Long-term effect of biodegradable vs. durable polymer everolimus-eluting stents on neoatherosclerosis in ST-segment elevation myocardial infarction: the CONNECT trial

Abstract

Background and Aims

Neoatherosclerosis is a leading cause of late (>1 year) stent failure following drug-eluting stent implantation. The role of biodegradable (BP) vs. durable polymer (DP) drug-eluting stents on long-term occurrence of neoatherosclerosis remains unclear. Superiority of biodegradable against durable polymer current generation thin-strut everolimus-eluting stent (EES) was tested by assessing the frequency of neoatherosclerosis 3 years after primary percutaneous coronary intervention (pPCI) among patients with ST-segment elevation myocardial infarction (STEMI).

Methods

The randomized controlled, multicentre (Japan and Switzerland) CONNECT trial (NCT03440801) randomly (1:1) assigned 239 STEMI patients to pPCI with BP-EES or DP-EES. The primary endpoint was the frequency of neoatherosclerosis assessed by optical coherence tomography (OCT) at 3 years. Neoatherosclerosis was defined as fibroatheroma or fibrocalcific plaque or macrophage accumulation within the neointima.

Results

Among 239 STEMI patients randomized, 236 received pPCI with stent implantation (119 BP-EES; 117 DP-EES). A total of 178 patients (75%; 88 in the BP-EES group and 90 in the DP-EES group) underwent OCT assessment at 3 years. Neoatherosclerosis did not differ between the BP-EES (11.4%) and DP-EES (13.3%; odds ratio 0.83, 95% confidence interval 0.33–2.04, P = .69). There were no differences in the frequency of fibroatheroma (BP-EES 9.1% vs. DP-EES 11.1%, P = .66) or macrophage accumulation (BP-EES 4.5% vs. DP-EES 3.3%, P = .68), and no fibrocalcific neoatherosclerosis was observed. Rates of target lesion failure did not differ between groups (BP-EES 5.9% vs. DP-EES 6.0%, P = .97).

Conclusions

The use of BP-EES for primary PCI in patients presenting with STEMI was not superior to DP-EES regarding frequency of neoatherosclerosis at 3 years.

Structured Graphical Abstract

Design and results of the CONNECT trial. Patients were randomized to 1:1 to biodegradable vs. durable polymer EES. The primary endpoint (frequency of neoatherosclerosis) was assessed by OCT 3 years after primary PCI with biodegradable or durable polymer EES. The frequency of neoatherosclerosis [defined as presence of fibroatheroma (arrow heads) or macrophages (bold arrows) or fibrocalcific plaques (lined arrows) within the neointima of a stented segment with a longitudinal extension of >1 mm] for each group as well as the frequency of each component is shown on the right. EES, everolimus-eluting stent; FUP, follow-up; OCT, optical coherence tomography; OR, odds ratio; PCI, percutaneous coronary intervention; STEMI, ST-segment elevation myocardial infarction.

Open in new tabDownload slide

Introduction

Neoatherosclerosis is a leading cause of late (>1 year) stent failure including in-stent restenosis (ISR)^1,2 and very late stent thrombosis (ST) in drug-eluting stents (DES),^3,4 and develops more frequently and earlier after implantation in DES as compared with bare metal stents (BMS).^5–7 Neoatherosclerosis is characterized by migration of macrophages and lymphocytes and entrance of oxidized low-density lipoprotein (LDL)-cholesterol into the neotintimal layer, triggered by stent implantation-related endothelial denudation and delayed endothelial recovery. The latter is more pronounced in DES due to the antiproliferative effect of the eluted drug and in the presence of persistent inflammation potentially related to the stent polymer.^8,9 Lesions in patients with acute myocardial infarction are frequently lipid rich and reportedly carry the highest risk for impaired healing and accelerated neoatherosclerosis formation.¹⁰

Long-term remnants of durable polymer (DP) of DES remaining in the vessel wall after drug elution are assumed to trigger local inflammation,⁹ which may contribute to accelerated neoatherosclerosis. Stents with biodegradable polymer (BP) that fully degrades within months were developed to reduce the risk of chronic vascular inflammation. However, the hydrolysis-process also stimulates biologic activity, potentially leading to a cellular reaction in the vessel wall.

