Bovine jugular covered stent-graft implanted in the swine inferior vena cava - a study of tissue response
(Portuguese PDF version)

Cristina Ribeiro Riguetti Pinto,1 Celso Luiz Muhlethaler Chouin,2 Gaudencio Espinosa Lopez3

1. MSc. student, Vascular Surgery Service, Department of General Surgery, Faculdade de Medicina, Universidade Federal do Rio de Janeiro, RJ, Brazil.
2. PhD student, Vascular Surgery Service, Department of General Surgery, Faculdade de Medicina, Universidade Federal do Rio de Janeiro, RJ, Brazil.
3. PhD, Associate Professor, Vascular Surgery Service, Department of General Surgery, Faculdade de Medicina, Universidade Federal do Rio de Janeiro, RJ, Brazil.

Correspondence:
Cristina Ribeiro Riguetti Pinto
Rua Raul Pompéia, 144/704 - Copacabana
CEP 22080-000 - Rio de Janeiro, RJ - Brazil
Tel: +55 (21) 9873.3328
Fax: +55 (21) 2249.2346
E-mail:crisriguetti@uol.com.br

ABSTRACT

Objective: To evaluate tissue response to a bovine jugular vein covered stent when implanted in the swine inferior vena cava.

Method: We developed a self-expanding stent, using a segment of L-hydro conserved bovine jugular vein, which was trimmed and sutured to a 316L stainless steel stent. We used the Taheri-Leonhardt delivery system for aortic stent-graft deployment (Florida, USA). Ten handmade stent-grafts were implanted in 10 swine inferior venae cavae. All animals were submitted to perioperative venography. At necropsy, 2 months later, the stent-grafts were removed en bloc and histopathologic analysis was undertaken, in order to analyze its patency, adherence to neighboring tissues and incorporation to the venous wall, as well as tissue response.

Results: All stent-grafts were patent and adherent to venous wall, but six presented with gross trabeculation and four had some degree of perivascular fibrosis at macroscopy. Three animals developed lymphocele, one in the retroperitoneal space and the others in the abdominal wall. At histopathology, we observed chronic inflammatory reaction with foreign body granulomatous response in all cases, with prevalence of the tunica media (80%).

Conclusion: The model presented low thrombogenicity, which corroborates the efficacy of the chosen means of preservation and material. However, there was low compatibility, probably due to the immunological obstacle of xenografts and exaggerated tissue response of the venous territory.

Key-words: Stent-graft, bioprosthesis, jugular veins, cattle, inferior vena cava, histocompatibility.

J Vasc Bras. 2006;5(2):81-8

Article submitted April 18, 2006, accepted June 12, 2006.

Treatment of traumatic lesions and stenotic or venous occlusive processes is a difficult issue that generates therapeutic controversy.1-4 Because it is a vascular territory whose lesion presents lower clinical impact and slower progress in relation to the arterial lesion, there is a lack of experience and consensus in the literature. With the advent of endovascular surgery, a minimally invasive technique, indication for this type of intervention in the venous territory has been growing.2,11

Percutaneous venous transluminal angioplasty may be efficient for some months, but stenoses are generally very resistant, with a high level of primary inefficacy and short- and medium-term recurrent stenosis.12 Stent-grafts represent a certain progress, but not a definitive one. Most patients, except in cases of palliative treatment for neoplastic stenoses, will have to undergo several endovascular interventions to maintain patency of their venous system and to prevent possible thrombosis, which may have drastic consequences in case it is a severe one. However, outcomes, even being inferior to arterial outcomes, are similar to those obtained with conventional surgeries of great vessels of the superior and inferior vena cava systems.12,13

Biological materials to develop stent-grafts have only recently been investigated.14 The vein was one of the first grafts to be used, in its autologous or denatured heterologous form.15 Although autologous venous grafts present the best results, they require surgical procedure for collection and depend on patient's venous capital.16 Nevertheless, biological grafts offer some qualities that synthetic polymers do not have, such as resistance to infection, low profile, increased complacency, and higher biocompatibility.16

