Antithrombotic therapy for stroke prevention in atrial fibrillation and mechanical heart valves
Abstract
Cardioembolic strokes account for one-sixth of all strokes and are an important potentially preventable cause of morbidity and mortality. Vitamin K antagonists (e.g., warfarin) are effective for the prevention of cardioembolic stroke in patients with atrial fibrillation (AF) and in those with mechanical heart valves but because of their inherent limitations are underutilized and often suboptimally managed. Antiplatelet therapies have been the only alternatives to warfarin for stroke prevention in AF but although they are safer and more convenient they are much less efficacious. The advent of new oral anticoagulant drugs offers the potential to reduce the burden of cardioembolic stroke by providing access to effective, safe, and more convenient therapies. New oral anticoagulants have begun to replace warfarin for stroke prevention in some patients with AF, based on the favorable results of recently completed phase III randomized controlled trials, and provide for the first time an alternative to antiplatelet therapy for patients deemed unsuitable for warfarin. The promise of the new oral anticoagulants in patients with mechanical heart valves is currently being tested in a phase II trial. If efficacy and safety are demonstrated, the new oral anticoagulants will provide an alternative to warfarin for patients with mechanical heart valves and may also lead to increased use of mechanical valves for patients who would not have received them in the past because of the requirement for long term warfarin therapy. Am. J. Hematol. 2012. © 2012 Wiley Periodicals, Inc.
Introduction
Stroke is the second most common cause of death worldwide and a major cause of disability. Cardioembolic strokes accounts for approximately one-sixth of all strokes and are potentially preventable with the use of effective antithrombotic therapies [1]. Atrial fibrillation (AF) affects 1% of the population and is the most common cause of cardioembolic stroke [2]. Thrombus formation on the surface of mechanical heart valves is also a cause of cardioembolic stroke [3].
Antithrombotic treatment guidelines recommend long-term anticoagulant therapy for prevention of cardioembolic stroke in patients with AF [4, 5] and in patients with a mechanical heart valve [6, 7]. Vitamin K antagonists (e.g., warfarin) have been the only orally active anticoagulants available for clinical use for more than 60 years and, although effective, they have many limitations [8]. Warfarin has a slow onset of action, variable, and unpredictable anticoagulant effect because of genetic variability and numerous dietary and drug interactions, and is associated with a high risk of bleeding. Consequently, patients treated with warfarin require lifestyle restrictions and routine coagulation monitoring to maintain the international normalized ratio (INR) within a narrow target range [9]. These limitations contribute to the underutilization of warfarin for stroke prevention in patients with AF and contribute to suboptimal control of oral anticoagulant therapy.
Prompted by the need for effective, safe, and more convenient alternatives to warfarin, several new orally active anticoagulants have undergone evaluation in phase III trials for the prevention of cardioembolic stroke in patients with AF [10-13]. At least one trial of a new oral anticoagulant is presently ongoing in patients with a mechanical heart valve. Because of their demonstrated excellent efficacy and safety profile and improved convenience compared with warfarin, the new oral agents offer the potential to substantially reduce morbidity and mortality due to cardioembolic stroke. The objective of this article is to critically review the efficacy and safety of currently available and emerging antithrombotic strategies for cardioembolic stroke prevention in patients with AF and in those with mechanical heart valves.
Atrial Fibrillation
Risk stratification
Antithrombotic treatment guidelines recommend that physicians and patients base decisions on the choice of antithrombotic therapy in AF on an assessment of risks and benefits [4, 5].
For prediction of stroke risk, the CHADS2 scoring system (one point each for congestive heart failure, hypertension, age ≥75, diabetes mellitus; two points for prior stroke) has gained greatest acceptance because of its simplicity and broad applicability [14]. The 2006 American College of Cardiology/American Heart Association/European Society of Cardiology guidelines recommended anticoagulant therapy for patients with a CHADS2 score of 2 or above, either aspirin or anticoagulant therapy in those with a CHADS2 score of 1 (warfarin preferred over aspirin), and aspirin or no treatment in those with a CHADS2 score of 0 [15]. Improved risk prediction in patients with a CHADS2 score of 0 or 1 seems desirable because some of these patients are at sufficiently high risk of stroke that they would benefit from long-term anticoagulant therapy while others are at very low risk and are unlikely to benefit. The CHA2DS2VASc scheme has been proposed as a modification to the CHADS2 scheme in an attempt to improve stroke risk stratification in patients with a CHADS2 score of 0 or 1 (Table I) [4]. The most recent ESC guidelines recommend oral anticoagulation in patients with a CHA2DS2VASc score of 2 or above, oral anticoagulation or aspirin in those with a CHA2DS2VASc score of 1, and nothing or aspirin in those with a CHA2DS2VASc score of 0 [4]. A clear advantage of CHA2DS2VASc compared with CHADS2 for stroke risk stratification remains to be demonstrated.
| CHADS2 | CHA2DS2VASc | ||||
|---|---|---|---|---|---|
| Risk factor | Score | Risk factor | Score | ||
| C | Congestive heart failure | 1 | C | Congestive heart failurea | 1 |
| H | Hypertension | 1 | H | Hypertension | 1 |
| A | Age ≥75 | 1 | A | Age ≥75 | 2 |
| D | Diabetes mellitus | 1 | D | Diabetes mellitus | 1 |
| S | Stroke | 2 | S | Stroke/TIA/SE | 2 |
| V | Vascular diseaseb | 1 | |||
| A | Age 65 to 74 | 1 | |||
| S | Sex (female) | 1 | |||
| 6 (max) | 9 (max) | ||||
- Adapted from Camm et al. [4] with permission.
