Net clinical benefit of adding aspirin to warfarin in patients with atrial fibrillation: Insights from the J-RHYTHM Registry

Net clinical benefit of adding aspirin to warfarin in patients with atrial fibrillation: Insights from the J-RHYTHM Registry

    Net clinical benefit of adding aspirin to warfarin in patients with atrial fibrillation: Insights from the J-RHYTHM registry Eiichi W...

313KB Sizes 0 Downloads 25 Views

    Net clinical benefit of adding aspirin to warfarin in patients with atrial fibrillation: Insights from the J-RHYTHM registry Eiichi Watanabe, Mayumi Yamamoto, Itsuo Kodama, Hiroshi Inoue, Hirotsugu Atarashi, Ken Okumura, Takeshi Yamashita, Gregory Y.H. Lip, Eitaro Kodani, Yuji Okuyama, Akiko Chishaki, Ken Kiyono, Hideki Origasa PII: DOI: Reference:

S0167-5273(16)30419-3 doi: 10.1016/j.ijcard.2016.03.008 IJCA 22137

To appear in:

International Journal of Cardiology

Received date: Revised date: Accepted date:

20 October 2015 22 February 2016 12 March 2016

Please cite this article as: Watanabe Eiichi, Yamamoto Mayumi, Kodama Itsuo, Inoue Hiroshi, Atarashi Hirotsugu, Okumura Ken, Yamashita Takeshi, Lip Gregory Y.H., Kodani Eitaro, Okuyama Yuji, Chishaki Akiko, Kiyono Ken, Origasa Hideki, Net clinical benefit of adding aspirin to warfarin in patients with atrial fibrillation: Insights from the J-RHYTHM registry, International Journal of Cardiology (2016), doi: 10.1016/j.ijcard.2016.03.008

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

IP

T

R1 Net clinical benefit of adding aspirin to warfarin in patients with atrial fibrillation: Insights from the J-RHYTHM Registry Eiichi Watanabe MD1, Mayumi Yamamoto MD1, Itsuo Kodama MD2, Hiroshi Inoue MD3,

SC R

Hirotsugu Atarashi MD4, Ken Okumura MD5, Takeshi Yamashita MD6, Gregory Y.H. Lip MD7, Eitaro Kodani MD4, Yuji Okuyama MD8, Akiko Chishaki MD9, Ken Kiyono PhD10 and Hideki Origasa PhD11

NU

on behalf of the J-RHYTHM Registry Investigators. (1) Department of Cardiology, Fujita Health University School of Medicine, Toyoake, Japan

MA

(2) Nagoya University, Nagoya, Japan

(3) Second Department of Internal Medicine, Toyama University Hospital, Toyama, Japan (4) Department of Cardiology, Nippon Medical School, Tama-Nagayama Hospital, Tokyo, Japan

D

(5) Department of Cardiology, Hirosaki University Graduate School of Medicine, Hirosaki,

TE

Japan

(6) Department of Cardiovascular Medicine, The Cardiovascular Institute. England, UK.

CE P

(7) University of Birmingham Institute of Cardiovascular Sciences, City Hospital, Birmingham, (8) Department of Advanced Cardiovascular Therapeutics, Osaka University Graduate School of Medicine, Osaka, Japan. Japan.

AC

(9) Department of Health Sciences, Faculty of Medical Sciences, Kyushu University, Fukuoka, (10) Division of Bioengineering, Graduate School of Engineering Science, Osaka University, Toyonaka, Japan

(11) Division of Biostatistics and Clinical Epidemiology, University of Toyama, Toyama, Japan All authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.

Address for Correspondence Eiichi Watanabe MD Department of Cardiology Fujita Health University School of Medicine 1-98 Dengakugakubo, Kutsukake-cho Toyoake, Aichi 470-1192, JAPAN e-mail: [email protected] TEL: 81-562-93-2312, FAX: 81-562-93-2315 R1_1

ACCEPTED MANUSCRIPT

Acknowledgement of grant support This study was planned by the Japanese Society of Electrocardiology and supported by a grant

T

from the Japan Heart Foundation, Tokyo, Japan. (UMIN Clinical Trials Registry UMIN

SC R

IP

000001569)

Disclosure of conflicts of interest

Dr. Watanabe received lecture fees from Bayer Healthcare and Boehringer Ingelheim; Dr. Inoue reports receiving research fund from Boehringer Ingelheim and Daiichi-Sankyo, and

NU

remuneration from Daiichi -Sankyo, Bayer Healthcare and Boehringer Ingelheim; Dr. Atarashi, receiving research fund from Daiichi-Sankyo and Boehringer Ingelheim, and lecture fees from

MA

Bayer Healthcare and Boehringer Ingelheim; Dr. Okumura, receiving research fund from Boehringer Ingelheim and Daiichi -Sankyo, and remuneration from Boehringer Ingelheim, Bayer Healthcare, Daiichi-Sankyo, and Pfizer; Dr. Yamashita, receiving research fund from Boehringer Ingelheim and Daiichi -Sankyo, and remuneration from Boehringer Ingelheim,

D

Daiichi -Sankyo, Bayer Healthcare, Pfizer, Bristol-Myers Squibb, and Eisai; Dr. Origasa,

TE

receiving lecture fees from Daiichi -Sankyo; Dr. Okuyama, receiving lecture fees from Boehringer Ingelheim; Dr. Chishaki, receiving lecture fees from Bayer Healthcare. Dr Lip;

CE P

Chairman, Scientific Documents Committee, European Heart Rhythm Association (EHRA). Reviewer for various guidelines and position statements from the ESC, EHRA, NICE etc. Steering Committees/trials: Includes steering committees for various Phase II and III studies, Health Economics & Outcomes Research, etc. Investigator in various clinical trials in

AC

cardiovascular disease, including those on antithrombotic therapies in atrial fibrillation, acute coronary syndrome, lipids, etc. Consultant for Bayer/Jensen J&J, Astellas, Merck, Sanofi, BMS/Pfizer, Biotronik, Medtronic, Portola, Boehringer Ingelheim, Microlife and Daiichi-Sankyo. Speaker for Bayer, BMS/Pfizer, Medtronic, Boehringer Ingelheim, Microlife, Roche and Daiichi-Sankyo. No other potential conflicts of interest relating to this article were reported.

