| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Experimental Studies |
1 From the Departments of Diagnostic Radiology (K.S., P.N., J.M., A.K, R.W.G.), Pathology (B.K.), and Textile and Macromolecular Chemistry (D.K.), University of Technology, Pauwelsstrasse 30, D-52057 Aachen, Germany; Harvard-MIT Divison of Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge (J.L.); and Department of Diagnostic and Interventional Radiology, Ingolstadt Clinic, Germany (D.V.). Received August 28, 2002; revision requested November 6; revision received January 16, 2003; accepted March 3. Supported by a grant from the Interdisciplinary Center for Clinical Research on Biomaterials ("IZKF Biomat") of the Medical Faculty of the University of Technology, Aachen, Germany. Address correspondence to K.S.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Physical stability, hemocompatibility, and cytotoxicity of the coating were first assessed in vitro. Thirty-six self-expanding nitinol (Memotherm), 24 stainless steel balloon-mounted (Palmaz), and 12 self-expanding nitinol (ZA) stents were coated with PHPX by using chemical vapor deposition polymerization. Seventy-two coated and 72 uncoated stents were placed into iliac arteries of 36 sheep. Sheep were classified into three groups of 12 animals each. In each group, six sheep were killed after 1 month; six, after 6 months. In each sheep, two uncoated stents were placed into one limb; two coated stents of the same type, into the opposite limb. In groups 1 and 2, Palmaz and Memotherm stents were used; in group 3, Memotherm and ZA stents were used. In groups 1 and 3, arteries were healthy. In group 2, arteries were pretreated with a Fogarty maneuver. Stent patency was measured with intravascular ultrasonography (US) and histologic analysis. Cellular response to coated and uncoated stents was assessed. Measurements were compared (Wilcoxon test).
RESULTS: In vitro, PHPX coating was stable; hemocompatibility and cytotoxicity were similar to those of stainless steel. In vivo, patency of coated and uncoated Palmaz and ZA stents was not different (P > .05). Patency of coated and uncoated Memotherm stents did not differ in four of six follow-up subgroups, but it was significantly reduced in group 2 after 6 months (intravascular US, P = .03; histologic analysis, P = .01) and in group 3 after 1 month (histologic analysis, P = .01). Histologically, the cellular response to coated and uncoated stents was not different (P > .05).
CONCLUSION: PHPX coating had good physical stability and biocompatibility in vitro and in vivo. Performance of coated and uncoated Palmaz and ZA stents was similar. Patency of Memotherm stents was similar in four of six follow-up subgroups. Materials effects did not result in severely enhanced neointimal hyperplasia.
© RSNA, 2003
Index terms: Animals • Arteries, grafts and prostheses, 986.1268 • Arteries, stenosis or obstruction, 986.721 • Experimental study
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The primary purpose of our study was to evaluate the biologic response to poly(hydroxymethyl-p-xylylene-co-p-xylylene) (PHPX)–coated stents in vitro and in vivo. The secondary purpose was to determine whether the coating provides a useful interface that can, with future study, be exploited for immobilization of drug molecules.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Basic physicochemical properties of the PHPX coating were determined from coated stainless steel foils (Goodfellow, Bad Nauheim, Germany). The chemical composition of the coating was in good accordance with expected chemical structures, as confirmed by using x-ray photoelectron spectroscopy and surface-sensitive infrared spectroscopy (24,25). The wettability of the PHPX coating was determined with the water-air contact angle according to the sessile drop method with pure water at room temperature (26). Measurements revealed that the coating was more hydrophilic (contact angle, 32° ± 4 [SD]) than bare stainless steel (contact angle, 42° ± 2). The zeta potential (electrokinetic potential), an indicator of the surface charge of a coating, was negative (-38 mV ± 4 at pH 7.0) (25,26). Atomic force microscopy indicated homogeneously distributed features with a diameter smaller than 50 nm. The PHPX coating was stable for several weeks in solvents such as ethanol, water, or hexane. Integrity of the PHPX coating was not affected by storage in water at a temperature of 70°C for 7 days, extraction in hexane and ethanol (79:21 wt/wt), or treatment in an ultrasound bath (120 W, 2 hours) at room temperature. Adhesion of the PHPX coating to stainless steel and nitinol was examined by pressing a 1 cm2 area of a piece of adhesive tape (Scotch; 3M, St Paul, Minn) onto the polymer coating. After the tape was peeled off, the sample was examined by using optical microscopy and infrared spectroscopy and was mechanically and chemically intact.
