Disseminated intravascular coagulation (DIC) profoundly increases the morbidity and mortality of patients who have sepsis. Both laboratory and clinical research advanced the understanding of the biology and pathophysiology of DIC. This, in turn, gave rise to improved therapies and patient outcomes. Beginning with a stimulus causing disruption of vascular integrity, cytokines and chemokines cause activation of systemic coagulation and inflammation. Seemingly paradoxically, the interplay between coagulation and inflammation also inhibits endogenous anticoagulants, fibrinolytics, and antiinflammatory pathways. The earliest documented and best-studied microbial cause of DIC is the lipopolysaccharide endotoxin of Gram-negative bacteria. Extensive microvascular thrombi emerge in the systemic vasculature due to dysregulation of coagulation. The result of this unrestrained, widespread small vessel thromboses multiorgan system failure. Consumption of platelets and coagulation factors during this process can lead to an elevated risk of hemorrhage. The management of these patients with simultaneous hemorrhage and thrombosis is complex and challenging. Definitive treatment of DIC, and attenuation of end-organ damage, requires control of the inciting cause. Currently, activated protein C is the only approved therapy in the United States for sepsis complicated by DIC. Further research is needed in this area to improve clinical outcomes for patients with sepsis. Clin Trans Sci 2012; Volume 5: 85–92
Acute disseminated intravascular coagulation (DIC) is an acquired activation of coagulation, initiated by the presence of an unrelenting, pro-thrombotic stimulus. Recognition of DIC denotes an urgent need to identify the underlying cause.1–3 DIC can be an associated complication of sepsis, trauma, obstetric emergencies, solid tumors, and acute leukemias.4 The clinical presentation is closely aligned with the primary cause. Chronic DIC, associated with malignancy, is more likely to present with thrombosis, and is reviewed in other articles.5,6
Patients with DIC and septic shock exhibit a constellation of clinicopathologic findings including: thrombocytopenia, prolonged coagulation times, consumption of fibrinogen over time, and microangiopathic hemolytic anemia. In addition to vigorous acceleration of the coagulation cascade, there is also a downregulation of endogenous anticoagulant pathways and inhibition of fibrinolysis. Because of this dysregulation of hemostasis, inappropriate fibrin deposition occurs in small blood vessels. Microvascular thrombi are found in histologie examination of tissues in multiorgan system failure.7
This review will discuss the pathogenesis of DIC in sepsis, and current best practices for its management.
The incidence of sepsis in the United States is greater than 500,000 patients per year.1 DIC is diagnosed in approximately 35% of all patients with severe sepsis,2,3 and DIC severity directly correlates with morbidity and mortality.8 Mortality is estimated to be 40% in sepsis patients with DIC, as compared to 27% in sepsis patients without DIC.9,10 DIC associated with sepsis has a higher mortality rate than trauma-associated DIC.11 Infection leading to septic shock includes both Gram-positive and Gram-negative organisms.12,13 Identification of the microbial source is not required for the diagnosis of sepsis.
Hemostasis is a meticulously orchestrated system for maintenance of vascular integrity. Initiation of local hemostasis at the site of endovascular injury simultaneously induces auto-inhibitory feedback mechanisms, and coordination of fibrinolysis to prevent catastrophic thrombosis or hemorrhage.
The first event in hemostasis is the formation of a platelet plug at the site of blood vessel injury14 Integrins on the surface of platelets interact with endothelial cells and each other via von Willebrand factor in adhesion and fibrinogen in aggregation.15 However, platelet-mediated plug formation alone is not sufficient for stable hemostasis. This requires the additional contribution by the coagulation cascade to generate a fibrin meshwork that links adjacent platelets, and provides stabilization and structural support to the blood clot.16,17
The Coagulation Cascade
Historically, hemostasis was understood to comprise two distinct and simultaneous pathways of coagulation factor activation, termed the “intrinsic” and “extrinsic” pathways. A more contemporary model is that hemostasis involves both an initiation phase (Figure 1A) and an amplification phase (Figure IB).
