Author + information
- Received April 26, 2011
- Revision received July 5, 2011
- Accepted July 12, 2011
- Published online October 18, 2011.
- Jung-Won Suh, MD, PhD⁎,
- Roxana Mehran, MD⁎,†,⁎ (, )
- Bimmer E. Claessen, MD⁎,
- Ke Xu, PhD⁎,
- Usman Baber, MD†,
- George Dangas, MD, PhD†,
- Helen Parise, ScD⁎,
- Alexandra J. Lansky, MD⁎,‡,
- Bernhard Witzenbichler, MD§,
- Cindy L. Grines, MD∥,
- Giulio Guagliumi, MD¶,
- Ran Kornowski, MD#,
- Jochen Wöhrle, MD⁎⁎,
- Dariusz Dudek, MD††,
- Giora Weisz, MD⁎,‡ and
- Gregg W. Stone, MD⁎,‡
- ↵⁎Reprint requests and correspondence:
Dr. Roxana Mehran, Mount Sinai Hospital, Cardiovascular Research Foundation, 111 East 59th Street, New York, New York 10022
Objectives We aimed to investigate the long-term prognosis of patients with in-hospital major bleeding (IHMB).
Background The effect of IHMB on the long-term prognosis of patients undergoing primary percutaneous coronary intervention (PCI) for ST-segment elevation myocardial infarction is unknown.
Methods Primary PCI was performed in 3,345 (92.9%) of 3,602 patients in the HORIZONS-AMI (Harmonizing Outcomes With Revascularization and Stents in Acute Myocardial Infarction) trial; in-hospital protocol-defined non–coronary artery bypass graft–related major bleeding developed in 231 (6.9%). We examined medication use at discharge, mortality, and major adverse cardiovascular events (composite of death, reinfarction, stroke, or ischemic target vessel revascularization) at 3-year follow-up in patients with and without IHMB.
Results At 3-year follow-up, patients with IHMB had higher mortality (24.6% vs. 5.4%, p < 0.0001) and major adverse cardiovascular events (40.3% vs. 20.5%, p < 0.0001). The deleterious effect of major bleeding was observed within 1 month, between 1 month and 1 year, and between 1 and 3 years. IHMB was an independent predictor of mortality (hazard ratio: 2.80; 95% confidence interval: 1.89 to 4.16, p < 0.0001) at 3-year follow up.
Conclusions Patients with IHMB after primary PCI have significantly increased 3-year rates of morbidity and mortality. Further investigation is warranted to understand the mechanisms underlying this relationship and to further improve outcomes in patients with ST-segment myocardial infarction. (Harmonizing Outcomes With Revascularization and Stents in Acute Myocardial Infarction [HORIZONS-AMI]; NCT00433966)
Although potent antiplatelet agents and early revascularization have improved the prognosis for patients with acute myocardial infarction (MI), bleeding complications may result in substantial morbidity or mortality. Major bleeding has been identified as an independent risk factor for mortality after acute MI (1,2). Previous reports have suggested that major bleeding in patients with acute coronary syndromes treated with an early invasive strategy confers an increased risk for both short-term and mid-term mortality (3–5). However, few studies have evaluated the impact of in-hospital major bleeding (IHMB) on longer-term outcomes in these patients.
Accordingly, we sought to evaluate the clinical correlates and impact of IHMB on the 3-year outcomes after primary percutaneous coronary intervention (PCI) for acute MI among randomized participants in the HORIZONS-AMI (Harmonizing Outcomes With Revascularization and Stents in Acute Myocardial Infarction) trial.
