Comparison of a Polymer-Based Paclitaxel-Eluting Stent With a Bare Metal Stent in Patients With Complex Coronary Artery Disease:  A Randomized Controlled Trial FREE

Drug-eluting stents have revolutionized the treatment of atherosclerotic coronary artery disease. The demonstration that both sirolimus-eluting and paclitaxel-eluting stents safely reduce clinical and angiographic restenosis compared with bare metal stents^1- 4 has resulted in an estimated 160 000 of these devices currently being implanted per month worldwide (unpublished data, Boston Scientific Corp, Natick, Mass), signifying drug-eluting stents as by far the most widely used permanent bioprosthesis. Enrollment in the pivotal randomized trials,^1- 4 however, was restricted to relatively simple stenoses (vessel diameter of 2.5-3.75 mm with lesion length ≤30 mm). More than 55% of lesions currently treated with these bioactive devices may fall outside this range.⁵ The efficacy of drug-eluting stents has not been established for small vessels (in which the utility of stents as a class is still uncertain),⁶ large vessels (in which outcomes with bare metal stents are favorable),^7- 11 or in long lesions requiring multiple stents (a complex subset with increased periprocedural complications and reduced efficacy).^7- 12 Therefore, we performed a prospective, multicenter, randomized trial to investigate the safety and efficacy of a paclitaxel-eluting stent in a patient population with more complex lesions than previously studied.

Patients aged 18 years or older with stable or unstable angina or provokable ischemia undergoing percutaneous coronary intervention of a single de novo lesion in a native coronary artery were considered for enrollment. Clinical exclusion criteria included previous or planned use of intravascular brachytherapy or any antirestenotic drug-eluting stent in the target vessel; myocardial infarction (MI) within 72 hours or creatine kinase-MB level higher than 2 times the upper limit of normal; left ventricular ejection fraction of less than 25%; stroke within 6 months; planned coronary artery bypass graft surgery within 9 months; hemorrhagic diatheses or contraindications or allergy to aspirin, thienopyridines, paclitaxel, stainless steel, or anaphylaxis to iodinated contrast; chemotherapy within 12 months, or planned use of paclitaxel, rapamycin, or colchicine within 9 months; serum creatinine level higher than 2.0 mg/dL (>176.8 μmol/L), leukocyte count lower than 3.5 × 10⁹/L, or platelet count lower than 100 000/mm³ or higher than 750 000/mm³; female with a recent positive pregnancy test, lactating, or planned procreation within 3 months; comorbid conditions limiting life expectancy to less than 24 months or that could affect protocol compliance; planned procedure requiring antiplatelet therapy withdrawal within 6 months; and current participation in other investigational trials. The study was approved by the institutional review boards at each participating center, and consecutive, eligible patients signed informed, written consent.

Angiographic eligibility required a single target lesion with visual reference vessel diameter of 2.25 to 4.0 mm and lesion length of 10 to 46 mm. Exclusion criteria included left main or ostial lesion; excessive vessel or lesion calcification, tortuosity or angulation; bifurcation disease; target lesion occlusion or thrombus; and planned atherectomy. Enrollment was permitted after successful treatment of 1 to 2 nonstudy lesions in a nonstudy vessel prior to randomization.

To enrich the study population with complex lesions, the protocol specified randomizing 200 or more patients requiring 2.25-mm stents; 200 or more patients requiring 4.0-mm stents; and 300 or more patients with lesions longer than 26 mm in length requiring multiple overlapping stents. The protocol likewise specified the inclusion of no more than 350 lesions with diameters equal to or larger than 2.5 mm and less than or equal to 3.5 mm and with a lesion length of 26 mm or less requiring only a single stent. For purposes of the present analysis, lesions requiring 2.25-mm stents, 4.0-mm stents, or multiple stents, or combinations thereof, were defined as complex or previously unstudied.