All physiologic clotting originates with tissue factor (TF). The complex of TF and coagulation Factor Vila, (FVIIa)18 directly converts FX to FXa in the initiation phase. However, this initial step is short-lived because of the rapid inhibition of TF-FVIIa by the TF pathway inhibitor (TFPI). An alternate pathway, also activated by TF-FVIIa, but not inhibited by TFPI, serves as the amplification pathway of coagulation. Factor IX (FIX) is converted to FIXa, and along with FVIIIa, converts FX to FXa.
FXa next forms a complex with Factor Va (FVa), platelet phospholipid surfaces, and calcium that converts prothrombin into thrombin. In turn, thrombin then cleaves fibrinogen into fibrin. Activated Factor XIII cross-links fibrin thrombi, which provides additional stabilization. Fibrin clots are ultimately degraded by another protease called plasmin.
Pathogenesis of DIC
Any disruption in the balance of hemostatic forces can tip the equilibrium toward uncontrolled bleeding or thrombosis. In Intensive Care Unit patients, this disruption is almost always due to an infection or cancer, although crush trauma is another cause. These disorders that instigate DIC results in acceleration of the clotting cascade, downregulation of endogenous anticoagulants, and impaired fibrinolysis. The net effect is microthrombi formation in the systemic vasculature.
The most extensive studied cause of DIC is Gram-negative sepsis. In this disorder, bacterial lipopolysaccharide (LPS) endotoxin binds directly to the CD14/TLR4/MD2 receptor complex on monocytes and macrophages.19 This can induce several signaling cascades that involve either phosphatidylinositol-3-kinase (PI3K) or a variety of mitogen-activated protein kinases (MAPKs), such as ERK1/2, p38, or JNK1/2.20 These signaling pathways contribute to the production of DIC by activating transcription factors that lead to the expression of TF on the cell surface of these mononuclear cells. Alternatively, LPS-induced activation of these innate immunity cells can also lead to the production of cytokines such as tumor necrosis factor-α (TNF-α) or interleukin-6 (IL-6). In turn, these cytokines stimulate endothelial cells or other monocytes to express TF. Regardless of whether the effect of LPS is direct, or indirect and requires the production of cytokines, it leads to the intravascular exposure of TF that initiates inappropriate coagulation. The events that lead to DIC in patients with Gram-positive or viral infections, or in cancer patients are less well characterized. But, these disorders probably also induce DIC by the production of TNF-α or IL-6, and the induction of intravascular TF expression on the surface of endothelial and mononuclear cells.
Rapid acceleration of the clotting cascade causes brisk consumption of platelets and coagulation factors, which exceeds the repletion ability of the bone marrow and the liver, respectively.7 As a result, concomitant hemorrhage can be a complicating issue.
Accelerated Thrombus Formation via Tissue Factor
Under physiologic conditions, expression of TF is tightly regulated21 and limited to sub endothelial surfaces. In DIC, TF is exposed on cytokine-activated endothelial cells, due to changes in permeability or direct cellular disruption.22 However, the predominant location for TF expression in sepsis is on the surface of peripheral monocytes.18,23–25 The importance of TF exposure in the development of DIC has been evidenced by experiments which showed that inhibition of TF by a monoclonal antibody, lessened the intensity of DIC in a baboon model of gram-negative sepsis.26,27
Impaired Function of Endogenous Anticoagulants
The endogenous anticoagulants and their biologic targets are listed in Table 1. The three principle inhibitory regulators of coagulation are TFPI, protein C, and antithrombin, previously termed antithrombin III (AT). Their function is impaired in DIC through rapid consumption during accelerated coagulation, impaired hepatic synthesis, removal via the reticuloendothelial system, and leakage from capillaries3,7,28 Poor prognosis in patients with sepsis is associated with low antithrombin levels,29,30 low protein C levels,31 and elevated plasminogen-activator-inhibitor-1 levels (PAI-1).32
Table 1. Endogenous regulatory proteins of coagulation and their targets.
