evaluationoftheage,biomarkers,andclinicalhistory ...€¦ · hypertension 8961 (86.6) 4153 (89.5)...

Original Investigation | Cardiology

Evaluation of the Age, Biomarkers, and Clinical History–Bleeding Risk Scorein Patients With Atrial Fibrillation With Combined Aspirin and AnticoagulationTherapy Enrolled in the ARISTOTLE and RE-LY TrialsZiad Hijazi, MD, PhD; Jonas Oldgren, MD, PhD; Johan Lindbäck, MSc; John H. Alexander, MD, MHS; Marco Alings, MD, PhD; Raffaele De Caterina, MD, PhD;John W. Eikelboom, MD; Michael D. Ezekowitz, MBChB, DPhil; Claes Held, MD, PhD; Kurt Huber, MD; Elaine M. Hylek, MD, MPH; Christopher B. Granger, MD;Renato D. Lopes, MD, PhD; Dragos Vinereanu, MD, PhD; Agneta Siegbahn, MD, PhD; Lars Wallentin, MD, PhD

Abstract

IMPORTANCE Most patients with atrial fibrillation (AF) and coronary artery disease have indicationsfor preventing stroke with oral anticoagulation therapy and preventing myocardial infarction andstent thrombosis with platelet inhibition.

OBJECTIVE To evaluate whether the recently developed ABC (age, biomarkers, and clinical history)–bleeding risk score might be useful to identify patients with AF with different risks of bleeding duringconcomitant aspirin and anticoagulation therapy.

DESIGN, SETTING, AND PARTICIPANTS The biomarkers in the ABC-bleeding risk score (growthdifferentiation factor 15, hemoglobin, and troponin) were measured in blood samples collected atrandomization between 2006 and 2010 in the ARISTOTLE (Apixaban for Reduction in Stroke andOther Thromboembolic Events in Atrial Fibrillation) trial and between 2005 and 2009 in the RE-LY(Randomized Evaluation of Long-term Anticoagulation Therapy) trial, both of which weremultinational randomized clinical trials. The trials were reported 2011 and 2009, respectively. A totalof 24 349 patients with AF (14 980 patients from the ARISTOTLE trial and 9369 patients from theRE-LY trial) were analyzed in the present cohort study. The median (interquartile range) length offollow-up was 1.8 (1.3-2.3) years in the ARISTOTLE cohort and 2.0 (1.6-2.3) years in the RE-LY cohort.Data analysis was performed from February 2018 to June 2019.

EXPOSURES Concomitant aspirin treatment during study follow-up.

MAIN OUTCOMES AND MEASURES Time to first occurrence of a major bleeding was determinedaccording to International Society on Thrombosis and Hemostasis definition. Hazard ratios wereestimated with Cox models adjusted for ABC-bleeding risk score and randomized treatment.

RESULTS The median (interquartile range) age was 70 (63-76) years in the ARISTOTLE cohort and72 (67-77) years in the RE-LY cohort (5238 patients [35.6%] in the ARISTOTLE cohort and 3086patients [36.4%] in the RE-LY cohort were women). The total number of patients with a first majorbleeding event was 651 (207 with aspirin and 444 without) in ARISTOTLE and 463 (238 with aspirinand 225 without) in RE-LY. For both cohorts, in those with a low ABC-bleeding risk score, the absolutebleeding rate was low even with concomitant aspirin treatment, whereas in those with a higherABC-bleeding risk score, the rate of bleeding was higher with concomitant aspirin compared with oralanticoagulation alone (ARISTOTLE, hazard ratio, 1.65; 95% CI, 1.40-1.95; P < .001; RE-LY, hazard ratio,1.70; 95% CI, 1.42-2.04; P < .001). Thus, a low annual ABC-bleeding risk (eg, 0.5% without aspirinuse) would with concomitant aspirin result in an annual rate of 0.8%, and a high estimatedABC-bleeding risk (eg, 3.0%) would result in a substantially higher rate of 5.0%.

(continued)

Key PointsQuestion Is the recently developedABC (age, biomarkers, and clinical

history)–bleeding risk score for patients

with atrial fibrillation (AF) useful to

identify patients with different risks of

bleeding during concomitant aspirin and

anticoagulation therapy?

Findings This cohort study examined atotal of 24 349 patients with AF from 2

randomized clinical trials and found that

the ABC-bleeding risk score identified

patients with different risks of bleeding

when combining aspirin and oral

anticoagulation.

Meaning The ABC-bleeding risk scoremay be a useful tool for decision support

concerning intensity and duration of

combination antithrombotic treatment

in patients with AF and coronary

artery disease.

+ Invited Commentary+ Supplemental contentAuthor affiliations and article information arelisted at the end of this article.

Open Access. This is an open access article distributed under the terms of the CC-BY License.

JAMA Network Open. 2020;3(9):e2015943. doi:10.1001/jamanetworkopen.2020.15943 (Reprinted) September 16, 2020 1/13

Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 06/05/2021

https://jama.jamanetwork.com/article.aspx?doi=10.1001/jamanetworkopen.2020.16126&utm_campaign=articlePDF%26utm_medium=articlePDFlink%26utm_source=articlePDF%26utm_content=jamanetworkopen.2020.15943https://jama.jamanetwork.com/article.aspx?doi=10.1001/jamanetworkopen.2020.15943&utm_campaign=articlePDF%26utm_medium=articlePDFlink%26utm_source=articlePDF%26utm_content=jamanetworkopen.2020.15943

Abstract (continued)

CONCLUSIONS AND RELEVANCE These findings suggest that the ABC-bleeding risk scoreidentifies patients with different risks of bleeding when combining aspirin and oral anticoagulation.The ABC-bleeding risk score may, therefore, be a useful tool for decision support concerning intensityand duration of combination antithrombotic treatment in patients with AF and coronaryartery disease.

