Author + information
- Received February 25, 2010
- Revision received March 30, 2010
- Accepted April 7, 2010
- Published online July 20, 2010.
- Dean J. Kereiakes, MD*,* (, )
- Louis A. Cannon, MD†,
- Robert L. Feldman, MD‡,
- Jeffrey J. Popma, MD§,
- Raymond Magorien, MD∥,
- Robert Whitbourn, MBBS¶,
- Ira M. Dauber, MD#,
- Abram C. Rabinowitz, MD**,
- Michael W. Ball, MD††,
- Barry Bertolet, MD‡‡,
- Ameer Kabour, MD§§,
- Michael C. Foster, MD∥∥,
- John C. Wang, MD¶¶,
- Paul Underwood, MD## and
- Keith D. Dawkins, MD##
- ↵*Reprint requests and correspondence:
Dr. Dean J. Kereiakes, The Christ Hospital Heart and Vascular Center and The Carl and Edyth Linder Center for Research and Education at The Christ Hospital, 2123 Auburn Avenue, Suite 424, Cincinnati, Ohio 45219
Objectives The aim of this study was to evaluate the safety and efficacy of the novel platinum chromium TAXUS Element paclitaxel-eluting stent (PES) compared with the TAXUS Express PES (Boston Scientific, Natick, Massachusetts) in treating coronary artery stenoses.
Background The TAXUS Element is a novel thin-strut (81 μm), platinum chromium alloy PES designed to improve radial strength, radiopacity, and deliverability, while safely providing comparable restenosis benefit compared with a previous-generation PES.
Methods The PERSEUS (Prospective Evaluation in a Randomized Trial of the Safety and Efficacy of the Use of the TAXUS Element Paclitaxel-Eluting Coronary Stent System) Workhorse (WH) trial is a prospective, randomized (3:1), controlled, multicenter study of the TAXUS Element (vs. TAXUS Express) PES for the treatment of de novo coronary atherosclerotic lesions ≤28 mm in length in reference vessels ≥2.75 to ≤4.0 mm in diameter. The primary end point was the 12-month rate of target lesion failure, and the secondary end point was 9-month angiographic in-segment percentage diameter stenosis. The study was powered to demonstrate noninferiority to TAXUS Express for both end points.
Results The intent-to-treat analysis included 1,262 patients (320 TAXUS Express, 942 TAXUS Element). The TAXUS Element was noninferior to TAXUS Express with respect to both the incidence of target lesion failure (5.57% vs. 6.14%, respectively; difference: 0.57%; 95% credible interval: 1.85%; Bayesian posterior probability of noninferiority = 0.9996) and percentage diameter stenosis (ln[%DS] 3.09 vs. 3.12, respectively; difference: 0.03; 95% credible interval: 0.11; Bayesian posterior probability of noninferiority = 0.9970). No differences in clinical outcomes to 12 months were observed between stent treatments, and stent thrombosis was infrequent (0.3% Express, 0.4% Element).
Conclusions At 1 year, the TAXUS Element is comparable in efficacy to the TAXUS Express PES. Furthermore, no safety concerns related to the novel platinum chromium alloy or stent design were observed. (A Prospective Evaluation in a Randomized Trial of the Safety and Efficacy of the Use of the TAXUS Element Paclitaxel-Eluting Coronary Stent System for the Treatment of De Novo Coronary Artery Lesions; NCT00484315)
Paclitaxel-eluting stents (PES) reduce clinical and angiographic restenosis compared with bare-metal stents. The TAXUS Express (1,2) and TAXUS Liberté stents (3) (Boston Scientific, Natick, Massachusetts) have demonstrated durable benefit across a broad spectrum of patient risk cohorts (1,4–7).
The TAXUS Element PES (Boston Scientific) incorporates a novel, thin-strut (81 μm) platinum chromium metal alloy (8) platform designed to enhance radiopacity, radial strength, and conformability (Table 1).Although the polymer and drug are similar to those of prior TAXUS PES platforms, the platinum chromium alloy has increased bend fatigue resistance, greater conformability, and increased radial strength versus 316L stainless steel (Table 1). Because platinum chromium is denser than 316L stainless steel (9), radiopacity is enhanced despite thinner struts (10,11).
