Abstract
Inflammatory mediators and hemostatic markers were evaluated in patients enrolled in a phase-2b study evaluating the safety and efficacy of recombinant thrombomodulin (ART-123) in patients with sepsis and suspected disseminated intravascular coagulation (DIC). In contrast to controls, patients with sepsis and suspected DIC showed an increase in the circulating levels of inflammatory and fibrinolytic markers. The levels of procalcitonin (PCT), interleukin 6 (IL-6), interleukin 10 (IL-10), anaphylatoxin C5a, plasminogen activator inhibitor 1 (PAI-1), and myeloperoxidase were higher in the patients with sepsis and suspected DIC, whereas protein C (PrC) exhibited a significant decrease. When the patients with overt and nonovert DIC were compared, the PrC level was lower, and PCT, PAI-1, IL-6, and IL-10 levels were higher in the patients with overt DIC. These results indicate that inflammation is elevated in sepsis and suspected DIC, and inflammation, impairment of fibrinolysis, and overconsumption of PrC may play a key role in the pathogenesis of DIC.
Keywords
Disseminated intravascular coagulation (DIC) occurs as a complication of many underlying diseases. The DIC is characterized by a systemic activation of the coagulation and fibrinolytic systems, which can lead to microvascular thrombosis causing organ failure. The ongoing activation of coagulation can exhaust platelets and coagulation factors that results in a hypercoagulable state and bleeding. The DIC is more commonly associated with gram-negative bacterial infections but can also occur with gram-positive infection and with systemic infections with other organisms. 1,2
The International Society of Thrombosis and Hemostasis classifies DIC into 2 categories, overt and nonovert. 3,4 Overt DIC is characterized by a decompensated hemostatic system, whereas nonovert DIC is characterized by a stressed but compensated hemostatic system. The International Society on Thrombosis and Hemostasis (ISTH) criteria for overt DIC consists of a 5-step diagnostic algorithm in which a number is assigned to each laboratory test depending on the severity of the abnormality found. A cumulative score of 5 or more indicates a diagnosis of DIC. 4
Sepsis is frequently associated with hemostatic changes. These abnormalities in the hemostastic system range from mild to severe activation. The DIC is a complication in 25% to 50% of the patients with sepsis. 5 Mortality from sepsis is 2-fold higher when it is complicated by DIC. 5
Inflammation and hemostasis are closely linked. Leukocytes release several mediators of inflammation, namely, interleukin 1 and 6 (IL-1) (IL-6), and tumor necrosis factors (TNFs), which act as initiators for DIC. 1,2 Inflammatory mediators also amplify thrombin generation, upregulate procoagulant pathways, downregulate physiologic anticoagulants, and suppress fibrinolysis, which lead to the loss of localized hemostasis and microvascular thrombosis. 2 The cytokines, which result from inflammation, trigger tissue factor overexpression on several cell types such as neutrophils, endothelial cells, and monocytes. 1,2,6–9 Inflammatory cytokines can activate platelets as well. 10,11
In addition, levels of coagulation inhibitors are altered by inflammatory cytokines. Antithrombin (AT) levels are reduced due to consumption, impaired hepatic synthesis, and degradation by neutrophil elastase. 5,8 In DIC, the levels of activated protein C (PrC) are also decreased due to consumption, impaired synthesis, decreased levels of protein S, and downregulation of thrombomodulin expression at the endothelial surface. 2,5–8 This downregulation of thrombomodulin is regulated by proinflammatory cytokines TNF-α and interleukin 1β (IL-1β). 5,6,8
At the same time, inflammatory cytokines are capable of stimulating the production of plasminogen activator inhibitor 1 (PAI-1), which results in simultaneous suppression of fibrinolysis along with coagulation activation. 5–8 An important mechanism by which coagulation proteases influence inflammation is through their interaction with protease-activated receptors (PARs). Binding of the tissue factor–factor VII complex to PAR 2 results in the upregulation of inflammation. 5
Similarly, cross talk between the anticoagulant and the inflammatory pathways has also been demonstrated. For example, AT acts as a mediator of inflammation by binding to neutrophils and stimulating cytokine expression. 1,5,7 In addition, activated PrC has been shown to inhibit endotoxin-induced cytokine release and leukocyte activation. 1,5,7 Furthermore, several coagulation proteases and inhibitors are capable of inducing signaling pathways by inducing specific cell receptors. 7
The purpose of this investigation was to study the level of hemostatic activation and inflammation in patients with sepsis and suspected DIC enrolled in the ART-123, phase 2b study. 12
Materials and Methods
Study Patients
This study evaluated markers of hemostatic activation and inflammation in patients who were enrolled in the randomized, double-blind, placebo-controlled phase 2b study to evaluate the safety and efficacy of recombinant human soluble thrombomodulin, ART-123, in patients with sepsis and suspected DIC. This study was conducted in compliance with the ethical principles of the Declaration of Helsinki and in compliance with all International Conference on Harmonization Good Clinical Practice Guidelines.
