2018 Volume 82 Issue 12 Pages 2983-2991
Background: The combination of oral anticoagulant (OAC) and antiplatelet drug (APD) increases the bleeding risk in atrial fibrillation (AF). Non-vitamin K antagonist OAC (NOAC) have been increasingly used since 2011. We investigated current status, time trends and outcomes of AF patients using combination therapy in 2011–2017.
Methods and Results: The Fushimi AF Registry is a community-based prospective survey of AF patients in Fushimi-ku, Kyoto, Japan. Of 2,378 patients with OAC at enrollment, 521 (22%) received combination therapy, while 1,857 (78%) received OAC alone. When compared with OAC alone, combination therapy patients had more comorbidities, but approximately 30% had no atherosclerotic disease. From 2011 to 2017, the prevalence of combination therapy decreased from 26% to 14%. The prevalence of NOAC increased in those on combination therapy. Off-label under-dosing of NOAC increased year by year, especially in combination therapy. During follow-up, the incidence of major bleeding (hazard ratio [HR], 1.42; 95% CI: 1.03–1.95) and stroke/systemic embolism (HR, 1.48; 95% CI: 1.09–2.00) was higher in the combination therapy than in the OAC alone group.
Conclusions: In Japanese AF patients receiving OAC, the prevalence of combination therapy decreased, with the proportion of NOAC use increasing in 2011–2017. Many patients, however, received off-label NOAC under-dosing, especially in the combination therapy group. Patients with combination therapy had higher incidences of major bleeding as well as stroke/systemic embolism, compared with OAC monotherapy.
Stroke prevention is central to the management of atrial fibrillation (AF), and oral anticoagulants (OAC) are recommended in guidelines to reduce the risk of ischemic stroke and mortality.1 Many AF patients have concomitant atherosclerotic disease such as coronary artery disease (CAD) and peripheral artery disease (PAD).2,3 Therefore, the combination of OAC and antiplatelet drugs (APD) is often prescribed, although this is likely to increase the risk of bleeding.4–6
In the 1 year after percutaneous coronary intervention (PCI) in patients requiring OAC, triple therapy (OAC plus dual APD) was inferior to double therapy (OAC plus single APD) in a randomized trial in patients receiving warfarin.7 In AF patients receiving a combination of non-vitamin K antagonist OAC (NOAC) and single APD, bleeding risk was reduced compared with conventional triple therapy (warfarin plus dual APD).8,9 Beyond 1 year, current European guidelines recommend OAC monotherapy for AF patients with stable vascular disease.10 There is no evidence, however, based on randomized controlled trials regarding OAC monotherapy in those patients.
Vitamin K antagonists (VKA) including warfarin had been the only available OAC before 2011; since then, 4 NOAC have been introduced for stroke prevention in AF patients, and have been used in a growing number of patients. There are limited data, however, regarding NOAC and APD combination therapy in AF patients in contemporary clinical practice in Japan.
The purpose of this study was therefore to investigate the current status, time trends and outcomes of AF patients using combination therapy from 2011 to 2017, using data from the Fushimi AF Registry, a community-based prospective survey of AF patients.
The Fushimi AF Registry, a community-based prospective survey, was designed to enroll all of the AF patients who visited the participating medical institutions in Fushimi-ku, Kyoto, Japan. The detailed study design, patient enrollment, participating institutions, the definition of the measurements, and patient baseline clinical characteristics for the Fushimi AF Registry have been previously described (UMIN Clinical Trials Registry: UMIN000005834).11 The inclusion criterion for the registry is the documentation of AF on 12-lead electrocardiogram or Holter monitoring at any time. There were no exclusion criteria. The participating institutions consisted of 2 cardiovascular centers (National Hospital Organization Kyoto Medical Center and Ijinkai Takeda Hospital), 9 small- and medium-sized hospitals, and 69 primary care clinics. Patient enrollment started in March 2011. All of the participating institutions attempted to enroll all consecutive patients with AF under regular outpatient care or under admission. Clinical patient data were registered in the Internet Database System (https://edmsweb16.eps.co.jp/edmsweb/002001/FAF/top.html) by the doctors in charge at each institution. Data were automatically checked for missing or contradictory entries and values out of the normal range. Additional editing checks were performed by clinical research coordinators at the general office of the registry. The study protocol conformed to the ethics guidelines of the 1975 Declaration of Helsinki, and was approved by the ethics committees of the National Hospital Organization Kyoto Medical Center and Ijinkai Takeda General Hospital.
