Author + information
- Received February 5, 2003
- Revision received August 13, 2003
- Accepted August 25, 2003
- Published online January 21, 2004.
- Miles Dalby, MD*,
- Gilles Montalescot, MD, PhD*,* (, )
- Claire Bal dit Sollier, MD†,
- Eric Vicaut, MD, PhD‡,
- Thierry Soulat, MD†,
- Jean-Philippe Collet, MD, PhD*,
- Rémi Choussat, MD*,
- Vanessa Gallois, MSc*,
- Gérard Drobinski, MD, PhD*,
- Ludovic Drouet, MD† and
- Daniel Thomas, MD*
- ↵*Reprint requests and correspondence:
Dr. Gilles Montalescot, Institut de Cardiologie, Bureau 2-236, Pitié-Salpétrière University Hospital, 47 Boulevard de l'Hôpital, 75013 Paris, France.
Objectives The present study hypothesis was that eptifibatide offered further antiplatelet efficacy above clopidogrel in non–ST-elevation myocardial infarction (NSTEMI) patients before an expeditive coronary intervention.
Background Although thienopyridines and glycoprotein (GP) IIb/IIIa antagonists are often co-prescribed in the context of NSTEMI, the antiplatelet interaction of these agents is poorly described and the superiority of GP IIb/IIIa antagonists above thienopyridine treatment alone is not clear.
Methods Thirty-two NSTEMI patients treated with aspirin and enoxaparin were studied using flow cytometry to define parameters of platelet activation with a panel of agonists before clopidogrel, after clopidogrel, and during an eptifibatide infusion following the clopidogrel load.
Results After platelet activation with adenosine diphosphate, thrombin receptor-activating peptide, or U46-619, relative reductions in conformationally activated GP IIb/IIIa receptor expression (evaluated with PAC-1) of 48%, 43%, and 33%, respectively (all p < 0.0001), were seen with clopidogrel, but further 80%, 78%, and 72% (all p < 0.0001) reductions were seen with eptifibatide. With the same agonists, fibrinogen binding was significantly reduced after clopidogrel by 70%, 64%, and 81% (all p < 0.0001) and again further reduced with eptifibatide by 90%, 95%, and 69% (all p < 0.0001). The total number of GP IIb/IIIa receptors (measured as P2 expression) and P-selectin expression fell after clopidogrel, after ex vivo stimulation with the same agonists; however, both parameters increased slightly during the eptifibatide infusion.
Conclusions The activated GP IIb/IIIa expression and fibrinogen binding findings indicate that eptifibatide provides significant potent antiplatelet activity above aspirin and clopidogrel, suggesting additive immediate protection in the treatment of NSTEMI. The P2 and P-selectin findings suggest the possibility of a partial agonist and/or pro-inflammatory effect.
The traditional antithrombotic management of unstable angina and non–ST-elevation myocardial infarction (NSTEMI) is based on aspirin and heparin. A number of newer, more potent antiplatelet agents have been developed separately and incorporated into clinical practice. In addition to aspirin, the adenosine diphosphate (ADP) receptor antagonist clopidogrel has been shown to be of clinical benefit in the treatment of acute coronary syndrome patients (1). Similar positive results have been found with small-molecule glycoprotein (GP) IIb/IIIa antagonists in the medical stabilization of unstable angina/NSTEMI patients, especially in diabetic and troponin-positive patients (2,3). Because they have been developed and trialed independently, however, data on the antiplatelet interaction of these drugs are very limited, and it is not known whether additive effects can be expected, as no study has ever compared a double antiplatelet regimen of aspirin and clopidogrel with a triple regimen of aspirin, clopidogrel, and GP IIb/IIIa antagonist in this situation. Glycoprotein IIb/IIIa antagonists remain expensive drugs requiring IV bolus plus infusion delivery, and it is therefore important to know whether they provide a significant benefit above that of clopidogrel during the period of medical stabilization preceding catheterization. We investigated the effects of clopidogrel and eptifibatide prescribed successively in addition to aspirin and enoxaparin in a cohort of NSTEMI patients before early angiography. The primary objective was to establish whether better control of GP IIb/IIIa receptor expression and fibrinogen (Fg) binding could be achieved with both agents than with clopidogrel alone in the context of the expeditive care of NSTEMI patients. The secondary objective was to evaluate the early effect of a loading dose of clopidogrel on top of aspirin on the same parameters.
