Abstract
Anticoagulation with warfarin in Asian patients with atrial fibrillation (AF) often has been decreased as an international normalized ratio (INR) of prothrombin time 1.6-2.6 due to fear of bleeding, although universal criteria recommend an INR of 2.0-3.0. In this randomized, open-label trial, low-intensity anticoagulation (INR 1.6-2.6) was compared with standard-intensity anticoagulation (INR 2.0-3.0) with warfarin. A total 616 patients with AF and at least 1 risk factor for stroke were randomized to low-intensity anticoagulation (n = 308) and standard-intensity anticoagulation (n = 308) groups. The intention-to-treat analysis was performed to determine differences. The baseline characteristics of the two groups were comparable. New-onset stroke occurred in 2 patients (0.44% per year) in the low-intensity group and 5 patients (1.05% per year) in the standard-intensity group (HR 0.42, 95% CI 0.08-2.15). Major bleeding occurred in 4 patients (0.89% per year) in the low-intensity group and 5 patients (1.06% per year) in the standard-intensity group (HR 0.84, 95% CI 0.22-3.11). The rate of the net clinical outcome (composite of stroke, systemic embolism, major bleeding, and death) was 1.33% per year in the low-intensity group compared with 2.12% per year in the standard-intensity group (HR 0.63, 95% CI 0.23-1.72). In Asian patients with AF, clinical outcomes were not different between low-intensity and standard-intensity anticoagulation with warfarin.
Introduction
Stroke prevention with oral anticoagulants (OACs) constitutes for the management of atrial fibrillation (AF).1–3 Although direct oral anticoagulants (DOACs) are preferred to vitamin K antagonists (eg, warfarin) in patients with AF and high stroke risk, warfarin still constitutes a kind of OACs because of its easy availability and cost-effectiveness. 4 With the use of warfarin, an optimal titration of anticoagulation intensity is critical to achieve a balance between the prevention of thromboembolism and minimization of bleeding complications. Most guidelines suggest an international normalized ratio (INR) 2.0 to 3.0 as the optimal anticoagulation intensity for warfarin.1,5 However, these recommendations were mostly based on Western populations with minor consideration of Asian populations. In addition, these recommendations have not been proven prospectively in Asian populations.
The current Japanese pharmacotherapy guidelines for non-valvular AF recommend an INR of 1.6 to 2.6 as the optimal anticoagulation intensity with warfarin regardless of age, based on the Japanese Nonvavular Atrial Fibrillation-embolism Secondary Prevention Cooperative Study and J-RHYTHM registry.2,6–9 Other retrospective or registry studies in Asia also suggest that lower intensity anticoagulation with warfarin might be more suitable than standard intensity anticoagulation.10–12 A recent Korean retrospective analysis for comparing between low-intensity and standard-intensity anticoagulation with warfarin demonstrated similar clinical outcomes. 13
Nevertheless, it is hard to conclude that low intensity anticoagulation with warfarin is better than or similar to standard intensity anticoagulation considering the limited data from randomized trials or retrospective studies. Therefore, in this randomized, prospective, open-label trial, low-intensity anticoagulation (INR 1.6 to 2.6) was compared with standard-intensity anticoagulation (INR 2.0 to 3.0) with warfarin in Korean non-valvular AF patients.
Methods
Study Population
The Korean Anticoagulation with Warfarin in Atrial Fibrillation (KAWAF; Clinical Research Information Service number KCT0007994) trial is a multi-center, randomized, open-label trial conducted at 8 tertiary teaching hospitals. The study was supported by a grant from the Korean Society of Cardiology. The Heart Center of Chonnam National University Hospital and Research Institute of Medical Science of Chonnam National University coordinated the trial, managed the database, and analyzed the primary results.
Patients were randomly assigned to receive standard-intensity (INR 2.0 to 3.0) or low-intensity (INR 1.6 to 2.6) anticoagulation with warfarin. Randomization was stratified according to baseline comorbidities. Stratification variables included components of CHA2DS2-VASc score as age, sex, hypertension, diabetes mellitus, heart failure, stroke, and vascular disease. Computerized, on-line system allocated subjects to each group after entering stratification variables by simple randomization within strata. The inclusion criteria were as follows: ≥18 years old, CHA2DS2-VASc score ≥1, and warfarin medication as antithrombotic treatment. The criteria for exclusion included patients with mitral stenosis with more than moderate severity, prosthetic mitral valve replacement or repair, AF from reversible causes, acute cerebral infarction within 7 days, dual anti-platelet therapy with aspirin and clopidogrel, warfarin hypersensitivity, and any increased bleeding diathesis. Patients with any change in the oral anticoagulant class (from warfarin to DOACs, from DOACs to warfarin), consent withdrawal, and follow-up loss were excluded from the primary analysis.