Optical coherence tomography (OCT) is a high axial resolution (10–20 µm) technique allowing in vivo characterization of neointimal tissue and detection of neoatherosclerosis, specifically fibroatheroma, fibrocalcific plaque, or macrophage accumulation.¹¹ Long-term data investigating the impact of the polymer are largely missing to date.

To address this gap in evidence, this randomized controlled trial aimed to compare the frequency of neoatherosclerosis between a DP thin-strut everolimus-eluting stent (EES) and a BP thin-strut EES as assessed by intracoronary OCT 3 years after primary percutaneous coronary intervention (pPCI) in ST-segment elevation myocardial infarction (STEMI) patients, a patient population with a presumed higher risk for neoatherosclerosis.

Methods

Study design and population

CONNECT (a randomized COmparison of loNg-term vascular healiNg between biodegradable polymer vs. durable polymer everolimus-eluting stents in acute ST-elevation myocardial infarCTion) is a prospective, multicentre, open-label, assessor-blind, randomized, controlled trial testing superiority of a BP-EES over a DP-EES, both thin-strut stents, in patients with STEMI undergoing pPCI. The trial was conducted at seven centres in two countries (Japan and Switzerland).

Patients 18 years or older were eligible for the study, if they presented with STEMI for pPCI within 24 h of symptom onset with a culprit lesion that qualified for DES implantation. Major exclusion criteria were haemodynamic instability, left ventricular ejection fraction ≤ 20% or mechanical complications of acute myocardial infarction, myocardial infarction secondary to ST or restenosis, known chronic kidney disease with estimated glomerular filtration rate < 30 mL/min, life expectancy < 3 years, and culprit lesions not suitable for OCT imaging. The full list of inclusion and exclusion criteria is provided in the Supplementary material.

All subjects provided informed consent prior to randomization. In view of the presence of STEMI, a specific informed consent process was applied, allowing for preliminary oral consent upon oral informed consent information given by the operating physician, followed by written consent after pPCI. The process is detailed in the Supplementary material. The study protocol was approved by the responsible ethics committees. The trial was conducted in accordance with Good Clinical Practice guidelines and is reported following the Consolidated Standards of reporting Trials (CONSORT) 2010 guideline statement for reporting parallel-group randomized trials statement.¹²

Randomization and study devices

After successful recanalization of the acute infarct artery target lesion with the guidewire, patients fulfilling all eligibility criteria were randomly allocated in a 1:1 ratio to either receiving BP-EES or DP-EES. Web-based randomization was performed based on computer-generated random numbers and stratified according to trial site. Participants and operators were not blinded to the allocation.

The BP-EES (Synergy, Boston Scientific, Marlborough, MA, USA) is a platinum chromium back-boned thin-strut EES (74–81 μm) with a biodegradable poly-(DL-lactide-co-glycolide acid) polymer coating (4 μm) applied to the abluminal surface with <4 month absorption time. The comparator DP-EES (Xience Alpine/Sierra, Abbott Vascular, Santa Clara, CA, USA) is a thin-strut (81 μm) EES based on a cobalt chromium platform, which is circumferentially coated with an inactive permanent biocompatible polymer/co-polymer (poly-n-butyl methacrylate and vinylidene fluoride and hexafluoropropylene monomers). Detailed characteristics of the study stents are summarized in Supplementary data online, Figure S1.

Study procedures

Randomized patients underwent pPCI as per local standard of care. The culprit lesion was treated with one or more stents of the allocated stent type. Clinically relevant non-culprit lesions were allowed to be treated during index procedure or as a staged PCI (within 3 months) at the discretion of the operating physician (use of the same allocated stent type was recommended) and were not part of the primary endpoint assessment. Post-procedural use of OCT was recommended to all operators but the protocol did not foresee any specific guidance criteria, i.e. the use of any corrective measures was left at the discretion of the operator, and were not recorded. Periprocedural and subsequent antithrombotic therapy was according to standard practice with country-specific differences choice and dosing of the P2Y₁₂ inhibitor (see Supplementary material). Treatment with high-intensity statin was recommended (Japan: rosuvastatin ≥ 10 mg daily, atorvastatin ≥ 20 mg daily or pitavastatin ≥ 4 mg daily; Switzerland rosuvastatin ≥ 20 mg daily or atorvastatin ≥ 40 mg daily) to achieve an LDL-cholesterol level < 70 mg/dL (1.8 mmol/L) recommended by guidelines at the time of the enrolment period.¹³

After hospital discharge, patients underwent telephone visits at 1 and 2 years, and repeat angiography with OCT assessment of the stents in the culprit lesion after 3 years.

