Author + information
- Received April 25, 2008
- Revision received July 25, 2008
- Accepted September 12, 2008
- Published online December 1, 2008.
- Mark A. Turco, MD⁎,⁎ (, )
- John A. Ormiston, MBChB†,
- Jeffrey J. Popma, MD‡,
- Jack J. Hall, MD§,
- Tift Mann, MD∥,
- Louis A. Cannon, MD¶,
- Mark W.I. Webster, MBChB#,
- Gregory J. Mishkel, MD⁎⁎,
- Charles D. O'Shaughnessy, MD††,
- Thomas F. McGarry, MD‡‡,
- Lazar Mandinov, MD§§,
- Keith D. Dawkins, MD§§ and
- Donald S. Baim, MD§§
- ↵⁎Reprint requests and correspondence:
Dr. Mark A. Turco, Center for Cardiac & Vascular Research, Washington Adventist Hospital, 7600 Carroll Avenue, Takoma Park, Maryland 20912
Objectives The TAXUS ATLAS Small Vessel (SV) and Long Lesion (LL) multicenter studies compared the performance of the thin-strut (0.0038 inch) TAXUS Liberté 2.25-mm stent (Boston Scientific; Natick, Massachusetts) and the TAXUS Liberté 38-mm long stent (Boston Scientific; Natick, Massachusetts) with the earlier paclitaxel-eluting TAXUS Express (Boston Scientific) stent that has identical polymer, drug dosage, and release kinetics but different stent geometry and thicker struts (0.0052 inch).
Background The TAXUS Liberté stent was designed with thinner and more even strut spacing to provide more uniform drug distribution, as well as increased flexibility and conformability. Clinical benefits of the new stent design have not been evaluated.
Methods The TAXUS ATLAS SV and LL studies are nonrandomized studies comparing outcomes of the TAXUS Liberté 2.25 mm (N = 261) and TAXUS Liberté 38 mm (N = 150) stents to TAXUS Express historical control groups derived from the TAXUS IV and V trials. Inclusion/exclusion criteria for TAXUS Express and Liberté groups were similar in both studies.
Results Each study met its primary end point of noninferiority of 9-month in-segment diameter stenosis. Furthermore, TAXUS Liberté 2.25 mm, when compared with TAXUS Express, significantly reduced the rate of both 9-month angiographic restenosis (18.5% vs. 32.7%, p = 0.0219) and 12-month target lesion revascularization (6.1% vs. 16.9%, p = 0.0039). In addition, TAXUS Liberté 38 mm significantly reduced the risk of 12-month myocardial infarction compared with TAXUS Express (1.4% vs. 6.5%, p = 0.0246).
Conclusions The thinner-strut TAXUS Liberté stent improved outcomes compared with the earlier TAXUS Express stent in both SVs and LLs (A Study of the TAXUS Liberté Stent for the Treatment of de Novo Coronary Artery Lesions in Small Vessels; NCT00371748; A Study of the TAXUS Liberté Stent for the Treatment of Long De Novo Coronary Artery Lesions; NCT00371475).
Compared with bare-metal stents, drug-eluting stents significantly reduce the incidence of restenosis and the need for repeat revascularization in patients with relatively simple de-novo coronary lesions (1,2). More recently, the safety and effectiveness of the first-generation TAXUS (Boston Scientific Corporation, Natick, Massachusetts) and Cypher (Cordis/Johnson and Johnson; Miami Lakes, Florida) drug-eluting stents have been evaluated in the more challenging subsets of smaller vessels and longer lesions (3,4). Small vessel (SV) diameter remains a significant predictor of restenosis or stent thrombosis after drug-eluting stent implantation (5–7), while long lesions (LL), particularly those requiring multiple stents, may be associated with an increased risk of periprocedural myocardial infarction (MI) (4).
Stent design is recognized as a determinant of stent performance (8). Strut thickness, metal composition, and radial strength can each influence short- and long-term outcomes (9,10). The new-generation TAXUS Liberté stent was developed using the same proven polymer-based, paclitaxel-eluting technology as on TAXUS Express but applied to the newer Liberté stent. This new stent design has thinner struts and more uniform distribution of struts than the earlier TAXUS Express to improve deliverability, conformability, and more homogenous drug distribution (11). The TAXUS ATLAS study demonstrated that in noncomplex lesions TAXUS Liberté and TAXUS Express performance was similar (12). The current report describes the outcomes of the nonrandomized, historically controlled TAXUS ATLAS SV and TAXUS ATLAS LL trials that compared the performance of the thin-strut TAXUS Liberté to the TAXUS Express in patients with SVs or LLs, respectively.
