OUP user menu

Acute management of bleeding in patients on novel oral anticoagulants

Deborah M. Siegal, Mark A. Crowther
DOI: http://dx.doi.org/10.1093/eurheartj/ehs408 489-498 First published online: 7 December 2012


Novel oral anticoagulants that directly inhibit thrombin (dabigatran) or factor Xa (rivaroxaban, apixaban) are currently available for prevention of venous thromboembolism (VTE) after orthopaedic surgery, treatment of acute VTE, and prevention of arterial thromboembolism in non-valvular atrial fibrillation. These agents offer advantages over VKAs, including rapid onset, shorter half-lives, fewer drug interactions, and lack of need for routine monitoring. However, there are no established agents to reverse their anticoagulant effect. We review the risk of bleeding with the novel oral anticoagulants and the limitations of conventional coagulation assays for measuring anticoagulant effect. We provide an approach to the management of patients with bleeding complications with evidence for various interventions for reversal, where available.

  • Novel oral anticoagulants
  • Bleeding
  • Management


Novel oral anticoagulant agents that directly inhibit thrombin (dabigatran) or factor Xa (rivaroxaban, apixaban) are currently available for the prevention and treatment of thromboembolism and have been shown to be more effective and/or safer than warfarin and low molecular weight heparin (LMWH) for some indications. Advantages of these agents include rapid onset of action, shorter half-lives, fewer drug interactions, and predictable pharmacokinetics that preclude the need for routine monitoring. However, there are no antidotes to reverse their anticoagulant effect. In this narrative review of novel oral anticoagulants, we summarize published rates of major bleeding, present the use and limitations of coagulation testing, review potential reversal agents, and provide an approach to the management of acute bleeding associated with their use.

Mechanism of action and pharmacokinetics

Thrombin and factor Xa are targets for anticoagulant treatment due to their important role in coagulation. Thrombin catalyses the final step in the coagulation cascade with conversion of fibrinogen to fibrin.1,2 It amplifies its own production through feedback mechanisms and is a potent platelet activator. Factor Xa is an effector in the final common pathway of coagulation which, in conjunction with factor Va, mediates activation of prothrombin to thrombin.

Following oral administration, dabigatran etexilate is rapidly converted to dabigatran, which inhibits thrombin through interaction with its active site.3 Plasma levels peak in 1.25–3 h.4 It is excreted predominantly by the kidneys (80%) and has a half-life of 12–14 h.4 Rivaroxaban selectively and competitively inhibits free and prothrombinase/clot-associated factor Xa through reversible interactions with its active site.3 It achieves peak plasma levels ∼2–4 h after oral administration.5 It has a half-life of 9–13 h and is partially excreted by the kidneys (66%).5,6 Apixaban is an oral, reversible, direct active-site inhibitor of free and clot-bound factor Xa.7 It achieves peak plasma levels within 1–3 h after oral administration, has a half-life of 10–14 h, and is partially excreted by the kidneys (25%).8,9

Novel oral anticoagulants are substrates of the P-glycoprotein (P-gp) transport protein leading to important drug interactions with P-gp inhibitors (e.g. amiodarone, verapamil, ketoconazole, quinidine, clarithromycin, grapefruit juice) and inducers (e.g. carbamazepine, rifampicin, St John's wort, trazodone).3,10 Rivaroxaban and apixaban are metabolized by the CYP3A4 enzyme and co-administration of inhibitors (e.g. azole amtimycotic agents, HIV protease inhibitors) or inducers (e.g. phenytoin, rifampin, phenobarbital, carbamazepine) of this enzyme should be avoided.7,10,11

Indications for use

Dabigatran is approved for the prevention of arterial thromboembolic events in non-valvular atrial fibrillation (Europe, USA, Canada) and VTE prophylaxis following hip and knee-replacement surgery (Europe, Canada). Rivaroxaban is approved for stroke prevention in non-valvular atrial fibrillation (Europe, USA, Canada), prevention of VTE following hip and knee-replacement surgery (Europe, USA, Canada) and treatment of deep vein thrombosis (Europe, Canada). Apixaban is approved for stroke prevention in non-valvular atrial fibrillation (Europe) and VTE prevention after hip and knee replacement surgery (Europe, Canada).

