OUP user menu

Will direct thrombin inhibition change the boundaries of oral anti-coagulation?

O. Lomas, R.A. Argyle, B.D. Prendergast
DOI: http://dx.doi.org/10.1093/qjmed/hcq024 429-434 First published online: 15 March 2010


There are many indications for anti-coagulation, such as the prevention of stroke in atrial fibrillation (AF), treatment of venous thrombo-embolic disease and thrombo-prophylaxis for mechanical heart valves. Therapy may be required for many months, if not lifelong, and safe oral preparations are therefore desirable.1 Until recently, the choice of oral anti-coagulant treatment has been confined to vitamin K antagonists, such as warfarin, and these are prescribed for ∼1% of the UK population as a whole and 8% of those aged over 80 years.2

Warfarin acts as an anti-coagulant by inhibiting the vitamin K-dependent modification of proenzymes (Factors II, VII, IX and X) to active serine proteases of the coagulation cascade.3 Treatment with warfarin is, however, complicated by a narrow therapeutic range, numerous interactions with food and other drugs, and a highly variable dose-response relationship.1 The aim of this article is to describe how a novel class of anti-coagulant, the direct thrombin inhibitors, may change the management of patients with AF and following surgery.

Direct thrombin inhibitors

The limitations of existing anti-coagulants have stimulated the search for novel therapies that target specific steps in the coagulation cascade, particularly factors Xa and IIa (thrombin). Thrombin is the final enzyme of the clotting cascade and thus represents an excellent therapeutic target. Inhibition of thrombin prevents the formation of fibrin, blocks the feedback activation of factors V, VIII and XI, and attenuates the aggregation of platelets. Inhibition of thrombin may be indirect, using heparin (and, in effect, factor Xa inhibitors); or direct, using parenterally administered hirudin or novel oral agents, such as ximelagatran and dabigatran. Direct thrombin inhibitors are small, synthetic molecules that offer rapid, reversible, competitive inhibition of thrombin, few interactions with food or other drugs, no metabolism by the cytochrome P450 system and a predictable pharmacodynamic effect that does not require monitoring of anti-coagulant status.4–6 There are, however, some limitations to the current class of direct thrombin inhibitors, including a lack of safety data in pregnancy and a specific antidote. In the event of serious haemorrhage, possible measures include dialysis to eliminate the drug and administration of fresh frozen plasma.7 Despite these deficiencies, recently available data for the commercially available agent, dabigatran, herald a major transformation in the clinical approach to the prevention of thrombo-embolism.

AF: current strategies and the need for change

AF is characterized by uncoordinated atrial contraction, which generates local zones of static blood8 and a pro-thrombotic state9 that together encourage the formation of thrombus within the left atrium, with risk of systemic embolism, particularly to the cerebral circulation. Such a process accounts for ∼10% of all ischaemic strokes.10 The prevalence of AF is ∼1% in the general population, rising to >10% in those aged 85 years and over.11 With an increasingly elderly population, the prevalence of AF is likely to at least double over the next 50 years.12 The management of AF and its consequences therefore constitutes a major and increasing burden on healthcare resources.

Among patients with AF, anti-coagulation with warfarin to an international normalized ratio (INR) of 2.0–2.9 confers a relative reduction in the risk of stroke by 62% compared to placebo, equating to an absolute risk reduction of 2.7% per year for primary prevention and 8.4% per year for secondary prevention of stroke.13,14 The narrow therapeutic range of warfarin is determined by the increased risk of ischaemic stroke associated with an INR <2.015 and the increased risk of serious haemorrhage with an INR > 3.0.16 Retrospective analysis of over 2000 patients treated with warfarin for the prevention of stroke in non-valvular AF demonstrated an INR outside the target range for approximately one-third of the time.17 Strict control of INR is vital to minimize adverse outcomes in patients with AF who are treated with warfarin.18 This control requires regular blood tests and frequent changes of dose, which is onerous for the patient and the healthcare system. A more consistent method of anti-coagulation may therefore be expected to improve upon the already clear benefits of warfarin in this setting.

The direct costs of the management of AF accounted for ∼1% of the budget of the UK National Health Service in 2000,19 of which 20% was consumed by medications, including warfarin and its associated monitoring requirements. Complications of warfarin therapy, such as major haemorrhage,20 make a further hidden contribution to costs. Moreover, immobility, sensory impairment and cognitive impairment may make the use of warfarin unsafe and impractical, and many patients (and physicians) hold an exaggerated perception of the risks associated with treatment.21 Such problems are particularly common in the elderly, the very population at highest risk of ischaemic stroke.22 As a result, many patients do not benefit fully from effective protection against the thrombo-embolic consequences of AF.

