REVIEWS Oral anticoagulants for Asian patients with atrial fibrillation Ian Sabir, Kaivan Khavandi, Jack Brownrigg and A. John Camm Abstract | Anticoagulation is the most-important intervention to prevent stroke in patients with atrial fibrillation (AF). Despite a lower point prevalence of AF in Asian communities and Asian countries than in other populations, individuals of Asian ethnicity are at a disproportionately high risk of stroke and have greater consequent mortality. Warfarin and other vitamin K antagonists are conventionally used for anticoagulation, and demonstrably reduce the risk of stroke and all-cause mortality in patients with AF. The use of warfarin in Asian countries is suboptimal, primarily owing to the universal challenge of achieving controlled anticoagulation with an unpredictable drug as well as concerns about the particularly high-risk of haemorrhage in Asian patients. Instead, antiplatelet therapy has been favoured in Asian communities, this strategy is neither safe nor effective for stroke prevention in these individuals. The non-vitamin K antagonist, oral anticoagulant drugs offer a solution to this challenge. The direct thrombin inhibitor dabigatran, and the direct factor Xa inhibitors apixaban, edoxaban, and rivaroxaban, have demonstrated noninferiority to warfarin in the prevention of stroke and systemic embolism in international, randomized, controlled trials. Importantly, some of these drugs are also associated with a significantly lower incidence of major haemorrhage, and all result in lower rates of intracranial haemorrhage and haemorrhagic stroke than warfarin. In this article, we review the use of the non‑vitamin K antagonist anticoagulants in the management of AF in Asian populations. Sabir, I. et al. Nat. Rev. Cardiol. 11, 290–303 (2014); published online 11 March 2014; corrected online 2 September 2014; doi:10.1038/nrcardio.2014.22
Introduction
Rayne Institute, St Thomas’ Hospital, London SE1 7EH, UK (I.S., K.K.). St George’s Vascular Institute, Division of Cardiovascular Sciences (J.B.), Cardiovascular Sciences Research Centre (A.J.C.), St George’s University of London, Cranmer Terrace, London SW17 0RE, UK Correspondence to: I.S. [email protected]
Atrial fibrillation (AF) is the most common cardiac arrhythmia, affecting approximately 1% of the world’s population.1 AF is a supraventricular tachyarrhythmia characterized by unco-ordinated atrial activity and rapid ventricular responses, leading to the absence of defined P waves and irregularly irregular R–R intervals on the electrocardiogram. Functionally, AF leads to reduced cardiac output as a result of mechanical dissociation of atrial and ventricular contraction, and is an independent risk factor for multiple adverse outcomes, including transient ischaemic attack (TIA), stroke, dementia, heart failure, and cardiovascular and all-cause death.2,3–5 The mechanisms underlying the pathogenesis of AF remain under active investigation. Aberrant automaticity originating in the pulmonary veins is hypothesized to trigger the initiation, perpetuation, and maintenance of AF in myocardial tissue rendered susceptible to this arrhythmia by genetic and environmental factors. 6 Hypertension, cardiac failure, ischaemic and valvular Competing interests A.J.C. declares associations with the following companies: Actelion Pharmaceuticals, ARYx Therapeutics, Bristol-Myers Squibb Company, Cardiome Pharma Corporation, CV Therapeutics, Daiichi Sankyo, Johnson & Johnson Pharmaceutical Research & Development, Menarini Group, Merck & Co., Novartis Pharmaceuticals Corporation, Pfizer, Sanofi, SERVIER, and Xention Limited. See the article online for full details of the relationships. The other authors declare no competing interests.
290 | MAY 2014 | VOLUME 11
heart disease, diabetes mellitus, and advanced age are all well-established risk factors for the development of AF.7,8 Novel risk factors for AF, such as sleep apnoea, obesity, and metabolic syndrome, have also been identified.9,10 A heritable component of AF pathogenesis has been postulated; parental AF is an independent risk factor for the development of AF in offspring,11 and several candidate genes contributing to AF susceptibility have been identified in genome-wide association studies.12 The consequences of AF are serious; fibrillation of the atria can cause turbulent blood flow and stasis, increasing the risk of thrombus formation and systemic embolization, particularly from the left atrial appendage. AF raises the risk of ischaemic stroke almost fivefold, a risk magnified in individuals aged between 80 and 89 years, among whom AF has been found to be the major cardiovascular morbidity exerting an independent effect on the incidence of stroke.2 AF is estimated to account for around 15% of all strokes, rising to 25% in the elderly (age ≥70 years),13 and also independently raises the risk of congestive heart failure and all-cause mortality.4,14 Conventionally, warfarin and other vitamin K antagonists (VKAs) have been used to reduce the risk of thromboembolism in patients with AF. Effective treatment with warfarin relies on careful monitoring of patients to maintain anticoagulation within a target international normalized ratio (INR) range. However, achieving appropriate anticoagulation is complicated by pharmacogenomics,
www.nature.com/nrcardio © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Key points ■■ Although few prospective studies of atrial fibrillation (AF) epidemiology in Asian countries have been conducted, AF prevalence is thought to be lower in Asians than in white populations ■■ However, Asian individuals seem to be at a higher risk of stroke, and have greater stroke-related morbidity and mortality, than white individuals ■■ Asian individuals seem to be at particularly high risk of haemorrhagic stroke and intracranial haemorrhage, which has discouraged the widespread uptake of warfarin in Asian patients with AF ■■ Antiplatelet agents have been overused for primary stroke prevention in Asian patients with AF, in whom these drugs are neither safe nor effective ■■ Non-vitamin K antagonist anticoagulants are noninferior to warfarin for stroke prevention in AF, and are associated with a reduced incidence of haemorrhagic stroke and intracranial haemorrhage compared with warfarin ■■ The efficacy and safety profile of non-vitamin K antagonist anticoagulants in randomized, controlled trials suggest these drugs have great potential as anticoagulants in Asian individuals with AF
Table 1 | Prevalence of AF in Asia by country 1 Country
Community-based studies
Hospital-based studies
Prevalence of AF (%)
Number of studies
Prevalence of AF (%)
Number of studies
China
0.8–2.8
4
NA
NA
India
0.1
1
NA
NA
Japan
0.4–0.6 (women); 1.0–1.7 (men); 0.6–1.6 (overall)
10
12.2–14.0
2
Malaysia
NA
NA
2.8
1
Singapore
1.4
1
NA
NA
South Korea
0.7
1
NA
NA
Taiwan
1.1
1
NA
NA
Thailand
0.4–2.2
2
NA
NA
Abbreviations: AF, atrial fibrillation; NA, not available.
