| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Experimental Studies |
1 From the Departments of Diagnostic Radiology (J.M.A.M., K.S., A.B., F.M., A.K., R.W.G.), Nuclear Medicine (B.N., U.B.), Biometry (N.H.), and Textile and Macromolecular Chemistry (T.G.), University of Technology, Pauwelsstrasse 30, Aachen D-52074, Germany. Received April 6, 2002; revision requested June 17; final revision received February 25, 2003; accepted April 7. Supported by the Interdisciplinary Center for Clinical Research in Biomaterials and Tissue-Material Interaction in Implants (BMBF Project No. 01/KS 9503/9). Address correspondence to J.M.A.M.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: A self-expanding nitinol stent (30 mm in length, 8 mm in diameter) coated with a functionalized polymer layer was radiolabeled with 188Re. Fifty prostheses, 25 of which were radioactive (mean radioactivity, 20 MBq ± 3.8 [SD]) and 25 of which were nonradioactive, were implanted into the external iliac arteries of 25 sheep. Stent patency was assessed with angiography. Neointimal formation was assessed with intravascular ultrasonography and histologic examination 1 month (in all sheep) and 3 months (in 12 sheep) after implantation. The results were analyzed by using repeated-measures analysis of variance with two repeated factors and paired t tests for comparison at each measuring point.
RESULTS: All stents were placed successfully. Data in one animal had to be excluded from the study. After 3 months, a mean neointimal area reduction of 70 mm2 ± 55 (SD) was observed inside the radioactive stents, and a mean lumen reduction of 126 mm2 ± 39 was observed inside the nonradioactive control stents (P = .022). An edge effect was observed in the radioactive stents in that they showed an amount of neointimal formation at the edges that was similar to that seen in control stents. This neointimal formation accounted for the maximum lumen loss in the vascular segment with the stent.
CONCLUSION: As compared with a nonradioactive stent, a ß particle–emitting stent, through endovascular irradiation, significantly inhibits neointimal formation inside the stent but not at the stent edges.
© RSNA, 2003
Index terms: Animals • Arteriosclerosis, 984.721, 986.721 • Coronary vessels, stents and prostheses, 54.1269 • Dosimetry • Stents and prostheses • Stents and prostheses, radiation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Different irradiation techniques have been proposed, including brachytherapy with a radioactive source after stent implantation, external beam irradiation, and the use of radioactive stents (1,2,4,10–13). No late in-stent restenosis was observed in patients treated with phosphorus 32 (32P) radioactive ß particle–emitting coronary stents with initial radioactivity of 3–12 µCi (0.111–0.444 MBq) (14).
However, there was a high restenosis rate owing to lumen loss at the stent edges. The suspected reason was a reduction in radioactivity at the stent margins combined with high-pressure balloon injury (14). A clinical study in which stents with higher radioactivity levels (ie, 12–21 µCi [0.444–0.777 MBq]) were used and in which vessel overdilation was carefully avoided at stent placement did not reveal a solution to the problem of edge restenosis (15). To combat the edge effect, various modifications in technique have been proposed (16,17). One such modification is the use of a self-expanding radioactive stent.
We developed a method that allows dip coating of a self-expanding nickel-titanium alloy (nitinol) stent with radioactive rhenium 188 (188Re). Compared with 32P, 188Re is characterized by a shorter half-life (17 hours vs 14.3 days) and a higher mean ß energy (2.1 MeV vs 645 keV). The purpose of our study was to evaluate a self-expanding 188Re radiochemically labeled radioactive stent in sheep.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Stent Activation Process and Properties
Surface coating of the nitinol stent by means of chemical vapor deposition was performed (T.G.) to create a polymerized surface that provides functional NH2 groups for covalent immobilization of a chelating agent (18–21). Polymerization of 4-amino [2.2]-paracyclophane monomers to poly(amino p-xylylene) directly onto the stent surface was performed by using a novel modification of the chemical vapor deposition method. Chemical vapor deposition coating of nitinol stents resulted in completely closed polymer layers of about 50 nm in thickness being firmly attached to the metal stent surface. As determined at scanning electron microscopy, the integrity of these polymer layers was not affected by stent-loading and -expanding procedures. After integration of hexamethylene-diisocyanate as a spacer, the chelating agent diethylenetriaminepentaacetic acid was covalently bound by the functional NH2 groups (18,21) (Fig 1).