Key words: arrhythmia, warfarin, aspirin, stroke and hemorrhage.

R1_2

ACCEPTED MANUSCRIPT

T

Abstract

IP

Background: Concomitant use of vitamin K antagonist (VKA) and aspirin (ASA) is becoming

SC R

increasingly prevalent among atrial fibrillation (AF) patients. We quantified the net clinical benefit of adding ASA to a VKA using nationwide prospective AF registry data. Methods: We studied 6074 patients (VKA monotherapy: 83% and VKA+ASA: 17%) between

NU

January 2009 and July 2009, and followed them for a mean follow-up period of 2 years. The risk of strokes and bleeding was calculated by the CHA2DS2-VASc and HAS-BLED scores. The

MA

net clinical benefit was defined as the annual rate of ischemic strokes and systemic emboli

by an impact weight of 1.5.

D

prevented by VKAs minus intracranial hemorrhages attributable to the VKA+ASA, multiplied

TE

Results: Patients on a VKA+ASA were older with more medical comorbidities than those on

CE P

VKA alone. Using VKA monotherapy as a reference, higher major bleeding rates and all-cause death were evident in those on VKA+ASA. The net clinical benefit of VKA+ASA for the overall cohort was –0.1%/year (95% confidence interval, –0.74% to 0.46%). There was a trend

AC

toward a negative net clinical benefit from VKA+ASA in patients with a CHA2DS2-VASc>2 and HAS-BLED<2 (–1.17%/year). The VKA+ASA yielded a positive net clinical benefit in patients with a CHA2DS2-VASc>2 and HAS-BLED>3 (1.16%/year). The result patterns were relatively constant using impact weight of 1.0 and 2.0. Conclusions: Our estimates of the net clinical benefit can provide a useful anchoring point for adding ASA to VKA in patients with AF.

R1_3

ACCEPTED MANUSCRIPT

T

Introduction

IP

Atrial fibrillation (AF) is becoming increasingly prevalent among elderly patients, and is a

SC R

significant risk factor for strokes and death [1]. Vitamin K antagonists (VKAs) such as warfarin are efficacious for the prevention of strokes and systemic thromboembolisms in AF patients at moderate to high risk of thromboembolic events [2]. In recent AF trials and registries,

NU

approximately 30% of the patients had a concomitant use of aspirin (ASA) due to coexisting

MA

vascular diseases [3, 4].

ASA is a cornerstone of the secondary prevention in patients with coronary artery disease. Once

D

being started, withdrawal of ASA is known to be associated with a three-fold higher risk of

TE

major adverse events [5]. However, many studies have reported that the addition of an

CE P

antiplatelet therapy to VKA therapy is not associated with a reduction in the risk of thromboembolisms, whereas the risk of bleeding is increased significantly [4, 6-9]. To date, few reports have quantified the net clinical benefit balancing thromboembolisms against intracranial

AC

hemorrhages (ICHs) in patients on VKA+ASA compared to VKA alone and did not differentiate between the patients according to the bleeding risk.

To date, several major guidelines on AF have been updated to address the management of this complex clinical scenario [10-12]. Much of the available data are based on small and retrospective cohorts, or are derived from post-hoc analyses of administrative datasets from white Caucasian cohorts [6, 8, 13, 14]. Thus, more evidence from prospective studies is still required to provide a firm guidance for the clinical decision-making, particularly amongst Asian cohorts.

R1_4

ACCEPTED MANUSCRIPT

The aim of this study was to assess the rates of thromboembolisms, bleeding, and mortality in a

T

nationwide prospective cohort of AF patients treated with VKA+ASA as compared to VKA

IP

monotherapy [15-17]. We then estimated the net clinical benefit of adding ASA to a VKA using

NU

Methods

SC R

a stroke and bleeding risk strata [18].

MA

Patients

The J-RHYTHM Registry is an observational, prospective cohort study that enrolled 7,937

D

Japanese patients with AF. This registry was a nation-wide collaboration of cardiologists and

TE

electrophysiologists. The detailed methods and main results of the J-RHYTHM Registry have been published previously [15-17]. In brief, consecutive AF patients were recruited at the

CE P

outpatient departments of 150 institutions from January 2009 to July 2009 and were followed for 2 years. The AF data capture included their demographics, cardiovascular risk factors,

AC

diagnosis, type of AF, treatment strategy, antithrombotic therapy and monitoring, concomitant medications and doses, and outcomes. Pre-defined outcomes of interest included strokes or systemic embolisms, all-cause mortality, major bleeding, and specific anticoagulation outcomes (e.g. time in the therapeutic range). Antithrombotic drugs and their dosages were determined by the physicians’ discretion. In this post-hoc study, we excluded patients with mitral stenosis or that had undergone a mechanical valve replacement, and those taking antiplatelet therapy other than ASA because the use of antiplatelet drugs other than ASA was infrequent. The study protocol conformed to the 1975 Declaration of Helsinki and was approved by the institutional review board of the participating institutions and all patients gave their written informed consent.

R1_5

ACCEPTED MANUSCRIPT

Endpoints

T

A local investigator ascertained the events and the members of the physician outcomes review

IP

committee reviewed all outcomes. The endpoints were thromboembolisms including ischemic

SC R

strokes and systemic embolisms, major bleeding, and all-cause death. Major bleeding was defined as bleeding requiring hospitalization, and consisted of ICHs, gastrointestinal bleeding, and other causes of bleeding. We defined a validated ischemic stroke as a sudden neurologic

NU

deficit lasting >24 hours, corresponding to a vascular territory in the absence of a primary

MA

hemorrhage, that was not explained by other causes such as trauma or an infection. The diagnosis of ischemic strokes and ICHs was made with computed tomography or magnetic

D

resonance imaging, and adjudicated by an independent neurologist on the review committee.

TE

The gastrointestinal bleeding was defined as a visible bleeding or positive endoscopic evaluation. All analyses of the rates of the endpoints were based on the first event during

CE P

follow-up.

