Stroke Prevention in Patients with Atrial Fibrillation

Source: Hart RG, et al. Ann Intern Med 1999;131:492-501.

In this paper, hart and associates performed a meta-analysis of data from 16 randomized clinical trials of antithrombotic therapy for patients with nonvalvular atrial fibrillation. The 16 trials included a total of 9874 participants, including 2239 patients who were assigned to placebo.

Six trials involving 2900 patients compared adjusted-dose warfarin with placebo or control. The mean age of participants at study entry was 69 years, with 29% of the patients being women. The control stroke rate was 4.6% and 12.3% per year for primary and secondary prevention trials, respectively. Meta-analysis showed that therapy with adjusted-dose warfarin reduced the relative risk of stroke by 62%. The absolute risk reduction for all strokes was 2.7% per year in primary prevention and 8.4% per year in secondary prevention. All-cause mortality was decreased by 26% in patients who received warfarin.

Six trials compared antiplatelet therapy vs. placebo. Approximately 90% of total follow-up exposure during antiplatelet therapy was with aspirin alone. The aspirin dosage ranged from 25 mg twice daily to 1300 mg daily. The mean age of participants was 70 years, with 38% women. The average rate of stroke among participants assigned to placebo was 5.2% per year for primary prevention and 12.9% per year for secondary prevention.

Meta-analysis of all six trials showed that aspirin reduced the incidence of stroke by 22%. The absolute risk reduction was 1.5% per year for primary prevention and 2.5% per year for secondary prevention.

Adjusted-dose warfarin was compared to aspirin in five trials involving 2837 patients. The mean age of participants was 71 years and 38% were women. The results of these comparisons were variable but meta-analysis showed that adjusted-dose warfarin reduced overall relative risk for stroke by 36%. The paper also lists data from trials that compared adjusted-dose warfarin with fixed doses of warfarin and aspirin and a number of other antithrombotic regimens. However, these trials were sufficiently different to prevent meaningful use of meta-analysis to combine results. Hart et al then present a table giving the estimated size of treatment effects according to risk status. The greatest benefit is seen with warfarin compared to aspirin in high-risk patient groups, with a 48% reduction with warfarin seen for secondary prevention and a 24% reduction for primary prevention in these subgroups.

Hart et al conclude that there is conclusive evidence from a large number of randomized trials that antithrombotic agents reduce the risk of stroke in nonvalvular atrial fibrillation.

Table-ACCP Recommendations
Age High Risk Antithrombic
< 65 yrs. No Aspirin
Yes Warfarin
65-75 yrs. No Aspirin or Warfarin
Yes Warfarin
> 75 yrs. All Warfarin


Comment by John P. DiMarco, MD, PhD

This paper summarizes two decades of clinical trials in the field of stroke prevention among patients with atrial fibrillation. By combining data from these trials in a meta-analysis, Hart et al give physicians a better estimate of the magnitude of treatment effect of such therapy.

The most valuable parts of this meta-analysis are the comparisons of warfarin vs. placebo, aspirin vs. placebo, and warfarin vs. aspirin. It is clear that warfarin is superior to aspirin or aspirin plus low, fixed-dose adjusted warfarin in high-risk populations. Here, the magnitude of treatment effect is substantial and the risk of excess bleeding relatively modest. Controversy about the role of warfarin still exists, however, for relatively low-risk patients. Here, there is only a modest benefit achieved with warfarin over aspirin or even over placebo, but the excess risk of hemorrhage is still present. A possible conclusion from these data is that physicians should still exercise judgment when considering warfarin for patients at low risk for stroke. A conservative approach may be appropriate for patients at low risk but physicians should be aggressive in the use of warfarin in patients at high risk. High-risk factors include: older women, patients with prior stroke or transient ischemic attack (TIA), and those with hypertension and diabetes.

It is disappointing that this meta-analysis does not allow us to show any regimen superior to adjusted-dose warfarin. Although a large number of alternative antithrombotic strategies have been tested, none has proved to be superior to warfarin. Anticoagulation with warfarin is still difficult and one hopes that in the future some effective alternative strategy may be identified. Until then, the recommendations of the American College of Chest Physicians (Chest 1998) (See Table) are worth considering.

Dr. DiMarco is Professor of Medicine, Division of Cardiology, University of New Mexico, Albuquerque.