- Max, maximum; SE, systemic embolism; TIA, transient ischemic attack.
- a Includes left ventricular dysfunction as documented by echocardiography, radionuclide ventriculography, cardiac catheterization, or cardiac magnetic resonance imaging
- b Prior myocardial infarction, peripheral artery disease, or aortic plaque.
Several risk stratification schemes have been proposed for prediction of bleeding risk. The HAS-BLED scoring system has recently gained popularity but none of the currently available risk prediction tools are sufficiently accurate that they can be recommended for routine use [16]. Despite lack of adequate validation, the HAS-BLED scoring system has been incorporated into the most recent ESC guidelines [4].
Antithrombotic therapy for stroke prevention
Both antiplatelet drugs and anticoagulants are effective for stroke prevention in patients with AF although anticoagulants are substantially more effective than antiplatelet drugs [10-13, 17-20]. Evidence for the effectiveness of antiplatelet therapy comes from randomized controlled trials (RCTs) comparing aspirin to placebo/no antiplatelet therapy [18] and comparing the combination of aspirin and clopidogrel to aspirin [20]. Evidence for the effectiveness of anticoagulant therapy comes from RCTs comparing warfarin to placebo/no anticoagulant therapy [18], a new anticoagulant to warfarin [10, 12, 13] or an anticoagulant (warfarin or apixaban) to antiplatelet therapy (aspirin or aspirin plus clopidogrel) [11, 17, 19].
Antiplatelet therapy
Aspirin vs. placebo/no antiplatelet therapy.
Aspirin has been compared with placebo/no antiplatelet therapy in at least seven RCTs involving a combined total of 3,990 patients (Table II) [21-27]. A meta-analysis of the placebo-controlled trials has demonstrated that aspirin reduces the risk of any stroke (ischemic and hemorrhagic) by 22% (95% CI: 2–39%) [18, 28]. When these results are combined with the results of two additional trials comparing aspirin to control (no antiplatelet drug), the pooled relative risk reduction for any stroke is 19% (95% CI: −1 to 35%) [18]. The absolute benefits of aspirin are greater in patients with a history of prior stroke (2.5 events prevented per 100 patients treated for 1 year) than in patients without a history of stroke (0.8 events prevented per 100 patients treated for 1 year). The cost in terms of bleeding is an excess of 0.2 major extracranial bleeds for every 100 patients treated for 1 year [18]. Aspirin was not found to significantly increase intracranial bleeding in the AF trials but this likely reflects the relatively modest sample sizes. A meta-analysis of trials of aspirin for primary prevention of cardiovascular disease demonstrated that aspirin increased intracranial bleeding by 0.01 event per 100 patients treated for 1 year [29].
| Trial | Stroke rates (ischemic or hemorrhagic) | |||
|---|---|---|---|---|
| Aspirin (per 100 pts/yr) | Control (per 100 pts/yr) | RRR (%) | ARR (per 100 pts/yr) | |
| vs. placebo | ||||
| AFASAK [21] | 3.9 | 4.8 | 17 | 0.9 |
| SPAF [22] | 3.5 | 6.0 | 44 | 2.5 |
| EAFT [23] | 10.3 | 12.2 | 11 | 1.9 |
| ESPS-2 [24] | 13.8 | 20.7 | 29 | 6.9 |
| UK-TIA (300mg) [25] | 5.8 | 6.7 | 17 | 0.9 |
| UK-TIA (1,200 mg) [25] | 6.0 | 6.7 | 14 | 0.7 |
| Pooled (five trials) | 22 (2–39) | 1.9 (primary) 2.5 (secondary) | ||
| Vs. no antiplatelet | ||||
| LASAF (125 mg) [26] | 2.7 | 2.2 | −17 | −0.5 |
| LASAF (125 mg/2 days) [26] | 2.7 | 0.6 | 67 | 1.6 |
| JAST [27] | 2.2 | 2.0 | −10 | −0.2 |
| Pooled (seven trials) | 19 (−1 to 35) | 0.8 (primary) 2.5 (secondary) | ||
- Data from Hart et al. [18], adapted from Cairns et al. [28] with permission.
- ARR, absolute risk reduction; pts, patients; RRR, relative risk reduction; TIA, transient ischemic attack; UK, United Kingdom; US, United States; vs., versus; yr, year.
Several trials have evaluated antiplatelet strategies other than aspirin alone for stroke prevention in patients with AF. The treatments tested included aspirin plus an ineffective dose of warfarin, aspirin plus dipyridamole or dipyridamole alone, the trials collectively included less than 1,000 patients, and produced results that are consistent with those of the aspirin trials [18]. When the results of all 10 trials are combined the reduction in stroke is 22% (95% CI: 6–35%) [18].
Aspirin plus clopidogrel vs. aspirin.
