Platelet glycoprotein IIb/IIIa blockers during percutaneous coronary intervention and as the initial medical treatment of non-ST segment elevation acute coronary syndromes

  • Conclusions changed
  • Review
  • Intervention

Authors


Abstract

Background

During percutaneous coronary intervention (PCI), and in non-ST segment elevation acute coronary syndromes (NSTEACS), the risk of acute vessel occlusion by thrombosis is high. Glycoprotein IIb/IIIa blockers strongly inhibit platelet aggregation and may prevent mortality and myocardial infarction. This is an update of a Cochrane review first published in 2001, and previously updated in 2007 and 2010.

Objectives

To assess the efficacy and safety effects of glycoprotein IIb/IIIa blockers when administered during PCI, and as initial medical treatment in patients with NSTEACS.

Search methods

We updated the searches of the Cochrane Central Register of Controlled Trials (CENTRAL) on The Cochrane Library (Issue 12, 2012), MEDLINE (OVID, 1946 to January Week 1 2013) and EMBASE (OVID, 1947 to Week 1 2013) on 11 January 2013.

Selection criteria

Randomised controlled trials comparing intravenous IIb/IIIa blockers with placebo or usual care.

Data collection and analysis

Two authors independently selected studies for inclusion, assessed trial quality and extracted data. We collected major bleeding as adverse effect information from the trials. We used odds ratios (OR) and 95% confidence intervals (CI) for effect measures.

Main results

Sixty trials involving 66,689 patients were included. During PCI (48 trials with 33,513 participants) glycoprotein IIb/IIIa blockers decreased all-cause mortality at 30 days (OR 0.79, 95% CI 0.64 to 0.97) but not at six months (OR 0.90, 95% CI 0.77 to 1.05). All-cause death or myocardial infarction was decreased both at 30 days (OR 0.66, 95% CI 0.60 to 0.72) and at six months (OR 0.75, 95% CI 0.64 to 0.86), although severe bleeding was increased (OR 1.39, 95% CI 1.21 to 1.61; absolute risk increase (ARI) 8.0 per 1000). The efficacy results were homogeneous for every endpoint according to the clinical condition of the patients, but were less marked for patients pre-treated with clopidogrel, especially in patients without acute coronary syndromes.

As initial medical treatment of NSTEACS (12 trials with 33,176 participants), IIb/IIIa blockers did not decrease mortality at 30 days (OR 0.90, 95% CI 0.79 to 1.02) or at six months (OR 1.00, 95% CI 0.87 to 1.15), but slightly decreased death or myocardial infarction at 30 days (OR 0.91, 95% CI 0.85 to 0.98) and at six months (OR 0.88, 95% CI 0.81 to 0.96), although severe bleeding was increased (OR 1.29, 95% CI 1.14 to 1.45; ARI 1.4 per 1000).

Authors' conclusions

When administered during PCI, intravenous glycoprotein IIb/IIIa blockers reduce the risk of all-cause death at 30 days but not at six months, and reduce the risk of death or myocardial infarction at 30 days and at six months, at a price of an increase in the risk of severe bleeding. The efficacy effects are homogeneous but are less marked in patients pre-treated with clopidogrel where they seem to be effective only in patients with acute coronary syndromes. When administered as initial medical treatment in patients with NSTEACS, these agents do not reduce mortality although they slightly reduce the risk of death or myocardial infarction.

Résumé scientifique

Inhibiteurs des glycoprotéines plaquettaires IIb-IIIa au cours d'une intervention coronarienne percutanée et en tant que traitement médical initial des syndromes coronariens aigus sans élévation du segment ST

Contexte

Au cours d'une intervention coronarienne percutanée (ICP) et dans les syndromes coronariens aigus sans élévation du segment ST (SCA non ST), il existe un risque élevé d'occlusion aiguë des vaisseaux par thrombose. Les inhibiteurs IIb/IIIa ont un puissant effet inhibiteur de l'agrégation plaquettaire et peuvent prévenir la mortalité et l'infarctus du myocarde (IM). Cette étude est une mise à jour d'une revue Cochrane publiée pour la première fois en 2001 et précédemment mise à jour en 2007 et 2010.

Objectifs

Évaluer les effets et l'innocuité des inhibiteurs IIb-IIIa lorsqu' ils sont administrés pendant l'ICP et en tant que traitement médical initial chez les patients atteints de SCA non ST.

Stratégie de recherche documentaire

Nous avons mis à jour les recherches dans le registre Cochrane des essais contrôlés (CENTRAL) dans la Bibliothèque Cochrane (Numéro 12, 2012), MEDLINE (OVID, de 1946 à la 1ère semaine de janvier 2013) et EMBASE (OVID, de 1947 à la semaine 1 2013) le 11 janvier 2013.

Critères de sélection

Les essais contrôlés randomisés comparant des inhibiteurs IIb-IIIa intraveineux à un placebo ou à des soins habituels.

Recueil et analyse des données

Deux auteurs ont sélectionné les études à inclure, évalué la qualité des essais et extrait les données de manière indépendante. Nous avons recueilli les informations concernant les hémorragies graves en tant qu'effet indésirable. Des rapports des cotes et des intervalles de confiance (IC) à 95% ont été utilisés pour mesurer l'effet.

Résultats principaux

Soixante essais impliquant 66 689 patients ont été inclus. Pendant l'ICP (48 essais totalisant 33 513 participants), les inhibiteurs des glycoprotéines IIb-IIIa réduisaient la mortalité à 30 jours, toutes causes confondues (rapport des cotes de 0,79, IC à 95% 0,64 à 0,97), mais pas à 6 mois (RC 0,90, IC à 95% 0,77 à 1,05). Les décès de toutes causes confondues ou les infarctus du myocarde étaient réduits à 30 jours (rapport des cotes de 0,66, IC à 95% 0,60 à 0,72) et à 6 mois (RC 0,75, IC à 95% 0,64 à 0,86), mais les hémorragies sévères augmentaient (rapport des cotes de 1,39, IC à 95% 1,21 à 1,61; augmentation du risque absolu (IRA) 8,0 pour 1 000). Les résultats de l'efficacité étaient homogènes pour tous les critères de jugement en fonction de l'état clinique des patients, mais étaient moins prononcés chez les patients prétraités au clopidogrel, en particulier chez les patients sans SCA.

En tant que traitement médical initial de SCA non ST (12 essais totalisant 33 176 participants), les inhibiteurs IIb-IIIa ne réduisaient pas la mortalité à 30 jours (RC 0,90, IC à 95% 0,79 à 1,02) ou à 6 mois (RC 1,00, IC à 95% 0,87 à 1,15), mais réduisaient légèrement les décès ou les infarctus du myocarde à 30 jours (rapport des cotes de 0,91, IC à 95% 0,85 à 0,98) et à 6 mois (RC 0,88, IC à 95% 0,81 à 0,96), avec toutefois une augmentation des hémorragies graves (RC 1,29, IC à 95% 1,14 à 1,45; IRA 1,4 pour 1 000).

Conclusions des auteurs

Lorsqu' ils sont administrés pendant l'ICP, les intraveineuses glycoprotéines plaquettaires IIb-IIIa réduisent le risque de décès toutes causes confondues à 30 jours et à 6 mois, et réduisent le risque de décès ou d'infarctus du myocarde à 30 jours et à six mois, mais sont associés à une augmentation du risque d'hémorragie grave. L'efficacité est homogène mais moins prononcée chez les patients prétraités au clopidogrel, à l'exception des patients atteints de syndrome coronarien aigu. Lorsqu'ils sont administrés en tant que traitement médical initial chez des patients atteints de SCA non ST, ces agents ne réduisent pas la mortalité, mais réduisent légèrement le risque de décès et d'IM.

Plain language summary

Platelet glycoprotein IIb/IIIa blockers during percutaneous coronary intervention and as the initial treatment of acute coronary syndromes

During the last two decades, doctors have been looking for the best treatment to prevent clots in the coronary arteries of patients with coronary heart disease. This review summarises the results of 60 studies which used a potent class of intravenous antiplatelet drugs - glycoprotein IIb-IIIa blockers - in 66,689 participants. This treatment was tested in two different conditions: in patients undergoing coronary angioplasty (inserting a deflated small balloon in the artery and expand it to open the vessel) with or without inserting a stent (a thin metal expandable tube or sleeve to scaffold the artery open); and as the initial treatment of patients hospitalised for unstable angina and non-ST segment elevation acute myocardial infarction (prolonged coronary chest pain with or without small myocardial infarction). The evidence is up to date as January 2013 and the overall quality of the body of evidence was good.

Overall, the use of these drugs during coronary angioplasty with or without inserting a stent reduced the risk of death at 30 days, and the risk of death or myocardial infarction at 30 days and at six months. The results were similar for stable and for unstable patients with coronary artery disease, but there was comparatively less benefit for patients previously treated with clopidogrel, an oral antiplatelet drug. On the other side, these drugs only slightly reduced the risk of death or myocardial infarction when administered as initial medical treatment in patients with unstable angina or non-ST-elevation myocardial infarction. The benefits of glycoprotein IIb/IIIa blockers need to be balanced against the increased risk of bleeding.

Résumé simplifié

Inhibiteurs des glycoprotéines plaquettaires IIb-IIIa au cours d'une intervention coronarienne percutanée et en tant que traitement médical initial des syndromes coronariens aigus sans élévation du segment ST

Depuis ces vingt dernières années, les médecins cherchent à identifier le traitement le plus efficace pour éviter la formation de caillots dans les artères coronaires des patients atteints de maladie coronarienne. Cette revue résume les résultats de 60 études qui utilisaient une classe de médicaments antiplaquettaires intraveineux puissants, les inhibiteurs des glycoprotéines IIb-IIIa - chez 66 689 participants. Ce traitement était testé dans deux pathologies différentes: chez les patients soumis à une angioplastie coronaire (pose d'un petit ballonnet dégonflé dans l'artère puis gonfler pour ouvrir le vaisseau) avec ou sans pose d'une endoprothèse vasculaire métallique (tube fin expansible pour maintenir l’artère ouverte); et en tant que traitement initial chez des patients hospitalisés pour cause d'angine instable et d'infarctus aigu du myocarde sans élévation du segment ST (coronaires prolongées des douleurs thoraciques avec ou sans infarctus du myocarde de petite taille). Les preuves sont à jour en janvier 2013 et la qualité globale de l'ensemble des preuves était bonne.

Dans l'ensemble, l'utilisation de ces médicaments pendant une angioplastie coronaire, en introduisant ou non une endoprothèse vasculaire, réduisait le risque de décès à 30 jours, et le risque de décès ou d'infarctus du myocarde à 30 jours et à six mois. Les résultats étaient similaires chez les patients atteints d'une maladie coronarienne stable et instable, mais les bénéfices étaient comparativement inférieurs chez les patients préalablement traités au clopidogrel, un nouveau médicament antiplaquettaire oral. Néanmoins, ces médicaments ne réduisaient que légèrement le risque de décès ou d'infarctus du myocarde lorsqu' ils étaient administrés en tant que traitement médical initial chez des patients atteints d'angine instable ou d'infarctus du myocarde sans élévation du segment ST. Les bénéfices des inhibiteurs des glycoprotéines IIb-IIIa doivent être mis en rapport avec le risque accru d'hémorragie sévère.

Notes de traduction

Traduit par: French Cochrane Centre 23rd March, 2014
Traduction financée par: Instituts de Recherche en Sant� du Canada, Minist�re de la Sant� et des Services Sociaux du Qu�bec, Fonds de recherche du Qu�bec-Sant� et Institut National d'Excellence en Sant� et en Services Sociaux

Laienverständliche Zusammenfassung

Glycoprotein IIb/IIIa-Hemmer bei perkutanen Koronarinterventionen und als initiale Behandlung akuter Koronarsyndrome

Während der letzten zwei Jahrzehnte haben Ärzte nach der besten Behandlung zur Vorbeugung von Blutgerinnseln in den Koronararterien von Patienten mit koronarer Herzkrankheit gesucht. Dieser Review fasst die Ergebnisse von 60 Studien zusammen, die eine starke Gruppe von intravenösen Thrombozytenaggregationshemmern - Glykoprotein-IIb-IIIa-Blocker - bei 66.689 Teilnehmern eingesetzt haben. Diese Behandlung wurde für zwei verschiedene Erkrankungen getestet: bei Patienten nach Koronarangioplastie (Einführen von einem nicht aufgeblasenen kleinen Ballon in die Arterie, der erweitert wird, um die Gefäße zu öffnen) mit oder ohne Einsetzen eines Stents (ein dünnes erweiterbares Metallrohr, um die Arterie offen zu halten); und als initiale Behandlung von Patienten, die mit instabiler Angina und akutem Nicht-ST-Hebungsinfarkt (verlängerte Schmerzen in der Brust mit oder ohne kleinen Myokardinfarkt) ins Krankenhaus eingeliefert wurden. Die Evidenz ist auf dem Stand von Januar 2013 und die Gesamtqualität der Evidenz war gut.

Insgesamt verringerte die Verwendung dieser Arzneimittel während einer koronaren Angioplastie mit oder ohne Einsetzen eines Stents das Sterberisiko nach 30 Tagen, und das Sterberisiko oder das Risiko für einen Myokardinfarkt nach 30 Tagen und sechs Monaten. Die Ergebnisse waren ähnlich für stabile und instabile Patienten mit koronarer Herzkrankheit, aber es gab vergleichsweise weniger Nutzen für Patienten, die zuvor mit Clopidogrel, einem oralen Thrombozytenaggregationshemmer behandelt wurden. Andererseits, reduzierten diese Medikamente das Sterberisiko oder das Riskio für einen Herzinfarkt nur geringfügig, wenn sie als initiale medikamentöse Behandlung bei Patienten mit instabiler Angina oder Nicht-ST-Hebungsinfarkt verabreicht wurden. Die Nutzen von Glykoprotein IIb/IIIa-Hemmern müssen gegen ein erhöhtes Blutungsrisiko abgewogen werden.

Anmerkungen zur Übersetzung

K. Kunzweiler und I. Töws, Koordination durch Cochrane Schweiz.

Background

Description of the condition

Cardiovascular diseases were the direct cause of more than four million deaths in Europe around the year 2000 (1.9 million in the European Union), accounting for 43% of all deaths at all ages in men and 55% in women. Coronary artery disease (CAD) accounts for half of this mortality burden and depends mostly on the occurrence of acute coronary syndromes (ACS) (ESC Guidelines CV Prevention 2012).

According to the current European and American guidelines, most patients with an ACS should be referred for coronary angiography, as they recommend percutaneous coronary intervention (PCI) as the preferred treatment option for both patients with ST-segment elevation myocardial infarction (STEMI), and patients with non-ST elevation ACS (NSTEACS) at moderate to high risk of ischaemic events (ACCF/AHA Guideline NSTEACS 2013; ESC Guidelines NSTEACS 2011). In addition, PCI is the preferred revascularisation technique for patients with stable CAD. As a consequence, the number of PCI procedures is rapidly increasing (ACCF/AHA Guideline revascularization 2012; ESC Guidelines revascularization 2010).

The mechanism of PCI using balloon angioplasty, with or without stent implantation, includes profound vessel injury and plaque rupture, all of which triggers an immediate activation of the coagulation cascade, and adhesion, activation and aggregation of platelets (ACCF/AHA Guideline revascularization 2012; ESC Guidelines revascularization 2010). Pre-treatment with aspirin, ticlopidine and heparin has been shown to reduce by 70% the risk of acute vessel occlusion and of myocardial infarction in these patients (ATC 2002). Bivalirudin, a new anticoagulant drug that is a direct thrombin inhibitor, has also been shown to have similar efficacy to unfractionated heparin with less bleeding (ACUITY 2006; HORIZONS-AMI 2008). When a stent is implanted, the addition of clopidogrel instead of ticlopidine, before the procedure and during follow-up, has also been shown to be of benefit (PCI-CURE 2001). Thus, treatment with aspirin, clopidogrel and heparin or bivalirudin has been the standard antithrombotic treatment for this procedure (ACCF/AHA Guideline NSTEACS 2013; ESC Guidelines NSTEACS 2011). Recently, new P2Y12 receptor antagonists, like prasugrel and ticagrelor, have been demonstrated to provide quicker, greater and more consistent inhibition of platelet activation than clopidogrel (Patrono 2011), and to decrease the incidence of ischaemic events in patients with ACS submitted for PCI (PLATO 2009; TRITON TIMI-38 2007).

The pathophysiology of non-ST segment elevation ACS (NSTEACS), i.e. unstable angina and non-ST segment elevation myocardial infarction, involves the rupture or erosion of an atherosclerotic coronary plaque (Falk 1995), activation of the coagulation cascade, and adhesion, activation and platelet aggregation. Treatment with aspirin and fractionated or unfractionated heparin has been shown to reduce the risk of cardiac events by 50% in patients with this syndrome (ATC 2002; Mehta 2003). The addition of clopidogrel has also been shown to decrease further the risk of vascular events by 20% in these patients (CURE 2001). Finally, fondaparinux, a new anticoagulant, has been shown to have similar efficacy with less bleeding effects than heparin (OASIS-5 2006). Thus, the current standard antithrombotic treatment for patients with NSTEACS is the administration of aspirin and clopidogrel with heparin or fondaparinux as recommended in current clinical guidelines (ACCF/AHA Guideline NSTEACS 2013; ESC Guidelines NSTEACS 2011). Also, ticagrelor has been shown to be more effective than clopidogrel in reducing ischaemic events in patients with ACS treated invasively or conservatively (PLATO 2009), while prasugrel has been shown to be effective in patients with NSTEACS treated invasively (TRITON TIMI-38 2007).

Description of the intervention

The glycoprotein IIb/IIIa integrin present in platelets mediates the final common pathway in platelet aggregation, spawning the development of glycoprotein IIb/IIIa receptor blockers (Phillips 1988). Intravenous glycoprotein IIb/IIIa inhibitors rapidly block up to 80% of the glycoprotein IIb/IIIa platelet receptors and inhibit fibrinogen-mediated platelet aggregation (Gurbel 2005; Patrono 2011).

The three glycoprotein IIb/IIIa receptor inhibitors approved for clinical use are intravenous agents belonging to different classes: abciximab is a monoclonal antibody fragment; eptifibatide is a cyclic peptide; and tirofiban is a peptidomimetic molecule. All of them have a short half-life and are eliminated by the kidney, which is the reason why they are only indicated in patients without severe renal impairment (glomerular filtration rate < 30) . Glycoprotein IIb/IIIa receptor inhibitors have been shown to be effective in reducing ischaemic events in patients undergoing PCI (ACCF/AHA Guideline revascularization 2012; ESC Guidelines revascularization 2010), while tirofiban, lamifiban and eptifibatide have been tested as initial medical treatment in patients with NSTEACS (ACCF/AHA Guideline NSTEACS 2013; ESC Guidelines NSTEACS 2011). Current guidelines give a class IIa recommendation to these drugs for their administration during PCI with a level of evidence B (data derived from a single randomised clinical trial or large non-randomised studies) (ACCF/AHA Guideline NSTEACS 2013; ACCF/AHA Guideline revascularization 2012; ESC Guidelines NSTEACS 2011; ESC Guidelines revascularization 2010).

How the intervention might work

In patients submitted for PCI and also in patients presenting with a NSTEACS, intravenous glycoprotein IIb/IIIa blockade induces strong platelet inhibition (Gurbel 2005; Patrono 2011) and may prevent mural and intraluminal thrombus formation. This effect can prevent acute coronary occlusion and embolisation of plaque thrombi to the distal microvasculature resulting in acute myocardial infarction and death.

Why it is important to do this review

In spite of their known effects in preventing myocardial infarction, some doubts remain as to the efficacy of these drugs in decreasing mortality, the homogeneity of their effects in different subgroups of patients or PCI techniques used, and their additional benefit when patients are pre-treated with clopidogrel (NICE 2010). In addition, since glycoprotein IIb/IIIa receptor blockers induce profound platelet inhibition, the risk of bleeding complications is increased, particularly when administered concomitantly with other antiplatelet drugs and with high-dose heparin treatment (Quinn 2002). Thus, safety is an important issue in the management of patients with these drugs, especially during PCI.

This is an update of a Cochrane review first published in the year 2001, and previously updated in 2007 and in 2010. In the last update, the main conclusion was that when administered during PCI, intravenous IIb/IIIa blockers reduce the risk of all-cause death and of death or myocardial infarction at 30 days and at six months, but increase the risk of severe bleeding. In contrast, when administered as initial medical treatment in patients with NSTEACS, these drugs do not reduce all-cause mortality, slightly reduce the risk of death or myocardial infarction, and increase the risk of severe bleeding.

In the last three years, several studies have been performed with these drugs in the setting of PCI or in patients presenting with NSTEACS (ASSIST 2009; Bhattacharya 2010; INSTANT 2012; ITTI 2012; Yan 2009), while information on other small studies has become available (DANTE 2004; De Luca 2005; De Luca 2008; Gasior 2003; Kim 2005; Kurowski 2005; Prati 2005). In addition, higher doses of clopidogrel as a loading dose (i.e. 600 mg) to achieve a more rapid and potent antiplatelet effect are being used, and new anticoagulant drugs (i.e. bivalirudin and fondaparinux, OASIS-5 2006) have been incorporated in the usual medical treatment of patients with ACS and patients referred for PCI (ACCF/AHA Guideline NSTEACS 2013; ESC Guidelines NSTEACS 2011). Since the efficacy of intravenous IIb/IIIa blockers and risk of bleeding could be different with higher doses of clopidogrel, and in the presence of these new drugs, an update of this systematic review is justifiable.

Objectives

The aim of this systematic review was to assess the efficacy and safety effects of glycoprotein IIb/IIIa blockers given in addition to standard medical treatment when administered:

  1. during PCI;

  2. as the initial medical treatment of patients with NSTEACS ('upstream treatment').

The first indication relates to a procedure rather than a health problem, where the studied treatment is administered a few minutes prior to PCI. The second indication relates to a clinical condition that includes a wide range of patient groups at different risk, where the studied treatment is administered at admission in patients with or without coronary angiography and PCI during hospitalisation. Both questions are addressed in the same review to better understand the efficacy of IIb/IIIa antagonists in patients undergoing PCI, including in patients with stable ischaemic heart disease and in patients with acute coronary syndromes with or without ST-segment elevation. The standard medical treatment considered was aspirin and ticlopidine or clopidogrel, plus heparin or bivalirudin during PCI; and aspirin with or without clopidogrel, and heparin or fondaparinux in the initial medical treatment of patients with NSTEACS.

Methods

Criteria for considering studies for this review

Types of studies

We sought to identify all randomised controlled clinical trials, with or without blinding, studying intravenous glycoprotein IIb/IIIa blockers and in which at least one of the pre-defined primary outcomes was measured. Specifically, we analysed trials performed when IIb/IIIa blockers were administered during PCI in patients with or without an ACS, and trials performed in patients with NSTEACS treated from admission with glycoprotein IIb/IIIa antagonists as the initial medical management.

We did not consider trials:

  • on oral glycoprotein IIb/IIIa blockers, since clinical research has indicated that their administration is associated with increased mortality, and they are not available for clinical use;

  • performed in patients with ST-segment elevation acute myocardial infarction (STEMI) treated with thrombolytics or with facilitated or rescue PCI after thrombolytic treatment;

  • comparing two different glycoprotein IIb/IIIa blockers;

  • comparing the timing of the glycoprotein IIb/IIIa blockers administration (i.e. pre-hospital administration versus hospital administration or administration in the emergency room versus in the catheterisation laboratory); or

  • comparing the route of administration (i.e. intracoronary versus intravenous).

Types of participants

The following participants were considered for each of the two types of studies that we considered in this review:

  1. Studies that randomised adults (18 years and older), male or female, with or without an acute coronary syndrome, who underwent elective or urgent PCI, with or without stent implantation.

  2. Studies that included adult patients with NSTEACS in whom glycoprotein IIb/IIIa antagonists were administered at the time of hospital admission as part of the initial medical management.

For both indications there were no exclusion criteria regarding patients' comorbidities, the presence of renal dysfunction or the risk of bleeding.

Types of interventions

Intravenous glycoprotein IIb/IIIa blockers administered as a bolus followed by 12 to 96 hours infusion, at any dose, on top of usual care (aspirin and ticlopidine or clopidogrel, and heparin or bivalirudin).

Comparators

A control group with or without placebo, on top of usual care (aspirin and ticlopidine or clopidogrel, and heparin or bivalirudin).

For every major endpoint, two subgroup analyses have been performed in the PCI group:

  • glycoprotein IIb/IIIa blockers compared with placebo (double-blind studies) on top of usual care;

  • glycoprotein IIb/IIIa blockers compared with usual care (not blinded studies).

Types of outcome measures

Primary outcomes
  • All-cause mortality at 30 days and at six months.

  • All-cause death or non-fatal myocardial infarction at 30 days and at six months.

All-cause mortality was chosen instead of cardiovascular mortality or fatal myocardial infarction because although most deaths occurring early after an ACS or a PCI procedure are due to fatal myocardial infarction, total mortality does not depend on physician interpretation of the cause. In addition, in all of the reviewed studies total mortality was considered.

The combined endpoint of all-cause death or myocardial infarction was chosen because most deaths occurring early after a NSTEACS or a PCI procedure are due to an acute coronary occlusion leading to a myocardial infarction.

Because definitions of myocardial infarction may vary among studies, especially in those performed during PCI, when the data were obtainable we used as a definition of post-procedural myocardial infarction an elevation more than three times the upper limit of normal of the biomarker of necrosis used (creatine kinase-MB fraction (CK-MB) or cardiac troponins) occurring within 48 hours of the procedure. Spontaneous or non-post-procedural myocardial infarction was defined in the presence of signs or symptoms of ischaemia not related to or appearing more than 48 hours after a PCI procedure and with elevation more than two times the upper limit of normal of the biomarker used to assess of necrosis.

Secondary outcomes
  • Need for an urgent revascularisation procedure at 30 days and at six months for patients who underwent PCI.

Urgent revascularisation was chosen as a secondary endpoint as this is a consequence of refractory ischaemia and because its indication is often clinician-driven and, in consequence, not entirely objective.

  • All-cause death, non-fatal myocardial infarction or urgent coronary revascularisation at 30 days and at six months.

All of the components of this combined endpoint share the same pathophysiology (i.e. acute vessel occlusion) also as a secondary endpoint.

  • Safety: severe bleeding at 30 days.

This was the most important safety outcome that was appropriately described in all studies. To decrease definition variability (Steinhubl 2007) and wherever reported, the Thrombolysis In Myocardial Infarction (TIMI) classification was used to define severe bleeding (Bovill 1991); the investigators' definition was used otherwise.

Search methods for identification of studies

Electronic searches

We searched the Cochrane Central Register of Controlled Trials (CENTRAL) on The Cochrane Library (2012, Issue 12), MEDLINE OVID (1946 to January Week 1 2013) and EMBASE Classic + EMBASE OVID (1947 to 2013 Week 01) on 11 January 2013. We checked and revised search strategies for the updates. See Appendix 1; Appendix 2; Appendix 3 and Appendix 4 for details of the search strategies. We loaded all records into Reference Manager and removed duplicates (see Acknowledgements).

Searching other resources

We carried out handsearches (to January 2013) of abstracts from conferences published in Circulation (American Heart Association Annual Meeting), Journal of the American College of Cardiology (American College of Cardiology Foundation Annual Congress), European Heart Journal (European Society of Cardiology Annual Congress) and online at Theheart.org, TIMI.org. We also searched Clinicaltrials.org. We also examined the references sections of reviews and meta-analysis addressing glycoprotein IIb/IIIa inhibitors.

No language restrictions were applied to the search. English, French and Spanish language papers were read by the authors. The Cochrane Heart Group helped us by translating two Chinese articles (see Acknowledgements). All published studies could be retrieved from the bibliographic search.

Data collection and analysis

Selection of studies

Two authors screened and independently extracted results. A paper was rejected only when both review authors agreed that the article did not meet the inclusion criteria. Two review authors scored each manuscript independently using a pre-designed in/out form. Each manuscript was scored 1 (definitely 'in', i.e. the study met the inclusion criteria); 2 (maybe 'in', i.e. small sample size or difficult to determine the type of control or of diagnosis); or 3 (definitely 'out', i.e. definitely did not meet the inclusion criteria). All manuscripts which scored at least one '1' or one '2' were analysed in common by both review authors.

Data extraction and management

Two authors independently extracted data from original reports of trial results and resolved differences by discussion. We abstracted data on characteristics of participants (age, sex, presence of diabetes, prior myocardial infarction, ACS and type), interventions (PCI and type, stent and type, pre-treatment with clopidogrel), outcomes, trial quality characteristics (i.e. sequence generation, allocation concealment, blinding, incomplete outcome report, incomplete data addressing, selective reporting and other) onto this form. In addition, we collected data on potential confounding factors including participants' baseline risk and characteristics, trial duration, intensity of intervention (dosing and duration of treatment), type and dose of concomitant medication used and revascularisation procedures including stent implantation.

Assessment of risk of bias in included studies

Two authors independently assessed risk of bias in the included studies using The Cochrane Collaboration's 'Risk of bias' tool (Higgins 2011). We created a 'Risk of bias' table for each study by adding the degree of bias for the primary and secondary endpoints according to sequence generation, allocation concealment, blinding, incomplete outcome report, incomplete data addressing, selective reporting and other.

In addition, we further analysed the risk of bias by including two tables with the summary assessment of the risk of bias for major and minor endpoints (across key domains) within and across studies. We selected allocation concealment and blinding as the key domains for this analysis.

Only the effect of the intravenous route of administration (i.e. bolus plus infusion) was examined and compared with placebo/control with usual care (i.e. aspirin, ticlopidine/clopidogrel and heparin). To prevent attrition bias, we analysed all patients allocated to active treatment in this group regardless of whether they received it or not (i.e. intention-to-treat analysis). We used intention-to-treat data in the pooled analyses of this review as it was possible to obtain them from the publications of all the studies.

Data synthesis

We calculated pooled odds ratio (OR) of the individual and combined endpoints. Fixed-effect meta-analyses are presented unless a Chi2 test for heterogeneity was statistically significant at a 5% level, in which case random-effects meta-analyses are presented. We calculated risk differences after pooling together the studies for each meta-analysis.

Subgroup analysis and investigation of heterogeneity

In the analysis of the effects of glycoprotein IIb/IIIa blockers administered during PCI, we performed six separate subgroup analyses:

1. According to the clinical condition of the patients:

  • Stable coronary artery disease in which PCI was performed as an elective procedure

  • NSTEACS

  • Primary PCI in patients with STEMI

2.The technique used:

  • Balloon angioplasty alone

  • PCI with stent implantation

3. Pre-treatment with clopidogrel

  • Patients with ACS

  • Patients without ACS

For all subgroup analyses we selected studies in which at least two-thirds of patients had the focused condition or procedure. For the analysis of patients pre-treated with clopidogrel and according to the known dose-related onset of action of this drug (Hochholzer 2005), we selected only trials in which patients were receiving chronic clopidogrel treatment (five days or more), and trials in which a loading dose of 300 mg was administered at least six hours before the procedure or 600 mg at least two hours before the procedure.

We assessed heterogeneity of studies by clinical judgement according to differences in type of patients enrolled, study quality, interventions and outcome. We also evaluated funnel plots for every major endpoint.

Sensitivity analysis

In the PCI meta-analysis, we made two different analyses in the overall group according to the presence or absence of a high risk of bias for the primary endpoint of the review. Group 1 were studies at low to intermediate risk of bias (i.e. comparison with placebo, double-blind and with adequate allocation concealment); and Group 2 were studies at high risk of bias (i.e. comparison with usual care, without placebo and not blinded). We made a sensitivity analysis by comparing the results for each group and for the two groups together.

Results

Description of studies

Results of the search

The search strategy was designed to be very sensitive and yielded 4106 documents from the original and the 2007, 2010 and 2013 updates of the review. We screened all references from relevant reviews and meta-analyses (see Additional references) for studies possibly missed by our search strategy. In total 184 articles were retrieved for further examination; 20 of these were from the current update. One hundred and two references to 92 studies were excluded (see Characteristics of excluded studies). Thirteen new RCTs were identified from the current update. The previous version of the review included 48 RCTs so that in total 60 trials (66,689 participants) were included in the review. Of these, 48 trials (33,513 participants) were on glycoprotein IIb/IIIa blockers administered during PCI, and 12 trials (33,176 participants) tested these drugs as initial medical treatment in patients with NSTEACS (Figure 1).

Figure 1.

Study flow diagram.

Included studies

The three glycoprotein IIb/IIIa receptor inhibitors approved for clinical use are intravenous agents that belong to different classes: abciximab is a monoclonal antibody fragment; eptifibatide is a cyclic peptide; and tirofiban is a peptidomimetic molecule.

Most of the studies were multicentre, international studies and were performed mostly in North America and Europe, Canada, the Netherlands, France, Belgium, Germany, Spain, Poland and Italy. An important number were also performed in south American countries, Australia, New Zealand and the Czech Republic. Only eight studies were performed in low to middle-income countries (ASIAD 2005; Bhattacharya 2010; Chen 2000; Fu 2008; ITTI 2012; Kim 2005; Shen 2008; Yan 2009).

During percutaneous coronary intervention

These studies were performed in stable or unstable coronary patients undergoing elective or urgent PCI with or without stent implantation, and IIb/IIIa blockers were administered during the procedure. In all the studies performed during elective PCI, randomisation was performed when the exact coronary anatomy was known and inclusion criteria required the presence of one or more > 70% stenosis in a coronary segment amenable to PCI.

Most of the studies excluded patients with renal dysfunction, since this condition is associated with excess dosing, and those at high risk of bleeding.

PCI was the condition studied in 48 studies with 33,513 patients. Forty-four studies reported outcomes at 30 days while 34 studies reported outcomes at 6 to 12 months.
Among the 48 studies and according to the clinical condition of the patients, there were 18 studies in patients with stable CAD, seven studies in patients with NSTEACS and 11 studies performed during primary PCI in patients with ST-segment elevation myocardial infarction (STEMI); the other 12 studies were performed in a mixed population. According to the technique used, there were 11 studies performed during balloon angioplasty (10 of them performed before year 2000), 33 during PCI with stent implantation and four with mixed techniques. Finally, 17 studies including 8925 patients were performed in patients pre-treated with clopidogrel, three in patients with NSTEACS (ISAR-REACT 2 2006; Kurowski 2005; Yan 2009), six in patients with STEMI (ASSIST 2009; BRAVE-3 2009; Fu 2008; ITTI 2012; On-TIME 2004; Shen 2008) and eight in patients with stable CAD (3T/2R 2009; Claeys 2005; Cuisset 2008; INSTANT 2012; ISAR-REACT 2004; ISAR-SWEET 2004; ISAR-SMART-2 2004; OPTIMIZE-IT 2009; TOPSTAR 2002).

All patients included in trials with stent implantation received heparin, aspirin and ticlopidine or clopidogrel during and after the procedure except for the ERASER 1999 study in which ticlopidine was left to the investigator's discretion. In only two studies (CLEAR PLATELETS-2 2009; OPTIMIZE-IT 2009) a drug-eluting stent was used in most of the patients. Abciximab was used in 28 trials, tirofiban in 13 and eptifibatide in seven. The doses varied among studies.

As initial medical treatment of patients with non-ST segment elevation acute coronary syndromes

These studies were performed in patients with a recent (< 24 hours) chest pain at rest associated with ischaemic changes in the admission ECG (ST-segment depression > 0.5 mm or transient elevation > 1 mm, > 1 mm T-wave inversion) or CK-MB or troponin elevation above the upper limits of normality in each participant institution. Most of these trials were performed in patients managed conservatively. All of the studies excluded patients with moderate to severe renal dysfunction and those at high risk of bleeding

Twelve studies with 33,176 patients concerned glycoprotein IIb/IIIa use as initial medical treatment in patients with NSTEACS. All these studies presented 30-day follow-up results and six presented six months follow-up. Abciximab was used in one study, eptifibatide in three, tirofiban in five and lamifiban in three. No subgroup analyses were performed since all of the studies included patients with the same global condition.

PRISM 1998 differed from the other trials in that the glycoprotein IIb/IIIa blocker was given without heparin. PRISM Plus 1998 and PARAGON A 1998 also included an arm with glycoprotein IIb/IIIa blocker without heparin. Seven studies were performed following a conservative management while only three (ELISA-2 2006; PRACTICE 2007; PRISM Plus 1998) were performed on an invasive basis with most of the patients scheduled for early coronary angiography.

