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 Disseminated Intravascular Coagulation

Jean-Louis Vincent, MD, PhD, FCCM; Mayakonda K. Ramesh, MS; David Ernest, MBBS; Steven P. LaRosa, MD; Jan Pachl, MD, PhD; Naoki Aikawa, MD, DMSc, FACS; Eric Hoste, MD, PhD; Howard Levy, MB, BCh, PhD; Joe Hirman, PhD; Marcel Levi, MD, PhD; Mradul Daga, MD, FCCP; Demetrios J. Kutsogiannis, MD, MHS; Mark Crowther, MD, MSc, FRCPC; Gordon R. Bernard, MD; Jacques Devriendt, MD; Joan Vidal Puigserver, MD; Daniel U. Blanzaco, MD; Charles T. Esmon, PhD; Joseph E. Parrillo, MD; Louis Guzzi, MD, FCCM; Seton J. Henderson, MB, ChB; Chaicharn Pothirat, MD, FCCP; Parthiv Mehta, MD; Jawed Fareed, PhD, FAHA; Deepak Talwar, MD, DM, DNB; Kazuhisa Tsuruta, PhD; Kenneth J. Gorelick, MD, FCCP; Yutaka Osawa, MPharm; Inder Kaul, MD, MPH


Crit Care Med. 2013;41(9):2069-2079. 

In This Article


In this phase 2b study, the administration of ART-123 when combined with standard care was associated with lower D-dimer, F1.2, and TATc levels compared with placebo in critically ill patients with sepsis and activation of coagulation as indicated by a modified DIC score. Importantly, the administration of ART-123 was not associated with excess bleeding, thromboembolic events, or increased development of new infections. The difference in 28-day mortality compared with standard care alone met the prespecified level of significance for suggestion of efficacy in this phase 2 trial, although it did not meet the target for proof of efficacy. This phase 2b study has, thus, demonstrated that ART-123 is a safe intervention in critically ill patients with coagulopathy due to sepsis; evidence suggestive of efficacy was observed, supporting further development.

Despite improvements in the general management of patients with sepsis over the past 30 years, 28-day mortality rates in sepsis remain unacceptably high, and no targeted therapy is currently available. The Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis (PROWESS) study demonstrated a survival benefit in patients with severe sepsis treated with recombinant activated protein C,[18] and post hoc analyses of these data suggested that patients with overt DIC (using a different modification of the ISTH score to that used in the present study) may have benefited in particular.[19] The recent PROWESS-SHOCK study (20), however, failed to confirm the survival benefit of recombinant activated protein C, possibly in part because of the lower placebo mortality rate compared with the original PROWESS study making it more difficult to demonstrate an effect of the trial agent, and this drug has now been withdrawn.

Thrombomodulin is a thrombin receptor present on the endothelial cell surface that is of key importance in regulating physiological anticoagulation.[21,22] Binding of thrombin to thrombomodulin not only scavenges active thrombin but also creates a complex that activates protein C, thereby effectively converting thrombin from a procoagulant to an anticoagulant. In sepsis, thrombomodulin expression and function are markedly down-regulated, contributing to the procoagulant state. Although serum levels of thrombomodulin increase in patients with sepsis and sepsis-associated DIC or multiple organ failure (23–25), endothelial thrombomodulin expression is decreased, suggesting that the increased serum thrombomodulin levels are the result of shedding from the endothelium following endothelial injury (26). Interestingly, thrombomodulin has been found to have other anti-inflammatory effects not related to its coagulation modulatory activity, including interfering with the activation of complement and inactivating high-mobility group protein B1.[7–9]