The Atrial Fibrillation Clopidogrel Trial with Irbesartan for Prevention of Vascular Events (ACTIVE) A trial evaluated the efficacy and safety of the combination of aspirin and clopidogrel (75 mg per day) vs. aspirin alone in patients with AF and at least one additional risk factor for stroke who were deemed unsuitable for warfarin [30]. The recommended dose of aspirin was 75–100 mg per day. Among 7,554 patients with AF and at least one additional risk factor for stroke, the rate of the composite primary outcome, stroke, myocardial infarction, non-central nervous system systemic embolism or death from vascular causes was reduced by 11% with the combination of aspirin and clopidogrel compared to clopidogrel alone. This difference was primarily due to a 28% reduction in the rate of stroke (2.4% vs. 3.3% per year). The reduction in stroke was achieved at the cost of a 57% increase in major bleeding (2.0% vs. 1.3%). The relative increase in intracranial bleeding was 87%, corresponding to an absolute excess of 0.2 events per every 100 patients treated for 1 year [30].
The results of the ACTIVE A trial indicate that the combination of aspirin and clopidogrel is a modestly more effective alternative to aspirin in patients with AF and at least one additional risk factor for stroke. The reduction in stroke must be balanced against an increase in intracranial and extracranial bleeding as there was no difference in vascular or total mortality. The results of a net clinical benefit analysis that weighted thrombotic and bleeding events according to their impact on death or on death and disability suggest a possible benefit of combined aspirin and clopidogrel when compared with aspirin alone that may be greatest in patients with a history of prior stroke [31].
Summary of antiplatelet trials.
Trials of aspirin and other antiplatelet therapies provide conclusive evidence of their efficacy for stroke prevention in patients with AF. Most of the trials comparing aspirin with placebo/no antiplatelet therapy did not independently demonstrate a benefit of aspirin but the pooled data demonstrated about a one-fifth reduction in stroke, a more modest benefit than that of anticoagulants but still worthwhile for patients who are unable to be treated with anticoagulants. The addition of clopidogrel to aspirin provides additional benefit when compared with aspirin alone but at the cost of more bleeding, including an increased risk of intracranial bleeding.
Vitamin K antagonists
Warfarin vs. placebo/no anticoagulant therapy.
Six RCTs have compared warfarin to placebo/no anticoagulant therapy in patients with AF (Table III) [21-23, 32-34]. Most of these trials were conducted during the 1980s and several included a third group who were randomized to receive antiplatelet therapy alone. Four of the trials were stopped early by their Data Safety Monitoring Boards (DSMBs) because interim analyses suggested a large benefit of warfarin. A fifth trial was stopped early because of the positive results of other trials. Five of the six trials were performed in primary prevention and the sixth was performed in patients with a history of stroke [18, 28].
| Stroke rates (ischemic or hemorrhagic) | ||||
|---|---|---|---|---|
| Trial | Warfarin (per 100 pts/yr) | Control (per 100 pts/yr) | RRR (%) | ARR (per 100 pts/yr) |
| BAATAF [32] | 0.6 | 3.0 | 78 | 2.4 |
| CAFA [33] | 2.5 | 3.7 | 33 | 1.2 |
| SPINAF [34] | 1.4 | 4.8 | 70 | 3.3 |
| AFASAK [21] | 2.2 | 4.8 | 54 | 2.6 |
| SPAF [22] | 3.8 | 7.8 | 60 | 4.7 |
| EAFT [23] | 3.6 | 12.0 | 68 | 8.4 |
| Pooled (six trials) | 64 | 2.7 (primary) 8.4 (secondary) | ||
- Data from Hart et al. [18]; adapted from Cairns et al. [28] with permission.
- ARR, absolute risk reduction; pts, patients; RRR, relative risk reduction; vs., versus; yr, year.
A meta-analysis of the six trials involving a combined total of 2,900 patients demonstrated that adjusted-dose warfarin when compared with placebo/no anticoagulant therapy reduced the risk of stroke (ischemic or hemorrhagic) by 64% (95% CI: 49–74%) [18, 28]. The relative treatment effect was consistent in primary and secondary prevention settings but the absolute benefit in secondary prevention (8.4 fewer strokes per 100 patients treated for 1 year) was much greater than that in primary prevention (2.7 events prevented per 100 patients treated for 1 year). Warfarin when compared with placebo/no anticoagulant was associated with an excess of extracranial bleeding (0.3 events per every 100 patients treated for 1 year), a relatively modest increase when considering the absolute increase in bleeding with warfarin seen in the venous thromboembolism trials (two to three events per 100 patients treated for 1 year). These differences might be explained by the highly selected patient population included in the AF trials and by differences between the trials in definitions of bleeding. As with the aspirin vs. placebo/no antiplatelet therapy trials, no significant excess of intracranial bleeding was detected with warfarin. More recent RCTs evaluating different antithrombotic therapies for stroke prevention in AF have consistently demonstrated an excess of intracranial bleeding with warfarin [10, 12, 13, 17] (discussed in more detail below).
RCTs comparing warfarin to placebo/no anticoagulant therapy were conducted in an era when INR monitoring was less well standardized and routine cardiovascular prevention therapies such as blood pressure- and lipid-lowering were much less widely used [35]. Contemporary anticoagulant management practices may result in even greater benefits of warfarin with fewer bleeding complications than seen in the early AF trials. On the other hand, the more widespread use of effective secondary prevention strategies in recent trials may lower the risk of stroke and thereby reduce the absolute benefits of anticoagulant therapy for stroke prevention in patients with AF [35].