The design of the HORIZONS-AMI trial has been previously described (4). Endpoints for the current analysis included all-cause mortality, the composite endpoint of major adverse clinical events (MACE) (composite of death, reinfarction, target vessel revascularization for ischemia, or stroke), and stent thrombosis (ST, definite, or probable) at 3-year follow-up. Major bleeding was defined as the occurrence of any of the following: intracranial bleeding, intraocular bleeding, retroperitoneal bleeding, access site hemorrhage requiring surgery or radiologic or interventional procedure, hematoma ≥5 cm in diameter at the puncture site, reduction in hemoglobin concentration of ≥4 g/dl without an overt source of bleeding, reduction in hemoglobin concentration of ≥3 g/dl with an overt source of bleeding, reoperation for bleeding, or use of any blood product transfusion (4). Bleeding complications were adjudicated as either related or unrelated to coronary artery bypass (CABG) graft surgery. An independent clinical events committee adjudicated all primary endpoints and ST events throughout the 3-year follow-up period.
For the current analysis, we compared the 3-year clinical outcomes in patients with a non–CABG-related major bleeding event as their first in-hospital adverse event after the index procedure with patients who did not have major bleeding during their hospital course.
Categorical variables are presented as percentages and were compared with the chi-square test or Fisher exact test. Continuous variables are presented as medians with interquartile ranges and were compared using Mann-Whitney U test.
Predictors of IHMB were identified with logistic regression analyses. The following variables were entered: age, female sex, body mass index, hypertension, hyperlipidemia, smoking, diabetes, previous MI, previous CABG, previous angina, previous heart failure, baseline thrombocytopenia, baseline anemia, renal insufficiency (defined as estimated glomerular filtration rate <60 ml/min), use of pre-randomization heparin, antiplatelet medication use in the past 5 days, pharmacologic randomization arm, loading dose of clopidogrel, symptom to balloon time, and culprit lesion in left anterior descending artery.
Rates of clinical endpoints were compared between patients with and without IHMB cumulative to 3 years and within 3 separate time intervals (0 to 30 days, 30 days to 1 year, and 1 to 3 years). The results are displayed using Kaplan-Meier estimates and compared using the log-rank test.
Cox proportional hazards analysis was used to identify independent predictors of 3-year mortality. The multivariate model was built by stepwise variable selection with entry and exit criteria set at the p = 0.1 level, and we selected all relevant variables from previous studies as candidate variables for this model. Variables included in the multivariable model were age, sex, diabetes mellitus, Killip class, baseline anemia, white blood cell count, renal insufficiency, hypertension, hyperlipidemia, previous MI, smoking, pharmacologic randomization arm, loading dose of clopidogrel, culprit lesion in left anterior descending artery, symptom to first balloon time, and final Thrombolysis In Myocardial Infarction flow grade 3. IHMB and in-hospital ST were forced into the outcomes model.
As a secondary analysis, Cox proportional hazards analysis was repeated for in-hospital survivors only. In addition to the same variables used in the previous model, we included prescription of a statin, beta-blocker, or a thienopyridine at discharge as variables in this model. A p value <0.05 was considered statistically significant.
Primary PCI was performed in 3,345 (92.9%) of 3,602 randomized patients; in-hospital protocol-defined non–CABG-related major bleeding developed in 231 (6.9%). Thirteen cases were excluded because the bleeding events developed after thrombotic complications (reinfarction, n = 1; stroke, n = 1; ischemic target vessel revascularization, n = 2; or ST, n = 9). Table 1 shows the baseline clinical, angiographic, and procedural characteristics of patients. Patients with IHMB were older, more often female, more often had a history of congestive heart failure, had a higher Killip class, and had a lower body mass index. Moreover, they were more likely to have diabetes mellitus, anemia, and renal insufficiency. IHMB occurred less frequently in patients randomized to bivalirudin. Table 2 shows antithrombotic medication use in patients. A 600-mg loading dose of clopidogrel and bivalirudin was administered less frequently, and glycoprotein IIb/IIIa inhibitors were more often used in patients with bleeding. Independent predictors of IHMB are presented in Figure 1.
Medications use at and after discharge
As shown in Table 2, there were no significant differences in the prescription rate of aspirin between patients with and without IHMB at discharge, 1 month, and 3 years. At 1 year, however, patients with bleeding were less likely to be taking aspirin. There were no significant differences in thienopyridine use between both groups at discharge, 1 year, and 2 years. However, patients with IHMB were more likely to be taking a thienopyridine at 3 years. Patients with IHMB were less likely to receive a beta-blocker and statin at hospital discharge, but were more likely to receive diuretics, digoxin, and antiarrhythmic agents.
