NPC Archive Item: Dabigatran▼ vs. warfarin for prevention of stroke in atrial fibrillation: the RE-LY study

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
On The Horizon Rapid Review NPC Logo

27 October 2009,

The RE-LY study found that, in patients with atrial fibrillation with at least one other risk factor for stroke, dabigatran was non-inferior to adjustable dose warfarin in preventing strokes, and reduced the incidence of major bleeding. Further studies are required to establish long-term safety.

Level of evidence:
Level 2 (limited quality, patient-orientated evidence) according to the SORT criteria

Should dabigatran receive a license for use in AF it may offer an effective alternative for those patients who have difficulty maintaining anticoagulation control or who cannot take warfarin for other reasons. The advantages demonstrated in this study for efficacy (high dose only) and for reduction of bleeding (low dose only) are small in absolute terms. Questions remain regarding dabigatran’s long-term safety and how it would compare in patients who are well controlled on warfarin. It is likely to have a considerably higher drug cost than warfarin.

A NICE appraisal of dabigatran in AF is expected to begin during early March 2010, in order to be available at or around the expected launch of dabigatran later in the year, should its licence be approved. Cost-effectiveness relative to warfarin should be an important consideration. Local decision making bodies on medicines are advised to engage with stakeholders and agree a protocol for its use if/when dabigatran is licensed for this indication. This includes identifying those patients for whom the drug may be appropriate and planning for possible NICE guidance.

What is the background to this?
NICE recommends thromboprophylaxis for people with AF with either warfarin or aspirin depending on their level of risk, and after consideration of individual patient factors. Dose-adjusted warfarin is the most effective and preferred option for people who are at high risk of stroke. However, it is associated with a higher bleeding risk than aspirin, and is not appropriate for those at low risk of stroke.

For more information on the background and treatment options for AF, including the evidence supporting the use of warfarin and aspirin, please refer to the Cardiovascular section of NPC.

Dabigatran (more correctly called dabigatran etexilate – a prodrug of dabigatran) is an orally active antithrombotic agent. It is a direct thrombin inhibitor, which has the potential advantage over warfarin of not requiring blood monitoring, and may have fewer clinically important drug interactions. It is currently licensed for short-term use (maximum 35 days) in the primary prevention of venous thromboembolism in adults undergoing elective total hip or knee surgery, and is recommended by NICE as one of the options to consider in this indication.

Dabigatran is also being considered for the prevention of strokes in people with AF. The RE-LY study was a Phase III clinical study. It had a prospective, randomized, open-label, blinded endpoint (PROBE) design. It evaluated the non-inferiority of two doses of dabigatran compared with warfarin in people with AF who were at moderate to high risk of stroke. The primary efficacy endpoint of the trial was incidence of stroke (including haemorrhagic) and systemic embolism. The primary safety endpoint was major bleeding.

What does this study claim?
Over a median two-year follow-up, the mean rates of the primary end point were 1.53%/year for low-dose dabigatran (110mg twice daily), 1.11% per year for the high-dose dabigatran (150mg twice daily) and 1.69% per year for warfarin. The study found that the lower dose of dabigatran was non-inferior to warfarin at reducing the risk of stroke and systemic embolism in people with AF (relative risk [RR] 0.91; 95% confidence interval [CI] 0.74 to 1.11; P<0.001 for non-inferiority). The higher dose was found to be statistically significantly more effective than warfarin (RR 0.66; 95%CI 0.53 to 0.82; P<0.001; number needed to treat [NNT] 172 over one year).

The mean rates for major bleeding were 2.71% per year for low dose dabigatran, 3.11% per year for high dose dabigatran and 3.36%/year for warfarin. Whereas low-dose dabigatran was associated with a reduced risk of major bleeding (P=0.003; NNT 154 over one year), there were no significant differences between the high-dose dabigatran and warfarin in this respect.

