Vascular and Interventional Radiology
Changes in Wall Mechanics After Endovascular Stenting in the Rabbit Aorta
Comparison of Three Stent Designs
Audio Available | 
Share
OBJECTIVE. The aim of this study was to evaluate changes in the wall mechanics of small-diameter arteries after endovascular placement of three different stents.
SUBJECTS AND METHODS. Five self-expandable stents (Wallstent), five balloon-expandable noncovered Palmaz stents, and three balloon-expandable covered stents (Jostent) were placed in the infrarenal aorta of 13 New Zealand white rabbits. Systolic blood pressure changes, blood-flow velocity, systolic diameter, and diameter changes were measured and used to calculate the diameter compliance, the distensibility coefficient, and the pulsatility index.
RESULTS. Compliance (10-3 mm kPa-1) was 75.3 ± 20.1 before stenting and reached 94.7 ± 42.2 upstream, 38.8 ± 14.2 at the stent level (p < 0.05), and 70.8 ± 23.2 downstream from the stent. Distensibility (10-3 kPa-1) was 24.3 ± 6.3 before stenting and reached 27.8 ± 10.3 upstream, 10.5 ± 4.4 at the stent level (p < 0.001), and 21.9 ± 8.6 downstream from the stent.
Compliance and distensibility were significantly lower at the stent level than upstream and downstream (p < 0.05). Aortic diameter increased significantly at the stent level from 3.11 ± 0.40 mm before to 3.76 ± 0.42 mm after stenting. No significant difference was found among the three stent designs for all the studied data.
CONCLUSION. Regardless of the three tested stent designs, endovascular stenting produces a significant decrease in arterial wall compliance of the rabbit aorta.
| Introduction |   |
|---|
Currently, angioplasty is the first choice treatment for arterial stenosis and short occlusions. Stenting improves both initial and long-term results of angioplasty but does not avoid restenosis [1, 2]. The native-artery diameter appears to be a major factor in restenosis, for which the rate is dramatically higher in small-diameter arteries. For instance in coronary arteries, the actual estimated rate of restenosis is approximately 30% [1,2,3]. Compliance mismatch between native arteries and synthetic vascular substitutes is often reported as the main possible contributing mechanical factor of graft failure, especially in small-caliber arteries [4,5,6,7,8]. Abrupt compliance change and concomitant nonlaminar flow patterns may contribute to the early accumulation of thrombus and later neointimal thickening [4, 7, 8]. The number of stent designs available to improve feasibility and to decrease immediate and late restenosis has increased recently [9]. Presently, components, size and geometry, and method of implantation characterize stents. In vitro evaluation of the mechanical characteristics of stents has been extensively reported [10,11,12,13]. Unfortunately, in vivo measurement of the mechanical properties of small-caliber arteries and of interposed vascular substitutes or endoprostheses still requires invasive procedures and cannot be performed in humans. Less in vivo experimental data are available regarding the mechanical properties of stented arteries [14, 15]. The research of Rolland et al. [14] in 1999 described impressive results concerning hemodynamics and arterial wall mechanics in large-diameter arteries after stenting (swine iliac arteries). These researchers showed marked, but varying or even opposite, changes depending on the stent design. Because the stent-wire caliber does not decrease linearly with the stent diameter, the consequences of stenting on arterial wall mechanics may be expected to be greater in small-diameter arteries. Therefore, the main goal of our study was to evaluate the compliance mismatch between native and stented parts of small-diameter arteries. Our secondary goal was to compare three different stent designs: the balloon-expandable rigid Palmaz stent (Johnson & Johnson, Warren, NJ), the self-expandable flexible Short Magic Wallstent (Schneider, Bülach, Switzerland), and the balloon-expandable covered Jostent (Jomed, Rangendingen, Germany).
| Subjects and Methods | |
|---|
Thirteen New Zealand white male rabbits weighing 3-4.3 kg were preanesthetized with ketamine (50 mg/kg of body weight). Anesthesia was maintained with IV infusion of ketamine in a 0.9% saline solution (0.1 mg/kg per minute).
