Translation in cardiovascular stents and occluders: From biostable to fully degradable

Abstract Cardiovascular disease is a major cause of morbidity and mortality, especially in developed countries. Most academic research efforts in cardiovascular disease management focus on pharmacological interventions, or are concerned with discovering new disease markers for diagnosis and monitoring. Nonpharmacological interventions with therapeutic devices, conversely, are driven largely by novel materials and device design. Examples of such devices include coronary stents, heart valves, ventricular assist devices, and occluders for septal defects. Until recently, development of such devices remained largely with medical device companies. We trace the materials evolution story in two of these devices (stents and occluders), while also highlighting academic contributions, including our own, to the evolution story. Specifically, it addresses not only our successes, but also the challenges facing the translatability of concepts generated via academic research.


| I N T R O D U C T I O N
Advances in nonpharmacological management of cardiovascular disease are largely driven by clever use of established biomaterials, or the development of new ones. When we talk of advances in management, we refer to patient benefits such as reducing the complexity of the procedures used, shorter healing periods and arresting the progression of disease. The complexities and morbidities associated with open heart surgery are well-known. Any advance that results in minimally invasive deployment is of substantial benefit to the patients and for the healthcare system as a whole. This is significant since management of cardiovascular disease is the primary healthcare cost for the developed world. Pharmacological intervention in post-Myocardial Infarction (post-MI) patients deals predominantly with blood-pressure management and anticlotting medication that ironically is necessitated by the presence of the blood-contacting device (stent, in this case); the search for the elusive "plaque-buster" continues but seems to have lost momentum in recent times. Thus, in cardiology, implanted devices dominate in disease management, and in this paper, we review advances in biomaterials-based design of implanted devices, more specifically stents and occluders, including highlighting our own contributions to design and performance enhancements. The underlying theme of our review is translatability of concepts, preferably as demonstrated in the clinic or in well-accepted animal models.
Both the design and the materials of implanted devices such as heart valves, coronary stents, and occluders have undergone dramatic changes driving the transition from surgical implantation to percutaneous deployment. This is a tremendous leap in the use of device technology, and confers obvious benefits to the patient: the procedure is minimally invasive, it is relatively painless and can be done as an outpatient procedure; and the subsequent healing period is drastically shortened. Both patient and healthcare system costs are thus dramatically reduced.
The revolution in management of MI started with the advent of balloon angioplasty in the late 1970s; this was made possible via the invention of the balloon catheter, by Dr. Gruntzig. 1,2 This catheter incorporated a remarkable segment (the "balloon") that could be expanded in situ at the stenosed section of the coronary artery by hydrostatic pressure, while guided by X-ray fluoroscopy to the site of blockage. Most importantly, the catheter was inserted through an artery in the leg or arm, and contained radio-opaque segments that made it visible in X-ray. Although balloon angioplasty replaced bypass surgery as the first option for patients with a single blocked artery, the procedure still led to re-blockage of the stenosed artery, resulting in reintervention rates of 40-50% that were unacceptable.
As one of the reasons for the restenosis was vessel recoil, the insertion of a slotted stainless-steel tube (stent) concurrent or subsequent to balloon angioplasty was approved in 1993. This concept of a coronary stent, envisaged by a cardiologist named Julio Palmas was so revolutionary that he could not find anyone willing to fund it, until a Texas restauranteur named Phil Romano put up some initial funding of $250,000. Following successful human trials this stent known as the Palmas stent was approved in 1994. 3 From such humble beginnings, the bare metal stent has undergone significant design and material changes to the present-day Co-Cr stent, which has a strut thickness of about 90 microns. 4 Ironically, while the bare metal stent was being deployed more and more widely, it was noted that restenosis rates were still around 20-30%; this restenosis was triggered largely by uninhibited growth of smooth muscle cells (SMCs) in response to the trauma caused by the stenting procedure.
This led to the invention of a drug-eluting stent (DES), a combination of pharmaceutical/device action, first developed by Johnson & Johnson and approved in 2003. 5 For a while, DES enjoyed the lion's share of the stent market until late-stage thrombotic events began to be noticed in long-term studies following deployment. These thrombosis events were traced to incomplete or delayed endothelialization of the stent, caused ironically by the use of drugs such as paclitaxel and sirolimus that did not discriminate between proliferating SMCs and ECs. Although such thrombotic events were rare, they were fatal. This led a few cardiologists to think about fully resorbable stents, as evidence mounted that stents were really needed only for about 6-9 months. 6,7 After this period, stents were not only redundant but also a liability, as their presence necessitated the use of anticlotting drugs and their presence was counter-indicated for certain procedures such as MRI imaging.
The pace of research progress in this area is astounding, and largely led by medical-device industry R&D up to the point where the first notion of fully resorbable stents was mooted. The fully absorbable stent concept arose in a medical school, and was developed by Japanese researchers, Dr. Igaki and Tamai, and eventually licensed to Kyoto Medical Planning. This nondrug eluting stent was first reported to have undergone human trials in 2000. 8 The Igaki-Tamai stent used a hydrolytically degradable polymer called Poly (L-lactide) which had been used as a suture material in surgery, and is known to degrade over a period of 12-24 months.
Our involvement in the fully degradable stent research started in 2003, with a grant from Singapore's Ministry of Education. Concurrently, we also worked on a metal stent coated with a biodegradable polymer that delivered two drugs, an antiproliferative (sirolimus) and antithrombotic (triflusal). 9 We describe our efforts to translate these concepts to the clinic in Sections 2.2.1 and 3.1.1. All abbreviations used are listed in Table 5.