Delivery of coated stents.—PHPX-coated stents were reloaded to the delivery system by the manufacturers and were delivered in vitro. These stents included two types of self-expanding nitinol stents (Memotherm (n = 2), Bard-Angiomed, Karlsruhe, Germany; ZA (n = 2), Cook, Bjaeverskov, Denmark) (7). In addition, two stainless steel stents (Palmaz; Johnson & Johnson, Hamburg, Germany) were mounted on an 8-mm-diameter balloon and were expanded (1). Scanning electron micrographs (S 360; Leica, Bensheim, Germany) of the delivered and expanded stents were acquired.
For both the in vivo and in vitro studies, stents were 8 mm in diameter and 3 cm in length, except that the Palmaz stent was 2.9 cm long. The thickness of the PHPX coating was adjusted to be 200 nm ± 8%, as determined with a profilometer. For the measurement, a well-defined scratch was applied to the surface of a sample, and a stylus that was brought into direct contact with the probe followed height variations as the sample was moved. The height variations were converted into electrical signals, which produced a profile from which the thickness of the film was extracted.
Hemocompatibility testing.—Hemocompatibility testing included qualitative determination of platelet adhesion and semiquantitative determination of fibrinogen adsorption, activation of the complement system (factor C5a generation), and thrombin-antithrombin complex generation.
Blood or plasma samples used for the hemocompatibility tests were obtained from healthy human donors who had not ingested any medication for at least 2 weeks. Samples were obtained under the auspices of the Interdisciplinary Center for Clinical Research on Biomaterials of the Medical Faculty of the University of Technology ("IZKF Biomat"), Aachen, Germany. All donors had to undergo routine medical and serologic examinations to prove their health before they were included. To analyze cytotoxicity of the PHPX coating, standard cell culture experiments in direct contact and a 5-bromo-2'-deoxyuridine (BrdU) test (Boehringer, Mannheim, Germany) were performed. All samples included in the hemocompatibility and cytotoxicity testing and all stents implanted during the in vivo study were sterilized with ethylene oxide and stored for at least 3 weeks for degassing prior to use. All in vitro experiments were performed and/or supervised by two authors (J.L., D.K.).
Platelet adhesion.—The glutardialdehyde-induced fluorescence technique was used (27). In addition, scanning electron microscopy was performed. Briefly, circular uncoated (n = 6) and coated (n = 6) foils with a diameter of 25 mm were put into a self-designed incubation chamber (27). Polypropylene foils (Hüls, Marl, Germany), known to induce a strong platelet adhesion, served as positive controls. After the samples were rinsed with 0.9% saline for 1 hour, they were incubated for 30 minutes with 0.8 mL platelet-rich plasma (platelet concentration, 300,000/µL [300 x 109/L] ± 10%) at 37°C. The supernatant was removed. After the samples were washed with phosphate-buffered saline, they were fixed for 30 minutes at room temperature with 1 mL of 1.5% glutardialdehyde solution in phosphate-buffered saline.
For the glutardialdehyde-induced fluorescence technique, the samples were mounted on a glass slide, and one drop of fluorescent mounting medium (Dako, Hamburg, Germany) was added. Adherent platelets were studied with a fluorescence microscope (Leica). The excitation wavelength interval ranged from 450 to 490 nm; the emitted light was detected at a wavelength of 515 nm. For scanning electron microscopy, samples were washed with phosphate-buffered saline and dehydrated with ethanol in increasing concentrations from 20% to 100% in steps of 20% for 15 minutes. The samples were removed from the incubation chamber, vacuum dried for 24 hours, and further processed for scanning electron microscopy in the usual way. With both modalities, the number of adherent platelets was estimated, and their configuration (ie, whether they were round or discoid, which meant they were inactivated, or spreading and pseudopod forming, which meant they were activated) was qualitatively evaluated.
