Author + information
- Received May 19, 2014
- Revision received June 21, 2014
- Accepted July 1, 2014
- Published online October 7, 2014.
- Martial Hamon, MD∗,†,
- Gilles Lemesle, MD‡,§,‖,
- Olivier Tricot, MD¶,
- Thibaud Meurice, MD#,
- Michel Deneve‡,
- Xavier Dujardin, MD∗∗,
- Jean Michel Brufau, MD††,
- Jerome Bera, MD#,
- Nicolas Lamblin, MD‡,§,‖ and
- Christophe Bauters, MD‡,§,‖∗ ()
- ∗Centre Hospitalier Universitaire de Caen, Caen, France
- †Faculté de Médecine de Caen, Caen, France
- ‡Centre Hospitalier Régional et Universitaire de Lille, Lille, France
- §Inserm U744, Institut Pasteur de Lille, Université Lille Nord de France, Lille, France
- ‖Faculté de Médecine de Lille, Lille, France
- ¶Centre Hospitalier de Dunkerque, Dunkerque, France
- #Polyclinique du Bois, Lille, France
- ∗∗Centre Hospitalier de Boulogne-Sur-Mer, Boulogne-Sur-Mer, France
- ††Polyclinique Vauban, Valenciennes, France
- ↵∗Reprint requests and correspondence:
Dr. Christophe Bauters, Hôpital Cardiologique, CHRU de Lille, Boulevard Prof. Leclercq, 59037 Lille Cedex, France.
Background Although there is evidence that patients who experience major bleeding after an acute coronary event are at higher risk of death in the months after the event, the incidence and impact on outcome of bleeding beyond 1 year of follow-up in patients with stable coronary artery disease (CAD) are largely unknown.
Objectives The goal of this study was to assess the incidence, source, determinants, and prognostic impact of major bleeding in stable CAD.
Methods We prospectively included 4,184 consecutive CAD outpatients who were free from any myocardial infarction (MI) or coronary revascularization for >1 year at inclusion. Follow-up was performed at 2 years, with major bleeding defined as a type ≥3 bleed using the Bleeding Academic Research Consortium (BARC) definition.
Results There were 51 major bleeding events during follow-up (0.6%/year). Most events were BARC type 3a bleeds with 12 fatal bleeds (type 5). In most cases (54.9%), the site of bleeding was gastrointestinal. Major bleeding was significantly associated with mortality (adjusted hazard ratio: 2.89; 95% confidence intervals: 1.73 to 4.83; p < 0.0001). The increased risk of bleeding associated with vitamin K antagonist (VKA) treatment was particularly evident when VKA was combined with an antiplatelet therapy (APT). In contrast, the risk of cardiovascular death, MI, or nonhemorrhagic stroke did not differ in patients who received VKA + APT versus patients on VKA alone.
Conclusions In patients with stable CAD (i.e., >1 year, with no acute events), major bleeding events are rare, but such events are an independent predictor of death. When oral anticoagulation is required, concomitant APT should not be prescribed in the absence of a recent cardiovascular event.
- antiplatelet agent
- coronary artery disease
- major bleeding
- oral anticoagulant
- vitamin K antagonist
Bleeding has emerged as an important outcome in patients with coronary artery disease (CAD). There is evidence that patients who experience major bleeding when presenting with acute coronary syndrome or who are undergoing percutaneous coronary intervention are at a higher risk of death in the following months (1–5). Several mechanisms explain the deleterious effects of major bleeding, including the cessation of antithrombotic therapies, which induces an increased risk of acute stent thrombosis, myocardial infarction (MI), and cardiovascular death, blood transfusion, the prevalence of comorbidities in patients who bleed, and the deleterious effect of anemia (6–8). In contrast, there are limited data on the incidence, source, and prognostic impact of bleeding beyond 1 year of follow-up (i.e., in patients with stable CAD). However, these patients, who may be very long past or who may never have sustained an acute event, constitute the majority of CAD patients in the outpatient setting. We therefore designed the present study to assess the importance of bleeding and its impact on outcomes in patients with stable CAD who are at least 1 year past any coronary events (any MI or revascularization procedure).
The CORONOR (Suivi d’une cohorte de patients COROnariens stables en region NORd-Pas-de-Calais) study was a prospective multicenter study that enrolled 4,184 consecutive outpatients with stable CAD (9). Patients were prospectively included between February 1, 2010 and April 30, 2011 by 50 cardiologists. The cardiologists were selected on the basis of geographic distribution to provide a representative sample of the area’s current cardiology practices in public universities, public institutions, and private centers. To be eligible for the CORONOR study, each patient had to fulfill the inclusion criteria. There were no exclusion criteria. This study was approved by the French medical data protection committee CCTIRS (Comité consultatif sur le traitement de l'information en matière de recherche dans le domaine de la santé) and authorized by the Commission nationale de l’informatique et des libertés for the treatment of personal health data. All patients consented to the study after being informed in writing of the study’s objectives and treatment of data, as well as of their rights to object, of access, and of rectification.
