Update on Direct Oral Anticoagulants in the Emergency Department

Authors: Megan Donohue MD, MPH, Erica Bates MD, Robert Brown MD, Christine Carter MD, Hannah Goldberg MD.
Editors: Kami M. Hu MD FAAEM, Kelly Maurelus, MD FAAEM
Originally Published: Common Sense May/June 2018

Questions: What new information do we have regarding direct oral anticoagulants (DOACs), specifically:

  1. What is the most recent safety data?
  2. Do our standard coagulation assays provide an accurate measure of anticoagulant activity?
  3. Where do we stand with reversal agents?

Use of DOACs such as rivaroxaban (Xarelto®), apixaban (Eliquis®), edoxaban (Savaysa®), and dabigatran (Pradaxa®) has increased dramatically since dabigatran first hit the market in 2009. Data from four major trials (ARISTOTLE, ROCKET-AF, RE-LY, ENGAGE AF-TIMI 48) supported the safety of DOACs and promoted their FDA approval.[1,2,3,4] Their use presents some challenges in the management of the therapeutically-anticoagulated patient. This article reviews recent data on safety metrics, assessment of coagulopathy, and reversal strategies for these drugs.

Xu Y, Schulman S, Dowlatshahi D, et al. Direct Oral Anti-coagulation or Warfarin-Related Major Bleeding Characteristics, Reversal Strategies, and Outcomes from a Multicenter Observational Study. CHEST. 2017;152(1):81-91.

Because the early trials have been criticized for limited generalizability due to relatively low rates of therapeutic INR and reversal in the warfarin groups, the authors of this study sought to determine if their findings were reproducible in a more generalizable population with a higher rate of attempted warfarin reversal. To achieve this, the Bleeding Effected by Direct Oral Anticoagulants (BLED-AC) study looked at post-market surveillance data; which is used to asses and monitor safety in the real-world setting.

The study included data from five tertiary care hospitals in Ontario, Canada, on adults over 66 years old with atrial fibrillation or flutter who were taking warfarin, rivaroxaban, apixaban, or dabigatran and were hospitalized for major bleeding. Outcomes included location of bleeding, treatment strategies, and survival rates. Patients were excluded if they had a prosthetic heart valve or if their initial presentation did not involve hemorrhage.

The data included 460 patients using DOACs and 1,542 on warfarin. The mean age was 81 years old. There were no significant differences between the two study groups in general demographics, blood pressure, or hemoglobin levels at presentation. There was an increased prevalence of chronic kidney disease (CKD) and INR elevation among the warfarin group which was expected given the contraindications to use of various DOAC in CKD and the known general lack of effect on INR. The warfarin group had more intracranial hemorrhages (19.0% vs 13.3%, p < 0.01) and the DOAC group had more gastrointestinal bleeds (31.1% vs 17.3%, p < 0.01), which translated to a higher rate of packed red blood cell transfusions in the DOAC group (52% vs 39.5%, RR 1.32, 95% CI 1.18-1.46). The warfarin group received more vitamin K, fresh frozen plasma, and prothrombin complex concentrate. Survival comparison showed lower in-hospital mortality in the DOAC group (9.8% vs 15.2%, p<0.0001), but no statistically significant difference in all-cause mortality at 30 days (12.6% vs 16.3%, 95% CI 0.61-1.03).

Despite inherent limitations of retrospective chart reviews, this study addressed criticisms of previous studies by comparing outcomes to a control group representative of current clinical practice. Other strengths included a large sample size and its longitudinal nature. The authors conclude that their results confirm the relative safety profile of the DOACs in comparison with warfarin.

Ebner M, Birschmann I, Peter A, et al. Emergency Coagulation Assessment During Treatment with Direct Oral Anticoagulants: Limitations and Solutions. Stroke. 2017;48(9):2457-63.

DOACs pose a challenge to EPs because, unlike for warfarin, there is no rapid, reliable test to determine how anticoagulated a patient is. This fact hinders decision-making in conditions such as ischemic stroke, where thrombolysis might otherwise be offered, or hemorrhagic stroke, where reversal agents may be necessary. Ebner et al. evaluated samples collected as part of a separate prospective observational study, the POCT-DOAC trial (Point-of-Care Testing of Coagulation in Patients Treated with Direct Oral Anticoagulants), which determined that point-of-care testing allows the rapid identification of clinically-significant concentrations of dabigatran and rivaroxaban (but not apixaban) in nonbleeding patients. For this investigation, they looked to see if the prothrombin time (PT), activated partial thromboplastin time (aPTT), and thrombin time (TT) could reliably screen for clinically significant concentrations of the same DOACs.

