The study protocol was reviewed and approved by the Institutional Review Board (IRB) of Seoul National University Bundang Hospital (IRB No. E-1803/456-002) and the other recruiting institutions. This study was registered in the clinical trial list of Ministry at the Food and Drug Safety of Korea for medical device approval (registered No. 873). Subjects in this study voluntarily signed the consent form.

This clinical trial was a prospective, multicenter, parallel, single-blind, randomized clinical trial. We compared Osstem Cardiotec Centum versus Abbott's “Xience Alpine®” based on data collected for up to 270 days after PCI in 6 medical institutions of South Korea. The subjects were randomly assigned to one of the stents at a ratio of 1:1. Subjects visited for follow-up at 30, 180, and 270 days after the procedure, respectively. They were evaluated for efficacy and safety according to the schedule at each visit.

Subjects were enrolled if they met the inclusion/exclusion criteria during screening. At the screening visit, the baseline data were investigated and evaluated. Inclusion criteria included adults aged over 19 years who were suitable for PCI due to myocardial ischemia such as stable angina, unstable angina, and asymptomatic ischemia. The criteria also started that patients should have stenosis in the de novo lesion located in the native coronary artery (>50% diameter stenosis [DS] on coronary angiography and thrombolysis in myocardial infarction flow grade 1 or more, a diameter of 2.25 mm or more and 4.00 mm or less, and a lesion length of 34 mm or less during coronary angiography).

Patients were excluded if they had presented with acute myocardial infarction within a year from the enrolment, with experiencing cardiogenic shock, reduced left ventricular ejection fraction of less than 30%, restenosis lesions, lesion located on the saphenous vein graft, a history of bleeding tendency or blood clotting disorder, the expected life expectancy of 1 year or less due to accompanying disease, allergy or a contraindication to everolimus, aspirin, clopidogrel cobalt chromium, PLA/PLGA or contrast media, etc., total occlusion, a lesion of a blood vessel that had undergone other PCI procedures (balloon angiography, stent, cutting balloon atherectomy, etc.) or brachytherapy within 9 months (registration is possible if the procedure was performed on a vessel other than the target vessel.). It also excluded patients having a lesion with bifurcation or ostial lesion of >2 mm in diameter and occluded by 50% or more, a case in which a procedure/operation was scheduled to stop double antiplatelet therapy (DAPT) within 12 months after the procedure, or with a platelet count of <100,000 cells/mm³ or> 700,000 cells/mm³, a leukocyte count of <3,000 cells/mm³, a serum creatinine level ≥2.5 mg/dL, or on dialysis. Patients who had the lesion present in the left main coronary artery, were pregnant, or lactating mothers, planned to become pregnant within 12 months of the procedure, or disagreed with the use of appropriate contraception during the clinical trial were also excluded. Lastly, patients were excluded in the case of clinically significant abnormal values of hematological results and urinalysis, which are laboratory test items performed at the time of screening. or if they had reduced immunity.

The registered subjects were hospitalized on the day before or on the day of the scheduled procedure according to the institution's standard procedure and received pre-procedure treatment, including an aspirin and a P2Y12 inhibitor (clopidogrel, ticagrelor, or prasugrel) as DAPT. The subjects were discharged from the hospital according to the institution's standard procedure after PCI. The subjects were then hospitalized for 270 days after PCI for coronary angiography and follow-up data.

According to the judgment of the investigator and the method of use suggested by the manufacturer, a stent of the optimal size was selected in considering of the reference vessel diameter (RVD) and the length of the lesion. After pre-dilatation, the stent was delivered to the lesion using a balloon catheter system. After deploying the stent in the target lesion, the stent was expanded by applying a nominal pressure. For post-dilatation, a noncompliant balloon of the same size as the stent was used.

Quantitative angiographic analysis was performed by operators not involved in the intervention using CAAS® software (Pie Medical Imaging B.V, Maastricht, the Netherlands). A guiding catheter was used for magnification calibration as a standard index. In-segment/in-stent %DS, in-segment/in-stent minimal lumen diameter (MLD), and in-segment/in-stent reference lumen diameter were obtained from this system. These measurements were taken using the same end-diastolic angiographic frames acquired pre- and post-stent implantation. Acute gain was computed as the difference between the baseline MLD and the final post-stenting MLD of the procedure day. Late lumen loss (LLL) was computed as the difference between the final post-stenting MLD and the MLD found at follow-up angiography.

The primary efficacy endpoint was in-segment LLL on follow-up coronary angiography at 270 days after PCI by an independent evaluator. In-segment was defined as a section including a segment corresponding to a location 5.0 mm away from both ends of the stent. LLL was calculated by subtracting the MLD at 270 days after PCI from the MLD immediately after stenting.

The secondary endpoints were target lesion failure (TLF) at 270 days after PCI and clinical success rate. TLF was defined as a composite of revascularization of the target lesion, target vessel myocardial infarction, and cardiac death. Clinical device success was defined as the successful application of the stent to the target lesion, the delivery system successfully removed, and residual stenosis found to be less than 30% on angiography. Clinical procedure success was defined as the case in which clinical device success was achieved and TLF did not occur for 24 hours after PCI.