AC

We quantified the net clinical benefit of the VKA+ASA therapy defined as the annual rate of thromboembolic events (TE rate) that included ischemic strokes and systemic emboli prevented by VKAs minus the ICH rate attributable to the VKA+ASA, as follows: Net Clinical Benefit=(TE rate VKA–TE rate VKA+ASA) – weighted factor × (ICH rate VKA+ASA–ICH rate VKA). We assigned our base case a weighted factor of 1.5, and also provided additional sensitivity analyses using weighted factors of 1.0 and 2.0 [18]. In addition, we compared the net clinical outcome with regard to a combined end point consisting of thromboembolisms, major bleeding, and all-cause death for patients on VKAs alone with the VKA+ASA. Further, we quantified the net clinical benefit in a subgroup of 542 patients who had AF and coronary artery disease (including those that underwent either coronary intervention or coronary artery bypass surgery).

R1_6

ACCEPTED MANUSCRIPT

Definitions

T

Components of the CHA2DS2-VASc score [19] were defined by congestive heart failure,

IP

hypertension, age ≥75 (doubled), diabetes, strokes (doubled), vascular disease, an age 65–74.

SC R

and the sex category (female). We modified the “VA” criterion to include coronary artery disease only, because no data were available regarding peripheral artery disease and aortic plaque. Components of the HAS-BLED score [20] were defined by hypertension, abnormal

NU

renal/liver function, strokes, a bleeding history or predisposition, a labile international normalized ratio, elderly (>65 years), and drugs/alcohol concomitantly. We modified the “D”

MA

criterion because we did not collect information on the nonsteroidal anti-inflammatory drugs

TE

Statistical analysis

D

and alcohol abuse.

CE P

The baseline variables for each drug exposure group are presented as the number and frequency or mean ± standard deviation (SD) values, or median, and interquartile range. Differences between the two groups were evaluated using a Student’s t-test or Mann-Whitney test for

AC

continuous variables and the χ2 test or Fisher’s exact probability test for categorical data. We calculated incidence rates of the outcome as the number of events per 1000 person-years of follow-up. The relative risk was assessed for the endpoints. The cumulative probability of each endpoint was estimated as a function of the time using the Kaplan-Meier method and comparisons between the groups were performed using a log-rank test. The 95% confidence interval (CI) of the net clinical benefit was calculated using a bootstrap sample of 1000 replications. The net clinical outcome analysis was performed between the VKA alone and VKA+ASA with the use of a Cox proportional-hazards model with 2 years of follow up. We adjusted for the treatment groups and age (continuous), sex, hypertension, diabetes, and heart

R1_7

ACCEPTED MANUSCRIPT

failure. The hazard ratio (HR) and 95%CI were given. The patients were categorized according

T

to a low to intermediate TE risk (CHA2DS2-VASc <1) and high TE risk (CHA2DS2-VASc >2),

IP

and low to intermediate bleeding risk (HAS-BLED <2) and high bleeding risk (HAS-BLED >3)

SC R

for the net clinical benefit analysis and net clinical outcome analysis. The statistical analyses were performed with JMP 10.0.2 software (SAS Institute, USA) or R-project

MA

NU

(https://www.r-project.org/). A two-tailed p-value of <0.05 was considered significant.

D

Results

TE

The baseline characteristics of the patients dichotomized by the antithrombotic therapy are shown in Table 1. A total of 5,046 patients (83%) were taking VKA monotherapy while 1,025

CE P

patients (17%) were taking VKA+ASA. The patients in the VKA+ASA group were older, more likely to be male, and had higher CHA2DS2-VASc, and HAS-BLED scores. A previous stroke or

AC

transient ischemic attack (TIA) was observed in 864 patients (14.2%). In the VKA+ASA group, the prevalence of coronary artery disease or a stroke/TIA was 30.7% and 23.4%, respectively; 52.0% of the patients receiving a VKA+ASA did not have any obvious atherosclerotic disease. The prevalence of a CHA2DS2-VASc <1, and HAS-BLED <2 was 22.6%, and 68.5%, respectively. The rate of previous bleeding did not significantly differ between the two groups and gastrointestinal bleeding accounted for 52%. The median dose of the VKA was 3.0 mg/day for the VKA monotherapy, and for the combination therapy, the dose of the VKA was 2.5 mg/day and of the ASA 100 mg/day. Patients with VKA alone had a statistically higher median and average INR values.

R1_8

ACCEPTED MANUSCRIPT

Follow-up and outcomes

T

Using the VKA monotherapy as a reference, the relative risk of the study endpoints is

IP

summarized in Table 2. There was no significant difference in the risk of thromboembolisms

SC R

between the two groups. The combination therapy of VKA+ASA significantly increased the risk of major bleeding and gastrointestinal bleeding by 67% and 101%, respectively, when compared to VKA alone, but this was non-significant for an ICH. The VKA+ASA patients had a higher

MA

of the endpoints are shown in Figure 1.

NU

mortality rate by 50% compared to the VKA alone patients. The Kaplan-Meier survival curves

Net clinical benefit

D

The net clinical benefit analyses stratified by the CHA2DS2-VASc score and HAS-BLED score

TE

are summarized in Table 3. The net clinical benefit of VKA+ASA for the overall cohort was

CE P

–0.1%/year (95%CI, –0.74% to 0.46%). There was a trend toward a negative net clinical benefit with a VKA+ASA in patients with a CHA2DS2-VASc >2 and HAS-BLED <2 (–1.17%/year). A VKA+ASA yielded a positive net clinical benefit in patients with a higher thromboembolic risk

AC

and bleeding risk: a CHA2DS2-VASc >2 and HAS-BLED >3 (1.16%/year). We then quantified the net clinical benefit in a subgroup of 542 patients who had AF and coronary artery disease (either coronary intervention or coronary artery bypass surgery) using a weighted factor of 1.5 (Table 4). The net clinical benefit of the VKA+ASA for the overall cohort was –0.57%/year (95%CI, –2.08% to 0.95%). There was a trend toward a negative net clinical benefit with the VKA+ASA in all strata. The patterns of the results were relatively constant when weighted factors of 1.0 and 2.0 were used (Supplementary File).

