Rescue PCI refers to PCI that is performed early (usually within 12 hours after fibrinolysis) in patients with evidence of failed reperfusion—thus with a goal of opening the persistently occluded infarct-related artery. There is now a good evidence base of randomized trials to support the addition of rescue PCI into standard clinical care.

The first randomized trial was conducted by Ellis and colleagues, which studied anterior MI patients with a documented occluded artery (8). Those randomized to a rescue PCI strategy done within 8 hours after the onset of symptoms in patients had a lower mortality rate and decreased frequency of a composite endpoint of death or CHF (8). This trial, and observational studies supporting these benefits (9,10), led to recommendations that rescue PCI be considered following thrombolysis, but its uptake in clinical practice has been relatively haphazard.

Two more recent studies have lent much clearer support to the routine incorporation of rescue PCI into critical pathways. One trial randomized 181 patients with STEMI and evidence of failed reperfusion who had been referred for coronary angiography and had TIMI flow grade ≤2 (11). Patients were randomize to undergo coronary stenting or balloon angioplasty, with a primary endpoint of myocardial salvage index, defined as the proportion of initial scintigraphic perfusion defect salvaged by rescue intervention, as assessed 7 to 10 days after PCI. Myocardial salvage index was significantly greater in the stenting group (0.35 versus 0.25; P = 0.005). One-year mortality tended to favor the stenting group, 8% versus 12% (relative risk, 0.6; P = 0.35). This study concluded that patients with failed thrombolysis benefit from rescue PCI in terms of myocardial salvage, with greater myocardial salvage with coronary stenting (11).

Most importantly, however, the recent REACT trial has reaffirmed the benefit of rescue PCI. This trial studied 427 acute MI patients within 6 hours of pain onset that had failed thrombolysis (diagnosed by <50% resolution of ST elevation on ECG at 90 minutes) (12). They were randomized to conservative treatment, repeat thrombolysis, or rescue PCI. The rate of death, MI, stroke, or severe heart failure at 6 months was significantly reduced with rescue PCI of 15.3% versus 29.8% and 31.0% for the conservative and repeat thrombolysis groups, respectively (each P <0.002) (Fig. 11-1) (12). Thus, rescue PCI should be a part of standard post-thrombolysis care if there are indications of failed thrombolysis. For timing, although benefit of rescue PCI has been seen in these trials when performed on average 6 to 10 hours following thrombolysis, the earlier reperfusion can be achieved the better, and thus efforts are important to shorten time to intervention, as they are for fibrinolysis and primary PCI.

The use of full-dose fibrinolytic therapy as the “facilitating” pharmacologic agent before PCI was tested in several trials, but most definitively in the ASSENT-4 PCI trial (20). A total of 1,667 patients with large STEMIs presenting within 6 hours were randomized to either standard PCI or full-dose tenecteplase (TNK) followed by PCI (that was carried out quickly—on average approximately 1 hour postrandomization). They all received aspirin and unfractionated heparin; GP IIb/IIIa inhibitors were discouraged in the thrombolytic arm, but allowed in the primary PCI arm. The trial was stopped early by the Data and Safety Monitoring Board because of a higher rate of mortality in the pretreatment arm (6% compared with 3%). The primary endpoint—death, congestive heart failure, or shock within 90 days—occurred significantly more frequently in the facilitated PCI arm (18.8% versus 13.7%) (20) (Fig. 11-2). In addition, higher rates of reinfarction (6% versus 4%), repeat target revascularization (7% versus 3%), and stroke (1.8% versus 0%), half of which were intracranial hemorrhages, were seen in the facilitated PCI group. Thus, contrary to the hypothesis, the use of a thrombolytic to open the artery appeared to worsen outcomes, potentially by the known prothrombotic effects of thrombolysis.

A related strategy that might overcome the prothrombotic effects of thrombolysis has also been tested with the use of half-dose fibrinolytic therapy combined with GP IIb/IIIa inhibition before PCI. This unfortunately has not looked promising thus far.

A small randomized trial, the BRAVE study, evaluated whether early administration of the combination of half-dose reteplase plus abciximab produced better results compared with abciximab alone in patients with acute MI referred for PCI (21). The primary outcome measure was assessment of the final infarct size according to a single-photon emission computed tomography perfusion imaging performed between 5 and 10 days after randomization in the 228 patients studied. As expected, patients who received combination therapy had higher coronary patency rates than those treated with abciximab alone (40% vs. 18%, respectively) at the time of initial angiography. This angiographic advantage, however, did not translate into improved myocardial salvage as measured by nuclear scintigraphy. Moreover, bleeding rates were higher in the group receiving the combination of reteplase and abciximab prior to PCI (21). Nearly identical results were seen in another small trial, the ADVANCE MI trial. This trial of just 151 patients also found higher rates of mortality or heart failure with the combination therapy (half-dose fibrinolytic and GP IIb/IIIa inhibitor) than for treatment with a GPIIb/IIIa inhibitor alone (22). The rate of major bleeding was also substantially increased (22). Additional studies are ongoing, but to date this strategy is not encouraging. Thus, based on current evidence, it appears that one needs to choose a single strategy—thrombolysis or primary PCI, but not both.