The odds ratio estimate of each study is indicated with a square. The size of the square represents the weight of the corresponding study in the meta-analysis.
The odds ratio estimate of each study is indicated by a square whose size represents the weight that the corresponding study had in the meta-analysis.
Individual and pooled odds ratios for stent thrombosis and bleeding complications. The odds ratio estimate of each study is indicated by a square whose size represents the weight that the corresponding study had in the meta-analysis. BARC bleeding indicates bleeding at Bleeding Academic Research Consortium categories 2 through 5.
eTable. Inclusion and exclusion criteria of the RCTs.
eFigure 1. Flow diagram for study selection
eFigure 2. Risk of bias of included randomized controlled trials
eFigure 3. Sensitivity analysis for major adverse cardiac events, target lesion revascularization, target vessel revascularization and myocardial infarction using fixed- fixed-effect models
eFigure 4. Sensitivity analysis for all-cause mortality, cardiac mortality and BARC (2-5) bleeding using fixed-effect models
Customize your JAMA Network experience by selecting one or more topics from the list below.
Shah R, Rao SV, Latham SB, Kandzari DE. Efficacy and Safety of Drug-Eluting Stents Optimized for Biocompatibility vs Bare-Metal Stents With a Single Month of Dual Antiplatelet Therapy: A Meta-analysis. JAMA Cardiol. 2018;3(11):1050–1059. doi:10.1001/jamacardio.2018.3551
What is the optimal stent strategy for coronary intervention in patients with high risk for bleeding in whom a short course of dual antiplatelet therapy is preferred?
In this meta-analysis of randomized clinical trials involving 3943 patients, we found that coronary intervention with drug-eluting stents decreased the risk for myocardial infarction, target vessel revascularization, ischemia-driven target lesion revascularization, and stent thrombosis compared with bare-metal stents used with 1 month of dual antiplatelet therapy.
In patients with high bleeding risk, coronary interventions with drug-eluting stents optimized for biocompatibility appear preferable over those with bare-metal stents.
A significant number of patients receive bare-metal stents (BMSs) instead of drug-eluting stents (DESs) to shorten the duration of dual antiplatelet therapy (DAPT). Emerging evidence suggests that new-generation DESs, particularly those optimized for biocompatibility, may be more efficacious and safer than BMSs, even with a single month of DAPT after stent implantation.
To evaluate the efficacy and safety of DESs compared with BMSs for coronary intervention with a single month of DAPT.
Human studies found in PubMed, the Cochrane databases through April 2018, and reference lists of selected articles.
Randomized clinical trials were included if they enrolled patients undergoing percutaneous coronary intervention and randomly assigned each patient to treatment with either DESs or BMSs. The additional inclusion criterion was use of only 1 month of DAPT poststent implantation.
Data Extraction and Synthesis
Two reviewers independently extracted the data. Odds ratios (ORs) were calculated using random-effects models.
Main Outcomes and Measures
The efficacy end points were major adverse cardiac events, myocardial infarction, target vessel revascularization, ischemia-driven target lesion revascularization, cardiac mortality, and all-cause mortality at 1 year. The safety outcomes were stent thrombosis and bleeding complications.
Data from 3 randomized clinical trials involving 3943 patients were included (2457 men [62.3%]; mean [SD] age ranging from 75.7 [9.3] years to 81.4 [4.3] years per trial subgroup). Coronary intervention with DESs reduced the rates for major adverse cardiac events (OR, 0.68 [95% CI, 0.57-0.82]; P < .001), target lesion revascularization (OR, 0.38 [95% CI, 0.22-0.67]; P = .001), target vessel revascularization (OR, 0.50 [95% CI, 0.38-0.65]; P < .001), and myocardial infarction (OR, 0.51 [95% CI, 0.31-0.83]; P = .01) compared with BMSs at 1 year. The incidence of stent thrombosis was also lower with DESs compared with BMSs (1.8% vs 2.8%), but this difference was not statistically significant in the random-effects model. Additionally, the 2 stent types did not differ in the risks of all-cause mortality, cardiac mortality, and bleeding.