Inclusion criteria of the coronor study
Patients were categorized with stable CAD if they had documented CAD and were free from any MI and/or any coronary revascularization (either percutaneous coronary intervention or coronary artery bypass graft surgery) for at least 1 year at inclusion. Documented CAD was defined as all patients with a history of MI, history of coronary revascularization, and/or the presence of a coronary stenosis >50% on a coronary angiogram. To represent the real-life spectrum of stable CAD, patients with other cardiovascular or noncardiovascular illnesses or comorbidities were not excluded from the study.
At the initial visit, attending physicians prospectively completed the case record forms, which contained information regarding demographic and clinical characteristics of the patients, including usual cardiovascular risk factors and treatments. During the outpatient visit, the investigators reviewed the patients' current drug treatment and entered all prescribed drugs on the case record form.
Objective, follow-up, definitions, and endpoints
The objectives of this analysis were to assess the incidence, source, determinants, and prognostic impact of major bleeding in patients with stable CAD. A 2-year clinical follow-up was performed at outpatient visits or by contacting the general practitioner. We collected data on death, MI, stroke, and major bleeding. All clinical events were adjudicated blindly by 2 investigators, or by 3 investigators in case of disagreement, according to pre-specified definitions. The cause of death was determined after a detailed review of the circumstances of death and was classified as cardiovascular or noncardiovascular. Deaths from unknown causes were considered cardiovascular. Bleeding events were classified using the Bleeding Academic Research Consortium (BARC) definition (10). Information on BARC type 1 and 2 bleeds was not available in our registry. Although not in the BARC definitions, for the purpose of this study, we arbitrarily defined major bleeding as all BARC type ≥3 events. To assess ischemic risk, a composite endpoint was defined as cardiovascular death, MI, or nonhemorrhagic stroke.
Continuous variables are expressed as mean ± SD. Categorical variables are expressed as absolute numbers and percents. We used the chi-square test or the Fisher exact test for the comparison of categorical variables and the Student unpaired t test for the comparison of continuous variables as appropriate. Cumulative event rates were estimated using the Kaplan-Meier method and compared using the log-rank test. Cox proportional hazard analyses were performed to calculate hazard ratios (HRs) and 95% confidence intervals (CIs). For each variable, the proportional hazards assumption was tested visually using Kaplan-Meier curves and by examining a plot of –ln [–ln(survival time)] against the ln(time). In addition, the proportional hazard was assessed and satisfied by including an interaction time-dependent term in the multivariable Cox regression analysis. All statistical analyses were performed using STATA 9.2 software (STATA Corporation, College Station, Texas). Statistical significance was assumed at p value <0.05.
The baseline characteristics of the patients included in the CORONOR study have been reported previously (9) and are summarized in Table 1. Most patients were men, with a mean age of 66.9 ± 11.5 years. Most patients underwent at least 1 coronary revascularization procedure before inclusion. There was wide prescription of major secondary prevention drugs (i.e., beta-blockers, statins, and angiotensin system antagonists). Almost all patients received at least 1 antithrombotic treatment (99.3%). The majority received antiplatelet therapy (APT) alone (n = 2,798; 67.4%; aspirin, n = 2,025; clopidogrel, n = 773). Although these patients had no recent coronary events, a substantial proportion received dual antiplatelet treatment (i.e., aspirin and clopidogrel; n = 861; 20.8%). Altogether, 11.1% received a vitamin K antagonist (VKA). Among these patients, most (n = 342) received the combination of a VKA and an antiplatelet agent (VKA + APT) (aspirin, n = 308; clopidogrel, n = 34), whereas 119 received a VKA alone. Of note, no patient received prasugrel, ticagrelor, or new oral anticoagulants at inclusion.
At follow-up, there were 271 deaths (119 cardiovascular and 152 noncardiovascular), 91 MIs, and 48 strokes. The combined endpoint of cardiovascular death, MI, or nonhemorrhagic stroke occurred in 226 patients. There were 51 major bleeding events during the 2-year follow-up (0.6 per 100 patient-years). As shown in Table 2, most events were BARC type 3a bleeds; there were 12 fatal bleeds (type 5). In most of the cases, the site of bleeding was gastrointestinal (54.9%). The site of fatal bleeding was intracranial in 6 patients, gastrointestinal in 4 patients, retroperitoneal in 1 patient, and pericardial in 1 patient. Figure 1 illustrates the timing of the bleeding events and shows that the risk of bleeding was constant throughout the follow-up. During the follow-up, 6 additional deaths occurred among the patients who experienced major bleeding, giving a total of 18 (35.3%) deaths. The time interval between the bleeding event and death was 12 ± 19 days for the 12 patients with type 5 bleeding, and 174 ± 142 days in the 6 patients who died after type 3 to 4 bleeding. When used as a time-dependent variable, major bleeding was associated with a significant increase in mortality (unadjusted HR: 6.78; 95% CI: 4.2 to 10.94; p < 0.0001); major bleeding remained significantly associated with mortality (HR = 2.89; 95% CI 1.73 to 4.83; p < 0.0001) when adjusted for age, sex, diabetes mellitus, previous heart failure, and the estimated glomerular filtration rate (eGFR).