Warfarin vs. antiplatelet therapy.
Warfarin has been compared to aspirin in nine RCTs involving a combined total of 4,620 patients (Table IV) [21, 23, 36-42]. A meta-analysis of the nine trials demonstrates that adjusted-dose warfarin when compared with aspirin reduced the risk of stroke (ischemic or hemorrhagic) by 39% (95% CI: 19–53%) [18, 19]. The absolute benefits of warfarin when compared with aspirin were greater in secondary prevention (seven events prevented for every 100 patients treated for 1 year) than in primary prevention (0.9 events prevented for every 100 patients treated for 1 year). Warfarin compared with aspirin was associated with an excess of 0.2 extracranial bleeding events for every 100 patients treated for 1 year [18, 19]. One of the nine trials was conducted in patients aged at least 75 years of age (mean age 81.5 years) [42]. The results of this study demonstrated a 52% reduction in stroke, intracranial hemorrhage, or arterial embolism with warfarin when compared with aspirin with no increase in extracranial bleeding and indicate that warfarin is very effective and might not increase bleeding when compared with aspirin in carefully selected elderly patients.
| Trial | Stroke rates (ischemic or hemorrhagic) | |||
|---|---|---|---|---|
| Warfarin (per 100 pts/yr) | Aspirin (per 100 pts/yr) | RRR (%) | ARR (per 100 pts/yr) | |
| AFASAK [21] | 2.2 | 3.9 | 45 | 1.7 |
| SPAF-II, age ≤75 [36] | 1.7 | 1.9 | 10 | 0.2 |
| SPAF-II, age >75 [36] | 5.0 | 5.5 | 10 | 0.5 |
| EAFT [23] | 3.9 | 10.9 | 67 | 7.0 |
| AFASAK-II [37] | 3.1 | 2.5 | −23 | −0.6 |
| PATAF [38] | 0.7 | 1.0 | 20 | 0.3 |
| Vemmos et al. [39] | 0 | 4.0 | 100 | 4.0 |
| Chinese ATAFS [40] | 1.7 | 2.9 | 43 | 1.2 |
| WASPO [41] | 0 | 0 | – | – |
| BAFTA [42] | 2.5 | 4.9 | 47 | 2.3 |
| Pooled (10 trials) | 39 (19–53) | 0.9 (primary) 7.0 (secondary) | ||
- Data from Hart et al. [18,19]; adapted from Cairns et al. [28] with permission.
- ARR, absolute risk reduction; pts, patients; RRR, relative risk reduction; vs., versus; yr, year.
The ACTIVE W trial compared warfarin (target INR 2.0–3.0) with the combination of clopidogrel (75 mg per day) and aspirin (recommended dose 75–100 mg per day) in 6,006 patients with AF and at least one additional risk factor for stroke [17]. The overall results of the trial demonstrated that warfarin when compared with aspirin and clopidogrel reduced the risk of the composite of stroke, non-CNS embolism, myocardial infarction, and vascular death by 31% (95% CI: 15–43%) and stroke by 42% (95% CI: 19–74%). There was no significant difference in major bleeding but warfarin when compared with clopidogrel and aspirin doubled the risk of intracranial bleeding. In a post hoc analysis, warfarin was superior to the combination of aspirin and clopidogrel only at centers that achieved time-in-therapeutic range (TTR) for warfarin above the median for all centers participating in the trial; warfarin and the combination of aspirin and clopidogrel were similarly effective when center TTR was below the median [43].
The pooled results from RCTs comparing warfarin to aspirin are consistent with pooled estimates from trials comparing warfarin to any antiplatelet therapy (including ACTIVE W) [19]; a meta-analysis of all 12 trials involving 12,721 patients demonstrated that warfarin compared to antiplatelet therapy reduced the risk of stroke by 39% (95% CI: 27–49%) [19].
Summary of warfarin trials.
Initial RCTs evaluating the efficacy and safety of warfarin for stroke prevention in patients with AF included highly selected patients, targeted various INR ranges, and provided lower quality anticoagulant management than in more recent trials. Despite these limitations, the trials provided overwhelming evidence for the effectiveness of warfarin and have led guideline panels to strongly recommend their use in patients with AF and additional risk factors for stroke. Vitamin K antagonists are substantially more effective than antiplatelet therapies, although poorly managed warfarin may not provide any efficacy advantages over dual antiplatelet therapy.
New oral anticoagulants.
Potential pharmacological advantages of the new oral anticoagulants when compared with warfarin include a more rapid onset of action, low propensity for food, alcohol, and drug interactions and a more predictable anticoagulant effect that obviates the need for routine coagulation monitoring (Table V) [8]. Potential disadvantages of the new oral anticoagulants when compared with warfarin include the need for twice daily administration (dabigatran, etexilate, and apixaban), less established methods of coagulation monitoring, partial renal clearance, which can lead to accumulation in patients with severe renal insufficiency, and lack of a specific antidote (Table VI) [8]. The acquisition costs of the new oral anticoagulants are also higher than for warfarin.
| Feature | Warfarin | New oral anticoagulants |
|---|---|---|
| Onset | Slow | Rapid |
| Dosing | Variable | Fixed |
| Food effect | Yes | No |
| Interactions | Many | Few |
| Monitoring | Yes | No |
| Features | Warfarin | New agents |
|---|---|---|
| Frequency | Once daily | Twice dailya |
| Monitoring | INR | Not well established |
| Clearance | Non-renal | Renal 25–80% |
| Antidote | Vitamin K, FFP, PCC | Nil |
- FFP, fresh frozen plasma; PCC, prothrombin complex concentrates
- a Dabigatran etexilate and apixaban.