Fibrinogen adsorption.—Fibrinogen (Sigma-Aldrich, Steinheim, Germany) was adsorbed at PHPX-coated metal foils (n = 5) of 1.77 cm2. Tissue culture polystyrene (TCPS; Nunc, Roskilde, Denmark [n = 5]) was chosen as positive reference material. Rabbit antifibrinogen (Sigma-Aldrich) was used as the primary antibody in a dilution of 1:2,500 for 30 minutes. The secondary antibody was conjugated with the enzyme horseradish peroxidase. After reaction with the secondary antibody for 1 hour in a dilution of 1:10,000, the enzymatic reaction with diammonium-1,2-azinobis(3-ethyl-benzothiazolinsulfonate) (Sigma-Aldrich) was performed for 20 minutes. Finally, the absorbance was photometrically measured at a wavelength of 405 nm.
Activation of complement factor C5a and thrombin-antithrombin complex formation.—Blood samples were exposed to heparin (1 IU of heparin per milliliter of blood) prior to use. Uncoated (n = 3) and coated foils (n = 3) of 25-mm diameter and positive controls—glass for thrombin-antithrombin complex and regenerated cellulose (Cuprophane; Akzo Faser, Wuppertal, Germany) for complement factor C5a—were placed in the custom-made incubation chamber (27) and were incubated for 30 minutes at 37°C with 0.8 mL blood (complement factor C5a) or plasma (thrombin-antithrombin complex). The complement factor C5a and thrombin-antithrombin complex concentrations were determined relative to positive controls with commercially available enzyme-linked immunosorbent assay kits (TAT-Enzygnost Mikro, Enzygnost C5a Mikro; Behring, Marburg, Germany).
Cytotoxicity testing.—For cell vitality and cell morphology tests, uncoated (n = 3) and PHPX-coated foils (n = 3) were seeded with immortalized L-929 murine fibroblasts (German collection of microorganisms and cell cultures, Braunschweig, Germany) with a primary cell density of 50,000/mL and were incubated for 24 hours at 37°C. A biocompatible polyester foil (Biofolie; Heraeus, Tuttlingen, Germany) served as negative control; cytotoxic polyvinyl chloride, as positive control.
For cell vitality and morphology tests, 0.8 mL of cell medium was added after incubation, and incubation was continued for 24 hours. The supernatant was removed, and samples were incubated with fluorescein diacetate solution and ethidium bromide solution for 2 minutes. Fluorescein diacetate passes the cell membrane of vital cells and is hydrolyzed by cytoplasmic enzymes to yield fluorescein. Because of its polarity, fluorescein cannot pass the cell membrane, and it stains vital cells green (emitting a wavelength of 510 nm), whereas dead cells do not show fluorescence. Ethidium bromide can pass the cell membrane of vital cells only very slowly, whereas it can pass the membranes of dead cells easily because of reduced membrane integrity. Ethidium bromide forms a complex with DNA, which stains the nuclei of dead cells red. Samples were studied by using fluorescence microscopy.
Cell morphology was examined with light microscopy after formaldehyde fixation and hemalum staining (with Mayer hemalum solution) of incubated cells (J.L.). Cell morphology, density, and spreading on the test surfaces were compared. If a surface provides a favorable cellular environment, cells spread and the contact area with the surface increases, whereas if a surface provides an unfavorable environment, cells become spindle shaped and their contact area with the underlying material decreases.
