Atrial Fibrillation and Anticoagulation

Patients with Nonvalvular Atrial Fibrillation at Low Risk of Stoke During Treatment with Aspirin: Stroke Prevention in Atrial Fibrillation III Study SPAF III Investigators JAMA 279:1273-7, 4/22/98 OF N=892 low risk patients (absence of following risk factors: recent CHF or LVEF < 25%, previous thromboembolism, systolic BP > 160 mmHg, or female sex at age older than 75) given aspirin 325 mg qd and followed for about 2 years. The rate of primary events (ischemic stroke and systemic embolism) was 2.2% per year (ischemic stroke 2.0% per year, disabling ischemic stroke 0.8% per year). Those with HTN had higher rate of primary events (3.6% per year), while those without HTN had only 1.1% events per year. The rate of major bleeding was 0.5% per year. Comments: There does seem to be a place for aspirin in treating those at low risk of complications from AF.

Adjusted-Dose Warfarin Versus Low-Intensity, Fixed-Dose Warfarin Plus Aspirin in High-Risk Patients with Atrial Fibrillation: Stroke Prevention in AF III Randomized Clinical Trial Lancet 348:633-8, 9/7/96 (JW 16:159, 10/15/96) N=1044 with AF at high risk of stroke (women > age 75, or patients with systolic HTN, CHF or previous thromboembolism) randomized to INR of 1.2-1.5 and then fixed at that dose with the addition of aspirin 325 mg qd vs warfarin adjusted to INR of 2-3. It didn't work. More strokes (7.9 % vs 1.9% after 1 year) with no difference in bleeds.

An Analysis of the Lowest Effective Intensity of Prophylactic Anticoagulation for Patients with Nonrheumatic Atrial Fibrillation Hylek et al NEJM 335:540-6, 8/22/96 OF N=74 patients with nonrheuamtic AF who presented with CVA while taking warfarin compared with 222 controls (also with nonrheumatic AF on warfarin). They compared the INR at the time of stroke to the INR among the controls closest to the date of the CVA in the case. The risk of stroke rose sharply at INR < 2.0; at an INR of 1.7, the adjusted OR for stroke was 2.0; at an INR of 1.5, the OR was 3.3; at an INR of 1.3, the OR was 6.0. Other risk factors for stroke included prior stroke (OR 10.4), DM (OR 2.9), HRN (OR 2.5), and current smoking (OR 5.7). The authors recommend maintaining INR between 2.0 and 3.0.

Cost-Effectiveness of Warfarin and Aspirin for Prophylaxis of Stroke in Patients with Nonvalvular Atrial Fibrillation Gage et al JAMA 274:1839-45, 12/20/95 OF Decision and cost-effectiveness analyses using probabilities from published controlled trials. For patients with nonvalvular atrial fibrillation (NVAF) and additional risk factors for stroke (previous stroke or TIA, diabetes, hypertension, or heart disease, including CHF, angina, or prior MI), warfarin therapy led to greater quality-adjusted survival and to cost savings. For those with NVAF and one additional risk factor, warfarin therapy cost $8000 per quality-adjusted life-year saved; for 65-year-old patients with NVAF alone, warfarin cost about $370,000 per quality-adjusted life-year saved as compared with aspirin therapy. However, for a 75-year-old with NVAF alone, warfarin cost $110,000 per quality-adjusted life-year saved. For patients not given warfarin, aspirin was preferred to no therapy.

Optimal Oral Anticoagulant Therapy in Patients with Nonrheumatic Atrial Fibrillation and Recent Cerebral Ischemia European AF Trial Study Group NEJM 333:5-10, 7/6/95 OF/E N=214 patients in trial who were randomized to oral anticoagulant therapy. The optimal INR was found to lie between 2.0 and 3.9. No treatment effect was apparent with anticoagulation below INR of 2.0 (at INR < 2.0, there were 18 ischemic events per 100 person-years, compared with 2-4 per 100 person-years at INR of 2.0-3.9); the rate of thromboembolic events was lowest at INR of 2.0-3.9, with most major bleeding episodes occuring at INR >= 5.0. The authors suggest the target INR be set at 3.0. Note that most patients in this trial received acenocoumarol or phenprocoumon which may be associated with a higher risk of bleeding than warfarin

Risk Factors for Stroke and Efficacy of Antithrombotic Therapy in Atrial Fibrillation: Analysis of Pooled Data from Five Randomized Controlled Trials AF Investigators Arch Int Med 154:1449-57, 7/11/94 (JW 14:28, 8/15/94) Known risk factors for stroke in AF include DM, HTN, increasing age, and prior TIA or CVA. Among patients with AF not receiving antithrombotic therapy, the annual risk of stroke was 1% for those under age 65 with no risk factors vs 8.1% for those over age 75 with at least 1 other risk factor. Warfarin dramatically reduced the risk of stroke, by 84% in women and by 60% in men. The risk reduction with aspirin was 36%; no benefit was found among patients without a history of HTN. The annual rate of major bleeding was 1.3% with warfarin vs 1.0% with aspirin and controls.