In pre-clinical testing, vascular compatibility was similar among bare-metal, polymer-only, and drug-coated Element stents (10). Furthermore, strut coverage and endothelialization were more rapid in Element compared with the Express and Liberté stents (12).
The TAXUS PERSEUS (Prospective Evaluation in a Randomized Trial of the Safety and Efficacy of the Use of the TAXUS Element Paclitaxel-Eluting Coronary Stent System) Workhorse (WH) randomized clinical trial was designed to evaluate the safety and efficacy of the novel thin-strut platinum chromium alloy TAXUS Element stent compared with the TAXUS Express stent in single de novo coronary stenoses.
The TAXUS Element stent is laser-cut from 81-μm platinum chromium alloy. Controlled slow release of paclitaxel, incorporated at 1 μg/mm2into poly(styrene-b-isobutylene-b-styrene) polymer is provided with release kinetics similar to those of the previous generation TAXUS Express (2,6) and Liberté (3) 316L stainless steel stent systems (11).
The PERSEUS WH trial design has been reported previously (11). The PERSEUS WH trial is a prospective, randomized (3:1), single-blind controlled trial powered to evaluate noninferiority of the TAXUS Element compared with the TAXUS Express2PES. Eligible subjects with a single, native vessel, de novo target lesion of ≥50% diameter stenosis with length ≤28 mm and reference vessel diameter ≥2.75 to ≤4.0 mm were enrolled between July 16, 2007, and October 1, 2008, at 90 clinical sites (11). Randomization to treatment group was stratified for the presence or absence of medically-treated diabetes. Subjects (n = 330) were randomly assigned to 9-month quantitative coronary angiographic (QCA) follow-up throughout the course of trial enrollment. The study primary end point is the rate of target lesion failure (TLF) (ischemia-driven target lesion revascularization [TLR], myocardial infarction [MI] related to target vessel, or cardiac death related to target vessel) at 12 months after index procedure, and the key secondary end point is 9-month in-segment percentage diameter stenosis by QCA. Clinical follow-up was scheduled at 30 days, 9 months, 12 months, 18 months, and annually to 5 years.
Planned use of multiple stents was prohibited in both treatment arms, and additional study stents in the target lesion were allowed only for suboptimal angiographic results, which compromised vessel integrity or threatened vessel closure (i.e., “bailout”). Treatment of 1 lesion in a nontarget vessel during the index procedure was allowed before treatment of the target lesion, provided that nontarget lesion treatment was successful angiographically and did not require additional unplanned stents. Staged PCI or subsequent planned coronary artery bypass graft procedures were not allowed. Other inclusion and exclusion criteria have been reported (11).
Thienopyridine treatment was required by protocol for at least 6 months, preferably to 12 months, in subjects not at high risk of bleeding (13), and daily aspirin therapy was required indefinitely. Other periprocedural adjunctive pharmacotherapy was administered per standard practice of the participating institution (11).
The study protocol was approved by all participating ethics review committees, and all patients provided written informed consent. An independent clinical events committee adjudicated all stent thrombosis and major adverse cardiac events (MACE). An independent data monitoring committee provided oversight of aggregate safety data. Central core laboratory analysis of all angiographic studies was performed by the CardioVascular Institute at Beth Israel Deaconess Medical Center (Boston, Massachusetts) with standard qualitative morphologic criteria similar to those used in the TAXUS Express and Liberté clinical trials (14).
