Stent Scrutiny

Since the inception of balloon angioplasty in the early 1980s, the field of interventional cardiology has enjoyed explosive growth. It is estimated that more than 1.3 million percutaneous coronary revascularization procedures were performed worldwide in 1999, considerably more than the approximately 700,000 coronary artery bypass graft operations.¹ Important limitations of balloon angioplasty, however, include the risk of uncontrolled plaque disruption, leading to periprocedural coronary occlusion and myocardial infarction, and a 20% to 40% incidence of recurrent narrowing (restenosis) during the 6 to 12 months following successful revascularization.² A variety of new technologies have been developed to overcome these limitations, the most successful of which has been the coronary stent.

A stent is a coil-like metal scaffold implanted within the arterial lumen to provide radial support to the diseased vascular wall. The adoption of stenting into clinical practice was made possible by a crucial evolution in deployment technique and adjunctive pharmacologic therapy. Initial clinical experience was marked by prohibitively high subacute thrombosis rates.³ Intensive anticoagulation regimens including aspirin, dipyridamole, dextran40, intravenous heparin sodium, and warfarin sodium somewhat diminished the risk of occlusion,⁴ but resulted in frequent and often serious vascular and bleeding complications, prolonged hospitalizations, and increased costs. Thus, although early clinical data suggested that stents were effective in reversing coronary occlusions and reducing restenosis,^{5 - 6} few physicians routinely used these devices.

Evidence then emerged that thrombosis may arise primarily at sites of inadequate stent apposition against the arterial wall,⁷ prompting investigation of high-balloon pressure techniques of stent deployment.⁸ Randomized trials also demonstrated that enhanced antiplatelet therapy with aspirin and ticlopidine hydrochloride following high-pressure stent implantation produced remarkably lower rates of ischemic and hemorrhagic complications than traditional warfarin anticoagulation,⁹ findings that served as the impetus for acceptance of stenting as the dominant means of percutaneous coronary revascularization. The proportion of interventional cardiology procedures using stents has accordingly increased from only 10% to 15% in 1994 to nearly 80% this year, with projections of continued growth.¹

In this issue of THE JOURNAL, Suwaidi and colleagues¹⁰ provide a comprehensive review of the clinical data comparing stents with balloon angioplasty. During long-term follow-up in randomized trials, rates of repeat revascularization were reduced by 35% to 75% by stenting, coupled with 20% to 55% reductions in angiographic indices of restenosis. The settings for which benefits of stenting have been demonstrated are delineated in this article, as are the limitations of the current clinical data set for other indications. In a separate article, Mehilli and colleagues¹¹ analyzed the interaction of sex with stenting among more than 4000 consecutive patients treated at 2 institutions, demonstrating that clinical outcome at 1-year follow-up was the same in men and women. These 2 articles are important additions to the evidence supportive of stenting to prevent or treat abrupt coronary occlusion and limit the risk of restenosis.

Widespread acceptance of stenting has been driven not only by clinical data, however, but also by more subjective factors, including an appealing theoretical rationale and gratifying immediate angiographic results. Stenting is typically faster and less cumbersome to perform than alternative percutaneous techniques, with fewer catheter exchanges or prolonged balloon inflations. Angiographic and clinical results are more consistent than with balloon angioplasty, particularly for complex lesions. The technical challenge of stenting is usually less than that of atherectomy or other devices, engendering a sense of confidence (at times, misleading) among less experienced operators who may attempt to take on difficult cases.

Considerations such as these have contributed to an overwhelming enthusiasm for stenting, which has outpaced the accrual of rigorous supportive clinical trial results. Until recently, reductions in restenosis by stenting had been demonstrated among only highly select groups of patients,^{5 - 6} with extrapolation in clinical practice to patient and lesion subsets for which little or no objective data had been generated. While subsequent studies have substantiated the efficacy of stenting among patients with a broader spectrum of clinical syndromes and coronary lesion morphologies, stent technology continues to be applied to many settings in which important limitations exist.

Acute procedural results with stenting have improved considerably with newer low-profile and flexible stent designs. Nevertheless, stent deployment may still be problematic in some lesions. In tortuous or heavily calcified and inflexible vessels, for example, difficulties may be encountered in passing the stent to the lesion site, with consequent risks of embolization or incomplete expansion. Coronary stenoses at major vessel bifurcations continue to pose an appreciable challenge, in that stent placement into 1 branch in effect "jails" the origin of the other, with further compromise if plaque shifts and a "snowplow" effect occurs. Treatment of diffusely diseased vessels may result in plaque disruption at the ends of a stent, necessitating placement of multiple sequential stents. Revascularization of degenerated saphenous vein grafts is frequently complicated by extrusion of friable atheromatous material through the interstices of stents and distal coronary embolization.

