Author Affiliations: From the Department of Angiology, Herz-Zentrum Bad Krozingen, Bad Krozingen, Germany (TZ) and the Department of Diagnostic and Interventional Radiology, University Tuebingen, Germany. Correspondence: Dr. Thomas Zeller, Herz-Zentrum Bad Krozingen, Südring 15, D-79189 Bad Krozingen, Germany. E-mail address: firstname.lastname@example.org. Manuscript submitted July 17, 2008, provisional acceptance given August 18, 2008, manuscript accepted September 5, 2008. Disclosure: The authors report no financial relationships or conflicts of interest regarding the content therein. _________________________ Abstract In the coronaries, the role of drug-eluting stents (DES) is already well defined. However, in the peripheral arteries very limited data exists about their potential benefit. A small randomized study on superficial femoral artery (SFA) lesions comparing bare nitinol stents with short-time sirolimus-eluting nitinol stents resulted in no benefit for the DES group. Two different approaches are still under evaluation, one registry and randomized study investigating the effect of short time release of paclitaxel (the Zilver PTX global registry and the Phase 2 Zilver PTX clinical randomized trial), and another registry evaluating the outcome of SFA treatment using long-term drug release of everolimus (STRIDES). Due to different study protocols, these studies will be difficult to compare. The experience with DES below-the-knee (BTK) is also still limited. Four single-center reports suggest a beneficial outcome of the drug-eluting Cypher stent regarding patency and clinical outcome. Data from multicenter randomized studies are still pending. A randomized study investigating the value of short-time paclitaxel release using a drug-coated balloon gave promising results, with significantly improved patency rates compared to plain balloon angioplasty in femoro-popliteal lesions. Below-the-knee, this promising concept is still under evaluation (PICCOLO study). This article gives an overview of published data and ongoing trials on drug-releasing concepts in the peripheral arteries. Drug-Eluting Stents and Drug-Coated Balloons in the Superficial Femoral Artery Percutaneous transluminal balloon angioplasty for revascularization of the superficial femoral artery (SFA) has an initial technical success rate of more than 95%.1 However, restenosis occurs in 40–60% of the treated segments after 6 to 12 months.2–4 Compared to other vascular beds, such as the coronary arteries and the renal arteries, the restenosis rate is much higher in the SFA.5 Furthermore, the benefit of a stent depends on the type of stent, and long-term patency rates will be confirmed last (but not least) in the lower extremity, because of stent fractures.6–9 Several other attempts to increase long-term patency have failed. Drug-eluting stents. The only study which reported local drug delivery in the SFA was the Sirolimus-Coated Cordis Self-Expandable Stent (SIROCCO) trial, in which sirolimus-coated stents were not significantly superior to uncoated stents.8–10 In the SIROCCO I trial, including 36 patients randomized either to the Smart nitinol (bare stent) or to the sirolimus-eluting Smart stent, Duda et al showed that it is safe and feasible to use sirolimus-eluting Smart (Cordis, Johnson & Johnson, Miami Lakes, Florida) stents for the treatment of obstructive SFA disease.8 After 6 months, the primary endpoint angiographic binary instent restenosis rate was 0% for the drug-eluting stent (DES) versus 17.6% for the bare nitinol stent. This difference was not statistically significant, due to a much better performance of the bare nitinol stent group, as expected. This study was the first one to raise the issue of stent fractures in lesions treated with 3 overlapping stents. Consequently, in the SIROCCO II trial, in which the investigators evaluated the efficacy of the slow-release, sirolimus-eluting stent (SES) in a larger number of patients (n = 54), only two stents were allowed to be placed, leading to a reduction of the maximum lesion length from 20 cm in SIROCCO I to 14.5 cm (average lesion length was 8.1 cm). After 6 months, again, the DES group did extremely well, with a binary instent restenosis rate of 0%, representing the primary study endpoint. This was probably due to the reduced maximum lesion length. The bare nitinol stent group resulted in a binary restenosis rate of only 7.7%, again failing to show any statistically significant difference in any of the predefined study endpoints.9 After 24 months, primary patency rates for both study groups were virtually the same (DES group: 76%, bare nitinol stent group: 75%).10 Interestingly, stent occlusions occurred only in the bare nitinol stent group. By limiting the number of stents used per lesion from 3 to 2, the stent fracture rate after 6 months was reduced from 18% in the SIROCCO I study to 8% in the SIROCCO II study. Reasons for the failed concept of the reduction of restenosis rate using SES are multiple. The most obvious ones are the use of the coronary Cypher stent (Cordis), adapted