Study organization

This was an investigator-initiated study supported by an unrestricted research grant from Boston Scientific, Marlborough, MA, USA and intramural grants of Bern University Hospital, Bern, Switzerland. The funding source was not involved in the design of the study, data collection, and data management, and had no role in the analysis and interpretation of the study data and was not involved in the decision to publish this report. This trial was conducted in accordance with the Helsinki Declaration. The study was registered at www.clinicaltrials.gov (Identifier: NCT03440801).

Patient data were collected in a web-based data entry system hosted at Research Anchor Japan, KIC Co. Ltd, Tokyo, Japan. On-site monitoring was performed independently by the Site Support Institute Co., Ltd, Tokyo, Japan at Japanese sites and by the CTU Bern, Bern, Switzerland at the Swiss site. Optical coherence tomography and quantitative coronary angiography (QCA) were analysed by core laboratories (OCT: Bern University Hospital, Bern, Switzerland; QCA: Cardiocore Japan, KIC Co. Ltd, Division Corelab, Tokyo, Japan) by experienced analysts blinded to stent type allocation and clinical and procedural characteristics of the patients. Statistical analyses were independently performed at CTU Bern, Bern, Switzerland.

Acquisition and analysis of intracoronary optical coherence tomography imaging at 3 years

Details of OCT acquisition and offline analysis are described in the Supplementary material. Briefly, follow-up OCT 3 years after pPCI was performed using a frequency-domain OCT system (OPTIS ILUMIEN or ULTREON; Abbott Vascular, Santa Clara, CA, USA). Core laboratory analysis of offline data was performed blinded to treatment allocation using proprietary software (QCU-CMS version 4.69, LKEB, Leiden, The Netherlands) at the OCT Corelab of Bern University Hospital. Cross-sections at 0.4 mm longitudinal intervals within the stented segment were analysed.

Outcomes

The primary endpoint was the frequency of neoatherosclerosis, which was defined as the presence of a fibroatheroma or fibrocalcific plaques or macrophage accumulation within the neointima of a stented segment with a longitudinal extension of >1 mm at 3 years (i.e. at least three consecutive analysed frames with evidence of lipid, calcium, or macrophage accumulation).^11,14,15

Fibroatheroma was characterized as a signal-poor region displaying high attenuation (to differentiate from layered neointima) with diffuse borders and a lateral extension of at least 45°. Calcification was defined as low signal intensity areas with sharply delineated borders with a minimal lateral extension of 45°. Macrophage accumulation was characterized as bright spots or lines with increased signal intensity within the plaque accompanied by heterogeneous backward shadow.^14,16 In case of fibroatheroma, minimal cap thickness, longitudinal length, and maximal lipid angle extension were determined. Maximal angle of macrophage accumulation was assessed.

Additional quantitative measures such as neointimal thickness, lumen area, stent area, neointimal area and qualitative measures such as stent strut coverage, stent strut malapposition, and neointimal healing score were assessed and definitions shown in the Supplementary material. Clinical outcomes included death, myocardial infarction, repeat revascularization, and stroke, and target lesion failure (composite of cardiac death, target-vessel myocardial infarction, and clinically indicated target lesion revascularization) as a device-oriented endpoint at 3 years. Definitions of clinical endpoints are provided in the Supplementary material. All events were assessed by an event adjudication committee.

Sample size calculation

On the basis of the results from the CVPath DES Registry⁸ and the SIRTAX-LATE OCT study,¹⁷ we assumed a frequency of neoatherosclerotic lesions including fibroatheroma or calcified plaque or macrophage accumulation of 35% in the DP-EES at 3 years. The frequency in the BP-EES was anticipated to be half of the comparator’s. The rationale for these assumptions is provided in the Supplementary material. With a 1:1 allocation ratio and a two-sided alpha of .05 with an attrition rate of 20% at 3 years, enrolment of a total of 240 patients (120 patients per treatment group) would provide 80% power to detect superiority of BP-EES over DP-EES.

