Optimization of Local Methylprednisolone Delivery to Inhibit Inflammatory Reaction and Neointimal Hyperplasia of Coated Coronary

Metallic coronary stents have shown to be essential in the treatment of (sub)acute vessel closure after balloon angioplasty and multicenter randomized trials have shown a decreased restenosis rate in selected patient subgroups compared to conventional balloon angioplasty. However, in-stent restenosis caused by an increased neointimal hyperplasia remains the major limitation of coronary stents.1,2 Restenosis is related to vessel injury caused by stent implantation and foreign body response induced by stenting, resulting in thrombosis, inflammation, dedifferentiation, migration and proliferation of smooth muscle cells (SMC), and finally neointimal hyperplasia. Activated inflammatory cells secrete different mitogens that play a crucial role in SMC dedifferentiation, migration and proliferation. Modulation of this inflammatory response is an interesting target to decrease in-stent restenosis. Stent-based local drug delivery has been proposed to prevent in-stent restenosis. Recent clinical studies showed very promising results.3,4 In this work, we attempted to give more insight on the development of drug-eluting stents. Methylprednisolone (MP), a potent anti-inflammatory drug, was chosen as the model. The purpose of the study was to evaluate the effect of increasing total drug loading by using different coating methods and prolonged local drug delivery, using barrier coatings on the coated-stent induced inflammatory response and neointimal hyperplasia. A fluorinated polymethacrylate, polyfluoroalkylmethacrylate (PFM-P75), was developed as coating material for stainless-steel stents. The surface characteristics of different coated stents were examined. The efficacy of local methylprednisolone release to inhibit peri-strut inflammatory response and neointimal hyperplasia was evaluated in a porcine coronary model. METHODS Stents and stent coating. Custom-designed, balloon-expandable, stainless-steel coil stents were made of a 0.18 mm stainless-steel wire, folded in a zigzag shape over a 6 French tubular device.5 These stents can easily be mounted onto any conventional angioplasty balloon and deployed with a pressure of 6 atmospheres. Experimental work has demonstrated that these custom-designed stents can be safely deployed in porcine coronary arteries resulting in similar neointimal hyperplasia compared to the commercially available Palmaz-Schatz stent. Clinical experience also showed beneficial results.6 For the current study, 0.18 mm thick and 16 mm long stents were used. PFM-P75 was synthesized by radical polymerization of octafluoropentyl-methacrylate (75 mol%) and 2-ethylhexylacrylate (EHA, 25 mol%). Methylprednisolone was obtained from Sigma chemicals (Bornem, Belgium). For spray coating, stents were treated with a continuous airspray of the drug/polymer mixture to obtain a homogeneous film on the stent wires. For barrier coating, the spray-coated stents were treated with a 1% (g/v) PFM-P75 coating also applied by the spray coating technique. For the in vitro studies, PFM-P75 polymer mixed with 9%, 33% and 50% methylprednisolone was prepared and applied by spray coating. A 1% PFM-P75 barrier coating was added to the 50% methylprednisolone-loaded stent. For the in vivo studies, PFM-P75 polymer mixed with 10% and 50% methylprednisolone was coated on stent. The barrier coating with 1% PFM-P75 was added to both stents. The stents were sterilized by gamma irradiation (25 kGy) before implantation in porcine coronary arteries. The drugs showed chemical stability during the coating and sterilization processes. In vitro drug release. To measure the in vitro release, a methylprednisolone-loaded, PFM-P75 spraycoated stent or a barrier-coated, methylprednisolone-loaded PFM-P75 spray-coated stent was incubated in a 5 ml 0.1 M phosphate buffer solution (pH 7.4 at 37 °C). At regular time intervals, the medium was analyzed for the concentration of the drug by means of high-performance liquid chromatography (HPLC) by ultra violet (UV) detection. A Kontron 420-UV(432)-HPLC (Kontron AG, Zurich, Switzerland) equipped with a bio Sil C-18 column was used. The surface characteristics of the stents were examined by scanning electron microscopy (SEM). Stent implantation. Domestic cross-bred pigs of both sexes, weighing 20–25 kg were used. They were fed with a standard natural grain diet without lipid or cholesterol supplementation throughout the study. All animals were treated and cared for in accordance with the National Institute of Health Guide for the