The study was approved by the Ethics Committee of Chonnam National University Hospital, Gwangju, South Korea (CNUH-2017-010). All patients provided written informed consent.
Definition
The primary efficacy outcome was stroke or systemic embolism. Stroke was defined as the sudden onset of a focal neurologic deficit in a location consistent with the territory of a major cerebral artery and categorized as ischemic, hemorrhagic, or transient ischemic attack (TIA). Systemic embolism was defined as the acute vascular occlusion of an extremity or organ, documented through imaging or surgery.
The primary safety outcome was major bleeding, which was defined according to the International Society on Thrombosis and Haemostasis (ISTH) criteria as clinically overt bleeding accompanied by a decrease in the hemoglobin level of 2 g/dL or transfusion of at least 2 units of packed red blood cells, occurring at a critical site, or resulting in death. Minor bleeding was defined as clinically overt bleeding that did not meet the major bleeding criteria. Net clinical outcome was defined as the composite of stroke, systemic embolism, major bleeding, and death.
The degree of anticoagulation was measured using the INR. The intensity of anticoagulation was calculated as the mean INR value and the mean time in the therapeutic range (TTR) using a Rosendaal method of linear interpolation between each pair of measured INR values. 14 The therapeutic range of the INR incorporated into the TTR was 2.0 to 3.0 in the standard-intensity group and 1.6 to 2.6 in low-intensity group. Target INR intervention either in low or standard-intensity utilized standardized warfarin dosing regimen, which modified warfarin dosing algorithm.15,16
Statistical Analysis
The primary hypothesis was that low-intensity anticoagulation would be non-inferior to standard-intensity anticoagulation, which requires 444 patients in each group to provide a power of 80% with a non-inferiority margin of 0.009 at half of the lower limit of the 95% confidence interval in a placebo-controlled trial with a two-sided beta level of 0.2. 17 For continuous variables, differences between groups were evaluated by unpaired t-test or Mann-Whitney rank-sum test. For discrete variables, differences between groups (expressed as counts and percentages) were analyzed by chi-square test or Fisher's exact test as appropriate. Mean INR values were divided into 5 groups (<1.6, 1.6-2.99, 2.0-2.59, 2.6-2.99, ≥3.0). The primary analysis was performed as intention-to-treat analysis with censoring at the time of last follow-up for the end-points. Kaplan-Meier estimation by log-rank test was performed for the assessment of differences in clinical outcomes. Logistic regression was used to analyze hazard ratios (HRs) as estimates for clinical outcomes. All statistical analyses were performed using SPSS 21.0 (Statistical Package for the Social Sciences, SPSS-PC Inc., Chicago, IL). All analyses were 2-tailed, with clinical significance defined as P < .05.
Results
Patients and Follow-Up
A total of 616 patients with AF who had taken warfarin from Jan 2017 to Dec 2019 were randomized. A total of 308 patients were assigned to the standard-intensity anticoagulation group (INR 2.0 to 3.0), and 308 patients were assigned to the low-intensity anticoagulation group (INR 1.6 to 2.6). All enrolled patients were followed for 2 years or until the first occurrence of any study outcome from the date of enrollment. During follow-up, 49 (15.9%) patients changed OAC class, 1 (0.3%) patient withdrew consent, and 19 (6.2%) patients were follow-up loss in the standard intensity group. In the low-intensity group, 51 (16.5%) patients changed OAC class, 1 (0.3%) patient withdrew consent, and 29 (9.4%) patients were follow-up loss.
