Evolution of drug-eluting coronary stents: a back-and-forth journey from the bench to bedside

Abstract

Coronary stents have revolutionized the treatment of coronary artery disease. Compared with balloon angioplasty, bare-metal stents (BMSs) effectively prevented abrupt vessel closure but were limited by in-stent restenosis (ISR) due to smooth muscle cell proliferation and neointimal hyperplasia. The first-generation drug-eluting stent (DES), with its antiproliferative drug coating, offered substantial advantages over BMSs as it mitigated the risk of ISR. Nonetheless, they had several design limitations that increased the risk of late stent thrombosis. Significant advances in stent design, including thinner struts, enhanced polymers’ formulation, and more potent antiproliferative agents, have led to the introduction of new-generation DES with a superior safety profile. Cardiologists have over 20 different DES types to choose from, each with its unique features and characteristics. This review highlights the evolution of stent design and summarizes the clinical data on the different stent types. We conclude by discussing the clinical implications of stent design in high-risk subsets of patients.

Drug-eluting stents, Percutaneous coronary intervention, Coronary artery disease, Ischaemic heart disease, Stent thrombosis, In-stent restenosis

1. Introduction

Percutaneous coronary intervention (PCI) is one of the most common therapeutic interventions in contemporary clinical practice. It was initially introduced by Andreas Gruentzig who performed the first coronary angioplasty in 1977. This breakthrough procedure was frequently complicated by acute vessel closure and often required urgent surgical intervention.¹ The introduction of bare-metal stents (BMSs), with a unique cylindrical design, solved the problem by preserving luminal patency and integrity.^2,3 However, the hyperplastic neointimal response to BMS materials was associated with high rates of in-stent restenosis (ISR) and repeat revascularization, which prompted the search for alternative technologies.^4–6 Drug-eluting stents (DESs) were developed to reduce ISR risk, as the direct delivery of antiproliferative agents to the arterial wall remarkably limited neointimal tissue growth.^7–9 Yet, a few years after their introduction, two independent meta-analyses presented at the European Society Cardiology (ESC) 2006 congress revealed that DES potentially increase the mortality risk.^10,11 The so-called ESC Firestorm ignited concerns across the globe and resulted in a comprehensive re-evaluation of the safety profile of these coronary devices.¹² New evidence emerged from reassessment studies and reiterated the focus on the complex relationship between DES design (i.e. platform, polymer, and delivered drug) and arterial healing. Studies have shown that drug release kinetics, polymer biocompatibility, and strut thickness directly influence long-term stent-related outcomes.¹³ These observations guided the design of new-generation DES, with robust platform materials, decreased polymer hypersensitivity reactions, and highly effective drug formulations.

This review discusses the design of coronary stents, the evolution from early to new-generation DES, and the clinical evidence supporting their use in contemporary clinical practice.

2. Coronary vessel healing, restenosis, and thrombosis

2.1 Arterial healing

Although angioplasty aims to restore coronary anatomy and reinstate laminar flow, it inevitably causes a mechanical injury to the endothelium and triggers a cascade of inflammatory reactions. The coronary stent itself is a foreign body that induces neointimal growth as a general healing process, mediated by the migration and proliferation of smooth muscle cells within the arterial wall and the activation of chemokines and cytokines.¹⁴ Although DESs were initially designed to limit cellular proliferation and neointimal hyperplasia, the persistence of polymer coatings led to impaired healing, chronic inflammation, and local hypersensitivity reactions. In addition to pathology studies, a significant amount of our understanding of coronary healing following stent implantation came from optical coherence tomography studies conducted over the last two decades.^15–17 Theoretically, the stent strut should be covered by a fully functional endothelium. However, strut constituents induce inflammation through cellular infiltration and fibrin clot formation at different time points of the healing process. As a result, thickening of the intima and progression to ISR or stent thrombosis (ST) may occur.¹⁸

2.2 In-stent restenosis

ISR refers to the progressive narrowing (>50% of the cross-sectional area of the vessel lumen) of the stented arterial segment as an exaggerated coronary healing response to the mechanical insult and foreign material.¹⁹ The inflammatory cascade occurs over three main stages: (i) activation of thrombotic and inflammatory pathways (early phase), (ii) formation of granulation tissue by smooth muscle cell proliferation and migration (intermediate phase), and (iii) tissue remodelling (late phase).^14,18,20,21 In the early phase, endothelial cells are merely crushed, promoting activation and aggregation of platelets, infiltration of circulating leucocytes, and release of growth factors and cytokines. Although different mechanisms reduce the coronary lumen, neointimal hyperplasia is the most notable through an uncontrolled proliferation of smooth muscle cells.^22,23 This process is evident within the first year after stent implantation,^4,5 while the accumulation of proteoglycan–collagen matrix becomes more relevant in the long term. Another important but rare mechanism revealed is the development of neo-atherosclerosis, especially after 2 years of stent placement.²² Indeed, this process, mainly due to endothelial dysfunction and hypersensitivity reaction to polymer components, may be associated with aneurysm formation and late strut malapposition, which in turn amplifies flow disturbance between stent surface and platelets.²⁴ The incidence of neo-atherosclerosis in first- and second-generation DES appears to be accelerated compared with BMS.²⁵ Furthermore, the prevalence of neo-atherosclerosis increases with time in both DES and BMS and leads to stent failure.^23,26

