Background
For more than 60 years, the only oral anticoagulants available for the prevention of ischaemic stroke in patients with non-valvular atrial fibrillation (AF) have been the vitamin K antagonists (VKAs), such as warfarin, phenprocoumon and acenocoumarol [
1]. However, the need for coagulation monitoring and dose adjustments, as well as concerns about drug–drug or diet–drug interactions and the risk of bleeding, have restricted the use of VKAs for ischaemic stroke prevention in patients with AF [
2,
3]. Several direct oral anticoagulants (DOACs) have been approved for this indication: the oral direct thrombin inhibitor dabigatran, and oral direct factor Xa inhibitors (e.g. rivaroxaban, apixaban or edoxaban) [
4,
5]. Compared with VKAs, DOACs produce a more predictable anticoagulant effect and can be given in fixed doses without routine coagulation monitoring. The results of randomised controlled trials (RCTs) and observational studies indicate that DOACs are at least as effective as VKAs for stroke prevention in patients with AF, with reduced rates of intracranial bleeding and, at certain doses, a reduction in life-threatening bleeding [
6‐
10].
Dabigatran, the active moiety of dabigatran etexilate, has a rapid onset of action, and the plasma concentration peaks within 0.5–2.0 hours of administration [
11]. This anticoagulant has a half-life of 7–17 hours [
12] and is eliminated predominantly via renal excretion (80 %) [
13]. Dabigatran is licensed in many countries for several indications, including: primary prevention of venous thromboembolism (VTE) in patients who have undergone elective total hip or knee arthroplasty (150 mg or 220 mg once daily); prevention of ischaemic stroke and systemic embolism in adult patients with non-valvular AF (110 or 150 mg twice daily, except in the USA where 75 and 150 mg twice-daily doses are approved); and treatment of acute deep vein thrombosis (DVT) and pulmonary embolism (110 and 150 mg twice daily, except in the USA where only the 150 mg twice-daily dose is approved) [
14,
15].
Most episodes of bleeding in patients treated with dabigatran can be managed with supportive measures and by temporarily withholding the drug. However, additional strategies may be needed in patients with life-threatening bleeding and those who require urgent or emergency surgery or other invasive procedures for which haemostasis is necessary. Supportive care is also not sufficient in patients with intracranial bleeding, where outcome may be directly associated with the time needed for coagulation reversal [
16].
A specific reversal agent for dabigatran, idarucizumab, has been approved by the US Food and Drug Administration. Animal models and phase I–III clinical data show that idarucizumab achieves predictable, complete and sustained reversal of dabigatran, with the potential for significant reductions in blood loss [
17‐
19]. In an interim analysis of the phase III RE-VERSE AD study [
19], involving 90 dabigatran-treated patients with serious bleeding or in need of an urgent surgical procedure, idarucizumab completely reversed the effects of dabigatran within minutes. Idarucizumab appears to be well tolerated, and it has no direct effects on procoagulant or anticoagulant activity [
17,
19]. However, idarucizumab is yet to be approved in many countries and, pending its widespread availability, multiple therapeutic options have been suggested for emergency reversal of dabigatran’s anticoagulant effects. These options include prothrombin complex concentrates (PCCs) and activated PCCs (aPCCs), as well as recombinant activated factor VII (rFVIIa) [
20‐
24].
In this article we review pre-clinical and clinical evidence for the use of PCCs and aPCCs to restore haemostasis in dabigatran-treated patients with either haemorrhage or the need for an urgent surgical procedure, and review the role of laboratory coagulation assessments in this setting [
20‐
24].
General bleeding management of the dabigatran-anticoagulated patient
Expert and professional society guidelines are available regarding emergency bleeding management in patients receiving dabigatran treatment [
28‐
32]. There are variations in the specifics but the principles are consistent across different guidelines. The general approach to managing non-emergency bleeding complications is similar to that with pre-surgical management, i.e. discontinue dabigatran therapy temporarily and wait for the elimination of dabigatran. In this review, we will focus on major or life-threatening bleeding (including intracranial bleeding) and emergency surgery, where a need for more rapid reversal of the effects of dabigatran necessitates a different strategy. Exogenous coagulation factor repletion with PCCs or aPCCs has been suggested as a potential treatment option in these settings [
33].