The TAXUS Liberté-SR Stent (Boston Scientific Corporation) consists of a balloon-expandable Liberté stent with a polymer coating containing 1 μg/mm2 of paclitaxel in a slow-release formulation. Drug dosing and release kinetics are identical to that of the TAXUS Express-SR Stent (Boston Scientific Corporation). Both stents are made from 316L stainless steel, but the Liberté platform has a more uniform strut pattern and thinner struts (0.0038 inches or 0.097 mm) than the Express platform (0.0052 inches or 0.132 mm).
Patient selection and procedures
The study protocols were approved by local ethics committees, and all patients provided written informed consent before enrollment. The TAXUS ATLAS SV and LL study designs are summarized in Table 1. A total of 261 patients were enrolled in the TAXUS ATLAS SV study, and 150 patients were enrolled in the TAXUS ATLAS LL study. General inclusion and exclusion criteria were the same as in the TAXUS ATLAS and TAXUS IV and V trial protocols and have been described previously (2,4,12). The only difference among these protocols was related to target lesion length and target vessel size.
The historical control groups were derived from the TAXUS IV and V trials (2,4). The TAXUS ATLAS SV study results were compared with those seen in a control group of 75 TAXUS V trial patients with reference vessel diameter (RVD) ≤2.5 mm, lesion length ≤28 mm, and receiving a single TAXUS Express 2.25-mm stent. The TAXUS ATLAS LL study results were compared with those seen in a control group of 145 TAXUS IV and V trial patients with RVD ≥2.5 and ≤4.0 mm, lesion length ≥26 and ≤34 mm and receiving ≥1 TAXUS Express stents. Multiple stenting was only allowed for bailout purposes in the TAXUS ATLAS LL study and in the TAXUS IV trial. Planned multiple stenting was permitted in the TAXUS V trial. Patient flow through 12-month follow-up is shown in Figure 1.
Dual antiplatelet therapy
Loading doses of clopidogrel (300 mg) or ticlopidine (500 mg) and aspirin (300 mg) were administered before the procedure. After stent implantation, clopidogrel (75 mg daily) or ticlopidine (250 mg twice daily) was mandated for a minimum of 6 months, and aspirin therapy (≥100 mg daily), mandated for 9 months, was recommended indefinitely.
Quantitative angiographic analysis
Identical angiographic methods were used for the TAXUS ATLAS studies and TAXUS IV and V trials. Quantitative coronary angiography analysis, performed by the same Core Laboratory, used a validated automated edge detection algorithm (Medis CMS, Leiden, the Netherlands). Minimum lumen diameter (MLD) was measured within the stent (in-stent), within the adjacent 5-mm proximal and distal edges of the stent, and within the segment between the proximal and distal reference vessel (in-segment). Diameter stenosis (DS) was defined as (1 − [MLD/RVD]) × 100%. Binary restenosis was defined as a ≥50% DS.
Major adverse cardiac events (MACE) and stent thromboses were adjudicated by an independent clinical events committee that was similar for all TAXUS ATLAS studies and the TAXUS IV and V trials. The clinical and angiographic end point definitions were identical to those in the TAXUS ATLAS study and the TAXUS IV and V trials (2,4,12). Stent thrombosis was adjudicated according to both the protocol definition and the Academic Research Consortium definition (13).
The primary end point of both the TAXUS ATLAS SV and LL trials was noninferiority of 9-month in-segment DS versus the historical TAXUS Express control subjects. Secondary end points in both studies included quantitative coronary angiography and clinical end points.
Baseline, post-procedural, and follow-up data were summarized using descriptive statistics and presented as proportions (%, count/sample size) or mean ± standard deviation. The student t test was used for comparing continuous variables. The chi-square or Fisher exact tests, as appropriate, were used for comparing proportions. The Kaplan-Meier product-limit method and log-rank test were used to assess time-to-event end points between groups.