pharmacokinetics, drug–drug interactions, and patient adherence to drug regimens. In the Asian population, the increased prevalence of haemorrhagic stroke has led to under use of warfarin as an anticoagulant, in favour of antiplatelet agents. The non-VKA anticoagulants provide new therapeutic options for the management of thromboembolic risk related to AF in Asian populations. In this Review, the evidence supporting the use of non-VKA anticoagulants and their bleeding risk profiles in Asian patients is discussed, and we compare non-VKA anticoagulants to current management options for these individuals. We set these discussions in the context of the rising burden of stroke in Asian populations.
AF and stroke epidemiology in Asians
The global epidemiology of AF is an area of active research, particularly as the burden of AF and associated health-care costs increase.15 The white population of the western world has been comprehensively and prospectively studied. Pooled data suggest that the prevalence of AF in the general population of white individuals is between 0.5% and 2.0%, with men at higher risk than women, although the prevalence increases with age in both sexes.16–19 In one European population (Rotterdam,
Netherlands) of individuals aged ≥55 years, the lifetime risk of developing AF was 23.8% in men and 22.2% in women,17 estimates that are supported by data from the Framingham heart study.17,20 The epidemiology of AF within Asian populations has become increasingly clear. However, analysis of disease epidemiology by race, ethnicity, or geography is methodologically challenging, as study populations and definitions of ethnicity and race are heterogeneous.1 In addition, caution should be exercised in the interpretation of epidemiological data in migrant populations, as the relative contributions of environmental risk from their original countries and the countries to which they migrate might differ. Nevertheless, research has consistently demonstrated that the prevalence of AF is lower in Asian populations than in white populations.1 The Global Anticoagulant Registry in the Field (GARFIELD) study,21 in which 2,940 patients (27.7% of the total cohort) originated from the Asia–Pacific region, will greatly improve our understanding of this issue. A summary of AF prevalence in Asian countries is provided in Table 1. The point prevalence of AF has been reported to be 1.2% in the white population of South and East London compared with 0.22% in the South Asian community of the same area.22 In a small study conducted in the UK, the prevalence of AF in Indo-Asian patients aged >50 years was estimated to be 0.6%.23 These data are supported by studies performed in China and Japan, in which the age-standardized prevalence of AF was 0.65% and 0.56%, respectively.24,25 Moreover, prospective data collected using implanted cardiac devices demonstrated that new-onset AF lasting >6 h was less prevalent in Chinese and Japanese patients than in Europeans.26 Reduced prevalence of AF in Asian populations does not, however, translate into reduced stroke risk. Broadly, stroke is subclassified into ischaemic (that is, atherothrombotic, cardioembolic, and lacunar) and haemorrhagic aetiology. Asian populations have consistently been shown to be at a higher overall risk of stroke than white populations.27–29 The REACH registry 30,31 showed that, although Eastern Europe was the region in which patients were at highest risk of composite cardiovascular and cerebrovascular outcomes, individuals in Japan had the highest risk of haemorrhagic stroke. The relative burden of stroke compared with ischaemic heart disease is significantly greater in Asian countries, particularly China, compared with the West, as is the cost in disability-adjusted life years.28 This issue was highlighted in a systemic review in which the age-standardized incidence of a first stroke was higher among Chinese individuals (residing in China) than white individuals in European countries.27 South Asian individuals tend to present with acute first stroke at a younger age than East Asian and white individuals.32,33 The increased risk of stroke in South Asians is associated with greater mortality.29 A systemic review showed that the standardized mortality ratio for ischaemic stroke in South Asians was 55% and 41% greater in men and women, respectively, compared with the white population, a difference
NATURE REVIEWS | CARDIOLOGY
VOLUME 11 | MAY 2014 | 291 © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Contact activation (intrinsic) pathway
XII
Tissue factor (extrinsic) pathway
XIIa XI
XIa IX
IXa
Apixaban Betrixaban Edoxaban Rivaroxaban
Tissue factor VIIa
Warfarin (indirect via VKORC1)
VII
VIIIa Tissue factor Warfarin (indirect via VKORC1) Warfarin (indirect via VKORC1)
Xa
X
X
Va Prothrombin (II)
V Fibrinogen (I)
Thrombin (IIa)
Dabigatran
Fibrin (Ia) XIIIa Cross-linked fibrin clot
Figure 1 | Interaction between anticoagulant drugs and the coagulation cascade. Warfarin indirectly inhibits the synthesis of factors II, VII, IX, and X via inhibition of VKORC1. The direct thrombin (dabigatran) and direct factor Xa (apixaban, betrixaban, edoxaban, and rivaroxaban) inhibitors discussed in this article are also shown. Abbreviation: VKORC1, vitamin K epoxide reductase complex subunit 1.
