In our cohort study between January 2004 and December 2011, 124 multivessel CAD patients who were candidates for CABG but received PCI with multiple coronary stents were enrolled. Making the decision to receive PCI instead of CABG may be either due to a physician's opinion that the patient is not a good candidate or because it is the patient's preference. Patients were excluded if 1) with cardiogenic shock before PCI, 2) had contraindications for dual antiplatelet therapy, 3) had been diagnosed with cancer before they received PCI, or 6) PCI details, or follow-up data were missing (Fig. 1).

Significant coronary lesions that required PCI were defined as those with a stenosis of ≥70% in a segment with a reference diameter of ≥2.0 mm as shown by quantitative coronary angiography. For patients who presented with myocardial infarction with ST elevation, the non-infarct-related lesions were treated by separate interventions within 3 months. All complex lesions, including left main lesions, bifurcation lesions, calcified lesions, long lesions, chronic total occlusion, and ostial lesions of the 3 major coronary arteries had been limited to be treated with DES. Because DES is not covered by Taiwan national insurance, other significant lesions may been treated by bare-metal stents (BMS) according to patient preferences and physician treatment practices. The types of DES implanted were as follows: 1) sirolimus-eluting stents; 2) paclitaxel-eluting stents; 3) everolimus-eluting stents; 4) biolimus A9-eluting stents; and 5) zotarolimus-eluting stents. The choice regarding the use of a specific type of DES was left to the operator's discretion. Sirolimus-eluting and paclitaxel-eluting stents are classified as a first generation DES. Procedural success was defined as a final diameter stenosis of <20% by quantitative coronary angiography (QCA) with normal flow according to the thrombolysis in myocardial infarction (TIMI)-3 criteria. Complete revascularization was defined as successful angiographic results in the major epicardial vessels and their major branches with a reference diameter of ≥2.0 mm, and it had intentionally been done. Follow-up coronary angiography was not routinely performed except for patients with recurrent angina, myocardial ischemia revealed by noninvasive stress tests, or left main CAD.

Data regarding the demographics, clinical, angiographic, and procedural characteristics, and outcomes were collected from the electronic medical records system of the hospital, supplemented by information from patients and their physicians. Chronic kidney disease (CKD) stage ≥3 was defined as an estimated glomerular filtration rate (eGFR) of <60 mL/min/1.73 m², which was calculated by a 4-variable modification of diet in a Renal Disease study equation.7) The SYNTAX score was calculated to estimate the complexity of the patient's CAD.8) The subjects were divided into 2 groups: those who had major adverse cardiovascular events (MACEs) and those who did not. The primary end-point was the incidence of MACEs during follow-ups, defined as the composite of all-cause mortality, nonfatal myocardial infarction (MI), repeat revascularization, or stent thrombosis. The diagnosis of MI was defined according to the American College of Cardiology criteria.9) Stent thrombosis was defined as definite or probable according to the Academic Research Consortium definitions.10) Subjects were followed up from the date of index PCI to the date of the first MACE, date of death, or December 31, 2015, whichever came first. The study conformed to the principles outlined in the Declaration of Helsinki. The research protocol was approved by the local Institutional Review Board.

Continuous variables were summarized as mean±standard deviation if normally distributed and as a median and interquartile range if non-normally distributed. Continuous variables were compared using Student's t-test. Categorical variables (presented as absolute values and percentages) were compared using the chi-square test (or Fisher's exact test when indicated). The data were collected at baseline and during follow-ups. Significant factors (p<0.05) were included in the Cox regression analysis, which was used to evaluate time-to-event associations with MACEs. Kaplan-Meier curves were plotted, and the log-rank test was used to compare differences between groups. For all analyses, a p<0.05 was considered to be statistically significant. Statistical analysis was performed using the Statistical Package for Social Sciences statistical software, version 19 for Windows (SPSS Inc., Chicago, IL, USA).

Among 107 study patients, 38 patients had experienced MACEs and 69 patients had no MACEs during follow-ups in the present study. The MACE rate was 36%, and the composite MACEs included 9 deaths, 28 repeated revascularizations, and 1 stent thrombosis. Baseline characteristics for patients of the study are shown in Table 1. The mean age was 62.3±11.2 years. No significant differences were found in age, sex, smoking, hypertension, diabetes mellitus, dyslipidemia, old MI, history of CABG, low density cholesterol (LDL-C), total cholesterol, and discharge medications including aspirin, clopidogrel, angiotensin-converting-enzyme inhibitor/angiotensin II receptor blockers, or beta-blockers between the event and no event groups. There were significant differences in left ventricular ejection fraction (LVEF), serum creatinine, and eGFR. Patients who experienced MACE had a lower LVEF (53.8±15.4% vs. 59.6±13.1%, p=0.049) and worse renal function (serum creatinine: 2.2±2.8 vs. 1.2±1.3 mg/dL, p=0.016; eGFR: 62.7±32.3 vs. 76.0±24.6 mL/min/1.73 m², p=0.019) than did patients without MACEs.

Table 2 shows the QCA and PCI data. We compared angiographic lesion characteristics, including bifurcation lesions, chronic total occlusions, ostial lesions, left main artery involvement, SYNTAX score, and number of vessels showing CAD between patients with and without MACEs. The procedural details including number, length, diameter, and type of stent (DES or BMS) were also analyzed. All of these angiographic and procedural characteristics did not differ significantly between patients with and without MACEs.

In a Cox regression analysis, LVEF and CKD stage ≥3 were significant predictors for MACEs. After an adjustment, only CKD stage ≥3 remained an independent predictor of MACEs (HR: 2.97, 95% CI: 1.50-5.90) (Table 3). Kaplan-Meier curves were plotted and showed a higher MACE rate in patients with CKD than without CKD (Fig. 2).