TFPI is found in endothelial cells, megakaryocytes, monocytes, vascular smooth muscle cells, and in platelets. Knockout of the TFPI gene in mice is a lethal mutation, and heterozygote mice exhibit a phenotype with recurrent thromboses.33 TFPI mRNA transcription in peripheral blood monocytes increases upon exposure of these cells to LPS and localizes to the cells’ surface.34 Despite an increase in TFPI as an acute phase reactant, its activity is overwhelmed in DIC. This raises the possibility that giving exogenous TFPI would be of benefit.
After forming a complex with thrombomodulin, thrombin is capable of cleaving Protein C thereby generating Activated Protein C. This process is facilitated 5- to 20-fold by the endothelial cell surface receptor.35,36 Under normal physiologic conditions, activated protein C curbs the maintenance phase of coagulation by catabolizing FVIIIa and FVa. The presence of the inflammatory cytokines TNF-α and IL-Iβ can both lead to downregulation of the expression of thrombomodulin on endothelial cells and protein S.37,38 Because the concentration of these inflammatory cytokines is increased during sepsis, the anti-thrombotic functions of protein C are largely impaired during overwhelming infections. Conversely, activated protein C inhibits TNF-α and pro-inflammatory interleukins. Heterozygous protein C deficiency exacerbates sepsis-induced multiorgan system failure in mice.39 Activated protein C was shown to regulate pro-coagulant protein synthesis in peripheral monocytes from healthy volunteers during stress.40
In the presence of heparin, AT binds irreversibly to thrombin, which leads to hepatic clearance. AT also directly inhibits Factor Xa. Laboratory studies suggest that AT can also regulate TF-FVIIa, and Factor IXa, although the biologic significance of these studies remains to be established.41,42 AT facilitates the release of prostacyclin (PGI2) from endothelial cells, inhibits NF-?β in monocytes and endothelial cells,43 and inactivates neutrophils’ ability to release pro-inflammatory cytokines.44In vitro, AT prevents chemotaxis, TF expression, and pro-inflammatory cytokine release from macrophages.44 Thus during DIC, impaired AT function limits its ability to act as an antiinflammatory mediator.
Cytokine-Mediated Inflammation and Coagulation Pathways
The bidirectional influence45–47 among coagulation and inflammation, existing in an exquisitely regulated balance under normal physiologic conditions, is skewed toward pro-coagulant and pro-inflammatory signals during DIC. Primate models of sepsis48 have shown that initial release of IL-6, followed by TNF-α and IL-149 are the prevailing inducers of DIC and subsequent impairment of fibrinolysis via up-regulation of PAI-1. Several pro-inflammatory cytokines, including IL-6, TNF-α, IL-1, IL-2, and IL-12, induce mononuclear cells to express TF (Figure 1), which commences coagulation.
Activated platelets provide the required phospholipid surface on which the maintenance phase of coagulation is propagated. Thus, endotoxin-induced or IL-6-induced platelet activation provides an additional cross-link between systemic inflammatory responses and accelerated coagulation.37 Conversely, thrombin accelerates IL-1 production from macrophages, and thrombin, FXa, and fibrin can all induce the production of IL-6 and IL-8,50 which enables a perpetual positive feedback loop of sustained inflammation and coagulation.