consistent across Indian, Pakistani, and Bangladeshi ethnicities.29 Moreover, the prevalence of stroke in the South Asian community has increased by an estimated threefold to fivefold over the past 40 years, a phenomenon thought to be a consequence of the increasing prevalence of hypertension and hyperlipidaemia in this population.29 The CHADS2 and CHA2DS2-VASc scores, originally derived from white populations as tools for the stratification of stroke risk in patients with AF, have also been shown to be predictive of future stroke in some Asian populations.34–36 Interestingly, increased CHADS2 score was predictive of future stroke even in the absence of AF in a Japanese cohort.34 Although AF raises the relative risk (RR) of stroke fivefold in white populations,2 this risk seem to be more modest in Asians, ranging from a 2.8-fold to a 3.7-fold increase depending on the population studied.24,37–39 Nevertheless, CHADS2 or CHA2DS2VASc scores comparable with or significantly higher than those in white populations have been found to identify individuals at high risk of stroke in Asian populations22,26 despite the lower prevalence of AF. Interestingly, a study conducted in Birmingham, UK showed that a significantly lower percentage Indo-Asian patients with AF who were resident in this area were aware of the risk of stroke associated with AF (26% versus 70%; P 50%
Hepatic: predominantly CYP3A4 Substrate of P‑gp
30–120 min
60–120 min
8–15 h
~27%
Strong inhibitors and inducers of CYP3A4 and P‑gp
Edoxaban
Factor Xa
Once daily
45–61%
Hepatic: predominantly CYP3A4
30–60 min
60–120 min
9–11 h
35%
Strong P‑gp inhibitors
Betrixaban
Factor Xa
Once daily
34%
Minimal hepatic metabolism Substrate of P‑gp
180–240 min
180–240 min
20 h
17%
Strong P‑gp inhibitors
*Half-life is extended to 11–13 h in elderly patients (age ≥70 years). Abbreviations: ABCG2, ATP-binding cassette sub-family G member 2 (also known as breast cancer resistance protein); CYP, cytochrome P450; P‑gp; P‑glycoprotein; VKORC1, vitamin K epoxide reductase complex subunit 1.
event compared with 3.24% in Western Europe and 2.69% in North America.64 Failure to prescribe warfarin to individuals at high risk of thromboembolic events is of particular concern in Asian populations. Evidence suggests that war farin therapy is prescribed half as often in Asia as it is in Europe, irrespective of stroke risk as determined by CHA2DS2-VASc score.65 The main reason for under prescription is the perceived increased bleeding risk with warfarin; studies in the Japanese population have shown a higher incidence of haemorrhage than in white individuals, even among groups of patients with a low target INR.47 Instead, aspirin is commonly prescribed as an alternative to warfarin among Asians, despite evidence that low-dose aspirin is neither safe nor beneficial in Japanese patients with lone AF for stroke prevention.66 Antiplatelets do, however, have a role in secondary stroke prevention in Asian populations.67,68 Poor patient compliance might also be an issue; in a UK study, Indo-Asians were the ethnic group most likely to describe themselves as ‘careless’ when taking warfarin.40 Maximizing the percentage of time in the therapeutic range (TTR; that is, INR 2–3 for nonvalvular AF) is vital for war farin users to maximize the therapeutic effect for stroke prevention and minimize the associated risk of major haemorrhage.69 However, in a systematic review of 36 studies, the mean TTR varied between 59% and 67%, depending on the trial design.69 The mean time above the target INR range was 14%, and the mean time below the target INR range varied between 20% and 25%.69 In one US‑based study, an estimated one-fifth of all 294 | MAY 2014 | VOLUME 11
thromboembolic events occurred during a period of low INR (≤1.5).70 The risks associated with warfarin therapy have prompted great interest in a group of drugs known collectively as the non-VKA oral anticoagulants in the general and Asian populations.