R1_9

ACCEPTED MANUSCRIPT

Net clinical outcome

T

The annualized rate of the net clinical outcome (thromboembolisms, major bleeding, and

IP

all-cause death) was significantly lower with VKA than with VKA+ASA (2.7%/year with VKA,

SC R

as compared with 3.8%/year with VKA+ASA, p=0.01). The results from the adjusted Cox-proportional hazard regression models stratified by the CHA2DS2-VASc score and HAS-BLED score are shown in Table 5. An adjusted Cox hazard regression revealed that

NU

VKA+ASA use showed a non-significant trend towards a higher risk of net clinical outcome in

MA

the overall cohort (HR 1.15, 95%CI 0.89 to 1.48). The VKA+ASA use had a significant excess risk of the net clinical outcome in patients with a CHA2DS2-VASc >2 and HAS-BLED <2 (HR

D

1.97, 95%CI 1.14 to 3.20). When we analyzed the net clinical outcome in the subgroup of 542

TE

patients who had AF and coronary artery disease (Table 6), an adjusted Cox hazard regression showed a non-significant trend towards a higher risk of the net clinical outcome with

AC

CE P

VKA+ASA use (HR 1.15, 95%CI 0.68 to 1.95).

Discussion

In this J-RHYTHM Registry subanalysis, we found that 52% of the patients receiving VKA+ASA did not have a valid indication for ASA. Second, the VKA+ASA use had no net clinical benefit in the entire cohort. Third, VKA+ASA use had a trend toward higher risk of the net clinical outcome in the overall cohort.

In this study approximately 50% of the patients receiving a VKA+ASA did not have obvious atherosclerotic disease. This was evident in the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT AF) study [4], which is a large US registry of R1_10

ACCEPTED MANUSCRIPT

outpatients in a stable condition with AF, which demonstrated that nearly 40% of the patients

T

who received VKA+ASA therapy had no evidence of coronary artery disease. In our study, for

IP

patients receiving VKA+ASA, the risk of major bleeding increased by 67% as compared to a

SC R

VKA alone, and the level of the risk was similar in direction and magnitude as the recent large registries from Denmark [8, 9] and ORBIT AF [4]. Indeed, appropriate selection of patients for a combination therapy is important and physicians should consider the expected benefits and

NU

risks carefully before prescribing ASA with a VKA, for example, those patients with recent

MA

acute coronary syndrome or those with recent revascularization [6, 21, 22].

As compared with VKA, VKA+ASA use was associated with higher rates of major bleeding,

D

particularly with gastrointestinal bleeding. A recent review has shown that a history of a

TE

previous stroke was more prevalent in the Asian AF population who are more prone to develop

CE P

ICH when treated with a VKA [23]. The absolute rates of an ICH during the VKA use in AF patients in the recent studies ranged from 0.3%/year to 0.6%/year [23], which was comparable to our data of 0.37%/year. In this study there was no excess risk of an ICH by adding ASA to a

AC

VKA. This observation is similar to the meta-analysis by Dentali et al.[21], but counter to the Danish cohort study [6].

Net clinical benefit A risk assessment that incorporates both the risk for thromboembolisms and the risk for an ICH provides a more quantitatively informed basis for the decision on antithrombotic therapy in AF patients [18]. In this study VKA+ASA had no net clinical benefit in the overall cohort. Of those, we found that a VKA+ASA yielded a relatively negative net clinical benefit in patients with a higher thromboembolic risk, but with a low to intermediate bleeding risk because increase in

R1_11

ACCEPTED MANUSCRIPT

ICH outweighed the increase in the thromboembolism in the VKA+ASA patients. We show that

T

VKA+ASA yielded a positive net clinical benefit in patients with a higher thromboembolic risk

IP

and high bleeding risk because their gain in thromboembolism reduction associated with a lower

SC R

ICH in the VKA+ASA patients. A similar net clinical benefit analysis has been reported by Olesen et al., [13] who show a positive net clinical benefit in patients on VKA +ASA with CHA2DS2-VASc >2 and HAS-BLED ≥3 when compared to patients with no antithrombotic

NU

treatment. This result is somewhat unexpected, but the observation by Olesen et al. [13]

MA

suggests that VKA+ASA may perhaps be suitable for patients with a higher risk of a thromboembolism, as well as a higher bleeding risk. Connolly et al. proposed another method to

D

assess the net clinical benefit to consider ischemic events (ischemic strokes or myocardial

TE

infarctions) and hemorrhagic events (hemorrhagic strokes or subdural or extracranial bleeding), weighted by the hazard ratio for death (or death or disability) after an event relative to death (or

CE P

death or disability) following an ischemic stroke [24]. In an analysis of the Atrial Fibrillation Clopidogrel Trial with Irbesartan for Prevention of Vascular Events (ACTIVE) trial data set

AC

using this method, they showed that adding clopidogrel to ASA therapy prevented 0.57 ischemic stroke equivalents per 100 patient-years of treatment suggesting that adding clopidogrel to aspirin therapy resulted in a modest net benefit among the patients with AF for whom warfarin was unsuitable. In the subgroup analysis, a relatively positive net clinical benefit was observed in patients with CHADS2 score of 3 to 6 (1.03%/year, 95%CI –0.64 to 2.22).

Net clinical outcome In this study, an adjusted Cox hazard regression revealed that the VKA+ASA use tended to yield an excess risk of the net clinical outcome in the overall cohort. In addition, the net clinical outcome showed an increased risk of the VKA+ASA in patients with a higher thromboembolic

R1_12

ACCEPTED MANUSCRIPT

risk, but with low to intermediate bleeding risk. This observation is broadly similar to the results

T

from the net clinical benefit analysis. Previous clinical studies that compared the rates of

IP

adverse events between a VKA and ASA in patients with coronary disease (but without AF)

SC R

showed that a VKA to be at least as cardioprotective as ASA [25, 26]. Theoretically, if a patient with coronary disease and AF is treated with a VKA, ASA might not be needed given that the

NU

VKA protects against strokes and lowers the cardiovascular risk.

MA

We demonstrated that the VKA+ASA treatment resulted in a negative net clinical benefit and increased risk of a net clinical outcome compared to VKA alone in the patients with AF and

D

coronary artery disease. In accordance with the European joint consensus document, patients

TE

with AF and stable vascular disease (defined as being free from any acute ischemic events or repeat revascularization during the preceding year) should be managed with oral anticoagulant

CE P

alone [12].