We defined OAC as warfarin, dabigatran, rivaroxaban, apixaban, and edoxaban. APD included aspirin, ticlopidine, clopidogrel and cilostazol. The primary endpoint in the current analysis was the incidence of major bleeding during the follow-up period. Other clinical endpoints included the incidence of stroke or systemic embolism (SE), all-cause death, cardiac death, myocardial infarction (MI) and composite of cardiac death, stroke and MI during follow-up. Major bleeding was defined according to the International Society on Thrombosis and Haemostasis criteria: fatal bleeding, reduction in hemoglobin (Hb) ≥2 g/dL, transfusion of ≥ units of blood, or symptomatic bleeding in a critical area or organ.12 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 it was confirmed on computed tomography or magnetic resonance imaging. SE was defined as an acute vascular occlusion of an extremity or organ. Transient ischemic attack was defined as a transient episode of neurological dysfunction caused by focal brain, spinal cord, or retinal ischemia, without acute infarction. CHADS2 score, CHA2DS2-VASc score and HAS-BLED score were calculated as previously described.13–15 The “L” in HAS-BLED score (labile international normalized ratio [INR]), however, was not incorporated in this analysis, because we checked INR only at enrollment and could not evaluate the lability of INR.
Dose selection of each NOAC was evaluated based on the manufacturer labeling recommendations in Japan (standard dose: dabigatran, 300 mg/day; rivaroxaban, 15 mg/day; apixaban, 10 mg/day; edoxaban, 60 mg/day; reduced dose: dabigatran, 220 mg/day; rivaroxaban, 10 mg/day; apixaban, 5 mg/day; edoxaban, 30 mg/day). Compliance with on-label-dosing was evaluated according to whether the dose was adjusted in accordance with current package insert labeling in Japan. Dabigatran dose is reduced when the patient has any one of the following: age ≥70 years, creatinine clearance 30–50 mL/min, history of major bleeding, and use of p-glycoprotein inhibitors (verapamil), but dabigatran does not have definite dose reduction criteria. Rivaroxaban, apixaban and edoxaban have dose reduction criteria: low-dose rivaroxaban is indicated in patients with creatinine clearance 15–50 mL/min, and apixaban dose is reduced with any 2 of the following: body weight ≤60 kg, age ≥80 years, and serum creatinine ≥1.5 mg/dL; edoxaban dose is reduced when the patient has any one of the following: body weight ≤60 kg, creatinine clearance 15–50 mL/min, and use of p-glycoprotein inhibitors (verapamil).
Continuous variables are expressed as mean±SD. Categorical variables are presented as numbers and percentages. We compared categorical variables using the chi-squared test when appropriate; otherwise, we used Fisher’s exact test. We compared continuous variables using Student’s t-test or the Wilcoxon rank sum test on the basis of the distribution. Data were analyzed as crude and stratified by prescription of APD at study enrollment. Multivariable logistic regression analysis was used to assess factors associated with concomitant APD use. The Kaplan-Meier method was used to estimate the cumulative incidence of clinical events, and the log-rank test was used to compare survival across the groups. We carried out multivariate analysis using a Cox proportional hazards model. The covariates chosen for stroke/SE were components of the CHA2DS2-VASc score,14 excluding “Sc” (sex category). The covariates chosen for major bleeding were components of the HAS-BLED score15 excluding “L” (labile INR).
Data analysis was performed with JMP version 13 (SAS Institute, Cary, NC, USA). Two-sided P<0.05 was considered statistically significant.