A series of complex cellular and molecular processes contributes to the healing of stent-grafts. These healing processes reflect the tissue response to the implantation of a foreign body in the vascular system in four types of tissue formation: thrombus, neo-intima, endothelium, and inflammatory cell infiltrate.16,17 Extension and location of each of the four tissue types are influenced by a number of mechanical factors related to the stent-graft implantation, location of the stent metal arms, type of polymer or biological material of the stent-graft, and conservation method.16

This study was carried out to evaluate tissue response to the implantation of a self-expanding stent covered by a segment of heterologous vein preserved in a special solution by the L-hydro process.

MATERIAL AND METHOD

This study was carried out at the research center of Labcor Laboratórios, in Santa Luzia (MG, Brazil), and was approved by the ethics committee for laboratory animal research of the Department of Surgery at UFRJ and by the committee for animal care and use of the research center of Labcor Laboratórios, in Belo Horizonte (MG, Brazil).

Stent-grafts

The stent-graft was structured by a stent composed of three wire rings with round section of 0.018", made of 316 L stainless steel, zigzag folded like a "napkin strap," and longitudinally interconnected by a wire made of the same material and profile (connecting bar). Each stent ring was approximately 12 mm long, with a total of 10 equidistant folds, resulting in five crests and five troughs. The whole device was in average 63 mm (47 mm-80 mm) long and 15 mm (10 mm-22 mm) in diameter. Stent diameters were appropriate to the available bovine jugular segments. On the other hand, stent-graft diameters, chosen for each animal, were determined according to the protocol established by Gomez-Jorge et al. and empirically confirmed by the evaluation of the vena cava external diameter under direct visualization.18

Bovine jugular segments were removed from animals killed at an adult age, at a slaughterhouse in Santa Luzia (MG, Brazil), after approval by the Ministry of Agriculture and Health Surveillance. After being separated, in order to eliminate tissue excess, and set up in support, the segments received a preservation treatment by the L-hydro process.

The L-hydro preservation process is divided into three different stages. The first one combines the extraction of antigens, under controlled conditions, without the use of detergents, surfactants or digestive enzymes, with masking of remaining antigens by polyglycol, under controlled chemical oxidation. This stage is performed under specific physical conditions, which protect the extracellular components, such as collagen and elastin. The second stage consists of a process of incorporating a non-steroid anti-inflammatory agent (equivalent to the acetylsalicylic acid) and an anti-thrombotic agent (equivalent to heparin) to the tissue. The third and last stage is to perform the tissue sterilization using vapor-phase hydrogen peroxide (H2O2).

To implant the stent-graft, we used the Taheri-Leonhardt delivery system for aortic stent-graft deployment (Florida, USA), composed of a coaxial sheath with an internal pusher rod (Figure 1). The systems used varied in caliber (18-24F), being chosen according to the rule "n + 4" developed in vitro by Gomez-Jorge et al..18

click hereFigure 1 - Stent-graft assembled in the deployment system


C = sheath cone; R = stent rings; P = pusher rod.

Sample

Ten hybrid female swine (crossing between Landrace and Large White) were used, originated from Fazenda Córrego Fundo, Brumadinho (MG, Brazil). All animals were healthy, aged 3 months and had approximate mean weight of 26 kg, ranging from 20-35 kg. The animals were given swine ration for growth Socil Guyomarc'h, Cevadil 12, 600 mg/day and water ad libitum.

Surgical procedure

The animals were submitted to a 12-hour fasting for solid diet and a 6-hour fasting for liquid diet. Venous access was the marginal vein of the right ear, and the following drugs were used for anesthesia: intramuscular (IM) atropine sulphate 0.04 mg/kg 20 minutes before induction; IM ketamine chloride 7.8 mg/kg; IM xylazine chloride 2.2 mg/kg; endovenous sodium thiopental 8 mg/kg and adjusted as required; and succinylcholine 1 mg at the beginning of the procedure.