Restenosis within a stent is a formidable problem and remains the principal factor limiting the long-term benefit of percutaneous coronary revascularization. Stents reduce restenosis relative to balloon angioplasty through improved immediate procedural outcome and prevention of elastic recoil and negative vascular remodeling. Yet restenosis within a stent can still occur by neointimal ingrowth between stent struts.¹² While this process is usually responsive to repeat balloon dilation, rates of recurrence (re-restenosis) have been as high as 60% to 80% in some series.¹³

Pharmacologic therapies have not yet been shown to reduce the incidence of in-stent restenosis, nor have atherectomy techniques. Intracoronary radiation appears to be effective in this regard,¹⁴ although this modality is technically cumbersome and long-term results and complications have yet to be defined. Potentially promising new approaches in early stages of clinical evaluation include coated stents, which elute antiproliferative agents such as rapamycin or paclitaxel.

Certain characteristics have been associated with a high risk for in-stent restenosis.¹⁵ Rates of target vessel revascularization in these settings may be 2 to 4 times greater than with stenoses meeting the restricted criteria of the original stent vs angioplasty trials. Unfortunately, lesion subsets at high risk for in-stent restenosis are also those that are most difficult to treat by balloon angioplasty. Diffuse atherosclerotic disease may require multiple or long stents, and the likelihood of restenosis is proportional to the length of the treated segment. Small vessel diameter is associated with more frequent restenosis; it is therefore notable that 2.5-mm diameter stents are frequently used in clinical practice, but have not been demonstrated to reduce restenosis. Patients with bifurcation disease are at high risk for restenosis after stenting, with the additional complication that the branch vessel may be relatively inaccessible through the side of the stent. Saphenous vein grafts have a rate of failure as high as 50% over 2 years or more following stent placement,¹⁶ reflecting diffuse degeneration throughout the graft, which is not modified by focal stent implantation. In addition, diabetes mellitus remains a potent risk factor for long-term ischemic complications after percutaneous revascularization, and stenting has not yet been convincingly shown to reduce this adverse risk profile.

In addition to the clinical considerations, the unrestrained growth of stenting has important economic implications. With an average of 1.3 to 1.8 stents used per procedure and a mean price of more than $1400 per stent, the worldwide stent market was estimated to be $2.2 billion in 1999.¹ Although stents have usually been shown to be relatively cost-effective over the long-term, the approximately $1900 incremental increase in procedural costs remains a burden to hospitals in the current environment of fixed or diminishing medical reimbursements.¹⁷

Enthusiasts of stenting may erroneously equate excellent angiographic results achieved by this technique with optimal clinical outcomes, thereby withholding effective new adjunctive pharmacologic therapies from patients with stents because they are under the misconception that no further improvements in outcome are possible. Importantly, clinical benefit observed with stenting has been confined to reductions in recurrent ischemia or the need for revascularization procedures. To date, no trial has shown a reduction by stents of the risk of death or myocardial infarction. In fact, trends toward higher rates of long-term mortality have been observed with stenting compared with balloon angioplasty during acute myocardial infarction.¹⁸

Clinical outcome in patients receiving stents can be improved considerably, however, by intensified adjunctive antiplatelet therapy with glycoprotein IIb/IIIa inhibitors. Periprocedural administration of abciximab has been associated with 50% to 60% reductions in the risk of acute ischemic complications and long-term (1 year) mortality following stenting.^{17 ,19} Moreover, glycoprotein IIb/IIIa blockade during stenting for acute myocardial infarction improves indices of microvascular flow and tissue level reperfusion.²⁰ Nevertheless, despite consistent findings of clinical efficacy and safety with these agents, as well as incremental cost and cost-effectiveness comparable with that of stenting, pharmaceutical industry data show that only 30% to 40% of patients undergoing a percutaneous revascularization procedure receive one of these agents. Antiembolization devices represent another approach, currently under early investigation, to improving clinical outcomes with stenting.

Stents are the most important nonpharmacologic advance in interventional cardiology since the introduction of balloon angioplasty and have had a profound impact on the clinical efficacy of percutaneous coronary revascularization. Notwithstanding the demonstrated benefits of stenting, optimal application of this technology to clinical practice demands critical understanding of its limitations, ongoing assessment of outcome in high-risk settings, refined understanding of appropriate indications for the procedure, continued improvement in stent and catheter designs, recognition of the complementarity with enhanced adjunctive pharmacologic strategies, and development of effective means of preventing in-stent restenosis.