Author + information
- Received December 7, 2018
- Revision received February 19, 2019
- Accepted February 26, 2019
- Published online June 24, 2019.
- Sonia S. Anand, MD, PhDa,∗ (, )@DrSoniaAnand1,
- John W. Eikelboom, MBBS, MSca,
- Leanne Dyal, MScb,
- Jackie Bosch, MSc, PhDc,
- Christoph Neumann, PhDd,
- Petr Widimsky, MD, DSce,
- Alvaro A. Avezum, MD, PhDf,
- Jeffrey Probstfield, MDg,
- Nancy Cook Bruns, MDd,
- Keith A.A. Fox, MBChB, BSch,
- Deepak L. Bhatt, MD, MPHi,
- Stuart J. Connolly, MDa,
- Salim Yusuf, MBBS, DPhila,
- for the COMPASS Trial Investigators
- aDepartment of Medicine, Population Health Research Institute, McMaster University, Hamilton Health Sciences, Hamilton, Ontario, Canada
- bDepartment of Statistics, Population Health Research Institute, McMaster University, Hamilton Health Sciences, Hamilton, Ontario, Canada
- cSchool of Rehabilitative Science, Population Health Research Institute, McMaster University, Hamilton Health Sciences, Hamilton, Ontario, Canada
- dBayer AG, Wuppertal, Germany
- eDepartment of Cardiology, University Hospital Kralovske Vinohrady and Third Faculty of Medicine, Charles University, Prague, Czech Republic
- fResearch Division, Instituto Dante Pazzanese de Cardiologia, São Paulo, Brazil
- gDepartment of Medicine and Cardiology, University of Washington, Seattle, Washington
- hCenter for Cardiovascular Science, University of Edinburgh, Edinburgh, United Kingdom
- iBrigham and Women’s Hospital Heart and Vascular Centre, Harvard Medical School, Boston, Massachusetts
- ↵∗Address for correspondence:
Dr. Sonia S. Anand, Population Health Research Institute, 30 Birge Street, Hamilton, Ontario 8L 0A6, Canada.
Background The COMPASS (Cardiovascular Outcomes for People Using Anticoagulation Strategies) trial showed that the combination of low-dose rivaroxaban and aspirin reduced major vascular events in patients with stable vascular disease.
Objectives The purpose of this study was to identify subsets of patients at higher risk of recurrent vascular events, which may help focus the use of rivaroxaban and aspirin therapy.
Methods COMPASS patients with vascular disease were risk stratified using 2 methods: the REACH (REduction of Atherothrombosis for Continued Health) atherothrombosis risk score and CART (Classification and Regression Tree) analysis. The absolute risk differences for rivaroxaban with aspirin were compared to aspirin alone over 30 months for the composite of cardiovascular death, myocardial infarction, stroke, acute limb ischemia, or vascular amputation; for severe bleeding; and for the net clinical benefit.
Results High-risk patients using the REACH score were those with 2 or more vascular beds affected, history of heart failure (HF), or renal insufficiency, and by CART analysis were those with ≥2 vascular beds affected, history of HF, or diabetes. Rivaroxaban and aspirin combination reduced the serious vascular event incidence by 25% (4.48% vs. 5.95%, hazard ratio: 0.75; 95% confidence interval: 0.66 to 0.85), equivalent to 23 events prevented per 1,000 patients treated for 30 months, at the cost of a nonsignificant 34% increase in severe bleeding (1.34; 95% confidence interval: 0.95 to 1.88), or 2 events caused per 1,000 patients treated. Among patients with ≥1 high-risk feature identified from the CART analysis, rivaroxaban and aspirin prevented 33 serious vascular events, whereas in lower-risk patients, rivaroxaban and aspirin treatment led to the avoidance of 10 events per 1,000 patients treated for 30 months.
Conclusions In patients with vascular disease, further risk stratification can identify higher-risk patients (≥2 vascular beds affected, HF, renal insufficiency, or diabetes). The net clinical benefit remains favorable for most patients treated with rivaroxaban and aspirin compared with aspirin.