We chose the retroperitoneal approach, using an arcuate oblique abdominal incision with medial concavity in the right lower quadrant. After exposure, dissection, and repair of the external, internal and common iliac veins, the animal was anticoagulated with sodium heparin 350 IU/kg. The right common iliac vein was punctured and a short 5F sheath was introduced using the Seldinger technique. Phlebography was also performed to establish proximal (renal artery) and distal (iliac bifurcation) limits for graft placement. After distal clamping of vessels, a deployment system with J-tipped Teflon-coated guide wire 0.035" x 145 cm was introduced under fluoroscopy by longitudinal phlebotomy. After graft deployment, a control phlebography was performed by the deployment system distally retracted and further phleborrhaphy with 7-0 polypropylene suture (Figure 2). Anticoagulation was reverted with protamine chloride at 1 ml for each 1,000 IU of sodium heparin, and the abdominal wall was sutured by layers.

click hereFigure 2 - Stent-graft implanted into the infrarenal vena cava


LRV = left renal vein; RRV = right renal vein; 1, 2, 3 = stent rings.

Postoperative period

After the surgery, the animals were kept under room temperature in individual cages during the first days, and were roomed in for the rest of the period. They received anti-inflammatory medication for 3 consecutive days (flunixin meglumine 50 mg/day) and antibiotics for 7 days (enrofloxacin), 5 mg/kg/day.

Sacrifice

Two months after graft implantation, the animals were anesthetized using the same technique and sacrificed with intravenous injection of 10% potassium chloride in bolus dose of 20 ml. The inferior vena cava segment was removed en bloc and fixated with a solution of 10% formaldehyde.

Macroscopic analysis

According to the evaluation of internal and external tissue characteristics, a scoring scheme was established to compare the final version of the results. Reaction of tissue fibrosis with adherence to adjacent structures was scored as 0 when it did not occur, 1 when it was moderate, and 2 when it was intense. Luminal patency was scored as 0 when it was total, 1 when there were trabeculations, and 2 when there was total vessel occlusion. Stent-graft incorporation received 0 when it was totally incorporated into the native vessel, 1 when it was partially incorporated (detachable during handling), and 2 when it was not incorporated (detached).

Microscopic analysis

The stents were deployed using microsurgical technique, and cross-sectional cuts were performed in the segments in contact with the stent ring (intersegments) and outside the stent rings (free segments). The segments were included in blocks of paraffin and were later submitted to histological sections of 4 µm and stained with hematoxylin-eosin.

The macroscopic analysis quantified the inflammatory reaction in the intersegments and free segments, being considered mild when ≤ 30%, moderate when > 30% and ≤ 70% and severe when > 70%.

RESULT

Postoperative evolution

All 10 stent-grafts were successfully deployed. There was an unexpected placement, with the most distal ring in the right common iliac vein, but it remained asymptomatic. All animals remained alive until sacrifice. One animal presented abdominal wall abscess, which regressed under antibiotic therapy. Three animals presented lymphocele, one moderate with spontaneous drainage, one giant in the abdominal wall, and another giant lymphocele in the retroperitoneal cavity; two animals presented granuloma in the distal and external tissue wall, and one of them developed perigraft lymphorrhea.

Macroscopy

At necropsy, there was no case of vascular occlusion, and most tissues (60%) were classified as partially patent (Figure 3), and the others (40%) as totally patent, as shown in Table 1. We found a good dissection plan in more than half (60%) of the group under investigation: in two animals it presented light adherence, in both cases close to the iliac bifurcation; in two other animals it presented intense adherence, as shown in Table 1. Although all stent-grafts seemed to be totally incorporated into the venous wall, three of them proved to be detachable during handling, i.e., partially incorporated, according to Table 1.

click hereTable 1 - Macroscopic findings

Number of the animalPatencyAdherenceIncorporation
1110
2100
3100
4000
5001
6120
7001
8120
9000
10111
Patency: 0 = complete; 1 = partial; 2 = occlusion; adherence to perivascular fibrosis: 0 = none, 1 = mil, 2 = intense; graft incorporation to the vessel: 0 = total, 1 = partial, 2 = none.