care and use of laboratory animals. A total of 30 pigs were used for this study. PFM-P75 spray-coated stents (PFM-P75), 10% (g/g) methylprednisolone-loaded PFM-P75 spray-coated stents (PFM-P75 + 10% MP), 50% (g/g) methylprednisolone-loaded PFM-P75 spray-coated stents (PFM-P75 + 50% MP), barrier-coated 10% (g/g) methylprednisolone-loaded PFM-P75 spray-coated stents (PFM-P75 x 2 + 10% MP) and barrier-coated 50% (g/g) methylprednisolone-loaded PFM-P75 spray-coated stents (PFM-P75 x 2 + 50% MP) were used. Each group included 6 pigs. Stent implantation in the right coronary artery and left anterior descending coronary artery was performed randomly according to the method described by De Scheerder et al.5,7 The arterial segment was selected to obtain a 1.2:1 stent to artery ratio. Four weeks after implantation, control angiography of the stented vessels was performed after administration of 0.25 mg of nitroglycerin. Pigs were sacrificed using an intravenous bolus of 20 ml oversaturated potassium chloride. For these follow-up studies, the instrumentation of the pigs and angiographic techniques were identical to those used during the implantation procedure. MEASUREMENTS Quantitative coronary angiography. Angiographic analysis of stented vessel segments was performed before stenting, immediately after, and at follow-up using the polytron 1000® system as described previously by De Scheerder et al.5,7 The polytron 1000® system was previously validated in vitro and in vivo with a metal bar as a calibration device. The diameters of the vessel segments were measured before, immediately after stent implantation and at follow-up. The degree of oversizing was expressed as measured maximum balloon size minus selected artery diameter divided by selected artery diameter. Recoil was expressed as measured maximum balloon size minus the minimal stent lumen diameter measured 15 minutes after stent implantation divided by measured maximum balloon size. SEM images of stented artery. Scanning electron microscopy was used to evaluate the presence of platelet adhesion, fibrin disposition, endothelium coverage and its maturity. Tissue segments for SEM were fixed in 2.5% glutaraldehyde and post-fixed in osmium tetroxide, both kept at PH = 7.2 with phosphate buffer. Subsequently, they were dehydrated through a series of increasing concentrations of acetone, and, finally, critical point dried with carbon dioxide and sputtered with gold. The tissue was examined in a Philips XL40 scanning electron microscope. Histopathology and morphometry. At 4-week follow-up, the pigs were sacrificed and the stented coronary artery was fixed using a 10% formalin solution at 80 mmHg. Coronary segments were carefully dissected together with a 1 cm minimum vessel segment both proximal and distal to the stent. The segment was fixed in a 10% formalin solution. Each segment was cut into a proximal, middle and distal stent segment for histomorphometric analysis. The part between the proximal and middle stent segment was harvested for SEM evaluation. Tissue specimens were embedded in a cold-polymerizing resin (Technovit 7100, Heraus Kulzer GmbH and Wehrheim, Germany). Sections, 5 microns thick, were cut with a rotary heavy-duty microtome HM 360 (Microm, Walldorf, Germany) equipped with a hard metal knife and stained with hematoxylin-eosin, masson’s trichrome, elastic stain and a phosphotungstic acid hematoxylin stain. Light microscopic examination was performed by an experienced pathologist who was blinded to the type of stent used. Injury of the arterial wall due to stent deployment was evaluated for each stent filament site and graded as described by Schwartz et al.8 Inflammatory reaction at every stent filament site was carefully examined searching for inflammatory cells, and scored as followed: 1 = sparsely located histolymphocytes around the stent filament; 2 = more densely located histolymphocytes covering the stent filament, but no lymphogranuloma and/or giant cells formation found; 3 = diffusely located histolymphocytes, lymphogranuloma and/or giant cells, also invading the media. Appearance of thrombus was evaluated for every stent filament on the phosphotungstic acid hematoxylin stained slides and graded as follows: 1 = small thrombus adjacent to the stent filament; 2 = more pronounced, covering the stent filament; 3 = big thrombus resulting in an area stenosis of 50%. Mean score = sum of score for each filament divided by the number of filaments present. Morphometric analysis of the coronary segments harvested was performed on 3 slices (proximal, middle and distal stent part) by using a computerized morphometry program (Leitz CBA 8000). Measurements of lumen area, lumen area inside the internal elastic lamina, and lumen inside the external elastic lamina were performed. Furthermore, area stenosis and neointimal hyperplasia area were calculated. Statistics. Arteriographic measurements before, immediately after and 4 weeks after stent deployment were compared using a paired t-test. For comparison among different groups, a non-paired t-test was used. Data are presented as mean values ± standard deviation. A p-value SEM images of the coated stents. The surface of PFM-P75 spray-coated stents was smooth (Figure 1A). 