Telephone randomization was performed after mandatory predilatation in random blocks of 2 to 4 patients, stratified by presence or absence of medically treated diabetes and by lesion length (<18 or ≥18 mm). Patients were assigned using random serial numbers to treatment with either the slow rate-release polymer-based paclitaxel-eluting TAXUS stent or a visually indistinguishable bare metal Express² stent (both Boston Scientific Corp). Stents were available in 8- to 32-mm lengths and in diameters of 2.25 to 4.0 mm. Coverage of 3 mm of the normal reference segment at both the proximal and distal lesion margins was recommended. When multiple stents were required, 4 mm of stent overlap was specified. Antithrombin and glycoprotein IIb/IIIa inhibitor use and performance of postdilatation were at operator discretion.

Patients were given 325 mg/d of aspirin prior to stent implantation and were required to continue taking this dose indefinitely. A 300-mg loading dose of clopidogrel administered more than 6 hours prior to the procedure, followed by a dose of 75 mg/d for at least 6 months was recommended. Clinical follow-up was scheduled at 1, 4, and 9 months and yearly thereafter for 5 years. Follow-up angiography was scheduled in all patients at 9 months.

Independent study monitors verified all data from the case report forms onsite. Major adverse cardiac events were adjudicated by an independent committee blinded to treatment allocation after review of original source documentation. The clinical and angiographic endpoint definitions were identical to those in TAXIS IV, as previously described.⁴ A data and safety monitoring committee periodically reviewed safety data; the committee recommended that the study continue without modification at each review period. Blinded core angiographic laboratory analysis was performed using validated quantitative methods.¹³ Measures were reported separately within the stent, within 5-mm proximal and distal to each edge, and over the entire analysis segment. The investigators had unrestricted access to the database. The manuscript was prepared by the principal investigator (G.W.S.) and revised after coauthor review.

The primary end point was the 9-month incidence of ischemia-driven target vessel revascularization. Enrolling 1172 patients and allowing for 10% attrition and using a 2-sided test for differences in independent binomial proportions with an α level of .05 yielded 90% power to detect a reduction in the primary end point from an anticipated 18% after implantation with bare metal stents to 10.8% after implantation with paclitaxel-eluting stents. For the major secondary end point of follow-up angiographic diameter stenosis, allowing for 25% attrition for noncompliance or technical failures with an expected mean (SD) diameter stenosis in the control group of 32.6% (16.9%), the minimum detectable difference with 80% power afforded was 3.3%, which is a 10.1% change relative to control.

Categorical variables were compared using the Fisher exact test. Continuous variables are presented as mean ±1 SD and were compared using the t test. The influence of baseline variables on 9-month categorical end points was evaluated with logistic regression using the Wald χ² test; all baseline clinical and angiographic features, randomization assignment, and procedural parameters were entered. The statistical analysis plan specified that the primary intent-to-treat population would consist of all consenting patients in whom an attempt was made to implant a study stent. All P values are 2-sided. All statistical analyses were performed using SAS software version 8.2 (SAS Institute Inc, Cary, NC).

Between February 27, 2003, and March 29, 2004, a total of 1172 patients at 66 US centers were randomly assigned to receive either paclitaxel-eluting stents (n = 586) or bare metal stents (n = 586) (Figure 1). Sixteen patients (1.4%) in whom no attempt was made to implant a study stent were excluded. Therefore, the analysis population consisted of 1156 patients: 577 assigned to paclitaxel-eluting stents and 579 to bare metal stents (control). Baseline characteristics appear in Table 1. Diabetes mellitus was present in 31% of patients and 78% of lesions were American College of Cardiology/American Heart Association class type B₂/C. Initial procedural results appear in Table 2. Of the 1156 lesions analyzed, 664 (57.4%) were complex or previously unstudied as defined above. Stents of 2.25 mm and 4.0 mm in diameter were used in 18% and 17% of lesions, respectively; multiple stents were used in 33% of lesions.

