The major complication of atrial fibrillation (AF) is ischemic stroke, leading to high morbidity and mortality rates in AF patients worldwide. Clinical trials carried out in western countries have demonstrated an effectiveness of warfarin treatment in reducing the risk of stroke in patients with nonvalvular AF.1–5 The current strategy of preventing stroke in AF patients is use of anticoagulation therapy. The use of warfarin for stroke prevention among patients with AF in China is very low. It has been reported that in China only 1.7% of AF patients were treated with warfarin therapy and 37.9% were treated with aspirin, while others were not treated with either warfarin or aspirin.6 Trials of stroke prevention in the Chinese non-valvular atrial fibrillation (NVAF) patients are lacking though the prevalence of AF in China is 0.61%.7 The general lacking of a standard anticoagulation clinical practice, the low patient compliance rate, and the concerns of cost are well recognized plausible causes of the under utilization of warfarin in this patient population.
In this prospective, multi-center, randomized clinical trial on patients with AF and increased risk of stroke, we aimed to compare three treatment strategies: aspirin (200 mg/d) vs. lower dose warfarin (international normalized ratio (INR) 1.6–2.0) vs. standard dose warfarin (INR 2.1–2.5). Our central hypothesis is that warfarin is more effective than aspirin in stroke prevention in Chinese patients, as has been shown in multiple trials conducted in the US and Europe.
Between November 2002 and December 2004, 786 patients with age between 50 to 80 years were enrolled into the study in 75 institutions in China. All patients had NVAF lasting ≥1 month that was confirmed and documented from their medical history, electrocardiogram (ECG) and/or Holter recordings. Each patient also met the following inclusion criteria: (1) at least one of the following conditions: age ≥60 years; well controlled mild to moderate hypertension (systolic blood pressure (SBP) <180 mmHg and diastolic blood pressure (DBP) <110 mmHg); transient ischemic attack (TIA), ischemic stroke or systemic embolism after 6 months; left ventricular dysfunction (NYHA II-III); and/or diabetes mellitus. (2) were able to complete the entire study period and cooperate with the follow-up. (3) were not presently participating in any other clinical trials.
Patients were excluded if any of the following conditions applied: (1) presently taking warfarin or aspirin (≥200 mg/d) for any reason. (2) cardiovascular factors including cardioversion was planned within 3 months of enrollment; cardiac valvular disease documented by echocardiography; severe left ventricular dysfunction (NYHA IV); myocardial infarction within 6 months; coronary artery bypass graft (CABG) surgery within 6 months; percutaneous coronary intervention (PCI) within 6 months; unstable angina pectoris (new onset of angina pectoris or symptom aggravated within 1 month); thrombus in the left sided heart chamber documented by echocardiography; uncontrolled severe hypertension (SBP ≥180 mmHg or DBP ≥110 mmHg); or Wolff-Parkinson-White syndrome. (3) non-cardiac factors including allergic to warfarin or aspirin; severe lung disease; TIA, ischemic stroke or systemic embolism within 6 months; history of a hemorrhagic stroke (including an intracranial hemorrhage and subarachnoid hemorrhage); requirement for treatment with other nonsteroidal anti-inflammatory drugs (NSIADs) due to non-cardiac diseases.
Recruitment and randomization
The study was approved by the Institutional Review Board of each participating center. After giving a signed informed consent, patients who met the inclusion criteria were enrolled and randomly allocated to one of three study groups according to a stratified block randomization: the standard-intensity warfarin anticoagulation group (INR 2.1–2.5), low-intensity warfarin anticoagulation group (INR 1.6–2.0) and aspirin group (200 mg/d). Randomization was blocked according to group.
Anticoagulation therapy and follow-up
In the warfarin groups, an initial dose of 1–3 mg/d of of warfarin was prescribed after the baseline INR values were measured. Then the INR values were measured every 1–2 days after the initial dose based on which the next dose was adjusted. The frequency of the INR measurements was reduced to once a week when a stable target value was achieved and was further reduced to once a month following the first month. The INR measurement could be performed at any time according to the clinical necessity during follow-up. In the aspirin group, a fixed dose of 200 mg/d of aspirin was used. All patients were evaluated by physicians at 1, 3, 6, 9, 12, 15, 18, 21 and 24 months after randomization to obtain a patient questionnaire, physical examination and related laboratory tests.
