Author + information
- Received January 17, 2017
- Revision received February 27, 2017
- Accepted March 9, 2017
- Published online June 5, 2017.
- Ion S. Jovin, MD, ScDa,b,∗ (, )
- Rachit M. Shah, MBBSa,
- Dhavalkumar B. Patel, MBBS, MPHa,
- Sunil V. Rao, MDc,
- Dmitri V. Baklanov, MDd,
- Issam Moussa, MDe,
- Kevin F. Kennedy, MSd,
- Eric A. Secemsky, MD, MScf,g,
- Robert W. Yeh, MD, MScf,
- Michael C. Kontos, MDa and
- George W. Vetrovec, MDa
- aVirginia Commonwealth University, Richmond, Virginia
- bMcGuire VAMC, Richmond, Virginia
- cDuke University, Durham, North Carolina
- dMid America Heart Institute, Kansas City, Missouri
- eRutgers Robert Wood Johnson Medical School, New Brunswick, New Jersey
- fHarvard Medical School, Beth Israel Deaconess Medical Center, Boston, Massachusetts
- gMassachusetts General Hospital, Boston, Massachusetts
- ↵∗Address for correspondence:
Dr. Ion S. Jovin, Virginia Commonwealth University/McGuire VAMC, 1201 Broad Rock Boulevard, 111J, Richmond, Virginia 23249.
Objectives The aim of this study was to compare bivalirudin with heparin as anticoagulant agents in patients with ST-segment elevation myocardial infarction treated with radial primary percutaneous coronary intervention (PCI).
Background Recent studies in which PCI was performed predominantly via radial access did not show bivalirudin to be superior to heparin.
Methods Outcomes were compared in patients with STEMI included in the National Cardiovascular Data Registry CathPCI database from 2009 to 2015 who underwent primary PCI via radial access and who were anticoagulated with bivalirudin or heparin.
Results The sample included 67,368 patients, of whom 29,660 received bivalirudin and 37,708 received heparin. The 2 groups of patients did not differ significantly in their mean age or percentage of men. The unadjusted comparison showed no significant difference in the rate of the composite endpoint of death, myocardial infarction, or stroke (4.6% vs. 4.7%; p = 0.47) and a significantly higher rate of acute stent thrombosis (1.00% vs. 0.60%; p < 0.001) with bivalirudin compared with heparin. After adjusting for multiple variables, including a propensity score reflecting the probability of receiving bivalirudin, the odds ratio of the composite endpoint of death, myocardial infarction, or stroke for bivalirudin versus heparin was 0.95 (95% confidence interval: 0.87 to 1.05; p = 0.152), and the odds ratio for acute stent thrombosis was 2.11 (95% confidence interval: 1.73 to 2.57) for bivalirudin versus heparin. Major bleeding rates were not significantly different.
Conclusions In patients undergoing primary PCI via transradial access anticoagulated with bivalirudin or heparin, there was no difference in the composite endpoint of death, myocardial infarction, or stroke.
- percutaneous coronary intervention
- ST-segment elevation myocardial infarction
The use of the direct thrombin inhibitor bivalirudin during percutaneous coronary intervention (PCI) has proved to be an effective alternative to standard unfractionated heparin therapy (alone or in combination with glycoprotein [GP] IIb/IIIa inhibitors) with similar ischemic adverse events but with a reduction in major bleeding (1). In the setting of primary PCI for ST-segment elevation myocardial infarction (STEMI), the HORIZONS-AMI (Harmonizing Outcomes With Revascularization and Stents in Acute Myocardial Infarction) trial showed that the use of bivalirudin was associated with a significant reduction in bleeding events and also mortality at 30 days (2) and until 3-year follow-up (3) compared with patients treated with GP IIb/IIIa inhibitors plus heparin. However, most of the interventional procedures included in HORIZONS-AMI were performed through the transfemoral approach, and data on patients who underwent treatment through the transradial approach are sparse (4,5). The transradial approach for PCI has been associated with a significant reduction in bleeding complications and in some studies also with a significant reduction in mortality in patients with STEMI (6,7) compared with the transfemoral approach. The improved outcomes with bivalirudin have been questioned by more recent studies, such as EUROMAX (European Ambulance Acute Coronary Syndrome Angiography Trial) (8), which enrolled more patients treated via the radial access, and also by 3 other recent studies (HEAT-PPCI [Unfractionated Heparin Versus Bivalirudin in Primary Percutaneous Coronary Intervention] , NAPLES [Novel Approaches for Preventing or Limiting Events] III , and BRAVE 4 [Bavarian Reperfusion Alternatives Evaluation] ). In contrast, a network meta-analysis found an improved survival in patients treated with bivalirudin compared with those treated with heparin (12). Finally, 2 recently published studies in all-comers showed no benefit with bivalirudin compared with heparin when PCI is performed via radial access (13,14). Thus, whether bivalirudin is truly superior to heparin with or without GP IIb/IIIa inhibitors and can reduce bleeding events and improve outcomes in patients treated through the transradial approach during primary PCI is still not well understood (15,16).
