- Email: [email protected]
Bioabsorption qualities of chitosan-absorbable vascular templates1
GARY P. WRATTEN SURGICAL SYMPOSIUM: PART 2
Bioabsorption Qualities of ChitosanAbsorbable Vascular Templates Mohamad I. Haque, CPT, MC, USA,* Alec C. Beekley, CPT, MC, USA,* Anna Gutowska, PhD,§ Ruth Ann Reardon, CPT, MC, USA,† Paul Groo, LTC, MC, USA,† Sean P. Murray, MAJ, MC, USA,‡ Charles Andersen, COL, MC, USA,* and Kenneth Azarow, LTC, MC, USA* Departments of *Surgery, †Pathology, and ‡Radiology, Madigan Army Medical Center, Tacoma, Washington and the §United States Pacific Laboratory Northwest/Battelle Corporation, Richland, Washington PURPOSE: The scope of endovascular surgical techniques has
expanded to include the treatment of diseases considered at one time to be amenable only to surgical treatment. The development of the biodegradable template follows as an extension of current permanent stent technology. The goal of our project is to develop and test chitosan as an absorbable template for the vascular system. METHODS: Ultrapure chitosan, heparin sodium salt and ly-
sozyme, and contrast agents MD-76R and Oxilan-350 were used to give radioopaque quality. Prototype chitosan vascular templates were obtained by a dip coating method in which alternate layers of chitosan were coagulated with nonsolvents or heparin. The amount of loaded and released heparin was determined using Azure II colorimetric assay. In vitro enzymatic degradation of templates was evaluated using lysozyme solutions in phosphate buffered saline. Mechanical properties were analyzed using the Dynamic Mechanical Analyzer, DMA-7 (Perkin Elmer, Foster City, Calif.). The microstructure of freeze-dried templates was investigated by field emission scanning electron microscopy (FE SEM) using an LEO 982 electron microscope (Zeiss, Thornwood, NY). In vivo deployment of the templates was undertaken in 10 full-sized pigs (Sus scrofa). After open expose and control of the iliac artery, a closed balloon catheter technique was used to advance and place the balloon catheter and template. The balloon was then expanded, deploying a Palmaz stent with a chitosan template anchored distally. Patency and deployment of the stent-template complex was confirmed by an arteriogram. The animals were sacrificed at 1, 2, 3, 4, and 5 weeks poststent placement, and arterial sections were taken for microscopic
Correspondence: Inquiries to Kenneth Azarow, LTC, MC, USA, Madigan Army Medical Center, MCHJ-SGY, MAMC, Tacoma, WA 98431; fax: (253) 968-0232. Support: This study was supported by a grant from the National Medical Test Bed. This article contains the opinions of the authors only and does not represent the opinion of the United States Department of Defense or the United States Army.
CURRENT SURGERY • Published 2001 by Elsevier Science Inc.
analysis. The amount of chitosan remaining was estimated to determine an in vivo rate of absorption. RESULTS: On hematoxilyn and eosin staining of the section arterial samples, a marked inflammatory response was noted and progressed with duration of in vivo contact. A giant cell foreign body reaction coupled with intense intimal hyperplasia and organized thrombus was also noted and progressed with duration of time in vivo. Also noted was the degradation of the template material with only small remnants of material noted within the giant cell by week 4. Clinically, none of the pigs developed limb ischemia or evidence of thromboembolic events. CONCLUSIONS: In this in vivo study, the chitosan template
proved to be biodegradable but elicited an intense thrombotic and foreign body reaction despite heparin bonding. Further investigation is ongoing as to decreasing the thrombogenic and antigenic qualities of the template materials by either alteration of the base material or addition of bioactive side chains. (Curr Surg 58:77-80) KEY WORDS: chitosan, endovascular surgery, vascular stents
INTRODUCTION Over the last 3 decades, the explosion in the field of endovascular surgery has led to rapid advances in the biomaterials used for intraluminal deployment. The mainstay of this technology has focused on permanent, expandable metallic stents for use virtually anywhere in the vascular tree. This technology has been applied to the therapy of both occlusive and aneurysmal disease of the coronary, carotid, renal, iliac arteries, and, most recently, aortic repair. Since the early 1980s, little work has been done in the area of degradable stents (templates) for use in the repair of arterial or venous injuries. A single study from 1982 described the unsuccessful attempts at vicryl stents in porcine coronary arteries.1 Stent grafts with harvested saphenous vein, polytetrafluoroethelene, or bioengineered endothelial cells have been devel0149-7944/01/$20.00 PII S0149-7944(00)00422-0
77
FIGURE 1. Chitosan: (1,4)-linked 2-amino-2-deoxy-B-D-glucan, negatively charged organic compound.