How does this relate to other studies?
To reduce the risk of stroke in people with AF, the NICE guideline on the management of AF recommends thromboprophylaxis with either adjusted-dose warfarin (target international normalised ratio [INR] 2.5, range 2.0 to 3.0) or aspirin depending on the level of stroke risk. A stroke risk stratification algorithm is provided in the NICE guideline for this purpose. For those at high risk, warfarin is recommended, although aspirin can be considered where warfarin is contraindicated. For those at moderate risk of stroke either warfarin or aspirin can be considered, and for people at low risk of stroke (i.e. not at moderate or high risk) aspirin is recommended. However, owing to lack of sufficient clear-cut evidence, treatment may be decided on an individual basis, and the physician must balance the risks and benefits of warfarin versus aspirin. The potential bleeding risks of long-term anticoagulation are an important consideration.

Clopidogrel has also been considered for stroke prevention in people with AF. However it is not licensed, or recommended for use, alone or with aspirin for this indication. Studies to date in people with AF, have not demonstrated any significant benefits over existing treatments. The ACTIVE W demonstrated clear superiority of warfarin over clopidogrel in preventing vascular events. In the ACTIVE A study (see Blog) clopidogrel plus aspirin reduced the risk of a vascular composite end point compared with aspirin alone in people with AF who were at increased risk of stroke, and for whom warfarin-like drugs were unsuitable. However, this benefit has to be balanced against a significant increased risk of major bleeding where the NNT to cause a major bleed was similar to that to prevent a major vascular event – around 42 over 3.6 years.

So what?
The RE-LY study aimed to demonstrate non-inferiority for dabigatran compared with adjustable dosed warfarin. The design allowed for superiority to be tested once non-inferiority was identified, which it was for the primary endpoint.

Dabigatran was shown to be at least as effective as warfarin in preventing strokes, particularly haemorrhagic strokes, in people with AF who are at moderate or high risk of strokes. This finding, taken together with no greater risk of major bleeding, suggests a possible role as an alternative to warfarin in such patients.

However, there are a number of issues raised by the study which need to be considered when putting the results in the context of normal clinical practice. These relate to the level of anticoagulant control in the warfarin group, the long-term safety and tolerability of dabigatran, and limitations in the study methodology.

In the study, the INR was within the therapeutic range for 64% of the time. Although, this seems low, this is similar to other contemporary trials of warfarin and, in this trial, may reflect the high proportion of people in the study who had not received warfarin previously. Nevertheless, some patients will have been more controlled than others, and the study does not address the issue of whether dabigatran would be as effective as warfarin in those people who were well controlled on warfarin.

More patients discontinued treatment with dabigatran than warfarin during the study, which might be due to poorer tolerability. A higher incidence of discontinuations that were a result of serious side effects supports this view. However, as patients and physicians knew which treatment (dabigatran or warfarin) were being received this may have raised their perception of possible side effects from the newer drug, and decisions to discontinue may have been taken more readily.

Although, major bleeding was no more frequent between groups overall, the higher risk of GI side effects (both doses) and GI bleeding with dabigatran at the 150mg dose compared with warfarin raises questions about its use in people who are at high risk of these effects.

The study only considered dabigatran treatment for a median period of two years, and thus long-term safety is unclear. Serious hepatic side-effects were the reason for the withdrawal of the license for ximelegatran (another thrombin inhibitor) for a similar indication. Although there were no indications of a difference between treatments with regard to hepatic side effects in the present study, patients with a creatinine clearance of less than 30ml/min were excluded. It is not known whether monitoring of liver enzymes will be required with long-term use of dabigatran.

The significantly greater rate of MIs with high-dose dabigatran serves as a signal of potential long-term safety which will need to be considered. The absolute differences in this study were small; results suggest that 476 patients, like those in this study, would need to be treated with dabigatran for one year for one of them to have a myocardial infarction who would not have done if they had received warfarin. Nevertheless, this raises particular concerns about the use of dabigatran in people who are at high risk of coronary heart disease.

The studies main limitation, in terms of design, is its open-label nature. Both patients in the study and the physicians allocating treatment were aware of the treatment allocated (although allocation of dabigatran doses was blinded); this has potential to introduce considerable bias into the study. We have already mentioned how this might affect discontinuation. The independent blinded assessment of outcomes mitigates this effect to some degree, at least for the objective outcomes.

