Randomized Comparison Between Polymer-Free Versus Polymer-Based Paclitaxel-Eluting Stent

Study Design and Population

This clinical trial was a prospective, randomized, single-blinded, multicenter, controlled trial comparing PF-PES with PB-PES in the treatment of coronary lesions (see Data Supplement for a list of the participating centers and the investigators). Eligible patients with a diagnosis of stable or unstable angina or non–ST-elevation myocardial infarction or silent ischemia were included if they were aged ≥18 years and volunteered for follow-up. Patients were eligible if they had atherosclerotic coronary artery disease, including native and de novo lesions with a diameter stenosis of minimally 50% of reference vessel diameters ranging from 2.25 to 4.0 mm by quantitative coronary angiogram (QCA) with objective evidence of ischemia. Patients with acute coronary syndrome <72 hours before admission or creatin kinase twice compared with the upper normal limit, previous brachytherapy or DES implantation in the target lesion, restenotic lesion, allergy to aspirin or thienopyridine, bypass graft lesions, true bifurcated lesions, severe renal insufficiency (creatinin clearance <30 mL/min), severe liver failure (glutamyl oxaloacetic transaminase and glutamyl pyruvic transaminase >3× the upper normal limit), or life expectancy <1 year because of other pathologies were excluded from the study. Patients were requested to receive clinical follow-up at 1, 9, 12, and 24 months and angiographic follow-up at 9 months. Incidence of in-stent restenosis and late lumen loss at 9 months were verified by QCA. Intravascular ultrasound (IVUS) was performed at postprocedure and angiographic follow-up to verify incidence of LISA and neointimal hyperplasia area (NIHA) at 9 months. Patients were assigned to undergo percutaneous coronary intervention with either PF-PES or PB-PES. Randomization was performed using a predefined computer-generated tabulated random order, which ensured near-equal numbers in each of the 2 arms before the attempt of stent implantation. Patients who met all the inclusion criteria and none of the exclusion criteria were randomized in order that they qualified. Time zero was defined as the time of randomization. Patients were considered enrolled in the study and eligible for the final intention-to-treat analysis at the time of randomization. The same randomly assigned stent had to be implanted in all lesions in those patients who required stenting in multiple lesions, and the use of >1 stent per lesion was also allowed. A total of 5 academic institutions participated in this trial. This was an investigator-initiated trial with no official sponsor involved in the study. All centers submitted and received the approval of their medical ethics committee for the protocol and for the informed consent. The study was performed in compliance with the protocol, the Declaration of Helsinki, British Standard European Norm International Standardization Organization 14155 Part 1 and Part 2, and applicable local requirements. All patients provided written informed consent.

Study Procedures

At least 2 hours before percutaneous coronary intervention, patient received a loading dose of 600 mg clopidogrel and 325 mg aspirin. Immediately after the decision to perform the intervention, patient was given 100 IU/kg intravenous unfractionated heparin. Device implantation was performed according to standard procedure techniques. If an additional stent was required to cover the lesion, additional stents had to be identical to the randomized treatment group. Procedure success was defined as a residual stenosis in stent <30% by QCA without stent edge dissection and Thrombolysis in Myocardial Infarction >2 coronary flow. Electrocardiograms and cardiac biomarkers were to be evaluated before and after the index procedure.

After the procedure, both aspirin 100 mg per day and clopidogrel 75 mg per day were prescribed for ≥6 months. Other medications included β-blockers, statins, angiotensin-converting enzyme inhibitors, and angiotensin receptor blockers according to clinical practice. Clinical information at 1, 9, 12, and 24 months was obtained by either a hospital visit or telephone contact with the patient or the referring physician. At 9 months, control angiography was mandated to assess the long-term angiographic results. If clinically warranted, revascularization was performed at the physician’s discretion. Relevant data were collected and entered into a computer database https://www.e-capture.net/LISA/.

Study Devices

The PF-PES (Axxion) is a balloon-expandable, 316 L stainless steel stent with 117 μm strut thickness. The whole surface of the stent is coated with synthetic glycocalix, which is a permanent biocompatible carbohydrate of glycoproteins and glycolipid. This surface modification mimics naturally occurring endothelial surface glycocalix and acts as a hemocompatible protective barrier similar to that found on native endothelium, resulting in less inflammation, less immunologic responses, and prevention of thrombosis.¹⁶ Paclitaxel, which inhibits smooth muscle cell proliferation and migration, secretion of extracellular matrix, and inflammation, is spray-coated onto the synthetic glycocalix substrate on the abluminal surface of stent only. The thickness of coating is ≈2 nm, and the dosage of drug is ≈2.7 μg/mm². This process assures that the high concentration of paclitaxel is delivered directly to the vessel intima. Approximately 40% to 50% of drug is released within the first week, and the remainder is released in the next 3 weeks.¹⁷

