Date: April 5th, 2019

Reference: Connolly et al. Full Study Report of Andexanet Alfa for Bleeding Associated with Factor Xa Inhibitors. NEJM 2019

Guest Skeptic: Dr. Ryan Radecki is an Emergency Physician at Kaiser Permanente NW, co-host of the Annals of Emergency Medicine podcast and Journal Club section editor.

Case: You are caring for a 72-year-old man who comes in after having slipped on the ice.  A routine evaluation finds only minor bumps and bruises, including a rather nasty one on his occiput where he struck a step.  He reports he has been taking apixaban to prevent stroke in the context of atrial fibrillation, which you easily recognize as one of the modern oral anti-Factor Xa inhibitors.  You order a non-contrast CT to rule out hemorrhage. It demonstrates a 7mm subdural hematoma with 3mm of midline shift.  As you are reassessing your patient and treatment plan, the question presents itself – how should we reverse his anticoagulation?

Background: More and more patients are being treated with direct oral anticoagulants (DOACs). This number will probably increase since the AHA/ACC/HRS 2019 updated guidelines for atrial fibrillation guidelines. It now contains the following recommendation:

NOACs (dabigatran, rivaroxaban, apixaban, and edoxaban) are recommended over warfarin in NOAC-eligible patients with AF (except with moderate-to-severe mitral stenosis or a mechanical heart valve). Level A

One of the concerns clinicians had with DOACs was there was no way to reverse these new anti-coagulants when they were introduced. In contrast, protamine could be used for heparin and LMWH reversal and vitamin K, fresh frozen plasma and prothrombinase complex concentrate could be used to reverse coumadin (Hunt and Levi BMJ 2018).

This changed in 2015 when the Food and Drug Administration (FDA) approved idarucizumab for the reversal of dabigatran. Dabigatran is a direct thrombin inhibitor. We covered the interim analysis of 90 patients included in a prospective cohort study by Pollack et al NEJM 2015 on SGEM#139. Our bottom line for that episode was that idarucizumab is here (USA) and probably works but its patient-oriented efficacy and safety are still pending.

The full study cohort of 503 patients has since been published (Pollack et al NEJM 2017) and we are in the same place we were in 2015. Idarucizumab clearly and effectively removes dabigatran from circulation and this ought to be occasionally clinically useful. I would certainly exhaust all potential supportive and expectant management options first, as well as try to definitively confirm dabigatran as the culprit for abnormal hemostasis (EM Lit of Note).

The FDA granted accelerated approval for andexanet alfa (Andexxa) in May of 2018. Andexxa is an antidote for factor Xa inhibitor like rivaroxaban, apixaban and edoxaban. It acts as a decoy and binds to the factor Xa inhibitors.

The American College of Cardiology has published a fact sheet to provide some guidance for the use of anticoagulation reversal agents.

Clinical Question: Should andexanet alfa be used to treat serious bleeding events in patients taking Factor Xa inhibitors?

Reference: Connolly et al. Full Study Report of Andexanet Alfa for Bleeding Associated with Factor Xa Inhibitors. NEJM 2019

  • Population: Adult patients presenting with an acute major bleed, and had received a DOAC (apixaban, rivaroxaban, or edoxaban) at any dose or enoxaparin at a dose of at least 1 mg per kilogram of body weight per day within the previous 18 hours.
    • Acute Major Bleed: Bleeding having one or more of the following features: potentially life-threatening bleeding with signs or symptoms of hemodynamic; bleeding associated with a decrease in the hemoglobin level of at least 2 g per deciliter or bleeding in a critical area or organ.
    • Exclusions: There were a number of exclusions listed in supplemental material but the key ones were as follows: Planned surgery within 12hrs; ICH in a patient with a GCS<7 or an estimated hematoma volume of more than 60 cc; expected survival of <1 month; the occurrence of a thrombotic event within 2 weeks before enrollment; or use of vitamin K antagonist, dabigatran, prothrombin complex concentrate, recombinant factor VIIa, whole blood, or plasma within the previous 7 days.
  • Intervention: Andexanet intravenous (IV) bolus over a period of 15 to 30 minutes, followed by a 2-hour infusion of the drug.
  • Comparison: None
  • Outcome:
    • Co-Primary Efficacy Outcome: The percent change in anti–factor Xa activity after andexanet treatment and the percentage of patients with excellent or good hemostatic efficacy at 12 hours after the end of the infusion.
    • Primary Safety OutcomesDeath, thrombotic events, and the development of antibodies to andexanet or to native factor X and factor Xa at 30 days.