Study limitations

AC

This study is a prospective nationwide cohort study but was not a randomized or blinded study; thus the limitations include patient selection and reporting biases. A randomized control study would be ideal but impossible to perform for ethical reasons, considering the increased risk of bleeding, without any beneficial effects of preventing thromboembolic events by the VKA and ASA treatment, as reported by previous studies [6, 21]. No adjustment was done despite the major differences between the 2 groups noted in Table 1 in almost all variables because of the small number of endpoints. We had to modify the “VA” criterion when calculating the patients’ thromboembolic risk, and this is an important limitation. We included coronary artery disease only, because we did not collect information on peripheral artery disease or aortic plaque. We

R1_13

ACCEPTED MANUSCRIPT

might have underestimated the thromboembolic risk. This study was conducted with patients of

T

Asian origin only, and therefore some differences may be expected in patients of different

IP

origins. An important limitation was that we could not assess the severity of the consequences

SC R

of individual thromboembolisms or ICH events in our study and had to rely on a weight of 1.5 or 2 as was adopted in the previous studies [13, 14]. Recent evidence has quantified the severity of the consequences of ischemic strokes and ICHs. Kamitani et al. [27] reported an

NU

approximately 100% increase in the in-hospital mortality (7.2% vs. 16.3%), and 47% increase in

MA

severe disabilities or death (defined by the modified Rankin Scale 4 to 6) in those with ICHs compared to those with ischemic strokes (55% vs. 37.5%). Iihara et al. [28] have shown that

D

ICH patients had an approximately 100% increase in the in-hospital mortality rates compared to

TE

ischemic strokes (16.8% vs. 7.8%) using data from 53,170 emergency-hospitalized patients. These observations suggest that ICHs have more disastrous effects than ischemic strokes and

CE P

the severity of the consequences of thromboembolisms or ICH events should be assessed to quantify the net clinical benefit in future studies. We should be aware of the limitations of the

AC

HAS-BLED score. There is a considerable overlap between the HAS-BLED and thromboembolic risk scores such as the CHA2DS2-VASc score, since stroke risk is closely related to the bleeding risk in patients with AF. Recent studies reported that HAS-BLED score had significantly better predictive value for major bleeding than CHA2DS2-VASc (or CHADS2) [29-31]. Of note, the HAS-BLED score has been validated in several independent cohorts [29, 32-35] and correlates well with the ICH risk. The HAS-BLED score was used as a measure of categorizing the subgroup of patient at high bleeding risk, so we could calculate the net clinical benefit. In this particular study, the HAS-BLED score differentiated ICH risk in patients treated with VKAs, but not for the VKA+ASA treatment. The latter is very probably due to the small number of ICH patients in our cohort. While we recognize that some results may seem

R1_14

ACCEPTED MANUSCRIPT

T

counter-intuitive, they are hypothesis generating for further studies.

IP

Conclusions

SC R

In conclusion, we provide quantitative assessments of the net clinical benefit of VKA+ASA among Asian AF patients stratified by the thromboembolic risk and bleeding risk strata. The VKA+ASA yielded no net clinical benefit in the overall cohort, but it had a positive net clinical

NU

benefit in patients with a higher risk of thromboembolic events and bleeding. Patients with AF

AC

CE P

TE

D

MA

with stable coronary artery disease should be managed with VKA alone.

R1_15

ACCEPTED MANUSCRIPT

References

IP

T

[1] Lloyd-Jones DM, Wang TJ, Leip EP, Larson MG, Levy D, Vasan RS, et al. Lifetime risk for development of atrial fibrillation: the Framingham Heart Study. Circulation. 2004;110:1042-6.

SC R

[2] Camm AJ, Kirchhof P, Lip GY, Schotten U, Savelieva I, Ernst S, et al. Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of

NU

the European Society of Cardiology (ESC). Eur Heart J. 2010;31:2369-429. [3] Nieuwlaat R, Capucci A, Camm AJ, Olsson SB, Andresen D, Davies DW, et al. Atrial

MA

fibrillation management: a prospective survey in ESC member countries: the Euro Heart Survey on Atrial Fibrillation. Eur Heart J. 2005;26:2422-34.

D

[4] Steinberg BA, Kim S, Piccini JP, Fonarow GC, Lopes RD, Thomas L, et al. Use and

TE

associated risks of concomitant aspirin therapy with oral anticoagulation in patients with atrial

CE P

fibrillation: insights from the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF) Registry. Circulation. 2013;128:721-8. [5] Biondi-Zoccai GG, Lotrionte M, Agostoni P, Abbate A, Fusaro M, Burzotta F, et al. A

AC

systematic review and meta-analysis on the hazards of discontinuing or not adhering to aspirin among 50,279 patients at risk for coronary artery disease. Eur Heart J. 2006;27:2667-74. [6] Hansen ML, Sorensen R, Clausen MT, Fog-Petersen ML, Raunso J, Gadsboll N, et al. Risk of bleeding with single, dual, or triple therapy with warfarin, aspirin, and clopidogrel in patients with atrial fibrillation. Arch Intern Med. 2010;170:1433-41. [7] Doyle BJ, Rihal CS, Gastineau DA, Holmes DR, Jr. Bleeding, blood transfusion, and increased mortality after percutaneous coronary intervention: implications for contemporary practice. J Am Coll Cardiol. 2009;53:2019-27. [8] Lamberts M, Gislason GH, Olesen JB, Kristensen SL, Schjerning Olsen AM, Mikkelsen A, et al. Oral anticoagulation and antiplatelets in atrial fibrillation patients after myocardial R1_16

ACCEPTED MANUSCRIPT

infarction and coronary intervention. J Am Coll Cardiol. 2013;62:981-9. [9] Lamberts M, Gislason GH, Lip GY, Lassen JF, Olesen JB, Mikkelsen AP, et al. Antiplatelet

IP

T

therapy for stable coronary artery disease in atrial fibrillation patients taking an oral anticoagulant: a nationwide cohort study. Circulation. 2014;129:1577-85.

SC R

[10] Verma A, Cairns JA, Mitchell LB, Macle L, Stiell IG, Gladstone D, et al. 2014 focused update of the canadian cardiovascular society guidelines for the management of atrial

NU

fibrillation. Can J Cardiol. 2014;30:1114-30.

