Effect of Short Procedural Duration With Bivalirudin on Increased Risk of Acute Stent Thrombosis in Patients With STEMI: A Secondary Analysis of the HORIZONS-AMI Randomized Clinical Trial

Key Points

Question  Do “fast” procedures (<45 minutes) have a higher risk of acute stent thrombosis?

Findings  In this ad hoc analysis of the Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction (HORIZONS-AMI) clinical trial, patients who were randomized to bivalirudin, underwent stent placement, and had shorter (<45 minutes) procedures were at statistically higher risk of acute stent thrombosis than patients who underwent longer procedures.

Meaning  Short-acting medications, such as bivalirudin, may not allow for adequate antithrombotic effect of oral antiplatelets in fast procedures; additional studies are needed to evaluate antithrombotic therapies with different pharmacokinetics.

Importance  Bivalirudin has been associated with reduced bleeding and mortality during primary percutaneous coronary intervention for ST-segment elevation myocardial infarction (STEMI). However, increased rates of acute stent thrombosis (AST) have been noted when bivalirudin is discontinued at the end of the procedure, which is perhaps related to this medication’s short half-life.

Objectives  To evaluate the clinical effect of procedure duration on AST when either bivalirudin or heparin plus glycoprotein IIb/IIIa receptor inhibitor (GPI) is used.

Design, Setting, and Participants  An ad hoc analysis of the Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction (HORIZONS-AMI) clinical trial was performed between March 1, 2015, and April 30, 2016, on patients who underwent primary percutaneous coronary intervention with stents and were randomized 1:1 to bivalirudin or heparin plus GPI. Defined as the difference between the patient’s arrival at the catheterization laboratory and the patient’s final angiogram. Participants included 3602 patients with STEMI, aged 18 years or older, who were undergoing primary percutaneous coronary intervention and presenting less than 12 hours from symptom onset.

Main Outcomes and Measures  Clinical events committee–adjudicated definite AST (occurring ≤24 hours after percutaneous coronary intervention).

Results  Among patients included in this analysis, procedure time was identified in 1286 receiving bivalirudin and 1412 receiving heparin plus GPI. Shorter procedures were defined as the lowest quartile of duration (<45 minutes). Patients undergoing shorter procedures were younger and less likely to be hypertensive and smokers. Shorter procedures were less complicated with fewer stents implanted, less multivessel stenting, less thrombus, and less no-reflow. An increased risk of definite AST was associated with shorter than with longer procedures with bivalirudin (7 [2.1%] vs 7 [0.7%]; relative risk, 2.87; 95% CI, 1.01-8.17; P = .04) but not with heparin plus GPI (0 vs 3 [0.3%]; P = .30).

Conclusions and Relevance  Despite less procedural complexity, shorter primary percutaneous coronary intervention time was associated with an increased risk of AST in patients treated with bivalirudin but not patients treated with heparin plus GPI, possibly because of the rapid offset of bivalirudin’s antithrombotic effect during a window of limited oral antiplatelet action.

Trial Registration  clinicaltrials.gov Identifier: NCT00433966

In the Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction (HORIZONS-AMI) randomized clinical trial, bivalirudin, compared with heparin and glycoprotein IIb/IIIa receptor inhibitor (GPI), had similar rates of ischemic complications and reduced bleeding and mortality after primary percutaneous coronary intervention (PPCI) for ST-segment elevation myocardial infarction (STEMI).¹ Definite stent thrombosis was more common with bivalirudin.^2-4 Preprocedural heparin and high-dose clopidogrel therapy reduced subacute stent thrombosis but not acute stent thrombosis (AST).² It was postulated that (1) in shorter procedures supported by procedural bivalirudin, AST rates would be higher than in longer procedures because the therapeutic antiplatelet effect was not achieved before antithrombin inhibition resolves, and (2) this phenomenon would not be observed with heparin plus GPI because of longer antithrombin and more potent antiplatelet effect.^1,2

The HORIZONS-AMI randomized clinical trial, conducted between March 25, 2005, and May 7, 2007, was a prospective, open-label, multicenter trial that included 3602 patients with STEMI, aged 18 years or older, who were undergoing PPCI and presenting fewer than 12 hours from symptom onset. The study design and primary results of this trial have been published.¹ The trial was approved by the institutional review board or ethics committee at each participating site, and all participants provided written informed consent prior to participation. Patients were randomized 1:1 to bivalirudin or heparin plus GPI and then 3:1 to the TAXUS-Express paclitaxel-eluting stent (Boston Scientific Corporation) or identical bare metal stents. Clinical follow-up occurred at 30 days, 6 months, 1 year, 2 years, and 3 years. This ad hoc analysis was conceived and prepared between March 1, 2015, and April 30, 2016.