For noninferiority testing of the primary end point in the TAXUS ATLAS SV study, enrollment of 260 patients conferred 89% power with the following assumptions based on the TAXUS IV and VI trials: a noninferiority margin of 10.0%, a 9-month in-segment DS of 37.1%, and 30% attrition.
For noninferiority testing of the primary end point in the TAXUS ATLAS LL study, the required enrollment of 150 subjects conferred 80% power with the following assumptions based on the TAXUS IV trial: a noninferiority margin of 6.9%, a 9-month in-segment DS of 29.0%, and 30% attrition.
Primary end points in both studies were adjusted by propensity score analysis to control for differing baseline characteristics. Propensity score analysis was performed using a hierarchical logistic regression model with a stepwise selection process and an entry/exit criterion of 0.10. The per protocol population excluded patients who did not receive the study stent and was the main population of interest for the primary end point.
Multivariate analysis was used to determine predictors of 1-year clinical outcomes. All covariates were modeled by univariate and multivariate analysis using a stepwise procedure in a Cox regression model. The significance level for entry/exit was set at 0.10. Linear regression and logistic regression models were used to determine predictors of continuous and categorical angiographic variables, respectively. All outcomes were adjusted by propensity score analysis. All baseline variables considered for both propensity score adjustment and multivariate analysis are outlined in Table 2. Stent type was also considered for multivariate modeling. Statistical significance was accepted at a value of p < 0.05.
Baseline characteristics in the TAXUS ATLAS SV and LL trials
Baseline patient and lesion characteristics for both the TAXUS ATLAS SV and LL studies are shown in Table 2. In the TAXUS ATLAS SV study, baseline characteristics were well matched between the groups; however, the target lesion was significantly longer in the control patients. In the TAXUS ATLAS LL study, TAXUS Liberté patients were more likely to have stable rather than unstable angina, hyperlipidemia, longer lesions, calcification, and American College of Cardiology/American Heart Association type B2 or C lesions but were less likely to have thrombus than the control patients.
Procedural characteristics in the TAXUS ATLAS SV and LL trials
In both studies, procedure time in patients without a nontarget lesion treated was significantly shorter for TAXUS Liberté than TAXUS Express (Table 2). Technical success, defined as stent delivery with no device malfunction, was identical for the 2 stents: 97.3% in the TAXUS ATLAS SV study and 100.0% in the TAXUS ATLAS LL study. Clinical success was similar for both stents in the TAXUS ATLAS SV study but significantly higher for the TAXUS Liberté in the TAXUS ATLAS LL study (Table 2). In both studies, stented length was significantly longer for TAXUS Liberté than TAXUS Express, although the stent-to-lesion length ratio was lower for TAXUS Liberté. The frequency of multiple stenting was similar in the TAXUS ATLAS SV study but higher for the TAXUS Express in the TAXUS ATLAS LL study reflecting the lack of a 38-mm length for the TAXUS Express.
Primary end point in the TAXUS ATLAS SV and LL trials
The primary noninferiority end point was met in both the TAXUS ATLAS SV and LL studies (Fig. 2). In both studies, the upper 1-sided 95% confidence interval (CI) of the difference in in-segment DS between groups was significantly less than the pre-specified margin, thus demonstrating the noninferiority of the TAXUS Liberté 2.25 mm and the TAXUS Liberté 38 mm versus TAXUS Express.
Outcomes in SVs
At 9 months, the TAXUS Liberté 2.25-mm stent exhibited significant reductions in in-segment DS and late loss compared with the TAXUS Express stent. These differences became statistically nonsignificant after propensity score adjustment although a clear trend remained (Table 3). Both in-stent and in-segment binary restenosis rates were significantly reduced with TAXUS Liberté but no longer met statistical significance after propensity score adjustment (Table 3). The TAXUS Liberté stent predominantly reduced focal in-stent restenosis, with no apparent edge effect (Tables 3 and 4).⇓
Multivariate linear regression models demonstrated that treatment with the TAXUS Liberté (vs. TAXUS Express) was a significant predictor of both lower DS (coefficient −6.11, p = 0.0415) and less late loss (coefficient −0.16, p = 0.0171). In addition, a Cox regression model identified TAXUS Liberté as a significant predictor of reduced binary restenosis (odds ratio: 0.39, 95% CI: 0.18 to 0.83, p = 0.0148) (Fig. 3).
