The use of drug eluting stents for coronary artery disease
Peter Barlis, MBBS, MPH,
FRACP Jun Tanigawa, MD,
Carlo Di Mario, MD, PhD
Department of Invasive Cardiology, Royal Brompton Hospital, London, UK
Address for correspondence:
Department of Invasive Cardiology
Royal Brompton Hospital, Sydney Street
London SW3 6NP, UK.
Tel: +44-20-7351-8616 Fax: +44-20-7351-8614
The growth in the availability and use of drug eluting stents has seen the interventional cardiologist tackle more lesions with increasing complexity and in higherrisk patients. The single digit rates of restenosis observed have helped propel these novel devices well ahead of their bare metal counterparts, with applications in bifurcation lesions, chronic total occlusions and unprotected left main disease. Not all has come without cost, however, and the problem of late stent thrombosis has cast a cloud on long-term outcomes and the optimal duration of dual anti-platelet therapy. This review will describe the role drug eluting stents have in contemporary interventional cardiology with a focus on their shortcomings and possible future refinements.
Since the introduction of drug eluting stents (DES), interventional cardiology has witnessed tremendous growth in the types and numbers of coronary lesions treated percutaneously. The advent of these novel devices has propelled percutaneous coronary intervention to the forefront of the management of patients with coronary artery disease. Globally, the DES market, and in particular the two leading manufacturers (Cordis Corp. [Johnson & Johnson] and Boston Scientific, both USA), have found their way into almost all modern catheterisation laboratories and provided opportunities for interventional cardiologists to safely tackle complex coronary lesions even in high-risk patients.
More than 20 million DES have been implanted worldwide to date and these novel devices continue to dominate the interventional literature. More recently, evidence is mounting to support their use in complex coronary lesions including unprotected left main stenosis [1-6] where a percutaneous strategy may be an alternative to coronary artery bypass grafting (CABG), although results of current randomised trials are eagerly awaited. DES have furthermore become a vital part of the armamentarium in treating chronic total occlusions and bifurcation lesions. The only limitation to more widespread use has been their direct and indirect costs. These stents remain in the vicinity of three to four times more expensive than bare metal stents (BMS), and concerns about the risk of late stent thrombosis have prolonged the duration of dual anti-platelet therapy including 75 mg of clopidogrel to an average of 9 to 12 months. This review will provide a practical framework for the role of DES in contemporary practice and give a glimpse into what the future may hold for these innovative devices.
DES for de novo coronary lesions
A number of studies [7-11] have convincingly demonstrated a reduction in the need for target vessel revascularisation (TVR) with DES, but the benefits in reducing myocardial infarction (MI) and death remain blurred (see Table 1). A meta-analysis of 10 randomised trials concluded that both sirolimus- and paclitaxeleluting stents were equivalent to BMS in terms of mortality and overall MI risk for the first year of followup . Still, these trials deal with single-vessel, singlelesion treatment in a super-selected patient population; results can be different when complex-lesion and multivessel disease is treated in high-risk patients.
DES for the treatment of bare metal in-stent restenosis
Studies such as ISAR-DESIRE  and TROPICAL  have confirmed significantly improved angiographic outcomes, led primarily by reductions in the need for TVR with restenosis rates less than 10% using DES for BMS in-stent restenosis (ISR).
In evaluating the feasibility of DES compared with vascular brachytherapy (VBT), Barlis et al  identified a rate of major adverse cardiac events (MACE) significantly lower in the DES group (n=29) compared with the VBT group (n=25) (14% versus 40%, p=0.03). More recently, two randomised controlled trials comparing the use of DES with VBT confirmed superior results out to 9 months of follow-up for a strategy of repeat stenting with a DES [16,17]. The TAXUS V ISR (instent restenosis) study  enrolled 396 patients and showed that, compared with VBT, implantation of paclitaxel-eluting stents (PES) reduced the 9-month rate of TVR from 17.5% to 10.5% (p=0.046) and target lesion revascularisation (TLR) rate from 13.9% to 6.3% (p=0.01). The study also demonstrated an 11.5% rate of MACE for the PES group, compared with 20.1% for VBT.