Conclusions and Relevance
In the limited number of randomized clinical trials comparing DESs with BMSs with shortened DAPT durations in patients who have high bleeding risk or are uncertain candidates for prolonged DAPT, coronary intervention with specific DESs optimized for biocompatibility is not only safe but also efficacious, even with only 1 month of DAPT.
Current evidence supports the superiority of drug-eluting stents (DESs) over bare-metal stents (BMSs) for coronary stenosis in most clinical settings.1 However, both stent types induce platelet adhesion and activation, leading to stent thrombosis; therefore, dual antiplatelet therapy (DAPT) is the cornerstone treatment for patients with coronary artery stents.2 Because the risk of stent thrombosis is greatest during the first 30 days after BMS implantation, current guidelines recommend DAPT for at least 1 month in patients with stable ischemic heart disease.2 In the past, safety concerns led to the recommendation for 12 months of DAPT after implantation of a first-generation DES.3,4 Although stent thrombosis risk has significantly decreased with new-generation DESs, current guidelines continue to recommend at least 6 months of DAPT after DES implantation.1,2,5 Estimates show that at least 15% of these patients are not candidates for prolonged DAPT, either because of perceived bleeding risk or because compliance with DAPT is doubted for medical, social, or economic reasons.6-10 Therefore, default management, supported by current guidelines, favors the use of a BMS followed by 1 month of DAPT.2,3 Emerging evidence suggests that new-generation DESs, particularly those optimized for biocompatibility, might not only be safe but also more efficacious compared with BMSs, even when followed by a single month of DAPT.7,9,11-13 To confirm this, a meta-analysis of RCTs was performed, comparing the efficacy and safety of DESs and BMSs when each are followed by a single month of DAPT.
This meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines for systematic reviews and meta-analyses.14 Computerized literature searches of the PubMed and Cochrane databases from their respective inceptions through March 2018 were conducted without language restrictions to locate relevant studies, and relevant articles were cross-referenced. Searches were performed using various combinations of the terms drug-eluting stents, bare-metal stents, percutaneous coronary intervention, antiplatelet therapy, and clinical trial.
Randomized clinical trials (RCTs) were included if they enrolled patients undergoing percutaneous coronary intervention and randomly assigned each patient to treatment with either a DES or BMS. One additional inclusion criterion was use of only 1 month of DAPT after stent implantation. Duplicate reports were excluded, as were those that compared DAPT use that was not of the same length for both the groups receiving DESs or BMSs; where necessary, subgroup data were extracted to meet this criterion.
Two investigators (R.S. and S.B.L.) independently extracted data pertaining to study characteristics, design, and outcomes. The efficacy end points were major adverse cardiac events (MACEs), myocardial infarction (MI), target vessel revascularization (TVR), ischemia-driven target lesion revascularization (TLR), cardiac mortality, and all-cause mortality at 1 year. Study definitions were used for the MACEs outcome.
The safety outcomes were stent thrombosis (definite or probable, as defined by the Academic Research Consortium) and bleeding complications (according to the Bleeding Academic Research Consortium [BARC], classifications 2 through 5).15 In these trials, independent clinical-events committees, whose members were unaware of study group assignments, adjudicated the end points. The potential risk of bias in RCTs was appraised using the Cochrane Collaboration guidelines.16
A standard pairwise meta-analysis was performed according to the Comprehensive Meta-Analysis system, version 3 (Biostat Inc). Pooled odds ratios (ORs) were calculated using random-effects models, because it is the most conservative methodology to account for between-trial heterogeneity. However, in random-effects models, power can be low even when tens of thousands of patients are included if the number of studies is small and between-study variance is nontrivial.17 Therefore, an additional sensitivity analysis was performed using a fixed-effects mode.17 Heterogeneity across trials was evaluated using the Cochran Q test and the Higgins I2 test.18 Publication bias was not assessed because the number of included trials was inadequate (<10) to properly assess a funnel plot or to use more advanced regression-based assessments.19
Three RCTs that included 3943 patients satisfied the inclusion criteria.7,11,13 The search flow diagram is shown in eFigure 1 in the Supplement. The bias assessment for each RCT is shown in eFigure 2 in the Supplement. These studies were high-quality trials based on Cochrane Collaboration guidelines (eFigure 2 in the Supplement). The eTable in the Supplement shows the inclusion and exclusion criteria for each trial.