Study inclusion criteria included men or non-pregnant women >18 years hospitalized with NSTEMI, defined as angina at rest or crescendo angina, and electrocardiographic (ECG) ST-T-segment changes (ST-segment depression or transient elevation of at least 0.1 mV or T-wave changes in at least 2 ECG leads). A positive troponin I test was a requirement (≥0.1 μg/ml), and coronary angiography had to be indicated. Exclusion criteria were contraindications to aspirin, clopidogrel, or eptifibatide; severe renal insufficiency (creatinine clearance <30 ml/min); vitamin K antagonist treatment in the preceding five days; and treatment with other anticoagulants or thrombolytics.
Study design and hypothesis
The trial was a single-center, platelet end-point study in which each patient was their own control. The NSTEMI patients treated on admission with aspirin, 500 mg IV (followed by 75 mg/day orally), and enoxaparin, 1 mg/kg twice a day subcutaneously, were considered for enrollment in the Platelet activity Extinction in non–Q-wave myocardial infarction with Aspirin, Clopidogrel and Eptifibatide (PEACE) study. Appropriate anti-anginal treatment (beta-blocker, nitrate, calcium antagonist) was used at the discretion of the clinician. When patients were eligible for enrollment, and after giving informed, written consent, the first blood sample (baseline, or T1) was taken on aspirin and enoxaparin treatment. Immediately after this, an oral loading dose (300 mg) of clopidogrel was given. The second sample (clopidogrel effect, or T2) was taken at least 3 h after clopidogrel administration. Immediately after T2, an IV bolus of eptifibatide, 180 μg/kg IV, followed by an infusion of 2 μg/kg per minute, was administered. A third sample was taken >12 h after the start of the eptifibatide infusion (eptifibatide effect, or T3), just before cardiac catheterization.
The primary purpose of the study was to evaluate whether eptifibatide had an additive effect to that of clopidogrel. The secondary objective was to evaluate the effect of clopidogrel. The null hypothesis was that eptifibatide offered no significant antiplatelet effect beyond that of clopidogrel.
Biologic measurements at T1, T2, and T3
Samples of venous blood were collected both on EDTA for platelet count and on PPACK to avoid calcium chelation (75 μmol/l, Haematologic Technologic Inc., Essex Junction, Vermont) for flow cytometry and transferred immediately to the laboratory for analysis. PPACK was chosen as an anticoagulant rather than trisodium citrate because of previous observations that Ca2+chelation afforded by citrate may alter the interaction between GP IIb/IIIa antagonists and platelet receptor by a number of possible mechanisms (4–6). The platelet count and volume were determined using the STKS automatic blood cell counter (Coultronics France SA, Margency, France). Platelet parameters were immediately measured on whole blood (diluted 1/4 with saline for GPs, non-diluted for microparticles). Fluorescein isothiocyanate (FITC) or phycoerythrin (PE)-conjugated monoclonal antibodies (MAb) were from Beckman Coulter (Roissy, France) and Becton Dickinson (Le Pont de Claix, France). Preliminary studies have defined the optimal amount of each antibody that would provide maximum antigen labeling and minimum unspecific binding. The flow cytometer was a Calibur fluorescence-activated cell sorter from Becton Dickinson. Flow cytometry was used to measure the degree of activation of circulating platelets in the resting state and after ex vivo stimulation by 0.6 mmol/l arachidonic acid (Biodata Corp., Horsham, Pennsylvania), 5 μmol/l ADP (Coulter, Margency, France), 25 μmol/l thrombin receptor-activating peptide (TRAP) (Neosystems, Strasbourg, France), and 1 μmol/l U46-619, a representative agonist for the thromboxane A2receptor (Calbiochem, Merck, Darmstadt, Germany). Regarding the doses, a relatively low dose of ADP was chosen