|
The radioactive solution consisted of 4 mL of 188Re perrhenate in 0.9% saline (200 MBq/mL in the in vitro experiment), 1 mL total of SnCl2 · 2 H2O (80 mg/mL) and sodium tartrate (16 mg/mL), and 50 µL of sodium acetate (100 mg/mL). A more highly concentrated solution with 260 MBq of 188Re perrhenate per milliliter was used in the animal experiment. The concentration was calculated, on the basis of results of the in vitro experiment and the physical decay rate, to achieve a radioactivity of 20 MBq on the stent. Expanded stents were incubated in this solution for 120 minutes at 85°C and washed four times in 20 mL of 0.9% saline. The radioactive stents were sterilized by their own radioactivity and with 80% ethanol. The control stents were sterilized in the usual fashion by the manufacturer. The radioactivity of labeled stents was measured with a calibrated radioisotope scintillator (CRC-15R; Capintec, Ramsey, NJ) directly before implantation.
Plastic Scintillator Dosimetry System
One radioactive stent was transferred for dosimetry to Essen University Hospital in Essen, Germany. The dose distribution for one radioactive stent was measured in a water phantom by using a plastic scintillation detector that was coupled to an optical fiber conveying the scintillation light to a photomultiplier tube (22,23). This dosimetry system allows three-dimensional measurement of the dose distributions in water of ß sources appropriate for use in vascular brachytherapy. The scintillator dosimeter system was calibrated in terms of absorbed dose rate to water with a precision of ±15% (2
) at the National Institute of Standards and Technology of Germany. This system fulfills the recommendations for dosimetry of cardiovascular brachytherapy sources published by Task Group 60 of the American Association of Physicists in Medicine (23).
Stability Testing of Chemical Fixation
Stability of radiolabeling was tested by using an in vitro blood circulation model (ie, a Chandler loop). Six radiolabeled 8 x 30-mm stents were placed in plastic rings of 20 cm in length and 7 mm in inner luminal diameter. The rings were filled with heparinized human blood (10 mL per ring) and mounted on a rotating device. Pulsating forward and backward rotations of 120°, with 60 rotations per minute, in a heated chamber (37°C) were performed for 48 hours. During the rotations, the stents were continuously covered with blood.
The plastic rings were placed on the collimator of a gamma camera (Multispect-3; Siemens, Erlangen, Germany) after 1, 4, and 48 hours so that we could measure radioactive washout from the stent in the blood. During image acquisition, the head of the gamma camera was rotated in a slightly oblique position (10° from the horizontal plane) to ensure separation of the fixed stent on one side and blood on the other side of the ring. Acquisition parameters were as follows: high-spatial-resolution collimator; matrix, 256 x 256; zoom, 1.0; photopeak, 155 keV (energy window width, 15%); and acquisition time, 15 minutes. Using regions of interest, we calculated the relative remaining stent radioactivity (%RSR) as an indicator of radioactive washout by comparing the total counts of the stent (tcstent) and those of blood (tcblood) as follows:
|
Animal Model and Implantation of Prostheses
A sheep animal model was used. The study was approved by the regional authorities. Regulations for animal care were met throughout the time of the experiments. Throughout the experiments, the animals were fully anesthetized with a standard technique (24).
In 25 male sheep weighing 35–57 kg (mean, 45 kg ± 5 [SD]), a total of 50 prostheses—25 radioactive stents with a polymer layer (mean radioactivity, 20 MBq ± 3.8) and 25 nonradioactive stents coated only with a polymer layer—were implanted into the external iliac arteries (J.M.).