Assessment of end points
The primary endpoint was defined as a thromboembolic event including ischemic stroke, TIA or systemic embolism. An ischemic stroke was defined as rapidly developed clinical signs of focal (or global) disturbance of cerebral function lasting more than 24 hours. A TIA was defined as focal neurological symptoms lasting less than 24 hours. Systemic embolism was clinically defined as abrupt vascular insufficiency associated with clinical or Doppler ultrasound evidence of arterial occlusion in the absence of other likely mechanisms (e.g., atherosclerosis, instrumentation). All ischemic strokes were confirmed and documented by cranial computed tomography (CT) or magnetic resonance imaging (MRI) while a TIA event was only confirmed by focal deficits lasting less than 24 hours. Medical records from all potential events were further reviewed by a 5-physician clinical outcomes committee.
The secondary endpoints were all-cause death and hemorrhagic complications. All-cause mortality was ascertained from each hospital database and household committee. All hemorrhagic events were recorded and reviewed by the committee. A severe hemorrhagic event was defined as intracranial bleeding, fatal bleeding, or bleeding leading to the transfusion of four or more units of blood. Other bleeding events were defined as mild hemorrhages.
Estimation of required sample size was calculated using one-sided alpha error of 5% and 90% power, and based on the assumption that the warfarin anticoagulation treatment was superior to aspirin in preventing thromboembolic events; assuming a 5.5% annual risk of thromboembolic event in patients using aspirin and a 2.0% annual risk of thromboembolic event in patients using warfarin (AFASAK1 study).
All analyses were based on intention-to-treat and were completed using SPSS 11.0 software (SPSS Inc., USA). Continuous variables were expressed as mean ± standard deviation (SD) and analyzed with one way analysis of variance (ANOVA). Categorical variables were analyzed with the chi-square test. The outcomes in the three treatment groups were displayed with Kaplan-Meier curves and compared with use of the log-rank test. A 2-sided P < 0.05 was considered statistically significant.
A total of 786 patients who met the criteria were enrolled, 261 patients allocated to standard warfarin group, 265 patients allocated to low warfarin group and 260 patients allocated to aspirin group. After randomization, 96 patients withdrew the study because they did not want to do any follow-up, 22 in standard warfarin group, 15 in low warfarin group and 59 in aspirin group. The patients who were allocated to aspirin group did not need to do blood test, so more patients withdrew the study than those in other two groups. The remaining 690 patients were analyzed, 239 in standard warfarin group, 250 in low warfarin group and 201 in aspirin group. Baseline patient characteristics are summarized in Table 1. There were no significant differences between the three groups, except that a history of a systemic embolism was more common in patients randomized to warfarin (P=0.026). The mean follow-up period was 15.0 months, varying from 1.0–24.0 months. The follow-up periods of the 3 groups were 15.0, 16.5 and 15.0 months respectively, with no significant difference (P=0.982).
Warfarin anticoagulation treatment
There were 489 patients received a warfarin anticoagulation treatment. During the follow-up period, 78% of the INR values were in the range of 1.6–2.5, 8% were above 2.5 and 14% were below 1.6 according to the INR measurements. In the standard-intensity warfarin anticoagulation group, 51.2% were within the target range (2.1–2.5), 10.8% were above the target range and 38.0% were below the target range. In the low-intensity warfarin anticoagulation group, 66.7% were within the target range (1.6–2.0), 15.9% were above the target range, and 17.4% were below the target range (Figure 1). There were 158 patients who discontinued the medication during the study period; 73 (30.5%) in the standard-intensity warfarin anticoagulation group, 60 (24.0%) in the low-intensity warfarin anticoagulation group, and 25 (12.4%) in the aspirin group. There were more patients that discontinued the medication in the warfarin groups than in the aspirin group (P <0.001) and the specific causes are shown in Table 2.
Thromboembolic events occurred in 32 patients, 7 in the standard-intensity warfarin group, 9 in the low-intensity warfarin group and 16 in the aspirin group. The annual event rates of thromboembolism were 2.6%, 3.1% and 6.9% respectively (P=0.027). Compared to aspirin, the rate of thrombolism was reduced by standard intensity warfarin (hazard ratio 0.36, 95% CI 0.15–0.88; P=0.018) and by low intensity warfarin (hazard ration 0.45, 95%CI 0.20–1.01; P=0.044). Standard intensity warfarin did not significantly reduce thrombolism compared to low intensity warfarin (hazard ratio 0.81, 95% CI 0.30–2.18; P=0.676). Among the 32 thromboembolism events, 12 were ischemic strokes, 19 were TIAs and 1 was a systemic embolism (Table 3). The ischemic stroke event rates in the warfarin anticoagulation groups were much lower than that in the aspirin group (P=0.018), while there was no significant difference in TIAs among the three groups. The Kaplan-Meier survival curves are shown in Figure 2.