2.3 Stent thrombosis

While ISR occurs gradually, ST is the sudden occlusion of a stented coronary artery by a thrombus within or near the stented segment. Although relatively rare, ST is a severe complication associated with high mortality rates and accounting for up to one-fifth of all myocardial infarction cases in patients with known coronary artery disease.^27,28 The Academic Research Consortium classified ST according to its degree of diagnostic certainty (definite, probable, or possible) and the time of occurrence as early—within 30 days of PCI (further stratified into acute—within 24 h of PCI—and subacute—from 24 h to 30 days), late—between 31 and 365 days of PCI, and very late—after 1 year of PCI.¹⁹ The mechanisms leading to ST are multifactorial and include patient-, lesion-, procedure-, and device-related factors (Figure 1). In particular, clinical risk factors [i.e. diabetes mellitus (DM), malignancy, heart failure, etc.] and complex lesions (e.g. bifurcation treated with two stent techniques, complex lesions, small vessel size, postprocedural TIMI flow <3) are well-established predictors of ST.^29,30 Interestingly, following stent implantation, the association between ST and some risk factors is time-dependent. For instance, diabetes and kidney disease correlate with late ST, whereas treatment of ISR and small coronary vessels are predictors of early ST.³¹

Figure 1

Factors associated with ST following DES implantation.

3. Coronary stent design

3.1 Early studies with animal models

Animal models (mainly large elastic rabbit arteries and porcine coronary arteries) were essential to understanding the impact of stent design on its interaction with the arterial wall. Using infrarenal abdominal aortas of 10 healthy rabbits undergoing balloon angioplasty, it was found that stents with gaps in-between the coils have reduced neointimal thickness within the device itself, decreasing the ISR rates compared with stents without these gaps.³² Furthermore, a linear relationship between neointimal formation after stent implantation and arterial injury inflicted by the stent strut was described by Schwartz in a porcine coronary model.³³ Nonetheless, these models had their limitations owing to the inherent physiological differences between animals and humans. For example, the atherosclerotic rabbit iliac artery model is limited by the fact that the atheroma developed by the hypercholesterolaemic rabbit is different in texture and composition (i.e. soft and constituted mainly of lipid-laden macrophages) from the one observed in humans.^32,34,35 Similarly, despite the fact that the healing response to vascular injury in pig models is very similar to that observed in humans, the time course of events is substantially different.^35,36 Neointimal growth after BMS implantation in porcine coronary arteries peaks at 1 month and then decreases over the next 3–6 months. In contrast, it peaks at 6–9 months in humans and then decreases over 3 years.^37,38 In other words, healing after BMS implantation takes at least two to three times longer in humans than in pig models. Nevertheless, pig models are generally more representative of atherosclerosis in humans than rabbit models.

Preclinical animal studies revealed significant reductions in neointimal formation with first-generation DES vs. BMS at the expense of chronic inflammation, delayed arterial healing, and incomplete endothelialization.^39,40 Yet, clinical trials in humans showed the superiority of first-generation DES over BMS in reducing clinical and angiographic restenosis.⁷ The discordance between preclinical studies in animals and clinical trials in humans was attributed to individual patient sensitivity to polymers or antiproliferative drugs, different healing speeds, heterogeneity in plaque morphologies, and drug-release kinetics. While the animal models provided minimal insights into the efficacy of DES, they were quintessential to evaluating the safety of these devices and directly contributed to their refinement.

3.2 Evolution in stent design

Over the past years, the evolution in stent design has mainly focused on improving deliverability and increasing radial strength while reducing the strut thickness and easing side-branch access. However, this came at the expense of a decreased stent capacity to resist longitudinal force (primarily related to the interaction between the balloon and the stent) and maintain stent’s structure.^41,42 The radial strength depends on several factors such as strut width and thickness, the type of alloy used, and the architecture of the stent itself. There are two main stent architectures, each with its advantages and disadvantages: (i) closed-cell stents (connected on all sides) and (ii) open-cell stents (not connected on all sides). While the first offers stronger radial force and greater plaque coverage, the latter (i.e. open-cell design, which is used for most contemporary DES) confers more flexibility allowing better deliverability and side-branch access.

DES comprises three main components: (i) a stent platform, (ii) a polymer coating, and (iii) an antiproliferative drug (Figure 2).

1. Early DESs were made of stainless-steel alloys (iron, nickel, and chromium) which were relatively thick and occasionally associated with allergic reactions. To overcome these limitations and enhance safety, new metal alloys (primarily cobalt–chromium and platinum–chromium) were used in newer-generation devices. Furthermore, bioresorbable stents were developed with an utterly resorbable structure over time. These devices were made of materials such as polycarbonate polymers, poly-lactic acid, salicylic acid polymers, and magnesium. The main ‘theoretical’ advantage of bioresorbable stents was restoring normal vascular motion and avoiding any permanent metallic material within the vessel wall. However, initial results were disappointing because of an excessive risk of late ST.

2. Single or multiple polymer layers can be stacked on the stent surface: an adhesion layer at the base followed by another layer that stores and releases the antiproliferative agent directly (by contact transfer) into the arterial wall or indirectly into the bloodstream where it is subsequently absorbed into the arterial wall. Occasionally, these are topped by a third layer that controls drug release and prolongs its antiproliferative effect. Most early-generation DES used durable coatings, which appeared to be associated with inflammatory reactions in the long term, and some possibly led to thrombosis. As a result, joint biomedical and engineering research efforts led to the development of new coating approaches or no polymer coating (i.e. polymer-free), as discussed in the upcoming sections.

3. The drug released from the polymer inhibits the proliferation of smooth muscle cells and thus limits neointimal growth within the stent and its surrounding. Since most neointimal hyperplasia is induced by inflammation, immunosuppressants and antiproliferative drugs are typically used (Figure 3). The immunosuppressive effect is particularly important in polymer-coated stents to suppress hypersensitivity reactions that trigger ST or late restenosis.

Figure 2

The three main components of contemporary DESs.

Figure 3

Major drugs used in DESs.

