Due to their remarkable reduction of the need for subsequent revascularization, and yet despite the fact that drug-eluting stents (DES) have not been shown to improve long-term patient survival or freedom from myocardial infarction (MI), the use of DES has been very warmly embraced by the interventional community. Most U.S. centers report utilization in approximately 90% of their stentreceiving patients. For the physician in the relatively costunconstrained environment, the principal question then becomes “For whom are DES contraindicated on the basis of increased risk?” For the physician operating in an environment where there is more focus on cost, an additional important question is “For whom are DES relatively contraindicated due to marginal ‘value’?”

The field of DES continues to evolve. Despite several moderate-sized randomized trials and registries, important questions germane to the question of who should receive DES remain. Several reports now highlight the small but significantly increased hazard of late stent thrombosis with the utilization of DES compared to bare metal stents (BMS). The absolute increase appears to be about 0.5% between 6 and 24 months after stenting (1). This seems to occur principally during periods of antiplatelet therapy withdrawal, and especially during the stress of surgery. The exact risk for such patients is difficult to ascertain at present, due to our inexact knowledge of precisely which medications patients are taking long after enrollment into randomized trials, but the risk may be as high as 5% with dual antiplatelet therapy withdrawal for periods of a week or more in conjunction with the stress of noncardiac surgery (2). Whether or not both CYPHER and TAXUS stents pose equal risk has not been clarified yet. Early reports suggest also that patients resistant to aspirin, clopidogrel, or both also are at increased risk (3). If validated, and particularly if these tests can be standardized, these reports suggest that patients considered for elective DES use should have resistance to antiplatelet medications tested before stent implantation. These populations represent, then, the principal group for whom DES may well be contraindicated. To this group, of course, must be added those occasional patients who are truly allergic to either sirolimus or paclitaxel.

The cost-effectiveness of DES relates principally to the absolute magnitude of reduction in clinical restenosis. Generally speaking, DES reduce restenosis to a relatively equal extent across various patient populations; hence absolute reduction in risk is directly related to initial risk with BMS. This is known to be particularly high in patients with diabetes, long lesions, small reference dimensions, previous restenosis at the treatment site, and saphenous vein graft (SVG) intervention. Conversely, in the absence of these findings, the relative benefit is diminished. This can be seen graphically in the target vessel revascularization (TVR) rate versus lesion length/reference vessel diameter graft from the Taxus series (4) (Fig. 15A.1) and more formally in the work from Cohen (5) depicted in Figure 15A.2.

Another group of patients for whom DES may be relatively contraindicated are those patients who, by virtue of complex intervention and the location of the lesion, may be at greater risk of subacute thrombosis (SAT), with catastrophic results. This would include, in particular, patients undergoing “crush” or other high metal/polymer/drug density bifurcation intervention in the distal left main
(when that vessel is unprotected) and in the proximal left anterior descending (LAD)/large diagonal setting. The approximate doubling of SAT risk in this setting must be balanced against the benefit of reducing restenosis.

In summary, the physician operator should generally elect to utilize DES in preference to BMS except when the risk of stent thrombosis is particularly high or when the relative clinical benefit with regard to clinical restenosis reduction is marginal. We have considerably more data regarding the latter issue than the former. Whether we can better identify patients at heightened SAT risk utilizing aspirin or clopidogrel resistance, or perhaps specific forms of genetic testing, remains to be ascertained.

Restenosis is not a benign condition. Although it does not carry an increased mortality risk (1), recurrent angina and repeat revascularization resulting from restenosis decrease quality of life and erode patient satisfaction, while adding significantly to health care expenditure (2). It has persisted as the Achilles’ heel of percutaneous coronary intervention (PCI), affecting about one-third of all PCI patients (3), and limiting the application of PCI in lesion subsets with unacceptably high rates of restenosis. Despite a long history of failed innovations to prevent restenosis, the recent development of drug-eluting stents (DES) has revolutionized the application of PCI by achieving previously unseen reductions in the rate of restenosis (4,5). The resulting paradigm shift has been compared to the transition from balloon angioplasty (BA) to coronary stenting in terms of its magnitude (6). However, the universal application of DES is currently limited by high DES cost (7) and lack of published evidence in so-called “real world” patient cohorts and complex lesion subsets (8). Because of these limitations, the selective use of DES is currently advocated by some authors (9).

The purpose of this review is to discuss patient and lesion subsets likely to benefit or be harmed by DES, and to provide a framework for rational patient selection for
DES use. A number of different stent platforms have been compared to bare metal stenting (BMS) in randomized, controlled trials; however, only sirolimus- (Cypher, Cordis) and paclitaxel- (Taxus, Boston Scientific) eluting stents have achieved U.S. Food and Drug Administration (FDA) approval and widespread use. This review focuses only on peer-reviewed publications regarding paclitaxeland sirolimus-eluting stents.