Warfarin plus Aspirin after Myocardial Infarction or the Acute Coronary Syndrome: Meta-Analysis with Estimates of Risk and Benefit FREE

Continued thrombin generation persists for several months after acute cardiac events.

Data from 10 randomized trials involving 5938 patients with the acute coronary syndrome who were not stented showed that, compared with aspirin alone, warfarin plus aspirin decreased annual rates of myocardial infarction, ischemic stroke, and revascularization and increased major bleeding rates. In patients with low or average bleeding risks, numbers of cardiovascular events prevented by warfarin plus aspirin exceeded numbers of major bleeding episodes caused by it.

Benefits of warfarin plus aspirin may exceed harms in patients with the acute coronary syndrome who are not stented and do not have high bleeding risks.

–The Editors

Myocardial infarction is a leading cause of illness and death in the United States (1). Patients with a history of myocardial infarction are at increased risk for recurrent infarction, stroke, and death (2 - 3). Several interventions have proven beneficial in the secondary prevention of myocardial infarction, including β-blockers (4), angiotensin-converting enzyme inhibitors (5), lipid-lowering therapy (6), and aspirin (7). In addition, after acute cardiac events, a marked thrombin generation persists for months after clinical stabilization (8), suggesting a role for anticoagulation beyond the initial use of low-molecular-weight heparin (9). Although some studies have shown that addition of warfarin to aspirin decreases subsequent risk for cardiovascular events (10 - 14), other studies have not (15 - 16). In a meta-analysis, Anand and Yusuf (17) suggested that only moderate- to high-intensity anticoagulation, with a target international normalized ratio (INR) more than 2.0, decreased cardiovascular events more than aspirin alone. A large randomized trial conducted in a primary care setting showed that adding warfarin to aspirin decreased combined cardiovascular end points of myocardial infarction, stroke, and death by 29% (12). Despite these findings, warfarin use has not been widely adopted, perhaps because of concern that the increased bleeding risk may counter the cardiovascular benefit (18) or that the benefits are too small to justify the inconvenience (19).

The risks and benefits that are probably associated with warfarin therapy are not equally distributed. Patients at the highest risk for recurrent cardiovascular events should derive the most benefit, which may be offset in those with increased bleeding risk. To better quantify the risks and benefits of warfarin therapy for individuals, we conducted a meta-analysis of randomized trials and then calculated the expected benefit for groups at varying risk for cardiovascular disease and bleeding.

We searched MEDLINE using the following Medical Subject Heading (MeSH) terms and text words: anticoagulant or anticoagulation or warfarin and aspirin, combined with myocardial infarction, unstable angina, coronary arteriosclerosis, or coronary artery disease. We limited our search to original, English-language articles published between 1 January 1990 and 1 October 2004; earlier trials were not likely to reflect current standards of care for secondary prevention, including lipid therapy and β-blockers.

Two reviewers scanned the titles independently for relevant randomized trials. We checked abstracts and then manuscripts of all potentially appropriate references. We included studies if they were randomized, controlled trials of warfarin and aspirin in patients with an acute coronary syndrome. We excluded studies of percutaneous coronary stenting (for which warfarin is generally not considered appropriate therapy) (20) and low-intensity warfarin therapy (target INR < 2.0). We assessed studies for proper randomization, blinding of patients and investigators to treatment allocation, and completeness of follow-up.

The 2 reviewers extracted the following data independently: study size and duration; aspirin dose; target INR; and demographic characteristics of trial participants, including age, sex, number of patients with diabetes, smoking status, systolic blood pressure, and whether the patient had a Q-wave infarction or received thrombolysis. We evaluated 7 end points: myocardial infarction, ischemic stroke, revascularization, death, minor bleeding, major bleeding, and intracranial hemorrhage. Each study's authors defined major and minor bleeding, and most studies reported all 7 end points. When results were presented only as combined end points or were ambiguous, we contacted the study authors to obtain primary data.

In the Organization to Assess Strategies for Ischemic Syndromes (OASIS) study (13), a multicenter, international trial, investigators noted 35 days into the trial that compliance as measured by achievement of target INR was better in some countries than in other countries and designated a subset analysis of countries with good compliance before data collection was complete. We included only the compliant countries in our analysis.

