Disseminated intravascular coagulation (DIC) is a syndrome characterized by the systemic activation of blood coagulation, which generates intravascular fibrin, leading to thrombosis of small and medium-sized vessels, and eventually organ dysfunction 1. DIC may occur as a complication of infections, solid cancers, hematologic malignancies, obstetric diseases, trauma, aneurysm, and liver diseases, among others, and each type of DIC presents characteristic features related to the underlying disorder. The diagnosis and treatment of DIC must therefore take into consideration these underlying etiologic features. Patients with DIC resulting from sepsis, hematologic malignancy or obstetric disease can be successfully treated for DIC, whereas DIC associated with solid cancers may not respond to standard treatments 2-5. There were no significant differences in outcome between solid cancer patients with and without DIC 5. Three guidelines for DIC 2-4 have been published in the literature from the British Committee for Standards in Haematology (BCSH), the Japanese Society of Thrombosis and Hemostasis (JSTH), and the Italian Society for Thrombosis and Hemostasis (SISET). Although they are broadly similar, there are variations in the recommendations (Tables 1 and 2). This communication is an attempt to harmonize the three guidelines for DIC by the active members of the subcommittee for DIC of the Scientific and Standardization Committee (SSC) of the ISTH. The quality of the evidence and the definitions for recommendations were evaluated by use of a modified GRADE system 6 (Table 3), and agreed among all active members of the ISTH/SSC subcommittee for DIC between 1 July 2011 and 10 December 2012. Drugs that are frequently used in clinical practice were considered under the category of ‘Potential Recommendation’ in cases of disagreement between the three guidelines or insufficient randomized controlled trial (RCT) results. It needs to be stressed that this harmonization may not be appropriate for all patients, and individual patient circumstances may dictate the selection of alternative approaches. B1: treatment has moderately HQ of Ed, or it has HQ of Ed but the CU is not significant. B2: treatment does not have HQ of Ed, but it has few deleterious effects and it is carried out clinically Recommendations: The underlying clinical condition can have an influence on the laboratory tests that are usually performed to diagnose DIC. To facilitate the diagnostic process, scoring systems are recommended by the three different guidelines 3-5. Different diagnostic criteria have been established by the ISTH/SSC 1, the Japanese Ministry of Health, Labor and Welfare (JMHLW) 7, and the Japanese Association of Acute Medicine (JAAM) 8. The ISTH overt DIC score is useful for the diagnosis of DIC resulting from infective and non-infective etiologies 9, 10. The JMHLW score predicts the outcome in DIC 11, and the JAAM score in septic patients also correlates with the ISTH and JMHLW scores and outcomes 8, 10. A prospective study in Japan reported no significant difference in the odds ratio for the prediction of DIC outcomes for these three diagnostic criteria 12. A combination of tests repeated over time in a patient with suspected DIC can be used to diagnose the disorder with reasonable certainty in most cases 13-15. A template for the non-overt DIC scoring system, including global tests, changes in global tests, and hemostatic molecular markers, has been proposed 1, 16, 17. Screening assays (global coagulation tests), such as the prothrombin time (PT), fibrinogen, platelet count, and fibrin-related markers (FRMs), provide important evidence of the degree of coagulation factor activation and consumption. Although PT is prolonged in approximately 50–60% of cases of DIC at some point during the course of the disease 13, the abnormalities are often observed in patients with concomitant liver disease or warfarin treatment. A reduction in the platelet count or a clear downward trend in subsequent measurements is a sensitive (although not specific) sign of DIC 2. A reduced fibrinogen level is valuable for the diagnosis of DIC, but is not observed in most DIC patients 2. The elevated FRMs, such as fibrinogen and fibrin degradation products 18, D-dimers 19, and soluble fibrin (SF), reflect thrombin formation. SF 20 assays offer theoretical advantages in DIC, in that they give a closer reflection of thrombin action on fibrinogen. However, it is important to remember that many conditions, such as trauma, recent surgery, or venous thromboembolism (VTE), are associated with elevated FRMs. Reductions in the levels of natural anticoagulants such as antithrombin (AT) and protein C (PC) are common