Baseline Clinical Characteristics
The two groups were well balanced in terms of baseline characteristics (Table 1). The median age was 55 years, 25.8% of the patients were female, and the mean CHA2DS2-VASc score was 1.9. Approximately 66%, 16%, 19%, and 8% of the patients had hypertension, diabetes mellitus, heart failure, and previous stroke or TIA, respectively. A total of 51.6% patients had paroxysmal AF. The initial symptoms at initial enrollment were similar between the two groups. Dyspnea, palpitation, chest pain, and dizziness were observed in 48%, 43%, 24%, and 33% of patients, respectively. The initial blood pressure and heart rate were comparable between the groups. Co-medications were comparable between the groups. Specifically, all types of anti-arrhythmic drugs were comparable between the groups.
Baseline Clinical Characteristics.
†Median (25% to 75% percentiles); comparison made using Mann-Whitney test. ACEI = angiotensin converting enzyme inhibitor; AF = atrial fibrillation; ARB = angiotensin receptor blocker; BP = blood pressure; CABG = coronary artery bypass grafting; DHP CCB = dihydropyridine calcium channel blocker; MI = myocardial infarction; INR = international normalized ratio; PCI = percutaneous coronary intervention; TIA = transient ischemic attack.
Achievement of Anticoagulation Intensity
The median frequency of INR measurement was 12 (10-14) times with an interval of 43.5 days. The median value of the INR was significantly lower in the low-intensity group than in the standard-intensity group (2.1 vs 2.2, P = .001). The target mean INR value was achieved in 67.7% of patients in the standard-intensity group (INR 2.0 to 3.0) whereas 80.8% of patients in the low-intensity group (INR 1.6 to 2.6). In comparison with the standard intensity group, more patients in the low intensity group achieved a mean INR value of 1.6 to 2.0 (21.8% vs 31.3%, P = .008). On the other hand, more patients in the standard intensity group achieved a mean INR value of 2.6 to 3.0 (9.4% vs 4.2%, P = .011). In addition, more patients in the standard intensity group achieved a mean INR value of 2.0 to 3.0 (67.7% vs 53.7%, P < .001) (Table 2).
Achievement of Anticoagulation Intensity.
INR = international normalized ratio; TTR = time in the therapeutic range.
In the low intensity group, the median value of the TTR was significantly higher (56.6% vs 77.4%, P = .001), and more patients achieved a TTR of ≥70% (30.6% vs 60.7%, P < .001) (Table 2).
Clinical Outcomes According to Randomization
Stroke occurred in 2 patients (0.44% per year) in the low-intensity group and 5 patients (1.05% per year) in the standard-intensity group (HR 0.42, 95% confidence interval [CI] 0.08-2.15, log-rank P = .281) (Table 3 and Figure 1A). The incidence rates of ischemic stroke, hemorrhagic stroke, and TIA were not different between the two groups. In the as-treated safety population, major bleeding occurred in 4 patients (0.89% per year) in the low-intensity group and 5 patients (1.06% per year) in the standard-intensity group (HR 0.84, 95% CI 0.22-3.11, log-rank P = .787) (Table 3 and Figure 1B). The causes of major bleeding were comparable between the two groups. In addition, the incidence rates of minor bleeding and death were comparable between the groups. The rate of the net clinical outcome (composite of stroke, systemic embolism, major bleeding, and death) was 1.33% per year in the low-intensity group compared with 2.12% per year in standard-intensity group (HR 0.63, 95% CI 0.23-1.72, log-rank P = .360) (Table 3 and Figure 1C).
Kaplan-Meier estimation for clinical outcomes according to anticoagulation intensity. (A) Cumulative incidence of stroke. (B) Cumulative incidence of major bleeding. (C) Cumulative incidence of the net clinical outcome. Standard intensity anticoagulation targets an international normalized ratio (INR) of 2.0 to 3.0, and low-intensity anticoagulation targets an INR of 1.6 to 2.6. Analysis by intention-to-treat protocol.
Clinical Outcomes According to Anticoagulation Intensity.
TIA = transient ischemic attack.
Clinical Outcomes According to International Normalized Ratio (INR) Groups
The incidence rate of stroke was the highest in the patients with a mean INR of <1.6 and ≥3.0. The incidence rate of major bleeding was the highest in the patients with a mean INR of ≥3.0. The net clinical outcome rate was less developed in patients with a mean INR of 1.6-1.99, 2.0-2.59, and 2.6-2.99 (Supplemental Table 1). When the mean INR was less than 1.6 or more than 3.0, the net clinical outcome rate was significantly higher (Figure 2, log-rank P = .019).