Stents with the same polymer coating as in our study have been tested in a previous sheep study (Schuermann K and Lahann J, unpublished data, 1999): The Memotherm stents coated with the polymer showed no difference in neointima formation compared with the non–polymer-coated control stents.
Heparin (Liquemin N5000; Roche, Grenzach, Germany) (5,000 IU) was administered intraarterially before stent implantation and at each follow-up examination. No further systemic anticoagulation therapy was administered. One radioactive stent was placed in one external iliac artery, and one nonradioactive stent was placed in the contralateral external iliac artery—caudal to the origin of the deep circumflex iliac artery—in each animal. The side (left or right) in which the radioactive stent was placed was randomized.
Implantation of prostheses was achieved without complications. All sheep were followed up for 1 month and the remaining 12 for 3 months before they were sacrificed by administering an overdose of barbiturate (Narcoren; Merial, Hallbergmoos, Germany) to the fully anesthetized animals.
Sterilized polymer-coated and radioactive stents were mounted, by using a manually operated crimping device, onto the original delivery catheter, which contained as its only modification a removed tip for easier stent mounting. During the procedure, the operator was shielded by a lead brick wall. Each stent was measured before implantation to ensure that the correct radioactivity was applied. The stents were brought from the radiochemical laboratory to the animal laboratories; all laboratories were located in the same building.
After bilateral percutaneous femoral artery puncture, a standard 7-F introducer sheath was placed with the Seldinger technique. The cutaneous puncture site was 2–3 cm distal to the inguinal fold, which is easily recognized in sheep. A 7-F pigtail catheter with radiopaque markers was positioned above the aortic bifurcation, and digital subtraction angiography was performed. The diameter and cross-sectional area of the external iliac artery lumina were determined (K.S.) with intravascular ultrasonography (US) (Sonos Intravascular; Hewlett-Packard, Andover, Mass).
Before implantation, the lumen diameter was measured twice at 90° angles and the lumen area was determined in five equidistant steps below the origin of the deep circumflex iliac artery. Measurements were performed on screen with intravascular US software operated with a built-in trackball, and a hard copy of the results was printed. Intravascular 6.2-F 20-MHz US catheters (Sonicath; Boston Scientific, Watertown, Mass) were used. Fogarty balloon (EMB 80; Baxter, Irvine, Calif) denudation was performed in the external iliac arteries immediately before implantation with a standard arterial embolectomy catheter (Baxter). The Fogarty balloon was filled to its maximum liquid capacity (1.5 mL) with contrast medium (iopamidol, Solutrast 300; Byk-Gulden, Konstanz, Germany). According to the manufacturer, this corresponds to a nominal inflated balloon diameter of 11 mm. The true diameter of the balloon was slightly smaller when expanded in the artery owing to the flexibility of latex. This maneuver was repeated three times. No relevant vasospasm occurred during angiography. Angiographic and intravascular US examinations were repeated immediately after stent insertion. After removal of the introducer sheaths, hemostasis was achieved by manually compressing each puncture site for at least 20 minutes.
Venous blood samples (1.0 mL) were collected (A.K.) from five animals at 30 minutes and 24 and 48 hours after stent implantation. Measurement of radioactivity in these probes was performed (B.N.) with a well counter (FH 412; Frieseke & Hoepfner, Erlangen, Germany).
Scintigraphy of one sheep 4 hours after stent placement was performed by using a large-field-of-view gamma camera (E.CAM; Siemens). Acquisition parameters were as follows: high-spatial-resolution collimator; matrix, 256 x 256; zoom, 1.0; photopeak, 155 keV (energy window width, 15%); and acquisition time, 15 minutes.
Initial and Follow-up Examinations
Immediately after stent insertion and at each follow-up examination, 11 sets of intravascular US measurements were obtained in steps of 5 mm from point 1, located 10 mm proximal to the stent, to point 11, located 10 mm distal to the stent. The lumen area measurements were performed by one author (K.S.), who was blinded to the type of stent used. At each site, the patent and total lumen areas were determined. The total lumen area was defined as the patent lumen area plus any thrombotic or neointimal layer. Inside the stent, the total lumen area was measured as a circle adjacent to the inner margins of the stent struts. Outside the stent, the total lumen area was measured along the inner rim of the circular media layer. The patent lumen area was measured as the area free of echo inside the artery. Each area was measured twice.
