| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Experimental Studies |
1 From the Department of Radiology, University Hospital Schleswig-Holstein–Campus Kiel, Arnold-Heller-Strasse 9, 24105 Kiel, Germany (S.M.H., P.S., J.H., C.L., T.J., M.H.); and Institute of Anatomy, Christian-Albrecht-University, Kiel, Germany (F.P.). Received December 5, 2003; revision requested February 12, 2004; final revision received July 13; accepted August 17. Address correspondence to S.M.H. (e-mail: muehue@rad.uni-kiel.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: The local animal experimentation committee did not require its approval for this study. With a benchtop vascular model (flow rate, 470 mL/min; dicrotic pulsatile flow, 76 pulses per minute; pressure, 115/67 mm Hg [mean pressure, 91 mm Hg]) into which 85 porcine internal carotid arteries (ICAs) were inserted, five different protection devices (Angioguard [Cordis/Johnson & Johnson, Miami, Fla], Filterwire EX [Boston Scientific, Natick, Mass], Trap [Microvena, White Bear Lake, Minn], Neuroshield [Abbott Laboratories, Redwood City, Calif], and Percusurge [Abbott Laboratories]) were evaluated. Adverse movement (1 cm up, 2 cm down, and 1 cm up again) of the activated devices (deployed filters or inflated balloons [Percusurge only]) was simulated, and the device was retrieved. For each of these steps (deployment, movement, retrieval) the amount of debris from the vessel wall in the effluent of the ICA was determined by using a 100-µm filter. The Mann-Whitney test was used to test for differences, and a correction for multiple comparisons was made. P < .05 was considered to indicate a significant difference. The authors attempted to determine whether there was a notable association between the total amount of debris captured and the classification of damage at microscopy. Carotid arteries were analyzed histologically with light and scanning electron microscopy.
RESULTS: All examined protection devices caused dislodged debris, which was captured in the effluent filter. There were significant differences among the devices in terms of the total amount of debris captured in the filters (lowest amounts of debris, 4.75 mg [Angioguard] and 5.02 mg [Filterwire EX]; highest amount, 7.51 mg [Trap]; P 
.001 for all). All devices caused histologically visible wall damage, with the degree of intimal denudation correlating with the mass of the debris. The Trap device caused the most severe intimal and subintimal wall damage. Adverse movement resulted in no increased debris dislodgment as compared with the debris dislodged during deployment and retrieval of the devices.
CONCLUSION: On the basis of the data obtained, cerebral protection devices themselves have a potential influence on embolization rates by causing debris to be dislodged during carotid stent placement.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Cerebral protection devices have been shown to reduce the rate of thromboembolic complications in carotid angioplasty and stent placement procedures. A systematic review of the literature from 1996 to 2002 reveals that the use of protection devices significantly lowers the risk of the occurrence of a stroke (1), which is usually caused by emboli dislodged from the lumen of the artery. However, although there have been substantial efforts to determine the degree of embolization caused by carotid angioplasty and stent placement procedures, to our knowledge there are no substantial data that describe restenosis or vessel trauma caused by the protection devices themselves.
Thus, the purpose of our study was to determine the extent of vessel wall damage caused by cerebral protection devices designed for carotid angioplasty by using ex vivo porcine carotid arteries.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The carotid arteries were prepared carefully by a well-experienced butcher according to our recommendations (P.S., S.M.H.) immediately after sacrifice of the animals. The vessels ranged from 4.0 to 5.5 mm in diameter and matched the size of the protection devices used. Prepared carotids were cooled in the kind of icebox used for transplanted organs and were used within 6 hours. After the experiments, the samples were prepared for light or scanning electron microscopy.
Flow Model
In a benchtop model made with silicone tubing, dicrotic pulsatile flow (76 pulses per minute) was delivered by a microprocessor-controlled peristaltic roller pump (MCP-Standard ISM 404; Ismatec, Wertheim-Mondfeld, Germany) at a flow rate of 470 mL/min. This is the estimated flow rate in patent carotid arteries required to guarantee sufficient blood supply to the brain when the contralateral carotid artery is occluded. The resulting pressure was 115/67 mm Hg (mean pressure, 91 mm Hg) as measured with a specific sensor (Sirecust 620; Siemens, Danvers, Mass). Saline solution (0.9%) was used as a flow medium and was not recirculated (2,3).