in DIC. Although measuring AT activity may be useful for achieving the full efficacy of heparin 21, it is not quickly and easily measured in all hospitals. A reduction in ADAMTS-13 activity and elevations of soluble thrombomodulin (TM), plasminogen activator inhibitor-1 (PAI-1) and von Willebrand factor propeptide are often observed in DIC, and these have been shown to have prognostic significance 22-24. An atypical light transmittance profile on the activated partial thromboplastin time (APTT) as the biphasic waveform has been associated with DIC, and appears to have a positive predictive value for DIC 25, 26. Although many attractive markers have been reported for DIC, there is no single marker that is able to diagnose DIC. Therefore, these three guidelines 2-4 recommend that DIC should not be diagnosed with a single marker, but with a combination of laboratory markers. Recommendation: The most important treatment for DIC is the specific and vigorous treatment of the underlying disorder, such as administration of antibiotics or surgical drainage. All three guidelines 3-5 agree on this point, although there is no high-quality evidence. The DIC will spontaneously resolve in many cases, when the underlying disorder is properly managed. However, some cases require additional supportive treatment, specifically aimed at the abnormality in the coagulation system. In an RCT of all-trans-retinoic acid (ATRA) as compared with conventional chemotherapy in patients with acute promyelocytic leukemia (APL), the mortality rate of APL was significantly lower in the ATRA group than in the conventional chemotherapy group (overall survival: 69% vs. 45%, respectively) 27. ATRA exerts not only differential effects on APL progression, but also anticoagulant and anti-fibrinolytic effects 28. Similarly, several RCTs on the treatment of sepsis have shown a parallel improvement in coagulation derangement and DIC, although data have not been always concordant 29-34. Recommendations: Low levels of platelets and coagulation factors may increase the risk of bleeding. However, blood component therapy should not be instituted on the basis of laboratory results alone, but reserved for those with active bleeding, requiring an invasive procedure, and who are otherwise at high risk for bleeding complications. The three guidelines 3-5 recommended the administration of platelet concentrate and FFP in patients with active bleeding or at high risk of bleeding. The threshold for transfusing platelets depends on the clinical state of the patient. In general, platelet transfusions are administered to patients who are actively bleeding and who have a platelet count of <50 × 109 L−1. A much lower threshold of 10–20 × 109 L−1 is adopted in non-bleeding patients, on the basis of RCTs in patients with thrombocytopenia following chemotherapy. Platelets may be administered at higher levels than this in patients perceived to be at high risk of bleeding on the basis of other clinical and laboratory features 35. It may be necessary to use large volumes of plasma to correct the coagulation defects shown by prolonged APTT or PT, or a decreased fibrinogen level. Initial doses of 15 mL kg−1 of FFP are suggested, although there is evidence that a dose of 30 mL kg−1 produces more complete correction of coagulation factor levels. In this regard, the consequences of volume overload may have to be considered. Smaller volumes of PCC may be useful in this setting, although these products lack certain essential coagulation factors, such as factor V. Specific deficiencies in fibrinogen can be corrected by administration of purified fibrinogen concentrates or cryoprecipitate. The response to component therapy should be monitored both clinically and by repeating platelet counts and coagulation tests following administration of these components. The efficacy and safety of recombinant FVIIa in DIC with life-threatening bleeding are unknown, and it should be used with caution, or as part of a clinical trial. Recommendations: There are several differences in the recommendations for heparin among the three guidelines (Table 1) 3-5. Anticoagulant treatment may be a rational approach based on the notion that DIC is characterized by extensive activation of coagulation. Although experimental studies have shown that heparin can at least partly inhibit the activation of coagulation in DIC 36, there are no RCTs demonstrating that the use of heparin in patients with DIC results in an improvement in clinically relevant outcomes. A small RCT showed that LMWH was better than UFH in the treatment of DIC (low quality)37. Patients with DIC are at high risk of VTE, and VTE prophylaxis with UFH, LMWH and/or mechanical methods