Follow-up intravascular US and angiographic examinations were performed by means of a bilateral percutaneous femoral approach in the technique described above. Use of a standard 7-F introducer sheath was necessary for inserting the 6.2-F intravascular US catheter. Follow-up examinations were performed in all sheep at 1 month. In the group that survived for 3 months, angiographic and intravascular US follow-up examinations were repeated at 3 months.
Autopsy and Histologic Evaluation
Autopsy and preparation of specimens.—At autopsy, the external iliac arteries—including the distal abdominal aorta and the proximal segments of the femoral arteries—were exposed, major side branches were ligated, the distal abdominal aorta was cannulated antegradely, and the external iliac arteries were flushed with 0.9% saline solution. The external iliac arteries, including the distal abdominal aorta and proximal femoral arteries, were removed en bloc and fixed in buffered formalin (5%) for at least 2 weeks. Vessel segments with implanted stents were completely cut into cross-sectional specimens of 5 mm in width with a diamond blade saw (Exakt, Norderstedt, Germany). Autopsies and histologic evaluation were performed by one author who was blinded to the details of treatment (F.M.). Histologic specimens were embedded alternatingly in paraffin and in plastic. Standard histologic slices for hematoxylin-eosin and elastica-van Gieson staining were prepared.
Measurement of neointimal thickness.—The neointima of vascular prostheses is considered to be the layer that grows on the luminal side of the internal elastic membrane. Therefore, we defined the neointima as the tissue between the lumen surface and the inner border of the medial layer. A light microscope combined with computer-based analysis software (Quantimet 600; Leitz, Wetzlar, Germany) was used (F.M.) to perform the measurements.
Statistical Analysis
Data were expressed as means ± SDs. Results were analyzed by using repeated-measures analysis of variance with two repeated factors (stent and measuring point) and paired t tests for comparisons at each measuring point. A P value less than or equal to .05 was considered to indicate a statistically significant difference. Because of the explorative capacity of the P values, no 
adjustment was made. For comparing time points (ie, 1 month vs 3 months), additional models for each stent status with the repeated factors time point and measuring point were considered. All statistical analyses were performed by using SAS V8.2 software (SAS Institute, Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Dosimetry
Figure 2 shows two-dimensional dose rate profiles measured along the long axis of the 8 x 30-mm stent labeled with a radioactivity of 13 MBq 188Re at the time of dosimetric data acquisition. Dose rate profiles at various radial distances ranging from 0.5 to 3.0 mm from the stent surface are shown. The half-value thickness for 188Re was calculated as 0.81 mm.
|
Including correction for washout, infinite integration of the dose rate at a 1-mm radial distance from the stent surface is shown in Figure 3 for the 13 MBq 8 x 30-mm stent. Owing to the physical half-life of 188Re (16.98 hours), 95% of the accumulated dose was delivered to the target tissue within the first 72 hours. After an infinite time interval, a mean dose of 7.4 Gy ± 1.1 (SD) would be supplied to the soft tissue at a 1-mm radial distance from the stent surface.
|
Decay-corrected mean counts per second in venous blood samples (1 mL each) from five animals after stent implantation decreased from 2.30 ± 0.73 (SD) at 30 minutes to 1.07 ± 0.22 at 24 hours and to 0.21 ± 0.04 at 48 hours. Figure 4 shows scintigraphic images of a radioactive stent in vivo that were obtained 4 hours after stent placement.
|
Relative patent lumen area over time.—The percentage of patent lumen area to total lumen area calculated with the intravascular US measurements showed no decrease immediately after stent insertion (ie, patent lumen area and total lumen area both equaled 100%). One animal had severe clinical symptoms of unilateral arterial stenosis at day 2 after stent insertion. The left femoral pulse was not palpable. At angiography by means of right femoral access, the nonradioactive control stent on the left side was observed to be occluded. This animal was sacrificed after angiography, and its data were not included in the statistical analysis. None of the other animals had clinical symptoms of significant arterial stenosis. Mean patent lumen area and mean total lumen area for all animals were 111% ± 12 according to intravascular US data but were 100% at histologic examination owing to a phenomenon known as methodical shrinkage caused by histologic fixation of the tissue.