The qualitative BrdU test is a colorimetric immunoassay that is based on the measurement of 5-bromo-2'-deoxyuridine (BrdU) incorporation during DNA synthesis (28). For this test, L-929 cells were cultured in the presence of uncoated (n = 3) and coated foils (n = 3) for 24 hours at 37°C. The positive control was a polyvinyl chloride foil. A biocompatible foil served as negative control. BrdU solution (10 µL, 0.1 mol/L) was added to the cell culture medium, and incubation continued for 2 hours. During this labeling period, the pyrimidine analogue BrdU was incorporated into the DNA of proliferating cells in place of thymidine. The labeling medium was removed, the cells were fixed, and the DNA was denatured by adding 200 µL of FixDenat solution (included in the test kit) that was removed after 30 minutes incubation at room temperature. Anti–BrdU-antibody solution (100 µL) was added that was labeled with a peroxidase that binds to the BrdU in newly synthesized DNA. Incubation was continued for 90 minutes at room temperature. Samples were washed three times with 300 µL of washing solution (included in the test kit). The substrate 3,3',5,5'-tetramethylbenzidin was added. Formation of the reaction product was stopped after 10 minutes by adding 25 µL of 1 mol/L of H2SO4. Samples were analyzed with fluorescence microscopy at 450 nm. Proliferating cells fluoresce green, whereas nonproliferating cells are neutral.
In Vivo Study
Study design.–All experiments were approved by the regional authorities (Bezirksregierung, Cologne, Germany) and were performed according to the national general guidelines for animal experiments. All interventions were performed with general anesthesia and a technique described elsewhere (7).
In sheep, there are no common iliac arteries. The abdominal aorta divides into two long external iliac arteries and a short trunk from which the internal iliac arteries branch off. The external iliac arteries are separated into a cranial and caudal segment at the origin of the deep circumflex iliac artery.
Stents were implanted, according to a fixed protocol, into the external iliac arteries of 36 male sheep (Merinolandschaf) with an average weight of 42 kg ± 10. Animals were divided into three groups of 12 animals each (Table 1). In group 1, Memotherm and Palmaz stents were placed into healthy arteries. In group 2, Memotherm and Palmaz stents were placed into arteries pretreated with a Fogarty maneuver. In group 3, Memotherm and ZA stents were placed into healthy arteries.
|
Follow-up examinations were performed in all animals after 1 month (n = 36), and in half of the animals after 6 months (n = 18). In each of the three groups, half of the animals (n = 6) were killed after 1 month, and the remaining half (n = 6) were killed after 6 months. This plan yielded an overall of six follow-up subgroups (groups 1—3, with 1- and 6-month follow-up subgroups). In the 1- and 6-month follow-up subgroups of each group, the distribution of stents to the four possible locations (right or left limb, cranial or caudal position) was equal.
Stent implantation.—Both femoral arteries were punctured retrogradely, and a 7-F sheath was introduced. Heparin (5,000 IU) was administered. A 7-F pigtail measuring catheter (Quanticor; Cordis, Haan, Germany) was advanced to the level of the aortic bifurcation, and digital subtraction angiographic images in the posteroanterior and the right and left anterior oblique view were obtained to determine and document the anatomy of the iliac arteries.
Prior to stent implantation, the diameter and the cross-sectional area of the iliac arteries were measured at three equidistant points above and at four equidistant points below the origin of the deep circumflex iliac artery with intravascular US (Hewlett Packard, Andover, Mass). All measurements were made freehand by one author (K.S.) with a built-in trackball.
In group 2 (pretreated arteries), a 5-F Fogarty balloon catheter (Baxter, Irvine, Calif) was positioned at the origin of the external iliac artery, inflated with 1 mL of a mixture of saline and contrast agent (Solutrast 300; Byk Gulden, Konstanz, Germany) in a ratio of 1:1, and pulled back to the femoral bifurcation three times. Palmaz stents were manually crimped onto a 4-cm-long balloon (Smash; Boston Scientific, Haan, Germany). The diameter of the balloon was chosen according to the result of the intravascular US measurements. If the diameter was smaller than 7 mm, a 7-mm-diameter balloon was used; if the diameter was larger than 7 mm, an 8-mm-diameter balloon was used. Palmaz stents were dilated to the measured diameter with a manometer syringe (Boston Scientific). All stents were implanted by one author (K.S.). Three-plane postimplantation angiographic images similar to the preimplantation angiographic images were acquired. Hemostasis was achieved with manual compression of the puncture sites. No systemic anticoagulation therapy was administered during follow-up.
