Factor XIa antibody non-inferior to standard thromboprophylaxis after knee surgery

Effect of Osocimab in Preventing Venous Thromboembolism Among Patients Undergoing Knee Arthroplasty - The FOXTROT Randomized Clinical Trial

Literature - Weitz JI, Bauersachs R, Becker B et al., - JAMA 2020;323(2):130-139. doi:10.1001/jama.2019.20687

Introduction and methods

Anticoagulants are often administered postoperatively to prevent venous thromboembolism (VTE). Examples include factor Xa inhibitors such as enoxaparin (which also inhibits thrombin), and apixaban. These agents are effective at lowering VTE risk, but are also associated with a risk of bleeding. Emerging evidence suggest that targeting factor XI attenuates thrombosis with minimal disruption of hemostasis. Persons with congenital factor XI deficiency show a lower risk of VTE than those without it, but rarely experience spontaneous bleeding. Bleeding may, however, occur with surgery in these persons [1,2]. A study that knocked down factor XI with an antisense oligonucleotide, starting 35 days before knee arthroplasty, showed that it lowered the risk of postoperative VTE without increasing the bleeding risk [3].

This randomized non-inferiority phase 2 study compared the effect of osocimab with enoxaparin and apixaban (exploratory analysis) on prevention of VTE in patients undergoing total knee arthroplasty. Osocimab is a fully human monoclonal antibody that binds adjacent to the active site of factor XIa [4]. Osocimab has a half-life of 30 to 44 days, enabling a single-dose administration for surgical thromboprophylaxis.

The study consisted of two phases: the first part evaluated four osocimab doses (0.3, 0.6, 1.2 and 1.8 mg/kg), given on the day after surgery. In the second phase, 2 pre-operative osocimab doses (0.3 and 1.8 mg/kg), given on the day before surgery, were assessed. Each dose was administered as a single, 60-minute, IV infusion in a dose-blinded manner. Enoxaparin 40 mg was administered subcutaneously once daily either in the evening before surgery or 6 to 8 hours postoperatively and apixaban 2.5 mg twice daily 12 to 24 hours postoperatively. Both enoxaparin and apixaban were to be continued for at least 10 days or until venography was performed 10-13 days postoperatively.

The primary efficacy endpoint was incidence of the composite of asymptomatic deep vein thrombosis (DVT), objectively confirmed symptomatic DVT or nonfatal pulmonary embolism (PE), documented fatal PE, or unexplained death for which PE could not be excluded at 10 to 13 days postoperatively. The primary efficacy analysis was done in the per-protocol population, which consisted of those who received at least 1 dose of study medication and who had an evaluable venogram at 10 to 13 days postoperatively.

Main results

  • 813 Patients at 54 centers in 13 countries underwent randomization, of whom 600 formed the per protocol population.
  • In patients receiving osocimab postoperatively, primary endpoint events occurred in 18 (23.7%) persons on 0.3 mg/kg, 8 (15.7%) on 0.6 mg/kg, 13 (16.5%) on 1.2 mg/kg and 14 (17.9%) on 1.8 mg/kg.
  • In those receiving it postoperatively, 23 (29.9%) events were seen in those on 0.3 mg/kg, and 9 (11.3%) in those receiving 1.8 mg/kg osocimab. 20 Patients (26.3%) on enoxaparin and 12 (14.5%) on apixaban experienced a primary outcome event.
  • One-sided significance-level analysis showed that postoperative osocimab met non-inferiority criteria compared with enoxaparin, with risk differences of 10.6% (95%CI: -1.2% to ∞) in the 0.6 mg/kg group, 9.9% (95%CI: -0.9% to ∞) in the 1.2 mg/kg group, and 8.4% (95%CI: -2.6% to ∞) in the 1.8 mg/kg group.
  • Two-sided 90% CI analysis then demonstrated that preoperative osocimab at the 1.8 mg/kg dose met criteria for superiority compared with enoxaparin (risk difference 15.1%, 90%CI: 4.9% to 25.2%, P=0.007).
  • No PE or death was observed between randomization and completion of venography.
  • Major or clinically relevant nonmajor bleeding up to 10 to 13 days postoperatively occurred in 2% of patients on 0.3 mg/kg, 1% receiving 1.2 mg/kg and in 3% on 1.8 mg/kg of osocimab, and 1.9% patients receiving 0.3 mg/kg and 4.7% on 1.8 mg/kg of osocimab preoperatively. 5.9% Of those receiving enoxaparin and 2% of those receiving apixaban suffered such a bleeding event.
  • No intracranial bleeding or bleeding events into other critical sites until 10-13 days occurred postoperatively.
  • Of 585 patients taking any dose of osocimab, 414 (70.8%) had an adverse event, and 18 (3.1%) a serious adverse event. For enoxaparin, the respective percentages were 73.5% and 5.9% and for apixaban they were 63% and 1%.


