Author + information
- Received March 23, 1998
- Revision received July 10, 1998
- Accepted August 6, 1998
- Published online December 1, 1998.
- ↵*Address for correspondence: Robert A. Harrington, MD, FACC, Duke Clinical Research Institute, 2024 West Main Street, Durham, NC 27705
Objectives. The trial was designed to assess the safety, pharmacodynamics and effects on reperfusion of the platelet glycoprotein (GP) IIb/IIIa inhibitor lamifiban when given with thrombolysis to patients with ST segment elevation acute myocardial infarction.
Background. Studies of fibrinolytic agents in acute myocardial infarction have demonstrated a direct relationship between early complete reperfusion and survival. Blockade of the platelet GP IIb/IIIa receptor complex inhibits platelet aggregation and may speed reperfusion when given in conjunction with thrombolysis to patients with acute myocardial infarction.
Methods. Patients with ST segment elevation presenting within 12 h of symptom onset who were treated with either tissue-plasminogen activator or streptokinase were enrolled in this three-part Phase II dose exploration study. In Part A, all patients received the GP IIb/IIIa inhibitor lamifiban in an open-label, dose escalation scheme. Parts B and C were a randomized, double-blind comparison of a bolus plus 24-h infusion of lamifiban versus placebo with patients randomized in a 2:1 ratio. The goal was to identify a dose(s) of lamifiban that provided >85% adenosine diphosphate (ADP)-induced platelet aggregation inhibition. A composite of angiographic, continuous electrocardiographic and clinical markers of reperfusion was the primary efficacy end point, and bleeding was the primary safety end point.
Results. Platelet aggregation was inhibited by lamifiban in a dose-dependent manner with the highest doses exceeding 85% ADP-induced platelet aggregation inhibition. There was more bleeding associated with lamifiban (transfusions in 16.1% lamifiban-treated vs. 10.3% placebo-treated patients). Lamifiban induced more rapid reperfusion as measured by all continuous electrocardiographic (ECG) parameters.
Conclusions. Lamifiban given with thrombolytic therapy appears to be associated with more rapid and complete reperfusion than placebo. As expected in this small sample, there were no obvious clinical benefits to lamifiban over placebo. Reconciliation of ECG monitoring with clinical outcomes will require a larger, adequately powered clinical trial.
Numerous studies of fibrinolytic agents in acute myocardial infarction have demonstrated a direct relationship between early complete reperfusion and survival (1,2). There are significant limitations to reperfusion therapy, however, and although the majority of patients experience clinical reperfusion, it has been postulated that only a minority achieve optimal reperfusion with current thrombolytic regimens (3). Fibrinolytic therapy has a potent platelet aggregating effect as does the exposure of the ruptured atherosclerotic plaque. Given the key role of platelets in the pathogenesis of acute myocardial infarction (4), even a weak antiplatelet agent such as aspirin is beneficial when given alone or in conjunction with reperfusion (5,6). Such therapy may improve early coronary flow rates as well as stabilize or maintain subsequent perfusion and provide an incremental improvement in clinical outcomes.
Platelet aggregation and formation of a platelet-rich thrombus involves fibrinogen ligand binding to the platelet glycoprotein (GP) IIb/IIIa receptor complex (7). Blockade of this receptor with monoclonal antibody fragments (8), peptides and nonpeptides (9)has proved an effective therapeutic strategy for patients undergoing percutaneous intervention (10–14). Such therapy also may speed reperfusion when given in conjunction with thrombolysis to patients with acute myocardial infarction (15,16).
Lamifiban (F. Hoffmann-La Roche, Ltd., Basel, Switzerland) is a highly selective, nonpeptide antagonist of the platelet GP IIb/IIIa receptor (17)that provides dose-dependent inhibition of platelet aggregation in response to all agonists (18). The PARADIGM (Platelet Aggregation Receptor Antagonist Dose Investigation and Reperfusion Gain in Myocardial Infarction) trial was designed as a Phase II dose-finding study to assess the safety, pharmacokinetics, pharmacodynamics and effects on reperfusion of lamifiban when given with thrombolysis to patients with ST segment elevation acute myocardial infarction.