PB-PES (TAXUS Express) is a balloon-expandable, 316 L stainless steel stent with a permanent polymer. The strut thickness is 132 μm, and the thickness of coating is ≈16 μm. The coating is composed of 1 μg/mm² antiproliferative drug paclitaxel and a poly(styrene-b-isobutylene-b-styrene), which provides controlled biphasic release. The first release of paclitaxel (≈38% of the 10-day dose) occurs during the first 48 hours after implantation, and in the second phase, paclitaxel is slowly released during the next 10 days.¹⁸

Angiographic Analysis

QCA data were recorded into compact discs, and analysis was performed offline by an expert analyst in the independent core laboratory of the Hospital Clinic, blinded to the stent type, using automated edge-detection algorithms (CMS version 6.0; Medis Medical Imaging Systems, Nuenen, the Netherlands). Lesion complexity was analyzed according to the modified American Heart Association/American College of Cardiology grading classification.¹⁹ In each lesion, the coronary segment, including the stent and 5 mm proximal and distal to the stent edge, was analyzed at baseline and at 9-month follow-up. Late loss was estimated as the difference between the minimal lumen diameter at postimplantation and at follow-up using matched angiographic views. The pattern of in-stent restenosis was classified on the basis of Mehran classification.²⁰ To evaluate intraobserver variability, the observer repeated the analysis 3 months later.

IVUS Analysis

IVUS imaging was obtained at poststenting and at 9-month follow-up and performed after 100 to 200 μg of intracoronary nitroglycerin injection using a motorized transducer pullback system and a commercial scanner that consisted of a rotating 40-MHz transducer within a 2.5F imaging sheath (Galaxy2; Boston Scientific Corporation). The ultrasound catheter was advanced ≤10 mm beyond the stent into the distal vessel, and the transducer was withdrawn at an automatic pullback speed of 0.5 mm/s to a point ≤10 mm proximal to the stent. All IVUS data were recorded on compact discs and analyzed by an expert analyst blinded to stent type in the independent core laboratory of the Hospital Clinic. Cross-sections at 1-mm intervals within the stent segment were analyzed. The external elastic membrane, lumen, and stent contours were detected using offline software (QCU version 2.0; Medis Medical Imaging Systems). Vessel area and lumen area were automatically drawn, and stent area was drawn at the endoluminal site of the metallic struts. Minor modifications were made whenever necessary. NIHA was calculated as difference between stent area and lumen area at follow-up. NIHA was normalized per stent size as %NIHA and calculated as (NIHA/stent area)×100. The peri-stent area was derived by (vessel area–stent area) and normalized per stent size as percent peri-stent area and calculated as (peri-stent area/vessel area)×100. Incomplete stent apposition was defined as a separation of at least a single stent strut from the intimal surface of the arterial wall. Incomplete stent apposition was classified into the following 3 groups on the basis of serial assessment¹: resolved—incomplete stent apposition present after the procedure but no longer present at follow-up²; persistent—incomplete stent apposition present both after the procedure and follow-up³; and LISA not present at baseline but present at follow-up.^21–23 To evaluate intraobserver variability, the observer repeated the analysis 3 months later.

Study End Point

The primary end point of this study was late loss at 9 months by QCA. Regarding clinical end points, the efficacy end point was any target lesion revascularization at 2 years on patient basis, whereas the safety end point was a composite of all-cause death, any myocardial infarction, and target vessel revascularization (TVR) at 2 years. Target lesion revascularization was defined as either percutaneous coronary intervention or coronary artery bypass grafting owing to restenosis or other complication of the target lesion. All-cause death included noncardiac and cardiac death. Cardiac death was defined as any death because of cardiac cause (myocardial infarction, low-output failure, and fatal arrhythmia), death related to the procedure, or death of unknown cause. TVR was defined as any repeat percutaneous intervention or surgical bypass of any segment of the target vessel. The target vessel is defined as the entire major coronary vessel proximal and distal to the target lesion. Stent thrombosis was assessed according to Academic Research Consortium criteria.²⁴ All adverse events were adjudicated by an event committee blinded to treatment allocation.

Statistical Analysis

Categorical variables were presented as numbers and percentages and compared between groups with χ² or Fisher exact tests, when appropriate. Continuous data were shown as mean±SD and compared using the Mann–Whitney test. Comparison of data at lesion level has been performed accounting for clustering within patient.²⁵ Intraobserver reproducibility of QCA and IVUS was calculated by intraclass correlation coefficient for repeated measurement. Survival and event-free status were assessed using the methods of Kaplan–Meier. A sample size of 105 patients per group was calculated with a statistical power of 90% and a 2-sided type 1 error of 0.05 to detect a difference in late loss of 0.19 mm (0.39 mm for PB-PES in TAXUS IV²⁶ and 0.20 mm for PF-PES; mean value between 0.29 mm in Asian Paclitaxel-Eluting Stent Clinical Trial⁹ [ASPECT] and 0.11 mm in European Evaluation of Taxol Eluting Stent¹⁰ [ELUTES]). Expecting that ≈25% to 30% of the patients would have been lost at follow-up, a total of 270 patients were planned to be enrolled. A 2-sided P value <0.05 was considered statistically significant. All statistical analyses were performed using SPSS (version 11.0.1; SPSS Inc, Chicago, IL).