This phase 2 non-inferiority clinical trial in patients undergoing knee arthroplasty showed noninferiority of osocimab at postoperative doses of 0.6, 1.2 and 1.8 mg/kg, as compared with enoxaparin in prevention of VTE up to 10-13 days postoperatively. In addition, the preoperative dose of 1.8 mg/kg was superior to enoxaparin. Due to the lack of clinically relevant bleeding events after the 0.6 and 1.2 mg/kg doses of osocimab, these appear to be the most promising doses to further evaluate.

These data suggest that upstream inhibition of factor XIa prevents thrombosis to a similar extent as downstream inhibition of factor Xa or thrombin, but more research should be done to establish the efficacy and safety of this approach relative to standard thromboprophylaxis.

Editorial comment

Recent attention has been directed at targeting factor XII, a component of the contact activation pathway, and factor XI in the intrinsic coagulation cascade, as antithrombotic strategies. Studies in nonhuman primates have suggested that inhibiting factor XI may reduce platelet and fibrin deposition more effectively than inhibiting factor XII. It is anticipated that factor XI-inhibiting agents will be as effective as current anticoagulants, but with a lower bleeding risk.

About the current study by Weitz and colleagues, Bauer [5] notes that ‘thromboprophylaxis following elective total knee replacement is a time-honored way to perform dose-finding studies with respect to efficacy and safety’. He sees, however, several limitations of this approach. First: the thrombotic endpoint in most patients is asymptomatic distal DVT, which is a surrogate for symptomatic thrombotic events, which is much less common. Second, there was no placebo group that did not receive anticoagulation, thus the baseline rate of bleeding from surgery could not be assessed. Bauer notes that all dose groups showed adjudicated bleeding rates of 0 to 3%, except for the highest dose of osocimab preoperatively, in which it was 4.7%. While these results should not be overinterpreted because the number of bleeding events was small, but this group was exposed to the most severe hemostatic challenge, namely receiving the highest dose just before surgery.

The current data do not allow to draw conclusions about the bleeding risk of osocimab relative to currently available anticoagulants, nor that it is more effective in preventing VTE than apixaban. The bleeding rate in the highest preoperative dosing group signals that a high degree of factor XI inhibition can still provoke bleeding risk. This is in line with observations that individuals with hereditary factor XI deficiency can have bleeding following major surgery or trauma. The bleeding phenotype in these individuals can be highly variable, with little correlation between plasma factor XI activity levels and bleeding. The biological mechanism behind this variability is poorly understood.

The FOXTROT did not show that osocimab was safer than low-molecular-weight heparin or apixaban after total knee replacement. Current antithrombotic agents including aspirin are commonly used for this purpose, with low rates of symptomatic VTE and bleeding. Higher risk populations who need safer anticoagulant agents may include older patients, those with renal dysfunction or those with prior major bleeding. Moreover, an unmet need exists for acutely ill patients who require extracorporeal membrane oxygenation or left ventricular assist devices, in whom contact between artificial surfaces and blood can induce coagulation activation. Agents inhibiting factor XI or XII could be particularly useful in these settings. Trials in such high-risk populations will be more challenging than those that led to approval of direct oral anticoagulants, so it may be important to have reversal agents for factor XI inhibitors ready when these drugs gain regulatory approval.


1. Duga S, Salomon O. Congenital factor XI deficiency: an update. Semin Thromb Hemost. 2013; 39(6):621-631.

2. Salomon O, Steinberg DM, Zucker M, et al. Patients with severe factor XI deficiency have a reduced incidence of deep-vein thrombosis. Thromb Haemost. 2011;105(2):269-273.

3. Büller HR, Bethune C, Bhanot S, et al; FXI-ASO TKA Investigators. Factor XI antisense oligonucleotide for prevention of venous thrombosis. N Engl J Med. 2015;372(3):232-240.

4. Schaefer M, Buchmueller A, Dittmer F, et al. Allosteric inhibition as a new mode of action for BAY 1213790, a neutralizing antibody targeting the activated form of coagulation factor XI [published online October 24, 2019]. J Mol Biol.

5. Bauer K. How Effective and Safe Is Factor XI Inhibition in Preventing Venous Thrombosis? JAMA. 2020;323(2):121-122

Find this article online at JAMA

Facebook Comments


We’re glad to see you’re enjoying PACE-CME…
but how about a more personalized experience?

Register for free