The SISR trial  randomised 384 patients with ISR to receive either VBT (n=125) or a sirolimus-eluting stent (SES, n=259). In this trial target vessel failure occurred in 22% patients who underwent VBT and 12% of those who received a SES. Overall, the TLR rate was 19% in the VBT group compared with 8.5% in the SES group (relative risk for VBT versus SES, 2.3, 95% CI, 1.3-3.9 p=0.004). The rate of MACE was 19% in the VBT group versus 10% in the SES group (p=0.015).
DES for complex coronary interventions:
unprotected left main stenosis
Restenosis in an unprotected left main (ULM) may mean sudden death and this sword of Damocles has prevented widespread diffusion of percutaneous treatment with conventional BMS despite immediate results being better than surgery. Improvements in stent deployment techniques using high balloon pressures, IVUS guidance and glycoprotein IIb/IIIa receptor antagonists have dramatically reduced complications such as sub-acute stent thrombosis [18-21]. Now, a number of registries and non-randomised studies (Table 2) have emerged suggesting the effectiveness of DES for ULM treatment.
Table 1. An overview of four randomised trials of drug eluting stents (DES) in de novo coronary lesions.
In a series of 102 patients with ULM, Park et al  reported excellent clinical and angiographic outcomes with elective SES implantation. Six-month angiographic follow-up was completed in 84.3% of patients, with an overall restenosis rate of 7.0% and 2.0% requiring TLR. Oneyear mortality was 0% and MACE-free survival 98%. In looking at a more representative population, Chieffo et al  enrolled 85 patients with ULM with high mortality risk scores (EuroSCORE >6 and/or Parsonnet >15) present in 45%. The 6-month cardiac mortality and MACE-free survival rates were 3.5% and 80%, respectively. With the inclusion of emergency procedures due to ST-elevation MI (STEMI) and cardiogenic shock, Valgimigli et al  reported a mortality rate of 11% at 12 months. Notably, all deaths within 30 days occurred in patients with STEMI and cardiogenic shock.
On multivariate analysis, the authors identified use of DES, Parsonnet classification, troponin elevation at entry, distal left main location and reference vessel diameter as independent predictors of MACE .
Recently two groups have published non-randomised studies of DES for ULM compared with consecutive patients undergoing CABG during the same time period. Lee et al  compared 50 patients having DES percutaneous coronary intervention (PCI) for ULM with 123 CABG patients. High-risk patients (Parsonnet score >15) comprised 46% of the CABG group and 64% of the PCI group (p=0.04). Thirty-day mortality was lower in the PCI group (2% versus 5% for CABG) with 6-month follow-up demonstrating a non-significant survival advantage in the PCI group (96% versus 87% for CABG). Price et al  studied 50 patients with surveillance angiography performed at 3 and 9 months’ follow-up. The group predominantly consisted of patients with distal bifurcation lesions (94%). TLR occurred in 19 patients (38%) over a mean follow-up of 276 ± 57 days. There were two acute stent thromboses and five deaths at 1 year. Angiographic follow-up at both 3 and 9 months revealed angiographic restenosis in 23% of left main to left anterior descending (LAD) stents and 35% in the left circumflex (LCx) with an overall angiographic restenosis rate of 42% in any vessel. Although alarming, this study confirms the need for meticulous surveillance of patients receiving DES for ULM even if they remain asymptomatic. Results from multicentre randomised trials (e.g. SYNTAX) comparing the percutaneous approach with CABG are eagerly awaited.
Table 2. Clinical and procedural characteristics and outcome of drug eluting stent (DES) implantation for left main stenosis in five registries.
