline [rho = 0.25, P = 0.027]), blood cell mobilization (platelet count [Fig. 1H] and leukocyte count [rho = 0.34, P = 0.002]), inflammation (IL-6, Fig. 1G) and hyperfibrinolysis and coagulopathy (D-dimer [Fig. 1F], APTT [rho = 0.25, P = 0.030]).

image

Figure 1.  Correlations between soluble CD40L (sCD40L) and markers of injury severity, shock, endothelial damage, hyperfibrinolysis, inflammation and platelet count on admission in 80 trauma patients. Rho and P-values are shown for correlations between sCD40L (pg mL−1) and each of the investigated variables: (A) sCD40L vs. Injury Severity Score (ISS), (B) sCD40L vs. adrenaline (pg mL−1), (C) sCD40L vs. pH, (D) sCD40L vs. Annexin V (ng mL−1), (E) sCD40L vs. Syndecan-1 (ng mL−1), (F) sCD40L vs. D-dimer (ng mL−1), (G) IL-6 (pg mL−1) and (H) sCD40L vs. platelet count (× 109 L−1).

Download figure to PowerPoint

In order to identify variables associated independently with sCD40L, we performed linear regression analyzes with sCD40L as a dependent variable and explanatory variables reflecting the platelet level (platelet count), tissue injury (ISS, hcDNA), shock (pH, adrenaline), endothelial glycocalyx damage (syndecan-1) and inflammation (IL-6), based on the a priori hypothesis that these variables/(pathophysiologic) conditions could result in platelet activation and hence sCD40L release. Significant univariate variables were ISS, hcDNA, pH, adrenaline, syndecan-1 and IL-6 whereas platelet count was only borderline significant (P = 0.074) (Table 3). When these variables were included in a multivariate regression model (platelet count was kept in the multivariate model as platelets are the main source of circulating sCD40L [15,18]), the variables independently associated with higher sCD40L were higher platelet count, ISS, circulating adrenaline and lower pH (similar results for a full and a backward model) (Table 3).

Table 3.   Univariate and multivariate linear regression analysis of variables associated with sCD40L in trauma patients upon admission to a Level I Trauma Center
 UnivariateMultivariate (R2 = 0.51)
 Unitβ (SE)t-valueP-valueβ (SE)t-valueP-value
  1. Regression coefficients (β) with standard errors (SE), t- and P-values and R2 displayed for the multivariate model. P-values are shown in bold for variables with P < 0.05. Predicted change in sCD40L (pg mL−1) associated with one unit increase in platelet count, injury severity score (ISS), histone-complexed DNA fragments (hcDNA), circulating adrenaline, pH, syndecan-1 and IL-6.

Platelet count10 × 109 L−17.47 (4.13)1.80.0748.96 (3.46)2.60.012
ISSPoint8.11 (1.70)4.8< 0.0016.29 (2.56)2.50.017
hcDNA%3.85 (1.36)2.80.006−1.20 (1.69)−0.7ns
Adrenalinepg mL−10.08 (0.02)5.3< 0.0010.04 (0.02)2.00.049
pH0.1 U−116.4 (22.3)−5.2< 0.001−76.1 (23.5)−3.20.002
Syndecan-1ng mL−12.15 (0.56)3.9< 0.0010.64 (0.63)1.0ns
IL-6pg mL−11.36 (0.45)3.00.003−0.74 (0.53)−1.4ns