The rabbits received an IV injection of 200,000 U of penicillin G, 0.4 mg/kg of dexamethasone, and 500 U of heparin. The rabbits were intubated but breathed spontaneously and were placed in a supine position on a heating blanket. After subcutaneous injection of lidocaine (Xylocaine; Roger Bellon, Neuilly-sur-Seine, France), the rabbits underwent right femoral arteriotomy, and a 5-French introducer sheath with check-valve and side-arm (Radiofocus; Terumo, Tokyo, Japan) was introduced and placed in the abdominal aorta. The sidearm of the introducer was used to measure arterial blood pressure. The ECG and internal temperature were also monitored. Using a right transverse laparotomy, we exposed the infrarenal aorta, and the periaortic tissue was carefully dissected to place the sonographic transducers for diameter and blood flow measurement on the aorta. A 3.5 Short Magic Wallstent adapted to a mean vessel diameter of 3 mm was implanted in five rabbits, a balloon-expandable Palmaz stent with a 3- to 5-mm diameter and 7-mm length was implanted in five rabbits, and a balloon-expandable Jostent coronary stent graft with a 2.5- to 5-mm diameter range and 9-mm length was implanted in three rabbits. In all cases, the diameter of the stent and of the balloon catheter was adapted to the previously measured infrarenal aorta: a 3-mm diameter (n = 6) or a 3.5-mm diameter (n = 2). A Skinny dilatation catheter (Scimed, Maple Grove, MN) inflated at 6 atm was used to expand the eight balloon-expandable stents. Via the right femoral introducer, the delivery system of the Wallstent or the balloon-expandable stent mounted on the dilatation catheter was introduced over a 0.014-inch-diameter wire, and the stents were deployed in the infrarenal aorta with visual and biomicroscopic control. Repeated diameter and flow measurements were then obtained.
Diameter and systolic diameter changes were measured in the 13 rabbits with a 20-MHz emitting frequency 1-mm-diameter ultrasound echo tracking microprobe (DMT201N; Crystal Biotech, Hopkinton, MA) mounted on a laboratory-made silastic support connected by electric wire to an eight-channel sonometry system (CBI 8000, Crystal Biotech) with a WT20 wall-tracker module and a HVPD-20 pulsed Doppler module (Crystal Biotech). Spatial resolution of the wall-tracker module, according to the manufacturer's specifications, is 1/8 of the wavelength (i.e., 0.0077 mm at 20-MHz operating frequency). ECG, blood pressure, diameter, and diameter changes were digitized with an MP 100 WSW acquisition module (Biopac Systems, Goleta, CA) and processed with dedicated software (AcqKnowledge version 3.5, Biopac Systems) on an IBM-compatible computer.
Aortic blood-flow velocity was measured in 12 rabbits with a 20-MHz pulsed Doppler probe. The mean arterial pressure was calculated as the time averaged-mean arterial pressure. The pulsatility index was calculated as P/M (P = total amplitude of the velocity curve, M = time averaged—mean velocity). Diameter compliance was calculated as 2Δd/ΔP (ΔP = systolic — diastolic arterial blood pressure or systolic pressure change) [10]. The distensibility coefficient was defined as 2Δd/ΔP/d. Measurements were obtained before and after stent placement, 3 mm upstream, at the stent level, and 3 mm downstream from the stent as shown in Figure 1.
 View larger version (16K) | Fig. 1. —Drawing shows location of diameter and flow measurements on rabbit aorta, at stent level (S), upstream (U), and downstream (D) from stent. |
Descriptive statistics hereafter concern the pooled data of the 13 rabbits, unless significant differences among groups were found. Aortic diameter, compliance, distensibility coefficient, and pulsatility index change after stenting, and differences among measurement sites (respectively upstream, at the stent level, and downstream from the stent) were evaluated with the paired t test. We compared the three stent-design groups by analysis of variance with a Bonferroni test (Prism version 3.0; GraphPad, San Diego, CA) (p values < 0.05 were considered significant).
The animal care complied with the “Principles of Laboratory Animal Care,” as formulated by the National Society for Medical Research, and the “Guide for the Care and Use of Laboratory Animals” [16].
| Results | |
|---|
In the entire rabbit population, the mean aortic diameter was 3.11 ± 0.40 mm before stenting, and after stenting was 3.39 ± 0.69 mm upstream, 3.76 ± 0.42 mm at the stent level, and 3.33 ± 0.50 mm downstream from the stent. There was no significant difference in aortic diameter among the three groups before and after stenting. Aortic diameter increased at the stent level after stenting (p < 0.05) in all groups, whereas there was no significant diameter change upstream or downstream from the stent and no significant difference among sites after stenting.