Bleeding risk with novel oral anticoagulant use

Comparison of bleeding rates between agents is hampered by the lack of a common definition of major bleeding. In clinical trials of novel oral anticoagulants, major bleeding rates were generally low and comparable to (or lower than) those with LMWH or warfarin (Table 1). Differences in indication, drug dosage, duration of treatment, and patient characteristics likely account for variability across studies. For example, the study population in ROCKET AF (rivaroxaban) had a higher baseline thromboembolic risk profile (older age, worse renal function) compared with the RE-LY (dabigatran) and ARISTOTLE (apixaban) studies.1214

View this table:
Table 1

Major bleeding rates in published clinical trials of dabigatran, rivaroxaban, and apixaban

Trial acronym First authorIndicationCriteria for major bleeding eventsDose of novel oral anticoagulantComparatorTreatment durationMajor bleeding eventsEstimate of treatment effect (95% CI)
Novel oral anticoagulant, %Comparator, %
 Meta-analysis (RE-MODEL, RE-MOBILIZE, RE-NOVATE) Wolowacz et al.81VTE prevention orthopaedic surgery
  • Clinically overt bleeding with ≥20 g/L fall in Hb or transfusion ≥2 units packed cells or whole blood

  • Fatal bleeding

  • Retroperitoneal, intracranial, intraocular, or intraspinal bleeding

  • Bleeding warranting treatment cessation or reoperation

220 mg once dailyEnoxaparin 40 mg once daily or 30 mg twice daily6–35 days0.6 1.4 2.01.4 0.9 1.6RR 0.42 (0.15–1.19) RR 1.14 (0.46–2.78) RR 1.29 (0.70–2.37) RR 0.94a (0.51–1.75)
 RE-LY, Connolly et al.13ATE prevention atrial fibrillation
  • Reduction in Hb ≥ 20 g/L

  • Transfusion ≥ 2 units blood

  • Symptomatic bleeding in a critical area or organ

110 mg twice dailyWarfarin (INR 2–3)2 years (median)2.7/year3.4/yearRR 0.80 (0.69–0.93)
150 mg twice dailyWarfarin (INR 2–3)3.1/year3.4/yearRR 0.93 (0.81–1.07)
 RE-COVER, Schulman et al.82VTE acute treatmentISTH major bleeding150 mg twice dailyWarfarin (INR 2–3)6 months1.61.9HR 0.82 (0.45–1.48)
Clinically overt bleeding:
 RECORD 1–4, Turpie et al.83VTE prevention orthopaedic surgery
  • Fatal

  • Involving a critical organ

  • Necessitating re-operation

  • Outside of the surgical site and associated with a fall in Hb ≥ 20 g/L, or required ≥ 2 units of blood

10 mg once dailyEnoxaparin 40 mg once daily or 30 mg twice daily12±2 days0.30.2OR 1.62 (0.77–3.53)
30–35 day0.40.2HR 1.84 (0.94–3.62)
 ROCKET AF, Patel et al.12ATE prevention atrial fibrillationISTH major bleeding and/or bleeding causing permanent disability20 mg once dailyWarfarin (INR 2–3)707 days3.6/year3.4/yearHR 1.04 (0.90–1.20)
 EINSTEIN, Bauersachs et al.84DVT acute treatmentISTH major bleeding15 mg twice daily for 3 weeks then 20 mg dailyEnoxaparin 1 mg/kg twice daily then VKA (INR 2–3)3, 6, or 12 months0.81.2HR 0.65 (0.33–1.30)
DVT continued treatment20 mg dailyPlacebo6 or 12 months0.70
 EINSTEIN-PE, Buller et al.85PE acute treatmentISTH major bleeding15 mg twice daily for 3 weeks then 20 mg dailyEnoxaparin 1 mg/kg twice daily then VKA (INR 2–3)3, 6, or 12 months1.12.2HR 0.49 (0.31–0.79)
 Agnelli et al.86VTE acute treatmentISTH major bleeding10 mg twice dailyEnoxaparin 1 mg/kg twice daily then VKA (INR 2–3)12 weeks1.70
20 mg twice daily1.7
30 mg twice daily3.3
40 mg daily1.7
 ATLAS ACS 2, Mega et al.36ACS secondary preventionTIMI major bleeding2.5 mg twice dailyPlacebo13 months (mean)2.1 (combined)0.6HR 3.96 (2.46–6.38)
5 mg twice daily
Acute clinically overt bleeding with ≥1 of:
ADVANCE-1, Lassen et al.87VTE prevention orthopaedic surgery
  • Decrease in Hb ≥ 20 g/L in 24 h