This trial was conducted to prove the non-inferiority of the Osstem Cardiotec Centum to the Xience-Alpine stent with respect to the primary endpoint at 270 days. We assumed a SD of in-segment LLL (0.35 mm in each stent group derived from 5 recent clinical trials).7)8)9)10) To verify the non-inferiority of Osstem Cardiotec Centum compared to Xience stents in this clinical trial, we set 0.195 mm as the non-inferiority margin of LLL at 9 months.7) We calculated the appropriate sample size with a 1-sided significance level of 0.025 and 80% power. The sample size was 61 in each treatment arm in consideration of a 16% drop out rate. Sample size calculation in detail is represented in Supplementary Data 1.

The primary endpoints and secondary endpoints were evaluated by intention-to-treat analysis. Additionally, in order to analyze safety issues that are generally reflected by actual biological and biochemical responses, the secondary endpoints for safety evaluation were analyzed using the as-treated set and these were described at Supplementary Figure 1 and Supplementary Table 3. Demographic data, basic characteristics, and safety data were obtained using the safety set.

Statistical significance was established at a 2-sided test significance level (α) of 5%. However, the non-inferiority test was performed using the 97.5% 1-sided confidence interval (CI; the upper limit of the 95% 2-sided CI). For normally distributed continuous data, the 2 sample t-test is used, whereas the Wilcoxon rank sum test was used for non-normally distributed continuous data. A chi-squared test was used for categorical, and when cells with an expected frequency of less than 5 were 20% or more of all cells, the significance was examined using Fisher's exact test. Descriptive statistics for LLL were presented in the Osstem Cardiotec Centum and Xience Alpine groups to analyze differences in LLL as the primary endpoint. Since a 95% 2-sided CI for the mean difference between the 2 groups (LLL mean in the test group − LLL mean in the Xience Alpine group) did not have normality on normality test, we had obtained the 95% CI of the median difference between the LLL values of the 2 groups based on the Hodges-Lehmann estimation as a nonparametric test. From this, if the upper limit of the 95% bilateral CI was smaller than the preset non-inferiority criterion (0.195),7) the test device (the Osstem Cardiotec Centum) was judged to be non-inferior to the control device (Xience Alpine).

Among patients who had visited outpatient clinics in 6 institutions from September 21, 2018 to July 3, 2020, those who required PCI due to de novo lesions were screened. Out of a total of 125 subjects, we randomly assigned 59 subjects to the Osstem Cardiotec Centum group and 62 subjects to the Xience Alpine group after excluding 4 screening dropouts (Figure 1). All randomized subjects were treated with the allocated stents except one subject who was randomly assigned to the Osstem Cardiotec Centum group; however, an error occurred, and hence, the Xience Alpine was applied. Therefore, 58 subjects were subjected to Osstem Cardiotec Centum while 63 subjects were subjected to the Xience Alpine stent.

The number of subjects who failed to complete the follow-up until 270 days after the procedure and dropped out was 4 in the Osstem Cardiotec group and 4 in the Xience group, hence totaling to 8 subjects. Therefore, data from 55 subjects in the Osstem Cardiotec group and 58 subjects in the Xience group were analyzed with the intention-to-treat primary endpoint. All subjects completed the clinical follow-up for 1 year. Data from a total of 121 subjects (58 in the Osstem Cardiotec group and 63 in the Xience group as treated) were analyzed for safety endpoints.

The demographic and clinical characteristics of the 121 subjects were well balanced between the 2 groups (Table 1). The mean age was 67.03 in the Osstem Cardiotec group and 66.68±8.39 in the Xience group. The risk factors for coronary artery disease such as current smoking, diabetes mellitus, hypertension, and dyslipidemia were not different. The most clinical diagnoses for PCI were stable angina (59% in the Osstem Cardiotec group and 67% in the Xience group) and unstable angina (45% and 40%, respectively). DAPT was prescribed to all subjects in both groups.

Lesion and procedure characteristics of 113 subjects who underwent follow-up angiography are shown in Table 2. The target lesion location and lesion complexity were not significantly different between the groups. The stent diameter and length were comparable between the groups without statistical significance.

Pre-procedural MLD and RVD were not significantly different between the groups and thus DS was comparable (Table 2). Post-procedural MLD, DS and acute gain were not different between the groups. In the 9 months angiographic data, MLD and DS were similar between the groups (Table 3). However, LLL of the Osstem Cardiotec group was 0.09±0.13 mm and that of the Xience group was 0.12±0.14 mm (p=0.034). Figure 2 presents the cumulative frequency of in-segment of LLL. The average difference between the 2 groups was −0.03 mm (95% CI: −0.05, −0.01), and the upper limit of the CI was −0.01, which was found to be smaller than the non-inferiority margin of 0.195, which demonstrated the non-inferiority of the Osstem Cardiotec Centum stent.

The success rate of the device or procedure was 100% in both groups (Table 4). As for the TLF, one ischemic target lesion revascularization occurred in the Xience group. In the Osstem Cardiotec group, one patient died due to exacerbation of underlying COPD and pneumoconiosis.