sirolimus dose, and release kinetics. The SES contained 90 mg of sirolimus per square centimeter of stent surface area within a copolymer matrix (30% sirolimus, 70% copolymer) that was 5 mm thick. However, the distance between the stent struts of the Smart stent was much larger compared to the Cypher stent, leading to a lower drug dose in the SFA compared to the coronary arteries. In addition, the release kinetics of the sirolimus-coated Smart stent was much faster than with the Cypher stent. With sirolimus especially, a longer sustained release might be much more of an issue than paclitaxel. Proinflammatory side effects of the polymeric matrix of the drug-coated Smart stents were also discussed as a factor for the reduced benefit of the local drug delivery. Another likely underestimated aspect is the different vessel-wall-to-injury response of a balloon-expandable coronary stent and a self-expanding nitinol stent. Where a balloon-expandable stent is placed with its nominal diameter resulting in a predictable remodeling period of about 28 days, a self-expanding nitinol stent is usually oversized by 1 to 2 mm compared to the reference vessel diameter. This results in ongoing stent expansion and motion of the stent through the vessel wall layers over weeks, with a prolonged vessel wall response, such as smooth-muscle cell proliferation and macrophage migration from the adventitia. Based on these considerations, it will be interesting if one of the different concepts of the 2 ongoing DES trials using different drugs and drug-release concepts will be successful in restenosis suppression. Currently, one concept is investigating paclitaxel, a highly lipophilic antineoplastic drug delivered within hours from a polymer-free nitinol stent platform (3 µg/mm2 dose density and maximum 880 µg total dose on largest stent). More than 1000 patients received the paclitaxel-eluting Zilver stent (Zilver PTX, Cook Medical, Inc., Bloomington, Indiana). In a randomized trial that recently terminated the enrollment, 420 patients were randomized either to be treated with Zilver PTX stents or with plain balloon angioplasty, and optional stenting in a 2:1 randomization protocol. The lesion length in this randomized trial is limited to 14 cm. In addition, a large registry focusing on safety issues of the Zilver PTX stent included 760 patients and completed enrollment in the beginning of 2008. Also, in this registry, long lesions (up to 28 cm in length, accommodating 4 stents at maximum) and instent restenoses were treated with the DES. No other endovascular tools such as atherectomy were allowed for pretreatment. A completely different approach is used in the STRIDES registry: 104 patients received an everolimus-eluting DYNALINK-E stent (Abbott Vascular, Diegem, Belgium) with a permanent polymer coating, resulting in a 6-month release kinetic with a dose density of 225 µg/mm2. Enrollment of this registry was finished in February 2008. The trials with DES in the SFA are summarized in Table 1. Drug-coated balloons. Several studies in cell culture and swine11–15 have demonstrated sustained inhibition of smooth-muscle cell proliferation when cells or tissues were exposed to paclitaxel for up to a few minutes. This inhibitory effect was seen in animals, both when paclitaxel was added to the X-ray contrast agent used to visualize the arteries during the intervention and when it was coated on angioplasty balloons. Clinical trials suggested significant inhibition of re-restenosis following treatment of coronary in-stent restenosis by paclitaxel-coated angioplasty balloons.16 The tolerance of paclitaxel dissolved in a contrast agent was confirmed in a dose-escalating phase I study in humans undergoing coronary angiography.17 The Local Taxane with Short Exposure for Reduction of Restenosis in Distal Arteries (THUNDER) trial was designed to investigate the effect of paclitaxel on restenosis following angioplasty of stenotic or occluded superficial femoral or popliteal arteries — de novo lesions and restenoses — using the above-described methods of administration in comparison to conventional balloon angioplasty.18 Overall, 154 patients were randomized in a double-blinded fashion. Balloons were either uncoated or coated with paclitaxel at a dose of 3 µg/mm2 of balloon surface, depending on the treatment group assignment. Balloons were inflated to restore the reference diameter of the vessel with a maximum of 12 atmospheres for a standardized inflation time of 1 minute. The average paclitaxel dose per patient on 1 or more balloons was 5 mg (range 1–16 mg), and almost uniformly, 17 mg in the group treated with paclitaxel added to the contrast agent. In patients treated with the drug-coated balloons, maximum paclitaxel concentration in plasma remained below the detection limit of the method (Drug-Eluting Stents and Drug-Coated Balloons Below the Knee Although the results of interventional therapy are promising, the occurrence of restenosis and re-occlusions may result in insufficient clinical improvement. Endovascular stenting using