Translation Pathway for Coronary Stent Development
Preclinical Evaluation of Coronary Stents
Laura E. Leigh Perkins, DVM, PhD, DACVP
From their humble beginnings as stainless-steel platforms to open occluded coronary arteries, coronary stents have evolved rapidly since their first clinical use nearly 30 years ago. Most significant in their history has been the approval of the first drug-eluting stent (DES) in the early 2000s, a landmark event that has allowed percutaneous coronary intervention (PCI) to become one of the most common medical procedures performed today (1). Today coronary stents, as compared to the earliest predecessors, have matured into a diversified group to include novel metal alloys, durable and bioresorbable polymer coatings, fully bioresorbable platforms (scaffolds), reservoir drug delivery, novel drugs and drug formulations (e.g., amorphous versus crystalline), combination drugs, and biologics, all with the aim of improving and advancing clinical safety and efficacy. Importantly, the many variables in stent design, from platform to polymer to drug, interplay and influence the biocompatibility, safety, and efficacy of a coronary stent.
With the ever-advancing variability and complexity in coronary stent design, calculated preclinical evaluation has become increasingly important to ensure that a device’s targeted performance and safety will translate to the clinical setting. The objective of this review is to outline the requirements and means for evaluation of the in vivo performance and safety of an implantable coronary device (stent or scaffold) destined for regulatory approval. (Editor’s note: While for simplicity referred to as stent or device throughout, this review is applicable to both permanent stents [bare-metal and drug-eluting] as well as to transient [bioresorbable] scaffolds implanted for the treatment of occlusive coronary artery disease.) The preclinical evaluation of coronary stents also can have alternate objectives, such as the assessment of a therapy for a specific efficacy endpoint and the evaluation of the pathogenesis of risks related to coronary stents (e.g., restenosis, thrombosis, and neoatherosclerosis). Studies designed to meet these alternate objectives often require the use of less conventional or specialized animal models, such as those with injury or endothelial denudation, dietary alteration, or induced or intrinsic disease processes and/or require the use of ex vivo models or alternative modalities of assessment. These preclinical studies are outside the scope of this review, and supplementary references regarding these types of studies are provided (2-8).
Under the guidance of the three R’s (replacement, reduction, refinement) for humane animal research, before initiating in vivo testing, there should be sound reasoning in place to ensure that a coronary stent will perform successfully. This assurance should be obtained through a composite of in vitro testing of the bioactive substance to identify specific cellular targets or pathways, in vitro release testing, biocompatibility testing of stent components (individually or en toto), in vitro degradation, and/or performance bench testing as applicable to the particular stent. With reasonable assurance being established that a stent can perform as intended, small in vivo proof-of-concept studies can be initiated for confirmation, followed by full-scale in vivo studies executed to fulfill the requirements necessary for regulatory approval. While ideally these in vivo studies would involve the final device intended for clinical use, any device changes that occur between preclinical testing and clinical use should be documented and justified. The following outlines the key considerations in the execution of these preclinical studies by addressing details regarding model selection, study objectives and design, and parameters for safety assessment.
There is no perfect animal model of human coronary artery disease. As such, there is no animal model that will perfectly predict the safety or efficacy of a device designed for the treatment of diseased human coronary arteries. But the accumulation preclinical research conducted over the history of coronary stents has built a solid foundation to show that the demonstration of safety in the preclinical setting provides a reasonable assurance of clinical safety. The demonstration of efficacy through preclinical models, on the other hand, remains elusive. For example, for DES designed with the intent of preventing restenosis through the suppression of neointimal hyperplasia, preclinical studies have shown that efficacy of such suppression may be observed early up to 28 days; however, at 90 days and longer time points, this effect is often no longer evident (3,9-12).
Several species have been used to assess coronary stents, including rodents, rabbits, swine, sheep, dogs, and non-human primates, with each offering advantages and disadvantages (2-5,13). Of these animal models, normal, nonatherosclerotic swine is the standard species used in the evaluation of coronary stents. Rabbits also are commonly used, though the inability to accurately assess morbidity or mortality related to coronary complications is a major deficit of this species (3,13).
