Drug-eluting stents (DES) have revolutionized the field of interventional cardiology. The US Food and Drug Administration (FDA) is expecting that the total number of DES implanted in 2010 will surpass 5 million worldwide. Although initial results seemed promising, longer-term follow-up in a wider range of patients revealed some concerns. Enhanced platelet aggregation, delayed neointimal growth, a local hypersensitivity reaction against the polymer coating, and a failure of sirolimus-, paclitaxel-, and tacrolimus-eluting stents to reduce neointimal hyperplasia at 90 and 180 days in animals when the drug was completely eluted from the stent are the main concerns.1–8
A number of stents have been and are currently being tested in animal and human trials. They are coated with drugs that aim to inhibit inflammation and neointimal proliferation. However, the process of restenosis is a sequence of complex events that has been partly tested and understood over the last two decades.9 Locally released drug interferes with the various mechanisms responsible for each step in the restenotic cascade, and a wide variety of agents are currently available. Although only everolimus-eluting stents (EES), sirolimus-eluting stents (SES), and paclitaxel-eluting stents (PES) have received FDA approval to date, multiple alternative devices are attempting to find a way to achieve the same goal, namely the reduction of re-stenosis and the need for repeat interventions.
Existing and Under Investigation Drugs
Six Limus-family-related drugs are currently being studied in DES, namely sirolimus, everolimus, biolimus A9, zotarolimus, tacrolimus, and pimecrolimus. Sirolimus, everolimus, biolimus A9, and zotarolimus all bind to the FKBP12-binding protein, which subsequently binds to the mammalian target of rapamycin (mTOR) and thereby blocks the cell cycle, mainly of the smooth-muscle cell from the G1 to S phase. The mechanisms of action of tacrolimus and pimecrolimus are different. Both drugs bind to FKBP506. The tacrolimus/pimecrolimus FKBP506 complex subsequently inhibits the calcineurin receptor, which leads to decreased cytokine expression on the cell surface membrane and results in an inhibition of T-cell activation and lower smooth-muscle cell selectivity (see Figures 1 and 2). A non-Limus-family-related drug widely studied for its efficacy in coronary stents is paclitaxel. Its effect has been mainly explained by its ability to stabilize microtubules and thereby inhibit cell division in the G0/G1 and G2/M phases.
Limus Family
Sirolimus
The first of the Limus family drugs used on endovascular prostheses was sirolimus, a natural macrocyclic lactone that is able to inhibit mTOR.10,11 Sirolimus proved to have potent antiproliferative and immunosuppressive effects. Several successive studies proved the efficacy of the SES: the Randomized study with the Sirolimus-eluting IL = Velocity balloon-expandable stent in the treatment of patients with de novo native coronary artery Lesions (RAVEL); SirolImus-coated Bx Velocity balloon-expandable stent in the treatment of patients with de novo coronary artery lesions (SIRIUS), Canadian [C]-SIRIUS, and European [E]-SIRIUS.12–18 As a result of the polymer, 75% of the drug is slowly released over the first 10 days. Nevertheless, the antirestenotic properties of the SES proved to persist for much longer.19 As there was no significant change in neointimal thickening between two and four years in the first-in- man (FIM) trial, and given the continued clinical superiority of SES after four years in a pooled analysis of the four pivotal randomized Cypher trials (RAVEL, SIRIUS, C-SIRIUS, and E-SIRIUS), it seems reasonable to rule out a late catch-up in restenosis—at least so far, because both the clinical and angiographic end-points continue to slowly accrue over time.14–20
Everolimus
A second member of the Limus family is everolimus, a sirolimus analog with a single minimal alteration in its molecular structure (position 40), without a chemical modification of the mTOR-binding domain21 (see Figures 1 and 2). Of interest is that, when implanted in rabbit iliac arteries, a more rapid endothelialization was observed in the everolimus-eluting stent than with sirolimus-, zotarolimus-, or paclitaxel-eluting stents, demonstrated by a complete endothelialization of the struts with exhibition of cd31 (antigen surface marker of good endothelial functionality) in the cells at 14 days (R Virmani, MD, unpublished data, 2006).