Non-VKA oral anticoagulants in Asians
The non-VKA oral anticoagulant drugs have been developed to provide alternative anticoagulation that is not hampered by the unpredictability and impracticalities associated with warfarin. The ideal anticoagulant therapy would be orally administered, achieve effective anti coagulation, with a favorable adverse-effect profile, predictable pharmacokinetics, and a wide therapeutic index precluding the need for monitoring and dose adjustment. Currently, two classes of non-VKA anticoagulants have reached phase III clinical trial and beyond—the direct thrombin inhibitors and the direct factor Xa inhibitors (Table 3). Unlike warfarin, which inhibits multiple targets in the coagulation cascade, the non-VKA anticoagulants are small-molecule, targeted inhibitors of individual coagulation factors. Both classes of drug lead to inhibition of the final common pathway of the coagulation cascade. The results of the clinical trials involving non-VKA anticoagulants are summarized in Table 4, and the results for Asian subgroups are shown in Table 5.
Direct thrombin inhibitors The development of orally administered, direct thrombin inhibitors has intensified since the elucidation of the 3D crystal structure of thrombin in 1989.71
www.nature.com/nrcardio © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Table 4 | Results from phase III trials of non-VKA drugs vs warfarin for the prevention of stroke and systemic embolism in nonvalvular AF Trial
Drug
Dose*
Study population (n)
Age (years)
Mean CHADS2 score
RR (95% CI) for non-VKA drugs versus warfarin Stroke or systemic embolism
Major haemorrhage
Intracranial haemorrhage
Death
RE-LY82
Dabigatran
150 mg twice daily
18,113
71‡
2.1
0.65 (0.52–0.81)
0.93 (0.81–1.07)
0.41 (0.28–0.60)
0.88 (0.77–1.00)
RE-LY82
Dabigatran
110 mg twice daily
18,113
71‡
2.1
0.90 (0.74–1.10
0.80 (0.70–0.93)
0.30 (0.19–0.45)
0.91 (0.80–1.03)
ROCKET-AF62
Rivaroxaban
20 mg once daily
14,264
73
3.5
0.88 (0.75–1.03)
1.04 (0.90–1.20)
0.67 (0.47–0.93)
0.85 (0.70–1.02)
J-ROCKET-AF97
Rivaroxaban
15 mg once daily
1,280
71
3.3
0.49 (0.24–1.00)
0.85 (0.50–1.43)
NA
NA
ARISTOTLE109
Apixaban
5 mg twice daily
18,201
70
2.1
0.79 (0.66–0.96)
0.69 (0.60–0.80)
0.42 (0.30–0.58)
0.89 (0.80–0.99)
ENGAGE AF‑TIMI 48120
Edoxaban
30 mg once daily
21,105
72
2.8
1.13 (0.96–1.34)
0.47 (0.41–0.55)
0.30 (0.21–0.43)
0.87 (0.79–0.96)
ENGAGE AF‑TIMI 48120
Edoxaban
60 mg once daily
21,105
72
2.8
0.87 (0.73–1.04)
0.80 (0.71–0.91)
0.47 (0.34–0.63)
0.92 (0.83–1.01)
*Dose adjustments were made in all trials on the basis of renal function, and in some trials on the basis of body mass and concurrent administration of other drugs. ‡Mean, all other ages are median. Abbreviations: AF, atrial fibrillation; NA, not available; non-VKA, non-vitamin K antagonist; RR, relative risk.
The enzymatic cleavage of prothrombin (factor II) to thrombin (factor IIa) by factor Xa initiates the final common pathway of the coagulation cascade (Figure 1). Thrombin and factor Xa are, therefore, essential effectors of haemostasis and attractive therapeutic targets for anticoagulants. Thrombin is a serine protease that, once activated, proteolytically cleaves factors V, VII, XI, and XIII to their active forms to result in further thrombin generation, and cleaves fibrinogen to fibrin to enable the development of a polymeric protein structure around which a clot can form. The protein structure of thrombin contains three domains—an active enzymatic site, and exosites 1 and 2 that interact with fibrinogen and heparin, respectively. Several parenterally administered anticoagulants that target thrombin directly or indirectly to achieve anticoagulation are licensed for a range of clinical uses and are extensively reviewed elsewhere.72,73 In this section, we include only those direct thrombin inhibitors that have been evaluated in Asian patients. Therefore, a discussion of ximelagatran is outside the scope of this Review. Dabigatran Dabigatran, a highly specific small-molecule peptido mimetic that specifically and directly inhibits thrombin, is the most-extensively studied orally administered direct thrombin inhibitor. This drug is administered as the lipophilic prodrug dabigatran etexilate, which is rapidly converted by ubiquitous plasma esterases to the active, hydrophilic drug dabigatran. The cytochrome P450 enzymes are not involved in the metabolism of dabigatran, and the risk of drug–drug interactions is considered low; however, interactions with p‑glycoprotein inhibitors are common and clinically releavant.82 Dabigatran was first investigated in the phase II PETRO study,83 in which 502 patients with nonvalvular AF and coronary artery disease plus another CHADS2
risk factor were randomly assigned to receive 50 mg, 150 mg, or 300 mg dabigatran twice daily with concurrent aspirin or to dose-adjusted warfarin alone.83 No conclusions could be drawn from the PETRO study regarding the efficacy of dabigatran in stroke preven tion. Howe ver, the upper limit of tolerability of dose escalation was identified. The PETRO study was shortly followed by the RE‑LY trial,58 a large, multicentre, international, prospective, randomized, open-blinded, end-point trial. Patients (n = 18,113) with nonvalvular AF plus one or more stroke risk factor (CHADS2 score >1) were allocated to 110 mg or 150 mg dabigatran twice daily or to open-label, doseadjusted warfarin with a target INR of 2.0–3.0. RE‑LY demonstrated the noninferiority of dabigatran in the prevention of stroke or systemic embolism both for 110 mg and 150 mg compared with warfarin. Furthermore, 150 mg dabigatran twice daily was proven to be statistically superior to warfarin for the prevention of stroke or systemic embolism (RR 0.66, 95% CI 0.53–0.82, P