Conventionally, domestic farm swine have been used for shorter-term studies (<6 months duration), and due to their rapid growth and large body weight potential (>400 kg), mini-swine, such as Yucatan, Sinclair, Göttingen, and Hanford strains, are used for longer-term studies. The cardiovascular anatomy of swine is similar to humans with regards to size, morphology, relative collateral arterial supply, and the presence of a well-developed vasa vasorum, and the general physiology and coagulation system of swine closely parallel that of humans (14,15). The coronary vasculature is suitable for the performance evaluation of the stent, guide wire, and catheter just as they are designed for clinical use, thus offering an environment of testing that directly simulates the clinical setting. Furthermore, the flow dynamics of porcine coronary arteries closely parallel that of humans, allowing for greater confidence in the accuracy of pharmacokinetic studies. With the muscular coronary arteries being the implant site as in the clinical setting, the neointimal response is of a similar histology to that of human coronary arteries, although the degree of restenosis observed in normal swine often is not of clinical significance.
A key consideration with regard to animal models is the rate of healing following implant of a coronary stent. While there are differences between species and arterial beds, overall preclinical models demonstrate healing faster than that of diseased, mature human coronary arteries. Estimations for healing of porcine versus human coronary arteries suggest a 1:6 ratio (1 month porcine equates to 6 months human) for bare-metal stents, though for DES this ratio of healing becomes more ambiguous due to the effects of the eluted drug (12). This disparity in healing rate is an important consideration in the translation of a stent’s performance from the preclinical to the clinical setting.
Study Objectives and Design
The core study objectives involved in preclinical evaluation of coronary stents include dosing, pharmacokinetics (PK), degradation, and safety. Studies should be executed to evaluate only one objective, though certainly outcomes of dosing, PK, and degradation studies, such as morbidity, mortality, clinical chemistry, and necropsy results, can and should play into the collective safety assessment. In the design of in vivo studies, simplicity is the best policy; trying to meet too many main objectives, fulfill too many requirements, evaluate too many study arms or time points can complicate the interpretation of final outcomes and may necessitate repeating a study.
Variability is an inherent factor of any biological system. In designing and executing a preclinical study for coronary stents, every attempt should be made to minimize the potential for variability, especially between test and control articles to ensure accurate ‘head-to-head’ comparisons. Means by which such variability can be minimized include: selecting a uniform population of test systems; having well-defined and routinely executed implant procedures; evaluating only a finite number of test and control articles per study; ensuring uniformity in size of test and control articles; maintaining detailed specifications regarding appropriate anatomical sites; and randomizing test and control articles within each animal. This latter aspect of randomization is especially important for studies conducted in swine, a species known to be prone to developing inflammation that can follow a host-dependent (as opposed to a device-dependent) distribution (3,10,16,17).
Studies should be designed to mimic the clinical applications, utilizing similar techniques for implantation, similar device configurations (e.g., single and overlapping configurations, bifurcations), and similar regimens for antiplatelet therapy. This is particularly true for safety studies in swine to allow for assessment of potential coronary complications; dosing, PK, and degradation studies can utilize both coronary and internal thoracic arteries to reduce the total number of animals required to meet appropriate sample numbers. Implantation of coronary stents using 10% overstretch (1.1:1 balloon:artery ratio) helps to ensure appropriate apposition and embedment of the struts into the arterial wall without inducing excessive injury. Due to anatomical limitations, including tapering and side branches, it can be challenging to assess longer length devices in porcine coronary arteries with ease and consistency. However, shorter lengths that are representative of longer lengths based on drug dose density and/or overlapping device studies may be able to be leveraged to justify safety of longer-length stents.
Fundamental to development of a stent that delivers bioactive agents is determining the dose range from sub-therapeutic to toxic levels. In vivo dose-ranging studies that assess both PK and safety should be performed to select the optimal (i.e., minimum efficacious) dose for clinical use. The high dose is useful in determining a safety margin, for in clinical practice a patient can receive multiple longer length stents at one time, and it is necessary to establish the maximum drug dose allowable for clinical use. It is recommended to include stents with the maximum drug (and polymer) load possible (three to 10 times over the drug dose density of the intended final product) and/or overlapping stents to adequately establish the safety margin.