ART-123 is a recombinant thrombomodulin currently approved for the treatment of DIC in Japan where more than 28,000 patients have received the drug. In a double-blind randomized controlled phase 3 study conducted in Japan, patients with DIC and infection or hematological malignancy were randomized to receive treatment with ART-123 (0.06 mg/kg for 30 min, once daily) or heparin sodium (8 U/kg/hr for 24 hr) for 6 days.[10] DIC resolved in two thirds of the ART-123 patients compared with one half of the heparin-treated patients. Patients treated with ART-123 had fewer bleeding-related adverse events than those treated with heparin (p = 0.0487). There were no differences in 28-day mortality rates, but the study was powered to detect a difference in the rate of resolution of DIC rather than in mortality. Because the study included patients with DIC caused by hematological malignancy or infection, a retrospective subanalysis was conducted, excluding the patients with malignancy and analyzing the data from the 80 patients with DIC and infection.[27] The results indicated a trend to better outcome in the ART-123 group compared with the heparin group (DIC resolution rates 67.5% vs 55.6%; mortality rates 21.4% vs 31.6%). However, patients enrolled in that study appeared more ill than those enrolled in the present study, as indicated by the overall placebo mortality rate of 34.6% versus 21.6% in the current study.

The lack of a mortality benefit in the subgroup population of patients with overt DIC in the present study warrants discussion. Multiple possibilities could explain this finding. First, this subgroup may simply have been too small to see an effect. Second, the drug may truly not be efficacious or have only limited efficacy in DIC. Third, compared with the Japanese Ministry of Health criteria used in the study by Saito et al (10, 27), overt DIC as defined using the ISTH criteria appears to represent a late-stage, far end of the spectrum (28, 29) at which the coagulopathy may be too far advanced to respond to thrombomodulin; patients without overt DIC are also at high risk of death but may be more able to respond to therapy. Fourth, the window for randomization in the current study was up to 36 hours, a period during which supportive care could result in resolution of the coagulopathy prior to treatment with ART-123. Indeed, when a greater degree of coagulopathy was required after protocol amendment 4, with confirmation within just 4 hours of randomization, not only did the placebo mortality increase but so did the absolute reduction in mortality with ART-123, suggesting that ART-123 may be more beneficial for subjects with greater coagulation abnormality. Finally, mechanisms of action other than the anticoagulant activity of ART-123 are likely to be involved.

The population enrolled in the present study had less severe disease (13.2% overt DIC, 30.5% of patients with no organ dysfunction) than anticipated during study design. ART-123 may be efficacious even without organ dysfunction in the presence of coagulopathy; however, the low placebo mortality in the subgroup with no organ dysfunction (10%) makes it difficult to evaluate a treatment benefit of ART-123 on 28-day mortality in these patients. Limiting enrollment to patients with higher ISTH DIC scores (overt DIC) may have resulted in higher rates of death and secondary endpoints, which may have enabled the effects of ART-123 on these endpoints to be better characterized. Alternatively, use of a shorter window for definition of coagulopathy may enable a more responsive population of patients with sepsis and coagulopathy to be selected. Finally, the post hoc analyses raise the possibility of a greater beneficial effect of ART-123 in specific groups of patients, notably patients with a PT-INR > 1.4 and the presence of organ failure, suggesting that presence of organ dysfunction should be included in the target population for further testing of recombinant thrombomodulin.

In conclusion, administration of ART-123 at a dose of 0.06 mg/kg/d for a 6-day treatment period was associated with clear pharmacologic effects (lower D-dimer, F1.2, and TATc levels) and evidence suggestive of efficacy. ART-123 is a safe intervention in critically ill patients with coagulopathy due to sepsis, even in settings where patients are administered prophylactic doses of heparin, with no difference in bleeding complications between ART-123-treated and placebo-treated patients. These findings provide support for undertaking a phase III study, based on the population identified from the post hoc analyses, to extend these observations. Future trials using ART-123 should also be designed to better characterize the complex interactions present within the coagulation cascade in septic patients and to enable a description of the interactions between ART-123 and heparin or low–molecular weight heparins and between ART-123 and the emerging prophylactic/therapeutic anticoagulants (thrombin inhibitors and Xa inhibitors) used in critically ill patients.