50% (g/g) and 33% (g/g) methylprednisolone-loaded PFM-P75 spray coating, however, resulted in an irregular stent surface (Figures 1C and 1D). This effect was not observed with the 9% (g/g) methylprednisolone-loaded PFM-P75 spray-coated stent (Figure 1B). Barrier coating of the 50% (g/g) methylprednisolone-loaded PFM-P75 spray-coated stent could dramatically decrease the surface irregularities observed after spray coating (Figure 1E). In vitro drug release. The total amount of methylprednisolone encrusted in 1 single spray-coated stent loaded with 9%, 33% and 50% methylprednisolone was calculated to be 100–150 µg, 400–450 µg, and 700–1000 µg, respectively. Figure 2 depicts the percent of release of methylprednisolone in function of time. Within 48 hours, 20%, 50% and 80% of the total amount of methylprednisolone was released from the 9%, 33% and 50% methylprednisolone-loaded PFM-P75 spray-coated stents, respectively. The barrier coating on the 50% methylprednisolone-loaded spray-coated stent could significantly slow down the methylprednisolone release. Within 48 hours, only 13% of the methylprednisolone was released (Figure 3). SEM evaluation of stented artery. The endothelialization of luminal surface was confirmed morphologically with SEM. At 4 weeks, the luminal surface was smooth and totally covered by endothelium. The majority of endothelial cells were spindle shaped and oriented in the direction of blood flow. Immature endothelial cells with bulging nuclei were present (Figure 4). No significant differences of endothelium coverage and endothelium immaturity, fibrin deposition and platelet adhesion were found among groups. In vivo studies. All stent deployment procedures were successful. All pigs had arteriographically patent arteries immediately after stent implantation. All pigs survived the follow-up period. Quantitative coronary angiography. Before stenting, the selected arterial segments of the groups were very similar in size. After the stent deployment, luminal diameter was also comparable. No significant difference in oversizing and recoil ratio was found among groups. At 4-week follow-up, the minimal luminal stent diameter and late loss were also not significantly different among the groups. Histopathology (Table 1). For each group, the targeted right coronary artery and left anterior descending showed an identical histopathologic response. Severe disruption of the internal elastic membrane and partial disruption of the external elastic membrane were found in the PFM-P75 group. Macrophages, lymphocytes and giant cells adjacent to the stent filaments were observed. In addition, extension of the inflammation process to the adventitia was also frequently found (Figure 5). The averages of proximal, middle and distal parts of stented segments were used for statistical comparisons. Mean inflammatory response scores of the stent filaments of the PFM-P75 group were significantly higher compared to the methylprednisolone-loaded groups [2.34 ± 0.73 for PFM-P75 versus 0.73 ± 0.62 for PFM-P75 + 10% MP; 0.52 ± 0.47 for PFM-P75 + 50% MP; 0.46 ± 0.54 for PFM-P75 x 2 + 10% MP; and 0.42 ± 0.32 for PFM-P75 x 2 + 50% MP (p Morphometry (Table 2). After 4-week follow-up, the lumen areas of the methylprednisolone-coated groups were larger than in the PFM-P75 group [1.34 ± 0.84 for PFM-P75 versus 1.92 ± 0.97 for PFM-P75 + 10% MP (p = not significant); 2.22 ± 1.26 for PFM-P75 + 50% MP (p = not significant); 2.21 ± 0.70 for PFM-P75 x 2 + 10% MP (p Stent-based local drug delivery.The advantage of stent-mediated local drug delivery is the direct delivery of a high dose of drug directly to the target organ, the diseased and injured vascular tissue. Polymers have been used as matrices for drug incorporation and elution. However, the biocompatibility of the polymers used, the insufficient drug delivery amounts and suboptimal local drug release period have limited their efficacy. Biocompatible polymers with a high drug capacity and slow drug release would be optimal candidates