has become the standard of care in patients with DIC (moderate and high quality, respectively) 38, 39. A recent large trial in patients with severe sepsis showed a non-significant benefit of low-dose heparin on the 28-day mortality, and underscored the importance of not stopping heparin in patients with DIC and abnormal coagulation parameters 21. Although it is not easy to quickly measure the AT activity level in all hospitals in order to decide whether to administer urgent heparin treatment, measuring this might be useful for achieving the full efficacy of heparin. Anti-FXa agents such as Fondaparinux are recommended for the prophylaxis of deep vein thrombosis after orthopedic surgery, but there is little evidence in critically ill patients. Danaparoid sodium has been used in Japan, but no RCTs have shown a reduction in the mortality rate or an improvement in the resolution rate in DIC. Synthetic protease inhibitors such as Gabexate mesilate and Nafamostat have been used and evaluated in Japan 9, 40, 41, but there are no RCTs showing a reduction in the mortality rate or an improvement in the resolution rate in DIC. Recommendations: The use of agents that are capable of restoring the dysfunctional anticoagulant pathways in patients with DIC has been extensively studied. The three guidelines 3-5 give different recommendations for the use of anticoagulant factor concentrates (Table 1). A large-scale multicenter RCT performed to directly assess the effect of AT concentrate on mortality in patients with severe sepsis showed no significant reduction in those treated with AT concentrate (high quality) 28. Interestingly, the subgroup of patients who had DIC and who did not receive heparin showed a remarkable survival benefit (moderate quality) 42, but this finding requires prospective validation. The clinical efficacy of recombinant human APC (rhAPC) in severe sepsis was demonstrated in a large RCT 29, but a prospective trial in septic patients with relatively low disease severity did not show any benefit of rhAPC 30. The withdrawal of rhAPC from sepsis treatment was proposed after an RCT for sepsis shock failed to show any benefit 31. Plasma-derived APC improved the outcome in a small RCT 43 in Japan, but it was not approved for DIC treatment. There are no useful RCTs on the use of PC concentrate for treating sepsis or DIC. An RCT comparison of rhTM with UFH 33 showed that rhTM significantly increased the resolution rate in DIC, but did not significantly decrease the mortality rate. In another RCT on severe sepsis, rhTM reduced the mortality and significantly reduced organ failure as compared with placebo 44. Another RCT in severe sepsis showed that a recombinant human tissue factor pathway inhibitor showed no significant benefit 34. Antifibrinolytic agents are effective in bleeding patients, but the use of these agents in patients with bleeding resulting from DIC is generally not recommended 45. An exception may be made in cases where hyperfibrinolysis dominates the clinical picture, as seen in the coagulopathy associated with acute promyelocytic leukemia (AML-M3) and in some cases of malignancy (e.g. prostate carcinoma). The three guidelines show some differences in the recommendations here (Table 1). A study in AML-M3 has shown a beneficial effect of antifibrinolytic agents in this situation (low quality) 46, but cases complicated with severe thrombosis owing to the combined use of ATRA and tranexamic acid have been documented 47. A recent RCT (moderate quality) 48 showed that tranexamic acid significantly reduced the mortality of patients with trauma. In these cases, antifibrinolytics should be administered in the early part of management, before the levels of PAI-1 and other endogenous antifibrinolytics are elevated. H. Wada was chairman and explained the JSTH guidelines. J. Thachil was co-chairman and explained the BCSH guidelines. M. Di Nisio explained the SISET guidelines and differences among three guidelines. P. Mathew examined the difference among three guidelines. S. Kurosawa was co-chairman and examined the differences among three guidelines. S. Gando was co-chairman and explained the JSTH guidelines. H.K. Kim was co-chairman and examined the diagnosis. J.D Nielsen was co-chairman and examined the treatments. C-E. Dempfle was co-chairman and examined the diagnosis. M. Levi explained the BCSH guidelines and examined the treatments. C.H Toh explained the BCSH guidelines and GRADE system. All members discussed to harmonize the three guidelines. The authors state that they have no conflict of interest.

Guidance for diagnosis and treatment of disseminated intravascular coagulation from harmonization of the recommendations from three guidelines