Warfarin vs Aspirin for Prevention of Thromboembolism in Atrial Fibrillation: Stroke Prevention in Atrial Fibrillation Trial II Study Lancet 343:687-91, 3/19/94 (JW 13:65, 5/1/94) N=1100 randomized to 325 mg/d aspirin vs warfarin to INR of 2.0-4.5. In those < 75, the annual rate of ischemic stroke or systemic embolism was 1.9% for aspirin vs 1.3% for warfarin; among older patients, the rate was 4.8% with aspirin vs 3.6% with warfarin (neither difference was statistically significant). Event rates were significantly higher in patients with HTN, previous thromboembolism, or heart failure. In older patients, the risk of major hemorrhage was signficantly higher with warfarin (4.2% vs 1.6%). In JW 14:8, 7/1/94, they point out that the annual risk of stroke in those with lone AF of 2.5%; risk rises to 7.2% in patients with HTN, recent heart failure, or previous thromboembolism; if 2 of these risk factors is present, the risk rises to 17.6%/year. Conclude: In those with higher risk of stroke, anticoagulation with warfarin seems warranted if not contraindications exist; in those at lower risk (younger, no other risk factors), aspirin 325 mg per day seems to be the best alternative. For those at high risk of bleeding with warfarin, low-dose aspirin would be a reasonable alternative.

Warfarin in the Prevention of Stroke Associated with Nonrheumatic Atrial Fibrillation Ezekowitz NEJM 327:1406-12, 11/12/92 OF (JW 10:86, 12/1/92) N=8000 men, 7% of whom met inclusion criteria (40% were excluded secondary to contraindications to anticoagulation). The 538 men were followed for 1.7 years on warfarin (to PT of 1.2-1.5 time control) or placebo. Among those with no prior history of stroke, annual rates of cerebral infarction were 0.9% with warfarin vs 4.3% with placebo (among those with a prior history of stroke, the rates were 6.1% with warfarin vs 9.3% with placebo). Warfarin benefitted all age groups, including those > 70. Warfarin treated subjects had no increased risk of cerebral hemorrhage. NNT = 30.

Predictors of Thromboembolism in Atrial Fibrillation: Clinical Features of Patients at Risk, and Echocardiographic Features of Patients at Risk Stroke Prevention in Atrial Fibrillation Investigators AIM 116:1-12, 1/1/92 (JW 9:18, 2/1/92) N=568 with nonrheumatic AF who were treated with placebo in the SPAF study. In multivariate analysis of 21 factors, only recent CHF, HTN, and previous arterial thromboembolism were independently associated with stroke. Patients with none of these risk factors had a 2.5% yearly rate of thromboembolism (those with 2 or 3 of these risks, had a rate of 17.6%/year). On analysis of 14 echocardiographic variables, global LV dysfunction and LAE emerged as significant risk factors for stroke. In the 26% of patients with neither the clinical nor the echocardiographic risks, the yearly rate of thromboembolism was 1%. Comments: This group is probably better treated with aspirin rather than Coumadin.

Stroke Prevention in Atrial Fibrillation Study: Final Results Stroke Prevention in AF Investigators Circulation 84:527-39, 8/91 (JW 8:35, 9/1/91) N=627 with nonrheumatic AF randomized to warfarin, ASA (325 mg/d) or placebo. An additional 703 who were not candidates for warfarin were randomized to aspirin or placebo. During the average follow-up of 1.3 years, 6.3% on placebo had cerebrovascular or embolic event; taking aspirin reduced this by 42%; taking warfarin (adjusted to PT 1.3-1.8 x) reduced the risk of events by 67%. The risk of bleeding complications were the same in all groups. Note that aspirin was not beneficial in the Danish AFASAK study, but the different results here may be due to dose or patient selection.

Placebo-Controlled, Randomized Trial of Warfarin and Aspirin for Prevention of Thromboembolic Complications in Chronic Atrial Fibrillation: The Copenhagen AFASAK Study Lancet 1:175, 1989 NOF N=1007 outpatients aged 38-91 with chronic nonrheumatic atrial fibrillation. One-third randomized to warfarin in a nonblind fashion (anticoagulation maintained at 4.2-2.8 INR), while the other two-thirds were blindly randomized to aspirin (75 mg qd) vs placebo. F/U for 2 yrs. TIA, fatal or nonfatal stroke, or emboli to viscera or extremities occurred in 5 (4 cerebral infarcts, 1 fatal intracerebral bleed) on warfarin, 20 on aspirin and 21 on placebo (p < 0.05). Deaths: 3/12/15 (p < 0.02). Of the four cerebral infarcts in the coumadin group, 2 occurred when treatment had been discontinued, and one occurred after randomization by prior to therapy. The annual incidence of thromboembolic events was 2% in the warfarin group vs 5.5% in the other groups. 21 warfarin-treated patients had nonfatal bleeding complications (9/21 lead to the discovery of malignant or inflammatory diseases in the GU or GI tracts). There were 2 bleeding episodes in the aspirin group and none in the placebo group. Comments: I have seen only the abstract of this article, but it is quite impressive. Colin (FPN 4:36, 5/1/89) abstracts the editorial by Ronthal who reminds us of the criteria for those with atrial fibrillation who have already had a stroke: 1) no anticoagulation for nonembolic cerebral infarction, 2) immediate anticoagulation for small embolic cerebral infarction, and 3) delayed (5-7 d) anticoagulation for patients with large embolic infarcts. This study also serves to remind us of the risks of even temporary withdrawal of coumadin in some patients (2/5 events in the coumadin group occurred after therapy had been stopped for evaulation of hematuria and peroperatively).