| Background on drug types
The first generation of DES includes the sirolimus-eluting stent (Cypher®) and the paclitaxel-eluting stent (Taxus®), with both demonstrating impressive reductions in restenosis from 50 to 20/30% compared with bare metal stents (BMS). However, due to incomplete healing, the incidence of late stent thrombosis (LST) marred their longterm use, especially after discontinuation of dual antiplatelet therapy. 5 Paclitaxel is a lipophilic molecule with potent antiproliferative and antimigratory activities. The drug is a microtubule-stabilizing agent, which enhances formation of microtubular polymerized structures and thus, decreases the concentration of tubulin required for new microtubule formation. Paclitaxel affects primarily the M phase of the cell cycle inhibiting growth factor-induced DNA synthesis and cell proliferation, and leads to apoptosis or cell death. Concerns regarding cardiotoxicity of paclitaxel has been one of the reasons for its slow eclipse.
Currently, it appears that the limus-family of eluting stents dominates the market, as shown in Table 1 share an almost identical lipophilic chemical structure and bind to their major cytosolic FK-506 binding protein-12 (FKBP12), forming a complex which subsequently inhibits the mTOR. The major cellular effects include a decrease of the positive (blockage of the p70S6 kinase pathway of the cyclin-dependent kinases) and an increase of the negative (through inhibitor p27 kip1) regulators of the cell cycle 11 ; they stop the cell cycle at the G0/G1 phase inhibiting both cell (mainly smooth muscle cells) proliferation and migration, so the mechanism of action is cytostatic rather than cytotoxic. Tacrolimus and pimecrolimus are not analogs of the archetypal rapamycin; after they bind intracellularly to FKBP12, the complex in turn binds to and blocks calcineurin, and in this way inhibits the T-cell transduction pathways and the synthesis of pro-inflammatory cytokines. 12 In vitro cell work indicates that tacrolimus allows earlier endothelial regeneration than sirolimus; however, inhibitory activity on human vascular SMCs with tacrolimus is much less than sirolimus. 13 In this context, both everolimus-eluting stents (EES) and biolimus-eluting stents (BES) are the front-runners, and likely to be approved (or already approved) as noninferior products.

| Drug-eluting stents in research
Long term follow-up studies of DES have shown increased incidence of (sometimes fatal) stent thrombosis probably due to delayed endothelialization by the currently used drugs or delayed hypersensitivity reaction caused by the durable polymer; or by degraded polymer products currently used in DES. [14][15][16] Researchers have been hunting for selective drugs that inhibit SMCs without affecting ECs, as well as developing dual-drug-eluting stent (DDES). However, these efforts have not met with much success until recently.

