Stent design matters: insights from virtual bench testing

Bare-metal stent design influences clinical outcome

Stent design matters when using bare-metal stents as it affects both immediate and long-term clinical outcome. This clear conclusion was drawn by a randomized clinical study that compared five different uncoated stent designs, all having the same material composition [1]. In this study, the event-free survival rate for the included stent types ranged from 69.4 to 82.4% at 1-year follow-up. However, besides the significant impact of different stent strut patterns, more subtle differences have also been identified as important factors. For example, the Intracoronary Stenting and Angiographic Results: Strut Thickness Effect on Restenosis Outcome (ISAR-STEREO) trial evaluated two stents with a very similar design and composed of the same material, but the strut thicknesses of the two devices were 50 and 140 µm [2]. The thin-strut design led to a significantly lower re-intervention rate of 8.6% compared with 13.8% for the thick-strut group.

Drug-eluting stent design: who cares?

Stent design received less attention after the introduction of drug-eluting coatings, partially because of the fact that the involvement of multiple components (coating, drug and design) complicated the analysis of one individual factor. However, more importantly, these drug-eluting stents lowered restenosis rates in the single-digit range, with small differences between the available products [3], suggesting a dominant impact of the drug. However, this observation should in no way reduce the efforts to continue the investigation and optimization of the current drug-eluting stent platforms because of multiple motivations. The design determines the arterial injury [4], for example, caused by mechanical factors such as dogboning and foreshortening, and although the drug may suppress any excessive tissue response, the first aim should be to minimize the injury. In addition, the clinical outcome can still be further improved in terms of efficacy and safety. It is likely that this goal will be achieved through further optimization of both the drug delivery system and the underlying stent platform (i.e., design and material).

Virtual bench testing

A promising approach to study and optimize stents and medical devices in general is the use of simulation methods, which can be divided into two major categories: computational fluid dynamics and finite element analysis. Computational fluid dynamics research typically focuses on the interaction of a medical device or artificial organ with blood [5], while finite element analysis provides information about deformations and stresses within a material. In the case of stents, finite element analysis can, for example, be used to investigate stent and artery deformations and resulting vessel wall stresses. This research domain, which could be defined as virtual bench testing, has undergone a thorough transformation during the last few years thanks to increased computer power and innovative simulation strategies. The initial virtual bench testing studies were primarily focused on obtaining realistic simulation results for stent deployments, but nowadays simulations can be used to accurately predict stent deformations occurring during complex bifurcation stenting techniques. Therefore, we could say that virtual bench testing has changed from a purely engineering tool to a broader research method that may provide useful insights in clinical practice.

Recently, we compared the mechanical behavior of three second-generation drug-eluting stents using computer simulations [6]. The three stents, Cypher Select^® (Cordis, NJ, USA), Endeavor^® (Medtronic, MN, USA) and Taxus^® Liberte^™ (Boston Scientific, MA, USA), were virtually implanted in the curved main branch of a patient-specific bifurcation model, with considerable vessel straightening being observed for all stents. The vessel wall stress patterns caused by the different stents were very similar: high stresses in the regions with a smaller initial lumen and at the stent ends. These high stresses, which can be interpreted as a degree of arterial injury, are logical and inevitable in the central stent region where it is the aim to enlarge the vessel lumen. The observed stress peaks at the stent ends are a result of the abrupt change in longitudinal and circumferential stiffness from the stented to the unstented vessel. These ‘edge stress peaks’, in combination with a reduced vessel scaffolding and drug delivery at these locations, may explain the frequent incidence of edge dissections and stent edge restenosis [7,8]. Future-generation drug-eluting stents should therefore have a platform that avoids such stress peaks near the stent ends. As illustrated in [6], virtual bench testing may be used for this purpose, as it allows for the investigation of different design variants without having to manufacture every intermediate design. When applied with sound engineering judgment, virtual bench testing can therefore reduce costs, timescales and risks associated with the development of a new design.

Future perspective

Much progress has been made in the field of virtual bench testing in recent years, but a better integration of virtual bench testing with in vivo studies is crucial to increase the credibility of the simulations and to relate the observed mechanical performance with clinical outcomes. In addition, continued efforts to improve the efficiency and level of realism are required in order to increase the impact of this tool, both on medical device development and on current clinical practice. For example, the bifurcation model we used is based on patient-specific angiographic data. Using angiographic data as input allows the generation of a 3D computer model with realistic vessel curvatures, but on the other hand, requires assumptions regarding the vessel wall thickness and composition because of the absence of tissue information. Enhanced geometric models of coronary arteries could be obtained by combining angiographic information with, for example, intravascular ultrasound data [9]. Furthermore, the composition of the vessel wall and some information regarding the local tissue elasticity could be derived from clinical images by using virtual histology, intravascular palpography or other available techniques. The current study only considered one bifurcation model, but including multiple bifurcations in these virtual studies is necessary to compare different stents using common statistical methods. This will, however, require automation of the different steps involved. In particular, the generation of the arterial models starting from patient-specific images is currently quite time consuming and should be accelerated.

Although not fully justified, the introduction of drug-eluting coatings forced stent design into a secondary role. However, design will become more important than ever before during the biodegradable stent era, also called the fourth revolution in interventional cardiology. The polymeric and metallic degradable materials do not have the same mechanical strength compared with the currently used stainless steel or cobalt chromium alloys. As a result, innovative stent designs are required in order to obtain an acceptable mechanical behavior in terms of recoil and radial strength. However, besides the adequate initial mechanical properties, the evolution of the design should also be considered. For example, how will the partially degraded stent structure resist the remaining compressive forces exerted by the vessel wall? Virtual bench testing is perfectly suited to address such mechanical questions, and therefore it seems likely that the fourth revolution in interventional cardiology will also lead to a more prominent role for virtual bench testing.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.