Authors’ Conclusions: “In patients with acute major bleeding associated with the use of a factor Xa inhibitor, treatment with andexanet markedly reduced anti–factor Xa activity, and 82% of patients had excellent or good hemostatic efficacy at 12 hours, as adjudicated according to prespecified criteria.”

Quality Checklist for Observational Study:

  1. Did the study address a clearly focused issue? No
  2. Did the authors use an appropriate method to answer their question? Unsure
  3. Was the cohort recruited in an acceptable way? No
  4. Was the exposure accurately measured to minimize bias? Yes
  5. Was the outcome accurately measured to minimize bias? Yes/No
  6. Have the authors identified all-important confounding factors? No
  7. Was the follow up of subjects complete enough? Yes
  8. How precise are the results? Unsure
  9. Do you believe the results? Yes
  10. Can the results be applied to the local population? No
  11. Do the results of this study fit with other available evidence? Yes

Key Results: They enrolled 352 patients into the study. The mean age was 77 years and 80% of patients were taking anticoagulation for atrial fibrillation. Two-thirds of the patients had ICH and one-quarter had GI bleeds.

Andexanet decreased anti-factor Xa activity and four out of five patients had excellent or good hemostasis.

  • Co-Primary Outcome:
    • Excellent or good hemostasis occurred in 204 of 249 patients (82%)
    • Decreased in anti-factor Xa activity from 149.7ng/ml to 11.1ng/ml (92% reduction; 95% CI: 91% to 93%)
  • Primary Safety Outcomes:
    • Death: 49/352 (14%)
    • Thrombotic Events: 34/352 (10%)
    • Development of antibodies to andexanet or to native factor X and factor Xa: None

We are not going to specifically talk about the fact this was an industry funded study and the authors listed multiple conflicts of interest. I have said multiple times it does not make the data wrong but should make us more skeptical of the results and the interpretation.

1) RecruitmentAn important first step to any study is the recruitment of subjects. There were multiple problems with their recruitment, most prominently including multiple changes to their inclusion and exclusion criteria.  The authors amended these criteria four times over the duration of the study enrolled between 2015 and 2018.  One key change was during 2016 and 2017 when the authors restricted to only patients with intracerebral hemorrhage in order to “enrich” the study with these patients.

The intent to “enrich” the study population with intracranial hemorrhage could be in response to observations regarding its lack of ability to improve control of extracranial hemorrhage.  After all, the underlying theory behind its efficacy is that the short period of restored hemostasis will allow for clot formation that is sustained following cessation of the infusion and subsequent drop in anti-Factor Xa activity.

This is probably more likely to be efficacious in a small volume bleed into a closed space, such as intracranially.  However, is this any better than alternative strategies recommended by the neurocritical care societies, such as those using prothrombin concentrate complexes?  This exact question is being addressed by a clinical trial that has only just now started enrollment, and our data regarding andexanet’s true clinical effectiveness will remain uncertain until its completion.

There were other substantial changes made during the study in the recruitment of patients making it hard to generalize the results to the entire cohort.

2) Co-Primary Outcome: As the SGEM listeners know, there can be only ONE primary outcome. The authors amended their study to have two “coprimary” outcomes.  One of which is ostensibly a surrogate for the other. The percent change in factor Xa activity is a simple pharmacokinetic evaluation and not typically the purpose of a Phase III clinical trial.

The original primary outcome was rate of hemostatic efficacy assessed at 24 hours from the start of the andexanet bolus. The change in anti-factor Xa activity was considered as a secondary efficacy endpoint. It was elevated to one of two coprimary outcomes and the hemostatic efficacy was changed from 24 hours to 12 hours.  They state in the supplementary appendix that these changes were made based on regulatory feedback.

Another thing about their efficacy outcome was that they found no relationship between hemostatic efficacy and the reduction observed in anti-factor Xa activity in all patients. This seriously calls into question the clinical impact of andexanet. There was a weak association with the subgroup analysis of patients suffering from ICHs. This should be viewed as hypothesis generating.

