Our prospective, non-randomised study included consecutive adult patients (>18 years of age) with BMS-ISR (≥50% diameter stenosis; DS) who were treated with seal-wing PEB (Protége) in the Cathlab of University Hospital Ostrava in 2013–2015. The patients were followed 12 month after intervention and the study was finished in December 2016. The control group comprised patients with BMS-ISR who were treated using iopromide-coated PEB (Sequent Please) and EES (Promus; Pt/Cr metallic platform) in the previous randomized part of the TIS study [3]. The main exclusion criteria were concomitant diseases carrying expected survival times of <12 months or limiting the possibility of control coronary angiography (e.g., advanced renal failure), or inability to undergo long-term dual antiplatelet treatment (e.g., due to aspirin or clopidogrel allergy, bleeding complications, etc.).

The primary end-point was in-segment LLL at 12 months as measured by quantitative control angiography (QCA) [5]. Secondary end-points were the incidence of binary ISR (≥50% DS) and the overall incidence of 12-month major adverse cardiac events (MACE), including cardiovascular death, non-fatal acute myocardial infarction (MI), and target vessel revascularization (TVR).

PCI was performed under standard conditions from the radial or femoral approach, using a 6F guiding catheter and an Axiom X-ray system (Siemens AG, Forchheim, Germany).The patients were pre-treated with aspirin and clopidogrel (600 mg loading dose). Full anticoagulation was achieved by administering 100 IU/kg non-fractionated heparin, with a target activated clotting time of 250–300 s.

To pre-dilate the lesions and prevent any edge dissection, we used relatively shorter or scoring balloon catheters. Following predilatation, the PEB Protége (Blue Medical, Helmond, the Netherlands) was inflated for 30s at the recommended pressure. The seal-wing PEB Protége has paclitaxel (3 μg/mm²) tightly bound directly to the balloon catheter surface between the wings and hydrophilic coating prior to folding. This coating prevents releasing particles during bending of the balloon or transition to the stenosis. Paclitaxel, not coating, is only released when the balloon touches the vessel wall [6, 7].

Post-dilatation was performed using a non-compliant balloon catheter in the case of a suboptimal outcome, and another bailout stent was implanted in the case of edge dissection. All patients received standard therapy after coronary intervention. Dual antiplatelet therapy with aspirin (100 mg) and clopidogrel (75 mg) was administered daily for three months following PEB dilatation.

Clinical follow-up was performed at 6 and 12 months. Angiographic follow-up was performed at 12 months (±2 months) unless needed earlier. All deaths were considered cardiac-related unless clearly from non-cardiac causes. Acute myocardial infarction was defined according to the third universal definition of myocardial infarction of the ESC [8], and stent thrombosis was defined based on ARC criteria [9].

Prior to imaging, patients were administered intracoronary isosorbide dinitrate (1 mg). Imaging was performed using appropriate orthogonal projections to best avoid potential shortening or overlap of the reporting segment and lateral branches. Similar projections were used at the 12-month coronary angiography. Lesion type and ISR were evaluated in an in-segment section ±5 mm from the proximal and distal edges of the stent using ACC/AHA criteria [10] and Mehran’s classification [11]. Angiographic parameters were evaluated off-line using syngo Quantification software version 2007 (Siemens AG, Forchheim, Germany). The following parameters were measured: minimum lumen diameter (MLD), reference lumen diameter (RefD = ½ proximal + distal diameter), acute gain, lesion length, diameter of the stenosis (%DS), and late lumen loss (LLL = MLD post-intervention − MLD control). Binary ISR was defined as DS ≥50% in the stented segment.

The sample size calculation was based on the previous TIS study. The 12-month difference in LLL 0.24 (±0.27) mm with alpha type I error of 5%, and beta test strength of 80% was used to determine the required group size of 64 per arm.

Normally distributed continuous variables are presented as mean and standard deviation, and were compared using Student’s two-sample t-test. Continuous variables with non-normal distribution are presented as median and range, and were compared using the non-parametric Mann–Whitney U test. Categorical variables are presented as count and percentage, and were compared using the chi-square or Fisher’s exact test. Bonferroni correction was used for multiple comparisons. Kaplan-Meier curves are used to display time-to-event data, which were compared using the log-rank test. Multiple logistic regression (stepwise forward method) was used to identify the most significant predictive factors for repeated binary restenosis, with adjustment for diabetes mellitus and other possible confounding factors. Each odds ratio (OR) is expressed with a 95% confidence interval (CI). A p value of <0.05 was considered significant. Evaluation was based on intention-to-treat. All statistical analyses were performed using IBM SPSS Statistics version 22.

Our study has several limitations. In particularly, it was a non-randomized study that compared active treatment groups with control arms from the previous TIS study. Nevertheless, selection bias likely did not play a major role, since the patient cohorts (particularly the seal-wing PEB and iopromide-coated PEB groups), did not differ with respect to main baseline parameters.

Diferences in clinical outcomes between the study arms should be taken with caution, as our study was not powered for clinical endpoints. The results of the logistic regression analysis in different subgroups should also be interpreted with caution.