| Stents eluting a selective antiproliferative peptide
Lack of selectivity in inhibition has been a major drawback of current synthetic drugs used in the DES. The presence of these immunosuppressive and cytotoxic drugs invariably also delays endothelial healing of the vessel, [17][18][19] which gives rise to LST. Researchers have been scouting for "selective drugs." Researchers at Mayo Clinic reported on a chimeric peptide termed CD-NP or cenderitide that has interesting vasodilating properties without renal effects. 20 Our collaborative efforts resulted in an observation that released CD-NP inhibits human coronary artery smooth muscle cells proliferation but did not hamper human umbilical vein endothelial cells proliferation in vitro, 21 which is not surprising given their endothelial cell origins. Based on this in vitro work, we designed a peptideeluting stent (or cenderitide-eluting stent, CES) that could deliver cenderitide in a controlled fashion from a Co-Cr stent coating matrix composed substantially of biodegradable poly(Ɛ-caprolactone) (PCL).
An array of slow, moderate and fast release profiles were attained from the addition of poly(ethylene glycol) and its copolymers in formulations for 30 days. 21 Following this, a 4-week pig study was carried out with a total of 32 stents implanted into 16 pigs. 22 The in vivo results demonstrated significantly higher plasma levels of CD-NP in the CES, and greater endothelial coverage of the stent struts compared with a control group (BMS and polymer-coated stent) at poststenting, as shown in Figure 1.
We believe that ensuring unperturbed endothelium regeneration is crucial to avert further progression of in-stent restenosis and LST.
However, the CES group was not able to show a significantly different outcome in terms on stenosis, when compared to BMS. This led us to believe that 2 lg/day over 28 days dosing of CD-NP used in this study was not sufficient to activate sufficient cGMP locally or systemically to have a significant effect on SMC proliferation. Further detailed studies, incorporating higher doses and frequent sampling, are therefore warranted.

| Fully bioresorbable stents in the market
It has been recognized for some time, that a fully resorbable stent is a preferred option for unblocking coronary arteries, deployed via percutaneous coronary intervention (PCI). The concept of a temporary multifunctional stent that enables revascularisation of the artery via a mechanical support like a metallic stent; prevention of restenosis by sustained-delivery of an antirestonotic like a DES; and allowing the restoration of normal vasomotion after resorption, has been considered the "holy-grail" of a fourth-generation stenting technology. 24 However, despite the many positive attributes that bioresorbable stents are expected to bring, there has been so far only two companies obtaining regulatory approval for such implants.
The world's first bioresorbable stent that found its way into humans was the implanted Igaki-Tamai stent at the turn of this century. 8 The stent was made with poly-L-lactide (PLLA) and expanded in the artery via a heated delivery balloon. The first-in-human (FIM) study involved a total of 25 such stents implanted in 15 patients. Together with a second trial, clinical results were promising, but usage was not extended to PCI due to the complexity involving the thermal balloon. 25 Although the stent was deployed percutaneously in the study, it needed to be heated to about 708C briefly, in order to overcome the natural recoil of the viscoelastic PLLA on balloon expansion. This problem in deployment got us interested in using shape memory concepts in deploying the sent. 26 The leader in the field of fully absorbable coronary stents is Abbott

| The amaranth story: An example of academic translation
Our early efforts in the area of bioabsorbable stents were funded by Singapore's Ministry of Education. Our concept of a percutaneously deployable fully bioabsorbable stent was based on self-expansion built into a multilayered construct. 29 The radial strength of these constructs was studied and reported. 30 Other studies included biodegradation, drug release from coatings, and the viscoelasticity and shape memory properties of the materials used, as shown in Figure 3.
Our initial focus was on ureteric stents, where the ureter was also subject to stenosis due to various causes; a fully bioabsorbable stent had to degrade in about 4-6 weeks in this case. 33  Further development, carried out by Amaranth led to several patented refinements including the process of making and sterilizing the stents. 29 The first-generation stent was named FORTITUDE TM . It exhibited excellent mechanical properties, comparable to or better than the Abbott BVS stent VISION TM , which was ahead in terms of clinical results ( Figure 4). As explained in a European PCR presentation in 2017, 35 the FORTITUDE stent used a specially made ultrahigh molar mass PLLA of very low crystallinity, which resulted in a tougher stent that showed very few balloon-expansion-related fractures.
In terms of degradation, the special PLLA polymer (custom-synthesized for Amaranth) behaves as shown in the Figure 5 below: The  The IDEAL BioStent by Xenogenics Corporation (Woonsocket, Rhode Island) is a sirolimus-eluting bioresorbable stent based on a polyanhydride ester mixed with salicylic acid. 25 The stent has both antiproliferative and anti-inflammatory properties attributable to its dual release capability for salicylic acid and sirolimus. An FIM study on an earlier version enrolled 11 patients and results were unsatisfactory due to excessive neointimal growth. 25 The current design is under development and in clinical trials, but not much information has been released thus far. Details of these are summarized in Table 2.