Patients between 21 and 75 years old with ischemic chest pain (or its equivalent) lasting more than 20 min and beginning within 12 h of presentation were eligible for enrollment (19). They needed to have ST segment elevation ≥1 mm in two limb leads or ≥2 mm in two contiguous precordial leads to be eligible for thrombolytic therapy. Major exclusion criteria included uncontrolled hypertension, renal insufficiency (serum creatinine >2.0), recent major trauma or surgery, peptic ulcer or gastrointestinal bleeding within 6 months, known coagulopathy or thrombocytopenia (platelets <100,000/μl), major comorbid illnesses limiting survival, recent transient ischemic attack, stroke, cerebral aneurysm or a history of intracranial hemorrhage.
At the investigators’ discretion, all patients received either a weight-adjusted regimen of accelerated tissue-plasminogen activator (t-PA) (100 mg over 90 min) (20)or streptokinase (SK) (1.5 million U over 1 h). All patients also received aspirin (160 to 325 mg/day). Patients given t-PA received intravenous heparin for at least 24 h after thrombolysis; patients given SK did not receive heparin for the first 24 h. Heparin was given as a 5,000-U bolus followed by 1,000 U/h and titrated to maintain the activated partial thromboplastin time between 60 and 85 s. In Part A, all patients received open-label lamifiban beginning with a low dose chosen from previous Phase II work in unstable angina (18)and proceeding to a second dose chosen by the study’s Executive Committee based on safety and platelet aggregation studies. The goal of Part A was to identify a dosing strategy that resulted in 85% to 95% adenosine diphosphate (ADP)-induced inhibition of platelet aggregation. Part B was a randomized, double-blind comparison of a bolus followed by a 24-h infusion of lamifiban versus placebo with patients being randomized in a 2:1 ratio. After review by the Executive Committee of the pharmacodynamics, safety and efficacy data obtained in Part B, the length of infusion was increased to 48 h in Part C; patients continued to be randomized in a 2:1 ratio of lamifiban versus placebo. Lamifiban was given as close to the start of thrombolytic administration as possible, with the goal being to start within 30 min of lytic administration.
The primary goal of PARADIGM was to identify a dose or doses of lamifiban that provided the desired pharmacodynamic effect of ≥85% ADP-induced platelet aggregation inhibition without a major increase in serious bleeding, especially intracranial hemorrhage. For platelet aggregation testing at selected centers, blood samples were obtained 60 to 90 min and 6 to 18 h after the start of study drug. Samples were also obtained 2 to 6 hours after study drug termination. Each center participating in this substudy was required to submit normal platelet aggregation curves on three to five subjects using their center’s standard technique, which served as their baseline. All aggregation curves were compared to each center’s baseline curves, and the level of inhibition was calculated as a percentage of normal aggregation. Platelet aggregation was induced by ADP (final concentration 10 μmol/L) and thrombin receptor agonist peptide (TRAP, final concentration 25 μmol/L). In the case of TRAP-induced aggregation, all values were corrected as previously described (18).
The primary safety end point was bleeding during hospitalization. Major or life-threatening bleeding was defined as any intracranial bleeding or other bleeding that led to hemodynamic instability requiring intervention. Intermediate bleeding required blood transfusion but was not associated with hemodynamic instability. Thrombocytopenia was defined as a drop in platelet count of more than one third from baseline and below 100,000/μl.