Variability

The potential sources of heterogeneity among the studies may include the variability in patient characteristics. In the PCI group the mean age ranged from 59 years to 70 years with a median of 61 years, while in the group of studies on initial medical treatment of NSTEACS the mean age ranged from 60 to 65 years. Considering all studies, the proportion of males ranged from 61% to 95%, and the proportion of patients with prior myocardial infarction, which was described in most studies, ranged from 10% to 67% (see Characteristics of included studies table).

In the PCI analysis, 12 trials were performed in patients with STEMI, 11 of them during primary PCI. The frequency of this diagnosis was of 0% in 26 other studies and ranged from 3% to 41% in the other 11 trials. A specific analysis of those 11 trials has been performed.

In the analysis of glycoprotein IIb/IIIa blockers administered as initial medical treatment in patients with NSTEACS, the prevalence of positive markers of myocardial necrosis ranged from 14% to 100%. Thirty-two to 80% of patients had ST-segment depression at enrolment. Nine studies were performed following a conservative management with less than 14% of patients having in-hospital PCI. In the other three studies, coronary angiography was performed in 60% to 90% of patients, and PCI from 31% to 61% of patients during drug infusion which was administered for 24 to 72 hours after enrolment.

The daily doses of aspirin ranged in the studies from 50 mg to 500 mg and those of heparin were typically aimed at maintaining an activated clotting time of > 200 seconds or an activated partial thromboplastin time between 50 and 85 seconds, or twice that of laboratory control. Only in one study (Schulman 1996) was aspirin not allowed in the treatment group but administered in the placebo group.

Some studies had arms of active treatment with lower heparin doses (EPILOG 1997) or no heparin at all (PARAGON A 1998; PRISM Plus 1998; PRISM 1998). All patients from these studies were included in the present update since a prior meta-analysis performed with individual patient data reported similar results by including or excluding those patients (Boersma 2002). Patients from the control groups of all studies received heparin.

Dosing

Where trials had intervention arms with varying doses of glycoprotein IIb/IIIa inhibitor drugs, data from such intervention arms were pooled. In the EPIC 1994 study, one intervention arm had glycoprotein IIb/IIIa blocker administered as a bolus alone, i.e. with no subsequent perfusion; the results of that arm were not included.

Endpoints

All studies followed more than 95% of patients at 30 days and more than 90% at six months. In general the methodological quality of the 60 selected RCTs was good, but differed in PCI and non-PCI studies. Ten of the 49 PCI trials were considered to be at high risk of bias (Table 1). However, these 10 studies with 3863 participants represented only 11.5% of the total and have been studied separately in this review for each major endpoint. In contrast, among studies on IIb/IIIa blockers administered as initial medical treatment in patients with NSTEACS, only one small study was considered to have a high risk of bias (Table 2)

Table 1. Summary assessment of the risk of bias (allocation concealment and blinding) for major endpoints within and across PCI studies
  1. Allocation concealment and blinding were the 2 selected key domains.

    Bias for each endpoint: NA: Not applicable; Low: Plausible bias unlikely to seriously alter the results; Unclear: Plausible bias that raises some doubt about the results; High: Plausible bias that seriously weakens confidence in the results.

    Bias within a study: Low: Low risk of bias for all key domains; Unclear: Unclear risk of bias for one or more key domains; High: High risk of bias for one or more key domains.

    Bias across studies: Low: Most information is from studies at low risk of bias; Unclear: Most information is from studies at low or unclear risk of bias; High: The proportion of information from studies at high risk of bias is sufficient to affect the interpretation of the results.

StudyN30-day mortality6-month mortality30-day death or MI6-month death or MIMajor bleedingWithin study
EPIC 19942099LowLowLowLowLowLow
Simoons 199460LowNALowNALowLow
IMPACT 1995150LowNALowNAUnclearLow
Kereiakes 199693LowNALowNALowLow
RESTORE 19972141LowLowLowLowLowLow
IMPACT-II 19974010LowNALowNALowLow
EPILOG 19972792LowLowLowLowLowLow
CAPTURE 19971265LowLowLowLowLowLow
EPISTENT 19982399LowLowLowLowLowLow
RAPPORT 1998483LowLowLowLowLowLow
ERASER 1999215LowLowLowLowLowLow
Galassi 1999106HighNAHighNAHighHigh
Chen 200042UnclearNAUnclearNAUnclearUnclear
ESPRIT 20002064LowLowLowLowLowLow
ISAR-2 2000401UnclearNAHighNAHighHigh
PRIDE 2001127UnclearNAUnclearNAUnclearUnclear
ADMIRAL 2001300LowLowUnclearUnclearUnclearUnclear
Tamburino 2002107HighHighHighHighHighHigh
TOPSTAR 200296LowLowUnclearUnclearUnclearUnclear
Juergens 2002894LowLowLowUnclearUnclearUnclear
CADILLAC 20022082LowLowHighHighHighHigh
ACE 2003400HighHighHighHighHighHigh
Gasior 200341UnclearUnclearUnclearUnclearUnclearUnclear
ADVANCE 2004202LowLowUnclearUnclearUnclearUnclear
DANTE 200496LowLowLowLowUnclearLow
ISAR SMART-2 2004502NALowNALowLowLow
ISAR-REACT 20042159LowLowLowLowLowLow
ISAR-SWEET 2004701LowLowLowLowLowLow
Claeys 2005200UnclearUnclearHighHighHighHigh
ASIAD 2005254LowLowUnclearUnclearUnclearUnclear
De Luca 2005122NAUnclearNAUnclearNAUnclear
Kurowski 200550LowLowUnclearUnclearNAUnclear
Prati 2005140NANAUnclearNANAUnclear
ISAR-REACT 2 20062022LowLowLowLowLowLow
FU 2008150NAUnclearNAUnclearUnclearUnclear
Cuisset 2008149LowNAHighNAHighHigh
Shen 2008172UnclearUnclearHighHighHighHigh
On-TIME 2 2008984LowNALowNALowLow
De Luca 2008132LowLowLowLowLowLow
OPTIMIZE-IT 200946LowLowHighHighHighHigh
JEPPORT 2009973LowNALowNALowLow
CLEAR PLATELETS-2 2009200LowLowHighHighHighHigh
BRAVE-3 2009800LowNALowNALowLow
3T/2R 2009263LowNALowNALowLow
ASSIST 2009400LowLowLowLowLowLow
Yan 2009240LowLowLowLowUnclearLow
INSTANT 201291NALowNAUnclearUnclearUnclear
ITTI 2012100NALowNAUnclearUnclearUnclear
ACROSS STUDIES33,474LowLowUnclearUnclearUnclear 
Table 2. Summary assessment of the risk of bias (allocation concealment and blinding) for major endpoints within and across studies on initial treatment of patients with NSTEACS
  1. Allocation concealment and blinding were the 2 selected key domains.

    Bias for each endpoint: NA: Not applicable; Low: Plausible bias unlikely to seriously alter the results; Unclear: Plausible bias that raises some doubt about the results; High: Plausible bias that seriously weakens confidence in the results.

    Bias within a study: Low: Low risk of bias for all key domains; Unclear: Unclear risk of bias for one or more key domains; High: High risk of bias for one or more key domains.

    Bias across studies: Low: Most information is from studies at low risk of bias; Unclear: Most information is from studies at low or unclear risk of bias; High: The proportion of information from studies at high risk of bias is sufficient to affect the interpretation of the results.

StudyN30-day mortality6-month mortality30-day death or MI6-month death or MIMajor bleedingWithin study
CANADIAN 1996365LowNALowNALowLow
Schulman 1996227LowNALowNALowLow
PURSUIT 199810,948LowLowLowLowLowLow
PRISM 19983232LowNALowNALowLow
PARAGON A 19982282LowLowLowLowLowLow
PRISM Plus 19981915LowLowLowLowLowLow
GUSTO-IV 20017800LowNALowNALowLow
PARAGON B 20025225LowNALowNALowLow
Kim 2005160LowLowUnclearUnclearUnclearUnclear
ELISA-2 2006328LowNAHighNAHighHigh
PRACTICE 2007393LowLowLowLowLowLow
Bhattacharya 2010301UnclearNAUnclearNAUnclearUnclear
ACROSS STUDIES33,176LowLowLowLowLow 

We pooled three studies describing endpoints occurring within follow-up periods shorter than 30 days (i.e. in-hospital, seven-day, etc.) with 30-day follow-up studies (ERASER 1999; Kereiakes 1996; Simoons 1994) due to the fact that most adverse events occur within the first week after the procedure. Also, one study (ADVANCE 2004) that only reported events at six months was also included in the 30-day follow-up analysis in order not to miss important information, on the basis that > 80% of mortality and myocardial infarction and > 95% of severe bleeding occurred during the first 30 days. Finally, we pooled three studies (ISAR-REACT 2004; ISAR-REACT 2 2006; ISAR-SMART-2 2004) that reported one-year follow-up alone with the six-month follow-up analysis for the same reason.

The definition of primary and secondary outcomes varied among studies. However, it was possible to obtain or calculate the number of cases with the primary and secondary endpoints of this review in most studies. Myocardial infarction was part of the composite effectiveness endpoint of all trials, but the applied myocardial infarction definition was different, especially regarding the increased levels of the MB fraction of creatine kinase required in studies performed before the year 2005 and of troponin levels thereafter. This point was especially important regarding the definition of post-PCI myocardial infarction. Because of this, and when data were obtainable, we used as a definition of post-procedural myocardial infarction an elevation more than three times the upper limit of normal of the biomarker of necrosis used (creatinine kinase-MB fraction (CK-MB) or cardiac troponins).

The secondary endpoint 'major bleeding' was assessed by the Thrombolysis In Myocardial Infarction (TIMI) classification when described (bleeding was classified as major if it involved intracranial haemorrhage or cardiac tamponade or if it was associated with a decrease in haemoglobin concentration of more than 50 g/L regardless of whether or not a bleeding site had been identified (Bovill 1991). Otherwise, we used the 'major' bleeding described in each study. When no description of major bleeding existed at all, we selected intracranial haemorrhages and need for transfusion as 'major' bleeding.

In some instances, it was not specified whether the nature of PCI during follow-up was urgent or not. In such cases, we considered overall revascularisations.

Excluded studies

Most of the excluded studies were not RCTs, did not compare a IIb/IIIa blocker with placebo or control, did not report clinical events or were performed in conditions other than during PCI or as initial medical treatment of patients with NSTEACS (Figure 1; see Characteristics of excluded studies).

Risk of bias in included studies

It is important to note that most of the trials excluded patients at high risk of bleeding and patients with renal failure. In addition, the mean age of the studied population was lower (61 years) than what it is usually observed in clinical practice (CRUSADE 2006; GRACE 2007).

Allocation

Reporting of adequate sequence generation was obtained in 70% of the studies and adequate allocation concealment in 54% of them. Only in 10% of the studies was the allocation concealment inadequate (or considered to be at high risk of bias (Higgins 2011) (Figure 2; Figure 3).

Figure 2.

Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.

Figure 3.

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

Blinding

Patients and clinicians were adequately blinded to treatment (Higgins 2011) in most of the trials. Some studies were performed on an open-label basis while in others (Claeys 2005; ISAR-2 2000) the treatment strategy precluded blinding of the used drugs. Overall, for the primary endpoint of the review, blinding of patients and investigators was adequate in 55% of the studies, unclear in 15% and inadequate in 30% of them. For the secondary endpoints, blinding was adequate in 50%, unclear in 15% and inadequate in 35% (Figure 2; Figure 3).

Regarding allocation concealment and blinding as the two major key domains, the risk of bias across PCI studies was low for mortality and unclear for death or myocardial infarction, and for major bleeding (Table 1). In studies performed with the use of IIb/IIIa blockers as initial medical treatment in patients with NSTEACS, the risk of bias across studies was low for every major endpoint (Table 2).

Incomplete outcome data

Outcome data for the primary endpoint was adequately addressed (Higgins 2011) in 85% of the studies and for the secondary endpoints in 80% of them (Figure 2; Figure 3)

Selective reporting

Eighty-five per cent of the studies were free of suggestion of selective outcome reporting (Higgins 2011) (Figure 2).

Other potential sources of bias

Almost all studies were industry funded, enhancing the risk of bias (Als-Nielsen 2003; Higgins 2011). In fact, only 30% of them were apparently free of other problems that could put them at a risk of bias. We considered 50% of the studies to have an unclear risk of other sources of bias and 20% were at high risk because of the relationship between investigators and industry, premature stopping of the study or other reasons (Figure 2; Figure 3).

Effects of interventions

1. Glycoprotein IIb/IIIa blockers during percutaneous coronary intervention (PCI)

Primary endpoints

Of the 48 PCI trials with 33,513 participants, data from 43 trials with 31,744 patients (95%) were available on 30-day mortality, and 44 trials with 31,880 participants on 30-day mortality or myocardial infarction. Twenty-eight of these trials with 27,205 patients (86%) were blinded studies with a placebo group and considered to be at low risk of bias. Six-month data on mortality and on mortality or myocardial infarction were available from 33 trials with 24,845 patients (74%), of which 20 trials (20,329 patients) were blinded and at low risk of bias (Table 2).

All-cause mortality

Mortality occurred in 0.92% of patients in the treatment group versus 1.33% in controls at 30 days, and in 2.44% and 2.94% respectively at six months. Treatment with intravenous glycoprotein IIb/IIIa blockers was associated with a significant reduction in the odds of mortality at 30 days (odds ratio (OR) 0.79, 95% confidence interval (CI) 0.64 to 0.97). The results were homogeneous (I2 = 0%) and similar for blinded (0.78, 95% CI 0.61 to 1.00) and unblinded studies (0.81, 95% CI 0.54 to 1.21); Analysis 1.1 and Figure 4. The absolute risk reduction (ARR) per thousand treated patients was of 4.0 and the number needed to treat (NNT) to save a life was 249. The results were also similar for all the different subgroups that were considered according to the clinical condition of the patients (Analysis 2.1; Analysis 3.1; Analysis 4.1) and the technique used (Analysis 5.1; Analysis 6.1), although they were less marked for patients pre-treated with clopidogrel (OR 0.83, 95% CI 0.57 to 1.20; Analysis 7.1; Table 3).

Figure 4.

Funnel plot of comparison: 1 During PCI (all patients), outcome: 1.1 30-day mortality.

Table 3. Main results for the primary outcomes
Intervention30-day mortality 6-month mortality 30-day death or non-fatal MI6-month death or non-fatal MI
  OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)
1. During PCI (all patients)0.79 (0.64 to 0.97)0.90 (0.77 to 1.05)0.66 (0.60 to 0.72)0.75 (0.64 to 0.86)
1.1. Subgroup analysis by patient's condition
Patients with stable CAD0.71 (0.34 to 1.46)0.88 (0.63 to 1.22)0.68 (0.55 to 0.84)0.77 (0.64 to 0.92)
Patients with NSTEACS0.77 (0.46 to 1.28)1.01 (0.74 to 1.38)0.68 (0.56 to 0.83)0.77 (0.65 to 0.92)
Primary PCI in patients with STEMI0.88 (0.64 to 1.21)0.88 (0.68 to 1.15)0.78 (0.61 to 1.00)0.77 (0.51 to 1.16)
 
1.2. Subgroup analysis by technique
Balloon angioplasty0.79 (0.55 to 1.14)1.06 (0.75 to 1.50)0.65 (0.56 to 0.75)0.78 (0.65 to 0.94)
PCI with stent placement0.78 (0.58 to 1.04)0.87 (0.71 to 1.07)0.67 (0.59 to 0.76)0.72 (0.60 to 0.87)
 
1.3. Subgroup analysis by pre-treatment with clopidogrel0.83 (0.57 to 1.20)0.95 (0.75 to 1.21)0.81 (0.67 to 0.98)0.87 (0.66 to 1.14)
Patients with ACS0.84 (0.56 to 1.25)0.97 (0.72 to 1.31)0.77 (0.61 to 0.96)0.77 (0.46 to 1.27)
Patients without ACS0.78 (0.29 to 2.09)0.92 (0.63 to 1.35)0.91 (0.65 to 1.28)0.94 (0.73 to 1.21)
 
2. As initial medical treatment of NSTEACS0.90 (0.79 to 1.02)1.00 (0.87 to 1.15)0.91 (0.85 to 0.98)0.88 (0.81 to 0.96)
MI, myocardial infarction; PCI, percutaneous coronary intervention; CAD, coronary artery disease; NSTEACS, non-ST segment elevation acute coronary syndrome; STEMI, ST-segment elevation acute myocardial infarction; ACS, acute coronary syndromes; OR, odds ratio; CI, confidence interval

However, at six months treatment with glycoprotein IIb/IIIa inhibitors was not associated with a reduction of mortality in the overall group (OR 0.90, 95% CI 0.77 to 1.05, I2 = 0%; Analysis 1.2; Figure 5) and in each subgroup; Table 3).

Figure 5.

Funnel plot of comparison: 1 During PCI (all patients), outcome: 1.2 6-month mortality.

All-cause death or myocardial infarction

The rate of death or myocardial infarction at 30 days and six months was of 5.05% versus 7.44% and 7.45% versus 10.05% in the treatment and control groups respectively. Glycoprotein IIb/IIIa blockers were associated with a significant decrease in the odds of death or myocardial infarction at 30 days (OR 0.66, 95% CI 0.60 to 0.72, P < 0.00001; I2 = 22%) without significant differences between blinded and unblinded studies (Analysis 1.3; Figure 6). The ARR per thousand treated patients was of 23.9 and the NNT was 42. The results were similar in all subgroups but were less marked in patients pre-treated with clopidogrel (OR 0.81, 95% CI 0.67 to 0.98), especially in patients without ACS (OR 0.91, 95% CI 0.65 to 1.28; Analysis 7.3; Table 3).

Figure 6.

Funnel plot of comparison: 1 During PCI (all patients), outcome: 1.3 30-day mortality or myocardial infarction.

The results at six months showed marked heterogeneity (I2 = 47%) but were similar to those obtained at 30 days (OR 0.75, 95% CI 0.64 to 0.86, P < 0.0001; Analysis 1.4). The ARR per thousand treated patients was of 26.0 and the NNT was 39. Again, the results were similar in all the subgroups of patients considered in the review but were less marked for patients pre-treated with clopidogrel (OR 0.87, 95% CI 0.66 to 1.14; Analysis 7.4; Table 3).

Secondary endpoints

Data on urgent coronary revascularisation and the combined endpoint of death, myocardial infarction or urgent revascularisation at 30 days were available from 42 trials with 31,555 patients (94% of patients included in the overall review) and from 32 trials with 21,548 patients (64% of patients included in the overall review) at six months.

Urgent revascularisation

Urgent coronary revascularisation was performed in 2.04% of patients in the treatment group versus 3.44% in controls at 30 days, and in 12.54% and 15.31% respectively at six months. Treatment with intravenous glycoprotein IIb/IIIa blockers was associated with a reduction in the risk of urgent revascularisation at 30 days (OR 0.62, 95% CI 0.54 to 0.71, P < 0.00001; I2 = 18%) (Analysis 1.5) and at six months (OR 0.85, 95% CI 0.79 to 0.93, P < 0.0001; I2 = 9%) (Analysis 1.6). At 30 days, the results were less marked for patients with stable coronary artery disease (CAD) (OR 0.84, 95% CI 0.54 to 1.32) and for those pre-treated with clopidogrel (OR 0.89, 95% CI 0.62 to 1.28), especially in patients without ACS; while at six months, the results were similar in all subgroups (Table 4).

Table 4. Main results for the secondary outcomes
Intervention30-day urgent revasc 6-month revasc 30-day death, MI or revasc6-month death, MI or revasc
  OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)
1. During PCI (all patients)0.62 (0.54 to 0.71)0.85 (0.79 to 0.93)0.65 (0.57 to 0.75)0.80 (0.73 to 0.88)
1.1. Subgroup analysis by patient's condition
Patients with stable CAD0.84 (0.54 to 1.32)0.93 (0.81 to 1.08)0.74 (0.60 to 0.91)0.87 (0.77 to 0.98)
Patients with NSTEACS0.68 (0.52 to 0.89)0.91 (0.79 to 1.05)0.69 (0.58 to 0.81)0.85 (0.76 to 0.96)
Primary PCI in patients with STEMI0.61 (0.45 to 0.83)0.75 (0.63 to 0.90)0.69 (0.49 to 0.98)0.84 (0.73 to 0.97)
 
1.2. Subgroup analysis by technique
Balloon angioplasty0.58 (0.49 to 0.70)0.81 (0.60 to 1.10)0.63 (0.51 to 0.76)0.84 (0.75 to 0.94)
PCI with stent placement0.75 (0.58 to 0.98)0.87 (0.78 to 0.97)0.69 (0.57 to 0.84)0.78 [0.69 to 0.89)
 
1.3. Subgroup analysis by pre-treatment with clopidogrel0.89 (0.62 to 1.28)0.87 (0.76 to 0.98)0.88 (0.65 to 1.20)0.88 (0.79 to 0.98)
Patients with ACS0.82 (0.54 to 1.24)0.77 (0.64 to 0.94)0.80 (0.50 to 1.26)0.81 (0.69 to 0.95)
Patients without ACS1.17 (0.54 to 2.53)0.94 (0.80 to 1.11)1.08 (0.76 to 1.52)0.94 (0.81 to 1.10)
revasc, revascularisation; MI, myocardial infarction; PCI, percutaneous coronary intervention; CAD, coronary artery disease; NSTEACS, non-ST segment elevation acute coronary syndrome; STEMI, ST-segment elevation myocardial infarction; ACS, acute coronary syndromes; OR, odds ratio; CI, confidence interval
Death, myocardial infarction or urgent revascularisation

The combined endpoint of death, myocardial infarction or urgent revascularisation at 30 days and six months was 6.70% versus 9.71% and 19.01% versus 23.22% in the treatment and control groups respectively. IIb/IIIa blockers were also associated with a lower risk of death, myocardial infarction or urgent revascularisation both at 30 days (OR 0.65, 95% CI 0.57 to 0.75) (Analysis 1.7) and at six months (OR 0.80, 95% CI 0.73 to 0.88) (Analysis 1.8), although the analysis showed a marked heterogeneity of the results both at 30 days (I2 = 40%) and at six months (I2 = 35%). Again, the global results were similar to those obtained in all subgroups but were less marked for the subgroup of patients pre-treated with clopidogrel (30-day OR 0.88, 95% CI 0.65 to 1.20; six-month OR 0.88, 95% CI 0.79 to 0.98), especially in patients without ACS (Table 4).

Safety

Data were available from 40 trials with 31,430 patients (94% of patients included in the overall review). Major bleeding occurred in 3.03% of patients in the treatment group versus 2.23% in controls. Treatment with intravenous glycoprotein IIb/IIIa blockers was associated with an increased risk of severe bleeding (OR 1.39, 95% CI 1.21 to 1.61; P < 0.0001; I2 = 12%; Analysis 1.9). The absolute risk increase per thousand treated patients over 30 days was 8.0 and the number needed to harm (NNH) was 125. The results were homogeneous in blinded and in unblinded studies (Analysis 7.9), and in all subgroups (Table 5).

Table 5. Main results for safety outcomes
Intervention30-day major bleeding (OR, 95% CI)
1. PCI (all patients)1.39 (1.21 to 1.61)
1.1. Subgroup analysis by patient's condition
Patients with stable CAD1.83 (1.10 to 3.03)
Patients with NSTEACS1.41 (1.03 to 1.93)
Primary PCI in patients with STEMI1.54 (1.12 to 2.11)
 
1.2. Subgroup analysis by technique
Balloon angioplasty1.38 (1.02 to 1.86)
PCI with stent placement1.38 (1.05 to 1.82)
 
1.3. Subgroup analysis by pre-treatment with clopidogrel1.30 (0.93 to 1.83)
Patients with ACS1.27 (0.87 to 1.86)
Patients without ACS1.41 (0.67 to 2.97)
 
2. As initial medical treatment of patients with NSTEACS1.29 (1.14 to 1.45)
d, day; PCI, percutaneous coronary intervention; NSTEACS, non-ST elevation acute coronary syndrome; STEMI, ST elevation myocardial infarction; CAD, coronary artery disease; OR, odds ratio; CI, confidence interval

2. Glycoprotein IIb/IIIa blockers as initial medical treatment in patients with non-ST segment elevation acute coronary syndromes

Primary endpoints

Data from 12 trials with 31,490 patients (94% of patients included in this meta-analysis) were available on 30-day mortality and myocardial infarction, while data from only six trials but with 19,396 patients (58% of patients included in the overall review) were available on six-month mortality and myocardial infarction.

Mortality occurred in 3.32% of patients in the treatment group versus 3.62% in controls at 30 days, and in 6.26% and 6.23% respectively at six months. Death or myocardial infarction occurred in 10.50% and 11.89% at 30 days, and 13.25% and 14.53% at six months respectively.

Treatment with intravenous glycoprotein IIb/IIIa blockers was not associated with a reduction in the odds of all-cause mortality at 30 days (OR 0.90, 95% CI 0.79 to 1.02; I2 = 9%; Analysis 8.1) or at six months (OR 1.00, 95% CI 0.87 to 1.15; I2 = 0%; Analysis 8.2). However, these agents reduced the risk of all-cause mortality or myocardial infarction both at 30 days (OR 0.91, 95% CI 0.85 to 0.98; I2 = 28%; Analysis 8.3) and at six months (OR 0.88, 95% CI 0.81 to 0.96; I2 = 0%; Analysis 8.4). The ARR per thousand treated patients was 13.9 at 30 days and 12.7 at six months, and the corresponding NNTs were of 75 and 79 respectively.

Safety

Data were available from 12 trials with 31,099 patients (93.7% of patients included in the overall review). Treatment with intravenous glycoprotein IIb/IIIa blockers was associated with an increase in the incidence of major bleeding at 30 days (OR 1.29, 95% CI 1.14 to 1.45, P < 0.0001; I2 = 24%; Analysis 8.5). Major bleeding occurred in 3.76% and in 3.56% of patients in the treatment and control groups, respectively.

Discussion

Glycoprotein IIb/IIIa blockers during percutaneous coronary intervention (PCI)

This systematic review has identified that glycoprotein IIb/IIIa blockers may be safe and effective when administered during PCI with or without stent implantation. This is based on data from 48 trials including over 33,000 patients. Overall, the administration of IIb/IIIa blockers as a bolus immediately before the intervention followed by a 12 to 24-hour infusion is beneficial. Although associated with an increased risk of severe bleeding (8.0 per 1000), this hazard may be considered to be offset by the reduction of 30-day mortality (4.0 patients per 1000 treated), mortality or non-fatal myocardial infarction (23.9 patients per 1000) and urgent revascularisation (14 patients per 1000 treated). The early benefit of glycoprotein IIb/IIIa blockers on mortality seems to be lost at six months, although the lower number of included studies and enrolled patients with six-month follow-up may account for these different results. The short antiplatelet effects of these intravenous drugs may also account for this apparent decrease in efficacy at six months, although the initial benefit on mortality or myocardial infarction is maintained at six months.

Most of the beneficial effects of these drugs were in the prevention of peri-procedural myocardial infarction, a condition whose definition has varied over the years and between studies partially because troponins have only been included in their diagnosis in studies performed after the year 2005. In addition, their diagnosis is especially difficult in patients presenting with a myocardial infarction. Therefore, no study applied the current third universal definition of a post-procedural myocardial infarction in which an elevation more than five times the upper limit of normal or a 20% increase from a previously abnormal level is required (Universal definition MI 2012). Although we tried to limit the variability of the definition by using a uniform criteria of an elevation of more than three times the upper limit of normal of the biomarker of necrosis used when the data were obtainable, the variability of the myocardial infarction definition could explain the heterogeneity observed in the analysis of this endpoint.

The beneficial effect of these drugs is homogeneous in different subgroups of patients according to their clinical condition (i.e. stable coronary artery disease (CAD), non-ST segment elevation acute coronary syndromes (NSTEACS) or ST-segment elevation myocardial infarction (STEMI)) and the technique used (i.e. balloon angioplasty or PCI with stent), although 30-day and six-month mortality was only reduced when administered in procedures with stent implantation. The use of drug-eluting stents has been reported to be associated with a higher risk of thrombosis. Since no study on IIb/IIIa blockers has been performed specifically in patients with drug-eluting stents, and since in only four of the 24 reviewed studies (BRAVE-3 2009; CLEAR PLATELETS-2 2009; ISAR-REACT 2 2006; OPTIMIZE-IT 2009) a drug-eluting stent was used in more than 40% of the cases, the results of this meta-analysis only apply to patients in which a bare metal stent was implanted.

The administration of clopidogrel before PCI in addition to aspirin and heparin, and during the first year following PCI, has been shown to reduce the risk of acute coronary occlusion and of mortality, myocardial infarction or recurrent ischaemia (PCI-CURE 2001), and is currently the standard medical treatment for patients subjected to this procedure (ACCF/AHA Guideline NSTEACS 2013; ESC Guidelines NSTEACS 2011). In recent years, the administration of clopidogrel in addition to aspirin and heparin has also been shown to be of benefit as initial medical treatment of patients with NSTEACS (ACCF/AHA Guideline NSTEACS 2013; CURE 2001; ESC Guidelines NSTEACS 2011; PCI-CURE 2001). Controversy currently exists about the effectiveness of glycoprotein IIb/IIIa blockers in patients submitted for PCI on long-term clopidogrel treatment or in patients pre-treated with a loading dose from the time of hospital admission or at least two to six hours before PCI. The results of this systematic review show that IIb/IIIa blockers are less efficacious in decreasing major and minor events in patients pre-treated with clopidogrel, and suggest that they retain a beneficial effect only in patients with ACS. These results are in agreement with the higher risk of thromboembolic events in patients with ACS compared with patients with stable CAD. On the other hand, the risk of severe bleeding is not enhanced in patients pre-treated with clopidogrel.

Glycoprotein IIb/IIIa blockers as initial medical treatment in patients with non-ST segment elevation acute coronary syndromes

This systematic review also identified that glycoprotein IIb/IIIa antagonists are safe but much less effective when administered as an initial medical treatment to patients with NSTEACS than in patients undergoing PCI. This conclusion is based on data from over 33,000 patients. Overall, the administration of intravenous glycoprotein IIb/IIIa blockers as an initial bolus followed by a continuous infusion for 24 to 72 hours resulted in a modest benefit at 30 days (13.9 deaths or myocardial infarctions prevented per 1000 patients treated) and at six months. This benefit was obtained in spite of a very acceptable excess of severe bleeding (2 per 1000). However, the treatment provided no significant benefit on all-cause mortality at 30 days or six months.

These results contrast with those mentioned above in the overall population submitted for PCI, and also in the subgroup of patients with NSTEACS that underwent PCI. It is worth noting that except for two studies (PRACTICE 2007; PRISM 1998) the beneficial effect obtained was higher in trials with a high use rate of PCI procedures than in trials with a low frequency of these procedures. In addition, in two trials (PRISM Plus 1998; PURSUIT 1998) patients that underwent PCI 24 to 72 hours after admission obtained greater benefit from glycoprotein IIb/IIIa antagonists after PCI than before the procedure, and in one trial (PARAGON B 2002) a benefit was observed only among patients that underwent PCI during drug infusion. These results strongly suggest the existence of a positive interaction between PCI and the effect of glycoprotein IIb/IIIa blockers. In fact, in previous meta-analysis using individual data, a more marked effect of the intervention was observed among patients that underwent PCI during the administration of IIb/IIIa antagonists that among patients that did not undergo PCI (Boersma 1999; Boersma 2002). Finally, because the overall treatment effect of glycoprotein IIb/IIIa inhibitors when administered as initial medical management of patients with NSTEACS is small and these drugs are expensive, the best cost-effectiveness ratio may be obtained when they are administered in high-risk patients scheduled for early PCI. In this sense, one trial has shown similar effects but excess bleeding when administering one of these drugs (eptifibatide) since hospital admission (upstream treatment) rather than during early PCI (downstream treatment) (EARLY-ACS 2009).

Characteristics and limitations of the review

Heterogeneity of studies was statistically important only in 19 of the 96 analyses performed, all of them related to PCI and 11 regarding secondary endpoints. Such heterogeneity is likely to be due to the subjective nature of urgent revascularisation, and the inclusion of small studies. Differences in patients' characteristics such as age, gender, history of myocardial infarction and proportion of patients with acute coronary syndromes, although important, did not result in significant statistical heterogeneity. It is unlikely that other factors such as drug dosages or important concomitant treatments may have affected homogeneity, particularly heparin and aspirin.

We did not perform a cost-effectiveness analysis since this was out of the scope of our review. Some of the analysed studies performed a retrospective analysis on cost-effectiveness, most of them with data from a specific country and applying the results of the overall study. It is of note that this kind of analysis is difficult to perform in multicentre trials in which participating countries have major differences in local practices and public health policies and economies.

It should be noted that the studied population may not be representative of all patients undergoing PCI or with NSTEACS. In the group of patients treated during PCI the mean age of patients was lower than what it is usually observed in clinical practice, as well as the proportion of other co-morbidities (CRUSADE 2006; GRACE 2007). However, the subgroup analysis performed on these patients showed similar results to those obtained in the global analysis, and in those obtained for patients with or without stable CAD. In the group of patients with NSTEACS treated medically, the inclusion was limited to patients with ST-segment changes during the admission ECG or with positive biological markers of myocardial necrosis. These features are present in three-quarters of patients and are known to select high-risk patients. In fact, in most centres patients with these characteristics are submitted for coronary angiography within 48 hours as currently recommended (ACCF/AHA Guideline NSTEACS 2013; ESC Guidelines NSTEACS 2011). In addition, some studies have shown a significant differential treatment effect when glycoprotein IIb/IIIa blockers have been administered in patients with positive or negative troponin levels at admission (Boersma 2002). On the other hand, all of these randomised controlled trials excluded patients with significant renal impairment, cerebrovascular disease and also any patient with a moderate to high risk of bleeding complications. For these reasons, the generalisability of the findings of this review is limited to a moderate to high-risk population with a low risk of bleeding complications.

Authors' conclusions

Implications for practice

Intravenous IIb/IIIa blockers administered during percutaneous coronary intervention (PCI) reduce the risk of death at 30 days but not at six months, and of death or myocardial infarction at 30 days and at six months, at a price of an increase in the risk of severe bleeding. The efficacy effects are homogeneous for all subgroups of patients, although they are less marked in patients pre-treated with clopidogrel, where they seem to be effective only in patients with acute coronary syndromes.

When administered as initial medical treatment in patients with NSTEACS, these agents do not reduce all-cause mortality, but slightly reduce the risk of death or myocardial infarction at 30 days and at six months, and the risk of severe bleeding.

Implications for research

Since the analysis of patients that underwent PCI after pre-treatment with clopidogrel showed less benefit than in the main analysis, and since new oral antiplatelet agents such as prasugrel and ticagrelor have been shown to be more efficacious than clopidogrel but with a higher risk of bleeding (PLATO 2009; TRITON TIMI-38 2007), further trials are warranted in patients pre-treated with these drugs. Also, further research is needed to analyse whether the favourable effects observed in patients in whom a bare metal stent is implanted will also be observed in patients with drug-eluting stents.

In addition, considering the cost of these drugs, prospective cost-effectiveness analyses in patients managed according to current recommendations (ACCF/AHA Guideline NSTEACS 2013; ESC Guidelines NSTEACS 2011) will be desirable. In addition, patient-centred outcomes such as quality of life have not been studied and are particularly warranted.

Acknowledgements

We would like to thank Margaret Burke and Nicole Martin from the Cochrane Heart Group for updating the bibliographic search and providing help with obtaining difficult to find references; Joey Kwong, Isaac Fung and Aleksandra Mas-Stachurska for their help in translating to English two Chinese and one Polish study; and the Editors of the Cochrane Heart Group for their helpful comments through the review and its updates. We also thank Dr. Pablo Loma-Osorio for his help in calculating absolute risk reductions and NNT in the 2007 update of the review.

