Coagulation Management after Trauma in the Presence of Direct Oral Anticoagulants

Major trauma in the elderly and the use of direct oral anticoagulants (DOACs) have increased simultaneously in the past decade. The probability of encountering an elderly trauma patient treated with DOACs is therefore considerable. Because randomized trials in trauma are challenging, establishing animal models to evaluate additional potential therapeutic interventions may help to examine potential novel management strategies.

In this issue of the Journal, the article by Rayatdoost et al. provides novel data from an established polytrauma pig model.¹  In brief, a high rivaroxaban plasma concentration (approximately 400 ng/ml) was established in anesthetized pigs before inflicting a standardized major trauma of blunt liver injury and bilateral femur fractures. The authors administered increasing doses of prothrombin complex concentrate (PCC; 12.5, 25, and 50 U/kg) along with combinations of PCC plus tranexamic acid and fibrinogen concentrate, and determined total blood loss and survival over 4 h. In comparison to controls (saline), treatment with PCC 25 and 50 U/kg and PCC 12.5 U/kg plus tranexamic acid and fibrinogen concentrate reduced bleeding and achieved 100% survival versus 0% in controls.¹ 

The important findings of this study include the following. First, the impact of a high rivaroxaban plasma concentration on rotational thrombelastometry (ROTEM) showed a prolonged extrinsically activated clotting time, while maximum clot formation was unaltered. Likewise, thrombin generation was significantly reduced, while fibrinogen concentration remained stable. Bleeding during the first 12 min after trauma and before coagulation treatment was extensive (about 800 ml) but comparable in all groups, reflecting the severity of the trauma and the high degree of standardization of the experimental model. The exact mechanism of how PCC restores hemostasis under rivaroxabvan is not fully understood yet, but results suggest that coagulation factors have to overcome rivaroxaban-induced thrombin inhibition in order to restore thrombin generation. This is underlined by the finding that after the injury and while bleeding was ongoing, thrombin generation showed a progressive and dose-dependent improvement toward baseline in all treated groups. In contrast, fibrinogen levels showed a progressive decrease, with exception of the group receiving PCC 12.5 U/kg plus tranexamic acid and fibrinogen concentrate, thereby maintaining fibrinogen at approximately 1.5 g/l. ROTEM showed a progressive normalization toward baseline in all treated groups, while maximum clot formation decreased in all groups except the one treated with PCC 12.5 U/kg plus tranexamic acid and fibrinogen concentrate. In the PCC 25 and 50 U/kg–treated group, maximum clot formation decrease was at most 20%. The preservation of maximum clot formation may be explained by higher platelet counts and in particular a higher fibrinogen concentration in the group receiving PCC 12.5 U/kg plus tranexamic acid and fibrinogen concentrate. Accordingly, treatment reduced the blood loss after treatment to less than half in these groups versus controls and resulted in a 100% survival versus 0% in the control group.

Another important finding is the absence of thrombotic events in the heart, lungs, liver, or kidneys despite meticulous macroscopic and histological examinations. However, safety assessment is significantly limited by the observation time of 4 h and the low number of animals (n = 48).

Despite the interesting findings reported, can we extrapolate this information to humans, and what are the issues facing clinicians who manage bleeding patients receiving DOACs? In bleeding trauma patients, when their history is not available, the first challenge is assessing whether they are anticoagulated at all and if so, with which anticoagulant. As suggested by the European guideline on management of major bleeding and coagulopathy after trauma,²  this information can be achieved by three widely available routine assays: prothrombin time/international normalized ratio, anti-Xa activity (calibrated for low-molecular-weight heparin; table 1), and thrombin time.²  Interestingly, the plasma concentration of all direct Xa inhibitors can be measured based on a standard anti-Xa assay.³ 

European trauma treatment guidelines recommend reversing the effects of all antithrombotic agents when bleeding is ongoing.²  In vitamin K antagonist–treated patients, this is achieved by four-factor PCC plus vitamin K1, and in dabigatran-treated patients by idarucizumab.²  Recommendations regarding patients taking Xa inhibitors are less specific: if bleeding is life-threatening, the administration of PCC (25 to 50 U/kg) is suggested “until specific antidotes are available.”² 

Andexanet alfa, a recombinant Xa decoy protein, is now available as a specific antidote to Xa inhibitors. It is approved in the United States and Europe for the treatment of life-threatening bleeding in patients anticoagulated with rivaroxaban or apixaban but has not been formally studied in or approved for surgical patients. Whether a PCC or andexanet alfa–based strategy is superior is difficult to tell since a direct comparison is lacking. Also, approval of andexanet alfa is based on data from 352 patients. Of these, 64% were included due to intracranial and 26% due to gastrointestinal bleeding⁴  but no typical cases of major trauma. Hemostasis was deemed excellent in 69% and good in 13% of patients.⁴  Thromboembolic complications occurred in 10%, and the mortality rate was 14% within 30 days.

Similar populations of patients bleeding while anticoagulated with apixaban or rivaroxaban were investigated in Canada and Sweden.^5,6  Treatment was based on four-factor PCC. The bleeding was again predominantly intracranial (65%) and gastrointestinal (19%), and no typical major trauma was included. PCC was deemed effective in 69%, thromboembolism occurred in 5%, and the mortality rate was 24% within 30 days across both cohorts.

When choosing among andexanet alfa and four-factor PCC for the reversal of Xa inhibitors, one might expect a comparable effectiveness and rate of thromboembolic complications. However, there are other aspects to consider: cost, duration of treatment, and coagulation monitoring. Andexanet alfa is at least 10 times more expensive than PCC. Importantly, the dose of andexanet alfa comprises an initial bolus and a subsequent 2-h infusion. Due to the short duration of reversal by andexanet alfa, the Xa inibitor’s plasma concentration increases again after 2 to 3 h,⁷  and an effect on blood coagulation can be assumed. This leads to the question of whether andexanet alfa should be infused for a longer time, and if so, for how long. Commercial anti-Xa assays are used for the routine monitoring of direct Xa inhibitors. However, after andexanet alfa, they overestimate the residual anti-Xa activity, thereby underestimating the effectiveness of the reversal. A modified test setup with a reduced sample dilution is required.⁸  Viscoelastic testing may be helpful in this situation. In ROTEM, we expect a prolonged clotting time in the extrinsically activated thromboelastometry (EXTEM) assay that is related to the Xa inhibitor’s plasma concentration,⁹  and in thrombelastography a prolonged reaction time.¹⁰ 

Although typical major trauma patients were absent in the mentioned andexanet alfa studies,^4–6  reversal of anticoagulation may be indicated in such patients. It is unlikely that one single drug will ever be the “magic bullet” of coagulation management. Effective coagulation management is achieved by an individualized goal-directed coagulation algorithm,²  which reduces transfusion requirement and contributes to a favorable outcome.¹¹  If a DOAC with an elevated plasma concentration is present, effective management would include reversal of its effect, subsequent reassessment of coagulation, and treatment of trauma-related coagulopathy if present.

Last but not least, we would like to mention the option to delay surgery for several hours in stable patients without life-threatening injuries, thus allowing the rivaroxaban plasma concentration to decrease to a safe level without reversal.