Author + information
- Received February 6, 2007
- Revision received March 26, 2007
- Accepted March 28, 2007
- Published online July 24, 2007.
- Sebastian Grundmann, MD⁎,1,
- Niels van Royen, MD, PhD⁎,1,
- Gerard Pasterkamp, MD, PhD†,
- Nieves Gonzalez, PhD‡,2,
- Edze J. Tijsma, PhD‡,2,
- Jan J. Piek, MD, PhD⁎ and
- Imo E. Hoefer, MD, PhD⁎,†,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Imo E. Hoefer, Laboratory of Experimental Cardiology, Room G02.523, University Medical Center, Heidelberglaan 100, 3584 CX Utrecht, the Netherlands.
Objectives The purpose of this study was to develop a cytokine-eluting stent to stimulate collateral artery growth (arteriogenesis) in the peripheral circulation of the rabbit via local transforming growth factor (TGF)-β1 release.
Background The stimulation of arteriogenesis with cytokines is a potential new treatment option for patients suffering from vascular occlusive diseases. However, the lack of a delivery platform for continuous intra-arterial application of pro-arteriogenic compounds has hampered the clinical implementation of this promising therapeutic strategy.
Methods Different polymer coatings were tested regarding their suitability for cytokine release. Fifty-four rabbits underwent implantation of bare-metal stents (BMS), polymer-only coated stents (PDLLA), polymer-coated TGF-β1–eluting stents (TGF) in the iliac artery, or bolus infusion of TGF-β1 and subsequent femoral artery ligation. Collateral artery growth was assessed with fluorescent microspheres, angiography, and histological quantification of the proliferation marker Ki67. In-stent neointima formation was measured in histological sections of plastic-embedded stents.
Results A TGF-β1–loaded coating based on poly(D,L-lactide) released up to 2.4 μg active TGF-β1 over a period of 24 h. Perfusion measurements revealed a significant increase in collateral conductance after TGF-β1 stent implantation compared with the control groups ([ml/min/100 mm Hg] BMS: 47.8 ± 2.5; PDLLA: 44.1 ± 3.9; TGF: 91.3 ± 32.6). Bolus infusion of TGF-β1 had no effect. Collateral arteries showed a higher proliferation activity in the TGF-treated group. At 7 days, no significant difference in in-stent neointima formation was observed.
Conclusions We first describe the use of a cytokine-releasing stent to stimulate collateral artery growth. These results show that intra-arterial cytokine-releasing devices might serve as a novel platform for the delivery of compounds affecting biological processes downstream of the site of implantation.
The development of a collateral circulation is an important natural defense mechanism that can partially or completely compensate for occluded arteries in the heart and the peripheral circulation. Unfortunately, in most patients the capacity of the collateral circulation falls short of the nutritional needs of tissue distal to the occluded arteries. However, experimental studies have shown that a naturally developed collateral circulation can be further stimulated to expand, resulting in the formation of larger collateral arteries that can improve tissue perfusion and restore organ function (1).
Stimulation of arteriogenesis is of potential benefit to the growing number of patients with peripheral or coronary artery disease that cannot be adequately treated by conventional measures like bypass surgery or percutaneous intervention.
Pro-arteriogenic factors display their largest effect when infused intra-arterially over a prolonged period of time (2,3). For practical reasons, in clinical trials these substances have usually been delivered either as intra-arterial bolus infusions (4), subcutaneous injections (5), or combinations of intra-arterial and intra-venous infusions (6). The limited success of clinical trials on pharmacological modulation of collateral artery growth so far might be explained by these suboptimal modes of delivery. Thus, a strong need exists for a clinically applicable intra-arterial delivery platform for pro-arteriogenic growth factors (7,8).