Data and analyses

Download statistical data

Comparison 1. During PCI (all patients)
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 30-day mortality4331744Odds Ratio (M-H, Fixed, 95% CI)0.79 [0.64, 0.97]
1.1 Blinded studies with a placebo group2827205Odds Ratio (M-H, Fixed, 95% CI)0.78 [0.61, 1.00]
1.2 Non-blinded studies and without placebo154539Odds Ratio (M-H, Fixed, 95% CI)0.81 [0.54, 1.21]
2 6-month mortality3424431Odds Ratio (M-H, Fixed, 95% CI)0.90 [0.77, 1.05]
2.1 Blinded studies with a placebo group2020329Odds Ratio (M-H, Fixed, 95% CI)0.90 [0.76, 1.08]
2.2 Non-blinded studies and without placebo144102Odds Ratio (M-H, Fixed, 95% CI)0.90 [0.65, 1.23]
3 30-day mortality or myocardial infarction4431880Odds Ratio (M-H, Fixed, 95% CI)0.66 [0.60, 0.72]
3.1 Blinded studies with a placebo group2827205Odds Ratio (M-H, Fixed, 95% CI)0.66 [0.60, 0.73]
3.2 Non-blinded studies and without placebo164675Odds Ratio (M-H, Fixed, 95% CI)0.64 [0.48, 0.85]
4 6-month mortality or myocardial infarction3324845Odds Ratio (M-H, Random, 95% CI)0.75 [0.64, 0.86]
4.1 Blinded studies with a placebo group2020329Odds Ratio (M-H, Random, 95% CI)0.72 [0.62, 0.84]
4.2 Non-blinded studies and without placebo144516Odds Ratio (M-H, Random, 95% CI)0.82 [0.57, 1.18]
5 30-day urgent revascularisation4231555Odds Ratio (M-H, Fixed, 95% CI)0.62 [0.54, 0.71]
6 6-month urgent revascularisation3221548Odds Ratio (M-H, Fixed, 95% CI)0.85 [0.79, 0.93]
7 30-day mortality, myocardial infarction or urgent revascularisation4131433Odds Ratio (M-H, Random, 95% CI)0.65 [0.57, 0.75]
8 6-month mortality, myocardial infarction or urgent revascularisation3322432Odds Ratio (M-H, Random, 95% CI)0.80 [0.73, 0.88]
9 30-day major bleeding4031430Odds Ratio (M-H, Fixed, 95% CI)1.39 [1.21, 1.61]
9.1 Blinded studies with a placebo group2727037Odds Ratio (M-H, Fixed, 95% CI)1.38 [1.19, 1.60]
9.2 Non-blinded studies and without placebo134393Odds Ratio (M-H, Fixed, 95% CI)1.53 [0.99, 2.37]
Analysis 1.1.

Comparison 1 During PCI (all patients), Outcome 1 30-day mortality.

Analysis 1.2.

Comparison 1 During PCI (all patients), Outcome 2 6-month mortality.

Analysis 1.3.

Comparison 1 During PCI (all patients), Outcome 3 30-day mortality or myocardial infarction.

Analysis 1.4.

Comparison 1 During PCI (all patients), Outcome 4 6-month mortality or myocardial infarction.

Analysis 1.5.

Comparison 1 During PCI (all patients), Outcome 5 30-day urgent revascularisation.

Analysis 1.6.

Comparison 1 During PCI (all patients), Outcome 6 6-month urgent revascularisation.

Analysis 1.7.

Comparison 1 During PCI (all patients), Outcome 7 30-day mortality, myocardial infarction or urgent revascularisation.

Analysis 1.8.

Comparison 1 During PCI (all patients), Outcome 8 6-month mortality, myocardial infarction or urgent revascularisation.

Analysis 1.9.

Comparison 1 During PCI (all patients), Outcome 9 30-day major bleeding.

Comparison 2. Subgroup of PCI in patients with stable CAD
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 30-day mortality156419Odds Ratio (M-H, Fixed, 95% CI)0.71 [0.34, 1.46]
2 6-month mortality126323Odds Ratio (M-H, Fixed, 95% CI)0.88 [0.63, 1.22]
3 30-day mortality or myocardial infarction156419Odds Ratio (M-H, Fixed, 95% CI)0.68 [0.55, 0.84]
4 6-month mortality or myocardial infarction126326Odds Ratio (M-H, Fixed, 95% CI)0.77 [0.64, 0.92]
5 30-day urgent revascularisation146156Odds Ratio (M-H, Fixed, 95% CI)0.84 [0.54, 1.32]
6 6-month urgent revascularisation126326Odds Ratio (M-H, Fixed, 95% CI)0.93 [0.81, 1.08]
7 30-day mortality, myocardial infarction or urgent revascularisation146156Odds Ratio (M-H, Fixed, 95% CI)0.74 [0.60, 0.91]
8 6-month mortality, myocardial infarction or urgent revascularisation126326Odds Ratio (M-H, Fixed, 95% CI)0.87 [0.77, 0.98]
9 30-day major bleeding166510Odds Ratio (M-H, Fixed, 95% CI)1.83 [1.10, 3.03]
Analysis 2.1.

Comparison 2 Subgroup of PCI in patients with stable CAD, Outcome 1 30-day mortality.

Analysis 2.2.

Comparison 2 Subgroup of PCI in patients with stable CAD, Outcome 2 6-month mortality.

Analysis 2.3.

Comparison 2 Subgroup of PCI in patients with stable CAD, Outcome 3 30-day mortality or myocardial infarction.

Analysis 2.4.

Comparison 2 Subgroup of PCI in patients with stable CAD, Outcome 4 6-month mortality or myocardial infarction.

Analysis 2.5.

Comparison 2 Subgroup of PCI in patients with stable CAD, Outcome 5 30-day urgent revascularisation.

Analysis 2.6.

Comparison 2 Subgroup of PCI in patients with stable CAD, Outcome 6 6-month urgent revascularisation.

Analysis 2.7.

Comparison 2 Subgroup of PCI in patients with stable CAD, Outcome 7 30-day mortality, myocardial infarction or urgent revascularisation.

Analysis 2.8.

Comparison 2 Subgroup of PCI in patients with stable CAD, Outcome 8 6-month mortality, myocardial infarction or urgent revascularisation.

Analysis 2.9.

Comparison 2 Subgroup of PCI in patients with stable CAD, Outcome 9 30-day major bleeding.

Comparison 3. Subgroup of PCI in patients with NSTEACS
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 30-day mortality65778Odds Ratio (M-H, Fixed, 95% CI)0.77 [0.46, 1.28]
2 6-month mortality55716Odds Ratio (M-H, Fixed, 95% CI)1.01 [0.74, 1.38]
3 30-day mortality or myocardial infarction75914Odds Ratio (M-H, Fixed, 95% CI)0.68 [0.56, 0.83]
4 6-month mortality or myocardial infarction55716Odds Ratio (M-H, Fixed, 95% CI)0.77 [0.65, 0.92]
5 30-day urgent revascularisation55728Odds Ratio (M-H, Fixed, 95% CI)0.68 [0.52, 0.89]
6 6-month urgent revascularisation55716Odds Ratio (M-H, Fixed, 95% CI)0.91 [0.79, 1.05]
7 30-day mortality, myocardial infarction or urgent revascularisation55728Odds Ratio (M-H, Fixed, 95% CI)0.69 [0.58, 0.81]
8 6-month mortality, myocardial infarction or urgent revascularisation55716Odds Ratio (M-H, Fixed, 95% CI)0.85 [0.76, 0.96]
9 30-day major bleeding55728Odds Ratio (M-H, Fixed, 95% CI)1.41 [1.03, 1.93]
Analysis 3.1.

Comparison 3 Subgroup of PCI in patients with NSTEACS, Outcome 1 30-day mortality.

Analysis 3.2.

Comparison 3 Subgroup of PCI in patients with NSTEACS, Outcome 2 6-month mortality.

Analysis 3.3.

Comparison 3 Subgroup of PCI in patients with NSTEACS, Outcome 3 30-day mortality or myocardial infarction.

Analysis 3.4.

Comparison 3 Subgroup of PCI in patients with NSTEACS, Outcome 4 6-month mortality or myocardial infarction.

Analysis 3.5.

Comparison 3 Subgroup of PCI in patients with NSTEACS, Outcome 5 30-day urgent revascularisation.

Analysis 3.6.

Comparison 3 Subgroup of PCI in patients with NSTEACS, Outcome 6 6-month urgent revascularisation.

Analysis 3.7.

Comparison 3 Subgroup of PCI in patients with NSTEACS, Outcome 7 30-day mortality, myocardial infarction or urgent revascularisation.

Analysis 3.8.

Comparison 3 Subgroup of PCI in patients with NSTEACS, Outcome 8 6-month mortality, myocardial infarction or urgent revascularisation.

Analysis 3.9.

Comparison 3 Subgroup of PCI in patients with NSTEACS, Outcome 9 30-day major bleeding.

Comparison 4. Subgroup of primary PCI in patients with STEMI
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 30-day mortality96525Odds Ratio (M-H, Fixed, 95% CI)0.88 [0.64, 1.21]
2 6-month mortality94887Odds Ratio (M-H, Fixed, 95% CI)0.88 [0.68, 1.15]
3 30-day mortality or myocardial infarction96525Odds Ratio (M-H, Fixed, 95% CI)0.78 [0.61, 1.00]
4 6-month mortality or myocardial infarction94887Odds Ratio (M-H, Random, 95% CI)0.77 [0.51, 1.16]
5 30-day urgent revascularisation96527Odds Ratio (M-H, Fixed, 95% CI)0.61 [0.45, 0.83]
6 6-month urgent revascularisation94889Odds Ratio (M-H, Fixed, 95% CI)0.75 [0.63, 0.90]
7 30-day mortality, myocardial infarction or urgent revascularisation96527Odds Ratio (M-H, Random, 95% CI)0.69 [0.49, 0.98]
8 6-month mortality, myocardial infarction or urgent revascularisation95677Odds Ratio (M-H, Fixed, 95% CI)0.84 [0.73, 0.97]
9 30-day major bleeding106675Odds Ratio (M-H, Fixed, 95% CI)1.54 [1.12, 2.11]
Analysis 4.1.

Comparison 4 Subgroup of primary PCI in patients with STEMI, Outcome 1 30-day mortality.

Analysis 4.2.

Comparison 4 Subgroup of primary PCI in patients with STEMI, Outcome 2 6-month mortality.

Analysis 4.3.

Comparison 4 Subgroup of primary PCI in patients with STEMI, Outcome 3 30-day mortality or myocardial infarction.

Analysis 4.4.

Comparison 4 Subgroup of primary PCI in patients with STEMI, Outcome 4 6-month mortality or myocardial infarction.

Analysis 4.5.

Comparison 4 Subgroup of primary PCI in patients with STEMI, Outcome 5 30-day urgent revascularisation.

Analysis 4.6.

Comparison 4 Subgroup of primary PCI in patients with STEMI, Outcome 6 6-month urgent revascularisation.

Analysis 4.7.

Comparison 4 Subgroup of primary PCI in patients with STEMI, Outcome 7 30-day mortality, myocardial infarction or urgent revascularisation.

Analysis 4.8.

Comparison 4 Subgroup of primary PCI in patients with STEMI, Outcome 8 6-month mortality, myocardial infarction or urgent revascularisation.

Analysis 4.9.

Comparison 4 Subgroup of primary PCI in patients with STEMI, Outcome 9 30-day major bleeding.

Comparison 5. Subgroup of balloon angioplasty
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 30-day mortality1113378Odds Ratio (M-H, Fixed, 95% CI)0.79 [0.55, 1.14]
2 6-month mortality45291Odds Ratio (M-H, Fixed, 95% CI)1.06 [0.75, 1.50]
3 30-day mortality or myocardial infarction1113378Odds Ratio (M-H, Fixed, 95% CI)0.65 [0.56, 0.75]
4 6-month mortality or myocardial infarction45291Odds Ratio (M-H, Fixed, 95% CI)0.78 [0.65, 0.94]
5 30-day urgent revascularisation1113378Odds Ratio (M-H, Fixed, 95% CI)0.58 [0.49, 0.70]
6 6-month urgent revascularisation45291Odds Ratio (M-H, Random, 95% CI)0.81 [0.60, 1.10]
7 30-day mortality, myocardial infarction or urgent revascularisation1113378Odds Ratio (M-H, Random, 95% CI)0.63 [0.51, 0.76]
8 6-month mortality, myocardial infarction or urgent revascularisation56229Odds Ratio (M-H, Fixed, 95% CI)0.84 [0.75, 0.94]
9 30-day major bleeding1013285Odds Ratio (M-H, Random, 95% CI)1.38 [1.02, 1.86]
Analysis 5.1.

Comparison 5 Subgroup of balloon angioplasty, Outcome 1 30-day mortality.

Analysis 5.2.

Comparison 5 Subgroup of balloon angioplasty, Outcome 2 6-month mortality.

Analysis 5.3.

Comparison 5 Subgroup of balloon angioplasty, Outcome 3 30-day mortality or myocardial infarction.

Analysis 5.4.

Comparison 5 Subgroup of balloon angioplasty, Outcome 4 6-month mortality or myocardial infarction.

Analysis 5.5.

Comparison 5 Subgroup of balloon angioplasty, Outcome 5 30-day urgent revascularisation.

Analysis 5.6.

Comparison 5 Subgroup of balloon angioplasty, Outcome 6 6-month urgent revascularisation.

Analysis 5.7.

Comparison 5 Subgroup of balloon angioplasty, Outcome 7 30-day mortality, myocardial infarction or urgent revascularisation.

Analysis 5.8.

Comparison 5 Subgroup of balloon angioplasty, Outcome 8 6-month mortality, myocardial infarction or urgent revascularisation.

Analysis 5.9.

Comparison 5 Subgroup of balloon angioplasty, Outcome 9 30-day major bleeding.

Comparison 6. Subgroup of stent implantation
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 30-day mortality2815827Odds Ratio (M-H, Fixed, 95% CI)0.78 [0.58, 1.04]
2 6-month mortality2513754Odds Ratio (M-H, Fixed, 95% CI)0.87 [0.71, 1.07]
3 30-day mortality or myocardial infarction2915963Odds Ratio (M-H, Fixed, 95% CI)0.67 [0.59, 0.76]
4 6-month mortality or myocardial infarction2513757Odds Ratio (M-H, Random, 95% CI)0.72 [0.60, 0.87]
5 30-day urgent revascularisation2615516Odds Ratio (M-H, Fixed, 95% CI)0.75 [0.58, 0.98]
6 6-month urgent revascularisation2413663Odds Ratio (M-H, Fixed, 95% CI)0.87 [0.78, 0.97]
7 30-day mortality, myocardial infarction or urgent revascularisation2615516Odds Ratio (M-H, Random, 95% CI)0.69 [0.57, 0.84]
8 6-month mortality, myocardial infarction or urgent revascularisation2513758Odds Ratio (M-H, Random, 95% CI)0.78 [0.69, 0.89]
9 30-day major bleeding2715643Odds Ratio (M-H, Fixed, 95% CI)1.38 [1.05, 1.82]
Analysis 6.1.

Comparison 6 Subgroup of stent implantation, Outcome 1 30-day mortality.

Analysis 6.2.

Comparison 6 Subgroup of stent implantation, Outcome 2 6-month mortality.

Analysis 6.3.

Comparison 6 Subgroup of stent implantation, Outcome 3 30-day mortality or myocardial infarction.

Analysis 6.4.

Comparison 6 Subgroup of stent implantation, Outcome 4 6-month mortality or myocardial infarction.

Analysis 6.5.

Comparison 6 Subgroup of stent implantation, Outcome 5 30-day urgent revascularisation.

Analysis 6.6.

Comparison 6 Subgroup of stent implantation, Outcome 6 6-month urgent revascularisation.

Analysis 6.7.

Comparison 6 Subgroup of stent implantation, Outcome 7 30-day mortality, myocardial infarction or urgent revascularisation.

Analysis 6.8.

Comparison 6 Subgroup of stent implantation, Outcome 8 6-month mortality, myocardial infarction or urgent revascularisation.

Analysis 6.9.

Comparison 6 Subgroup of stent implantation, Outcome 9 30-day major bleeding.

Comparison 7. Subgroup of PCI in patients pre-treated with clopidogrel
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 30-day mortality138048Odds Ratio (M-H, Fixed, 95% CI)0.83 [0.57, 1.20]
1.1 Patients with ACS74634Odds Ratio (M-H, Fixed, 95% CI)0.84 [0.56, 1.25]
1.2 Patients without ACS63414Odds Ratio (M-H, Fixed, 95% CI)0.78 [0.29, 2.09]
2 6-month mortality147526Odds Ratio (M-H, Fixed, 95% CI)0.95 [0.75, 1.21]
2.1 Patients with ACS83934Odds Ratio (M-H, Fixed, 95% CI)0.97 [0.72, 1.31]
2.2 Patients without ACS63592Odds Ratio (M-H, Fixed, 95% CI)0.92 [0.63, 1.35]
3 30-day mortality or myocardial infarction138048Odds Ratio (M-H, Fixed, 95% CI)0.81 [0.67, 0.98]
3.1 Patients with ACS74634Odds Ratio (M-H, Fixed, 95% CI)0.77 [0.61, 0.96]
3.2 Patients without ACS63414Odds Ratio (M-H, Fixed, 95% CI)0.91 [0.65, 1.28]
4 6-month mortality or myocardial infarction147529Odds Ratio (M-H, Random, 95% CI)0.87 [0.66, 1.14]
4.1 Patients with ACS83934Odds Ratio (M-H, Random, 95% CI)0.77 [0.46, 1.27]
4.2 Patients without ACS63595Odds Ratio (M-H, Random, 95% CI)0.94 [0.73, 1.21]
5 30-day urgent revascularisation117737Odds Ratio (M-H, Fixed, 95% CI)0.89 [0.62, 1.28]
5.1 Patients with ACS64586Odds Ratio (M-H, Fixed, 95% CI)0.82 [0.54, 1.24]
5.2 Patients without ACS53151Odds Ratio (M-H, Fixed, 95% CI)1.17 [0.54, 2.53]
6 6-month urgent revascularisation137435Odds Ratio (M-H, Fixed, 95% CI)0.87 [0.76, 0.98]
6.1 Patients with ACS83936Odds Ratio (M-H, Fixed, 95% CI)0.77 [0.64, 0.94]
6.2 Patients without ACS53499Odds Ratio (M-H, Fixed, 95% CI)0.94 [0.80, 1.11]
7 30-day mortality, myocardial infarction or urgent revascularisation117737Odds Ratio (M-H, Random, 95% CI)0.88 [0.65, 1.20]
7.1 Patients with ACS64586Odds Ratio (M-H, Random, 95% CI)0.80 [0.50, 1.26]
7.2 Patients without ACS53151Odds Ratio (M-H, Random, 95% CI)1.08 [0.76, 1.52]
8 6-month mortality, myocardial infarction or urgent revascularisation137381Odds Ratio (M-H, Fixed, 95% CI)0.88 [0.79, 0.98]
8.1 Patients with ACS73786Odds Ratio (M-H, Fixed, 95% CI)0.81 [0.69, 0.95]
8.2 Patients without ACS63595Odds Ratio (M-H, Fixed, 95% CI)0.94 [0.81, 1.10]
9 30-day major bleeding148239Odds Ratio (M-H, Fixed, 95% CI)1.30 [0.93, 1.83]
9.1 Patients with ACS74734Odds Ratio (M-H, Fixed, 95% CI)1.27 [0.87, 1.86]
9.2 Patients without ACS73505Odds Ratio (M-H, Fixed, 95% CI)1.41 [0.67, 2.97]
Analysis 7.1.

Comparison 7 Subgroup of PCI in patients pre-treated with clopidogrel, Outcome 1 30-day mortality.

Analysis 7.2.

Comparison 7 Subgroup of PCI in patients pre-treated with clopidogrel, Outcome 2 6-month mortality.

Analysis 7.3.

Comparison 7 Subgroup of PCI in patients pre-treated with clopidogrel, Outcome 3 30-day mortality or myocardial infarction.

Analysis 7.4.

Comparison 7 Subgroup of PCI in patients pre-treated with clopidogrel, Outcome 4 6-month mortality or myocardial infarction.

Analysis 7.5.

Comparison 7 Subgroup of PCI in patients pre-treated with clopidogrel, Outcome 5 30-day urgent revascularisation.

Analysis 7.6.

Comparison 7 Subgroup of PCI in patients pre-treated with clopidogrel, Outcome 6 6-month urgent revascularisation.

Analysis 7.7.

Comparison 7 Subgroup of PCI in patients pre-treated with clopidogrel, Outcome 7 30-day mortality, myocardial infarction or urgent revascularisation.

Analysis 7.8.

Comparison 7 Subgroup of PCI in patients pre-treated with clopidogrel, Outcome 8 6-month mortality, myocardial infarction or urgent revascularisation.

Analysis 7.9.

Comparison 7 Subgroup of PCI in patients pre-treated with clopidogrel, Outcome 9 30-day major bleeding.

Comparison 8. As initial medical treatment in patients with NSTEACS
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 30-day mortality1231490Odds Ratio (M-H, Fixed, 95% CI)0.90 [0.79, 1.02]
2 6-month mortality514171Odds Ratio (M-H, Fixed, 95% CI)1.00 [0.87, 1.15]
3 30-day mortality or myocardial infarction1231490Odds Ratio (M-H, Fixed, 95% CI)0.91 [0.85, 0.98]
4 6-month mortality or myocardial infarction619396Odds Ratio (M-H, Fixed, 95% CI)0.88 [0.81, 0.96]
5 30-day major bleeding1231099Odds Ratio (M-H, Fixed, 95% CI)1.29 [1.14, 1.45]
Analysis 8.1.

Comparison 8 As initial medical treatment in patients with NSTEACS, Outcome 1 30-day mortality.

Analysis 8.2.

Comparison 8 As initial medical treatment in patients with NSTEACS, Outcome 2 6-month mortality.

Analysis 8.3.

Comparison 8 As initial medical treatment in patients with NSTEACS, Outcome 3 30-day mortality or myocardial infarction.

Analysis 8.4.

Comparison 8 As initial medical treatment in patients with NSTEACS, Outcome 4 6-month mortality or myocardial infarction.

Analysis 8.5.

Comparison 8 As initial medical treatment in patients with NSTEACS, Outcome 5 30-day major bleeding.

Appendices

Appendix 1. Search strategies for 2013 update

CENTRAL

#1 MeSH descriptor: [Platelet Glycoprotein GPIIb-IIIa Complex] explode all trees
#2 glycoprotein near/6 inhibit*
#3 glycoprotein near/6 block*
#4 glycoprotein near/6 antagonist*
#5 gpiib*
#6 abciximab
#7 sibrafiban
#8 tirofiban
#9 lamifiban
#10 aggrastat
#11 eptifibatide
#12 xemilofiban
#13 lotrafiban
#14 orbofiban
#15 fradafiban
#16 fibrinogen next receptor next antagonist*
#17 roxifiban
#18 (#1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12 or #13 or #14 or #15 or #16 or #17)
#19 MeSH descriptor: [Angioplasty, Balloon, Coronary] this term only
#20 ptca
#21 coronary near/6 angioplasty
#22 pci
#23 percutaneous next coronary next intervention*
#24 MeSH descriptor: [Angina, Unstable] explode all trees
#25 angina
#26 stent*
#27 MeSH descriptor: [Myocardial Infarction] explode all trees
#28 myocardial next infarction
#29 heart next infarction
#30 coronary next syndrome*
#31 non next st next segment
#32 non next st next elevation
#33 without next st next segment
#34 MeSH descriptor: [Acute Coronary Syndrome] this term only
#35 #19 or #20 or #21 or #22 or #23 or #24 or #25 or #26 or #27 or #28 or #29 or #30 or #31 or #32 or #33 or #34
#36 #18 and #35

MEDLINE OVID

Sensitivity maximising Cochrane RCT filter (Lefebvre C, Manheimer E, Glanville J. Chapter 6: Searching for studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. www.cochrane-handbook.org)

1. Platelet Glycoprotein GPIIb-IIIa Complex/
2. (glycoprotein adj5 (inhibit$ or block$ or antagonist$)).tw.
3. gpIIb$.tw.
4. abciximab.tw.
5. sibrafiban.tw.
6. tirofiban.tw.
7. lamifiban.tw.
8. aggrastat.tw.
9. eptifibatide.tw.
10. xemilofiban.tw.
11. lotrafiban.tw.
12. orbofiban.tw.
13. roxifiban.tw.
14. or/1-13
15. Angioplasty, Balloon, Coronary/
16. ptca.tw.
17. (coronary adj5 angioplasty).tw.
18. exp Angina, Unstable/
19. angina.tw.
20. exp Stents/
21. stent$.tw.
22. exp Myocardial Infarction/
23. myocardial infarction.tw.
24. coronary syndrome$.tw.
25. Acute Coronary Syndrome/
26. pci.tw.
27. percutaneous coronary intervention$.tw.
28. or/15-27
29. 14 and 28
30. randomized controlled trial.pt.
31. controlled clinical trial.pt.
32. randomized.ab.
33. placebo.ab.
34. drug therapy.fs.
35. randomly.ab.
36. trial.ab.
37. groups.ab.
38. 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37
39. exp animals/ not humans.sh.
40. 38 not 39
41. 29 and 40
42. 29 and 41
43. (200910* or 200911* or 200912* or 2010* or 2011* or 2012* or 2013*).ed.
44. 42 and 43 

EMBASE OVID

Cochrane RCT filter (Lefebvre C, Manheimer E, Glanville J. Chapter 6: Searching for studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from www.cochrane-handbook.org)

1. fibrinogen receptor/
2. (glycoprotein adj5 (inhibit$ or block$ or antagonist$)).tw.
3. gpIIb$.tw.
4. abciximab.tw.
5. sibrafiban.tw.
6. tirofiban.tw.
7. lamifiban.tw.
8. aggrastat.tw.
9. eptifibatide.tw.
10. xemilofiban.tw.
11. lotrafiban.tw.
12. orbofiban.tw.
13. roxifiban.tw.
14. or/1-13
15. exp transluminal coronary angioplasty/
16. ptca.tw.
17. (coronary adj5 angioplasty).tw.
18. exp unstable angina pectoris/
19. angina.tw.
20. coronary stent/
21. stent$.tw.
22. exp heart infarction/
23. myocardial infarction.tw.
24. coronary syndrome$.tw.
25. pci.tw.
26. percutaneous coronary intervention$.tw.
27. acute coronary syndrome/
28. or/15-27
29. 28 and 14
30. random$.tw.
31. factorial$.tw.
32. crossover$.tw.
33. cross over$.tw.
34. cross-over$.tw.
35. placebo$.tw.
36. (doubl$ adj blind$).tw.
37. (singl$ adj blind$).tw.
38. assign$.tw.
39. allocat$.tw.
40. volunteer$.tw.
41. crossover procedure/
42. double blind procedure/
43. randomized controlled trial/
44. single blind procedure/
45. 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 38 or 39 or 40 or 41 or 42 or 43 or 44
46. exp animals/ or exp invertebrate/ or animal experiment/ or animal model/ or animal tissue/ or animal cell/ or nonhuman/
47. human/ or normal human/ or human cell/
48. 46 and 47
49. 46 not 48
50. 45 not 49
51. 29 and 50
52. (20094* or 20095* or 2010* or 2011* or 2012* or 2013*).em.
53. 51 and 52

Appendix 2. Search strategies for 2010 update

CENTRAL

#1 MeSH descriptor Platelet Glycoprotein GPIIb-IIIa Complex explode all trees
#2 (glycoprotein in All Text near/6 inhibit* in All Text)
#3 (glycoprotein in All Text near/6 block* in All Text)
#4 (glycoprotein in All Text near/6 antagonist* in All Text)
#5 gpiib* in All Text
#6 abciximab in All Text
#7 sibrafiban in All Text
#8 tirofiban in All Text
#9 lamifiban in All Text
#10 aggrastat in All Text
#11 eptifibatide in All Text
#12 xemilofiban in All Text
#13 lotrafiban in All Text
#14 orbofiban in All Text
#15 fradafiban in All Text 1
#16 fibrinogen next receptor next antagonist* in All Text
#17 roxifiban in All Text
#18 (#1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 or #10)
#19 (#11 or #12 or #13 or #14 or #15 or #16 or #17)
#20 (#18 or #19)
#21 MeSH descriptor Angioplasty, Transluminal, Percutaneous Coronary explode all trees
#22 ptca in All Text
#23 (coronary in All Text near/6 angioplasty in All Text)
#24 pci in All Text
#25 percutaneous next coronary next intervention* in All Text
#26 MeSH descriptor Angina, Unstable explode all trees
#27 angina in All Text
#28 stent* in All Text
#29 MeSH descriptor Myocardial Infarction explode all trees
#30 myocardial next infarction in All Text
#31 heart next infarction in All Text
#32 coronary next syndrome* in All Text
#33 non next st next segment in All Text
#34 non next st next elevation in All Text
#35 without next st next segment in All Text
#36 MeSH descriptor acute coronary syndrome this term only
#37 (#21 or #22 or #23 or #24 or #25 or #26 or #27 or #28 or #29 or #30)
#38 (#31 or #32 or #33 or #34 or #35 or #36)
#39 (#37 or #38)
#40 (#20 and #39)

MEDLINE (on Ovid)

1 Platelet Glycoprotein GPIIb-IIIa Complex/
2 (glycoprotein adj5 (inhibit$ or block$ or antagonist$)).tw.
3 gpIIb$.tw.
4 abciximab.tw.
5 sibrafiban.tw.
6 tirofiban.tw.
7 lamifiban.tw.
8 aggrastat.tw.
9 eptifibatide.tw.
10 xemilofiban.tw.
11 lotrafiban.tw.
12 orbofiban.tw.
13 roxifiban.tw.
14 or/1-13
15 exp Angioplasty, Transluminal, Percutaneous Coronary/
16 ptca.tw.
17 (coronary adj5 angioplasty).tw.
18 exp Angina, Unstable/
19 angina.tw.
20 exp Stents/
21 stent$.tw.
22 exp Myocardial Infarction/
23 myocardial infarction.tw.
24 coronary syndrome$.tw.
25 Acute Coronary Syndrome/
26 pci.tw.
27 percutaneous coronary intervention$.tw.
28 or/15-27
29 14 and 28
30 randomized controlled trial.pt.
31 controlled clinical trial.pt.
32 Randomized controlled trials/
33 random allocation/
34 double blind method/
35 single-blind method/
36 or/30-35
37 exp animal/ not humans/
38 36 not 37
39 clinical trial.pt.
40 exp Clinical Trials as Topic/
41 (clin$ adj25 trial$).ti,ab.
42 ((singl$ or doubl$ or trebl$ or tripl$) adj (blind$ or mask$)).ti,ab.
43 placebos/
44 placebo$.ti,ab.
45 random$.ti,ab.
46 research design/
47 or/39-46
48 47 not 37
49 38 or 48
50 29 and 49
51 (2006$ or 2007$ or 2008$ or 2009$).em.
52 51 and 50

EMBASE (on Ovid)

1 exp Fibrinogen Receptor/
2 (glycoprotein adj5 (inhibit$ or block$ or antagonist$)).tw.
3 gpIIb$.tw.
4 abciximab.tw.
5 sibrafiban.tw.
6 tirofiban.tw.
7 lamifiban.tw.
8 aggrastat.tw.
9 eptifibatide.tw.
10 xemilofiban.tw.
11 lotrafiban.tw.
12 orbofiban.tw.
13 roxifiban.tw.
14 or/1-13
15 exp Transluminal Coronary Angioplasty/
16 ptca.tw.
17 (coronary adj5 angioplasty).tw.
18 exp Unstable Angina Pectoris/
19 angina.tw.
20 exp coronary stent/
21 stent$.tw.
22 exp Heart Infarction/
23 myocardial infarction.tw.
24 coronary syndrome$.tw.
25 pci.tw.
26 percutaneous coronary intervention$.tw.
27 acute coronary syndrome/
28 or/15-27
29 28 and 14
30 controlled clinical trial/
31 random$.tw.
32 randomized controlled trial/
33 follow-up.tw.
34 double blind procedure/
35 placebo$.tw.
36 placebo/
37 factorial$.ti,ab.
38 (crossover$ or cross-over$).ti,ab.
39 (double$ adj blind$).ti,ab.
40 (singl$ adj blind$).ti,ab.
41 assign$.ti,ab.
42 allocat$.ti,ab.
43 volunteer$.ti,ab.
44 Crossover Procedure/
45 Single Blind Procedure/
46 or/30-45
47 (exp animals/ or nonhuman/) not human/
48 46 not 47
49 29 and 48
50 (2006$ or 2007$ or 2008$ or 2009$).em.
51 50 and 49

Appendix 3. Search strategies for 2007 update

Database: Ovid MEDLINE(R) <1966 to June 2006>

1 Platelet Glycoprotein GPIIb-IIIa Complex/
2 (glycoprotein adj5 (inhibit$ or block$ or antagonist$)).tw.
3 gpIIb$.tw.
4 abciximab.tw.
5 sibrafiban.tw.
6 tirofiban.tw.
7 lamifiban.tw.
8 aggrastat.tw.
9 eptifibatide.tw.
10 xemilofiban.tw.
11 lotrafiban.tw.
12 orbofiban.tw.
13 roxifiban.tw.
14 or/1-13
15 exp Angioplasty, Transluminal, Percutaneous Coronary/
16 ptca.tw.
17 (coronary adj5 angioplasty).tw.
18 exp Angina, Unstable/
19 angina.tw.
20 exp Stents/
21 stent$.tw.
22 exp Myocardial Infarction/
23 myocardial infarction.tw.
24 coronary syndrome$.tw.
25 pci.tw.
26 percutaneous coronary intervention$.tw.
27 or/15-26
28 14 and 27
29 randomized controlled trial.pt.
30 controlled clinical trial.pt.
31 Randomized controlled trials/
32 random allocation.sh.
33 double blind method.sh.
34 single-blind method.sh.
35 or/29-34
36 exp animal/ not human/
37 35 not 36
38 clinical trial.pt.
39 exp Clinical trials/
40 (clin$ adj25 trial$).ti,ab.
41 ((singl$ or doubl$ or trebl$ or tripl$) adj (blind$ or mask$)).ti,ab.
42 placebos.sh.
43 placebo$.ti,ab.
44 random$.ti,ab.
45 research design.sh.
46 or/38-45
47 46 not 36
48 37 or 47
49 28 and 48
50 limit 49 to yr=2001 - 2005

Database: EMBASE <1980 to 2006 Week 15>

1 exp Fibrinogen Receptor/
2 (glycoprotein adj5 (inhibit$ or block$ or antagonist$)).tw.
3 gpIIb$.tw.
4 abciximab.tw.
5 sibrafiban.tw.
6 tirofiban.tw.
7 lamifiban.tw.
8 aggrastat.tw.
9 eptifibatide.tw.
10 xemilofiban.tw.
11 lotrafiban.tw.
12 orbofiban.tw.
13 roxifiban.tw.
14 or/1-13
15 exp Transluminal Coronary Angioplasty/
16 ptca.tw.
17 (coronary adj5 angioplasty).tw.
18 exp Unstable Angina Pectoris/
19 angina.tw.
20 exp coronary stent/
21 stent$.tw.
22 exp Heart Infarction/
23 myocardial infarction.tw.
24 coronary syndrome$.tw.
25 pci.tw.
26 percutaneous coronary intervention$.tw.
27 or/15-26
28 14 and 27
29 random$.ti,ab.
30 factorial$.ti,ab.
31 (crossover$ or cross over$ or cross-over$).ti,ab.
32 placebo$.ti,ab.
33 (double$ adj blind$).ti,ab.
34 (singl$ adj blind$).ti,ab.
35 assign$.ti,ab.
36 allocat$.ti,ab.
37 volunteer$.ti,ab.
38 Crossover Procedure/
39 Double Blind Procedure/
40 Randomized Controlled Trial/
41 Single Blind Procedure/
42 or/29-41
43 exp animal/
44 nonhuman/
45 exp animal experiment/
46 or/43-45
47 exp human/
48 46 not 47
49 42 not 48
50 49 and 28
51 limit 50 to yr=2000 - 2005
52 from 51 keep 1-569

Appendix 4. MEDLINE search strategy for original review (2001)

1 Clinical trial/ or Phase 1 clinical trial/ or Phase 2
clinical trial/ or Phase 3 clinical trial/ or Phase 4
clinical trial/ or Randomized controlled trial/
2 Randomization/
3 Double blind procedure/ or Meta analysis/ or Single blind
procedure/
4 exp controlled study/
5 Placebo/
6 ["150".tg.]
7 ["197".tg.]
8 (clinic$ adj10 trial).ti,ab.
9 (clinic$ adj10 trial$).ti,ab.
10 (controlled adj trial$).ti,ab.
11 ((singl$ or doubl$ or trebl$ or tripl$) adj10 (blind$ or
mask$)).ti,ab.
12 (placebo$ or random$).ti,ab.
13 1 or 2 or 3 or 4 or 5 or 6 or 7 or 9 or 10 or 11 or 12
14 limit 13 to human
15 ("glycoprotein IIb/IIIa" or "glycoprotein IIb-IIIa" or "glycoprotein-IIb/IIIa" or "Platelet IIb/IIIa" or "GP IIb-IIIa" or "GP IIb/IIIa" or "IIb" or "IIIa").mp. [mp=Title, Abstract, registry number word, mesh subject heading]
16 (Abciximab or 7E3 or Sibrafiban or Tirofiban or MK-383 or lamifiban or Aggrastat or Eptifibatide or Xemilofiban or Sibrafiban or Orbofiban or Lefradafiban or Integrilin or Integrelin or Fradafiban or Lefradafiban).mp. [mp=Title, Abstract, registry number word, mesh subject heading]
17 (Inhibitor$ or block$ or antagonist$).mp. [mp=Title, Abstract, registry number word, mesh subject heading]
18 14 and (15 or 16) and 17

What's new

DateEventDescription
23 October 2013AmendedSummary of findings table was replaced with additional tables to summarise risk of bias. Minor correction was made to the 'Implication for practice' section.
9 May 2013New citation required and conclusions have changed

Twelve new studies were added, 10 in the PCI group and two in the group of patients with NSTEACS that were treated with IIb/IIIa blockers as initial medical treatment.