The initial steps of bleeding management algorithms typically consist of local/surgical haemostasis where appropriate, qualitatively assessing anticoagulant activity (e.g. by activated partial thromboplastin time (aPTT)), and general measures such as volume replacement and blood product transfusion. Medical history including anticoagulant intake is integral to the early phase of treatment. In many cases, however, anticoagulation may not be contributing meaningfully to the bleeding and so reversal of dabigatran is not usually a first-line priority. In the RE-VERSE AD study [
19], 22 out of 90 enrolled patients did not have prolonged diluted thrombin times (dTTs) due to the natural clearance of dabigatran. Acquired coagulopathy can develop secondary to blood loss, loss and consumption of coagulation factors and haemodilution owing to excessive fluid replacement. This is a major risk factor for progression from initial bleeding to severe haemorrhage. There is some evidence suggesting that ‘restrictive’ or goal-directed as opposed to ‘liberal’ fluid resuscitation strategies may reduce morbidity and lengths of hospital stay [
34,
35]. However, other investigations have cast doubt on the benefits of a restrictive approach [
36] and, importantly, hypovolaemia can cause acidosis, thereby exacerbating coagulopathy. In the presence of substantial tissue injury and haemorrhagic shock, activation of protein C and subsequent hyperfibrinolysis may also aggravate coagulopathy [
37]. Early intervention with haemostatic therapy (e.g. fibrinogen concentrate, cryoprecipitate, fresh frozen plasma, platelets) may be critical for preventing complex coagulopathies and progression to severe, life-threatening haemorrhage—especially in patients who have bled so that they have acquired coagulopathy in addition to anticoagulant therapy. Multimodal therapy should therefore be administered as early as possible in life-threatening bleeding under dabigatran anticoagulation [
38].
Laboratory evaluation of dabigatran concentration
Although patients taking dabigatran do not need to undergo routine coagulation monitoring, rapid assessment of whether or not the patient is actively anticoagulated is important in an emergency situation. This information can help determine the contribution of anticoagulation to the bleeding, the need for a reversal/repletion strategy and whether an invasive procedure should be delayed [
39,
40].
A variety of tests have been explored for the detection or quantification of plasma dabigatran activity, but limitations are common. For example, the prothrombin time (PT) and the international normalised ratio (INR) exhibit low sensitivity to anticoagulation with dabigatran, with therapeutic doses having minimal effect; these tests are therefore not recommended [
31]. Thrombin-generation assays and viscoelastic testing parameters have also been considered, but thrombin-generation assays are mostly restricted to research settings and viscoelastic testing has not been validated for the monitoring of dabigatran in a clinical setting [
40]. The aim of testing for dabigatran activity is often not to provide precise quantification of dabigatran, but simply to detect the presence or absence of drug in plasma. This is particularly important in patients with acute ischaemic stroke, for whom thrombolytic therapy or procedures requiring high-dose antiplatelet agents are being considered. Candidate tests are now described in more detail and are summarised in Table
1.
Table 1
Clinical and research tests available for the assessment of plasma dabigatran concentration
Thrombin time | Linear dose response; oversensitive | Widely available | Recommended for detection of presence or absence of dabigatran activity |
Activated partial thromboplastin time (aPTT) | Non-linear dose response | Widely available | Recommended for semi-quantitative estimation of dabigatran activity; normal aPTT does not always exclude presence of dabigatran |
Direct thrombin inhibition assays | Linear dose response; sensitive | Usually available in specialised centres | Recommended for measurement of plasma concentration; ECT usually a local assay because commercial kits not available; not validated for dabigatran |
Diluted thrombin time |
Ecarin clotting time (ECT) |
Thromboelastometry (ROTEM®)/thrombelastography (TEG®) assays EXTEM/Rapid TEG | Sensitive | Limited availability | Potential measurement of effectiveness of treatment for dabigatran-induced anticoagulation; not validated for dabigatran |
Thrombin-generation assays | Sensitive (lag time) | Limited availability | Potential measurement of effectiveness of treatment for dabigatran-induced anticoagulation; not validated for dabigatran |
Prothrombin time or international normalised ratio | Low sensitivity | Widely available | Not recommended in this setting |
The aPTT exhibits a non-linear dose response with increasing concentrations of dabigatran, plateauing at higher concentrations [
31]. In addition, in some patients taking dabigatran, normal-range aPTT values have been reported when dabigatran is at trough levels [
41]. This test is therefore not suitable for precise quantification, particularly at high or low dabigatran concentrations, but can provide an approximate indication of dabigatran levels [
42]. The aPTT is readily available and therefore is commonly used to evaluate dabigatran activity or possible drug ingestion; normal aPTT results reduce the likelihood of therapeutic anticoagulation.