Hemorrhagic complications and deaths
Severe hemorrhagic events occurred in 15 patients, 7 in the standard-intensity warfarin anticoagulation group, 7 in the low-intensity warfarin anticoagulation group and 1 in the aspirin group. The annual event rates of severe bleeding were 2.6%, 2.4% and 0.4% respectively. Severe hemorrhagic events were more common in two warfarin anticoagulation groups, but the differences did not reach the statistical significance (P=0.101). Mild hemorrhagic events occurred in 40 patients. The mild hemorrhagic event rates in the warfarin anticoagulation groups were significantly higher than that in the aspirin group, with 7.7% in the standard-intensity warfarin anticoagulation group, 5.2% in the low-intensity warfarin anticoagulation group and 1.7% in the aspirin group (P=0.010). The annual event rates of total hemorrhages were 10.2%, 7.6% and 2.2% respectively (P=0.001) in the 3 groups (Table 4). Compared to aspirin, the rate of total hemorrhage was increased by standard intensity warfarin (hazard ratio 4.58, 95% CI 1.77–111.86; P=0.002) and by low intensity warfarin (hazard ratio 3.39, 95% CI 1.28–8.94; P=0.014). Compared to low intensity warfarin, standard intensity warfarin did not significantly increase severe hemorrhage (hazard ratio 1.06, 95% CI 0.37–3.02; P=0.916), mild hemorrhage (hazard ratio 1.49, 95% CI 0.77–2.89; P=0.238) and total hemorrhage (hazard ratio 1.35, 95% CI 0.77–2.36; P=0.289).
Seventeen patients died during the follow-up period, 5 in the standard-intensity warfarin anticoagulation group; 6 in the low-intensity warfarin anticoagulation group and 6 in the aspirin group (Table 5).
In this study on Chinese patients, the warfarin was considerably more effective than aspirin. Rates of severe and mild hemorrhage were higher with warfarin, and there was no significant difference in thromboembolic events or in hemorrhage between standard and low intensity warfarin, but the trial was not designed to show either superiority or inferiority of two warfarin intensities.
Comparison with previous studies
Several large randomized, controlled clinical trials evaluating the effects of warfarin anticoagulation treatment on the prevention of ischemic strokes in patients with AF have been completed in western countries. A meta analysis8 including 5 clinical trials (AFASAK,1 BAATAF,2 SPAF,3 SPINAF,4 and CAFA5) showed that the warfarin anticoagulation treatment could reduce the stroke rate by 68%, mortality rate by 33%, and combined endpoint event (stroke, systemic embolism and death) rate by 48%; aspirin could reduce the stroke rate by 36%; while severe hemorrhagic event rates in the placebo group, aspirin group and warfarin group were 1.0%, 1.0% and 1.3%, respectively, with no statistical difference. The conclusion was that the warfarin anticoagulation treatment could decrease the stroke rate significantly in NVAF patients with an acceptable increase in severe hemorrhagic event rate. Hu9 reported that adjusted dose of warfarin (INR 2.0–3.0) was effective and safe for the moderate to high risk nonvalvular atrial fibrillation. The results of our study were similar to those of the above mentioned trials that the warfarin anticoagulation treatment was superior to aspirin in preventing thromboembolic events, with an increase in hemorrhage similar to previous trials. But the difference between our study and the study reported by Hu9 is INR level. We compared three treatment strategies: aspirin (200 mg/d) vs. lower dose warfarin (INR 1.6–2.0) vs. standard dose warfarin (INR 2.1–2.5). It is the first time to compare two adjusted doses of wafarin in atrial fibrillation patients. We highlight studies that have suggested different warfarin targets may be reasonable for different races and that low INR is preferred to reduce bleeding events. Our study is limited by a large number of patients that discontinued in the warfarin groups. The combination of the study size and length of follow-up does not permit definite conclusion about the relative efficacy of the two wafarin groups. But the study does confirm previous work that adjusted warfarin therapy is more efficacious than aspirin for prevention of embolic events in NVAF. The 100 mg of aspirin was usually used in our clinical practice, but 200 mg of aspirin was selected in this study. We figured that increased dose of aspirin may be more efficacious. In our study, thromboembolic events occurred in 32 patients, 7 (2.6%) in the standard-intensity warfarin anticoagulation group, 9 (3.1%) in the low-intensity anticoagulation group and 16 (6.9%) in the aspirin group (P=0.027). The thromboembolism rates in both the standard-intensity warfarin anticoagulation group and low-intensity warfarin anticoagulation group were much lower than that in the aspirin group (P=0.018 and P=0.044). The ischemic stroke event rates in the warfarin anticoagulation groups were much lower than that in the aspirin group (P=0.018), but there was no significant difference in the TIAs among the 3 groups, which implied that the main effect of aspirin was to prevent cerebral arteriosclerosis embolisms. The results of the Warfarin-Aspirin Recurrent Stroke Study (WARSS)10 also support our conclusions.