Kaplan-Meier estimation for the net clinical outcome according to international normalized ratio (INR) groups. The net clinical outcome was defined as the composite of new-onset stroke, systemic embolism, major bleeding, and death.
Discussion
The KAWAF trial is the first randomized trial to compare low-intensity anticoagulation with a target INR of 1.6 to 2.6 and standard-intensity anticoagulation with a target INR of 2.0 to 3.0 in non-valvular AF patients with warfarin. The trial demonstrated that clinical outcomes were similar between a target INR of 1.6 to 2.6 and a target INR of 2.0 to 3.0. Most of the previous recommendations were based on registry subgroup analysis of the rate of events according to the INR range.9,18–20 Moreover, randomized trials defined low-intensity anticoagulation as therapy with a target INR of 1.5 to 2.0,6,7,21 which is considerably different from the current guideline recommendations. To align with current guidelines for the warfarin anticoagulation intensity, a target INR of 1.6 to 2.6 instead of 1.5 to 2.0 was selected for the low-intensity group, rather 1.5 to 2.0 at the stage of KAWAF study design. Therefore, the KAWAF trial can effectively address issues of contention regarding low-intensity anticoagulation in Asian AF patients, which has the largest sample size as a randomized trial.
In previous meta-analyses and randomized trials in Asia, low-intensity anticoagulation with a target INR of 1.5 to 2.0 showed similar clinical outcomes to those of standard-intensity anticoagulation with a target INR of 2.0 to 3.0 for primary or secondary stroke prevention.22,23 Meanwhile, an INR in the range of 2.0 to 2.6 consistently showed the lowest incidence rates for stroke or major bleeding.9,19 Therefore, expanding the target INR from 1.5 to 2.0 to 1.6 to 2.6 might be reasonable.
Remnant conflict for the risk/benefit balance occur in with an INR interval of 2.6 to 3.0. In the J-RHYTHM Registry, intracranial hemorrhage rate was substantially increased regardless of age when a mean INR was higher than 2.6. 19 In a Korean retrospective analysis based on the INR range, major bleeding rate was less developed with a mean INR less than 2.6. 13 However, a dichotomous comparison between an INR of 2.0 to 3.0 and an INR of <2.0 or >3.0 showed a similar major bleeding rate. In the present study, the risk of major bleeding was similar between an INR of 1.6 to 1.99 and an INR of 2.6 to 2.99. Also, there is no randomized prospective study to evaluate the risk/benefit for an INR interval of 2.6 to 3.0. Therefore, targeting a target INR interval of 2.6 to 3.0 is worthwhile to assess for risk/benefit as INR interval continuum as standard-intensity INR 2.0 to 3.0 in prospective randomized study design.
The incidence rates of stroke and major bleeding were similar between the low-intensity and standard-intensity group in the present study. Furthermore, the rates of ischemic stroke, hemorrhagic stroke, and minor bleeding were comparable between the two groups. In addition, the net clinical outcome was not different between the two groups. Analysis by INR interval revealed that the stroke rate was the highest when the INR was less than 1.6 and more than 3.0. The major bleeding rate was increased when the INR was more than 3.0. From this result, the following statement might be drawn at least: INR less than 1.6 and more than 3.0 should be avoided to minimize deleterious clinical outcomes.
Higher achievement of target INR range and TTR in the low-intensity group were also important findings in the present study. An INR of 1.6 to 2.6 was achieved in 80.8% of patients in the low-intensity group, whereas an INR of 2.0 to 3.0 was achieved in 67.7% of patients in the standard-intensity group. The median value of the TTR was higher in the low-intensity group. More patients achieved a TTR of ≥70% in the low-intensity group compared with the standard-intensity group (60.7% vs 30.6%, P < .001). The results suggest that low-intensity anticoagulation may make it easier to achieve a high TTR in real clinical practice. Inadequate TTR control is associated with adverse clinical outcomes in AF patients taking warfarin. 24 Patients with a TTR less than 50% were found to be more likely to switch to DOACs. 25 In another study, there was no difference between high TTR warfarin treatment and DOACs in terms of stroke prevention. 26 The major reason why patients are reluctant to switch to DOACs is economic limitations; thus, warfarin still has an advantage in this regard. Our study showed that low-intensity warfarin therapy with a higher achievement of TTR would be a suitable option for patients with difficulty converting to DOACs.