Statistical analysis

All randomized patients undergoing stent implantation were included in the analyses according to the intention-to-treat principle. All analyses were performed at patient level. Values are presented as mean ± standard deviation for continuous variables or count (%) for categorical variables. Procedural characteristics were statistically compared between arms using Student’s t-tests or Fisher’s exact tests. Optical coherence tomography variables, including the primary endpoint (frequency of neoatherosclerosis in the culprit vessel), could only be assessed in patients in which follow-up OCT results were available. These were compared between arms using logistic regressions for categorical variables or linear models for continuous variables. We report odds ratios (OR) or between-arm differences and their associated 95% confidence intervals (CIs). We used logistic models to compare clinical outcomes between arms. Pre-specified stratified analyses of the primary endpoint were performed between Japanese and Swiss patients, and according to diabetes status. All analyses were two-sided and the significance threshold was set to 5%. All analyses were performed in R version 4.3.2.

Results

Patient characteristics

Between June 2017 and June 2020, a total of 239 patients with STEMI were randomized after preliminary consent from the patients. Clinical characteristics are shown in Table 1. Three patients did not undergo any stent implantation after randomization and were withdrawn by the operator from the study. The patient flow is shown in Figure 1 and enrolment per site in Supplementary data online, Table S1. All 236 undergoing pPCI received exclusively stents of the allocated group (119 in the BP-EES group and 117 in the DP-EES group). Angiographic and procedural characteristics are presented in Table 2. Lesion preparation strategies were not significantly different between the stent groups (BP-EES vs. DP-EES: thrombus aspiration in 25.2% vs. 38.5%; predilation in 65.5% vs. 56.4%; primary stenting in 9.2% vs. 5.1%, P = .06), as were the number of stents implanted in the culprit lesion (BP-EES 1.27 ± 0.53 vs. DP-EES 1.29 ± 0.53 mm, P = .75), total stent length (BP-EES 32.2 ± 16.2 vs. DP-EES 32.0 ± 16.6 mm, P = .95), and mean stent diameter (BP-EES 3.13 ± 00.48 vs. DP-EES 3.12 ± 0.50 mm, P = .87). Post-procedural intracoronary imaging was used in 98.3% of patients in both groups (P = 1.00). Post-procedural TIMI 3 flow post-PCI was achieved in 99.2% (BP-EES) and 98.3% (DP-EES, P = 1.00).

A total of 178 patients (75.4%, 88 in the BP-EES group and 90 in the DP-EES group) underwent 3-year follow-up angiography with successful OCT imaging. Patients who underwent follow-up OCT were younger (63.4 ± 10.9 years at enrolment) than those without follow-up OCT (66.6 ± 11.0 years, P = .05) and less frequently had diabetes mellitus (12.4% vs. 23.0%, P = .04), and history of stroke (2.8% vs. 11.5%, P = .01, Supplementary data online, Tables S2 and S3). Medication at discharge and at 3-year follow-up was comparable between stent groups with no difference in the dual antiplatelet therapy (DAPT) treatment regimen (see Supplementary data online, Table S4). Of note, 89% of patients were discharged on DAPT with a potent P2Y₁₂ inhibitor with similar use of ticagrelor (BP-EES 42.9% vs. DP-EES 41.9%, P = .90) and prasugrel in both groups (BP-EES 45.4% vs. 47.9%, P = .79). Among patients undergoing follow-up OCT after 3 years, 80.7% in the BP-EES group and 78.9% in the DP-EES group (P = .65) were on high-intensity statin therapy. PCSK9 inhibitors were prescribed in two patients. LDL-cholesterol at 3 years amounted to 68 ± 29 mg/dL in the BP-EES group and 63 ± 36 mg/dL in the DP-EES group (P = .23).

Primary outcome

BP-EES was not superior to DP-EES as the prevalence of neoatherosclerosis at 3 years did not differ between the BP-EES (11.4%) and DP-EES group (13.3%, OR 0.83 [95% CI 0.33–2.04], P = .69), with no significant difference in the presence of fibroatheroma (BP-EES 9.1% vs. DP-EES 11.1%; OR 0.80 [95% CI 0.29–2.13], P = .66), and macrophage accumulation (BP-EES 4.5% vs. DP-EES 3.3%; OR 1.38 [95% CI 0.30–7.28], P = .68), and with no cases of fibrocalcific neoatherosclerosis (Table 3 and Structured Graphical Abstract). Pre-specified subgroup analysis showed no significant interaction with respect to the country of enrolment (Japan vs. Switzerland) and diabetes status (see Supplementary data online, Table S5). A post hoc analysis was performed to assess the impact of LDL-C lowering therapy regimen revealed lower frequency of neoatherosclerosis in patients treated with high-intensity statin therapy (8.5%) as compared to those without (27.8%, P = .004, Supplementary data online, Table S6).