for stent-mediated local drug delivery. In this study, PFM-P75 was used to coat stainless-steel stents with methylprednisolone. PFM-P75 is characterized by an enrichment of fluorinated units at the surface, minimizing surface free energy. Dynamic contact angle (DCA) measurements confirmed the hydrophobic nature of the polymer. The addition of ethylhexylacrylate (EHA) reduces the polymer glass transition temperature and allows the production of very elastic materials. From competitive adhesion of blood protein experiments, PFM-P75 absorbs quickly and irreversibly a protein layer. It belongs to the group of polymers with high dispersion energy and low surface free energies, characteristics known to be more blood biocompatible than polymers with low dispersion and high surface free energy.9 However, in our porcine coronary coated-stent model, PFM-P75 induced an important inflammatory response around the coated stent struts, leading to the conclusion that this polymer like so many others is not sufficiently biocompatible. Spray-coated PFM-P75 stents produced a profound inflammatory response. Macrophages, lymphocytes and giant cells were present around the stent struts. Extensive inflammation was found from the intima to the adventitia. This inflammatory reaction resulted in an important neointimal hyperplasia. The excellent biocompatibility of PFM-P75 found in vitro testing could not be confirmed in the porcine coronary artery model. The inconcordance of in vitro and in vivo studies with polymer coatings was also reported in other studies.10 Coronary blood flow and the tissue environment are a major challenge for polymer materials. Stent coating using the dip coating technique has the advantage of resulting in a very thin and smooth coating. This minimizes the adhesion and activation of blood cells. However, this technique is hampered by the limited drug amounts that can be applied to the stent. Our studies with PFM-P75 dip coating found that for a 5% and 10% methylprednisolone/PFM-P75 solution, 10–15 µg and 20–25 µg of methylprednisolone, respectively, were released into a buffer medium in vitro.11 Although it is not known how much drug is needed to be locally released to obtain a significant inhibitory effect, these drug amounts with dip coating are quite small. To overcome this limitation, spray coating has been proposed. Our study showed that by spray coating, almost a hundred-fold total drug amount could be incorporated in a stent coating. Spray coating therefore offers the possibility to achieve much higher local drug concentrations. The release rate of methylprednisolone in vitro was dependent on the methylprednisolone/polymer ratio used. Respectively, 20%, 50% and 80% of the total amount of methylprednisolone was released from the 9%, 33% and 50% methylprednisolone spray-coated stents in 48 hours. In addition, when higher methylprednisolone concentrations were used, the surface of the coating became more and more irregular, related to the methylprednisolone concentration in the polymer used for spray coating. Therefore, higher drug amounts with spray coating resulted in a progressively more irregular coating surface and faster drug release from the stent. To overcome these disadvantages, a thin polymer film was added on the top of spray-coated stents (barrier coating). Adding the barrier coating decreased the surface irregularities of methylprednisolone-loaded PFM-P75 spray-coated stents. The barrier coating could furthermore dramatically slow down the drug release from 80% to 13% for 50% (g/g) methylprednisolone spray-coated stents in the first 48 hours. It was concluded that the use of spray coating enabled the use of much higher drug amounts. However, this resulted in a progressive irregular surface on the coated stents and fast methylprednisolone release. Adding a barrier coating could significantly slow down the drug release. Corticosteroid-eluting stents. Implantation of a coronary stent results in an important vascular injury and vessel wall overstretch. This activates the healing response. Circulating inflammatory cells adhere to the site of injury. The stent, being a foreign body, will further activate the inflammatory reaction. Inflammatory cells release various chemotactic and growth factors, playing a major role in vascular repair, but they are also responsible for neointimal hyperplasia. In animal models, the extent of inflammatory reaction determines the amount of neointimal formation.12 Clinical studies also demonstrated that inflammatory markers could predict the rate of restenosis and late complications in patients undergoing coronary