balloon-expandable coronary stents has recently been shown to be potential ly valuable in improving the results of infrapopliteal angioplasty for patients with CLI and lifestyle-limiting claudication.19,20 Even so, re-obstruction rates of coronary bare-metal stents in tibial arteries are exceeding 50% after 12 months.1 Nevertheless, there is a growing acceptance that endovascular treatment of BTK lesions and critical limb ischemia (CLI) will become the primary treatment strategy in the near future. This is partly due to the disproportionate levels of comorbidity that exist in many patients and the decrease in the number of patients who are considered to be suitable for treatment by surgery. At the same time, we have witnessed the introduction of new endovascular techniques for the treatment of BTK lesions causing CLI. The use of carbon-coated stents are approved for the treatment of endovascular indications, such as CLI, at least in Europe (InPeria Stent, Sorin Biomedica Cardio Corporation, Saluggia, Italy). Excimer laser-assisted angioplasty, the use of local anesthesia, the potential to reduce the cost of treatment, and reduced hospitalization are all contributing factors in the growth of endovascular therapy in this challenging patient subset. Furthermore, several investigators have demonstrated the potential benefits to be gained from the use of stents to treat infrapopliteal lesions.21-23 Therefore, with a view to restricting restenosis following infrapopliteal interventions and improving clinical outcome of BTK CLI, the use of SES such as the Cypher stent after sub-optimal and/or complicated angioplasty may inhibit neointimal hyperplasia, resulting in superior angiographic and clinical outcomes over bare-metal stents. The Cypher stent is currently CE marked in Europe for peripheral use in CLI patients. Drug-eluting stents below the knee. Four trials including 166 patients treated with Cypher stents were published.24–27 The mean age was 71.7 years (68.8–73.8), and just over half had diabetes (27.8–83.3%). On entering their respective trial, all patients were classified as Rutherford category 3–6. Six months technical outcome data from the 4 trials are summarized in Table 2. The major difference in the 4 studies was that in the Siablis trial,27 only spot stenting was performed; in the other 3 trials, full lesion coverage was obtained. The group of Siablis was the only one reporting 1-year outcome data with a sustained benefit of Cypher stent implantation compared to bare-metal stent use.28 The fate of patients who underwent treatment with Cypher stents for their BTK CLI in the 4 selected studies can be compared to the known natural history of patients with BTK CLI. Patients undergoing successful intervention with Cypher stents had a 100% freedom from major amputation, compared with historic 1-year amputation rates of 30%.1 These outcomes also compare favorably with the 96% limb salvage rates in patients with initial technical success treated by primary stent-supported angioplasty for the treatment of BTK CLI.21 Furthermore, the need for target vessel revascularization across the 4 studies was only 4% compared with surgical revision rates as high as 49%.1 Therefore, the fact that nearly all patients in the 4 studies demonstrated sustained clinical improvements suggests that treatment with Cypher stents for BTK CLI is a promising alternative to tibial bypass surgery. The recently started ACHILLES trial randomly compares patency rates following balloon angioplasty and Cypher stent implantation in BTK lesions with a length up to 9 cm. A second trial (YUKON) randomly compares the on-site, sirolimus-coated Yukon stent with the bare stent platform in BTK lesions, with a maximum length of 45 mm. The enrollment for this study was finished in February 2008. Drug-coated balloons below the knee. No data has been published or presented on the use of drug-coated balloons BTK. Just recently, a randomized trial comparing the angiographic outcome of paclitaxel-coated balloons with plain balloon angioplasty in patients suffering from CLI with a lesion length up to 20 cm started enrollment (PICCOLO). Conclusion Currently, the use of antiproliferative agents, either exposed by stents or balloon catheters in preventing restenosis in infrainguinal arteries, is still investigational. The only finished trial investigating the concept of sirolimus elution via a self-expanding stent platform in SFA lesions (SIROCCO) failed to prove superiority compared to the bare nitinol stent.8–10 First, small single-center reports comparing SES and bare stents BTK give hope in reducing restenosis rates comparable to the coronary circulation.24–27 Another concept using short-time release of paclitaxel coated on a balloon catheter in femoropopliteal lesions (THUNDER) confirmed earlier positive results in instent coronary lesions. Further research on both concepts is mandatory and already on the way. Within the next 2 years, more data will be available to evaluate the future role of local drug delivery for restenosis suppression in peripheral arteries.