The stroke prevention in atrial fibrillation III study: rationale, design, and patient features

Clinical trials have consistently shown warfarin's high degree of efficacy for stroke prevention in patients with nonvalvular atrial fibrillation (AF). However, aspirin therapy may be sufficient for large subgroups of low-risk AF patients. In addition, many elderly AF patients cannot receive or sustain adjusted-dose warfarin because of bleeding (minor and major), drug interactions, and other disutility. For such patients, a safer and easier to administer anticoagulation regimen is desirable. The Stroke Prevention in Atrial Fibrillation (SPAF) III Study has two components, based on stratification of patients as high-risk or low-risk for thromboembolism. In high-risk patients, adjusted-dose warfarin (INR 2.0 to 3.0) is compared with low-intensity, fixed-dose warfarin (INR 1.2 to 1.5 initial dose adjustment) plus aspirin (325 mg/d) in a randomized trial. Patients categorized as low-risk all receive aspirin 325 mg/d and are followed-up to assess the reliability and durability of the risk stratification scheme. Medications are administered open-label. Primary events (ischemic strokes and systemic emboli) are assessed by a local neurologist and verified by an Events Committee, neither having knowledge of assigned treatment. Transesophageal echocardiography is obtained at entry, when possible. High-risk criteria for the 1,044 patients in the randomized trial (mean age,=72 years) are impaired left ventricular function (45%), systolic blood pressure >160 mm Hg (32%), prior thromboembolism (38%), and female gender over age 75 years (24%). The mean INR is 2.4 for adjusted-dose and 1.3 for fixed-dose warfarin, with mean daily doses of 3.9 mg and 2.1 mg, respectively. Compared with high-risk patients, those categorized as low-risk (n=892) are younger (mean age, 67 years; P<.001), with lower mean systolic blood pressures, (P<.001), and less ischemic heart disease (P<.001), but similar diastolic blood pressures. Given the large variation in stroke rate among AF patients, blanket recommendations for antithrombotic prophylaxis of all AF patients are not optimal. Patient age, gender, left ventricular function, and systolic blood pressure may be useful markers of the inherent thromboembolic risk and may influence selection of antithrombotic therapy to prevent stroke in AF patients.