because of the high sensitivity of the P2Y12 (thienopyridine-sensitive) ADP receptor subtype (7). Although TRAP is a thrombin agonist, it is very much less potent than thrombin (8), and this dose induced optimal subtotal stimulation. Epinephrine was not used because of the marked heterogeneity of interindividual response (9). Platelet parameters were quantified as follows: total GP IIb/IIIa complexes using P2 (anti-CD41) MAb, activated GP IIb/IIIa complexes using PAC-1 MAb; platelet Fg binding using anti-human Fg chicken MAb (Biopool International, Ventura, California); and P-selectin expression using anti-CD62P MAb. The fluorescence intensity was determined using a flow cytometer (FACScalibur, 3 colors with loader, Becton Dickinson) and Cell Quest software. For quantification, calibration beads in latex were used coated with mouse immunoglobulins (3 μm diameter; 350, 7800, 22,000, and 53,000 immunoglobulins/bead). Flow cytometry defined four peaks of fluorescence intensity, corresponding to the four immunoglobulin densities, and allowed a calibration curve to be drawn (fluorescence intensity as a function of the number of epitopes for the antibody). For platelet analyses, all fluorescence intensities were translated into a number of epitopes per platelet.
Platelet-derived microparticles were assayed using a PE-labeled anti-GP IIb/IIIa MAb (clone P2, anti-CD41) associated with FITC-labeled annexin-V, which binds to activated cellular membrane phospholipids and is not inhibited by the presence of eptifibatide. Additionally, microparticles were discriminated by size using a specific gate defined with 0.8-μm diameter latex beads (Sigma, Lyon, France). The number of microparticles is expressed as a function of the total whole blood volume and as a function of the platelet count.
The main comparison considered in the present study was defined, a priori, as the comparison between T3 and T2, thus exploring the additive effect of eptifibatide after clopidogrel administration. A comparison between T2 and T1 (i.e., effect of clopidogrel) was considered, a priori, as secondary. In order to limit the inflation of the alpha risk, a conservative approach was used. First, any possible difference between the three treatments (i.e., corresponding to 3 different times within the same patient) was tested by analysis of variance (ANOVA). The conservative Huyn-Feldt criteria (10)were used, rather than the classic F test, to estimate the significance level. Only if ANOVA allowed rejection of the null hypothesis of an absence of a difference between the three treatments, were 2 × 2 comparisons made using appropriate contrasts. Because a hierarchy of comparisons was made, a priori, defining a single main comparison, no adjustment of the alpha risk was required at this level. Version 8.1 of SAS (from the SAS Institute) was used in the present study. Results are expressed as the mean value ± SEM. For differences between time points T1-T2 and T2-T3, the mean values of the differences for individual patients are used in the analysis (Δ), and the significance value is calculated for Δ. The relative reduction (RR) or relative increase (RI), calculated as the mean reduction/increase in Δ, is also given. Graphically, the mean values of all patient data at T1, T2, and T3 are presented with corresponding SEMs. A significance level of <0.05 was considered significant for individual comparisons, as is conventionally the case in studies of this kind. However, when correcting for multiple comparisons to increase the robustness of the conclusions, the significance value is p < 0.0036. When the Bonferonni correction results in a significant difference (p < 0.05) becoming non-significant, “corrected NS” is indicated in parenthesis and subsequently discussed.
The protocol was approved by the Pitié-Salpétrière Ethics Committee, and written, informed consent was obtained from all patients.