angioplasty.13,14 A beneficial effect on neointimal hyperplasia has been demonstrated using agents that inhibit the inflammatory response after angioplasty.15,16 Corticosteroids potently inhibit monocyte and macrophage function, and may stabilize lysosomal membranes, thereby limiting cellular injury. The production and release of several soluble mediators of cell-mediated immunity, including macrophage aggregating factors and migration inhibitory factors, are markedly diminished by steroids. In addition, steroids have been shown to inhibit the formation of platelet activating factor and may exert an antiplatelet effect. Thus, steroids possess profound anti-inflammatory and in high doses even immunosuppressive effects.17,18 Three clinical randomized trials with corticosteroids have been performed on patients undergoing percutaneous coronary intervention. Single intravenous administration of methylprednisolone or double intramuscular methylprednisolone injections followed by oral prednisone administration for 7 days have not shown a positive effect on restenosis.19–21 Since continuous systemic administration of hydrocortisone did reduce the neointimal hyperplasia in a rabbit arterial injury model,22 the given dose and treatment duration could have contributed to the lack of beneficial effect in these clinical trials. Loading dexamethasone into polymer-coated stents has been shown to inhibit the inflammatory reaction, although the effect on neointimal hyperplasia was controversial. Strecker et al. showed positive results in the prevention of neointimal hyperplasia using a dog femoral artery model. However, the dose of dexamethasone in the study (8,000 µg/stent) was ten-fold that of another study, resulting in a negative result reported by Lincoff et al. (800 µg/stent) using a pig coronary model.23,24 Furthermore, Suzuki et al. reported that a 350 µg dexamethasone-loaded stent did not improve neointima or inflammatory score in a pig coronary artery model.25 In the present study, a normal endothelial surface appearance was observed in all groups. The lack of complete endothelialization of endoluminal prosthesis is a major factor for the development of intimal hyperplasia and late complications. Methylprednisolone released from the polymer matrix did not retard endothelial cell regeneration at the stented site. Histopathologic analysis showed that the inflammatory response was significantly lower in the groups treated with methylprednisolone-loaded stents. Introducing a barrier coating did not induce an increased inflammatory response. Morphometry showed that the lumen areas of the methylprednisolone-loaded stent groups were larger than the control group. Decreased neointimal hyperplasia in the methylprednisolone-loaded groups was observed as compared to the control group. Furthermore, neointimal hyperplasia of the groups with barrier coating was further decreased compared to the non-barrier-coated stents. It was concluded that methylprednisolone was able to abolish the inflammatory response induced by the PFM-P75 coating, resulting in a significantly lower inflammatory score and a decreased neointimal hyperplasia. The response to methylprednisolone was related to the dose used and the release time of the drug. Study limitations. In this study, normal porcine coronary arteries were used to evaluate the effect of stent-based methylprednisolone delivery on inflammation and neointimal formation. We used an artificial model with a biodegradable polymer that elucidated an important inflammatory response. In this model, local delivery of methylprednisolone effectively inhibited inflammation and neointimal hyperplasia. However, it remains uncertain whether this beneficial effect on inflammation and neointimal hyperplasia can be translated to human atherosclerotic coronary lesions. Since we used plastic-embedded anatomopathological samples, immunostaining was not performed in this study. Specific staining of smooth muscle cell markers, macrophages and other inflammatory cells may contribute to a better understanding and interpretation of the results. Despite these limitations, our study provides evidence that with spray coating plus a barrier coating, it is possible to load a high dose of methylprednisolone on a stent and obtain prolonged drug release from the stent. Furthermore, this study demonstrates a significant effect on inflammation and neointimal hyperplasia. Acknowledgments. We wish to thank T. Stassen and D. De Coux for technical assistance. Ivan De Scheerder is holder of the Andreas Gruentzig Chair for Interventional Cardiology, sponsored by Medtronic AVE.

