Percutaneous Treatment of Abdominal Aortic Aneurysm in a Swine Model
Understanding the Behavior of Aortic Aneurysm Closure Through a Serial Histopathological Analysis
Background Previous studies used covered stent grafts to treat abdominal aortic aneurysms; however, such devices block flow into aortic side branches. We used uncovered stents with and without additional embolization coils to treat abdominal aortic aneurysm in a swine model and examined serial histological changes in the aneurysms over a 6-month period.
Methods and Results We examined aneurysms in 9 control and 9 treated pigs (5 received stents alone and 4 received stents and coils). Aneurysms were surgically created with abdominal fascia. Three days later, we percutaneously placed a self-expandable uncovered stent across the aneurysm. Coils were implanted through the stent into the aneurysm lumen. An aortogram immediately after stent placement showed no significant change in aneurysm lumen; however, in pigs that had aortograms between 6 weeks and 6 months after treatment, the diameter decreased (28% to 65%) in 4 of 5 pigs, and 1 had no discernible aneurysm. Three treated pigs died, but only 1 from rupture. In contrast, 7 untreated aneurysms ruptured (2 pigs died of other causes). Histological examination revealed that the aneurysm lumen was reduced after treatment by collagen production. This healing process was accelerated in aneurysms treated with both stents and coils. In contrast, only limited amounts of new collagen were found in untreated, ruptured aneurysms. Instead, the fascia was disrupted and there was evidence of collagen degradation.
Conclusions We found that uncovered stents reduce the likelihood of aneurysm rupture in a swine model without blocking arterial branches. The presence of coils enhanced filling of the lumen by collagen.
Ever since Dubost et al1 described surgical aneurysmectomy in 1952, this procedure has been the mainstay of treatment for abdominal aortic aneurysm (AAA). In recent years, transluminal implantation of endovascular grafts has been tested in animal models that may not fully replicate the pathophysiological conditions found in human aortic aneurysms.2 3 4 5 6 On the basis of such animal studies, investigators have initiated limited clinical trials, with mixed results.7 8 9 10 11 The majority experience in animals and humans has been accumulated with covered devices; however, radiological exclusion of an aneurysm may not be equivalent to an effective therapy. Furthermore, covered devices will interrupt blood flow into arteries covered by the graft.
In a previous study, we demonstrated the ability of uncovered stents to gradually exclude surgically created AAA during short-term follow-up in a canine model with no incidence of rupture.5 The canine fibrinolytic system is very active compared with that of humans,12 a potential disadvantage in the evaluation of aneurysm rupture after placement of uncovered devices. In the present study, we used swine as the experimental animal because their fibrinolytic and coagulation systems are closer to humans than are those of dogs.12 Our aim was to assess the long-term radiological and histological response to uncovered stents with and without the additional deployment of embolization coils in the aneurysm lumen.
The study was conducted according to the principles outlined in the “Position of the American Heart Association on Research Animal Use,” adopted November 11, 1984. The protocol was approved by the animal research committee of Loma Linda University.
Eighteen Yorkshire swine (15 to 31 kg) were assigned to two groups, 9 controls and 9 in the study group. In the study group, 5 pigs received stent alone, and 4 had embolization coils deployed in the aneurysm cavity in addition to stent placement. We first surgically created an “aneurysm” in the abdominal aorta in each pig and then deployed a stent in the study group 3 days after aneurysm creation. The study group animals were euthanized at 4 weeks, 6 weeks, 3 months, and 6 months after treatment. Tissue from control pigs was collected at the time of death.
Creation of Aneurysm Model
The surgical creation of AAA has been described previously.13 Briefly, we took a piece of rectus abdominis fascia about 80×100 mm and cut it in half. After exposing the abdominal aorta, we sutured the two pieces of fascia on the external surface of the aorta in a longitudinal fashion and then sutured them to each other along their edges, which were left partially open. We placed aortic cross-clamps below the renal arteries and at the aortic bifurcation and proceeded to resect the entire aortic wall within the pouch. We then clamped the pouch at a point adjacent to the aorta and unclamped the aorta to allow blood flow through it while we sutured the roof of the pouch. After closing the roof, we unclamped the pouch and allowed blood flow into the aneurysm.