The results of this study are not directly applicable to those patient groups who were excluded from the study (e.g. those with recent strokes). However, in general, the patients included, who were at moderate to high risk of stroke, are the types of patients for which warfarin can be considered according to the NICE guideline.

All statistical analyses were conducted according to intention-to-treat principles. A per-protocol analysis, which includes patients who dropped out of the study, may have provided support to the non-inferiority status (as per EMEA recommendations).

In conclusion, recognising the limitation above, what does this study tell us about dabigatran’s place in therapy for the management of atrial fibrillation?

Any advantages in efficacy and safety were small in absolute terms. However, dabigatran has the considerable advantage that, unlike warfarin, it does not require regular anticoagulant monitoring, and there would be reductions in non-drug costs associated with provision of this service to people prescribed dabigatran instead of warfarin. Dabigatran may, therefore, be an appropriate option for those patients who cannot take warfarin, or undergo the monitoring required, or where control of anticoagulant status is poor, despite best efforts. However, may of the costs associated with current warfarin services will be fixed due to the need to maintain the existing infrastructure for patients well-established on warfarin. Therefore it seems unlikely that there will be real cost savings associated with the development of alternatives to warfarin, should the evidence for their wider use become more established.

Offset against the advantages for dabigatran, are questions about its long-term safety, and cost could be an issue. The drug’s acquisition cost is likely to be considerably higher than warfarin. According to the BNF, the current costs for dabigatran 110mg is £2.10 per capsule; this equates to about £1500 for a year’s course when given twice daily. The cost of warfarin is 100 times less; a year’s supply of warfarin 5mg costs about £16. However, the cost per daily dose of dabigatran for prevention of stroke may be less that its current price. NICE is due to review the use of dabigatran for AF in 2010, and cost-effectiveness should be an important consideration.

The increased drug cost may be offset by avoiding additional non-drug costs associated with anticoagulant monitoring.

Although dabigatran interacts with a number of medicines (see SPC), it has fewer clinically important food and drug interactions than warfarin. However, unlike warfarin its antithrombotic effect is not rapidly reversible, and there is no antidote. This may be of particular concern in patients who are at high risk of bleeding.

Study details

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

Multicentre, international, clinical study with PROBE design

The study included 18,113 patients with AF and at least one other risk factor for stroke. Exclusion criteria included the presence of severe heart-valve disorder, stroke with 14 days or severe stroke with six months before screening, a creatinine clearance of less than 30ml/min, or active liver disease. The mean age was 71 and 64% were men. At baseline, the mean CHAD score was 2.1; 20% had had a previous stroke or TIA, 16–17% a previous MI, and 32% had heart failure. Half were receiving vitamin K antagonists and 39–41% were receiving aspirin.

Intervention and comparison
Patients were randomised to one of three groups. Dabigatran 110mg or 150mg (blinded) twice daily or warfarin (unblinded) adjusted to an INR of 2.0 to 3.0. The INR was within the therapeutic range for 64% of the study period. Aspirin was used continuously during the treatment period in 20–21% of patients in all three groups.

Outcomes and results
After a median of 2 years, complete follow-up was achieved in 99.9% of patients. The rates of discontinuation at two years were 17% for warfarin and 21% for both dabigatran groups (P<0.001 for both dabigatran doses vs. warfarin).

The primary outcome of stroke or systemic embolism occurred at a rate of 1.53% 1.11% per year, and 1.69% per year, for dabigatran 110mg, dabigatran 150mg, and warfarin, respectively. The 110mg dabigatran dose was non-inferior to warfarin but not superior (RR 0.91; 95%CI 0.74 to 1.11; P<0.001 for non-inferiority, P=0.34 for superiority). The 150mg dose was superior to warfarin in relation to the primary outcome (RR 0.66; 95%CI 0.53 to 0.82; P<0.001; NNT172). The primary outcome events were mainly strokes — pulmonary emboli occurred at rates of 0.12% per year, 0.15% per year, and 0.09% per year for dabigatran 110mg, dabigatran 150mg, and warfarin, respectively (no significant difference between groups).