 

click hereFigure 3 - Macroscopy of a tissue longitudinally opened (infrarenal vena cava), showing the luminal trabeculations, equivalent to the partial patency

Microscopy

We observed the presence of a chronic inflammatory process with foreign-body granulomatous reaction in all tissues (Figure 4). The inflammatory response was predominantly located in the medial layer in 80% of the cases, and there was absence of the intimal layer in the last four cases, as shown in Table 2. The results of response quantification showed a slight difference between the intersegmental and free segmental sections, being severe in 40% of free segments and in 50% of intersegments; mild in 30% of free segments and 20% of intersegments; and equally moderate in 30% of both segments, as shown in Table 2. Therefore, severe response prevailed in one case of intersegmental sectioning, and mild response prevailed in one case of free segmental sectioning.

click hereFigure 4 - Microscopy with giant cells imaging (arrows)

click hereTable 2 - Microscopic findings

Number of the animalFree segmentsIntersegmentsPredominant sitePresence of intima
1sevsevM-Ayes
2sevmodMyes
3modsevMyes
4sevsevM-Ayes
5sevsevIyes
6modmodIyes
7milmodMno
8milmilMno
9modsevMno
10milmilMno
Quantification of inflammatory reaction in the segments in contact with the stent ring (intersegments) and with no contact (free segments): severe (sev), moderate (mod), mild (mil) response; predominant site of reaction: intima (I), media (M), adventitia (A), presence (yes) or absence (no) of the intimal layer.

DISCUSION

Animal models have been used for researching vascular disease since the beginning of the 20th century. Nevertheless, there is no ideal one, since no model perfectly reproduces the human response to the disease. Histopathology suggests that the peripheral vessels of domestic pigs and rabbits are those that are most similar to the ideal model, since they present similar size and vascular access to human's, besides showing substantial amounts of elastin and having comparable intima, media, and adventitia.19 However, healing responses are less intense in humans when compared to other species.20 In addition, pigs have a tendency to hypercoagulability and a fibrinolytic system that is not so active.21

In our study, we used pigs due to the anatomical, morphological, and histological similarity of the vascular system, besides being easy to acquire. We had no problems with body weight, since the experiment was planned to last for 2 months, which did not cause any difficulties in handling with the pigs. Although pigs have a higher tendency to hypercoagulability and a less active fibrinolytic system when compared to humans, there was no case of severe graft thrombosis. At the end of the experiment, all grafts were patent or partially patent. In our opinion, this fact demonstrated that the bioprosthesis used is not very thrombogenic.

We performed a qualitative and quantitative approach, describing occurrence, type and intensity of the inflammatory response, considering it was an initial assessment of a new stent-graft, and not of a device being clinically used. Therefore, establishing a 2-month period for the research allowed the analysis of early mortality, complications related to the procedure, and macro- and microscopic alterations in the tissue, such as patency and tissue response.

In feasibility studies of pharmacological coronary stents, devices that are already in regular clinical use, evaluation of the degree of inflammation is required, including an injury scaling at each site with metal, description of the inflammation (absent or cellular type and location), and inflammation scaling for the whole vessel, making a distinction between layer and region.19,21 In our study of venous tissue response, and not of the feasibility of a device in clinical use, we observed the degree of inflammation by applying an inflammatory reaction scaling in the intersegments and free segments. In studies on coronary devices, presence of inflammation or fibrin deposit must be in acceptable levels, occurring in up to 20% of the sections and described as mild or moderate, without accelerating or causing substantial vascular lesion or stenosis.19 In our study, we observed foreign-body granulomatous inflammatory reaction in all the cases, being severe in half of the intersegments and in less than half (40%) of the free segments. Vascular lesions, luminal gross trabeculations, characterized as stent-grafts with partial patency, were found in 60% of the tissues. Therefore, the degrees of inflammatory response were not in accordance with acceptable degrees reported in the literature for coronary disease, macroscopically resulting in gross vascular lesions in more than half of the cases.