The aim of this study was to compare the outcomes of patients with acute STEMI who underwent primary PCI through the transradial approach and who were anticoagulated with either bivalirudin or with heparin with or without GP IIb/IIIa inhibitors.
The National Cardiovascular Data Registry (NCDR) CathPCI registry, which is cosponsored by the American College of Cardiology and the Society for Cardiovascular Angiography and Interventions, has been previously described (17,18). The CathPCI registry collects data on patient and hospital characteristics, clinical presentation, treatments, and outcomes for PCI procedures from more than 1,800 sites across the United States. Data are entered into NCDR-certified software at participating institutions and exported in a standard format to the American College of Cardiology. There is a comprehensive data quality program, including both data quality report specifications for data capture and transmission and an auditing program. A comprehensive description of NCDR data elements and definitions is available at http://www.ncdr.com/webncdr/cathpci/home/datacollection.
The study population consisted of patients with STEMI treated with immediate primary PCI from July 1, 2009, to September 30, 2015. Patients were excluded if they underwent PCI through access of a nonfemoral and nonradial artery (i.e., ulnar, brachial), if they had unknown data on medications used for anticoagulation, or if they were treated with thrombolytic therapy. After these exclusions, 622,709 patients from 1,584 sites were included in the analysis of outcomes according to access site. The primary analysis consisted of 74,502 patients with STEMI accessed via the radial artery for primary PCI from 1,328 sites.
Patient and hospital characteristics were compared by approach (radial vs. femoral and overall). Patient characteristics, including demographics, history and risk factors, coronary anatomy, PCI procedure, lesions and devices, laboratory values, intraprocedure and post-procedure events, and hospital characteristics, were compared. The primary endpoint was a composite of in-hospital death, myocardial infarction (MI), and stroke. Secondary endpoints were bleeding and bleeding requiring transfusion, length of stay, and acute stent thrombosis. Acute stent thrombosis was defined as repeat PCI on the previously implanted stent during the index admission. Categorical variables are presented as frequency (percentage), and differences between the 2 groups were assessed using the chi-square test or the Wilcoxon summed rank test. Continuous variables are presented as mean ± SD and were compared using the Student t test.