Secondary optical coherence tomography and angiographic outcomes at three years

The percentage of uncovered (BP-EES 3.7 ± 5.1 vs. DP-EES 4.8 ± 7.3%, P = .25), malapposed struts (BP-EES 1.2 ± 2.6 vs. DP-EES 1.0 ± 2.7%, P = .69) and neointimal healing score (BP-EES 0.089 [interquartile range 0.343] vs. DP-EES 0.116 [0.297], P = .50) did not differ between stent types (Table 3). BP-EES tended to show a higher neointimal thickness (BP-EES 140.3 ± 74.8 vs. DP-EES 121.8 ± 64.2 µm, P = .08) and had a significantly larger neointimal area (BP-EES 1.04 ± 0.70 vs. DP-EES 0.83 ± 0.59 mm², P = .03) at follow-up. However, this did not translate into a smaller angiographic in-stent late lumen loss (BP-EES 0.22 ± 0.39 vs. DP-EES 0.22 ± 0.40, P = .95, Supplementary data online, Table S7).

Clinical outcomes

Clinical 3-year outcomes were available from 88.9% and are presented in Supplementary data online, Table S8. Rates of target lesion failure were identical between treatment groups (BP-EES 5.9% vs. DP-EES 6.0%, P = .97), rates of cardiac death and target-vessel myocardial infarction were very low and without significant difference between groups, and no definite or probable ST occurred. Clinically indicated target lesion revascularization occurred in 4.2% of 119 BP-EES and 5.1% of 117 DP-EES-treated patients. The majority of target lesion revascularizations occurred during the study-related repeat angiography (12/17). In patients undergoing TLR at 3 years, neoatherosclerosis was identified as underlying cause in 41.7% (see Supplementary data online, Table S9).

Discussion

In this trial, long-term occurrence of neoatherosclerosis assessed by OCT 3 years after pPCI among patients with STEMI was not significantly different between a thin-strut EES with BP and a thin-strut EES with DP. The findings of the primary endpoint are corroborated by similar patterns of neoatherosclerosis (fibroatheroma and macrophage accumulation) and the magnitude of neoatherosclerosis formation including length and lateral extension in both stent types. Fibroatheroma was the most frequent form of neoatherosclerosis, while neo-calcifications were not observed.

This study is by far the largest and providing the longest intracoronary imaging follow-up ever conducted to assess the impact of current generation BP-DES vs. DP-DES on the development of neoatherosclerosis confirming no difference between the two stent platforms with the main difference being the type of polymer.

Despite lower rates of ST and ISR compared to balloon angioplasty, BMS, or early generation DES, stent failure still occurs following the implantation of newer generation DES,^18,19 reportedly requiring target lesion revascularization in 20%–30% throughout 10 years and thus limiting the efficacy of PCI.²⁰ Neoatherosclerosis was consistently identified as one of the major morphological substrates for stent failures occurring beyond the first year of stent implantation.^1,3,4 Higher rates and more rapid development of neoatherosclerosis have reportedly been shown in DES-ISR when compared with BMS-ISR.^5–7 Durable polymer remaining in the vessel wall after drug elution may trigger persistent local inflammation,⁹ potentially contributing to accelerated neoatherosclerosis. BP-DES, in which hydrolysis of the polymer results in a ‘BMS situation’ after 3–15 months, were developed to reduce the risk of chronic vascular inflammation. Despite this hypothetical advantage of BP-DES, meta-analyses of clinical trials did not confirm superiority of BP-DES as compared to DP-DES with respect to target lesion failure or ST.^19,21,22 A caveat is that these clinical outcome studies had limited follow-up between 2.3 and 2.6 years. Thus, potential consequences of slowly growing neoatherosclerotic plaques inside stents may not yet be fully apparent in clinical studies. Indeed, longer time from implantation is an important predictor for neoatherosclerosis.^1,5,23 However, this trial failed to demonstrate superiority of BP-EES over DP-EES with respect to the frequency neoatherosclerosis. An exception to the above mentioned clinical data represents the BIOSTEMI study, which showed superiority of a BP-DES (ultra-thin strut sirolimus-eluting) as compared with a DP-DES (i.e. same device as in our study) throughout 5 years with respect to rates of target lesion failure among patients with STEMI.²⁴ The difference in target lesion failure, however, emerged within the first 2 years according to a landmark analysis, suggesting no difference in late occurring stent failure thereafter. The absence of a clinical effect on late stent failure beyond the second year thus supports our observation of no difference in the frequency of neoatherosclerosis between biodegradable and durable polymer stents with otherwise comparable design.

