Chronic heart disease is the leading cause of death in both men and women in the US, and coronary artery disease (CAD) constitutes the number one cause among them. Acute ST-segment-elevation myocardial infarction (STEMI) usually occurs when a thrombus forms on a ruptured atheromateus plaque and occludes an epicardial coronary artery. Patient survival depends on several factors, the most important being restoration of brisk antegrade coronary flow, the time taken to achieve this and the sustained patency of the infarct-related coronary artery. Thrombolysis and percutaneous coronary intervention (PCI) constitute the two approaches for achieving acute reperfusion of the occluded artery and thus reducing infarct size, minimising myocardial damage, preserving left ventricular function and decreasing morbidity and mortality.1 Effective and rapid reperfusion is the most important goal in the treatment of patients with acute STEMI.2,3 When feasible and when performed in a timely and expert fashion, primary PCI is the preferred strategy for reperfusion in the treatment of STEMI because it has been shown to produce superior clinical outcomes compared with fibrinolytic therapy.4–14 However, primary PCI has not become the treatment of choice in many centres because of logistical difficulties, including the inability to offer this treatment strategy in a timely fashion. The time to treatment with primary PCI is an important determinant of the clinical outcome among patients who have had an acute MI,15–19 and current guidelines from the American College of Cardiology (ACC) and the European Society of Cardiology (ESC) recommend a time of less than 90 minutes from the first medical contact to inflation of the balloon.20 Conceptually, the door-to-balloon time may be most important for patients with potentially large infarcts who present early, as they have the most myocardium to salvage.21
Rationale for Early Antithrombotic Therapies and Strategies
The Importance of Saving Time
Many series of investigations published in the last few years have demonstrated that time to treatment is a relevant matter in primary angioplasty, with a significant impact on mortality.17,22–26 Over the last few years, much interest has been focused on pharmacological facilitation with administration of early antithrombotic therapies aiming at early recanalisation.
In fact, the vast majority of STEMI patients currently present to a non-PCI hospital with the need for transportation with subsequent longer delay to treatment; therefore, a pharmaco-invasive approach seems an attractive strategy in the treatment of STEMI patients, particularly among high-risk patients and within the first hours from symptom onset.27
Several reports have demonstrated the prognostic impact of re-infarction after STEMI in patients treated with thrombolysis or primary angioplasty.21,28 In-stent thrombosis after coronary stenting in primary angioplasty is not as low as commonly believed; in fact, it seems that a larger unrestricted use of coronary stenting is associated with a poorer outcome in terms of re-infarction, particularly when glycoprotein (GP) IIb/IIIa inhibitors (GPIs) are not administered,29,30 ranging between 5 and 10%.
Despite successful mechanical revascularisation, suboptimal reperfusion may occur, which results in an unfavourable outcome.31–33 In the last few years, growing interest has focused on the role of distal embolisation as a major determinant of poor reperfusion. Based on the histological analysis of retrieved debris, the Enhanced Myocardial Efficacy and Recovery by Aspiration of Liberated Debris (EMERALD) trial34 showed visible debris in 78% of patients. Henriques et al.35 reported that the incidence of angiographically detectable distal embolisation was 16%, and this was associated with poor reperfusion, larger infarct size and unfavourable five-year survival compared with patients without angiographic signs of distal embolisation.
Inter-individual Variability in the Response to Conventional Antiplatelet Therapies
High inter-individual variability in response to antiplatelet therapies has been demonstrated.36–42 The percentage of non-responders ranges between 5 and 50% (according to different laboratory tests) for aspirin and between 20 and 30% for clopidogrel.36–42 Several mechanisms have been proposed.43–58 The question of the optimal pharmacological therapy for reperfusion before and in conjunction with primary PCI, especially when there is a delay in the initiation of therapy, remains unanswered.