In the entire rabbit population, diameter compliance (10-3 mm kPa-1) was 75.3±20.1 before stenting and reached 94.7 ± 42.2 upstream, 38.8 ± 14.2 at the stent level (p<0.05), and 70.8 ± 23.2 downstream from the stent. At the stent level, diameter compliance decreased from 77.63 ± 24.30 to 30.14 ± 8.24 in the Wallstent group, from 69.10 ± 19.25 to 37.46 ± 8.57 in the Palmaz group, and from 80.88 ± 18.61 to 55.25 ± 17.83 in the Jostent group (p≤0.01 for all groups). After stenting, diameter compliance was significantly lower at the stent level than upstream (98.48 ± 53.19 for the Wallstent, 85.36 ± 33.83 for the Palmaz stent, and 103.13 ± 48.00 for the Jostent; p < 0.0005) or downstream (59.04 ± 13.83 for the Wallstent, 79.28 ± 30.80 for the Palmaz stent, and 75.53 ± 19.42 for the Jostent; p < 0.0001). Diameter-compliance values measured upstream from the stent for each stent design were not significantly greater than those obtained before stenting. Diameter compliance values measured downstream from the stent for the Wallstent and for the Jostent were not significantly smaller than those obtained before stenting.
In the entire rabbit population, the mean distensibility coefficient (10-3 kPa-1) was 24.3 ± 6.3 before stenting and reached 27.8 ± 10.3 upstream, 10.5 ± 4.4 at the stent level (p < 0.001), and 21.9 ± 8.6 downstream from the stent. There was a significant decrease in distensibility at the stent level after stenting (p < 0.0001). The distensibility coefficient was significantly lower after stenting at the stent level than upstream (p < 0.0001) or downstream (p < 0.0001).
The distensibility coefficient upstream from the stent (27.47 ± 12.42 for the Wallstent, 28.49 ± 12.38 for the Palmaz stent, 27.59 ± 3.61 for the Jostent) for each stent design was not significantly greater than that before stenting (22.73 ± 5.02 for the Wallstent, 24.80 ± 8.49 for the Palmaz stent, 25.14 ± 6.99 for the Jostent). The distensibility coefficient downstream from the stent (15.92 ± 3.03 for the Wallstent, 24.86 ± 7.3 the Jostent) was not significantly lower than that before stenting.
After stenting, the Doppler signal revealed no detectable change in flow-velocity profile and no turbulence downstream from the stent. Pulsatility index was 2.14 ± 0.54, mean arterial pressure was 7.89 ± 1.71 kPa, and systolic pressure change was 3.53 ± 0996 kPa before stenting and showed no significant change at any site or among sites after stenting.
Comparison of the three stent designs for the pulsatility index, diameter compliance, distensibility coefficient, aortic diameter, and diameter changes showed no significant difference before or after stenting at any site (Fig. 2A,2B,2C).