Adult patients (≥18 years) admitted to 1 of the 233 study centers in 17 countries with sepsis and suspected DIC, 12 identified locally at each center using a score based on platelet count and prothrombin time international normalized ratio derived from the ISTH DIC score, 13,14 were screened for inclusion. Overt DIC was diagnosed when the ISTH score was ≥5. Patients were excluded from the study if any of the following criteria were present: inability to obtain informed consent; presence of any disorder other than sepsis that could alter coagulation; recent history of significant bleeding or increased risk of bleeding (eg, surgery within 12 hours of screening); presence of a disorder requiring anticoagulation; use of drotrecogin α (activated) within the 24 hours prior to enrollment or intended use; use of anticoagulants, antiplatelet agents, antithrombotics, and thrombolytics within the 72 hours prior to study dosing, or intended use, with the exception of heparin locks/flushes or deep vein thrombosis prophylaxis; platelet count <20 000; life expectancy <90 days; or current use of any chemotherapy agent.
Baseline blood samples were taken from patients who were diagnosed with sepsis and suspected DIC at individual centers and frozen at −70°C. In all, 750 patients were randomized to the study, and 741 patients were administered with study medication. Of the 741 patients, 617 had blood samples taken as well as confirmed ISTH DIC score at baseline. Of the 617 patients, 98 were diagnosed with overt DIC according to the ISTH criteria, and the remaining patients were categorized as nonovert DIC. In addition to the patients from the randomized trial, normal healthy volunteers (n = 30; 15 male and 15 female) were also included as normal controls.
Sample Analysis
Citrate- and EDTA-anticoagulated whole blood was drawn from patients, and plasma was prepared and frozen at −70°C and shipped to the central laboratory, Covance (Princeton, New Jersey). Samples were then organized into complete patient sets and shipped to Loyola on dry ice for analyses. Plasma samples were analyzed using the commercially available enzyme-linked immunosorbent assay kits for Asserachrom PrC and PAI-1 (Stago, Parsippany, New Jersey), IL-6 and IL-10 (R&D Systems, Minneapolis, Minnesota), C5a anaphylatoxin (BD Sciences, San Diego, California), and myeloperoxidase (MPO; Assay Design, Ann Arbor, Michigan). In addition, functional assays for PrC (Stago, Parsippany, New Jersey), protein C inhibitor (PCI; Technoclone, Vienna, Austria), and procalcitonin (PCT; Brahm’s Diagnostics, Annapolis, Maryland) were also performed. Samples were analyzed in batches, and results were compared to healthy normal controls.
Boxplots of the results from the normal patients as well as the baseline results from the patients with overt and nonovert DIC were produced. The top and bottom of the boxes represent the 75% and 25% percentiles, and the “whiskers” extend from these to the most extreme nonoutlier result. Outliers were plotted individually. The mean is represented within the figures as a triangle. Due to the extreme skewness of the data, some boxplots have been truncated.
Results
Baseline samples from patients with sepsis and suspected DIC included in the ART-123 phase 2b clinical trial were analyzed for specific markers of hemostatic activation and inflammation. The summary statistics from the overt, nonovert, and normal healthy volunteers are shown on Table 1. The PCT has been shown to be a marker of sepsis. Figure 1 shows the PCT results from patients with overt DIC and nonovert DIC compared to normal controls. As can be seen, the PCT levels are markedly elevated in patients with overt DIC and nonovert DIC compared to the normal controls. The overt DIC group has higher levels with a wider spread compared to the nonovert DIC group. The boxplot for PCT shows a median value of 11.2 ng/mL (first quartile 4.1 ng/mL and third quartile 36.4 ng/mL) in the patients with overt DIC having sepsis and a median of 5.3 ng/mL (first quartile 1.3 ng/mL and third quartile 17.5 ng/mL) in the nonovert group, compared to a median value of 0.3 ng/mL (first quartile 0.19 ng/mL and third quartile 0.32 ng/mL) in the normal controls. Several patients demonstrated very high levels of PCT, which are indicative severe septic infections that could lead to septic shock, while others demonstrated moderate to slight elevations.
Baseline biomaker results by group; procalcitonin.
Summary Statistics of Patients With Suspected Disseminated Intravascular Coagulation.
Abbreviations: aPC, activated protein C; DIC, disseminated intravascular coagulation; hs-CRP, high-sensitive C-reactive protein; IL, interleukin; MPO, myeloperoxidase; PAI-1, plasminogen activator inhibitor 1; PCI, protein C inhibitor; prot, protein.