Description of Devices
We used a self-expandable uncovered stent made from 0.15-mm-diameter superelastic nitinol wire (Cook Inc). The stent was constructed with 8 wires to form a woven mesh tube, and each cell size was 2.03×4.72 mm (9.58 mm2). The wires at both ends were rounded to prevent damage to the vessel wall. Three stent diameters were used: 8, 10, and 12 mm. The unconstrained length of all devices was 100 mm. The commercially available embolization coils measured 50×5 mm (Cook Inc).
Stent Placement Procedure
All 9 pigs in the study group were taken to the catheterization laboratory 3 days after AAA creation. Under general anesthesia, an 8F (1 French unit=0.33 mm) sheath was placed percutaneously into the left carotid artery. A 6F angiographic pigtail catheter, advanced over a Benson wire (0.889 mm), was then placed in the descending aorta 50 mm above the aneurysm. Before stent implantation, we documented the angiographic size and location of the created aneurysm from an abdominal aortogram (30 mL of contrast at 20 mL/s). The stent, premounted on the delivery catheter, was then advanced to the desired position and deployed under fluoroscopic guidance. In the group of pigs treated with both stents and coils, the coils were deployed with a 6F Judkins-type RCA catheter and pushed out with a 0.889 mm Teflon stainless fixed wire. Two pigs received 1 coil, 1 pig received 5 coils, and another pig received 10 coils. An aortogram was done after deployment of the stent or coils in all pigs. No anticoagulants or antiplatelet drugs were used before, during, or after this procedure.
Angiographic Measurement of Aneurysm Size
Aneurysm size was measured angiographically with a computerized digital analytical system (AngioComm StatVIEW Digital Recorder, ImageComm Systems, Inc). The largest transverse diameter of the contrast-filled aneurysm was measured. The reference aortic diameter was determined as the average of the normal aortic diameter immediately above and below the aneurysm. Two measurements were taken, and the mean value was calculated.
The control group animals were observed until death occurred, at which time the abdominal aorta was removed. The study group pigs, except for unexpected early deaths, were followed to 4 weeks, 6 weeks, 3 months, or 6 months after device implantation and then euthanized under deep sodium pentobarbital anesthesia by injection of potassium chloride solution. The entire descending aorta was then removed.
After fixation in a solution of 10% neutral buffered formalin, tissue from each aneurysm was processed for paraffin embedding and sectioned at a thickness of 5 μm. In addition, we also cut sections from the fascia used to construct the aneurysms. The sections were stained with hematoxylin-eosin and with picrosirius red.14
The hematoxylin-eosin–stained sections were examined by bright-field microscopy for the presence of inflammatory cells. The picrosirius red–stained sections were viewed with polarized light to examine the collagen fibers. We qualitatively assessed the thickness of the collagen fibers. The color of collagen fibers stained with picrosirius red and viewed with polarized light depends on the thickness of the fibers. As fiber thickness increases, the color changes from green to yellow to orange.15 In addition, we quantitatively examined three parameters: collagen fiber orientation, collagen content, and the optical properties of the collagen.
Collagen Fiber Orientation
Collagen fibers have been found to become, in time, increasingly aligned in healing tissue. Such an increase in organization has been reported in dermal wound healing16 and after acute myocardial infarction.17 We speculated that the healing process in our aneurysm model might be similar and therefore that examination of collagen fiber organization in the aneurysms could provide a method of assessing the progression of healing. We measured the two-dimensional orientation of collagen fibers in the aneurysm wall using a previously described method.18 Briefly, picrosirius red–stained sections were examined on the rotating stage of a polarizing microscope with a ×40 objective lens. We measured the orientation of 50 collagen fibers at three different locations in each sample. The fibers selected for measurement were those intersected by the grid of an eyepiece reticule. The orientation data obtained were analyzed by use of circular statistics.19 Specifically, we calculated the mean orientation and the angular deviation (the circular statistics equivalent of SD) of each distribution. We expressed the orientation of the fibers relative to the tangent angle to the aneurysm wall at the point at which the measurements were taken.
We measured the area occupied by collagen fibers in three regions in which new collagen fibers were present (ie, not in the fascia). This analysis was performed at a magnification of ×20 on picrosirius red–stained sections according to previously published methods.20 Briefly, we used an image analysis system to subtract an image of the background (a field of view with no tissue present) from a circularly polarized image of the tissue. This has the effect of producing an image of gray collagen fibers on a black background. The interstitial space was made black by the subtraction process (pixel brightness in the image was expressed on a 0 to 255 gray scale, with 0=black and 255=white). Thus, any pixel with a gray level >0 represented collagen. Collagen content was expressed as the area fraction of pixels with a gray-scale level >0.