DES for complex coronary interventions:
The use of DES for bifurcation lesions has seen improved angiographic and clinical outcomes compared with previous attempts using BMS. Still, restenosis of the side branch ostium and late thrombosis [22,23] following DES implantation continue to hamper the treatment of this lesion subset while casting doubts on the best long-term strategy. In fact, the problem of focal restenosis at the side branch ostium has not been overcome by the use of DES and the best treatment for bifurcation lesions remains a subject of controversy.
In deciding on an appropriate strategy for PCI to a bifurcation lesion, the operator will chose between a simple (DES implantation only at the main vessel with optional balloon angioplasty or stenting at the side branch) or complex (DES implantation at the main vessel and the side branch) technique.
In one of only a few randomised trials looking at these bifurcation strategies, Colombo et al enrolled 85 patients (86 lesions) to assess the feasibility and safety of two SES using a “T” technique at true bifurcation lesions (>50% stenosis in both main vessel and ostium of side branch) versus the implantation of a single SES in the main vessel with balloon dilatation across the stent struts for the side branch . There was an extremely high crossover rate (22/43, 51%) in the provisional stent group who received two stents. At 6-month follow-up, the total restenosis rate was 25.7%, and it was not significantly different between the doublestenting (28.0%) and the provisional side branchstenting (18.7%) groups. In another study, Pan et al  randomised 91 patients to stenting of the main branch and balloon dilatation for the side branch (n=47) and compared this group to a technique of stenting both branches (n=44). At 6-month follow-up, there were no differences in clinical outcomes between the two groups: restenosis of the main vessel was observed in one (2%) patient from the single stent arm and four (10%) from the double stent group (p=ns). Restenosis of the side branch developed in two (5%) patients from the single stent group and six (15%) of those receiving two stents. Although the optimal strategy for bifurcation stenting remains undetermined, stenting in the DES era must ensure complete lesion coverage, especially at the side branch ostium with well apposed stents to limit complications such as restenosis or thrombosis .
DES for complex coronary interventions: chronic total occlusions
Chronic total occlusions (CTO) are identified in up to 30% of patients with significant coronary artery disease on angiography [26,27] and represent 10-15% of cases treated by PCI [28,29]. The primary success rate remains relatively low, mainly due to inability to cross the occlusion with the guide wire , while the recurrence rate is higher than that of subtotal stenoses particularly with BMS . Moreover, the overall procedure and fluoroscopy times are longer and equipment use higher than with PCI of non-occluded vessels . A successfully recanalised CTO can improve anginal status, left ventricular (LV) function and survival, but this is hindered by the high rate of restenosis or reocclusion .
Recently, reports comparing DES with BMS have shown a dramatic decrease in restenosis rate after CTO treatment with relatively low complexity [31,32]. In the PRISON II study  the binary restenosis rate (in-stent) for the BMS group was 41% versus 11% for the sirolimus group at 6-month angiographic follow up (p<0.0001). The MACE rate was also significantly reduced in the sirolimus group (4% versus 20% for the BMS group; p<0.001). Despite the significant advantage of DES in reducing ISR, this is not eliminated completely with predictors of ISR post-CTO treatment similar to those seen with BMS including smaller reference vessel diameter and longer stent length [6,33]. There is no doubt, however, that when available DES should be the first-line treatment following recanalisation of a CTO.
DES for ST-elevation myocardial infarction (STEMI)
The use of DES in STEMI remains a controversial area with limited data currently available to definitively support the widespread use in such cases. In the STRATEGY trial,  the first randomised trial looking at DES and acute MI (AMI), 175 patients were assigned to receive single high-dose bolus tirofiban plus SES versus abciximab plus BMS. The primary composite endpoint of death, re-infarction, stroke, and angiographic binary restenosis at 8 months occurred more frequently in patients allocated to receive abciximab plus BMS (50% versus 19%, p<0.001). The main driver was a higher restenosis rate in the BMS group (36% versus 9%, p=0.002). There were no differences in death, re-infarction or stroke at 8 months, and no episodes of stent thrombosis in the SES group. Similarly, the RESEARCH registry  compared the outcomes of 186 consecutive patients with STEMI treated with DES against those of 183 patients treated with BMS. The composite rate of death and re-infarction was 9% at 10 months with a long-term TVR rate of 5%.