Among the identified trials, we extracted subgroup data from 2 so that inclusion criteria could be met. Specifically, the Zotarolimus-Eluting Endeavor Sprint Stent in Uncertain DES Candidates (ZEUS) trial prespecified DAPT use based on inclusion criteria11,12; therefore, we used data from the subgroup prespecified to receive the 30-day regimen.11 In the Synergy II Everolimus Eluting Stent in Patients Older Than 75 Years Undergoing Coronary Revascularisation Associated With a Short Dual Antiplatelet Therapy (SENIOR) trial,13 DAPT duration was 6 months for most of the group with acute coronary syndrome and 1 month for the group with stable ischemic heart disease13; we used data only from the group that received DAPT for 1 month.20
The MACES outcome varied between trials. It was defined as a composite of cardiac death, MI, or stent thrombosis in the Prospective Randomized Comparison of the BioFreedom Biolimus A9 Drug-Coated Stent vs the Gazelle Bare-Metal Stent in Patients at High Bleeding Risk (LEADERS FREE) trial7; a composite of all-cause death, MI, or TVR in the ZEUS trial11; and a composite of all-cause mortality, MI, TLR, or stroke in the SENIOR trial.13
The Table shows the basic characteristics of the individual trials. These trials included patients with a broad spectrum of coronary artery diseases, including stable angina, silent ischemia, and acute coronary syndrome. All included trials were multinational, blinded randomized clinical trials. Most patients were male (2457 of 3943 participants [62.3%]) and elderly (with a mean [SD] age ranging by trial and subgroup from 75.7 [9.3] years to 81.4 [4.3] years). Various types of DESs were used in these trials, as shown in the Table. Clopidogrel was the predominant P2Y12 inhibitor used (used in 90% of cases in the LEADERS FREE trial7 and 88% of cases in the SENIOR trial13).
The LEADERS FREE trial7 was the first trial reported in this field, to our knowledge, that was designed to evaluate the efficacy and safety of polymer-free Biolimus A9-coated stents compared with BMSs in patients with increased bleeding risk followed by a single month of DAPT in both groups.7 At 390 days, the primary safety end point (a composite of cardiac death, MI, and stent thrombosis) was lower with patients who had received DESs (112 events in 1221 patients [9.4%]) compared with patients who had received BMSs (154 events in 1211 patients [12.9%]; P = .01), driven by a lower rate of MI (DESs: 72 events in 1221 patients [6.1%]; BMSs: 104 events in 1211 patients [8.9%]; P = .01). The stent thrombosis rate was similar between the 2 groups. Similarly, the primary efficacy end point (clinically driven TLR) was significantly lower with patients with DESs (59 events in 1221 patients [5.1%]) compared with those with BMS (113 events in 1211 patients [9.8%]; P < .001). The second reported trial in this field, the ZEUS trial,11 was designed to evaluate the combined efficacy and safety of zotarolimus-eluting endeavor sprint stents (ZESs) compared with BMSs in uncertain candidates for DESs. In this trial, DAPT duration was prespecified based on inclusion criteria.11,12 Patients with high bleeding risk were prespecified to receive a 30-day DAPT regimen irrespective of the stent type.11 For this cohort, MACE rates at 12 months were lower with DESs (96 events in 424 patients [22.6%]) compared with BMSs (117 events in 404 patients [29%]) driven by both a lower MI rate (DESs: 15 events in 424 patients [3.5%]; BMSs: 42 events in 404 patients [10.4%]; P < .001) and a lower TVR rate (DESs: 25 events in 424 patients [5.9%]; BMSs: 46 events in 404 patients [11.4%]; P = .01) in the group receiving DESs. In addition, definite or probable stent thrombosis (DESs: 11 events in 424 patients [2.6%]; BMSs: 25 events in 404 patients [6.2%]; P = .02) and definite, probable, or possible stent thrombosis (DESs: 28 events in 424 patients [6.6%]; BMSs: 43 events in 404 patients [10.6%]; HR, 0.61 [95% CI, 0.38-0.98]; P = .04) were lower with DESs. Finally, in the most recently reported SENIOR trial,13 the safety and efficacy of DESs and BMSs were compared in patients 75 years or older.13 One of the prespecified subgroups was patients with stable angina or silent ischemia who had received only 1 month of DAPT, irrespective of stent type. In this trial, MACE rates were lower with DESs (30 events in 297 patients [9.4%]) compared with BMSs (52 events in 312 patients [15.7%]). The rates of stent thrombosis and mortality were not different between the 2 stent types.