When warfarin is considered unsuitable for a patient, oral anti-platelet agents represent an alternative therapeutic approach. Aspirin confers modest protection [∼20% relative risk reduction, (RRR)] against ischaemic stroke in patients with non-valvular AF, although this effect is inferior to that of warfarin (∼60% RRR).14 Clopidogrel is a purinoceptor antagonist that inhibits platelet function—in combination with aspirin, it reduces the risk of ischaemic stroke in comparison with aspirin alone (2.4 vs. 3.3% per year) but increases the risk of haemorrhage (2.0 vs. 1.3% per year).23,24 Even so, the combination is less effective than warfarin for the prevention of stroke.25 Therefore, an anti-coagulant that is more predictable in action without the need for monitoring would help more patients with AF to benefit from effective thrombo-prophylaxis.


The stroke prevention using an oral thrombin inhibitor and AF (SPORTIF) III and V trials, prospectively, measured the incidence of stroke in over 7000 patients with non-valvular AF who were randomized to either adjusted dose warfarin or fixed dose ximelagatran.26 Ximelagatran displayed non-inferiority to warfarin for the prevention of stroke and systemic embolism, with no significant difference in the rates of major bleeding between the two groups.27,28 Despite these promising results, elevated liver enzyme alanine aminotransferase levels that exceeded three times the upper limit of normal were found in 6% of ximelagatran recipients compared with 0.8% of patients who received warfarin.27 Associated hepatotoxicity, even after cessation of ximelagatran therapy, subsequently led to an enquiry by the United States Food and Drug Administration with subsequent withdrawal of the drug from the market.29


Despite these setbacks, development of other direct thrombin inhibitors, such as dabigatran, continued. Ximelagatran was developed from structural data of human thrombin bound to an irreversible inhibitor d-Phe–Pro–Arg-chloro-methyl-ketone (PPACK).30 However, dabigatran has been developed using a structure-activity approach based on X-ray crystallographic data of a thrombin inhibitor N-a-naphthylsulphonylglycyl-4-amidinophenylalanine piperidine (a-NAPAP), which was bound to bovine thrombin.6 The different origins of these compounds may account for their different pharmacological characteristics, including the apparently fewer side effects noted with dabigatran compared to ximelagatran.

The PETRO study,31 a phase II randomized comparison of three doses of dabigatran (with or without aspirin) compared to warfarin alone demonstrated that dabigatran 150 mg twice daily possessed similar efficacy and safety to adjusted dose warfarin. More recently published, the landmark RE-LY study32 was a randomized, prospective trial of over 18 000 patients which compared blinded dabigatran 110 or 150 mg twice daily against unblinded, adjusted-dose warfarin. Therapeutic range was achieved for 64% of the time for those on warfarin, consistent with previous experience.17 The primary outcome was non-inferiority with respect to stroke or systemic embolism. The population of the trial was recruited from many centres in Europe, North and South America, Asia and Australasia and mean age of the patients was 71 years (approximately two-thirds male). The mean CHADS2 score (a method of stroke risk stratification in the context of AF)33 was 2.1–2.2 (out of six) across subgroups of the trial, corresponding to an anticipated 4% annual risk of stroke.33 Mean follow-up of 2 years was achieved with almost no loss of outcome data.

For the primary endpoint of stroke or systemic embolism, lower dose dabigatran (110 mg twice daily) displayed non-inferiority to warfarin, but higher dose dabigatran (150 mg twice daily) proved superior [1.1 vs. 1.7% per year; number needed to treat (NNT) = 169]. Rates of major bleeding were equivalent with higher dose dabigatran and warfarin but reduced with lower dose dabigatran. Importantly, intracranial haemorrhage was less frequent with both doses of dabigatran. The composite outcome of major vascular events, major bleeding and death (used to evaluate net clinical benefit) did not differ significantly between warfarin and lower dose dabigatran. However, the higher dose dabigatran was associated with a small net benefit per year (6.91 vs. 7.64%; NNT = 137 per year compared to warfarin).

One concern despite these impressive results is the high rate of discontinuation of dabigatran compared to warfarin (20 vs. 17% at 2 years). Dyspepsia was the only significant adverse effect more common in the dabigatran group (11 vs. 6% of warfarin recipients). Dabigatran was also associated with a small increase in the incidence of myocardial infarction (0.72 and 0.74% per year with 110 and 150 mg, respectively, vs. 0.53% per year with warfarin), though the clinical significance of this difference remains unclear.

The striking results of the RE-LY trial raise the possibility of a re-appraisal of the boundaries of stroke prevention by making anti-coagulation more acceptable to the patient. However, further development of this class of drug may be necessary, perhaps to obtain once daily dosing and a more tolerable side-effect profile to improve compliance.

