Author + information
- Received May 3, 2013
- Accepted May 9, 2013
- Published online September 1, 2013.
- George D. Dangas, MD∗,†∗ (, )
- Patrick W. Serruys, MD, PhD‡,
- Dean J. Kereiakes, MD§,
- James Hermiller, MD‖,
- Ali Rizvi, MD‖,
- William Newman, MD¶,
- Krishnankutty Sudhir, MD, PhD#,
- Robert S. Smith Jr., PhD#,
- Sherry Cao, MS#,
- Kleanthis Theodoropoulos, MD∗,†,
- Donald E. Cutlip, MD∗∗,
- Alexandra J. Lansky, MD†† and
- Gregg W. Stone, MD†,‡‡
- ∗Division of Cardiology, Mount Sinai Medical Center, New York, New York
- †Cardiovascular Research Foundation, New York, New York
- ‡Cardialysis and Thoraxcenter, Rotterdam, the Netherlands
- §The Christ Hospital Heart and Vascular Center/Lindner Research Center, Cincinnati, Ohio
- ‖The Heart Center of Indiana, Indianapolis, Indiana
- ¶Division of Cardiology, Wake Medical Center, Raleigh, North Carolina
- #Abbott Vascular, Santa Clara, California
- ∗∗Division of Cardiology, Harvard Clinical Research Institute, Boston, Massachusetts
- ††Division of Cardiology, Yale University Medical Center, New Haven, Connecticut
- ‡‡Division of Cardiology, Columbia University Medical Center/New York-Presbyterian Hospital, New York, New York
- ↵∗Reprint requests and correspondence:
Dr. George D. Dangas, Mount Sinai Medical Center, Cardiovascular Research Foundation, 111 East 59th Street, 11th Floor, New York, New York 10022.
Objectives This study sought to investigate whether the everolimus-eluting stent (EES) is superior to the paclitaxel-eluting stent (PES) with respect to long-term individual clinical outcomes.
Background Individual studies have indicated a clinical advantage of coronary EES compared with PES with respect to restenosis and the composite endpoint of major adverse cardiac events. However, these trials were not powered for superiority in low-frequency event rates and have reported limited data beyond 1-year follow-up.
Methods We conducted a meta-analysis of the final 3-year results from the international SPIRIT (Clinical Evaluation of the Xience V Everolimus Eluting Coronary Stent System in the Treatment of Patients With De Novo Native Coronary Artery Lesions) II, III, and IV clinical trials. Individual patient data from 4,989 patients who were prospectively randomized to treatment with EES (n = 3,350) or PES (n = 1,639) were pooled for analysis.
Results At 3-year follow-up, EES was superior to PES in reducing the following event rates: target lesion failure (8.9% vs. 12.5%, hazard ratio [HR]: 0.71, 95% confidence interval [CI]: 0.59 to 0.85; p = 0.0002), all-cause mortality (3.2% vs 5.1%, HR: 0.65, 95% CI: 0.49 to 0.86; p = 0.003), myocardial infarction (3.2% vs. 5.1%, HR: 0.64, 95% CI: 0.48 to 0.85; p = 0.002), cardiac death or myocardial infarction (4.4% vs. 6.3%, HR: 0.70, 95% CI: 0.54 to 0.90; p = 0.005), ischemia-driven target lesion revascularization (6.0% vs. 8.2%, HR: 0.72, 95% CI: 0.58 to 0.90; p = 0.004), stent thrombosis (0.7% vs. 1.7%, HR: 0.45, 95% CI: 0.26 to 0.78; p = 0.003), and major adverse cardiac events (9.4% vs. 13.0%, HR: 0.71, 95% CI: 0.60 to 0.85; p = 0.0002). No interaction was present between stent type and the 3-year relative rates of target lesion failure across a broad range of subgroups, with the exception of diabetes and vessel (left anterior descending vs. other).
Conclusions In this large dataset with 3-year follow-up, coronary implantation of EES compared with PES resulted in reduced rates of all-cause mortality, myocardial infarction, ischemia-driven target lesion revascularization, stent thrombosis, and target lesion failure. Further research is warranted to characterize possible interactions between stent type, diabetes, and vessel.