The thrombin time (TT) exhibits a very steep, linear dose response with increasing concentrations of dabigatran. This test can be considered excessively sensitive for the detection of dabigatran activity because samples may not clot at dabigatran concentrations above 100 ng/ml, a level within the expected clinical range [
41,
43]. Therefore, the use of this assay for quantifying therapeutic concentrations of dabigatran is limited. However, the TT is a very useful assay to determine the presence of low levels of dabigatran, with TT values within the normal range suggesting the absence of dabigatran.
The TT and aPTT assays are useful tests in the clinical setting and can be used in combination, with the TT detecting the presence or absence of drug and the aPTT providing an approximate indication of the plasma dabigatran concentration. These assays are readily accessible and can provide information rapidly.
Tests that are used principally for research have also been considered for assessing dabigatran anticoagulation. For example, dedicated direct thrombin inhibition assays have a linear relationship over a wide range of plasma dabigatran concentrations [
44]. The dTT is one type of direct thrombin inhibition assay, in which plasma is diluted in buffer and then supplemented with normal human plasma; clotting is then initiated with thrombin. This methodology compensates for any coagulation factor deficiencies and only the effect of the direct thrombin inhibition is measured [
45]. Ecarin tests, representing another type of direct thrombin inhibition assay, are either clotting based (ecarin clotting time (ECT)) or chromogenic (ecarin chromogenic assay (ECA)) [
31,
45]. Ecarin is snake venom that directly activates prothrombin to initiate clotting, thereby bypassing upstream coagulation and enabling direct measurement of the influence of the direct thrombin inhibition. Some ECA tests are supplemented with prothrombin to allow targeted direct thrombin inhibition quantification independent of any coagulopathies in the patient sample. In a recent study, the dTT, ECT and ECA all showed linear correlations with dabigatran concentrations over a broad range [
41]. Although these tests are widely available, they are not used in all hospitals [
46].
The thrombin-generation parameter ‘lag time’ and, to some extent, the endogenous thrombin potential (ETP) have been shown to correlate with plasma dabigatran concentrations [
47]. Recent studies have also shown that therapeutic dabigatran doses have a significant effect on the lag time, ETP and peak height, when measured by the calibrated automated thrombogram (CAT) [
48]. However, thrombin-generation tests are time-consuming and their availability is generally limited to research laboratories and centres specialising in haemostasis. Thrombin-generation parameters therefore cannot be recommended for routine evaluation of plasma dabigatran concentration.
Viscoelastic tests, including thromboelastometry (ROTEM
®) or thrombelastography (TEG
®), can have faster turnaround times than standard laboratory coagulation tests because they are whole-blood based and are often performed in surgical or emergency rooms. This can be beneficial in emergency situations. The ROTEM
® EXTEM and Rapid TEG (r-TEG) assays measure tissue-factor-initiated extrinsic coagulation and pre-clinical data have shown that they are sensitive to dabigatran [
49]. There is anecdotal evidence that dabigatran prolongs the activated clotting time (ACT) in r-TEG, when all other standard coagulation tests were within the normal range [
50]. However, neither EXTEM nor r-TEG has been calibrated for measuring dabigatran levels; in many centres these assays are not used routinely and there is currently no clinical evidence supporting the use of these tests for this purpose.
Prothrombin complex concentrates (activated and non-activated)
If the patient continues to sustain major blood loss after initial intervention with haemostatic therapy, or in the setting of intracranial haemorrhage, treatment algorithms recommend further interventions to counteract the anticoagulant effects of dabigatran (Table
2) [
28‐
30,
32]. These interventions include PCCs and aPCCs.