Venous thrombo-embolism: clinical evidence

Venous thrombo-embolism (VTE) is a frequent complication of general surgery and anti-coagulant prophylaxis has been recommended since the 1980s.34 Orthopaedic surgery such as hip and knee replacement is a particularly strong risk factor for VTE, with 10–30% of patients having evidence of a proximal deep vein thrombosis (DVT) 7–14 days after surgery.35 This risk remains elevated for at least 2 months following surgery and most symptomatic episodes therefore occur after hospital discharge, despite the widespread use of in-patient thrombo-prophylaxis.36

Two approaches to prevention are widely used: vitamin K antagonists (VKA), such as warfarin, and the heparins. Oral administration of VKAs makes them a popular choice in North America, though the requirement for post-discharge monitoring of INR has made them unpopular in Europe, where low-molecular weight heparin (LWMH) is preferred. LWMH does appear to have greater efficacy than VKAs but is limited by the need for parenteral administration.35 Out-patient prophylaxis is therefore inconvenient, especially since 4 weeks of post-operative treatment are recommended following hip surgery in the UK.37 Fondaparinux, a factor Xa inhibitor, has efficacy against VTE that is similar to LMWH38 but also requires parenteral administration. Direct thrombin inhibitors may therefore represent an alternative approach to anti-coagulation that provides safe and convenient thrombo-prophylaxis following orthopaedic and other major surgery.

Prior to market withdrawal, the efficacy of ximelagatran was evaluated in this setting using large, randomized studies after total hip and knee replacement and found to be non-inferior to LMWH for the prevention of VTE.39–41 This potential has since been explored in three recent studies using dabigatran: RE-NOVATE,42 RE-MODEL43 and RE-MOBILIZE.44

The RE-NOVATE trial42 was a randomized, double-blind, non-inferiority study which compared the efficacy of dabigatran (220 or 150 mg once-daily) vs. the LMWH, enoxaparin (40 mg s/c), in a cohort of ∼3500 patients undergoing elective hip replacement. The mean duration of treatment was 33 days and primary outcome was a composite of total VTE events and all-cause mortality. The trial demonstrated non-inferiority of dabigatran at either dose compared to once daily enoxaparin. Both doses of dabigatran possessed similar adverse event profiles to enoxaparin and there was no significant difference in the rate of bleeding. The RE-MODEL trial43 was a randomized, double-blind, non-inferiority study which compared the efficacy of dabigatran (220 or 150 mg daily) and enoxaparin (40 mg s/c) used for 6–10 days in ∼2000 patients undergoing elective knee replacement. The risks of VTE, death and bleeding were not significantly different in each of the three treatment arms.

The RE-MOBILIZE44 trial was a randomized, double-blind, non-inferiority study which compared dabigatran (220 or 150 mg once daily) against lower dose enoxaparin (30 mg s/c twice daily, as opposed to 40 mg once daily in RE-MODEL and RE-NOVATE) in the setting of elective knee replacement. In this study, enoxaparin proved superior to dabigatran for the composite outcome of VTE and death. However, none of the secondary outcomes (major VTE and related mortality or adverse haemorrhage) differed significantly between the three treatment groups.

Meta-analysis of these three trials45 has demonstrated equivalent risks of VTE and all-cause mortality using either dose of dabigatran or enoxaparin. However, economic analysis has strongly suggested that dabigatran is more cost-effective, principally due to the lower costs of drug administration.46 These findings and the greater convenience of oral administration have led to the approval of dabigatran for thrombo-prophylaxis following hip and knee surgery in Europe and Canada.

The efficacy of dabigatran in the treatment of acute venous thrombo-embolic disease has recently been evaluated by the randomized, double-blind trial of 150 mg dabigatran twice-daily compared to dose-adjusted warfarin in the RE-COVER trial.47 Six months of either oral anti-coagulant therapy revealed no difference in the primary end-point of recurrent VTE. The incidences of death, acute coronary syndrome and deranged liver function tests were similar between both groups. The rate of discontinuation due to dyspepsia was greater in the dabigatran group compared to warfarin, which is consistent with observations made in the RELY trial.32 Dabigatran therefore appears to represent an effective alternative to warfarin therapy in the setting of acute venous thrombo-embolic disease. However it remains to be seen whether dabigatran extends the use of short-term anti-coagulation in the setting of venous thrombo-embolic disease, as compared to the lifelong need for anti-coagulation in AF, which is often complicated by long-term concerns about compliance and tolerability of the anti-coagulant regime.

Valve disease and prosthetic valves

Patients with valve disease associated with AF or those who have undergone valve surgery using a metallic prosthesis are at high risk of thrombo-embolic complications and lifelong anti-coagulation is usually recommended.48 Dabigatran may have a future role in these groups although no specific trials are complete or anticipated, reflecting the high risks associated with any treatment found to be inferior to current regimes.


The renewed availability of direct thrombin inhibitors presents an opportunity to improve outcomes for patients and reduce healthcare costs. A more consistent and reliable anti-coagulant such as dabigatran may surpass the already impressive ∼60% reduction in relative risk of stroke associated with warfarin, not only through direct improvements in pharmacological efficacy, but also by providing a more acceptable and safe method of anti-coagulation that allows more candidates for therapy to receive effective thrombo-prophylaxis. Moreover, dabigatran appears to be cost-effective in the context of thrombo-prophylaxis following orthopaedic surgery and allows improved compliance with treatment after discharge from hospital. With a wide range of compounds currently being explored, the field of anti-coagulation appears to be rapidly expanding. The days of cumbersome and unpredictable warfarin therapy may well be drawing to a close.


View Abstract