Introduction
Rayne Institute, St Thomas’ Hospital, London SE1 7EH, UK (I.S., K.K.). St George’s Vascular Institute, Division of Cardiovascular Sciences (J.B.), Cardiovascular Sciences Research Centre (A.J.C.), St George’s University of London, Cranmer Terrace, London SW17 0RE, UK Correspondence to: I.S. [email protected]
Atrial fibrillation (AF) is the most common cardiac arrhythmia, affecting approximately 1% of the world’s population.1 AF is a supraventricular tachyarrhythmia characterized by unco-ordinated atrial activity and rapid ventricular responses, leading to the absence of defined P waves and irregularly irregular R–R intervals on the electrocardiogram. Functionally, AF leads to reduced cardiac output as a result of mechanical dissociation of atrial and ventricular contraction, and is an independent risk factor for multiple adverse outcomes, including transient ischaemic attack (TIA), stroke, dementia, heart failure, and cardiovascular and all-cause death.2,3–5 The mechanisms underlying the pathogenesis of AF remain under active investigation. Aberrant automaticity originating in the pulmonary veins is hypothesized to trigger the initiation, perpetuation, and maintenance of AF in myocardial tissue rendered susceptible to this arrhythmia by genetic and environmental factors. 6 Hypertension, cardiac failure, ischaemic and valvular Competing interests A.J.C. declares associations with the following companies: Actelion Pharmaceuticals, ARYx Therapeutics, Bristol-Myers Squibb Company, Cardiome Pharma Corporation, CV Therapeutics, Daiichi Sankyo, Johnson & Johnson Pharmaceutical Research & Development, Menarini Group, Merck & Co., Novartis Pharmaceuticals Corporation, Pfizer, Sanofi, SERVIER, and Xention Limited. See the article online for full details of the relationships. The other authors declare no competing interests.
290 | MAY 2014 | VOLUME 11
heart disease, diabetes mellitus, and advanced age are all well-established risk factors for the development of AF.7,8 Novel risk factors for AF, such as sleep apnoea, obesity, and metabolic syndrome, have also been identified.9,10 A heritable component of AF pathogenesis has been postulated; parental AF is an independent risk factor for the development of AF in offspring,11 and several candidate genes contributing to AF susceptibility have been identified in genome-wide association studies.12 The consequences of AF are serious; fibrillation of the atria can cause turbulent blood flow and stasis, increasing the risk of thrombus formation and systemic embolization, particularly from the left atrial appendage. AF raises the risk of ischaemic stroke almost fivefold, a risk magnified in individuals aged between 80 and 89 years, among whom AF has been found to be the major cardiovascular morbidity exerting an independent effect on the incidence of stroke.2 AF is estimated to account for around 15% of all strokes, rising to 25% in the elderly (age ≥70 years),13 and also independently raises the risk of congestive heart failure and all-cause mortality.4,14 Conventionally, warfarin and other vitamin K antagonists (VKAs) have been used to reduce the risk of thromboembolism in patients with AF. Effective treatment with warfarin relies on careful monitoring of patients to maintain anticoagulation within a target international normalized ratio (INR) range. However, achieving appropriate anticoagulation is complicated by pharmacogenomics,
www.nature.com/nrcardio © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Key points ■■ Although few prospective studies of atrial fibrillation (AF) epidemiology in Asian countries have been conducted, AF prevalence is thought to be lower in Asians than in white populations ■■ However, Asian individuals seem to be at a higher risk of stroke, and have greater stroke-related morbidity and mortality, than white individuals ■■ Asian individuals seem to be at particularly high risk of haemorrhagic stroke and intracranial haemorrhage, which has discouraged the widespread uptake of warfarin in Asian patients with AF ■■ Antiplatelet agents have been overused for primary stroke prevention in Asian patients with AF, in whom these drugs are neither safe nor effective ■■ Non-vitamin K antagonist anticoagulants are noninferior to warfarin for stroke prevention in AF, and are associated with a reduced incidence of haemorrhagic stroke and intracranial haemorrhage compared with warfarin ■■ The efficacy and safety profile of non-vitamin K antagonist anticoagulants in randomized, controlled trials suggest these drugs have great potential as anticoagulants in Asian individuals with AF
Table 1 | Prevalence of AF in Asia by country 1 Country
Community-based studies
Hospital-based studies
Prevalence of AF (%)
Number of studies
Prevalence of AF (%)
Number of studies
China
0.8–2.8
4
NA
NA
India
0.1
1
NA
NA
Japan
0.4–0.6 (women); 1.0–1.7 (men); 0.6–1.6 (overall)
10
12.2–14.0
2
Malaysia
NA
NA
2.8
1
Singapore
1.4
1
NA
NA
South Korea
0.7
1
NA
NA
Taiwan
1.1
1
NA
NA
Thailand
0.4–2.2
2
NA
NA
Abbreviations: AF, atrial fibrillation; NA, not available.