They evaluated 481 samples of blood from 96 patients taking dabigatran, rivaroxaban, and apixaban; serial blood samples were drawn after medication administration to obtain samples with different drug concentrations. Patients with baseline coagulopathy or recent use of other anticoagulants were excluded. PT, aPTT, and TT were performed for each sample and were compared to drug concentrations obtained using ultraperformance liquid chromatography-mass spectrometry, the gold standard test to determine the actual plasma concentrations of DOACs. A concentration of 30 ng/ml was used as a cutoff for “safe” levels in patients undergoing possible thrombolysis or surgery. Sensitivity and specificity were calculated for each DOAC, and receiver operating characteristic (ROC) analysis was used to calculate optimized PT, aPTT, and TT cutoff values that were >95% specific for concentrations <30 ng/ml.

For dabigatran, a normal aPTT was 91% sensitive but only 49% specific for a level <30 ng/ml. Adding a normal PT did not significantly increase specificity. A normal TT was 100% specific for a level <30 ng/ml, but only 26% sensitive. Increasing the cutoff levels for TT increased sensitivity to 95-99% but decreased specificity to 90%. For rivaroxaban, a normal PT was 77% specific with 94% sensitivity, and as with dabigatran, addition of PT values did not improve accuracy. The coagulation assays were unreactive to apixaban; normal test results had a less than 20% specificity for apixaban concentrations <30 ng/ml. The modified cutoff values obtained with ROC analysis for dabigatran and rivaroxaban improved specificity but markedly decreased sensitivity (to 22% and 47%, respectively). Enhanced cutoffs could not be established for apixaban.

Overall, this study demonstrated that these standard coagulation tests are not sufficient to reliably identify patients with DOAC serum concentrations safe for surgery or thrombolysis. While using the modified cutoffs can increase identification of patients with higher dabigatran and rivaroxaban concentrations, the low sensitivities decrease the identification of patients who do not necessarily require the time and expense of additional reversal. Strengths of this study include investigation of the more commonly used oral anticoagulants and application of tests commonly and rapidly available in emergency departments nationwide. Despite the determination of reagent-specific optimized cutoff ranges, it remains unclear whether the specific reagents used for this study are standardized or rapidly available.

Pollack CV, Reilly PA, van Ryn J, et al. Idarucizumab for Dabigatran Reversal –- Full Cohort Analysis. N Engl J Med. 2017;377(5):431-41.

The increased use of dabigatran led to an increased incidence of dabigatran-related hemorrhage, requiring the creation of its reversal agent idarucizumab (Praxbind®). In 2015, interim analyses of the REVERSE AD (A Study of the RE-VERSal Effects of Idarucizumab on Active Dabigatran) trial were published, resulting in the licensing of idarucizumab use in several countries. This paper reports results from the full cohort analysis of the REVERSE AD trial.

Pollack et al. designed a multicenter, prospective, single-cohort study to evaluate the reversal effects of idarucizumab in patients known to be on dabigatran who presented with life-threatening bleeding (group A) or need for non-bleeding related emergent surgery within 8 hours (group B). “Life-threatening bleeding” was defined as fatal bleeding, symptomatic intracranial bleeding, reduction in hemoglobin of at least 5g/dL, transfusion of at least 4 units of blood or packed cells, bleeding associated with hypotension requiring the use of inotropic agents, or bleeding necessitating surgical intervention. The primary endpoint was the maximum percentage reversal of anticoagulant effects within 4 hours of administering idarucizumab 5g intravenously. Labs were monitored both on site and at a central laboratory, but results from the latter were not revealed to treating physicians. Locally, complete reversal was defined as normalization of diluted TT or ecarin clotting time. Centrally, aPTT and concentration of unbound dabigatran were also reviewed. Additionally, clinicians performing the surgery or procedure evaluated and rated hemostasis as either normal or mildly, moderately, or severely abnormal.

There were 301 patients in group A and 202 in group B. The majority of bleeding was gastrointestinal (39%) intracranial (35%), and trauma-related (18%). The median maximum percentage reversal within 4 hours after the administration idarucizumab was 100% (95% CI, 100 to 100). Among the non-ICH Group A patients (n=203), bleeding cessation was confirmed in the first 24 hours and median time to hemostasis was 2.5 hours (95% CI, 2.2-3.9). In Group B, peri-procedural hemostasis was reported by treating clinicians as normal (93.4%), mildly abnormal (5.1%) and moderately abnormal (1.5%). Four of the 503 patients experienced potential hypersensitivity reactions, although two of these patients were on separate medications that could also have been responsible. Other adverse events were deemed by the authors to actually be worsening of the index event or sequelae of critical illness rather than secondary to idarucizumab administration (e.g. delirium, cardiac arrest, septic shock).

The main limitation of this study is the lack of a control group. The authors argue, however, that in the setting of no other effective reversal agents it would be unethical to withhold the reversal agent and randomize patients to a placebo or treatment with PCC only.

Connolly SJ, Milling TJ, Eikelboom JW, et al. Adnexanet Alfa for Acute Major Bleeding Associated with Factor Xa Inhibitors. N Engl J Med. 2016;375(12):1131-41.