4. Types of drug-eluting stents

4.1 Early-generation drug-eluting stents

4.1.1 First-generation sirolimus and paclitaxel-eluting stents

Enhanced knowledge of the pathophysiological mechanisms underlying neointimal hyperplasia has allowed the identification of molecular compounds that limit smooth muscle cell proliferation. Two different classes of antiproliferative agents, with distinct mechanisms of action, were used in early-generation DES: (i) mammalian target of rapamycin (mTOR) inhibitors (i.e. sirolimus) and (ii) taxans (i.e. paclitaxel).^43–45 At the molecular level, sirolimus binds to the cytosolic protein FK-binding protein 12 (FKBP12) and forms the sirolimus–FKBP12 complex.⁴⁶ The latter acts as an inhibitor of the mTOR protein, a downstream target of the phosphatidylinositol-3 kinase pathway that regulates diverse cellular processes, including proliferation and survival.⁴⁶ Conversely, paclitaxel stabilizes the microtubule polymers and prevents disassembly, thus inhibiting cell division in the G0/G1 and G2/M phases. These drugs are stored and then released in a timely manner from polymers mounted on the stainless steel (316L) platform of early-generation DES. Poly(styrene-b-isobutylene-b-styrene) polymer is used in the paclitaxel-eluting stent (PES, TAXUS™, Boston Scientific Corporation, USA), while the polyethylene-co-vinyl acetate and poly(n-butyl methacrylate) polymers are used in the sirolimus-eluting stent (SES, CYPHER™, Cordis Corporation, USA). The effectiveness of either of the two stents is directly related to the drug release kinetics and the delivered drug concentration. Indeed, angiographic and intravascular ultrasound findings at 4-year follow-up among patients who underwent SES implantation showed higher luminal patency in slow-release (up to 60 days post-stent implantation) vs. fast-release (within 7 days of stent implantation) Cypher stents.⁴⁷ In contrast, few differences were found between slow- and moderate-release PES with respect to in-stent net volume obstruction at 6 months as measured by intravascular ultrasound.⁴⁸ Importantly, the drug release kinetics had a larger effect on the inhibition of the neointimal hyperplasia than the drug dose.⁴⁹

Clinical trials investigating first-generation DES established the efficacy of these devices over BMS in terms of reduction of restenosis rates.⁵⁰ A collaborative network meta-analysis of 38 trials (n = 18 000) directly comparing sirolimus and paclitaxel DES with each other or with BMS revealed no mortality differences between the three stent types up to 4-year follow-up.⁵¹ However, SES was associated with a reduced need for target lesion revascularization compared with PES, and both DES types were superior to BMS.⁵¹ Nevertheless, there was an apparent increase in the rate of late ST with SES and PES.^10,11

Several plausible biological mechanisms related to the inhibition of neointimal growth and development of neo-atherosclerosis explain the increased risk of ST with first-generation DES. Upon exposure to the non-endothelialized stent surface, mainly in the early phase after stent implantation, thrombosis is primarily driven by platelet activation.⁵² Furthermore, localized hypersensitivity reactions triggered by the host response to the foreign polymer increase the risk of late and very late ST.²⁴ Indeed, autopsy findings revealed aneurysmal dilation of the stented coronary segment and extensive infiltration of macrophages, T cells, eosinophils, and plasma cells into the three layers of the arterial wall.²⁴ These inflammatory reactions are evident 3 months after stent implantation, which correlates with the time needed to elute sirolimus fully. Similar findings were observed in a porcine model in which implantation of the Cypher stent decreased neointimal formation at 1 month but was associated with cellular proliferation and delayed intimal thickening at long-term follow-up, when compared with BMS.⁵³ These observations, supported by data from human pathology studies and clinical trials, suggested that outcomes may be improved by decreasing excessive inflammation and risk of ST through changes in polymers’ composition to limit the activation of platelets and immune cells.^22,54,55

4.1.2 Actinomycin-D-eluting stents

Actinomycin-D is an antibiotic with potent antiproliferative properties that inhibit ribonucleic acid transcription by forming a stable complex with double-stranded deoxyribonucleic acid at the transcription initiation complex, thus preventing ribonucleic acid synthesis.⁵⁶ The safety of the actinomycin-D-eluting stent was first demonstrated in a porcine coronary model, where the histologic reaction was comparable to BMS at 180 days. The ACTION (ACTinomycin-eluting stent Improves Outcomes by reducing Neointimal hyperplasia) trial randomized 360 patients to receive a DES (2.5 or 10 μg/cm² of actinomycin-D) or BMS.⁵⁷ Despite similar in-hospital and 1-month outcomes, increased rates of restenosis, late lumen loss, and target lesion revascularization were noted in the DES arm at 6 months. Although actinomycin-D-eluting stents were not successful, they provided great insights into the development of newer-generation DES.

4.2 Newer-generation drug-eluting stents

4.2.1 Durable-polymer drug-eluting stents

The cobalt–chromium (CoCr) Xience (Abbott) and the platinum chromium (PtCr) PROMUS Element (Boston Scientific) everolimus-eluting stents (EESs) contain highly fluorinated polymers that decrease platelet activation and adhesion.⁵⁸ These polymers prolong everolimus release, with 25% of the drug released during the first day and up to 75% during the first month.⁵⁹ In the PLATINUM trial, comparing CoCr-EES with PtCr-EES in patients with one or two de novo lesions, both stent types showed similar rates of target lesion failure (3.2% in CoCr-EE vs. 3.5% in PtCr-EES, P = 0.72) and very low definite or probable ST (0.4 vs. 0.4%, P = 1.00) at 12-month follow-up.⁶⁰