JAMA Network Open. 2020;3(9):e2015943. doi:10.1001/jamanetworkopen.2020.15943

Introduction

In patients with atrial fibrillation (AF) and coronary artery disease (CAD), there are indications forpreventing stroke with oral anticoagulation therapy and preventing myocardial infarction (MI) andstent thrombosis with platelet inhibition.1,2 However, as the antithrombotic therapy intensifies, therisk of major bleeding increases.3 This is a well-known, and rather frequently encountered, dilemmain clinical practice, because approximately 20% to 30% of patients with AF also have concomitantCAD.4-6 International guidelines1,2 recommend an individual risk-benefit assessment to tailor theantithrombotic treatment and its duration to an optimally balanced risk of ischemic events vs risk ofbleeding. However, current bleeding risk scores provide limited guidance in the setting of AF andconcomitant CAD.5 Recently, the ABC (age, biomarkers, and clinical history)–bleeding risk score wasdeveloped and validated in patients with AF receiving oral anticoagulation therapy for estimation ofmajor bleeding events.7 The biomarker-based ABC-bleeding risk score outperformed other riskscores based on clinical variables, such as the HAS-BLED (hypertension, abnormal renal and liverfunction, stroke, bleeding, labile international normalized ratio, elderly, drugs or alcohol) score.Importantly, the ABC-bleeding risk score was also shown to be well-calibrated in several cohorts ofpatients with AF.7,8 The ABC-bleeding score has, therefore, been mentioned in the recent EuropeanAF and Dual Antiplatelet Therapy guidelines1,2 as a possible tool to estimate risk of major bleeding inpatients with AF. In this cohort study, we evaluated whether the ABC-bleeding risk score might beuseful to identify patients with different risks of bleeding during concomitant antiplatelet (aspirin)and anticoagulation therapy based on 9369 patients from the RE-LY (Randomized Evaluation ofLong-term Anticoagulation Therapy) trial9 and 14 980 patients with AF from the ARISTOTLE(Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation) trial.10 Wehypothesized that the biomarker-based ABC-bleeding risk score can differentiate between thebleeding risk in patients with AF taking oral anticoagulation and concomitant aspirin.

Methods

The ARISTOTLE trial randomized 18 201 patients with AF and an increased risk of stroke to warfarinor apixaban and was reported in 2011.10,11 Biomarker samples were available at baseline for 14 697participants, with a median (interquartile range) length of follow-up of 1.8 (1.3-2.3) years. The RE-LYtrial randomized 18 113 patients with AF to dabigatran or warfarin and was reported in 2009.9,12

Biomarker samples were available at baseline for 8468 participants, with a median (interquartilerange) length of follow-up of 2.0 (1.6-2.3) years. Overall, both the ARISTOTLE and RE-LY biomarkercohorts were representative of each respective study cohort and have been described in detailelsewhere.13,14 Institutional review board or ethics committee approval was obtained at all sites forboth trials, and all patients provided written informed consent, including consent for the biomarkersubstudy program. This study follows the Strengthening the Reporting of Observational Studies inEpidemiology (STROBE) reporting guideline.

JAMA Network Open | Cardiology ABC-Bleeding Risk Score and Antithrombotic Therapy in Patients With AF



https://jama.jamanetwork.com/article.aspx?doi=10.1001/jamanetworkopen.2020.15943&utm_campaign=articlePDF%26utm_medium=articlePDFlink%26utm_source=articlePDF%26utm_content=jamanetworkopen.2020.15943http://www.equator-network.org/reporting-guidelines/strobe/

Outcome AssessmentIn both the ARISTOTLE and the RE-LY trials, the primary safety outcome was major bleeding definedaccording to the International Society on Thrombosis and Hemostasis.9,10 Blinded clinical eventcommittees reviewed and centrally adjudicated the outcome events.

Biochemical MethodsIn both the ARISTOTLE and RE-LY studies, blood samples were collected at randomization into EDTAtubes and immediately centrifuged, frozen in aliquots, and stored at −70 °C until analyzed centrallyat the Uppsala Clinical Research Center laboratory in Uppsala, Sweden. High-sensitive cardiactroponin T and N-terminal prohormone of brain natriuretic peptide (NT-proBNP) were analyzed withhigh-sensitivity immunoassays on the Cobas Analytics e601 (Roche Diagnostics), and growthdifferentiation factor–15 (GDF-15) levels were detected with the Elecsys growth differentiationfactor–15 precommercial assay kit P03 with the same standardization as the recently introducedroutine reagent (Roche Diagnostics). All analyses were performed according to the instructions of themanufacturer and have been described elsewhere.15-17

Statistical AnalysisAll patients with no missing values for the variables included in the ABC-bleeding score (prior strokeor transient ischemic attack, age, high-sensitive cardiac troponin, and N-terminal prohormone ofbrain natriuretic peptide) measured at baseline were included in the analyses. The ABC-bleedingscore was defined in this study as the estimated 1-year major bleeding risk.

The consistency of the ABC-bleeding score with concomitant aspirin use was assessed using aCox regression model. The model included aspirin use and the estimated 1-year ABC-bleeding risk asa restricted cubic spline, with 4 knots placed at the 5th, 35th, 65th, and 95th sample percentiles,and the interaction between aspirin status and the linear part of the ABC-bleeding risk. If the test ofthe interaction was not significant, a model excluding the interaction term was fitted to estimate therelative hazard of bleeding according to aspirin use. The fitted model was also used to estimate theabsolute increase in bleeding for 2 examples of low (0.5%) and high (3.0%) ABC-bleeding riskswithout aspirin. These examples were chosen because there are no established cutoffs for low orhigh annual major bleeding risks in patients with AF; previously, risks below 1% and above 2% havebeen indicated, although not established.18

In the ARISTOTLE trial, aspirin use was recorded continuously during follow-up and wastherefore included as a time-updated covariate in the model. In the RE-LY trial, a patient wasassumed to be taking aspirin if she or he was taking aspirin at any time during follow-up. The modelsdid not include any further adjustment for confounders because the ABC-bleeding risk score alreadyincorporates the main risk factors for major bleeding and further adjustment would not be expectedto improve the model.7

Nevertheless, in a sensitivity analysis of the ARISTOTLE data, a Cox regression model includingpotential baseline and time-updated confounders was fitted. Baseline variables included age, sex,diabetes, hypertension, history of MI, history of peripheral arterial disease, and history of vasculardisease. Postrandomization variables included whether the patient was currently taking the studydrug, MI, angina, percutaneous coronary intervention, coronary stenting, nonmajor bleeding andinteractions between prior MI and nonmajor bleeding, MI and nonmajor bleeding, study drug andangina, study drug and percutaneous coronary intervention, and nonmajor bleeding. The resultswere materially unaltered compared with the unadjusted model; therefore, we only present theresults from the simpler model. Furthermore, the 3-way interaction between the ABC-bleeding riskscore, aspirin use, and study treatment was evaluated by fitting a Cox model including theABC-bleeding risk score, a treatment combination variable (aspirin or no aspirin in combination withwarfarin or apixaban in ARISTOTLE, and warfarin or dabigatran [110 mg or 150 mg] in RE-LY) and theinteraction between the ABC-bleeding risk score and the treatment combination variable.




Results from all models are presented graphically as plots of hazard ratios (HRs) with anarbitrary reference point set to 2.0% 1-year ABC-bleeding risk in the no-aspirin group. TheABC-bleeding and HAS-BLED risk scores were calculated according to previouslydescribed methods.7

All tests were 2-sided, with P < .05 denoting statistical significance. All analyses were doneusing R statistical software version 3.6.1 (R Project for Statistical Computing). Data analysis wasperformed from February 2018 to June 2019.