We previously demonstrated the pro-arteriogenic effects of continuous intra-arterial infusion of transforming growth factor (TGF)-β1 (9). This multifunctional growth factor is secreted by a large variety of cell types including endothelial cells and monocytes and is known to play a pivotal role during vascular proliferation (10). In the present study we now report the use of an endovascular stent coated with a bio-erodable polymer that continuously releases TGF-β1 into the arterial lumen for stimulation of arteriogenesis in a rabbit model of peripheral artery disease.
Polymer-coating and cytokine-loading of Cobalt-Chrome alloy stents
Prototypes were developed with 3 different polymers and evaluated in terms of amount of coating, durability, and release profiles. The polymers used in the coating preparation were poly(D,L-lactide) (PDLLA, MW 75000-12000, supplied by Sigma-Aldrich, St. Louis, Missouri), poly(ethylene-co-vinyl acetate) (PEVA, 60:40, MW 120000) and poly(n-butyl methacrylate) (PBMA, MW 180000) (both from Polysciences, Eppelheim, Germany), and rhTGF-β1 carrier free with a purity >97% (R&D Systems, Minneapolis, Minnesota). The stents were Medtronic Driver 15 mm medium vessel stents (3.0 mm) with a polyurethane based primer (Medtronic, Minneapolis, Minnesota).
A dip coating technique was developed and optimized for coating of stents with cytokine-loaded polymer. Coating was evaluated by SEM observation, and durability was tested after crimping the stents for 8 s at 40 psi and expanding them on a balloon catheter at 9 atm.
Testing of cytokine release and TGF-β–specific activity of stent eluate
In vitro release experiments were performed by incubating the coated stents in phosphate buffered saline (PBS) buffer with 0.1% bovine serum albumin at 37°C for time periods up to 7 days. Samples were taken periodically, and concentration of active TGF-β1 was determined by enzyme linked immunosorbent assay (ELISA) (Quantikine kits, R&D Systems).
To verify a TGF-β1–specific biochemical activity of the cytokine-polymer eluate, 2 TGF-β1 loaded stents were incubated in PBS buffer for 6 h at 37°C. Human umbilical endothelial cells (HUVECS) were cultured routinely (endothelial basic medium with 10% fetal calf serum and antibiotics) at passages 3-5, serum-deprived and stimulated with different dilutions of the stent eluate for 45 min. Phosphorylation of the TGF-family specific signaling proteins SMAD1 and SMAD2 in response to stimulation with stent eluate was measured in Western blot technique as previously described (11). rhTGF-β1 served as a positive control.
This study involved 54 male New Zealand white rabbits of 3 ± 0.25 kg bodyweight. All animal experimental protocols were approved by the local authorities and conformed to the “Position of the American Heart Association on Research Animal Use.” After induction of anesthesia, vascular access was obtained at the right femoral artery, and a 3.0 balloon, mounted with the appropriate stent, was advanced into the external iliac artery, where the stent was deployed between the branching of the internal iliac artery and the deep femoral artery. Stents were expanded with a single 30-s balloon inflation of 9 atm.
Animals received either bare-metal stents (BMS), polymer coated stents, or polymer coated stents loaded with TGF-β1 (n = 15/group) and a single oral bolus of 10 mg/kg clopidogrel before stent implantation. The site of vascular access in the superficial femoral artery was occluded by double ligation, and the animals were allowed to recover and to move freely for an observation period of 1 week. For bolus infusion, animals underwent femoral artery ligation only, and received a single bolus of 2.5 μg TGF-β1 proximal to the ligation site (n = 6). For investigation of local TGF-β1 plasma levels, animals received both a TGF-β1–eluting stent and a stent coated with the polymer only (n = 3).
In vivo TGF-β1 release
Blood samples were collected in ethylenediamine tetraacetic acid (EDTA)-containing tubes from the ear artery before stent implantation and 1 and 3 days after stent implantation. Plasma was obtained according to the manufacturer’s instructions to minimize platelet contamination, and concentration of active TGF-β1 was determined by ELISA technique. In the subgroup receiving both a TGF-β1–eluting stent and a polymer-only stent in the right and the left iliac artery, a 1-ml blood sample was withdrawn from the tibial arteries distal to the stent implantation sites 24 h after stent deployment.