The most common mechanism of stroke, myocardial infarction (MI), and acute limb ischemia is thrombosis superimposed on a ruptured or eroded atherosclerotic plaque, or embolism from a proximal source (1,2). This process of atherothromboembolism can be reduced by using medical therapies that prevent atherosclerosis, stabilize plaques or prevent thromboembolic complications. The mainstay of antithrombotic treatment in patients at risk of ischemic events has been use of a single antiplatelet agent (3). However, the residual risk of ischemic vascular events remains considerable, and this has fueled the search for more effective antithrombotic therapies. We recently reported that among individuals with established vascular disease, the combination of rivaroxaban and aspirin compared with aspirin alone reduced the risk of (CV) death, stroke, MI, acute limb ischemia, and vascular amputation (4). Although the combination also increased bleeding, there was no significant excess of severe bleeding, and mortality was reduced. The purpose of this analysis is to identify whether in this already high-risk patient cohort, further risk stratification can identify subsets of the highest-risk patients with the greatest net clinical benefit, and also whether subgroups who have a lower risk of recurrent CV events still benefit from combination therapy with rivaroxaban and aspirin.
Trial design and population
COMPASS (Cardiovascular Outcomes for People Using Anticoagulation Strategies) was a multicenter, double-blind, randomized, placebo-controlled trial that enrolled 27,395 high-risk patients with a clinical history of coronary and/or peripheral artery disease (PAD), and compared the combination of rivaroxaban 2.5 mg twice daily and aspirin or rivaroxaban 5 mg twice daily (with aspirin placebo) to aspirin alone (with rivaroxaban placebo). The primary outcome was CV death, MI, or stroke. The details of inclusion and exclusion criteria, adjudication process, and definitions of outcomes have been previously published (4–7). The trial was stopped prematurely due to overwhelming benefit of rivaroxaban and aspirin over aspirin only, and this analysis is limited to these trial participants (n = 18,278).
We expanded the pre-specified primary outcome of the trial, which was the composite of CV death, MI, and stroke, to include other clinically important vascular outcomes including acute limb ischemia (ALI) and vascular amputation. The main safety outcome in this analysis is severe bleeding, pre-defined as bleeding that was fatal or symptomatic into a critical organ (4). The net clinical benefit outcome is the composite of cardiovascular death, stroke, MI, ALI, vascular amputation, fatal bleeding, or symptomatic bleeding into a critical organ. Data for the CV death, stroke, or MI and major bleeding outcomes are shown in Online Tables 1 and 2.
We used 2 methods to classify patients by risk of vascular events. In the first, we applied the validated REACH (REduction of Atherothrombosis for Continued Health) registry risk score to the COMPASS trial patients (8). The REACH registry recruited and followed a large cohort of outpatients with known vascular disease or at high risk of developing ischemic vascular events. We used the published scoring system, which includes sex, age, smoking, diabetes, low body mass index, number of vascular beds affected, CV event in the past year, heart failure (HF), atrial fibrillation, statin therapy, aspirin therapy, and low estimated glomerular filtration rate (eGFR), and grouped countries into regions defined as low, intermediate, or high risk (Online Table 3). Some adaptations to the scoring algorithm were required. Specifically, the original scoring classified patients by use of “statin therapy,” which was replaced with “lipid-lowering therapy” as collected in the COMPASS trial. Individuals with atrial fibrillation were excluded; therefore, all patients were given a score of 0 for this criterion. Aspirin use at baseline was replaced by “any antiplatelet therapy use” at baseline. Countries in the COMPASS trial were grouped into 3 risk regions for the outcome of CV death, MI, and stroke and were then divided into high-, intermediate-, and low-risk regions. We also identified low eGFR <60 ml/min as a predictor of major adverse cardiovascular events, and these patients were given a +2.0 risk score. The maximum REACH risk score in the COMPASS population is 27; and the median value is 12. We used this cutpoint to divide the patients into low- (<13) versus high-risk (13+) groups (Online Table 4).
For the second method of risk stratification, we used a CART (Classification and Regression Trees) survival analysis to identify independent groups of high-risk individuals in the aspirin-treated patients of the COMPASS trial. The CART method uses a series of algorithms to divide the population into subgroups that are best separated with respect to the incidence risks over time. Groups are split until no further variable splits significantly, and group sizes ≥100 are considered (9,10).