The carotids were inserted into the flow model end-to-end; one end was the common and the other was the internal carotid artery. An approach to the carotid artery was made with a 10-cm-long 9-F sheath (Terumo, Tokyo, Japan) that was placed in the silicone tube in front of the inserted common carotid artery. Flow was directed through the carotid artery for 3 minutes to flush out remaining postmortem debris; the control arteries were then removed for sample preparation and light and scanning electron microscopy. The study protocol consisted of the following three steps:
Step 1: placement.—Flow was directed through the carotid artery for 3 minutes to flush out remaining postmortem debris. The protection device was then inserted and placed according to the manufacturer’s recommendations, and the amount of dislodged debris in the 100-µm polyethylene effluent filter (Schimmel, Nordheim, Germany) was determined (Fig 1).
|
Step 3: retrieval.—Finally, the protection device was retrieved according to the instructions for its use, flow was adjusted and maintained for another 3 minutes, and the amount of debris was determined. P.S. and S.M.H. worked together in consensus to perform all procedures involving the flow model.
The amount of dislodged debris (in milligrams) in the effluent filters was determined. The occurrence of dislodged debris was measured separately for each step. The masses of the filters were determined before they were used. After the procedure was completed, the captured vessel wall fragments were weighed again after the filters were dried to assess the presence of peripheral emboli.
Protection Devices
Five protection devices were investigated: four filter devices (Angioguard [Cordis/Johnson & Johnson, Miami, Fla], Filterwire EX [Boston Scientific, Natick, Mass], Trap [Microvena, White Bear Lake, Minn], and Neuroshield [Abbott Laboratories, Redwood City, Calif]) and one balloon occlusion device (Percusurge [Abbott Laboratories]) (Fig 2). The Angioguard and Filterwire EX devices were each used in 15 of the 85 arteries, the Trap and Percusurge devices were each used in 14 arteries, and the Neuroshield device was used in 12 arteries. These systems were meant to constitute a representative choice of currently available systems. Technical details of these devices have been described previously (3,4).
|
Light and Scanning Electron Microscopy
Carotid segments from the middle of the vessel (corresponding to the position of the protection device within the vessel before and after adverse movement) were evaluated histologically by using light and scanning electron microscopy. For each step of the procedure, the conditions of the vessels in which devices had been used were compared with the conditions of a control group of undamaged porcine carotids (in which no protection devices had been used). Owing to the use of different microscopy modalities, the preparations of the vessels were slightly different for light microscopy and scanning electron microscopy.
Sample preparation and light microscopy.—The porcine carotid arteries were fixed in 4% formalin immediately after the experiment. They were decalcified in 20% ethylenediaminetetraacetic acid at 37°C for several days until x-ray examination revealed complete decalcification. The carotids were then embedded in paraffin and sliced in a horizontal plane. Thirty 7-µm sections from each carotid (middle vessel segment) were stained with toluidine blue (pH, 8.5) and the Goldner method (5) and were examined with a light microscope (Standard 25 CSI; Carl Zeiss, Heidenheim, Germany).
Scanning electron microscopy.—After immediate fixation in 4% formalin, the carotids were cut longitudinally and opened into halves so that the epithelial lining could be examined. One 10 x 6-mm tissue block was prepared from each middle-vessel segment. All tissue blocks were then impregnated in 2.5% tannic acid for 2 days. Counter-fixation in 2% osmium tetroxide for 4 hours was followed by dehydration in ethanol and drying in a critical point dryer. Preparations were coated with gold and analyzed by using a scanning electron microscope (XL 20; Philips, Kassel, Germany).
The classification of vessel wall injury and denudation was determined at light microscopy; scanning electron microscopy was used only to confirm the classification and distribution of the injury. P.S., S.M.H., and F.P. worked together in consensus and performed the light and scanning electron microscopy; F.P. had more than 12 years of experience in evaluating vessel disease. The extent of damage to the vessel wall was classified as follows:
Group 0: The inner endothelial layer was intact and undamaged, as observed in the control group.
Group 1: Partial endothelial damage of less than 50% was present—there was a loss of endothelial monolayer in less than 50% of the evaluated vessel surface at light microscopy that was proved with scanning electron microscopy.
Group 2: Partial endothelial damage of 50% or more was present—there was a loss of endothelial monolayer in 50% or more of the evaluated vessel surface.
Group 3: Subendothelial destruction was present—there was a partial loss of the media layer.
Statistical Analysis
Data are presented as means ± standard deviations. A P value of less than .05 was considered to indicate a significant difference. The amounts of debris captured during the different steps of the procedure were compared, and the performance of the different protection devices was evaluated. Correlation between the total amount of debris captured and the classification of the damage seen with microscopy was attempted; correlation coefficients significant at the .01 level were identified with the nonparametric measures of Kendall (
b) and, alternatively, those of Spearman (
). All statistical tests for differences were based on the Mann-Whitney test (SPSS version 2.0; SPSS, Chicago, Ill). J.H. and C.L. performed the analyses.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
When the masses of the debris captured during the individual steps were added (ie, when we looked at the complete procedure), we observed significant differences among the devices: Angioguard performed best (4.75 mg ± 0.35), followed by Filterwire EX (5.02 mg ± 0.20), while Trap performed worst (7.51 mg ± 0.35) (P .001 for all) (Table 1, Table 2, Fig 3). In terms of the individual steps of the complete procedure, the following results were obtained:
|
|
|
Step 2: Adverse Movement
The greatest mass of dislodged material occurred with the Trap device (P < .02, as compared with all other devices except the Neuroshield device). There were no significant differences between the Angioguard, Filterwire EX, and Percusurge devices in terms of the mass of debris dislodged during adverse movement.