Hemodynamic perfusion measurements
Collateral conductance measurements were performed 1 week after femoral artery ligation as previously described (12) in 6 animals/group. In brief, a pump-driven shunt between the left carotid artery and abdominal aorta was installed. The pressure gradient between the place of the aortic cannula insertion and the saphenous artery just above the ankle was measured invasively. A blood flow reference sample was withdrawn from the left femoral artery. With a roller-pump, 6 different pressure levels were maintained, and maximal vasodilatation was achieved by continuous adenosine infusion into the shunt system. At each pressure level, differently labeled fluorescent microspheres were injected. Tissue samples were processed for flow cytometric quantification of microspheres/sample. Flow to each tissue sample was calculated from the number of microspheres/sample. The resulting collateral blood flows were plotted against their respective pressure gradients. Collateral conductance equals the slope of the resulting curve of the 6 different pressure-flow relations.
In 3 animals/group, the abdominal aorta was cannulated and the hindlimbs were perfused with contrast medium based on bismuth and gelatin for X-ray imaging as previously described (13). Collateral arteries were defined according to Longland as vessels with a stem, midzone, and re-entry and marked after counting.
Flow cytometric analysis of monocyte activation by eluted TGF-β1
The effect of the eluted TGF-β1 on the expression of the leukocyte adhesion molecule Mac-1 was analyzed in flow cytometry. Whole blood samples from healthy volunteers were collected in heparinized tubes. After addition of either PBS, different dilutions of stent eluate, or rhTGF-β1 as positive control, samples were incubated for 2 h. Stent eluate was obtained as previously described. The buffer solution was then changed after 24 h to discard the first burst release of cytokine, and stents were again eluted for 1 h in 1 ml PBS. A dilution of 1:10 of stent eluate was used for full blood stimulation. After stimulation, the samples were stained for CD14 and the alpha-subunit of the activation marker Mac-1 (CD11b) as previously described (14) and analyzed in a flow cytometer.
Histological studies of collateral arteries
Immunohistological studies of rabbit collateral arteries 7 days after stent implantation were performed as previously described (14) in 6 animals/group. Frozen 5-μm tissue sections were fixed in ice-cold acetone, incubated with a cross-reactive mouse anti-rat antibody against Ki67 (Clone MIB-5, Dako, Glastrup, Denmark). Vascular smooth muscle cells (VSMC) were visualized with an FITC-conjugated antibody against alpha-smooth muscle actin (Sigma, St. Louis, Missouri). Perivascular macrophages were visualized with a cross-reactive antibody against CD68 (mouse anti-human, Clone EBM11, Dako). A Cy3-labeled anti-mouse antibody (Amersham Biosciences, Uppsala, Sweden) was used as a secondary reagent, and nuclei were stained with Hoechst 33342 dye (Molecular Probes, Eugene, Oregon).
Only cells with a positive CD68 staining and an identifiable nucleus were counted as macrophages. Vascular smooth muscle cell proliferation was quantified in a double staining for Ki67 and alpha smooth muscle actin as the percentage of Ki67+ VSMC. All cell counts were performed on collateral arteries in the vastus intermedius of the quadriceps muscle, where 2 defined collateral arteries span from the circumflex femoral artery to the re-entry vessels at the knee region. Imaging was performed on an Olympus (Lake Success, New York) BX60 microscope, and fluorescent images were adjusted (filter overlay, background subtraction) with analySIS software 3.0 (Soft-Imaging, Muenster, Germany).