Results

Baseline CharacteristicsA total of 24 349 patients with AF were analyzed in the present cohort study. The median(interquartile range) age was 70 (63-76) years in the ARISTOTLE cohort and 72 (67-77) years in theRE-LY cohort; 5238 patients (35.6%) in the ARISTOTLE cohort and 3086 patients (36.4%) in theRE-LY cohort were women. The characteristics according to concomitant aspirin use or no aspirin useare shown in Table 1 for the ARISTOTLE and RE-LY cohort. The proportion of patients takingconcomitant aspirin was 31.0% (4638 patients) at day 1 in the ARISTOTLE cohort and approximately20% on any day after day 1.4 In the RE-LY cohort, the proportion of patients taking concomitantaspirin at any time during follow-up was 36.4% (3413 patients). In both cohorts, a higher proportionof patients receiving concomitant aspirin had vascular disease (eg, prior MI, CAD, or peripheral arterydisease). The presence of diabetes was also more common in patients with concomitant aspirin.Patients with prior vitamin K antagonist use were more common in the group withoutconcomitant aspirin.

Risk of Major Bleeding With Concomitant Aspirin Use and ABC-Bleeding Risk ScoreARISTOTLE CohortThe total number of patients with a first major bleeding event was 651 (207 while taking aspirin and444 while not taking aspirin) during 24 903 person-years of follow-up. The annual rate of majorbleeding during follow-up was higher in patients taking aspirin than in those not taking aspirin(4.04% vs 2.24%) (Table 2). The bottom quarter of the patients had an estimated ABC-bleeding riskless than 1.1% annually. In these patients, the absolute bleeding risk was low even with concomitantaspirin treatment (Figure 1A). In those in the top quarter with an ABC-bleeding risk greater than2.5%, the absolute risk of bleeding was higher with concomitant aspirin compared with oralanticoagulation alone (Figure 1A). The P value for the test of the null hypothesis of no multiplicativeinteraction was .07. In a model without the interaction term, adjusted for ABC-bleeding risk score andrandomized treatment, concomitant aspirin treatment increased bleeding significantly (HR, 1.65;95% CI, 1.40-1.95; P < .001). Thus, a low ABC-bleeding risk score without aspirin (eg, 0.5% annually)would with concomitant aspirin result in an annual rate of 0.8%, and a high estimated ABC-bleedingrisk score without aspirin (eg, 3.0% annually) would result in a substantially higher absolute annualrate for major bleeding of 5.0%.

The observed event rate for major bleeding as a function of the ABC-bleeding score by thedifferent combinations of antithrombotic treatment is shown in Figure 2. At low estimated bleedingrisk with the ABC-bleeding score, the model showed good calibration. With higher estimatedbleeding risk, concomitant use of aspirin conferred higher than estimated bleeding rates.

RE-LY CohortThe total number of patients with first major bleeding events was 463 (238 with aspirin and 225without) during 16 212 person-years of follow-up. The annual rate of major bleeding during follow-upwas higher in patients with aspirin than without aspirin (4.18% vs 2.14%) (Table 2). Overall, the resultsin the RE-LY cohort were similar to those of the ARISTOTLE cohort. The bottom quarter of thepatients had an estimated ABC-bleeding risk less than 1.3%, whereas the top quarter had an




ABC-bleeding risk greater than 2.8%. The relative hazard of major bleeding associated with aspirinuse was similar across the full range of the ABC-bleeding risk score (Figure 1B; P for interaction = 0.2).In Cox models adjusted for ABC-bleeding risk score and randomized treatment, concomitant aspirintreatment increased bleeding significantly (HR, 1.70; 95% CI, 1.42-2.04; P < .001). Thus, a lowABC-bleeding risk score without aspirin (eg, 0.5% annually) would with concomitant aspirin result in

Table 1. Baseline Characteristics in the ARISTOTLE Cohort by Aspirin Status on Day 1 (Day of Randomization) and in the RE-LY Cohort by Aspirin StatusDuring Follow-up, Anytime During the Study, Including Biomarkers Used in the ABC-Bleeding Score

Variable

Patients, No. (%)

ARISTOTLE RE-LY

No aspirin (n = 10 342) Aspirin (n = 4638) No aspirin (n = 5956) Aspirin (n = 3413)Randomized treatment: warfarin 5193 (50.2) 2289 (49.4) 1985 (33.3) 1144 (33.5)

Received dabigatran

110 mg NA NA 1972 (33.1) 1137 (33.3)

150 mg NA NA 1999 (33.6) 1132 (33.2)

Age, y, median (IQR) 70.0 (62.0-76.0) 70.0 (63.0-76.0) 72.0 (67.0-77.0) 72.0 (67.0-78.0)

Female 3769 (36.4) 1562 (33.7) 2275 (38.2) 1133 (33.2)

Body mass index, median (IQR)a 28.5 (25.3-32.5) 28.6 (25.3-32.8) 28.0 (25.1-31.6) 27.9 (25.1-31.2)

Missing values, No. 49 22 3 4

Systolic blood pressure, mm Hg, median (IQR) 130.0 (120.0-140.0) 130.0 (120.0-140.0) 130.0 (120.0-142.0) 130.0 (120.0-140.0)


Diabetes 2454 (23.7) 1243 (26.8) 1196 (20.1) 883 (25.9)

Hypertension 8961 (86.6) 4153 (89.5) 4660 (78.2) 2731 (80.0)

Current smoker 870 (8.4) 349 (7.5) 467 (7.8) 254 (7.4)

Missing values, No. 12 2

Alcohol 264 (2.6) 114 (2.5) 866 (14.5) 456 (13.4)

Permanent or persistent atrial fibrillation 8865 (85.7) 3846 (82.9) 4255 (71.5) 2078 (60.9)


Prior stroke or transient ischemic attack 1948 (18.8) 861 (18.6) 1159 (19.5) 664 (19.5)

Prior bleeding 1701 (16.4) 736 (15.9) 678 (11.4) 505 (14.8)

Anemia 665 (6.4) 342 (7.4) 678 (11.4) 505 (14.8)


Heart failure 3115 (30.1) 1536 (33.1) 1685 (28.3) 1026 (30.1)


Prior myocardial infarction 1008 (9.7) 918 (19.8) 747 (12.5) 842 (24.7)


Prior peripheral arterial disease 456 (4.4) 274 (5.9) 186 (3.1) 159 (4.7)


Prior vascular disease 2028 (19.6) 1695 (36.5) 882 (14.8) 937 (27.5)

Warfarin within 7 d of randomization 6455 (62.5) 1582 (34.2) 4397 (73.8) 1609 (47.1)


Estimated glomerular filtration rate, mL/min, median (IQR) 74.5 (57.0-96.0) 73.1 (56.1-93.8) 65.5 (54.5-77.4) 63.7 (52.7-76.4)


Growth differentiation factor 15, ng/L, median (IQR) 1348.0 (965.0-1992.0) 1464.0 (1005.0-2216.0) 1476.0 (1092.0-2124.0) 1592.0 (1142.2-2338.0)


High-sensitive cardiac troponin T, ng/L, median (IQR) 10.7 (7.4-16.3) 11.6 (7.8-17.6) 11.7 (7.5-18.8) 13.1 (8.2-20.6)


Hemoglobin, g/dL, median (IQR) 14.3 (13.2-15.3) 14.2 (13.1-15.3) 14.3 (13.3-15.3) 14.2 (13.1-15.3)


Abbreviations: ABC, age, biomarkers, and clinical history; ARISTOTLE, Apixaban forReduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation; IQR,interquartile range; NA, not applicable; RE-LY, Randomized Evaluation of Long-termAnticoagulation Therapy.