Fibrinolytic therapy alone was found to be harmful among patients in the Assessment of the Safety and Efficacy of a New Treatment Strategy with PCI (ASSENT-4 PCI) trial, possibly owing to the deleterious effects of early activation of platelets by the fibrinolytic agents without effective antiplatelet treatment or plaque haemorrhage at the time of PCI. GPIs alone, and especially in conjunction with fibrinolytic therapy, have been evaluated in relatively small numbers of patients. The results of these studies have been inconclusive,59–64 although there is evidence of an increase in major bleeding, particularly among the elderly.65
Aggressive antithrombotic therapy carries a risk of bleeding and blood transfusion. Although the true incidence of bleeding depends on the population studied (i.e. clinical trial versus registry) and the definition used,66,67 it is clear that bleeding is associated with an increased risk of adverse outcomes including MI and death.66,68 Therefore, therapies that provide an adequate level of anticoagulation to reduce ischaemia while simultaneously minimising the risk of bleeding and transfusion have the potential to improve outcomes among patients with STEMI, especially in those patients at higher risk of bleeding complications, such as those with low bodyweight, female gender and impaired renal function.66–69
Glycoprotein IIb/IIIa Inhibitors
Several randomised trials have been conducted in primary angioplasty, the vast majority of them on abciximab.31,70–82 In the largest trial, Controlled Abciximab and Device Investigation to Lower Late Angioplasty Combinations (CADILLAC),79 a total of 2,082 patients were randomised to stent or balloon with or without peri-procedural administration of abciximab. Abciximab did not improve myocardial perfusion as evaluated by myocardial blush grade and ST-segment resolution.81 Some benefits in mortality with abciximab were observed in patients undergoing balloon angioplasty only, whereas no benefits were observed in terms of re-infarction. Abciximab did not increase the risk of bleeding complications; however, a major limitation of this study was the relatively low-risk population. In fact, in trials without strict patient selection, as conducted by Antoniucci et al., abciximab was associated with benefits in terms of death and re-infarction.79
A recent meta-analysis of randomised trials has shown that periprocedural abciximab administration is associated with a significant reduction in mortality and re-infarction, without an increased risk of major bleeding complications.70 However, data from the Bavarian Reperfusion AlternatiVes Evaluation (BRAVE-3) trial83 showed no benefits in infarct size and 30-day mortality when a clopidogrel-loading dose of 600mg was administered. Keeping in mind the relationship between the risk profile and mortality benefits from abciximab administration, it could be claimed that the absence of benefits would have been expected in a population with the mortality lower than 3%, as observed in the BRAVE-3 trial.83
Limited data have been reported on eptifibatide and tirofiban.82,84–88 Steen et al.82 did show in a small randomised trial (53 patients) significantly improved epicardial and myocardial perfusion by adjunctive tirofiban. In the randomised Zwolle study,84 the data showed that high-dose tirofiban was associated with better platelet inhibition compared with abciximab or a standard dose of tirofiban. In the STRATEGY trial,86 no difference in death and/or re-infarction was observed between high-dose tirofiban and abciximab. Data from the Multi-STRATEGY trial87 have shown a similar outcome (non-inferiority) between tirofiban and abciximab in 745 STEMI patients undergoing primary angioplasty, and no difference in major bleeding complications. In the eptifibatide versus abciximab in primary PCI for acute ST elevation MI (EVA-AMI) trial,88 400 STEMI patients were randomly assigned to peri-procedural administration of eptifibatide or abciximab, with a similar outcome between the two molecules. The major limitation of the study is that the primary end-point (ST resolution at 60–90 minutes) was available in only 50% of patients.
Further benefits may be expected by adjunctive intracoronary administration of GPIs. A small randomised trial89 showed that selective intracoronary administration of abciximab distally to the occlusion (through an over-the-wire balloon) was associated with a significant improvement in myocardial perfusion and smaller infarct size.
Several randomised trials have been conducted to evaluate the benefits of early GPI administration in patients undergoing primary angioplasty.60,77,91–101 In the ongoing tirofiban in MI evaluation (On-TIME) trial,92 a total of 507 STEMI patients transferred to a PCI centre were randomised to early, pre-hospital initiation of tirofiban (early) or to its initiation in the catheterisation laboratory (late). Early tirofiban was associated with a better pre-procedural thrombolysis in MI (TIMI) grade 2–3 flow (43 versus 34%; p=0.04) and myocardial perfusion (myocardial blush grade 2–3: 30 versus 22%; p=0.04). However, no benefits were observed in post-procedural TIMI grade 3 flow, myocardial perfusion, mortality (4.5 versus 3.7%; p=0.66) and re-infarction (2.4 versus 3.7%; p=0.43) at one-year follow-up. Similar results have been observed in the Time to Integrilin Therapy in Acute MI (TITAN)-TIMI,34 where 316 STEMI patients were randomised to early or late eptifibatide.93
In the on-TIME 2 trial,102 the data pertaining to 984 patients with STEMI were examined. In the ambulance en route to receiving primary PCI, patients were randomised to receive either tirofiban (25μg/kg bolus plus 0.15μg/kg/minute infusion for 18 hours) or a placebo, in addition to aspirin (500mg), heparin (5,000IU) and clopidogrel (600mg). The median time from symptom onset to initial electrocardiogram (ECG) diagnosis was 76 minutes, and the median time from pre-hospital drug administration to angioplasty was 55 minutes. The baseline clinical and angiographic characteristics of the two patient arms were similar.
Early tirofiban was associated with a better median ST deviation one hour after angiography/PCI (4.8 versus 3.6mm; p=0.003), but there was no difference in 30-day mortality (2.3 versus 4%; p=NS), recurrent MI (2.75 versus 2.9%) and major bleeding (4 versus 2.9%). Indeed, there was a difference in thrombotic bail-out (19.9 versus 28.5%; p=0.002) (see Table 1).
