Efficacy and safety of tifacogin (recombinant tissue factor pathway inhibitor) in severe sepsis: a randomized controlled trial

Context  The expression and release of tissue factor is a major trigger for the activation of coagulation in patients with sepsis. Tissue factor pathway inhibitor (TFPI) forms a complex with tissue factor and blood protease factors leading to inhibition of thrombin generation and fibrin formation. Objectives  To determine if administration of tifacogin (recombinant TFPI) provides mortality benefit in patients with severe sepsis and elevated international normalized ratio (INR) and to assess tifacogin safety in severe sepsis, including patients with low INR. Design and Setting  A randomized, double-blind, placebo-controlled, multicenter, phase 3 clinical trial conducted from March 21, 2000, through September 27, 2001, in 245 hospitals in 17 countries in North America, Europe, and Israel. Patients  The primary efficacy population consisted of 1754 patients (18 years) with severe sepsis and a high INR (1.2) randomly assigned to intravenous infusion of either tifacogin (0.025 mg/kg per hour for 96 hours, n = 880) or placebo (arginine citrate buffer, n = 874), and 201 patients with a low INR (<1.2) randomly assigned to receive the same dose of either tifacogin or placebo. Main Outcome Measure  All-cause 28-day mortality. Results  Overall mortality at 28 days in the tifacogin-treated group (n = 880) vs the placebo group (n = 874) for high INR was 34.2% vs 33.9%, respectively (P = .88, Pearson ² test; P = .75, logistic regression model). None of the protocol-specified secondary end points differed between the tifacogin vs placebo groups. An analysis on the first 722 patients demonstrated a mortality rate of 38.9% for placebo vs 29.1% for tifacogin (P = .006, Pearson ² test). Tifacogin significantly attenuated prothrombin fragment 1.2 and thrombin:antithrombin complex levels (P<.001, 2-sample t test) in patients with high and low INR. Overall mortality was lower in the tifacogin response in patients with low INR (12%; n = 83) vs placebo (22.9%; n = 118) (P = .051, Pearson ² test; P = .03, logistic regression model). There was an increase in serious adverse events with bleeding in the tifacogin group in both cohorts (6.5% tifacogin and 4.8% placebo for high INR; 6.0% tifacogin and 3.3% placebo for low INR). Conclusions  Treatment with tifacogin had no effect on all-cause mortality in patients with severe sepsis and high INR. Tifacogin administration was associated with an increase in risk of bleeding, irrespective of baseline INR.