Figure 2A shows the results of the PrC antigen levels in these patient groups compared to normal healthy volunteers. The PrC antigen levels were decreased in patients with both nonovert and overt DIC in comparison to the normal controls. There were wide variations in PrC antigen levels in the patients with DIC. A majority of the patients with DIC demonstrated low levels; however, in the patients with overt DIC, lower levels were measured. The median PrC value was 31.0% Ag (first quartile 20.0% and third quartile 45.0%) for the patients with overt DIC, 47.0% Ag (first quartile 35.0% and third quartile 63.0%) for patients with nonovert DIC, and 87.5% Ag (first quartile 83.9% and third quartile 92.4%) for normal controls. Figure 2B shows the result of the functional PrC levels. Similar trends were observed. The median PrC functional level was 28.3% (first quartile 17.3% and third quartile 38.5%) in patients in the overt DIC group, 37.0% (first quartile 24.1% and third quartile 60.0%) in patients in the nonovert DIC group, and 91.8% (first quartile 82.0% and third quartile 107.4%) in normal healthy volunteers. Similar to the PrC antigen, the functional PrC was decreased in both the overt and the nonovert DIC groups, but the overt DIC showed lower levels.
A, Baseline biomaker results by group; protein C-Ag. B, Baseline biomaker results by group; functional protein C.
As shown in Figure 3, a much wider scatter of PCI in the patients with both overt DIC and nonovert DIC was noted than that of the normal healthy volunteers; however, the lowest levels were observed in the nonovert DIC group, and there was no much difference between the 3 groups. The median value for PCI in patients with overt DIC was 71.5.0% normal human plasma (NHP; first quartile 43.0% NHP and third quartile 121.0% NHP) compared to the patients with nonovert DIC with a median 58.0% NHP (first quartile 34.0% NHP and third quartile 104.0% NHP) and normal control group with a median of 81.80% NHP (first quartile 71.7% NHP and third quartile 95.1% NHP).
Baseline biomaker results by group; protein C inhibitor (PCI).
Figure 4 shows the PAI-1 levels. The PAI-1 levels were markedly elevated in the overt group and slightly elevated in the nonovert group, pointing to the inhibition of fibrinolysis in patients with sepsis and suspected DIC. The median for patients with overt DIC was 137.5 ng/mL (first quartile 67.0 ng/mL and third quartile 294 ng/mL) compared to the patients with nonovert DIC with a median of 76.0 ng/mL (first quartile 45.0 ng/mL and third quartile 163.0 ng/mL) and normal controls with a median of 18.7 ng/mL (first quartile 10.8 ng/mL and third quartile 31.6 ng/mL). A 7.0-fold and 4-fold higher PAI-1 levels were measured in patients with overt DIC and in the nonovert group, respectively, compared to the normal controls.
Baseline biomaker results by group; plasminogen activator inhibitor 1 (PAI-1) antigen.
Several markers of inflammation including IL-6, IL-10, MPO, and C5a were also analyzed in the plasma of both the overt and the nonovert DIC groups and normal healthy controls. Figures 5 and 6 show the results of IL-6 and IL-10. Both IL-6 and IL-10 were elevated in patients with overt and nonovert DIC in comparison to the normal controls. The levels of the interleukins in the overt DIC patients were much higher than those in the nonovert patients. Both IL-6 and IL-10 were elevated in patients with overt and nonovert DIC in comparison to the normal controls. The levels of the lls in the overt DIC patients were much higher than those in the nonovert patients. Although both the markers of inflammation were elevated, the elevations in IL-6 were higher than IL-10. For IL-6 the median levels were 374.4 pg/mL (first quartile 84.7 pg/mL and third quartile 2242.2 pg/mL), 174.4 pg/mL (first quartile 45.1pg/mL and third quartile 568.3 pg/mL), and 6.1pg/mL (first quartile 4.7 pg/mL and third quartile 8.9 pg/mL) for overt DIC, nonovert DIC, and normal controls, respectively. For IL-10 the corresponding results were median 46 pg/mL (1st quartile 21.0 g/mL, 3rd quartile 152.0 pg/mL), median 31 pg/mL (1st quartile 18.0 g/mL, 3rd quartile 67.0 pg/mL) and median 3.5 pg/ml (1st quartile 0 pg/ml, 3rd quartile 18.0 pg/mL). Figure 7 shows the MPO levels in the patients with overt and nonovert DIC in comparison to the normal controls. Similar to the ILs, the median MPO levels in both the overt and the nonovert groups were markedly elevated in comparison to normal controls. Patients with overt DIC had a median of 131.6 ng/mL (first quartile 68.2 ng/mL and third quartile 226.8 ng/mL) compared to patients with nonovert DIC, 92.2 ng/mL (first quartile 55.1 ng/mL and third quartile 163.3 ng/mL), and the normal controls, 0.48 ng/mL (first quartile 0.5 ng/mL and third quartile 5.2 ng/mL). Complement activation was measured in terms of C5a levels as shown in Figure 8. The C5a is a potent mediator of inflammation, which was 6-fold as higher in the patients with overt and 4-fold higher in the patients with nonovert DIC compared to the normal controls. The median for the patients with overt DIC was 19.5 ng/mL (first quartile 9 ng/mL and third quartile 45.0 ng/mL) compared to the patients with nonovert DIC with a median of 12.0 ng/mL (first quartile 8 ng/mL and third quartile 35.0 ng/mL) and normal controls with a median of 3.4 ng/mL (first quartile 2.5 ng/mL and third quartile 4.0 ng/mL). The median value was higher in patients in the overt group compared to the patients in then onovert group. The elevation of all 4 of these inflammatory markers is indicative of a strong inflammatory component in sepsis and suspected DIC.