We tried to minimize the device patency by reducing the mechanical trauma using microsurgical technique of graft preparation, not performing predilatation, and using a solution of low oncotic pressure associated with antithrombotic therapy through the L-hydro conservation technique. However, some studies seem to demonstrate that neointimal hyperplasia occurs more evidently in the presence of low flow velocity and at a lower degree with the circumferential deformation of the vascular wall.22,23 Flow velocity would be directly related to the shearing force of the blood and vessel wall, which is a factor that determines the adherence probability and duration of blood elements to the intima.22 In our study, the stent-graft implantation was performed in the swine venous system, with low shearing force and lower flow velocity, which probably promoted high local proliferative response.

There are several types and quality of vascular bioprosthesis that are available and being tested, as well as several preservation methods. They are basically divided into autografts, allografts, and xenografts. Autografts and allografts, although having better histocompatibility, are less available, which restricts their use. Much attention has thus been given to xenografts and to different preservation methods. Phylogenetic distance and, therefore, degree of disagreement between antigens with higher histocompatibility seem to be determinant factors in the intensity of their rejection reaction.24 Non-human primates are, at first sight, the best donors; however, the swine donor seems to be the most appropriate candidate for a source of vascular xenografts, since it is easy to procreate and be genetically manipulated, besides presenting similarity with the human cardiovascular system and lower incidence of transmitted zoonoses.24

The recurrent stenosis related to stent implantation in the arterial bed seems to have a strong correlation with the lesion severity and the subsequent neointimal thickening and stenotic percentage.25 This is a deep medial lesion, being predominantly induced in the endovascular procedure. Events that occur minutes after this procedure are: platelet adhesion and aggregation and in situ thrombosis.26 Hypertrophy and proliferation of medial smooth muscle cells (SMC) occur after 24-48 hours.25,26 Although the period is species-dependent, approximately 4-14 days after the lesion, the migration of these cells to the intima and their proliferation occur under the effect of SMC mitogens.26 Deposition of extracellular matrix and intimal thickening characterizes the last stage of pathogenesis, ranging from 14 days to 3 months.25,26

Factors that would cause recurrent stenosis seem to be multifactorial, including the implant technique, hemorheologic factors, stent material and size, geometry of stent struts, and location of the disease in the vascular system. Venous wall lesion due to stent struts may be one of the related factors.27 SMC proliferation is proportional to the degree of arterial lesion caused by a balloon or stent.27 Although the association between vascular wall lesion and SMC proliferation is well known, little is known about strategies to minimize the stress under the tunica media and internal elastic lamina during an endovascular procedure.

With regard to types of stent, self-expanding stents have been associated with lower intimal trauma and consequent neointimal hyperplasia, when compared to balloon-expanding stents.28 In this study, the radial force of the self-expanding stent was intentionally used to reduce the luminal stenosis produced by the thickness of the venous graft cover, as well as to cause a better adaptation to the highly complacent venous environment.

Toyota et al., comparing external and internal graft coverings in relation to the stent, noted that the internal covering presented worse outcome, with major occlusion or stenosis as a consequence of intense intimal hyperplasia.29 This intimal hyperplasia has led to a higher inflammatory reaction and to a late and incomplete endothelization, when compared to the stent-graft with external graft. In the present study, we opted for an internal fixation of the venous xenograft, which might have accounted for the poor graft endothelization, for the three cases of incomplete adherence to the vena cava wall with detachment during handling, and for the prevalence of partial patency or gross trabeculations.