Collectively, our observation of an equal risk of neoatherosclerosis development as late as 3 years is an important indication of balanced risks of very late occurring stent failure events among patients treated with either BP- or DP-EES, providing a forecast of equal clinical outcomes with the two devices during long-term follow-up.

Similar to our trial, there were no differences in the risk of neoatherosclerosis in the TRANSFORM-OCT comparing BP-EES and a DP zotarolimus-eluting stent after 18 months (11.6% vs. 15.9%, P = .56).²⁵ However, the number of patients undergoing follow-up OCT was half as compared to the present study, only a small proportion of STEMI patients were included and follow-up duration may have been too short to detect a potential long-term benefit of BP-DES after the complete degradation of the polymer as compared to the DP-DES with polymer persistence.

The frequency of neoatherosclerosis after 3 years in this study was relatively low compared to previous trials reporting numbers between 11% and 15.9% in patients assessed by OCT between 1 and 3 years,^25,26 or between 20% and 35% at 5 years.^17,27 This is particularly remarkable, since culprit lesions in acute myocardial infarction patients are reportedly at a higher risk for delayed endothelial healing due to longer persistence of the lipophilic antiproliferative drug in the necrotic core of the frequently underlying unstable plaque, absence of enough viable tissue at the culprit site, and/or greater pro-inflammatory paracrine milieu.¹⁰ The low frequency in our study is largely explained by the routine use of high-intensity statin therapy in 80% of patients as per protocol with an excellent on treatment LDL-cholesterol of 65 mg/dL, reportedly lower as in other trials such as for example the aforementioned TRANSFORM-OCT study.²⁵ Indeed, patients on high-intensity statin therapy were significantly less affected by neoatherosclerosis at 3 years (8.5%) as compared to patients on no, low or moderate-intensity statin therapy (27.8%), notably independent of stent type implanted. This finding highlights the relevance of guideline-endorsed high-intensity statin therapy on the prevention of late occurring stent failure, i.e. the direct consequence of neoatherosclerosis.

Interestingly, the primary outcome findings were identical in Swiss and Japanese patients, suggesting no difference in the formation of accelerated in-stent atherosclerosis between Asian vs. European ethnicity. Similarly, no significant interaction emerged for diabetes status.

Consistent with previous clinical data,^28,29 there were no differences in clinical restenosis and angiographic late lumen loss between treatment groups. There was, however, a trend towards a higher neointimal thickness and a significantly higher neointimal area in BP-EES as compared to DP-EES-treated patients, suggesting a measurable but clinically non-significant difference in the antiproliferative effect of BP-EES as compared to DP-EES. Likewise, in the randomized MECHANISM trial,³⁰ STEMI patients treated with the BP-EES as compared to the DP-EES reportedly had a greater neointimal area (1.44 ± 0.15 vs. 0.86 ± 0.14 mm², P < .001) 12 months after pPCI. The greater neointimal growth may be attributable to the BP-EES design, where the everolimus-carrying polymer is limited to the abluminal strut surface, leaving a bare-metal surface allowing a more rapid endothelial healing at the adluminal face. The degree of neointimal hyperplasia was observed to be associated with the incidence of neoatherosclerosis.³¹ In our study, this difference in neointimal area did not result in different rates of neoatherosclerosis between the stent types. Other healing aspects include malapposition and stent strut uncoverage. Using a neointimal healing score that weights both parameters at a lesion level, we found no difference between treatment groups.