SI conversion factor: To convert hemoglobin to grams per liter, multiply by 10.0.a Body mass index is calculated as weight in kilograms divided by height in

meters squared.




an annual rate of 0.85%, and a high estimated ABC-bleeding risk score without aspirin, (eg, 3.0%annually) would result in a substantially higher annual rate for major bleeding of 5.1%.

Sensitivity Analyses According to Different Oral Anticoagulant TherapiesAlthough limited by a smaller number of events, sensitivity analyses were performed in relation todifferent oral anticoagulant therapies. The results were similar between the ARISTOTLE cohort(warfarin, HR, 1.62 [95% CI, 1.30-2.02]; apixaban, HR, 1.69 [95% CI, 1.32-2.18]) (Figure 3A) and theRE-LY cohort (warfarin, HR, 1.77 [95% CI, 1.29-2.42]; dabigatran 110 mg, HR, 1.50 [95% CI, 1.08-2.07];dabigatran 150 mg, HR, 1.84 [95% CI, 1.34-2.52]) (Figure 3B).

Comparison With the HAS-BLED Score and Discriminatory ValueThe risk of major bleeding according to the HAS-BLED score in patients with oral anticoagulationtaking aspirin compared with those not taking aspirin are shown in eFigure in the Supplement. Thediscriminatory ability to estimate the risk of major bleeding in patients with AF treated with oralanticoagulation and concomitant aspirin was 0.62 and 0.63 for the HAS-BLED score in theARISTOTLE and RE-LY cohorts, and 0.68 and 0.72 for the ABC-bleeding score, respectively.Additionally, the information gained from these risk scores was also assessed in models containingboth scores (eTable in the Supplement). This showed that the ABC-bleeding risk score remainedsignificant in the model, whereas the HAS-BLED score was no longer significant (eFigure in theSupplement).

Discussion

The main findings in this study are that the ABC-bleeding risk score is able to stratify patients with AFreceiving the combination of aspirin and oral anticoagulation according to their risk of bleeding. Alow estimated risk of major bleeding according to the ABC-bleeding risk score showed a low bleedingrisk even with concomitant aspirin treatment, whereas a higher ABC-bleeding risk score wasassociated with higher bleeding risk. These findings suggest that the ABC-bleeding risk score may bea useful tool for decision support concerning intensity and duration of combination antithrombotictreatment in patients with AF and CAD to improve the risk-benefit balance with differentantithrombotic strategies.

The added value of biomarkers for improved risk estimation in patients with AF have so far beenrather consistent in several cohorts, including randomized clinical trial cohorts, as well as lessselected observational registry cohorts.19-21 The ABC-bleeding risk score, which includes the mostrelevant variables, including both biomarkers and clinical data, was developed and validated usingrobust methods and very large cohorts of patients with AF.7,22,23 The ABC-bleeding risk scoreoutperforms traditional risk scores for stroke, death, and major bleeding, in comparison with riskscores based on clinical variables. In addition, the ABC-bleeding risk score has shown to be wellcalibrated, with estimated risk aligning well with observed risk during follow-up.7,22,23 Biomarkers

Table 2. Event Rates of Major Bleeding

Groupa Patients, No. Events, No. Person-years, No.Incidence rate/100person-years (95% CI)

ARISTOTLE

No aspirin 11 943 444 19 781 2.24 (2.04-2.46)

Aspirin 4987 207 5122 4.04 (3.51-4.63)

Total 14 697 651 24 903 2.61 (2.42-2.82)

RE-LY

No aspirin 5372 225 10 519 2.14 (1.87-2.44)

Aspirin 3096 238 5693 4.18 (3.67-4.75)

Total 8468 463 16 212 2.86 (2.60-3.13)

Abbreviations: ARISTOTLE, Apixaban for Reduction inStroke and Other Thromboembolic Events in AtrialFibrillation; RE-LY, Randomized Evaluation of Long-term Anticoagulation Therapy.a ARISTOTLE data are based on time-updated aspirin

data. RE-LY data are based on aspirin data fromanytime during the study.




https://jama.jamanetwork.com/article.aspx?doi=10.1001/jamanetworkopen.2020.15943&utm_campaign=articlePDF%26utm_medium=articlePDFlink%26utm_source=articlePDF%26utm_content=jamanetworkopen.2020.15943https://jama.jamanetwork.com/article.aspx?doi=10.1001/jamanetworkopen.2020.15943&utm_campaign=articlePDF%26utm_medium=articlePDFlink%26utm_source=articlePDF%26utm_content=jamanetworkopen.2020.15943https://jama.jamanetwork.com/article.aspx?doi=10.1001/jamanetworkopen.2020.15943&utm_campaign=articlePDF%26utm_medium=articlePDFlink%26utm_source=articlePDF%26utm_content=jamanetworkopen.2020.15943

and the ABC-bleeding risk scores in patients with AF have, therefore, been suggested as potentialnew tools for improved risk estimation in the latest European guidelines for management of AF.1

Similarly, the European Dual Antiplatelet Therapy guidelines refer to, among others, theABC-bleeding risk score as a possible tool to estimate bleeding risk in patients with AF with anindication for added antiplatelet therapy to assess the bleeding risk and make informed decisionsregarding the intensity and duration of concomitant antiplatelet therapy.2 Patients with AF and CADconsidered to have a high ischemic risk (eg, because of acute coronary syndrome or other anatomicalor procedural characteristics) should, according to the guidelines, be considered for a more intenseand prolonged antithrombotic strategy. If these patients are considered to have a high risk ofbleeding, a less intense and shorter duration of concomitant antiplatelet therapy is suggested in theguidelines. To our knowledge, the present study is the first to evaluate the ABC-bleeding risk scorein patients with AF and chronic CAD receiving concomitant single-antiplatelet therapy vs oral

Figure 1. Relative Hazard of Major Bleeding Comparing Patients With OralAnticoagulation Taking Aspirin With Patients Not Taking Aspirinby Estimated 1-Year Risk According to the ABC (Age, Biomarkers,and Clinical History)–Bleeding Risk Score

10

3

1

0.3

0.1

0.03

Rela

tive

haza

rd

1-y ABC bleeding risk0 0.02 0.04 0.06 0.08 0.10

AspirinNo aspirin

ARISTOTLE cohortA

10

30

3

1

0.3

0.1

0.03

Rela

tive

haza

rd

1-y ABC bleeding risk0 0.02 0.04 0.06 0.08 0.10

AspirinNo aspirin

RE-LY cohortB

Graphs show relative hazards (lines) and 95% CIs (shaded areas) for theARISTOTLE (Apixaban for Reduction in Stroke and Other ThromboembolicEvents in Atrial Fibrillation) (A) and RE-LY (Randomized Evaluation of Long-termAnticoagulation Therapy) (B) cohorts. The bottom of each panel shows thedistribution of the patients with ABC-bleeding risk scores as a density plot foreach group at baseline.




anticoagulation, and our findings support that the ABC-bleeding risk score is useful for estimatingbleeding risk in such patients.