However, the internal covering would prevent the luminal contact with the metal, thus promoting a smooth internal surface, with improvement in long-term results.15,17 Nevertheless, the main limitation is the thickness of the venous graft, which causes venous stenosis. In our study, protection against contact of venous flow with the stent was probably the explanation for the 0% occlusion rate. On the other hand, the contact of the stent with the venous wall resulted in a higher tissue response on these sites. We also had difficulty assembling the graft in the deployment system due to the thickness of the venous graft.

CONCLUSION

In this study, we attempted to associate a xenograft to a stent in order to increase efficacy in the treatment of venous diseases, thus promoting a less invasive approach. However, we came across the immunological obstacle of xenografts, which has not been overcome yet, besides the great challenge of the venous territory with tissue responses exaggerated by the low flow and decreased shearing force. We verified foreign-body granulomatous inflammatory reaction in all the cases, being severe in most segments in direct contact with the metal (intersegments). Besides the exaggerated tissue response, we macroscopically verified gross trabeculations in 60% of the cases, which compromised the lumen of the segments under investigation, thus characterizing partial patency in most of them. The internal fixation of the graft to the endoprosthesis resulted in poor endothelization, less adherence to the vena cava wall, and has probably affected patency. Although there has been no occlusion due to thrombosis or neointimal hyperplasia, which corroborates the efficacy of the chosen means of preservation and material, results point to a low biocompatibility of the device under investigation.

 

REFERENCES

1. Hurlbert SN, Rutherford RB. Subclavian-axillary vein thrombosis. In: Rutherford RB. Vascular surgery. 5th ed. Philadelphia: WB Saunders; 2000. p. 1208-21.

2. Raju S, Owen S Jr, Neglen P. The clinical impact of iliac venous stents in the management of chronic venous insufficiency. J Vasc Surg. 2002;35:8-15.

3. Yamakado K, Nakatsuka A, Tanaka N, et al. Portal venous stent placement in patients with pancreatic and biliary neoplasms invading portal veins and causing portal hypertension: clinical experience. Radiology. 2001;220:150-6.

4. Yilmaz S, Erdogan A, Luleci E. Transvenous embolization and stent placement for an internal iliac arteriovenous fistula with central iliac vein occlusion. J Vasc Interv Radiol. 2004;15:399-404.

5. Cronin P, McPherson SJ, Meaney JF, Mavor A. Venous covered stent: successful occlusion of a symptomatic internal iliac arteriovenous fistula. Cardiovasc Intervent Radiol. 2002;25:323-5.

6. Dheer S, Joseph AE, Drooz A. Retroperitoneal hematoma caused by a ruptured pelvic varix in a patient with iliac vein compression syndrome. J Vasc Interv Radiol. 2003;14:387-90.

7. Kaneoka Y, Yamaguchi A, Isogai M, Hori A. Intraportal stent placement combined with portal vein embolization against advanced gallbladder carcinoma. Surg Today. 1998;28:862-5.

8. Kreienberg PB, Chang, BB, Darling RC 3rd, et al. Long-term results in patients treated with thrombolysis, thoracic inlet decompression, and subclavian vein stenting for Paget-Schroetter Syndrome. J Vasc Surg. 2001;33:S100-5.

9. Savader SJ. Inferior vena cava filters. In: Trerotola SO, Savader SJ, Durham J. Venous Interventions. Fairfax: SCVIR Syllabus; 1995. p. 276-92.

10. Sharafuddin MJ, Sun S, Hoballah JJ, Youness FM, Sharp WJ, Roh BS. Endovascular management of venous thrombotic and occlusive diseases of the lower extremities. J Vasc Interv Radiol. 2003;14:405-23.

11. Yamakado K, Nakatsuka A, Tanaka N, Fujii A, Terada N, Takeda K. Malignant portal venous obstructions treated by stent placement: significant factors affecting patency. J Vasc Interv Radiol. 2001;12:1407-15.