Clinical follow-up was available in 1127 patients at 9 months (97.5%) and follow-up angiography at 9 months was completed in 990 patients (85.6%). Compared with bare metal stents, implantation of paclitaxel-eluting stents reduced the 9-month rate of target lesion revascularization from 15.7% to 8.6% (P<.001) and target vessel revascularization from 17.3% to 12.1% (P = .02) (Table 3). Among patients receiving the paclitaxel-eluting stent compared with a bare metal stent, the rate of in-stent restenosis was reduced from 31.9% to 13.7% and analysis segment angiographic restenosis was reduced from 33.9% to 18.9% (both P<.001; Figure 2). Restenosis was usually focal with the paclitaxel-eluting stent but was typically diffuse or proliferative with bare metal stents. By multivariate analysis, randomization to the paclitaxel-eluting stent was an independent predictor of freedom from 9-month target lesion revascularization (odds ratio [OR], 2.23; 95% confidence interval [CI], 1.49-3.34; P<.001), target vessel revascularization (OR, 1.66; 95% CI, 1.16-2.39; P = .006), and restenosis (OR, 2.89; 95% CI, 2.07-4.05; P<.001). These benefits were achieved with comparable safety in both groups, with similar rates of cardiac death, MI, and stent thrombosis at 1 and 9 months. Although uncommon, a statistically nonsignificant trend toward greater late-acquired aneurysm formation was seen in patients who received paclitaxel-eluting stents (Table 3).

A total of 664 patients had complex or previously unstudied lesions (requiring 2.25-mm, 4.0-mm, or multiple stents, or some combination thereof). Clinical and angiographic follow-up were available for 648 (97.6%) and 582 (87.7%) of these patients. The use of paclitaxel-eluting stents compared with bare metal stents in this group reduced the rate of target lesion and vessel revascularization from 19.0% to 9.9% (P = .001) and from 21.2% to 13.9% (P = .02), respectively (Table 4). A test for statistical heterogeneity among the 3 subsets of complex lesions (2.25-mm stents, 4.0-mm stents, and multiple stents) was negative, confirming that the beneficial effect of the paclitaxel-eluting stent on target vessel revascularization was consistent for all 3 types of complex lesions. The rate of angiographic restenosis was also significantly reduced for all complex lesion subsets (Table 4 and Table 5).

Clinical features and results in the individual complex and previously unstudied lesion subsets appear in Table 5. Important baseline characteristics were equally distributed between the 2 study groups with the exception of diabetes mellitus, which was present in a larger proportion of patients receiving the 2.25-mm paclitaxel-eluting stent. As would be anticipated, the rates of clinical and angiographic restenosis were high in patients receiving 2.25-mm stents and for those receiving multiple stents and low among patients receiving 4.0-mm stents. Subgroup analysis did not show a significant reduction in target vessel revascularization with the paclitaxel-eluting stent among patients receiving 2.25-mm or 4.0-mm stents (probably a consequence of the relatively small numbers of patients in these subgroups), although target lesion revascularization and angiographic restenosis rates were reduced.

Multiple stents were implanted in 379 patients (planned in 281 and “bail out” for complications or suboptimal results in 98), in whom the mean (SD) lesion and stent length were 25.3 (10.0) mm and 43.9 (10.1) mm, respectively. In patients requiring multiple stents, paclitaxel-eluting stent use was associated with an increased incidence of MI at 30 days (8.3% vs 3.3%; P = .047), most of which were non–Q-wave MIs (Table 5). Infarction rates were numerically increased with both planned (6.3% vs 2.9%) and unplanned (14.0% vs 4.2%) use of multiple paclitaxel-eluting stents. Blinded core laboratory angiographic analysis in the multiple stent cohort demonstrated more frequent occurrence of progressive side-branch narrowing to more than 70% diameter stenosis or to total occlusion with the paclitaxel-eluting stent (42.6% vs 30.6%; P = .03) and a greater likelihood of reduced side-branch thrombolysis in MI flow (41.9% vs 28.6%; P = .02). No differences were present in main vessel thrombolysis in MI flow, acute occlusion, stent thrombosis, distal embolization, or other angiographic complications. Reductions in clinical and angiographic restenosis rates were present at 9 months in patients who received multiple stents and who had been assigned to paclitaxel-eluting rather than bare metal stents, with similar rates of cardiac death, MI, and stent thrombosis.

