Title of Trial
Patients with nonvalvular atrial fibrillation at low risk of stroke during treatment with aspirin: Stroke Prevention in Atrial Fibrillation III Study
Date and Citation of Primary Publication(s) or References
The SPAF III Writing Committee for the Stroke Prevention in Atrial Fibrillation Investigators. JAMA. 279(16):1273-7, 1998 Apr 22-29
Purpose of the Trial
The purpose of this study was to test the effectiveness of a method of risk stratification in identifying patients with atrial fibrillation and low stroke rates when treated with aspirin alone.
Study Category
Atrial fibrillation. Prospective multicenter cohort study.
Patient Population
Outpatients with non-valvular atrial fibrillation thought to be low risk for thromboembolism.
Inclusion/Exclusion Criteria
To be eligible for this trial adult subjects had to have AF documented in the prior 6 months, without prosthetic heart valves, mitral stenosis, or other conditions requiring anticoagulation. In addition, they had to have no contraindication for aspirin or any of the following risk factors for thromboembolism:
- Reduced LV function with CHF symptoms in the prior 6 months or fractional shortening of < 25% by m-mode echocardiography
- SBP > 160mm Hg at study entry on two prior separate occasions
- Prior ischemic stroke, TIA or systemic embolism
- Female sex and age greater than 75 years.
Patients younger than 60 years without clinical cardiovascular disease or hypertension, who had a normal echocardiogram ("lone AF"), were excluded. Patients with intermittent AF were eligible for enrollment.
Study Design
Patients were recruited at 20 centers in the US and Canada. All patients were given enteric coated aspirin 325 mg/day. Patients were seen in follow up at 3, 6 and every 6 months thereafter. Telephone interviews were conducted every 3 months. Patients were withdrawn from the study, but followed, if any of the risk factors in the exclusion criteria developed during the course of the study.
Endpoints
Measured endpoints included stroke, systemic emboli and TIA.
Baseline Characteristics
| | All
(n=892) | History of HTN
(n=406) | No History of HTN
(n=486) |
| Male (%) | 78 | 77 | 79 |
| Age (y) | 67 | 67 | 67 |
| Intermittent AF (%) | 28 | 25 | 30 |
| History of HTN (%) | 46 | - | - |
| Mean SBP (mm Hg) | 130 | 135 | 127* |
| Mean DBP (mm Hg) | 78 | 80 | 76* |
| Diabetes (%) | 13 | 19 | 7* |
| Ischemic Heart Disease | 16 | 20 | 13* |
| Fractional Shortening (%) | 37 | 37 | 37 |
*p < 0.1 comparison of patients with history of HTN vs. those without history of HTN Time to Follow-Up
Patients were followed for a mean of 2.0 years. No patients were lost to follow-up.
Results
The incidence of the primary endpoints was 2.2% per year with no change in the annual rate of events during follow up. The rate of ischemic stroke was 2.0% per year. The rate of endpoint occurrence was significantly higher for those with hypertension when compared to those patients without hypertension (3.6% vs 1.1% per year, p < 0.001). The rate of disabling stroke was also higher among those with a history of hypertension (1.4% vs. 0.5% per year, p = 0.05). A history of hypertension was a significant predictor of primary outcomes with a relative risk of 3.3 (95% C.I. 1.7-6.9, p=0.001). The only other predictor of risk was age with a relative risk increase of 1.7 per 10 years (95% C.I 1.1-2.6, p = 0.01).
Sponsor
Grant from the Division of Stroke and Trauma, National Institute of Neurological Disorders and Stroke, Bethesda, MD.
Trial Status: Completed
Review written by: Jonathan Weinstock, MD