12. Picard JD, Barrelier MT, Joffre F. Veia normal. In: Picard JD, Barrelier MT, Joffre F. A veia - sua imagenologia. São Paulo: Lemos; 2003. p. 14-40.

13. Stock KW, Jacob AL, Proske M, Bolliger CT, Rochlitz C, Steinbrich W. Treatment of malignant obstruction of the superior vena cava with the self-expanding wallstent. Thorax. 1995;50:1151-6.

14. Schachner T, Zou Y, Oberhuber A, et al. Local application of rapamycin inhibits neointimal hyperplasia in experimental vein grafts. Ann Thorac Surg. 2004;77:1580-5.

15. Schellhammer F, Haberstroh J Wakhloo AK, Gottschalk E, Schumacher M. Vein graft coated vascular stents: a feasibility study in a canine model. Cardiovasc Intervent Radiol. 1998;21:158-64.

16. Dolmatch BL. Healing response to vascular stent-grafts. J Vasc Surg. 2000;31:1285-9.

17. Schellhammer F, Wahkloo AK, Nagursky H, Schumacher M. Vein graft-coated stents for endovascular occlusion of canine experimental arteriovenous fistulae. Invest Radiol. 1999;34:22-7.

18. Gomez-Jorge J, Venbrux AC, Magee C. Percutaneous Deployment of a valved bovine jugular vein in the swine venous system: a potential treatment for venous insufficiency. J Vasc Interv Radiol. 2000;11:931-6.

19. Schwartz RS, Edelman ER, Carter A, et al. Preclinical evaluation of drug-eluting stents for peripheral applications: recommendations from an expert consensus group. Circulation. 2004;110:2498-505.

20. Mason RG, Read MS. Some species differences in fibrinolysis and blood coagulation. J Biomed Mater Res. 1971;5:121-8.

21. Narayanaswamy M, Wright KC, Kandarpa K. Animal models for atherosclerosis, restenosis, and endovascular graft research. J Vasc Interv Radiol. 2000;11:5-17.

22. Dobrin PB, Littooy FN, Endean ED. Mechanical factors predisposing to intimal hyperplasia and medial thickening in autogenous vein grafts. Surgery. 1989;105:393-400.

23. Morinaga K, Eguchi H, Miyazaki T, Okadome K, Sugimachi K. Development and regression of intimal thickening of arterially transplanted autologous vein grafts in dogs. J Vasc Surg. 1987;5:719-30.

24. Miller VM, Reigel MM, Hollier LH, Vanhoutte PM. Endothelium-dependent responses in autogenous femoral veins grafted into the arterial circulation of the dog. J Clin Invest. 1987;80:1350-7.

25. Ip JH, Fuster V, Badimon L, Badimon J, Taubman MB, Chesebro JH. Syndromes of accelerated atherosclerosis: role of vascular injury and smooth muscle cell proliferation. J Am Coll Cardiol. 1990;15:1667-87.

26. Phillips-Hughes J, Kandarpa K. Restenosis: pathophysiology and preventive strategies. J Vasc Interv Radiol. 1996;7:321-33.

27. Rogers C, Edelman ER. Endovascular stent design dictates experimental restenosis and thrombosis. Circulation. 1995;91:2995-3001.

28. Wakhaloo AK, Schellhammer F, de Vries J, Haberstroh J, Schumacher M. Self-expanding and ballon-expandable stents in the treatment of carotid aneurysms: an experimental study in a canine model. AJNR Am J Neuroradiol. 1994;15:493-502.

29. Toyota N, Pavcnik D, VanAlstine W, et al. Comparison of small intestinal submucosa-covered and noncovered nitinol stents in sheep iliac arteries: a pilot study. J Vasc Interv Radiol. 2002;13:489-98.


J Vasc Bras. - Official Publication of the Brazilian Society of Angiology and Vascular Surgery