In several individual clinical trials, everolimus-eluting stents (EES) were shown to have advantages over paclitaxel-eluting stents (PES), including reduced rates of restenosis, target lesion revascularization (TLR), and composite endpoints such as major adverse cardiac events (MACE) (1–8). In addition, recent meta-analyses indicated that EES have lower rates of stent thrombosis than other types of drug-eluting and bare-metal stents do (9–11). However, few of these studies have reported follow-up beyond 1 year, and no study has been individually powered to investigate differences in low-frequency event rates such as death or myocardial infarction (MI).
We therefore conducted the present meta-analysis from the prospective, multicenter, SPIRIT (Clinical Evaluation of the Xience V Everolimus Eluting Coronary Stent System in the Treatment of Patients With De Novo Native Coronary Artery Lesions) clinical trials program (SPIRIT II, III, and IV) that included randomization of 4,989 patients to EES versus PES, with follow-up over 3 years.
The SPIRIT II, III, and IV trial designs and primary results have been previously described (1–8,12). In summary, all 3 trials were randomized, single-blind, active-control designs comparing EES (an 81-μm strut thickness cobalt chromium metallic stent that elutes everolimus from a durable fluorocopolymer [Xience V, Abbott Vascular, Santa Clara, California]) versus PES (a 132-μm strut thickness stainless steel stent that elutes paclitaxel from a durable polymer [Taxus Express, Boston Scientific, Natick, Massachusetts]) (6). Exclusion criteria for high-risk patients and lesions were: acute or recent MI; left ventricular ejection fraction <30%; lesions that were in a bypass graft conduit; occluded vessels; bifurcations (minor bifurcations were included in SPIRIT IV); ostial lesions (ostial right coronary artery lesions were included in SPIRIT IV); and severely calcified or tortuous vessels. The studies were individually approved by the institutional review board at each participating center, and consecutive, eligible patients signed informed, written consent. All 3 trials used the same data element definitions, enabling pooling of individual patient data for the present meta-analysis. Table 1 shows the key methodological characteristics of these 3 trials.
Clinical follow-up was scheduled at 30, 180, 270, and 365 days and then yearly; complete 3-year follow-up is available from all 3 trials and, therefore, data until this time point were used in the present analysis. Although the operators were by necessity unblinded during the stent procedure, the patient and staff involved in follow-up assessments remained blinded, with a standardized follow-up interview script used to reduce bias. In addition, all angiograms were reviewed by dedicated core laboratories that were blinded to the treatment assignment and followed established analytical methods (13).
Independent study monitors verified all case report form data. Reportable clinical events were adjudicated by an independent committee blinded to treatment allocation. The data safety and monitoring committee of each trial had periodically reviewed blinded safety data, each time recommending the study to continue without modification. Independent core angiographic laboratory analyses were performed by technicians blinded to treatment assignment as previously described (14).
Endpoints and definitions
The composite endpoint of ischemia-driven target lesion failure (TLF) was defined as the occurrence of cardiac death, target-vessel MI, or ischemia-driven TLR by either percutaneous coronary intervention (PCI) or bypass graft surgery (15). MACE was defined as the composite of cardiac death, MI, or ischemia-driven TLR. Target vessel failure was defined as all-cause death, MI, or ischemia-driven target vessel revascularization. Stent thrombosis was prospectively protocol-defined as an acute coronary syndrome with angiographic thrombus within or adjacent to a stent, or in the absence of angiography, any unexplained death or acute MI in the target lesion distribution within 30 days. Stent thrombosis was also adjudicated using the Academic Research Consortium definite or probable criteria (16).
Categorical variables were compared by the Fisher exact test. Continuous variables are presented as mean ± SD and were compared by the Student t test. All data are presented in the intent-to-treat population, consisting of all patients randomized, regardless of the treatment actually received. Time-to-event curves using all available follow-up data were also constructed using Kaplan-Meier estimates and compared by log-rank test. Subgroup analyses were conducted to further investigate the potential impact of baseline variables on the TLF composite endpoint using Cox proportional hazards regression with formal interaction testing. Multivariable analyses were also conducted using Cox regression to identify independent predictors of 3-year clinical outcomes; candidate variables included in each model are listed in the footnote of their respective table. A 2-sided alpha = 0.05 was used for all superiority testing. All statistical analyses were performed using SAS (version 9.1.3, SAS Institute, Cary, North Carolina).