 View larger version (27K) | Fig. 2A. —Comparative presentations of diameter (d), systolic diameter change (Δd), blood pressure (P), and ECG, at level of stent, before and after stenting. Recordings show Wallstent (Schneider, Bülach, Switzerland) (rabbit 2) (A), Palmaz stent (Johnson & Johnson, Warren, NJ) (rabbit 6) (B), and Jostent (Jomed, Rangendingen, Germany) (rabbit 12) (C). Note decrease in systolic diameter change after stenting in all instances. |
 View larger version (30K) | Fig. 2B. —Comparative presentations of diameter (d), systolic diameter change (Δd), blood pressure (P), and ECG, at level of stent, before and after stenting. Recordings show Wallstent (Schneider, Bülach, Switzerland) (rabbit 2) (A), Palmaz stent (Johnson & Johnson, Warren, NJ) (rabbit 6) (B), and Jostent (Jomed, Rangendingen, Germany) (rabbit 12) (C). Note decrease in systolic diameter change after stenting in all instances. |
 View larger version (26K) | Fig. 2C. —Comparative presentations of diameter (d), systolic diameter change (Δd), blood pressure (P), and ECG, at level of stent, before and after stenting. Recordings show Wallstent (Schneider, Bülach, Switzerland) (rabbit 2) (A), Palmaz stent (Johnson & Johnson, Warren, NJ) (rabbit 6) (B), and Jostent (Jomed, Rangendingen, Germany) (rabbit 12) (C). Note decrease in systolic diameter change after stenting in all instances. |
| Discussion | |
|---|
The theoretic hemodynamic consequences of an artery—graft compliance mismatch include increased impedance to flow, decreased distal perfusion, and disturbed flow or turbulence [4,5,6,7,8, 17]. Intimal hyperplasia has been shown to develop more frequently at the distal anastomosis or downstream connection to the native artery compared with the proximal anastomosis [6, 7]. A compliance mismatch—related flow disturbance increases the residence time of growth factors or activated particles at the distal anastomosis, leading to intimal hyperplasia and ultimately restenosis [4, 5]. Studying the intimal response to stenting in animal models, Barth et al. [16] and Sutton et al. [18] found a larger thickness of neointima at the distal extremity of the stent than at the proximal or middle part of the stent. Therefore, the preferential distal location of intimal hyperplasia after stenting could be explained by a mechanical theory such as compliance mismatch. The amount of intimal hyperplasia was the main focus of most studies comparing the in vivo properties of differently designed stents, but different or even opposite results have been reported [19,20,21]. Few in vivo studies have reported the change in mechanical properties after stenting. Back et al. [13] first established that stent placement induces a decreased arterial wall compliance of the stented segment in dogs [14]. In research published in 1999, Rolland et al. [14] compared the wall mechanics of large-caliber (iliac) arteries in swine after placement of six types of stents. They found that the changes in arterial compliance and wall mechanics strongly depended on the stent design. There is, to our knowledge, no previous study of wall mechanics in stented arteries and no comparison of different stents specially designed for small-caliber arteries, for which the rate of clinical failure is the highest (<5 mm).
Stenting kept the arterial pulsatility index unchanged downstream and upstream from the stented arterial segment and produced no downstream turbulence, showing that stents did not significantly impede blood flow in small-caliber arteries.
Like the results of Rolland et al. [14] for large arteries, we found a compliance decrease in the stented segment of small-caliber arteries in the rabbit aorta with the Wallstent and the Palmaz stent, whereas we found no significant change upstream or downstream from the stented segment. Therefore, independent of the three stent designs we tested, there was a marked compliance mismatch between the stented and nonstented aorta. Unlike the results reported by Rolland et al., no statistically significant diameter compliance or distensibility coefficient difference among the three types of tested stents was found before or after stenting at all the studied levels. There are, indeed, marked differences between the study carried out by Rolland et al. and our study. We Measured all parameters before and 15 min after stenting in the same animal, without any significant change in hemodynamic parameters during the entire procedure. Because we investigated much smaller arteries than those of Rolland et al., we used a different technique with a higher ultrasound frequency (20 instead of 10 MHz), ensuring a 0.01-mm spatial resolution, and echo tracking instead of ultrasound time of flight, avoiding any mechanical constraint on the artery, whereas Rolland et al. used a silicone clip supporting two transducers positioned face to face around the vessel. By using this different technique, we were able to show subtle changes and differences in arterial wall mechanics in even smaller arteries (rabbit iliac arteries) after laser-assisted microanastomosis compared with manual microanastomosis [22]. On the other hand, to avoid catheter-related flow disturbances, we did not perform in situ intraarterial blood pressure measurements. Therefore, we could not evaluate additional mechanical properties like hysteresis.
The increase in aortic diameter after stenting is not the only factor responsible for the decreased compliance at the stent level, as shown by the significant decrease of the distensibility coefficient. The stent size was chosen to fit the aortic diameter and to avoid overdilation. Some authors purposely overdilated the stents by 30%, arguing that stretching of the arterial wall is probably what usually occurs during stent placement for occlusive vascular lesions in clinical conditions [20]. In our study, we mimicked the clinical practice with small-caliber arteries when the nonstenotic artery upstream and downstream from the stenotic segment was covered with the stent.
