Original Contribution

Optimal Stent Design for High Bleeding Risk Patients: Evidence From a Network Meta-Analysis

Juan G. Chiabrando, MD1,2,3;  Giovanni M. Vescovo, MD4
Marco Giuseppe Del Buono, MD5;  Marco Lombardi, MD5
Massimiliano Camilli, MD5;  Krishna Ravindra, MD1;
Rocco Antonio Montone, MD5,6;  Giampaolo Niccoli, MD5,6;
Giuseppe Biondi-Zoccai, MD, MStat7,8

Juan G. Chiabrando, MD1,2,3;  Giovanni M. Vescovo, MD4
Marco Giuseppe Del Buono, MD5;  Marco Lombardi, MD5
Massimiliano Camilli, MD5;  Krishna Ravindra, MD1;
Rocco Antonio Montone, MD5,6;  Giampaolo Niccoli, MD5,6;
Giuseppe Biondi-Zoccai, MD, MStat7,8

Abstract: Objective. To determine the best stent design for high bleeding risk (HBR) patients. Background. Polymer-free (PF) drug-eluting stent (DES) devices have a proven benefit over bare-metal stent (BMS) devices in previous trials. It is unknown, however, whether polymer-based (PB)-DES devices are as safe as PF-DES devices. Methods. A network meta-analysis including all randomized controlled trials (RCTs) that compared different stent technology in HBR patients with a 1-month course of dual-antiplatelet therapy (DAPT) was performed. The main efficacy outcome was major adverse cardiac event (MACE) rate, defined as the composite of all-cause mortality, myocardial infarction (MI), and target-lesion revascularization (TLR). Secondary efficacy events included all-cause and cardiac mortality, MI, stroke, TLR, and target-vessel revascularization (TVR). Safety outcomes included all bleeding, major bleeding, and stent thrombosis (ST). Results. A total of 4 RCTs with 6456 patients were included. PF-DES and PB-DES yielded a reduced rate of MACE, MI, TLR, and TVR events compared with BMS (all P<.05). ST events were reduced in PB-DES compared with BMS (P=.01). No differences were found in all-cause death, cardiac death, or stroke events in PF-DES and PB-DES compared with BMS. Furthermore, no differences were found between PF-DES and PB-DES regarding any of the outcomes. Conclusion. DES devices were associated with lower MACE and TVR rates compared with BMS, whereas there were no statistical differences in other efficacy endpoints. Also, PB-DES were associated with fewer ST events compared with BMS. There were no statistical differences between PB-DES and PF-DES with regard to any of the endpoints.

J INVASIVE CARDIOL 2021;33(3):E182-E190. Epub 2021 January 21. 

Key words: drug-eluting stents, high bleeding risk, network meta-analysis

Percutaneous coronary intervention (PCI) is the preferred therapeutic approach to treat coronary artery disease (CAD) worldwide.1 Stent placement requires the implementation of dual-antiplatelet therapy (DAPT) for a period of time, in order to reduce the risk of stent thrombosis (ST) and recurrent ischemic events.2 

It is estimated that at least 15% of patients undergoing PCI are at high risk of bleeding events (HBR) and therefore might not be candidates for a prolonged DAPT course.3 Due to a high risk of late ST when DAPT was prematurely suspended in first-generation drug-eluting stent (DES) devices, patients considered to be at HBR often received bare-metal stent (BMS) devices in order to shorten their DAPT duration.4  

Recent evidence suggested that in select patients (ie, HBR patients), shorter DAPT duration may provide benefits in terms of bleeding events without a significant increase in ischemic events. In the STOPDAPT-2 (Short and Optimal Duration of Dual Anti Platelet Therapy After Everolimus-Eluting Cobalt-Chromium Stent) trial, a 1-month DAPT course in an all-comer population reduced bleeding events without increasing cardiovascular events when compared with a 12-month DAPT course.5 

Only recently, patients considered to be at HBR undergoing PCI were included in randomized clinical trials (RCTs) that showed better cardiovascular outcomes with DES (especially polymer-free [PF] stents) compared with BMS in a 1-month DAPT course.6–9 Thus, the paradigm that BMS must be implanted in patients with a reduced DAPT course is no longer valid. PF-DES were introduced with the aim to overcome the risks of late safety and efficacy outcomes associated with the preceding generations of stents.6 Recently, however, a polymer-based (PB) zotarolimus-eluting stent was found to be non-inferior to PF-DES at 1 year with regard to both safety and efficacy among patients at HBR treated with 1 month of DAPT.9 

As differences in both polymer and struts arise between stent designs, we aim to compare differences in clinical outcomes regarding stent types in HBR patients who underwent PCI with a short DAPT duration of 1 month.


This systematic review and meta-analysis is in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement. 

Data sources and searches. We conducted a systematic search of PubMed, Google Scholar, reference lists of relevant articles, and Medline. The search utilized the following terms: “drug eluting stent,” “dual antiplatelet therapy,” “high bleeding risk,” and “high thrombotic risk.” The search for articles compatible with our inclusion and exclusion criteria was performed from inception through March 2020 and returned a combined total of 126 articles. One additional article was included, yielding a total of 127 articles.

Study abstracts were screened for established inclusion and exclusion criteria. Studies thought to be relevant to our search were downloaded and the full manuscripts were reviewed. A thorough search of cited articles within the reviewed manuscripts was assessed for studies not previously identified from the initial database search.

Study selection. We included the articles that satisfied the following inclusion criteria: (1) RCTs comparing different stent technology in HBR patients who underwent 1 month of DAPT; (2) reported follow-up beyond 1 year of treatment; (3) reported cardiovascular outcomes; and (4) reported in English. We excluded non-randomized studies, retrospective cohorts or editorials, and articles that were not in English. The definition of HBR differed among studies, but must have met at least 1 of the following criteria: age ≥75 years; on oral anticoagulation; renal failure, liver disease; recent cancer (<3 months); anemia or transfusion; thrombocytopenia; stroke or intracranial hemorrhage; and hospitalization for bleeding. With the intent of increasing the strength to provide differences among PB-DES and PF-DES, we included bioabsorbable polymer (eg, Synergy; Boston Scientific) and durable polymer (eg, Onyx; Medtronic) as part of the same group (ie, PB-DES). 

Outcomes and definitions. Main efficacy outcomes of interest were major adverse cardiovascular event (MACE) rate, defined as a composite of total death, myocardial infarction (MI), and target-lesion revascularization (TLR). Secondary efficacy outcomes were all-cause death, cardiac death, MI, stroke, TLR, and target-vessel revascularization (TVR).

Safety outcomes included all bleeding events as defined by the Bleeding Academic Research Consortium (BARC 1-5), major bleeding events (BARC 3-5), and ST. ST was considered definite or probable as defined by the Academic Research Consortium (ARC).10

Data extraction and quality assessment. Two investigators (JGC and ML) independently reviewed study titles, abstracts, and articles. Those that satisfied the inclusion criteria were retrieved for full text evaluation. Discrepancies regarding data incorporation to the database were resolved through consensus among the authors. The following data from each selected study were extracted: number of participants; demographics; procedure strategies; and cardiovascular clinical outcomes of interest. Furthermore, we appraised the studies according to the Risk of Bias Assessment Tool, version 2 (RoB 2), as recommended by the Cochrane Collaboration (Supplemental Figure S1).11

Data synthesis and analysis. For inferential purposes, frequentist fixed-effect network meta-analysis was used to estimate the incidence rate ratio (IRR) for incidence of cardiovascular clinical outcomes. A random-effect analysis was conducted when heterogeneity was detected among studies. Heterogeneity values are reported as a percentage in the supplemental tables and figures.12

Descriptive statistics on baseline characteristics of the patients in the studies are provided. Dichotomous variables were reported as counts and percentages, and continuous variables as mean ± standard deviation or as median with or without interquartile range (IQR) if the values were not normally distributed. All P-values reported are two-sided and all confidence intervals (CIs) are calculated at the 95% level. The network meta-analysis was performed with R statistical software (R project for statistical computing, version 3.3.3) using R package “netmeta.” 

Heterogeneity across studies was assessed with Cochran’s Q method. I2 testing was also performed to evaluate the magnitude of the heterogeneity between studies, which was considered substantial when it was >50%. We evaluated the probability (P) scores in order to identify the best-to-worst treatment, taking into account precision and accuracy of effect.


Our initial search retrieved 127 titles, of which 4 studies were ultimately included in our systematic review and network meta-analysis, summarizing the data of 6456 patients at a median follow-up of 1 year (Figure 1). The four RCTs that met the inclusion criteria were: the ZEUS‐HBR subanalysis (Zotarolimus-Eluting Versus Bare-Metal Stents in Uncertain Drug-Eluting Stent Candidates); LEADERS FREE (Polymer-free Drug-Coated Coronary Stents in Patients at High Bleeding Risk); SENIOR (Synergy II Everolimus-Eluting Stent in Patients Older Than 75 Years Undergoing Coronary Revascularization Associated With a Short Dual-Antiplatelet Therapy) subgroup with stable ischemic disease who received 1 month of DAPT (drug-eluting stents in elderly patients with coronary artery disease); and ONYX ONE (Polymer-Based or Polymer-Free Stents in Patients at High Bleeding Risk). Baseline characteristics are presented in Table 1.6-9 Most of the trials included elderly patients with complex coronary disease, and radial artery access was the most commonly used approach. All articles were published between 2015 and 2019. 

Network meta-analysis. Statistical inconsistency and heterogeneity were not significant among the main outcomes of interest (all I2<50% and P-values >.05) (Supplemental Table S2). The evidence network geometry and P-scores are shown in Supplemental Figures S2, S3, and S4). 

Regarding the efficacy endpoints, in the IRR analysis (Table 2 and Figure 2), both newer-generation PB-DES and PF-DES yielded a significant reduction in MACE (IRR, 0.66; 95% CI, 0.57-0.76; P<.001 and IRR, 0.69; 95% CI, 0.61-0.78; P<.001, respectively), MI (IRR, 0.51; 95% CI, 0.34-0.76; P<.001 and IRR, 0.61; 95% CI, 0.41-0.9; P<.01, respectively), TLR (IRR, 0.42; 95% CI, 0.29-0.61; P<.001 and IRR, 0.54; 95% CI, 0.41-0.71; P<.001, respectively), and TVR (IRR, 0.44; 95% CI, 0.31-0.62; P<.001 and IRR, 0.56; 95% CI, 0.43-0.73; P<.001, respectively) when compared with BMS. There were no differences among the stent technology in all-cause death, cardiac death, and stroke (all P>.05). Moreover, there were no statistical differences between PB-DES and PF-DES among any of the efficacy endpoints at a median of 1-year follow-up. 

Regarding the safety endpoints, among the different stent technologies, there were no differences regarding bleeding (BARC 1-5) and major bleeding events (BARC 3-5). PB-DES yielded a reduction in ST events compared with BMS (IRR, 0.59; 95% CI, 0.4-0.88; P=.01). Moreover, there was a trend toward lower rates of ST events with PB-DES compared with PF DES; however, it did not reach statistical significance (IRR, 0.64; 95% CI, 0.39-1.05; P=.08) (Table 3 and Figure 3).


Our network meta‐analysis represents, to the best of our knowledge, the first network comparison between different stent technologies for the assessment of safety and efficacy in HBR patients. Of importance, our analysis demonstrated that both PB-DES and PF-DES are associated with lower MACE rates (primary endpoint) and a lower TVR rate (secondary endpoint) compared with BMS in HBR patients. In particular, the lower MACE rate in the DES cohort was primarily driven by lower MI and TLR, whereas no significant differences were observed for all-cause death, cardiovascular death, or stroke between DES and BMS. There were no statistical differences between PB-DES and PF-DES among any of the efficacy endpoints. Although no differences in major and total bleeding events were found among the 3 groups, PB-DES yielded a significant reduction in ST events compared with BMS (Figure 4). 

These findings are relevant due to the fact that HBR patients were generally under-represented in previous trials. In fact, HBR patients are not only at risk of bleeding, but also ischemic events. The mean pooled age of patients in our cohort was 66 years, with the following comorbid conditions: diabetes mellitus (32.2%); arterial hypertension (79.4%); previous MI (22.5%); multivessel coronary disease (44.7%); chronic kidney disease (30%); and stable coronary artery disease (46.4%). All of these factors portend a higher ischemic risk. Regarding bleeding risk, 39% had previous atrial fibrillation events, 12.2% had anemia, 29.3% were on concomitant anticoagulation, and 7.9% had a history of cancer. This is in line with previous research, in which HBR patients compared with non-HBR patients display more comorbidities, higher lesion complexity, and a higher risk of 4-year mortality.13 Although HBR definitions varied among trials included in our analysis, we did not find heterogeneity or inconsistency in our results.

We chose a 1-month DAPT therapy as the cut-off, because most trials with HBR patients included the use of 1-month DAPT as well. Furthermore, in HBR patients, extending the DAPT course may increase bleeding events without an overt benefit in reducing ischemic events. In the STOPDAPT-2 trial, the 1-month time frame provided benefit in both ischemic and bleeding events.

Few studies have compared DES with BMS for HBR patients.14,15 Unfortunately, the included trials did not have enough power to adequately examine differences in lower-frequency secondary outcomes, such as ST, MI, and TLR. By using a network meta-analysis framework, we were able to show the benefit of DES in reducing the rate of MI and TLR compared with BMS in this population, in part because both the LEADERS FREE and ZEUS-HBR trials showed a reduction in MI and TLR with DES compared with BMS. Thus, our results confirm that routine use of BMS for HBR patients has no benefit with the availability of current-generation DES options. 

Moreover, the inclusion of the BMS group in the analysis allowed us to make direct and indirect comparisons, thus providing further strength in assessing differences between PB and PF stents. Although no difference was found among PF-DES and PB-DES, a trend toward fewer ST events was found in the latter (IRR, 0.64; 95% CI, 0.39-1.05; P=.08). Stent composition differences might have a role in our results (Table 4 details the differences among stent designs). As polymer persistence was thought to be responsible for late restenosis events, PF-DES devices were initially introduced in order to overcome efficacy and safety issues seen in previous stent designs.6 As a drawback, these PF stents have thicker struts (120 microns) compared with the newer, more biocompatible stent polymers (79-91 microns).16 Also, improvements in material alloys, such as the inclusion of chrome, cobalt, and platinum, may compensate for the problems seen with previous-generation DES options.17 

Another key element when suspending DAPT is the risk of developing ST events, due to incomplete strut coverage by early endothelialization. Previous prospective trials have shown a 78.5% strut coverage in Synergy stent deployment at 1 month.18 Our results are in line with the recent ONYX ONE trial, which compared a zotarolimus PB-DES with a biolimus A9 PF-DES, yielding non-inferiority in the primary outcome, defined as a composite of cardiac death, MI, or ST at 1 year, between the two groups (17.1% and 16.9%, respectively).9 We have included bioabsorbable polymer (ie, Synergy) and durable polymer (ie, Onyx) in a single group for inferential purposes. We made this decision based on previous RCTs (ie, Evolve I and II) in which Synergy stents were non-inferior to everolimus-eluting stents.19,20 Several meta-analyses have compared bioabsorbable polymer stents with conventional DES in non-HBR patients.21–23 Late ST occurred less often with bioabsorbable polymer stents compared with first-generation DES (odds ratio, 0.43; 95% CI, 0.24-0.79; P<.01), whereas the risk of late ST was similar between BP-DES and second-generation DES (odds ratio, 0.95; 95% CI, 0.30-3.02; P=.93). There were no significant differences between PB-DES and either first-generation or second-generation DES devices for overall death, MI, or acute/subacute ST. 

Study limitations. Our analysis has several limitations. First, since this is a study-level meta-analysis, no further meta-regression analysis could be done in order to account for differences that arise among different subpopulations included in the analysis. Second, because this is an exclusive RCT network meta-analysis, our results come from only 4 RCTs, limiting the amount of comparisons made. Third, we extrapolate the results to other new-generation DES devices, although they were not included in the analysis, as we did not account for RCTs with these stent technologies. Fourth, some data from our network meta-analysis came from subgroup analyses prespecified in RCT protocols. 


In HBR patients, PB-DES and PF-DES were both associated with lower MACE rates and lower TVR rates compared with BMS devices in patients who underwent a short course of DAPT. These findings clearly suggest that new-generation DES options should represent the gold standard for PCI in HBR patients requiring a short DAPT period. Moreover, our results show that further studies are warranted in order to ascertain the best DES technology in this challenging subset of patients.


1. Levine GN, Bates ER, Blankenship JC, et al. 2015 ACC/AHA/SCAI focused update on primary percutaneous coronary intervention for patients with ST-elevation myocardial infarction an update of the 2011 ACCF/AHA/SCAI guideline for percutaneous coronary intervention and the 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction. J Am Coll Cardiol. 2016;67:1235-1250. Epub 2015 Oct 21.

2. Palmerini T, Benedetto U, Biondi-Zoccai G, et al. Long-term safety of drug-eluting and bare-metal stents: evidence from a comprehensive network meta-analysis. J Am Coll Cardiol. 2015;65:2496-2507. 

3. Rymer JA, Harrison RW, Dai D, et al. Trends in bare-metal stent use in the United States in patients aged ≥65 years (from the CathPCI registry). Am J Cardiol. 2016;118:959-966. Epub 2016 Jul 18.

4. Pfisterer M, Brunner-La Rocca HP, Buser PT, et al. Late clinical events after clopidogrel discontinuation may limit the benefit of drug-eluting stents. An observational study of drug-eluting versus bare-metal stents. J Am Coll Cardiol. 2006;48:2584-2591. 

5. Watanabe H, Domei T, Morimoto T, et al. Effect of 1-month dual antiplatelet therapy followed by clopidogrel vs 12-month dual antiplatelet therapy on cardiovascular and bleeding events in patients receiving PCI: the STOPDAPT-2 randomized clinical trial. JAMA. 2019;321:2414-2427. 

6. Urban P, Meredith IT, Abizaid A, et al. Polymer-free drug-coated coronary stents in patients at high bleeding risk. N Engl J Med. 2015;373:2038-2047. 

7. Varenne O, Cook S, Sideris G, et al. Drug-eluting stents in elderly patients with coronary artery disease (SENIOR): a randomised single-blind trial. Lancet. 2018;391:41-50. 

8. Ariotti S, Adamo M, Costa F, et al. Is bare-metal stent implantation still justifiable in high bleeding risk patients undergoing percutaneous coronary intervention? A pre-specified analysis from the ZEUS trial. JACC Cardiovasc Interv. 2016;9:426-436. 

9. Windecker S, Latib A, Kedhi E, et al. Polymer-based or polymer-free stents in patients at high bleeding risk. N Engl J Med. 2020;382:1208-1218. Epub 2020 Feb 12.

10. Cutlip DE, Windecker S, Mehran R, et al. Clinical end points in coronary stent trials: a case for standardized definitions. Circulation. 2007;115:2344-2351. 

11. Sterne JAC, Savović J, Page MJ, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366:14898. 

12. Biondi-Zoccai G. Network Meta-Analysis: Evidence Synthesis With Mixed Treatment Comparison. Biondi-Zoccai G, ed. Nova Publishers, New York; 2014. 

13. Sorrentino S, Claessen BE, Chandiramani R, et al. Long-term safety and efficacy of durable polymer cobalt-chromium everolimus-eluting stents in patients at high bleeding risk: a patient-level stratified analysis from four post-approval studies. Circulation. 2020;141:891-901. Epub 2020 Jan 29.

14. Neupane S, Khawaja O, Edla S, et al. Meta-analysis of drug eluting stents compared with bare metal stents in high bleeding risk patients undergoing percutaneous coronary interventions. Catheter Cardiovasc Interv. 2019;94:98-104. 

15. Shah R, Rao S V, 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;3:1050-1059. 

16. Busch R, Strohbach A, Rethfeldt S, et al. New stent surface materials: the impact of polymer-dependent interactions of human endothelial cells, smooth muscle cells, and platelets. Acta Biomater. 2014;10:688-700. Epub 2013 Oct 19.

17. Kolandaivelu K, Swaminathan R, Gibson WJ, et al. Stent thrombogenicity early in high-risk interventional settings is driven by stent design and deployment and protected by polymer-drug coatings. Circulation. 2011;123:1400-1409. 

18. Laine M, Dabry T, Combaret N, et al. OCT analysis of very early strut coverage of the Synergy stent in non-ST segment elevation acute coronary syndrome patients. J Invasive Cardiol. 2019;31:10-14. 

19. Meredith IT, Verheye S, Dubois CL, et al. Primary endpoint results of the EVOLVE trial: a randomized evaluation of a novel bioabsorbable polymer-coated, everolimus-eluting coronary stent. J Am Coll Cardiol. 2012;59:1362-1370. Epub 2012 Feb 15.

20. Kereiakes DJ, Meredith IT, Windecker S, et al. Efficacy and safety of a novel bioabsorbable polymer-coated, everolimus-eluting coronary stent. Circ Cardiovasc Interv. 2015;8:1-8. 

21. Lupi A, Gabrio Secco G, Rognoni A, et al. Meta-analysis of bioabsorbable versus durable polymer drug-eluting stents in 20,005 patients with coronary artery disease: an update. Catheter Cardiovasc Interv. 2014;83:E193-E206. Epub 2014 Feb 10.

22. Navarese EP, Kubica J, Castriota F, et al. Safety and efficacy of biodegradable vs durable polymer drug eluting stents: evidence from a meta-analysis of randomized trials. EuroIntervention. 2011;7:985-994. 

23. Wang Y, Liu S, Luo Y, et al. Safety and efficacy of degradable vs. permanent polymer drug-eluting stents: a meta-analysis of 18,395 patients from randomized trials. Int J Cardiol. 2014;173:100-109. Epub 2014 Feb 22.

From the 1VCU Pauley Heart Center, Virginia Commonwealth University, Richmond, Virginia; 2Interventional Cardiology Service, Hospital Italiano de Buenos Aires, Buenos Aires, Argentina; 3Health Science Statistics Applied Laboratory (LEACS), Pharmacology, and Toxicology Department, School of Medicine, University of Buenos Aires, Buenos Aires, Argentina; 4Department of Cardiac Thoracic, Vascular Sciences and Public Health, University of Padua, Padua, Italy; 5Department of Cardiovascular and Thoracic Sciences, Catholic University of the Sacred Heart, Rome, Italy; 6Department of Cardiovascular and Thoracic Sciences, Fondazione Policlinico Universitario A. Gemelli IRCCS, Rome, Italy; 7Department of Medical-Surgical Sciences and Biotechnologies, Sapienza University of Rome, Latina, Italy; and 8Mediterranea-Cardiocentro, Napoli, Italy.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Biondi-Zoccai reports consultant income from Cardionovum, Innovheart, Meditrial, and Replycare. The remaining authors report no conflicts of interest regarding the content herein.

Final version accepted June 22, 2020.

Address for correspondence: Juan Guido Chiabrando, MD, Department of Interventional Cardiology, Hospital Italiano de Buenos Aires, Peron 4190, Buenos Aires, Argentina. P.O. Box C1181ACH. Email: juan.chiabrando@hospitalitaliano.org.ar