The Endeavor zotarolimus-eluting stent (ZES) has a polymer similar to the phospholipid phosphorylcholine molecules found in cellular membranes; the polymer releases up to 95% of zotarolimus within the first 2 weeks after implantation. Nevertheless, the lack of antiproliferative drugs after this short timeframe resulted in unexpected high rates of restenosis and the need for repeat revascularization.⁶¹ To enhance drug release kinetics and ensure effective inhibition of neointimal hyperplasia, the Endeavor ZES was revamped to include BioLinx™, a new polymer with a hydrophilic surface that repels thrombogenic plasma proteins. The newly designed stent, known as Resolute ZES, was designed to overcome the limitations of the Endeavor ZES by prolonging the release of zotarolimus, with 85% of the drug released within 2 months and the remaining 15% within 6 months.⁶² The RESOLUTE randomized trial, enrolling patients with few exclusion criteria, showed that ZES was non-inferior to EES for target lesion failure at 12 months.⁶³ Similarly, other large randomized clinical trials confirmed the similar safety and efficacy of EES and ZES.^64–67

As confirmed by human autopsy analyses, improvements in new polymers’ conformity with the biologic milieu were thought to decrease inflammation and fasten vessel healing and re-endothelization.⁶⁸ Pathological findings correlate with clinical outcomes as both EES and ZES were found to be superior to PES and at least non-inferior to early-generation SES for target lesion failure.^69–72

4.3 Biodegradable and polymer-free DESs

A durable polymer on the stent surface has been associated with delayed arterial healing and chronic inflammation in human autopsy findings, which led to the development of DES with either a biodegradable polymer or no polymer.^22,73 This is intended to maximize early healing and minimize the risk of late and very late ST, ultimately allowing short-duration dual antiplatelet therapy (DAPT) regimens.

4.3.1 Biodegradable polymer drug-eluting stents (BP-DES)

Biodegradable polymer drug-eluting stents (BP-DES) are typically coated with poly-D-lactide acid, poly-L-lactide acid, or poly-lactic-co-glycolic acid, which are metabolized, after drug elution, into water and carbon dioxide molecules, leaving behind only a metallic structure. Although there is a high degree of similarity among various BP-DES, some essential variations exist. The most notable differences are the time required for the polymer to degrade, the time needed for complete elution of the antiproliferative agent, and the strut thickness. Table 1 lists the characteristics of currently available BP-DES. We discuss here two widely used BP-DES.

Table 1

Summary of biodegradable polymer DESs characteristics

Stent name	Stent material	Strut thickness	Drug–polymer carrier	Eluting drug	Additional features
Synergy	Platinum chromium alloy	74 μm; 79 μm; 81 μm	Poly (Dl-lactide-co-glycolide)	Everolimus
Synergy megatron	Platinum chromium alloy	89 μm	Poly (Dl-lactide-co-glycolide)	Everolimus
Orsiro	Cobalt chromium alloy	60 μm; 80 μm	Bioabsorbable poly-L-lactide	Sirolimus	Passive proBIO luminal silicon carbide layer minimizing contact between stent struts and circulating blood
Ultimaster	Cobalt chromium	80 μm	Poly (Dl-lactide-co-caprolactone)	Sirolimus	Gradient polymer coating with lack of polymer in areas of high physical stress
BioMatrix	Flex: Stainless steel	120 μm	Poly-lactic acid	Biolimus A9
	Alpha: Cobalt chromium
COMBO	Stainless steel	100 μm	Abluminal	Sirolimus	Endothelial progenitor cell capture, with circumferential coating of anti-CD34 antibodies
MiStent	Cobalt chromium	64 μm	Polylactide-co-glycolic acid	Crystalline sirolimus	Patented supercritical fluid technology allows a rigorously controlled drug/polymer coating to be applied to the bare-metal stent
BioMime	Cobalt chromium	65 μm	Poly-L-lactic and poly-lactic-co-glycolic acids	Sirolimus	Closed cells at the edges and large open cells in the middle
Supraflex	Cobalt chromium	60 μm	Poly(L-lactide), poly(Dl-lactide-co-glycolide), and polyvinylpyrrolidone	Sirolimus
Yukon Choice PC	Stainless steel	87 μm	Biodegradable polymer Resomer R202S on the abluminal surface with top coat of shellac resin	Sirolimus	Microporous PEARL surface for better endothelial cell adhesion
Firehawk	Cobalt chromium	86 μm	Abluminal polymer, poly-lactic acid	Sirolimus	Abluminal recesses with drug–polymer matrix
Nobori	Stainless steel	120 μm	Abluminal polymer poly-lactic acid	Biolimus A9

The Synergy stent (Boston Scientific) consists of a biodegradable everolimus-eluting polymer mounted on a PtCr platform. The everolimus drug content of the stent is 1 μg/mm² and is released over 3 months from an abluminal coating of poly-lactic-co-glycolic acid polymer that degrades after 4 months. The Synergy stent is characterized by a thin strut (74 μm) that eases its delivery, minimizes its intrusion into the arterial lumen, and improves endothelialization 1 year after implantation, compared with a durable polymer (DP) DES.⁷⁴ The EVOLVE trial revealed statistically and clinically significant decreases in the rates of ST and target lesion revascularization with the biodegradable polymer Synergy EES than with the PROMUS Element DP-EES.⁷⁵

The Orsiro (Biotronik) biodegradable polymer sirolimus-eluting stent has a CoCr platform with an ultra-thin strut (60 μm) coated circumferentially with poly-L-lactide that fully degrades within 2 years of implantation. Another unique feature of the Orsiro stent is its luminal surface which includes an amorphous silicon carbide layer capable of minimizing contact with circulating platelets. The BIOFLOW V randomized trial showed a significant decrease in the primary endpoint of target lesion failure at 12 months with the Orsiro biodegradable polymer SES than the durable polymer Xience EES [6 vs. 10%; 95% confidence interval (CI) (−6.84 to −0.29), P = 0.0399],⁷⁶ with benefits largely confirmed in subsequent studies.⁷⁷