We calculated the Kaplan-Meier estimates of the 30-month incidence risk of the main efficacy outcome in the aspirin group for individuals with a score ≥13. We then identified any REACH score components that had a 30-month incidence risk at least equivalent to those with a REACH score of 13+ (i.e., ≥2 vascular beds, low eGFR, or history of HF). The CART (survival tree) analysis split the data into independent groups of patients with high risk for vascular events. The treatment effect of rivaroxaban and aspirin compared with aspirin was conducted on an intention-to-treat basis, and hazard ratios and corresponding 95% confidence intervals (CIs) were obtained from stratified Cox proportional-hazards models for the main efficacy, safety, and net-clinical benefit outcomes. For subgroup analyses, a test for interaction between treatment group and subgroup factors was conducted. The number of events prevented or caused by using the rivaroxaban and aspirin combination was determined by calculating the absolute risk difference for each subgroup over 30 months. The CART analysis was performed using the “partykit” package in R 3.2.5 (R Foundation, Vienna, Austria), and all other analyses were conducted using SAS 9.4 (SAS Institute, Cary, North Carolina).
A total of 18,278 patients were randomized to the combination of rivaroxaban and aspirin or aspirin in the COMPASS trial, and the baseline characteristics of patients with and without a vascular outcome event are shown in Table 1. The REACH risk score of 13+ had a concordance index of 0.625 (SE: 0.001) for the outcome of CV death, MI, stroke, ALI, and vascular amputation. This concordance index provides a global assessment of the predictive value of the fitted survival model. In the aspirin-treated patients, the incidence risk of CV death, MI, stroke, ALI or vascular amputation after 30 months is 8.0%, whereas for patients with a REACH score of 13+, the Kaplan-Meier incidence risk is 11.6% over 30 months, which is 2-fold higher compared with those with a REACH score below the median (5.4%).
Aspirin-treated patients with a 30-month incidence risk of a similar magnitude to a REACH score 13+ included patients with a history of vascular disease affecting 2 or more vascular beds, a history of HF, or renal insufficiency defined as an eGFR <60 ml/min. The CART analysis identified the independent higher-risk patients as those with either a history of vascular disease affecting 2 or more vascular beds, a history of HF, or a history of diabetes (Figure 1). Each of these 2 approaches (REACH and CART) identified higher-risk patient subsets who have a 2-fold higher 30-month incidence risk compared with the patient subsets without these high-risk features, although the incidence risk in those at lower risk remains ≥5% over 30 months (Table 2).
Treatment effects of rivaroxaban and aspirin
Comparing rivaroxaban and aspirin patients with aspirin alone patients, the observed hazard ratio for the main efficacy outcome is 0.75 (95% CI: 0.66 to 0.85), with a relative nonsignificant increase in severe bleeding of 1.34 (95% CI: 0.95 to 1.88) (Tables 2 and 3). Importantly, no significant interactions in the efficacy or safety effects were observed comparing subsets of patients by risk (e.g., ≥2 vascular beds affected vs. <2 beds affected) (Online Figures 1A and 1B).
Applying the REACH scoring method to identify higher-risk patients, subsets with a 30-month incidence risk score 13+ included patients with a history of ≥2 vascular beds affected, history of HF, or a lower eGFR. The proportions of patients with ≥1 high-risk feature by the REACH score accounted for 50.2% of patients, who experienced 66.5% of the vascular events. From the CART analysis, the higher-risk patients were those with ≥2 vascular beds affected, a history of HF, or a history of diabetes, and accounted for 59.6% of patients, who experienced 74.7% of the vascular events (Tables 2 and 3).
Given the higher baseline risk, the absolute risk reductions are expectedly greater in the higher-risk subsets. For example, comparing patients with ≥2 vascular beds versus 1 vascular bed, the absolute risk reduction is 6.02% versus 1.36%, which translates into 60 events prevented versus 14 events prevented per 1,000 patients treated over 30 months (Online Figure 1). In general, the number of vascular events prevented in the higher-risk patient subsets ranged from 30 to 60 per 1,000 patients treated and followed for 30 months. Aspirin-treated patients who had ≥1 high-risk feature by REACH score (i.e., 2 vascular beds affected, HF, or renal insufficiency) have a 30-month incidence risk of 11%, and considering the absolute risk reduction with rivaroxaban and aspirin treatment of 3.64% results in 36 vascular events being prevented per 1,000 patients treated for 30 months (Figure 2). However, although the event rates are lower in the lower-risk patient subsets, over time they too represent considerable residual risk. For example, among the remainder of patients (i.e., those without any high-risk REACH features), the incidence risk is lower yet substantial with a 30-month incidence risk of 5.0% in aspirin-treated patients, and treatment with rivaroxaban and aspirin results in prevention of 11 events per 1,000 patients over a 30-month period (Table 2). Patients who had ≥1 high-risk feature by CART analysis (i.e., 2 vascular beds affected, HF, or diabetes) have a 30-month incidence risk of 10.4% in the aspirin-treated patients, and with rivaroxaban and aspirin-treated patients resulted in 33 vascular events being prevented per 1,000 patients treated for 30 months. However, even in the lower-risk patient subsets, the 30-month incidence risk is 4.6% in aspirin-treated patients, and treatment with rivaroxaban and aspirin results in prevention of 10 events per 1,000 patients over a 30-month period (Table 2, Central Illustration).