| Review of occlusion devices in the market
Atrial septal defect (ASD) and patent ductus arteriosus (PDA) are two major types of congenital heart diseases. ASD is defined as a persistent communication between the right and left atria. The incidence of ASD is about 1 in 2000 newborns 40 and this accounts for 7-10% of all congenital heart defects in adults. 41 Patients with unrepaired ASD can lead to   44 Nonetheless, serious complication of device erosion has been reported, with an estimated worldwide implant rate of 0.2-0.5%. 45 The mechanism is thought to be compliance mismatch between the left atrial disk wire mesh and the atrial wall.
The HSO device assumes a double disk configuration following deployment of a helical nitinol wire frame covered with expanded poly- devices are that the nitinol wire frame is coated with either titanium oxide or nitride; that is associated with less thrombosis, nickel ion elution, and improved endothelial tissue growth. 48 The occlusion devices for ASD in the market are summarized in Table 3. pletely occlude the ductus. 49 Besides Ginaturco coil, the Nit-Occlud PDA system is also utilized for closure of small to medium-sized PDAs.
It is similar in appearance with Ginaturco coil but has many more loops of varying diameters that stack to form a cone. Only the anterior loops of this device are positioned in the pulmonary artery while the remainder loops are generally placed in the aortic ampulla of the ductus.
For the closure of large PDAs, Amplatzer duct occluder (ADO) is the preferred choice. ADO is made of nitinol wire mesh that is configured into a cylindrical plug shape with a retention skirt to better anchor the device within the ductus and reduce embolization. Polyester fabric is sewn into the occluder to induce the thrombosis that closes the communication. In term of safety and efficacy, the successful deployment rate of ADO is 99.2% in a total of 390/393 patients with overall a 5.8% adverse event. Complete closure of the ductus was 98.4% at the 6-month interval and 98.6% at the 12-month interval. 50 A second generation of ADO, Amplatzer Duct Occluder II (ADO II), was granted FDA approval on 2013.
ADO II consists of fabric-free symmetrical multilayered nitinol mesh and dual articulating discs providing high conformability to treat the nonconical ducts, small infants with larger duct diameters while achieving compete closure from an aortic or pulmonary artery deployment approach. In a series of clinical trials, it demonstrates the safety and efficacy of the ADO II to occlude PDA with diameter that is 2 mm with successful deployment rate of 93.3% in a total of 56/60 pediatric patients. [51][52][53] The occlusion devices for PDA in the market are also summarized in Table 3.

| Occlusion devices in preclinical research
Device closure of appropriately indicated ASD and PDA is standard of care in most countries. All current commercial devices use a metallic frame and occlusive patch material. Metallic frameworks are recognized as the major cause of serious complications such as erosion,   54 The Chinese lantern occluder consists of soft portion ("head," "waist," and "tail" films) and structural skeleton (lock, and head tubes, and wires) with unique pull-fold mechanism to achieve folding and sealing. On retraction of the loop wire, the head films and tail films will fold into the working structure (Figure 7b) with the waist film length being adjustable corresponding to the septum thickness. In term of material, blend of PLC, PCL, and BaSO4 were utilized for the occluder fabrication. Preclinical study was also conducted in swine model, with satisfactory X-ray visibility. At the 1 month follow-up, there was no shunt from right atrium to the left atrium and complete endothelialization of the device. 60 Also, our group has developed a fully bioresorbable PDA occluder.
Build on the double umbrella concept, our PDA occluder consists of three parts: an anchoring arm, a stem and an umbrella with spokes with pull-fold mechanism in order to be repositionable and retrievable during the deployment (Figure 7c). The umbrella serves to cover the PDA and needs to recover as fast as possible postdeployment to ensure a fast delivery procedure. Therefore, PLC is used as the umbrella material for it is elastomeric and possesses good recovery property. The spokes on the umbrella serve as a structural support during deployment as blood pressure turbulence is experienced in the aortic side of the heart. Since both   to withstand high stress: therefore, the blend of PCL and BaSO4 is chosen as the material for it has the highest modulus among the blends. Furthermore, with the addition of BaSO4 into the stem and anchoring arm parts, the PDA occluder is radiopaque so that it can be easily monitored by fluoroscopy and accurately placed. 61 Our in vivo feasibility study demonstrated that the PDA device was able to recover within 2-3 min in vivo for immediate PDA closure and tissue overgrowth on the device at 1-month follow-up result, indicating the possibility of usage in human being. 62

| C ONC LUD I NG RE MARKS
It is now recognized that opening and closing of selected body vessels or defects is preferably done with a temporary scaffold rather than a biostable one, whose permanence may lead to serious side effects. Our translational efforts in two such fully bioabsorbable cardiovascular implants were seed-funded by academic grants; we believe these con-