To assess the potential impact of lamifiban on reperfusion, the primary efficacy outcome was a composite of angiographic, continuous electrocardiographic and clinical markers of reperfusion failure occurring by hospital discharge or 30 days, whichever came first. Components of this composite included the following: lack of Thrombolysis in Myocardial Infarction (TIMI) 3 flow on any predischarge angiogram or 60- to 90-min angiogram in selected centers; persistent or recurrent ST elevation by continuous 12-lead digital ST segment monitoring (21); death, reinfarction, refractory ischemia or the need for nonelective revascularization (rescue angioplasty or any intervention performed for hemodynamic instability with ischemia or for acutely refractory ischemia). Refractory ischemia was defined as chest pain, or its symptomatic equivalent, occurring after resolution of the initial infarction symptoms, having associated electrocardiographic evidence of ischemia, and lasting for at least 20 min despite maximal antianginal therapy.
Continuous 12-lead ST segment monitoring
All patients were monitored with a continuous 12-lead digital electrocardiographic monitor (ST-100, Mortara Instrument, Milwaukee, WI) as previously described (22)for 24 h following administration of thrombolytic therapy. Data were analyzed at the Ischemia Monitoring Core Laboratory by an experienced cardiologist blinded to drug assignment and angiographic and clinical information. Continuously updated 12-lead ST segment recovery was analyzed for the following end points: early recurrent ischemia, time from onset of therapy to stable ST recovery, frequency of stable ST recovery by 2 h, peak ST deviation, late ST reelevation after stable ST recovery, and a noninvasive 90-min patency assessment based on the resolution and stability of the ST segments. The 90-min patency variable uses previously published methods correlating ST segment recovery with simultaneous angiographic assessment of infarct artery patency (15,23).
Data collection and core laboratories
Demographic and clinical outcome data were collected on case report forms that were 100% source documented for key variables by clinical monitors. Platelet aggregation data were kept and submitted to the Coordinating Center separate from clinical data to protect study assignment blinding. An independent core laboratory at the Cleveland Clinic Foundation blinded to treatment assignment reviewed all cineangiograms and determined TIMI flow grades.
From historical data it was estimated that a sample size of approximately 10 patients receiving each dose of lamifiban in Part A should allow determination of the median ADP-induced platelet aggregation with a standard error <10%. Assuming a composite efficacy event rate of 50% in the placebo arm and 30% in the highest dose lamifiban arm, a sample size of approximately 360 patients was estimated to provide 86% power for detection of the difference with a one-sided alpha equal to 5%.
The medians (with 25th and 75th interquartile ranges) are given for continuous variables, and frequencies are given for dichotomous variables. If a patient was missing data regarding one of the components of the composite efficacy end point, that component was assumed to be a nonevent in defining the composite.
Patients were enrolled from the United States, Canada, Iceland and Belgium. The Duke Clinical Research Institute was the overall study Coordinating Center. The Canadian Coordinating Center was located in Edmonton, Alberta, and the Cleveland Clinic Foundation served as the Executive Center. Representatives from the coordinating centers and from the sponsor served on the Executive Committee. A Clinical Events Committee with no knowledge of the randomization data reviewed and classified all efficacy and safety events. An independent Data Safety and Monitoring Board oversaw the safety of the trial.
There were 353 patients enrolled in the trial at two doses in Part A (300-μg bolus and 1.0-μg/min infusion [n = 15]; 400-μg bolus and 2.0-μg/min infusion for 24 h [n = 15]), one dose in Part B (400-μg bolus and 1.5-μg/min infusion for 24 h) or placebo (n = 112 and 61, respectively) and one dose in Part C (400-μg bolus and 2.0-μg/min infusion for 48 h) or placebo (n = 94 and 56, respectively). Baseline demographics and prerandomization cardiac history are seen in Table 1.