To assess the impact of bivalirudin (vs. heparin) on outcomes, we used a hierarchical multivariate model clustered by site with a random intercept. To account for patient differences between groups, we developed a propensity score to receive bivalirudin using a saturated logistic regression model. This model included a variety of demographic, history, and risk factors (age, sex, race, insurance, glomerular filtration rate, smoking, hypertension, dyslipidemia, family history of coronary artery disease, prior MI, prior heart failure, valve surgery, prior PCI, prior coronary artery bypass grafting, current dialysis, prior cardiovascular disease, prior peripheral artery disease, claudication, diabetes mellitus, antianginal medications, heart failure in the previous 2 weeks, prior shock, prior cardiac arrest, and GP IIb/IIIa inhibitors). The C statistic for the propensity score was 0.70 (p < 0.001), suggesting good discriminatory power for the score. In addition to adjusting for the propensity to receive bivalirudin as a covariate, we additionally adjusted for the published models reflecting probability of bleeding (19) and of mortality (20) as derived from the CathPCI registry. Additionally, to account for the temporal changes over the study cohort, we tested the interaction between bivalirudin (vs. heparin) and year of procedure. Also, because there was a suggestion of the effect of bivalirudin being related to less use of GP IIb/IIIa inhibitors, we analyzed the temporal trends of GP IIb/IIIa inhibitors. Finally, we performed a sensitivity analysis in which we compared patients receiving bivalirudin with and without GP IIb/IIIa inhibitors with those receiving heparin with and without GP IIb/IIIa inhibitors. The final comparison between bivalirudin and heparin is shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p value <0.05 was considered to indicate statistical significance. SAS version 9.4 (SAS Institute, Cary, North Carolina) was used for all analyses.
The number of patients included in the study and the exclusion criteria are depicted in Figure 1. The characteristics of the 622,709 patients with STEMI in the NCDR database undergoing primary PCI via femoral and radial access between the third quarter of 2009 and the third quarter of 2015 are depicted in Table 1. Only 67,368 patients (10.8%) underwent the procedure via the radial approach. These patients tended to be younger and more often male and were less likely to have prior coronary artery bypass grafting, MI, or PCI. They were also less likely to present with shock or cardiac arrest.
The clinical characteristics of the 67,368 patients who underwent primary PCI via the radial approach are detailed in Table 2. Of these patients, 29,660 (44%) were anticoagulated with bivalirudin, and the remaining 42,158 received unfractionated heparin. There were no significant differences in the age or proportion of men between the 2 groups. Patients who received bivalirudin were more likely to not have insurance, more likely to be on antianginal medications, and less likely to have histories of recent heart failure or to have experienced cardiac arrest in the previous 24 hours.
The procedural characteristics of the 2 groups are shown in Table 3. The patients who received bivalirudin received a slightly but statistically significant higher dose of contrast, were markedly less likely to receive GP IIb/IIIa inhibitors or clopidogrel, and were more likely to receive prasugrel as an antiplatelet agent than patients who received heparin. The use of GP IIb/IIIa inhibitors decreased markedly over time in patients treated with heparin (from 79% in 2009 to 47% in 2016), whereas it remained relatively unchanged in patients treated with bivalirudin (from 23% to 21%, respectively). Because heparin is often used in the emergency department in patients with STEMI and in the cardiac catheterization laboratory in the spasmolytic cocktail administered during radial procedures, two-thirds of the patients anticoagulated with bivalirudin during PCI received a bolus of heparin beforehand.
The unadjusted outcomes are shown in Table 4. The main composite outcome of death, MI, and stroke was not significantly different in the bivalirudin group compared with the heparin group (4.6% vs. 4.7%, respectively, p = 0.47) although there were fewer deaths in the bivalirudin group. There were fewer bleeding episodes (6.8% vs. 8.1%, respectively, p < 0.001) but an increased incidence of acute stent thrombosis in the group of patients who received bivalirudin (1.03% vs. 0.62; p < 0.001). This group also had a slightly but statistically significant shorter length of stay.
The patient and procedural variables were included in a multivariate model to evaluate the relationship between them and the composite endpoint of death, MI, or stroke (Figure 2). The administration of bivalirudin was not significantly associated with the outcome, whereas female sex, age, prior cardiovascular disease and peripheral vascular disease, a history of diabetes and a history of heart failure, cardiogenic shock, and cardiac arrest in the period preceding the index event were.
After adjusting for multiple variables and the propensity score, the OR for the main composite of death, MI, and stroke was not statistically significant (OR: 0.95; 95% CI: 0.87 to 1.05; p = 0.358) for the patients receiving bivalirudin compared with the patients receiving heparin (Figure 3). The OR for bleeding was also not statistically significant (OR: 0.98; 95% CI: 0.91 to 1.05; p = 0.57). The OR for acute stent thrombosis remained statistically significant (OR: 2.11; 95% CI: 1.73 to 2.57; p < 0.001), whereas the adjusted length of stay was not statistically significant. Because the number of procedures performed via the radial approach is increasing constantly in the United States, as more operators are adopting the technique, we also tested for interaction between the endpoints and time. We found no statistically significant interaction with time for any of the adjusted endpoints.