In Vivo Pharmacokinetics
While DES are intended to deliver drug locally to the treatment site, local, regional, and systemic drug deliveries need to be considered and evaluated in vivo. In the evaluation of in vivo PK, a minimum of five time points should be included to determine the half-life (when half of original amount of drug remains on the stent) and the time to complete elution of drug from the stent. In vitro release kinetics can provide valuable information in the appropriate determination of these time points for in vivo sampling and developing in vitro-in vivo correlation curves can prove useful in future device iterations as an alternative to additional in vivo testing. Typically, a sample range of three to six stents per time point should be included. In addition to evaluating the release kinetics, drug tissue levels also should be determined, including the implanted segment, proximal and distal arterial segments, subjacent and/or downstream myocardium, blood, and organ levels (e.g., lung, liver, kidney, and spleen). It is important to ensure that the time points included when assessing drug tissue levels are inclusive of acute elution (hours), the time of expected peak tissue concentrations, and the time when drug has been mostly to fully eluted. Special care and consideration may be required for PK evaluation of stents with bioresorbable coatings, which can delaminate during tissue removal and yield inaccurate results.
In Vivo Degradation
In vivo degradation studies for coronary stents having biodegradable/biocorrodible components are intended to characterize the degradation profile. This degradation profile should then be aligned to the assessment of the device’s safety and performance. These studies should characterize the material content over the full duration of the device’s indwelling from implant (t0) to complete resorption, which can be highly variable: from as little as 1 month for bioresorbable polymer-coated DES to prolonged duration for polymeric bioresorbable scaffolds (>2 years). Depending on the total duration to complete resorption, similar to PK, at least three to five interim time points should be selected for determining the amount of remaining material. Of these time points, inclusion of the time at which the resorption is expected to be at its maximum rate is recommended so as to align to the safety assessment for vascular compatibility. In vitro degradation studies can prove useful in determining these time points prior to execution of an in vivo study. Ideally, a sample size of three to six implanted arteries should be assessed at each time point.
In Vivo Safety
For the demonstration of in vivo safety, the potential risks related to a coronary stent need to be considered and included as endpoints in the evaluation. This includes the standard risks associated with coronary stents, and for drug-eluting and stents with biodegradable components, risks related to the drug pharmacokinetics and the material degradation. Thus, the PK and degradation profiles should weigh into the selection of the optimum time points for safety evaluation. While the standard recommended time points for safety evaluation of coronary stents have included acute (3 or 7 days), 1 and 6 months, and 1 year (17,18), the performance characteristics, known risks, and predicated outcomes of the device should ultimately be the determining factors in the selection of the assessment time points. A time point congruent with the time of peak arterial tissue concentration of drug and a time during the peak rate of material resorption/corrosion, for example, are requisite time points to include for an adequate assessment of safety. As these critical time points vary between coronary stents, additional time points may therefore be required as alternates to and/or in addition to those that are the standard recommendations. Further, for bioresorbable devices that may have a prolonged duration of degradation, in vivo safety should be assessed at time points throughout this degradation process to near to complete resorption. Overall, the selection of time points for safety evaluation should be rationalized and tailored to the coronary stent.
Typically for safety evaluation, 10 to 12 test articles per time point is sufficient to ensure an adequate sample size for statistical comparison to the control article. The control article, whether bare-metal, polymer-only, or an already approved stent, should be included in each animal to account for potential host-dependent variability in the vascular response. The selection of a control article should be thoroughly considered and justified. In the porcine model, the implant of study articles in two to three of the main coronary arteries in six to eight animals, each with one control article and one or two test articles per time point is often suitable to ensure sufficient sample size for statistical comparison. However, whether rabbits or swine, mortality should be accounted for by including additional animals, especially for longer-term studies.
Table 1 outlines both the fundamental safety assessments included in routine safety evaluation of coronary stents as well as supplemental assessments that may prove valuable for further elucidating device performance, the vascular response, and potential clinical correlation. As outlined in Table 1, the recommended parameters for safety assessment can be divided into the general categories of systemic endpoints, imaging endpoints, and endpoints related to the vascular response.