The pharmacological characteristics of dabigatran etexilate, rivaroxaban, and apixaban are summarized in Table VII [8]. Shared features of the new agents include half-life of around 12 h in patients with preserved renal function, a sufficiently predictable anticoagulant effect that obviates the need for routine coagulation monitoring, and the potential for interactions with drugs that are metabolized by P glycoprotein. Potentially important differences among the agents include their dosing frequency (rivaroxaban is given once-daily whereas dabigatran etexilate and apixaban are given twice daily) and their extent of renal clearance (80% for dabigatran etexilate, 66% for rivaroxaban, and 25% for apixaban). Other differences that are less relevant in clinical practice include their pharmacological target (Factor Xa vs. thrombin), whether they are a pro-drug (dabigatran etexilate) or direct acting (apixaban and rivaroxaban) and their bioavailability (6% for dabigatran etexilate vs. 60–80% for rivaroxaban and apixaban) [8].
| Feature | Dabigatran etexilate | Apixaban | Rivaroxaban |
|---|---|---|---|
| Coagulation target | Thrombin | Factor Xa | Factor Xa |
| Prodrug | Yes | No | No |
| Bioavailability | 6% | 60% | 80% |
| Dosing frequency | b.i.d. (o.d.) | b.i.d. | o.d. (b.i.d.) |
| Half-life | 12–14 h | 12 h | 7–11 h |
| Renal clearance | 80% | 25% | 33% |
| Routine monitoring | No | No | No |
| Drug interactions | P-gp | 3A4/P-gp | 3A4/P-gp |
- Adapted from Eikelboom et al. [8] with permission.
- P-gp, p-glycoprotein.
The theoretical advantages and disadvantages of Factor Xa when compared with thrombin as a target for new oral anticoagulants has been the subject of much debate [44, 45]. However, phase III trial results have demonstrated that both are valid targets; thus the choice of anticoagulant has to be based on efficacy and safety as well as cost and convenience rather than on theoretical considerations of target.
Dabigatran vs. warfarin.
The Randomized Evaluation of Long-term Anticoagulant Therapy (RELY) trial when compared with dabigatran etexilate, given at a dose of either 150 mg bid or 110 mg bid, to warfarin (target INR 2.0–3.0) for stroke prevention in patients with AF who had at least one additional risk factor for stroke [10]. The comparison between warfarin and dabigatran was open (i.e., patients and care providers were not blinded to treatment allocation) to facilitate optimal anticoagulant management in the warfarin control arm (median TTR was 67%) [46] and maximize ease of study conduct. The comparison between dabigatran doses was, however blinded and multiple strategies were put in place to minimize the potential for ascertainment and reporting biases. These strategies included screening of patients for outcome events at each visit, the use of hard (objectively confirmed, irreversible) outcomes to assess efficacy and safety, and adjudication of primary outcome events by two adjudicators who were blinded to treatment allocation [47]. Dabigatran 150 mg bid when compared with warfarin reduced the risk of stroke or systemic embolism by 34% (95% CI: 18–47%) whereas dabigatran 110 mg bid when compared with warfarin non-significantly reduced the risk of stroke or systemic embolism by 9% (95% CI: −11 to 26%) [10]. Both doses of dabigatran met the pre-specified criteria for noninferiority and dabigatran 150 mg bid met the criteria for superiority over warfarin. Dabigatran 150 mg bid when compared with warfarin was associated with a nonsignificant 7% reduction in major bleeding (95% CI: −7 to 19%), whereas dabigatran 110 mg bid when compared with warfarin was associated with a 20% reduction in major bleeding (7–31%). Both doses of dabigatran reduced intracranial hemorrhage by about two-thirds. The effects of both doses of dabigatran when compared with warfarin were consistent for the primary efficacy outcome in all subgroups examined [10]. For the outcome of major bleeding, there was a significant interaction between age and treatment such that patients over the age of 75 had similar or increased major bleeding whereas those below the age of 75 had similar or reduced major bleeding with dabigatran when compared with warfarin [48]. This interaction appeared to be confined to extracranial bleeding; intracranial bleeding was consistently reduced with both doses of dabigatran when compared with warfarin irrespective of age.