The analysis of diameter changes after stenting must consider structural and mechanical differences among stents. In experimental models and clinical trials, elastic recoil in vivo accounts for the loss of 10% of the initial post-stenting arterial diameter [23]. The Wallstent is an elastic stent in contrast with the plastic Palmaz and Jostent stents. Except for the recoil, the size of a stent correlates with the size of the balloon angioplasty catheter used for its deployment. Balloon angioplasty catheters designed for the deployment of expandable stents are not compliant and support high inflation pressures. This feature accounts for a correct expansion of the balloon despite the parietal strain usually encountered in clinical practice. The stent diameter will not increase unless additional angioplasty with a larger balloon diameter is performed. In contrast, because of the elastic properties of the Wallstent, there is a spectrum of possible sizes for a given diameter. With the Wallstent, the final diameter depends on the strain applied by the arterial wall itself and surrounding tissues.
The possible decrease in parietal strain due to anesthesia and periaortic dissection could explain the fact that we found a larger (although not statistically significant) aortic diameter after stenting with the Wallstent than with the Palmaz stent or the Jostent. Balloon-expandable stents appear to offer the advantage of an expansion precisely adjusted to the limits of the delivery balloon.
A higher rate of both acute thrombosis and late restenosis has been reported when placing covered stents instead of noncovered stents in the rabbit aorta [24]. For Tepe et al. [24], stent failure could result from a decreased biocompatibility of the coverage. Covered stents seem to exhibit the same mechanical properties as uncovered stents, but their number in our study was too low to draw any definitive conclusion.
The differences in stent design concern the type and size of the wires, the number of strut—strut intersections, the length and surface of the stent, and its radial strength and longitudinal flexibility [23, 25]. Moreover, some parameters cannot be studied independently, so it is not possible to incriminate one single parameter for a given observed difference. However, our study reflects the clinical practice in which shorter Wallstents are not available and long Palmaz stents are too rigid to reach some lesions in tortuous vessels. Such differences related to stent design concern all comparative studies performed with the stents currently used in clinical practice [9]. In a 1999 publication, researchers established the stented segment length as an independent predictor of restenosis [26]. In our study, the Wallstent stents we placed in the aorta were nearly twice as long as the two other types of stents. This difference may be reflected by the insignificant decrease in compliance and pulsatility in our study. Nevertheless, the length of the stented segment was always less than 2 cm in our study so that all our cases corresponded to the same group in the classification used in the study of Kobayashi et al. [26].
Our study did not assess the histologic changes, including atrophy of the media and the neointima formation, that occur from 1 day to 4 weeks after stent placement, whereas the maximum restenosis rate occurs 8 weeks after stent placement and remains unchanged thereafter [18]. Although chronic studies would be interesting, stenting is the first causal factor, and its immediate mechanical consequences are thought to be the initial cause of histologic changes that may, in turn, induce or worsen mechanical changes [6].
The invasive procedure we used for the placement of ultrasound probes represents another limitation of our study and would not be suitable for chronic studies. As previously reported, both the anesthesia and the surgical dissection of the aorta modify wall mechanics [27, 28]. We are now developing and validating a new method for chronic studies using transparietal B mode sonography and a dedicated image-analysis software to noninvasively evaluate the diameter changes.
Finally, the rabbit aorta is an elastic artery, and the results of our study cannot be directly extrapolated to the treatment of small-caliber arteries in humans. Nevertheless, similar histologic reactions to stenting have been reported by Robinson et al. [29] in both the rabbit aorta and the pig or dog coronary artery, and the rabbit aorta is widely used as a model for the study of stent-induced histologic changes.
To our knowledge, ours is the first in vivo study comparing the effects of different stent designs on wall mechanics of small-caliber arteries. We found no significant differences among the Wallstent, the Palmaz, and the Jostent. At implantation, they reduce the compliance of the stented arterial segment and induce a compliance mismatch. As previously suggested, this compliance mismatch may be one of several factors promoting restenosis. However, long-term studies are mandatory for a thorough understanding of the underlying mechanisms.
Address correspondence to M. Dauzat.
We thank Margaret Manson for her help in revising the English manuscript.
| References | |
|---|