Table 2
Recommendations and algorithms for the management of bleeding patients with dabigatran-induced anticoagulation
| Discontinue treatment until bleeding resolves | Sequential treatment: | aPCC (50 IU/kg) |
(1) PCC (40 IU/kg) | If unavailable, give PCC (40 IU/kg) or rFVIIa (90 μg/kg) |
(2) aPCC (50 IU/kg) |
(3) rFVIIa (90 μg/kg) |
(4) Haemodialysis for 6–8 h or charcoal filtration |
Faraoni et al., 2015 [ 29] | No recommendation given | No recommendation given | (1) Monitor blood loss and perform coagulation assays |
(2) Standard resuscitation with fluid therapy, tranexamic acid (1 g), RBCs and massive transfusion protocola
|
(3) Four-factor PCC (25–50 IU/kg), aPCC (FEIBA; 30–50 IU/kg) |
| Maintain diuresis | Same recommendation as for mild bleeding | PCC 50 U/kg (additional 25 U/kg if clinically needed)aPCC 50 U/kg (maximum 200 U/kg/day)rFVIIa (90 μg/kg)Idarucizumab 5 g intravenously |
Local haemostatic measures |
Fluid replacement | |
RBC substitution if necessary |
Platelet substitution if necessary | |
FFP as plasma expander (not as reversal agent) | |
Consider tranexamic acid or desmopressin |
Consider dialysis |
| No recommendation given | No recommendation given | PCC, aPCC or rFVIIa may be used as non-specific antagonists |
A number of PCCs are commercially available. Detailed analysis of constituent differences between these products has been published previously [
51]. All PCCs contain the vitamin K-dependent factors II, IX and X, and are standardised according to their factor IX content. In addition, they contain differing amounts of factor VII; products with low or high quantities of factor VII are known as three-factor or four-factor PCCs, respectively. Some PCCs also contain anticoagulation proteins such as protein C, protein S, protein Z, antithrombin and heparin. Furthermore, aPCCs are available which contain non-activated factors II, IX and X, and activated factor VII.
VKAs (e.g. warfarin) produce their anticoagulant effects by inhibiting the synthesis of vitamin K-dependent coagulation factors II, VII, IX and X. In patients with life-threatening bleeding, rapid replacement of these coagulation factors is required and therefore PCCs are a reasonable option. Both three-factor and four-factor PCCs have been investigated for the reversal of VKAs; four-factor PCCs are more commonly used, because three-factor PCCs do not provide adequate reductions in INR owing to the low levels of factor VII [
52]. The first four-factor PCC was approved in the United States in 2013, specifically for this purpose [
53]. Nevertheless, in VKA-associated bleeding, data for PCCs are based principally on laboratory rather than clinical endpoints, meaning that the evidence may be considered not to be at the highest level.
For the treatment of dabigatran-induced anticoagulation, neither PCCs nor aPCCs act as specific reversal agents for dabigatran or any other DOAC. Instead, they raise levels of the vitamin K-dependent coagulation factors, notably prothrombin, and thrombin generation is consequently increased. In the case of dabigatran, the plasma concentration of thrombin is increased to a stoichiometric excess vs the drug, and therefore levels of free (unbound) dabigatran and the antithrombotic effect of the drug are minimised. Data from pre-clinical and clinical studies of PCCs and aPCCs for reversal of dabigatran (see later sections) are consistent with this mechanism of action.
Although a number of suitable tests for the monitoring of plasma dabigatran concentration have been identified, it does not necessarily follow that these are the most appropriate tests for monitoring the reversal of dabigatran-induced anticoagulation by PCCs or aPCCs. For example, the aPTT may be useful in providing an approximate indication of dabigatran levels, but appears to be insensitive to the reversal effects of PCCs and aPCCs [
49]. There is some evidence to suggest that the EXTEM assay parameters clotting time (CT) and clot formation time (CFT) are sensitive to the effects of PCCs and aPCCs on dabigatran anticoagulation; therefore, these parameters may potentially provide a means of monitoring dabigatran reversal, although clinical studies are warranted [
49].