The patients who experienced major bleeding were older and more frequently had angina, a history of hypertension, diabetes mellitus, and multivessel CAD. The patients who bled also more frequently underwent drug-eluting stent implantation, aortic or peripheral intervention, were hospitalized for decompensated heart failure, had a lower left ventricular ejection fraction and eGFR, more often received VKA, and were less frequently on statin therapy (Table 1). As shown in Table 3, 4 variables were identified as independent predictors of major bleeding by multivariate analysis: treatment with VKA, diabetes mellitus, age (positive association), and eGFR (negative association). The impact of antithrombotic treatments on the risk of major bleeding is shown in Figure 2. The increased bleeding risk associated with VKA treatment was particularly evident in patients who received the combination of VKA + APT, whereas there was no significant increase in bleeding in patients who received VKA alone. In contrast, as shown in Figure 3, there was no difference in the risk of the combined endpoint of cardiovascular death, MI, or nonhemorrhagic stroke in patients treated with VKA + APT versus patients treated with VKA alone (adjusted HR: 1.15; 95% CI: 0.58 to 2.27; p = 0.697).
There is now clear evidence that bleeding is a key event with important prognostic implications in patients with acute manifestations of CAD (1–3) and in patients undergoing myocardial revascularization procedures (4,5). Much less is known regarding bleeding events that occur in patients with stable and chronic CAD at a chronological distance from any MI or revascularization. Data from the CHARISMA (Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance) trial provided meaningful information on bleeding rates according to antiplatelet regimens (11). However, because the patients included in randomized trials are highly selected (patients who required VKA or patients believed to be at particularly high risk of bleeding were excluded from the CHARISMA study), the ability to extrapolate these results to the overall stable CAD population is limited. An analysis of the REACH (Reduction of Atherothrombosis for Continued Health) registry reported risk factors for serious bleeding in a large group of outpatients with various manifestations of atherothrombosis (CAD, cerebrovascular disease, peripheral arterial diseases) or with at least 3 atherosclerosis risk factors in 2003 to 2004 (12). However, to the best of our knowledge, the present study is the first to focus on a contemporary cohort of patients with stable CAD and the first to provide information on major bleeding in a real-life setting.
Our results showed that a BARC type ≥3 bleed is a relatively rare event in stable CAD patients (0.6%/patient-year). It is, however, important to point out that this rate applies to a study population included in an outpatient setting, and that all patients were >1 year from a cardiovascular hospitalization for MI or coronary revascularization at inclusion in the study. The most frequent finding was the incidence of major bleeding episodes related to gastrointestinal bleeding. The data also demonstrated that the occurrence of major bleeding in a stable CAD patient was an independent predictor of death. All physicians following stable CAD patients should, therefore, pay special attention to the determinants of bleeding. Several independent risk factors for major bleeding were identified. As observed in patients with unstable CAD (2,13), increasing age, diabetes, and renal insufficiency were associated with an increased risk of bleeding in stable CAD. Long-term oral anticoagulation was the strongest risk factor for bleeding in this stable setting; in particular, we found that the combination of oral anticoagulation with an antiplatelet agent was associated with a major increase in the number of bleeding episodes.
A significant proportion of patients with stable CAD are treated with oral anticoagulation (>10% in the present study), mainly because of atrial fibrillation. Although common practice is to treat such patients with VKA + 1 APT drug (usually aspirin), the 2010 European Society of Cardiology guidelines on atrial fibrillation suggest that, in patients with stable vascular disease (i.e., >1 year with no acute events), VKA monotherapy may be considered, and that in the absence of a subsequent cardiovascular event, concomitant APT should not be prescribed (14). This was on the basis of evidence that the addition of an APT is associated with a substantial increase in bleeding, with no clear evidence for a decrease in ischemic events (15–17). It should be noted, however, that data in patients with stable CAD and modern management (i.e., wide use of coronary revascularization, including drug-eluting stent implantation) were lacking. The American College of Cardiology and the American Heart Association endorsed a similar statement (18). Recruitment in the present study was performed at the time these guidelines were released. We observed that three-quarters of patients who received a VKA were also treated with an antiplatelet agent; this is concordant with the recent EuroObservational Research Programme Atrial Fibrillation Pilot Registry report, in which APT was still commonly prescribed with oral anticoagulation when there was coexistent CAD (19). This illustrates that physicians remain hesitant to stop all APTs in stable CAD patients who need oral anticoagulation. However, our follow-up data suggest that this attitude leads to an increased risk of bleeding that does not appear to be balanced by a concomitant decrease in the risk of ischemic complications.