Assessment of neointima formation
Seven days after implantation, stents were explanted, fixed in 4% formalin for 24 h, and processed through ethanol 50%, ethanol 70%, ethanol 96%, ethanol 100%, xylene, and methyl-methacrylate (MMA) (Merck, Darmstadt, Germany) for at least 1 h each. The samples were embedded in fresh MMA containing lucidol (Akzo Nobel, Arnhem, the Netherlands). Air was removed under vacuum pressure (1 h) with an exsiccator. Next, the catalyzer dimethyl-p-toluidine (Merck) was added, and the samples were mixed and left to polymerize in their containers in a 96% ethanol bath at −20°C. Full polymerization took 4 days, after which the samples were removed from their glass containers; 30-μm sections were cut from 3 segments of the stent, with a Leica SP1600 saw microtome (Leica Microsystem, Bannockburn, Illinois). Sections were HE stained. Neointima area was determined by subtracting the lumen area from the area encircled by the internal elastic lamina following the protocol of Schwartz et al. (15). Percent luminal stenosis was calculated as [1−(lumen/IEL area)] × 100.
Data are described as mean ± SD and analyzed with SPSS 12.01 (SPSS Inc., Chicago, Illinois). Differences between all treatment groups were assessed with analysis of variance with Bonferroni correction of all p values for multiple comparisons between independent groups. Bonferroni-adjustments were made for 6 comparisons for perfusion measurements (4 groups), for 3 comparisons for histological and angiographic analyses (3 groups), or for 36 comparisons for flow cytometry (9 groups). A p value <0.05 was considered to be statistically significant.
PDLLA-coated stents release up to 2.4 μg TGF-β1
A dip-coating method was optimized for coating of cobalt-chrome alloy stents, allowing cytokine loadings of several micrograms. Because the stability of TGF-β1 is critical, a cold coating technique was used, consisting in the immersion of the stents in chloroform solutions with polymer and TGF-β1 in a controlled rate and a following drying step. Solutions with different polymer concentrations were tested, achieving the highest amounts of coating with a 7% concentration (w/v). The amounts of coating obtained with this method were 0.47 ± 0.07 mg polymer at the second dip. A bio-erodable coating based on PDLLA released the largest amount of active cytokine compared with PEVA and PBMA-polymers in 24 h ([μg TGF released] PDLLA: 0.401 ± 0.003; PEVA: 0.26 ± 0.001; PBMA: 0.067 ± 0.001). Cytokine elution increased with the concentration of TGF-β1 in the coating solution ([μg TGF released in 24 h] 0.75% TGF: 0.401 ± 0.003; 1.2% TGF: 1.714 ± 0.074; 1.8% TGF: 2.461 ± 0.099).
Scanning electron microscopy showed smooth surfaces with limited bridging between stent struts and a satisfactory durability without fragmentation of the polymer surface for the PDLLA-coating after crimping and expansion of the stents (Appendix, Online Supplement 1).
For the animal experiments, the coatings were prepared with 7% PDLLA solution in chloroform with a concentration of 1.8% TGF-β. With a theoretical load of 8.5 μg TGF-β1/stent, almost 2.4 μg active TGF-β1 were released over a period of 24 h. Time course measurements demonstrated a strong burst release in the first hour, followed by the continuous elution of smaller doses during the first 24 h (Table 1).After 72 h, the amount of cytokine released decreased to <0.5 ng/h. Sterilization of coated stents by ethylenoxide exposure at a concentration of 819 mg/l during 75 min, a temperature of 50°C, and a following aeration step of approximately 12 h decreased the release of active cytokine to about 30% (data not shown). Therefore the preparation of the coated stents was performed under sterile conditions, and all equipment used in the preparation of coatings was treated with Etoxa-clean 1% solution (Sigma Chemical Co.). To confirm the sterility of the stents, 3 coated stents were incubated in bacterial growth medium for 7 days at 37°C. None of the samples showed bacterial growth. Eluate from 3 polymer-coated stents was tested for pyrogen contamination and did not induce an unspecific inflammatory reaction (Appendix, Online Supplement 2).