At 12 months, patients receiving TAXUS Liberté 2.25 mm had significantly greater freedom from target lesion revascularization (TLR) compared with TAXUS Express patients (Fig. 4). The TLR rate was statistically significantly lower for TAXUS Liberté patients versus TAXUS Express patients (6.1% vs. 16.9%, p = 0.0039). After propensity score adjustment, this difference was no longer statistically significant, although a trend for reduced TLR remained (5.8% vs. 13.2%, p = 0.08) (Table 5). Multivariate analysis revealed that treatment with the TAXUS Liberté significantly reduced the risk of TLR versus the TAXUS Express (hazard ratio: 0.34, 95% CI: 0.16 to 0.72, p = 0.0048) (Fig. 3), consistent with the risk reduction observed for binary restenosis.
Safety of TAXUS LIBERTÉ 2.25 mm in SVS
In-hospital (1.9% vs. 2.7%, p = 0.66) and 30-day MACE rates (1.9% vs. 4.1%, p = 0.38) were similar for the TAXUS Liberté and the TAXUS Express groups, respectively. The 12-month MACE rate was significantly lower for the TAXUS Liberté compared with that of the TAXUS Express, mainly attributable to the lower TLR rates (Table 5). Through 12 months, there were no significant differences in overall or cardiac mortality, MI, and stent thrombosis rates, either unadjusted or adjusted (Table 5).
Outcomes in LLs
In the TAXUS ATLAS LL study, all 9-month angiographic outcomes, whether adjusted or unadjusted, were comparable for TAXUS Liberté 38 mm and TAXUS Express (Table 3). Accordingly, the 12-month TLR rates were similar between the groups (Table 5).
Safety of TAXUS LIBERTÉ 38 mm in LLS
MACE were significantly reduced in the TAXUS Liberté patients compared with those in TAXUS Express patients both in-hospital (0% vs. 4.1%, p = 0.0134) and at 30 days (0% vs. 4.9%, p = 0.0061). All 30-day MACE in TAXUS Express patients were exclusively MI. The 12-month MACE rates were similar between the TAXUS Liberté and the TAXUS Express groups (Table 5), but freedom from MI through 12 months remained significantly improved for patients receiving TAXUS Liberté compared with those receiving TAXUS Express. The MI rate was significantly lower for TAXUS Liberté than TAXUS Express (1.4% vs. 6.5%, p = 0.0246). Although this difference at 12 months was no longer statistically significant after propensity score adjustment, a trend toward lower MI rate remained (1.7% vs. 5.1%, p = 0.11) (Table 5). Multivariate analysis demonstrated that treatment with TAXUS Liberté significantly reduced the risk of MI by 81% versus treatment with TAXUS Express (Fig. 5).
To further characterize the reduced risk of MI, we analyzed patients with single and multiple stents separately; the difference in MI rate was attributable to the single stent cohort only (Fig. 6B). Moreover, the lower MI rate with TAXUS Liberté was predominantly due to a significant reduction in the incidence of periprocedural (in-hospital) non–Q-wave MI with TAXUS Liberté versus TAXUS Express (Fig. 6C).
No stent thromboses were reported for TAXUS Liberté. The 12-month cardiac mortality, adjusted or unadjusted, was significantly lower for the TAXUS Liberté patients than the control subjects. Multivariate analysis did not demonstrate a significant risk reduction for cardiac death by treatment with TAXUS Liberté (Fig. 5).
The TAXUS ATLAS SV and LL trials were designed to assess the safety and efficacy of the second-generation TAXUS Liberté stent that uses identical polymer and paclitaxel release but more uniform strut distribution and thinner struts than the first-generation TAXUS Express stent in challenging patients with SVs or LLs.
The TAXUS ATLAS SV study demonstrated that the TAXUS Liberté 2.25-mm stent was noninferior versus the TAXUS Express in terms of 9-month in-segment DS, and in addition, reduced binary restenosis at 9 months and the need for TLR at 1 year in subjects with SVs (<2.5 mm). Indeed, multivariate regression analysis confirmed that use of TAXUS Liberté was an independent predictor of reduced restenosis and TLR in SVs. In this regard, the TAXUS Liberté showed improved efficacy compared with that seen with the TAXUS Express.