4.3.2 Polymer-free drug-eluting stents

Polymer-free drug-eluting stent (PF-DES) allows a controlled release of the antiproliferative drug from the stent surface without the need for a polymer by modifying the stent surface and drug-matrix formulations. There are two commercially available polymer-free DES in the USA: (i) The BioFreedom stent (Biosensors) and (ii) the Cre8 stent (CID-Alvimedica). The BioFreedom stent was initially made of a 316L stainless platform (strut thickness = 112 µm), then replaced by CoCr in the newer generation (BioFreedom Ultra, strut thickness = 84 µm). The platform is coated with Biolimus A9 (also called umirolimus), a highly lipophilic derivative of sirolimus. Around 90% of the antiproliferative drug is released from the stent within 2 days, with the remaining 10% over 28 days. In the EGO-BIOFREEDOM study, neointimal strut coverage increased from 85.8% at 1 month to 99.6% at 9 months.⁷⁸ The median neointimal volume was 13% at 9 months, and angiographic late lumen loss was 0.21 ± 0.30 mm.⁷⁸ At 6 months, an equivalent early and a superior late reduction in intimal proliferation was observed with the BioFreedom stent compared with an SES in a porcine model.⁷⁹ The BioFreedom stent was tested in multiple large randomized trials involving patients at high-bleeding risk (HBR) (discussed below). The SORT OUT IX (Scandinavian Organization for Randomized Trials With Clinical Outcome IX) trial randomized 3151 patients to either the BioFreedom stent or the Orsiro stent.⁸⁰ BioFreedom stent did not meet the criteria for non-inferiority to Orsiro in terms of 1-year major adverse cardiac events, mainly due to increased rates of target lesion revascularization with the BioFreedom than with the Orsiro stent [3.5 vs. 1.3%, rate ratio: 2.77, 95% CI (1.66–4.62), P < 0.0001].

The polymer-free amphilimus-eluting Cre8 stent comprises a CoCr alloy platform with a thin strut (89 μm) and two platinum markers at the stent ends. The antiproliferative agent amphilimus, composed of sirolimus and an organic fatty acid, is eluted from drug reservoirs in a sustained and stable fashion. To limit the risk of allergic reactions and platelet adhesion, the stent surface is coated with an ultra-thin (<0.3 μm) layer of carbon (BioInducer surface) covalently bonded to the CoCr platform. A large proportion of the drug is released within the first few days of implantation with a complete elution by 90 days.⁸¹ Clinical data on Cre8 stents are limited to a few randomized trials and registries.^82–84 In the ReCre8 trial (n = 1502), Cre8 was non-inferior to Resolute Integrity ZES for target lesion failure at 12 months (5.6% in Resolute vs. 6.2% in Cre8 patients, P for non-inferiority 0.0086).⁸²

4.4 Bioresorbable vascular scaffolds

The bioresorbable vascular scaffolds (BVS) technology is one of the most controversial inventions in interventional cardiology. Although BVS theoretically restores blood flow as any coronary stent, its design is fundamentally different since the device completely resorbs over time.⁸⁵ The BVS hypothesis was developed on the assumption that the absence of a metallic platform would enhance complete restoration of vasomotion, facilitate luminal expansion, and mitigate lesion-related events compared with permanent metallic DES. Although several BVS have undergone clinical testing, the Absorb BVS (Abbott Vascular, Santa Clara, CA, USA) is the only one that was extensively evaluated and used worldwide. The Absorb BVS is made of a poly-L-lactide backbone coated with an amorphous matrix of poly-D, L-lactide that releases everolimus as an antiproliferative drug. The poly-L-lactide is a semi-crystalline polymer made of crystal lamella interconnected with several polymer chains creating an amorphous segment.⁸⁶ The first version of the scaffold, Absorb BVS 1.0, has a strut thickness of 150 µm, a crossing profile of 1.4 mm, and out-of-phase zigzag hoops linked together by thin and straight bridges. The device has to be stored at –20°C to maintain its integrity. The second generation, Absorb BVS 1.1, has a similar strut thickness, in-phase zigzag hoops linked by bridges, and can withstand room temperature.⁸⁷ In addition, the polymer in the newer device’s version has slower hydrolysis rates, a feature that strengthens the radial force of the scaffold.

Hydrolysis constitutes the primary chemical process by which BVS degrade in vivo. Water can easily penetrate the loosely packed amorphous regions that bind the crystallites. The timing of the hydrolysis process is a key determinant factor for its performance. Indeed, the mechanical integrity that prevents recoil should be maintained over at least 6 months until the biological processes leading to restenosis become of less significance.⁸⁸ To note, the polymer is typically replaced by a provisional proteoglycan-based matrix followed by new connective tissue formation.⁸⁹ The optical coherence tomography analysis conducted within the ABSORB Cohort A study has shown that the complete resorption of the scaffold takes up to 3 years after implantation.⁹⁰ After the resorption process is complete, ‘vascular restoration’ occurs through tissue adaptation and cellular reorganization.⁹¹

Clinical evidence generated so far regarding the safety and efficacy of BVS has not met the expectations. In 2008, Ormiston et al.⁸⁶ demonstrated the feasibility of implanting the Absorb BVS 1.0, with a satisfactory in-stent late loss, acceptable neointimal hyperplasia, and minimal stent area obstruction at 6-month follow-up by imaging (i.e. angiography and intravascular ultrasound). Of note, BVS implantation should be considered a challenging procedure as the poly-L-Lactide is not inherently opaque and hence is hard to visualize by fluoroscopy, which renders stent delivery and deployment difficult. In addition, these stents were mainly used in simple lesions and large vessels as delivery to more complex lesions and narrower vessels is technically challenging.