The landmark non–vitamin K antagonist oral anticoagulant (NOAC) trials24 have showed thatpatients taking NOACs have similar or lower risks of major bleeding in comparison with patientstaking warfarin. Furthermore, substudies from these NOAC trials have shown that the risk of majorbleeding increases with concomitant use of aspirin, although with preserved safety with NOACscompared with warfarin; for example, even if the risk of major bleeding increases with concomitantantiplatelet therapy in both the NOAC as well as the warfarin groups, the relative risk is still lower withNOACs.4,25-27 Nonetheless, in some patients, the risk of ischemic events may warrant a moreaggressive antithrombotic treatment strategy. Recent randomized clinical trials have repeatedlyhighlighted the changing balance between benefits and risks with increasing intensity ofantithrombotic treatment. For instance, the COMPASS trial,28 which compared the use of dualtherapy (low-dose rivaroxaban in combination with aspirin) with single-antithrombotic therapy(aspirin alone) in patients with stable CAD without AF, found that the combination of low-doserivaroxaban with aspirin significantly reduced the risk of ischemic cardiovascular events. However,this comes at a cost of increased risk of major bleeding events. A similar pattern is also evident earlyafter coronary events (ie, percutaneous coronary interventions and/or acute coronary syndromes)when comparing triple- and dual-antithrombotic therapy strategies in patients with AF.29,30 Thesestudies consistently show that the risk of clinically relevant, as well as major, bleeding events aresubstantially lower with less-aggressive antithrombotic treatment strategies (ie, dual- vs triple-antithrombotic therapy).29,30 However, this might, to varying degrees, come at the price of anincreased risk of ischemic events, including stent thrombosis.29-31 Risk score models that improve therisk stratification concerning bleeding events would thus be useful to better identify patients for

Figure 2. One-Year Risk of Major Bleeding Rate as a Function of the ABC(Age, Biomarkers, and Clinical History)–Bleeding Score by DifferentTreatment Combinations

32

2

4

8

16

0.10

0.25

0.50

1

Obs

erve

d ev

ent r

ate,

%/y

1-y ABC bleeding risk, %0.5 1 2 4 8

Warfarin with aspirinWarfarin without aspirinApixaban with aspirinApixaban without aspirin

Lines denote hazard ratios. Shaded areas denote 95% CIs. The solid black lineindicates perfect calibration. The bottom of the graph shows the distribution ofthe patients with ABC-bleeding risk scores as a density plot for each group atbaseline.




whom more intensive antithrombotic therapy is suitable. By providing an individualized treatmentstrategy, these models would also improve patient outcomes.

In the present study, we expanded on previous findings by showing that the biomarker-basedABC-bleeding risk score successfully identifies patients with different risks of bleeding whencombining single-antiplatelet therapy with aspirin and oral anticoagulation. The ABC-bleeding riskscore may, therefore, be a useful tool for decision support concerning the intensity and duration ofcombination antithrombotic treatment in patients with AF and stable CAD. However, it is importantto keep in mind that concomitant use of aspirin will offset the calibration of the ABC-bleeding score,because the observed risk of bleeding will increase by adding concomitant antiplatelet therapy.

LimitationsThis study has limitations that should be addressed. The patients in the ARISTOTLE trial wereclinically stable, and the biomarkers used in the ABC-bleeding risk score, in particular troponin, will beaffected in patients with acute coronary syndromes. Thus, the ABC-bleeding risk score shouldpreferably be assessed at a later time point, such as during routine outpatient postdischargecheck-up, for the most accurate ABC-bleeding risk estimation. Because antiplatelet therapy was notrandomized, there may be confounding by indication, and thus evaluation of ischemic risks was notperformed. For the ARISTOTLE data, it was possible to use an advanced statistical model in whichaspirin was used as a time-updated covariable. In the RE-LY cohort, the status of concomitantantiplatelet therapy was defined as any time during the study and may not completely reflect themedication during study follow-up. Nevertheless, the results were very consistent between the 2cohorts. Furthermore, as with every risk score, a prospective evaluation of the risk score such as the

Figure 3. Relative Risk of Major Bleeding by Study Treatment

10

30

3

1

0.3

0.1

0.03

Rela

tive

haza

rd

1-y ABC bleeding risk


Warfarin

0 0.02 0.04 0.06 0.08 0.10


Apixaban

ARISTOTLE trialA

0 0.02 0.04 0.06 0.08 0.10

30

10

3

1

0.1

0.3

0.03

Rela

tive

haza

rd

WarfarinRE-LY trialB

0 0.02 0.04 0.06 0.08 0.10


Dabigatran 150 mg

0 0.02 0.04 0.06 0.08 0.10


Dabigatran 110 mg

0 0.02 0.04 0.06 0.08 0.10

AspirinNo aspirin

Graphs show data for the ARISTOTLE (Apixaban for Reduction in Stroke and OtherThromboembolic Events in Atrial Fibrillation) (A) and RE-LY (Randomized Evaluation ofLong-term Anticoagulation Therapy) (B) cohorts. Lines denote hazard ratios. Shaded

areas denote 95% CIs. The bottom of each panel shows the distribution of the patientsABC (age, biomarkers, and clinical history)–bleeding risk scores as a density plot for eachgroup at baseline.




ongoing ABC-AF study (ClinicalTrials.gov identifier NCT03753490) is desirable to fully assess the clinicalusefulness of the ABC-bleeding score for better tailoring of antithrombotic treatment strategies.

Conclusions

The ABC-bleeding risk score accurately identifies patients with different risks of bleeding whencombining aspirin and oral anticoagulation; low ABC-bleeding risk scores were associated with lowbleeding risk, and higher ABC-bleeding risk scores were associated with higher bleeding risk. Thesedata suggest that the ABC-bleeding risk score may provide useful information concerning intensityand duration of combination antithrombotic treatment in patients with AF and stable CAD.

ARTICLE INFORMATIONAccepted for Publication: June 24, 2020.

Published: September 16, 2020. doi:10.1001/jamanetworkopen.2020.15943

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2020 Hijazi Z et al.JAMA Network Open.