The TAXUS ATLAS LL study demonstrated that the new TAXUS Liberté 38-mm stent was noninferior versus the TAXUS Express in terms of 9-month in-segment DS. Importantly, however, TAXUS Liberté reduced the rate of MI at 1 year due to a significant reduction in the rate of periprocedural MI in the single stent cohort. Multivariate regression analysis showed that the TAXUS Liberté 38 mm was an independent factor for reducing the MI risk, confirming the improved safety profile of the TAXUS Liberté compared with that of the TAXUS Express in LLs (26 to 34 mm).
Taken together, these 2 studies demonstrate that the thin-strut TAXUS Liberté stent may have specific clinical benefits for patients with SVs or LLs, beyond simply improved deliverability. Given that the 2 stents use the same stent material, polymer coating, drug dose, and release kinetics, the observed clinical benefits are likely the result of the more uniform strut pattern and thinner strut design.
Several previous studies have demonstrated that thinner-strut stents are associated with a significant reduction of clinical and angiographic restenosis than thick-strut stents (9,10,14). Although the mechanism for this improved outcome is not clearly understood, it may relate to reduced deep vessel injury during implantation or improved healing (9). The relationship between stent strut thickness and restenosis remains particularly evident in smaller vessels (9,10). A subanalysis of the ISAR-STEREO (Intracoronary Stenting and Angiographic Results: Strut Thickness Effect on Restenosis Outcomes) trial indicated that the differences between thin- and thick-strut stents were only apparent in patients with complex lesions (14). More recently, a randomized trial of the sirolimus-eluting Cypher stent versus a thin (0.0032 inch) strut bare-metal stent in a population at medium risk of restenosis confirmed the efficacy of Cypher in terms of reduced late loss and restenosis (15). However, the absolute differences for late loss (0.40 mm) and angiographic restenosis (13%) were much less between these 2 stents than the differences observed between Cypher and a thick-strut (0.14 mm) bare-metal stent in the SIRIUS (Sirolimus-Eluting Stent in Coronary Lesions) trial (late loss 0.57 mm; angiographic restenosis 27.4%) (1).
The TAXUS ATLAS trial is the first program to compare thick- and thin-strut paclitaxel-eluting stents. The pivotal TAXUS ATLAS trial did not show any differences in restenosis rates between TAXUS Liberté and TAXUS Express in workhorse lesions (12) and reiterates the findings of lack of differences in outcomes between thin-strut and thick-strut bare-metal stents in simple lesions in the ISAR-STEREO trial (14). Furthermore, the TAXUS ATLAS LL study, excluding SVs (<2.7 mm), failed to show any difference in clinical and angiographic restenosis between thin-strut and thick-strut paclitaxel-eluting stents. Importantly, the restenosis reduction with TAXUS Liberté versus TAXUS Express in SVs is similar to the finding for thin- versus thick-strut bare-metal stents in the ISAR-STEREO trial. This differing outcome with thick and thin struts was ascribed to less vessel injury during implantation of the thinner struts in SVs (9), and a similar mechanism may be responsible for improved outcomes with TAXUS Liberté versus TAXUS Express in SVs. If so, this may explain part of the benefit of the thinner-strut Xience V (Abbott Vascular, Santa Clara, California)/PROMUS (Boston Scientific Corporation) everolimus-eluting stent compared with the thick-strut TAXUS Express in SVs of the SPIRIT III (Everolimus-Eluting Stent in Coronary Lesions) trial (16), although the differences in the active drug and polymer must still be considered.