This clinical trial (TAXUS V) evaluated the use of paclitaxel-eluting stents compared with bare metal stents in a patient cohort significantly more complex than previously studied. In comparison with the prior pivotal trials of paclitaxel-eluting and sirolimus-eluting stents, lesion length in the current trial was substantially greater (17.2 mm vs 13.4 mm and 14.4 mm, respectively), more lesions were type B₂/C (78% vs 53% and 56%), more patients had diabetes (31% vs 24% and 26%), and 2.25-mm and 4.0-mm stents had not been evaluated.^2,4

In the entire study cohort, implantation of the polymer-based, slow-release paclitaxel-eluting stent was associated with similar rates of death, MI, and stent thrombosis at 1 and 9 months compared with an otherwise identical bare metal stent. Rates of clinical and angiographic restenosis were significantly reduced with the paclitaxel-eluting stent. Of note, however, the absolute and relative efficacy of the paclitaxel-eluting stent in this trial were somewhat diminished compared with earlier studies in patients with less complex lesions.^3- 4 However, the upper bound of the 30% observed relative reduction in the primary end point of target vessel revascularization with the paclitaxel-eluting stent was 47%, which is greater than the 40% improvement anticipated, demonstrating that a significantly larger trial would have been required for a more accurate point estimate.

In this trial, we enrolled several lesion subgroups that had not previously been studied with drug-eluting stents but are commonly treated in clinical practice. Whether stent implantation improves outcomes compared with balloon angioplasty alone in small vessels is still a matter of debate.^6- 11 In prior studies, clinical and angiographic outcomes with 4.0-mm bare metal stents have been favorable,^7- 11 suggesting that a large vessel drug-eluting stent may not offer significant incremental value. Finally, long lesions requiring multiple stents represent a unique challenge, with increased periprocedural complications and reduced efficacy.^7- 12

Nonetheless, the rates of both clinical and angiographic restenosis were significantly reduced with the paclitaxel-eluting stent compared with the bare metal stent in the 664 patients with complex or previously unstudied lesions as defined by these criteria. A test for heterogeneity demonstrated that the benefit of paclitaxel stents over bare metal stents was consistent across all 3 subsets.

With a mean (SD) diameter of 2.08 (0.32) mm, vessels receiving 2.25-mm stents represent the smallest coronary arteries studied to date in patients in a randomized drug-eluting stent trial. Compared with patients who received bare metal stents, restenosis rates were safely reduced with paclitaxel-eluting 2.25-mm stents, although angiographic restenosis and target lesion revascularization still occurred in approximately 30% and approximately 10% of vessels, respectively. Target vessel revascularization at sites remote from the target lesion was also required in nearly 9% of 2.25-mm stent patients (compared with 1.4% in the simpler lesion cohort in TAXUS IV⁴), reflecting the diffuse and progressive nature of coronary atherosclerosis in these patients. Whether stents that further suppress the arterial response to injury may improve outcomes in such small vessels is unknown. In this regard, sirolimus-eluting stents were found in a randomized trial to markedly reduce clinical and angiographic restenosis in small coronary arteries.¹⁵ However, the lesions treated were mostly focal (mean [SD] length, 11.8 [6.2] mm) in this study and coverable with a single stent as large as 2.75 mm in diameter, making direct comparison with the present study difficult.

In prior studies, clinical and angiographic outcomes with 4.0-mm bare metal stents have been favorable,^7- 11 questioning the need for a large vessel drug-eluting stent. Indeed, in the present study, target lesion revascularization was required in only 5% of patients receiving 4.0-mm bare metal stents, and angiographic restenosis occurred in only 14.4% of lesions. Nonetheless, paclitaxel-eluting stents further reduced angiographic restenosis in this cohort by an additional 76% (to 3.5%), and no patient who received a 4.0-mm stent required target lesion revascularization or developed stent thrombosis within 9 months. As coronary vessels accommodating 4.0-mm stents typically supply a large amount of myocardium, minimizing restenosis in this lesion subset is desirable. However, further studies are required to determine the cost-effectiveness of drug-eluting stents in large vessels.