A total of 4,875 patients were enrolled in the Fushimi AF Registry by November 2017. Of 4,760 patients who were enrolled 1 year before, follow-up data (collected every year) were available for 4,325 (follow-up, 90.9%) as of November 2017. Of these 4,325 patients, we excluded 19 patients whose prescription data were unavailable and 1,926 patients who did not receive OAC. Analyses were performed on 2,378 patients with OAC. The median follow-up period was 1,386 days (IQR, 725–2,160 days).
Clinical characteristics at baseline between patients with OAC alone (OAC; n=1,857, 78%) and those with OAC and APD combination therapy (OAC+APD; n=521, 22%) are listed in Table 1. Compared with OAC alone, patients with combination therapy were more often male, were older, had more comorbidities, including heart failure, diabetes mellitus, dyslipidemia and chronic kidney disease. When compared with the OAC alone group, the combination therapy group had more prevalent atherosclerotic disease, such as CAD, PAD or cerebrovascular disease. Of note, 29.9% of AF patients in the combination therapy group did not have a history of atherosclerotic disease, while 27.1% of those in the OAC alone group had a history of atherosclerotic disease. Those with combination therapy had higher mean CHADS2, CHA2DS2-VASc and HAS-BLED scores. When compared with the OAC alone group, the combination therapy group received warfarin more commonly, with 82.3% taking aspirin and 12.5% using multiple APD.
Data given as n (%) or mean±SD. †Atherosclerotic disease includes CAD, PAD and CVD. ‡0, patients without CAD, PAD or CVD; 1, patients with CAD alone, PAD alone or CVD alone; 2, patients with CAD+PAD, CAD+CVD or PAD+CVD; 3, patients with CAD+PAD+CVD. AF, atrial fibrillation; APD, antiplatelet drug; BMI, body mass index; CABG, coronary aorta bypass graft; CAD, coronary artery disease; CKD, chronic kidney disease; CrCl, creatinine clearance; CVD, cerebrovascular disease; DBP, diastolic blood pressure; MI, myocardial infarction; NOAC, non-vitamin K antagonist oral anticoagulant; OAC, oral anticoagulant; PAD, peripheral artery disease; PCI, percutaneous coronary intervention; SBP, systolic blood pressure; SE, systemic embolism; TIA, transient ischemic attack.
Factors associated with combination therapy on multivariable analysis are listed in Table 2. Strong positive effect estimates were observed for PAD, previous PCI, previous coronary artery bypass graft, cerebrovascular disease, CAD and dyslipidemia.
Hb, hemoglobin; INR, international normalized ratio. Other abbreviations as in Table 1.
Time trends in anti-thrombotic therapy since 2011 are shown in Figure 1. The proportion of patients on combination therapy decreased over time: 26% in 2011 to 14% in 2017 (Figure 1A). A similar trend was seen irrespective of the presence of atherosclerotic disease (Figure 1B,C). In 2011, 13% of patients without atherosclerotic disease and 45% of patients with atherosclerotic diseases received combination therapy. In 2017, the corresponding figures were 5% and 34%, respectively.
Change in anti-thrombotic therapy (A) in the entire cohort and (B,C) according to the presence of atherosclerotic disease. APD, anti-platelet drug; OAC, oral anticoagulant.
The proportion of patients with NOAC prescriptions is shown in Figure 2. In 2011, when the first NOAC, dabigatran, was released, 4% of patients with OAC alone and 4% of those with combination therapy received NOAC prescription. The proportion of patients with NOAC prescription increased year by year. In 2017, 61% of patients with OAC alone and 52% of those with combination therapy received NOAC. After 2012, the proportion of patients with NOAC prescription was lower in the combination therapy group compared with the OAC alone group.
Change in the proportion of patients with no-vitamin K antagonist oral anticoagulant (NOAC) prescription from 2011 to 2017. APD, anti-platelet drug; OAC, oral anticoagulant.