Follow-up.—All follow-up examinations included angiography and intravascular US. After bilateral retrograde puncture of the femoral arteries, a 7-F sheath was placed. Heparin (5,000 IU) was administered. Three-plane angiographic images were obtained with a 7-F measuring catheter (K.S.). Three sets of intravascular US measurements, all performed by one author (K.S.), were obtained inside each stent: in the proximal segment within 5 mm of the proximal end of the stent, in the middle segment, and in the distal segment within 5 mm of the distal end of the stent. The patent and total cross-sectional areas were measured. The patent area was defined to be the cross-sectional area around the US probe that was hypoechogenic or without echogenicity, which indicated free flow. The total cross-sectional area was defined to be the entire cross-sectional area and included a neointimal layer on the luminal side of the stent struts that may have built up during follow-up. From both patent and total area measurements, the percentage of luminal patency, or PLP, was calculated according to the following equation: PLP = PL/TL · 100, where PL is patent lumen and TL is total lumen. After the final examination, the animals were killed with intravenous administration of a barbiturate (pentobarbital sodium, Narcoren; Rhone Merieux, Laupheim, Germany) and an overdose of 1 mol/L (7.45 g/100 mL) potassium solution.
Autopsy.—The external iliac arteries, distal abdominal aorta, and proximal segments of the femoral arteries were exposed, and major side branches were ligated. The distal abdominal aorta was antegradely cannulated, and the iliac arteries were flushed with 0.9% saline. Special focus was placed on the degree of periprosthetic soft-tissue thickening that was macroscopically estimated and compared between coated and uncoated stents. Exposed arteries were completely removed and fixed in buffered formalin for at least 2 weeks. Stents containing iliac segments were cut into cross-sectional specimens with a diamond blade saw (Exakt; Exakt Apparatebau, Norderstedt, Germany). Autopsy and cutting of specimens was performed by two authors (K.S., A.K.). Three specimens of about 5–7-mm thickness—one from each of the proximal, the middle, and the distal segments of each stent—were processed for en bloc embedding in polymethylmethacrylate. From each segment of each stent, two consecutive histologic slices were prepared for Giemsa staining.
Preparation of histologic specimens.—Standard histologic slices stained with hematoxylin-eosin were prepared from the segments between the specimens taken for embedding in polymethylmethacrylate. Stent struts were carefully removed. Scanning electron microscopic images of three randomly selected struts of each type of stent were obtained. Since tissue was adherent to the struts, even if the struts appeared to be clean after macroscopic removal of tissue, the struts were conditioned by means of incubation in aqueous NaOH solution (5 mol/L) for 7 days.
Neointimal area measurements.—A stereomicroscope (M3C; Wild, Heerbrugg, Switzerland) with computer-based measuring software (Quantimet 600 S; Leica) was used for area measurements from histologic slices. The neointima was defined to be the tissue that included the stent struts luminal to the media. A total of six slices per stent, two from each segment (proximal, middle, and distal), were analyzed. In each slice, the patent luminal area and the total luminal area (ie, patent luminal area plus neointimal area) were measured electronically after the border of the patent and the total luminal area were manually marked with a mouse pointer (7). Finally, the neointimal area was calculated with subtraction of the patent luminal area from the total luminal area. The two measurements per segment were averaged. All measurements were performed by one author (A.K.).
Histologic evaluation of the cellular response.—One randomly chosen Giemsa-stained and one hematoxylin-eosin–stained slice from the middle segment of each stent were evaluated in a blinded study. The analysis was focused to the vicinity of the stent struts because the cellular response was mainly confined to this area. A pathologist (B.K.) quantitatively assessed the cellular response by counting the number of macrophages, polymorphonuclear and eosinophilic granulocytes, myofibroblasts, epithelioid cells, and polynuclear giant cells by means of light microscopy with a 10 x 10 grid at a magnification of x200. Per slice, 10 square fields with an area of 400 µm2 each were analyzed. Fields were randomly chosen along the circumference of the stent. The number of cells was averaged per stent.