The sensitivity analysis revealed different results depending on the constellation chosen. Thus, comparing patients who received bivalirudin without GP IIb/IIIa inhibitors with patients receiving heparin without GP IIb/IIIa inhibitors (Online Table 1, Online Figure 1) showed a significantly lower rate of the composite of death, MI, or stroke for patients anticoagulated with bivalirudin, which remained significant after adjustments (OR: 0.84; 95% CI: 0.74 to 0.96; p = 0.004). Comparing patients who received bivalirudin without GP IIb/IIIa inhibitors with patients receiving heparin with GP IIb/IIIa inhibitors (Online Table 2, Online Figure 2) showed no significant difference in the composite endpoint, which remained not significant after adjustments (OR: 0.89; 95% CI: 0.79 to 1.01; p = 0.052). Both subgroup analyses showed a significantly lower risk for bleeding with bivalirudin. Finally, comparing patients who received bivalirudin with and without GP IIb/IIIa inhibitors with patients receiving heparin without GP IIb/IIIa inhibitors (Online Table 3, Online Figure 3) showed a significantly lower rate of the composite of death, MI, or stroke for patients anticoagulated with bivalirudin, which remained significant after adjustments (OR: 0.89; 95% CI: 0.79 to 1.00; p = 0.049). The risk for bleeding was not significantly different between the 2 groups.
In a large cohort of real-world patients in the United States, we found that there was no significant difference in the rate of a composite of death, MI, and stroke in patients undergoing primary transradial PCI for STEMI whether they were anticoagulated with bivalirudin or unfractionated heparin. A significant strength of our study is the sample size, which is the largest reported in a study comparing the 2 anticoagulation strategies specifically in patients with STEMI undergoing primary PCI via the radial approach. In addition, our data are representative of current clinical practice in the United States, as more than 80% of PCI centers participate in the CathPCI registry.
Our results are in line with recently published trials such as the HEAT-PPCI trial (9), which randomized 1,829 patients to primary transradial PCI using bivalirudin or unfractionated heparin with provisional GP IIa/IIIa inhibitors. Adverse events occurred less often in the group of patients randomized to heparin compared with those randomized to bivalirudin. However, the use of GP IIb/IIIa inhibitors was significantly lower in both arms of the trial compared with our cohort. In the BRIGHT (Bivalirudin in Acute Myocardial Infarction vs. Glycoprotein IIb/IIIa and Heparin: A Randomised Controlled Trial) trial (21), net clinical outcomes were better with bivalirudin compared with heparin and heparin with tirofiban in patients with acute MI treated with PCI, but this was due primarily to a reduction in bleeding in the patients receiving bivalirudin, as the rates of major adverse cardiac events were similar between the groups.
The MATRIX (Minimizing Adverse Hemorrhagic Events by Transradial Access Site and Systemic Implementation of Angiox) investigators compared bivalirudin versus heparin and femoral access versus radial access (22,23) in patients presenting with acute coronary syndrome in a 2 × 2 design study. They found that the rates of major adverse cardiovascular events and net adverse clinical events were not significantly lower in patients treated with bivalirudin compared with those treated with heparin. Interestingly, they found that this held true for the patients who had the procedure done via the radial approach and for those who had the procedure done via the femoral approach.
This is in contradistinction to the findings of the HORIZONS AMI study (2), in which patients with STEMI had significantly better outcomes if they were anticoagulated with bivalirudin compared with heparin. The majority of procedures in the HORIZONS AMI study were done via the transfemoral approach, but a post hoc analysis (4) revealed that the transradial approach was associated with reduced major bleeding and improved event-free survival. The rate of major adverse cardiovascular events and the rate of death or reinfarction at 30 days was the same, respectively, in the patients who were treated via the transradial approach and who received either bivalirudin or heparin with provisional GP IIb/IIIa inhibitors. However, the total number of patients done via the transradial approach was only 200 (of 3,340).