Recent data103 combining the on-TIME 2 patients with 414 open-label STEMI patients randomised to tirofiban or placebo showed one-year survival benefit for those who ultimately went on to have primary PCI (n=1,555, 97.6 versus 94.5%; p=0.007). As the rate of bail-out tirofiban was 20% in the tirofiban group and only 29% in the placebo group, one can argue that the on-TIME 2 trial does not prove the concept of early GPIs, but merely the class effect of this type of therapy in STEMI patients. This is confirmed by the Comparison of the Pre-hospital or Cath-lab Administration of High Dose Tirofiban in Patients Undergoing Primary Angioplasty (AGIR 2) data.104
This study randomised 320 STEMI patients who were being transferred for primary PCI to the GPI tirofiban (loading dose of 25μg/kg followed by an infusion of 0.15μg/kg/min) during ambulance transfer (n=164) or in the catheterisation laboratory (n=156). On average, patients in the pre-hospital group received tirofiban 48 minutes earlier than the catheterisation laboratory group. All patients were also given clopidogrel (600mg loading dose), aspirin and heparin. Baseline and procedural characteristics were similar between the two groups, with the exception of previous PCI, which was more common in the catheterisation laboratory group (14.7 versus 7.9%; p=0.05).
The proportion of patients with TIMI grade 2–3 flow at initial angiography (primary end-point) was similar in both groups (44.2% in the pre-hospital group versus 39.7% in the catheterisation laboratory group; p=0.42). In fact, no differences were observed between the groups in TIMI flow grade either before or after PCI. Patients given pre-hospital tirofiban were more likely to have ST-segment resolution >70% on admission to the catheterisation laboratory, although the difference was not significant (15.2 versus 8.7%; p=0.10). Indeed, the difference had disappeared one-hour post-PCI (52.6% in the pre-hospital group versus 55.4% in the catheterisation laboratory group; p=0.64). Troponin and creatine kinase (CK) levels were similar in the two groups. In-hospital clinical events were also similar, with death and severe bleeding casting further negative light on pre-hospital tirofiban (see Table 2).
The length of time between symptom onset and treatment or the duration of tirofiban infusion did not significantly influence TIMI flow grade or ST-segment resolution. Although the trial was not powered for clinical events, the data do not support pre-hospital administration of tirofiban provided that patients are pre-treated with a high loading dose of clopidogrel, aspirin and heparin in the pre-hospital setting and that they receive tirofiban on admission to the catheterisation laboratory.
The largest study on the concept of facilitated PCI is the Facilitated Intervention with Enhanced Reperfusion Speed to Stop Events (FINESSE) trial,105 designed to compare the efficacy and safety of the early administration of reduced-dose reteplase and abciximab in combination or abciximab alone followed by PCI with abciximab only administered just before PCI. A total of 2,452 patients were enrolled, all presenting within six hours after the onset of signs and symptoms of cardiac ischaemia.
All patients had ST-segment elevation suggestive of an acute MI, were eligible for fibrinolytic therapy or primary PCI and the estimated time to diagnostic catheterisation was one to four hours after randomisation. These patients were randomly assigned through a central randomisation centre in a 1:1:1 ratio to receive reteplase plus abciximab (combination-facilitated PCI), abciximab alone (abciximab-facilitated PCI) or placebo (primary PCI). Immediately after randomisation, patients in the combination-facilitated PCI group received intravenous doses of abciximab (0.25mg/kg) and reteplase (two 5U boluses separated by 30 minutes for those <75 years of age, or one 5U dose for those ≥75 years of age); patients in the abciximab-facilitated PCI group received an intravenous bolus of abciximab at a dose of 0.25mg/kg.
The primary end-point was a composite of death from all causes, ventricular fibrillation occurring more than 48 hours after randomisation, cardiogenic shock and congestive heart failure requiring re-hospitalisation or an emergency room visit over a 90-day follow-up. Major secondary end-points were complications of MI over 90 days (as in the primary end-point), death from all causes over 90 days and ST-segment resolution of more than 70% from baseline as assessed at 60–90 minutes after randomisation. Major safety end-points were non-intracranial major or minor bleeding as assessed by the TIMI classification and intracranial haemorrhage up to discharge or day seven, whenever was sooner.
The baseline characteristics of the patients and the initial treatments received were similar across all groups. At randomisation, 66.7% of the patients were considered to be at high risk. The median door-to-balloon time for all patients was 2.2 hours (interquartile range 1.8–2.8), and 92% of the patients underwent PCI (see Figure 1). Significantly more patients in the combination-facilitated PCI group than in the abciximab-facilitated PCI group or the primary PCI group had ST-segment resolution that was greater than 70% in 60–90 minutes (43.9 versus 33.1 and 31%, respectively; p=0.003 for combination-facilitated PCI versus primary PCI; p=0.01 for combination-facilitated PCI versus abciximab-facilitated PCI). Significantly more patients in the group that received reteplase plus abciximab than in the group that received abciximab alone or underwent primary PCI had a TIMI flow grade of 3, as determined by the site investigator, before PCI was performed (32.8 versus 14.1 and 12%, respectively; p<0.001 for both comparisons). No substantial difference among treatment groups was seen for TIMI flow grade after PCI or for ST-segment resolution at 180–240 minutes.