Response to injury in patients with high vs. low sCD40L

We recently reported that the biomarker response to increasing ISS differed markedly in trauma patients stratified according to high vs. low circulating syndecan-1 [28]. Given the potentially hazardous effects of sCD40L on the endothelium [17–19], we investigated discrepancies in the hemostatic and biomarker response to increasing ISS (correlations between ISS and biomarkers) in patients with high vs. low sCD40L. Only in patients with high sCD40L, did higher ISS correlate with protein C consumption (protein C, Fig. 2A), endothelial cell damage (sTM, Fig. 2C) and coagulopathy (INR, Fig. 2F). Notably, ISS did not correlate with PF1.2 or TAT in patients with high sCD40L whereas ISS correlated positively with both PF1.2 and TAT in patients with low sCD40L (Fig. 2D–E). ISS correlated positively with and tended to correlate with APC in patients with low vs. high sCD40L, respectively (Fig. 2B). Furthermore, only in patients with low sCD40L did ISS correlate with adrenaline (rho = 0.34, P = 0.034), hemoglobin (rho = −0.31, P = 0.049), platelets (rho = −0.42, P = 0.007), neutrophils (rho = 0.37, P = 0.019) and D-dimer (rho = 0.65, P < 0.001) probably reflecting a considerably increased level of several of these variables (adrenaline, platelets, neutrophils [data not shown], D-dimer) in patients with high sCD40L (Table 2). In both groups did higher ISS correlate positively with syndecan-1, hcDNA and IL-6 and negatively with antithrombin, fibrinogen and FXIII (data not shown).

image

Figure 2.  Correlations between Injury Severity Score (ISS) and protein C, activated protein C, soluble thrombomodulin, prothrombinfragment 1 and 2 (PF1.2), thrombin–antithrombin complex (TAT) and international normalized ratio (INR) on admission in 80 trauma patients stratified according to sCD40L [(high sCD40L > median) vs. low sCD40L (≤ median age)]. Rho and P-values are shown for correlations between ISS and the investigated variables in patients with high sCD40L (black circles, filled lines) or low sCD40L (white circles, dashed lines): (A) ISS vs. protein C (%), (B) ISS vs. activated protein C (ng mL−1), (C) ISS vs. soluble thrombomodulin (ng mL−1), (D) ISS vs. PF1.2 (nmol L−1), (E) ISS vs. TAT (ng mL−1) and (F) ISS vs. INR (ratio).

Download figure to PowerPoint

Mortality and sCD40L

To investigate the association between sCD40L and mortality in trauma patients, variables with previously reported prognostic value [27–29] were tested using univariate and multivariate logistic regression analysis with 30-day mortality as a main endpoint. In the unadjusted analysis, higher age, ISS, pre-hospital GCS, pH, APTT and sCD40L were all associated with increased 30-day mortality (Table 4) but after including these variables in a multivariate model, only higher age, pre-hospital GCS and sCD40L remained associated, statistically independently, with increased 30-day mortality (Table 4). The strong predictive value of CGS pre-hospital may in part be related to the fact that trauma patients with severe traumatic brain injuries (sTBI) tended to have a higher mortality as compared with non-sTBI patients (32% vs. 14% 30-day mortality in sTBI vs. non-sTBI patients, P = 0.086). It should, however, be noted that the level of sCD40L was similar in patients with or without sTBI (P = 0.423).

Table 4.   Univariate and multivariate predictors of 30-day mortality in trauma patients upon admission to a Level I Trauma Center
 UnitUnivariate OR (95% CI)χ2P-valueMultivariate OR (95% CI)χ2P-value
  1. Odds ratios with 95% confidence intervals [OR (95% CI)], χ2 and P-values are shown for all univariate variables (only P-values < 0.20 are shown for multivariate variables), with P-values in bold for variables with P < 0.05. OR (95% CI) associated with a one unit increase in the specified variables.

AgeYears1.04 (1.01–1.07)5.50.0191.10 (1.04–1.17)9.60.002
ISSScore1.11 (1.04–1.18)10.40.001-2.3ns
GCS (pre-hospital)Score0.77 (0.66–0.89)11.9< 0.0010.59 (0.42–0.82)9.70.002
pH0.1 U0.52 (0.28–0.96)4.40.036-0.2ns
APTTS1.13 (1.03–1.24)6.20.013-1.6ns
sCD40L100 pg mL−11.46 (1.07–2.00)5.60.0181.84 (1.05–3.23)4.50.034