The Clinical Evaluation of the XIENCE™ V Everolimus Eluting Coronary Stent System in the Treatment of Patients with de novo Native Coronary Artery Lesions First (SPIRIT) trial proved the superiority of everolimus embedded in a durable polymer on a cobalt–chromium stent compared with bare-metal stents (BMS).22,23 In the recently completed SPIRIT-II trial, the everolimus-eluting XIENCE V stent proved to be superior to the PES for reduction of both late loss and binary restenosis.24
Subsequently, the SPIRIT-III trial randomized 1,002 patients in the US to treatment with either an everolimus XIENCE V stent or a PES. As part of the SPIRIT-III study, additional patients will also be enrolled in four registry arms in Japan, one each for stents that are 38, 2.25, and 4mm long. Additionally, the SPIRIT-IV and SPIRIT-V studies will provide further data.
Zotarolimus
A third descendant of the Limus family that also has a change on position 40 and that is used on coronary stents is zotarolimus (ABT- 578, Abbott Pharmaceuticals, Abbott Park, Ill), which likewise has antiproliferative and anti-inflammatory effects, but zotarolimus is suggested to have higher tissue retention than the SES (see Figures 1 and 2). Of note, data on endothelial function after stent placement in porcine coronaries showed a normally functioning endothelium one and three months after zotarolimus-eluting stent implantation, whereas a dysfunctional endothelium was observed after both Cypher and Taxus implantation.25
In the Randomized Comparison of Zotarolimus- and Paclitaxel-eluting Stents in Patients With Coronary Artery Disease (ENDEAVOR) I and II trials, the phosphorylcholine polymer-based cobalt-alloy Driver coronary stent (Medtronic Vascular, Santa Rosa, California) loaded with zotarolimus proved to be superior to BMS in both angiographic and clinical end-points.26,27 Recently presented two- and three-year follow-up data of the ENDEAVOR I and II trials proved sustained superiority in the reduction of target lesion revascularization (TLR), with remarkably low rates of total stent thrombosis (0.3%) and no cases of stent thrombosis after 30 days. The ENDEAVOR III trial was a prospective randomized comparison of the Endeavor zotarolimus-eluting stent and the SES (n=436). At eight months, the Endeavor stent failed to meet its non-inferiority end-point in terms of late lumen loss. Of note, the rates of death, myocardial infarction (MI), and target vessel revascularization were equal in both groups.28 A possible explanation for the lack of non-inferiority of the Endeavor stent might be the rate of elution. The Cypher stent elutes 75% of its drug within the first 10 days; in the Endeavor stent this took only two days.
Another zotarolimus-eluting device is the Zomaxx TriMaxx™ stent (Abbott Pharmaceuticals, Abbott Park, Illinois), which has a trilayer pharmacoat that consists of a phosphorylcholine basecoat and topcoat wrapped around a zotarolimus laye with elution rates similar to the Cypher stent. The platform used was a stainless steel/ tantalum/stainless steel triplex stent. Although the four-month results of the Zomaxx-intravascular ultrasound (IVUS) trial (n=40), which showed a late loss of 0.20mm, were promising, its manufacturer Abbott recently announced it would discontinue the Zomaxx program after the disappointing results of the Zomaxx-I trial. At nine months, the Zomaxx stent was associated with a significantly higher late loss and binary restenosis rate compared with the Taxus stent.29
Paclitaxel
Although not a member of the Limus family, the PES (Taxus, Boston Scientific, Natick, Mass) was the second DES to receive FDA approval one year after the SES. Paclitaxel was first found by The National Cancer Institute (NCI) in a search for naturally occurring agents with strong antiproliferative qualities. Paclitaxel stabilizes microtubules and thereby inhibits cell division in the G0/G1 and G2/M phases (see Figure 1). The randomized TAXUS-I trial (2003) was designed as a FIM phase I feasibility study and proved that a polymer-coated PES was superior to BMS at six- and 12-month follow-up.30 Thereafter, the TAXUS family trials expanded with the II, IV, V, and VI trials and confirmed the superiority of PES compared with BMS in more complex patients and lesions.36–39 Recently, the TAXUS-V-ISR (in-stent restenosis) trial compared the efficacy of a slow-release polymer-based PES with brachytherapy for in-stent restenotic lesions. At nine months, the use of PES was associated with lower rates of clinical and angiographic restenosis and improved event-free survival.40 The TAXUS clinical trial program, which assessed the TAXUS Express stent system from single to complex lesions, was followed by the TAXUS ATLAS and Olympia programs, which transferred the established polymer drug combination to a new stent platform, the Liberté stent.41,42
