Reversal Strategies for Non-vitamin K Antagonist Oral Anticoagulants: A Critical Appraisal of Available Evidence and Recommendations for Clinical Management

A Joint Position Paper of the European Society of Cardiology Working Group on Cardiovascular Pharmacotherapy and European Society of Cardiology Working Group on Thrombosis

Alexander Niessner; Juan Tamargo; João Morais; Lorenz Koller; Sven Wassmann; Steen Elkjær Husted; Christian Torp-Pedersen; Keld Kjeldsen; Basil S. Lewis; Heinz Drexel; Juan Carlos Kaski; Dan Atar; Robert F. Storey; Gregory Y. H. Lip; Freek W. A. Verheugt; Stefan Agewall


Eur Heart J. 2017;38(22):1710-1716. 

In This Article

Risk of Bleeding With Non-vitamin K Antagonist Oral Anticoagulants

Bleeding is the main adverse effect when using oral anticoagulants (OACs). The risk of major bleeding with vitamin K antagonists (VKAs) is dependent on the quality of anticoagulation control and estimated to be ~1.3/100 patients/year in patients with INR 2.0–3.0.[1,2] Multiple limitations of VKAs stimulated the development of non-vitamin K antagonist oral anticoagulants (NOACs) acting directly on coagulation factors, including factor FIIa (thrombin) and FXa. Pharmacological properties of the NOACs relevant for bleeding are summarized in the supplement including Supplementary material online, Table S1 and their main effects are summarized in Supplementary material online, Table S2 The risk of bleeding in patients managed with NOACs was properly assessed in the four landmark Phase III trials in patients with atrial fibrillation (AF; Table 1). Major bleeding was the principal safety outcome in RE-LY,[3,4] ARISTOTLE,[5] and ENGAGE AF.[6] In ROCKET-AF, the primary safety endpoint was the composite of major and non-major clinically relevant bleeding.[7] The risk of major bleeding was significantly reduced with dabigatran 110 mg b.i.d., apixaban and both doses of edoxaban[3–6] while the bleeding rates for dabigatran 150 mg b.i.d. and rivaroxaban were similar to those with warfarin (Table 1).

Gastrointestinal (GI) bleeding accounts for ~90% of major extracranial haemorrhages in patients with AF receiving VKAs.[8] With the NOACs, dabigatran 150 mg b.i.d., rivaroxaban, and edoxaban 60 mg o.d. significantly increased rates of GI bleeding (~1.5-fold) compared with warfarin.[3,6,7] While rivaroxaban increases upper and lower GI bleeding to a similar extent, dabigatran 150 mg b.i.d. prominently increased lower GI bleeding.[9] Apixaban was associated with GI bleeding rates comparable with those with warfarin.[5] Edoxaban 30 mg o.d. demonstrated significantly lower rates of GI bleeding compared with warfarin (Table 1).[6]

Intracranial haemorrhage (ICH) is the major concern in patients on OACs and this is the field where NOACs demonstrated major clinical improvement (Table 1). In the four trials, the hazard ratio for ICH was between 0.31 and 0.64, with meta-analysis showing a hazard ratio of 0.48 overall.[10] Furthermore, the sequelae of ICH may be less severe when it has occurred during NOAC treatment.[11] Major bleeding and ICH rates in 'real-world' post-marketing data for NOACs also seems to be generally reassuring, although GI bleeding rates appear to be higher for dabigatran (150 mg b.i.d.) compared with warfarin.[12–14]

While this paper focuses on the reversal of the effects of NOACs due to an emergency situation, prevention of bleeding remains of utmost importance. The risk of bleeding may be estimated using the well-validated HAS-BLED score.[15] A high HAS-BLED score is there to 'flag up' patients at higher risk of bleeding for more careful review and follow-up, as well as attention to the potentially reversible risk factors for bleeding, such as uncontrolled hypertension (systolic blood pressure >160 mmHg), labile INRs (for VKA user), concomitant use of aspirin/NSAIDs, alcohol excess. Among AF patients, the HAS-BLED score has been validated to be predictive of bleeding with VKAs, aspirin, and NOACs.[16–19]

The need to reverse the effect of NOACs occurs in case of life-threatening bleeding, bleeding which cannot be controlled by withdrawal of OAC and usual care, or medical emergencies other than bleeding requiring immediate surgery (Figure 1). Intracranial haemorrhage, with the highest likelihood for fatal outcome, is one of the primary aims for reversal agents. A timely administration of reversal agents is likely to be of major relevance for an improved clinical outcome in case of ICH. However, bleeding at other sites may also be life-threatening and risk stratification is more difficult but essential for these bleeding events. A subanalysis of the RE-LY trial showed that attempts to reverse the effect of dabigatran with clotting factor concentrates (CFCs) were observed in <2% of all major bleeding events.[20] Despite the lack of a specific reversal agent, patients in the NOAC arm showed a trend for reduced mortality associated with major bleeding compared with VKA, most likely due to a shorter half-life and a lower rate of ICH.[20,21] At the present time, available guidance for reversal of uncontrolled bleeding in patients treated with NOACs is based on expert opinion rather than on clinical trials.[22–25]

Figure 1.

Practical management of the reversal of non-vitamin K antagonist oral anticoagulants. *: requiring normal haemostasis (for other conditions than bleeding); **: when approved by respective regulatory authorities; idarucizumab has been approved for reversal of the effect of dabigatran by the European Medicines Agency and the U.S. Food and Drug Administration; ***: antagonization of concomitant anti-platelet therapy: • Reconsider combination therapy and hold off anti-platelet therapy until safe/appropriate to restart. • Consider platelet transfusion if receiving irreversible anti-platelet drug (aspirin, clopidogrel, and prasugrel) or thrombocytopenia. • Consider specific antidote (MEDI2452) if receiving ticagrelor and when antidote approved by respective regulatory authorities and available.