  • Transfusion ≥ 2 units of packed red cells

  • Bleeding at a critical site (i.e. intracranial, intraspinal, intraocular, pericardial, or retroperitoneal bleeding)

  • Bleeding into operated joint, requiring an additional operation or intervention

  • Intramuscular bleeding with compartment syndrome

  • Fatal bleeding

2.5 mg twice dailyEnoxaparin 30 mg twice daily10–14 days0.71.4RR 0.50c (0.24–1.01)
Acute clinically overt bleeding with ≥1 of:
ADVANCE-2, Lassen et al.88VTE prevention orthopaedic surgery
  • Decrease in Hb ≥ 20 g/L in 24 h

  • Transfusion ≥ 2 units packed red blood cells

  • Bleeding at a critical site (intracranial, intraspinal, intraocular, pericardial, or retroperitoneal bleeding)

  • Bleeding into operated joint, needing reoperation or intervention

  • Intramuscular bleeding with compartment syndrome

  • Fatal bleeding

2.5 mg twice dailyEnoxaparin 40 mg once daily10–14 days0.60.9RR 0.65c (0.29–1.45)
Acute, clinically overt bleeding with ≥1 of:
ADVANCE-3, Lassen et al.89VTE prevention orthopaedic surgery
  • Decrease in Hb ≥ 20 g/L in 24 h

  • Transfusion ≥ 2 units packed red cells

  • Bleeding at a critical site (intracranial, intraspinal, intraocular, pericardial, and retroperitoneal bleeding)

  • Bleeding into operated joint, necessitating reoperation or intervention

  • Intramuscular bleeding with the compartment syndrome

  • Fatal bleeding

2.5 mg twice dailyEnoxaparin mg once daily35 days0.80.7RR 1.21c (0.65–2.26)
AVERROES, Connolly et al.90ATE prevention atrial fibrillationISTH major bleeding5 mg twice dailyAspirin (81 to 324 mg daily)1.1 year (mean)1.4/year1.2/yearHR 1.13 (0.74–1.75)
ARISTOTLE, Granger et al.14ATE prevention atrial fibrillationISTH major bleeding5 mg twice dailyWarfarin (INR 2–3)1.8 years (median)2.1/year3.1/yearHR 0.69 (0.60–0.80)
APPRAISE-2, Alexander et al.37ACS secondary preventionTIMI major bleeding5 mg twice dailyPlacebo240 days (median)bTIMI: 2.4/yearTIMI:0.9/yearHR 2.59 (1.50–4.46)
ISTH major bleedingISTH: 5.1/yearISTH: 2.0yearHR 2.48 (1.72–3.58)
ADOPT, Goldhaber et al.91VTE prevention medically ill patients
  • Fatal bleeding

  • Overt with ≥1 of the following:

  • Decrease in Hb ≥ 20 g/L over 24 h

  • Transfusion ≥ 2 units of packed red cells

  • Intracranial, intraspinal, intraocular, pericardial, or retroperitoneal bleeding

  • Bleeding that occurred in an operated joint that required reoperation or intervention

  • Intramuscular bleeding with the compartment syndrome

2.5 mg twice dailyEnoxaparin 40 mg once dailyApixaban 30 days0.50.2RR 2.58 (1.02–7.24)
Enoxaparin 6–14 days
Levine et al.92bVTE prevention in cancer patients receiving chemotherapyISTH major bleeding5 mg once dailyPlacebo12 weeks03.4
10 mg once daily0
20 mg once daily6.3RR 1.81c (0.25–13.5)
  • Major bleeding events associated with the use of dabigatran, rivaroxaban, and apixaban as published in completed clinical trials.

  • ACS, acute coronary syndrome; ATE, arterial thromboembolism; CI, confidence interval; DVT, deep vein thrombosis; HR, hazard ratio; INR, international normalized ratio; ISTH, International Society on Thrombosis and Haemostasis; OR, odds ratio; PE, pulmonary embolism; RR, relative risk; TIMI, Thrombolysis in Myocardial Infarction; VKA, vitamin K antagonist; VTE, venous thromboembolism.

