Stenting Small Coronary Arteries

Abstract

INTERVENTIONAL CARDIOLOGY HAS A LONG AND RICH TRAdition of randomized clinical trials, the results of which have dramatically improved patient care. In the past few years, attention has particularly focused on drugeluting stents, which have quickly become predicate devices (ie, against which new stents are compared). In the case of sirolimus-eluting stents, device approval was based in large measure on the 2 initial randomized clinical trials of RAVEL and SIRIUS. Based on the dramatic improvement in reducing restenosis with these devices demonstrated in these 2 trials, patients and physicians alike have embraced this new technology and physicians have used drug-eluting stents in subsets of patients for whom the data from trials were very limited. Early randomized trials of any device or drug typically target restricted “ideal” patient groups; this was certainly the case with the early drugeluting stent trials. An important subset of patients that were not the focus of these early trials were those with clinically important stenoses of small coronary arteries. The randomized trial by Ardissino and colleagues published in this issue of JAMA serves to fill in important gaps. This multicenter, single-blind trial randomly assigned patients with de novo native coronary arterial lesions in vessels of size 2.75 mm or less to either a sirolimus-eluting stent or a bare-metal stent. The primary end point was 8-month binary in-segment restenosis as assessed by quantitative coronary angiography while secondary end points included procedural success and 8-month major adverse cardiac and cerebrovascular events. The authors hypothesized that they would demonstrate a 66% reduction in restenosis from 30% to 10% and the sample size was calculated accordingly. Two hundred sixty patients were enrolled; 3 were not randomized because of exclusion criteria that became apparent only after enrollment so that 129 patients were enrolled in the sirolimus-eluting stent group and 128 in the bare-metal stent group. There were several important differences between the 2 groups, which could have affected subsequent outcome. The sirolimus-eluting stent group had fewer patients with diabetes mellitus (9% vs 30%, P=.055) a known risk factor for increased restenosis and more patients with acute coronary syndromes (49% vs 36%). In addition, patients in the sirolimus-eluting stent group had longer lesions (13 vs 11 mm) and required the use of longer stents (17 vs 15 mm). Three important aspects of trial design deserve mention. First, the authors did achieve their goal of treating small vessels with a mean (SD) reference diameter of only 2.2 (0.28) mm. Second, mandatory balloon dilatation was required before stent placement. There has been a suggestion that this may result in increased in-segment restenosis because of damage to the arterial wall during predilatation, which is not completely covered by the drug-eluting stent. And third, administration of glycoprotein IIb/IIIa antagonists was encouraged but was only used in 7.4% of the patients. Procedural success was excellent in both groups; more importantly for this angiographic trial, the immediate postprocedural minimal lumen diameter in both the insegment zone and the in-stent zone was not different. Inhospital clinical outcomes including death, myocardial infarction, target lesion revascularization, cerebrovascular accident, and stent thrombosis, as well as 8-month mortality were virtually identical in both groups. Patients receiving sirolimus-eluting stents had fewer myocardial infarctions and fewer episodes of in-stent thrombosis; however, the overall numbers of these clinical events were small and future study involving larger trial cohorts will be needed. The most important finding in the trial was the primary end point with dramatic reductions in long-term percentage diameter stenosis, minimal lumen diameter, late lumen loss, and late loss index in both the in-segment and the in-stent zones in patients who were randomized to sirolimus-eluting stents. The late lumen loss at 0.16 mm was similar to that observed in the SIRIUS trial. The frequency of binary in-segment restenosis was 9.8% in patients receiving a sirolimus-eluting stent vs 53.1% in patients receiving an uncoated stent, although the frequency of binary instent restenosis was 4.9% and 49.1%, respectively. This marked improvement in angiographic outcome was mirrored in the clinical outcome with target lesion revascularization performed in only 7% of patients treated with the