The area under the curve for CK was significantly reduced with combination-facilitated PCI (1,625IU per litre per hour) compared with both abciximab-facilitated PCI (1,782IU per litre per hour) and primary PCI (1,860IU per litre per hour) (p=0.01 and p<0.001, respectively).
The 90-day primary composite end-point occurred in 9.8% of the patients in the combination-facilitated PCI group, in 10.5% of the patients in the abciximab-facilitated PCI group and in 10.7% of the patients in the primary PCI group (hazard ratio [HR] in the combination-facilitated-PCI group compared with the primary PCI group 0.91, 95% confidence interval [CI] 0.67–1.23). Complications of MI occurred in 7.4, 7.5 and 9% of patients in the three groups, respectively, with no significant differences. The individual components of the primary end-point did not differ significantly among the combination-facilitated PCI, abciximab-facilitated PCI and primary PCI groups: the respective rates were 5.2, 5.5 and 4.5% for death from all causes, 0.6, 0.2 and 0.4% for ventricular fibrillation occurring more than 48 hours after randomisation, 5.3, 4.8 and 6.8% for cardiogenic shock and 1.9, 2.9 and 2.2% for re-hospitalisation or an emergency room visit for congestive heart failure. The Kaplan-Meier estimate of the composite primary end-point is shown in Figure 2.
Non-intracranial TIMI major or minor bleeding, assessed on the basis of the TIMI classification, occurred in 14.5, 10.1 and 6.9% of the patients in the combination-facilitated PCI, abciximab-facilitated PCI and primary PCI groups, respectively (p<0.001 for the comparison of combination-facilitated PCI with primary PCI). Intracranial haemorrhage occurred in 0.6% of the patients in the combination-facilitated PCI group (four patients with cerebral haemorrhage and one with cerebral infarction and major haemorrhage), no patients in the abciximab-facilitated PCI group and 0.1% of the patients in the primary PCI group.
Previous studies have suggested that the addition of a combination of reduced-dose fibrinolytic therapy and GPIs would improve early TIMI flow and ST-segment resolution compared with treatment with aspirin and antithrombin therapy only.60–63,106 The results of this trial confirm these observations. Significantly greater TIMI 3 flow before PCI and ST-segment resolution 60–90 minutes after the initiation of treatment were seen in the group that received treatment with abciximab plus reteplase compared with the groups that received placebo or abciximab only.
Therefore, the question remains why enhanced early reperfusion did not significantly improve clinical outcomes. Differences in the time to reperfusion may affect major myocardial salvage during only approximately the first two hours after the onset of infarction, an interval that is shorter than that in which many patients can be treated; after two hours, the time-dependency of PCI-mediated salvage may be considerably attenuated. Treatment was initiated three hours or less after the onset of symptoms in only 60% of the patients in this study, and among these patients there was a very modest trend towards more clinical benefit from treatment with abciximab plus reteplase (HR for the composite end-point 0.83, 95% CI 0.56–1.21) compared with patients who presented later (HR 1.12, 95% CI 0.67–1.86).
On the other hand, only high-risk patients are likely to have a major benefit from early reperfusion with PCI. Indeed, among the patients in our study who were classified as being at high risk (those >70 years of age and those with anterior MI, Killip class >I or a presenting heart rate >100 beats per minute), accounting for 67% of all of the patients in the study, there was a weak trend towards a greater benefit of combined therapy compared with treatment with primary PCI with abciximab administered in the catheterisation laboratory immediately before PCI (HR for the composite end-point 0.84, 95% CI 0.60–1.17), whereas there was no benefit for lower-risk patients (HR 1.22, 95% CI 0.56–2.63). However, these subgroup analyses should not be used to justify the use of combination-facilitated PCI in higher-risk patients, given the absence of an effect on mortality and the excess bleeding observed in this study.
In summary, the use of a facilitated pharmacological strategy for reperfusion, with either abciximab only or abciximab plus reduced-dose reteplase, in anticipation of urgent PCI for patients with a STEMI cannot be justified. Primary PCI with abciximab administered in the catheterisation laboratory provides a better risk–benefit ratio than the two facilitated strategies among patients with STEMI who can undergo PCI within four hours after the first medical contact. At the same time, the data suggest further study may be needed in certain high-risk groups, and there is a need to find therapies that improve reperfusion without greatly increasing bleeding risk.