[11] January CT, Wann LS, Alpert JS, Calkins H, Cigarroa JE, Cleveland JC, Jr., et al. 2014

MA

AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: executive summary: a report of the American College of Cardiology/American Heart Association Task

D

Force on practice guidelines and the Heart Rhythm Society. Circulation. 2014;130:2071-104.

TE

[12] Heidbuchel H, Verhamme P, Alings M, Antz M, Diener HC, Hacke W, et al. Updated

CE P

European Heart Rhythm Association Practical Guide on the use of non-vitamin K antagonist anticoagulants in patients with non-valvular atrial fibrillation. Europace. 2015;17:1467-507. [13] Olesen JB, Lip GY, Lindhardsen J, Lane DA, Ahlehoff O, Hansen ML, et al. Risks of

AC

thromboembolism and bleeding with thromboprophylaxis in patients with atrial fibrillation: A net clinical benefit analysis using a 'real world' nationwide cohort study. Thromb Haemost. 2011;106:739-49.

[14] Friberg L, Rosenqvist M, Lip GY. Net clinical benefit of warfarin in patients with atrial fibrillation: a report from the Swedish atrial fibrillation cohort study. Circulation. 2012;125:2298-307. [15] Inoue H, Okumura K, Atarashi H, Yamashita T, Origasa H, Kumagai N, et al. Target international normalized ratio values for preventing thromboembolic and hemorrhagic events in Japanese patients with non-valvular atrial fibrillation. Circ J. 2013;77:2264-70. [16] Atarashi H, Inoue H, Okumura K, Yamashita T, Kumagai N, Origasa H, et al. Present status R1_17

ACCEPTED MANUSCRIPT

of anticoagulation treatment in Japanese patients with atrial fibrillation: a report from the J-RHYTHM Registry. Circ J. 2011;75:1328-33.

IP

T

[17] Atarashi H, Inoue H, Okumura K, Yamashita T, Origasa H, Investigators JRR. Investigation of optimal anticoagulation strategy for stroke prevention in Japanese patients with atrial

SC R

fibrillation--the J-RHYTHM Registry study design. J Cardiol. 2011;57:95-9. [18] Singer DE, Chang Y, Fang MC, Borowsky LH, Pomernacki NK, Udaltsova N, et al. The

NU

net clinical benefit of warfarin anticoagulation in atrial fibrillation. Ann Intern Med. 2009;151:297-305.

MA

[19] Lip GY, Nieuwlaat R, Pisters R, Lane DA, Crijns HJ. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based

D

approach: the Euro Heart Survey on atrial fibrillation. Chest. 2010;137:263-72.

TE

[20] Pisters R, Lane DA, Nieuwlaat R, de Vos CB, Crijns HJGM, Lip GYH. A novel

CE P

user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation. Chest. 2010;138:1093-100. [21] Dentali F, Douketis JD, Lim W, Crowther M. Combined aspirin-oral anticoagulant therapy

AC

compared with oral anticoagulant therapy alone among patients at risk for cardiovascular disease: a meta-analysis of randomized trials. Arch Intern Med. 2007;167:117-24. [22] Dewilde WJ, Oirbans T, Verheugt FW, Kelder JC, De Smet BJ, Herrman JP, et al. Use of clopidogrel with or without aspirin in patients taking oral anticoagulant therapy and undergoing percutaneous coronary intervention: an open-label, randomised, controlled trial. Lancet. 2013;381:1107-15. [23] Chiang CE, Wang KL, Lip GY. Stroke prevention in atrial fibrillation: an Asian perspective. Thromb Haemost. 2014;111:789-97. [24] Connolly SJ, Eikelboom JW, Ng J, Hirsh J, Yusuf S, Pogue J, et al. Net clinical benefit of adding clopidogrel to aspirin therapy in patients with atrial fibrillation for whom vitamin K R1_18

ACCEPTED MANUSCRIPT

antagonists are unsuitable. Ann Intern Med. 2011;155:579-86. [25] van Es RF, Jonker JJ, Verheugt FW, Deckers JW, Grobbee DE, Antithrombotics in the

IP

T

Secondary Preventionof Events in Coronary Thrombosis-2 Research G. Aspirin and coumadin after acute coronary syndromes (the ASPECT-2 study): a randomised controlled trial. Lancet.

SC R

2002;360:109-13.

[26] Hurlen M, Abdelnoor M, Smith P, Erikssen J, Arnesen H. Warfarin, aspirin, or both after

NU

myocardial infarction. N Engl J Med. 2002;347:969-74.

[27] Kamitani S, Nishimura K, Nakamura F, Kada A, Nakagawara J, Toyoda K, et al.

MA

Consciousness level and off-hour admission affect discharge outcome of acute stroke patients: a J-ASPECT study. J Am Heart Assoc. 2014;3:e001059.

D

[28] Iihara K, Nishimura K, Kada A, Nakagawara J, Ogasawara K, Ono J, et al. Effects of

TE

comprehensive stroke care capabilities on in-hospital mortality of patients with ischemic and

CE P

hemorrhagic stroke: J-ASPECT study. PLoS One. 2014;9:e96819. [29] Gallego P, Roldan V, Torregrosa JM, Galvez J, Valdes M, Vicente V, et al. Relation of the HAS-BLED bleeding risk score to major bleeding, cardiovascular events, and mortality in

AC

anticoagulated patients with atrial fibrillation. Circ Arrhythm Electrophysiol. 2012;5:312-8. [30] Roldan V, Marin F, Manzano-Fernandez S, Gallego P, Vilchez JA, Valdes M, et al. The HAS-BLED score has better prediction accuracy for major bleeding than CHADS2 or CHA2DS2-VASc scores in anticoagulated patients with atrial fibrillation. J Am Coll Cardiol. 2013;62:2199-204. [31] Apostolakis S, Lane DA, Buller H, Lip GY. Comparison of the CHADS2, CHA2DS2-VASc and HAS-BLED scores for the prediction of clinically relevant bleeding in anticoagulated patients with atrial fibrillation: the AMADEUS trial. Thromb Haemost. 2013;110:1074-9. [32] Lip GY, Frison L, Halperin JL, Lane DA. Comparative validation of a novel risk score for R1_19

ACCEPTED MANUSCRIPT

predicting bleeding risk in anticoagulated patients with atrial fibrillation: the HAS-BLED (Hypertension, Abnormal Renal/Liver Function, Stroke, Bleeding History or Predisposition,

IP

T

Labile INR, Elderly, Drugs/Alcohol Concomitantly) score. J Am Coll Cardiol. 2011;57:173-80. [33] Olesen JB, Lip GY, Hansen PR, Lindhardsen J, Ahlehoff O, Andersson C, et al. Bleeding

SC R

risk in 'real world' patients with atrial fibrillation: comparison of two established bleeding prediction schemes in a nationwide cohort. J Thromb Haemost. 2011;9:1460-7.