Safety of PCCs and aPCCs
The procoagulant/prothrombotic risks of treatment with PCCs and aPCCs must be weighed against the benefits. It should be noted that neither PCCs nor aPCCs are currently licensed for the treatment of DOAC-induced anticoagulation; therefore, any such use is off-label. The majority of evidence relating to the safety of PCCs has been obtained from VKA reversal or non-anticoagulated patients with perioperative bleeding. These are very different situations from the emergency treatment of DOAC-induced anticoagulation, and more evidence is needed regarding the safety of PCCs in this setting.
Historically, PCCs/aPCCs have been associated with a risk of thrombotic complications when used for the treatment of haemophilia or VKA reversal [
64]. Composition adjustments, such as the inclusion of coagulation inhibitors, reduced use of activated factors and improved balance of coagulation factor content, have been implemented with the aim of improving the safety of PCCs [
65]. Relative levels of factor II (prothrombin) and the key inhibitor antithrombin have been identified as the major cause of thrombogenicity [
66]. In an observational study in trauma patients, PCC was shown to increase the ETP for 3–4 days following treatment (i.e. approximately the half-life of prothrombin) [
67]. In addition, patients receiving PCC had low levels of antithrombin. It has been suggested that PCCs should be labelled according to prothrombin content, rather than factor IX content [
66]. Overall, the available safety data indicate that there are possible risks when using PCCs.
An in-vivo animal study assessed the safety of PCC for the treatment of dabigatran-induced anticoagulation [
56]. In the absence of dabigatran, high-dose PCC (300 IU/kg) produced low-grade pulmonary emboli in 5/5 (100 %) of animals. However, when the same dose of PCC was administered to animals previously treated with dabigatran, the frequency of pulmonary emboli was decreased in the presence of dabigatran in a dose-dependent manner (2/5 (40 %) at a dabigatran dose of 75 μg/kg, 1/5 (20 %) at 200 μg/kg and 0/5 (0 %) at 450 μg/kg) [
56].
In a porcine model of coagulopathy with blunt liver injury and no anticoagulation, administration of a four-factor PCC (50 IU/kg) resulted in protracted elevation of thrombin–antithrombin complexes and D-dimers, and formation of thromboemboli and pulmonary fibrinogen deposits in some animals [
68]. Signs of disseminated intravascular coagulation were also shown in 44 % of animals. In contrast, a PCC dose of 35 IU/kg safely improved coagulation parameters and halted blood loss. Furthermore, survival and total blood loss were significantly improved in both PCC groups when compared with control animals [
68].
In contrast to the previous study, the administration of PCCs or aPCCs to pigs after blunt liver injury under high-dose dabigatran appears not to be associated with thromboembolic events. Histopathological assessments showed that there was no thrombus formation in the heart, lungs, liver and kidneys after administration of PCC (25, 50 or 100 IU/kg) [
58] or aPCC (25 or 50 IU/kg) [
57].
Clinical perspective and conclusion
DOACs such as dabigatran have proven effective in decreasing the risk of ischaemic stroke in patients with AF, and in the prevention and long-term treatment of VTE. Although the bleeding risk associated with dabigatran is low, any anticoagulant can cause bleeding. With life-threatening bleeding in dabigatran-treated patients, urgent reversal of the thrombin inhibitory effects of dabigatran should be considered. The available data indicate that PCCs and aPCCs may be able to reverse dabigatran-induced anticoagulation in a dose-dependent manner. However, we do not have high-level evidence to support the use of PCCs/aPCCs in this setting, so the recommendation to use them is based on haematological principles, animal studies, healthy volunteer studies, an ex-vivo study of ‘real-world’ patients, and outcomes in a few case reports.