Another new device coated with paclitaxel is the Asian Infinnium (Sahajanand Medical Technologies, Gujarat, India) stent. The stent has a biodegradable hemocompatible polymer coating and lower strut thickness (0.084mm compared with 0.14mm for Cypher) designed to reduce vessel trauma. The coating consists of slow-, medium-, and fast-release polymer layers. The multicenter open-label registry Safety and Efficacy of Infinnium: A Paclitaxel-Eluting Stent (SIMPLE-I) trial (n=282) was the first to test the efficacy of this new device. The SIMPLE-I trial was followed by the multicenter, single-arm, prospective SIMPLE-II trial (n=103) to further investigate the safety and efficacy of the Infinnium stent.43 Certainly the late loss at six months was low (0.38mm) and the binary restenosis rate was 7.3%; this helped this device to be the first designed and evaluated ‘low-cost’ DES from Asia to receive CE mark approval.
New Polymer
After disappointing results with the use of carbon-, platinum-, and gold-coated stents, the polymer was hypothesized to be an appealing alternative carrier to reduce restenosis and thrombosis and to guarantee controlled drug-release kinetics.44 Soon, the first-generation polymer-coated SES and PES proved to be more effective than their non-polymer-coated counterparts.45–48 Nevertheless, a major limitation is that many polymer coatings are not entirely inert, and hypersensitivity reactions against the polymer have been frequently reported.1,3,49 In line with these data, long-term adverse effects such as increased inflammation of the vessel wall, a thrombogenic response, and induced apoptosis of smooth-muscle cells have been described.49–51
To reduce the inflammatory reaction, which is partially caused by the polymer, Medtronic recently developed a novel co-polymer, the ‘Endeavor Resolute,’ for the extended release of zotarolimus in a next-generation DES. The new BioLinx polymer system contains a C10 polymer, which is lipophilic/hydrophobic and stimulates a controlled drug release, a C19 polymer, which is primarily hydrophilic and thus more biocompatible and helpful in drug elution, and finally polyvinyl pyrrolidone, which is hydrophilic, increases the initial drug burst, and enhances the elution rate. Porcine coronary implants (n=25) showed no difference in inflammatory response after implantation of the Endeavor Resolute or a control BMS and a 100% re-endothelialization. Four-month follow-up in the first 30 patients revealed an in-stent late loss of 0.12mm, and no target vessel revascularization or stent thrombosis was observed.52
In a search for more biocompatible coatings, hydroxyapatite was found to be a valuable alternative as a polymer surrogate. Hydroxyapatite is a well-known and excellent bioceramic that closely resembles biological apatite (bone); it is biocompatible, bioactive, and bioresorbable, and it forms the basis of a polymer that is only 200nm thick.
Furthermore, its porous structure makes it an ideal drug carrier. Likewise, a new spongious nanocarbon coating constructed of porous, glassy, pyrolytic carbon, which guarantees extraordinary elasticity, was recently developed. Lastly, the Intracoronary Stenting and Angiographic Restenosis-Test Equivalence Between 2 Drug-Eluting Stents (ISAR-TEST) study recently showed that a polymer-free, microporous, sirolimus-coated Yukon stent (Translumina, The Drug-Eluting System Company, Hechinger, Germany) was not inferior to a polymer-based PES in the reduction of restenosis.54
Another alternative for the polymer is a heparin coating. Heparin-coated stents proved to be superior to both balloon angioplasty and BMS. In 1996, the Belgium Netherlands Stent II (BENESTENT II) randomized trial proved the superiority of heparin-coated stents compared with balloon angioplasty. A more recent registry even proved a significant reduction in stent thrombosis in the heparin-coated stent compared with a standard BMS.55
A completely new concept is heparin coupled with poly-L-lactic acid to create a so-called absorbable ‘heparinized’ polymer, which, in turn, can serve as a drug reservoir.