Statistical Analysis
Side-by-side comparison of intravascular US and histologic data about stent patency was conducted separately for each group (groups 1–3) and separately for each time of follow-up (immediately after stent placement [only intravascular US data], after 1 month and after 6 months [intravascular US and histologic data], and after 1 and 6 months combined [intravascular US and histologic data]) by using the Wilcoxon test. Intravascular US and histologic data of each stent segment (proximal, middle, distal) were included separately.
In the cellular response analysis, cell counts per stent were compared separately per group and type of cell (macrophages, polymorphonuclear granulocytes, and myofibroblasts). Measurements of the 1- and 6-month follow-up groups were combined. A difference with a P value less than .05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
Fibrinogen adsorption of coated samples was increased by about 50% compared with that of uncoated stainless steel foils (Table 2) but was reduced by about 25% relative to the reference tissue culture polystyrene.
|
Thrombin-antithrombin complex generation in the PHPX-coated samples was reduced by about 40% compared with that in stainless steel foils and by about 75% relative to the reference glass (Table 2).
In the direct contact cytotoxicity test, PHPX-coated foils were as densely populated with L-929 cells as the biocompatible foils; PHPX-coated foils were slightly more densely populated with L-929 cells than were stainless steel foils; polyvinyl chloride foils were not populated. Cells on PHPX-coated foils were round and had numerous pseudopods, which indicated enhanced spreading. Morphology and proliferation of L-929 cells on PHPX-coated foils were similar to those on biocompatible foils. Compared with findings with stainless steel foils, the number of spreading cells on PHPX-coated foils and their degree of spreading was mildly increased.
In the BrdU test, qualitative fluorescence microscopy showed that the vast majority of L-929 cells incubated with PHPX-coated and biocompatible foils had produced DNA—that is, proliferated. However, almost no DNA production was found in cells incubated with polyvinyl chloride foils. The number of proliferating green fluorescent cells was similar for the PHPX-coated foils and the biocompatible foils. In summary, both cytotoxicity tests indicated that the PHPX coating was nontoxic.
In Vivo Study
Follow-up with angiography and intravascular US.—Intravascular US measurements before stent insertion revealed no significant difference (P > .05) in the cross-sectional luminal area of the iliac arteries on either side. Immediately after stent placement, patency of uncoated and coated stents did not differ significantly (P > .05).
In animal 14 (group 2), thrombotic occlusion of an uncoated Palmaz stent occurred immediately after deployment for unknown reasons. An attempt at recanalization by simply moving a 0.0035-inch guide wire back and forth inside the stent failed, and the thrombosis did not resolve. The animal did not develop a gait disorder and was followed up for 1 month.
In animal 20 (group 2), the first three stents were placed without complications. The fourth stent, a coated Palmaz stent that was to be placed in the right caudal position, dislodged partly from the balloon catheter during insertion. Although the stent was finally placed correctly, thrombus had formed in both iliac arteries. The animal had to be killed 4 days later because of an ischemic gait disorder that resulted from bilateral thrombotic stent occlusion. Data about animals 14 and 20 were excluded from statistical analysis.
After 1 month and after 6 months, patency of uncoated and PHPX-coated stents was not significantly different in groups 1 and 3 (Tables 3, 4; Fig 2).
|
|
|
In animal 19 (group 2), more than 70% area stenosis in a coated Memotherm stent in the left caudal position was observed at 1-month follow-up. After 6 months, the stent was occluded. Histologic analysis proved that the occlusion was a result of excessive neointimal formation.
Autopsy and histologic analysis.—At autopsy, the periprosthetic soft tissue looked similar for all stents. Macroscopically, no signs of an enhanced inflammatory response to the PHPX coating were observed.