Within the 3 months of follow-up, a mild neointimal proliferation (mean area, 27 mm2 ± 36 at 1 month and 70 mm2 ± 55 at 3 months; P = .003) was observed inside the central part of the radioactive stents, while a greater neointimal proliferation (mean area, 61 mm2 ± 70 at 1 month and 126 mm2 ± 39 at 3 months; P = .081) was seen with the control stents (Fig 5). A reduction in neointimal area at each location inside the stent was observed at each time point during follow-up for the radioactive stents as compared with the control stents, except for point 5 (5 mm proximal from the center of the stent) at 3 months (Fig 6). The difference between radioactive stents and control stents was significant at each follow-up time point for all measurement points (P < .001 at 1 month, P = .041 at 3 months) and for each single measurement point in the inner part of the stent except for point 5 (5 mm proximal from the center of the stent) at 3 months (Fig 6). A maximum lumen reduction of more than 20% was observed in four of the 24 radioactive stents (in two followed up for 1 month and two followed up for 3 months). A maximum lumen loss of more than 20% was observed in 12 of the 24 control stents (in seven followed up for 1 month and five followed up for 3 months).
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
We developed a radiochemical method for on-site labeling of a self-expanding commercially available nitinol stent with 188Re. The process of activating 32P stents is completely different from the process of radiochemically labeling 188Re stents. The 32P stent is activated through ion implantation of 32P onto its metallic surface. This sophisticated method requires a reactor. The activated 32P stents have to be shipped from the reactor to the interventional center, with the risk of radioactivity loss owing to delays in transportation or implantation. In our study, 188Re stents were activated by radiochemical labeling within 2 hours before stent placement in a nuclear medicine laboratory. The simple radiolabeling procedure used in this study enabled us to succeed in fixing sufficient amounts of 188Re reproducibly and stably onto the stent surface and provides a useful tool for endovascular brachytherapy.
Washout from the radioactive stent, as determined in vitro with the Chandler loop, showed considerable release into the circulation during only the first 4 hours. These in vitro results were confirmed in the animal experiment, which revealed a decrease in blood radioactivity to nearly zero 48 hours after stent placement. Scintigraphic examination of one animal also revealed that radioactive release after 4 hours was negligible compared with background activity.
Dosimetric results revealed that, at a 5-mm distance from the stent surface, only 0.04 Gy ± 0.01 per megabecquerel of 188Re was supplied (with 2% of the dose supplied at a 1-mm distance). So, there was negligible contribution of the stent wire to the therapeutic dose delivered across the vessel to the opposite vessel wall. Our dosimetric results with the 8-mm-diameter stent can be extrapolated to 5–10-mm-diameter stents with negligible errors. Our dose measurements predict that peaks in dose distribution will be observed in areas adjacent to the stent wire.
Implantation of 188Re-labeled radioactive stents resulted in a significant reduction of in-stent neointimal formation compared with that seen in nonradioactive stents; however, an edge effect (with increased neointima immediately proximal to, distal to, and at both ends of the stent) was observed in the radioactive stents that was similar to that seen in nonradioactive control stents. Unfortunately, this new method for creating a radioactive stent did not enable us to overcome the major limitation of radioactive and nonradioactive stents: the high rate of restenosis at the stent edges.