Indeed, the most commonly studied biodegradable polymers are derived from lactic and glycolic acid. Biodegradation is achieved by hydrolytically unstable linkages (esters) in the backbone of the polymer, which results in surface erosion. The rate of erosion or biodegradation can be altered by the molecular weight of the polymer and the number of unstable linkages. Furthermore, the drug-release profile can be adjusted by alteration of the biodegradation profile of the polymer.
Dual-elution Concept
Endothelial progenitor cells have been identified as a key factor in the reendothelialization process after stent implantation.56 To accelerate the process of endothelialization and thereby reduce the risk of thrombosis and restenosis, the Genous Bioengineered R stent (OrbusNeich, Fort Lauderdale) was developed. The Healthy Endothelial Accelerated Lining Inhibits Neointimal Growth (HEALING)-FIM (n=16) was the first clinical study to evaluate the use of an endothelial progenitor cell (EPC)-captured stent, which was developed with immobilized antibodies targeted at EPC surface antigens.
Six-month angiographic outcomes showed a binary restenosis rate of 13.3% with an associated late loss of 0.63±0.52. Nine-month outcomes showed that its use was safe and feasible (major adverse cardiac event and cerebrovascular event rate was 6.3%).57 The HEALING-II study (n=63) extended these results in a non-randomized multicenter trial. The initial results reported a zero incidence of major adverse cardiac events at 30 days and six-month in-stent restenosis rates of 17.2% with an associated in-stent late luminal loss of 0.78±0.39. Of interest is the late loss at 18 months, which decreased to 0.59±0.06mm.58 Of note, two things have to be mentioned. First, the patient’s total number of circulating EPCs was shown to be of critical importance for the efficacy of the EPC-captured stent. This can be illustrated by the results of HEALING-I: late loss in patients found to have low levels of circulating EPCs was more than double that of patients with normal circulating EPC levels. Second, the total number of circulation EPCs can be increased by an optimal usage of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins).59
Following the dual-elution trend, recent developments have been made on an EPC–DES combination: the concept is a stent with a biodegradable, abluminally focused drug on a Genous-coated platform with an additional drug component integrated throughout the polymer backbone. The stent should be able to enhance drug delivery at the abluminal site, to inhibit neointimal proliferation, and to simultaneously possess CD34 endothelial cell capture activity at the endoluminal site to enhance re-endothelialization.
Stent Thrombosis
ST has emerged as an important safety concern after stent implantation, although the rates have decreased from 20% after Wall stent implantation in the early 1990s to 0.2–1.8% after the implantation of current-generation BMS and DES.60
In an attempt to identify predictors of ST, several studies identified >15 patient- and procedure-related factors associated with early ST. Whereas in the early days the early pattern was hypothesized to be related mainly to technical aspects of stent implantation such as underexpansion and dissections, two recent large-scale registries showed that patient-related factors such as age, hypertension, smoking, renal failure, acute coronary syndrome at presentation, left ventricular function, and female gender were also independently associated with early ST.69,70
Unfortunately, the lack of consistent data and the overall low number of events make it difficult to interpret these predictors. Conversely, the late form has been related to delayed endothelialization and a hypersensitivity reaction to the drug or polymer.71,73 The most common predictors of late ST proved to be acute coronary syndrome at presentation, diabetes, and stent implantation of the left anterior descending coronary artery.69,70 A worldwide controversy is currently ongoing as to whether (late) ST indeed occurs more frequently after DES implantation. Two meta-analyses showed early ST rates between 0.51 and 0.9% for patients treated with BMS.61–74 Ong et al.75 published a series of 2,512 unselected patients who underwent stenting. Early ST proved to be equal in patients treated with BMS, SES, and PES and occurred in 1–1.5% of all patients, depending on the definition. The concerns for late ST (>30 days) after DES implantation originate from a report of a patient in the European SirolImus-coated Bx Velocity balloon-expandable stent in the treatment of patients with de novo coronary artery lesions (E-SIRIUS) study, who developed late ST 18 months after SES implantation.72 As a result of the presence of polymer fragments surrounded by giant cells and eosinophils, the authors concluded that this might have been the cause of the thrombotic event rather than the drug itself, which is no longer present in the vessel wall after 60 days. Additionally, McFadden et al.76 reported four cases of late ST when antiplatelet therapy was interrupted after elective implantation of either an SES (n=2) or a PES (n=2).