A recent randomised study by Lee et al  using the two commercially available DES (Cypher® [SES] and Taxus® [PES]) enrolled 217 patients randomly assigned to PES (n=108) or SES (n=109). The PES group showed higher angiographic in-segment restenosis (16% versus 4%, p=0.04) and higher in-segment late loss (0.35 ± 0.69 mm versus 0.05 ± 0.42 mm, p<0.001) than the SES group. Ten-month MACE rates were also significantly higher in the PES group (92 ± 3% versus 82 ± 4%, p=0.04). The PASSION trial  randomised 619 patients with STEMI to receive either a PES or a BMS. The use of PES was not associated with a difference in the primary composite endpoint of death, MI or TLR when compared with BMS at 1 year . The TYPHOON trial  compared the SES with BMS in STEMI and showed it to be associated with a significant reduction in target vessel failure compared with BMS (3.7% versus 12.6% p<0.0001) . This was largely driven by a reduction in TLR with no difference in death or MI between the two groups .
Drug eluting stent thrombosis
Drug eluting stent thrombosis rates as high as 3.6% have been reported in patients with complex lesions including bifurcations . Such complications have also cast a cloud on the appropriate duration of dual anti-platelet therapy, particularly given the increased bleeding risks of longterm therapy . Ong et al  studied 2,006 patients treated with DES and identified an early (<30 day) incidence of stent thrombosis of 1%. At a mean follow-up of 1.5 years, there were eight angiographically confirmed cases of late stent thrombosis, three with SES (at 2, 25 and 26 months) and five with PES (at 6, 7, 8, 11 and 14.5 months). Three cases were related to complete cessation of anti-platelet therapy, two cases occurred while patients were on aspirin therapy within 1 month of cessation of clopidogrel and three cases occurred at a time when patients were apparently clinically stable on aspirin monotherapy . There were no cases of late stent thrombosis in patients who were on dual anti-platelet therapy.
More recently, a study by Joner et al  reviewing autopsies of 23 DES cases compared these with 25 autopsies from matched patients with BMS implantation. Of 23 patients with DES >30 days old, 14 (61%) had evidence of late stent thrombosis. Cypher® and Taxus® DES showed greater delayed healing characterised by persistent fibrin deposition and poorer endothelialisation (55.8 ± 26.5%) compared with BMS (89.8 ±20.9%, p=0.0001). In five of 14 patients suffering late stent thrombosis, antiplatelet therapy had been withdrawn with procedural and pathologic risk factors for the development of late stent thrombosis being: 1) local hypersensitivity reaction to the polymer; 2) ostial and/or bifurcation stenting; 3) malapposition/incomplete apposition; 4) restenosis; and 5) strut penetration into a necrotic core . In an accompanying editorial, Colombo et al  reflected on many of these pathological findings, which they postulated may be a manifestation of a number of factors including the individual’s healing response to sirolimus or paclitaxel, the drugs themselves, the polymer or a complex interplay of all factors. Barlis et al  have also confirmed the multi-factorial nature of stent thrombosis with findings of poor stent expansion and eosinophilic infiltrates in thrombotic material extracted from a patient successfully treated for late acute stent thrombosis, 12 months after SES implantation.
Recently presented findings from the BASKET-LATE trial,  designed to assess the incidence of stent thrombosis in patients treated with DES (n=502) versus BMS (n=244) following cessation of clopidogrel therapy, identified rates of death and MI that were higher in the DES group. These were more frequently related to late stent thrombosis which carried a 4-times greater risk of cardiac death/MI compared to non-thrombosis related events. Although the complete 18-month data did not confirm these initial findings, anecdotal evidence of stent thrombosis several months or years after DES implantation is worrisome and involves both Cypher and Taxus trials. These observations have also led the FDA to issue a statement on coronary DES in which they support their use for approved indications but will more stringently focus their attention on monitoring late thrombotic risks to minimize future patient harm .