Limitations

The study has some limitations to be considered. First, the study had an open-label design. However, the primary endpoint was assessed by Corelab assessors unaware of treatment allocation. As a strength of this study, a differentiation of the stent types at the Corelab level was impossible due to the similar strut thickness of the two stent platforms. Second, the observed frequency of neoatherosclerosis at 36 months was less than anticipated in our sample size calculation, based on neoatherosclerosis prevalence reported from autopsy and a long-term follow-up OCT study assessing earlier-generation DES. The use of high-intensity statin in most patients in this study (∼80% at 3 years) has contributed to lower occurrence of neoatherosclerosis. Third, the attrition rate was slightly higher (24%) than expected (20%), which was partially related the COVID-19 pandemic. The attrition rate was higher among diabetic patients (38%), but did not differ between countries (26% in Japan and 24% in Switzerland) or between stent groups (26% in the BP-EES group and 23% in the DP-EES group). The power to detect between-arm differences in the frequency of neoatherosclerosis was reduced due to both, the lower-than-expected frequency of neoatherosclerosis and the higher-than-expected attrition rate, accordingly, we cannot exclude a type II error. To address the concern on the 61 patients not undergoing follow-up, we conducted a sensitivity analysis using multiple imputation and found primary endpoint results to be robust (see Supplementary data online, Table S10). However, given that the observed neoatherosclerosis frequency was numerically similar between groups, it is unlikely that a larger sample size would have revealed clinically relevant differences between stent types. Fourth, two patients in each treatment arm received a target lesion revascularization prior to the 3-year follow-up and did not undergo OCT. As three of these four target lesion revascularizations occurred within the first year after primary PCI, the presence of neoatherosclerosis as reason for the restenosis is unlikely. Finally, this study was a mechanistic study not designed to evaluate differences in clinical outcomes.

Conclusions

This intercontinental (Japanese–Swiss) randomized trial showed no difference in the long-term occurrence of neoatherosclerosis assessed by OCT 3 years after pPCI in STEMI patients with a BP vs. DP current generation thin-strut EES. While it failed to demonstrate superiority of the BP-EES, its findings thus expand current evidence of similar long-term vascular responses of both, BP-EES and DP-EES up to 3 years. The results must be interpreted in the light of a lower than expected occurrence of neoatherosclerosis at 3 years and therefore limited power. Accordingly, we cannot exclude a type II error and larger studies may be warranted.

Supplementary data

Supplementary data are available at European Heart Journal online.

Declarations

Disclosure of Interest

M.T. received research grants to the institution by Boston Scientific. R.K. received consulting fee from Infraredx USA, speaker fee from Abbott Medical Japan, Boston Scientific Japan, Philips Japan, and Orbusneich Medical, and manuscript writing fee from Orbusneich Medical and Philips Japan. K.A. received speaker/consultation fees from Abbott Medical Japan, Japan Lifeline, Medtronic Japan, and Biotronik Japan. S.S. has received research grants to the institution from Edwards Lifesciences, Medtronic, Abbott Vascular, and Boston Scientific and consultant fees from Boston Scientific and Teleflex. J.L. received speaker fees to the institution from Abbott and Edwards Lifesciences. Y.U. reported grants from Astellas Pharma, and personal fees from Abbott Vascular, Amgen, Bayer, Daiichi Sankyo, Kowa, NIPRO, and Novartis, outside the submitted work. S.L. is employed by the CTU Bern, University of Bern, which has a staff policy of not accepting honoraria or consultancy fees. However, CTU Bern is involved in design, conduct, or analysis of clinical studies funded by not-for-profit and for-profit organizations. In particular, pharmaceutical and medical device companies provide direct funding to some of these studies. For an up-to-date list of CTU Bern’s conflicts of interest, see https://www.ctu.unibe.ch/research_projects/declaration_of_interest/index_eng.html. LR received research grants to the institution by Abbott, Biotronik, Boston Scientific, Infraredx, Sanofi, and Regeneron and personal fees by Abbott, Canon, Gentuity, Medtronic, Novo Nordisc, and Occlutech. All other authors have no competing interests to declare.

Data Availability

The data set will be available from the corresponding author on reasonable request.

Funding

CONNECT is an investigator-initiated trial supported by an unrestricted research grant to the institution by Boston Scientific, Marlborough, MA, USA and intramural grants of Bern University Hospital. The authors are solely responsible for the design and conduct of this study and all study analyses. The funding source was not involved in the design of the study, data collection, and data management, and has and will have no role in the analysis or interpretation of the study data.

Ethical Approval

The study protocol was approved by the responsible ethics committees. The trial was conducted in accordance with Good Clinical Practice guidelines and is reported following the Consolidated Standards of reporting Trials (CONSORT) 2010 reporting guideline statement for parallel-group randomized trials statement.

Pre-registered Clinical Trial Number

The study was registered at www.clinicaltrials.gov (Identifier: NCT03440801).

References

Author notes