The WARSS compared the effects of the secondary prevention of strokes caused by noncardiac thromboembolism with warfarin and with aspirin; it also showed that aspirin mainly prevented strokes caused by noncardiac thromboembolism. Severe hemorrhagic events occurred in 15 patients in our study, 7 in the standard-intensity warfarin anticoagulation group, 7 in the low-intensity warfarin anticoagulation group and 1 in the aspirin group (P >0.05). The warfarin anticoagulation treatment was safe and effective, and the effect was superior to that of aspirin.
The INR values in the clinical trials carried out in the western countries were mostly between 2.0 and 3.5. However, a Japanese clinical trial6 on the secondary prevention of strokes in patients with NVAF showed that the low-intensity anticoagulation treatment (INR 1.5–2.1) was safer than the conventional-intensity anticoagulation treatment (INR 2.2–3.5) in old patients (mean age (66.7±6.5) years). Another Japanese study11 showed that the ischemic events increased when the INR values were ≤1.59 and the severe hemorrhagic events increased when the INR values were ≥2.60, especially in elderly patients. We observed there was no statistical difference in the thromboembolic events between the standard-intensity warfarin group (2.6%) and low-intensity warfarin group (3.1%). However the trial was not powered to detect such a difference. During the entire treatment period, the patients' compliance with the INR measurements and medication in the low-intensity warfarin group was superior to that in the standard-intensity group. In the standard-intensity warfarin group, 51.2% of the INR values were in the target range, which was lower than that in the low-intensity warfarin group (66.7%). As for the all-cause discontinuation of medications, 73 (30.5%) occurred in the standard-intensity warfarin group, and 60 (24.0%) occurred in the low-intensity warfarin group.
The stroke prevalence in the AF patients was 24.8% from a case-control study carried out in 18 hospitals in China.12 Ma et al13 showed that the annual rate of ischemic strokes in the NVAF patients was 5.3% in a study carried out in Beijing, and was similar to that in the western countries. The main problem with anticoagulation treatment of AF in China was that the strength of the prevention of thromboembolism was far from adequate, because there was no optimal intensity of the anticoagulation treatment (optimal INR range) for the Chinese people, thus clinicians tended to worry about the hemorrhagic complications. Many physicians used low dose aspirin for the AF patients with high risk factors, instead of the warfarin anticoagulation treatment guided by INR measurements, and thus the thromboembolic events were not be prevented effectively.
Randomized trials have demonstrated a clear benefit of anticoagulation treatment in patients with atrial fibrillation at risk of stroke, and guidelines recommend that those patients should receive anticoagulation treatment,15,16 but the proportion of eligible patients who are treated with warfarin remains low, especially in China. The present study showed severe hemorrhagic events were more common in two warfarin anticoagulation groups and there was a significant increase in the number of minor hemorrhages. This increase may have important implications for patients' perceptions of benefit from warfarin treatment, which would affect compliance. The other obstacles to implement effective warfarin therapy is cost and inconvenience, as patient commitment to monitoring INR values requires significant time and money.
There are several limitations in this study. First, a lot of patients in the standard warfarin arm below 2.0 (and some in the other arm above 2.0) and more than 25% of patients stopped their warfarin completely so that there was little difference between the two warfarin groups. Another weakness is that the TIA was included in the primary outcome, though it is more usual to only include stroke and embolism. In our study, the number of event of stroke and embolism was very low (2, 3 and 8 respectively), given the sample size it was not powered to detect any statistically significant difference.