Optical Properties of the Collagen
In the same regions in which collagen content was measured, we also measured the brightness of the fibers. Several groups have found that collagen fiber brightness in scar tissue increases with time.20 21 Thus, we anticipated that measurements of this parameter would indicate the degree of healing. We examined circularly polarized images of picrosirius red–stained tissue using a ×20 objective lens. Again, an image of the background was first subtracted from the circularly polarized image, and then a histogram of pixel intensity was plotted, from which we calculated the average gray level of collagen fibers in the tissue. The average gray level of the fibers in the three regions was then calculated.
Stent and Coil Placement
The stents were successfully deployed and placed across the aneurysms in all animals. In addition to stent placement, pigs 6 through 9 also received coil implantation. An aortogram taken before stent placement revealed an average aneurysm diameter of 61±15 mm (Table 1⇓; Figs 1⇓ left and 2 left). The aortogram obtained immediately after stent placement showed relatively faint contrast filling the aneurysm cavity, but the angiographic cavity size was unchanged or slightly decreased (Figs 1⇓ middle and 2 middle). In one pig (pig 7), an aortogram before stent placement revealed extensive retroperitoneal extravasation of contrast, consistent with aneurysm rupture (Fig 3A⇓). A stent was immediately placed across the aneurysm, followed by coil deployment. An aortogram after placement of the stent and coil showed a significant decrease in lumen size with no evidence of leaking and only residual contrast stain retained in the retroperitoneal space (Fig 3B⇓, 3C⇓, and 3D⇓). This pig survived the procedure and was euthanized 4 weeks later.
Control group. All pigs in the control group died. One pig died of distal bowel perforation at 2 days after surgery, and another died of undetermined cause (probably sepsis) at 3 days. The remaining 7 pigs died of spontaneous aneurysm rupture between 4 and 43 days (median, 8 days) after the initial surgery. The rupture site was found to be at the thinnest part of the aneurysm wall and none at a suture line. The aneurysm size was much larger (>200%) than that of the original fascia pouch, an observation consistent with rapid expansion.
Study group. Pig 1 died of aneurysm rupture 8 days after stent placement. The aneurysm diameter in this animal at the time of stent implantation was 87 mm, six times larger than the reference aorta. Pig 2 died of sepsis secondary to intra-abdominal abscess 29 days after stent placement. Pig 6, which received both stent and coil, died of anesthesia complications 12 hours after the procedure.
Follow-up aortograms were performed on 5 pigs: at 6 weeks (pig 3), at 3 months (pigs 4 and 8), and at 6 months (pigs 5 and 9); however, no follow-up aortogram was available in pig 7. In all cases, the lumen of the stent was angiographically patent, with no filling defects, and all arterial side branches were patent, with vigorous flow through these arteries (Figs 1⇑ right and 2 right). A decrease (28% to 65%) in aneurysm cavity size was observed in 4 of 5 pigs undergoing repeat aortogram. The 1 remaining pig (pig 9) treated with stent and coils had no angiographic evidence of an aneurysm at 6 months.
No migration of the stent was observed after deployment in any of the pigs. All stents were patent, with no fracture and no thrombus in the lumen. The stents were either partially or entirely covered by neointima; however, the orifice of aortic arterial branches remained patent.
In the group of 5 pigs that received stent alone, one pig (pig 1) died 8 days after the procedure. On autopsy, we found that the aneurysm had ruptured even though the stent had completely crossed the aneurysm. In another pig (pig 2), the gross external aneurysm diameter was 8 cm at 4 weeks after stent placement, which was the same as the angiographic diameter (8 cm) before stent placement. In the remaining 3 pigs (pigs 3 through 5), the aneurysm wall increased in thickness over time, decreasing the lumen volume (Fig 4⇓).
There was no aneurysm rupture in the 4 pigs that received stent and coils. The aneurysm lumen in pig 6 (dead of anesthesia complications 12 hours after the procedure) was completely thrombosed, and the coil was embedded in the thrombus; however, the intraluminal stent was patent and free of clots. In the remaining 3 pigs (7 through 9), we found that the external size of the aneurysms was smaller than that of the original pouch. The aneurysm wall was thick at 4 weeks after stent and coil placement (pig 7) (Fig 5⇓ top) and even thicker at 3 months (pig 8). We were unable to identify an obvious aneurysm in pig 9 at 6 months after the procedure, and all that remained was a small collagenous pouch containing the coils (Fig 5⇓ bottom).