Conclusions about IIb/IIIa blockers during PCI were slightly changed: treatment with IIb/IIIa blockers was not associated with a significant reduction in mortality at six months.

Conclusions about IIb/IIIa blockers as initial treatment of NSTEACS were unchanged.

History

Protocol first published: Issue 2, 2000
Review first published: Issue 4, 2001

DateEventDescription
21 May 2013New search has been performedSearches re-run in January 2013.
12 July 2010New citation required and conclusions have changed

Change in authors.

Ten new studies in the PCI group were added, five performed during primary PCI in patients with STEMI, four in patients with stable CAD and one in a mixed population. Four of these 10 studies were performed in patients pre-treated with clopidogrel.

No new studies were added to the group of patients with non-STEACS in which IIb/IIIa blockers were given as initial medical treatment. However, two studies that were previously analysed in the PCI group based on preliminary data from congress reports have been finally included in this group.

Conclusions about IIb/IIIa blockers during PCI were slightly changed to underline that the results were homogeneous for all the subgroups that have been analysed except for patients pretreated with clopidogrel where these drugs seem to be effective only in patients with an ACS.

Conclusions about IIb/IIIa blockers as initial treatment of NSTEACS were unchanged.

9 September 2008AmendedConverted to new review format.
22 December 2006New citation required and conclusions have changedSubstantive amendment.

Contributions of authors

Xavier Bosch originated and was primarily responsible for planning, designing and carrying out the primary version of the review and the 2007, 2010 and 2013 updates. He was the principal author and prepared the results and the clinical discussion of the findings.
Juan Sanchis participated in reviewing the studies for the 2010 and 2013 updates and in the discussion of the results.
Jaume Marrugat participated in the design, the methodology and the discussion of the primary version of the review and in the discussion of the 2007 update.
In each version, two of the three authors participated in the study selection, review of pre-selected studies, data extraction and in the preparation of the manuscript.

Declarations of interest

Xavier Bosch participated as investigator in the PRISM Plus 1998, GUSTO-IV 2001 and EARLY-ACS 2009 studies. He has no other potential conflicts of interest to declare.

Jaume Marrugat has no potential conflict of interest to declare.

Juan Sanchis participated as investigator in the EARLY-ACS 2009 study.

Sources of support

Internal sources

  • University of Barcelona, Spain.

  • Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Spain.

  • Institut Municipal d'Investigació Medica, Barcelona, Spain.

External sources

  • FIS PI050120, Ministerio de Sanidad y Consumo, Instituto de Salud Carlos III, Spain.

  • FIS PI07/0640. Ministerio de Sanidad y Consumo, Instituto de Salud Carlos III, Spain.

  • Red HERACLES RD06/0009, Ministerio de Sanidad y Consumo, Instituto de Salud Carlos III, Spain.

  • EUPHORIC Project (Ref. 2003/134), Spain.

Differences between protocol and review

None

Notes

None

Characteristics of studies

Characteristics of included studies [ordered by study ID]

3T/2R 2009

Methods

Method of treatment allocation: an independent study nurse at each site performed assignments of study treatments via a procedure using sealed envelopes, in preselected blocks of 6

Double-blinded?: yes

Stratification: yes, according to the presence of stable or unstable coronary artery disease and poor responsiveness to aspirin, clopidogrel, or both

Placebo: yes

Sample size calculation: no

Intention-to-treat analysis: yes

Funding: partially supported by a research grant from Merck, USA and Iroko, USA

Participants

Location: 10 centres in Italy, Belgium, France and Spain

Timeframe: from February 2006 to June 2008

Follow-up: 30 days

Eligibility criteria: 263 patients > 18 ys scheduled for coronary angiography, PCI or both who presented with stable or troponin-negative NSTEACS, and showed poor ex-vivo response to aspirin or clopidogrel

Exclusion criteria: any evidence of myocardial damage as witnessed by a rise of cardiac specific injury markers and ongoing MI, defined as the presence of ST-segment elevation at ECG or new or presumably new left bundle-branch block

Mean age: 68 y, 74% male, 26% diabetes, 43% prior MI

ACS: 32% (unstable angina: 32%, non-STEMI: 0%, STEMI: 0%)

PCI: 100% (balloon angioplasty: 7%, stent: 93%, drug-eluting stents: ?%), pre-treatment with clopidogrel: 100%

Interventions

Tirofiban (bolus of 25 µg/kg plus a 14 to 24-h infusion of 0.15 µg/kg/min) for 14 to 24 hours versus placebo (bolus and infusion)

93 patients were low-responders to aspirin, 147 to clopidogrel and 23 to both. Screening for clopidogrel response was undertaken in patients at steady state for aspirin provided at least 1 of the following 2 requirements was fulfilled: the patient received a 600 or 300 mg loading dose ≥ 2 or 6 hours before, respectively, or the patient received a 75 mg maintenance clopidogrel dose for ≥ 7 consecutive days

Outcomes

Primary: rate of periprocedural MI

Secondary: 30-day occurrence of minor myocardial injury, the composite of death, MI, or urgent target vessel revascularisation, and the incidence of stent thrombosis

NotesScreening for clopidogrel response was undertaken in patients at steady state for aspirin provided at least 1 of the following 2 requirements was fulfilled: the patient received a 600 or 300 mg loading dose >2 or 6 h before, respectively, or the patient received a 75 mg maintenance clopidogrel dose for ≥ 7 consecutive days
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRefers to a random number generator
Allocation concealment (selection bias)Low riskAn independent study nurse at each site performed assignments of study treatments via a procedure using sealed envelopes, in preselected blocks of 6.
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blinded study with a placebo group. Mortality and myocardial infarction adjudicated by an independent clinical events committee
Blinding (performance bias and detection bias)
Secondary
Low riskMajor bleeding assessed by TIMI criteria and adjudicated by an independent clinical events committee
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all the prespecified outcomes
Other biasUnclear riskThe study was partially supported by a research grant from Merck, USA and Iroko, USA

ACE 2003

Methods

Method of treatment allocation: by means of a computer-generated sequence, and assignments were made using a centralised telephone system

Double-blinded?: no (study not blinded, without a placebo group)

Stratification: no

Placebo: no

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: SORIN Biomedica and the ARCARD ONLUS Foundation

Follow-up: 30 days and 6 months

Participants

Location: 4 centres in Italy, Argentina and Germany

Timeframe: from January 2001 to August 2002

Eligibility criteria: 400 patients with ST-segment elevation acute myocardial infarction of less than 6 hours or between 6 and 24 hours if there was evidence of continuing ischaemia, including patients with cardiogenic shock

Exclusion criteria: previous administration of fibrinolytic or abciximab therapy, a history of bleeding diathesis or allergy to the study drug, major surgery within 15 days, active bleeding, participation in another study, and inability to obtain informed consent. Angiographic criteria: IRA reference diameter < 2.5 mm, a previously stented IRA; 3) < 70% stenosis of the IRA associated with Thrombolysis In Myocardial Infarction (TIMI) trial flow grade 3 (9); and 4) an inability to identify the IRA

Mean age: 66 y, male: 78%, diabetes: 18%, prior myocardial infarction: 11%
Acute coronary syndrome: 100% (unstable angina: 0%, non-STEMI: 0%, STEMI: 100%)
PCI: atherectomy: ?%, balloon angioplasty: 0%, stent: 99%, pre-treatment with clopidogrel: 0%

InterventionsUnblinded abciximab versus placebo. Patients randomised to abciximab received the drug immediately before the procedure as a bolus of 0.25 mg/kg body weight, followed by a 12-h infusion at a rate of 0.125 µg/kg/min
Outcomes

Primary: a composite of death from any cause, reinfarction, target vessel revascularisation, and stroke within 1 month of the index procedure

Secondary: ST-segment reduction, postprocedural corrected TIMI frame count, infarct size at 1 month, death from any cause at 6 months, reinfarction at 6 months, 6-month composite of death and reinfarction, TVR at 6 months, and angiographic restenosis of the IRA at 6 months

NotesAll patients treated with 325 mg of aspirin and heparin. Immediately after the procedure patients received 500 mg of ticlopidine or 300 mg of clopidogrel. Aspirin and clopidogrel 75 mg or ticlopidine 500 mg were maintained for 1 month
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskBy means of a computer-generated sequence
Allocation concealment (selection bias)Low riskUsing a centralised telephone system
Blinding (performance bias and detection bias)
Primary
High riskStudy not blinded and without a placebo group. Events were not adjudicated by an independent clinical events committee
Blinding (performance bias and detection bias)
Secondary
High riskStudy not blinded and without a placebo group. Events were not adjudicated by an independent clinical events committee
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published report includes all pre-specified outcomes
Other biasUnclear riskThe study was supported by SORIN Biomedica and the ARCARD ONLUS Foundation

ADMIRAL 2001

Methods

Method of treatment allocation: not stated (all the patients were randomly assigned, in the order in which they were enrolled, to receive either abciximab or placebo)

Double-blinded?: yes

Stratification: none stated

Placebo. yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: supported by Eli Lilly, Saint-Cloud, France and Indianapolis, and by Saint-Côme-Chirurgie, Marseilles, France

Follow-up: 30 days and 6 months

Participants

Location: 26 centres in France

Timeframe: from 12 July 1997 to 22 December 1998

Eligibility criteria: 300 patients with ST-segment elevation acute myocardial infarction of less than 12 h with coronary anatomy suitable for stent implantation

Exclusion criteria: bleeding diathesis, administration of thrombolytic agents for the current episode, neoplasm, recent stroke, uncontrolled hypertension, recent surgery, oral anticoagulant therapy, a limited life expectancy, childbearing potential and known contraindications to therapy with aspirin, ticlopidine or heparin

Mean age: 61 y, male: 82%, prior myocardial infarction: 11%
Acute coronary syndrome: 100% (unstable angina: 0%, NSTEMI: 0%, STEMI: 100%)
PCI: atherectomy: ?%, balloon angioplasty: 13%, stent: 87%, pre-treatment with clopidogrel: 0%

InterventionsAbciximab as a bolus of 0.25 mg per kilogram of body weight, followed by a 12-h infusion of 0.125 µg per kilogram per minute (maximum, 10 µg per minute) versus placebo
OutcomesPrimary: a composite of death, myocardial infarction or urgent revascularisation at 30 days
Secondary: a composite of death, myocardial infarction or any revascularisation at 30 days and at 6 months; death or myocardial infarction, death, myocardial infarction or urgent revascularisation at 6 months; TIMI flow grade; ejection fraction
Notes

All patients treated with aspirin, heparin and ticlopidine

The data were retained by Eli Lilly, where the analyses were performed. Independent statistical advice was provided by E. Vicant (Paris VII University)

One of the authors, Dr. Pinton, was an employee of and stockholder in Eli Lilly. Dr. Montalescot has served as a consultant to Eli Lilly

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk"All the patients were randomly assigned, in the order in which they were enrolled, to receive either abciximab or placebo"
Allocation concealment (selection bias)Unclear riskNot stated
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blinded study with a placebo group. "The patients, investigators, and sponsors of the study were blinded to the treatment assignments during the entire study".
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blinded study with a placebo group. "The patients, investigators, and sponsors of the study were blinded to the treatment assignments during the entire study".
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published report includes all pre-specified outcomes
Other biasHigh risk

The data were retained by Eli Lilly, where the analyses were performed. Independent statistical advice was provided by E. Vicant (Paris VII University)

One of the authors, Dr. Pinton, was an employee of and stockholder in Eli Lilly. The principal investigator has served as a consultant to Eli Lilly

ADVANCE 2004

Methods

Method of treatment allocation: by the use of computer-based 1:1 randomisation scheme by an independent study nurse

Double-blinded?:  yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: supported by a grant from Cassa dei Risparmi di Ferrara, Italy

Follow-up: 6 months

Participants

Location: 1 hospital in Italy (Arcispedale S. Anna Hospital, University of Ferrara)

Timeframe: from March 2002 to August 2003

Eligibility criteria: 202 coronary patients undergoing elective or urgent PCI + stent with clinical or angiographic high-risk features. Inclusion criteria were the presence of > 1 stenosis > 70% amenable to coronary stenting and the presence of diabetes mellitus, or a planned multivessel intervention, or the presence of non-ST-segment elevation ACS

Exclusion criteria: ST-segment elevation myocardial infarction, administration of any GP IIb/IIIa inhibitors during the previous 2 weeks, serum creatinine ?2.5 mg/dl, ongoing bleeding or bleeding diathesis, previous stroke in the last 6 months, major surgery within the previous 6 weeks, and platelet count < 100,000 per mm3

Mean age: 69 y, male: 68%, diabetes: 49%, prior myocardial infarction: 48%
Acute coronary syndrome: 56% (unstable angina: ?, NSTEMI: ?, STEMI: 0%)
ST-segment depression: ?, CK-MB elevation: ?, Troponin elevation: ?

PCI: atherectomy: ?, balloon angioplasty: 2%, stent: 98%, pre-treatment with clopidogrel: 0%

InterventionsHigh-dose bolus tirofiban (25 µg/kg per 3 min) and infusion (0.15 µg/kg/min for 24 to 48 h) versus placebo
OutcomesPrimary: a composite of death, nonfatal MI, urgent TVR and thrombotic bailout GP IIb/IIIa inhibitor therapy
Secondary: each component of the primary endpoint, the effects on troponin I release after the procedure and the effects on pre-specified subgroups: diabetes and ACS
NotesAll patients were pre-treated with aspirin and a thienopyridine (clopidogrel 300 mg orally 6 h before the procedure (63%) or ticlopidine 500 mg 48 h before (37%))
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskUsing computer random number generation
Allocation concealment (selection bias)Unclear risk"by an independent study nurse"
Blinding (performance bias and detection bias)
Primary
Unclear riskDouble-blinded, placebo-controlled study, but the preparation of the placebo was not stated. Some key study personnel were not blinded, and this condition could bias the assessment of myocardial infarction, since no clinical events committee adjudicated the endpoints
Blinding (performance bias and detection bias)
Secondary
High riskSome key study personnel were not blinded, and this condition could bias the assessment of subjective endpoints
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published report includes all pre-specified outcomes
Other biasLow riskThe study appears to be free of other sources of bias

ASIAD 2005

Methods

Method of treatment allocation: by means of a computer-generated randomisation numbers at each centre

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Guidant, Corporation

Participants

Location: 7 centres in China (5 in Hong Kong), India and Singapore

N: 254 patients

Timeframe: from January 2001 to October 2002

Follow-up: 6 months

Eligibility criteria: type 2 diabetic patients undergoing PCI with planned use of stents to treat de novo ?50% coronary lesions

Exclusion criteria: platelet count < 100,000/L, concurrent warfarin therapy, INR ?1.5, uncontrolled hypertension, stroke in previous 2 y, transient ischaemic attack in previous 6 months, intracranial neoplasm, major surgery within 6 weeks, gastrointestinal bleeding within 3 months, PCI within 3 months, myocardial infarction within 5 days, and target lesion or target vessel proximal to target lesion containing thrombus

Mean age: 61 y, 74% male, 100% diabetes, 41% prior MI

ACS: 46% (non-STEACS: 31%, STEMI: 15%).

PCI: 100% (balloon angioplasty: 0%, stent: 100%, drug-eluting stents: 0%)

InterventionsAbciximab (bolus of 0.25 mg/kg up to 60 minutes before the intervention followed by a continuous infusion of 0.125 µg/kg/ min for 12 h) versus placebo (bolus and infusion)
Outcomes

Primary: incidence of angiographic restenosis at 6 months follow-up

Secondary: major cardiac events (death, MI or target lesion revascularisation) at 6 months

NotesClopidogrel pre-treatment was administered as a 300 mg loading dose at least 12 h before the procedure, or a 75 mg daily dose 5 days before the intervention
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"By means of a computer-generated randomisation numbers at each centre"
Allocation concealment (selection bias)Low riskNot specifically stated but probably yes
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blinded, placebo-controlled study.: "patients and investigators were blinded to treatment allocation".
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blinded, placebo-controlled study: "patients and investigators were blinded to treatment allocation"
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published report includes all pre-specified outcomes
Other biasUnclear riskThe study was supported by Guidant, Corporation

ASSIST 2009

Methods

Method of treatment allocation: not stated

Double-blinded?: no

Placebo: no

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Schering-Plough Canada Inc. provided an unrestricted grant and free distribution of eptifibatide, and Medtronic Canada Ltd provided free stents for the study

Follow-up: 30 days

Participants

Location: 3 centres in Ottawa, Canada

Timeframe: August 2005 to March 2008

Eligibility criteria: 400 patients with STEMI referred for primary PCI after pre-treatment with 160 mg of aspirin, 600 mg of clopidogrel and 60 U/kg of IV unfractionated heparin

Exclusion criteria: active bleeding, stroke within 90 days, intracranial bleeding at any time, major surgery or trauma within 6 weeks, systolic blood pressure > 200 mm Hg or diastolic blood pressure > 110 mm Hg, prolonged cardiopulmonary resuscitation, PCI within 30 days, fibrinolytic agents within 7 days, any glycoprotein IIb/IIIa inhibitors within 7 days, low-molecular-weight heparin within 12 h, coagulation disorder, current warfarin treatment, intolerance to aspirin or clopidogrel, other illness likely to result in death within 12 months, pregnancy, creatinine > 200 µmol/L, cardiogenic shock and severe contrast allergy

Mean age: 60 y, male: 77%, prior myocardial infarction: 13%
Acute coronary syndrome: 100%
ST-segment depression: 0%, CK-MB elevation: 100%, troponin elevation: ?%
PCI: balloon PCI: 1%, stent: 93%, pre-treatment with clopidogrel: 100%

InterventionsPatients were randomly assigned in a 1:1 ratio to PCI with heparin plus eptifibatide or to PCI with heparin alone, in blocks of 10 at each site. Immediately after randomisation, patients assigned to the eptifibatide group received open-label drug, 180 µg/kg of body weight bolus, followed by a continuous infusion of 2.0 µg/kg per minute with a second bolus of 180 µg/kg administered 10 minutes after the first bolus. The infusion was reduced to 1.0 µg/kg per minute in patients with a creatinine clearance of < 50 mL/min. Eptifibatide was to be initiated before cardiac catheterisation and the infusion continued for 18 hours after PCI
OutcomesPrimary: a composite of death from any cause, recurrent myocardial infarction or recurrent severe ischaemia during the first 30 days after randomisation
Secondary: stroke, congestive heart failure and cardiogenic shock. Episodes of bleeding were classified by an independent adjudicator as major or minor according to the TIMI criteria
NotesSTEMI patients referred for primary PCI
All patients pre-treated with high-dose clopidogrel before the procedure
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNo information provided, but probably done: "Patients were randomly assigned in a 1:1 ratio to PCI with heparin plus Eptifibatide or to PCI with heparin alone, in blocks of 10 at each site"
Allocation concealment (selection bias)Low riskNo information provided, but probably done: "Patients were randomly assigned in a 1:1 ratio to PCI with heparin plus Eptifibatide or to PCI with heparin alone, in blocks of 10 at each site"
Blinding (performance bias and detection bias)
Primary
Low riskOpen-label study, but the outcome measurement is not likely to be influenced by lack of blinding. A blinded off-site independent Clinical Event Committee adjudicated all primary events
Blinding (performance bias and detection bias)
Secondary
Low riskOpen-label study, but the outcome measurement is not likely to be influenced by lack of blinding. Episodes of bleeding were classified by an independent adjudicator. In addition, an independent Data Safety Monitoring Committee reviewed the data
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published report includes all expected outcomes
Other biasLow riskFunding: Schering-Plough Canada Inc provided an unrestricted grant and free distribution of Eptifibatide, and Medtronic Canada Ltd provided free stents for the study. However, the study appears to be free of other sources of bias

Bhattacharya 2010

Methods

Method of treatment allocation: by a computer-generated randomised table

Double-blinded?: yes

Placebo: yes, normal saline

Sample size calculation: not stated

Intention-to-treat analysis: yes

Funding: not stated

Follow-up: 3 months

Participants

Institutions: 1 clinical centre from Kolkata, India

Timeframe: June 2007 to May 2009

Eligibility criteria: 301 patients with angina at rest within the last 48 h of presentation and a new ST depression of > 1 mm, or a positive biomarker (cardiac troponin T > 0.04 μg/L or creatine kinase myocardial band (CK-MB) elevation > the upper limit of normal), and with a TIMI score > 4

Exclusion criteria: age > 80 years, persistent ST-segment elevation, prior treatment with tirofiban within 3 months of presentation, PCI within the last 6 months, cardiogenic shock, contraindications for the use of anti-platelet drugs or LMWH, ACE inhibitors and beta blockers, or presence of valvular or congenital heart disease

Mean age: 62 y, 54% males, 45% diabetics, 26% with prior MI

Acute coronary syndrome: 100% (unstable angina: 55%, non-ST elevation myocardial infarction: 45%, STEMI: 0%)

ST-segment depression: 95%, troponin elevation: 45%

Coronary angiography: 0%, PCI: 0% (balloon angioplasty: 0%, stent: 0%, drug-eluting stent: 0%)

Interventions

Tirofiban was initially administered at the rate of 0.4 μg/kg/min for 30 minutes and thereafter, the infusion rate was reduced to 0.1 μg/kg/min. This was continued for the next 47 h and 30 minutes. The placebo was administered in bottles similar to those of tirofiban and consisted of 0.9% normal saline

Conventional therapy was defined as aspirin (375 mg started during admission followed by 75 mg once daily) and clopidogrel (300 mg loading dose, followed by 70 mg daily), with low-molecular weight heparin (LMWH) (enoxaparin 1 U/kg sc twice daily)

Outcomes

Primary: a composite of total death, MI or refractory ischaemia at 1 and 3 months

Secondary: major bleeding

Notes

Study performed in Kolkata, India

No patient underwent coronary angiography or PCI

Complete outcome assessment available only for the first 7 days of follow-up

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)High risk"By a computer-generated randomised table." An imbalance occurred in the number of patients allocated to the intervention (n = 136) and the control groups (n = 165). However, no differences were observed in the baseline characteristics of both groups
Allocation concealment (selection bias)Unclear riskNo information provided
Blinding (performance bias and detection bias)
Primary
Low risk"The patients, the investigators that administered the drugs and the person who evaluated the patients were blinded to the treatment administered"
Blinding (performance bias and detection bias)
Secondary
Low risk"The patients, the investigators that administered the drugs and the person who evaluated the patients were blinded to the treatment administered"
Incomplete outcome data (attrition bias)
Primary
High riskAn imbalance occurred in the number of patients allocated to the intervention (n = 136) and the control groups (n = 165). 9 patients were lost of follow-up but, in fact, complete outcome assessment is available only for the first 7 days of follow-up. The incidence of the secondary endpoint (bleeding events) was 0% in both groups
Incomplete outcome data (attrition bias)
Secondary
High riskAn imbalance occurred in the number of patients allocated to the intervention (n = 136) and the control groups (n =1 65). 9 patients were lost of follow-up but, in fact, complete outcome assessment is available only for the first 7 days of follow-up. The incidence of the secondary endpoint (bleeding events) was 0% in both groups
Selective reporting (reporting bias)High riskThe study protocol is not available and the incidence of bleeding events was 0% in both groups. Complete outcome assessment is available only for the first 7 days of follow-up. The incidence of the secondary endpoint (bleeding events) was 0% in both groups
Other biasLow riskThe study appears to be free of other sources of bias

BRAVE-3 2009

Methods

Method of treatment allocation: according to a computer-generated random sequence enclosed in sealed envelopes in the emergency room or intensive care unit of the 5 participating PCI centres. The size of the block was preselected by the statistician and was unknown to the investigators and medical staff caring for the patients

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: grants from Deutsches Herzzentrum, Munich, Germany

Participants

Location: 5 centres in Germany and 1 in Austria

Timeframe: from June 2003 to January 2008

Follow-up: 30 days

Eligibility criteria: 800 patients with acute STEMI presenting < 24 h after the onset of symptoms, or presumed new left bundle-branch block on surface ECG, who gave informed consent

Exclusion criteria: thrombolytic therapy for the index infarction; those with previous stroke within the last 3 months, active bleeding or bleeding diatheses, recent trauma or major surgery during the last month, suspected aortic dissection, oral anticoagulation therapy with coumarin derivatives within the last 7 days, recent use of glycoprotein IIb/IIIa inhibitors within the last 14 days, severe uncontrolled hypertension, haemoglobin < 100 g/L or haematocrit < 4%, platelet count < 100,000/L or > 600,000/L), malignancies, prolonged cardiopulmonary resuscitation, or cardiogenic shock; those > 80 or < 18 years of age; those with known or suspected pregnancy; and those who were allergic to study drugs

Mean age: 62 y, 74% male, 18% diabetes, 10% prior MI

ACS: 100% (non-STEACS: 0%, STEMI: 100%)

PCI: 100% (balloon angioplasty: 4%, stent: 96%, drug-eluting stents: 44%), pre-treatment with clopidogrel: 100%

InterventionsAbciximab (bolus of 0.25 mg/kg followed by a continuous infusion of 0.125 µg/kg/ min (up to a maximal dose of 10 µg/min) for 12 h) versus placebo (bolus and infusion)
Outcomes

Primary: infarct size in a SPECT study

Secondary: total death resulting from any cause, recurrent MI, stroke, urgent IRA revascularisation at 30 days, and in-hospital incidence of major and minor bleeding complications

NotesAll patients received 600 mg clopidogrel orally at the emergency room or intensive care unit of the admitting hospital.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"according to a computer-generated random sequence"
Allocation concealment (selection bias)Low risk"enclosed in sealed envelopes in the emergency room or intensive care unit of the 5 participating PCI centres"
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blinded, placebo-controlled study
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blinded, placebo-controlled study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasLow riskThe study appears to be free of other sources of bias

CADILLAC 2002

Methods

Method of treatment allocation: not stated (patients were randomly assigned in a balanced fashion to 1 of 4 interventional strategies of reperfusion with the use of a 2 by 2 factorial design: PTCA alone, PTCA plus abciximab, stenting alone or stenting plus abciximab)

Double-blinded?: no, open-label study

Stratification: none stated

Placebo: no

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: supported in part by Guidant, Lilly Research Laboratories and Mallinkrodt

Follow-up: 30 days and 1 year

Participants

Location: 76 centres in USA and 8 European countries

Timeframe: from November 1997 to September 1999

Eligibility criteria: 2082 patients with ST-segment elevation acute myocardial infarction of less than 12 h with a coronary stenosis no longer than 64 mm and with a reference diameter of 2.5 to 4 mm. Patients were randomised using a 2 by 2 factorial design to undergo PTCA alone, PTCA plus abciximab, stenting alone or stenting plus abciximab

Exclusion criteria: cardiogenic shock, history of bleeding diathesis or allergy to the study drug; major surgery within the preceding 6 weeks; gastrointestinal or genitourinary bleeding within the preceding 6 months; cerebrovascular event within the preceding 2 years or any permanent residual neurologic defect; history of leukopenia, thrombocytopenia, or hepatic or renal dysfunction; recent treatment with a thrombolytic agent; a non-cardiac illness associated with a life expectancy of less than 1 year; and participation in another study

Mean age: 60 y, male: 73%, diabetes: 16%, prior myocardial infarction: 14%
Acute coronary syndrome: 100% (unstable angina: 0%, NSTEMI: 12%, STEMI: 88%)
ST-segment depression: 0%, CK-MB elevation: 100%, troponin elevation: ?%

PCI: balloon angioplasty: 50%, stent: 50%, pre-treatment with clopidogrel: 0%

InterventionsAbciximab (bolus of 0.25 mg per kilogram of body weight, followed by a 12-h infusion at a rate of 0.125 µg per kilogram per minute (maximum, 10 µg/minute)) versus control
OutcomesPrimary: a composite of death, reinfarction, repeated intervention or revascularisation of the target vessel as a result of ischaemia or disabling stroke during the first 6 months after the index procedure
Notes

Al patients received 324 mg of aspirin before the procedure, 500 mg ticlopidine or 300 mg clopidogrel orally, a 5000 U bolus of heparin and a beta-blocker IV

All but 2 of the authors have received research support from, or have served as consultants to, companies manufacturing interventional devices or pharmaceutical agents relevant to this study. These companies include, but are not limited to, Guidant and Lilly Research Laboratories, 2 of the sponsors of the study

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"Patients were randomly assigned in a balanced fashion to one of four interventional strategies of reperfusion with the use of a 2-by-2 factorial design"
Allocation concealment (selection bias)High riskOpen-label study
Blinding (performance bias and detection bias)
Primary
High riskIn spite of the existence of a clinical events adjudication committee, the open-label nature of the study could bias the assignment of myocardial infarction
Blinding (performance bias and detection bias)
Secondary
High riskThe open-label nature of the study could certainly bias the assignment of subjective outcomes such as revascularisation and major bleeding
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasUnclear risk"Supported in part by Guidant, Lilly Research Laboratories, and Mallinkrodt"

CANADIAN 1996

Methods

Method of treatment allocation: by sequential numbers generated centrally and incorporated into double-blind labelling

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: no

Intention-to-treat analysis: yes

Funding: F. Hoffmann-La Roche

Follow-up: 30 days

Participants

Institutions: 15 Canadian centres

Timeframe: not stated

Eligibility criteria: 365 patients with unstable angina or myocardial infarction without ST-segment elevation

Exclusion criteria: age > 75 years (380 patients); unstable angina precipitated by identifiable factors (67 patients) or occurring within 6 months of coronary angioplasty or 2 months after bypass surgery (198 patients); a previous stroke (22 patients); a high bleeding risk including trauma, surgery or active bleeding within the previous month (89 patients); shock, congestive heart failure, left bundle-branch block, uncontrolled hypertension, a life-threatening concomitant illness, platelet count < 100 000/mm3, use of oral anticoagulants or of an investigational drug, potential for pregnancy, or the inability to obtain informed consent (270 patients)

Mean age: 60 y, male: 72%, prior myocardial infarction: 55%
ACS: 100% (unstable angina: 86%, non-STEMI: 14%, STEMI: 0%), ST-segment depression: 65%, CK-MB elevation: 14%, Troponin elevation: ?
PCI: in-hospital: not stated, during drug-infusion: not stated
Treatment with clopidogrel: 0%

InterventionsDose-ranging study in which patients were randomly assigned to 1 of 5 parallel study arms: 1 with placebo and 4 with different bolus plus infusion doses of lamifiban: 150 µg plus 1 µg/min, 300 µg plus 2 µg/min, 600 µg plus 4 µg/min, and 750 µg plus 5 µg/min
OutcomesStudy designed to obtain a first evaluation of lamifiban in patients with unstable angina and to determine an optimal dose of the drug for a subsequent pivotal trial on a large scale
NotesDose-ranging trial: all 4 lamifiban arms were grouped for the analysis
All patients were treated with aspirin
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"sequential numbers generated centrally"
Allocation concealment (selection bias)Low risk"sequential numbers incorporated into double-blind labelling"
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blinded, placebo-controlled trial. "Events were all classified by a Critical Event Committee before the study was unblinded. Relevant data and documents without patient/centre identifiers were reviewed independently by a panel of two for all patients with any complication or event."
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blinded, placebo-controlled trial. "Events were all classified by a Critical Event Committee before the study was unblinded. Relevant data and documents without patient/centre identifiers were reviewed independently by a panel of two for all patients with any complication or event".
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasLow riskProbably yes because the trial sponsor (F. Hoffmann-La Roche) and the investigators remained blinded to randomisation code and study results until all study endpoints had been agreed on by the Critical Event Committee

CAPTURE 1997

Methods

Method of treatment allocation: randomisation was obtained by telephone call to an independent service organised by the Department of Clinical Epidemiology of the University of Amsterdam

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: not stated

Follow-up: 30 days and 6 months

Participants

Location: 69 centres in 12 countries (Netherlands, Germany, France, Belgium, Spain, UK, Italy, Israel, Switzerland, Canada, Portugal, Austria)

Timeframe: from May 1993 to December 1995

Eligibility criteria: 1265 patients with refractory unstable angina who underwent percutaneous transluminal coronary angioplasty

Exclusion criteria: recent myocardial infarction, persisting ischaemia that would require immediate intervention; a greater than 50% occlusion of the left main coronary artery or a culprit lesion located in a bypass graft; bleeding risk factors such as surgery, gastrointestinal or genitourinary bleeding during the 6 weeks before enrolment, or a cerebrovascular accident within the previous 2 years; planned administration of oral anticoagulants or a thrombolytic agent before or during PTCA; underlying medical conditions such as persistent hypertension despite treatment; history of haemorrhagic diathesis; history of autoimmune disease, or a platelet count below 100 x 109/L

Mean age: 61 y, male: 72%, prior myocardial infarction: 50%
Acute coronary syndrome: 100% (unstable angina: 100%, NSTEMI: 0%, STEMI: 0%)
ST-segment depression: ?, CK-MB elevation: 0%, troponin elevation: ?
PCI: balloon angioplasty: 98%, stent: 1%, pre-treatment with clopidogrel: 0%

InterventionsAbciximab (0.25 mg/kg bolus followed by a continuous infusion of 10 µg/min) versus matching placebo
Outcomes

Primary: 30-day incidence of death (from any cause), myocardial infarction or an urgent intervention for treatment of recurrent ischaemia

Secondary: each component of the composite endpoint and bleeding

Notes

Intervention began 18 to 24 h before percutaneous transluminal coronary angioplasty and continued until 1 h after it. Permits assessment of the effect of GP IIb/IIIa blockers as initial medical treatment of unstable angina and also during percutaneous transluminal coronary angioplasty

The trial was discontinued on the recommendation of the Safety and Efficacy Monitoring Committee after interim analysis of 1050 patients (planned 1400 patients)

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskUsing a computer random number generator
Allocation concealment (selection bias)Low risk"Randomisation was obtained by telephone call to an independent service organised by the Department of Clinical Epidemiology of the University of Amsterdam"
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blinded, placebo-controlled trial
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blinded, placebo-controlled trial
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasHigh riskThe trial was discontinued after the third interim analysis of 1050 patients (1400 planned) by the Safety and Efficacy Monitoring Committee

Chen 2000

Methods

Method of treatment allocation: not stated

Double-blinded ?: yes

Stratification: no

Placebo: yes

Sample size calculation: no

Intention-to-treat analysis: no

Funding: not stated

Follow-up: 30 days

Participants

Location: 1 centre in Taiwan (The Veterans General Hospital, Taipei)

Timeframe: from January 1997 to July 1997

Eligibility criteria: 42 coronary patients scheduled to undergo coronary angioplasty. They were eligible if they had early postinfarction angina or unstable angina with >= 2 episodes of angina at rest associated with ECG changes during the previous 24 h; or clinical or angiographic characteristics indicating high risk according to the AHA/ACC criteria