Improvement in the BVS design and mechanical integrity led to the second-generation Absorb BVS 1.1, first approved in Europe in 2011 and 5 years later in the USA. Several randomized trials compared Absorb BVS with newer-generation metallic DP-EES in a broad range of patients with different risk profiles.^92,93 In a patient-level data pooled meta-analysis of the ABSORB trials (n = 3389), when compared with EES, BVS was associated with increased rates of ST and target lesion failure at 1 year and cumulatively through 3 years of follow-up.⁹⁴ Although experience with BVS implantation (i.e. intravascular imaging guidance coupled with routine pre-and post-dilatation) decreased device-related events in the ABSORB IV and COMPARE-ABSORB trials, concerns about scaffold thrombosis remained, and the use of the Absorb BVS is currently limited to clinical research.^95,96

The occurrence of early scaffold thrombosis has been documented in multiple reports.^97–99 It is possibly attributed to the strut thickness of ∼150 μm (almost twice that of new-generation DES), technical aspects of the procedure (i.e. lesion preparation and scaffold expansion), and post-stenting antithrombotic therapy. Similarly, late and very late scaffold thrombosis have also been described in the literature and are potentially related to intraluminal scaffold dismantling and restenosis during the resorption process.¹⁰⁰ Another relevant mechanism leading to very late scaffold thrombosis is neo-atherosclerosis, induced by chronic inflammation, suboptimal endothelial coverage, and blood flow turbulences linked to the scaffold architecture and strut size.^101,102 Indeed, a strong inflammatory reaction related to the absorption process facilitates the recurrence of atherosclerotic disease within the neointima. This phenomenon has been described in long-term (up to 5 years) follow-up imaging studies in patients who have undergone BVS implantation.¹⁰³

5. Drug-eluting stents in high-risk subsets of patients

5.1 Diabetes mellitus

Patients with DM usually have advanced atherosclerotic coronary disease, which increases the risk for recurrent ischaemic events and the need for repeat revascularization.^104,105 Despite significant improvements in coronary stent design, PCI outcomes in diabetic patients remain inferior to those observed in non-diabetic.^106–108 This has been mainly attributed to the pathophysiological mechanisms associated with DM, including endothelial dysfunction, vascular inflammation, and increased platelet activation, all of which lead to a prothrombotic state. The incidence of restenosis remains a critical issue in patients with DM. Hyperinsulinaemia and insulin resistance are likely triggers for restenosis and other stent-related events in diabetic patients treated with DES.

In a network meta-analysis including 3852 diabetic patients from 35 randomized trials, first-generation DESs were found to be as effective as BMS in patients with or without DM.¹⁰⁹ Nonetheless, PCI with BMS and first-generation DES have been shown to be inferior to cardiac surgery in patients with DM.^110–113 Newer-generation stent platforms have challenged these observations. A meta-analysis of 42 trials (n = 22 844) revealed that, when compared with BMS and other DES types, newer-generation DP-EES is the most efficacious (in terms of target vessel revascularization) and safest (low rates of ST) stent in patients with diabetes.¹⁰⁵ Nonetheless, the failure of early trials to establish the superiority of the newer-generation DP-EES over PES raised concerns about whether stents eluting sirolimus or its analogues (e.g. everolimus) could be less efficacious in these patients. It was hypothesized that concurrent hyperleptinaemia in diabetic patients overrides the mTOR inhibition of sirolimus and thus increases ISR risk.^114,115 To address this question, the Taxus Element vs. Xience Prime in a Diabetic Population (TUXEDO)–India trial randomized 1830 diabetic patients undergoing PCI to either a DP-EES or a PES.¹¹⁶ At 1-year follow-up, patients assigned to the DP-EES arm had significantly lower target vessel-related myocardial infarction, ISR, and target lesion revascularization rates.¹¹⁶ The new-generation ultra-thin Orsiro biodegradable sirolimus-eluting stent outperformed DP-EES in complex patient populations, as observed in the BIOFLOW V trial in which one-third of patients had DM.⁷⁶

Recently, the SUGAR trial (n = 1175) showed a 35% relative risk reduction in the primary endpoint of target lesion failure (defined as a composite of cardiac death, target vessel myocardial infarction, and target lesion revascularization) at 1 year among patients with DM randomized to the new-generation Cre8 EVO stent compared with Resolute Onyx stent.¹¹⁷

An innovative strategy based on the Abluminus technology, a unique biodegradable polymer SES mounted on a sirolimus-coated balloon that ensures uniform drug delivery at a high concentration at the edge of the DES, is currently being tested in a large, randomized trial (ABILITY Diabetes Global, NCT04236609). Around 3000 patients with DM and undergoing PCI are randomized in a 1:1 fashion to either Abluminus DES +SES or DP-EES. The primary outcome of interest includes target lesion failure or revascularization 1 year after PCI.

As more data emerge from ongoing clinical trials, our understanding of the safety and efficacy of new-generation DES in diabetic patients will keep growing. Consequently, we will be able to determine whether the choice of a particular stent platform should be guided by the presence or absence of DM.