Procedure time, defined as the difference between the patient’s arrival at the catheterization laboratory and the patient’s final angiogram, was identified in all patients (bivalirudin, n = 1286; heparin plus GPI, n = 1412) (eFigure in the Supplement). Total procedure time was divided into shorter (shortest procedure quartile) and longer (other 3 quartiles) procedures. Patients were excluded for bivalirudin or heparin infusion postprocedure (n = 185) (eFigure in the Supplement), missing procedure times (n = 16), or no stent implantation (n = 703).

The primary outcome was clinical events committee–adjudicated definite AST (occurring ≤24 hours after percutaneous coronary intervention), according to Academic Research Consortium criteria.⁵ Core laboratory angiographic analysis confirming stent thrombosis was performed. Propensity score matching (eTable in the Supplement) was employed to address known confounding factors and to quantify the time-dependence odds of AST according to treatment.

Median (interquartile range [IQR]) procedure time was 55 (49-77) minutes and was similar in the bivalirudin group and the heparin-plus-GPI group (median [IQR], 58 [45-76] minutes; P = .35). Median (IQR) procedure time was 38 (32-42) minutes in the shorter procedure quartile, while the other quartiles had a pooled median (IQR) of 68 (56-85) minutes. Baseline clinical characteristics are shown in Table 1. Patients who underwent shorter procedures were younger and less likely to be hypertensive and smokers.

Shorter procedures in both treatment groups were associated with shorter door-to-balloon times, less frequent use of radial access, and more frequent use of a 600-mg (rather than 300-mg) clopidogrel loading dose (Table 2). Rates of preprocedure heparin administration were similar for shorter and longer procedures. Shorter procedural times were associated with less complex procedures and less multivessel PPCI. There were no substantial differences between pre-PCI and post-PCI thrombolysis in myocardial infarction flow or peak activated clotting time measurements according to procedure time. Bailout use of GPI in patients treated with bivalirudin was more frequent among patients with longer procedures (Table 2). The most common indications for bailout use of GPI were sustained no-reflow (47 [32.1%]) and giant thrombus (39 [26.7%]).

Among patients treated with bivalirudin, definite AST was more frequent after shorter procedures than after longer procedures (7 [2.1%] vs 7 [0.7%]; 95% CI, 1.01-8.17; P = .04) (Table 3). All stent thromboses occurred postprocedure. The median (IQR) procedure time for bivalirudin-treated patients with AST was 49 (30-67) minutes and for bivalirudin-treated patients without AST was 60 (45-79) minutes (P = .04). Most AST occurred among patients treated with bivalirudin, with procedure times faster than 60 minutes. Among patients treated with heparin plus GPI, no AST occurred in shorter procedures compared with longer procedures (0 vs 3 [0.3%]; P = .30) (Table 3).

There were no substantial differences in stent thrombosis between the treatment groups from 1 to 30 days (subacute) and from 1 to 3 years (very late) (Table 3). Results of propensity score analysis (eTable in the Supplement) indicated most AST occurred in procedures faster than 60 minutes. There was an interaction of treatment assignment and procedure duration indicating this relationship was observed in patients treated with bivalirudin but not for those treated with heparin plus GPI.

Compared with the use of heparin plus GPI, bivalirudin use in PPCI was associated with less major bleeding, thrombocytopenia and mortality, and an increased risk of AST.^1,3,4 Shorter procedural duration (≤45 minutes) among patients treated with bivalirudin was associated with more definite AST despite being less complex than longer procedures. These findings were specific to patients treated with bivalirudin because this association was not observed in patients receiving heparin plus GPI.

The pharmacokinetics of bivalirudin and clopidogrel offer insight. When the PPCI procedure is completed rapidly, adequate antiplatelet effect may not have been achieved with oral agents, particularly if gastrointestinal tract absorption has been slowed in STEMI.⁶ In this setting, the short half-life of bivalirudin (25 minutes) leaves the newly implanted stent relatively unprotected against AST for short procedures. Conversely, the duration of antithrombin effect is longer and rapid antiplatelet effect is achieved with heparin plus GPI; no relationship between procedural duration and AST was found. Whether this phenomenon occurs with the use of unfractionated heparin without GPI is unknown.