NU

[34] Friberg L, Rosenqvist M, Lip GY. Evaluation of risk stratification schemes for ischaemic stroke and bleeding in 182 678 patients with atrial fibrillation: the Swedish Atrial Fibrillation

MA

cohort study. Eur Heart J. 2012;33:1500-10.

[35] Okumura K, Inoue H, Atarashi H, Yamashita T, Tomita H, Origasa H, et al. Validation of

D

CHA(2)DS(2)-VASc and HAS-BLED scores in Japanese patients with nonvalvular atrial

AC

CE P

TE

fibrillation: an analysis of the J-RHYTHM Registry. Circ J. 2014;78:1593-9.

R1_20

ACCEPTED MANUSCRIPT

Figure legends

IP

T

Figure 1. Kaplan-Meier survival curves for the endpoints.

Significant differences in the cumulative rates of major bleeding, GIB, and all-cause death were

SC R

observed. VKA: vitamin K antagonist, ASA: aspirin, ICH: intracranial hemorrhage, GIB:

AC

CE P

TE

D

MA

NU

gastrointestinal bleeding

R1_21

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

R1_22

ACCEPTED MANUSCRIPT

Male, (n) %

VKA+ASA

(n=5046)

(n=1025)

69.5 ± 9.8

71.7 ± 8.5

<0.001

756 (73.8)

<0.001

23.9 ± 3.6

0.01

3510 (69.6) 2

23.6 ± 3.5

SC R

Body mass index (kg/m )

IP

Age, years

VKA

T

Table 1. Baseline characteristics of patients by difference of antithrombotic therapy

P-value

Systolic blood pressure (mmHg)

125.5 ± 16.1

126.5 ± 15.6

0.06

Diastolic blood pressure (mmHg)

73.3 ± 11.2

73.1 ± 10.8

0.60

Type of AF, n (%)

1827 (36.2)

330 (32.2)

Persistent

787 (15.6)

136 (13.3)

Permanent

2432 (48.2)

559 (54.5)

MA

NU

Paroxysmal

Congestive heart failure, n (%)

<0.001

1427 (28.3)

348 (34.0)

<0.001

3002 (59.5)

705 (68.8)

<0.001

1675 (33.2)

419 (40.9)

<0.001

835 (16.5)

283 (27.6)

<0.001

624 (12.4)

240 (23.4)

<0.001

Coronary artery disease, n (%)

227 (4.5)

315 (30.7)

<0.001

COPD, n (%)

95 (1.9)

15 (1.5)

0.35

Malignancy, n (%)

381 (8.0)

77 (7.9)

0.96

CHA2DS2-VASc score (points)

2.6 ± 1.5

3.5 ± 1.6

<0.001

3 [1 – 4]

4 [2 – 5]

<0.001

Hypertension, n (%) Diabetes, n (%)

AC

CE P

TE

Previous stroke or TIA, n (%)

D

Age>75 years

<0.001 0

383 (7.6)

33 (3.2)

1

884 (17.5)

78 (7.6)

>2

3779 (74.9)

914 (89.2)

1.9 ± 0.9

3.2 ± 0.9

<0.001

2 [1 – 2]

3 [3 – 4]

<0.001

HAS-BLED score (points)

<0.001 <2

3920 (77.7)

240 (23.4)

>3

1126 (22.3)

785 (76.6)

Previous bleeding, n (%)

227 (4.8)

a

54 (5.5)

0.28

Intracranial

57 (1.2)

12 (1.2)

0.90

Gastrointestinal

120 (2.5)

26 (2.7)

0.76

Other sites

52 (1.1)

16 (1.6)

0.14

Laboratory data R1_23

ACCEPTED MANUSCRIPT

13.5 ± 1.8

<0.001

Platelet (x10 /uL)

19.6 ± 5.8

19.1± 5.5

0.01

Creatinine (mg/dL)

0.95 ± 0.54

1.01 ± 0.56

0.001

ClCr (mL/min)

69.4 ± 27.3

63.5 ± 25.0

<0.001

PT-INR

1.98±0.54

<0.001

1.92 [1.65 – 2.25]

1.91 [1.63 – 2.24]

<0.001

Antiarrhythmic drug

2556 (53.6)

491 (50.4)

0.10

ACE-I/ARB

2672 (53.0)

613 (59.8)

<0.001

Statin

1090 (21.6)

385 (37.6)

<0.001

SC R

1.99±0.52

T

13.7 ± 1.7

4

IP

Hemoglobin (g/dL)

MA

NU

Medications, n (%)

VKA = vitamin K antagonist, ASA=aspirin, AF = atrial fibrillation, TIA = transient ischemic

D

attack, COPD = chronic obstructive pulmonary disease, CHA2DS2-VASc =acronym of

TE

congestive heart failure, hypertension, age ≥75 (doubled), diabetes, stroke (doubled), vascular disease, age 65–74 and sex category (female)), HAS-BLED = acronym of hypertension,

CE P

abnormal renal/liver function, stroke, bleeding history or predisposition, labile international normalized ratio, elderly, drugs/alcohol concomitantly, ClCr = clearance of creatinine, PT-INR:

AC

international normalized ratio of prothrombin time, ACE-I = angiotensin converting enzyme inhibitor, ARB = angiotensin II type 1 receptor blocker. Data represent number, frequency, median [interquartile range] or means ± SD. a: 2 patients had both intracerebral bleeding and gastrointestinal bleeding.