The predominant antiinflammatory cytokines, IL-4, IL-10, IL-13, IL-1-receptor antagonist (IL-1ra), and soluble TNF receptors exist in response to pro-inflammatory signals from TNF-α, to act as negative regulators of an otherwise sustained inflammatory response.51 Their own regulation, in turn, is also tightly controlled, so as to avoid a tipped balance toward systemic immunosuppression, particularly during active infection. In DIC, the compensatory response of antiinflammatory cytokines is insufficient to overcome the widespread inflammatory response within multiple organs, due to variation in the endothelial and leukocyte milieu of various tissue compartments.52
Deficient fibrinolysis causes persistence and extension of thrombosis in DIC. Levels of PAI-1 are upregulated in DIC, which causes marked suppression of plasmin activity.3,7 Bacterial endotoxin given to healthy volunteers caused an increase in tissue plasminogen activator within one hour and a subsequent rise in alpha 1-plasminogen inhibitor by 3 hours and plasminogen-activator inhibitor-1 by 5 hours.53 The net result is impaired fibrinolysis. Mice without normal plasmin generation, as produced by plasminogen deficiency or defective plasminogen activators, develop widespread microvascular thrombosis, whereas knockout mice with the PAI-1 gene do not generate thromboses under conditions of sepsis. In cardiac arrest patients, prolonged hypoperfusion resulted in high levels of PAI-1, low t-PA, and impaired fibrinolysis, which was reversed when spontaneous circulation returned.54 A proposed mechanism for the development of acute respiratory distress syndrome (ARDS) in septic patients is hypoxia-induced-PAI-1 up-regulation, which causes fibrin deposition in the lungs.55
Clinical Presentation and Laboratory Abnormalities
Patients with DIC and sepsis exhibit multiorgan system failure. Overt deep vein thrombosis is often not present. Symptoms of hemorrhage are characterized by “oozing” from central lines and venipuncture sites, or mucocutaneous bleeding.
A constellation of laboratory abnormalities, rather than one single diagnostic finding, is observed in DIC.3 Serial measurements of diagnostic laboratory studies are often more useful than a single measurement. The most common finding is thrombocytopenia, which correlates in magnitude with thrombin production, and is a surrogate marker for DIC severity15,56 The risk of bleeding increases by four to five times when the platelet count falls below 50,000 × 109/L.56 The true platelet count in DIC may be overestimated by automated counters, which is due to nonplatelet particles, or size alterations in activated platelets.57
The thrombocytopenia of DIC in patients with sepsis must be distinguished from other common causes of thrombocytopenia in critically ill patients, particularly thrombotic thrombocytopenic purpura, which requires emergent plasmapheresis.58 Alternate explanations for thrombocytopenia and the discriminating characteristics are listed in Table 2.
Table 2. The differential diagnosis of thrombocytopenia in critically ill patients.
Alternate explanation for thrombocytopenia
Distinguishing features from DIC
Offending drug will be present, and thrombocytopenia resolves on withdrawal, coagulation studies will be normal.
Timing of thrombocytopenia 4–10 days after heparin exposure, fall in platelet count by 50%, nadir platelet count ∼50,000. No microangiopathic hemolytic anemia.
Thrombotic thrombocytopenia purpura
Fever, and neurologic symptoms will be present. Normal coagulation studies.
Acute or chronic liver disease
Factor VIII level normal or high, presence of portal hypertension, and stigmata of liver failure.
Patient is pregnant, hypertensive, and abnormalities resolve after delivery.
Thrombocytopenia in DIC is multifactorial and includes: platelet activation and consumption, clearance by an overstimulated reticuloendothelial system,59 and underproduction of platelets within the bone marrow.15 The thrombopoetin-enhanced stimulus of megakaryocyte proliferation is offset by accelerated phagocytosis of megakaryocytes by activated macrophages.60
The prothrombin time and activated partial thromboplastin time is prolonged in approximately 50% of patients with DIC, due to coagulation factor consumption, impaired hepatic production, or Vitamin K deficiency.56 Because fibrinogen is an acute phase reactant, it may be elevated rather than decreased as one would anticipate in a patient undergoing consumption of coagulation factor. Therefore, a single laboratory value can be misleading. Serial measurements may demonstrate a decrease in the concentration of plasma fibrinogen if the consumption is brisk.
Fibrin degradation products (FDPs) and the D-dimer are two markers of fibrinolysis that indicate thrombus formation in DIC. The D-dimer measures only the breakdown of fibrin that has been crosslinked by Factor XIII, whereas FDPs measure the breakdown of either fibrinogen or fibrin (Figure 2). Although the D-dimer has a higher specificity for diagnosing DIC than FDPs, the overall diagnostic specificity for either test is poor.1,3,56
Examination of the peripheral blood smear may reveal schizothymes to corroborate the presence of fibrin clots causing microangiopathic hemolytic anemia. However, this finding is only observed in 25–50% of patients with DIC.1,3 The anemia associated with DIC is usually not severe.