Three additional observations from the RELY trial merit further consideration. The first concerns dyspepsia, which was increased in patients treated with dabigatran when compared with warfarin and led to treatment discontinuation in approximately 2% of patients during the trial [10]. Dyspepsia in patients taking dabigatran is believed to be related to the tartaric acid formulation of the capsule and can be avoided in the majority of patients by taking the drug with a glass of water or food. The second observation concerns an increase in myocardial infarction in patients receiving dabigatran when compared with warfarin [49, 50]. This difference in myocardial infarction (MI) is most likely explained by superior efficacy of warfarin for MI prevention rather than a hazard of dabigatran, in the same way that the difference in stroke rate is most likely explained by superior efficacy of dabigatran for stroke prevention rather than a hazard of warfarin. The modest excess of MI is clearly outweighed by the reduction in stroke with dabigatran when compared with warfarin [51]. The third observation relates to gastrointestinal bleeding, which was increased in patients treated with the 150 mg bid dose of dabigatran when compared with warfarin [10]. This excess of gastrointestinal bleeding was evident with both doses of dabigatran but was confined to patients over the age of 75 [48].
Results of the RELY trial have prompted approval of dabigatran for stroke prevention in patients with AF in more than 75 countries worldwide. In most countries, both doses of dabigatran were approved but in the United States, only the 150 mg bid dose was approved based on the conclusion by the Food and Drug Administration (FDA) that there is no patient group in whom the 110 mg bid dose of dabigatran offers advantages over the 150 mg bid dose [52]. The FDA also approved the 75 mg bid dose of dabigatran for patients with severe renal impairment (Creatinine clearance 15–30 ml/min). In the majority of other countries, the availability of both doses provides clinicians with the opportunity to tailor dose according to patient's risk of bleeding. Thus, clinicians might prefer the 150 mg bid dose of dabigatran in most patients because it is the most effective for stroke prevention and reserve the 110-mg bid dose for patients deemed to be at high risk of bleeding (e.g., age >75, creatinine clearance 30–50 ml/min, concomitant antiplatelet therapies).
Rivaroxaban vs. warfarin.
The Rivaroxaban once daily, oral, direct factor Xa Inhibition Compared with vitamin K antagonist for prevention of stroke and Embolism Trial in AF (ROCKET-AF) trial investigators randomized 14,264 patients with AF and a prior history of stroke or at least two additional risk factors for stroke to receive rivaroxaban 20 mg once daily (15 mg once daily in those with a creatinine clearance of 30–49) or warfarin (target INR 2.0–3.0) [12]. The decision to perform a blinded trial minimized the potential for bias in the ascertainment and reporting of outcomes but may have contributed to suboptimal INR control (median TTR 58%). The pre-specified primary efficacy analysis was an on-treatment analysis, which demonstrated that rivaroxaban when compared with warfarin was associated with a 21% reduction in the risk of stroke or systemic embolism (95% CI: 4–34%), thus meeting the criteria for noninferiority. An intention-to treat analysis demonstrated a nonsignificant 12% reduction in stroke or systemic embolism (−3 to 26%). Rates of major bleeding were similar between the two groups but rivaroxaban was associated with a one-third reduction in intracranial bleeding. Efficacy results were consistent irrespective of center INR control and there was no excess of myocardial infarction in patients treated with rivaroxaban when compared with warfarin. Although various aspects of the design and analysis of the ROCKET-AF trial have been criticized [53], the FDA panel voted overwhelmingly in favor of approval. At the time of writing rivaroxaban has just been approved in the United States and Europe for stroke prevention in AF.
Apixaban.
Apixaban has been evaluated for stroke prevention in patients with AF in two phase III RCTs. One trial compared apixaban with aspirin in AF patients deemed unsuitable for warfarin [11] and the second compared apixaban with warfarin [13].
Apixaban vs. aspirin.
The Apixaban vs. Acetylsalicylic Acid to Prevent Strokes (AVERROES) trial investigators randomized 5,599 patients with AF and at least one additional risk factor for stroke who were deemed unsuitable for warfarin therapy to receive apixaban at a dose of 5 mg twice daily (2.5 mg twice daily in a subset of patients defined by age, creatinine, and body weight) or aspirin 81–324 mg daily [11]. Treatment allocation was blinded but patients could receive nonstudy aspirin at the discretion of the responsible physician. The AVERROES trial was stopped early because of overwhelming benefit of apixaban. Apixaban compared with aspirin was associated with a 54% reduction in stroke or systemic embolism (95% CI: 34–67%) with no excess of hemorrhagic stroke and no significant increase in major bleeding. Both efficacy and safety data were consistent in all subgroups examined. Patients randomized to receive apixaban when compared with aspirin also had a 12% lower rate of permanent study drug discontinuation (95% CI: 1–22%) [11].
Apixaban vs. warfarin.
The Apixaban for Reduction in Stroke and Other Thromboembolic Events in AF (ARISTOTLE) trial investigators randomized 18,201 patients with AF and at least one additional risk factor for stroke to receive apixaban 5 mg bid (2.5 mg bid in selected patients) or warfarin (target INR 2–3) in blinded fashion [13]. The median TTR was 66%, which is close to that achieved in the open RELY trial (67%) and reflects meticulous attention to anticoagulant control throughout the trial. Apixaban compared with warfarin reduced the risk of stroke or systemic embolism by 21% (95% CI: 5–34%) and major bleeding by 31% (95% CI: 20–40%) with a two-thirds reduction in intracranial bleeding. Efficacy results were consistent irrespective of center INR control and there was no excess of myocardial infarction in patients treated with apixaban when compared with warfarin [13].
Based on the results of the AVERROES and ARISTOTLE trials, apixaban is at the time of writing undergoing priority review by the FDA for stroke prevention in AF.