R1_24

T

ACCEPTED MANUSCRIPT

CR

IP

Table 2. Crude rates of endpoints by difference of antithrombotic therapy VKA

US

(n=5046)

VKA+ASA (n=1025)

(95% CI)

No. of event

Incidence¶

Thromboembolism

75

8.2 (6.4 to 10.2)

13

6.9 (3.7 to 11.8)

0.85 (0.47 to 1.53)

Major bleeding

92

9.9 (8.0 to 12.1)

31

16.5 (11.2 to 23.4)

1.67 (1.11 to 2.50)*

Intracranial hemorrhage

35

3.8 (2.6 to 5.2)

10

5.3 (2.5 to 9.5)

1.41 (0.70 to 2.85)

Gastrointestinal bleeding

27

2.9 (1.9 to 4.2)

11

5.8 (2.9 to 10.4)

2.01 (1.00 to 4.06)*

Other cause of bleeding

30

3.2 (2.1 to 4.6)

10

5.3 (2.5 to 9.5)

1.64 (0.81 to 3.35)

109

11.7 (9.6 to 14.1)

33

17.5 (12.1 to 24.6)

1.50 (1.02 to 2.21)*

MA N

TE D

Incidence rate: 1000-person /years. CI = confidence interval. Other abbreviations are as in Table 1. * p<0.05

AC



CE P

All-cause death

No. of event

Incidence¶

Relative risk

R1_25

ACCEPTED MANUSCRIPT

ICH

No. of events, (Incidence¶ )

No. of events, (Incidence¶ ) VKA

HAS-BLED >3

<2

>3

<2

5

0

0

0

4

(1.9)

(0)

(0)

(0)

44

26

3

10

(8.2)

(11.7)

(9.2)

(6.7)



MA N

<2

HAS-BLED

>3

<2

>3

<2

0

0

0

0.44*

(1.5)

(0)

(0)

(0)

(0.16 to 0.77)

11

20

3

7

–1.17

1.16*

0.01

(9.0)

(9.2)

(4.7)

(–3.13 to 0.60)

(0.17 to 2.13)

(–0.71 to 0.69)

–0.65

1.18*

–0.10

(–1.97 to 0.50)

(0.20 to 2.15)

(–0.74 to 0.46)

TE D

CE P

(2.1)

AC

Overall

VKA+ASA

HAS-BLED

CHA2DS2-VASc

>2

Net clinical benefit

US

VKA+ASA

CR

Thromboembolism

VKA

<1

IP

T

Table 3. Net clinical benefit of VKA vs. VKA+ASA in all patients

>3

NA

Overall

0.43* (0.17 to 0.76)

Incidence rate: 1000-person /years. ICH: intracranial hemorrhage. Values >0 favors treatment. *: p<0.05. The net clinical benefit was calculated using

a weighted factor of 1.5. Abbreviations are as in Table 1. NA: not available.

R1_26

ACCEPTED MANUSCRIPT

ICH

No. of events, (Incidence¶ )

No. of events, (Incidence¶ ) VKA

HAS-BLED <2

>3

<2

>3

<2

0

0

0

0

0

(0)

(0)

(0)

(0)

4

1

1

2

(2.5)

(1.6)

(1.2)

(0.9)



<2

>3

0

0

0

(0)

(0)

(0)

(0)

0

0

2

(0)

(0)

(2.4)

CE P

>3

AC

Overall

HAS-BLED

TE D

HAS-BLED

CHA2DS2-VASc

>2

VKA+ASA

MA N

VKA+ASA

Net clinical benefit

US

Thromboembolism

VKA

<1

CR

IP

T

Table 4. Net clinical benefit of VKA vs. VKA+ASA in patients with coronary artery disease

<2

>3

Overall

NA

NA

NA

3

-1.19

-0.62

-0.57

(1.3)

(-4.45 to 1.52)

(-2.37 to 1.64)

(-2.05 to 0.96)

-1.09

-0.62

-0.57

(-4.32 to 1.52)

(-2.54 to 1.52)

(-2.08 to 0.95)

Incidence rate: 1000-person /years. HR: hazard ratio, CI: confidence interval. ICH: intracranial hemorrhage. The net clinical benefit was calculated

using a weighted factor of 1.5. Other abbreviations are as in Table 1. Values >0 favors treatment. NA: not available.

R1_27

ACCEPTED MANUSCRIPT

VKA+ASA

No. of events, (Incidence¶ )

No. of events, (Incidence¶ )

HAS-BLED

HAS-BLED >3

<2

27

0

1

(10.4)

(0)

151

99

(28.3)

(44.7)



HAS-BLED

1

0.55

(7.0)

(15.0)

(0.03 to 2.57)

18

55

1.97*

0.71

1.15

(55.2)

(36.7)

(1.14 to 3.20)

(0.50 to 1.00)

(0.88 to 1.49)

1.68*

0.73

1.15

(1.01 to 2.67)

(0.51 to 1.11)

(0.89 to 1.48)

TE D CE P

Overall

(HR, 95%CI)

<2

AC

>2

Net clinical outcome

>3

MA N

<2

US

VKA

CHA2DS2-VASc <1

CR

IP

T

Table 5. Net clinical outcome of VKA vs. VKA+ASA in all patients.

>3

NA

Overall

0.84 (0.14 to 2.81)

Incidence rate: 1000-person /years. HR: hazard ratio, CI: confidence interval. Other abbreviations are as in Table 1. *: p<0.05, NA: not available.

R1_28

ACCEPTED MANUSCRIPT

VKA+ASA ¶



No. of events, (Incidence )

HAS-BLED

HAS-BLED >3

<2

0

0

0

(0)

(0)

(0)

14

8

(8.8)

(12.9)



TE D

0

(0)

(HR, 95%CI) HAS-BLED <2

>3

Overall

NA

NA

NA

14

23

1.98

1.22

1.15

(17.1)

(10.1)

(0.88 to 4.45)

(0.36 to 1.85)

(0.67 to 1.94)

1.98

1.22

1.15

(0.87 to 4.48)

(0.36 to1.85)

(0.68 to 1.95)

CE P

Overall

>3

MA N

<2

AC

>2

US

No. of events, (Incidence )

CHA2DS2-VASc <1

Net clinical outcome

CR

VKA

IP

T

Table 6. Net clinical outcome of VKA vs. VKA+ASA in patients with coronary artery disease

Incidence rate: 1000-person /years. HR: hazard ratio, CI: confidence interval. Other abbreviations are as in Table 1. NA: not available.

R1_29