Table 1. Standard Assessments for In Vivo Safety
Device Deployment and Acute Performance
|Systemic||Imaging (in vivo and ex vivo)||Vascular Response|
Over the course of a study, animals should be monitored daily for any possible changes in health status. Weighing and body condition scoring should be done intermittently over the study course to ensure that animals are maintaining or gaining weight as expected and that the health condition is maintained. Clinical pathology, including complete blood count and serum chemistry, should be performed at implant and termination to confirm that the animal maintained health and that the device, especially if drug eluting, has had no adverse effect on these clinical metrics.
While detailed here with regards to safety evaluation, foremost for any in vivo study is the inclusion of comprehensive necropsy of all animals enrolled in a study. This includes animals that survive to the time of scheduled termination as well as any that die or are euthanized prematurely; deaths that occur within 24 hours of implantation, often considered to be “procedural,” are no exception. Necropsy should be performed by qualified personnel to ensure thorough evaluation and collection of tissues with aberrations from normal.
The tissues routinely included for histological evaluation are the heart and implanted sites, lung, liver, and kidneys; spleen, brain, lymph nodes (thoracic), thymus, and bone marrow are also worthy of consideration, especially if safety studies are intended to be leveraged to satisfy chronic systemic biocompatibility requirements. Also to be included are any tissues in which gross lesions are identified. For coronary stents, care should be taken for complete assessment of any downstream effects. Thus, hearts should be perfusion flushed and fixed following excision. Once the heart is fully fixed, the implanted arteries can be removed, and the myocardium should be fully evaluated grossly by bread loafing at approximately 1 cm intervals for any gross evidence of thrombo-embolization. Any gross lesions as well as a well-distributed sampling of myocardium subjacent and distal to each implanted artery should be collected for histological evaluation.
Imaging of implanted arteries allows for the evaluation of in vivo performance and for drawing clinical relevance; angiography is the standard modality used for this assessment. Angiography allows for device evaluation during implant (e.g., placement, migration, and dissection); it should also be used at termination to assess changes that may have occurred in the stent or vessel response over time. Baseline and post-implant metrics of the reference vessel and implanted segment should be collected for comparison to those obtained at interim time points and/or termination.
While not requisite assessments, other imaging modalities can provide valuable insight into the in vivo performance of coronary stents and supplement the correlation to the clinical setting. Cardiac computed tomography (CT) angiography, magnetic resonance imaging, optical coherence tomography (OCT) and 3-dimensional OCT, and intravascular ultrasound (IVUS) and IVUS-virtual histology can complement safety assessments, optimize the use of research animals, and allow for more in-depth dimensional and functional assessments, such as vessel remodeling, pulsatility, and vasomotility when compared across time points. For radiolucent polymeric implants, OCT and IVUS have shown value for obtaining more accurate measurement of vessel dimensions and for the assessment of changes to the tissue and implant over time as compared to angiography (19-21). However, OCT and IVUS may impact the neointima over struts at explant (e.g., endothelial denudation) and therefore should not be performed on arteries designated for evaluation by scanning electron microscopy post-mortem.
Following euthanasia and both prior to and following excision of the implanted arteries, high-resolution radiography should be performed to confirm device placement, stent and arterial contours, and stent integrity, particularly for stents composed of or containing radiopaque materials. Polymeric devices are not radiopaque, and microCT can may serve as an alternate modality to evaluate the implant in the excised coronary arteries following validation.
The third realm for determining safety of a coronary stent is the evaluation of the implanted arteries by histology and scanning electronic microscopy (SEM). For histology, as a majority of coronary stents tend to be metallic, implanted arteries are often embedded in polymer resins (e.g., methyl methacrylate, glycol methacrylate), for histological evaluation (22). Polymeric scaffolds can be embedded in paraffin for sectioning, though polymer resins tend to allow for better maintenance of tissue integrity. Thin sections (2 to 5 µm) can usually be obtained with most coronary stent materials, though harder metals such as nitinol require grinding, thus resulting in thicker histological sections (15 to 50 µm). Standard histological stains for safety evaluation include hematoxylin and eosin and Movats pentachrome; with careful consideration, additional histochemical and immunohistochemical can be performed for supplemental assessments of the tissue response.