Summary of new oral anticoagulant trials.
Trials of new oral anticoagulant trials in AF were designed with the intent of demonstrating their noninferiority when compared with warfarin based on the premise that their greater convenience would lead them to replace warfarin even if they did not demonstrate superior efficacy. The trial results were unexpected, in two of the three trials (RELY, ARISTOLE), the new oral anticoagulant was statistically superior to warfarin for stroke prevention and in the third trial a consistent pattern of superiority was demonstrated (ROCKET-AF). In all three trials, the new oral anticoagulant was associated with numerically fewer deaths and markedly lower rates of intracranial bleeding (Fig. 1).
Intracranial hemorrhage is the most feared complication of vitamin K antagonist therapy in patients with AF and is responsible for the bulk of disability and death from anticoagulation-associated bleeding (Fig. 2). About 30% of major hemorrhages during warfarin therapy are intracranial, but these account for nearly 90% of deaths from hemorrhage and 75% of intracranial bleeds results in severe disability or death [54]. The risk of intracranial hemorrhage is often the deal-breaker in the risk vs. benefit decision for anticoagulation of AF patients. The reduction in intracranial hemorrhage in the phase III randomized trials comparing dabigatran, apixaban, and rivaroxaban with warfarin is key to their appeal, driving the overall stroke reduction and mortality benefit.
CT scan of intracerebral hemorrhage during warfarin anticoagulation. About half are fatal, and most survivors are left with severe disability,
For every 1,000 patients treated with warfarin, six to eight experience intracranial bleeding in the highly controlled setting of an RCT and the burden of intracranial bleeding may be even higher in registry studies [55, 56]. This risk is reduced by one- to two-thirds by the new oral anticoagulants, even in patients managed at centers with optimal INR control [46]. Most marked is the impact of the new oral anticoagulants on risk of intracerebral bleeding; event rates with the new agents are even lower than those seen with aspirin and approach the rates of intracerebral hemorrhage seen in patients of similar age not treated with any antithrombotic drugs (Table VIII) [10, 12, 13, 17-19].
| General population, age 70 yrs | 0.15%/yr |
| Dabigatran (either dosage) [10] | 0.1%/yr |
| Aspirin (any dosage) [18,19] | 0.2%/yr |
| Rivaroxaban in high-risk AF pts [12] | 0.33%/yr |
| Apixaban in AF pts [13] | 0.25%/yr |
| AF pts, age 70 yrs taking aspirin [18,19] | 0.3%/yr |
| Warfarin (INR 2–3) in AF pts [18,19] | 0.4–0.6%/yr |
Some experts have cautioned against the use of the new oral anticoagulants as first-line therapy for stroke prevention in patients with AF [57]. The authors believe, however, that the more favorable benefit/risk profile of the new oral anticoagulants when compared with warfarin, and, in particular, the markedly lower rates of intracranial bleeding means that they will likely replace warfarin as the preferred antithrombotic strategy in many patients with AF who do not have a contraindication (e.g., severe renal impairment) and can afford them. The highest priority to switch from warfarin to the new oral anticoagulants will be patients with poor INR control and those who cannot readily access the laboratory for monitoring. The new oral anticoagulants will also lower the threshold for the use of oral anticoagulants in patients at low-to-moderate risk for stroke who were not previously considered suitable for warfarin because of unfavorable benefit/risk balance.
Although only modestly effective for stroke prevention, aspirin is used as an alternative to warfarin in patients with AF because of its simplicity and low cost. The finding in the AVERROES trial that apixaban is much more effective than aspirin but does not increase major bleeding suggest that, once approved, it should be preferred over aspirin in the majority of AF patients.
Important unresolved issues remain. Outside the setting of an RCT, adherence to twice daily dosing regimen may be challenging. Because the new oral anticoagulants are partly renally cleared, they may accumulate in patients with severe renal impairment. The new anticoagulants lack an antidote, which poses challenges in patients with life-threatening bleeding and those who require urgent surgery. To optimize the benefits of the new oral anticoagulants for patients, physicians need education regarding patient and dose selection, potential drug interactions, appropriate follow-up, interruption of treatment for procedures, and management of bleeding. Anticoagulation clinics currently dedicated to monitoring warfarin treatment appear to be ideally positioned to take a major role in optimizing the management of the new oral anticoagulants in hospitals and the community.
Mechanical heart valves
Mechanical heart valves are thrombogenic because they activate coagulation (by exposing the blood to the artificial surfaces of the valve) and alter blood flow (by creating a region of stagnant blood flow) [3]. Consequently, patients with a mechanical heart valve require life-long anticoagulant therapy to prevent thromboembolic complications. Unlike patients with AF, there is near uniform uptake of vitamin K antagonist therapy in patients with mechanical heart valves. Eligibility for vitamin K antagonist therapy is assessed prior to making a decision about the choice of a mechanical or bioprosthetic valve; patients unwilling to take warfarin or deemed unlikely to be able to comply with dosing and monitoring requirements are more likely to receive a bioprosthetic valve.