Eluted TGF-β1 is biochemically active and induces specific signaling events in endothelial cells
To verify the specific biochemical activity of the eluted cytokine we added stent eluate to the medium of cultured HUVECS. In endothelial cells, TGF-β1 balances the cellular activation state via 2 distinct TGF-β1 type 1 receptors in a dose dependent manner (11). Different concentrations of stent eluate resulted in a dose-response pattern on SMAD phosphorylation after a 30-min stimulation (Supplement 3) similar to the results of a previous study using rhTGF-β1 (11). With increasing concentrations of stent eluate, SMAD phosphorylation switched from a SMAD2/3 dominated response toward a SMAD1/5 dominated activated status of the cells. This functional dose dependent effect on signaling proteins responsive to the TGF-family of growth factors verified the specific activity of the eluted cytokine.
TGF-β1–eluting stents increase local but not systemic TGF-β1 plasma levels
Systemic TGF-β1 concentrations were measured before implantation and 24 h and 72 h after implantation of a TGF-β1–eluting stent. No significant differences in systemic TGF-β1 concentrations after implantation of a TGF-β1–eluting stent were observed (active TGF-β1 in plasma [ng/ml]: before stenting: 0.34 ± 0.07; 24 h: 0.31 ± 0.03; 72 h: 0.30 ± 0.02). In animals receiving both a TGF-β1 eluting stent as well as a polymer-only coated stent in 1 of the external iliac arteries, blood withdrawn from the tibial arteries contained a higher concentration of active TGF-β1 downstream of the TGF-β1 eluting stent (active TGF-β1 in plasma [ng/ml]: systemic: 0.31 ± 0.02: PDLLA: 0.31 ± 0.04; TGF: 0.43 ± 0.05; p = 0.035 for TGF vs. systemic, p = 0.027 for TGF vs. PDLLA).
TGF-β1–eluting stents stimulate collateral artery growth downstream of the implantation site
As the main end point of this proof of concept study, we measured the effect of TGF-β1–eluting stent placement on collateral artery growth distal to the site of stent implantation.
One week after stent implantation, combined pressure and flow measurements revealed a significant increase in collateral conductance in the TGF-β1–eluting stent treated group compared with the control groups receiving a bolus infusion of TGF-β1, a BMS, or a polymer-only coated stent ([ml/min/100 mm Hg] BMS: 47.8 ± 2.5; PDLLA: 44.1 ± 3.9; bolus: 49.2 ± 8, TGF: 91.3 ± 32.6, p = 0.002 for TGF vs. bolus, p = 0.002 for TGF vs. BMS, p = 0.001 for TGF vs. PDLLA) (Fig. 1).Collateral conductance in the bolus treated group and the 2 stented control groups (BMS and PDLLA) was comparable to the natural time course of arteriogenesis in historical control animals undergoing femoral artery ligation only (14). Perfusion measurements of the left hindlimbs, which were acutely occluded at time of measurement and represent the pre-existing collateral circulation, served as an internal control and showed no significant differences in baseline collateral conductance among the treatment groups ([ml/min/100 mm Hg]: bolus: 10.0 ± 1.3 BMS: 9.9 ± 0.5; PDLLA: 9.1 ± 1.5; TGF: 10.1 ± 1.3, p = NS for all comparisons).
The functional increase in collateral conductance of the treated right hindlimbs was accompanied by a higher proliferative activity of VSMC in the collateral arteries of the quadriceps muscle (Fig. 2A).Immunohistochemical quantification in 6 animals/group revealed a significantly higher percentage of VSMC expressing the proliferation marker Ki67 in the TGF-β1–treated animals (Ki67 positive VSMC: BMS: 18.9 ± 1.3%; PDLLA: 18.4 ± 2.3%; TGF: 27.5 ± 8.0%, p = 0.039 for TGF vs. BMS, p = 0.022 for TGF vs. PDLLA).