For severe bleeding, the absolute risks are low overall, with a 30-month incidence risk of <1% in aspirin-treated patients (Table 3). However, when comparing aspirin-treated patients with ≥1 high-risk feature by REACH or CART to those without any high-risk features, there is a 1.7- to 1.8-fold increase in the incidence risk of severe bleeding (Table 3). Patients treated with the combination of rivaroxaban and aspirin have a numeric increase in the incidence risk of severe bleeding (an absolute risk increase) compared with aspirin-treated patients; however, these estimates are not robust given that no overall statistically significant increase in severe bleeding was observed (Table 3, Online Figure 1B).
We calculated the net clinical benefit by attributing the efficacy and safety events to the randomized treatment group at 30 months, and because the risk of bleeding events in the rivaroxaban and aspirin group appeared to occur earlier, we calculated the net clinical benefit at 12 and 24 months also. Figure 3 and Online Tables 5 and 6 show that the net clinical benefit, calculated as events prevented per 1,000 patients treated, favors rivaroxaban and aspirin combination in all patients, but is most apparent in the higher-risk subgroups, and the benefit increases the longer patients are treated with rivaroxaban and aspirin.
In the COMPASS trial, we enrolled a high-risk group of patients with coronary artery disease (CAD) or PAD. In this analysis, we explored whether further risk stratification using simple clinical criteria can refine risk, and we calculated the benefits of the rivaroxaban and aspirin combination in the higher-risk subsets. Patients with the highest risk of recurrent vascular events include those with disease in ≥2 vascular beds, a history of HF, renal insufficiency defined as an eGFR <60 ml/min, or diabetes. Individuals with ≥1 of these high-risk features (defined using the REACH or CART analysis) have a 2-fold increase in vascular events compared with patients without these high-risk features, and the combination of low-dose rivaroxaban and aspirin reduces the absolute risk of vascular complications by 3%, which is equivalent to approximately 30 events prevented for every 1,000 patients treated for 30 months.
Higher-risk compared with lower-risk patients have an increased incidence of severe bleeding, irrespective of whether they are treated with aspirin alone or the low-dose rivaroxaban and aspirin combination. Although there is a numeric increase in the incidence of severe bleeding among patients treated with low-dose rivaroxaban and aspirin compared with aspirin alone, this increase is not statistically significant. This results in a clearly favorable net clinical benefit of low-dose rivaroxaban and aspirin, which is especially apparent in higher-risk patient subsets, and increases as the duration of treatment with low-dose rivaroxaban and aspirin increases.
Taking all of this information together, a substantial number of clinical events, including CV death, MI, stroke, and vascular limb events, can be prevented by use of low-dose rivaroxaban and aspirin in the highest-risk groups. However, it is important to emphasize that in the remaining population without any high-risk features, the 30-month incidence risk of CV events is not inconsequential (5%) and the bleeding risks are low. Therefore, in addition to current use of secondary prevention treatments of vascular risk factors, treatment of these patients with low-dose rivaroxaban and aspirin can still be justified as the benefit risk ratio is also favorable, and as the data indicate, the majority of people enrolled into COMPASS benefit from combination treatment. However, in lower-risk patients, greater consideration to other factors such as patients’ values and preferences, and the cost of the treatment may be given.