Coronary intervention with DES (optimized for biocompatibility), compared with BMSs, reduced MACE rates at 1 year (OR, 0.68 [95% CI, 0.57-0.82]; P < .001; Figure 1). No between-trial heterogeneity was found (Q = 0.35; P = .83; I2 = 0%). Similarly, TLR rates were reduced with the use of DESs compared with BMSs (OR, 0.38 [95% CI, 0.22-0.67]; P = .001), as were TVR rates (OR, 0.50 [95% CI, 0.38-0.65]; P < .001; Figure 1). Additionally, MI risk was reduced in patients treated with DESs compared with those treated with BMSs (OR, 0.51 [95% CI, 0.31-0.83]; P = .01; Figure 2). However, no significant differences were found between the 2 stent types for the risks of all-cause mortality (OR, 0.87 [95% CI, 0.70-1.08]; P = .21) and cardiac mortality (OR, 0.84 [95% CI, 0.63-1.11]; P = .22; Figure 2). No statistically significant heterogeneity between the trials was found for these outcomes. Sensitivity analysis using a fixed-effects model did not change the summary results for any of these outcomes (eFigures 3 and 4 in the Supplement).
The incidence of stent thrombosis at 1 year was lower with DESs compared with BMSs (36 events in 1972 patients [1.8%] vs 56 events in 1971 patients [2.8%]). However, this difference was not statistically significant (OR, 0.56 [95% CI, 0.27-1.17]; P = .13) in the random-effects model, but was statistically significant in the fixed-effects model (OR, 0.64 [95% CI, 0.41-0.99]; P = .045) (Figure 3). In addition, moderate between-trial heterogeneity (Q = 4.09; P = .12; I2 = 51.1%) was found for the risk of stent thrombosis.
Finally, no significant differences were found between the 2 stent types for the risk of bleeding on Bleeding Academic Research Consortium scale levels 2 through 5 (OR, 0.84 [95% CI, 0.66-1.08]; P = .20; Figure 3). No between-trial heterogeneity was found (Q = 2.33; P = .31; I2 = 14.1%) for the bleeding complication. Sensitivity analysis using fixed-effects models did not change the summary results (eFigure 4 in the Supplement).
In this study of 3943 patients enrolled in 3 RCTs, we compared the efficacy and safety of DESs (optimized for biocompatibility) and BMSs with a single month of DAPT. No significant differences in the rates of cardiac mortality, all-cause mortality, or bleeding consistent with Bleeding Academic Research Consortium scale levels 2 through 5 were found between the 2 stent types. The rate of stent thrombosis was numerically lower with DES, but the difference was not statistically significant in the random-effects model. However, DES use significantly decreased risks of MACEs, MI, TVR, and TLR at 1 year compared with BMS use. These results suggest that guidelines recommendations regarding the routine use of BMSs for patients with high bleeding risk or those who are uncertain candidates for DAPT (with the sole goal of shortening DAPT duration) are no longer warranted with the availability of current-generation DES, particularly those optimized for biocompatibility.