The predictive value of the other investigated biomarkers was also investigated using univariate logistic regression with 30-day mortality as the main endpoint. The only biomarkers besides sCD40L that could predict 30-day mortality were (OR given for one unit increase) adrenaline (ng mL−1) (OR 1.74 [95% CI 1.18–2.57], P = 0.005); Protein C (%) (OR 0.97 [95% CI 0.94–0.99], P = 0.007) and sEPCR (ng mL−1) (OR 0.99 [95% CI 0.99–1.00], P = 0.036). Furthermore, sTM (ng mL−1) (OR 1.32 [95% CI 0.99–1.77], P = 0.058), D-dimer (ng mL−1) (OR 1.03 [95% CI 1.00–1.07], P = 0.069) and IL-6 (OR 1.01 [95% CI 1.00–1.03], P = 0.083) were borderline significantly associated with mortality. None of the significant/borderline biomarkers had a predictive value for mortality when confronted with the multivariate model (included age, pre-hospital GCS and sCD40L) and of the significant variables only adrenaline was capable of changing the significance level of sCD40L in this model (both sCD40L and adrenaline became non-significant).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References

The main finding of the present study was that high levels of sCD40L in trauma patients at hospital admission was associated with enhanced tissue and endothelial damage, shock and sympathoadrenal activation along with reduced thrombin generation, hyperfibrinolysis and increased APTT. Furthermore, higher ISS, adrenaline level and platelet count and lower pH were statistically independently associated with higher sCD40L. High sCD40L along with high age and low GCS pre-hospital were independent predictors of increased 30-day mortality.

Trauma patients with a high sCD40L displayed a clinical and biomarker profile indicating more severe injuries, tissue and endothelial damage, shock and sympathoadrenal activation along with coagulopathy. As 95% of circulating sCD40L is of platelet origin [15,18] this finding indicates that excessive platelet activation and degranulation is part of the early response to severe trauma. Importantly, in spite of a higher platelet count, thrombin generation was considerably reduced in patients with a high sCD40L, which may reflect a combination of enhanced natural anticoagulation (increased protein C activation [31–33], antithrombin activity or endogenous heparinization through glycocalyx degradation, high syndecan-1 [26,28,29]), coagulation factor depletion [34,35] and platelet exhaustion [36–38].

Considering platelet exhaustion, the finding in the present study that adrenaline, injury severity and shock were all statistically independently associated with sCD40L even after adjusting for platelet count, indicates that excessive tissue injury, shock and an exaggerated trauma-induced catecholamine surge may all contribute to extensive platelet mobilization and activation [39–41], and hence enhanced release of sCD40L. Given that sCD40L, like adrenaline [39,42–44], is a potent autocrine and paracrine platelet agonist [16,21,22], enhanced release of sCD40L may further promote platelet activation and ensuing exhaustion. This notion is in line with our recent finding that factors associated with Thrombelastography clot strength in trauma patients at admission (platelet count, fibrinogen and FXIII) change with injury severity as both platelet count and fibrinogen level predict clot strength in less severely injured patients (ISS ≤ 26) whereas only fibrinogen level predicts clot strength in the most severely injured patients (ISS > 26) [37]. Importantly, when including the sCD40L level in the model, sCD40L was negatively correlated with clot strength only in patients with the highest ISS (> 26) and adjusting for sCD40L made platelet count significantly positively correlated with clot strength also in these patients, indicating that sCD40L may be a surrogate for platelet exhaustion after excessive platelet activation. The notion that platelet count alone not simply reflects platelet activation may also explain the finding in the present study of a non-significant correlation between platelet count and sCD40L in the univariate regression analysis that however became significant after adjusting for adrenaline, injury severity and shock, variables and/or conditions that may induce platelet activation. This finding thus emphasizes that there may be a poor correlation between platelet number and overall platelet activation and/or function (assessed here as sCD40L) in severely injured and shocked patients.

Also considering platelet exhaustion, there is emerging evidence that platelets, beyond their role in hemostasis, are critical players in the immune continuum, with functions resembling that of innate immune cells [45–47]. In this regard we speculate that early excessive platelet activation and ensuing exhaustion may ultimately contribute to poor outcome in trauma patients through loss of platelet-mediated immune function and regulation [45], in line with the recent finding that even a normal platelet count after severe trauma may be insufficient as the probability of death at 24 h decreases with increasing platelet count [48].