Different types of radioactive stents for preventing in-stent restenosis are currently being evaluated, since results of animal studies have shown that 32P stents reduce the rate of restenosis after angioplasty (10–14,25). The use of 32P radioactive ß particle–emitting stents in patients with coronary arteriosclerotic disease has been shown to reduce intrastent neointimal hyperplasia in a dose-related manner (14,15). However, at the stent edges, intralesion restenosis was high. A higher dose combined with a nonaggressive stent implantation strategy did not solve the problem of the edge effect called the "candy wrapper effect" (14,15).
The pathophysiologic mechanism of stent edge restenosis is currently not completely understood. The edge effect could be secondary to low-dose radiation at the margin of the stent, be due to balloon injury at the stent edges, or stem from a combination of both factors (17). To combat the edge effect, various modifications have been proposed, including the use of (a) a square-shouldered balloon, (b) a "cold-end" stent, (c) a "hot-end" stent, (d) a hybrid radioactive stent and radioactive catheter–based system, (e) a 
particle–emitting stent, or (f) a self-expanding radioactive stent. Use of a cold-end stent has resulted in increased neointimal hyperplasia in the in-stent nonradioactive segments (26). The results of the present study indicate that the edge effect of the radioactive stents we used (which did not have cold ends) may be caused by a dose drop-off proximal and distal to the ends of the stents.
The major finding of our study is that neointimal formation was only moderately increased in the stent edge region compared with that observed in animal and patient studies that involved the use of 32P shunts (14–16) and was almost totally inhibited in the in-stent area. These results may be due to the different properties of 188Re (shorter half-life, higher ß energy) as compared with those of 32P.
Our results can be compared with those of the animal study conducted by Tepe et al (27), in which 186Re stents were used. The maximum ß energies of 32P (1.7 MeV) and 188Re (2.12 MeV) are much higher than that of 186Re (0.7 MeV), but both 186Re (3.8 days) and 188Re (17 hours) have shorter half-lives than 32P (14.3 days). The shorter application period of energy with 188Re and 186Re stents may be a reason for the reduced or absent edge stenosis observed when these stents are compared with 32P stents. Low-dose radiation at the stent margins combined with the low radiation over a long time period have been discussed for 32P stents (28,29). Only a few patients in the Isostent trial of 32P stents developed late thrombosis (28).
The release of 188Re raises the question of radiation exposure to the patient resulting from the released radioactivity. Kotzerke et al (30) performed dosimetric evaluations after intravenous administration of 188Re perrhenate in humans. On the basis of Medical Internal Radiation Dose Committee Dose Estimate Report No. 8, assuming the kinetic model for technetium 99m pertechnetate, the authors calculated an effective dose equivalent of 0.42 mSv/MBq for 188Re perrhenate, which decreased to 0.16 mSv/MBq after oral administration of 600 mg perchlorate. Therefore, a therapeutically effective stent labeled with 30 MBq would release approximately 9.3 MBq, resulting in an effective dose equivalent of 3.91 mSv without and 1.49 mSv with accompanied administration of perchlorate. Both effective dose equivalents are within the range of dose equivalents of standard diagnostic nuclear medicine procedures.
We used a sheep model for assessing the effect of radioactive stents on neointimal formation. This may limit direct comparison of our results with those of other animal studies of different radioactive stents, which primarily involved swine (12,13).
The effects of radioactive stents in the external iliac arteries of sheep can probably only approximate the effects of radioactive stents placed in human atherosclerotic iliac arteries. Edge restenosis is the major limitation of radioactive stents in clinical studies, a finding that was also observed in various animal models (such as swine or canine) for evaluation of radioactive stents (10–13). Therefore, at least some animal experiment results regarding neointimal formation in arteries should be applicable to humans.
Practical application: 188Re stents can be produced by chemical radiolabeling instead of complex activation in a reactor (the method used to produce P32 stents). The use of 188Re-labeled stents significantly reduced in-stent neointimal proliferation in a sheep model. Neointimal proliferation at the stent edges was not significantly different between radioactive and control stents. 188Re stents are useful as long arterial stents, in which edge stenosis has a relatively minor effect compared with the effect of the entire stent.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
] and 24 radioactive stents [
]) and (b) 3-month follow-up (in 12 control stents [