Exclusion criteria: patients > 80 y old, bleeding diatheses, major surgery within prior 6 weeks, stroke within prior 2 y, planned stent implantation or atherectomy

Mean age: 70 y, male: 95%, diabetes:35%, prior myocardial infarction: 46%
Acute coronary syndrome: 29% (unstable angina: 29%, NSTEMI: ?%, STEMI: 0%)
PCI: balloon angioplasty: 100%, stent: 0%, pre-treatment with clopidogrel: 0%

InterventionsAbciximab (bolus of 0.25 mg/kg, followed by an infusion of 10 µg/min for 12 hours) versus placebo
Outcomes

Primary: 30-day incidence of the composite endpoint of death, nonfatal myocardial infarction, unplanned surgical or repeated percutaneous revascularisation, or insertion of an intra-aortic balloon pump for refractory ischaemia

Secondary: each component of the composite endpoint and bleeding

NotesSmall study performed in 42 Chinese patients
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot stated
Allocation concealment (selection bias)Unclear riskNot stated: "Patients were randomly assigned in a double-blind fashion to one of the two treatment groups"
Blinding (performance bias and detection bias)
Primary
Unclear risk"Double-blinded, placebo-controlled study", but no specific methods reported
Blinding (performance bias and detection bias)
Secondary
Unclear risk"Double-blinded, placebo-controlled study", but no specific methods reported
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasUnclear riskTo few explanations in the study report

Claeys 2005

Methods

Method of treatment allocation: not stated

Double-blinded ?: no

Stratification: no

Placebo: no

Sample size calculation: no

Intention-to-treat analysis: no

Funding: Sanofi-Synthelabo, Belgium and Dade Behring, Belgium

Follow-up: 30 days and 6 months

Participants

Location: 1 centre in Belgium: Antwerp University Hospital, Antwerp, Belgium

Timeframe: from October 2001 to November 2003

Eligibility criteria: 200 patients scheduled for elective PCI with stent implantation and pre-treated with aspirin and a loading dose of clopidogrel (450 mg) were randomised just before coronary intervention to treatment with or without abciximab

Exclusion criteria: acute coronary syndromes requiring urgent coronary intervention or early treatment with GP IIb/IIIa receptor antagonists, recent myocardial infarction, intervention of lesions located in bypass grafts or near major side branches, presence of an angiographically visible intracoronary thrombus, patients with creatinine value > 2.0 mg/dL, with haemostatic disorders, or with a history of intolerance to thienopyridine or to abciximab

Mean age: 67 y, male: 70%, prior myocardial infarction: 19%
Acute coronary syndrome: 33% (unstable angina: 31%, NSTEMI: 0%, STEMI: 0%)
PCI: balloon angioplasty: 2%, stent: 98%, pre-treatment with clopidogrel: 100%

Interventions

Abciximab (bolus of 0.25 mg/kg, followed by an infusion of 0.125 µg/kg/min for 12 h) versus placebo. All patients treated with 160 mg ASA, unfractionated heparin and clopidogrel at randomisation

Abciximab was administered just before intervention as an intravenous bolus (0.25 mg/kg) followed by a 12-h infusion (10 µg/min)

OutcomesPrimary: level of platelet aggregation inhibition and of peri-procedural myonecrosis
Secondary: a composite endpoint of death, MI or urgent target-vessel revascularisation within 30 days of randomisation
NotesLow-risk patients scheduled for elective PCI with stent placement
All patients pre-treated with high-dose clopidogrel (300 mg) in the evening preceding PCI and 150 mg in the morning of the intervention
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot stated
Allocation concealment (selection bias)Unclear riskNot stated
Blinding (performance bias and detection bias)
Primary
High riskThe open-label nature of the study could influence the diagnosis of myocardial infarction
Blinding (performance bias and detection bias)
Secondary
High riskOpen-label study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasUnclear riskFunding: Sanofi-Synthelabo, Belgium and Dade Behring, Belgium

CLEAR PLATELETS-2 2009

Methods

Method of treatment allocation: by a computer-generated assignment

Double-blinded?: no, open-label study

Stratification: yes, according to clopidogrel therapy before PCI

Placebo: no

Sample size calculation: yes

Intention-to-treat analysis: no

Funding: a research grant from Integrated Therapeutics Group, Inc., a subsidiary of Schering-Plough Corporation

Participants

Location: 2 centres in USA

Timeframe: from March 2006 to December 2007

Follow-up: 30 days and 6 months

Eligibility criteria: 200 consecutive stable patients undergoing PCI

Exclusion criteria: age < 18 y, history of bleeding diathesis, MI within 48 h, elevated cardiac markers of necrosis, cerebrovascular event within 3 months, chronic vessel occlusion, visible thrombus, illicit drug or alcohol abuse, prothrombin time > 1.5 times control, platelet count < 100,000/mm3, haematocrit < 30%, creatinine > 2.0 mg/dl, and anticoagulation therapy or GP IIb/IIIa blocker use before the procedure

Mean age: 64 y, 61% male, 41% diabetes, 36% prior MI

ACS: 0%

PCI: 100% (balloon angioplasty: 3%, stent: 97%, drug-eluting stents: 72%), pre-treatment with clopidogrel: 100%

InterventionsEptifibatide plus bivalirudin (n = 98) versus bivalirudin alone (n = 102) stratified by prior clopidogrel treatment, thus providing 4 treatment groups: 1) 600 mg clopidogrel plus bivalirudin; 2) 75 mg clopidogrel plus bivalirudin; 3) 600 mg clopidogrel plus bivalirudin plus eptifibatide; and 4) 75 mg clopidogrel plus bivalirudin plus eptifibatide. Clopidogrel-naive patients (n = 128) received treatment with 600 mg clopidogrel in the catheterisation laboratory immediately after stenting, whereas patients currently on 75 mg (n = 72) did not receive a load. Eptifibatide was administered as a double bolus (180 µg/kg) followed by an infusion (2 µg/kg/min) for 18 h after the procedure
Outcomes

Primary: effects on platelet reactivity measured by turbidometric aggregometry and thrombin-induced platelet-fibrin clot strength measured by thrombelastography in PCI patients

Secondary: to study the relation of platelet aggregation and thrombin-induced platelet-fibrin clot strength to the occurrence of periprocedural infarction

Notes

All patients received clopidogrel therapy. Clopidogrel-naive patients received treatment with 600 mg clopidogrel in the catheterisation laboratory immediately after stenting

72% received a drug-eluting stent

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskBy computer-generated random numbers
Allocation concealment (selection bias)Low riskBy computer-generated assignment. Details previously published
Blinding (performance bias and detection bias)
Primary
High riskOpen-label study
Blinding (performance bias and detection bias)
Secondary
High riskOpen-label study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasUnclear riskFunding: a research grant from Integrated Therapeutics Group, Inc., a subsidiary of Schering-Plough Corporation

Cuisset 2008

Methods

Method of treatment allocation: not stated

Double-blinded?: no, open-label study

Stratification: no

Placebo: no

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: the Assistance Publique Hôpitaux de Marseille

Participants

Location: 1 centre in France

Timeframe: not stated

Follow-up: 30 days

Eligibility criteria: 149 patients older than 18 years with stable angina or a positive functional study with a planned PCI with stent implantation of a de novo lesion in a native coronary artery

Exclusion criteria: left ventricular ejection fraction < 30%, ACS in the previous month, prior MI in the target vessel related territory, positive biomarkers pre-PCI, platelet count < 100 g/l and history of bleeding diathesis

Mean age: 65 y, 75% male, 38% diabetes, ?% prior MI

ACS: 0%

PCI: 100% (balloon angioplasty: 0%, stent: 100%, drug-eluting stents: ?%), pre-treatment with clopidogrel: 100%

InterventionsAbciximab (0.25 mg/kg of body weight bolus, followed by a 0.125 μg/kg/min (maximum, 10 mg/min) infusion for 12 h) versus control
Outcomes

Primary: death from any cause, periprocedural myonecrosis, acute or subacute definite or probable stent thrombosis and recurrent ACS

Secondary: major bleeding

NotesStudy on patients non-responders to clopidogrel therapy. Antiplatelet therapy was administered with loading doses of 600 mg of clopidogrel and 250 mg of aspirin the day before the procedure
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot stated
Allocation concealment (selection bias)Unclear riskNot stated
Blinding (performance bias and detection bias)
Primary
High riskOpen-label study
Blinding (performance bias and detection bias)
Secondary
High riskOpen-label study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasUnclear riskInsufficient information

DANTE 2004

Methods

Method of treatment allocation: by use of sealed envelope

Double-blinded?: no, single-blinded

Stratification: no

Placebo: no

Sample size calculation: yes

Intention-to-treat analysis: yes

Losses to follow-up: no

Funding: research grant from Sociedade Brasileira de Cardiologia (SBC/FUNCOR)

Follow-up: 30 days and 1 year

Participants

Location: 1 centre in Brazil (Instituto Dante Pazzanese de Cardiologia, Sao Paulo)

Timeframe: from February 2001 to March 2002

Eligibility criteria: 96 patients with diabetes mellitus and a symptomatic or ischaemia-provoking de novo stenosis in a native coronary artery

Exclusion criteria: myocardial infarction within < 7 days, renal dysfunction (serum creatinine > 2.0 mg/dL) or standard contraindications to abciximab. Angiographic exclusions included target vessel < 2.5 mm or > 4.0 mm in reference diameter, target lesion > 18 mm in length, impaired left ventricular function (ejection fraction < 30%), unprotected left main coronary lesion > 50%, ostial lesion, total occlusion, or evidence of thrombus

Mean age: 60 y, male: 49%, diabetes mellitus: 100%, prior myocardial infarction: 50%
Acute coronary syndrome: 24% (none with acute myocardial infarction)
PCI: 100% (balloon PCI: 0%, stent: 100%, drug-eluting stents: ?%, pre-treatment with clopidogrel: 0%

InterventionsPatients were randomly assigned, by use of sealed envelopes, to stenting with versus without abciximab. Abciximab was administered as a bolus of 0.25 mg/kg, followed by a 12-h infusion (0.125 µg/kg per minute, maximum 10 µg/min). Patients received 200 mg of aspirin and 500 mg of ticlopidine started > 24 h before the procedure
Outcomes

Primary: the primary endpoint was the in-stent per cent volume obstruction as measured by IVUS at 6-month follow-up

Secondary: secondary angiographic endpoints included MLD, late loss and binary restenosis. Secondary clinical endpoints were the composite major adverse cardiac events (MACE) of death of any cause, nonfatal myocardial infarction and target lesion revascularisation (TLR) within 30 days and 1 year after the procedure

NotesStudy on diabetic patients undergoing PCI with stent implantation. All patients pre-treated with 500 mg of ticlopidine started > 24 h before the procedure
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo information provided but probably done: “Patients were randomly assigned”
Allocation concealment (selection bias)Unclear riskNot enough information was provided, just that it was performed “By use of sealed envelopes”
Blinding (performance bias and detection bias)
Primary
High riskOpen-label study. Mortality and myocardial infarction adjudicated by the investigators
Blinding (performance bias and detection bias)
Secondary
High riskOpen-label study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all the prespecified outcomes
Other biasLow riskThe published reports include all the prespecified outcomes

De Luca 2005

Methods

Method of treatment allocation: not stated, but number of patients different in the 2 groups

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: no

Intention-to-treat analysis: not stated

Losses to follow-up: no, but number of patients different in the 2 groups

Funding: not stated

Follow-up: 30 days and 12 months

Participants

Location: 1 centre in Rome, Italy

Timeframe: January 2002 to May 2003

Eligibility criteria: 122 patients with DM undergoing elective PCI with stenting of de novo coronary artery lesions

Exclusion criteria: if they require urgent procedures for an acute MI or haemodynamic instability, they are undergoing interventions in previously instrumented vessels, had a chronic occlusion (present for > 3 months), had a target lesion with angiographically visible thrombus, had trauma or major surgery in the preceding month, or had a stroke in the prior 3 months

Mean age: 62 y, male: 66%, diabetes: 100%, prior myocardial infarction: 12%
Acute coronary syndrome: 0%
PCI: balloon PCI: 0%, stent: 100% (10% DES), pre-treatment with clopidogrel: 0%

Interventions

Abciximab or placebo was administered in a double-blind manner. Abciximab was given as 0.25 mg/kg bolus and 0.125 µg/Kg/min infusion immediately before revascularisation, and continued for 12 hs. All patients were treated with aspirin before the procedure.

Heparin was dosed to achieve an activated clotting time of > 250 seconds. After stent implantation, all patients received either clopidogrel (loading dose of 300 mg, followed by 75 mg/d for at least 4 weeks) or ticlopidine (250 mg orally twice a day for at least 4 weeks)

OutcomesPrimary: MACE at 1 year. The diagnosis of MI was based on either the development of new pathological Q waves in ≥ 2 contiguous electrocardiogram leads or an elevation of creatine kinase or its MB isoenzyme to ≥ 3 times the upper limit of normal in ≥ 2 samples
Secondary: need for TVR (surgical or percutaneous) at 30 and 180 days
NotesDiabetic patients undergoing elective PCI
Clopidogrel (300 mg) or ticlopidine (250 mg) after PCI
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo information provided. “Abciximab or placebo was administered in a double-blind manner”, but probably not done since the number of patients allocated to each group were markedly different (69 versus 53)
Allocation concealment (selection bias)Unclear riskNo information provided
Blinding (performance bias and detection bias)
Primary
Unclear riskNot enough information provided, just that “Abciximab or placebo was administered in a double-blind manner “. No data provided on the placebo used and the method of administration
Blinding (performance bias and detection bias)
Secondary
Unclear riskNot enough information provided, just that “Abciximab or placebo was administered in a double-blind manner “. No data provided on the placebo used and the method of administration
Incomplete outcome data (attrition bias)
Primary
Unclear riskInsufficient information provided
Incomplete outcome data (attrition bias)
Secondary
Unclear riskInsufficient information provided
Selective reporting (reporting bias)Unclear riskInsufficient information to permit judgement
Other biasLow riskThe study appears to be free of other sources of bias

De Luca 2008

Methods

Method of treatment allocation: not stated

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: not stated.

Losses to follow-up: 4 (3%) of the 132 patients

Funding: not stated

Follow-up: 30 days and 6 months

Participants

Location: 1 centre in Rome, Italy

Timeframe: not stated

Eligibility criteria: 132 patients diabetic patients undergoing elective PCI with stenting of de novo coronary artery lesions

Exclusion criteria: need for urgent procedures for an acute MI or haemodynamic instability, undergoing interventions in previously instrumented vessels, had a chronic occlusion (present for longer than 3 months), target lesion with angiographically visible thrombus, trauma or major surgery in the preceding month or a stroke in the previous 3 months

Mean age: 63 y, male: 62%, diabetes: 100%, prior myocardial infarction: 12%
Acute coronary syndrome: 27% (unstable angina 11%, NSTEMI 16%)
PCI: balloon PCI: 0%, stent: 100% (10% DES), pre-treatment with clopidogrel: 0%

Interventions

Abciximab or placebo was administered in a double-blind manner. Abciximab was given as 0.25 mg/kg bolus and 0.125 mµg/Kg/min infusion immediately before revascularisation and continued for 12 hs. All patients were treated with aspirin before the procedure

Heparin was dosed to achieve an activated clotting time of > 250 seconds. Before or immediately after stent implantation, all patients received clopidogrel (loading dose of 300 mg, followed by 75 mg/d for at 9 months)

Outcomes

Primary: MACE at 30 days year: target-lesion revascularisation (TLR), death for every cause and MI. Periprocedural MI was defined as a raise in creatine kinase (CK)-MB of more than 3 times the upper limit of normal. The diagnosis of re-MI during follow-up required the development of new, pathologic Q-waves in 2 or more contiguous leads, with creatine kinase or creatine kinase MB levels elevated above the upper limit of the normal range

Secondary: bleeding complications

NotesDiabetic patients undergoing elective PCI
Clopidogrel (300 mg) immediately before or after PCI
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo information provided (“Abciximab or placebo was administered in a double-blind manner”)
Allocation concealment (selection bias)Unclear riskNo information provided
Blinding (performance bias and detection bias)
Primary
Unclear riskNot enough information provided, just that "Abciximab or placebo was administered in a double-blind manner". No data on the placebo used and the method of administration
Blinding (performance bias and detection bias)
Secondary
Unclear riskNot enough information provided, just that "Abciximab or placebo was administered in a double-blind manner". No data on the placebo used and the method of administration
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Unclear riskInsufficient information provided, but it seems that all the planned outcomes were reported
Other biasLow riskThe study appears to be free of other sources of bias

ELISA-2 2006

Methods

Method of treatment allocation: by a computerised randomisation procedure

Double-blinded ?: no, open-label study

Stratification: no

Placebo: no

Sample size calculation: yes

Intention-to-treat analysis: no

Funding: not stated

Follow-up: 30 days

Participants

Location: 3 centres in The Netherlands

Timeframe: not stated

Eligibility criteria: 328 patients with NSTEACS who underwent coronary angiography a median of 23 h after admission

Exclusion criteria: age > 80 y, persistent ST-segment elevation, previous PCI within the preceding 6 months, cardiogenic shock, or a contraindication for the use of triple antiplatelet therapy or invasive therapy

Mean age: 63 y, male: 71%, diabetes:18%, prior myocardial infarction: 21%
Acute coronary syndrome: 100% (unstable angina: 22%, NSTEMI: 78%, STEMI: 0%)
ST-segment depression: 61%, CK-MB elevation: ?, troponin elevation: 78%
PCI: in-hospital: 59%, balloon angioplasty: 10%, stent: 49%, pre-treatment with clopidogrel: 100%

InterventionsPatients were randomised to pre-treatment with either dual (aspirin, clopidogrel 600 mg) or triple antiplatelet therapy (aspirin, clopidogrel 300 mg and tirofiban 10 µg/kg bolus, 0.15 µg/kg/min maintenance)
Study medication was given in an open-label manner. All patients were scheduled for coronary angiography within 48 h after admission
OutcomesPrimary: enzymatic infarct size
Secondary: initial TIMI flow of the culprit vessel
Notes

Comparison of dual versus triple antiplatelet pre-treatment in patients with NSTEACS who were planned for early catheterisation.

All patients were pre-treated with 300 to 600 mg of clopidogrel

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskBy a computerised randomisation procedure
Allocation concealment (selection bias)Unclear riskNot stated
Blinding (performance bias and detection bias)
Primary
High riskOpen-label study
Blinding (performance bias and detection bias)
Secondary
High riskOpen-label study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasHigh risk44% of the patients in the control group received the intervention before (5%), during (22%) or after (17%) PCI

EPIC 1994

Methods

Method of treatment allocation: by telephone (patients were randomly assigned to 1 of the 3 treatment groups according to a double-blind study design)

Double-blinded ?: yes

Stratification: according to their study centre and whether they were having an acute evolving myocardial infarction

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: supported by a grant from Centocor, Inc., Malvern, Pa

Follow-up: 30 days

Participants

Location: 56 institutions in the United States

Timeframe: from November 1991 to November 1992

Eligibility criteria: 2099 patients with unstable angina, angina post myocardial infarction or clinically or angiographically high-risk patients undergoing percutaneous transluminal coronary angioplasty or atherectomy

Exclusion criteria: age > 80 years old, bleeding diathesis, major surgery within the preceding 6 weeks, stroke within the preceding 2 years

Mean age: 61 y, male: 72%, prior myocardial infarction: 56%
Acute coronary syndrome: 100% (unstable angina: 43%, NSTEMI: ?%, STEMI: 41%)
ST-segment depression: ?, CK-MB elevation: ?, troponin elevation: ?
PCI: atherectomy: 10%, balloon angioplasty: 90%, stent: 0%, pre-treatment with clopidogrel: 0%

InterventionsAbciximab bolus (0.25 mg/kg) + placebo infusion versus abciximab bolus + infusion (10 µg per minute) versus placebo bolus and infusion
Outcomes

Primary endpoint: the 30-day incidence of the composite endpoint of  death from any cause, nonfatal myocardial infarction, coronary-artery bypass grafting or repeat percutaneous intervention for acute ischaemia, and insertion of a coronary endovascular stent because of procedural failure or placement of an intra-aortic counterpulsation balloon pump to relieve refractory ischaemia.

Secondary endpoint: unplanned repeat angioplasty to treat recurrent ischaemia, urgent coronary surgery to treat recurrent ischaemia or failure of an angioplasty, placement of an intracoronary stent to treat imminent or complete abrupt closure of the vessel undergoing angioplasty, and placement of an intra-aortic balloon pump for recurrent ischaemia when a repeat revascularisation procedure was contraindicated

NotesExcluded bolus-alone arm from analysis. This was the only study using the bolus-alone arm
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskRandomisation by telephone
Allocation concealment (selection bias)Low riskNot specified but probably correct
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind, placebo-controlled trial with Clinical Events Committee
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blind, placebo-controlled trial with Clinical Events Committee
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data.
Incomplete outcome data (attrition bias)
Secondary
Unclear riskNo missing outcome data.
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasUnclear riskSupported by a grant from Centocor, Inc., Malvern, Pa

EPILOG 1997

Methods

Method of treatment allocation: by means of a central telephone hot line

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Centocor, Malvern, Pa. and Eli Lilly and Company, Indianapolis

Follow-up: 30 days

Participants

Location: 69 clinical sites in the United States and Canada

Timeframe: from 27 February 1995 to December 1995

Eligibility criteria: 2792 patients with ST-segment elevation acute myocardial infarction, unstable or stable angina undergoing percutaneous transluminal coronary angioplasty

Exclusion criteria: acute myocardial infarction or unstable angina, planned stent implantation or rotational atherectomy; percutaneous coronary intervention performed within the previous 3 months; a left-main coronary artery stenosis of more than 50% not protected by collateral vessels; concurrent warfarin therapy or a baseline prothrombin time more than 1.2 times the control value; cerebrovascular accident within the previous 2 years or a residual neurologic deficit; intracranial neoplasm, aneurysm, or arteriovenous malformation; history of vasculitis, known haemorrhagic diathesis, or active internal bleeding; hypertension, with a systolic blood pressure of more than 180 mm Hg or a diastolic blood pressure of more than 100 mm Hg; and major surgery, gastrointestinal bleeding, or genitourinary bleeding within the previous 6 weeks

Mean age: 60 y, male: 72%, prior myocardial infarction: %
Acute coronary syndrome: 69% (unstable angina: 48%, NSTEMI: ?, STEMI: ?
ST-segment depression: ?, CK-MB elevation: ?, troponin elevation: ?
PCI: atherectomy: 5%, balloon angioplasty: 95%, stent: 11%, pre-treatment with clopidogrel: 0%

InterventionsAbciximab + low-dose heparin versus abciximab + standard dose heparin versus placebo + standard dose heparin. For those receiving abciximab, a bolus of 0.25 mg per kilogram of body weight was administered followed by an infusion of 0.125 mg per kilogram per minute (maximum, 10 mg per minute) for 12 hours
Outcomes

Primary endpoint: a composite of death from any cause, myocardial infarction or reinfarction, or severe myocardial ischaemia requiring urgent coronary bypass surgery or repeated percutaneous coronary revascularisation within 30 days after randomisation

Secondary endpoints: a composite of death, myocardial infarction, or coronary bypass surgery or repeated percutaneous revascularisation (urgent or non-urgent) within 6 months after randomisation

NotesThe 2 abciximab groups have been grouped together for the analysis
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNot specified "by a central telephone", but probably correct
Allocation concealment (selection bias)Low riskBy means of a central telephone hot line
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind, placebo-controlled trial with Clinical Events Committee. "To preserve the blinding of all investigators and personnel involved in patient care, a heparin coordinator at each clinical site performed all measurements of activated clotting time and directed the administration of heparin"
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blind, placebo-controlled trial with Clinical Events Committee
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Unclear riskThe published reports include all pre-specified and expected outcomes
Other biasHigh risk

The study was stopped after enrolment of 2792 of the planned 4800 patients after the recommendation of the Data and Safety Monitoring Committee

Funding: Centocor, Malvern, Pa. and Eli Lilly and Company, Indianapolis

EPISTENT 1998

Methods

Method of treatment allocation: by a telephone hotline

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Centocor, Malvern PA, USA

Follow-up: 30 days

Participants

Institutions: 63 centres in the USA and Canada

Timeframe: not stated

Eligibility criteria: 2399 patients scheduled to elective or urgent coronary angioplasty or stent

Exclusion criteria: unprotected left-main stenosis, bleeding diathesis, intracranial neoplasm, a history of stroke in the previous 2 years, uncontrolled hypertension (systolic blood pressure > 180 mm Hg, diastolic > 100 mm Hg), recent surgery, or percutaneous coronary intervention within the previous 3 months, concurrent warfarin therapy or an INR > 1.5 at baseline

Mean age: 59 y, male: 75%, prior myocardial infarction: 35%
Acute coronary syndrome: 52% (unstable angina: 36%, NSTEMI: ?, STEMI: ?)

ST-segment depression: ?, CK-MB elevation: ?, troponin elevation: ?

PCI: balloon angioplasty: 29%, stent: 71%, pre-treatment with clopidogrel: 0%

Interventions

Abciximab + stent versus abciximab + balloon coronary angioplasty versus placebo + stent

Patients were randomly assigned stent plus placebo (n = 809), stent plus abciximab (n = 794) or balloon angioplasty plus abciximab (n = 796). Patients received abciximab 0.25 mg/kg body weight, followed by an infusion of 0.125 μg/kg every min (maximum 10 μg/min) for 12 hours

Outcomes

Primary endpoint: a combination of death from any cause, myocardial infarction or reinfarction, or severe myocardial ischaemia requiring urgent coronary artery bypass surgery or revascularisation within 30 days of intervention

Secondary endpoints: death or myocardial infarction, and death or large myocardial infarction

NotesAll patients treated with aspirin and heparin. The stent group received also ticlopidine 250 mg twice daily starting at the discretion of the investigator before the start of the study agent
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskBy computer generation
Allocation concealment (selection bias)Low riskRandomisation by telephone. "The study drug allocation was concealed from patients and investigators". "The masking of the study drug allocation was maintained through the 1-year follow-up"
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind, placebo-controlled trial with Clinical Events Committee
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blind, placebo-controlled trial with Clinical Events Committee
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasUnclear riskStudy sponsored by Centocor, Malvern PA, USA

ERASER 1999

Methods

Method of treatment allocation: into 1 of 3 groups by sealed envelopes provided by the co-ordinating centre

Double-blinded ?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: no

Funding: not stated.

Follow-up: hospitalisation and 6 months

Participants

Institutions: 17 centres from 9 countries (USA, Canada, Israel, Italy, Germany, France, Belgium, UK, Netherlands)

Timeframe: from May 1996 to February 1997

Eligibility criteria: 215 angina patients with 1 lesion suitable for stent implantation (>/ = 50% reduction in intraluminal diameter) in an artery with an intraluminal diameter from 2.75 to 3 mm. Excluded were patients with acute coronary syndrome < 72 h

Exclusion criteria: myocardial infarction within 72 h before randomisation, evident intracoronary thrombus, previous coronary intervention on a non-target lesion within the past 6 months, planned debulking before stent placement, expected inability to access the target lesion by IVUS, or standard contraindications to the use of abciximab

Mean age: 60 y, male: 79%, prior myocardial infarction: ?
Acute coronary syndrome: 56% (see exclusions) (unstable angina: 56%, NSTEMI: 0%, STEMI: 0%). ST-segment depression: ?, CK-MB elevation: ?, Troponin elevation: ?
PCI: balloon angioplasty: 0%, stent: 100%, pre-treatment with clopidogrel: 0%.

Interventions3 different treatment regimens: placebo bolus + 2 consecutive 12-h placebo infusions; abciximab 0.25 mg/kg bolus + 0.125 µg/kg/min (up to 10 µg/min maximum) continuous infusion for 12 hours followed by 12-h placebo infusion; or abciximab 0.25 mg/kg bolus + 2 consecutive 12-h 0.125 µg/kg//min (up to 10 µg/min maximum) infusions
Outcomes

Primary efficacy criterion: per cent in-stent volume obstruction of the target lesion, measured at 6 months by IVUS. Primary safety objectives: major bleeding not associated with bypass surgery, and mortality and intracranial haemorrhage through 6 months

Secondary efficacy objectives: target lesion mean and minimum lumen diameter, late loss and loss index by QCA at 6 months, and a composite of death, myocardial infarction and TLR within 6 months

NotesThe 12-h and 24-h infusion groups were grouped together for the analysis. In-hospital was considered 30-day in this review.
All patients treated with aspirin and heparin. Ticlopidine use was left to the investigator's discretion
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer random number generator
Allocation concealment (selection bias)Low riskBy sealed envelopes provided by the co-ordinating centre
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind, placebo-controlled trial
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blind, placebo-controlled trial
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasLow riskThe study appears to be free of other sources of bias

ESPRIT 2000

Methods

Method of treatment allocation: generated with random permuted blocks within each investigative site, with 1 to 1 allocation of treatments

Double-blinded?:  yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: COR Therapeutics, South San Francisco, CA, USA and Schering-Plough Corporation, Kenilworth, NJ, USA

Follow-up: 30 days

Participants

Institutions: 92 centres in USA and Canada

Timeframe: from 3 June 1999 to 4 February 2000

Eligibility criteria: 2064 patients with coronary artery disease undergoing stent implantation on a native coronary artery

Exclusion criteria: acute myocardial infarction; continuing chest pain precipitating urgent referral for PCI; PCI within the previous 90 days; previous stent implantation at the target lesion; anticipated staged PCI, treatment with a glycoprotein IIb/IIIa inhibitor or a thienopyridine, stroke or transient ischaemic attack within 30 days before randomisation; any history of haemorrhagic stroke; history of bleeding diathesis or evidence of abnormal bleeding within the previous 30 days; major surgery within the previous 6 weeks; uncontrolled hypertension; documented thrombocytopenia; or a serum creatinine greater than 350 mmol/L

Mean age: 62 y, male: 73%, prior myocardial infarction: 32%
Acute coronary syndrome: 18% (unstable angina: 14%, NSTEMI: ?, STEMI: 5%)
ST-segment depression: ?, CK-MB elevation: ?, Troponin elevation: ?
PCI: balloon angioplasty: 3%, stent: 96%, pre-treatment with clopidogrel: 1% (all patients pre-treated with a thienopyridine, either ticlopidine or clopidogrel, with use of a loading dose) on the day of randomisation

InterventionsEptifibatide was delivered as 2 boluses and an infusion. The first bolus of 180 μg/kg was immediately followed by initiation of a 2.0 µg/kg/min (or 1 µg/kg/min in patients with serum creatinine values greater than 177 (mol/L) continuous infusion. A second bolus of 180 µg/kg was given 10 min after the first. The infusion was continued until hospital discharge or up to 18-24 h
Outcomes

Primary endpoint: the composite of death, myocardial infarction, urgent target vessel revascularisation and thrombotic bailout glycoprotein IIb/IIIa inhibitor therapy within 48 h after randomisation

Secondary endpoint: the composite of death, myocardial infarction and urgent target vessel revascularisation within 30 days after randomisation

Notes

All patients treated with aspirin, heparin and clopidogrel or ticlopidine

Sample size calculation was based on a predicted reduction in the rate of the key 30-day secondary composite efficacy endpoint, rather than the primary endpoint

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer random number generator
Allocation concealment (selection bias)Low riskStudy drugs were packaged to be indistinguishable, irrespective of content
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind, placebo-controlled trial
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blind, placebo-controlled trial
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasHigh risk

The data and safety monitoring board recommended to stop the trial after 2064 patients were enrolled (2400 planned)

Funding: COR Therapeutics, South San Francisco, CA, USA and Schering-Plough Corporation, Kenilworth, NJ, USA

Fu 2008

Methods

Method of treatment allocation: not stated

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: no

Intention-to-treat analysis: no

Funding: a grant from the Natural Science Foundation of Hebei Province (no C2004000615)

Participants

Location: 1 centre in China

Timeframe: from January 2005 to May 2007

Follow-up: 6 months

Eligibility criteria: 150 patients with acute STEMI (or documented new left bundle-branch block) presenting < 12 hs after the onset of symptoms and with coronary anatomy suitable for PCI

Exclusion criteria: bleeding diathesis, administration of thrombolytic agents, neoplasm, recent stroke, uncontrolled hypertension, recent surgery, oral anticoagulant therapy and known contraindications to therapy with aspirin, clopidogrel, tirofiban or heparin

Mean age: 53 y, 68% male, 19% diabetes,?% prior MI

ACS: 100% (non-STEACS: 0%, STEMI: 100%)

PCI: 100% (balloon angioplasty: ?%, stent: ?%, drug-eluting stents: ?%), pre-treatment with clopidogrel: 100%

InterventionsTirofiban (bolus of 10 µg/kg over 3 minutes followed by continuous infusion of 0.15 µg/kg/min for 24 h) versus placebo (bolus and infusion). All patients treated with 300 mg of clopidogrel at enrolment
Outcomes

Primary: not clear (“to evaluate the effect and safety of tirofiban in STEMI patients undergoing PCI via transradial approach”)

Secondary: not stated

NotesChinese study. All patients treated with 300 mg of clopidogrel at enrolment
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot stated and probably not done since the number of patients in each group were different.
Allocation concealment (selection bias)Unclear riskNot stated.
Blinding (performance bias and detection bias)
Primary
Unclear risk

Patients were blinded with use of placebo (infusion of saline). Unclear information about blinding of investigators.

 

Blinding (performance bias and detection bias)
Secondary
Unclear riskPatients were blinded with use of placebo (infusion of saline). Unclear information about blinding of investigators
Incomplete outcome data (attrition bias)
Primary
Unclear riskNo missing outcome data seem to be missing present although the extremely good results obtained in the intervention arm raises serious doubts on about the methods used.
Incomplete outcome data (attrition bias)
Secondary
Unclear riskNo missing outcome data seem to be missing present although the extremely good results obtained in the intervention arm raises serious doubts on about the methods used.
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes.
Other biasHigh riskThe extremely good results obtained in the intervention arm raises serious doubts about the methods used in this trial.

Galassi 1999

Methods

Method of treatment allocation: with a standard list of random numbers

Double-blinded?: no, open-label study

Stratification: no

Placebo: no

Sample size calculation: no

Intention-to-treat analysis: no

Funding: not stated

Follow-up: 30 days

Participants

Location: 1 centre in Italy

Timeframe: from October 1996 to February 1998

Eligibility criteria: 106 patients with CAD, demonstrable ischaemia and a target de novo complex lesion stenosis > 70% in a native vessel scheduled for elective implantation of a > 20 mm stent or multiple stents

Exclusion criteria: acute myocardial infarction; bleeding diathesis, thrombocytopenia, history of stroke, active internal bleeding, severe uncontrolled hypertension, major surgery or trauma within 6 weeks

Male: 88%, mean age: 62 y, diabetes: 27%, previous MI: 67%, ACS: 0%

Balloon angioplasty: 0%, stent: 100%, drug-eluting stents: 0%

Pre-treatment with clopidogrel: 0%

InterventionsPatients were randomly assigned, in an open-label fashion, to receive either a combination of abciximab (bolus and a 12-h infusion) and weight-adjusted low-dose heparin or weight-adjusted heparin alone. All patients received 325 mg of aspirin the day before the procedure and daily thereafter. Ticlopidine 250 mg twice daily was started the day before the intervention and was given to all patients for the first 4 weeks
OutcomesNo primary outcome was specified
Outcomes: mortality, MI, urgent revascularisation, target lesion revascularisation, acute or subacute stent thrombosis
NotesPatients were 'randomly allocated'. There is no description of the allocation concealment or the primary outcome of the study
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskPatients were "randomly allocated. . . with a standard list of random numbers"
Allocation concealment (selection bias)High riskThere is no description of the allocation concealment; probably not used
Blinding (performance bias and detection bias)
Primary
High riskOpen-label study
Blinding (performance bias and detection bias)
Secondary
High riskOpen-label study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasHigh riskNo primary outcome was specified

Gasior 2003

Methods

Method of treatment allocation: not stated

Stratification (yes/no): not stated

Placebo (yes/no): no

Sample size calculation (yes/no): no

Intention-to-treat analyses (yes/no): not stated.

