Drug-Eluting or Bare-Metal Stents in ST-Elevation Myocardial Infarction: A Growing Dilemma or Closer to an Answer?

The decision of whether to implant a drug-eluting stent (DES) or bare-metal stent (BMS) in potentially the highest “risk” situation, primary percutaneous coronary intervention (pPCI) for acute ST-elevation myocardial infarction (STEMI), represents, in many ways, an evolution of science and practice. As our understanding of the risks and benefits of DES implantation has progressed, the pendulum of enthusiasm of use may have swung from one extreme to another. Now, the totality of evidence may well have left us somewhere in between, in a confusing space informed by a wide array of epidemiologic and mechanistic data. In this issue of the Journal of Invasive Cardiology, Pedersen et al¹ attempt to add yet another important piece to the puzzling question, “Which stent type is optimal in STEMI patients?”

First studied in low-risk populations, DES were initially met with significant fervor, even in the absence of data for acute coronary syndromes (ACS); as early as June 2004, over 80% of all non-ST elevation ACS patients in the CRUSADE registry were treated with DES.² Certainly, the benefit of DES with respect to reductions in target lesion revascularization (TLR) rates compared with BMS has been consistently demonstrated. However, as initial reports of the previously largely unrecognized entities of late (after 30 days) and very late (after 12 months) stent thrombosis began, concern grew for the risk of DES. Further, the devastating impact of stent thrombosis, with associated death rates of 45% presented in an early observational series by Iakovou et al, heightened concern, particularly in a non-clinical trial setting.³ In that regard, a flurry of mechanistic and epidemiologic studies ensued. Mechanisms including delayed or incomplete reendothelialization, inflammation from polymer, late, acquired malapposition of stent struts and fracture, have all been examined in differing indications for PCI with varying degrees of evidence. In STEMI indications in particular, pathologic autopsy data demonstrated significant delay in vessel healing with DES, when compared with stable angina with DES.⁴ This finding was recently nearly replicated in an optical coherence tomography substudy of the randomized trial, HORIZONS-AMI.⁵ In addition, higher degrees of late acquired stent malapposition in STEMI patients receiving DES were seen by intravascular ultrasound.⁶ In all, prior mechanistic data suggesting risk of DES in stable angina patients appear more germane in STEMI populations. How these pathologic data translate to patient outcomes is not well established, as the direct link between the pathology and epidemiology remains missing.

In this regard, an intial 2006 meta-analysis of the 9 pivotal Cypher and Taxus randomized controlled trials (RCT) established the epidemiologic evidence of risk of ST in DES, with demonstration of an increase in ST between 1 and 4 years, although with similar rates of death and MI.⁷ Since then, a number of large registry series, as well as limited randomized trials, have reported the benefits versus risks of DES, particularly in “off-label” indications such as STEMI, and most still found similar or lower rates of MI and death. A more recent meta-analysis of over 38,000 patients from 13 RCTs and 18 observational studies also demonstrated similar death and MI rates through 2 years of follow-up.⁸ Still, the data were not completely consistent, and difficulty weeding through biases inherent to one observational study, or limitations in statistical power of randomized trials for these rare, devastating events, created a somewhat muddy landscape. Larger, more definitive, trials and registries, with longer follow-up periods, particularly in STEMI, were needed and highly anticipated.

HORIZONS-AMI promised to be one such study, and in many respects, the present study is another. The HORIZONS-AMI trial randomized 3,602 patients to unfractionated heparin (UFH) plus a GPI or bivalirudin monotherapy; after angiography, 3,006 patients were then randomized again to either paclitaxel-eluting stents (n = 2257) or BMS (n = 749).⁹ Surprisingly, among all stented patients in this STEMI population, the rates of definite/probable ST were higher than previously reported, but also the late and very late ST rates were nearly identical by stent type (DES versus BMS: acute ST [0.9% versus 0.6%], subacute ST [1.3% versus 2.1%], late ST [1.1% versus 0.7%], and very late ST [1.2% versus 1.0%]). The present study by Pedersen et al gives us insight into whether these randomized results still apply in the “real world” setting, and at relatively long follow-up times of up to 4 years. When comparing this randomized comparison of stent use to the observational, operator-choice for stent type, driven data from the present study, remarkably, the two share many findings. There were similar definite ST rates, though more late ST in DES (DES versus BMS: early ST [0.8% versus 0.7%; p = 1.0], late ST [1.3% versus 0.5%; p = 0.20], very late ST [1.5% versus 0.2%; p = 0.03]). Probable ST rates (7.2% BMS versus 6.5% DES; p = 0.59) were also higher. In many ways, the present study demonstrates that it is possible to mirror key aspects of a well-conducted RCT, including exceptionally high rates of follow-up, and limited confounders due to exclusive use of pPCI in Denmark, high-volume operators, and consistent adjunctive pharmacology. Still, in the present study, unlike HORIZONS-AMI, operators chose which stent type to implant, presumably on a complex set of clinical and angiographic considerations, and thus this confounding by indication cannot be fully eliminated. For instance, not surprisingly, DES were used much more frequently in complex lesions, perhaps another independent risk factor for late events. Confounding by indication for DES may well have contributed to higher unadjusted rates of ST in DES patients in this study. In fact, when adjusted for differences in baseline characteristics, there was no clear hazard to DES, with a hazard ratio of over 1 to DES but with fairly wide confidence intervals crossing to include a risk of less than 1.