Best estimates of the efficacy of anticoagulants for the prevention of thromboembolic complications in patients with mechanical heart valves come from a systematic review of observational studies performed by Cannegieter et al. [58]. Among 460 patients not receiving anticoagulants who were followed for 1,225 patient years, the annual risk of valve thrombosis was 1.8%, major embolism 4.0%, and total embolism 8.6% [58]. In a parallel group of patients who received vitamin K antagonists, the risk of thromboembolic complications was reduced by approximately 75%. Vitamin K antagonists have not been compared with placebo or no anticoagulant treatment in an RCT but clinical experience provides overwhelming evidence that they are effective.
Several RCTs have evaluated the efficacy of vitamin K antagonists when compared with antiplatelet therapy [59], the optimal intensity of oral vitamin K antagonist therapy [60, 61], and optimal approaches to monitoring oral anticoagulant therapy in patients with mechanical heart valves [62-64]. Their results indicate that vitamin K antagonists are superior to antiplatelet therapy for the prevention of thromboembolic complications in patients who have completed 6 months of anticoagulant therapy, rates of thromboembolic complications in patients with a mechanical aortic valve replacement are similar when the vitamin K antagonist therapy is adjusted to achieve an INR of two to three when compared with a higher INR, and that self-monitoring of anticoagulation is at least as effective as conventional management by physicians or an anticoagulation clinic.
The efficacy of antiplatelet therapy when compared with placebo has been tested on top of vitamin K antagonist therapy in trials that included patients with bioprosthetic or mechanical heart valves. A meta-analysis of 11 such trials involving 2,428 patients demonstrated that the addition of antiplatelet therapy significantly reduced thromboembolic complications and mortality at the cost of an increase in bleeding [65]. Results were consistent irrespective of valve type. Based on these results, the ACC/AHA guidelines recommend the addition of aspirin (75–100 mg/day) in all patients with mechanical heart valves [7].
New oral anticoagulants
The success of the large phase III RCTs comparing new oral anticoagulants to warfarin for stroke prevention in patients with AF has prompted interest in their evaluation as an alternative to warfarin in patients with mechanical heart valves. If a new anticoagulant is shown to be effective and safe for this indication, they will likely become the preferred treatment because of their improved convenience (no dietary and drug restrictions, no need routine coagulation monitoring). New oral anticoagulants may also impact decisions regarding the choice of a mechanical or bioprosthetic valve, mechanical heart valves are more durable than biological valves, and will likely be used in a higher proportion of patients if there is an effective, safe, and more convenient alternative to warfarin.
Dabigatran.
Following successful in vitro and animal studies with the use of dabigatran, the currently ongoing Randomised, phase II study to Evaluate the sAfety and pharmacokinetics of oral dabiGatran etexilate in patients after heart valve replacement (RE-ALIGN) trial is comparing dabigatran to warfarin for prevention of thromboembolic complications in patients undergoing mechanical aortic or mitral valve replacement. Dabigatran is being tested at doses of 150, 220, and 300 mg twice daily, with dose-adjustment according to drug levels. A large phase III trial will be required to demonstrate the efficacy and safety of dabigatran as an alternative to warfarin for the prevention of thromboembolic complications in patients with a mechanical heart valve before the drug is introduced into clinical practice for this indication.
Conclusion
Cardioembolic strokes are an important potentially preventable cause of morbidity and mortality. Vitamin K antagonists are effective for the prevention of cardioembolic stroke but because of their inherent limitations are underutilized in patients with AF and often suboptimally managed. Antiplatelet therapies have been the only alternatives to vitamin K antagonists for stroke prevention in AF, but although they are safer and more convenient than warfarin, they are much less effective. The limitations of vitamin K antagonists also hamper the management of patients requiring valve replacement surgery because of the requirement for lifelong anticoagulant therapy for mechanical heart valves. The advent of new oral anticoagulant drugs as an alternative to both aspirin and warfarin offers the potential to substantially reduce the burden of morbidity and mortality due to cardioembolic stroke by providing access to effective, safe, and more convenient therapies and greater access to mechanical heart valves for patients who require a valve replacement.
Disclosures
Dr. Eikelboom has received honoraria and/or research support from companies that develop and market antithrombotic therapies for prevention of cardioembolic stroke in patients with AF and mechanical heart valves, including Bayer, BMS, BI, J&J, Pfizer, and Sanofi. Dr. Hart has been a consultant to Sanofi-Aventis for service on the stroke advisory and adjudication committee for a phase III randomized clinical trial testing a novel antiarrhythmic agent in patients with AF (2010–2011; PALLAS trial; <10,000 per year), a consultant to Bristol Myers&#150;Squibb for service on the operations, stroke advisory, and stroke adjudication committee for a phase III randomized clinical trial testing the novel oral anticoagulant apixaban in patients with AF (AVERROES trial; <10,000 per year; 2008–2010), a consultant to Biotronik for service on a data monitoring committee for a clinical trial assessing novel detection device for AF (IMPACT; <10,000 per year) and a consultant to Boehringer Ingelheim for analysis of intracranial hemorrhages occurring during a phase III trial testing dabigatran vs. warfarin in patients with AF (2010; RE-LY; <10,000). He is the co-principal investigator on an NIH/NINDS-sponsored phase III multi-center clinical trial testing antiplatelet agents and blood pressure control for secondary prevention of small subcortical lacunar&#148; strokes (SPS3) for which he receives salary support (>10,000 per year).