pharmacokinetics, drug–drug interactions, and patient adherence to drug regimens. In the Asian population, the increased prevalence of haemorrhagic stroke has led to under use of warfarin as an anticoagulant, in favour of antiplatelet agents. The non-VKA anticoagulants provide new therapeutic options for the management of thromboembolic risk related to AF in Asian populations. In this Review, the evidence supporting the use of non-VKA anticoagulants and their bleeding risk profiles in Asian patients is discussed, and we compare non-VKA anticoagulants to current management options for these individuals. We set these discussions in the context of the rising burden of stroke in Asian populations.
AF and stroke epidemiology in Asians
The global epidemiology of AF is an area of active research, particularly as the burden of AF and associated health-care costs increase.15 The white population of the western world has been comprehensively and prospectively studied. Pooled data suggest that the prevalence of AF in the general population of white individuals is between 0.5% and 2.0%, with men at higher risk than women, although the prevalence increases with age in both sexes.16–19 In one European population (Rotterdam,
Netherlands) of individuals aged ≥55 years, the lifetime risk of developing AF was 23.8% in men and 22.2% in women,17 estimates that are supported by data from the Framingham heart study.17,20 The epidemiology of AF within Asian populations has become increasingly clear. However, analysis of disease epidemiology by race, ethnicity, or geography is methodologically challenging, as study populations and definitions of ethnicity and race are heterogeneous.1 In addition, caution should be exercised in the interpretation of epidemiological data in migrant populations, as the relative contributions of environmental risk from their original countries and the countries to which they migrate might differ. Nevertheless, research has consistently demonstrated that the prevalence of AF is lower in Asian populations than in white populations.1 The Global Anticoagulant Registry in the Field (GARFIELD) study,21 in which 2,940 patients (27.7% of the total cohort) originated from the Asia–Pacific region, will greatly improve our understanding of this issue. A summary of AF prevalence in Asian countries is provided in Table 1. The point prevalence of AF has been reported to be 1.2% in the white population of South and East London compared with 0.22% in the South Asian community of the same area.22 In a small study conducted in the UK, the prevalence of AF in Indo-Asian patients aged >50 years was estimated to be 0.6%.23 These data are supported by studies performed in China and Japan, in which the age-standardized prevalence of AF was 0.65% and 0.56%, respectively.24,25 Moreover, prospective data collected using implanted cardiac devices demonstrated that new-onset AF lasting >6 h was less prevalent in Chinese and Japanese patients than in Europeans.26 Reduced prevalence of AF in Asian populations does not, however, translate into reduced stroke risk. Broadly, stroke is subclassified into ischaemic (that is, atherothrombotic, cardioembolic, and lacunar) and haemorrhagic aetiology. Asian populations have consistently been shown to be at a higher overall risk of stroke than white populations.27–29 The REACH registry 30,31 showed that, although Eastern Europe was the region in which patients were at highest risk of composite cardiovascular and cerebrovascular outcomes, individuals in Japan had the highest risk of haemorrhagic stroke. The relative burden of stroke compared with ischaemic heart disease is significantly greater in Asian countries, particularly China, compared with the West, as is the cost in disability-adjusted life years.28 This issue was highlighted in a systemic review in which the age-standardized incidence of a first stroke was higher among Chinese individuals (residing in China) than white individuals in European countries.27 South Asian individuals tend to present with acute first stroke at a younger age than East Asian and white individuals.32,33 The increased risk of stroke in South Asians is associated with greater mortality.29 A systemic review showed that the standardized mortality ratio for ischaemic stroke in South Asians was 55% and 41% greater in men and women, respectively, compared with the white population, a difference
NATURE REVIEWS | CARDIOLOGY
VOLUME 11 | MAY 2014 | 291 © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Contact activation (intrinsic) pathway
XII
Tissue factor (extrinsic) pathway
XIIa XI
XIa IX
IXa
Apixaban Betrixaban Edoxaban Rivaroxaban
Tissue factor VIIa
Warfarin (indirect via VKORC1)
VII
VIIIa Tissue factor Warfarin (indirect via VKORC1) Warfarin (indirect via VKORC1)
Xa
X
X
Va Prothrombin (II)
V Fibrinogen (I)
Thrombin (IIa)
Dabigatran
Fibrin (Ia) XIIIa Cross-linked fibrin clot
Figure 1 | Interaction between anticoagulant drugs and the coagulation cascade. Warfarin indirectly inhibits the synthesis of factors II, VII, IX, and X via inhibition of VKORC1. The direct thrombin (dabigatran) and direct factor Xa (apixaban, betrixaban, edoxaban, and rivaroxaban) inhibitors discussed in this article are also shown. Abbreviation: VKORC1, vitamin K epoxide reductase complex subunit 1.