Our data reflect the practice in a regional area in 2010 to 2011. Although this was not a population-based registry, the patients included were treated at research institutions and smaller community hospitals, as well as in private practice. It will have to be determined whether these findings are representative of practices in other parts of the world, and how these trends could change in the era of new drug-eluting stents that lower the risk of thrombosis. In the present study, anticoagulant therapy was limited to VKA, and antiplatelet agents were limited to aspirin and/or clopidogrel; therefore, our results cannot be directly extrapolated to novel oral anticoagulants and antiplatelet agents. In addition, our study only focused on BARC ≥3 bleeds; thus, less severe events that could be associated with prognosis were not analyzed. Further studies are needed to clarify the association of the different BARC classes with outcomes in the outpatient setting, as was recently reported for patients with ST-segment elevation MI (20).
As with any observational study, the association between treatment and outcomes cannot prove causality. The limitation of adjustment methodology is also acknowledged, and persistent unknown or unmeasured confounding factors may exist. Finally, we also acknowledge that our study had limited power to compare patients receiving VKA alone with patients receiving VKA + APT.
In patients with stable CAD, major bleeding events are rare, but they are an independent predictor of death (Central Illustration). When oral anticoagulation is required, the use of concomitant APT is associated with an important increase in bleeding, with no evidence of a decrease in ischemic events. For the clinician, the important message is to avoid long-term use of antiplatelet agents in combination with oral anticoagulation in stable CAD patients, and to consider co-prescription of a proton-pump inhibitor because many events in these patients are related to gastrointestinal bleeding. Lastly, it should be emphasized that our findings should not be extrapolated to patients within 1 year of MI or coronary revascularization.
COMPETENCY IN MEDICAL KNOWLEDGE: Antiplatelet drugs increase the risk of bleeding, but have little or no impact on the risk of ischemic events in patients with CAD who receive chronic anticoagulation and who have not experienced an acute ischemic event or undergone revascularization for >1 year. Major bleeding occurs rarely in patients with stable ischemic heart disease, but it does have important prognostic implications for mortality.
TRANSLATIONAL OUTLOOK: The relative risks and benefits of prolonged APT should be periodically reassessed in patients with stable ischemic heart disease who receive chronic anticoagulation, and additional studies are needed to delineate predictors of adverse outcomes.
The Fédération Française de Cardiologie, Paris, France, supported this study. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- antiplatelet therapy
- Bleeding Academic Research Consortium
- coronary artery disease
- confidence intervals
- estimated glomerular filtration rate
- hazard ratio
- myocardial infarction
- vitamin K antagonist
- Received May 19, 2014.
- Revision received June 21, 2014.
- Accepted July 1, 2014.
- American College of Cardiology Foundation
- Eikelboom J.W.,
- Mehta S.R.,
- Anand S.S.,
- et al.
- Steg P.G.,
- Huber K.,
- Andreotti F.,
- et al.
- Collet J.P.,
- Montalescot G.,
- Blanchet B.,
- et al.
- Silvain J.,
- Pena A.,
- Cayla G.,
- et al.
- Sabatine M.S.,
- Morrow D.A.,
- Giugliano R.P.,
- et al.
- Mehran R.,
- Rao S.V.,
- Bhatt D.L.,
- et al.
- Berger P.B.,
- Bhatt D.L.,
- Fuster V.,
- et al.,
- for the CHARISMA Investigators
- Ducrocq G.,
- Wallace J.S.,
- Baron G.,
- et al.,
- for the REACH Investigators
- Moscucci M.,
- Fox K.A.,
- Cannon C.P.,
- et al.
- European Heart Rhythm Association; European Association for Cardio-Thoracic Surgery,
- Camm A.J.,
- Kirchof P.,
- Lip G.Y.,
- et al.
- DiMarco J.P.,
- Flaker G.,
- Waldo A.L.,
- et al.,
- AFFIRM Investigators
- Flaker G.C.,
- Gruber M.,
- Connolly S.J.,
- et al.,
- for the SPORTIF Investigators
- Lip G.Y.
- Fuster V.,
- Ryden L.E.,
- Cannom D.S.,
- et al.
- Kikkert W.J.,
- van Geloven N.,
- van der Laan M.H.,
- et al.