Bayesian hierarchical modeling was used for the primary and key secondary end points, and modeling analyses (15) based on the per-protocol analysis cohort (excluding patients who did not receive the assigned study stent). For all other analyses, the intent-to-treat randomized treatment groups were compared with frequentist analysis methods with a Student ttest for continuous measures and a chi-square or Fisher exact test for discrete measures. Historical data from the TAXUS IV and V studies (2,6) were used in sample size calculations (11). Prior data were to be borrowed under the Bayesian framework only if the observed 12-month TLF rate in the PERESUS WH randomized TAXUS Express control group was >8.0%. Because the observed TLF rate in this group was only 6.1%, prior data were not borrowed for either the primary or secondary end point analyses in order to maintain a more conservative approach. For the secondary end point, a natural log (ln) transformation was used on the data to improve the normality of the distribution; analyses were performed on the transformed data. Noninferiority margins (Δ) of 4.1% and 0.20 were used for the primary and secondary end point analyses, respectively. Noninferiority of the TAXUS Element stent is accepted if the Bayesian posterior probability is at least 95% that the difference in 12-month TLF or ln(percentage diameter stenosis) between TAXUS Element and TAXUS Express was less than the pre-specified Δ. In this Bayesian approach, the posterior probability of noninferiority is similar to the complement of the p value used in the frequentist approach.
A sample size of 1,264 subjects was expected to provide 1,200 evaluable subjects (assuming 5% attrition) and was determined through simulations based on hierarchical modeling as well as discussions with the U.S. Food and Drug Administration (FDA). Although power and type I error do not apply to the PERSEUS WH trial in the frequentist sense, this sample size is the minimum sample size required to give approximately an 80% probability of correctly concluding noninferiority (over a range of assumed TAXUS Express TLF rates from 6% to 12%), under the assumption that TAXUS Element was noninferior to TAXUS Express.
To assess the validity of pooling across clinical sites in the PERSEUS WH trial, a frequentist logistic regression model was used to test the site × treatment interaction on the primary TLF end point (p > 0.99).
Baseline demographic data and lesion characteristics
Figure 1depicts patients included in the 9-month QCA and 12-month clinical follow-up. A total of 1,262 patients were included in the intent-to-treat analysis set (320 TAXUS Express, 942 TAXUS Element) with 97.7% 12-month clinical follow-up and 87.6% 9-month angiographic follow-up. Baseline demographic and lesion characteristics are shown in Table 2.No differences between randomized stent treatment groups were observed, with the exception that TAXUS Element stent-treated subjects were slightly younger.
Technical and procedural performance
Technical success (successful delivery and deployment of the study stent without balloon rupture or stent embolization) was comparable between treatment groups (Table 3).Other procedural characteristics including use of multiple stents, stent length/lesion length ratio, maximum deployment pressure, use of post-dilation, and maximum post-dilation pressures were all similar between randomized treatment groups (Table 3).
Antiplatelet medication compliance
Aspirin and thienopyridine use at 12 months after the procedure was similar for TAXUS Express- and Element-treated patients (aspirin 97.4% vs. 96.4%, respectively, and thienopyridine 91.3% vs. 91.6%, respectively).
Post-procedural QCA demonstrated a significant increase in in-stent minimal lumen diameter (MLD) for TAXUS Element compared with TAXUS Express stent-treated patients, which was supported by numerically increased in-stent acute gain, despite similar deployment and post-dilational pressures between stent treatment groups (Table 3).
Angiographic measures including MLD, percentage diameter stenosis, late loss (in-stent and in-segment), and binary (>50%) restenosis were all similar between QCA-subset treatment groups (Table 4).For the key secondary study end point of in-segment percentage diameter stenosis at 9 months, with a pre-specified margin of 0.20, the posterior probability of noninferiority was 99.7% (Fig. 2B),meeting the secondary end point.
Two TAXUS Element stent fractures (0.64%) were identified by the core laboratory among the 314 implanted stents in 294 patients of the randomized QCA subset (16). No fractures were observed in TAXUS Express study stents. No additional fractures were identified from among all available films (including non–QCA-subset patients who received angiography). There were no in-stent restenoses or aneurysms associated with either fractured stent, and neither patient experienced MACE or stent thrombosis events though 12-month follow-up.
The primary study end point of TLF to 12 months after procedure was similar between the randomly assigned stent platforms (5.57% TAXUS Element vs. 6.14% TAXUS Express; difference −0.57%; 95% credible interval: 1.85%) (Figs. 2A and 3).With a pre-specified margin of 4.1%, the Bayesian posterior probability of noninferiority was 99.96%, meeting the primary end point. Bayesian multivariate analysis demonstrated that use of unplanned stents and smaller final in-segment MLD was associated with an increased risk of TLF to 12 months (Online Appendix A). After adjustment for covariate predictors, the Bayesian posterior probability of noninferiority for TLF was maintained at 95.47%.