Corresponding Author: Ziad Hijazi, MD, PhD, Uppsala Clinical Research Center, Uppsala University, UppsalaScience Park, SE-752 37 Uppsala, Sweden ([email protected]).

Author Affiliations: Department of Medical Sciences, Cardiology, Uppsala University, Uppsala, Sweden (Hijazi,Oldgren, Held, Wallentin); Uppsala Clinical Research Center, Uppsala University, Uppsala, Sweden (Hijazi, Oldgren,Lindbäck, Held, Siegbahn, Wallentin); Duke Clinical Research Institute, Duke Medicine, Durham, North Carolina(Alexander, Granger, Lopes); University Medical Center Utrecht, Utrecht University, Utrecht, Netherlands (Alings);University Cardiology Division, University of Pisa, Pisa, Italy (De Caterina); Population Health Research Institute,Hamilton, Ontario, Canada (Eikelboom); Sidney Kimmel Medical College, Thomas Jefferson University,Philadelphia, Pennsylvania (Ezekowitz); Cardiovascular Medicine, Lankenau Institute for Medical Research,Wynnewood, Pennsylvania (Ezekowitz); Wilhelminenhospital, 3rd Department of Medicine, Cardiology andIntensive Care Medicine, Medical Faculty, Sigmund Freud University, Vienna, Austria (Huber); CardiovascularMedicine, Boston University School of Medicine, Boston, Massachusetts (Hylek); University and EmergencyHospital, University of Medicine and Pharmacy Carol Davila, Bucharest, Romania (Vinereanu); Clinical Chemistry,Department of Medical Sciences, Uppsala University, Uppsala, Sweden (Siegbahn).

Author Contributions: Dr Hijazi and Mr Lindbäck had full access to all of the data in the study and takeresponsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Hijazi, Oldgren, Lindbäck, Eikelboom, Ezekowitz, Hylek, Granger, Vinereanu, Wallentin.

Acquisition, analysis, or interpretation of data: Hijazi, Oldgren, Lindbäck, Alexander, Alings, De Caterina,Eikelboom, Ezekowitz, Held, Huber, Hylek, Lopes, Vinereanu, Siegbahn, Wallentin.

Drafting of the manuscript: Hijazi.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Lindbäck.

Obtained funding: Hijazi, Oldgren, Eikelboom, Siegbahn, Wallentin.

Administrative, technical, or material support: Ezekowitz, Siegbahn, Wallentin.

Supervision: Oldgren, De Caterina, Ezekowitz, Hylek, Vinereanu, Siegbahn, Wallentin.

Conflict of Interest Disclosures: Dr Hijazi reported receiving lecture fees from Boehringer Ingelheim, Bristol-Myers Squibb/Pfizer, and Roche Diagnostics and consulting fees from Boehringer Ingelheim, Bristol-Myers Squibb,Merck Sharp & Dohme, Pfizer, and Roche Diagnostics. Dr Oldgren reported receiving institutional research grantsand fees paid to his institution for advisory boards and lectures from Roche Diagnostics; fees paid to his institutionfor advisory boards, study steering committees, and lectures from AstraZeneca, Bayer, BMS, BoehringerIngelheim, Pfizer, and Sanofi; and fees paid to his institution for advisory boards, safety committees, and lecturesfrom Daichii-Sankyo. Dr Alexander reported receiving institutional research grants and consulting fees andhonoraria from Bayer, Bristol-Myers Squibb, CryoLife, and XaTek and consulting fees and honoraria from Pfizer andPortola. Dr Alings reported receiving consulting fees from Bayer, Boehringer Ingelheim, Bristol-Myers Squibb,Daiichi Sankyo, and Milestone Pharmaceuticals; personal fees from Pfizer; and a nonpersonal study grant fromSanofi. Dr De Caterina reported receiving grants for congress organizations and personal fees and honoraria for




https://clinicaltrials.gov/ct2/show/NCT03753490https://jama.jamanetwork.com/article.aspx?doi=10.1001/jamanetworkopen.2020.15943&utm_campaign=articlePDF%26utm_medium=articlePDFlink%26utm_source=articlePDF%26utm_content=jamanetworkopen.2020.15943https://jamanetwork.com/journals/jamanetworkopen/pages/instructions-for-authors#SecOpenAccess/?utm_campaign=articlePDF%26utm_medium=articlePDFlink%26utm_source=articlePDF%26utm_content=jamanetworkopen.2020.15943mailto:[email protected]

lecturing and advisory board participations from Amgen, Boehringer Ingelheim, Bayer, Bristol-Myers Squibb/Pfizer, Daiichi-Sankyo, Novartis, Protola, Roche, and Sanofi. Dr Eikelboom reported receiving institutional researchgrants and honoraria from AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb/Pfizer, Daiichi-Sankyo, Eli Lilly, Glaxo Smith Kline, Janssen, and Sanofi. Dr Held reported receiving an institutional research grantfrom GlaxoSmithKline, honoraria from Pfizer, and consultant and advisory board fees from AstraZeneca, Bayer,Boehringer Ingelheim, and Coala Life. Dr Granger reported receiving research grants and consulting and speakerfees from Boehringer Ingelheim, Bristol-Myers Squibb, Janssen Pharmaceutica Products, LP, and Pfizer; researchgrants from Daichii-Sankyo, AKROS, Apple, AstraZeneca, GlaxoSmithKline, US Food & Drug Administration,Medtronic Foundation, and Novartis Pharmaceutical Company; and consulting and speaker fees from Abbvie,Bayer Corp US, Boston Scientific Corp, CeleCor Therapeutics, Correvio, Espero BioPharma, Medscape, MedtronicInc, Merck, National Institutes of Health, NovoNordisk, Rhoshan Pharmaceuticals, and Roche Diagnostics. DrLopes reported receiving research grants and consulting fees from Bristol-Myers Squibb, GlaxoSmithKline,Medtronic PLC, Pfizer, and Sanofi and consulting fees from Bayer, Boehringer Ingelheim, Daichii-Sankyo, Merck,and Portola. Dr Vinereanu reported receiving grants and personal fees from Bayer, Boehringer Ingelheim, andPfizer. Dr Siegbahn reported receiving institutional research grants from AstraZeneca, Boehringer Ingelheim,Bristol-Myers Squibb/Pfizer, GlaxoSmithKline, and Roche Diagnostics and consultancy fees from Olink Proteomics.Dr Wallentin reported receiving institutional research grants from AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb/Pfizer, GlaxoSmithKline, Roche Diagnostics, and Merck & Co and consulting fees from Abbott andholding 2 patents involving GDF-15 licensed to Roche Diagnostics (EP2047275B1 and US8951742B2). No otherdisclosures were reported.