Thirty-two NSTEMI patients were recruited to the study. Patient characteristics were (mean ± SEM or number [%] of patients): age 64.3 ± 2.8 years, body mass index 26.0 ± 0.82 kg/m2, creatinine clearance 77.3 ± 5.7 ml/min, smoker (n = 15 [47%]), hyperlipidemia (n = 9 [28%]), hypertension (n = 15 [47%]), diabetes (n = 12 [38%]), previous myocardial infarction (n = 8 [25%]), previous coronary artery bypass graft surgery (n = 1 [3%]), previous percutaneous coronary intervention (PCI; n = 5 [16%]), and Killip class I (n = 25 [78%]), class II (n = 1 [3%]), and class III (n = 6 [19%]). The mean troponin I level at presentation was 6.9 ± 2.2 μg/ml.
Seventeen patients underwent PCI (53%), and all PCI procedures involved the deployment of at least one stent. Four patients underwent bypass surgery (13%). Of the remaining 11 patients, one died before angiography, one died before planned bypass surgery, and the remainder had anatomy deemed either not severe enough or not suitable for revascularization. By 30 days, two patients had died (6%) and three others had a myocardial infarction (9%).
In the basal state, reflecting the activation status of circulating platelets, PAC-1 (<1,500) and anti-Fg binding (<200) remained low at all time points (MAb binding expressed as the mean number of binding sites per platelets throughout). Despite these low levels of expression, both parameters were further reduced with clopidogrel and again with eptifibatide. Binding of PAC-1 fell significantly after clopidogrel (Δ = −130, p < 0.0001; RR 37%,) and further with eptifibatide (Δ = −29, p < 0.0001; RR 22%). Anti-Fg binding decreased significantly after clopidogrel (Δ = −22, p < 0.0006; RR 21%) but increased with eptifibatide (Δ = +9, p = NS; RI 37%). Expression of P2 in the basal state did not change significantly with either clopidogrel (Δ = −818, p = NS; RR 3%,) or eptifibatide (Δ = +3071, p = NS; RI 9%). P-selectin expression in the basal state was also low (all values <1,000) at all time points, with no significant differences after clopidogrel (Δ = −22, p = NS; RR 6%) or eptifibatide (Δ = +26, p = NS; RI 18%). Because baseline expression of PAC-1, Fg binding, and P-selectin expression were low in this population (as they were with the relatively weak agonist, arachidonic acid), the results were also expressed as percentage of platelets fixing the relevant antibody. The same comparisons were made, and the exact same results were found (data not shown).
Ex vivo arachidonic acid–induced expression of activation epitopes reflects the degree of aspirin blockade and low levels of both PAC-1 (92 ± 9) and anti-Fg (122 ± 49) seen. Although only 11 of the patients were receiving long-term aspirin therapy, all received 500 mg IV on admission, and the arachidonic acid data indicated a marked aspirin response and no evidence of aspirin resistance in the patients studied. However, further platelet inhibition was measured with a significant decrease in PAC-1 after clopidogrel (Δ = −11, p < 0.0001; RR 13%) and a further significant fall with eptifibatide (Δ = −45, p < 0.0001; RR 46%). Anti-Fg was unchanged after clopidogrel (Δ = −50 p = NS; RR 7%). There was no further change with eptifibatide (Δ = −5 p = NS; RI 9%). P2 expression remained unchanged from baseline (Δ = −26, p = NS; RR 1%) and with eptifibatide (Δ = +84 p = NS; RI 0%). P-selectin tended to decrease after clopidogrel (Δ = −21, p = NS; RR 9%) and tended to increase with eptifibatide (Δ = +25, p = NS; RI 21%).