The fascia used to construct the aneurysms consisted of three layers of collagen fibers. The middle layer was wider than the two outer layers and contained collagen fibers aligned perpendicular to those in the outer layers (Fig 6⇓ top). This characteristic trilayered structure proved useful in distinguishing the original fascia from new collagen fibers.
Observations in untreated aneurysms. In the cases in which rupture occurred within 1 week, we observed red blood cells, macrophages, and neutrophils within the fascia. In addition, the collagen bundles in the thick, middle layer of the fascia were frequently separated by red blood cells. There was always a thrombus adhering to the inside of the fascia; however, there was evidence of new collagen in only one specimen (6 days after surgery). In two cases, we observed collagen fibers in the fascia that were less bright than normal (ie, they had reduced birefringence when viewed with circularly polarized light) (Fig 7⇓). Such a reduction in birefringence is consistent with a breakdown of the subcellular structure of the collagen.
In two cases, rupture occurred relatively late, at 13 and 41 days. In both of these samples, we observed new collagen fibers. These fibers were green in the 13-day case, but some orange fibers were also present in the 41-day case. In both cases, we observed red blood cells, macrophages, and neutrophils within the wall.
Observations after stent placement. In the aneurysm that ruptured 8 days after stent placement, we saw no collagen in the aneurysm wall apart from that contained in the original fascia patch. In addition, we observed red blood cells, macrophages, and neutrophils that had infiltrated into the collagen of the fascia. At 4 weeks after stent placement, we found a very small amount of new collagen in the aneurysm wall. These fibers were green and were located at the border between the fascia and the adherent thrombus. At 6 weeks and 3 months, we observed more new collagen fibers both on the outside and on the inside of the fascia. These fibers were predominantly green, and although generally aligned circumferentially in the wall, were not particularly well organized (Fig 8⇓ top left). Fibroblasts could also be seen in the region of the border between thrombus and fascia. In contrast, at 6 months we observed considerable numbers of fibroblasts throughout a thick aneurysm wall. The collagen fibers were predominantly orange and were highly aligned in the circumferential direction (Fig 8⇓ top right). In addition, the collagen fibers were arranged in multiple layers (Fig 8⇓ top right).
Observations after stent and coil placement. In the pig that died 12 hours after placement of the stent and coil, there were red blood cells, macrophages, and neutrophils within the fascia. At 4 weeks, we found extensive new collagen both inside and outside the fascia layer. Most of these fibers were orange and were aligned in the circumferential direction. Collagen fibers were also found within the deepest regions of attached thrombus; however, these fibers were not so highly aligned and were generally organized into a lattice-like structure. We also observed fibroblasts and some neutrophils within the wall. At 3 months, we observed even more fibroblasts. The aneurysm wall was thick and contained several layers of organized orange collagen fibers (Fig 8⇑ bottom left). The aneurysm at 6 months no longer had the appearance of an aneurysm; instead, it consisted of a small nodule of connective tissue containing the coils. The histology of tissue from this sample revealed mainly highly aligned, orange collagen fibers (Fig 8⇑ bottom right).
Collagen fiber orientation. The average angular deviation of the collagen fiber orientation distributions decreased with time (Table 2⇓), which indicates that the fibers became progressively more aligned. Representative orientation distributions from the stent-treated cases are shown in Fig 9⇓. The average angular deviations of the distributions obtained from aneurysms treated with stents and coils were similar to or smaller than those obtained from aneurysms treated with stents alone (Table 2⇓). The average difference in orientation between the mean of each distribution and the tangent angle to the wall at the site at which the measurements were taken was 20±4° in the stent-treated aneurysms and 18±2° in the aneurysms treated with stents and coils. Therefore, the collagen fibers were generally aligned circumferentially in the aneurysm wall.
Collagen content. There was a progressive increase in the collagen content of the tissue with time in both stent-treated and stent-plus-coil–treated aneurysms (Table 2⇑). However, the collagen content in the aneurysms treated with both stents and coils was always higher, at any given time point, than in those treated with stents alone.