Angiograms showed an increase in visible collateral arteries as compared with baseline in all groups (Fig. 2B). Angiographic scoring demonstrated an insignificant trend (p = 0.15) toward more collateral arteries in the TGF-β1–treated group ([number collateral arteries] BMS: 14.3 ± 1.2; PDLLA: 14.7 ± 0.6; TGF: 16.0 ± 1.0).
Eluted TGF-β1 increases expression of the monocyte adhesion molecule Mac-1
Because it was previously shown that TGF-β1 activates monocytes, we tested the ability of TGF-β1 solution to increase monocyte adhesion molecule expression in flow cytometry. Incubation with rhTGF-β1 significantly increased expression of the Mac-1 adhesion molecule subunit CD11b in full blood of healthy volunteers in a dose dependent manner. The TGF-β1 eluted from polymer coated stents in a concentration of approximately 0.8 ng/ml after dilution induced a monocyte activation corresponding to the positive controls of 0.5 to 1 ng/ml rhTGF-β1 (Fig. 3).
Implantation of a TGF-β1 eluting stent stimulates perivascular macrophage infiltration around collateral arteries in vivo
Because infiltrating macrophages have been shown to be essential mediators of arteriogenesis, we quantified perivascular CD68+ cells around growing collateral arteries at day 7 after stent implantation (Fig. 4).Treatment with a TGF-β1–eluting stent resulted in a significantly higher number of perivascular macrophages compared with implantation of a BMS or a polymer-only coated stent ([CD68+ cells/mm2] BMS: 21.4 ± 6.5; PDLLA: 25.7 ± 2.3; TGF: 40.4 ± 6.4, p = 0.001 for TGF vs. BMS, p = 0.007 for TGF vs. PDLLA).
TGF-β1 loading of polymer coated stents did not increase short-term side effects
To assess short-term effects on in-stent neointima formation as a possible important side effect of cytokine loading, stents of 6 animals/group were processed for HE-staining. No case of stent thrombosis was observed, and neointima area did not differ significantly between the treatment groups ([mm2] BMS: 0.30 ± 0.10; PDLLA: 0.28 ± 0.11; TGF: 0.30 ± 0.13). This resulted in a comparable percent luminal stenosis in all 3 groups ([% stenosis] BMS: 7.5 ± 2.5; PDLLA: 6.5 ± 3.2; TGF: 6.0 ± 3.2) (Fig. 5).
In the present study we report for the first time on the successful use of an endovascular drug delivery platform for the effective release of a pro-arteriogenic compound. In vitro data show elusion of 2.4 μg active TGF-β1 over a time period of 24 h and a continuous release of smaller amounts in the first days after implantation. Whereas a single bolus infusion of the same dose had no effect on collateral artery growth in a rabbit hindlimb model of arteriogenesis, implantation of TGF-β1–releasing stent almost doubled the conductance of the collateral circulation after a 1-week treatment period as compared with control animals. This increase in collateral conductance is associated with a significant increase in smooth muscle cell proliferation in collateral arteries.
Stimulation of collateral artery growth or arteriogenesis was heralded as a promising new treatment option for large cohorts of patients with occlusive vascular disease (16). These high expectations stemmed from the positive results of experimental studies (17), which led to the initiation of several clinical trials. However, the majority of these studies did not achieve their primary goal, and the results remained disappointing (4,5). Most probably, sub-optimal delivery of compounds largely contributed to the failure of these trials (18). As it was shown in earlier experimental studies, intra-arterial delivery of factors to increase vascular growth is superior to other modes of administration (i.e., intravenous, intramuscular, or intrapericardial injection) (3). Furthermore, a prolonged exposure of the collateral circulation to the therapeutic agent is of critical importance (7). Optimally, the chosen factor for stimulation of arteriogenesis is delivered directly into a donor artery of the collateral circulation. Although technically feasible in the peripheral circulation, this is difficult to achieve in patients with obstructive coronary artery disease.