Several RCTs have shown that DESs, compared with BMSs, markedly decrease the incidence of restenosis and TVR. Thus, they have become the mainstay for native vessel percutaneous coronary intervention.1
However, because very late stent thrombosis with first-generation DES has a relatively high incidence, a prolonged course of DAPT was mandated after DES implantation.4 It is estimated that at least 15% of patients might not be candidates for prolonged DAPT because of high bleeding risk or were uncertain candidates owing to poor compliance.6-8 The use of a DES is an American Heart Association/American College of Cardiology class III (harmful) recommendation for those patients who are not likely to be able to tolerate and comply with prolonged DAPT or if this cannot be determined before stent implantation.3 Therefore, their default management, supported by current guidelines, favors the use of a BMS followed by 1 month of DAPT.2,3 A recent study using the CathPCI Registry showed that 1 in 5 contemporary percutaneous coronary intervention procedures in patients aged 65 years or older continue to use BMS implantation.10
Recent RCTs and meta-analyses have suggested that new-generation DESs have better safety and efficacy profiles than first-generation DESs.1 In addition, recent network meta-analyses suggest that new-generation DESs might be associated with lower rates of stent thrombosis compared with BMSs.1,5 A possible factor confounding the results of those network meta-analyses is their varying durations of DAPT between DESs and BMSs.1 However, additional sensitivity analyses suggest a significant difference in stent thrombosis rates between DESs and BMSs as few as 30 days after implantation, a period during which all patients are treated with DAPT irrespective of stent type.5 Therefore, based on these safety data, new RCTs have been conducted in patients with high bleeding risk to compare new-generation DESs with BMSs followed by a single month of DAPT.7,11-13
Consistent with these trials, our meta-analysis suggests that DESs are not only safe but more efficacious compared with BMSs with a single month of DAPT. It also shows that DESs not only improved soft outcomes (ie, revascularization) compared with BMSs, but also hard outcomes, such as MI, with this shorter duration of DAPT. The risk of stent thrombosis was lower with DESs in this meta-analysis, but this did not reach statistical significance in random-effects models, most likely because of a type II error (lack of power). In random-effects models, power can be low even if tens of thousands of patients are included when the number of studies is small.17 We also identified moderate heterogeneity for the outcome of stent thrombosis; this is probably driven by the ZEUS trial,11 given the rate of stent thrombosis was higher in this trial compared with the others. The higher stent thrombosis rate in the ZEUS trial11 can be explained by a much greater proportion of patients with ST elevation MIs (15.3%) in this trial compared with the others. Additionally, bare-metal in-stent restenosis is not always a benign clinical presentation: studies have shown that more than one-third of BMS restenoses can present as MI or unstable angina requiring hospitalization.21 A recent individual patient-level meta-analysis of 21 RCTs including 32 524 patients showed that nonemergent, uncomplicated TLR after percutaneous coronary intervention is an independent predictor of long-term mortality, driven partly by higher rates of MI.22 Therefore, in our meta-analysis, significantly lower MI rates with DESs are likely driven by lower stent thrombus-associated and restenosis-associated MI. Prevailing wisdom also suggests that in patients with high bleeding risk, DES implantation followed by a single month of DAPT will also be safer than the same with BMS implantation because the latter will be associated with higher rates of TVR requiring repeated DAPT. In the ZEUS trial, despite comparable protocol-mandated DAPT durations in both stent groups, cumulative treatment duration with DAPT was significantly longer with BMSs.11,12 This leads to a nearly 2-fold rate of cumulative bleeding end points in the BMS group compared with the DES group.11,12
Consistent with our findings, several observational studies have challenged the current contemporary practice of the preference for BMS use in patients who plan to undergo noncardiac surgery.23-25 Current evidence suggests that contemporary DES use is associated with better outcomes than BMS use in patients undergoing noncardiac surgery, even with early (1-6 months) DAPT discontinuation after noncardiac surgical procedures.23-25 Similarly, several recent RCTs, meta-analyses, and observational studies have consistently shown that new-generation DESs are associated with lower risks for stent thrombosis and revascularization compared with BMSs across a broad spectrum of patient populations.1,5,26-30 The lower risk of early stent thrombosis with DESs