We contacted the OASIS authors for select data on compliant countries from OASIS (13) and individual end points from the OASIS pilot study (21), which were published only as combined end points. We contacted the authors of Huynh and colleagues' study (22), because the published results for all end points were incorrect.

We used Stata, version 8.2 (Stata Corp., College Station, Texas), for all analyses. We calculated the rate ratios and 95% CIs for recurrent myocardial infarction, stroke, and bleeding (major or minor) for each study. We used a half-integer correction if no events occurred in 1 group of the study. We pooled rate ratios by using random-effects models that used weighting based on the inverse-variance model according to DerSimonian and Laird (23). We used the Mantel–Haenszel test to evaluate heterogeneity among trial outcomes. We considered statistical significance at a P value less than 0.05. We performed an influence analysis in which we computed the summary rate ratio, omitting the largest trial (Warfarin, Aspirin, Reinfarction Study [WARIS II] [12]), to assess for any single study dominance in the analysis. We used the Begg and Mazumdar adjusted rank correlation test (24) and the Egger regression asymmetry test (25) to evaluate publication bias.

To demonstrate the relative benefits of warfarin plus aspirin for individual patients at different degrees of risk, we calculated the number of myocardial infarctions, strokes, and major bleeding episodes expected to result from aspirin therapy, with or without warfarin, for 9 combinations of cardiovascular and bleeding risk. We calculated events in each risk group taking warfarin plus aspirin by multiplying the rate for the corresponding group of patients taking aspirin by the rate ratio derived from the meta-analysis. The absolute differences between these rates represent the number of myocardial infarctions and strokes prevented by adding warfarin to aspirin. Because the timing of events may be important in treatment decisions, we calculated outcomes at 3 months and 1 year. On the basis of 5059 patients in the Combination Hemotherapy and Mortality Prevention (CHAMP) study (16), we estimated that 51% of myocardial infarctions, 54% of strokes, and 50% of bleeding episodes in the first year after acute myocardial infarction would occur in the first 3 months. Event rates for all 3 outcomes continue to decrease over time. Thus, considering events in the first 2 years after an acute coronary syndrome, approximately 70% of all outcomes occur in the first year. To directly compare event rates across studies of differing duration, we converted all results into first-year rates by using the proportions just described.

Many published instruments are available for determining short- and long-term risk for reinfarction, stroke, and mortality. To demonstrate the clinical usefulness of estimating both cardiovascular and bleeding risk before prescribing warfarin, we chose clinical examples of low-, medium-, and high-risk patients. We derived our estimates of the rates for recurrent myocardial infarction and stroke for patients receiving aspirin by using the computed hazard ratios from a large population-based study of 2700 patients surviving hospitalization after a first myocardial infarction and followed for an average of 3.4 years (26). We chose this study for its large size, long follow-up period, emphasis on easily measured clinical predictors, and inclusion of stroke as an outcome. The following variables were associated with increased risk for recurrent myocardial infarction and stroke: age, treated diabetes, chronic congestive heart failure, angina, and serum creatinine level greater than 123.76 µmol/L (>1.4 mg/dL). We created examples of low-, medium-, and high-risk patients by applying the hazard ratios for specific risk variables to a baseline risk of 49 per 1000 patient-years of observation. Infarction rates per 1000 patient-years were 40 for a low-risk patient (no risk factors), 82 for a medium-risk patient (for example, diabetes and serum creatinine level >123.76 µmol/L [>1.4 mg/dL]), and 188 for a high-risk patient (for example, diabetes and congestive heart failure). Stroke rates for the same patients were 13, 32, and 79, respectively. Other risk stratification tools, based on the Thrombolysis In Myocardial Infarction (TIMI II) (27), Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto miocardico (GISSI) (28), Secondary Prevention Reinfarction Israeli Nifedipine Trial (SPRINT) (29), or Global Registry of Acute Coronary Events (GRACE) (30) studies, produce similar ranges of low-, medium-, and high-risk patients. A detailed calculator based on the GRACE model is available at http://www.outcomes-umassmed.org/grace.

Several instruments are also available for predicting bleeding risk for patients taking warfarin (31 - 34). We chose examples of low, medium, and high risk for bleeding by using the Outpatient Bleeding Risk Index devised by Landefeld and Goldman (35), which is the only prediction tool prospectively validated in clinical practice (36 - 37). Landefeld and Goldman identified 5 independent risk factors for major bleeding in outpatients: age 65 years or older, history of stroke, history of gastrointestinal bleeding, a specific comorbid condition (renal insufficiency, recent myocardial infarction, or severe anemia), and atrial fibrillation. Incidence of major bleeding in 2 prospective validation sets in the first 12 months was 1% for low-risk patients (no risk factors), 7% for medium-risk patients (1 or 2 risk factors), and 30% for high-risk patients (≥3 risk factors) (36 - 37). Approximately half of all bleeding events occurred in the first 3 months of therapy. We derived predicted rates of bleeding with aspirin alone by dividing the bleeding rates with warfarin and aspirin by the relative risk for warfarin plus aspirin versus aspirin. The difference between these rates represents the absolute number of bleeding episodes attributable to the addition of warfarin.

The initial search yielded 656 articles (Appendix Figure 1). Ten studies met the inclusion criteria, involving a total of 5938 patients (11 334 patient-years). All studies reported rates for myocardial infarction and major and minor bleeding. Nine studies reported rates for intracerebral hemorrhage and death, 8 studies reported revascularization rates, and 7 studies reported rates of ischemic stroke.

Appendix Table 1 shows the study designs. There was a mix of large and small studies, with 3 studies including more than 600 patients. Follow-up ranged from 3 months to 4 years; most studies lasted 12 months or less. Aspirin dose varied from 80 to 325 mg daily. Target INR for most studies was 2.0 to 2.5. All studies were intention-to-treat analyses, and compliance measured by target INR achieved ranged from 2.0 to 3.0. All studies used proper randomization techniques and had end points adjudicated by investigators blinded to treatment allocation. Because of the difficulty in maintaining placebo warfarin therapy, only 2 studies were double-blind. Seven studies had greater than 99% clinical follow-up, and all studies had follow-up of at least 90%.

The annualized rate of myocardial infarction in the aspirin group ranged from 0.03 to 0.93 (Table 1). Nine of 10 studies found a risk reduction attributable to warfarin, but only 2 studies were sufficiently powered for the reduction to reach statistical significance. Reductions in relative risk ranged from 29% to 100%, with an overall risk reduction of 44% (Figure 1). Excluding the 2 largest studies had little effect on the risk reduction, which increased to 45% (95% CI, 18% to 63%). Event-free survival curves continued to diverge for at least 5.5 years of therapy. Pairwise correlation analysis suggested that studies showing greater risk reductions with warfarin therapy tended to have more bleeding complications as well (r = 0.7; P < 0.036).

The annualized risk for ischemic stroke in the aspirin group ranged from 0.0 to 0.08, with a weighted average of 0.008. All 5 studies in which at least 1 stroke was reported found a risk reduction for the warfarin plus aspirin group, but only 1 risk reduction was statistically significant (Figure 2). Reductions in relative risk ranged from 50% to 100%, with an overall risk reduction of 54% (CI, 23% to 73%). Removing the largest study improved the risk reduction to 82% (CI, −1% to 97%).

Hemorrhagic strokes were counted separately and included in the estimates of major bleeding. Overall, 4 hemorrhagic strokes occurred in the warfarin group and 1 occurred in the aspirin group, translating to 1 additional intracranial hemorrhage per 1800 patient-years of combined anticoagulation.

The annualized risk for revascularization ranged from 0.076 to 1.3. Five of 7 studies showed decreased rates of percutaneous transluminal coronary angioplasty or coronary artery bypass grafting for patients receiving warfarin therapy, but only 1 rate reached statistical significance (Appendix Figure 2). Hazard ratios ranged from 0.51 to 1.70, with an overall relative risk reduction of 20% (CI, 5% to 33%).

No study showed a statistical difference in mortality (Figure 3). The combined studies showed a 4% decrease in overall mortality in the warfarin group, but this did not reach statistical significance.

Nine studies showed an increased risk for major bleeding associated with warfarin therapy (Figure 4). The annualized risk for major bleeding in the warfarin group ranged from 0.6% to 18.0%, with an overall risk of 1.5%. The relative risk for major bleeding with warfarin compared with aspirin was 2.5 (CI, 1.7 to 3.7). The relative risk for minor bleeding was 2.6 (CI, 2.0 to 3.3) (Appendix Figure 3).

The Mantel–Haenszel test showed no statistically significant heterogeneity for any end point among the studies (Appendix Table 2). An influence analysis excluding 1 study at a time and finally the 2 largest studies simultaneously (WARIS II [12] and OASIS [13]) had no substantial effect on the summary rate ratio. No evidence suggested publication bias for any end point, with the exception of stroke, and only the Egger regression asymmetry test suggested some bias. Stroke was not a primary end point in any trial.

As expected, studies lasting 3 months had the highest rates of myocardial infarction, stroke, and bleeding. Rates for all 3 outcomes decreased with increasing study duration. Projected first-year rates of myocardial infarction and major bleeding episodes varied considerably among the studies, but relative risk reductions from warfarin were the same for all outcomes, regardless of study duration or indication (Table 2).

Figure 5 shows the cardiovascular events averted and excess bleeding episodes caused by adding warfarin to aspirin for 1 year for different combinations of cardiovascular and bleeding risk. For most studies, the benefit of warfarin clearly outweighs the risk. Individuals at the highest risk for cardiovascular events (for example, a patient with diabetes and congestive heart failure) could avert 83 myocardial infarctions and 43 strokes per 1000 patient-years of warfarin therapy, while those at the lowest cardiovascular risk would avoid 18 myocardial infarctions and 7 strokes. Regardless of cardiovascular risk, adding warfarin to aspirin will cause an excess of 6 to 180 major bleeding episodes per 1000 patient-years, depending on the individual's bleeding risk. Because rates for all outcomes decrease over time, approximately half of the benefit and risk in the first year of therapy accrues in the first 3 months. Conversely, both benefits and risks are reduced by 60% in the second year of therapy.

Our meta-analysis bolsters the finding that after an acute coronary syndrome, warfarin plus aspirin decreases the rate of myocardial infarction or stroke more than aspirin alone. Although the probability of major bleeding is also increased, the benefits far outweigh the risks for many patients. For individuals at high cardiovascular risk but at low risk for bleeding (for example, a 58-year-old man with diabetes and congestive heart failure), adding warfarin to standard aspirin therapy could avert 83 myocardial infarctions and 43 strokes per 1000 patient-years of therapy at a cost of just 6 major bleeding episodes. The number needed to treat for 3 months to prevent 1 major cardiovascular event is 16, while the number needed to harm for the same period is 333. Even when it is not possible to fully separate bleeding risk from cardiovascular risk, warfarin still seems beneficial, as long as patients at the highest risk for bleeding are excluded from therapy.

We found no difference in relative risk reduction over time, but the absolute rates of all end points decrease steadily. As a result, the greatest benefits and risks occur in the first 3 months of therapy. Still, the curves for the combined end points continue to diverge for at least 5 years. For patients at high cardiovascular risk, the gains beyond 1 year are likely to be substantial, whereas those at low risk may choose to stop therapy after 3 months.

Despite an almost 3-fold increase in major bleeding, overall bleeding rates among patients taking warfarin and aspirin during 5500 patient-years of observation were low, similar to the rate seen in patients with atrial fibrillation or deep venous thrombosis (42 - 43). The low rate in WARIS II (0.6% per year) may have resulted from the intention-to-treat design (12). Some patients presumably did not have bleeding episodes because they received inadequate anticoagulation or had stopped taking warfarin altogether. An annual bleeding rate of 1.5% is probably more representative of adherent patients. Similarly, adherent patients could expect fewer myocardial infarctions and strokes than were seen in the intention-to-treat analysis, as was found in both the OASIS trial (13) and the Thrombosis Prevention Trial (44) when each was analyzed according to adherence to warfarin.

Our study has several limitations. Although the designs of the individual studies were similar, they were not identical. All meta-analyses contain some degree of heterogeneity, as do multicenter randomized trials. Despite minor variations in eligibility criteria, study duration, and aspirin dosage, the studies selected were remarkably similar. Although 2 large trials accounted for most patients, without these, the relative risk reduction for myocardial infarction was almost identical, emphasizing the homogeneity of the trials.

We chose to include only the compliant countries from the OASIS trial (13) because including countries where fewer than 70% of the patients took the therapy would bias our results heavily in favor of the null hypothesis. Although this violates the strictest interpretation of intention to treat, the designation of compliant countries was made only 35 days into the study and without examining clinical outcomes. Because entire countries were excluded, patient-specific bias could not occur. It is also encouraging that the effect size seen in compliant countries from OASIS (13) was almost identical to that seen in WARIS II (12).

All studies were conducted in the 1990s before coronary artery stenting became widespread. As a result of several comparative trials, warfarin is not considered a standard of care in stented patients (45 - 47). However, outside of large metropolitan areas in the United States, most patients with myocardial infarction still do not undergo stenting. Moreover, none of these studies examined long-term anticoagulation after the initial stenting period.

Finally, we assumed that the decrease in relative risk for myocardial infarction and stroke would be consistent across risk groups. Although this assumption has been found to be true with clopidogrel (48), simvastatin (49), and aspirin (50), additional trials of moderate- to high-risk patients, including those with diabetes and heart failure, would be helpful, especially because warfarin may not benefit diabetic patients (51) and aspirin may be harmful in patients with heart failure (52).

Despite the clear benefits demonstrated in WARIS II (12), no recommendations have been made for use of warfarin to prevent reinfarction and stroke. The most recent American College of Cardiology/American Heart Association guidelines (53) for treatment of myocardial infarction were published before the results of WARIS II were available. Three general objections have been raised to warfarin: The increased bleeding risk offsets any potential cardiovascular benefit (18), the cardiovascular gains are too small to justify the inconvenience of warfarin therapy (19), and studies have not demonstrated a mortality benefit (54).

On the basis of our analysis, the benefits of warfarin should outweigh the harms for most patients. Even those for whom the absolute numbers of major bleeding episodes and cardiovascular events are similar may benefit because cardiovascular events are usually more serious than bleeding events. Approximately 25% of reinfarctions are fatal, and among patients who survive, risk for congestive heart failure and disability increases. Ischemic stroke is even more likely to result in permanent disability. An atrial fibrillation study showed that warfarin therapy reduces not only the frequency but also the severity of strokes (55). Apart from intracranial hemorrhage, bleeding episodes rarely result in long-term sequelae. In our meta-analysis, warfarin resulted in 1 excess nonfatal intracranial hemorrhage per 1800 patient-years of therapy.

While the absolute benefit to participants in WARIS II was small because of their low overall risk (3% per year for myocardial infarction and 1% per year for stroke), patients were mostly young and nondiabetic and had Q-wave infarctions with preserved left ventricular function. Our study demonstrated that warfarin produces the same relative risk reduction in high-risk patients, such as those in the Antithrombotic Therapy in Acute Coronary Syndromes (ATACS) (38 - 39) and Antithrombotics in the Prevention of Reocclusion In Coronary Thrombolysis-2 (APRICOT-2) (11) studies, yielding considerable absolute benefit. Whether this benefit is worth the inconvenience of frequent blood draws and dietary restrictions is a decision that is best left to individual patients and their physicians.

Despite decreasing myocardial infarctions and strokes, warfarin had no demonstrable mortality benefit. However, mortality in all groups was relatively low, making it difficult to detect a difference. Moreover, study durations were short and mortality due to congestive heart failure may be delayed several years. A similar reduction in nonfatal myocardial infarction was shown in the Clopidogrel in Unstable Angina to Prevent Recurrent Events (CURE) trial (56), which also did not demonstrate a mortality benefit among 12 500 patients who were followed for up to 1 year.

Like insulin, warfarin engenders resistance in primary care physicians and patients alike. Despite ample evidence for more than a decade that warfarin reduces stroke in atrial fibrillation, as well as clear guidelines from the American College of Cardiology/American Heart Association, only one half of eligible patients receive warfarin therapy (57), an increase from one third in 1993 (58). In the face of similar resistance, acceptance of warfarin in patients with the acute coronary syndrome may take years, by which time less cumbersome, albeit more expensive, alternatives should be available. Routine use of clopidogrel for at least 9 months is quickly becoming the standard of care, despite studies that found the practice to be non–cost-effective (59 - 60). Moreover, the 23% reduction in myocardial infarction shown with clopidogrel is considerably smaller than the 44% reduction shown with warfarin (56). Although newer oral anticoagulants that do not require monitoring may someday replace warfarin, the benefits of warfarin therapy are available to patients today.