Figure 3A shows NOAC dose selection from 2011 to 2017, showing that the proportion of patients with NOAC under-dosing has been increasing year by year (with 17% NOAC under-dosing in 2017). Figure 3B,C shows temporal changes of NOAC dose selection in patients with NOAC and APD and those with NOAC alone. In patients with NOAC and APD, 29% received under-dose prescriptions in 2017. In patients with NOAC alone, the prevalence of under-dosing increased year by year, and 16% received under-dose prescriptions in 2017.
Compliance rate of no-vitamin K antagonist oral anticoagulant (NOAC) dose selection from 2011 to 2017 (A) in those taking NOAC, and (B,C) according to presence of antiplatelet drug (APD) prescription.
Clinical outcomes are shown in Figure 4. Major bleeding occurred more frequently in patients with combination therapy than those with OAC alone (hazard ratio [HR], 1.42; 95% CI: 1.03–1.95), stroke/SE and composite of cardiac death, stroke and MI also occurred more frequently in the combination therapy group than in those with OAC alone (HR, 1.48; 95% CI: 1.09–2.00 and HR, 1.82; 95% CI: 1.41–2.34, respectively). The incidence of all-cause death was not significantly different between the 2 groups, but tended to be higher in patients on combination therapy (HR, 1.23; 95% CI: 0.98–1.53).
Kaplan-Meier curves for the incidences of (A) major bleeding, (B) stroke or systemic embolism (SE), (C) all-cause death and (D) composite of cardiac death, stroke and myocardial infarction (MI) during the follow-up period in those taking oral anticoagulant-anti-platelet drug combination therapy (OAC+APD), vs. OAC alone.
Figure 5 shows Kaplan-Meier curves for incidence of major bleeding and composite of cardiac death, stroke and MI in both treatment groups stratified by presence of atherosclerotic disease. Major bleeding incidence was similar in the 2 groups regardless of presence of atherosclerotic disease (P=0.99 in patients without atherosclerotic disease and P=0.20 in those with atherosclerotic disease; Figure 5A,C). The composite of cardiac death, stroke and MI was not significantly different between the 2 groups, but tended to be higher in the combination therapy group regardless of the presence of atherosclerotic disease (P=0.08 in patients without atherosclerotic disease and P=0.08 in those with atherosclerotic disease; Figure 5B,D). For the subgroups of patients with atherosclerotic disease, CAD, PAD, cerebrovascular disease or no atherosclerotic disease, number of events, event rate and unadjusted HR are listed in Table S1. Further statistical comparisons of these groups were not performed due to the limited number of patients.
Kaplan-Meier curves for the incidences of (A,C) major bleeding and (B,D) composite of cardiac death, stroke and myocardial infarction (MI) according to the (A,B) absence or (C,D) presence of atherosclerotic disease during the follow-up period, in those taking oral anticoagulant–anti-platelet drug combination therapy (OAC+APD), vs. OAC alone.
Table 3 lists multivariate adjusted HR for stroke/SE and major bleeding. On multivariate Cox proportional hazard analysis, APD was not associated with major bleeding (HR, 1.32; 95% CI: 0.95–1.81) or stroke/SE (HR, 1.34; 95% CI: 0.95–1.85).
†Systolic blood pressure >160 mmHg. ‡Dialysis or serum creatinine >2.26 mg/dL. §Aspartate aminotransferase or alanine aminotransferase >3×upper limit of normal. ¶History of major bleeding or hemoglobin <11 g/dL. ††≥7 drinks/week. Abbreviations as in Tables 1,2.
The principal findings are as follows: (1) AF patients taking concomitant APD had a higher prevalence of previous atherosclerotic disease, but 30% of those patients had no history of atherosclerotic disease; (2) the proportion of patients on combination therapy decreased year by year and the proportion of patients with NOAC prescription increased year by year, especially in the OAC alone group; (3) off-label under-dosing of NOAC increased from 2011 to 2017, especially in the combination therapy group; and (4) major bleeding and stroke/SE occurred more frequently in the combination therapy group compared with OAC alone.