Single antiplatelet therapy (most common being aspirin) is the currently recommended and the most commonly used antithrombotic therapy for patients with stable vascular disease worldwide (3,11–14). COMPASS was a large randomized clinical trial with a broad inclusion of individuals with CAD and/or PAD, with a one-quarter reduction in vascular events and a one-sixth reduction in total mortality in favor of the rivaroxaban and aspirin combination compared with aspirin alone (4,6,7). Although individuals who were deemed to be at high risk for bleeding were excluded from the COMPASS trial, a small nonsignificant absolute increase in severe bleeding was observed in patients receiving rivaroxaban and aspirin. In applying the results of the COMPASS trial, the ideal candidates to treat with the combination of rivaroxaban and aspirin include patients with the highest risk of vascular events who will have the greatest absolute benefits. In Figures 2 and 3 and the Central Illustration among high-risk patients, the absolute and relative benefits of rivaroxaban and aspirin appear to increase over time while the severe bleeding risk remains low, which indicates that the net clinical benefit increases over time, likely beyond 30 months.
We used a previously validated risk score with some minor modifications to identify patients with an increased risk of vascular events. High-risk patient subsets were identified based on having a similar 30-month risk incidence when compared with those patients who have a REACH risk score greater than the median value. Each REACH score component was also placed into a CART analysis to identify how these high-risk features independently split the data by patient risk. This process largely identified the same high-risk patient subsets. A simplified algorithm of the patients who clearly have a benefit-risk profile favoring use of the rivaroxaban-aspirin combination includes those with 1 or more high-risk features such as presence of vascular disease in 2 or more vascular beds, history of HF, renal insufficiency, or diabetes. This approach to identifying high-risk patients is parsimonious and obviates the need for a detailed risk calculator to identify these patients. Our analyses provide physicians with guidance to assess patient benefits and risks by risk groups, which can assist them in deciding who should be considered for combined therapies.
Study strengths and limitations
Risk score modeling and the choice of threshold denote levels of risk that are arbitrary and there is no consensus on the optimal methodological approach. We used a previously validated risk score (8), and after some minor modifications, compared patients above and below the median risk score of the population. We then identified which patient subgroups had equivalent incidence risks over 30 months. Our findings were largely confirmed by CART analysis (9), and the results using 2 independent methods produced similar results in identifying the highest-risk subgroups. Second, this analysis represents a subgroup analysis of a large trial. However, as the overall treatment effect with rivaroxaban and aspirin compared with aspirin of 25% is robust, it can be applied to various types of high-risk patients with reasonably high confidence. The overall increase in severe bleeding is not statistically significant, and therefore, the bleeding risk in selected subgroups should be cautiously interpreted. Finally, as previously shown in the large international REACH registry, COMPASS-eligible patients represent a substantial fraction of stable CAD/PAD patients encountered in routine clinical practice suggesting wide external applicability of our findings (15).
In high-risk patients with vascular disease, further risk stratification can identify higher-risk patients (2 or more vascular beds affected, HF, renal insufficiency, or diabetes) in whom the benefits are substantial. However, even the lower-risk patient subsets have appreciable residual risk and benefit from more intensive treatment. The absolute risk of severe bleeding is low, and the net clinical benefit remains favorable for most patients treated with rivaroxaban and aspirin compared with aspirin alone.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: Patients with vascular disease affecting ≥2 vascular beds, HF, renal insufficiency, or diabetes derive the greatest benefit from the combination of low-dose rivaroxaban plus aspirin compared with aspirin alone.
TRANSLATIONAL OUTLOOK: The safety and efficacy of dual pathway inhibition of Factor Xa and platelets should be explored in other conditions with an increased atherothrombotic risk.