Guidelines for Diagnosis of DIC
The International Society on Thrombosis and Haemostasis (ISTH) scoring algorithm is summarized in Table 3.61 The reported per-patient sensitivity was 93% with a specificity of 98%.62 The platelet count and prothrombin time are particularly useful in predicting severity of illness and risk of mortality.63,64 This model offers the practical advantage of routine laboratory tests, without the need for specialized reference laboratories.
Table 3. The ISTH scoring algorithm. A score of ≥5 is indicative of overt DIC.
Elevated fibrin-related marker
Prolonged prothrombin time
In 2003, the ISTH criteria was compared to an historical Japanese criteria, which found a concordance rate of only 67.4% for patients with overt DIC.65 As a result, an updated guideline from the Japanese Association for Acute Medicine (JAAM) was proposed,66 and was prospectively validated.67 (Table 4)
Table 4. The JAAM scoring system for DIC. Four points or more is indicative of DIC.
*SIRS, systemic inflammatory response syndrome (temperature <36°C or >38°C, heart rate >90, respiratory rate >20, white blood cell count <4 × 109 cells/L or >12 × 109 cells/L or 10% band forms).
Platelet count (×109)
≥80 and <120 or >30% decrease within 24 hours
<80 or >50% decrease within 24 hours
Prothrombin time ratio (patient: normal)
Fibrin/fibrin degredation products (mg/L)
≥10 and <25
The immediate management of DIC is to identify and treat the underlying cause.56,68 Adjunct therapies have been proposed for use, in addition to supportive care of patients with sepsis. There are no randomized, controlled clinical trials that demonstrate a benefit from treatment-dose heparin,3,69 unless there is clinically significant deep vein or arterial thrombosis, purpura fulminans, or acral ischemia. However, the use of heparin at prophylactic doses is recommended in DIC treatment guidelines based on expert opinion in the United Kingdom and Europe.56 Heparins improved laboratory biomarkers of thrombosis in healthy volunteers70 and have shown a trend towards improvement in 28-day mortality, when compared with placebo in patients receiving other therapy for DIC.71 These data suggest that the use of prophylactic heparin does not worsen outcomes and may lessen the risk of deep venous thrombosis in critically ill patients with multiple concomitant risk factors for thrombosis.
In the absence of bleeding, repletion of platelets and coagulation factors is not necessary,3,56 as control and cessation of the underlying stimulus will usually correct the laboratory abnormalities. Patients with a platelet count above 10–20 × 109/L have an acceptably low risk of spontaneous hemorrhage, such that prophylactic platelet transfusions are not recommended.56 The increase in platelet count with allogeneic transfusion may not be as robust as expected, due to profound deficiency from consumption.
In the United States, cryoprecipitate is the recommended product for replacement of fibrinogen, with a target goal of 1 g/L.56 Worldwide, fibrinogen concentrate is also considered acceptable first-line therapy. Antifibrinolytic agents, such as aminocaproic acid are not recommended, as these agents would exacerbate the already-impaired fibrinolytic system.3,56 There is insufficient evidence to recommend the use of recombinant, Factor VIIa; because there are no randomized, controlled trials demonstrating its efficacy or safety.72
Replacement of Endogenous Anticoagulants
To date, manipulation of cytokines and their receptors as a treatment for patients with DIC and sepsis is not used as a therapeutic modality.73,74 The only FDA-approved therapy for DIC in patients with sepsis is Activated Protein C (rhAPC, drotrecogin alfa (activated), Xigris; Eli Lilly and Company, Indianapolis, Indiana, USA).75 In an animal model of sepsis,76 and in an early human Phase II study, it was found that rhAPC impeded organ toxicity, coagulopathy, and death.77 The PROWESS study,9 a Phase III, randomized, controlled double-blind, multicenter trial, reached its primary outcome of survival at 28 days. All-cause mortality for patients receiving rhAPC was 24.7% as compared to a mortality rate of 30.8% observed in the placebo-treated patients (p= 0.005). There was a nonsignificant trend to develop more bleeding over 28 days in patients receiving rhAPC, as compared to placebo (3.5% of rhAPC patients vs. 2.0% of control patients, p= 0.06). Notably, patients with platelet counts of <30,000 × 109/L were excluded from the study. There was no reported increase in the risk of thrombosis, new infection, or antibody formation.