Recently, the benefit of aspirin for the primary prevention of cardiovascular events has been questioned given the increased risk of bleeding when balanced against the prevention of cardiovascular events, and also the finding that vascular mortality was not reduced by treatment with aspirin.16 In contrast, treatment with APD for secondary prevention of cardiovascular events improves clinical outcomes.16,17 The benefit of concomitant APD for secondary prevention of cardiovascular events in AF patients receiving OAC, however, is less clear.
In the 1 year after PCI in patients requiring OAC, the WOEST trial (What is the Optimal Antiplatelet and Anticoagulant Therapy in Patients with Oral Anticoagulation and Coronary Stenting) demonstrated a reduction in not only bleeding events but also in all-cause mortality in patients with double therapy (OAC plus clopidogrel), compared with triple therapy (OAC plus aspirin and clopidogrel).7 Beyond 1 year after PCI, current guidelines recommend OAC monotherapy,10 based on a nationwide observational study from Denmark demonstrating that combination therapy in AF patients with stable CAD was not associated with reduction of risk of cardiovascular events as compared with OAC monotherapy, although combination therapy significantly increased risk of bleeding.18 These data may explain why the proportion of patients with combination therapy decreased year by year in the present study; but 30% of patients with atherosclerotic disease still received combination therapy in 2017. Those patients might have been judged to be at higher risk of cardiovascular events than of bleeding. We have no data, however, on the date of PCI or acute coronary syndrome and the type of coronary stent, and it is unknown how many patients should be recommended combination therapy.
In subgroup analyses of the clinical trials, concomitant APD use with NOAC increased the risk of major bleeding,19–22 although meta-analysis of those clinical trials showed that NOAC, compared with VKA, were both safer and more effective in AF patients with concomitant APD.23 In the present study, the proportion of patients with NOAC prescription has increased year by year in the combination therapy group and in those on OAC alone. The proportion of patients with NOAC use, however, was significantly lower in the combination therapy group. Warfarin could be prescribed at a low dose intentionally for low INR in the therapeutic range. That might have been the reason why the proportion of patients with NOAC prescription was lower in the combination therapy group.
Although NOAC prescription increased gradually, the number of patients receiving NOAC with off-label under-dosing also increased. In the USA, approximately 60% of patients receiving NOAC with reduced dose were prescribed an off-label under-dose.24 In Japan, approximately 50% of patients receiving reduced-dose NOAC were prescribed an off-label under-dose.25 In the ORBIT-AF II Registry (Outcomes Registry for Better Informed Treatment of Atrial Fibrillation II), approximately 10% of patients with NOAC received off-label under-dose, which was associated with an increased risk of stroke, cardiovascular hospitalization and mortality.26
In the present study, 20–30% of patients with combination therapy received off-label under-dose NOAC, and under-dosing was more prevalent in the combination therapy group. In the SAKURA AF Registry, prescription of APD was not associated with off-label under-dose NOAC.27 The present patients were older, and had lower body weight and higher prevalence of creatinine clearance <50 mL/min. Therefore, physicians might have reduced the NOAC dose as a precaution against bleeding events, especially in AF patients with combination therapy. In the PIONEER AF-PCI trial (Open-Label, Randomized, Controlled, Multicenter Study Exploring Two Treatment Strategies of Rivaroxaban and a Dose-Adjusted Oral Vitamin K Antagonist Treatment Strategy in Subjects with Atrial Fibrillation who Undergo Percutaneous Coronary Intervention), anti-thrombotic regimen including reduced-dose NOAC was associated with a reduction of bleeding events and no increase in thrombotic events.8 Appropriate NOAC dose for AF patients with concomitant APD use, however, has not been fully determined.
It is widely known that adding APD in AF patients receiving OAC increases the risk of bleeding.4–6 As expected, in the present study, the combination therapy group had a significantly higher incidence of major bleeding compared with OAC alone. On multivariate Cox proportional hazard analysis, however, the addition of APD was not associated with the incidence of major bleeding. Of the components of HAS-BLED score, “Bleeding”, “Elderly” and “Alcohol” were associated with major bleeding. Baseline characteristics rather than the addition of APD might have influenced the incidence of major bleeding.