A total of 4,989 patients were randomized to EES (3,350 patients with 4,174 treated lesions) versus PES (1,639 patients with 2,059 treated lesions); the patient contribution from each individual trial is shown in Table 1. Baseline clinical and angiographic characteristics were well matched between the groups (Table 2). There were trends toward fewer unstable angina patients and more American College of Cardiology/American Heart Association Class C lesions in the EES versus the PES group. Procedural outcomes of PCI are also summarized in Table 2, and they did not differ significantly between the 2 groups.
Three-year clinical event rates are detailed in Table 3 and Figures 1 to 4⇓⇓⇓. All point estimates were consistently indicating an advantage of treatment with EES versus PES. EES resulted in statistically significant reductions in the individual 3-year endpoints of all-cause mortality, MI (including the individual subcategories of Q-wave, non–Q-wave, and target vessel MI), ischemia-driven TLR, and stent thrombosis (both protocol- and Academic Research Consortium–defined). Stent thrombosis was reduced with EES compared with PES both within the first year after stent implantation, and between years 1 and 3 (Fig. 4). Superiority of EES was also present for the composite endpoints of death or MI, cardiac death or MI, MACE, TLF, and target vessel failure. By multivariable analysis, randomization to EES versus PES was an independent predictor of all individual and composite endpoints examined: all-cause mortality; MI; ischemia-driven TLR; TLF; and stent thrombosis (Table 4).
Subgroup analyses according to various baseline or procedure variables in relation to the TLF outcome at 3-year follow-up are shown in Table 5. The point estimates were consistently in favor of EES compared with PES, with no significant interactions detected except for subgroups according to diabetic status and target lesion location. With respect to the diabetic subgroup interaction, EES was superior to PES in nondiabetic patients with respect to TLF at 3 years, whereas diabetic patients had comparable event rates with both stents (p for interaction = 0.02). This interaction was most pronounced among the insulin-treated diabetic subgroup, whereas no significant interaction was observed when only the non-insulin-treated diabetic subgroup was considered. A large margin of superiority for EES versus PES in TLF at 3 years was also observed in the subgroup undergoing PCI of the left anterior descending (LAD) coronary artery, whereas both stents resulted in comparable results in non-LAD coronary arteries (p for interaction = 0.03).
Previous results of the SPIRIT II and III trials showed superiority of EES versus PES in angiographic restenosis (1,2), and noninferiority with respect to composite safety and efficacy endpoints at 1- and 2-year follow-up (3–8). Clinical superiority of EES over PES for TLF and other composite safety and efficacy endpoints was first documented in the SPIRIT IV trial at 1-year follow-up (3). The present meta-analysis further substantiates and extends these findings by demonstrating: 1) that the benefits of EES relative to PES were not only sustained but also continued to increase in magnitude over 3-year follow-up; and 2) significant reductions with EES compared with PES are also present for the safety endpoints of all-cause and cardiac death, MI, and stent thrombosis, findings made possible by the greater power of the combined dataset.
Differences in death and MI are rarely demonstrated in PCI trials in which 2 devices are compared (17). Therefore, the reduction in the individual endpoints of all-cause death and MI present with EES compared with PES from the current meta-analysis supports the contention that new-generation devices can indeed affect pathobiological processes that may afford improved outcomes in “hard endpoints” over time. We have previously highlighted that: 1) stent restenosis and thrombosis are distinct complications; 2) each one of them can affect patient outcomes; and 3) there is a tradeoff between these 2 events in patients treated with PES versus bare-metal stents that may minimize the clinical benefit of the drug-eluting stent (18). The observed combination of improved efficacy (reduced rates of TLR) and safety (reduced stent thrombosis) with EES versus PES maximizes the potential for net clinical benefit. In addition, the longer duration of follow-up in the present study compared with that of prior studies was required to demonstrate the improvement in freedom from all-cause and cardiac death with EES compared with PES. This mortality reduction may be due to the lower rate of very late (between 1 and 3 years of follow-up) stent thrombosis and MI with EES compared with PES.
The relative benefits of EES compared with PES were consistent across the multiple subgroups examined, with the possible exception of patients with diabetes mellitus and non-LAD PCI, in whom significant differences between EES and PES were not found. The mechanisms of restenosis and progressive atherosclerosis may indeed differ between diabetic and nondiabetic patients (19), and the restenosis implications in the LAD territory may be more clinically overt than in other vessel distributions. However, the interaction tests for these 2 subgroups were both borderline (p = 0.02 and 0.03, respectively) and were not corrected for multiple comparisons. Further studies are warranted to determine whether there are indeed differences between EES and PES according to these subgroups.