Virtually all patients received thrombolytic therapy (350/353) with the majority receiving t-PA given over 90 min, and the remainder streptokinase given over 1 h (266 vs. 85; one patient received both). There was good compliance with the study protocol with the majority of patients receiving study drug for 24 h in Parts A and B and for 48 h in Part C. The median time from thrombolytic therapy to the start of study drug was 0.48 h (0.28, 0.75) for placebo and 0.42 h (0.25, 0.67) for lamifiban. Study drug was terminated prematurely in more patients receiving lamifiban than placebo (17.2% vs. 13.7% for Parts B and C lamifiban vs. placebo). The most common reason for early discontinuation in the lamifiban group was an adverse event, bleeding (37.1% of early discontinuations compared with no early discontinuations for bleeding in the placebo patients).
Platelet aggregation data were available for 29 patients in Part A, 45 patients in Part B (29 in the lamifiban group) and 42 patients in Part C (27 in the lamifiban group). Platelet aggregation in response to ADP and TRAP was inhibited by lamifiban in a dose-dependent manner. Figure 1shows the percent inhibition of aggregation determined at steady state, 6 to 18 h after the start of study drug. Comparable inhibition of platelet aggregation was detected 60 to 90 min postbolus (Tables 2 to 4), ⇓⇓whereas at 2 to 6 h after stopping the infusion of lamifiban there were clearly lower levels of platelet inhibition (data not shown). With TRAP, a more potent platelet activator than ADP, higher lamifiban doses were required to achieve the same inhibition of aggregation as observed using ADP as the agonist (Tables 2 to 4). Platelet aggregation appears to have been more inhibited when SK was used as the thrombolytic agent instead of t-PA (Tables 2 to 4). However, the overall number of patients treated with SK in whom inhibition of aggregation was measured was quite small.
Safety end points
There were two intracranial hemorrhages, one in a patient receiving t-PA and lamifiban in Part B and one in a patient treated with SK and lamifiban in Part C. Bleeding and thrombocytopenia by treatment group are seen in Table 5. Intermediate and major bleeding were mainly due to gastrointestinal (5.6% lamifiban vs. 0.9% placebo), bypass surgery–related (3.4% lamifiban vs. 3.4% placebo) or femoral access site sources (2.6% lamifiban vs. 1.7% placebo). An increase in bleeding with lamifiban was seen with both t-PA and streptokinase. Patients who experienced a bleeding event were older (median 63 years vs. 59), and more likely to be female (42.9% vs. 18.6%), to have diabetes (26.5% vs. 13.6%) and to have undergone an invasive cardiac procedure during the index hospitalization (cardiac catheterization 89.8% vs. 66.4%; percutaneous coronary intervention 36.7% vs. 32.7%; coronary artery bypass surgery 36.7% vs. 7.3%) compared with those patients without any bleeding.
Efficacy end points
Clinical outcomes are displayed in Table 6. As would be expected in such a small sample, there were no significant differences among the treatment groups. Peak median creatine kinase levels were 1,643 IU/liter (781, 2,769) for the Parts B and C lamifiban patients and 1,312 IU/liter (675, 2,936) for the Parts B and C placebo patients (p = 0.73). Table 7shows the efficacy and safety end points when pooling all lamifiban versus all placebo patients.
ST segment monitoring data by treatment group are displayed in Table 8. The speed and stability of reperfusion were significantly enhanced in patients treated with lamifiban compared with placebo as seen in Table 9.
During the initial hospitalization, 69.7% (246/353) of patients underwent cardiac catheterization with only 34 undergoing the elective 90-min procedure. Given the very small numbers, no effect was seen on early TIMI 3 flow.
Inhibition of platelet aggregation
In this study of the nonpeptide platelet GP IIb/IIIa inhibitor lamifiban, median inhibition of platelet aggregation exceeded 85% in response to ADP stimulation. The goal was to identify a biologically active dose using the assumption that it would not be useful to increase the dose beyond that required to achieve 80% to 90% ADP-induced platelet aggregation inhibition. Although TRAP is a more potent stimulator of platelet aggregation, most previous trials in this field have relied on a measurement of ADP-induced aggregation for dosing selection.