1. Edelman ER, Rogers C. Pathologic responses to stenting. Am J Cardiol 1998;81(7A):4E–6E. 2. Moliterno DJ, Topol EJ. Restenosis and treatment. In: Topol EJ (ed). Textbook of Cardiovascular Medicine. Lippincott-Raven Press, Philadelphia: 1997. 3. Honda Y, Grube E, de la Fuente LM, et al. Novel drug-delivery stent: Intravascular ultrasound observations from the first human experience with the QP2-eluting polymer stent system. Circulation 2001;104:380–383. 4. Sousa JE, Costa MA, Abizaid AC, et al. Sustained suppression of neointimal proliferation by sirolimus-eluting stents: One-year angiographic and intravascular ultrasound follow-up. Circulation 2001;104:2007–2011. 5. De Scheerder IK, Wang K, Kerdsinchai P, et al. The concept of the home-made coronary stent: Experimental results and initial clinical experience. Cathet Cardiovasc Diagn 1996;39:191–196. 6. De Scheerder IK, Wang K, Kerdsinchai P, et al. Clinical and angiographic outcome after implantation of a home-made stent for complicated coronary angioplasty. Cathet Cardiovasc Diagn 1997;42:339–347. 7. De Scheerder I, Wang K, Wilczek K, et al. Experimental study of thrombogenicity and foreign body reaction induced by heparin-coated coronary stents. Circulation 95:1549–1553. 8. Schwartz RS, Huber KC, Murphy JG, et al. Restenosis and the proportional neointimal response to coronary artery injury: Results in a porcine model. J Am Coll Cardiol 1992;19:267–274. 9. Verweire I, Vansteenkiste S, Schacht EH, et al. Fluorinated polymers for biomedical applications synthesis and surface characterization. Doctoral Thesis, University of Ghent, Belgium. 10. van der Giessen WJ, Lincoff AM, Schwartz RS, et al. Marked inflammatory sequelae to implantation of biodegradable and nonbiodegradable polymers in porcine coronary arteries. Circulation 1996;94:1690–1697. 11. Huang Y, Verbeken E, Schacht E, De Scheerder I. Local drug delivery using drug eluting stents. In: Serruys PW, Leon MB, Colombo A, Kutryk MJB (eds). Coronary Lesions: A Pragmatic Approach. Martin Duntiz Ltd., London, 2002: pp. 319–335. 12. Rogers C, Edelman ER. Endovascular stent design dictates experimental restenosis and thrombosis. Circulation 1995;91:2995–3001. 13. Pietersma A, Kofflard M, de Wit LE, et al. Late lumen loss after coronary angioplasty is associated with the activation status of circulating phagocytes before treatment. Circulation 1995;91:1320–1325. 14. Gottsauner-Wolf M, Zasmeta G, Hornykewycz S, et al. Plasma levels of C-reactive protein after coronary stent implantation. Eur Heart J 2000;21:1152–1158. 15. Rogers C, Edelman ER, Simon DI. A mAb to the beta2-leukocyte integrin Mac-1 (CD11b/CD18) reduces intimal thickening after angioplasty or stent implantation in rabbits. Proc Natl Acad Sci USA 1998;95:10134–10139. 16. Feldman LJ, Aguirre L, Ziol M, et al. Interleukin-10 inhibits intimal hyperplasia after angioplasty or stent implantation in hypercholesterolemic rabbits. Circulation 2000;101:908–916. 17. Chervu A, Moore WS, Quinones-Baldrich WJ, et al. Efficacy of corticosteroids in suppression of intimal hyperplasia. J Vasc Surg 1989;10:129–134. 18. Fauci AS, Dale DC, Balow JE. Glucocorticosteroid therapy: Mechanism of action and clinical consideration. Ann Intern Med 1976;84:304–315. 19. Stone GW, Rutherford BD, McConahay DR, et al. A randomized trial of corticosteroids for the prevention of restenosis in 102 patients undergoing repeat coronary angioplasty. Cathet Cardiovasc Diagn 1989;18:227–231. 20. Pepine CJ, Hirshfeld JW, MacDonald RG, et al., for the M-HEART Group. A controlled trial of corticosteroids to prevent restenosis after coronary angioplasty. Circulation 1990;81:1753–1761. 21. Lee CW, Chae JK, Lim HY, et al. Prospective randomized trial of corticosteroids for the prevention of restenosis after intracoronary stent implantation. Am Heart J 1999;138:60–63. 22. Berk BC, Gordon JB, Alexander RW. Pharmacologic roles of heparin and glucocorticoids to prevent restenosis after coronary angioplasty. J Am Coll Cardiol 1991;17(Suppl B):111B–117B. 23. Lincoff AM, Furst JG, Ellis SG, et al. Sustained local delivery of dexamethasone by a novel intravascular eluting stent to prevent restenosis in the porcine coronary injury model. J Am Coll Cardiol 1997;29:808–816. 24. Strecker EP, Gabelmann A, Boos I, et al. Effect on intimal hyperplasia of dexamethasone released from coated metal stents compared with non-coated stents in canine femoral arteries. Cardiovasc Intervent Radiol 1998;21:487–496. 25. Suzuki T, Kopia G, Hayashi S, et al. Stent-based delivery of sirolimus reduces neointimal formation in a porcine coronary model. Circulation 2001;104:1188–1193.