In this study, patients with OAC-APD combination therapy had significantly higher incidences of stroke/SE. In a subanalysis of the J-RHYTHM Registry (the Japanese Rhythm Management Trial for Atrial Fibrillation), there was no significant difference in the incidence of thromboembolism between OAC-aspirin combination therapy and OAC monotherapy.28 This inconsistency may be due to the difference in the patient background between the 2 cohorts, but the present study was consistent with J-RHYTHM substudy in that combination therapy was not effective in patients with AF and atherosclerotic disease.
Even in patients with atherosclerotic disease, the incidence of the composite of cardiac death, stroke and MI was not significantly different, but tended to be higher in patients with combination therapy than in those with OAC alone. Although several studies demonstrated that APD for secondary prevention of cardiovascular events improved clinical outcomes,16,17 the additive efficacy of APD was not demonstrated in AF patients receiving OAC. This was consistent with the subanalysis of the J-RHYTHM Registry.28 In AF patients with atherosclerotic disease, the significance of combination therapy should be reconsidered, and it should be carefully performed, taking bleeding risk into consideration. The comparison of these outcomes, however, is limited by the differences in patient characteristics between the combination therapy and OAC monotherapy groups, meaning that we are unable to discuss the superiority of combination therapy and OAC monotherapy.
There were several limitations in this study. First, this was a prospective observational study. As aforementioned, given that the patient background was significantly different between the 2 groups, we assessed only associations, not causality. Second, we did not know when OAC or APD were discontinued because we collected prescription data every year. Third, we excluded patients without OAC at enrollment; and patients who were started on OAC or combination therapy during the follow-up period were not evaluated. Fourth, in patients receiving warfarin, we collected PT-INR data only at study enrollment, and we do not have data on time in therapeutic range or quality of warfarin control. Fifth, we do not have data on changes in renal function. Although creatinine clearance was necessary for evaluation of compliance with NOAC dose selection, this was calculated only at enrollment. Sixth, we do not have data on the date of PCI in patients with history of previous PCI at enrollment, and therefore it was unknown whether 1 year had elapsed after PCI in each patient.
In Japanese AF patients receiving OAC, the use of combination therapy decreased, with the proportion of patients with NOAC use increasing between 2011 and 2017. Many patients, however, received off-label under-dosing of NOAC, especially in the combination therapy group. Patients with combination therapy had a higher incidence of major bleeding as well as of stroke or SE than those with OAC monotherapy.
We sincerely appreciate the efforts of the clinical research coordinators (T. Shinagawa, M. Mitamura, M. Fukahori, M. Kimura, M. Fukuyama, C. Kamata). The Fushimi AF Registry was supported by research funding from Boehringer Ingelheim, Bayer Healthcare, Pfizer, Bristol-Myers Squibb, Astellas Pharma, AstraZeneca, Daiichi-Sankyo, Novartis Pharma, MSD, Sanofi-Aventis and Takeda Pharmaceutical. This research is partially supported by the Practical Research Project for Life-Style related Diseases including Cardiovascular Diseases and Diabetes Mellitus from Japan Agency for Medical Research and Development, AMED (18ek0210082 h0002, 18ek0210056 h0003).
M. Akao received lecture fees from Pfizer, Bristol-Myers Squibb, Boehringer Ingelheim, Bayer Healthcare and Daiichi Sankyo. G.Y.H.L. is a consultant for Bayer/Janssen, BMS/Pfizer, Biotronik, Medtronic, Boehringer Ingelheim, Novartis, Verseon and Daiichi Sankyo, and a speaker for Bayer, BMS/Pfizer, Medtronic, Boehringer Ingelheim, and Daiichi Sankyo. No fees are directly received personally. The other authors declare no conflicts of interest.
Supplementary File 1
Table S1. Clinical events in patients with atherosclerotic disease
Appendix S1. Institutions Participating in the Registry
Please find supplementary file(s);