3) Confounding Factors: This was an observational study so we must always consider confounders that could be responsible for the results seen.The authors consider a handful of factors influencing their pharmacokinetic and clinical outcomes, but there is substantial variation within the cohort observed.  They generally describe the influence on outcomes as relates to bleeding site and baseline anti-factor Xa levels, but otherwise do not include a detailed evaluation of other individual patient features predicting hemostatic efficacy.

4) Harm: When evaluating a new drug, we should always consider the harm. The safety outcomes were rather concerning, as well, including 34 thrombotic events within 30 days.  Half of these were cases of arterial thrombosis, such as myocardial infarction.  Then, there were 49 deaths within 30 days – 35 of cardiovascular causes.  This mismatch between the total cardiovascular deaths and their reports of only 7 myocardial infarction and 5 pulmonary embolism may indicate harms related to andexanet not fully captured by their counting of thrombotic events.

5) No Comparison Group: The lack of an active comparator is the biggest limitation of this study. How do we know if this is any better or worse than usual care? This study joins RE-VERSE AD as one of the least informative trials in recent memory.  Like RE-VERSE AD, featuring idarucizumab, this is a single-arm assessment of efficacy without a comparator.

However, unlike RE-VERSE AD and dabigatran, there are viable management strategies for management of anti-Factor Xa-associated bleeding, there is clinical equipoise for a comparative trial, and the impermanent underlying pharmacology of andexanet suggests it ought to be substantially less effective.  It might even be considered, rather than not having equipoise for a trial, it was unethical to perform this trial and expose participants to this experimental intervention without creating substantial generalizable knowledge.

Comment on Authors’ Conclusion Compared to SGEM Conclusion: The authors’ conclusion is factual and narrow, accurately reflecting the definitions in their protocol.  Andexanet alfa clearly binds circulating anti-Factor Xa during administration, and – as expected – rapidly diminishes afterwards.  According to their arbitrary criteria for classifying patients as having “good” or “excellent” hemostasis, patients met those criteria 82% of the time.  The clinical relevance of their criteria, as well as the utility of andexanet alfa, remain unclear.

SGEM Bottom Line: The routine use of andexanet alfa in the management of bleeding patients on factor Xa-inhibitors cannot be recommended at this time.

Case Resolution: The patient is treated, according to local protocol, with fixed-dose prothrombin concentrate complexes and admitted to the intensive care unit for ongoing monitoring of his neurologic status.  A repeat CT shows a small amount – <20% – increase in size of his subdural hematoma, representing “ ” hemostasis.  His neurologic status remains unchanged. Within 24-hours, the patient is discharged from the intensive care unit to the medical floor to complete planning for discharge.

Dr. Ryan Radecki

Clinical Application: There is no obvious clinical application for andexanet alfa.  At $25k to $50k per dose, it should first demonstrate superiority with regard to current management strategies of factor replacement using prothrombin concentrate complexes. There may be a subset of patients, specifically those with certain types of intracranial hemorrhage, whose hematoma growth is attenuated with andexanet alfa, but this cannot yet be determined from the evidence provided at present. The most important problem with clinically applying this study is the lack of a comparison group.

What Do I Tell My Patient? Unfortunately, the patient and family need to be informed of the presence of life-threatening intracranial bleeding and the limitations in our current management options.  We will be providing care using the best evidence and recommendations available, which involves treatment with PCCs and vigilant supportive care in the intensive care unit.

Keener Kontest: There were multiple correct answers to last keener questions. Clearly a lot of nerds listen to the SGEM. Marc Schieren knew the Turing Test was designed to determine, whether a computer shows signs of artificial intelligence (Imitation Game).

Listen to the podcast to hear this weeks’ trivia question. If you know the answer, send an email to with “keener” in the subject line. The first correct answer will receive a cool skeptical prize.

Other FOAMed:

  • First10EM: Andexanet Alfa – More garbage science in the NEJM
  • EM Lit of Note: Disutility, Thy Name is ANEXXA-4
  • EMCrit: I Have Issues with Andexanet by K. Kipp, PharmD
  • St. Emlyn’s: Reversal of DOACs with Andexanet Alfa. St.Emlyn’s
  • REBEL EM: ANNEXA-4: Andexanet Alfa and Factor Xa Inhibitors

Remember to be skeptical of anything you learn, even if you heard it on the Skeptics’Guide to Emergency Medicine.