Optical properties of the collagen. The average fiber brightness increased in both groups with time after treatment and, as with collagen content, the values were consistently higher at any given time point in the group treated with both stents and coils (Table 2⇑). There was a significant correlation between collagen content and average fiber brightness in both groups (P<.01, data not shown).
We have demonstrated the effectiveness of endovascular uncovered stent placement, with and without additional embolization coil, for treatment of an animal model of AAA. In contrast to previous methods,2 3 4 5 6 7 8 9 10 11 we used a stent with no covering material. We also demonstrated, through serial histological analysis, that the aneurysm lumen was gradually replaced by scar tissue after stent deployment. Furthermore, the combination of stent placement plus embolization of the aneurysm lumen with coils appeared to accelerate this healing process.
Animal Model of AAA
To test the effectiveness of any aneurysm treatment, it is desirable to do so in a model in which rupture is likely. Our finding of a 78% (7 of 9; the other 2 pigs died of other causes, but the aneurysm size significantly increased) incidence of rupture in the untreated pigs indicates the applicability of our fascia pouch model. However, we used exactly the same method to create aneurysms in canine aortas and found no rupture (although the aneurysms increased in size). We speculate that the difference between the two studies results from structural differences in the fascia. Although the fascias in the two models were similar in thickness and were composed of three distinct collagen layers, the collagen fibers in the middle layer were more organized and tightly packed in the dog fascia than in the pig (Fig 6⇑ bottom). Such organization in the dog probably confers greater structural integrity than that of the pig fascia, in which we found the collagen fibers of the middle layer to be disrupted and pulled apart by infiltration of red blood cells. Thus, we propose that the precise structural configuration of the material in the wall of the model aneurysm is a crucial determinant of outcome.
Even though we were able to create an animal model in which aneurysm rupture did occur, rupture of the aneurysm presents a paradox. Collagen is a strong and stiff material that should, under normal circumstances, be expected to easily withstand the hemodynamic forces exerted on it in the model. How could an aneurysm constructed of collagen expand and rupture? We observed collagen fibers in several ruptured aneurysms that had reduced birefringence when viewed with polarized light (they appeared less bright). This optical property is an indicator of the subcellular structure of a material; a material with an anisotropic subcellular structure is birefringent and therefore appears bright when viewed with polarized light, whereas isotropic materials appear dark. Thus, a reduction in collagen birefringence indicates a breakdown of its normal anisotropic structure. Similar optical changes have been found after in vitro degradation of collagen,22 and the presence of collagen fibers with reduced or no birefringence was associated with ventricular expansion after myocardial infarction.23 Therefore, we propose that our observation is consistent with collagen degradation and may explain the propensity for expansion and rupture in our model. Although our model was designed to evaluate stent treatment, it may also provide an opportunity to study the pathophysiology of aneurysm growth and rupture.
Aneurysm Treatment With Uncovered Stents
Deployment of the uncovered stent across the mouth of the aneurysm decreased the incidence of rupture in pigs from 78% to 20% (1 of 5; the angiographic size of this aneurysm was 87 mm). One major advantage of the uncovered stent is that it permits blood flow through all arterial branches crossed by the stent, and so there is no risk of preventing flow to vital organs. In contrast, a covered stent cannot be allowed to cross major arteries; otherwise, severe complications will develop. In an animal study, Mirich et al2 found that a covered device caused renal artery occlusion and death after the stent migrated in a canine aorta. In a human trial, Dake et al7 reported one death secondary to paraplegia (probably because of spinal artery occlusion) after using a covered stent graft to treat thoracic aortic aneurysm. Of course, this advantage also means that the aneurysm is not immediately excluded from the circulation, and there is still blood flow into the lumen. However, the amount of flow into the lumen decreased, as shown by the reduced density of contrast in the aneurysm lumen. Such a decrease would be accompanied by a reduction in the pulsatile hemodynamic forces acting on the aneurysm wall, which would lessen the likelihood of rupture.
Over the 6-month follow-up period, the aneurysm lumen was first filled with thrombus and finally by fibrosis. The appearance of a thrombus is consistent with the reduced flow into the aneurysm lumen suggested by the aortograms. However, the presence of large amounts of thrombus has been associated with a greater likelihood of rapid expansion.24 Vorp et al25 speculated that the thrombus would limit the amount of oxygen reaching the aneurysm wall by increasing the diffusion distance. Such a lack of oxygen might then lead to smooth muscle necrosis and hence make the wall less resistant to distension. However, our stent-treated aneurysms all contained extensive thrombus, and only one huge aneurysm ruptured. The lack of rupture in our model with extensive thrombus may be because of the previously mentioned reduction of pulsatile forces after stent placement.