Losses to follow-up (yes/no, and n (%)): yes, 1 patient in the control group

Funding: not stated

Participants

Location: 1 centre in Poland (Zabrze)

Time frame: from October 1998 to May 2000

Inclusion criteria: patients with all of the following characteristics: proximal or medial LAD chronic total occlusion with TIMI coronary blood flow  0 or 1, myocardial viability in LAD territory on a dobutamine stress echocardiography in patients with CCS II or III angina, angina at rest or during exercise, and no contraindication for abciximab

Exclusion criteria: LAD occlusion < 2 weeks or > 6 months (according to the last severe chest pain), contraindication for abciximab, previous revascularisation, failed attempt of recanalisation of the coronary vessel, or no indirect signs of viability in LAD territory

Male: 84%, mean age: 54 y, diabetes: 8%, previous MI: 72%, ACS: 15% (unstable angina 15%, AMI 0%)

Balloon angioplasty: 0%, stent: 100%, drug-eluting stents: 0%, pre-treatment with clopidogrel: 0%

Interventions

Abciximab group: AAS 75 to 150 mg/24 hs and ticlopidine 500 mg/24 hs at least 48 hs before PCI, and 500 mg/24 hs during 6 to 8 weeks. During PCI: abciximab bolus of 0.25 mg/kg + infusion of 0.125 μg/kg/min during 12 h

Control group: AAS 75 to 150 mg/24 h and ticlopidine 500 mg/24 h at least 48 h before PCI, and 500 mg/24 h during 6 to 8 weeks

Outcomes

Primary endpoint: recurrent ischaemia, urgent revascularisation, myocardial infarction, death

Secondary endpoints: coronary restenosis (> 50%), reocclusion, percent luminal diameter and of stenosis, improvement in left ventricular ejection fraction

Safety endpoints: severe bleeding

Notes55 patients were eligible but only 41 patients were randomised because of failure to cross the vessel complete occlusion with the guidewire. 2 patients (1 in each group) were  excluded after randomisation because of failure in stent implantation, resulting in 39 analysable patients, 20 in the abciximab group and 19 in the control group
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot stated
Allocation concealment (selection bias)Unclear riskNot stated
Blinding (performance bias and detection bias)
Primary
High riskNo
Blinding (performance bias and detection bias)
Secondary
High riskNo
Incomplete outcome data (attrition bias)
Primary
Unclear risk2 patients (1 in each group) were excluded after randomisation because of failure in stent implantation
Incomplete outcome data (attrition bias)
Secondary
High risk2 patients (1 in each group) were excluded after randomisation because of failure in stent implantation. The incidence of major and minor bleeding was unusually low (0% and 2.5% respectively)
Selective reporting (reporting bias)Low riskThe published reports include all pre-specified and expected outcomes
Other biasUnclear riskThe study appears to be free of other sources of bias although the incidence of major and minor bleeding was unusually low (0% and 2.5% respectively)

GUSTO-IV 2001

Methods

Method of treatment allocation: via a centralised interactive voice-response system

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Centocor Inc, Malvern, Pa

Follow-up: 30 days

Participants

Location: 458 hospitals in 24 countries: Australia, Austria, Belgium, Canada, Czech Republic, Finland, France, Germany, Greece, Ireland, Israel, Italy, Netherlands, New Zealand, Norway, Poland, Portugal, South Africa, Spain, Sweden, Switzerland, UK, USA

Timeframe: from 17 July 1998 to 21 April 2000

Eligibility criteria: 7800 average-risk patients with unstable angina or non-ST segment elevation myocardial infarction
Last episode of chest pain: < 24 h
Indicator of myocardial ischaemia: > 0.5 mm ST depression or > 0.5 mm transient ST elevation or troponin T or I elevation above ULN

Exclusion criteria: myocardial ischaemia precipitated by a disorder other than atherosclerotic CAD, STEMI, new left-bundle branch block; PCI within previous 14 days; planned PCI or coronary bypass surgery within 30 days after enrolment; active internal bleeding or history of haemorrhagic diathesis; major surgery, serious trauma or gastrointestinal or genitourinary bleeding of clinical significance within the previous 6 weeks; intracranial neoplasm or aneurysm, history of stroke within 2 years, or prior stroke with a residual neurological deficit; oral anticoagulation within the previous 7 days; platelet count of less than 100,000/mL; confirmed hypertension; history of vasculitis; puncture of non-compressible vessel within 24 h before enrolment; allergy to abciximab; weight more than 120 kg; a coexisting disorder associated with limited life expectancy; and participation in another investigational trial within 7 days

Mean age: 65 y, male: 62%, prior myocardial infarction: 31%, diabetes: 22%
Acute coronary syndrome: 100% (unstable angina: 72%, non-STEMI: 28%, STEMI: 0%), ST-segment depression: 80%, CK-MB elevation: 28%, troponin elevation: 59%
PCI: in-hospital: 19%, during drug-infusion: 1.6%, CABG: 11%

Treatment with clopidogrel: 0%

InterventionsAbciximab bolus + 48-h infusion versus abciximab bolus + 24-h infusion versus placebo
Dose:
a) 0.25 mg/kg bolus + 0.125 µg/kg/min infusion (maximum 0.10 µg/min) for 24 hs + heparin
b) 0.25 mg/kg bolus + 0.125 µg/kg/min infusion (maximum 0.10 µg/min) for 48 hs + heparin
c) placebo + heparin
Duration: 24 or 48 hs
Outcomes

Primary: death or myocardial infarction at 30 days
Secondary: death or myocardial infarction in patients with positive troponin levels; death, myocardial infarction, revascularisation or coronary angiography at 48 hs, 7 days and 30 days; death or myocardial infarction within 30 days

Required level of CK or CK-MB elevation in MI definition: 3 x ULN
Safety: intracranial haemorrhage; bleeding leading to decrease in haemoglobin concentration > 50 g/L

NotesGUSTO-IV Study
Both abciximab groups were analysed together in this review
All patients treated with 150 to 325 mg aspirin
Heparin was part of study regimen; initial dose weight-adjusted (maximum 5000 U bolus + 800 U/h infusion aiming for a PTT of 50 to 70 s; a subgroup treated with dalteparin (maximum dose 10,000 U)
Angiography was discouraged during infusion period
PCI was not scheduled (just 1.6% had PCI within 48 h)
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskUsing a computer random number generator
Allocation concealment (selection bias)Low riskVia a centralised interactive voice-response system
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind, placebo-controlled study. A clinical endpoint committee, the members of which were unaware of treatment assignment, adjudicated all possible cases of myocardial infarction and the cause of death
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blind, placebo-controlled study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is available and all of the study’s pre-specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre-specified way
Other biasLow risk

This study was supported by Centocor Inc, Malvern, Pa. "Full independence of the analyses and control over publication remain with the authors, along with the responsibility for any errors"

The study appears to be free of other sources of bias

IMPACT 1995

Methods

Method of treatment allocation: not stated

Double-blinded ?: yes

Stratification: no

Placebo: yes

Sample size calculation: no

Intention-to-treat analysis: yes

Funding: COR Therapeutics, Inc, South San Francisco, USA

Follow-up: 30 days

Participants

Institutions: 11 centres in USA

Timeframe: not stated

Eligibility criteria: 150 patients with coronary artery disease scheduled for elective percutaneous coronary revascularisation

Exclusion criteria: known history of bleeding disorder, recent gastrointestinal bleeding, major surgery within 6 weeks, history of stroke or other central nervous system structural abnormality, severe hypertension, pregnancy, elevation of baseline prothrombin time (> 1.2 times control), haematocrit < 30%, platelet count < 100,000/µL or creatinine > 4.0 mg/dl

Mean age: 62 y, male: 75%, prior myocardial infarction: 45%
Acute coronary syndrome: 75% (unstable angina: 58%, non-STEMI: ?%, STEMI: 17%)
ST-segment depression: ?%, CK-MB elevation: ?%, troponin elevation: ?%

PCI: atherectomy: 13%, balloon angioplasty: 100%, stent: 0%, pre-treatment with clopidogrel: 0%

InterventionsEptifibatide bolus (90 µg/kg) + 12-h infusion (1.0 µg/kg/min for 4 hs) versus eptifibatide bolus + 4-h infusion versus placebo
Outcomes

Primary: a composite of total death, myocardial infarction and urgent coronary revascularisation

Secondary: safety and bleeding complications

NotesA phase II study. The 2 eptifibatide groups were grouped together for the meta-analysis. All patients treated with aspirin and heparin
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNo information provided, but probably yes
Allocation concealment (selection bias)Unclear riskNo information provided
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind, placebo-controlled phase II clinical trial. "All clinical end points were adjudicated by blinded review"
Blinding (performance bias and detection bias)
Secondary
Unclear riskUncertain since the proportion of major bleeding was 4 times higher in patients in the placebo group than in the intervention group and inversely correlated with minor bleeding
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published report includes all expected outcomes, including those that were pre-specified
Other biasUnclear riskFunding: COR Therapeutics, Inc, South San Francisco, USA

IMPACT-II 1997

Methods

Method of treatment allocation: generated by computer in blocks of 9. Assignment was done at the same time as registration and enrolment by telephone call to the Duke Coordinating Center

Double-blinded?: yes

Stratification: yes, patients were stratified according to the perceived risk of ischaemic complications into high-risk and low-risk categories

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: COR Therapeutics, Inc (South San Francisco, California) and Schering-Plough, Inc (Kenilworth, New Jersey)

Follow-up: 30 days

Participants

Institutions: 82 centres in the USA

Timeframe: from 30 November 1993 to9 November 1994

Eligibility criteria: 4010 patients with coronary artery disease undergoing percutaneous revascularisation

Exclusion criteria: history of bleeding diathesis, severe hypertension, major surgery within the previous 6 weeks, history of stroke or other disorders of the central nervous system, pregnancy, gastrointestinal or genitourinary bleeding within the previous 30 days, or other major illness

Mean age: 61 y, male: 75%, prior myocardial infarction: 41%
Acute coronary syndrome: 42% (unstable angina: 38%, non-STEMI: ? %, STEMI: 4%)
ST-segment depression: ?%, CK-MB elevation: ?%, troponin elevation: ?%

PCI: atherectomy: 23%, balloon angioplasty: 92%, stent: 4%, pre-treatment with clopidogrel: 0%

InterventionsBolus of 135 μg/kg eptifibatide followed by an infusion of 0.5 mg/kg/min for 20 to 24 h versus a 135 mg/kg eptifibatide bolus followed by an infusion of 0.75 mg/kg/min for 20 to 24 h versus placebo bolus plus placebo infusion
Outcomes

Primary endpoint: the occurrence within 30 days of death, myocardial infarction, urgent or emergency repeat coronary revascularisation, or index placement of an intracoronary stent for abrupt closure

Secondary endpoint: the occurrence of the composite endpoint at the completion of drug infusion (24 h) and at 6 months, the composite endpoint as determined by the site principal investigators (rather than the Clinical Events Committee), outcomes by risk stratification and actual treatment received, and the frequency of angiographically documented abrupt closure

Notes2 arms with different doses of eptifibatide and a placebo arm. The 2 active treatment arms were grouped in our analysis. All patients received aspirin and heparin
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNo information provided, but probably yes
Allocation concealment (selection bias)Low riskNo information provided.
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind, placebo-controlled phase II clinical trial. "All clinical end points were adjudicated by blinded review"
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blind, placebo-controlled trial. Study drugs were packaged to be indistinguishable irrespective of content. All efficacy and safety events were adjudicated by consensus of the Clinical Events Committee, from which treatment assignment was concealed throughout the trial
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published report includes all expected outcomes, including those that were pre-specified.
Other biasUnclear riskFunding: COR Therapeutics, Inc, South San Francisco, USA.

INSTANT 2012

Methods

Method of treatment allocation: an online randomisation system

Double-blinded?: no

Placebo: no

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: GlaxoSmithKline, Verona, Italy

Follow-up: 6 months

Participants

Location: 1 centre in Verona, Italy

Timeframe: not stated

Eligibility criteria: 91 patients with stable coronary disease and diffuse disease involving a major epicardial coronary vessel undergoing PCI on a native coronary vessel with planned implantation of N33 mm of DES with a reference vessel diameter 2.25 to 4.0 mm

Exclusion criteria: 1) Clinical: women with childbearing potential, aged less than 18 years, ongoing or recent episode (< 48 h) of ACS, administration of any GP IIb/IIIa inhibitors during the previous 2 weeks, serum creatinine greater than 2.5 mg/dl, ongoing serious bleeding or bleeding diathesis, previous stroke in the last 6 months, major surgery within the previous 6 weeks, platelet count higher than 100,000 per mm3, ejection fraction below 30%, known hypersensitivity or contraindication to aspirin, heparin, clopidogrel, or sensitivity to contrast that cannot be adequately premedicated, haemodynamic instability requiring balloon counterpulsation or inotropic support, positive clinical history for intracranial neoplasia, arteriovenous malformation, aneurysm, international normalised ratio (INR) at least 2.0 or prothrombin time 1.2 times upper limit of normality, clinically manifested reduced liver function and programmed surgery within 1 month

2) Angiographical: DES implantation in a chronic total occlusion, vessel size less than 2.25 mm or greater than 5 mm, and previous implantation of a BMS or a DES in the target lesion

Mean age: 66 y, male: 81%, prior myocardial infarction: 30%
Acute coronary syndrome: 40%
ST-segment depression: ?, CK-MB elevation: ?%, troponin elevation: ?%
PCI: balloon PCI: 0%, stent: 100%, drug-eluting stent: 100%, pre-treatment with clopidogrel: 100%

InterventionsEnrolled patients were randomised to intravenous normal saline or intravenous eptifibatide (double bolus (180 μg/kg) followed by infusion (2 μg/kg per minute) for 18 to 24 h after the procedure). Concomitantly to study drug administration, an intravenous bolus of unfractionated heparin (60 IU/kg) was administered, and during the procedure, patients received intravenous boluses of heparin in sufficient doses to prolong the activated clotting time (≥ 250 seconds)
Outcomes

Primary: the rate of elevated postprocedural peak CK-MB mass ratio values

Secondary: in-hospital 1- and 6-month major adverse cardiovascular events (MACEs), defined as the composite of cardiac death, nonfatal MI, or urgent TVR

NotesOwing to difficulties in enrolling patients and funding issues due to the inability to meet enrolment milestones originally agreed upon with the sponsor, the study was prematurely stopped before any statistical analysis. Only 91 of the 320 planned patients were enrolled in the study
Pre-treatment with 600 mg of clopidogrel recommended in all patients
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear risk"An online randomisation system". Comment: the number of implanted stents per patient was much lower in the intervention group than in the control group, and the stent type was also different
Allocation concealment (selection bias)High riskThe method of concealment is not described. Comment: the number of implanted stents per patient was much lower in the intervention group than in the control group, and the stent type was also different
Blinding (performance bias and detection bias)
Primary
High riskSingle-blinded study, without placebo
Blinding (performance bias and detection bias)
Secondary
High riskSingle-blinded study, without placebo
Incomplete outcome data (attrition bias)
Primary
Unclear riskOnly 91 of the 320 planned patients were enrolled in the study; 5 patients were lost to follow-up
Incomplete outcome data (attrition bias)
Secondary
Unclear riskOnly 91 of the 320 planned patients were enrolled in the study; 5 patients were lost to follow-up
Selective reporting (reporting bias)High riskProbably correct for the primary endpoint. The incidence of bleeding was very low: 0% for major bleeding and just 2% for any bleeding
Other biasUnclear riskFunding: GlaxoSmithKline, Verona, Italy

ISAR-2 2000

Methods

Method of treatment allocation: sealed envelopes

Double-blinded?: no, patients but not physicians were blinded to the assignment of treatment

Stratification: no

Placebo: no

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Lilly, Deutschland

Follow-up: 30 days

Participants

Location: 1 hospital in Germany

Timeframe: not stated

Eligibility criteria: 401 patients with ST-segment elevation acute myocardial infarction undergoing angioplasty with stent implantation within 48 h after onset of chest pain

Exclusion criteria were inability to give informed consent and contraindications to 1 of the study drugs. All eligible patients who gave written, informed consent were randomised by means of sealed envelopes. Patients, but not physicians, were blinded to the assignment of treatment

Mean age: 61 y, male: 76%, prior myocardial infarction: ?%
Acute coronary syndrome: 100% (unstable angina: 0%, non-STEMI: 0%, STEMI: 100%)
ST-segment depression: 0%, CK-MB elevation: 100%, troponin elevation: ?%

PCI: balloon angioplasty: 0%, stent: 100%, pre-treatment with clopidogrel: 0%, but 100% with ticlopidine

InterventionsPatients were randomised to 1 of 2 treatment regimens: 1) a bolus of abciximab 0.25 mg/kg of body weight, followed by continuous infusion, 10 µg/min for 12 h plus an additional dose of heparin (2500 U intra-arterially) or heparin (10,000 U intra-arterially), followed by IV heparin infusion (1000 U/h), for the first 12 h after sheath removal
OutcomesPrimary: angiographic restenosis at 6 months
Secondary: clinical restenosis and a composite of death, myocardial infarction and target lesion revascularisation at 30 days
NotesNot placebo-controlled. All patients treated with aspirin, heparin and ticlopidine
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNot stated, but probably correct since all the trials from this group have a very good design
Allocation concealment (selection bias)Low riskSealed envelopes
Blinding (performance bias and detection bias)
Primary
High riskPatients, but not physicians, were blinded to the assignment of treatment
Blinding (performance bias and detection bias)
Secondary
High riskPatients, but not physicians, were blinded to the assignment of treatment
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published report includes all expected outcomes, including those that were pre-specified
Other biasUnclear riskFunding: Lilly, Deutschland

ISAR-REACT 2 2006

Methods

Method of treatment allocation: using sealed envelopes containing the block randomisation sequence for each participating centre

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: supported in part by the grant KKF 04-03 from Deutsches Herzzentrum, Munich, Germany

Follow-up: 30 days

Participants

Location: 5 centres in Germany, 1 in Netherlands and 1 in Brazil

Timeframe: from March 2003 through December 2005

Eligibility criteria: an episode of angina (with an accelerating pattern or prolonged or recurrent episodes at rest or with minimal effort) within the preceding 48 h, accompanied by an elevated troponin T level or a new finding of ST-segment depression of at least 0.1 mV or transient  ST-segment elevation of at least 0.1 mV or new or presumed new bundle-branch block; significant angiographic lesions in a native coronary vessel or venous bypass graft amenable to and requiring a PCI; and written informed consent from the patient

Exclusion criteria: ST-segment elevation acute MI; haemodynamic instability; pericarditis; malignancies with life expectancy less than 1 year; increased risk of bleeding (stroke within the previous 3 months, active bleeding or bleeding diathesis, recent trauma or major surgery in the last month, suspected aortic dissection); oral anticoagulation with a coumarin derivative within the previous 7 days; receipt of a GP IIb/IIIa inhibitor within the previous 14 days; uncontrolled hypertension; a haemoglobin level less than 100 g/L or haematocrit less than 34%, or platelet count less than 100,000 cells/?L or greater than 600,000 cells/?L; known allergy to the study medication; and pregnancy (present or suspected)

2022 patients with NSTEACS that underwent PCI in native coronary vessels within 6 h from diagnosis of ACS and after pre-treatment with 600 mg of clopidogrel > 2 h before PCI. Coronary stenting was the target PCI

Mean age: 66 y, male: 74%, diabetes 27%, prior myocardial infarction: 24%

ACS: 100% (unstable angina: 48%, NSTEMI: 52%, STEMI: 0%)

PCI: balloon angioplasty: 3%, stent: 97% (BMS: 48%, DES: 49%), pre-treatment with clopidogrel: 100%

InterventionsAbciximab (bolus of 0.25 mg/kg, followed by an infusion of 0.125 µg/kg/min) versus placebo. All patients treated with ASA, heparin and 600 mg of clopidogrel > 2 h before the procedure
OutcomesPrimary: all-cause death, MI or urgent target-vessel revascularisation within 30 days of randomisation
Safety: major and minor bleeding, and thrombocytopenia
NotesHigh-risk patients with NSTEACS treated with early (< 6 h) PCI after diagnosis
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated
Allocation concealment (selection bias)Low riskUsing sealed envelopes containing the block randomisation sequence for each participating centre
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blinded, placebo-controlled trial. Double-blinding was achieved by using vials of similar appearance in the
2 groups
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blinded, placebo-controlled trial. Double-blinding was achieved by using vials of similar appearance in the
2 groups
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is available and all of the study’s pre-specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre-specified way
Other biasLow riskThe study appears to be free of other sources of bias

ISAR-REACT 2004

Methods

Method of treatment allocation: patients underwent randomisation in a double-blind manner with the use of sealed envelopes containing the block randomisation sequence for each participating centre

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: supported by research grants from Deutsches Herzzentrum, Klinik an der Technischen Universität, Munich, Germany (67-00 and 04-01), and by an unrestricted educational grant from Bristol-Myers Squibb GmbH, Munich, Germany

Follow-up: 30 days and 1 y

Participants

Location: 4 institutions in Germany, 1 in The Netherlands and 1 in the USA

Timeframe: from May 2000 and February 2003

Eligibility criteria: 2159 patients with CAD that underwent elective PCI in native coronary vessels and had been pre-treated with 600 mg of clopidogrel > 2 h before the intervention. Coronary stenting was the target PCI

Exclusion criteria: recent myocardial infarction or unstable angina, a target lesion in a venous bypass graft; a chronic occlusion; a target lesion with angiographically visible thrombus; a left ventricular ejection fraction < 30%; haemodynamic instability,  insulin-dependent diabetes mellitus, pericarditis or cancer; stroke in the prior 3 months; active bleeding or bleeding diathesis; trauma or major surgery in the preceding month; suspected aortic dissection; oral anticoagulation therapy or glycoprotein IIb/IIIa inhibitor within the preceding 14 days; severe, uncontrolled hypertension; haemoglobin < 10 g/dl or a haematocrit < 34%; a platelet count < 100,000 or > 600,000; a known allergic reaction to the study medication; or child-bearing potential

Mean age: 66 y, male: 76%, prior myocardial infarction: 33%
Acute coronary syndrome: 0% (unstable angina: 0%, NSTEMI: 0%, STEMI: 0%)
ST-segment depression: ?, CK-MB elevation: 0%, troponin elevation: 0%
PCI: balloon angioplasty: 10%, stent: 90%, pre-treatment with clopidogrel: 100%

InterventionsAbciximab (bolus of 0.25 mg/kg, followed by an infusion of 0.125 µg/kg/min (maximum, 10 µg per minute) for 12 hours versus placebo. All patients treated with ASA and heparin
OutcomesPrimary: all-cause death, MI or urgent target-vessel revascularisation within 30 days of randomisation
Secondary: major and minor bleeding, and thrombocytopenia
NotesLow-risk patients scheduled for elective PCI with stent placement
All patients pre-treated with high-dose clopidogrel
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated
Allocation concealment (selection bias)Low riskPatients underwent randomisation in a double-blind manner with the use of sealed envelopes containing the block randomisation sequence for each participating centre
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blinded, placebo-controlled trial. Double-blinding was achieved by using vials of similar appearance in the
2 groups
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blinded, placebo-controlled trial. Double-blinding was achieved by using vials of similar appearance in the
2 groups. All events were adjudicated and classified by an event-adjudication committee whose members were unaware of the patients’ assigned treatment
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is available and all of the study’s pre-specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre-specified way
Other biasUnclear riskThe study appears to be free of other sources of bias

ISAR-SMART-2 2004

Methods

Method of treatment allocation: sealed envelopes containing the randomisation sequence generated by computer before initiation of the trial

Double-blinded?: yes

Stratification: no

Placebo: no

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: supported in part by an unrestricted research grant from Krauth medical KG, Hamburg, Germany

Follow-up: 1 year

Participants

Location: 3 centres in Germany

Timeframe: not stated

Eligibility criteria: 502 patients with stable angina pectoris or a positive exercise test that underwent elective PCI in native coronary vessels < 2.5 mm in size and after pre-treatment with 600 mg of clopidogrel > 2 h before PCI

Exclusion criteria: acute coronary syndrome, left main coronary artery, in-stent restenosis and contraindications to the antithrombotic medication used in the study

Mean age: 66 y, male: 73%, prior myocardial infarction: 37%
Acute coronary syndrome: 0%
ST-segment depression: ?, CK-MB elevation: 0%, troponin elevation: 0%
PCI: balloon angioplasty: 50%, stent: 50%, pre-treatment with clopidogrel: 100%

Interventions

Patients were randomly assigned to be treated with either PC-coated stents (n = 253) or PTCA (n = 249) and with either abciximab (n = 251) or placebo (n = 251) with the use of a 2 x 2 factorial design. Patients randomised to abciximab received a bolus of 0.25 mg/kg, followed by an infusion of 0.125 µg/kg/min (maximum, 10 µg per minute) for 12 hours

All patients treated with ASA, heparin and 600 mg of clopidogrel > 2 h before the procedure

OutcomesPrimary: angiographic restenosis at follow-up angiography
Secondary: combined incidence of all-cause death and MI as well as target vessel revascularisation during 1-year follow-up
NotesLow-risk patients scheduled for elective PCI with stent placement
All patients pre-treated with high-dose clopidogrel > 2 h before the procedure
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated
Allocation concealment (selection bias)Low riskSealed envelopes containing the randomisation sequence generated by computer before initiation of the trial
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind, placebo-controlled study
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blind, placebo-controlled study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published reports include all expected outcomes, including those that were pre-specified
Other biasLow riskThe study appears to be free of other sources of bias

ISAR-SWEET 2004

Methods

Method of treatment allocation: with the use of sealed envelopes containing the randomisation sequence for each participating centre

Double-blinded ?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: research grants by Deutsches Herzzentrum, Klinik an der Technischen Universität, Munich, Germany (H04-01)

Follow-up: 30 days and 1 y

Participants

Location: 3 German hospitals

Timeframe: between January 2001 and October 2003

Eligibility criteria: 701 diabetic patients with CAD scheduled for elective PCI in native coronary vessels and had been pre-treated with 600 mg clopidogrel at least 2 h before intervention. Coronary stenting was the target PCI

Exclusion criteria: myocardial infarction within the prior 14 days; ACS; target lesion with thrombus or  in a venous bypass graft; chronic coronary occlusion; a left ventricular ejection fraction < 30%, haemodynamic instability, pericarditis, malignancy, a stroke in the prior 3 months, active bleeding or bleeding diathesis, recent trauma or major surgery in the last month, a suspected aortic dissection, oral anticoagulation therapy, severe uncontrolled hypertension, haemoglobin < 100 g/L or haematocrit < 34%, thrombocytopenia, known allergic reaction to the study medication, had received a glycoprotein IIb/IIIa inhibitor within 14 days, or were pregnant (present or suspected)

Mean age: 67 y, male: 74%, prior myocardial infarction: 34%
Acute coronary syndrome: 0%
ST-segment depression: ?, CK-MB elevation: 0%, troponin elevation: 0%
PCI: balloon angioplasty: 10%, stent: 90%, pre-treatment with clopidogrel: 100%

InterventionsAbciximab (bolus of 0.25 mg/kg, followed by an infusion for 12 h of 0.125 µg/kg/min) versus placebo. All patients treated with ASA and heparin
Outcomes

Primary endpoint: the cumulative incidence of death from any cause and MI during the first 12 months after randomisation

Secondary endpoints: incidence of binary angiographic restenosis, and of target lesion revascularisation due to angiographic restenosis and symptoms or signs of ischaemia
Safety: major and minor bleeding, and thrombocytopenia

Notes

Patients with diabetes mellitus (29% treated with insulin) scheduled for elective PCI with stent

Patients of both study groups received clopidogrel 600 mg at least 2 h before the percutaneous coronary intervention

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated
Allocation concealment (selection bias)Low riskWith the use of sealed envelopes containing the randomisation sequence for each participating centre
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blinding was achieved with the use of vials that appeared similar in the 2 groups. All events were adjudicated and classified by an event adjudication committee blinded to the assigned treatment. All analyses were performed in a blinded manner regarding the randomly assigned treatment. Unblinding of the study groups was done after completion of the statistical analyses. No patient required unblinding because of clinical needs, and no cross-overs occurred
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blinding was achieved with the use of vials that appeared similar in the 2 groups. All events were adjudicated and classified by an event adjudication committee blinded to the assigned treatment. All analyses were performed in a blinded manner regarding the randomly assigned treatment. Unblinding of the study groups was done after completion of the statistical analyses. No patient required unblinding because of clinical needs, and no cross-overs occurred
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published reports include all expected outcomes, including those that were pre-specified
Other biasLow riskThe study appears to be free of other sources of bias

ITTI 2012

Methods

Method of treatment allocation: not stated

Stratification: yes, by the performance of thrombosuction procedure

Placebo: no

Sample size calculation: yes

Intention-to-treat analyses (yes/no): yes

Losses to follow-up: no

Funding: not stated

Follow-up: 6 months

Participants

Location: 5 centres in Taiwan

Time frame: not stated

Eligibility criteria: age ≧ 18 years, continuous chest pain ≧ 30 min, ST-segment elevation > 0.1 mV in ≧ 2 contiguous leads on a 12-lead electrocardiogram (ECG)

Exclusion criteria: cardiogenic shock (systolic BP < 80 mm Hg or need for inotropic agent), history of bleeding tendency, major operation within 6 weeks, hepatic or renal insufficiency, and contraindication to tirofiban use

Total recruited: a total of 100 STEMI patients undergoing primary PCI within 12 h of symptom onset were randomly allocated to 1 of the 4 intervention groups: standard PCI, initial thrombus aspiration (IT), tirofiban infusion (TI) and PCI with both treatments (IT + TI).

Mean age: 59 y, male: 74%, diabetes 26%, prior myocardial infarction: ?%
Acute coronary syndrome: 100% (STEMI 100%)
PCI: balloon angioplasty: 0%, stent: 100%, pre-treatment with clopidogrel: 100%

Interventions

Active group: tirofiban was administered with 10 μg/kg over 3 min as a bolus at cath lab before PCI procedure, followed by 0.15 μg/kg/min infusion for 24 h

Control group: all patients received aspirin (300 mg loading followed by 100 mg daily) and clopidogrel (300 mg loading followed by 75 mg daily) and unfractionated heparin 100 IU/kg before the procedure

Outcomes

Primary endpoint: occurrence of MBG 3 myocardial reperfusion

Secondary endpoints: complete ST-segment resolution, procedure time, occurrence of no-reflow, peak CK-MB level and time to peak, TIMI flow and corrected TIMI frame count, 6 months major adverse events including death, reinfarction, target lesion revascularisation and stroke

Safety endpoints: severe bleeding

NotesStudy performed in Taiwan. All patients received 300 mg of clopidogrel before the procedure
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot stated
Allocation concealment (selection bias)Unclear riskNot specified
Blinding (performance bias and detection bias)
Primary
High riskOpen study without placebo
Blinding (performance bias and detection bias)
Secondary
High riskOpen study without placebo
Incomplete outcome data (attrition bias)
Primary
Low riskAll data are reported
Incomplete outcome data (attrition bias)
Secondary
Unclear riskAll data are reported, but 2 patients in the tirofiban group crossed over to the tirofiban plus thrombectomy group
Selective reporting (reporting bias)Low riskAll data are reported
Other biasUnclear riskNone

JEPPORT 2009

Methods

Method of treatment allocation: not stated

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: not stated

Participants

Location: 88 centres in Japan

Timeframe: from May 1997 to April 2000

Follow-up: 30 days and 6 months

Eligibility criteria: 973 patients with ACS

Exclusion criteria: age > 75 y; > 100 kg of body weight; scheduled for primary stent placement, directional coronary atherectomy or rotablator; history of thrombocytopenia; bleeding symptoms or bleeding diathesis; undergone surgery in the previous 6 weeks; cerebrovascular disorder in the previous 2 y; ?50% stenosis in the left main trunk of the coronary artery; 3-vessel disease; uncontrolled hypertension or pulmonary hypertension; cardiogenic shock requiring cardiopulmonary resuscitation

Mean age: 61 y, 81% male, 31% diabetes, ?% prior MI

ACS: 100% (non-STEACS: 23%, STEMI: 77%)

PCI: 100% (balloon angioplasty: 75%, stent: 25%, drug-eluting stents: 0%), pre-treatment with clopidogrel: 0%

Interventions

Low-dose abciximab (0.20 mg/kg bolus + 0.125 µg/kg/min infusion for 12 h) versus high-dose abciximab (0.25 mg/kg bolus + 0.125 µg/kg/min infusion for 12 h) versus placebo (bolus and infusion)

Thrombolytic drugs, antiplatelet drugs other than aspirin, anticoagulant drugs other than heparin, PGE1 and its derivatives, dextran and low-molecular-weight dextran, were prohibited during the 6 months from the start of the investigation

Outcomes

Primary: 30-day occurrence of death, MI and/or urgent revascularisation for recurrence of ischaemia

Secondary: 6-month incidence of major coronary events

Notes

Aspirin administered before the procedure. Clopidogrel and ticlopidine were not allowed before and during the 6 month follow-up.

After the first 223 patients were enrolled in the study, “overseas authorities” stipulated that the approved intravenous drip infusion dose of abciximab be adjusted according to body weight. The study was resumed 1 year later with the new standard dose

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNot stated but probably yes because the baseline data were well balanced in the 3 groups
Allocation concealment (selection bias)Low riskNot stated but probably yes because the baseline data were well balanced in the 3 groups
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind, placebo-controlled trial
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blind, placebo-controlled trial
Incomplete outcome data (attrition bias)
Primary
Low risk3.6% of the patients excluded because of exclusion criteria or because they did not underwent PCI. However, excluded patients were well balanced in the placebo and intervention groups
Incomplete outcome data (attrition bias)
Secondary
Low risk3.6% of the patients excluded because of exclusion criteria or because they did not underwent PCI. However, excluded patients were well balanced in the placebo and intervention groups
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published reports include all expected outcomes, including those that were pre-specified
Other biasHigh riskThe study was stopped for 1 year and resumed with the dose of abciximab adjusted according to body weight. In total, the study took 8 years to be completed

Juergens 2002

Methods

Method of treatment allocation: the patient received the next consecutive ascending number allocated to the investigator

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Merck & Co, Inc

Follow-up: 30 days

Participants

Location: 59 centres in 24 countries (Argentina, Australia, Austria, Brazil, China, Colombia, Costa Rica, Ecuador, Greece, Lebanon, Malaysia, Mexico, New Zealand, Poland, Portugal, Singapore, Slovenia, South Africa, Spain, Switzerland, Taiwan, Turkey, United Kingdom and Venezuela)

Timeframe: from May 1998 to June 1999

Eligibility criteria: 894 patients scheduled to undergo PTCA with intracoronary stent placement

Exclusion criteria: thrombolytic therapy within 24 h of AMI, allergy to or unable to tolerate aspirin or heparin, prior treatment with abciximab within 14 days, ticlopidine, clopidogrel or low-molecular-weight heparin within 12 to 24 h, PTCA within 14 days or planned repeat PTCA as a staged procedure; unprotected left main stenosis; bleeding disorder within 3 months; persistent hypertension; history of stroke or other intracranial pathology within 1 year; recent major surgery, trauma, or cardiopulmonary resuscitation; active peptic ulcer disease, pericarditis, significant retinopathy, suspected aortic dissection, uncontrolled cardiac arrhythmia, other haemodynamically significant cardiac disease, or other clinically important medical illness that would make survival for the duration of the study unlikely; serum creatinine level > 2.5 mg/dL, haemoglobin < 11 g/dL, international normalised ratio > 1.5, or a platelet count < 150,000/mm3; or unable to give informed consent

Mean age: 59 y, male: 83%, prior myocardial infarction: 46%
Acute coronary syndrome: 46% (unstable angina: 46%, NSTEMI: ?, STEMI: 0%)
ST-segment depression: ?, CK-MB elevation: 0%, troponin elevation: 0%

PCI: balloon angioplasty: 2%, stent: 98%, pre-treatment with clopidogrel: 0%

InterventionsPatients were randomised in a 3:2 ratio to receive tirofiban as an intravenous bolus (10 µg/kg over 3 minutes) and maintenance infusion (0.10 µg/kg per minute for 36 h) or a bolus and infusion of placebo
Outcomes

Primary endpoint: proportion of patients with bleeding

Secondary endpoints: death, MI, urgent coronary artery bypass grafting for recurrent ischaemia and urgent repeat percutaneous intervention for recurrent ischaemia in the target vessel

NotesThis was primarily a tolerability study. 3 employees of Merck & Co, Inc, assisted in the preparation of the manuscript
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskThe patient received the next consecutive ascending number allocated to the investigator
Allocation concealment (selection bias)Low riskPatients and investigators were blinded to treatment assignment through the use of identical-appearing active treatment and placebo
Blinding (performance bias and detection bias)
Primary
Low riskPatients and investigators were blinded to treatment assignment through the use of identical-appearing active treatment and placebo. Cardiac events were reviewed and adjudicated by an external Event Classification Committee
Blinding (performance bias and detection bias)
Secondary
Unclear riskThere was no central adjudication of bleeding incidents despite bleeding being the primary endpoint. In addition, the number of major bleedings was 3 times lower in the intervention group than in the placebo group and was inversely correlated with minor bleedings
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published report includes all expected outcomes, including those that were pre-specified
Other biasHigh risk

4 employees of Merck & Co assisted the authors in the preparation of the manuscript.