The TAXUS ATLAS LL study also highlights the importance of stent design in LLs (26 to 34 mm) that often require the placement of multiple stents. Compared with TAXUS Express, the new TAXUS Liberté 38-mm stent decreased the frequency of multiple stent use in these LLs, and in addition significantly reduced the rate of periprocedural MI. While multiple stenting with either the TAXUS Express or the Cypher stent (≥3 stents) can increase the rate of periprocedural MI versus bare-metal stent controls (4,17), the MI reduction in the TAXUS ATLAS LL study was predominantly observed in the single stent cohorts and occurred despite the greater 38-mm length of the TAXUS Liberté compared with the longest TAXUS Express stent (32 mm). This benefit of TAXUS Liberté versus TAXUS Express seems to be specific for LLs (26 to 34 mm), which were not studied in the previous TAXUS ATLAS workhorse trial (12). Given the identical drug/polymer coating for the 2 TAXUS stents, this reduction in periprocedural MI seen with the long TAXUS Liberté stent likely also results from the improved stent geometry (potentially less side branch “jailing”) and thinner struts (less deep plaque injury and potential plaque embolization) than seen with the thicker-strut TAXUS Express design in these longer lesions.
Both the TAXUS ATLAS SV and LL studies are nonrandomized studies with historical control subjects drawn from the TAXUS IV and V trials. The differences in baseline characteristics observed between TAXUS Liberté and TAXUS Express patients were compensated for by the use of propensity score adjustment. Multivariate analysis provided additional compensation.
Besides the evident differences in strut thickness, it is possible that other differences in the mechanical and physical properties (e.g., radial strength, gap size or open areas between stent struts, and acute recoil) between TAXUS Liberté and TAXUS Express may have contributed to the improved performance of TAXUS Liberté. However, the designs of the TAXUS ATLAS SV and LL studies cannot distinguish the contributions of these individual factors on clinical and angiographic outcomes.
Finally, intravascular imaging in the TAXUS ATLAS SV study might have provided additional insights to differences in vessel injury and response but was not included in the protocol. Similarly, detailed side-branch analyses were not performed in the TAXUS ATLAS LL study to further define the specific mechanism of the reduction in periprocedural MI in LLs.
The new-generation paclitaxel-eluting TAXUS Liberté stent, using a more uniform geometry and thinner struts, significantly reduced the risk of restenosis in SVs and periprocedural MI in LLs compared with the earlier TAXUS Express platform. Since both stent platforms used otherwise identical drug/polymer coating, the improved performance and safety profile of the TAXUS Liberté is most likely the result of the thinner struts and improved stent geometry.
The authors thank the following individuals at Boston Scientific: Leslie Stolz, PhD for assistance with drafting this manuscript and Carola Alfaro, MS; Brian Johnson, MPH; and Joseph Bero, MS for statistical analyses. The authors also thank all investigators and investigational sites for their contributions to the TAXUS ATLAS Small Vessel and Long Lesion trials.
For a list of the TAXUS ATLAS Small Vessel/Long Lesion Stent Study Organization, please see the online version of this article.
This study was supported by Boston Scientific Corporation. Drs. Turco and Cannon have received consulting fees/honoraria and have been on the advisory board and the Speakers' Bureau for Boston Scientific Corporation; Dr. Ormiston has received consulting fees/honoraria and has been on the advisory board for Boston Scientific Corporation; Drs. Popma and Mishkel have received consulting fees/honoraria, have been on the Speakers' Bureau, and have received research grants Boston Scientific Corporation; Dr. Mann has been on the advisory board for Boston Scientific Corporation; Dr. Webster has received research grants Boston Scientific Corporation; Dr. O'Shaughnessy has received consulting fees/honoraria, has been on the Speakers' Bureau, has received research grants, and has been on the advisory board for Boston Scientific Corporation; and Drs. Mandinov, Dawkins, and Baim are full-time employees and stockholders of Boston Scientific Corporation.
- Abbreviations and Acronyms
- confidence interval
- percent diameter stenosis
- long lesion(s)
- major adverse cardiac events
- myocardial infarction
- minimum lumen diameter
- reference vessel diameter
- small vessel(s)
- target lesion revascularization
- Received April 25, 2008.
- Revision received July 25, 2008.
- Accepted September 12, 2008.
- American College of Cardiology Foundation
- Kastrati A.,
- Dibra A.,
- Mehilli J.,
- et al.
- Elezi S.,
- Dibra A.,
- Mehilli J.,
- et al.
- Marzocchi A.,
- Saia F.,
- Piovaccari G.,
- et al.
- Briguori C.,
- Sarais C.,
- Pagnotta P.,
- et al.
- Kastrati A.,
- Mehilli J.,
- Dirschinger J.,
- et al.
- Turco M.A.,
- Ormiston J.A.,
- Popma J.J.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.