5.2 HBR patients

Up to 40% of patients undergoing PCI are at HBR, which makes the management of antithrombotic therapy challenging.^118,119 A shortened DAPT duration of 1–6 months, depending on PCI indication (acute coronary syndrome vs. chronic coronary syndrome), is generally recommended in these patients.¹²⁰ Although DAPT decreases the risk of thrombotic complications, the risk–benefit ratio is often uncertain in HBR patients. Traditionally, BMS represented the device of choice in HBR patients, given the increased risk of ST after DAPT cessation with early-generation DES. As ST rates have significantly decreased with new-generation DES, it has been hypothesized that shorter DAPT durations might be as safe and efficacious as prolonged DAPT.¹²¹ Consequently, trials are now shifting to testing strategies that combine contemporary stent technology with short-duration DAPT regimens. Table 2 lists trials of new-generation coronary stents with short DAPT in HBR patients.^{61,67,122–131} The LEADERS-FREE trial randomized 2466 HBR patients undergoing PCI to either a polymer-free umirolimus-coated stent (BioFreedom) or the thick strut MultiLink BMS with a 1-month DAPT regimen.¹²² Significant reductions in the primary safety (a composite of cardiac death, myocardial infarction, or ST) and efficacy (clinically driven target lesion revascularization) endpoints were noted with the BioFreedom stent. Findings from the ONYX-ONE trial revealed similar safety and efficacy with BioFreedom as with Resolute Onyx among HBR patients maintained on 1-month DAPT after PCI.⁶⁷ In the XIENCE 28 and 90 studies, including 3652 HBR patients undergoing PCI with a CoCr-EES, 1- or 3-month DAPT resulted in similar rates of ischaemic events and lower incidence of ST, when compared with up to 12-month DAPT historical control after propensity score stratified analyses.¹²⁷ The risk of major bleeding was significantly lower in patients maintained on DAPT for 1 or 3 months.¹²⁷ Recently, the MASTER DAPT trial revealed that in an all-comers HBR cohort that underwent PCI with BP-SES (Ultimaster, Terumo) stent implantation, 1-month DAPT was non-inferior to 3 months or longer with respect to major adverse cardiac events and superior in terms of bleeding.¹²⁶ The findings of this trial emphasize the impact of new stent designs on DAPT duration. Perhaps, due to a rapid and smooth re-endothelization process with new-generation stents, 1–3 months of DAPT may be sufficient in most HBR patients. Nonetheless, it should be noted that the strategies investigated in these trials are only applicable to HBR patients who receive one of the aforementioned stent platforms; whether these findings can be extrapolated to those receiving other newer-generation DES warrants further investigation.

Table 2

ȃStudies evaluating coronary stents in high-bleeding risk patients

	Study type	Sample size	Experimental arm	Control arm	Main finding
LEADERS-FREE122	Randomized trial	2432	BioFreedom polymer-free umirolimus-coated stent	BMS	BioFreedom polymer-free umirolimus-coated stent was superior to BMS with respect to the primary safety (composite of cardiac death, MI, or stent thrombosis) and efficacy (clinically driven TLR) endpoints when used with a 1-month DAPT
ZEUS59	Randomized trial	1606	Endeavour durable-polymer zotarolimus-eluting stent		Compared with BMS, Endeavour zotarolimus-eluting stent implantation, combined with an abbreviated, tailored DAPT regimen, resulted in a lower risk of 1-year MACE
SENIOR123	Randomized trial	1200	Synergy bioabsorbable polymer everolimus-eluting stent		Synergy bioabsorbable polymer everolimus-eluting stent with short DAPT duration was superior to BMS with a similar DAPT duration with respect to the occurrence of all-cause mortality, MI, stroke, and TLR
DEBUT124	Randomized trial	243	SeQuent Please drug-coated balloon (coated with paclitaxel and iopromide)		SeQuent Please drug­-coated balloon was superior to BMS with respect to MACE
ONYX-ONE65	Randomized trial	1996	Resolute Onyx slow-release zotarolimus-eluting stent	BioFreedom polymer-free umirolimus-coated stent	Use of Resolute Onyx zotarolimus-eluting stents was non-inferior to BioFreedom polymer-free umirolimus-coated stent in terms of safety (composite of cardiac death, MI, or ST) and efficacy (composite of cardiac death, MI, or TLR) outcomes
COBRA-REDUCE125	Randomized trial	996	COBRA Polyzene-F polymer stent	Drug-eluting stent (Xience/Promus, Resolute, or Synergy)	COBRA Polyzene-F polymer stent did not meet the criteria for non-inferiority to a standard drug-eluting stent with respect to MACE
MASTER DAPT126	Randomized trial	4579	Ultimaster biodegradable-polymer sirolimus-eluting coronary stent (abbreviated DAPT)	Ultimaster biodegradable-polymer sirolimus-eluting coronary stent (standard DAPT)	Among patients receiving an Ultimaster biodegradable-polymer sirolimus-eluting coronary stent, 1-month DAPT was non-inferior to standard DAPT with respect to the occurrence of NACE and MACE
XIENCE 28/90127	Non-randomized trial	XIENCE 28: 2803 XIENCE 90: 2973	XIENCE cobalt–chromium everolimus-eluting stent	Objective performance goal	Among patients receiving XIENCE cobalt-chromium everolimus-eluting stents, DAPT for 1 or 3 months was non-inferior to 6 or 12 months of DAPT for MACE
ONYX-ONE Clear128	Non-randomized trial	1506	Resolute Onyx slow-release zotarolimus-eluting stent	Objective performance goal	Favourable safety and effectiveness between 1 and 12 months were observed among patients maintained on 1-month DAPT following Resolute Onyx DES implantation
EVOLVE Short DAPT129	Non-randomized trial	1487	Synergy bioabsorbable polymer everolimus-eluting stent	Objective performance goal	Favourable rates of ischaemic outcomes were observed among patients receiving SYNERGY everolimus-eluting stent and three-month DAPT (vs. standard DAPT)
LEADERS-FREE 2130	Non-randomized trial	1203	BioFreedom polymer-free umirolimus-coated stent	BMS (historical control)	When compared with BMS, BioFreedom polymer-free umirolimus-coated stent with 1-month DAPT resulted in a significant reduction in MI and TLR at 1 year, with similar rates of ST
POEM131	Non-randomized trial	443	Synergy bioabsorbable polymer everolimus-eluting stent	None	Among high bleeding risk patients undergoing PCI with the Synergy bioresorbable-polymer everolimus-eluting stent, a 1-month DAPT regimen is safe, with low rates of ischaemic and bleeding events

BMS, bare-metal stent; DAPT, dual antiplatelet therapy; DCS, drug-coated stent; DES, drug-eluting stent; MACE, major adverse cardiac events; MI, myocardial infarction; NACE, net adverse cardiac events; PCI, percutaneous coronary intervention; ST, stent thrombosis; TLR, target lesion revascularization.