For on-label use, pooled analyses of the randomized Cypher and TAXUS family trials demonstrated identical ST rates of 3.5% (using the new Academic Research Consortium definitions45 that take not only angiographically proven ST but also MI in the target region and sudden unexplained death into consideration) in both selected BMS and DES patients for up to four years.78 However, a trend seems to arise toward a higher rate of late ST (between 30 and 365 days) in patients treated with BMS that is partially related to target lesion revascularization, reaching its peak at six months, which is compensated for by a higher rate of very late (more than one year) ST in the DES patients not related to repeated interventions. An additional interesting finding from the meta-analyses of the Cypher and TAXUS trials was the association between intervening target lesion revascularization and ST.78
In the Cypher trials, six of 15 cases (40%) of definite or probable ST in the BMS group occurred after repeated target lesion intervention. Conversely, in the SES group, 13 of 13 (100%) of the ST was primary, without intervening target lesion revascularization. In the randomized TAXUS trials, a similar pattern was observed; five of 18 cases (28%) of definite or probable ST occurred after intervening target lesion revascularization in the BMS group compared with 21 of 22 cases (95%) of primary ST in the PES group. Remarkably, in pooled patient-level data of the ENDEAVOR-I, II, and III trials, although limited to two to three years of follow-up, the occurrence of very late ST was three times lower after zotarolimus-eluting stent implantation than after BMS implantation, and ST after repeated intervention occurred in only one patient in both the DES and BMS groups.79
Recently presented long-term follow-up data of 8,146 patients treated with DES in two academic institutions showed that ST, observed in 152 patients, occurred at a median of nine days and accrued at a steady rate of 0.6% per year between 30 days and three years of follow-up.70 It is uncertain whether these rates exceed those of unselected patients treated with BMS after many years of follow-up. To settle this issue, extremely large-scale randomized observations comparing DES and BMS in all patients are needed. Unfortunately, the likelihood that in the present era these trials will actually be performed is low, and even if they were started it would take at least another three to four years for valid conclusions about the long-term results to be drawn. For now, we need to rely on large-scale registries in which the BMS control groups often comprise lower-risk patients and lesions.
Did Stent Thrombosis Translate into Hard Clinical End-points?
Knowing that DES are able to reduce restenosis by 70%, one could subsequently expect a long-term benefit in survival.80–81 Instead, concerns were raised about a higher rate of death and MI after DES implantation, and even cancer was hypothesized to occur more frequently in DES-treated patients.82,83
Therefore, stent manufacturers put complete data sets of randomized trials with long-term patient-level-based follow-up at the disposal of independent researchers and statisticians for further analyses. After several intense scrutinizing exercises, long-term death and MI rates appeared to be similar in pooled analyses of the pivotal Cypher and TAXUS trials, including relatively low-risk patients.84
In real-world registries in which off-label DES use accounted for up to 60% of the population, the outcomes seem more at variance. Whereas the large DEScover and West Denmark registries and the three-year follow-up of the Rapamycin-eluting Stent Evaluated at Rotterdam Cardiology Hospital (RESEARCH) registry showed equal survival rates between DES and BMS, a large-scale Swedish registry highlighted a significantly lower survival rate in the DES group compared with the BMS group at 2.4 years of follow-up.85–87 Of note, the BMS control groups in the above-mentioned registries included significantly fewer complex patients because of either the sequential nature of the cohorts or substantial selection bias favoring the BMS control groups. It remains disputed whether comprehensive regression and propensity analyses are able to completely account for these differences.
It is to be expected that, in the long term, higher rates of very late ST after DES implantation will put our DES-treated patients at higher risk for death and MI. Although DES have been shown to be safe for up to four years for on-label use, the off-label long-term safety has not yet been determined, given the controversial findings of large real-world registries and the lack of properly powered randomized controlled trials.