Additional 12-month clinical end points are shown in Table 5(17). No differences in the rates of MACE, mortality (cardiac or noncardiac), revascularization, or stent thrombosis were observed between treatment groups. Although the overall incidence of MI was similar between stent types, non–Q-wave MI was numerically less frequent after TAXUS Element (1.6% vs. 2.9% TAXUS Express), driven by fewer periprocedural infarctions.
The PERSEUS WH trial provides the necessary assurance of safety and relative efficacy of the novel platinum chromium TAXUS Element PES platform, demonstrating noninferiority to the TAXUS Express PES with respect to the incidence of TLF at 12 months and the severity of residual angiographic stenosis at 9 months. These results were demonstrated within the framework of a Bayesian study design, which has been accepted by the FDA for medical device clinical trials (15) and might be considered a more intuitive and clinically meaningful method of analysis than the conventional frequentist approach (18).
Studies have suggested that bare-metal stents with thinner struts might provide reduced late loss compared with thicker-strut stents (19,20). Although this finding was not apparently replicated in the PERSEUS WH trial, the following points deserve mention. First, the increase in acute gain observed after TAXUS Element (vs. TAXUS Express) on post-procedural QCA would be expected to be accompanied by a greater late lumen loss in follow-up. The characteristic direct relationship between acute gain and late loss is demonstrated by the comparison of TAXUS Element with TAXUS Express and results in similar net gain as reflected by the incidence of binary angiographic restenosis. Second, the noninferiority study design was not powered to detect statistically significant differences between the 2 drug-eluting stent arms. Finally, the potential impact of reduced strut thickness on late lumen loss might be overcome by the relative impact of drug and polymer.
Importantly, no safety concerns were observed in the course of the trial. Although the study is underpowered for low-frequency events and the current duration of follow-up is limited (1 year), the incidence of death, MI, and stent thrombosis was low and similar between randomly assigned stent platforms. Only 2 stent fractures were noted in the study (0.64%), consistent with fracture rates in other PES. No adverse clinical events, in-stent restenoses, or aneurysms were associated with these fractures.
The ischemia-driven TLR rate in the nonangiographic cohort of the PERSEUS TAXUS Express control (3.4%) compares favorably to that observed in TAXUS Express-treated patients enrolled into the SPIRIT IV trial (4.5%) (21). Ischemia-driven TLR rates were numerically lower in the nonangiographic cohort of TAXUS Element patients in the PERSEUS trial (2.8%) and in everolimus-eluting stent-treated patients in the SPIRIT IV trial (2.3%). The TLR rates in the nonangiographic cohort of the ENDEAVOR IV study were 3.6% for zotarolimus-eluting stents and 3.2% for PES (22). Thus, the TAXUS Element PES seems to provide similar clinical efficacy to other available paclitaxel- or rapamycin-derived drug-eluting stents.
A potential limitation of the PERSEUS WH trial is that the control stent comparator is the first-generation TAXUS Element PES (1), which was the only commercially available PES in the U.S. at the time enrollment into the PERSEUS WH trial was initiated. Nevertheless, the safety and efficacy parameters observed after TAXUS Element in the PERSEUS WH trial compare favorably with those observed after deployment of the more recently FDA-approved TAXUS Liberté PES in the TAXUS ATLAS (A Multi-Center, Single-Arm Study of the TAXUS Liberté-SR Stent for the Treatment of Patients With de Novo Coronary Artery Lesions) trial (3).
The PERSEUS WH randomized clinical trial demonstrates comparable efficacy between the novel platinum chromium alloy TAXUS Element PES and the TAXUS Express PES. Importantly, no safety concerns regarding either the novel metal alloy or modified stent design were observed. The PERSEUS WH trial demonstrates successful transfer of PES technology from stainless steel to a platinum chromium alloy platform.
The authors thank Donald Baim, MD, for guidance, intellectual insights, and friendship, not only during the inception and course of the PERSEUS trial but throughout his career.