consistent across Indian, Pakistani, and Bangladeshi ethnicities.29 Moreover, the prevalence of stroke in the South Asian community has increased by an estimated threefold to fivefold over the past 40 years, a phenomenon thought to be a consequence of the increasing prevalence of hypertension and hyperlipidaemia in this population.29 The CHADS2 and CHA2DS2-VASc scores, originally derived from white populations as tools for the stratification of stroke risk in patients with AF, have also been shown to be predictive of future stroke in some Asian populations.34–36 Interestingly, increased CHADS2 score was predictive of future stroke even in the absence of AF in a Japanese cohort.34 Although AF raises the relative risk (RR) of stroke fivefold in white populations,2 this risk seem to be more modest in Asians, ranging from a 2.8-fold to a 3.7-fold increase depending on the population studied.24,37–39 Nevertheless, CHADS2 or CHA2DS2VASc scores comparable with or significantly higher than those in white populations have been found to identify individuals at high risk of stroke in Asian populations22,26 despite the lower prevalence of AF. Interestingly, a study conducted in Birmingham, UK showed that a significantly lower percentage Indo-Asian patients with AF who were resident in this area were aware of the risk of stroke associated with AF (26% versus 70%; P 50%
Hepatic: predominantly CYP3A4 Substrate of P‑gp
30–120 min
60–120 min
8–15 h
~27%
Strong inhibitors and inducers of CYP3A4 and P‑gp
Edoxaban
Factor Xa
Once daily
45–61%
Hepatic: predominantly CYP3A4
30–60 min
60–120 min
9–11 h
35%
Strong P‑gp inhibitors
Betrixaban
Factor Xa
Once daily
34%
Minimal hepatic metabolism Substrate of P‑gp
180–240 min
180–240 min
20 h
17%
Strong P‑gp inhibitors
*Half-life is extended to 11–13 h in elderly patients (age ≥70 years). Abbreviations: ABCG2, ATP-binding cassette sub-family G member 2 (also known as breast cancer resistance protein); CYP, cytochrome P450; P‑gp; P‑glycoprotein; VKORC1, vitamin K epoxide reductase complex subunit 1.
event compared with 3.24% in Western Europe and 2.69% in North America.64 Failure to prescribe warfarin to individuals at high risk of thromboembolic events is of particular concern in Asian populations. Evidence suggests that war farin therapy is prescribed half as often in Asia as it is in Europe, irrespective of stroke risk as determined by CHA2DS2-VASc score.65 The main reason for under prescription is the perceived increased bleeding risk with warfarin; studies in the Japanese population have shown a higher incidence of haemorrhage than in white individuals, even among groups of patients with a low target INR.47 Instead, aspirin is commonly prescribed as an alternative to warfarin among Asians, despite evidence that low-dose aspirin is neither safe nor beneficial in Japanese patients with lone AF for stroke prevention.66 Antiplatelets do, however, have a role in secondary stroke prevention in Asian populations.67,68 Poor patient compliance might also be an issue; in a UK study, Indo-Asians were the ethnic group most likely to describe themselves as ‘careless’ when taking warfarin.40 Maximizing the percentage of time in the therapeutic range (TTR; that is, INR 2–3 for nonvalvular AF) is vital for war farin users to maximize the therapeutic effect for stroke prevention and minimize the associated risk of major haemorrhage.69 However, in a systematic review of 36 studies, the mean TTR varied between 59% and 67%, depending on the trial design.69 The mean time above the target INR range was 14%, and the mean time below the target INR range varied between 20% and 25%.69 In one US‑based study, an estimated one-fifth of all 294 | MAY 2014 | VOLUME 11
thromboembolic events occurred during a period of low INR (≤1.5).70 The risks associated with warfarin therapy have prompted great interest in a group of drugs known collectively as the non-VKA oral anticoagulants in the general and Asian populations.
Non-VKA oral anticoagulants in Asians
The non-VKA oral anticoagulant drugs have been developed to provide alternative anticoagulation that is not hampered by the unpredictability and impracticalities associated with warfarin. The ideal anticoagulant therapy would be orally administered, achieve effective anti coagulation, with a favorable adverse-effect profile, predictable pharmacokinetics, and a wide therapeutic index precluding the need for monitoring and dose adjustment. Currently, two classes of non-VKA anticoagulants have reached phase III clinical trial and beyond—the direct thrombin inhibitors and the direct factor Xa inhibitors (Table 3). Unlike warfarin, which inhibits multiple targets in the coagulation cascade, the non-VKA anticoagulants are small-molecule, targeted inhibitors of individual coagulation factors. Both classes of drug lead to inhibition of the final common pathway of the coagulation cascade. The results of the clinical trials involving non-VKA anticoagulants are summarized in Table 4, and the results for Asian subgroups are shown in Table 5.