Measurement of the inhibition of platelet aggregation may be influenced by a variety of patient-specific factors, including concomitant medication usage, and by a number of technical limitations of the assay. No previous study has been able to sufficiently correlate levels of platelet inhibition with optimal patient outcomes, but ex vivo aggregation remains the standard of determining dosing of the platelet GP IIb/IIIa inhibitors (24). Although we measured other physiologic (electrocardiographic) and clinical end points to characterize the clinical consequences of this level of platelet inhibition, we recognized that PARADIGM was a small, Phase II dose-finding study without adequate statistical power to detect small to moderate meaningful differences in clinical outcomes.
In GUSTO-I, treatment with accelerated t-PA provided 90-min TIMI 3 flow rates in excess of 50%, which was significantly higher than other thrombolytic strategies (25), was highly associated with improvements in 30-day mortality and explained the superiority of accelerated t-PA over streptokinase (26). Since currently available thrombolytic agents provide only approximately 55% TIMI 3 flow rates, and since part of the resistance to fibrinolysis is felt to be due to the platelet-richness of the clot, we sought to combine potent antiplatelet therapy with fibrinolysis in hopes of achieving higher and more complete reperfusion rates.
Previous trials combining glycoprotein IIb/IIIa inhibition with fibrinolysis
There have been two previous studies of platelet GP IIb/IIIa inhibitors given with thrombolytic therapy. Kleiman et al. demonstrated the feasibility of combining the two therapies for the treatment of acute myocardial infarction in the TAMI 8 trial (16). More recently, Ohman et al. showed that use of the peptide GP IIb/IIIa inhibitor Integrilin (eptifibatide) (COR Therapeutics, South San Francisco, CA) could promote higher rates of early reperfusion than those achieved with thrombolysis alone as measured by TIMI 3 flow at 90-min catheterization (15).
Glycoprotein IIb/IIIa antagonism resulting in more than 80% inhibition of ADP-induced aggregation has been associated with favorable clinical outcomes in patients undergoing percutaneous coronary intervention after treatment with abciximab (27). A potentially important benefit of lamifiban with regard to safety was the reversibility of the antiplatelet effect after study drug termination. This feature might prove especially attractive in the acute myocardial infarction population being treated with fibrinolytic, antithrombin and antiplatelet combinations. What is unclear is the optimal dose of the thrombolytic agent that is combined with the antiplatelet therapy. Since thrombolysis has prothrombotic effects by virtue of exposing free thrombin and promoting platelet aggregation, a lower dose of the thrombolytic agent given in combination with platelet GP IIb/IIIa inhibition might result in high rates of reperfusion with acceptable bleeding risks. Additionally, the optimal dose and length of infusion of the antiplatelet agent in such a setting are unknown and will need to be studied in larger clinical outcome studies.
Measurement of reperfusion using continuous ST segment monitoring
Although the sample size was small, which limited definitive observations, there was no increase in angiographic TIMI 3 flow rates among patients treated with lamifiban compared with placebo. The use of lamifiban was, however, associated with significant improvement in reperfusion as measured by continuous electrocardiographic monitoring during the first 24 h of hospitalization, a consistent effect across all measured parameters. Although insufficient numbers of patients underwent 90-min catheterization in PARADIGM to correlate the individual TIMI flow rates with the electrocardiographic assessments of patency, previous work with continuous ST segment monitoring has demonstrated a strong correlation between angiographic patency and blinded electrocardiographic ST segment recovery analysis (23). In a series of 22 patients receiving thrombolysis and undergoing 90-min angiographic assessment, Ohman et al. showed that noninvasive patency assessment predicted coronary occlusion with 90% sensitivity and 92% specificity (15). Additionally, the time to steady state variable has been shown in other data sets to be predictive of adverse clinical outcomes at 30 days (15). The advantage of this noninvasive assessment of patency is the relative ease with which it can be employed and the absence of risk, so that reperfusion can be determined more rapidly in a greater number of patients than with conventional 90-min angiography. By all noninvasive measures, thrombolysis plus lamifiban had a favorable profile compared with thrombolysis alone.