The authors thank the following Boston Scientific employees for their contribution: Amy Britt, Andrey Nersesov, Manu Sondhi, Thomas Christen, and Stephen Mascioli for study design, management, and safety monitoring; Scott Wehrenberg and Peggy Pereda for statistical design and analysis; Alan Yu for assistance with Bayesian design; and Kristin Hood for medical writing. The authors also humbly thank the PERSEUS WH clinical staff, committees, and investigational sites for their support (Online Appendix B).
For supplementary methods and tables, please see the online version of this article.
Clinical and Angiographic Outcomes After Treatment of De Novo Coronary Stenoses With a Novel Platinum Chromium Thin-Strut Stent: Primary Results of the PERSEUS Trial (Prospective Evaluation in a Randomized Trial of the Safety and Efficacy of the Use of the TAXUS Element Paclitaxel-Eluting Coronary Stent System)
The PERSEUS WH trial is funded by Boston Scientific Corporation, Natick, Massachusetts. Dr. Kereiakes received research grants from Boston Scientific, Cordis, Medtronic, and Abbott Vascularand serves on the Advisory Boards for Boston Scientific and Abbott Vascular. Dr. Cannon serves on the Advisory Board or Speakers' Bureau for Medtronic, Abbott, and Boston Scientific and holds equity in Boston Scientific, Medtronic, and BioStar Ventures. Dr. Feldman has research grant support from and serves as a consultant for Boston Scientific. Dr. Popma has received personal and institutional research grants from Cordis Corporationand institutional research grants from Boston Scientificand Abbott Vascular. Dr. Magorien serves as a consultant to Boston Scientific and has received grants from Abbott Vascular. Dr. Whitbourn's institution has received a research grant from Boston Scientific. Dr. Rabinowitz has received funds from Boston Scientificto support expert testimony and investigator meeting travel. Dr. Kabour has received honoraria funds from Boston Scientific. Dr. Foster serves on the Speakers' Bureau for Volcano Corporation. Dr. Wang serves on the Medical Advisory Board for Boston Scientific and has received travel funds to attend Medical Advisory Board meetings. Drs. Underwood and Dawkins are full-time employees and stockholders of Boston Scientific Corporation.
- Abbreviations and Acronyms
- diameter stenosis
- Food and Drug Administration
- major adverse cardiac events
- myocardial infarction
- minimal lumen diameter
- paclitaxel-eluting stent
- quantitative coronary angiography
- target lesion failure
- target lesion revascularization
- target vessel revascularization
- Received February 25, 2010.
- Revision received March 30, 2010.
- Accepted April 7, 2010.
- American College of Cardiology Foundation
- Ellis S.G.,
- Stone G.W.,
- Cox D.A.,
- et al.
- Turco M.A.,
- Ormiston J.A.,
- Popma J.J.,
- et al.
- Mahmud E.,
- Ormiston J.A.,
- Turco M.A.,
- et al.
- Steiner R.
- Wilson G.J.,
- Huibregtse B.A.,
- Stejskal E.A.,
- et al.
- Allocco D.J.,
- Cannon L.A.,
- Britt A.,
- et al.
- Soucy N.V.,
- Feygin J.M.,
- Tunstall R.,
- et al.
- King S.B. III.,
- Smith S.C. Jr..,
- Hirshfeld J.W. Jr..,
- et al.
- Turco M.A.,
- Ormiston J.A.,
- Popma J.J.,
- et al.
- ↵Guidance for the Use of Bayesian Statistics in Medical Device Clinical Trials: U.S. Food and Drug Administration, Center for Devices and Radiological Health. http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm071072.htm. Accessed February 5, 2010.
- Nakazawa G.,
- Finn A.V.,
- Vorpahl M.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.
- Kaul S.,
- Diamond G.A.
- Pache J.,
- Kastrati A.,
- Mehilli J.,
- et al.
- Kastrati A.,
- Mehilli J.,
- Dirschinger J.,
- et al.
- Leon M.B.,
- Mauri L.,
- Popma J.J.,
- et al.