Current guidelines recommend anticoagulation therapy in AF patients with a CHA2DS2-VASc score ≥2 for preventing stroke and systemic embolism, and anticoagulation therapy is not recommended when the CHA2DS2-VASc score is 0.1,5 AF patients with a CHA2DS2-VASc score of 1 may be regarded as in the gray zone for anticoagulation therapy. 27 The mean CHA2DS2-VASc score was 1.9 in the present study. The number of patients with a CHA2DS2-VASc score ≥2 was 285 (46.3%) of the 616 patients in the randomized population. In addition, the transition rate from warfarin to DOACs was 15.7%. Therefore, the results of the present study should not be extrapolated to all non-valvular AF patients, especially patients with a higher thromboembolic risk. Nevertheless, the results of the present study suggest that a low-intensity anticoagulation strategy may be a potential alternative to standard-intensity anticoagulation with comparable efficacy and safety in non-valvular AF patients with an intermediate thromboembolic risk.
Limitations
The KAWAF trial is the first randomized trial with the largest sample size to compare low-intensity and standard-intensity anticoagulation strategy in Asian AF patients. However, the results of the present study should be interpreted cautiously due to certain limitations. First, the number of enrolled patients did not satisfy the superiority or non-inferiority margin between the two groups. Therefore, the results of the present study lacked statistical power to demonstrate non-inferiority, and at least demonstrate no difference in the clinical outcomes between the groups. Second, the mean CHA2DS2-VASc score was lower than that based on the existing anticoagulation criteria, which limits the application of the results to all thromboembolic risk groups. Nevertheless, CHA2DS2-VASc score ≥1 as part of the inclusion criteria still justifies the use of OACs for these patients. Third, the dropout rate was relatively high at 24% with a change in the anticoagulation strategy from warfarin to DOACs being the main cause, which reflect preference for OACs in the current era. Fourth, time interval between each INR measurement as 43.5 days was longer than recommended by most guidelines. It might influence accurate estimation of TTR. Fifth, quality of INR control as the achievement of target TTR was significantly lower in standard-intensity group. Better INR control as more frequent evaluation and meticulous dose adjustment is required to demonstrate clear difference between the 2 groups in further study.
Conclusions
Low-intensity anticoagulation with a target INR of 1.6 to 2.6 might be a potential alternative to standard-intensity anticoagulation with a target INR of 2.0 to 3.0 with warfarin in Korean patients with AF, especially patients with an intermediate thromboembolic risk. However, further randomized trials with sufficient population and good INR control are needed to clearly demonstrate that low-intensity anticoagulation is non-inferior to standard-intensity anticoagulation with warfarin.
Supplemental Material
sj-docx-1-cat-10.1177_10760296231171081 - Supplemental material for Standard-Intensity Versus Low-Intensity Anticoagulation with Warfarin in Asian Patients with Atrial Fibrillation: A Multi-Center, Randomized Controlled Trial
Supplemental material, sj-docx-1-cat-10.1177_10760296231171081 for Standard-Intensity Versus Low-Intensity Anticoagulation with Warfarin in Asian Patients with Atrial Fibrillation: A Multi-Center, Randomized Controlled Trial by Jeong Gwan Cho, Ki Hong Lee, Yoo Ri Kim, Sunah Kim, Jisoo Gwak, Eunbit Cho, Yourim Sin, Seung Yong Shin, Hyung Wook Park, Jum Suk Ko, Nam Ho Kim, Yae Min Park, Jung Myung Lee, Nam Sik Yoon, Sung Soo Kim, Jun Hyung Kim and Dong Min Kim in Clinical and Applied Thrombosis/Hemostasis
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethics Approval
Ethical approval to report this case was obtained from the ethics committee at Chonnam National University Hospital, Gwangju, South Korea (CNUH-2017-010).
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a grant (2021R1F1A1048115) of the National Research Foundation of Korea (NRF) funded by the Korea government (MSIT) and a grant (BCRI22021) of Chonnam National University Hospital Biomedical Research Institute.
Informed Consent
Written informed consent was obtained from the patients for their anonymized information to be published in this article.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