seems to be the result of reduced acute thrombogenicity from polymer coatings.31-33 It has been shown that well-designed polymer coatings serve as corrosive barriers and provide thromboresistance through modifications of certain properties such as surface potential, wettability, and roughness.32,33 On the other hand, the dominant mechanism for the lower rate of revascularization with DESs results from inhibition of in-stent neointimal hyperplasia from the antiproliferative and anti-inflammatory medication eluted by DESs.31,33 However, the lower risk of restenosis with first-generation DESs has been associated with increases in late ischemic events attribute to delayed endothelial healing, polymer hypersensitivity reactions, positive remodeling with late acquired malposition, and neoatherosclerosis.31,33 Those shortcomings have been addressed with the current generation of DESs through the use of biocompatible polymers, biodegradable polymers, or removal of the polymer, along with improved platform design, thinner struts, and novel antiproliferative drugs.31,33 Despite improvements in stent technology, late stent failure resulting from neoatherosclerosis remains a concern even with new-generation DESs.31,34 Neoatherosclerosis is the development of atherosclerotic plaque inside an implanted coronary stent, and it occurs more with DESs compared with BMSs.33,34 The exact mechanisms underlying the development of neoatherosclerosis remains unknown; however, the presence of an incompetent endothelium (with an impaired endothelial cell barrier) is more likely to lead to neoatherosclerosis after stenting with a DES.33,34 Therefore, long-term follow-up of patients in the reported trials will demonstrate whether benefits persist for several years.
This meta-analysis has several limitations. First, some data were based on subgroups, and the findings should be interpreted cautiously; however, all subgroup analyses were prespecified in the trial protocols. Second, we did not have access to individual-participant data; therefore, the data we analyzed were combined from various studies, each with its own protocol, inclusion and exclusion criteria, and definitions. Specifically, the criteria defining high bleeding risk varied across trials; however, most patients in these trials were enrolled based on older age as the high bleeding risk criteria. In addition, most of those included in this analysis were patients with stable coronary artery disease. Therefore, our findings might not be generalizable to younger patients with acute coronary syndrome. Third, in the ZEUS trial,11 the Endeavor ZES stent was used, but it is no longer commercially available. Furthermore, the BioFreedom stent was used in the LEADERS FREE trial,7 and it is based on a poorer stent platform in that the struts are thicker in comparison with most contemporary DES models. In addition, most patients in this meta-analysis received polymer-free or bioresorbable polymer DESs. Even so, our meta-analysis did not include all contemporary bioabsorbable polymer DESs available worldwide.35 Thus, our findings should not be generalized across all new-generation DES platforms (either with durable polymer or bioresorbable polymer platforms). Despite these limitations, this is the first meta-analysis addressing this topic, and it will assist physicians in deciding the best stent strategy for patients who have high bleeding risk or are uncertain candidates for prolonged DAPT.
Based on this meta-analysis, coronary intervention with DESs compared with BMSs decrease the risks of MACE, TLR, TVR, and MI with a single month of DAPT in patients with high bleeding risk or who are uncertain candidates for prolonged DAPT. No differences in the rates of all-cause mortality, cardiac mortality, or bleeding at Bleeding Academic Research Consortium levels 2 through 5 were found. Thus, based on these findings, in patients whose adherence to DAPT beyond 30 days is uncertain, specific DES models optimized for biocompatibility should be considered as the default therapy rather than BMSs. These data also suggest that current guidelines recommendations regarding the use of BMSs should be revisited.
Corresponding Author: Rahman Shah, MD, Division of Cardiovascular Medicine, School of Medicine, University of Tennessee, 1030 Jefferson Ave, Memphis, TN 38104 (email@example.com).
Accepted for Publication: September 11, 2018.
Published Online: October 31, 2018. doi:10.1001/jamacardio.2018.3551
Author Contributions: Dr Shah had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Shah.
Critical revision of the manuscript for important intellectual content: Rao, Latham, Kandzari.
Statistical analysis: Shah, Latham, Kandzari.
Administrative, technical, or material support: Shah.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Kandzari reports grants and personal fees from Medtronic CardioVascular outside the submitted work. No other disclosures were reported.
Create a personal account or sign in to: