Consecutive patients who underwent PCI with at least one second-generation DES for stable angina or ACS between April 2006 and December 2010 were enrolled.

Of these, patients at high-risk who were alive and did not have MI, repeat revascularization, stroke, or major bleeding at the 12-month follow-up after the index procedure were included. High-risk patients required at least one of adverse clinical factors or one of angiographic characteristics as described in previous studies.111314 Clinical high-risk criteria were an age of at least 65 years, female sex, ACS, established vascular disease, diabetes mellitus on medication, and chronic kidney disease.13 Angiographic high-risk criteria included multivessel coronary artery disease, a total stent length of more than 30 mm, a thrombotic target lesion, a bifurcation lesion treated with two stents, an obstructive left main or proximal left anterior descending (LAD) lesion, chronic total occlusion, in-stent restenosis and a calcified target lesion treated with atherectomy.81115 Patients were excluded if they received vitamin K antagonist or antiplatelet agents other than aspirin or clopidogrel. Second-generation DESs were defined as the Biomatrix (Biosensors Europe, Morges, Switzerland), Endeavor, Endeavor Resolute, Endeavor Sprint (Medtronic, Minneapolis, MN, USA), Nobori (Terumo Corporation, Tokyo, Japan), Promus, Promus Element (Boston Scientific, Natick, MA, USA), Xience, or Xience Prime (Abbott Vascular, Abbott Park, IL, USA), PICO Elite (AMG vascular product S.R.L, Winsen, Germany).

Data in regard to clinical outcome and PCI procedure were collected from the prospective PCI registry of Samsung Medical Center. Patients were routinely followed up at 1, 6, and 12 months after the index procedure and every year afterwards. Additional telephone contact or reviewing medical records were done to determine antiplatelet therapy duration and compliance of antiplatelet therapy, if necessary. Mortality data were confirmed by the National Population Registry of the Korea National Statistical Office.

Patients were categorized into two groups according to anti-platelet treatment status at 24 months after index PCI: the clopidogrel group (aspirin discontinued within 24 months) and the prolonged DAPT group (prolonged DAPT more than 24 months). Decisions on the discontinuation of aspirin after 12 month-DAPT following the index PCI were at the attending physicians' discretion. During this study, clopidogrel was the only P2Y12 inhibitor available in Korea.

The primary outcome was a major adverse cardiac and cerebrovascular event (MACCE), a composite of all-cause death, non-fatal MI, definite or probable stent thrombosis, or stroke over 24 months after landmark analysis (12–36 months after the index PCI). Secondary outcomes were the individual components of the MACCE, target vessel revascularization (TVR), target lesion revascularization (TLR), major bleeding and the composite of MACCE and major bleeding.

Cause of deaths were assumed to be cardiac unless an explicit non-cardiac cause was found. MI was defined as elevated cardiac enzyme (troponin I or MB fraction of creatine kinase) level, greater than the upper normal limit with ischemic symptoms or electrocardiographic changes suggesting myocardial ischemia or, MI that required hospitalization with a chief diagnosis of MI.16 Stent thrombosis was defined as the Academic Research Consortium (ARC) criteria.17 Stroke was defined as neurological deficit requiring hospitalization with clinically relevant lesions on brain imaging. TVR was defined as any repeat PCI in the target vessel, while TLR was defined as repeat PCI in the previous stent or occurring within 5 mm of the stent.18 Major bleeding was defined as Bleeding ARC (BARC) criteria type 3 or 5 bleeding. Minor bleeding was defined as BARC type 2 bleeding.19

Categorical variables were expressed as number and percentage and compared using χ² test. Normally distributed continuous variables were presented as mean and standard deviation and compared using unpaired t-test. Non-normally distributed continuous variables were presented as median and interquartile range (IQR) and compared by Wilcoxon rank sum test. Survival curves were drawn between groups using Kaplan-Meier model, and significant differences were calculated using log-rank tests. The Cox proportional hazard model was used to test whether prolonged DAPT was an independent predictor of clinical outcomes. To reduce treatment-selection bias for DAPT and potential confounding factors, we performed propensity score (PS) matching analysis. PSs were obtained by multiple logistic regression in which the antiplatelet regimen is used as the outcome variable, and the selected covariates as predictors. Co-variates which were used for propensity matching are following: age, sex, diabetes, hypertension, current smoking, chronic kidney disease, dyslipidemia, presentation as ACS on the index procedure, history of previous PCI, history of MI, history of previous stroke, PCI on left main artery or LAD artery, 2 or more target vessels, complex PCI, number of stent used, total stent length, stent diameter, use of statin, beta-blockers, angiotensin converting enzyme inhibitors or angiotensin receptor blockers, and calcium channel blockers. The propensity matching was considered to be adequate when overall balance was achieved, indicated by having a standardized mean difference < 0.2. The clopidogrel group and the DAPT group were matched by 1:1 nearest neighbor algorithm. The matched data set was compared by Cox regression model for clinical outcomes. Subgroup analyses were done to assess the treatment effects of prolonged DAPT on the primary end point, compared with clopidogrel monotherapy. The subgroups were age, ACS, ST-segment elevation MI, diabetes, chronic kidney disease, total stent length 28 mm or longer, 2 or more target vessels, and types of DES.

All P values were two-tailed, and P values < 0.05 were considered significant. All data were analyzed using the SPSS version 25 (IBM, Armonk, NY, USA) and R 3.3.3 (R foundation for Statistical Computing, Vienna, Austria) through SPSS custom dialogue.

The Samsung Medical Center Institutional Review Board approved this study and waived the requirement for written informed consent for access to an institutional PCI registry (reference number 2013-12-031). The study procedures followed the Declaration of Helsinki and the ethical standards of the responsible committee on human experimentation.

2,082 patients had PCI with at least one second-generation DES during the study period. Of these, 62 were excluded because of receiving vitamin K antagonist or antiplatelet agents other than aspirin or clopidogrel after index PCI. 308 patients with ischemic or bleeding events and 1,005 patients who stopped clopidogrel within 12 months after the index procedure were excluded. Seventy patients who did not meet the criteria for high-risk PCI at the index procedure were excluded. Finally, 637 patients were included in the present study. Patients were divided into the two groups according to antiplatelet regimens: the clopidogrel monotherapy group and the prolonged DAPT group (Fig. 1). Using the PS matching method, we created 246 matched pairs of patients.

Baseline clinical characteristics are presented in Table 1. There was no significant difference in baseline clinical characteristics, except stent type. Zotarolimus eluting stents were more frequently used in the prolonged DAPT group in the overall population. In the PS-matched population, no significant differences were observed in baseline characteristics between the clopidogrel monotherapy group and the prolonged DAPT group. Angiographic and procedural characteristics are presented in Table 2. In the overall population, multivessel PCI was performed more frequently in the clopidogrel monotherapy group compared with the prolonged DAPT group. The number of stents which were used at the index PCI were significantly greater in the clopidogrel monotherapy group. In the PS-matched population, no significant differences were observed in the angiographic or procedural data between the two groups. Median DAPT duration was 14.7 months (IQR, 5.5) in the clopidogrel monotherapy group and 36.0 months (IQR, 7.5) in the prolonged DAPT group (P < 0.0001) (Table 3).

In the overall population, the risk of MACCE was not significantly different between the clopidogrel monotherapy and prolonged DAPT groups (4.5% versus 4.6%; adjusted hazard ratio [HR], 0.87; 95% confidence interval [CI], 0.40–1.87; P = 0.714) (Fig. 2A and Table 4). In addition, no significant differences were also observed in all-cause death, cardiovascular death, non-fatal MI, stent thrombosis, stroke. There was no significant difference in terms of major bleeding events between the two groups, either (clopidogrel versus DAPT: 1.1% versus 0.3%, P = 0.353), while minor bleeding (BARC type 2) events were reported more frequently in the clopidogrel group (5.6% versus 1.1%; HR, 0.17; 95% CI, 0.05–0.55; P = 0.003) (Table 4).

In the PS-matched population, the risk of MACCE (4.5% versus 4.9%; HR, 1.21; 95% CI, 0.54–2.75; P = 0.643) was not significantly different between the two groups. All-cause mortality, cardiovascular death, non-fatal myocardial infarct, stent thrombosis, stroke, major bleeding, and the composite of MACCE and major bleeding were not significantly different between the two groups, although minor bleeding events were more frequently observed in the clopidogrel group (Fig. 2B and Table 4)

In subgroup analyses, prolonged DAPT did not show significant treatment effect in each subgroup, compared with clopidogrel monotherapy (Supplementary Table 1).