With ADP ex vivo platelet activation, the ADP responsiveness pathway is specifically tested, and there were significant reductions in PAC-1 expression (Δ = −9,128, p < 0.0001; RR 48%), Fg binding (Δ = −54,777, p < 0.0001; RR 71%) (Fig. 1), P2 expression (Δ = −5,717, p < 0.017 [corrected NS]; RR 9%), and P-selectin expression (Δ = −874, p < 0.001; RR 38%) (Fig. 2) after clopidogrel. With eptifibatide, marked further significant reductions were seen with both PAC-1 expression (Δ = −7,485, p < 0.0001; RR 80%) and Fg binding (Δ = −10,896, p < 0.0001; RR 90%) (Fig. 1). Both P2 (Δ = +14,321, p < 0.018 [corrected NS]; RI 34%) and P-selectin (Δ = +273, p < 0.012 [corrected NS]; RI 47%) expression increased significantly, however (Fig. 2).
With TRAP as the agonist, when testing the thrombin responsiveness pathway, there were significant reductions in PAC-1 expression (Δ = −8,677, p < 0.0001; RR 43%), Fg binding (Δ = −46,682, p < 0.0001; RR 64%) (Fig. 1), and P-selectin expression (Δ = −1,574, p < 0.0023; RR 15%) (Fig. 2) after clopidogrel, whereas there was a non-significant trend toward a fall in P2 (Δ = −4,137, p = NS; RR 5%). With eptifibatide, there was a further significant reduction in PAC-1 expression (Δ = −8,421, p < 0.0001; RR 78%) and Fg binding (Δ = −22,093, p < 0.0001; RR 95%), whereas there was a trend toward a rise in both P2 (Δ = +7,844, p = NS; RI 15%) and P-selectin (Δ = +1,114, p = NS; RI 29%), although neither reached significance (Fig. 2).
With U46-619 as the agonist, when testing the thromboxane A2responsiveness pathway, there were significant reductions in PAC-1 expression (Δ = −9,820, p < 0.0001; RR 33%), Fg binding (Δ = −97,252, p < 0.0001; RR 81%) (Fig. 1), P2 expression (Δ = −8,997, p < 0.026 [corrected NS]; RR 11%), and P-selectin expression (Δ = −4,071, p < 0.0001; RR 48%) (Fig. 2) after clopidogrel. With eptifibatide, there was a further significant reduction in PAC-1 expression (Δ = −11,116, p < 0.0001; RR 72%) and Fg binding (Δ = −33,821, p < 0.0001; RR 69%) (Fig. 1). However, P2 expression rose significantly with eptifibatide (Δ = +23,915, p < 0.0022; RI 53%), and P-selectin expression showed a marked trend toward increasing (Δ = +4,059, p = NS; RI 199%).
There was a non-significant fall in the platelet microparticles as a function of blood volume (Δ = −17,635, p = NS; RR 8.%) after clopidogrel, but a significant further fall with eptifibatide (Δ = −31,991, p < 0.001; RR 32%). The platelet microparticles as a function of platelet count did not change with clopidogrel (Δ = −109, p = NS; RR 8%), but it fell significantly with eptifibatide (Δ = −165, p < 0.0005; RR 32%) (Fig. 3).
There was no significant change in the baseline (pre-clopidogrel) platelet count or volume after clopidogrel or eptifibatide treatment (data not shown).
Low levels of conformationally active GP IIb/IIIa expression, as revealed by PAC-1 binding and Fg binding, were measured in the baseline state (in patients treated with aspirin and enoxaparin), indicating low but measurable activation levels of circulating platelets in the study patients with NSTEMI. These levels are similar to those seen in our previous study using similar methods, demonstrating low levels of peripheral blood platelet activation in such patients treated with aspirin and enoxaparin, indicating a consistency of technique (8). Other investigators have observed similar levels (11). The apparent discrepancy between the PAC-1 and Fg binding levels may be due to the fact that Fg can polymerize; thus, the anti-Fg MAb may bind several Fg ligands associated with a single GP IIb/IIIa receptor.