The present study demonstrating reduced rates of stent thrombosis with EES compared with PES are concordant with other direct and indirect (network) meta-analyses that examined other types of comparator stents (9–11). Potential reasons for the lower rates of stent thrombosis with EES versus PES may include: 1) thinner stent struts; 2) the fact that everolimus is completely eluted within 90 days, whereas paclitaxel is retained locally in the polymer for a prolonged period of time; 3) the different modes of action of the 2 drugs, with potentially more cytotoxicity with paclitaxel; 4) the presence of a thromboresistant fluoropolymer (8); and 5) greater conformability with a lower risk of strut fracture with the cobalt chromium EES platform compared with the thicker stainless steel device. The Taxus Liberté (Boston Scientific, Natick, Massachusetts), a thinner strut stainless steel PES, has been previously tested with 1- to 2-year safety results similar to those found in the SPIRIT trials with Taxus Express (6,7). Thus, reducing the metallic stent strut thickness alone was not sufficient to reduce stent thrombosis.
Despite the large size of the pooled population, differences in very low-frequency adverse events may have escaped detection. Similarly, subgroup testing is inherently underpowered and should be considered hypothesis-generating. Routine angiographic follow-up was performed in a subset of patients in the SPIRIT II and III trials, which may have increased the absolute difference in ischemia-driven TLR between the 2 stents, but would not have affected their relative efficacy, nor measures of stent safety (1–4). Finally, the SPIRIT trials excluded numerous high-risk categories of patients and lesions, including those with thrombus and ST-segment elevation MI, true bifurcation lesions, chronic total occlusions and saphenous vein grafts, and heavily calcified and tortuous lesions. However, additional data from an all-comers randomized trial suggest that EES may also be safer and more effective than PES in more complex subsets of patients and lesions (6).
The results from this large-scale meta-analysis from 3 similarly performed, prospective randomized trials with 3-year follow-up after the index coronary implantation of EES compared with PES demonstrate substantially reduced rates of all-cause mortality, MI, ischemia-driven TLR, stent thrombosis, TLF, target vessel failure, and MACE with EES. These data emphasize that stent selection decisions can markedly affect the long-term outcomes of patients with coronary artery disease undergoing PCI. Finally, the findings of the present analysis challenge the generalizability (to contemporary clinical decision making) of the results from previously completed large clinical trials that had used less safe and/or effective drug-eluting stent devices that are no longer in widespread use.
The SPIRIT II, III, and IV trials were sponsored and funded by Abbott Vascular. Dr. Dangas has reported that his spouse has received honoraria for serving on the advisory board of Abbott Vascular and Boston Scientific. Dr. Kereiakes has received consulting fees from Abbott Vascular, Boston Scientific, and REVA Medical. Dr. Hermiller has received consulting fees from Abbott Vascular and Boston Scientific. Dr. Sudhir, Dr. Smith, and Ms. Cao are full-time employees of and have stock and/or options in Abbott Vascular. Dr. Cutlip has received research support paid to his institution from Medtronic, Abbott Vascular, and Boston Scientific; and has received consulting fees from St. Jude Medical, Stentys, and Celonova. Dr. Stone has received consulting fees from Boston Scientific. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- everolimus-eluting stent(s)
- left anterior descending
- major adverse cardiac events
- myocardial infarction
- percutaneous coronary intervention
- paclitaxel-eluting stent(s)
- target lesion failure
- target lesion revascularization
- Received May 3, 2013.
- Accepted May 9, 2013.
- American College of Cardiology Foundation
- Stone G.W.,
- Midei M.,
- Newman W.,
- et al.,
- for SPIRIT III Investigators
- Lansky A.J.,
- Applegate R.,
- Hermiller J.B.,
- et al.
- Smits P.C.,
- Kedhi E.,
- Royaards K.J.,
- et al.
- Baber U.,
- Mehran R.,
- Sharma S.K.,
- et al.
- Palmerini T.,
- Kirtane A.J.,
- Serruys P.W.,
- et al.
- Lansky A.J.,
- Dangas G.,
- Mehran R.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.,
- for the Academic Research Consortium
- King S.B. 3rd.,
- Smith S.C. Jr..,
- Hirshfeld J.W. Jr..,
- et al.
- Stone G.W.,
- Ellis S.G.,
- Colombo A.,
- et al.
- Marso S.P.,
- Mercado N.,
- Maehara A.,
- et al.
- Rocic P.