Whether the lack of differences in the patient groups anticoagulated with bivalirudin and heparin found in this study is related mostly to the fact that we selected only patients who underwent primary PCI via the transradial approach, which is known to lead to fewer access site bleeding events (15), or whether it is related to a better understanding and more judicious use of anticoagulation in the setting of acute intervention is unclear, but our data are consistent with data from contemporary studies of patients with acute coronary syndrome and STEMI.
Our study shows that in transradial primary PCI, even with a high percentage of patients receiving GP IIb/IIIa inhibitors, the outcomes are not significantly different between patients anticoagulated with bivalirudin and heparin. This aspect is important because it was thought that the use of GP IIb/IIIa inhibitors was a significant contributor to the increased rate of bleeding and the poorer outcomes with heparin. Our data suggest that in transradial primary PCI, this may not be an issue.
Among the individual endpoints in our study, all-cause death seemed to favor bivalirudin in the unadjusted analysis, but there was no statistically significant association with improved mortality after adjusting for multiple variables and for the propensity score. The same finding applied to bleeding, which seemed to be reduced by bivalirudin compared with heparin in the unadjusted analysis. However, in the multivariate model, the OR did not reach statistical significance.
These results are consistent with those of some of the newer trials such as HEAT-PPCI, EUROMAX, and BRAVE 4, although in EUROMAX, there was a significant reduction of bleeding with bivalirudin compared with heparin.
It was thought that bleeding is an important contributor to major cardiovascular events in patients undergoing PCI in the setting of acute coronary syndrome and STEMI (24–26). The benefit of bivalirudin therapy was thought to be related primarily to the reduction in bleeding; however, an analysis of the HORIZONS AMI data showed that this did not completely explain the benefit (3). Some earlier studies (2,27) suggested that treatment with bivalirudin improved mortality in patients with STEMI, but a more recent meta-analysis comparing bivalirudin and heparin did not find a mortality advantage with bivalirudin (28). Our data also showed no significant reduction in mortality or transfusion rates in patients treated with bivalirudin.
We also found an increased incidence of acute stent thrombosis associated with bivalirudin compared with heparin. This remained significant after adjustments in the multivariate model. This finding is consistent among most trials and is thought to be related to the short half-life of bivalirudin. However, a pre-specified analysis of the MATRIX antithrombin therapy study did not find superior outcomes with prolonged infusions of bivalirudin compared with no post-PCI infusions (22). In our study, the increase in acute stent thrombosis was not associated with an increase in death or an increase in the composite endpoint.
The current American College of Cardiology/American Heart Association guidelines (29) give both bivalirudin and heparin a class I indication for anticoagulation of patients with STEMI undergoing primary PCI. Our data support the equipoise of the 2 anticoagulant agents in patients undergoing primary PCI via the radial approach and are consistent with the opinion of a number of interventional cardiologists surveyed about the comparison between heparin and bivalirudin (30). As the number of transradial procedures increases in the United States (31), this issue will become more important because of the cost difference between the 2 agents. Bivalirudin is easier to use: it has a rapid onset, it does not need activated clotting time monitoring during PCI, and, according to the findings of the MATRIX study, it does not need to be infused post-PCI (22). This will have to be balanced against its cost in a cost-effectiveness analysis.
Our sensitivity analysis provides some insights into direct comparisons of bivalirudin and heparin when GP IIb/IIIa inhibitors are forced out of the equation and suggests that in the direct comparison, bivalirudin may have superior outcomes. However, our study showed that in the real world, more than a third of patients with STEMI undergoing transradial PCI who receive heparin and about a fifth of patients who receive bivalirudin also receive GP IIb/IIIa inhibitors.