Although the biologic relevance of such low levels of epitopes could be questioned, both activated GP IIb/IIIa expression and Fg binding fell significantly after clopidogrel and fell further with eptifibatide, although this was only significant for PAC-1. The ex vivo arachidonic acid stimulation indicated a significant aspirin effect in all patients with profound inhibition of the arachidonic acid pathway and low levels of IIb/IIIa activation, Fg binding, and P-selectin activation. Interestingly, there was further suppression of IIb/IIIa expression and a trend toward reduced Fg binding with clopidogrel, suggesting further inhibition of the arachidonic acid pathway beyond the aspirin effect.
With ADP, TRAP, and U46-619 as the ex vivo agonists of platelet stimulation, clopidogrel reduced the induction of PAC-1 binding site expression and Fg binding. With eptifibatide, however, both parameters were markedly and significantly further reduced consistently with all three agonists. This major finding of the study indicates that GP IIb/IIIa activation induced by these potent agonists is only partially suppressed by clopidogrel alone and is far better controlled by the addition of eptifibatide therapy (Fig. 1). The similar pattern observed with platelet microparticles further confirms the profound added platelet inhibitory effect with eptifibatide when given in addition to aspirin and clopidogrel, and importantly, as it reflects in vivo change, it further supports the validity of the ex vivo stimulation findings. This additive antiplatelet effect may be particularly important in high-risk patients such as the study NSTEMI patients heading to the catheterization laboratory for angiography and often PCI, triggering further platelet activation. There was no temporal control in this study, because the patients were managed according to current clinical guidelines (12); however, we have shown that in acute coronary syndrome patients, there is no appreciable temporal change over 48 h when antiplatelet therapy is not modified (8). There was no group without eptifibatide because, as discussed, patients were managed according to current guidelines and standard local practice at our institution, allowing ethical approval by our Institutional Review Board. Equally, there was no group with eptifibatide and no clopidogrel, for the same reasons. Regarding the use of the ANOVA test before further analysis, this allows more robust conclusions and reinforces the conclusions of the present study. The secondary findings with P2 and P-selectin expression were unexpected. Although ex vivo–stimulated expression of both parameters with ADP, TRAP, and U46-619 decreased after clopidogrel, they both increased slightly during eptifibatide infusion. This reached statistical significance with ADP for both parameters and with U46-619 for P2, while the same trend was seen with TRAP. Analysis of the individual data points indicated that this was a group effect and not due to a few outliers. The statistical power was limited by the sample size, and we cannot exclude the possibility of chance findings. Furthermore, when applying Bonferonni correction, although there was a fall in P-selectin with all three agonists, the only significant rise was in P2 with U46; however, the same pattern was consistently seen with three separate agonists. Platelet degranulation could account for both P-selectin expression and increased P2 MAb binding due to release of granular GP IIb/IIIa receptors (not measured while intragranular). Alternatively, there could be a direct GP IIb/IIIa receptor upregulatory effect. The possibility of a partial agonist effect of GP IIb/IIIa antagonists inducing platelet activation and Fg binding has been raised in the past (13–16), although whether biologically important aggregation is induced in physiologic conditions at therapeutic doses is debated (17). However, our finding of increased P-selectin expression is in keeping with that of Cox et al. (16)in the OPUS-Thrombolysis In Myocardial Infarction (TIMI)-16 trial and similarly offers a possible mechanism for the observed detrimental effects of long-term oral GP IIb/IIIa antagonist therapy (18)and prolonged IV GP IIb/IIIa antagonist infusions without early PCI (19), the latter being potentially exaggerated by ongoing shear forces due to untreated coronary stenoses (20). One could further hypothesize that prolongation of the eptifibatide infusion in PEACE may have resulted in still further P-selectin expression and P2 increase. This raises the question of a possible pro-thrombotic and/or pro-inflammatory effect through platelet leukocyte aggregation (20,21), although these secondary findings regarding P2 and P-selectin must still be regarded as very preliminary and explorative rather than demonstrative in view of the small sample size of the study.