The finding in the present study of positive correlations between sCD40L and biomarkers of endothelial damage (sTM) and glycocalyx degradation (syndecan-1) after trauma is notable, given that sCD40L [15,19], similar to adrenaline [28,29,49,50], is one potential mediator of the downstream effects of tissue injury and shock [26] that may both activate and damage the vascular endothelium, including the glycocalyx [28,29]. Thus, we infer that the trauma-induced tissue injury, shock and catecholamine surge may, in addition to their direct negative effects on the endothelium, indirectly also contribute to endothelial cell damage and glycocalyx degradation, through exaggerated platelet activation and sCD40L release. These events may contribute to platelet exhaustion [36,38], coagulopathy and poor outcome [26,28,29], in alignment with the finding in the present study that the high sCD40L level was independently associated with increased mortality in trauma patients. However, it should be emphasized that the design of the present study does not allow any inferences about cause-effect relationships, so the hypotheses presented need to be tested in independent appropriately designed studies.

It may seem counterintuitive that injured and potentially bleeding trauma patients become progressively endogenously hypocoagulable in the fluid phase (circulating blood) of the vascular system with increasing injury severity [26,51], whether this is as a result of platelet exhaustion [36–38], excessive anticoagulation through e.g. protein C activation [31–33] or endogenous heparinization [26,28,29], factor consumption [34,35] or yet other factors. We recently hypothesized [26] that the progressive hypocoagulability observed in the circulating blood with increasing injury severity may reflect an evolutionary adapted response that counterbalances the potential procoagulant/detrimental effects of catecholamines on the solid phase (vascular endothelium) of the hemostatic system [28,29,49,50] so, from a systems biology perspective, the trauma-induced catecholamine surge may influence the endothelium and circulating blood in opposite directions with the overall aim to induce local hemostasis while preserving micro vascular perfusion and oxygen delivery [26]. In line with this notion, sCD40L [15,19] may be yet another factor contributing to the endothelial hit as well as to reduced hemostatic competence of the circulating blood [37], together with the tissue injury and shock. However, it should be emphasized that the results of the present study support the notion that CD40L levels reflect tissue injury, injury severity and shock as well as sympathoadrenal activation, and not one of these factors alone.

The discrepant biomarker response to increasing injury severity in patients with high vs. low sCD40L indicates that patients with high sCD40L were more likely to develop coagulopathy, with enhanced protein C consumption, endothelial damage and attenuated thrombin generation and ensuing increased INR. However, as patients with a high sCD40L also represented the most severely injured and shocked patients, the causality between high sCD40L and the development of coagulopathy must be questioned until this has been proven in a study designed to address this issue. Finally, it should be noted that in all patients, increasing injury severity was associated with evidence of glycocalyx degradation, inflammation and consumptive coagulopathy with low antithrombin, fibrinogen and FXIII, emphasizing that the extent of the injury, independent of the sCD40L level, either directly or indirectly promotes this (patho)physiologic response.

The results in the present study are subject to the limitations inherent to observational studies and, thereby, do not allow independent evaluation of a cause-and-effect relationship. Furthermore, as patients were selected to include the most severely injured patients, the findings cannot be generalized to a standard trauma population and should be confirmed in such. Finally, the low number of subjects and especially the limited number of severely injured patients and patients increases the risk of introducing a type II error, emphasizing that the findings should be confirmed in a larger cohort of patients, and this is currently underway.

In conclusion, the present study found high circulating levels of sCD40L upon admission in the most injured and shocked trauma patients, indicating that excessive platelet activation and degranulation is part of the early response to severe tissue injury. High sCD40L levels were associated with enhanced tissue and endothelial damage, shock, sympathoadrenal activation and coagulopathy and increased 30-day mortality. As sCD40L has pro-inflammatory activity and activates the endothelium, sCD40L may be involved in trauma-induced endothelial damage and coagulopathy.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References

The authors would like to thank K. Dyeremose and M. H. Andersson for their skilled technical assistance. The Danish Council for Independent Research (Medical Sciences), Aase and Ejnar Danielsens Foundation, L. F. Foghts Foundation, A. P. Møller and Chastine Mc-Kinney Møllers Foundation (Medical Sciences).

Disclosure of Conflict of Interest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References

The authors state that they have no conflict of interest.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosure of Conflict of Interest
  9. References