Thorough histological evaluation for safety of coronary stents includes both histomorphometry and histomorphology. Examples of parameters routinely assessed by histology, inclusive of both histomorphometry and histomorphology, are provided in Table 1. Histomorphometry is obtained through computerized planimetry and includes quantitative, continuous variables such as the areas of the external and internal elastic lamina and lumen. Conversely, histomorphology is semi-quantitative to qualitative and overall is more subjective in its evaluation. Importantly, attempts should not only be made to quantify or categorize the vascular response, but qualitative assessments should also be reported to provide insight into the pathogenesis underlying potentially adverse observations. In particular, this includes not only quantification of the amount of leukocytic infiltrates observed, but also a detailed descriptive of the types of cells, location (e.g., intimal, subintimal, circumstrut, adventitial, and transmural), and nature of cellular infiltrates. In the porcine model, strut-centered granulomatous responses with eosinophils are sporadically observed and do not necessarily indicate an adverse response specific for the implanted device (3,10,16,17).
SEM is used to characterize the surface characteristics of the implanted stent and to assess the extent and maturity of endothelial coverage and attachment of thrombus, platelets, and leukocytes to the lumen surface. Along with histology, evaluation of arteries implanted with test and control articles by SEM is recommended at any time point up to 6 to 12 months to ensure suitable and stable endothelial coverage and maturation. A sample size of at least two implanted arteries is sufficient to provide qualitative assessment.
In addition to the overview of recommendations provided herein, there are additional guidance and consensus documents that can serve as references in designing a preclinical program for the evaluation of coronary stents intended for regulatory approval (17,23-25), Importantly, the dynamic nature and multidimensional approaches of coronary stents in the treatment of occlusive coronary artery disease oblige tailoring the preclinical evaluation to suit the individual stent and its clinical application.
The preclinical evaluation of coronary stents provides valuable insight into a stent’s performance in vivo and can thus provide a reasonable assurance of safety in the clinical setting. However, the standard preclinical models applicable to studies intended for regulatory approval are not meant to determine a stent’s efficacy. The best estimation of efficacy comes through the use of specialized models and modalities of assessment, and even then, it is still just an estimation of clinical performance. Even with the amassment of preclinical data accrued on coronary stents, we continue to learn and gain better understanding of models, study design, and optimum assessments for drawing clinical relevance.
A shortcoming in the extensive research conducted to date in animal models is the lack of standardization in the means and parameters of assessment. The ability to standardize preclinical evaluation becomes even more complicated considering the rapid rate of evolution of coronary stent technologies and the rapid evolution of the technologies applicable to the assessment of in vivo performance. However, the ability to standardize will enhance our ability to draw direct correlations between preclinical observations and clinical outcomes, which potentially will allow for the design of preclinical studies that ultimately can enhance our ability to predict clinical safety and efficacy while enhancing our ability to follow the three R’s of humane animal research. This may be a pipe dream, but one we should pursue nonetheless.
- Stefanini GG, Holmes Drug-Eluting Coronary-Artery Stents. N Engl J Med. 2013;368:254-5.
- Suzuki Y, Yeung AC, Ikeno F. The pre-clinical animal model in the translational research of interventional cardiology. JACC Cardiovasc Interv. 2009;2:373-83.
- Perkins LEL. Preclinical models of restenosis and their application in the evaluation of drug-eluting stent systems. Vet Pathol. 2010;47:58-76.
- Touchard AG, Schwartz RS. Preclinical restenosis models: challenges and successes. Toxicol Pathol. 2006;34:11-8.
- Jackson CL. Animal models of restenosis. Trends Cardiovasc Med. 1994;4:122-130.
- Joner M, Nakazawa G, Finn AV, et al. Endothelial cell recovery between comparator polymer-based drug-eluting stents. J Am Coll Cardiol. 2008;52:333-42.