This was a retrospective analysis of the data from the NCDR CathPCI registry, not a randomized clinical trial; as such, our analysis may be prone to biases that remained despite our use of a propensity score to try to minimize them. A substantial number of patients anticoagulated with bivalirudin received some amount of heparin, but an analysis from the HORIZONS AMI study showed that these patients do not have worse outcomes than patients who are anticoagulated with heparin only (32). The data yield only in-hospital outcomes and do not allow us to make statements about longer term outcomes, but previous studies have shown that the first few days after a STEMI are the days when the patient is at the highest risk for events (33–35).
Our data suggest that anticoagulation with bivalirudin or with heparin during primary PCI via the transradial approach is associated with similar rates of adverse cardiovascular in-hospital outcomes, notwithstanding an increase in acute stent thrombosis in patients receiving bivalirudin.
WHAT IS KNOWN? Bivalirudin was thought to be associated with superior outcomes compared with heparin, but results from newer trials have questioned this assertion.
WHAT IS NEW? Our data from a large cohort of patients in the NCDR CathPCI registry suggest that in patients undergoing primary PCI via transradial access anticoagulated with bivalirudin or heparin, there was no difference in the composite endpoint of death, MI, or stroke.
WHAT IS NEXT? A randomized trial in patients treated exclusively via transradial primary PCI and anticoagulated with bivalirudin versus heparin as well as a cost-effectiveness analysis comparing heparin versus bivalirudin would help practitioners and hospitals make better decisions regarding anticoagulation in these patients.
For supplemental tables and figures, please see the online version of this article.
This work was supported by a grant from the Virginia Commonwealth University Pauley Heart Center to Dr. Jovin. Dr. Rao is a consultant to Medtronic, MedAxiom, Boston Scientific, and Svelte. Dr. Kontos is a consultant to AstraZeneca. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- confidence interval
- myocardial infarction
- National Cardiovascular Data Registry
- odds ratio
- percutaneous coronary intervention
- ST-segment elevation myocardial infarction
- Received January 17, 2017.
- Revision received February 27, 2017.
- Accepted March 9, 2017.
- Stone G.W.,
- Clayton T.,
- Deliargyris E.N.,
- Prats J.,
- Mehran R.,
- Pocock S.J.
- Genereux P.,
- Mehran R.,
- Palmerini T.,
- et al.,
- for the HORIZONS AMI Trial Investigators
- Romagnoli E.,
- Biondi-Zoccai G.,
- Sciahbasi A.,
- et al.
- Briguori C.,
- Visconti G.,
- Focaccio A.,
- et al.
- Schulz S.,
- Richardt G.,
- Laugwitz K.L.,
- et al.,
- for the Bavarian Reperfusion Alternatives Evaluation I
- Kinaird T.,
- Medic G.,
- Casella G.,
- et al.
- Perdoncin E.,
- Seth M.,
- Dixon S.,
- et al.
- Appleton D.L.,
- Cooke R.H.,
- Rao S.V.,
- Jovin I.S.
- Baklanov D.V.,
- Kim S.,
- Marso S.P.,
- Subherwal S.,
- Rao S.V.
- Brindis R.G.,
- Fitzgerald S.,
- Anderson H.V.,
- Shaw R.E.,
- Weintraub W.S.,
- Williams J.F.
- Shaw R.E.,
- Anderson H.V.,
- Brindis R.G.,
- et al.
- Eikelboom J.W.,
- Mehta S.R.,
- Anand S.S.,
- Xie C.,
- Fox K.A.,
- Yusuf S.
- Pinto D.S.,
- Ogbonnaya A.,
- Sherman S.A.,
- Tung P.,
- Normand S.L.
- Capodanno D.,
- Gargiulo G.,
- Capranzano P.,
- Mehran R.,
- Tamburino C.,
- Stone G.W.
- O’Gara P.T.,
- Kushner F.G.,
- Ascheim D.D.,
- et al.,
- for the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines
- Adamo M.,
- Byrne R.A.,
- Baumbach A.,
- Haude M.,
- Windecker S.,
- Valgimigli M.
- Baklanov D.V.,
- Kaltenbach L.A.,
- Marso S.P.,
- et al.
- Dangas G.D.,
- Mehran R.,
- Nikolsky E.,
- et al.