In Chinese patients with NVAF, the warfarin therapy (INR 1.6–2.5) for the prevention of thromboembolic events was superior to aspirin. No significant difference was observed between the low intensity anticoagulation treatment (INR 1.6–2.0) and the standard intensity anticoagulation treatment (INR 2.1–2.5). Large clinical trial is warranted to investigate the optimal intensity of warfarin therapy in Chinese patient population with NVAF.
1. Petersen P, Boysen G, Godtfredsen J, Andersen ED, Andersen B. Placebo-controlled, randomized trial of warfarin and aspirin for prevention of thromboembolic complications in chronic atrial fibrillation. The Copenhagen AFASAK study. Lancet 1989; 1: 175-179.
2. The Boston Area Anticoagulation Trial for Atrial Fibrillation Investigators. The effect of low-dose warfarin on the risk of stroke in patients with nonrheumatic atrial fibrillation. N Engl J Med 1990; 323: 1505-1511.
3. Stroke Prevention in Atrial Fibrillation Investigators. Stroke prevention in atrial fibrillation study final results. Circulation 1991; 84: 527-539.
4. Ezekowitz MD, Bridgers SL, James KE, Carliner NH, Colling CL, Gornick CC, et al. Warfarin in the prevention of stroke associated with nonrheumatic atrial fibrillation. N Engl J Med 1992; 327: 1406-1412.
5. Connolly SJ, Laupacis A, Gent M, Roberts RS, Cairns JA, Joyner C. Canadian Atrial Fibrillation Anticoagulation (CAFA) Study. J AM Coll Cardiol 1991; 18: 349-355.
6. Yamaguchi T. For Japanese Nonvalvular Atrial Fibrillation-Embolism Secondary Prevention Cooperative Study Group. Optimal intensity of wafarin therapy for secondary prevention of stroke in patients with nonvalvular atrial fibrillation: a multicenter, prospective, randomized trial. Stroke 2000; 31: 817-821.
7. Zhou ZQ, Hu DY, Chen J, Zhang RH, Li KB, Zhao XL. An epidemiological survey of atrial fibrillation in China. Chin J Intern Med (Chin) 2003; 43: 491-494.
8. Atrial Fibrillation Investigator. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Analysis of pooled data from five randomized controlled trials. Arch Intern Med 1994; 154: 1449-1457.
9. Hu DY. The efficacy and safety of antithrombotic therapy with warfarin in nonrheumatic atrial fibrillation. Chin J Intern Med (Chin) 2006; 45: 800-803.
10. Mohr JP, Thompson JL, Lazar RM, Levin B, Sacco RL, Furie KL, et al. A comparison of warfarin and aspirin for the prevention of recurrent ischemic stroke. Warfarin-Aspirin Recurrent Stroke Study Group. N Engl J Med 2001; 345:1444-1451.
11. Yasaka M, Minematsu K, Yamaguchi T. Optimal intensity of international normalized ratio in warfarin therapy for secondary prevention of stroke in patients with non-valvular atrial fibrillation. Intern Med 2001; 40: 1183-1188.
12. Hu DY, Sun YH, Zhou ZQ, Li KB, Ni YB, Yang G, et al. Rusk facts for stroke in Chinese
with non valvular artrial fibrillation: a case-control study. Chin J Intern Med (Chin) 2003; 42: 157-161.
13. Ma CS, Zhou YJ, Ma Y, Zheng ZG, Yang HB, Wang Y, et al. A follow-up study of incidence of stroke and risk factors in patients with nonvalvular atrial fibrillation in Beijing area. Chin J Cardiol (Chin) 2002; 30: 165-167.
14. Fuster V, Rydén LE, Cannom DS, Crijns HJ, Curtis AB, Ellenbogen KA, et al. ACC/AHA/ESC 2006 guidelines for the management of patients with atrial fibrillation. Circulation 2006; 114: 700-752.
15. Wann LS, Curtis AB, Ellenbogen KA, Estes NA, Ezekowitz MD, Jackman WM, et al. 2011 ACCF/AHA/HRS focused update on the management of patients with atrial fibrillation (update on dabigatran): a report of the American College of Cardiology Foundation/American Heart Association Task Force on practice guidelines. J Am Coll Cardiol 2011; 57: 1330-1337.