  • ISTH major bleeding in non-surgical patients: fatal bleeding, and/or symptomatic bleeding in a critical area or organ such as intracranial, intraspinal, intraocular, retroperitoneal, intra-articular, or pericardial, or intramuscular with compartment syndrome, and/or bleeding causing a fall in haemoglobin levels of 20 g L−1 (1.24 mmol L−1) or more, or leading to transfusion of two or more units of whole blood or red cells.93

  • TIMI major bleeding: fatal bleeding, and/or intracranial bleeding, and/or clinically overt bleeding associated with a fall in Hb ≥ 50 g/L.94

  • aRandom effects analysis.

  • bTrial stopped early due to slow accrual.

  • cFor studies which did not report an estimate of treatment effect, relative risk and 95% CIs were calculated using previously described methods.95,96

A recent meta-analysis examined the safety and efficacy of novel oral anticoagulants compared with warfarin for prevention of stroke and systemic embolism in atrial fibrillation.15 Patients receiving novel oral anticoagulants had a reduced risk of stroke and systemic embolism (RR 0.78, 95% CI 0.67–0.92) and all cause-mortality (RR 0.88, 95% CI 0.82–0.95), and a similar risk of major bleeding compared with those receiving warfarin (RR 0.88, 95% CI 0.71–1.09). Intracranial bleeding was reduced in patients receiving novel oral anticoagulants (RR 0.49, 95% CI 0.36–0.66).

Systematic large-scale documentation of novel oral anticoagulant use relative to other antithrombotic therapies is required to define bleeding risk outside the clinical trial setting.16 Currently available post-marketing reporting databases lack comparative vitamin K antagonist (VKA) data and do not reflect the total number of patients receiving novel agents thereby limiting interpretation. In November 2011, the EudraVigilance database reported 256 post-marketing episodes of serious bleeding resulting in death associated with use of dabigatran worldwide.17 The US Food and Drug Administration (FDA) recently concluded there is no evidence that bleeding rates are higher with dabigatran than seen with other anticoagulants.18

Several recently published case reports highlight physician concerns regarding the management of serious bleeding due to a lack of reversal agents.1923 However, the clinical impact of anticoagulant reversal in the event of serious bleeding is uncertain even when reversal agents are available such as for VKAs.2427

Elderly patients

Age is an established risk factor for anticoagulant-associated bleeding.28 Data regarding the impact of age on bleeding risk with novel agents are limited to subgroup analyses of large clinical trials not adequately powered to assess safety in the elderly. Patients >75 years receiving dabigatran (150 mg daily) for VTE prophylaxis following orthopaedic surgery had a similar risk of major bleeding compared with those receiving enoxaparin (40 mg once daily).29 In the RE-LY study, patients ≥75 years receiving dabigatran 110 mg twice daily for prevention of stroke and systemic embolism in atrial fibrillation had a similar risk of extracranial bleeding, while those receiving dabigatran 150 mg twice daily had a trend towards increased extracranial bleeding compared with warfarin.30 Dabigatran was associated with reduced risk of intracranial haemorrhage irrespective of age. Subgroup analyses of the ROCKET AF and ARISTOTLE trials showed no effect of age on bleeding risk.12,14 While further studies are needed, regulatory agencies in Europe and Canada recommend dose reduction when using dabigatran in patients ≥80 years.3,30

Renal failure

Renal function, a predictor of bleeding with anticoagulant use in general, should be assessed prior to initiation of therapy with novel oral anticoagulants and monitored during treatment.28,31 Monitoring is particularly important for dabigatran, which is excreted by the kidneys to a greater extent than rivaroxaban or apixaban.

Well-powered studies examining the impact of renal function on bleeding risk with novel oral anticoagulants are lacking. In a subgroup analysis of the RE-MODEL and RE-NOVATE data, patients with moderate renal impairment (CrCl 30–50 mL/min) taking dabigatran for VTE prophylaxis after orthopaedic surgery had fewer major bleeding events than those receiving enoxaparin (0% vs. 4.7%).29 In atrial fibrillation trials, patients with moderate renal dysfunction (CrCl 30–50 mL/min) had increased bleeding rates irrespective of treatment with novel oral anticoagulants or warfarin.13,14,32 In the ROCKET AF study, a reduced dose of rivaroxaban (15 mg once daily) was used for patients with moderate renal impairment due to pharmacokinetic considerations. Subgroup analyses of patients with CrCl 30–50 mL/min in the RE-LY and ROCKET AF studies showed a similar risk of bleeding with novel agents and warfarin.13,32 In a subgroup analysis of the ARISTOTLE study, a greater reduction in major bleeding was seen in patients with CrCl 25–50 mL/min receiving apixaban.14 Based on RE-LY study data, dose reduction of dabigatran is not currently recommended for patients with moderate renal impairment. However, subgroup analysis may lack sufficient power to detect differences and further well-powered studies are needed to determine the safety and efficacy of reduced doses in patients with moderate renal dysfunction. Patients with severe renal impairment (CrCl < 25–30 mL/min) were excluded from large clinical trials and should not receive novel oral anticoagulants.1214