Treatment with PCCs or aPCCs increases the concentrations of several coagulation factors, including prothrombin which has a half-life of 60–72 hours [
51]. Thrombin generation may therefore be enhanced for several days after the use of PCCs to treat major bleeding [
67]. This may be associated with an increased risk of thromboembolic events. The lack of high-level evidence with PCCs and aPCCs for dabigatran reversal makes it difficult to make dose recommendations, but it appears necessary to use the minimum effective dose. On the other hand, there is evidence that low doses may not be effective in reversing the anticoagulant effects of dabigatran, probably because of the need to increase the plasma concentration of thrombin to that of dabigatran [
58]. This scenario is complicated by the lack of specific coagulation tests that are routinely available and rapid to perform, with established sensitivity and specificity for guiding the required dose of PCC or aPCC and then monitoring the effects of treatment. There is a clear need for a test that fulfils these criteria. Against this background, it is unsurprising that there are variations between guidelines for the use of PCCs/aPCCs in the management of bleeding among dabigatran-treated patients. Treatment decisions will be made on a case-by-case basis according to clinical judgement, local hospital protocols and product availability. Because of the theoretical thromboembolic risk, it is reasonable to titrate PCCs/aPCCs according to the clinical condition of the patient, starting with an initial dose of 25 IU/kg. However, clinical data are needed to establish the optimum dosing strategy. Because the mechanisms of action of PCCs and aPCCs are similar (both act by increasing thrombin generation), there is no need to use first one and then the other in a step-wise approach to bleeding management. It is essential to remember that the anticoagulant effect of dabigatran may be only one aspect contributing to coagulopathy; the likelihood of coexistent hyperfibrinolysis, dilutional coagulopathy and loss of coagulation factors, etc. [
69] require a multi-therapeutic approach. In the future, once the specific reversal agent idarucizumab becomes widely available, this treatment will be considered preferable to PCCs and aPCCs for dabigatran reversal because it has not been associated with a risk of thromboembolic events and has shown no procoagulant effect in various laboratory analyses. The phase III RE-VERSE AD study of idarucizumab showed complete reversal of the anticoagulant effect of dabigatran within minutes, in patients with serious bleeding or who required an urgent procedure [
19]. However, idarucizumab may not be available at every hospital for quite some time. Also, there could conceivably be clinical circumstances under which PCCs or aPCCs might be valuable as part of multimodal therapy, such as when thrombin generation is impaired as a result of trauma-induced coagulopathy [
70], although this needs to be evaluated in clinical studies.
Competing interests
OG has received research funding from Novo Nordisk, Biotest, CSL Behring and Nycomed; he has also received honoraria for consultancy and/or travel support from CSL Behring, Boehringer Ingelheim, Bayer Healthcare and Portola. JA has received honoraria for consultancy from Boehringer Ingelheim and Portola. RB serves as a consultant, researcher, and speaker for Boehringer Ingelheim and Medtronic, and has served in those roles in the past for Pfizer/Bristol Myers and Daiichi-Sankyo. PG has received honoraria for serving as a speaker and providing consultancy for Boehringer Ingelheim, Pfizer/Bristol Myers, Daiichi-Sankyo, Bayer and AstraZeneca. MVH has received research funding from GSK, Actelion and Boehringer Ingelheim, and is on a steering committee for Boehringer Ingelheim. DGJ has received honoraria for consultancy from Boehringer Ingelheim, Bayer and Merck. JHL is on Steering Committees for Boehringer Ingelheim, CSL Behring, Grifols, Janssen and The Medicines Company. CVP has received honoraria for consultancy from Boehringer Ingelheim, Daiichi-Sankyo, Janssen and BMS-Pfizer. ACS has received research funding from Jannsen, and has received honoraria for consultancy from Boehringer Ingelheim, Daiichi-Sankyo, Bayer, Jansen, Portola and BMS-Pfizer. TS has received a research grant from Octapharma, consultancy fees from Daiichi-Sankyo, and speaker honoraria and consultancy fees from Bayer, Boehringer Ingelheim and BMS-Pfizer. GJdZ serves as a consultant or on advisory boards for Boehringer Ingelheim, Daiichi-Sankyo, Novartis and Remedy; he has also received funding for fundamental research projects from the National Institutes of Health, Boehringer Ingelheim and Novartis. JE has received honoraria and research support from Bayer, Boehringer Ingelheim, BMS, Daiichi-Sankyo, GSK, Janssen and Pfizer.
Authors’ contributions
OG conceived of this article and wrote the manuscript. All authors contributed to drafting and editing the manuscript, and all authors approved the final version.