- Otsuka F, Cheng Q, Yahagi K, et al. Acute thrombogenicity of a durable polymer everolimus-eluting stent relative to contemporary drug-eluting stents with biodegradable polymer coatings assessed ex vivo in a swine shunt model. JACC Cardiovasc Interv. 2015;8:1248-60.
- Granada JF, Kaluza GL, Wilensky RL, Biedermann BC, Schwartz RS, Falk E. Porcine models of coronary atherosclerosis and vulnerable plaque for imaging and interventional research. 2009;5:140-8.
- Schwartz RS, Chronos NA, Virmani R. Preclinical restenosis models and drug-eluting stents: still important, still much to learn. J Am Coll Cardiol. 2004;44:1373-85.
- Perkins LEL, Boeke-Purkis K, Wang Q, Stringer SK, Coleman LA. XIENCE VTM everolimus-eluting coronary stent system: a preclinical assessment. J Interv Cardiol. 2009;22:S28-S40.
- Wilson GJ, Polovick JE, Huibregtse BA, Poff BC. Overlapping paclitaxel-eluting stents: long-term effects in a porcine coronary artery model. Cardiovasc Res. 2007;76:361-72.
- Virmani R, Kolodgie FD, Farb A, Lafont A. Drug eluting stents: are human and animal studies comparable? 2003;89:133-8.
- Nakazawa G, Finn AV, Ladich E, et al. Drug-eluting stent safety: findings from preclinical studies. Expert Rev Cardiovasc Ther. 2008;6:1379-91.
- Lowe HC, Schwartz RS, Mac Neill BD, et al. The porcine coronary model of in-stent restenosis: current status in the era of drug-eluting stents. Catheter Cardiovasc Interv. 2003;60:515-23.
- Gross DR. Thromboembolic phenomena and the use of the pig as an appropriate animal model for research on cardiovascular devices. Int J Artif Organs. 1997;20:195-203.
- Wilson GJ, Nakazawa G, Schwartz RS, et al. Comparison of inflammatory response after implantation of sirolimus- and paclitaxel-eluting stents in porcine coronary arteries. 2009;120:141-9, 1-2.
- Schwartz RS, Edelman E, Virmani R et al. Drug-eluting stents in preclinical studies: updated consensus recommendations for preclinical evaluation. Circ Cardiovasc Interv. 2008;1:143-53.
- S. Food and Drug Administration. Draft Guidance for Industry: Coronary drug-eluting stents – nonclinical and clinical studies. 2008. Available at: https://www.fda.gov/media/71521/download. Accessed May 4, 2019.
- Nakatani S, Ishibashi Y, Sotomi Y, et al. Bioresorption and Vessel Wall Integration of a Fully Bioresorbable Polymeric Everolimus-Eluting Scaffold: Optical Coherence Tomography, Intravascular Ultrasound, and Histological Study in a Porcine Model With 4-Year Follow-Up. JACC Cardiovasc Interv. 2016;9:838-51.
- Gutierrez-Chico JL, Serruys PW, Girasis C, et al. Quantitative multi-modality imaging analysis of a fully bioresorbable stent: a head-to-head comparison between QCA, IVUS and OCT. Int J Cardiovasc Imaging. 2012;28:467-78.
- Slottow TLP, Pakala R, Okabe T, et al. Optical coherence tomography and intravascular ultrasound imaging of bioabsorbable magnesium stent degradation in porcine coronary arteries. Cardiovasc Revasc Med. 2008;9:248-54.
- Rousselle S, Wicks J. Preparation of Medical Devices for Evaluation. Toxicol Pathol. 2008;36:81-84.
- Schwartz RS, Edelman ER, Carter A et al. Drug-eluting stents in preclinical studies: recommended evaluation from a consensus group. 2002;106:1867-73.
- U.S. Food and Drug Administration. General Considerations for Animal Studies for Medical Devices: Draft Guidance for Industry and Food and Drug Administration Staff. 2015. Available at: https://www.fda.gov/media/93963/download. Accessed May 4, 2019.
- U.S. Food and Drug Administration. Guidance for Industry and FDA Staff: General Considerations for Animal Studies for Cardiovascular Devices. 2010. Available at: Available at: https://www.fda.gov/media/79366/download. Accessed May 4, 2019.