6. Knowledge gaps and future perspectives

The stent technology has reached satisfactory levels of safety and efficacy, compared with other cardiovascular devices. Refinements in stent technology will continue as far as our understanding of the biological impact of coronary stents on arterial wall healing keeps expanding. Most contemporary DESs show adequate endothelialization from 3 months onwards, though complete strut coverage may take up to 12 months. Over the past years, there has been an increased interest in intracoronary imaging to guide and optimize stent deployment during PCI. Optical coherence tomography and intravascular ultrasound have become well-established tools for lesion assessment, adequate stent sizing, appropriate apposition and expansion, and avoidance of edge dissection. Real-world registries have generated strong evidence that intracoronary imaging allows unprecedented discrimination of the mechanisms and predictors of ST at the time of the implantation.¹³² Ongoing studies will assess whether applying these mechanisms and predictors in the form of algorithms for the implantation of coronary devices is associated with a substantial reduction in early and late stent complications. Although there is compelling evidence supporting the use of intravascular imaging to decrease the risk of repeat revascularization, these tools are still underused in many parts of the world. Additional data are needed to generate more robust evidence supporting the use of intravascular imaging as a routine practice during PCI. Besides the conventional target vessel-related endpoints, studies should evaluate the role of optimal stent deployment as an endpoint using currently available state-of-the-art imaging technologies. Nevertheless, all these procedural aspects remain governed by multiple factors such as cost, reimbursement, and local availability, which are rarely if ever accounted for in clinical trials.

Future trials should determine the optimal stent choice and the adequate deployment strategy by lesion type. A thorough understanding of drug release kinetics, delivery, and retention in the diseased arterial tissue is needed. Arterial features such as porosity and tortuosity directly impact the drug transport and diffusion processes. Using advanced modelling techniques, computational studies can provide detailed predictions of fluid flow and drug kinetics in stented arteries.^133,134 Therefore, computational models of the transport of drugs eluted from DES within the diseased coronary artery wall hold the premise to considerably improve our understanding of the performance of each of these devices.

Early studies have shown that optimal stent deployment using intravascular imaging can decrease the need for prolonged antiplatelet therapy. This research area is constantly evolving in parallel to advances in device technologies.^135,136 The main challenge regarding the management of antithrombotic therapy is balancing protection against thrombotic events with excessive increases in bleeding risk. As the risk of stent-related complications is gradually decreasing, current research efforts are now focusing on mitigating the bleeding risk associated with prolonged DAPT. Several bleeding-avoidance strategies (i.e. short DAPT, P2Y12 inhibitor monotherapy, de-escalation of antiplatelet therapy, etc.) have emerged over the past few years and have been recently presented in a thorough review.¹³⁷ Nonetheless, patient-tailored DAPT approaches based on individualized risk–benefit assessments should focus on future research endeavours. These initiatives will help optimize the short- and long-term outcomes in patients undergoing PCI with DES implantation.

7. Conclusions

The evolution of coronary stents over the past 40 years of interventional cardiology has been phenomenal. New-generation DES have shown to be safer and more efficacious than older generations. A wide variety of new-generation DES is currently available on the market, each with its advantages and disadvantages. Durable-polymer stents, especially fluoropolymers, protect against ST in the early period after PCI. In contrast, biodegradable-polymer stents, through their abluminal coatings, exhibit fastened healing and re-endothelialization. Innovation in stent design and architecture will continue to improve deliverability and tolerability within the arterial wall. In turn, this will decrease the need for prolonged DAPT durations to prevent stent-related events. From this perspective, BP- or PF-DES offer particular advantages as they shorten drug action duration in the arterial wall and allow superior endothelial healing. The BVS technology, despite its disappointing preliminary results, may have promising applications if improvements in design occur.

In conclusion, it remains uncertain whether further changes in stent design will translate into improved outcomes since contemporary DES technologies have reached such high efficacy and safety standards that make it challenging to show the superiority of any new device. Future research in coronary stent technology will primarily focus on high-risk populations, imaging guidance for optimal stent deployment, and novel antithrombotic therapy strategies that minimize the bleeding risk.

Data availability

The data are available upon reasonable requests made to the corresponding author.

Author notes

Conflict of interest: R.M. reports institutional research payments from Abbott, Abiomed, Alleviant Medical, AM-Pharma, Applied Therapeutics, Arena, AstraZeneca, BAIM, Bayer, Beth Israel Deaconess, Biosensors, Biotronik, Boston Scientific, Bristol-Myers Squibb, CardiaWave, CellAegis, CeloNova, CERC, Chiesi, Concept Medical, CSL Behring, Cytosorbents, DSI, Duke University, Element Science, Faraday, Humacyte, Idorsia, Insel Gruppe AG, Magenta, Medtronic, Novartis, OrbusNeich, Philips, RenalPro, Vivasure, and Zoll; personal fees from Cine-Med Research, and WebMD; consulting fees paid to the institution from Abbott, Janssen, Medtronic, Novartis; Equity <1% in Applied Therapeutics, Elixir Medical, STEL, and CONTROLRAD (spouse); Scientific Advisory Board for AMA, ACC (BOT Member), SCAI (Women in Innovations Committee Member), and JAMA Associate Editor; Faculty CRF (no fee). The other authors have nothing to disclose.