Lesions requiring multiple stents for either planned or bailout use represented the most challenging subgroup enrolled in TAXUS V with an implanted stent length of more than 43 mm. In this cohort, paclitaxel-eluting stent assignment was associated with increased periprocedural myonecrosis due mostly to an excess of non–Q-wave MIs. Detailed angiographic analysis demonstrated the likely cause to be greater side-branch compromise with paclitaxel-eluting stents compared with bare metal stents. While the mechanism(s) underlying this observation remain speculative, possibilities include side-branch narrowing by thicker polymer-coated stent struts (possibly exacerbated by polymer webbing or clumping), transient platelet or thrombus deposition, or paclitaxel-induced spasm. Porcine studies with overlapping paclitaxel-eluting stents demonstrated delayed reendothelialization without incremental toxicity (unpublished data, Boston Scientific Corp). Ongoing analyses are attempting to address whether side-branch compromise is more frequent in the overlap zone or alternatively reflects implantation of long stents in diffusely diseased vessels. These data highlight the desirability of preserving major side-branch patency after drug-eluting stent implantation to prevent periprocedural MI. However, the prognosis following side-branch occlusion complicating stent implantation is in general favorable^16- 17 and excessive rates of death or stent thrombosis were not evident in patients treated with multiple paclitaxel-eluting stents. Moreover, among patients treated with multiple bare metal stents, angiographic and clinical restenosis by 9 months occurred in nearly 60% and 30%, respectively, remarkably high failure rates that were reduced by more than half with multiple paclitaxel-eluting stents. However, larger studies are required to determine whether the increase in periprocedural MI is associated with an increase in late mortality.

Several limitations of the present study should be acknowledged. First, the trial was powered to demonstrate a reduction in target vessel revascularization for the entire study population as a whole. Inferences from underpowered subset analyses should be considered hypothesis-generating only. However, pooled analysis of the cohort with complex and previously unstudied lesions demonstrated a significant reduction in the primary end point of target vessel revascularization as well as target lesion revascularization and angiographic restenosis with the paclitaxel-eluting stent, with no statistical heterogeneity between lesion subgroups, suggesting that the efficacy of the TAXUS stent in enhancing freedom from clinical and angiographic restenosis extends to these subgroups as well. However, the trial was underpowered to definitively examine low-frequency adverse events such as death and stent thrombosis, especially in subsets. Second, although the clinical follow-up rate in this study was quite high (97.5%), the results may have varied if no patients had withdrawn or been lost to follow-up. In a sensitivity analysis performed to evaluate the potential impact of these patients on the primary end point of target vessel revascularization, the results were generally unchanged and the treatment effect was robust. In addition, given the lack of complete angiographic follow-up, the restenosis rates reported in the present study likely overestimate the true frequency of recurrence¹⁸ (although the 85.6% rate of follow-up angiography achieved is high for a US-based investigation). Moreover, the performance of routine angiographic follow-up in most patients also likely resulted in some cases of repeat angioplasty based on the “oculostenotic reflex,”¹⁸ artificially elevating the reported rates of target lesion and vessel revascularization (despite attempts to adjudicate these occurrences). However, in a prior large-scale blinded investigation of paclitaxel-eluting and bare metal stents, this phenomenon affected both the active and control stent event rates to an equal degree,⁴ suggesting that any bias in the present study would be equally distributed between the 2 groups. Third, further trials are also required to address patients and lesions excluded from the present study, such as acute MI, major bifurcation involvement, and diseased saphenous vein grafts. In addition, the long-term efficacy of drug-eluting stents compared with coronary artery bypass graft surgery is unknown, with conflicting data regarding late freedom from adverse events recently reported in patients undergoing multivessel drug-eluting stent implantation.^19- 20 Finally, the results of the present trial apply only to the polymer-based, slow-release paclitaxel-eluting stent. Comparative trials with the sirolimus-eluting and emerging drug-eluting stents are required to determine the optimal platform for specific patient and lesion subtypes.

In conclusion, the TAXUS V trial investigated the use of paclitaxel-eluting stents in a patient population with more complex lesions than had been previously studied. Angiographic restenosis and target vessel revascularization were significantly reduced in the entire cohort, as well as in those patients with complex disease. Patients receiving multiple paclitaxel-eluting stents had a significant increase in the rate of early non–Q-wave MI, likely due to increased rates of side branch compromise. Although the long-term significance of this observation is unknown, treatment of long lesions with multiple paclitaxel-eluting stents compared with bare metal stents in the present study resulted in a marked reduction in clinical and angiographic restenosis with similar late survival.