Baseline biomaker results by group; interleukin 6 (IL-6).
Baseline biomaker results by group; interleukin 10 (IL-10).
Baseline biomaker results by group; myeloperoxidase (MPO).
Baseline biomaker results by group; C5a.
Discussion
This study investigated the level of hemostatic activation and inflammation in patients who were enrolled in a randomized, double-blind, placebo-controlled phase 2b study to evaluate the safety and efficacy of recombinant human soluble thrombomodulin, ART-123, in patients with sepsis and suspected DIC. The baseline characteristics of this patient population have been previously published. 12
There have been many reports on the abnormalities that occur in coagulation and fibrinolytic system in patients with DIC. 1–3 These range from subtle activation to strong activation and are patient dependent. These abnormalities can lead to either microvascular thrombosis and/or bleeding. 5–8 The microvascular thrombosis in these patients can result in multiple organ failure. 13 Bleeding results from the ongoing coagulation activation and the resulting consumption of the coagulation proteins. When there is activation of the coagulation system and inflammation, the risk of thrombosis and organ failures as observed in sepsis is higher. Levi et al has reviewed the literature on organ failure and mortality in patients with sepsis with and without DIC. 1,7 The literature suggests a correlation between severity of DIC and mortality in sepsis. 1,2,7 There is also increasing evidence of the interactions between coagulation system and inflammation; inflammation promotes coagulation while at the same time coagulation affects inflammation.
In sepsis, cytokines play a major role in the imbalance of the coagulation system. Therefore, the role of markers of coagulation activation, fibrinolysis, and inflammation were evaluated in this well-defined group of patients with sepsis and suspected DIC. Patients with both overt DIC and nonovert DIC were compared to normal healthy volunteers. Several of the markers were selected on the basis of their defined role in the mediation of sepsis and suspected DIC. The patient group included in this study demonstrated elevated levels of PCT, PAI-1, C5a, IL-6, IL-10, and MPO, whereas PrC levels were decreased. The PCI was similar between the overt and the nonovert groups. In the patient group with overt DIC , the levels of PCT, PAI-1, IL-6, and IL-10 were much higher than that of both the nonovert and the normal controls. In addition, the PrC antigen and functional levels were decreased in a larger extent in the overt group. The upregulation of the inflammatory cytokines as observed in this patient group contributes to the decreased levels of PrC. Cytokines may downregulate the production of endothelial thrombomodulin; therefore, there is insufficient levels of thrombomodulin to complex with thrombin to activate PrC. 15 In addition, PrC levels are low due to consumption and impaired synthesis. 15 Activation of fibrinolysis and upregulation of inflammatory mediators may induce the increased level of PAI-1 in these patients with sepsis having suspected DIC, especially in patients with overt DIC.
The PCT is a known marker of the systemic inflammatory response to infection. 16,17 In comparison to the control group, patients with sepsis and suspected DIC exhibited a marked elevation in the PCT levels indicating that a majority of these patients underwent a systemic inflammatory response due to sepsis.
These results clearly indicate that inflammation, coagulation activation, and impairment of fibrinolysis play a key role in the pathogenesis of sepsis and suspected DIC. The results of these studies also suggest that profiling of the mediators of inflammation and hemostatic aberration may be useful in the diagnostic and prognostic management of sepsis in patients with both overt and nonovert DIC.
Footnotes
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: K. Tsuruta, lnder Kaul and Yutaka Osawa are all employees of Asahi Kasei Pharma America Corporation, of which former body, Artisan Pharma Inc., sponsored the phase 2b study of ART-123. Joe Hirman has served as consultants to Asahi Kasei Pharma America Corporation.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This biomarker study was also partly funded by Artisan Pharma Inc.