Direct thrombin inhibitors The development of orally administered, direct thrombin inhibitors has intensified since the elucidation of the 3D crystal structure of thrombin in 1989.71
www.nature.com/nrcardio © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Table 4 | Results from phase III trials of non-VKA drugs vs warfarin for the prevention of stroke and systemic embolism in nonvalvular AF Trial
Drug
Dose*
Study population (n)
Age (years)
Mean CHADS2 score
RR (95% CI) for non-VKA drugs versus warfarin Stroke or systemic embolism
Major haemorrhage
Intracranial haemorrhage
Death
RE-LY82
Dabigatran
150 mg twice daily
18,113
71‡
2.1
0.65 (0.52–0.81)
0.93 (0.81–1.07)
0.41 (0.28–0.60)
0.88 (0.77–1.00)
RE-LY82
Dabigatran
110 mg twice daily
18,113
71‡
2.1
0.90 (0.74–1.10
0.80 (0.70–0.93)
0.30 (0.19–0.45)
0.91 (0.80–1.03)
ROCKET-AF62
Rivaroxaban
20 mg once daily
14,264
73
3.5
0.88 (0.75–1.03)
1.04 (0.90–1.20)
0.67 (0.47–0.93)
0.85 (0.70–1.02)
J-ROCKET-AF97
Rivaroxaban
15 mg once daily
1,280
71
3.3
0.49 (0.24–1.00)
0.85 (0.50–1.43)
NA
NA
ARISTOTLE109
Apixaban
5 mg twice daily
18,201
70
2.1
0.79 (0.66–0.96)
0.69 (0.60–0.80)
0.42 (0.30–0.58)
0.89 (0.80–0.99)
ENGAGE AF‑TIMI 48120
Edoxaban
30 mg once daily
21,105
72
2.8
1.13 (0.96–1.34)
0.47 (0.41–0.55)
0.30 (0.21–0.43)
0.87 (0.79–0.96)
ENGAGE AF‑TIMI 48120
Edoxaban
60 mg once daily
21,105
72
2.8
0.87 (0.73–1.04)
0.80 (0.71–0.91)
0.47 (0.34–0.63)
0.92 (0.83–1.01)
*Dose adjustments were made in all trials on the basis of renal function, and in some trials on the basis of body mass and concurrent administration of other drugs. ‡Mean, all other ages are median. Abbreviations: AF, atrial fibrillation; NA, not available; non-VKA, non-vitamin K antagonist; RR, relative risk.
The enzymatic cleavage of prothrombin (factor II) to thrombin (factor IIa) by factor Xa initiates the final common pathway of the coagulation cascade (Figure 1). Thrombin and factor Xa are, therefore, essential effectors of haemostasis and attractive therapeutic targets for anticoagulants. Thrombin is a serine protease that, once activated, proteolytically cleaves factors V, VII, XI, and XIII to their active forms to result in further thrombin generation, and cleaves fibrinogen to fibrin to enable the development of a polymeric protein structure around which a clot can form. The protein structure of thrombin contains three domains—an active enzymatic site, and exosites 1 and 2 that interact with fibrinogen and heparin, respectively. Several parenterally administered anticoagulants that target thrombin directly or indirectly to achieve anticoagulation are licensed for a range of clinical uses and are extensively reviewed elsewhere.72,73 In this section, we include only those direct thrombin inhibitors that have been evaluated in Asian patients. Therefore, a discussion of ximelagatran is outside the scope of this Review. Dabigatran Dabigatran, a highly specific small-molecule peptido mimetic that specifically and directly inhibits thrombin, is the most-extensively studied orally administered direct thrombin inhibitor. This drug is administered as the lipophilic prodrug dabigatran etexilate, which is rapidly converted by ubiquitous plasma esterases to the active, hydrophilic drug dabigatran. The cytochrome P450 enzymes are not involved in the metabolism of dabigatran, and the risk of drug–drug interactions is considered low; however, interactions with p‑glycoprotein inhibitors are common and clinically releavant.82 Dabigatran was first investigated in the phase II PETRO study,83 in which 502 patients with nonvalvular AF and coronary artery disease plus another CHADS2
risk factor were randomly assigned to receive 50 mg, 150 mg, or 300 mg dabigatran twice daily with concurrent aspirin or to dose-adjusted warfarin alone.83 No conclusions could be drawn from the PETRO study regarding the efficacy of dabigatran in stroke preven tion. Howe ver, the upper limit of tolerability of dose escalation was identified. The PETRO study was shortly followed by the RE‑LY trial,58 a large, multicentre, international, prospective, randomized, open-blinded, end-point trial. Patients (n = 18,113) with nonvalvular AF plus one or more stroke risk factor (CHADS2 score >1) were allocated to 110 mg or 150 mg dabigatran twice daily or to open-label, doseadjusted warfarin with a target INR of 2.0–3.0. RE‑LY demonstrated the noninferiority of dabigatran in the prevention of stroke or systemic embolism both for 110 mg and 150 mg compared with warfarin. Furthermore, 150 mg dabigatran twice daily was proven to be statistically superior to warfarin for the prevention of stroke or systemic embolism (RR 0.66, 95% CI 0.53–0.82, P
