Drug-Eluting Stents in Acute Myocardial Infarction: Title and subTitle BreakIs Science Catching Up With Practice?

In the last 3 to 4 years, the use of drug-eluting stents in clinical practice has revolutionized interventional cardiology. Two pivotal trials using sirolimus- and paclitaxel-eluting stents and reporting previously unheard-of single-digit restenosis rates^{1 - 2} heralded a new era for interventional cardiology, suggesting that the most vexing problem for intervention, namely, restenosis, had finally been tackled. These 2 trials predominantly studied patients undergoing elective procedures and found target-lesion revascularization rates of approximately 3% to 4% and postprocedure myocardial infarction rates that ranged from 2.8% to 3.5%.

In the United States, drug-eluting stents were approved by the Food and Drug Administration in April 2003, after approximately 90 000 sirolimus-eluting stents had been distributed outside the United States.³ Despite the fact that simple lesions in low-risk patients were studied in the clinical trials that led to approval of these devices in the United States, the adoption of these stents in clinical practice was dramatic compared with the relatively slow adoption in other countries, which was partially driven by economic considerations. In October 2003, a mere 6 months after approval, more than half of all the 450 000 drug-eluting stents distributed worldwide had been implanted in the United States.⁴ Recent data from the CRUSADE Quality Improvement Initiative, which examines US patients with high-risk non–ST-segment elevation myocardial infarction (NSTEMI), showed that average use of drug-eluting stents in the summer of 2004 was approximately 80%, with more than a quarter of the hospitals using drug-eluting stents in more than 85% of cases, even though no single randomized clinical trial has been undertaken to assess the use of these stents in patients with NSTEMI.⁵

This broad adoption of drug-eluting stents raises a number of important issues. Perhaps most important is whether the experience in stable patients can be extrapolated to unstable patients, including those with NSTEMI and ST-segment elevation myocardial infarction (STEMI). While it may seem reasonable to assume that drug-eluting stenting will have an outcome in patients with acute myocardial infarction (AMI) similar to the outcome in those with more stable coronary artery disease, there are a number of specific concerns, particularly in the thrombotic milieu of AMI. Data on late stent thrombosis (>30 days) recently reported by McFadden et al⁶ raise concerns about long-term safety, as the majority of these cases occurred when a drug-eluting stent was placed in a patient with acute coronary syndrome.

In this issue of JAMA, Iakovou and colleagues⁷ also provide a detailed observational report on late thrombosis with either sirolimus- or paclitaxel-eluting stents in 2229 patients at 3 high-volume referral European sites. As opposed to the clinical trials that led to approval of drug-eluting stents, the baseline and angiographic patient characteristics in the study by Iakovou et al showed a high prevalence of diabetes, multivessel disease, small reference-vessel diameter, and complex lesions in this observationalstudy. As expected, the overall stent thrombosis rates (1.3%) were higher than in the randomized trials^{1 - 2} and were associated with high mortality rates.

Virmani et al⁸ suggested that delayed healing and hypersensitivity to the polymer play a major role in the mechanism of late thrombosis after drug-eluting stent implantation. These findings have been compounded by anecdotal reports of cessation of aspirin and clopidogrel within the first 6 months for elective surgery procedures with dramatic rates of stent thrombosis.⁹ The report by Iakovou and colleagues⁷ further reinforces the importance of long-term aspirin and clopidogrel therapy. In their multivariable analysis, the strongest predictor of stent thrombosis was premature discontinuation of antiplatelet therapy, exceeding other independent predictors such as renal failure, bifurcation lesion, diabetes, and low ejection fraction.

The report of the STRATEGY trial by Valgimigli and colleagues,¹⁰ also in this issue of JAMA, provides the first randomized clinical trial data on drug-eluting stenting in patients with STEMI. In this well-conducted single-center study, a total of 175 patients were randomized to receive single high-dose bolus tirofiban plus sirolimus-eluting stenting vs abciximab plus bare-metal stenting. The rationale for this randomization scheme instead of a more traditional 2 × 2 factorial design was based on the hope that the lower cost of tirofiban would offset the cost of drug-eluting stenting and therefore have similar financial impact as a combination of abciximab plus bare-metal stenting. Intravenous antiplatelet therapy was started approximately 30 minutes before percutaneous coronary intervention (PCI). Reperfusion indicators, such as ST-segment resolution and epicardial flow, were similar in both groups.

The primary composite end point of death, reinfarction, stroke, and angiographic binary restenosis at 8 months occurred more frequently in patients allocated to receive abciximab plus bare-metal stenting (50% vs 19%, P<.001). The main driver was a higher restenosis rate in the bare-metal stent group (36% vs 9%, P = .002). There were no differences in death, reinfarction, or stroke at 8 months. The low 30-day death and reinfarction rate of 5% is consistent with previous studies of primary PCI.^{11 - 12} On the other hand, there were no episodes of stent thrombosis in the drug-eluting stent group.