Major bleeding and clinical outcomes
Kaplan-Meier estimates of the incidence of MACE and their individual components at 3 years and between 0 to 30 days, 30 days to 1 year, and 1 to 3 years are shown in Table 3. Patients with IHMB had significantly higher 3-year rates of mortality (24.6% vs. 5.4%, p < 0.0001) (Fig. 2A) and MACE (40.3% vs. 20.5%, p < 0.0001) (Fig. 2B).
Moreover, the rates of mortality and MACE were significantly higher in patients with IHMB within each time interval (Fig. 3). The 3-year cumulative rates of MI, ischemic target vessel revascularization, and stroke were also significantly higher in patients with IHMB. Rates of stroke were significantly higher in the bleeding group within 30 days and between 30 days to 1 year, but not thereafter (Table 3). The 3-year cumulative rates of ST were not significantly increased in patients with compared to those without IHMB.
Predictors of long-term mortality
IHMB was an independent predictor of 3-year mortality (hazard ratio [HR]: 2.80, 95% confidence interval [CI]: 1.89 to 4.16; p < 0.0001). Other independent predictors included in-hospital ST, age, male sex, diabetes mellitus, renal insufficiency, smoking, Killip classes 2 to 4 at admission, anemia, baseline white blood cell count, left anterior descending artery culprit lesion, and final Thrombolysis In Myocardial Infarction flow grade <3 (Table 4).
In the secondary analysis among in-hospital survivors (n = 3,260), IHMB remained an independent predictor of 3-year mortality (HR: 2.26, 95% CI: 1.43 to 3.54; p = 0.0004), while in-hospital ST was not a significant predictor (HR: 0.95; 95% CI: 0.13 to 6.93).
In the present analysis from the HORIZONS-AMI trial, we found that the occurrence of IHMB in ST-segment elevation myocardial infarction patients treated with primary PCI confers a sustained risk of both mortality and MACE for at least 3 years. We also identified several important differences in pharmacotherapy at the time of discharge between patients with and without IHMB that may in part explain the effect on late mortality.
Although the incidence of both mortality and MACE among bleeding patients was greatest in the first 30 days following PCI, the risk of late mortality after IHMB continued to monotonically accrue over time. While the deleterious impact of thrombotic complications is principally evidenced in the acute stage after their occurrence (6), our findings do not suggest a similar attenuation of risk over time in patients with IHMB in the setting of STEMI. Indeed, IHMB was associated with an approximate 3.5-fold increased risk of mortality between 1 and 3 years.
In this study, IHMB was adjudicated according to the protocol definition used in the HORIZONS-AMI and ACUITY (Acute Catheterization and Urgent Intervention Triage Strategy) trials (3,7). This report, therefore, validates the definition of major bleeding used in these studies as being clinically relevant and useful in predicting long-term prognosis.
The incidence of IHMB in the present study is higher than that reported from earlier analyses in other acute coronary syndrome studies (1,8). Differences in patient populations, treatment strategies, and bleeding definitions likely explain these discrepancies. Using the same protocol definition, the rate of IHMB among acute coronary syndrome patients treated with PCI in the ACUITY trial was 5.9% (9).
IHMB was associated with 3-year mortality and MACE after the exclusion of patients with large hematomas alone, and the influence of single large hematoma on subsequent mortality or MACE seems to be negligible, as in previous reports (10,11).
