Nonvalvular atrial fibrillation is well known as an independent risk factor for stroke, and millions of patients with this cardiac dysrhythmia are also at risk for cardiogenic embolism. Antithrombotics are the mainstay in stroke prevention among patients with atrial fibrillation. Hart and Halperin searched the MEDLINE database for recent advances in stroke prevention in patients with nonvalvular atrial fibrillation.

Dose-adjusted warfarin reduces the risk for stroke in patients with atrial fibrillation by about 60 percent compared with placebo. This risk reduction is similar for primary and secondary prevention and for disabling and nondisabling strokes. Aspirin reduces the risk for nondisabling stroke by about 20 percent compared with placebo, and warfarin reduces the risk by about 40 percent compared with aspirin. Warfarin protection is maximized at an International Normalized Ratio (INR) between 2.0 and 3.0. An INR between 1.6 and 2.5 is also associated with substantial embolism protection—about 90 percent of that achieved with the higher-intensity levels.

The target intensity of anticoagulation is a balance between prevention of stroke and an increased risk of major bleeding. Anticoagulation increases the frequency and severity of major intracranial and extracranial hemorrhage, predicted mostly by the intensity of anticoagulation and the patient age. Therefore, in patients younger than 75 years and for secondary prevention, the target INR is 2.5; for primary prevention in older patients, the target value is 2.0.

Aspirin may offer better protection against noncardioembolic stroke, but these strokes are generally less disabling in patients with atrial fibrillation. Therefore, warfarin therapy may be most beneficial in patients with atrial fibrillation who are at higher risk for disabling cardioembolic strokes and less beneficial in patients at lower risk. Combining low-dose warfarin (INR of less than 1.5) with aspirin confers little protection compared with aspirin alone. Adding aspirin to higher-intensity warfarin therapy may increase the risk for intracranial hemorrhage in elderly patients and is less useful than increasing anticoagulation intensity in patients who sustain a cardioembolic event while receiving low-intensity anticoagulation.

Stroke risk stratification studies have identified hypertension, advanced age, left ventricular dysfunction and previous stroke or transient ischemic attack as risk factors for cardioembolic embolism. Patients with these conditions are most helped by anticoagulation rather than antiplatelet therapy. Echocardiography, best done using a transesophageal technique (TEE), can locate cardiac thrombi, especially in the left atrial appendage, which are associated with increased stasis. It is unclear, however, whether observations from TEE add to the predictive value of the clinical risk factors for stroke. TEE may help distinguish cardioembolic from noncardioembolic mechanisms. Dose-adjusted anticoagulation is recommended in all cases involving secondary prevention. Risk factors for noncardioembolic embolism in patients with atrial fibrillation include aortic arch plaque, disturbances of hemostasis and postmenopausal estrogen replacement.

The authors conclude that dose-adjusted anticoagulation is highly effective in preventing stroke in patients with atrial fibrillation. Aspirin is modestly effective in reducing nondisabling, noncardioembolic strokes. Warfarin is more effective than aspirin alone, and low-intensity warfarin alone or in combination with aspirin offers little protection. All patients with atrial fibrillation should be evaluated for stroke risk factors because those at high or moderate risk clearly benefit from anticoagulation. Further study is needed to determine whether sustained control of hypertension reduces the risk for stroke. Other factors that need investigation include the impact of maintaining sinus rhythm on the risk for stroke, the role of cognition and quality of life on risk and the development of anticoagulants that are more easily administered than warfarin.

editor's note: In a related article, Okura and associates identified other sources of embolism unrelated to left atrial stasis that cause ischemic stroke in patients with atrial fibrillation. Left atrial thrombus, closely related to blood stasis and hypercoagulability in the left atrium, is the most common source of embolism. About one third of patients with nonvalvular atrial fibrillation who had had an ischemic stroke did not demonstrate left atrium blood stasis. Most had other sources of embolism not related to blood stasis, such as patent foramen ovale, atrial septal aneurysm and aortic atherosclerotic plaque. Atrial fibrillation may be a coincidental finding in these patient subsets. The authors conclude that sources of embolism other than blood stasis should also be sought in patients with nonvalvular atrial fibrillation, especially in the absence of left atrium spontaneous echo contrast. Therapeutic options for these other embolic sources need to be clarified.—r.s.