This article was an interim report of the Andexanet Alfa, a Novel Antidote to the Anticoagulation Effects of FXA Inhibitors (ANNEXA-4) trial, an ongoing multicenter, prospective, open-label, single-group study of andexanet alfa (andexanet), a recombinant modified human factor Xa decoy protein that specifically reverses the effects of both direct and indirect factor Xa inhibitors by binding to the inhibitors themselves.

Patients over the age of 18 years with acute major bleeding who were reported to have received a factor Xa inhibitor (apixaban, rivaroxaban, edoxaban, or enoxaparin) within the past 18 hours were enrolled. “Acute major bleeding” was defined as potentially life-threatening overt bleeding with signs or symptoms of hemodynamic compromise, decrease in hemoglobin of at least 2 g/dL, hemoglobin of ≤8 g/dL with no baseline available, or acute symptomatic bleeding in a critical area or organ. Exclusion criteria included use of an anticoagulant other than a factor Xa inhibitor within the prior seven days, intracranial hemorrhage with a GCS less than seven, major thrombotic event within the past two weeks, large cerebral hematoma volume, expected survival of less than one month, or need for emergent surgery within 12 hours of ED arrival. Patients received an andexanet bolus over 15-30 minutes followed by a 2-hour infusion. The primary outcomes included percent change in anti-factor Xa activity and rate of hemostatic efficacy. Visible bleeding hemostasis was considered “excellent” if there was cessation of bleeding within 1 hour and “good” if within 4 hours. Intracranial hemorrhage was considered to have excellent hemostasis if there was a less than 20% volume increase on repeat imaging at 1 and 12 hours. Analysis was performed on an efficacy population, a sample selected to represent ideal and controlled circumstances, in which baseline anti-factor Xa activity was 75 ng/ml or more.

Of the 47 patients studied, 26 were on rivaroxaban, 20 were on apixaban and 1 was on enoxaparin. Gastrointestinal (49%) and intracranial bleeding (42%) were the most common etiologies of hemorrhage. After administration of andexanet, anti-factor Xa activity decreased by 89% (95% CI 58-94) in patients on rivaroxoban and 93% (95% CI 87-94) in patients on apixiban. The results of a single enoxaparin patient are not significant nor than they be generalized.

Overall, 79% of patients had excellent or good hemostasis following andexanet infusion with no documented infusion reactions. During the 30-day follow-up period, thromboembolic events occurred in 18% of patients and there were 10 deaths. Although this study shows early preliminary evidence for the efficacy of andexanet alfa, its single cohort design with no control group or blinding, and its reliance on a non-widely available measurement of chemical activity are important limitations to recognize.

The use of DOACs will likely continue to increase as further evidence to support their safety is published. It is important to note that standard coagulation tests cannot be relied upon to identify DOAC activity or determine need for reversal strategies. Idarucizumab significantly reverses dabigatran’s anticoagulant effect without major adverse reactions and should be used in life-threatening dabigatran-related hemorrhage. Andexanet alfa is currently in development and shows promising benefit. A newer drug, aripazine (ciraparantag), has shown favorable results in reversal of multiple anticoagulant agents[5] but has yet to be examined in a clinical human study. Given the limitations of unblinded, single-cohort studies that measure chemical effect and subjective hemostasis as opposed to long term benefits to mortality, it is still important for physicians to follow the results of additional post-marketing studies. Further studies are needed to identify reliable markers of apixaban and edoxaban activity as well.


1. Analysis of post-market data continues to demonstrate that DOACs are likely no more dangerous than warfarin.

2. Standard coagulation tests cannot be reliably used to identify DOAC activity or to indicate when emergent interventions or procedures can proceed without anticoagulation reversal.

3. Idarucizumab should be given to patients who are on dabigatran and present with life-threatening bleeding or emergent need for surgery.

4. Preliminary analysis of the ANNEXA-4 trial shows adnexanet alfa can rapidly and successfully decrease anti-factor Xa levels and produce effective hemostasis in patients on apixaban and rivaroxaban.

Additional References:

1. Granger CB, Alexander JH, McMurray JJ, et al. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2011; 365(11):981-982.

2. Patel MR, Mahaffey KW, Garg J, et al. Rivaroxaban versus Warfarin in Nonvalvular Atrial Fibrillation. N Engl J Med. 2011; 365(10):883-891.

3. Connolly SJ, Ezekowitz MD, Yusuf S, et al. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009; 361(12):1139-51.

4. Guigliano RP, Ruff CT, Braunwald E, et al. Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2013; 369(22):2093-2104.

5. Tummala R, Kavtaradze A, Gupta A, et al. Specific antidotes against direct oral anticoagulants: A comprehensive review of clinical trials data. Int J Cardiol. 2016; 214:292-8.