Several secondary analyses followed. There was faster resolution of cardiovascular and respiratory failure in a 7-day period after infusion for all patients who received rhAPC.78 It was also shown that the greatest benefit was derived in patients with the most severe multiorgan system failure.79 Subset analyses of patients with the lowest scores on critical illness severity indices revealed that there was no mortality benefit to receiving rhAPC for this population of patients, and that there was a significantly increased risk of bleeding.80
In 2005, a pédiatric trial using rhAPC in the treatment of sepsis was halted, due to the projected lack of efficacy (organ failure resolution in 14 days) and increased rates of intracerebral hemorrhage, particularly for the youngest patients. Accordingly, rhAPC is only approved for use in adults. An international, prospective observational study identified demographic variables of patients who were most likely to receive, and benefit from, rhAPC.81 Practice patterns indicate that rhAPC is used most often in patients who are younger, have the most severe illness parameters, and have the fewest comorbid conditions. It appears that rhAPC is safest and most efficacious for this select patient population.
Early experiments of AT administration in animals with sepsis documented an improved morbidity and mortality.82–84 These observations led to the Kyber-Sept trial,85 an international, multicenter, double-blind, phase III study in humans. The study included adults with severe sepsis, and patients with a platelet count less than 30,000/μL were excluded. Patients were randomized to receive supraphysiologic doses of AT or placebo, with a primary endpoint of 28-day mortality. The mortality rate for patients receiving AT was 38.9%, as compared to a mortality rate of 38.7% for patients receiving placebo (p = 0.94). In a subgroup analysis of all patients receiving concomitant heparin prophylaxis, the addition of AT resulted in a bleeding rate of 23.8%, as compared to a bleeding rate of 13.5% for patients receiving placebo. For patients who did not receive heparin and who met ISTH criteria for overt DIC, there was a statistically significant improvement in the 90-day mortality rate with the addition of AT.86
Follow-up studies demonstrated that levels of thrombin and IL-6 decreased in healthy volunteers given an LPS infusion plus recombinant human AT.87 In a study of DIC patients with burn injuries, administration of AT reduced mortality.88 Improvement in measurable clinical outcomes for patients with sepsis remains to be determined. AT is not approved in the United States, but is used routinely in Japan.
Administration of exogenous TFPI improved DIC-related mortality in animal models, and in human Phase I/II trials.89,90 However, these findings were not validated in the randomized, placebo-controlled OPTIMIST trial of recombinant TFPI (tifacogin).91 This study was an international, multicenter, double-blind, randomized, placebo-controlled trial. Adult patients with sepsis and an International Normalized Ratio (INR) > 1.2 were included in Stage I of the trial, based on Phase II data. Stage II, approximately 1 year later, recruited patients with an INR < 1.2. There was no improvement in the overall 28-day mortality rate when patients who received TPFI were compared with patients receiving placebo, and there was a significant increase in overall bleeding. The bleeding risk did not decrease for patients in Stage II, with lower overall INR measurements.
Future Treatment Directions and Final Comments
Our current understanding of DIC can be heralded as a prototypical success story for translational science. First, laboratory research elucidated the complex regulatory mechanisms of physiologic coagulation at the molecular and cellular level. This, in turn, has directed clinical advances in the diagnosis and management of critically ill patients with DIC. The presence of accelerated microvascular coagulation requires prompt identification and treatment of the inciting cause.
Novel therapies, aimed at curtailing the hématologic and thrombotic complications of DIC, include restoration of anticoagulant and fibrinolytic pathways. Sepsis prevention initiatives, such as the Early Goal-Directed Therapy92 algorithm, appear to convey the most promise in improving overall survival outcomes. For intensivists and hematologists alike, timely recognition of DIC is the key to bedside management.
Conflict of Interest
Neither of the authors have any financial interest to disclose.