Drug-eluting stents seem to be an optimal platform to demonstrate the principle that pro-arteriogenic substances can be delivered directly into a donor artery of a collateral network. The improvement of materials and deployment technique has markedly reduced the threat of restenosis. Newer polymer-based and nonpolymer-based techniques permit the loading of stents with drugs in dosages up to several micrograms, comparable to dosages of cytokines that were shown to be effective in the stimulation of arteriogenesis in preclinical animal models (9,19). In the present proof of concept study, we used a coating based on the bio-erodable polymer poly(D,L-lactide), which was previously used for cytokine-loading of orthopedic implants (20) and showed the best combination of surface properties and cytokine release. Whereas intra-arterial bolus infusions expose the endothelium only for a few seconds to the therapeutic compound, implantation of a TGF-β1–eluting stent resulted in an elevation of local TGF-β1 plasma levels in the collateral circulation still detectable 24 h after stent implantation. After 72 h, the release decreased to <0.5 ng/h and can probably be considered insignificant. Although a more prolonged release of higher doses could have possibly resulted in a stronger pro-arteriogenic effect, the device used achieved an almost doubling of collateral conductance. Exposure time is an important determinant of TGF-β1 effects on endothelial cells (11) and could explain this difference between prolonged cytokine release from the stent and a single bolus infusion of the same dose TGF-β1. A stimulation of arteriogenesis by TGF-β1–eluting stents is also reflected on the anatomical level, resulting in an increased proliferation index of VSMC in collateral arteries. Although postmortem angiography detected a weak trend toward more visible collateral arteries, this difference was not statistically significant. However, conventional angiography shows poor correlation with gold-standard microsphere-based perfusion measurements and should not be used as a primary end point (21).
The TGF-β1 was previously shown to stimulate collateral artery growth via induction of the cardiac ankyrin repeat protein (carp) in collateral arteries (22) as well as via an activating effect on circulating monocytes (9). Indeed, in our study we found an up-regulation of the monocyte adhesion molecule Mac-1 in whole blood samples upon stimulation with TGF-β1, confirming previous results in isolated monocytes (9). Treatment with a TGF-β1–eluting stent resulted in a higher number of perivascular macrophages as key mediators of collateral enlargement. Whether this is a direct effect of TGF-β1 treatment or an indirect consequence of increased remodeling remains to be elucidated.
At this moment, stents eluting a pro-arteriogenic compound will not be considered for treatment of patients with noncomplex coronary artery disease. These patients will profit more from a direct revascularization of the arterial obstruction. Yet, significant percentages of the growing number of patients with multivessel disease are not candidates for standard revascularization or are not adequately revascularized by these approaches (16). These patients often undergo incomplete revascularization whereby re-opened vessels can serve as a donor artery for the collateral circulation to nonrevascularized perfusion territories. We envision that these patients are potential candidates for treatment with a stent that releases a pro-arteriogenic substance to improve blood flow to perfusion territories that were not adequately revascularized by conventional means.
One week after stent deployment, no significant differences in neointima formation were detected. However, neointima formation is incomplete at this time point, and our study was not designed to study in-stent restenosis as a primary end point. Also, potential effects on atherosclerotic plaque development could not be investigated in this animal model. Additional long-term studies are necessary to investigate these potential side effects of cytokine loading on stents. A treatment period of 1 week was chosen as a well established time frame to study collateral artery growth in the rabbit, because natural arteriogenesis in these young and healthy animals tends to mask beneficial effects at later time points (23). In addition, more detailed pharmacokinetic studies are necessary to further characterize the elution properties of our cytokine-eluting stent, because in vitro release models do not necessarily reflect the in vivo situation.