Step 3: Retrieval
When the devices were retrieved, Filterwire EX and Neuroshield yielded significantly smaller amounts of debris compared with the Trap device (P = .001). The remaining devices performed approximately equally. Results and statistical findings are summarized in Tables 1 and 2.
Light and Scanning Electron Microscopy
Visible wall damage, as compared with the condition of the walls in the control vessels, was caused by all devices (Figs 4, 5). The Trap device caused endothelial loss in each of the 14 samples and was the only device that caused subintimal wall damage (Fig 6). The degree to which the endothelial layer was affected by each device is shown in Table 3. The extent of intimal damage (as classified at microscopy) was related to the amount of detached debris, as indicated in Figure 7. Correlation coefficients significant at the .01 level were identified with the nonparametric measures of Kendall (b = 0.63) and, alternatively, with those of Spearman ( = 0.75). When these measurements were grouped into the damage levels listed in Table 3, both tests revealed a significant (linear) association between the considered parameters.
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The potential efficacy and limitations of protection devices have been documented in several in vitro and in vivo trials. Ohki and colleagues (6) described an embolic risk from the different stages of carotid bifurcation balloon angioplasty and stent placement. The phases before and after dilation of the lesion seemed particularly risky. Additionally, it has to be kept in mind that late embolization still occurs in a minority of patients and thus may account for the small but significant risk of delayed stroke (7).
There is still some controversy about the clinical relevance of cerebral microemboli caused during carotid angioplasty and stent placement. It has yet to be determined if the occurrence of cerebral emboli after carotid balloon angioplasty and stent placement is directly related to minor or major neurologic events. Evidence to support such a correlation was provided by Jaeger et al (8): In 29% of their unprotected procedures, stent implantation in the carotid artery was associated with new areas of cerebral ischemia detected at diffusion-weighted magnetic resonance imaging; the new lesions were clinically silent in all patients except one. Thus, it would be of great interest to determine the minimum size of emboli that need to be stopped to prevent a stroke.
To our knowledge, there is currently no animal model for the assessment of the amount of cerebral emboli dislodged during carotid artery stent placement. Thus, we constructed a benchtop model of a porcine internal carotid artery so that we could evaluate vessel wall damage and the potential for emboli formation resulting from the use of cerebral protection devices.
Embolization and Debris Capture in Vitro
Coggia et al (7), after performing an experimental study involving fresh human carotid specimens, reported that the majority of embolic particles during balloon angioplasty were smaller than 100 µm. Contrary to the model used by Ohki and colleagues (6), which involved real human plaque but steady flow, our benchtop model was made from elastic silicone and contained an internal carotid artery. Dicrotic pulsatile flow mechanics were initiated with a peristaltic computer-controlled pump, thus producing a physiologically more relevant situation. When Ohki et al tested the Neuroshield filter device by using a human carotid stent-placement model, they found that 88% of particles between 360 and 1000 µm were captured with the filter. Unlike ourselves (who used carotid arteries obtained from porcine cadavers), they obtained carotid plaques from carotid endarterectomy procedures to mimic as much as possible the scenario in patients.
With use of either a filter or a balloon placed in the internal carotid artery for cerebral protection, damage to the intimal monolayer due to unexpected movement of the inflated balloon or deployed filter basket might occur. Additionally, it is known from clinical experience that an abnormal major movement of the opened devices was very often observed during endovascular maneuvers with either an occlusion balloon or filters (9).
The present experimental evaluation revealed that all protection devices caused a measurable amount of debris when an abnormal major mobility situation was simulated. For the entire procedure, significant differences were found among the performance of the devices (lowest masses of captured debris, 4.75 mg [Angioguard] and 5.02 mg [Filterwire EX]; highest mass, 7.51 mg [Trap]; P .001 for all). This indicates that, when porcine arteries are used, a nitinol mesh serving either as a filter or as a retrieval device causes major vessel wall injuries, whereas currently available devices (which have been improved and redesigned in the latest generation without changing the working principle) seem more appropriate and cause less vessel wall damage. On the basis of our results in this experimental setting, we noted the following: The Angioguard and Filterwire EX devices performed better than the Neuroshield and Percusurge devices, while the Trap device, which caused vessel wall injury, had the poorest performance.