Rates of haemorrhagic stroke were 0.12% per year, 0.10% per year and 0.38% per year, for dabigatran 110mg, dabigatran 150mg and warfarin, respectively. The differences were statistically significant from warfarin for both doses of dabigatran 110mg (RR 0.31, 95%CI 0.17 to 0.56; P<0.001; NNT 384) and 150mg (RR 0.26; 95%CI 0.14 to 0.49; P<0.001; NNT 357).

The rate of MI was higher but not significantly different in the dabigatran 110mg group (0.72% per year) than in the warfarin group (0.53% per year) (RR 1.35; 95%CI 0.98 to 1.87; P=0.07) but significantly higher in the dabigatran 150mg group (0.74% per year) (RR 1.38, 95% CI, 1.00 to 1.91; P=0.048*; number needed to harm (NNH) 476).

Rates of death from any cause were 4.13% per year with warfarin, as compared with 3.75% per year with dabigatran 110mg (RR 0.91; 95%CI 0.80 to 1.03; P=0.13) and 3.64% per year dabigatran 150mg (RR 0.88; 95%CI, 0.77 to 1.00; P=0.051). Death from vascular causes was significantly lower with dabigatran 150mg compared with warfarin (2.28% vs. 2.69% per year; RR 0.85; 95%CI 0.72 to 0.99; P=0.04).

The rate of major bleeding was significantly lower in the dabigatran 110 mg group (2.71% per year) compared with the warfarin group (3.36%) (RR 0.80; 95%CI 0.69 to 0.93; P=0.003; NNT 154). There was no significant difference in major bleeding between the dabigatran 150mg group (3.11% per year) and the warfarin group (RR 0.93; 95%CI, 0.81 to 1.07; P=0.31).

Rates of life-threatening bleeding, intracranial bleeding, and major or minor bleeding were significantly higher with warfarin (1.80% per year, 0.74% per year, and 18.15% per year, respectively) than with either the 110mg dose of Dabigatran (1.22% per year, 0.23% per year, and 14.62% per year, respectively) or the 150mg dose of Dabigatran (1.45% per year, 0.30% per year, and 16.42% per year, respectively); RR and CIs for comparisons are available in the full article; all P-values were less than 0.05. Rates of major GI bleeding were higher with dabigatran 150mg (1.51% per year) compared with warfarin (1.02% per year, RR 1.50, 95%CI 1.19 to 1.89; P<0.001, NNH 204), but not dabigatran 110mg (1.12% per year). There were no significant differences between groups with regard to non-threatening major bleeding or extracranial bleeding.

Serious adverse events leading to discontinuation occurred more frequently with both doses of dabigatran (both 2.7%) than with warfarin (1.7%; P<0.001; NNH 100) Gastrointestinal symptoms (which included pain, vomiting and diarrhoea) were given as the reason for discontinuation more often with dabigatran 110mg (2.2%) and dabigatran 150mg (2.1%) than with warfarin (0.6%).

Dyspepsia occurred in 5.8% of patients in the warfarin group, 11.8% in the 110mg dabigatran  group and 11.3% in the 150mg dabigatran group (P<0.001 for both dabigatran doses vs. warfarin, NNHs 17 and 19, respectively).

The study was funded by Boehringer Ingelheim

More information on atrial fibrillation can be found on the cardiovascular section of NPC

Please comment on this blog in the NPC discussion rooms, or using our feedback form.

Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you.

*Amendment 14th December 2011.

The authors of the RE-LY study updated their article after identifying several additional primary efficacy and safety outcomes. This included an additional 4 cases of clinical myocardial infarction (MI) and 28 cases of silent MI. Silent MI is included in the RE-LY definition for cases of myocardial infarction.

When these cases were taken into account the rate of MI was numerically higher, but not statistically significantly higher, in the dabigatran 150 mg group (0.81% per year) than in the warfarin group (0.64% per year) (RR 1.27, 95% CI, 0.94-1.71; P=0.12).