Finally, adjunctive pharmacology clearly mitigates risk of DES-associated ST, and this is even more critical in STEMI. In fact, the crucial role of prolonged antiplatelet therapy was established early, after the 2006 BASKET-LATE trial demonstrated significantly higher cardiac death and MI rates in DES after clopidogrel was discontinued at 6 months.^10,11 As well, not surprisingly, 600 mg use of clopidogrel predicted lower acute ST rates in HORIZONS-AMI. In the present study, clopidogrel was used routinely for 12 months, with adherence in a remarkably high 90% of patients, and a loading dose of 300 or 600 mg. It is possible that improved adjunctive pharmacology in both of these trials may help explain the apparent disconnect between previously discussed mechanisms of action and outcomes data, and outcomes may have been even further improved with higher loading doses of clopidogrel in all patients. Certain immediate adjunctive anticoagulation choices may mitigate early ST. In HORIZONS-AMI, unfractionated heparin (UFH) use predicted lower acute ST rates compared with bivalirudin. Similarly, the operators in Denmark in the present study used 10,000 units of UFH, although with lower glycoprotein IIb/IIIa use. The RCT data from HORIZONS-AMI and the observational data of the present study raise a possible additional therapeutic consideration when using DES in STEMI, though this must be weighed against bleeding considerations.

Thus, when attempting to marry the mechanistic risks of DES in STEMI with the somewhat conflicted epidemiologic evidence, the study by Pedersen et al, when combined with recent RCT data, may help shape our newest opinions. Their findings further support the growing evidence that perhaps each device, BMS or DES, has its own set of differing risks that contribute to adverse events, both of which seem fairly similar in total. In addition, rates of late ST do in fact appear higher in STEMI patients, though whether only in DES or both DES and BMS is no longer clear. It is possible that the risk of clinically significant in-stent restenosis and TLR in BMS is similar to that of DES in conjunction with a smarter adjunctive pharmacology strategy. One could additionally hypothesize that the early benefits to DES are counterbalanced by late inflammatory pathway-mediated events. In all, with the requisite limitations of study designs, looking for a heightened DES risk in STEMI may seem like finding a “needle in a haystack.” One may suspect then, that this risk, while real, is unlikely significantly higher than for BMS, particularly when mindful of adjunctive pharmacology and stent implantation techniques. We may be left with using that knowledge as a scientific backdrop as we continue to practice on a case-by-case basis the decision to use DES versus BMS in STEMI.

References

1. Pedersen S, Galatius S, Mogelvang R, et al. Long-term clinical outcome in STEMI patients treated with primary PCI and drug-eluting or bare-metal stents — Insights from a high-volume single center registry. J Invasive Cardiol 2011;23:328–333.
2. Kandzari DE, Roe MT, Ohman EM, et al. Frequency, predictors, and outcomes of drug-eluting stent utilization in patients with high-risk non-ST segment elevation acute coronary syndromes. Am J Cardiol 2005;96:750–755.
3. Iakovou I, Schmidt T, Bonizzoni E, et al. Incidence, predictors, and outcome of thrombosis after successful implantation of drug-eluting stents. JAMA 2005;293:2126–2130.
4. Virmani R, Guagliumi G, Farb A, et al. Localized hypersensitivity and late coronary thrombosis secondary to a sirolimus-eluting stent: Should we be cautious? Circulation 2004;109:701–705.
5. Guagliumi G, Costa MA, Sirbu V, et al. Strut coverage and late malapposition with paclitaxel-eluting stents compared with bare-metal stents in acute myocardial infarction: Optical coherence tomography substudy of the Harmonizing Outcomes With Revascularization and Stents in Acute Myocardial Infarction (HORIZONS-AMI) trial. Circulation 2011;123:274–281.
6. Guo N, Maehara A, Mintz GS, et al. Incidence, mechanisms, predictors, and clinical impact of acute and late stent malapposition after primary intervention in patients with acute myocardial infarction: An intravascular ultrasound substudy of the Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction (HORIZONS-AMI) trial. Circulation 2010;122:1077–1084.
7. Stone GW, Moses JW, Ellis SG, et al. Safety and efficacy of sirolimus- and paclitaxel-eluting coronary stents. N Engl J Med 2007;356:998–1008.
8. Brar SS, Leon MB, Stone GW, et al. Use of drug-eluting stents in acute myocardial infarction: A systematic review and meta-analysis. J Am Coll Cardiol 2009;53:1677–1689.
9. Dangas GD, Caixeta A, Mehran R, et al. Frequency and predictors of stent thrombosis after percutaneous coronary intervention in acute myocardial infarction. Circulation 2011;123:1745–1756.
10. McFadden EP, Stabile E, Regar E, et al. Late thrombosis in drug-eluting coronary stents after discontinuation of antiplatelet therapy. Lancet 2004;364:1519–1521.
11. Pfisterer M, Brunner-La Rocca HP, Buser PT, et al., for the BASKET-LATE Investigators. Late clinical events after clopidogrel discontinuation may limit the benefit of drug-eluting stents. J Am Coll Cardiol 2006;48:2584–2591.

__________________________________

From Christiana Care Center for Heart & Vascular Health, Newark, Delaware.
Dr. Glaser discloses that she is principally funded by GlaxoSmithKline Research and Development.
Address for correspondence: Ruchira Glaser, MD, MSCE, FACC, Christiana Care Center for Heart & Vascular Health, 4755 Ogletown-Stanton Road, Newark, DE 19718. Email: ruglaser@gmail.com