Funding/Support: The ARISTOTLE trial was funded by Bristol-Myers Squibb and Pfizer, Inc, and coordinated bythe Duke Clinical Research Institute and Uppsala Clinical Research Center. The RE-LY trial was funded byBoehringer-Ingelheim and was coordinated by the Population Health Research Institute, Hamilton, Ontario,Canada, and Uppsala Clinical Research Center, Uppsala, Sweden. The current analyses were supported by grantRB13-0197 from the Swedish Foundation for Strategic Research. Roche Diagnostics provided the GDF-15 assay freeof charge. Dr Hijazi receives research grants from the Swedish Society for Medical Research (S17-0133) and theSwedish Heart-Lung Foundation (20170718).

Role of the Funder/Sponsor: The funding sources were given the opportunity to review and comment on the finalversion of the manuscript. The funding sources did not take part in the design and conduct of the study; collection,management, analysis, and interpretation of the data; preparation and approval of the manuscript; or the decisionto submit the manuscript for publication.

REFERENCES1. Kirchhof P, Benussi S, Kotecha D, et al; ESC Scientific Document Group. 2016 ESC guidelines for themanagement of atrial fibrillation developed in collaboration with EACTS. Eur Heart J. 2016;37(38):2893-2962.doi:10.1093/eurheartj/ehw210

2. Valgimigli M, Bueno H, Byrne RA, et al. 2017 ESC focused update on dual antiplatelet therapy in coronary arterydisease developed in collaboration with EACTS [in Polish]. Kardiol Pol. 2017;75(12):1217-1299. doi:10.5603/KP.2017.0224

3. Sørensen R, Hansen ML, Abildstrom SZ, et al. Risk of bleeding in patients with acute myocardial infarctiontreated with different combinations of aspirin, clopidogrel, and vitamin K antagonists in Denmark: a retrospectiveanalysis of nationwide registry data. Lancet. 2009;374(9706):1967-1974. doi:10.1016/S0140-6736(09)61751-7

4. Alexander JH, Lopes RD, Thomas L, et al. Apixaban vs. warfarin with concomitant aspirin in patients with atrialfibrillation: insights from the ARISTOTLE trial. Eur Heart J. 2014;35(4):224-232. doi:10.1093/eurheartj/eht445

5. Angiolillo DJ, Goodman SG, Bhatt DL, et al. Antithrombotic therapy in patients with atrial fibrillation undergoingpercutaneous coronary intervention: a North American perspective—2016 update. Circ Cardiovasc Interv. 2016;9(11):e004395. doi:10.1161/CIRCINTERVENTIONS.116.004395

6. Nieuwlaat R, Capucci A, Camm AJ, et al; European Heart Survey Investigators. Atrial fibrillation management:a prospective survey in ESC member countries—the Euro Heart Survey on Atrial Fibrillation. Eur Heart J. 2005;26(22):2422-2434. doi:10.1093/eurheartj/ehi505

7. Hijazi Z, Oldgren J, Lindbäck J, et al; ARISTOTLE and RE-LY Investigators. The novel biomarker-based ABC (age,biomarkers, clinical history)-bleeding risk score for patients with atrial fibrillation: a derivation and validation study.Lancet. 2016;387(10035):2302-2311. doi:10.1016/S0140-6736(16)00741-8

8. Berg DD, Ruff CT, Jarolim P, et al. Performance of the ABC scores for assessing the risk of stroke or systemicembolism and bleeding in patients with atrial fibrillation in ENGAGE AF-TIMI 48. Circulation. 2019;139(6):760-771.doi:10.1161/CIRCULATIONAHA.118.038312

9. Connolly SJ, Ezekowitz MD, Yusuf S, et al; RE-LY Steering Committee and Investigators. Dabigatran versuswarfarin in patients with atrial fibrillation. N Engl J Med. 2009;361(12):1139-1151. doi:10.1056/NEJMoa0905561




https://dx.doi.org/10.1093/eurheartj/ehw210https://dx.doi.org/10.5603/KP.2017.0224https://dx.doi.org/10.5603/KP.2017.0224https://dx.doi.org/10.1016/S0140-6736(09)61751-7https://dx.doi.org/10.1093/eurheartj/eht445https://dx.doi.org/10.1161/CIRCINTERVENTIONS.116.004395https://dx.doi.org/10.1093/eurheartj/ehi505https://dx.doi.org/10.1016/S0140-6736(16)00741-8https://dx.doi.org/10.1161/CIRCULATIONAHA.118.038312https://dx.doi.org/10.1056/NEJMoa0905561

10. Granger CB, Alexander JH, McMurray JJ, et al; ARISTOTLE Committees and Investigators. Apixaban versuswarfarin in patients with atrial fibrillation. N Engl J Med. 2011;365(11):981-992. doi:10.1056/NEJMoa1107039

11. Lopes RD, Alexander JH, Al-Khatib SM, et al; ARISTOTLE Investigators. Apixaban for reduction in stroke andother thromboembolic events in atrial fibrillation (ARISTOTLE) trial: design and rationale. Am Heart J. 2010;159(3):331-339. doi:10.1016/j.ahj.2009.07.035

12. Ezekowitz MD, Connolly S, Parekh A, et al. Rationale and design of RE-LY: randomized evaluation of long-termanticoagulant therapy, warfarin, compared with dabigatran. Am Heart J. 2009;157(5):805-810. doi:10.1016/j.ahj.2009.02.005

13. Hijazi Z, Oldgren J, Andersson U, et al. Cardiac biomarkers are associated with an increased risk of stroke anddeath in patients with atrial fibrillation: a Randomized Evaluation of Long-term Anticoagulation Therapy (RE-LY)substudy. Circulation. 2012;125(13):1605-1616. doi:10.1161/CIRCULATIONAHA.111.038729

14. Hijazi Z, Siegbahn A, Andersson U, et al; ARISTOTLE Investigators. High-sensitivity troponin I for riskassessment in patients with atrial fibrillation: insights from the Apixaban for Reduction in Stroke and otherThromboembolic Events in Atrial Fibrillation (ARISTOTLE) trial. Circulation. 2014;129(6):625-634. doi:10.1161/CIRCULATIONAHA.113.006286

15. Hijazi Z, Wallentin L, Siegbahn A, et al; ARISTOTLE Investigators. High-sensitivity troponin T and riskstratification in patients with atrial fibrillation during treatment with apixaban or warfarin. J Am Coll Cardiol. 2014;63(1):52-61. doi:10.1016/j.jacc.2013.07.093

16. Hijazi Z, Wallentin L, Siegbahn A, et al. N-terminal pro-B-type natriuretic peptide for risk assessment inpatients with atrial fibrillation: insights from the ARISTOTLE Trial (Apixaban for the Prevention of Stroke in SubjectsWith Atrial Fibrillation). J Am Coll Cardiol. 2013;61(22):2274-2284. doi:10.1016/j.jacc.2012.11.082