There was an increase in bleeding events in the patients treated with lamifiban compared with those receiving placebo. This difference was mainly apparent in Part C, due in large part to the lower incidence of bleeding in the Part C placebo group compared with the Part B placebo group. Like the other small molecule inhibitors of GP IIb/IIIa and unlike the antibody fragment, lamifiban’s effect at the actual receptor level is unknown. There are no data that have directly assessed the extent of integrin blockade with lamifiban, and so there may well be a toxic effect of lamifiban at higher doses. A similar effect of bleeding and worse clinical outcomes was seen with the highest dose of lamifiban, especially when combined with heparin, used in the PARAGON trial of non–ST segment elevation acute coronary syndromes (28).
Although an increase in bleeding is certainly important, the placebo group differences raise uncertainty concerning the generalizability of the observed differences. Much larger sample sizes will be required to accurately assess the bleeding risk of lamifiban when given with thrombolysis. Likewise, the low incidence of intracranial hemorrhage seen in this trial is reassuring, but with such small numbers of patients the estimates of event rates are not reliable. Smaller trials of the antithrombin agents have also had a low incidence of intracranial events, but higher rates were observed when larger numbers of patients were studied (29,30). Unlike the antithrombins, however, this class of drug has demonstrated no increased propensity for intracranial hemorrhage in the aggregated angioplasty and unstable angina trials reported to date.
No reduction in the composite clinical efficacy end point was observed with lamifiban versus placebo. Overall, the mortality rate was lower than the larger contemporary thrombolytic trials (20,31), although this probably reflects the age limitation and other more conservative exclusion criteria typical of a Phase II trial. Additionally, there were differences in event rates between placebo groups, likely reflecting both the small sample sizes and the sequential rather than parallel group enrollment. Reinfarction was higher with lamifiban than placebo in Part B but lower in Part C. Whether this was due to chance or to the effects of the lengthier lamifiban infusion is unknown. Refractory ischemia was lower with lamifiban than with placebo. There were too few 90-min catheterizations to interpret the early TIMI flow rates with confidence.
PARADIGM demonstrates the difficulties in trying to make decisions based on clinical outcomes with such small numbers of patients (32). Although the point estimates are similar, the wide confidence intervals about the composite point estimate do not preclude a possible important benefit to this therapy. The electrocardiographic findings support at least an early benefit to lamifiban. The major limitation of ST segment monitoring, imposed by the size of the monitor and the requirement for bed rest, is that it was limited to the first 24 h following thrombolysis, prohibiting electrocardiographic detection of ischemic events occurring thereafter.
Platelet inhibition, as measured by ADP- and TRAP-induced aggregations, was maximized by the studied lamifiban doses. Lamifiban given with thrombolytic therapy appears to be associated with more rapid and complete reperfusion than placebo, as determined by continuous ST segment monitoring for 24 h. In this small sample there were no obvious clinical benefits to lamifiban over placebo. There are uncertainties as to optimal doses when combining thrombolysis with GP IIb/IIIa inhibition, whether concomitant antithrombin therapy with heparin will be required, whether there are differences among GP IIb/IIIa inhibitors and what might be the optimal intensity and duration of platelet inhibition. Answers to these questions as well as reconciliation of the surrogate end point (ST monitoring) with the clinical outcomes and a more accurate determination of safety will require a larger, adequately powered trial (Appendix).
The authors thank Penny Hodgson for her expert editorial assistance in the preparation of the manuscript.