This work was supported by Bayer AG. Dr. Anand is supported by a Tier 1 Canada Research Chair in Ethnicity and Cardiovascular Disease and the Michael G. DeGroote Heart and Stroke Foundation Chair in Population Health; and has received speaking honoraria and consulting fees from Bayer. Dr. Eikelboom is supported by the Jack Hirsh Population Health Research Institute in Thrombosis and Atherosclerosis Mid-career award from the Heart and Stroke Foundation; and has received grants and personal fees from Bayer, Boehringer Ingelheim, Bristol-Myers Squibb/Pfizer, Daiichi-Sankyo, Janssen, AstraZeneca, Eli Lilly, GlaxoSmithKline, and Sanofi. Dr. Bosch has received fees for participating in an Advisory Board and as an Adjudication Committee Member from Bayer. Drs. Neumann and Cook-Bruns are employees of Bayer AG. Dr. Widimsky has received occasional speaker honoraria and honoraria for the COMPASS trial National Leader role from Bayer. Dr. Avezum has received honoraria and consulting fees from Boehringer Ingelheim, Bayer, and Pfizer. Dr. Probsfield has received honoraria from Sanofi. Dr. Fox has received grants and honoraria from Bayer, Janssen, and AstraZeneca; has served as a consultant for Sanofi/Regeneron and Verseon; has received grants and personal fees from Bayer/Janssen, AstraZeneca; and has received personal fees from Sanofi/Regeneron and Lilly. Dr. Bhatt has served on the Advisory Board of Cardax, Elsevier Practice Update Cardiology, Medscape Cardiology, PhaseBio, and Regado Biosciences; has served on the Board of Directors of Boston VA Research Institute, Society of Cardiovascular Patient Care, and TobeSoft; has served as Chair of the American Heart Association Quality Oversight Committee, NCDR-ACTION Registry Steering Committee, and VA CART Research and Publications Committee; has served on Data Monitoring Committees for Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute, for the PORTICO trial, funded by St. Jude Medical, now Abbott), Cleveland Clinic, Duke Clinical Research Institute, Mayo Clinic, Mount Sinai School of Medicine (for the ENVISAGE trial, funded by Daiichi-Sankyo), and the Population Health Research Institute; has received honoraria from the American College of Cardiology (Senior Associate Editor, Clinical Trials and News, ACC.org; Vice-Chair, ACC Accreditation Committee), Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute; RE-DUAL PCI clinical trial steering committee funded by Boehringer Ingelheim), Belvoir Publications (Editor-in-Chief, Harvard Heart Letter), Duke Clinical Research Institute (clinical trial steering committees), Medtelligence/ReachMD (CME steering committees), HMP Global (Editor-in-Chief, Journal of Invasive Cardiology), Journal of the American College of Cardiology (Guest Editor; Associate Editor), Population Health Research Institute (for the COMPASS operations committee, publications committee, steering committee, and USA national co-leader, funded by Bayer), Slack Publications (Chief Medical Editor, Cardiology Today’s Intervention), Society of Cardiovascular Patient Care (Secretary/Treasurer), and WebMD (CME steering committees); has served as Deputy Editor for Clinical Cardiology; has received research funding from Abbott, Amarin, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Chiesi, Eisai, Ethicon, Forest Laboratories, Idorsia, Ironwood, Ischemix, Lilly, Medtronic, PhaseBio, Pfizer, Regeneron, Roche, Sanofi, Synaptic, and The Medicines Company; has received royalties from Elsevier (Editor, Cardiovascular Intervention: A Companion to Braunwald’s Heart Disease); has served as Site Co-Investigator for Biotronik, Boston Scientific, St. Jude Medical (now Abbott), and Svelte; has served as a Trustee of the American College of Cardiology; and has performed unfunded research from FlowCo, Merck, PLx Pharma, and Takeda. Dr. Connolly has received grants and personal fees from Bayer, Boehringer Ingelheim, Bristol-Myers Squibb/Pfizer, and Daiichi-Sankyo, Janssen, AstraZeneca, Portola, and Sanofi. Dr. Yusuf is supported by the Heart & Stroke Foundation/Marion W. Burke Chair in Cardiovascular Disease; has received research grants, honoraria, and travel expenses for lectures from Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, AstraZeneca, and Sanofi; and has received research grants from Cadila. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Listen to this manuscript's audio summary by Editor-in-Chief Dr. Valentin Fuster on JACC.org.
- Abbreviations and Acronyms
- acute limb ischemia
- heart failure
- myocardial infarction
- peripheral artery disease
- Received December 7, 2018.
- Revision received February 19, 2019.
- Accepted February 26, 2019.
- 2019 The Authors
- Narula N.,
- Dannenberg A.J.,
- Olin J.W.,
- et al.
- Bosch J.,
- Eikelboom J.W.,
- Connolly S.J.,
- et al.
- Connolly S.J.,
- Eikelboom J.W.,
- Bosch J.,
- et al.
- Anand S.S.,
- Bosch J.,
- Eikelboom J.W.,
- et al.
- Breiman L.,
- Friedman J.,
- Stone C.J.,
- Olshen R.A.
- Aboyans V.,
- Ricco J.B.,
- Bartelink M.E.L.,
- et al.
- Gerhard-Herman M.D.,
- Gornik H.L.,
- Barrett C.,
- et al.
- Fox K.A.A.,
- Eikelboom J.W.,
- Anand S.S.,
- et al.
- Darmon A.,
- Bhatt D.L.,
- Elbez Y.,
- et al.