Aneurysm Treatment With a Combination of Stent and Embolization Coils
The deployment of coils after the stent was in position was feasible because the stent was uncovered and the size of the individual stent cells was sufficiently large to allow easy catheter-mediated placement. The rationale for the addition of coils was that their presence would accelerate thrombus formation and hence quickly eliminate the aneurysm lumen from the circulation.26 In fact, an extensive thrombus was found as early as 12 hours after treatment. Furthermore, the acute rupture that occurred in one pig (indicated by the leaking contrast media; Fig 3⇑) was successfully treated by stent and coil placement. There was no incidence of rupture in any of the four aneurysms subjected to this combined therapy, which further suggests that the presence of thrombus within the aneurysm lumen is not necessarily associated with an adverse outcome.
Long-term Histological Changes in Treated Aneurysms
We measured the orientation, content, and birefringence of the new collagen fibers that we observed to first infiltrate and then replace the thrombus. These parameters are known to increase in cases of wound healing; for example, in scar formation after acute myocardial infarction and in healing after tendon injury.17 21 In the treated aneurysms, the new collagen fibers became increasingly aligned, more densely packed, and thicker (denoted by both the color changes and the increased brightness seen with circularly polarized light). All of these changes will reduce the likelihood of rupture. An increase in fiber alignment will increase wall stiffness, whereas an increase in fiber content and thickness will increase wall strength. Furthermore, as the new fibers mature, there is an increase in intermolecular and intramolecular collagen cross-linking, leading to increase in tensile strength. In addition, as the wall thickens, there will be a decrease in wall stress according to Laplace’s law. Thus, we propose that our aneurysm treatment not only prevented early rupture but also allowed initiation of a healing process that further reduces the chances of expansion and rupture and will eventually result in complete replacement of the aneurysm lumen by fibrosis, as seen in one pig at 6-month follow-up. Although our sample size was too small to permit statistical comparison, the progression of healing appeared to be more advanced in aneurysms treated with stents and coils at all time points examined than in aneurysms treated with stents alone. This may indicate that early thrombus formation allows earlier onset of the repair process.
Even though healing was evident, the process appeared to be slow in comparison with other examples of wound healing. After acute myocardial infarction, complete replacement of the necrotic tissue by a firm, fibrous scar has occurred, even in large-animal models, after ≈6 weeks. We speculate that a lack of oxygen supply to the aneurysm, caused in part by limitation of oxygen diffusion from the circulation by the thrombus, is responsible. Similar slow healing has been reported in human aneurysms treated with coils. In two patients examined at necropsy 2 and 6 months after treatment, the resolution of thrombus within giant cerebral aneurysms (25 to 35 mm in diameter) was not complete.27 The authors speculated that the slow rate of healing was the result of the avascular nature of the aneurysm wall. Healing might be even slower in our model because the fascia lacks the vasa vasorum that would likely be present in a structure that contains former arterial tunica adventitia.
Relevance of the Fascia Pouch Model to Real Aneurysms
The fascia pouches, like actual aneurysms, have the capability of both expansion and rupture. In addition, the later appearance of the pouch, modified by the production of new collagen, is similar to that seen in human aneurysms. Specifically, unlike the normal artery wall, both the pouch model and real aneurysms are composed almost exclusively of collagen with little elastin or smooth muscle.28 The organization of this collagen into multiple distinct layers is also a common feature. Multiple layers of generally circumferentially aligned collagen fibers have been found in human cerebral aneurysms.29 30 We are unaware of published data regarding the organization of collagen fibers in human aortic aneurysms; however, in tissue obtained from two cases of aortic aneurysm resection, we found multiple layers of collagen fibers (Fig 10⇓ top). In addition, one of the pigs in our study was found to have a natural aortic aneurysm, which had a histological structure similar to that seen in the later stages of the treated model aneurysms (Fig 10⇓ bottom). These apparent similarities in structure indicate that our model may allow examination of the expansion, rupture, and healing phases of aneurysm. However, the dissimilarity between the pouch and normal aorta, which contains significant amounts of elastin and smooth muscle,28 means that the model has little relevance to the initiation of aneurysm formation.