If 1- and 6-month results were evaluated separately for each of the three study groups, neointimal area did not differ significantly between PHPX-coated and uncoated nitinol stents in four of the six subgroups (Fig 3; Tables 5, 6). A significantly more marked neointimal area was found in PHPX-coated Memotherm stents in group 2 after 6 months (P = .01) and in group 3 after 1 month (P = .01). The difference between PHPX-coated and uncoated Memotherm stents was significant in all three groups, if 1- and 6-month measurements were combined (group 1, P = .04; groups 2 and 3, P = .004). There were no significant differences in neointimal area between uncoated and coated Palmaz and ZA stents (P > .05).
|
|
|
|
|
|
|
The cellular response to uncoated and PHPX-coated stents was not significantly different in the light microscopic analysis, regardless of the type of stent (Table 8). In uncoated and coated stents, the number of eosinophilic cells was negligible. Cells that indicate a major foreign body reaction (epithelioid cells, polynuclear giant cells) were not observed.
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
In this study, stents coated with PHPX were evaluated in vitro and in vivo. In vitro experiment results indicated that the PHPX coating has a biocompatibility comparable to that of stainless steel. While more marked platelet adhesion and fibrinogen adsorption suggest an enhanced thrombogenicity of the PHPX coating, the reduced thrombin-antithrombin complex formation and factor C5a activation point to the opposite. In vivo experiment results confirmed the in vitro results. The cellular response to uncoated and PHPX-coated stents showed no significant difference regardless of the type of stent used. PHPX-coated stainless steel and ZA nitinol stents had similar patency rates compared with those of uncoated stents. The reason for the reduced patency of PHPX-coated Memotherm nitinol stents in two of six subgroups is not clear. It may be hypothesized that there is an unknown unfavorable interaction between the Memotherm stent and the PHPX coating that promotes neointimal growth. An interference of the PHPX coating with the nitinol wire of the Memotherm stent appears to be very unlikely.
The overall biocompatibility of PHPX-coated stents was regarded as sufficient, considering the fact that many other experimental polymer coatings caused local inflammation and high stent restenosis rates (17–19,31). van der Giessen and co-workers (17) tested five biodegradable polymers (among others polyglycolic acid/polylactic acid copolymer and polycaprolactone) and three nonbiodegradable polymers (polyurethane, silicone, polyethylene terephthalate). Polymer samples were deployed as stripes (thickness, 75–125 µm) longitudinally across 90° of the circumferential surface of coronary stainless steel stents in a pig animal model. All polymers induced a marked inflammatory reaction and restenosis rate. De Scheerder and co-workers (19,31) evaluated a biodegradable poly(organo) phosphazene coating and a nonbiodegradable polyurethane coating (thickness, 23 µm) of stainless steel stents in the iliac and coronary arteries of pigs. Stents coated with the biodegradable polymer were afflicted with a higher inflammatory response and higher rate of restenosis than were uncoated stents. Patencies of polyurethane-coated stents and uncoated stents were similar. Lincoff and co-workers (18) compared uncoated stents, polylactic acid (PLA)—coated stents, and PLA-coated dexamethasone-eluting tantalum stents that were implanted in the coronary arteries of pigs. PLA coatings of either low- (80 kDa) or high-molecular-weight (321 kDa) PLA (thickness, about 20 µm) were used. An intense inflammation and a reduced patency rate were encountered in response to the low-molecular-weight PLA-coated stents and—less pronounced—to the low-molecular-weight PLA-coated stents eluting cortisone. High-molecular-weight PLA-coated stents were comparable with uncoated stents.