Bleeding complications after acute MI may have detrimental effects on long-term prognosis due to several reasons. Patients with IHMB were discharged less often on beta-blocker and statin therapy, which are known to have a survival benefit after acute MI (12). Interestingly, we did not find a reduction in the use of antiplatelet agents in patients with IHMB, as has been reported in the PREMIER (Prospective Registry Evaluating Myocardial Infarction: Events and Recovery) registry (13). We speculate that this may be due to mandated, protocol-defined instructions to keep patients on dual antiplatelet therapy at discharge and for up to at least 1 year. Greater comorbidities in bleeding patients might have also contributed to higher cardiovascular risk. IHMB remained a strong and independent predictor of subsequent mortality, however, even after adjusting for these differences. Other mechanisms that might heighten cardiac risk after bleeding include hypovolemia, hypotension, anemia, impaired oxygen-carrying capacity, and inflammatory response caused by transfusion (3). These factors generate multiple hypotheses for future investigation.
The current analysis was not pre-specified in the original study. Multivariate analysis might not account for unmeasured confounders that are associated with mortality after acute MI.
IHMB confers an independent and sustained risk of both mortality and MACE in patients with ST-segment elevation myocardial infarction treated with primary PCI. Further investigation is warranted to understand the mechanisms underlying this relationship and whether the prognosis of these patients may be improved by tailored follow-up and detailed attention to adjunctive pharmacotherapy.
For supplementary figures, please see the online version of this article.
The main study (HORIZON-AMI) of this post-hoc analysis was supported by the Cardiovascular Research Foundation with grant support from Boston Scientific and the Medicines Company. Dr. Mehran reported receiving lecture fees from Boston Scientific and the Medicines Company; is a consultant for Abbott Vascular, AstraZeneca, and Regado Biosciences; and has received research support from BMS/Sanofi-Aventis. Dr. Dangas has received an institutional research grant from The Medicines Co., Sanofi-Aventis, and Bristol-Myers Squibb; has received speaker honoraria from Cordis; and his spouse is on the advisory board for Abbott and AstraZeneca. Dr. Witzenbichler has received lecture fees from The Medicines Co. and Boston Scientific. Dr. Grines is on the advisory board for Abbott Vascular. Dr. Guagliumi has received consulting fees from or serving on advisory boards for Abbott Vascular and Boston Scientific; and has received grant support from Medtronic and Boston Scientific. Dr. Stone has received consulting fees from Medtronic, GlaxoSmithKline, Eli Lilly-Daiichi Sankyo, Merck, AstraZeneca, Boston Scientific, Abbott Vascular, The Medicines Co., and Bristol-Myers Squibb-Sanofi; and has received grant support from Boston Scientific, the Medicines Company, and Abbott Vascular. Dr. Dudek has received research grants or served as a consultant/advisory board member for Abbott, Adamed, AstraZeneca, Biotronik, Balton, Bayer, BBraun, BioMatrix, Boston Scientific, Boehringer Ingelheim, Bristol-Myers Squibb, Cordis, Cook Eli Lilly, EuroCor, Glaxo, Invatec, Medtronic, The Medicines Co., MSD, Nycomed, Orbus-Neich, Pfizer, Possis, Promed, Sanofi-Aventis, Siemens, Solvay, Terumo, and Tyco.
All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- confidence interval
- hazard ratio
- in-hospital major bleeding
- major adverse cardiac event(s)
- myocardial infarction
- percutaneous coronary intervention
- stent thrombosis
- Received April 26, 2011.
- Revision received July 5, 2011.
- Accepted July 12, 2011.
- American College of Cardiology Foundation
- Eikelboom J.W.,
- Mehta S.R.,
- Anand S.S.,
- Xie C.,
- Fox K.A.,
- Yusuf S.
- Manoukian S.V.,
- Feit F.,
- Mehran R.,
- et al.
- Ndrepepa G.,
- Berger P.B.,
- Mehilli J.,
- et al.
- Mehran R.,
- Pocock S.J.,
- Stone G.W.,
- et al.
- Rao S.V.,
- Eikelboom J.A.,
- Granger C.B.,
- Harrington R.A.,
- Califf R.M.,
- Bassand J.P.
- Mehran R.,
- Pocock S.J.,
- Nikolsky E.,
- et al.
- White H.D.,
- Aylward P.E.,
- Gallo R.,
- et al.
- Wang T.Y.,
- Xiao L.,
- Alexander K.P.,
- et al.