17. Wallentin L, Hijazi Z, Andersson U, et al; ARISTOTLE Investigators. Growth differentiation factor 15, a marker ofoxidative stress and inflammation, for risk assessment in patients with atrial fibrillation: insights from the Apixabanfor Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation (ARISTOTLE) trial. Circulation.2014;130(21):1847-1858. doi:10.1161/CIRCULATIONAHA.114.011204

18. Camm AJ, Kirchhof P, Lip GY, et al; European Heart Rhythm Association; European Association for Cardio-Thoracic Surgery. Guidelines for the management of atrial fibrillation: the Task Force for the Management of AtrialFibrillation of the European Society of Cardiology (ESC). Eur Heart J. 2010;31(19):2369-2429. doi:10.1093/eurheartj/ehq278

19. Hijazi Z, Oldgren J, Siegbahn A, Wallentin L. Application of biomarkers for risk stratification in patients withatrial fibrillation. Clin Chem. 2017;63(1):152-164. doi:10.1373/clinchem.2016.255182

20. Inohara T, Kim S, Pieper K, et al; ORBIT AF Patients and Investigators. B-type natriuretic peptide, diseaseprogression and clinical outcomes in atrial fibrillation. Heart. 2019;105(5):370-377.

21. Hijazi Z, Wallentin L. Natriuretic peptide should be used as a routine tool for evaluation of patients with atrialfibrillation. Heart. 2019;105(5):353-354.

22. Hijazi Z, Lindbäck J, Alexander JH, et al; ARISTOTLE and STABILITY Investigators. The ABC (age, biomarkers,clinical history) stroke risk score: a biomarker-based risk score for predicting stroke in atrial fibrillation. Eur Heart J.2016;37(20):1582-1590. doi:10.1093/eurheartj/ehw054

23. Hijazi Z, Oldgren J, Lindbäck J, et al; ARISTOTLE and RE-LY Investigators. A biomarker-based risk score topredict death in patients with atrial fibrillation: the ABC (age, biomarkers, clinical history) death risk score. EurHeart J. 2018;39(6):477-485. doi:10.1093/eurheartj/ehx584

24. Ruff CT, Giugliano RP, Braunwald E, et al. Comparison of the efficacy and safety of new oral anticoagulants withwarfarin in patients with atrial fibrillation: a meta-analysis of randomised trials. Lancet. 2014;383(9921):955-962.doi:10.1016/S0140-6736(13)62343-0

25. Dans AL, Connolly SJ, Wallentin L, et al. Concomitant use of antiplatelet therapy with dabigatran or warfarin inthe Randomized Evaluation of Long-Term Anticoagulation Therapy (RE-LY) trial. Circulation. 2013;127(5):634-640. doi:10.1161/CIRCULATIONAHA.112.115386

26. Shah R, Hellkamp A, Lokhnygina Y, et al; ROCKET AF Steering Committee Investigators. Use of concomitantaspirin in patients with atrial fibrillation: findings from the ROCKET AF trial. Am Heart J. 2016;179:77-86. doi:10.1016/j.ahj.2016.05.019

27. Xu H, Ruff CT, Giugliano RP, et al. Concomitant use of single antiplatelet therapy with edoxaban or warfarin inpatients with atrial fibrillation: analysis from the ENGAGE AF-TIMI48 Trial. J Am Heart Assoc. 2016;5(2):e002587. doi:10.1161/JAHA.115.002587

28. Eikelboom JW, Connolly SJ, Bosch J, et al; COMPASS Investigators. Rivaroxaban with or without aspirin instable cardiovascular disease. N Engl J Med. 2017;377(14):1319-1330. doi:10.1056/NEJMoa1709118




https://dx.doi.org/10.1056/NEJMoa1107039https://dx.doi.org/10.1016/j.ahj.2009.07.035https://dx.doi.org/10.1016/j.ahj.2009.02.005https://dx.doi.org/10.1016/j.ahj.2009.02.005https://dx.doi.org/10.1161/CIRCULATIONAHA.111.038729https://dx.doi.org/10.1161/CIRCULATIONAHA.113.006286https://dx.doi.org/10.1161/CIRCULATIONAHA.113.006286https://dx.doi.org/10.1016/j.jacc.2013.07.093https://dx.doi.org/10.1016/j.jacc.2012.11.082https://dx.doi.org/10.1161/CIRCULATIONAHA.114.011204https://dx.doi.org/10.1093/eurheartj/ehq278https://dx.doi.org/10.1093/eurheartj/ehq278https://dx.doi.org/10.1373/clinchem.2016.255182https://www.ncbi.nlm.nih.gov/pubmed/30228248https://www.ncbi.nlm.nih.gov/pubmed/30314973https://dx.doi.org/10.1093/eurheartj/ehw054https://dx.doi.org/10.1093/eurheartj/ehx584https://dx.doi.org/10.1016/S0140-6736(13)62343-0https://dx.doi.org/10.1161/CIRCULATIONAHA.112.115386https://dx.doi.org/10.1016/j.ahj.2016.05.019https://dx.doi.org/10.1016/j.ahj.2016.05.019https://dx.doi.org/10.1161/JAHA.115.002587https://dx.doi.org/10.1056/NEJMoa1709118

29. Cannon CP, Bhatt DL, Oldgren J, et al; RE-DUAL PCI Steering Committee and Investigators. Dualantithrombotic therapy with dabigatran after PCI in atrial fibrillation. N Engl J Med. 2017;377(16):1513-1524. doi:10.1056/NEJMoa1708454

30. Lopes RD, Heizer G, Aronson R, et al; AUGUSTUS Investigators. Antithrombotic therapy after acute coronarysyndrome or PCI in atrial fibrillation. N Engl J Med. 2019;380(16):1509-1524. doi:10.1056/NEJMoa1817083

31. Lopes RD, Leonardi S, Wojdyla DM, et al. Stent thrombosis in patients with atrial fibrillation undergoingcoronary stenting in the AUGUSTUS Trial. Circulation. 2020;141(9):781-783. doi:10.1161/CIRCULATIONAHA.119.044584

SUPPLEMENT.eFigure. Relative Hazard of Major Bleeding Comparing Patients With Oral Anticoagulation on Aspirin (ASA) WithPatients Not on ASA in ARISTOTLE and RE-LY by Estimated HAS-BLED RiskeTable. ANOVA Table for the Cox Regression Model Including Aspirin, the ABC-Bleeding and HAS-BLED Scores,and the Interaction Between Aspirin and the Respective Bleeding Risk Scores



https://dx.doi.org/10.1056/NEJMoa1708454https://dx.doi.org/10.1056/NEJMoa1708454https://dx.doi.org/10.1056/NEJMoa1817083https://dx.doi.org/10.1161/CIRCULATIONAHA.119.044584https://dx.doi.org/10.1161/CIRCULATIONAHA.119.044584

evaluationoftheage,biomarkers,andclinicalhistory ...€¦ · hypertension 8961 (86.6) 4153 (89.5)...

Documents