In particular, the simulation of the adverse movement of an activated protection device clearly revealed the differences in the extent of wall injury that depended on the design of the device: The greatest amount of emboli was observed with the Trap device (P .05, compared with all devices except Neuroshield); there were no significant differences in terms of the amount of emboli between the Angioguard, Filterwire EX, and Percusurge devices; and there were no statistical differences for the Neuroshield device compared with all other devices. It may be hypothesized that the area of wall contact may be one factor that could be responsible for the severity of the vessel damage; for instance, the Neuroshield device has a 10-mm-long contact zone, while other devices also have circumferential contact but smaller extents in the craniocaudal direction. Second, bare metals like nitinol struts have a greater potential for injury than a filter membrane made of plastic mounted on a nitinol ring or skeleton. However, such a skeleton may cause localized areas of damage owing to the involvement of the struts responsible for filter expansion. Interestingly, no substantial differences between the amount of debris generated by a filter and that generated by a balloon device were observed.
There was a strong correlation between the degree of damage caused by the devices that was observed at microscopy and the amount of debris that was captured in the filter (Kendall b = 0.63, Spearman = 0.75). This supports the validity of our approach, but it also makes clear that optical measurements are not completely satisfactory for evaluating the potential danger of dislodged debris.
Study Limitations
There were some limitations associated with the design and construction of the benchtop model that we used: The carotid artery was not embedded in its usual surrounding tissue (fat, muscle), and the carotids had been separated from the porcine cadavers before use. However, the potential vessel wall damage caused by the use of cerebral protection devices should be limited to wall changes only, and this seems best evaluated in nondiseased, "clean" vessels. In spite of its limitations, this simple model offered a method for assessing the vessel wall damage and the amount of dislodged emboli in pulsatile flow conditions; the acquired data provide indications about the performance of the tested devices that might be expected in vivo. Additional aspects that are difficult to realize in vitro may have to be considered for future in vivo studies: For instance, the behavior of a native (ie, possibly calcified and severely stenosed) vessel might differ from that of an in vitro porcine carotid artery. Thus, these results should be extrapolated with care to the clinical setting.
In conclusion, the five cerebral protection devices tested caused visible intimal vessel wall damage. An adverse movement of an activated protection device caused more debris to detach from the vessel wall, in addition to the debris dislodged by the deployment and retrieval of the device.
Practical application: The use of cerebral protection devices during carotid angioplasty and stent placement may have a negative influence on embolization rates. This may require a redesign of current protection devices; further evaluations are warranted.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, S.M.H.; study concepts, S.M.H., P.S.; study design, S.M.H., F.P.; literature research, P.S., S.M.H.; experimental studies, F.P., S.M.H., P.S.; data acquisition, S.M.H., C.L., P.S.; data analysis/interpretation, P.S.; statistical analysis, J.H., C.L., P.S.; manuscript preparation, F.P., S.M.H., T.J., C.L.; manuscript definition of intellectual content and final version approval, S.M.H., C.L.; manuscript editing, S.M.H.; manuscript revision/review, M.H., C.L., S.M.H.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. K. Eskandari Preventable Complications of Carotid Stenting Perspectives in Vascular Surgery and Endovascular Therapy, March 1, 2008; 20(1): 17 - 25. [Abstract] [PDF]
|
|
|
|
 |
|
 H. S. Gurm, B. K. Nallamothu, and J. Yadav Safety of carotid artery stenting for symptomatic carotid artery disease: a meta-analysis Eur. Heart J., January 1, 2008; 29(1): 113 - 119. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Ribo, C. A. Molina, B. Alvarez, M. Rubiera, J. Alvarez-Sabin, and M. Matas Transcranial Doppler Monitoring of Transcervical Carotid Stenting With Flow Reversal Protection: A Novel Carotid Revascularization Technique Stroke, November 1, 2006; 37(11): 2846 - 2849. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Kastrup, T. Nagele, K. Groschel, F. Schmidt, E. Vogler, J. Schulz, and U. Ernemann Incidence of New Brain Lesions After Carotid Stenting With and Without Cerebral Protection Stroke, September 1, 2006; 37(9): 2312 - 2316. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. du Mesnil de Rochemont, S. Schneider, B. Yan, A. Lehr, M. Sitzer, and J. Berkefeld Diffusion-Weighted MR Imaging Lesions after Filter-Protected Stenting of High-Grade Symptomatic Carotid Artery Stenoses AJNR Am. J. Neuroradiol., June 1, 2006; 27(6): 1321 - 1325. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