Funding: Merck & Co, Inc.

Kereiakes 1996

Methods

Method of treatment allocation: 3 dose regimens of tirofiban were studied in 3 sequential panels. Patients within each panel were randomised to receive either tirofiban or placebo in a 3:1 randomisation design

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: not stated

Follow-up: hospitalisation

Participants

Institutions: 9 centres in USA

Timeframe: not stated

Eligibility criteria: men and women > 18 and < 75 years of age who were scheduled to undergo coronary angioplasty for treatment of 1) rest angina pectoris, 2) recurrent angina, or 3) complex coronary lesion morphology associated with a moderate to high risk of procedural failure; 93 patients were enrolled

Exclusion criteria: women of childbearing potential, thrombolytic therapy within 24 h of angioplasty, severed diffuse multivessel coronary atherosclerosis, uncontrolled cardiac arrhythmia, increased bleeding risk, history of stroke or other intracranial pathology, severe congestive heart failure or haemodynamic instability and allergy or intolerance to aspirin or heparin

Mean age: 59 y, male: 82%, prior myocardial infarction: 47%
Acute coronary syndrome: 52% (unstable angina: 39%, non-STEMI: 0%, STEMI: 13%)
ST-segment depression: ?%, CK-MB elevation: ?%, troponin elevation: ?%
PCI: atherectomy: 0%, balloon angioplasty: 100%, stent: 0%, pre-treatment with clopidogrel: 0%

InterventionsPatients received 1 of 3 graduated regimens of tirofiban intravenously with a bolus dose of 5, 10 and 10 µg/kg and continuous infusion doses of 0.05, 0.10 and 0.15 µg/kg per min, respectively
OutcomesPrimary composite: death, myocardial infarction and need for urgent revascularisation
NotesDose-ranging study
All tirofiban groups were grouped together for the analysis
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNot stated but probably correct
Allocation concealment (selection bias)Low riskNot stated but probably correct
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind, placebo-controlled dose-ranging study
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blind, placebo-controlled dose-ranging study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published report includes all expected outcomes, including those that were pre-specified
Other biasUnclear riskThe study appears to be free of other sources of bias

Kim 2005

Methods

Method of treatment allocation: not stated.

Double-blinded?: not stated.

Placebo: not stated.

Sample size calculation: not stated.

Intention-to-treat analysis: no

Funding: not stated.

Follow-up: 6 months.

Participants

Institutions: 1 One clinical centre from Korea (Chonnam National University Hospital).

Timeframe: March 2001 to December 2002.

Eligibility criteria: One hundred and sixty160 patients who had ST-segment depression more than 1 mm in 2 or more consecutive leads, or T wave inversions, or ST-segment elevation less than 20 min on ECG, and elevated troponin concentration.

Exclusion criteria: ST elevation greater than 20 min, thrombolytic therapy within 20 min, percutaneous coronary intervention within the past 6 months or previous coronary artery bypass graft, cerebrovascular accident within the past year, active bleeding or a high risk for bleeding, secondary angina, contra-indication for antithrombotic or anti-platelet therapy, previous history of thrombocytopenia, elevated creatinine (≥ 2.5 mg/dl) and thrombocytopenia (< 150,000/mm3).

Mean age: 61 ys, 65% males, 23% diabetics, ¿?% with prior MI.

Acute coronary syndrome: 100% (Uunstable angina: 50%, Nnon-ST elevation myocardial infarction: 50%, STEMI: 0%)

ST-segment depression: 14%, Ttroponin elevation: 100%.

Coronary angiography: 100%, PCI:¬71% (balloon angioplasty: ?%, stent: ?%, drug-eluting stent: ?%).

Interventions

Aspirin was administered as an oral loading dose of 300 mg followed by 100 mg daily for all patients. The patients were randomised into 4 groups; group I (n = 40): standard heparin alone; group II (n = 40): LMW heparin (dalteparin) alone; group III (n = 40): standard heparin + tirofiban (n = 40); group IV (n = 40): dalteparin + tirofiban

Tirofiban was administered as a bolus injection dose of 0.4 µg/kg//min for 30 min and then a maintenance dose of 0.1 µg/kg/min.

Unfractionated heparin was administered as a bolus injection of 5000 units and the 12 U/kg//h, keeping the activated partial thromboplastin time at 1.5 to 2.0-fold the normal control value. Dalteparin was injected at 120 U/kg twice daily. All therapeutic agents were administered between 48 and 96 h

Outcomes

Primary: a composite of cardiac death, MI or revasculariszation at 7, 30 and 180 days.

Secondary: Mmajor bleeding.

Notes

Patients underwent 1 month of combined therapy with aspirin and clopidogrel after stenting.

Study performed in Korea.

Unfractionated heparin or dDalteparin used as anticoagulant agent.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot described. "The patients were randomised into one of 4 groups and analysed prospectively". More patients in the intervention group than in the control group had a NSTEMI at baseline (61% versus 39%, P = 0.04)
Allocation concealment (selection bias)Unclear riskNot described. More patients in the intervention group than in the control group had a NSTEMI at baseline (61% versus 39%, P = 0.04)
Blinding (performance bias and detection bias)
Primary
Unclear riskNot described, probably the study was single-blinded
Blinding (performance bias and detection bias)
Secondary
Unclear riskNot described, probably the study was single-blinded
Incomplete outcome data (attrition bias)
Primary
High riskThe outcome data are reported only for the patients that underwent in-hospital revascularisation (75% of the patients)
Incomplete outcome data (attrition bias)
Secondary
High riskThe outcome data are reported only for the patients that underwent in-hospital revascularisation (75% of the patients)
Selective reporting (reporting bias)High riskThe study protocol is not available. The outcome data are reported only for the patients that underwent in-hospital revascularisation (75% of the patients)
Other biasLow riskThe study appears to be free of other sources of bias

Kurowski 2005

Methods

Method of treatment allocation: not stated

Double-blinded?: no

Stratification: no

Placebo: no

Sample size calculation: no

Intention-to-treat analysis: not stated

Losses to follow-up: no

Funding: a restricted educational grant from MSD GMBH, Germany

Follow-up: 30 days and a mean of 2 y

Participants

Location: 1 centre in Lubeck, Germany

Timeframe: June 2000 to February 2003

Eligibility criteria: 50 patients with stable angina or cardiac troponin T negative unstable angina. The culprit lesion was a critical saphenous vein graft stenosis determined by being the only significant lesion explaining the anginal symptoms or by appropriate objective evidence of ischaemia. All lesions had > 90% diameter stenosis and the SVGs were > 3 mm in diameter

Exclusion criteria: not stated

Mean age: 68 y, male: 84%, diabetes: 20%, prior myocardial infarction: 72%
Acute coronary syndrome: 68% (unstable angina); none had acute MI
PCI: balloon PCI: 0%, stent: 100%, pre-treatment with clopidogrel: 100%

InterventionsPatients randomised to tirofiban therapy received a bolus of 10 µg/kg body weight 4 to 6 h before PCI and a maintenance infusion of 0.15 µg/kg body weight per minute or 12 to 14 h. Immediately after inclusion, all patients received a clopidogrel loading dose of 300 mg with an oral maintenance dose of 75 mg/day for 4 weeks, an intravenous aspirin dose of 500 mg with an oral maintenance dose of 100 mg/day indefinitely, and a continuous heparin infusion to maintain the activated partial thromboplastin time at 1.5 to 2 times the upper normal range. The index procedure was performed on the following day
OutcomesPrimary: occurrence of myocardial necrosis as evidenced by an increase in the cTnT above the limit of detection (0.1 ng/ml) estimated quantitatively every 12 h for 72 h after the procedure
Secondary: the major adverse cardiac event rate (death, Q-wave myocardial infarction, or emergency revascularisation) at hospital discharge
NotesPatients with stenosis in saphenous vein grafts referred for PCI
All patients pre-treated with 300 mg of clopidogrel 24 h before the procedure
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo information provided, but probably done
Allocation concealment (selection bias)Unclear riskNo information provided
Blinding (performance bias and detection bias)
Primary
High riskOpen-label study. Mortality and myocardial infarction adjudicated by the investigators
Blinding (performance bias and detection bias)
Secondary
High riskOpen-label study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe published reports include all the prespecified outcomes
Other biasLow riskThe study appears to be free of other sources of bias

On-TIME 2 2008

Methods

Method of treatment allocation: Bby blinded sealed kits with study drug. All staff and study personnel were blinded to treatment. Kits were distributed among the ambulance services or referring centres in blocks of 4four.

Double-blinded? yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Merck (USA).

Participants

ocation: 24 centres in The Netherlands, Germany, and Belgium.

Timeframe: from 29 June 29,2006 to 13 November 13, 2007.

Follow-up: 30 days.

Eligibility criteria: 984 patients aged 21- to 85 y with acute STEMI presenting < 24 h after the onset of symptoms whio were candidates to undergo primary PCI.

Exclusion criteria: known severe renal dysfunction (glomerular filtration rate < 30 mL/min or serum creatinine > 2·.5 mg/dL), therapy- resistant cardiogenic shock (systolic blood pressure ?80 mm Hg for > 30 min), persistent severe hypertension, contraindication to anticoagulation or increased risk of bleeding, left bundle branch block, pregnant women or women who were breastfeeding, and patients with a life expectancy of less than 1 one year.

Mean age: 62 y, 76% male, 12% diabetes, 9% prior MI.

ACS: 100% (Nnon-STEACS: 0%, STEMI: 100%).

PCI: 100% (balloon angioplasty: 10%, stent: 90%, drug-eluting stents: 24%), pre-treatment with Cclopidogrel: 100%.

InterventionsPre-hospital treatment with tirofiban (25 μg/kg bolus and 0.15 μg/kg/min maintenance infusion for 18 h) or placebo (bolus plus infusion). All patients received at enrolment aspirin and a 600 mg loading dose of clopidogrel.
Outcomes

Primary: extent of residual ST-segment deviation at 1 h after PCI

Secondary: the composite of death, recurrent myocardial infarction, urgent target vessel revascularisation or blinded bail-out use of tirofiban at 30 days

NotesAll patients received 600 mg clopidogrel orally at enrolment.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated
Allocation concealment (selection bias)Low riskBy blinded sealed kits with study drug
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blinded, placebo-controlled study. All staff and study personnel were blinded to treatment. Kits were distributed among the ambulance services or referring centres in blocks of 4. A blinded, independent clinical endpoint committee adjudicated all clinical endpoints
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blinded, placebo-controlled study. All staff and study personnel were blinded to treatment. Kits were distributed among the ambulance services or referring centres in blocks of 4. An independent Data Safety
Monitoring Committee was responsible for identification of safety issues
Incomplete outcome data (attrition bias)
Primary
Low risk8% of the patients in the intervention group and 8% in the placebo group were excluded. In addition, 3.6% in both groups were lost to follow-up. However, these incomplete data were well balanced in the 2 groups and are expected in this kind of study in which patients with a presumed STEMI are randomised in the ambulance
Incomplete outcome data (attrition bias)
Secondary
Low risk8% of the patients in the intervention group and 8% in the placebo group were excluded. In addition, 3.6% in both groups were lost to follow-up. However, these incomplete data were well balanced in the 2 groups and are expected in this kind of study in which patients with a presumed STEMI are randomised in the ambulance
Selective reporting (reporting bias)Low riskThe study protocol is available and all of the study’s pre-specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre-specified way
Other biasUnclear riskFunding: Merck (USA). "The study was investigator initiated. The sponsors of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication".

OPTIMIZE-IT 2009

Methods

Method of treatment allocation: not clear (“by the use of computer-based 1:1 randomisation)

Double-blinded?: no, open randomised study

Stratification: no

Placebo: no

Sample size calculation: yes, but for an alpha error of 0.10

Intention-to-treat analysis: yes

Funding: not stated

Participants

Location: 1 centre in Italy

Timeframe: not stated

Follow-up: 6 months

Eligibility criteria: 46 diabetic patients with CAD undergoing elective PCI

Exclusion criteria: pre-menopause, severe renal failure (creatinine > 2 mg/dl), known haemorrhagic diathesis or thrombocytopenia, life expectancy < 1 year or raised troponin levels

Mean age: 66 y, 72% male, 100% diabetes, 17% prior MI

ACS: 0%

PCI: 100% (balloon angioplasty: 0%, stent: 100%, drug-eluting stents: 67%), pre-treatment with clopidogrel: ?%

InterventionsTirofiban (25 µg/kg bolus plus 0.15 µg/kg/min infusion for 8 hours versus placebo (bolus and infusion). Ticlopidine or clopidogrel (loading dose 300 mg) were administered > 24 hours before the procedure
Outcomes

Primary: incidence of MI and TIMI flow grade after PCI

Secondary: peak troponin levels and myocardial blush grade

NotesTiclopidine (?%) or 300 mg loading dose of clopidogrel (?%) were administered > 24 h before the procedure
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low risk"by the use of computer-based 1:1 randomisation"
Allocation concealment (selection bias)High riskNot stated and probably not
Blinding (performance bias and detection bias)
Primary
High riskOpen-label study
Blinding (performance bias and detection bias)
Secondary
High riskOpen-label study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published report includes all expected outcomes
Other biasLow riskThe study appears to be free of other sources of bias

PARAGON A 1998

Methods

Method of treatment allocation: not stated.

Double-blinded?:¬ yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Hoffman La-Roche (Basel, Switzerland).

Follow-up: 30 days and 6 months.

Participants

Institutions: 389 centres in 29 countries.

Timeframe: from February 1998 to June 1999.

Eligibility criteria: 5,225 patients with NSTEACS.

Exclusion criteria: active bleeding (particularly gastrointestinal bleeding within 1 month or history of active ulcer), impaired haemostasis (oral anticoagulation with international normalizsed ratio > 1.5, bleeding disorder such as von Willebrand disease, or thrombocytopenia [(< 100 000 platelets/µL)]), increased bleeding risk (stroke within 12 months, any prior intracranial haemorrhage, tumour or aneurysm, trauma or major surgery within 1 one month, blood pressure > 180/100 mm Hg despite treatment), contraindication to aspirin or heparin, planned fibrinolysis or GP IIb/IIIa inhibition, GP IIb/IIIa inhibition within 1 one week, left bundle-branch block or pacemaker use, estimated creatinine clearance < 30 ml/min, serious co-morbid disease likely to limit survival, and current enrolment in trials of other investigational drugs or devices.

Mean age: 64 y, male: 66%, prior myocardial infarction: 30%.
Acute coronary syndrome: 100% (unstable angina: 43%, Nnon-STEMI: 57%, STEMI: 0%). ST-segment depression: 44%, CK-MB elevation: 57%, tTroponin elevation: ?%.
PCI: Iin-hospital: 28%, during drug-infusion: 12%
Atherectomy: 4%, balloon angioplasty: 28%, stent: 21%.

Treatment with clopidogrel: 0%.

InterventionsLow-dose lamifiban versus low-dose lamifiban + heparin versus high-dose lamifiban versus high-dose lamifiban + heparin versus placebo + heparin
Dose:
a) 300 µg bolus + 1 µg/min infusion + random assignment to heparin or heparin-placebo
b) 750 µg bolus + 5 µg/min infusion + random assignment to heparin or heparin-placebo
c) placebo + heparin
Duration: 72 to 120 hours; median: 72 hours
OutcomesPrimary: a composite of death, myocardial infarction or severe recurrent ischaemia at 30 days.
Secondary: death or myocardial infarction.
Required level of CK or CK-MB elevation in MI definition: 2 x ULN in spontaneous MI; 3 x ULN in relation to PCI; 5 x ULN in relation to CABG.
Safety: Iintracranial haemorrhage; or bleeding leading to haemodynamic compromise requiring intervention.
NotesThe dose of lamifiban used was the one that had the best results in the previous PARAGON A study.
Performance of angiography and PCI at the discretion of treating physician.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated
Allocation concealment (selection bias)Low riskBy a central telephone
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blinded, placebo-controlled study. Matching heparin-placebo vials were supplied by the same manufacturer. "Systematic blinding of heparin administration and careful control of anticoagulation was achieved by use of a bedside aPTT device that produced encrypted results". A Clinical Events Committee, which consisted of practising cardiologists, was blinded to treatment assignment and adjudicated all clinical primary and main secondary endpoint events according to published predefined criteria
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blinded, placebo-controlled study. Matching heparin-placebo vials were supplied by the same manufacturer. "Systematic blinding of heparin administration and careful control of anticoagulation was achieved by use of a bedside aPTT device that produced encrypted results". A Clinical Events Committee, which consisted of practising cardiologists, was blinded to treatment assignment and adjudicated all clinical primary and main secondary endpoint events according to published predefined criteria
Incomplete outcome data (attrition bias)
Primary
Unclear riskStudy drug was given to 98.4% of the treatment group and 99.1% of the control group. Drug was terminated early in 13% of the control group and in 19% of the lamifiban-treated patients, most commonly for bleeding or planned surgical revascularisation
Incomplete outcome data (attrition bias)
Secondary
Unclear riskStudy drug was given to 98.4% of the treatment group and 99.1% of the control group. Drug was terminated early in 13% of the control group and in 19% of the lamifiban-treated patients, most commonly for bleeding or planned surgical revascularisation
Selective reporting (reporting bias)Low riskFollow-up was completed on 96.8% and 93.3% of patients at 6 months and 1 year, respectively, and were well balanced in the placebo and lamifiban groups
Other biasUnclear riskFunding: Hoffman La-Roche (Basel, Switzerland)

PARAGON B 2002

Methods

Method of treatment allocation: not stated

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Hoffman La-Roche (Basel, Switzerland)

Follow-up: 30 days and 6 months

Participants

Institutions: 389 centres in 29 countries

Timeframe: from February 1998 to June 1999

Eligibility criteria: 5225 patients with NSTEACS

Exclusion criteria: active bleeding (particularly gastrointestinal bleeding within 1 month or history of active ulcer), impaired haemostasis (oral anticoagulation with international normalised ratio > 1.5, bleeding disorder such as von Willebrand disease, or thrombocytopenia (< 100 000 platelets/µL)), increased bleeding risk (stroke within 12 months, any prior intracranial haemorrhage, tumour or aneurysm, trauma or major surgery within 1 month, blood pressure > 180/100 mm Hg despite treatment), contraindication to aspirin or heparin, planned fibrinolysis or GP IIb/IIIa inhibition, GP IIb/IIIa inhibition within 1 week, left bundle-branch block or pacemaker use, estimated creatinine clearance < 30 ml/min, serious co-morbid disease likely to limit survival and current enrolment in trials of other investigational drugs or devices

Mean age: 64 y, male: 66%, prior myocardial infarction: 30%
Acute coronary syndrome: 100% (unstable angina: 43%, non-STEMI: 57%, STEMI: 0%). ST-segment depression: 44%, CK-MB elevation: 57%, troponin elevation: ?%
PCI: in-hospital: 28%, during drug-infusion: 12%
Atherectomy: 4%, balloon angioplasty: 28%, stent: 21%

Treatment with clopidogrel: 0%

InterventionsLamifiban (72 h infusion) versus placebo
Dose:
a) 500 µg bolus + 1 to 2 µg/min infusion depending on creatinine clearance + heparin
b) placebo + heparin
Duration: 72 to 120 h
OutcomesPrimary: a composite of death, myocardial infarction or severe recurrent ischaemia at 30 days
Secondary: death or myocardial infarction
Required level of CK or CK-MB elevation in MI definition: 2 x ULN in spontaneous MI; 3 x ULN in relation to PCI; 5 x ULN in relation to CABG
Safety: intracranial haemorrhage; or bleeding leading to haemodynamic compromise requiring intervention
NotesThe dose of lamifiban used was the one that had the best results in the previous PARAGON A study
Performance of angiography and PCI at the discretion of treating physician
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskComputer-generated
Allocation concealment (selection bias)Low riskA central telephone
Blinding (performance bias and detection bias)
Primary
Low riskRandomised, double-blind, placebo-controlled trial. All suspected MIs and severe, recurrent ischaemic episodes were independently adjudicated by a clinical events committee (CEC)
Blinding (performance bias and detection bias)
Secondary
Unclear riskRandomised, double-blind, placebo-controlled trial. Revascularisation and bleeding outcomes were not adjudicated by a clinical events committee
Incomplete outcome data (attrition bias)
Primary
Low riskFollow-up data were 99.8% complete for the primary endpoint (99.9% for placebo, 99.7% for lamifiban)
Incomplete outcome data (attrition bias)
Secondary
Low riskFollow-up data were 99.8% complete for the primary endpoint (99.9% for placebo, 99.7% for lamifiban)
Selective reporting (reporting bias)Low riskThe study protocol is available and all of the study’s pre-specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre-specified way
Other biasUnclear risk

Following the Data and Safety Monitoring Board recommendation when 1639 patients were accrued, it was decided to augment the original sample by 1200 patients

This study was supported by F. Hoffman-La Roche Ltd, Basel, Switzerland

PRACTICE 2007

Methods

Method of treatment allocation: not stated (“using a prospective randomisation schedule”)..

Double-blinded?: yes

Stratification: yes

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis:?

Funding: Schering Plough.

Follow-up: 30 days and 6 months.

Participants

Institutions: 46 hospitals (34 in France, 5 in Israel, 4 in Spain, 2 in Denmark, and 1 in Germany).

Timeframe: September 2001 to July 2004.

Eligibility criteria: 393 patients with ischaemic chest pain at rest within last 24 h associated with ECG changes and elevated Tn I or T.

Exclusion criteria: Ppersistent ST-segment elevation, recent MI, prothrombin time > 1.2 times control, INR > 2, active bleeding within the previous 30 days, uncontrolled hypertension, major surgery or severe trauma within past 6six weeks, history of stroke, thrombocytopenia, creatinine clearance < 30 ml/min, concomitant use of other GP IIib/IIIa blocker, concomitant severe disease associated with shortened life expectancy or pregnancy.

Mean age: 63 y, 73% males, 22% diabetics, 19% with prior MI.

Acute coronary syndrome: 100% (unstable angina: 0%, Nnon-ST elevation myocardial infarction: 100%, STEMI: 0%)

ST-segment depression: 0%, Ttroponin elevation: 100%.

Coronary angiography: 94%, PCI:¬ 61% (balloon angioplasty: ?%, stent: 45%, drug-eluting stent: ?%).

InterventionsEptifibatide 180 mg/kg bolus + 2 μg/kg/min infusion for 72 h versus placebo. All patients received aspirin + clopidogrel (loading dose: 300 mg) from randomisation. An invasive strategy was planned within 6 to 48 hours after randomisation.
Outcomes

Primary: a composite of death, MI or urgent revascularizsation at 30 days.

Secondary: Iincidence of death, non -fatal MI, and recurrent ischaemia requiring urgent revasculariszation at hospital discharge and at 6six months.

Notes

First study to evaluate the efficacy of upstream administration of a IIb/IIIa antagonist in patients with NSTEACS pre-treated with aspirin and clopidogrel from the time of hospital admission.

Study stopped by the promoter because of low enrolment when 49% of planned patients were included.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNot stated (“using a prospective randomisation schedule”) but probably yes
Allocation concealment (selection bias)Unclear riskNot stated
Blinding (performance bias and detection bias)
Primary
Low riskDouble-blind randomised study
Blinding (performance bias and detection bias)
Secondary
Low riskDouble-blind randomised study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)High riskBecause of slow enrolment, the study was stopped by the promoter when 51% of the planned patients were enrolled
Other biasUnclear riskFunding: Schering Plough

Prati 2005

Methods

Method of treatment allocation: not stated.

Double-blinded?: no (open-label study).

Stratification: no

Placebo: no

Sample size calculation: no

Intention-to-treat analysis: yes

Losses to follow-up: yes, n = 4 (2.9%), all of them in the intervention group.

Funding: not stated.

Follow-up: 30 days.

Participants

Location: 1 centre in Rome, Italy.

Timeframe: not stated.

Eligibility criteria: 140 patients with non-ST elevation unstable anginaUA scheduled for target lesion PTCA.

Exclusion criteria: Ppre-procedural levels of troponin and/or the MB fraction of creatinine kinase (CK-MB) above the upper reference limit, previous coronary artery bypass grafts, congestive heart failure, myocardial infarction in the territory of the treated artery.

Mean age: 62 y, male: 83%, prior myocardial infarction: ?%.
Acute coronary syndrome: 100% (all of them with unstable angina).
PCI: balloon PCI: 0%, stent: 98%%. D, drug-eluting stent: 0%, pre-treatment with clopidogrel: 0%.

InterventionsPatients were randomised to receive abciximab administered as a bolus of 0.25 mg/kg followed by an infusion of 10 mg/min, or not to receive a GP IIb/IIIa inhibitor. After the insertion of the arterial sheath, they were all given 70– to 100 IU/kg of heparin and an additional bolus in order to maintain an activated clotting time of more than 250 sec.
Outcomes

Primary: coronary microcirculatory impairment as assessed by angiographic myocardial blush grade immediately after PCI

Secondary: in-hospital troponin I and CK-MB plasma levels after PCI

NotesTiclopidine 500 mg administered before the procedure. 4Four of the patients randomised to the abciximab group did not complete the study (2two withdrew their informed consent before the intervention and 2two experienced a large groin haematoma after sheath insertion that was interpreted as contraindicating the use of a GP IIb/IIIa inhibitor).
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNo information provided
Allocation concealment (selection bias)Unclear riskNo information provided
Blinding (performance bias and detection bias)
Primary
High riskOpen-label study. Mortality and myocardial infarction adjudicated by the investigators
Blinding (performance bias and detection bias)
Secondary
High riskOpen-label study
Incomplete outcome data (attrition bias)
Primary
High risk4 patients (5.7%) from the intervention group did not complete the study
Incomplete outcome data (attrition bias)
Secondary
High risk4 patients (5.7%) from the intervention group did not complete the study
Selective reporting (reporting bias)Unclear riskInsufficient information to permit judgement
Other biasUnclear riskInsufficient information to permit judgement

PRIDE 2001

Methods

Method of treatment allocation: not stated.

Double-blinded?:¬ yes

Stratification: no

Placebo: yes

Sample size calculation: no

Intention-to-treat analysis: no

Funding: COR Therapeutics, Inc., South San Francisco, California; and Schering-Plough Corp., Kenilworth, New Jersey.

Follow-up: 30 d.

Participants

Institutions: 14 centres in USA.

Timeframe: from September 1996 to June 1997.

Eligibility criteria: 127 coronary patients scheduled to undergo elective PCI.

Exclusion criteria: hHistory of a bleeding diathesis, severe hypertension (systolic blood pressure > 200 mm Hg or diastolic blood pressure > 100 mm Hg on therapy), major surgery within 6 weeks, history of stroke or other central nervous system disease, pregnancy, gastrointestinal or genitourinary bleeding within 30 days, and any other major co-morbid illness

Mean age: 59 y, male: ?, prior myocardial infarction: 51%.
Acute coronary syndrome: 0%.
ST-segment depression: 0%.
PCI: atherectomy: 0%, balloon angioplasty: 55%, stent: 45%, pre-treatment with clopidogrel: 0%.

InterventionsPatients randomised to 4 treatment regiments: 1) placebo bolus and infusion; 2) bolus of 135 µg/kg eptifibatide with a 0.75 µg/kg/min infusion; 3) bolus of 180 µg/kg eptifibatide with a 2 µg/kg/min infusion; 4) bolus of 250 µg/kg eptifibatide with a 3 µg/kg/min infusion.
OutcomesPrimary: to explore the pharmacodynamics of high doses of eptifibatide.
Secondary: Ssafety and the composite incidence at 30 days of death, myocardial infarction, or urgent revascularizsation.
NotesDose-ranging study.
45% of patients underwent stent implantation
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNot stated, but probably yes
Allocation concealment (selection bias)Unclear riskNot stated
Blinding (performance bias and detection bias)
Primary
Unclear riskA randomised, double-blind, placebo-controlled small dose-ranging study. The primary endpoints (eptifibatide pharmacokinetics and its effect on the pharmacodynamics of platelet function) were blinded. However, no clinical events committee adjudicated the events
Blinding (performance bias and detection bias)
Secondary
Unclear riskA randomised, double-blind, placebo-controlled small dose-ranging study. The primary endpoints (eptifibatide pharmacokinetics and its effect on the pharmacodynamics of platelet function) were blinded. However, no clinical events committee adjudicated the events
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published report includes all expected outcomes
Other biasUnclear riskSupported by COR Therapeutics, Inc., South San Francisco, California and Schering-Plough Corp., Kenilworth, New Jersey

PRISM 1998

Methods

Method of treatment allocation: not stated

Double-blinded?:  yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Merck Co.

Follow-up: 30 days

Participants

Institutions: 128 sites in 25 countries (Argentina, Australia, Austria, Belgium, Brazil, Canada, Colombia, Costa Rica, Finland, France, Germany, Greece, Israel, Italy, Mexico, Netherlands, New Zealand, Norway, Portugal, South Africa, Spain, Sweden, Switzerland, United Kingdom, United States)

Timeframe: from March 1994 to October 1996

Eligibility criteria: 3232 patients with unstable angina or non-ST segment elevation myocardial infarction

Exclusion criteria: prior thrombolytic therapy within the previous 48 h, allergy to or intolerance of heparin; serum creatinine > 2.5 mg/dL; active bleeding disorder; history of gastrointestinal bleeding; haematuria; a positive fecal occult blood test; known coagulopathy; a platelet disorder or a history of thrombocytopenia; persistent systolic blood pressure > 180 mm Hg, diastolic blood pressure > 110 mm Hg, or both; a history of haemorrhagic cerebrovascular disease or an active intracranial pathologic process; a history of cerebrovascular disease or transient ischaemic attack within the previous year; a major surgical procedure within the previous month; active peptic ulceration within the previous 3 months; or an invasive procedure within 14 days before enrolment that would substantially increase the risk of haemorrhage

Mean age: 62 y, male: 68%, prior myocardial infarction: 47%
Acute coronary syndrome: 100% (unstable angina: 75%, non-STEMI: 25%, STEMI: 0%)
ST-segment depression: 32%, CK-MB elevation: 24%, troponin elevation: ?%

PCI: in-hospital: 21%, during drug-infusion: 2%
Atherectomy: 0%, balloon angioplasty: 13%, stent: 8%

Treatment with clopidogrel: 0%

InterventionsTirofiban (0.6 µg/kg/min for 30 minutes + 0.15 µg/kg/min infusion for a mean of 47.5 hours + placebo heparin) versus placebo + heparin bolus and infusion
OutcomesPrimary: a composite of death, myocardial infarction or refractory ischaemia at 48 hours
Secondary: a composite of death, myocardial infarction and refractory ischaemia at 7 days
Required level of CK or CK-MB elevation in MI definition: 2 x ULN
Safety: intracranial haemorrhage; bleeding leading to decrease in haemoglobin concentration > 50 g/L; or cardiac tamponade
NotesAngiography was discouraged during the infusion period
PCI was not scheduled
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskUsing a computer random number generator
Allocation concealment (selection bias)Low riskCentral allocation
Blinding (performance bias and detection bias)
Primary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Blinding (performance bias and detection bias)
Secondary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is available and all of the study’s pre-specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre-specified way
Other biasUnclear risk

At the time of the second interim analysis, after 1350 patients had completed the study, the combined rate of clinical events comprised by the primary endpoint was lower than expected. Because of this, the steering committee and the data and safety monitoring committee recommended an increase in the sample size from the initial 2000 planned to 3200

Funding: Merck Co.

PRISM Plus 1998

Methods

Method of treatment allocation: locally by means of sealed envelopes.

Double-blinded?:¬ yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Merck COo.

Follow-up: 30 days

Participants

Institutions: 72 hospitals in 14 countries (Argentina, Australia, Austria, Canada, Chile, Colombia, Denmark, Finland, France, South Africa, Spain, Switzerland, United States),

Timeframe: from November 1994 to September 1996.

Eligibility criteria: 1,915 Hhigh- risk patients with unstable angina or non-ST elevation MI.

Exclusion criteria: ST-segment elevation lasting more than 20 minutes, coronary angioplasty within the previous 6 months or bypass surgery within the previous month, angina caused by identifiable factors, prior thrombolytic therapy in the previous 48 h, serum creatinine > 2.5 mg, an active bleeding disorder or a high risk of bleeding, a history of gastrointestinal bleeding, haematuria, known coagulopathy, a platelet disorder or a history of thrombocytopenia, stroke within the previous year, a history of haemorrhagic cerebrovascular disease or an active intracranial pathologic process.

Mean age: 63 y, Mmale: 67%, Pprior myocardial infarction: 43%.
Acute coronary syndrome: 100% (unstable angina: 55%, Nnon-STEMI: 45%, STEMI: 0%)
ST-segment depression: 58%, CK-MB elevation: 45%, Ttroponin elevation: ?%.

PCI: Iin-hospital: 31%, during drug- infusion: 25%.
Atherectomy: ?%, balloon angioplasty: ?%, stent: ?%.

Treatment with clopidogrel: 0%.

InterventionsPatients were randomised to receive one of three regimens: High-dose tirofiban (0.6 µg/kg/minfor 30 min + 0.15 µg/kg/min infusion) + placebo heparin; versus regular dose tirofiban (0.4 µg/kg/min for 30 min + 0.1 µg/kg/min infusion) + adjusted-dose heparin; versus placebo + heparin
Duration of infusion: 48 to 96 hours
Outcomes

Primary: a composite of death, myocardial infarction and refractory ischaemia at 7 days.
Secondary: the same composite endpoint at 48 h and 30 days; the components of the primary endpoint as separate measures, and a composite of death or myocardial infarction.
Required level of CK or CK-MB elevation in MI definition: 2 x ULN in spontaneous MI; 3 x ULN in relation to PCI.
Safety: Iintracranial haemorrhage; bleeding leading to decrease in haemoglobin concentration > 40 g/L or transfusion of > = 2 U blood; or requiring corrective surgery.

 

NotesAngiography was recommended after the first 48 h of randomisation and during the infusion period (48- to 96 h). PCI performed if indicated by angiography.
The study in the tirofiban-only group was stopped prematurely on the recommendation of the data and safety monitoring board at the time of the first interim efficacy analysis. This effect disappeared at 6-month follow-up. Both tirofiban groups were grouped together for the analysis.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskUsing a computer random number generator
Allocation concealment (selection bias)Low riskLocally by means of sealed envelopes
Blinding (performance bias and detection bias)
Primary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Blinding (performance bias and detection bias)
Secondary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is available and all of the study’s pre-specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre-specified way
Other biasHigh riskThe study in the tirofiban-only group was stopped prematurely on the recommendation of the data and safety monitoring board at the time of the first interim efficacy analysis because of an apparent mortality excess. At that time the sample size of the other 2 arms was increased to 735 patients per group
Funding: Merck Co.

PURSUIT 1998

MethodsMethod of treatment allocation: in a double-blind manner, by co-ordinating centres in the United States or the Netherlands.
Double-blinded?: yes
Stratification: no
Placebo: yes
Sample size calculation: yes
Intention-to-treat analysis: yes
Funding: COR Therapeutics and Schering-Plough Research Institute.
Follow-up: 30 days
ParticipantsLocation: 726 participating hospitals in 28 countries (United States, Argentina, Uruguay, Austria, Belgium, Canada, Chile, Colombia, Czech Republic, El Salvador, Finland, France, Germany, Greece, Guatemala, Hungary, Italy, Mexico, Norway, Poland, Portugal, Spain, Switzerland, the Netherlands, United Kingdom, Venezuela)

Timeframe: from November 1995 to January 1997.