Potential New Platforms
Several limitations and side effects have been associated with coronary stenting. First, stents cause permanent physical irritation with the risk of long-term endothelial dysfunction or inflammation.9 Second, stents possess a high thrombogenicity.88 Third, stents create an inability for the vessel to remodel and act in a normal physiological way.4 Finally, stents create difficulties for possible future bypass surgery and non-invasive imaging.
The first bioabsorbable stents, made of poly-L-lactic acid, were recently studied in porcine models.89 The first successful in-human experience with a poly-L-lactic acid stent was described by Tamai et al. in 2000.90 The study included 15 patients treated with a monopolymer poly-L-lactic acid Igaki- Tamai stent (Igaki Medical Planning Co, Ltd, Kyoto, Japan) with a zigzag helical coil pattern. The stent expanded by itself at a temperature of 37°C. Angiographic restenosis rate and TLR was 10.5%, which thereby proved that its use was feasible, safe, and effective in humans.
REVA Medical, Inc. (San Diego, California) is investigating a fully absorbable polymer stent with a ‘slide and lock’ design—sliding parts with monodirectional lockouts that are hypothesized to result in a nearly negligible stent recoil (see Figure 3). The stent consists of a radiopaque tyrosine-derived polycarbonate backbone. The composition of the polymer, comprising three basic components, allows the resorption time to be varied by a change in the ratio of these components. Both a bare and a paclitaxel-eluting version, in which the polymer is mixed with the drug, will become available. The Randomized Endovascular Study of the REVA Bioresorbable Stent (RESORB) clinical trial has recently been designed to assess the safety of this new platform.
The new generation of a bioabsorbable everolimus-eluting stent BVS (Abbott Vascular, Santa, Clara, CA) has shown no cardiac death and no restenosis at two years in 30 patients, and also no late stent thrombosis was recorded.94 The BVS has a bioabsorbable polylactic acid backbone with bioabsorbable polylactic acid coating that contains the antiproliferative drug evirolimus.94
Another alternative for the metallic backbone of the stent was found in magnesium. Magnesium, with antithrombotic, antiarrhythmic, and antiproliferative properties, is one of the first natural body components to be used as a basis for a bioabsorbable stent. Several experimental studies to evaluate the efficacy of a magnesium alloy stent degradable by biocorrosion have been performed. Heublein et al. described the use of a coronary stent prototype that consisted of the non-commercial magnesium-based alloy AE21 (contains 2% aluminum and 1% rare earth metals) with an expected 50% loss of mass within six months in 11 domestic pigs (see Figure 4). Quantitative angiography at follow-up showed a significant 40% loss of perfused lumen between 10 and 35 days caused by the loss of mechanical integrity of the stent.91 One year later, the use of a bioabsorbable magnesium-alloy-based stent with a controlled corrosion in 20 patients with critical limb ischemia was described. At nine months, a 90% vessel patency was observed.92 As a result of the successful FIM trial (n=5) by Erbel and colleagues, the enrollment in the larger worldwide Clinical Performance and Angiographic Results in Absorbable Metal Stents (PROGRESS-AMS) study has recently been completed.93 The four-month results showed a late loss of 1.08±-0.49 and an ischemia-driven TLR rate of 23.8%, which was similar to those reported with the use of BMS.
Recently, reservoir technology was used in the NEVO™ sirolimus-eluting stent (Cordis). The theoretical principle is to have the drug released from these reservoirs as opposed to coating the entire stent with the drug. By doing so 70% of the stent will remain bare-metal and within three months after full drug release will be 100% bare-metal. The results of the first trial (NEVO RES-1 study) comparing this technology with the paclitaxel-coated Liberte stent, presented at Euro PCR 2009, showed that at six months late loss was significantly lower in the NEVO stent arm (0.13 versus 0.36; p<0/001). However, this did not translate into a significant difference in TLR rate.
Conclusion
DES are effective in reducing the need for repeat revascularization. However, as with any new medical device, there are still unresolved issues regarding the platform, the drug kinetics, and the drug carrier. These are currently under development and we may be in for a new generation of platforms that will probably be bioabsorable and double-coated to enhance endothelialization and, at the same time, inhibit new intimal hyperplasia with a drug that is safe and carried by a non-toxic polymer. These new developments in safety must be meticulously scrutinized before being introduced into interventional cardiology.