Although for this study we employed drug-eluting stents as a cytokine delivery platform because of their availability, our results might further stimulate the developmental research for other intra-arterial drug delivery devices that do not primarily aim at the dilation of a stenotic lesion. Meier et al. (24) recently described an attenuation of collateral artery function after placement of a sirolimus- or paclitaxel-eluting stent. Our study, now for the first time, shows the potential of endovascular stents to therapeutically modulate biological processes at sites distant to the location of stent placement. In addition, several cytokines have recently been shown to be able to positively affect cardiovascular pathophysiology such as coronary atherosclerosis (25) or ischemia/reperfusion injury. In these scenarios, where coronary stents are commonly employed, additional properties of local cytokine release as demonstrated here for the stimulation of arteriogenesis could significantly improve patient outcome.
To view the supplementary figures and the assessment of pyrogen contamination of coated stents, please see the online version of this article.
Coating surface properties
The coating surface was evaluated by scanning electron microscopy, before and after crimping and expanding the stents and revealed a satisfactory durability of the PDLLA-polymer coating.
Assessment of pyrogen contamination of coated stents
Activity testing demonstrated biochemical activity of eluted TGF-β1
Although for the coating release-profile, eluted cytokine was measured with an ELISA against the active 25kDa form of TGF-β1, we wanted to verify the preserved biochemical activity on vascular cells. On cultured HUVECs, the eluted cytokine induced a strong phosphorylation response of Smad signaling proteins, which are phosphorylated upon binding of TGF-β1 to its receptor. TGF-β1 balances the activation state of the endothelium via two distinct type 1 receptors, resulting in the phosphorylation of Smad1 and Smad5 or Smad2 and Smad3. Increasing concentrations of eluted cytokine altered the balance towards an Smad1/5 dominated signaling response, verifying specific biochemical activity of the eluted TGF-β1 on endothelial cells.
- Abbreviations and Acronyms
- bare-metal stent(s)
- enzyme linked immunosorbent assay
- fibroblast growth factor
- hematoxyline and eosine
- human umbilical endothelial cells
- molecular weight
- poly(n-butyl methacrylate)
- phosphate buffered saline
- poly(ethylene-co-vinyl acetate)
- recombinant human transforming growth factor β1
- vascular endothelial growth factor
- vascular smooth muscle cells
- Received February 6, 2007.
- Revision received March 26, 2007.
- Accepted March 28, 2007.
- American College of Cardiology Foundation
- Ito W.D.,
- Arras M.,
- Winkler B.,
- Scholz D.,
- Schaper J.,
- Schaper W.
- Rajanayagam M.A.,
- Shou M.,
- Thirumurti V.,
- et al.
- Lazarous D.F.,
- Shou M.,
- Stiber J.A.,
- et al.
- Simons M.,
- Annex B.H.,
- Laham R.J.,
- et al.
- van Royen N.,
- Schirmer S.H.,
- Atasever B.,
- et al.
- Seiler C.,
- Pohl T.,
- Wustmann K.,
- et al.
- Post M.J.,
- Laham R.,
- Sellke F.W.,
- Simons M.
- van Royen N.,
- Hoefer I.,
- Buschmann I.,
- et al.
- Goumans M.J.,
- Valdimarsdottir G.,
- Itoh S.,
- Rosendahl A.,
- Sideras P.,
- ten Dijke P.
- Hoefer I.E.,
- Grundmann S.,
- Schirmer S.,
- et al.
- Fulton W.F.M.
- Grundmann S.,
- Hoefer I.,
- Ulusans S.,
- et al.
- Schwartz R.S.,
- Edelman E.R.,
- Carter A.,
- et al.
- Zbinden S.,
- Zbinden R.,
- Meier P.,
- Windecker S.,
- Seiler C.
- Srivastava S.,
- Terjung R.L.,
- Yang H.T.
- Simons M.
- Boengler K.,
- Pipp F.,
- Fernandez B.,
- Ziegelhoeffer T.,
- Schaper W.,
- Deindl E.
- Hoefer I.,
- van Royen N.,
- Buschmann I.,
- Piek J.,
- Schaper W.
- Meier P.,
- Zbinden R.,
- Togni M.,
- et al.
- Cipollone F.,
- Fazia M.,
- Mincione G.,
- et al.