Concurrent use of antiplatelet agents

Limited data suggest that aspirin and non-steroidal anti-inflammatory drugs (NSAIDs) may not increase bleeding in patients receiving dabigatran for VTE prophylaxis; in a post-hoc analysis of pooled data from the RE-MOBILIZE, RE-MODEL, and RE-NOVATE studies, patients receiving dabigatran and concomitant aspirin (<160 mg/day) or NSAIDs (half-life < 12 h) had a similar risk of major bleeding as those receiving dabigatran alone.33

When given in addition to therapeutic doses of dabigatran in patients with atrial fibrillation in the RE-LY study, aspirin increased the risk of intracranial bleeding (RR 1.6).34 In the phase II dose-finding PETRO trial, aspirin used in conjunction with dabigatran for stroke prevention in atrial fibrillation increased the risk of bleeding events in the group receiving doses of dabigatran higher than those currently approved (300 mg twice daily).35 There was no increased risk of bleeding when aspirin was added to lower doses of dabigatran (50 mg or 150 mg), but these analyses were limited by small sample size.

In patients with acute coronary syndrome, the addition of apixaban and rivaroxaban to dual antiplatelet therapy (aspirin and clopidogrel) increased the risk of major bleeding.36,37 However, there may be clinical benefit with concomitant therapy as shown by decreased cardiovascular mortality (HR 0.66; 95%CI 0.51–0.86) and all-cause mortality (HR 0.68; 95%CI 0.53–0.87) with rivaroxaban 2.5 mg twice daily compared with placebo in the ATLAS ACS 2 trial.36

Coagulation testing

Emergency situations (e.g. haemorrhage, overdose, urgent surgery/invasive procedure) necessitate rapid assessment of coagulation status. However, conventional coagulation assays have limitations when used to measure novel oral anticoagulant effect. Modified conventional tests and alternative assays have increased reliability and accuracy, but are not standardized or routinely available. Table 2 shows the effect of novel oral anticoagulants on coagulation tests.

View this table:
Table 2

Effect of novel oral anticoagulants on commonly used coagulation tests

Novel anticoagulantProthrombin time (PT)Activated partial thromboplastin time (aPTT)Thrombin clotting time (TCT)Ecarin clotting timeHaemoclot assayAnti-factor Xa activity
Dabigatran↑ or no change (low sensitivity, varies with reagents)↑ (varies with reagents)aND
Rivaroxaban↑ or no change (not sensitive at low concentrations, varies with reagents)↑ or no change (less sensitive than PT)a (sensitive and specific when calibration curve used)
Apixaban↑ or no change (other tests more sensitive, may vary with reagents)↑ or no change (other tests more sensitive, may vary with reagents)aa
  • ND, no data.

  • aPreferred test. Adapted from previously published review articles.41,59

Dabigatran prolongs the activated partial thromboplastin time (aPTT) and thrombin clotting time (TCT, also known as thrombin time or TT).3840 However, the aPTT varies with reagents used and does not accurately reflect the amount of dabigatran present.39,40 The TCT assay is the most effective test for detecting the presence of dabigatran, even in small quantities. It has a linear relationship with dabigatran concentration, but rapidly prolongs beyond the normal detection limits and is not routinely available.38 The ecarin clotting time uses a viper-venom-derived metalloprotease to determine the activity of direct thrombin inhibitors, but is not standardized or widely available.41,42 A commercial dabigatran calibrator, the Hemoclot direct thrombin inhibitor assay (HYPHEN BioMed, Neuville Sur Oise, France), provides accurate, reproducible measures of dabigatran anticoagulant activity and plasma concentrations using a dilute TT test in conjunction with a dabigatran calibration curve.43 It is licensed and approved for clinical use in Europe and Canada, but is approved only for research purposes in the USA.44 The prothrombin time (PT) has low sensitivity for measuring the effects of dabigatran and is not recommended for this purpose; however, a prolonged PT reliably predicts the presence of dabigatran in quantities sufficient to produce an anticoagulant effect.4,9,41