In the short term, the potent anti-aggregatory effect of GP IIb/IIIa receptor blockade is overwhelming and probably explains the marked efficacy of these agents in PCI (22–25), including when patients have been pretreated by a thienopyridine (26–30). Our data also strongly support the recently published post-hoc analysis of the patients in the Platelet IIb/IIIa Underpinning the Receptor for Suppression of Unstable Ischemia Trial (PURSUIT), demonstrating that the benefit of eptifibatide was confined to patients having very early intervention (31). It also supports the concept of a stronger double-bolus regimen and a shorter infusion of eptifibatide in PCI (26). Together, the results from PEACE and the most recent clinical data support the concept of “expeditive care” of NSTEMI patients with short infusions of GP IIb/IIIa receptor blockers and early PCI (22–25)for an extended long-term benefit (32,33).
In high-risk patients, the superior control of activated platelet GP IIb/IIIa receptor expression and Fg binding obtained with eptifibatide, in addition to clopidogrel, suggests that clopidogrel alone may not be sufficient as the “poor man's IIb/IIIa antagonist.” Our findings suggest that GP IIb/IIIa blockade may be particularly beneficial in the short-term treatment of high-risk patients undergoing early angiography, thus also avoiding the possible downsides of a prolonged infusion. Eptifibatide remains a potent, rapidly effective antiplatelet agent, whereas clopidogrel provides an intermediate, slower antiplatelet effect but also a long-term secondary prevention after NSTEMI and stenting.
This work will be presented in part at the Scientific Sessions of the American College of Cardiology, Chicago, Illinois, March 2003.
- adenosine diphosphate
- monoclonal antibody
- non–ST-elevation myocardial infarction
- percutaneous coronary intervention
- Platelet activity Extinction in non–Q-wave myocardial infarction with Aspirin Clopidogrel and Eptifibatide study
- relative increase
- relative reduction
- thrombin receptor-activating peptide
- Received February 5, 2003.
- Revision received August 13, 2003.
- Accepted August 25, 2003.
- American College of Cardiology Foundation
- Phillips D.R.,
- Teng W.,
- Arfsten A.,
- et al.
- Mousa S.A.,
- Bozarth J.M.,
- Forsythe M.S.,
- et al.
- Cattaneo M.,
- Zighetti M.L.,
- Lombardi R.,
- et al.
- Montalescot G.,
- Bal-dit-Sollier C.,
- Chibedi D.,
- et al.
- Kabbani S.S.,
- Watkins M.W.,
- Holoch P.A.,
- Terrien E.F.,
- Sobel B.E.,
- Schneider D.J.
- ↵Braunwald E, Antman EM, Beasley JW, et al. Guideline update for management of patients with unstable angina and non–ST-segment elevation myocardial infarction: ACC/AHA 2002. J Am Coll Cardiol 2002;40:366–74
- Honda S.,
- Tomiyama Y.,
- Aoki T.,
- et al.
- Peter K.,
- Schwarz M.,
- Ylanne J.,
- et al.
- Schneider D.J.,
- Taatjes D.J.,
- Sobel B.E.
- Cox D.,
- Smith R.,
- Quinn M.,
- et al.
- Frelinger A.L. 3rd.,
- Furman M.I.,
- Krueger L.A.,
- et al.
- Chew D.P.,
- Bhatt D.L.,
- Sapp S.,
- et al.
- Quinn M.J.,
- Plow E.F.,
- Topol E.J.
- Li N.,
- Hu H.,
- Lindqvist M.,
- et al.
- ↵The ESPRIT Investigators. Novel dosing regimen of eptifibatide in planned coronary stent implantation (ESPRIT): a randomised, placebo-controlled trial. Lancet 2000;356:2037–44
- Steinhubl S.R.,
- Ellis S.G.,
- Wolski K.,
- et al.
- Bonz A.W.,
- Lengenfelder B.,
- Strotmann J.,
- et al.
- Ronner E.,
- Boersma E.,
- Akkerhuis K.M.,
- et al.
- Anderson K.M.,
- Califf R.M.,
- Stone G.W.,
- et al.