Intravascular imaging in the DES era
Intravascular imaging has the ability to give the operator greater detail in both lesion assessment and in optimising stent implantation compared with angiography alone. Many of the DES studies incorporating intravascular ultrasound (IVUS) have shown that intimal hyperplasia volume is reduced to <10% of stent volume, but stent under-expansion continues to be a consistent finding in DES failures (restenosis and thrombosis) . Intravascular ultrasound also has a vital role to play in guiding the percutaneous treatment of complex lesion subsets including left main disease where optimal stent sizing and expansion is of paramount importance [20,45].
IVUS remains the gold standard for the assessment of stent strut apposition but is limited by low resolution (100-150 µm) and the artefacts induced by the stent struts. More recently, optical coherence tomography (OCT, LightLab®: Light Lab Imaging Inc., Westford, MA, USA), a relatively new imaging modality using infrared light, has been used with the distinct advantage of greater resolution (10-15 m) and less artefact compared with IVUS . With the improved image resolution, this modality can give more precise information on stent strut apposition, which may itself play a pivotal role in unravelling some of the reasons behind DES failures.
Some of the important parameters of coronary stents include the structure, geometry and dimensions. These play an important role in deliverability, visibility, scaffolding performance and procedural success. Refinements of currently available stents will come from improvements in flexibility and handling together with advances in polymer design, with attention to biodissolvable stent coatings  especially given that the stent polymer has come under scrutiny as a possible contributor to long-term DES failure. Polymers currently utilized for DES are either biodegradable or non-biodegradable.
While non-biodegradable polymers reside permanently on the stent surface, biodegradable polymers are released together with the drug and dissolve after a certain period of time. One such stent, the Biolimus A9 stent (Biosensors International, Singapore) is a novel DES that incorporates the S-Stent platform, a thin, stainless steel, laser-cut, tubular stent with 0.0054” strut thickness . Biolimus, a sirolimus analogue, is coated onto the S-Stent platform with a bio-absorbable, polylactic acid, polymer matrix that releases the drug (~70% eluted in 30 days); subsequently the polymer is absorbed over time locally into cells .
Novel gene eluting stents with plasmid encoding human vascular endothelial growth factor (VEGF) 2 coatings have shown promise in reducing ISR in animal models and may be used alone or in conjunction with other DES in the near future . Apart from sirolimus and paclitaxel, Tyrphostin AGL-2043, a potent tricyclic quinoxaline inhibitor of platelet-derived growth factor (PDGF) B receptor tyrosine kinase, has also demonstrated impressive results in porcine models with the ability to significantly reduce smooth muscle cell proliferation and migration and reduce neointimal formation .
A further novel technology aims to eventually do away with the stent altogether. These biodegradable stents are based on a magnesium alloy that allows controlled corrosion with release to the vessel wall and the blood stream of a natural body component such as magnesium with beneficial anti-thrombotic, anti-arrhythmic and anti-proliferative properties . Recently, a forthcoming study (ABSORB) using a fully bio-absorbable stent coated with everolimus was announced - this aims to enrol 60 patients and assess the safety profile of this technology . Another recent advance uses a balloon catheter coated with paclitaxel, thereby doing away with the stent altogether. This technology has been successfully applied to a small series of patients with in-stent restenosis  and may become a viable alternative for the treatment of de novo coronary disease in the future.
Drug eluting stents have been instrumental in broadening the complexity and type of coronary lesions treated by interventional cardiologists. With applications in ISR, bifurcation lesions, chronic occlusions and unprotected left main disease these novel devices have played a great part in giving rise to excellent clinical and angiographic outcomes. There remain, however, unanswered questions relating to long-term risk of stent thrombosis and the optimal duration of therapy with both aspirin and clopidogrel. Such complications may be addressed by future modifications in stent design and delivery systems that will no doubt continue to modify the ever-changing practice of interventional cardiology well into the future.
02-2007 BUY1154818/JB2548/MB002444/SC 13th edition