For rivaroxaban, the PT is useful for detecting the presence of drug if prolonged; however, the assay is not sensitive at low concentrations and the degree of prolongation does not reliably predict the amount of drug present.4547 The relationship between drug concentration and PT prolongation is influenced by the reagents used for testing. Commercially available chromogenic anti-factor Xa assays can be used to measure rivaroxaban levels, but sensitivity varies between assays.41,46 When used with a rivaroxaban calibration curve, chromogenic anti-factor Xa assays are sensitive and specific for rivaroxaban plasma concentrations.48

Apixaban may prolong the PT, but this assay lacks sensitivity in comparison to the dilute PT test (or modified PT).41,49 The HepTest (American Diagnostica), a commercially available clot-based anti-Factor Xa assay, is more sensitive than the PT and equally sensitive as the dilute PT to apixaban.41,49 Chromogenic anti-factor Xa assays also provide more accurate apixaban levels compared with the PT.46

Levels of all of the novel agents can be determined using chemical techniques such as liquid chromatography–mass spectrometry.8,50,51 These techniques are available only in selected biochemistry laboratories.

Despite their limitations, conventional coagulation assays may provide qualitative information regarding the presence of drug. A normal TCT in patients receiving dabigatran, normal PT in patients receiving rivaroxaban, and normal anti-factor Xa activity in patients receiving apixaban suggest very low drug levels and intact haemostatic function. Although it is less precise, aPTT can be used to determine dabigatran effect if TCT testing is unavailable. Test results should be interpreted relative to the time of drug administration and pharmacokinetics. Further research studies are required to confirm the predictive ability of these tests in the clinical setting. In the interim, procedures should be delayed, if possible, depending on procedural bleeding risk, drug pharmacokinetics, and renal function. Procedures with standard bleeding risk (e.g. uncomplicated laparoscopic procedures, colonoscopy without polyp removal) should be delayed 24 h in patients with normal renal function.52 A longer delay should be implemented for patients with impaired renal function (2–4 days) and/or procedures with high risk of bleeding (2–6 days; e.g. major cardiac surgery, epidural or spinal anaesthesia, neurosurgery, pacemaker/defibrillator insertion).38,52,53 Recently published data from the RE-LY study showed similar major bleeding event rates in patients requiring discontinuation of dabigatran or warfarin for invasive procedures (dabigatran 110 mg: HR 0.79, 95% CI 0.35–1.79; dabigatran 150 mg: HR 1.20, 95% CI 0.55–2.60) using a dabigatran discontinuation algorithm based on procedural bleeding risk and renal function.54

Anticoagulant reversal strategies

Specific reversal agents

A monoclonal antibody targeted against dabigatran is currently under development. It potently and specifically inhibited dabigatran anticoagulant activity in human plasma in vitro and in rats in vivo.55

Plasma-derived and recombinant factor Xa (pd-factor Xa and r-factor Xa) which lack catalytic and membrane-binding activity are being investigated as antidotes for factor Xa inhibitors.56,57 Preliminary studies showed that both pd-factor Xa and r-factor Xa are capable of neutralizing the coagulation test abnormalities induced by rivaroxaban and apixaban in vitro and in animal models.56,57

Coagulation factor replacement

To our knowledge, there are no data regarding the use of fresh frozen plasma (FFP) in patients with novel oral anticoagulant associated bleeding. In mice receiving high-dose dabigatran, FFP reduced the volume of intracerebral haemorrhage, but had no effect on mortality.58 FFP carries risks of volume overload and, rarely, allergic reactions, and infection.59