The STRATEGY trial investigators only used sirolimus-eluting stents, so a question can be raised whether similar outcomes can also be achieved with paclitaxel-eluting stents. To date there is no direct evidence for the use of paclitaxel-eluting stents in AMI. However, some recent data comparing different types of drug-eluting stents in more stable patients have yielded interesting information with results that may favor the sirolimus-eluting stents over paclitaxel-eluting stents.^{13 - 15} Although these studies show differences that are mostly limited to angiographic parameters, they may reflect differential cell biology effects of sirolimus and paclitaxel. On the other hand, the TAXUS program has studied the most complex lesions (outside STEMI) in which paclitaxel-eluting stents can be more easily delivered and rather surprisingly showed 2- to 3-times higher periprocedural myocardial infarction rates in the group of patients receiving overlapping paclitaxel-eluting stents.¹⁶ No such information is available for sirolimus-eluting stents and it was not recorded in the STRATEGY trial. Use of multiple overlapping stents is frequent practice in the treatment of complex lesions, many of which are generally seen in patients presenting with unstable acute coronary syndromes such as STEMI or NSTEMI.

The STRATEGY trial appears to extend the findings from a population with stable coronary artery disease to a more unstable group with favorable angiographic results during follow-up. Until now, only observational data were available on the use of drug-eluting stents in AMI. The largest published data set is the single-center RESEARCH registry from Rotterdam, which compared the outcomes of 186 consecutive patients with STEMI treated with drug-eluting stents with 183 patients treated with bare-metal stents before drug-eluting stents became available.¹⁷ In comparison with the STRATEGY trial, patients enrolled in the registry were slightly more likely to have previous history of myocardial infarction and to present with cardiogenic shock, but other markers of risk, such as age, diabetes, and time to treatment were similar. Short- and long-term outcomes were comparable in the registry and in the randomized trial in terms of death and myocardial infarction.¹⁷ The composite rate of death and reinfarction was 9% at 10 months in the RESEARCH registry and 13% at 8 months in the STRATEGY trial. In addition, the long-term rates of target-vessel revascularization in the RESEARCH registry (5%),¹⁷ the randomized SIRIUS trial (7%),¹ and the STRATEGY trial (7%) were similar. Thus, a consistent picture of low rates of restenosis has emerged with sirolimus-eluting stents. However, the TVR rate in the bare-metal stent group of the STRATEGY trial was 20%, which is substantially higher than in the bare-metal stent groups of the PAMI-STENT¹¹ and CADILLAC trials,¹² as well as in the bare-metal stent cohort of the RESEARCH registry. In these studies the 6-month rates of TVR were approximately 7% to 8%. The reason for this discrepancy is not readily apparent, but it highlights that more randomized trials of drug-eluting stents compared with bare-metal stents are needed in AMI, some of which are ongoing.

The STRATEGY trial is too small to address the most appropriate periprocedural pharmacological approach. The regimen used acutely was different in the 2 study groups, with the less expensive tirofiban linked with the more expensive drug-eluting stent and the more expensive abciximab linked with the less expensive bare-metal stent, in an attempt to preserve cost in a fixed health care expenditure setting such as Europe; however, somewhat surprisingly, the authors did not report a formal economic analysis. In addition, a more traditional approach would have been to use a 2 × 2 factorial design to explore the relative merits of the different glycoprotein IIb/IIIa blockers with drug-eluting or bare-metal stenting. Moreover, this trial used a higher tirofiban bolus dose (25 μg/kg) compared with the earlier trials. The TARGET trial showed that tirofiban at the current labeled bolus dose (10 μg/kg) is inferior to abciximab as adjunct to elective PCI, especially in unstable patients.¹⁸ Pharmacodynamic studies have subsequently shown that a tirofiban bolus dose of 10 μg/kg was insufficient to achieve optimal early platelet inhibition.^{19 - 20} While the results of small studies with higher bolus doses appear to be encouraging, the results of an ongoing phase 3 clinical trial may provide more definitive answers regarding safety and efficacy of tirofiban with a higher bolus dose.²¹ There is also limited data on tirofiban in the STEMI setting. Three smaller trials comparing early vs late administration of tirofiban before primary PCI showed conflicting results concerning Thrombolysis in Myocardial Infarction (TIMI) 3 flow restoration in the infarct-related vessel.²² Abciximab is the glycoprotein IIb/IIIa inhibitor with the most robust data in the primary PCI setting and is generally the recommended agent of choice in patients with STEMI,^{23 - 24} although the largest US experience, in the CADILLAC trial, showed less-obvious findings compared with many of the other trials.¹² The STRATEGY trial therefore falls short of providing data to help clarify several aspects regarding the acute antithrombotic regimen used, due to the study design and sample size.

Despite these reservations, the STRATEGY trial is an important step in exploring the use of drug-eluting stents in the setting of AMI. The data suggesting that the use of drug-eluting stents in AMI may be superior to the traditional bare-metal stent approach are encouraging. However, the focus should now shift to the long-term prevention of thrombotic complications through appropriate long-term antiplatelet therapies. It is remarkable to see how STEMI management has evolved over the last decade. With use of increasingly sophisticated stent technology, patients with STEMI are now enjoying substantial benefit by virtue of improved reperfusion with primary PCI, fewer ischemic complications, and lower rates of long-term restenosis. A goal that seemed very distant only a decade ago appears now to have been achieved.