These findings also provide insight into differential outcomes reported among STEMI trials comparing bivalirudin with heparin. Pooled results of the HORIZONS-AMI and EUROMAX (European Ambulance Acute Coronary Syndrome Angiography)⁷ studies demonstrated reduced rates of cardiac death (2.0% vs 2.9%; relative risk [RR], 0.70; 95% CI, 0.50-0.97; P = .03) and less major bleeding (4.2% vs 7.8%; RR, 0.53; 95% CI, 0.43-0.66; P < .0001).⁷ Similarly, among patients with STEMI in the MATRIX (Minimizing Adverse Hemorrhagic Events by Transradial Access Site and Systemic Implementation of Angiox) study,⁴ the 30-day rate of all-cause mortality was lower with bivalirudin treatment compared with heparin (2.1% vs 3.1%; RR, 0.68; 95% CI, 0.46-1.01; P = .05).^2,4 However, among patients treated with procedural-only bivalirudin in these trials (or with a low-dose post-PCI bivalirudin infusion of 0.25 mg/kg/h), AST was increased, compared with heparin with or without GPI; this is a finding also observed in the HEAT-PPCI (Unfractionated Heparin vs Bivalirudin in Primary Percutaneous Coronary Intervention) trial.⁸

In the EUROMAX and BRIGHT (Bivalirudin vs Heparin With or Without Tirofiban During Primary Percutaneous Coronary Intervention in Acute Myocardial Infarction) trials,^7-9 prolongation of the 1.75–mg/kg/h bivalirudin infusion for 3 to 4 hours after PCI appeared to mitigate the risk of AST regardless of whether clopidogrel or a more potent and rapid-acting oral antiplatelet agent, such as ticagrelor or prasugrel, was used. Similarly, AST occurred in 0.6%, 0.8%, and 0.2% of patients in the MATRIX trial⁴ with no, low-dose, and high-dose postprocedural bivalirudin infusions, confirming that prolongation of infusion likely does reduce the excess risk of stent thrombosis.^4,9-11 Thus, the US Food and Drug Administration has recently revised the instructions for use of bivalirudin to strongly recommend this prolonged high-dose infusion after PPCI.

Whether use of a rapid-acting potent intravenous antiplatelet agent lowers AST risk with procedural-only bivalirudin for short procedures requires further study. A 2-hour infusion of cangrelor in bivalirudin-treated patients undergoing percutaneous coronary intervention resulted in approximately 50% reduction in AST compared with periprocedural clopidogrel use, although few of these were patients with STEMI.¹² Bolus GPI led to rapid and sustained platelet inhibition with prasugrel in the FABOLUS-PRO (Facilitation Through Aggrastat by Dropping or Shortening Infusion Line in Patients With ST-Segment Elevation Myocardial Infarction Compared to or on Top of Prasugrel Given at Loading Dose) trial.¹³ Furthermore, preloading with ticagrelor has also shown to decrease risk of thrombotic events.^14,15

Our findings should be interpreted with care given the low event rates. Although HORIZONS-AMI was one of the largest PPCI trials ever, there was not sufficient statistical power to evaluate differences in various patient characteristics and technical factors for a low-frequency outcome such as AST, especially when stratifying by procedure time. However, most of the observed factors confounding the association of procedure time with AST, such as complexity of patients and procedures, would have biased toward negating the results observed in this study. Similarly, interpretation of the effect of nonrandomized treatments, such as bailout or upstream GPI use and prerandomization unfractionated heparin use, is limited. Differences in procedural duration may also reflect less meticulous technique, resulting in AST.

Shorter procedure time among patients with STEMI is associated with an increased risk of AST after PPCI with procedural bivalirudin, but not with heparin plus GPI. Additional studies evaluating strategies aimed at mitigating this effect, including a high-dose prolonged bivalirudin infusion or intravenous cangrelor, are warranted in order to achieve the benefits of bivalirudin during PPCI, including reducing major bleeding and mortality, without an increased risk of AST.

Corresponding Author: Duane S. Pinto, MD, MPH, Department of Cardiology, Beth Israel Deaconess Medical Center, 1 Deaconess Rd, Palmer 415, Boston, MA 02215 (dpinto@bidmc.harvard.edu).

Accepted for Publication: December 3, 2016.

Published Online: March 1, 2017. doi:10.1001/jamacardio.2016.5669

Author Contributions: Dr Pinto had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Tamez, Pinto, Kirtane, Dangas, Mehran, Deliargyris, Gibson, Stone.

Acquisition, analysis, or interpretation of data: Tamez, Pinto, Kirtane, Litherland, Yeh, Mehran, Deliargyris, Ortiz, Stone.

Drafting of the manuscript: Tamez, Pinto, Kirtane, Ortiz.

Critical revision of the manuscript for important intellectual content: Tamez, Pinto, Kirtane, Litherland, Yeh, Dangas, Mehran, Deliargyris, Gibson, Stone.

Statistical analysis: Tamez, Pinto, Litherland.

Obtained funding: Stone.

Administrative, technical, or material support: Tamez, Kirtane, Deliargyris.

Study supervision: Kirtane, Mehran, Gibson, Stone.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.

Funding/Support: This analysis was supported by internal funds of the Cardiovascular Research Foundation.

Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: Dr Kirtane is an associate editor of JAMA Cardiology. He was not involved in the editorial evaluation or decision to accept this article for publication.