Prothrombin complex concentrates (PCC) contain high doses of vitamin K-dependent coagulation factors and variable amounts of proteins C and S. Four-factor PCC (factors II, VII, IX, X) is commercially available in Europe and Canada, while 3-factor PCC (factors II, IX, X) is licensed for use in the USA. In healthy volunteers, a high dose of 4-factor PCC (50 U/kg) reversed PT prolongation due to rivaroxaban, but did not correct aPTT prolongation due to dabigatran.60 It is unclear whether correction of laboratory tests would correspond to amelioration of bleeding. In animal models, high-dose 4-factor PCC reduced intracranial haematoma expansion and 24 h mortality in mice pre-treated with dabigatran.58 Administration of 4-factor PCC to rivaroxaban-treated rabbits resulted in partial correction of laboratory parameters, but no correction of bleeding.61 PCC improved thrombin generation indices following the addition of apixaban in vitro to aliquots of blood from healthy donors.62 While there is no clinical evidence to support its use, a dose of 50 IU/kg of 4-factor PCC might be reasonable for severe/life-threatening bleeding in patients taking novel oral anticoagulants.52

PCC use is associated with an increased risk of thrombosis. A meta-analysis of observational studies found a 1.4% incidence of thromboembolic events in VKA-treated patients receiving PCC for bleeding or urgent surgical procedures.63

Prohaemostatic agents

Recombinant activated factor VII (rVIIa) was developed for the treatment of bleeding episodes in haemophiliac patients with inhibitors to factors VIII and IX.64 It has been used outside of its approved indication as a potential treatment of severe/life-threatening bleeding in non-haemophiliac patients with mixed results.6467 There is concern regarding thrombogenicity in non-haemophiliac patients.67,68 In a meta-analysis, arterial thromboembolic events were increased in non-haemophiliac patients treated with rVIIa compared with placebo (4.5 vs. 2.0%).69 As a result, the FDA issued a black box warning regarding the use of rVIIa outside its approved indications.70

RVIIa has no demonstrated efficacy for reversing bleeding complications associated with novel oral anticoagulants. In healthy volunteer subjects receiving the oral direct thrombin inhibitor melagatran, rVIIa had no effect on aPTT, thrombin generation, and platelet activation.71 In an animal model, rVIIa failed to stabilize induced intracerebral haematoma expansion in mice receiving dabigatran.58 Similarly, a randomized controlled animal model showed that rVIIa did not reverse bleeding and only partially corrected laboratory abnormalities induced by rivaroxaban.61 RVIIa improved thromboelastometry parameters after the addition of apixaban in vitro to blood from healthy donors.62 These results provide little support for the use of rVIIa to reverse bleeding associated with the novel oral anticoagulants. Because no human studies have been conducted to date, it is unclear whether this therapy is of benefit in emergent anticoagulant reversal.

Activated PCC [aPCC, commercially available as factor eight inhibitor bypassing activity (FEIBA), Baxter Bioscience, Vienna, Austria] is a coagulation complex containing factors II, VII, IX, and X which are activated during the manufacturing process.72 Like rVIIa, aPCC was developed for haemostatic control in haemophiliac patients with inhibitors to factors VIII or IX.72 Activated PCC has been shown to correct the anticoagulant effect of high-dose rivaroxaban in animal models.73,74 In human plasma incubated with dabigatran, aPCC reduced clot initiation time in vitro.75 Even at low doses (75–80 U/kg), aPCC corrected thrombin generation parameters in vitro in plasma from healthy volunteers receiving single doses of rivaroxaban or dabigatran and in blood from healthy volunteers following the addition of apixaban.62,76

Activated PCC safety data are limited to pharmacovigilance studies in haemophilia patients showing a low risk of thrombosis (4–8 per 105 infusions).77,78 However, a majority of thrombotic events (81%) occurred in patients with risk factors for thrombosis raising concerns about applying these data to non-haemophiliac patients requiring anticoagulant therapy for thrombosis prevention or treatment.77,79

Oral activated charcoal

An in vitro model using dabigatran suspended in acidic water demonstrated absorption of >99.9% of drug after the addition of activated charcoal, suggesting it may be effective for reducing dabigatran absorption following recent ingestion.38 To our knowledge, there are no data regarding the use of oral activated charcoal with rivaroxaban or apixaban.


Acute haemodialysis can be considered for removal of dabigatran in cases of severe bleeding, or in patients with renal impairment. In patients with end-stage renal disease up to 68% of active dabigatran was removed after 4 h of haemodialysis.80 Rivaroxaban and apixaban may not be dialysable due to a high degree of protein binding.7,11 Charcoal haemoperfusion removes highly protein-bound drugs; however, this intervention is not evidence-based, has limited availability and should only be considered in the context of a clinical trial.