Clinical Centers (Principal Investigator/Study Coordinator): Belgium: Frans Van de Werf/Anne Luyten; Canada: Benjamin Potkin/Nancy McIntosh-Yellin, Christopher Morgan/Kaye Feskiw, Keith Finnie/Shirley McCreery, Jean Diodati/Ellen Shalit, Sherryn Roth/Jennifer Smith, William Hui/Linda Kvill, Manohara Senaratne/Marlene Goeres, Paul Greenwood/Anne Prosser; Iceland: Arni Kristinsson/Valgerdur Runarsdottir, Gudmundur Oddsson/Hronn Hardardottir; United States: A. Bleakley Chandler/Marcia Edwards, Jerry Becker/Sheila Nalley, Richard Becker/Steven Ball, Eric Bates/Terri Johnson, Robert Harrington/Kirby Quintero, Alexander Paraschos/Leslie Paraschos, Kenneth Lerrick/Beth Spinogatti, N. Bhalodkar/Amanda Valeria, David Moliterno/Phyllis DeSantis, J. King White/Karen Andreatta, John Burks/Nancy Kohler, John Seaworth/Julie Jensen and Gloria Carreon, Thomas Nygaard/Joyce White, Joseph Puma/Laurie Jones.
Coordinating and Executive Centers: Canada: Paul Armstrong, Wanda Sutherland; United States (Cleveland Clinic Foundation): Eric Topol, David Moliterno; United States (Duke Clinical Research Institute): Robert Califf, Robert Harrington, Kristin Newby, Maria Peek, Cynthia Binanay, Ken Tinnin, Manju Bhapkar, Mary Dorsey, Judy Parker.
Clinical Events Committee: Barbara Tardiff, John Alexander, Brian Crenshaw, Cuttina Green, Greg Strand.
Ischemia Monitoring Core Lab: Mitchell Krucoff, Kathleen Trollinger, Cynthia Green.
Angiographic Core Lab: Stephen Ellis, Tom Ivanc, Darrell Debowey.
Data Safety and Monitoring Board: William Hillegass, Michael Lincoff, James Tcheng.
Hoffmann-La Roche, Ltd: Alexis Rames, Beat Steiner, Marianne Bokslag, Debbie Marcantuono, Erik Luedin, Thorkild Nielsen.
Presented in part at the 69th Scientific Sessions of the American Heart Association, New Orleans, Louisiana, November 10–13, 1996.
- adenosine diphosphate
- Platelet Aggregation Receptor Antagonist Dose Investigation and Reperfusion Gain in Myocardial Infarction
- Thrombolysis in Myocardial Infarction
- tissue-plasminogen activator
- thrombin receptor agonist peptide
- Received March 23, 1998.
- Revision received July 10, 1998.
- Accepted August 6, 1998.
- American College of Cardiology
- Anderson J.L.,
- Karagounis L.A.,
- Becker L.C.,
- Sorensen S.G.,
- Menlove R.L.
- Lincoff A.M.,
- Topol E.J.
- Coller B.S.
- Ferguson J.J.
- Ohman EM, Kleiman NS, Gacioch G, et al., for the IMPACT-AMI Investigators
- Kleiman N.S.,
- Ohman E.M.,
- Califf R.M.,
- et al.
- Kouns W.C.,
- Kirchhofer D.,
- Hadvary P.,
- et al.
- Theroux P.,
- Kouz S.,
- Roy L.,
- et al.
- Moliterno DJ, Harrington RA, Califf RM, Rapold HJ, Topol EJ, for the PARADIGM Investigators
- Krucoff M.W.,
- Wagner N.B.,
- Pope J.E.,
- et al.
- ↵Harrington RA, Kleiman NS, Granger CB, Ohman EM, Scarborough RM, Berkowitz SD. Relationship between inhibition of platelet aggregation and clinical outcomes. Am Heart J (in press).
- Simes RJ, Topol EJ, Holmes DR Jr, et al., for the GUSTO-I Investigators
- Tcheng J.E.,
- Ellis S.G.,
- George B.S.,
- et al.
- The PARAGON Investigators
- The GUSTO-IIa Investigators
- Antman E.M.
- Topol EJ, Califf RM, Van de Werf F, et al., for the VIGOUR Group