However, findings of most recent reports about polymer-coated vascular stents indicate that stent-coating techniques have improved (8,9,12). It has become clear that not only the polymer but also the way it is applied is important. Polycaprolactone and low-molecular-weight PLA stent coatings that produced poor results in the previously mentioned experiments were successfully used in recent studies. Alt and co-workers (8) reported their findings in regard to a very-low-molecular-weight (30 kDa) PLA-coated stent (thickness of the coating, 10 µm) that released hirudin and iloprost. The stent was evaluated in vitro (32) and in vivo in sheep and pig animal models and demonstrated no enhanced inflammatory response and superior patency rates compared with those of uncoated stents. The authors attributed this finding to an increase in stability and uniformity and a decrease in thickness of the PLA coating that resulted in an overall enhanced biocompatibility (33). Similar observations were made by another group of investigators for a rabbit animal model. The coated stents that were implanted in the iliac arteries released paclitaxel from a coating (thickness not given) with a biodegradable poly(lactide-co-3-caprolactone) copolymer (9). Finally, Sousa and co-workers (12) incorporated the drug sirolimus in a polymer coating of stents; the coating consisted of a mixture of nonerodible poly-n-butyl methacrylate and polyethylene–vinyl acetate copolymer. Sirolimus-eluting coronary stents demonstrated an enhanced patency rate in an animal experiment in pigs and in a clinical study.
This study had several limitations. In vitro experiments were mostly performed with coated stainless steel foils but not with coated nitinol foils and not with stents. Findings of stent expansion experiments demonstrated that the coating was stable and completely covered the stent struts without changing the original stent wire mesh design. Therefore, the physicochemical and biologic behavior of coated foils and that of coated stents was considered similar, regardless of the underlying metal backbone. However, the divergent in vivo performance of the Memotherm and the ZA stents makes this consideration questionable.
Destruction of the PHPX coating found in some samples was attributed to the NaOH treatment. It is impossible to prove in retrospect that the destruction did not occur during the follow-up period. On the other hand, the NaOH solution was so aggressive that it even dissolved some of the metal stent strut samples, and in vitro test results indicated excellent stability of the coating.
Two stents were inserted into the same artery. However, in a previous study it was proved that neighboring metal stents in the same iliac artery do not interact with regard to stent patency and neointimal hyperplasia (34). Placement of four instead of two stents per animal reduced the number of animals needed.
The sheep is a less common animal model for evaluation of endovascular devices than is the pig. Nevertheless, it was proved that the fibrinolysis and coagulation system of sheep is more similar to that of the human system than it is to that of the pig and most other species used for animal experiments in the vascular system (35,36). Our long-standing experience with the sheep animal model has shown that it is as feasible as pig, canine, or rabbit animal models. In addition, it must be pointed out that, regardless of the animal model used, the transfer of results to the human is generally limited (37).
Practical application: The PHPX coating provides a biocompatible stent coating with high mechanical stability, biologic and chemical inertness, homogeneity, and minimum thickness, with complete coverage of the stent wires without disturbance of the original stent design. Further improvements of its biocompatibility are possible through bonding of drugs to or releasing of drugs from the PHPX coating. Additional studies are warranted.
|
FOOTNOTES |
|---|
Abbreviations: BrdU = 5-bromo-2'-deoxyuridine, PHPX = poly(hydroxymethyl-p-xylylene-co-p-xylylene), PLA = polylactic acid
Author contributions: Guarantors of integrity of entire study, K.S., J.L.; study concepts, K.S., J.L., D.V., D.K.; study design, K.S., J.L.; literature research, K.S., J.L., P.N.; experimental studies, K.S., J.L., J.M., P.N.; data acquisition, K.S., J.L., A.K.; data analysis/interpretation, K.S., J.L., B.K.; statistical analysis, K.S., J.M., J.L.; manuscript preparation, K.S., J.L., B.K.; manuscript definition of intellectual content, K.S., J.L., D.V., D.K. ; manuscript editing and revision/review, K.S., J.L., R.W.G.; manuscript final version approval, K.S.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. Tang, X. Chen, S. Tang, T. LaLonde, and J. M Gardin Granulation encapsulated stent: a new therapeutic approach for vascular implantation Heart, February 1, 2007; 93(2): 238 - 243. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. W. Hamer, T. Finkenzeller, I. Borisch, C. Paetzel, N. Zorger, S. Feuerbach, and W. Nitz In Vivo Evaluation of Patency and In-Stent Stenoses After Implantation of Nitinol Stents in Iliac Arteries Using MR Angiography Am. J. Roentgenol., November 1, 2005; 185(5): 1282 - 1288. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