Adjunctive therapies

There are no clinical data regarding the efficacy of desmopressin (DDAVP) and antifibrinolytic agents (tranexamic acid and ɛ-aminocaproic acid) for the reversal of bleeding with the novel oral anticoagulants. They may be considered as adjunctive therapies in the event of severe bleeding. Use of desmopressin requires close monitoring of serum electrolytes as it is associated with clinically significant hyponatraemia which may lead to seizures.

Suggested management of the actively bleeding patient

Figure 1 provides a suggested approach to the management of acute bleeding in patients receiving novel oral anticoagulants. Actively bleeding patients should undergo rapid assessment with attention to haemodynamic stability and identification of bleeding source followed by risk stratification. Time elapsed since last dose and renal function (particularly for dabigatran) should be ascertained as they may influence the threshold for implementation of interventions described subsequently.

Figure 1

Suggested management of acute bleeding in patients taking novel oral anticoagulants

Minor bleeding (e.g. epistaxis, ecchymosis, menorrhagia) should be managed using local haemostatic measures. A short period of anticoagulant withdrawal may be considered, but should be balanced against individual thromboembolic risk and followed by re-initiation of drug.

In the event of moderate bleeding (e.g. upper or lower gastrointestinal bleeding), anticoagulants should be discontinued. An extended period of withdrawal may be required depending on the feasibility of local or mechanical haemostasis. Patients should be monitored for deterioration and referred for definitive haemostatic interventions. The addition of a low-dose parenteral anticoagulant may be considered for patients at high risk of thromboembolic events.

Severe/life-threatening bleeding requires transfer to an intensive care setting with provision of life-supporting therapies (e.g. volume replacement, vasopressors, mechanical ventilation) as required. Urgent referral should be made for procedural or surgical haemostasis. Anticoagulant medications should be withdrawn and haemostatic function and complete blood count assessed. Packed red blood cells (PRBC) should be transfused in response to actual or expected severe anaemia. Coagulation factor replacement with FFP should be administered to patients in whom coagulation factor levels are abnormally low (e.g. dilutional coagulopathy, disseminated intravascular coagulation). Platelet transfusions can be given to patients receiving concurrent antiplatelet therapies. A prohaemostatic agent may be a reasonable addition to maximum supportive therapies for severe/life-threatening bleeding. Based on available and methodologically limited data, we recommend 4-PCC (50 IU/kg) over aPCC (80 U/kg) for rivaroxaban or apixaban. For dabigatran, aPCC (80 U/kg) is preferred over 4-PCC (50 IU/kg). We acknowledge a lack of clinical data to guide the use of these agents and recommend consideration of the increased thrombosis risk prior to administration. Adjunctive therapies such as desmopressin or antifibrinolytic agents may be added. Haemodialysis can be considered for dabigatran removal if practically feasible.


The novel oral anticoagulants which directly inhibit thrombin (dabigatran) or factor Xa (rivaroxaban and apixaban) have significant advantages over VKAs and LMWH for the prevention and treatment of thromboembolism. Use of novel oral anticoagulants may avoid bleeding complications in some patients such as those with atrial fibrillation in whom intracranial haemorrhage is significantly reduced with these agents. However, there are no specific antidotes to reverse their anticoagulant effect in the event of bleeding. Currently, there is a paucity of clinical evidence regarding the efficacy of therapies such as FFP, PCC, aPCC, and rVIIa. However, administration of a 4-factor PCC or aPCC may be a reasonable addition to maximum supportive measures in cases of severe/life-threatening bleeding associated with these agents. Further studies are needed to clarify the role of these treatments, and others, for the reversal of the novel oral anticoagulants and management of bleeding complications. In the interim, aggressive supportive management and prompt consideration of procedural/surgical haemostatic intervention remain the mainstays of treatment.

Conflicts of interest: D.M.S. has no conflict of interest to declare. M.A.C. has sat on advisory boards for Leo Pharma, Pfizer, Bayer, Boehringer Ingelheim, Alexion, CSL Behring, and Artisan Pharma. M.A.C. has prepared educational materials for Pfizer, Octapharm, and CSL Behring. M.A.C. has provided expert testimony for Bayer. M.A.C. holds a Career Investigator award from the Heart and Stroke Foundation of Canada, and the Leo Pharma Chair in Thromboembolism Research at McMaster University. M.A.C.'s institution has received funding for research projects from Boehringer Ingelheim, Octapharm, Pfizer, and Leo Pharma.


View Abstract