Limitations of the Model
Despite the structural similarities described above, there are potentially important differences between human aneurysms and our model. Although we attempted to create a fusiform aneurysm to mimic the common appearance of human AAAs, their angiographic appearance tended to be more saccular in nature. Whether such differences will influence the effectiveness of stent treatment is unknown.6 Similarly, aneurysm size may be an important limitation of stent effectiveness. Although we substantially reduced the incidence of aneurysm rupture with stent treatment, it was not eliminated, and the one aneurysm that did rupture was very large. The addition of coils appeared to enhance stent treatment; however, our sample size was too small to definitively answer such questions. Even though our model appears to possess some features of human AAAs, no model can fully duplicate the human disease; for example, macrophages and lymphocytes rather than neutrophils are predominant in human AAAs.31 Thus, although our results are encouraging, an even longer follow-up period in a larger number of animals may be necessary. However, the limitations of the animal model may mean that once feasibility has been demonstrated, perhaps the only reasonable progression is evaluation in humans.6
Our study documents that uncovered stents, with and without additional embolization coils, can be used to effectively treat large AAAs in a swine model. In the untreated model aneurysms, the fascia pouch was disrupted and there was evidence of collagen degradation. Although we found some collagen fibers in the cases of late rupture, the degree of healing was minimal. In contrast, placement of stents prevented aneurysm expansion, reduced the incidence of rupture, and allowed healing to proceed so that the aneurysm wall became reinforced by multiple layers of highly aligned collagen fibers (Fig 11⇓). We speculate that the stents stabilize the aneurysms sufficiently to provide enough time for this relatively slow healing process to eventually fill the aneurysm lumen with fibrosis. The use of uncovered stents permits continued blood flow through arterial branches crossed by the stent and allows the additional deployment of embolization coils into the aneurysm lumen. This combination allowed us to successfully treat an acutely leaking aneurysm. Furthermore, the presence of embolization coils appeared to accelerate the rate of healing. We conclude that the long-term patency of both the uncovered stents and the arterial branches crossed by the stents, plus the reduction in the incidence of rupture via enhanced healing, support the initiation of clinical trial of this approach.
We thank Cook Inc, Bloomington, Ind, for both their grant support of the study and their contribution to the publication of the color photographs.
Reprint requests to Dr Carlos E. Ruiz, Division of Cardiology, Department of Medicine, Loma Linda University Medical Center and Children’s Hospital, Room 4002, 11234 Anderson St, PO Box 2000, Loma Linda, CA 92354-0200.
Presented in part at the 69th Scientific Sessions of the American Heart Association, New Orleans, La, November 11-14, 1996, and published in abstract form (Circulation. 1996;94[suppl I]:I-59).
- Received February 18, 1997.
- Revision received April 15, 1997.
- Accepted April 18, 1997.
- Copyright © 1997 by American Heart Association
Ruiz CE, Zhang HP, Douglas JT, Zuppan CW, Kean CJC. A novel method for treatment of abdominal aortic aneurysms using percutaneous implantation of a newly designed endovascular device. Circulation. 1995;91:2470-2477.
Ehrlich HP, Krummel TM. Regulation of wound healing from a connective tissue perspective. Wound Rep Reg. 1996;4:203-210.
Batschelet E. Circular Statistics in Biology. London, UK: Academic Press; 1981.
Whittaker P, Bougner DR, Kloner RA. Role of collagen in acute myocardial infarct expansion. Circulation. 1991;84:2123-2134.
Vorp DA, Federspiel WJ, Webster MW. Does laminated intraluminal thrombus within abdominal aortic aneurysm cause anoxia of the aortic wall? J Vasc Surg. 1995;23:540-541.
Massoud TF, Turjman F, Cheng J, Viñuela F, Guglielmi G, Gobin YP, Duckwiler GR. Endovascular treatment of fusiform aneurysms with stents and coils: technical feasibility in a swine model. Am J Neuroradiol. 1995;16:1953-1963.
Freestone T, Turner RJ, Coady A, Higman DJ, Greenhalgh RM, Powell JT. Inflammation and matrix metalloproteinases in the enlarging abdominal aortic aneurysm. Arterioscler Thromb Vasc Biol. 1995;15:1145-1151.