Eligibility criteria: 10,948 patients with unstable angina or non-ST segment elevation myocardial infarction

Exclusion criteria: Ppersistent ST-segment elevation of more than 1 mm, active bleeding or a history of bleeding diathesis, gastrointestinal or genitourinary bleeding within 30 days before enrolment, systolic blood pressure > 200 mm Hg or diastolic blood pressure > 110 mm Hg, a history of major surgery within the previous 6six weeks, a history of non-haemorrhagic stroke within the previous 30 days or any history of haemorrhagic stroke, renal failure, pregnancy, the planned administration of a platelet glycoprotein IIb/IIIa receptor inhibitor or thrombolytic agent, or the receipt of thrombolytic therapy within the previous 24 h.

Mean age: 64 y, Mmale: 65%, pPrior myocardial infarction: 32%.
Acute coronary syndrome: 100% (unstable angina: 54%, Nnon-STEMI: 46%, STEMI: 0%)
ST-segment depression: 50%, CK-MB elevation: 46%, Ttroponin elevation: ?%.

PCI: Iin-hospital: 24%, during drug-infusion: 11%
Atherectomy: ?%, balloon angioplasty: 12%, stent: 12%.

Treatment with clopidogrel: 0%.
InterventionsRegular dose eptifibatide (180 µg/kg bolus + 1.3 µg/kg/min infusion) + heparin versus high-dose eptifibatide (180 µg/kg bolus + 2.0 µg/kg/min infusion) + heparin, versus placebo + heparin
Duration of infusion: 72 to 96 h
OutcomesPrimary: a composite of death or non-fatal myocardial infarction at 30 days.
Secondary: Mmortality at 30 days; myocardial infarction at 30 days; death or myocardial infarction at 96 h and 7 d; bleeding complications.
Required level of CK or CK-MB elevation in MI definition: 1 x ULN in spontaneous MI; 3 x ULN in relation to PCI; 5 x ULN in relation to CABG.
Safety: Iintracranial haemorrhage; or bleeding leading to haemodynamic compromise requiring intervention.
NotesAngiography and PCI at the discretion of treating physician.

After 3218 patients had been randomly assigned to treatment groups, the independent data safety and monitoring committee recommended dropping the lower dose .
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskUsing a computer random number generator
Allocation concealment (selection bias)Low riskCentrally by co-ordinating centres
Blinding (performance bias and detection bias)
Primary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Blinding (performance bias and detection bias)
Secondary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is available and all of the studies pre-specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre-specified way
Other biasUnclear risk

After 3218 patients had been randomly assigned to treatment groups, the committee recommended dropping the lower dose of the intervention arm

Funding: COR Therapeutics and Schering-Plough Research Institute

RAPPORT 1998

Methods

Method of treatment allocation: not stated

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes
Funding: this study was supported by Centocor, Malvern, Pa, and Eli Lilly and Company, Indianapolis, Ind

Follow-up: 30 days and 6 months

Participants

Location: 36 centres in the USA

Timeframe: from 16 November 1995 to 2 February 1997

Eligibility criteria: 483 patients with myocardial infarction (< 12 h) candidates for primary percutaneous transluminal coronary angioplasty

Exclusion criteria: severe thrombocytopenia, baseline prothrombin time > 1.2 times control, ongoing internal bleeding or recent major surgery, previous stroke, severe uncontrolled hypertension, PTCA of the infarct artery within 3 months, cardiogenic shock or prolonged resuscitation, vasculitis, prior administration of abciximab or fibrinolytic therapy, or inability to give written informed consent

Mean age: 61 y, male: 72%, prior myocardial infarction: 21%
Acute coronary syndrome: 100% (unstable angina: 0%, non-STEMI: 0%, STEMI: 100%)
ST-segment depression: 0%, CK-MB elevation: 100%, troponin elevation: ?

PCI: atherectomy: 0%, balloon angioplasty: 85%, stent: 7% (unplanned), pre-treatment with clopidogrel: 0%

InterventionsAbciximab (0.25 mg/kg bolus followed by a 12-h infusion of 0.125 µg/kg/min infusion (maximum, 10 µg/min) versus placebo (bolus and infusion)
Outcomes

Primary: a composite of total death, myocardial infarction and any repeat target vessel revascularisation within 6 months

Secondary: major bleeding

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNot stated but probably yes because this was a multicentre, double-blind, placebo-controlled trial and the same investigators have performed other well-designed and performed trials
Allocation concealment (selection bias)Low riskNot stated but probably yes because this was a multicentre, double-blind, placebo-controlled trial and the same investigators have performed other well-designed and performed trials
Blinding (performance bias and detection bias)
Primary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Blinding (performance bias and detection bias)
Secondary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is available and all of the study’s pre-specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre-specified way
Other biasUnclear riskFunding: this study was supported by Centocor, Malvern, Pa and Eli Lilly and Company, Indianapolis, Ind

RESTORE 1997

Methods

Method of treatment allocation: not stated.

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: yes

Intention-to-treat analysis: yes

Funding: Merck & Co., Inc., Whitehouse Station, New Jersey.

Follow-up: 30 days

Participants

Location: 104 centres in USA and Europe.

Timeframe: from 9 January 9 to 1 December 1, 1995.

Eligibility criteria: 2,141 patients who were undergoing coronary interventions (balloon angioplasty or DCA) within 72 h of presentation with an acute coronary syndrome.

Exclusion criteria: thrombolytic therapy within 24 hours, contraindication to anticoagulation, history of a platelet disorder or thrombocytopenia, history of stroke or other intracranial pathology likely to predispose to bleeding, patients scheduled for elective stent placement or angioplasty using a rotablator or transluminal extraction catheter.

Mean age: 59 y, male: 72%, prior myocardial infarction: 35%.
Acute coronary syndrome: 100% (unstable angina: 67%, Nnon-STEMI: ?, STEMI: 32%).
ST-Ssegment depression: ?, CK-MB elevation: ?, Ttroponin elevation: ?.

PCI: atherectomy: 8%, balloon angioplasty: 92%, stent: 0%, pre-treatment with clopidogrel: 0%.

InterventionsTirofiban (10 mg/kg bolus + 0.15 mg/kg/min infusion for 36 hours versus placebo.
Outcomes

Primary endpoint: 30-day incidence of a composite endpoint of death from any cause, MI, CABG surgery owing to angioplasty failure or recurrent ischaemia, repeat target-vessel angioplasty for recurrent ischaemia or insertion of a stent owing to actual or threatened abrupt closure of the target artery.

Secondary endpoints: the incidence of all individual endpoints.

 

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNot stated but probably yes because this was a multicentre, double-blind, placebo-controlled trial and the same investigators have performed other well-designed and performed trials
Allocation concealment (selection bias)Low riskNot stated but probably yes because this was a multicentre, double-blind, placebo-controlled trial and the same investigators have performed other well-designed and performed trials
Blinding (performance bias and detection bias)
Primary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Blinding (performance bias and detection bias)
Secondary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published reports include all expected outcomes, including those that were pre-specified
Other biasLow riskFunding: Merck & Co., Inc., Whitehouse Station, New Jersey

Schulman 1996

Methods

Method of treatment allocation: not stated.

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: no

Intention-to-treat analysis: no

Funding: COR Therapeutics, Inc, San Francisco, California.

Follow-up: 30 days.

Participants

Location: 15 centres in USA.

Timeframe: not stated.

Eligibility criteria: 227 patients with unstable angina and ST-T changes on admission ECG, or known coronary artery disease.

Exclusion criteria: Ssuspected myocardial infarction in evolution, prior coronary artery bypass graft surgery within 6six months, coronary angioplasty within 72 h, thrombolytic therapy within 7 days, major surgery within 6six weeks, a history of cerebral vascular disease, major gastrointestinal or genitourinary bleeding within 30 days, significant thrombocytopenia (< 100 ,000/mm3), coagulopathy (receiving coumarin or bleeding time > 20 minutes), and if they presented with severe hypertension¬ or had renal insufficiency with a creatinine level > 4 mg/dL.

Mean age: 62 y, male: 63%, prior myocardial infarction: 55%.

ACS: 100% (unstable angina: 100%, Nnon-STEMI: 0%, STEMI: 0%).

ST-segment depression: 33%, CK-MB or Tn elevation: 0%.

PCI: not stated (Aatherectomy: ?%, balloon angioplasty: ?%, stent: ?%).

InterventionsLow-dose eptifibatide (45 µg/kg bolus over 3 minutes followed by a continuous 0.5 µg/kg//min infusion for 24 to 72 h) versus high-dose eptifibatide (90 µg/kg bolus plus infusion of 1 µg/kg/min) versus placebo bolus and infusion
Outcomes

Primary: number and duration of ischaemic episodes on continuous monitoring over the first 24 hours as well as for the entire duration of drug infusion.

Secondary: number and duration of symptomatic ischaemic episodes, ECG episodes of ischaemia after study drug withdrawal, and clinical events of death, myocardial infarction and refractory ischaemia.

NotesThe placebo group received aspirin and heparin while the eptifibatide group received only heparin.
Both active treatment groups have been grouped in our analysis
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNot stated but probably yes because this was a multicenter, double-blind, placebo- controlled trial and the same investigators have performed other well- designed and performed trials.
Allocation concealment (selection bias)Low riskNot stated but probably yes because this was a multicenter, double-blind, placebo- controlled trial and the same investigators have performed other well- designed and performed trials.
Blinding (performance bias and detection bias)
Primary
Low risk

Blinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken.

 

Blinding (performance bias and detection bias)
Secondary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published reports include all expected outcomes, including those that were pre-specified
Other biasUnclear riskFunding: COR Therapeutics, Inc, San Francisco, California

Shen 2008

Methods

Method of treatment allocation: by a 24-h computer-generated random-allocation system

Double-blinded?: no

Stratification: no

Placebo: yes

Sample size calculation: no

Intention-to-treat analysis: yes

Funding: a grant from the Shanghai Science and Technology Committee (no 05DZ19503)

Participants

Location: 1 centre in China

Timeframe: from January 2005 to June 2006

Follow-up: 30 days and 6 months

Eligibility criteria: 172 patients with STEMI presenting < 12 h after the onset of symptoms

Exclusion criteria: cardiogenic shock and known bleeding diathesis

Mean age: 66 y, 81% male, 27% diabetes, ?% prior MI

ACS: 100% (non-STEACS: 0%, STEMI: 100%)

PCI: 100% (balloon angioplasty: 0%, stent: 100%, drug-eluting stents: 100%), pre-treatment with clopidogrel: 100%

InterventionsTirofiban (bolus of 10 µg/kg plus a 36-h infusion of 0.15 µg/kg/min) in the emergency room versus tirofiban (bolus of 10 μg/kg plus a 36-h infusion of 0.15 μg/kg/min)  in the catheterisation laboratory versus placebo (bolus and infusion). Upon admission, loading doses of aspirin (300 mg) and clopidogrel (450 mg) were given for all patients in the emergency room
Outcomes

Primary: occurrence rate of major adverse cardiac events including death, nonfatal MI and target vessel revascularisation (either by PCI or coronary artery bypass surgery) at 30-day and 6-month follow-up

Secondary: haemorrhagic complications and thrombocytopenia

Notes

All patients treated with drug-eluting stents

Upon admission, loading doses of aspirin (300 mg) and clopidogrel (450 mg) were given for all patients in the emergency room

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskBy a 24-h computer-generated random list of numbers
Allocation concealment (selection bias)High riskUsing an open random allocation schedule
Blinding (performance bias and detection bias)
Primary
High riskOpen-label study
Blinding (performance bias and detection bias)
Secondary
High riskOpen-label study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published reports include all expected outcomes
Other biasUnclear riskThere may be a risk of bias, but there is insufficient information to assess whether an important risk of bias exists

Simoons 1994

Methods

Method of treatment allocation: not stated

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: no

Intention-to-treat analysis: yes

Funding: Centocor, Inc, Malvern, Pa

Follow-up: hospitalisation

Participants

Location: 7 centres in Europe

Timeframe: from September 1991 to July 1992

Eligibility criteria: 60 refractory unstable angina patients with chest pain at rest despite optimal treatment and a recent (< 24 h) coronary angiography showing a single culprit lesion suitable for PCI were enrolled, provided that a second coronary angiogram, followed by PCI, could be performed 18 to 24 h after the first (diagnostic) angiogram

Exclusion criteria: features of ongoing ischaemia requiring immediate intervention, prior PCI of the same coronary segment within 6 months, a previous myocardial Q-wave infarction within 7 days, female sex with childbearing potential, recent major trauma including resuscitation, surgery or gastrointestinal or genitourinary bleeding within the past 6 weeks, known hepatic or renal disorder, history of bleeding diathesis or a platelet count of < 100,000/mm3, and known autoimmune disorders

Mean age: 60 y, male: 73%, prior myocardial infarction: 18%

ACS: 100% (NSTEACS: 100%)

Transient ST-segment depression or elevation: 67%; CK-MB or Tn elevation: ?%

PCI: atherectomy: 0%, balloon angioplasty: 100%, stent: 0%

InterventionsAbciximab (0.25 mg/kg bolus + 10 µg/min infusion) versus placebo. The infusion started < 4 h after first coronary angiography and was continued until 1 h after PCI, which was scheduled between 18 and 24 h after the start of the infusion
Outcomes

Primary: a composite of death, myocardial infarction and recurrent ischaemia requiring urgent intervention (PCI, CABG or intra-aortic balloon pump)

Secondary: occurrence of all recurrent ischaemic episodes and angiographic endpoints

NotesAll patients treated with IV nitroglycerin infusion, aspirin and heparin
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskNot stated but probably yes because this was a multicentre, double-blind, placebo-controlled trial and the same investigators have performed other well-designed and performed trials
Allocation concealment (selection bias)Low riskNot stated but probably yes because this was a multicentre, double-blind, placebo-controlled trial and the same investigators have performed other well-designed and performed trials
Blinding (performance bias and detection bias)
Primary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken 
Blinding (performance bias and detection bias)
Secondary
Low riskBlinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published reports include all expected outcomes, including those that were pre-specified
Other biasUnclear riskFunding: Centocor, Inc, Malvern, Pa

Tamburino 2002

Methods

Method of treatment allocation: Bby a “standard list of random numbers” with the use of closed envelopes.

Double-blinded?: no

Stratification: no

Placebo: no

Sample size calculation: no

Intention-to-treat analysis: not stated.

Funding: not stated.

Follow-up: 30 days and 6 months.

Participants

Location: Hospital of Catania (Italy).

Timeframe: from October 1996 to February 1998.

Eligibility criteria: 107 patients with demonstrable reversible ischaemia and > 70% de novo native coronary stenoses requiring implantation of either a stent longer than 20 mm or of multiple overlapping stents.

Exclusion criteria: Ppatients with saphenous graft lesion, bleeding diathesis, thrombocytopenia, history of stroke, active bleeding, severe uncontrolled hypertension, major surgery or trauma within 6 weeks.

Mean age: 62 y, male: 88%, diabetes: 27%, prior myocardial infarction: 67%.
Acute Ccoronary Ssyndrome: 48% (unstable angina: 48%, Nnon-STEMI: ?%, STEMI: 0%).
ST-segment depression: ?%, CK-MB elevation: ?%, Ttroponin elevation: ?%.
PCI: atherectomy: 0%, balloon angioplasty: 0%, stent: 100%, pre-treatment with clopidogrel: 0%, but 100% with ticlopidine.

InterventionsAbciximab (bolus of 0.25 mg/Kkg, followed by an infusion of 0.125 µg/Kkg/min for 12 h) versus placebo. All patients treated with ASA and heparin. Ticlopidine 250 mg twice daily was started the day before the intervention and was prescribed to all patients for 4 weeks following the procedure.
OutcomesPrimary: safety (bleeding and vascular complications) and efficacy in reducing major in-hospital adverse cardiac events related to the procedure (death, MI and urgent revascularisation)
Secondary: reduction in death, MI, target lesion revascularisation and angiographic binary restenosis at 6 months
Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskBy a "standard list of random numbers"
Allocation concealment (selection bias)Unclear riskWith the use of closed envelopes but it remains unclear whether envelopes were sequentially numbered, opaque and sealed
Blinding (performance bias and detection bias)
Primary
High riskOpen-label study
Blinding (performance bias and detection bias)
Secondary
High riskOpen-label study
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published reports include all expected outcomes
Other biasHigh riskThe commercial name of the drug and the pharmaceutical company are listed in the Abstract

TOPSTAR 2002

Methods

Method of treatment allocation: Nnot stated (“The patients were randomised by an independent study nurse”).

Double-blinded?: yes

Stratification: no

Placebo: yes (0.9% NaCl solution).

Sample size calculation: no

Intention-to-treat analysis: not stated.

Funding: Ssupported by a grant from MSD (Merck, Sharp and Dohme) GmbH, Germany.

Follow-up: 30 days and 9 months

Participants

Location: University of Würzburg, Würzburg, Germany.

Eligibility criteria: 96 of 109 patients with stable CAD, a target lesion > 70% suitable for PCI, and thatwho underwent elective PCI after pre-treatment with 375 mg of clopidogrel at least 1 one day before PCI.

Exclusion criteria: acute coronary syndromes, stenosis located in venous or arterial bypass grafts; renal insufficiency; recent peptic ulcers or a history of bleeding, thrombocytopenia or thrombolytic therapy within the previous 24 hours; stroke during the past 2 years; severe hypertension; neoplasms; and previous or planned administration of a GP IIb/IIIa receptor antagonist.

Mean age: 65 years, male: 75%, prior myocardial infarction: 38%.
Acute coronary syndrome: 0%
ST-segment depression: ?, CK-MB elevation: ?, Ttroponin elevation: ?.
PCI: atherectomy: 0%, balloon angioplasty: 8%, stent: 92%, pre-treatment with clopidogrel: 100%.

InterventionsTirofiban bolus of 10 µg/kg + infusion of 0.15 µg/kg/min.
OutcomesPrimary: presence of post interventional release of troponin T after 24 hours
Secondary: incidence of death, MI or target vessel revascularisation
NotesAll patients pre-treated with clopidogrel 375 mg and ASA 500 mg at least 1 one day before PCI.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot stated ("The patients were randomised by an independent study nurse")
Allocation concealment (selection bias)Unclear riskNot stated ("The patients were randomised by an independent study nurse")
Blinding (performance bias and detection bias)
Primary
High riskBlinding of participants and key study personnel ensured during the study period (first 48 h after PCI) but likely that the blinding could have been broken afterwards
Blinding (performance bias and detection bias)
Secondary
High riskBlinding of participants and key study personnel ensured during the study period (first 48 h after PCI) but likely that the blinding could have been broken afterwards
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published reports include all expected outcomes
Other biasUnclear riskFunding: supported by a grant from MSD (Merck, Sharp and Dohme) GmbH, Germany

Yan 2009

  1. a

    ACE; Angiotensin converting enzyme

    AMI: Acute myocardial infarction

    ASA: Acetyl salicylic acid

    ACS: acute coronary syndrome
    AHA/ACC: American Heart Association/American College of Cardiology
    BP: blood pressure

    CABG: coronary artery bypass graft

    CCS: Canadian cardiovascular society
    CK: creatinine kinase
    CK-MB: creatinine kinase-MB fraction
    d: day/s

    DES: drug-eluted stent

    DM: diabetes mellitus
    ECG: electrocardiogram
    GP: glycoprotein
    h: hour/s
    INR: international normalised ratio

    IRA: Infarct-related artery
    IV: intravenous

    IVUS: intravascular ultrasound

    LAD: left anterior descending
    LMW: low molecular weight
    MACE: major adverse cardiovascular event
    MI: myocardial infarction

    MLD: minimal lumen diameter
    NSTEACS: non-ST segment elevation acute coronary syndromes
    NSTEMI: non ST-segment elevation myocardial infarction
    PCI: percutaneous coronary intervention

    PTCA: percutaneous transluminal coronary artery

    PTT: Prothrombin time
    s: second/s
    sc: subcutaneous
    STEMI: ST-segment elevation myocardial infarction

    SVG: single vein graft
    TIMI: thrombolysis in myocardial infarction
    TLR: target lesion revascularisation

    TVR: target vessel revascularisation

    ULN: upper limit of normal
    y: year/s

Methods

Method of treatment allocation: not stated

Double-blinded?: yes

Stratification: no

Placebo: yes

Sample size calculation: not stated

Intention-to-treat analysis: not stated

Funding: not stated

Follow-up: 30 days and 1 year

Participants

Location: 1 centre in China (Beijing Anzhen Hospital, Capital Medical University, Beijing, China)

Timeframe: from June 2005 to June 2006

Eligibility criteria: 240 high-risk patients with NSTEACS

Exclusion criteria: acute Q wave myocardial infarction, coronary artery bypass grafting within 1 month, drug allergy and/or anticoagulation contraindication

Mean age: 63 y, male: 73%, prior myocardial infarction: 15%
Acute coronary syndrome: 100%
ST-segment depression: ?, CK-MB elevation: ?%, troponin elevation:?%
PCI: balloon PCI: ?%, stent: ?%, pre-treatment with clopidogrel: 100%

Interventions

All patients orally took 300 mg of aspirin and 300 mg of clopidogrel before PCI, then took aspirin 100 mg/d and clopidogrel 75 mg/d.

During PCI, heparin was given in intravenous bolus of 5000 IU, then followed by continuous infusion at a rate of 1000 IU/h, to maintain the activated clotting time (ACT) in 300 to 400 seconds

Tirofiban (Wuhan Yuanda Co., Wuhan, China) was given as intravenous bolus of 10 μg/kg over 3 minutes immediately after PCI, then followed by continuous infusion at a rate of 0.15 μg/kg/min for 24 h

Outcomes

Primary: a combined endpoint of death from any cause; new myocardial infarction or repeat target vessel revascularisation

Secondary: platelet aggregation rate, ECG changes, CK-MB and Tn levels, and safety results (bleeding and thrombocytopenia)

Notes

High-risk patients with NSTEACS scheduled for PCI
All patients pre-treated with 300 mg of clopidogrel > 2 h before the procedure

Tirofiban was given as intravenous bolus of 10 μg/kg over 3 minutes immediately after PCI

Study performed in Beijing, China

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskNot explained. "All patients were divided into the study group and the control group according to the principle of randomised, double-blind, placebo-controlled trials"
Allocation concealment (selection bias)Unclear riskNot explained. "All patients were divided into the study group and the control group according to the principle of randomised, double-blind, placebo-controlled trials". No baseline differences were observed between groups
Blinding (performance bias and detection bias)
Primary
Unclear riskNot explained. "All patients were divided into the study group and the control group according to the principle of randomised, double-blind, placebo-controlled trials"
Blinding (performance bias and detection bias)
Secondary
Unclear riskNot explained. "All patients were divided into the study group and the control group according to the principle of randomised, double-blind, placebo-controlled trials"
Incomplete outcome data (attrition bias)
Primary
Low riskNo missing outcome data
Incomplete outcome data (attrition bias)
Secondary
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available but the published reports include all expected outcomes
Other biasLow riskThe study appears to be free of other sources of bias

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
  1. a

    ACS: acute coronary syndrome
    CCU: coronary care unit

    CRP: C-reactive protein
    GP: glycoprotein
    h: hour/s
    IV: intravenous
    MI: myocardial infarction
    NSTEACS: non-ST segment elevation acute coronary syndromes
    NSTEMI: non ST-segment elevation myocardial infarction
    PCI: percutaneous coronary intervention
    RCT: randomised controlled trial
    STEMI: ST-segment elevation myocardial infarction
    UA: unstable angina

ACUITY 2006Comparison of unfractionated heparin or enoxaparin plus any GP IIb/IIIa inhibitor versus bivalirudin plus any GP IIb/IIIa inhibitor versus bivalirudin alone
ADVANCE MI 2005Facilitated thrombolysis (eptifibatide + tenecteplase) versus facilitated PCI (eptifibatide) in patients with STEMI
Alexander 1999Sub-study of the PURSUIT trial on the effect of prior use of aspirin in GP IIb/IIIa inhibitors use in unstable angina
Azar 2010The aim was to assess the anti-inflammatory effect of tirofiban. No clinical events were reported
Batyraliev 2009Study on rescue coronary angioplasty after unsuccessful thrombolysis
Bellandi 2006Comparison of abciximab administered in the emergency room versus in the catheterisation laboratory
Bertrand 2006Study comparing bolus abciximab versus bolus + infusion abciximab
Blankenship 1998EPIC sub-study on local bleeding after GP IIb/IIIa inhibitors use
BOCHUM 2004Open-label pilot study to assess the practical application and safety of pre-hospital eptifibatide versus control in patients with suspected ACS. Patients were assigned eptifibatide or control in even/uneven days
Of the 356 patients included, 42% had a NSTEACS, 32% had a STEMI and 42% had a non-specific chest pain
Boehrer 1994EPIC sub-study on the effect of abciximab in coronary artery bypass surgery
Brener 1999RAPPORT sub-study on the pattern of reperfusion in myocardial infarction patients treated with abciximab
Cannon 1998Trial with an oral GP IIb/IIIa antagonist (TIMI 12)
Claeys 2002Open-label, non-randomised study. Comparison of the degree of inhibition of platelet aggregation after the administration of a loading dose of clopidogrel versus abciximab. 39 patients that underwent PCI with stent implantation
CLEAR PLATELETS 1b 2006Study on the effects of eptifibatide on top of aspirin and clopidogrel on platelet aggregation and clinical markers of inflammation and necrosis. No clinical endpoints reported
CLOTILDA 2005Comparison of tirofiban versus provisional abciximab
Costantini 2004Sub-study of the CADILLAC trial
Cutlip 2003Tirofiban versus control in the emergency room followed by any IIb/IIIa inhibitor during PCI a median of 90 min later
EARLY-ACS 2009Study comparing eptifibatide administered at admission versus at the catheterisation laboratory (a mean of 12 h later) in 9492 patients with NSTEACS
ELISA 2003The ELISA pilot study. This study compared 2 different PCI strategies in patients with UA/NSTEMI rather than 2 different treatments: immediate (median 6 h) PCI after randomisation without pre-treatment with tirofiban versus delayed (median 50 h) PCI after prolonged pre-treatment with tirofiban
Thus, although tirofiban administration was randomised, the basal conditions were different because of differences in timing of administration
Ellis 2008RCT on facilitated PCI in patients with STEMI comparing PCI after abciximab plus half-dose reteplase versus abciximab alone
Emre 2006Comparison of tirofiban administered in the emergency room versus in the catheterisation laboratory
ERAMI 2006Comparison of abciximab administered in the emergency room versus in the catheterisation laboratory
Ercan 2004Small study looking at differences in CRP at 48 to 72 h. No clinical events reported
EVEREST 2006Comparison of tirofiban administered in the CCU versus in the catheterisation laboratory
Ghaffari 1998EPILOG and EPIC joined sub-analysis
GRAPE 1999Pilot study performed in 60 patients with STEMI treated with primary PCI without a control group. Not a randomised study
Gunasekara 2006A non-randomised comparison of abciximab versus high-dose tirofiban
GUSTO V 2001Trial comparing the addition of a GP IIb/IIIa antagonist to the fibrinolytic treatment in patients with ST-segment elevation acute myocardial infarction
Hamm 1999Sub-study of the CAPTURE trial. Differential effects of abciximab in patients with refractory angina according to basal troponin levels
Hamza 2011Small study performed with 50 patients comparing tirofiban + clopidogrel 300 mg versus clopidogrel 600 mg. Just the Abstract from a Congress is available and there is not enough information. In addition, there were no deaths, MI, target-vessel revascularisation or major bleeding
Hanefeld 2002Pilot study of the BOCHUM trial
Heeschen 1999Sub-study of the PRISM trial. Effects of tirofiban in patients with UA/NSTEMI according to baseline troponin levels
HORIZONS-AMI 2008RCT comparing bivalirudin versus any GP IIb/IIIa inhibitor plus heparin in patients with STEMI submitted to primary PCI
IMPACT-AMI 1997RCT on the effect of GP IIb/IIIa inhibitors in patients with ST-segment elevation acute myocardial infarction treated with thrombolytics
INTAMI 2005Comparison of eptifibatide administered in the emergency room versus in the catheterisation laboratory in patients with STEMI submitted to primary PCI
ISAR-REACT 4RCT comparing abciximab plus unfractionated heparin with bivalirudin alone. Excluded because of no appropriate control
Ji 2012Abstract from a congress. Only 41 patients studied and not enough information on clinical events is available (mortality 0%, severe bleeding 0%)
Kereiakes 1997Oral GP IIb/IIIa inhibitor xemilofiban. It is not a randomised clinical trial
Kereiakes 1998aOral GP IIb/IIIa inhibitor xemilofiban
Kereiakes 1998bSub-study of the EPILOG trial. Sub-analysis in unplanned stent patients
Kleiman 1998EPILOG sub-analysis in patients with diabetes
Klootwijk 1998CAPTURE sub-study on silent ischaemia in GP IIb/IIIa inhibitors in unstable angina
Krause 1996Abstract from a congress. A phase II RCT with 3 escalating doses of IV fradafiban in 65 patients with stable angina submitted to elective PTCA
Aim: safety and antiplatelet effects. No clinical events reported
Lefkovits 1996EPIC sub-study on the effects of abciximab on outcomes after percutaneous transluminal coronary angioplasty for acute myocardial infarction
Lenderink 2003Sub-study of the CAPTURE trial
Lincoff 1997EPIC sub-study of prevention of ischaemic complications in patients with unstable angina undergoing percutaneous transluminal coronary angioplasty
Mahaffey 1999PURSUIT sub-study on stroke after GP IIb/IIIa inhibitors in unstable angina
Mak 1997EPIC sub-study on distal embolisation during coronary artery bypass surgery
McClure 1999PURSUIT sub-study on the significance of thrombocytopenia after non-ST-elevation in acute coronary syndromes
McElwee 1997Cost-effectiveness analysis review
Miller 1999GUSTO-III trial sub-analysis on effectiveness of GP IIb/IIIa inhibitors in patients in whom thrombolysis failed
Mockel 2005Comparison of pre-hospital tirofiban versus fibrinolysis before primary PCI in patients with STEMI
Morrow 2001Sub-study of the TACTICS trial
Muller 1997Analysis of the degree of platelet inhibition by an oral GP IIb/IIIa inhibitor fradafiban (ledrafiban is the active prodrug). Not a randomised clinical trial
Murdock 1997Non-randomised study of patients with ST-segment elevation acute myocardial infarction treated with GP IIb/IIIa inhibitors
Narins 1999EPIC sub-analysis on periprocedural myocardial infarction during percutaneous transluminal coronary angioplasty
Neumann 1998Sub-study on coronary flow and left ventricular ejection fraction after GP IIb/IIIa inhibitors in patients with ST-segment elevation acute myocardial infarction who underwent stent implantation
Newby 1999Design description of the SYMPHONY trial with oral GP IIb/IIIa inhibitor sibrafiban
Newby 2001A sub-study of the PARAGON-B study
Okmen 2004No data on the randomisation procedure. Patients with failed PCI were retrospectively excluded from the analysis (8 of 119 patients). No useful data on clinical events
Okmen 2006Comparison of tirofiban versus no tirofiban on QT dispersion in patients that underwent PCI. Patients with failed PCI were excluded and no clinical outcomes were reported
On-TIME 2004Pre-hospital tirofiban versus hospital (median delay 59 min) tirofiban during primary PCI in patients with ST-segment elevation acute myocardial infarction
PARADIGM 1998Trial on GP IIb/IIIa blockers in patients with ST-segment elevation acute myocardial infarction treated with thrombolytics
PARAGON-B 2001PARAGON-B sub study on the effects of lamifiban according to baseline troponin levels
Pels 2008Abciximab administered in the ambulance versus in the catheterization laboratory in patients with ST-segment elevation myocardial infarction undergoing primary PCI
Peng 2009Study performed to analyse the effect of tirofiban not before or during PCI but after primary PCI in patients with STEMI that had no ST-segment resolution
Petronio 2002Rescue PCI in STEMI after thrombolysis
PROLOG 1997Study on the effect of different doses of heparin in patients treated with abciximab during percutaneous revascularisation
Rakowski 2007Abciximab administered in the emergency room versus in the catheterisation laboratory in patients with ST-segment elevation myocardial infarction undergoing primary PCI. No clinical outcomes were reported
RELAx-AMI 2007Abciximab administered in the emergency room versus in the catheterisation laboratory in patients with ST-segment elevation myocardial infarction undergoing primary PCI
Ren 2012Study comparing high-dose tirofiban versus regular dose tirofiban versus control on platelet activity in patients with STEMI that underwent primary PCI. Paper in Chinese. In the English abstract no clinical events are reported
ReoPro-BRIDGING 2004Abciximab at admission versus abciximab immediately before primary PCI (mean difference 62 min) in patients with ST-segment elevation myocardial infarction
REPLACE-2 2003A comparison of planned treatment with heparin plus any GP IIb/IIIa inhibitor versus provisional treatment with bivalirudin plus any GP IIb/IIIa inhibitor only in case of complications during PCI
Roe 2003Treatment with eptifibatide versus placebo in the emergency department followed by open-label eptifibatide 12 to 24 h later
Shen 2007RCT comparing tirofiban versus control in 160 patients with STEMI that underwent primary PCI. The study was performed during the same dates as the Shen 2008 study performed with the same drug and the same type of patients. The authors were contacted to clarify if one study included the patients of the other study. Since they have not responded, this study was excluded from the review
Simpfendorfer 1997Controlled clinical trial with oral GP IIb/IIIa blockade with xemilofiban in patients with unstable angina undergoing percutaneous transluminal coronary angioplasty
SPEED P-St 2000Primary PCI with or without GP IIb/IIIa antagonist in patients with ST-segment elevation acute myocardial infarction treated with a thrombolytic (facilitated PCI)
Steen 2005Comparison of myocardial tissue perfusion with and without tirofiban in patients with STEMI. No clinical events reported
STOPAMI 2000Primary PCI with stent and abciximab versus thrombolysis in patients with STEMI
STOPAMI-2 2002Controlled clinical trial with GP IIb/IIIa blockade in patients with ST-segment elevation acute myocardial infarction, comparing primary PCI with stenting and abciximab versus fibrinolysis and abciximab
Svensson 2006Thrombolysis versus facilitated PCI with abciximab in patients with STEMI
SYMPHONY 2 2001Controlled clinical trial with oral GP IIb/IIIa blockade with sibrafiban in patients with acute coronary syndromes 7 days or more after admission
TAMI-8 1993Pilot study on the effects of abciximab in patients with STEMI treated with thrombolytics
TARGET 2001RCT comparing abciximab with tirofiban in patients submitted to PCI
Thiele 2005Patients with STEMI were randomised to either pre-hospital facilitated fibrinolysis (half-dose reteplase + abciximab) or pre-hospital facilitated fibrinolysis (half-dose reteplase + abciximab) plus PCI
TIGER-PA 2003Open-label randomisation of patients (n = 100) with ST-segment elevation MI to "early" administration of tirofiban in the emergency room versus "late" administration in the catheterisation laboratory immediately before primary PCI
TIMI 14 1999Thrombolysis with or without abciximab in patients with STEMI
TIMI 15A 2000A randomised open-label study of a new drug (RPR 109891) administered IV for 24 to 96 h in 91 patients. Patients were assigned to 1 of 9 regimens. No placebo group was included
TITAN-TIMI 34 2006Comparison on the administration of eptifibatide in the emergency room versus provisional eptifibatide in the catheterisation laboratory
Valgimigli 2005Comparison of tirofiban and an drug eluting stent versus abciximab + bare metal stent during primary PCI in patients with STEMI
van den Brand 1999CAPTURE sub-study on angiographic assessment of GP IIb/IIIa inhibitor use
van den Merkhof 1999Study on the TIMI perfusion grade of 60 patients with STEMI treated with abciximab in the emergency department. Not a RCT
Wong 2003Small study (n = 32) on the coronary flow reserve before and after stenting in patients receiving tirofiban versus control. No data on clinical events
Zajdel 2002Abstract from a congress, written in Polish, with preliminary data. No clinical events
Zeymer 2008Study comparing 2 IIb/IIIa receptor blockers: eptifibatide versus abciximab
Zhang 2011The text of the published article does not mention if this is really a randomised study
Zhao 1999PRISM plus sub-study on angiographic results with tirofiban