oped and used in clinical trials; however, these leave permanent material within the lumen of the vessel.2,3 The application for a vascular template with reliable bioabsorption qualities has a particular application in the setting of vascular trauma, especially in blood vessels that are difficult to approach through open surgical techniques. In this setting, endovascular techniques could be used to place a template across the injured vessel and to serve as a bridge for regeneration of the intimal surface while preventing exanguinating hemorrhage. As the vessel heals, in vivo serum enzymes degrade the template material until it is completely degraded or incorporated into the vessel wall . The goal of this project was to study the bioabsorption qualities of a novel template material within the vascular tree of a porcine model. Gutowska describes the development of the chitosan templates.4 Beekley describes the porcine model and endovascular technique for deployment of these templates.5
MATERIALS AND METHODS Ultrapure chitosan was purchased from Carbomer (Westborough, Mass) (Fig. 1). Heparin sodium salt and lysozyme was purchased from Sigma Chemical Company (St. Louis, Mo). Oxilan-350 nonionic x-ray contrast was obtained. The templates were produced by alternate dipping of thin glass rods (2-mm diameter) in an ultrapure chitosan solution and a dilute acetone solution. This was then alternated with a heparin salt solution to produce a heparin-impregnated template. The templates were then exposed to in vitro enzymatic degradation studies in a lysozyme solution. Before phase 2, a series of templates was then created using a sterile technique and soaked in Oxilan-350 until the moments before in vivo deployment.4 Ten adult pigs (Sus scrofa) were brought to a veterinary operating suite with fluoroscopic capability, placed under general anesthesia, systemically heparinized, and surgical exposure of the left external iliac artery was obtained. Through a transverse arteriotomy, sequential dialators were introduced distally until a 10 French sheath could be easily introduced. The balloon catheter was then loaded with a Palmaz (Johnson & Johnson, Warren, NJ) stent (2.5 mm ⫻ 15 mm) and the chitosan template advanced over the stent such that 5 mm of the templatestent complex overlapped, and this complex was then deployed under fluorscopic guidance. Predeployment and postdeployment arteriograms were performed. The animals were then sacrificed at weeks 1, 2, 3, 4, and 5 postdeployment, and the entire stent-template complex was removed en block and preserved in formalyn.5 Sections were then taken 2 to 4 mm distal to the 78
FIGURE 2. Intense foreign body giant cell reaction: eosinophilic chitosan (a) within the hypercellular granuloma and foreign body giant cell at medial aspect of the granuloma (b).
palpable end of the Palmaz stent and analyzed by hematoxylin and eosin staining.
RESULTS Clinically, none of the animals studied displayed evidence of sepsis, acute limb ischemia, or thromboembolic phenomenon. Nine of 10 stent-template arterial sections were recovered and studied microscopically. Study of these templates revealed an intense foreign body reaction (Fig. 2) and severe intimal hyperplasia (Fig. 3). These templates were degraded in a nonuniform manner (Fig. 4, left and right). It appears that the entire resorption took place at the week 5 time frame. The only stent and template not recovered was at week 5. Qualitative analysis of chitosan detected in lumen reflected as percent degradation (Fig. 5). Although the absorption over time was unequal throughout the circumference and length of the template, the rate of total resorption was equivalent between each pair of animals sacrificed at the specific intervals measured.
FIGURE 3. Intimal hyperplasia: proliferative effect of chitosan on intima of vessel (a) in a 5th-week harvest (no intraluminal chitosan noted). CURRENT SURGERY • Volume 58/Number 1 • January/February 2001
FIGURE 4. Organized thrombus: extent of organized thrombosis within the lumen of the vessel with classic evidence for organized thrombosis. (Left) Two weeks in vivo. (Right) Four weeks in vivo.
DISCUSSION Chitosan as an absorbable vascular template material proved to have many desirable qualities in the in vitro stage of the study. Its biodegradability and saturation quality with heparin are novel qualities to current stent technologies and initially allowed for deployment and confirmation of arterial patency immediately post-template deployment. Furthermore, the negative charge of this organic compound contributes to its antithrombogenic quality. Unlike in commercially expandable metallic stents, well described by Palmaz, these stents carry an obligate positive charge that elicits an irregular clot between the strut indentations and is attributed to early stent thrombosis or restenosis at the edge of the intimal-stent interface.6 Chitosan’s negative charge carries the theoretical advantage of repelling positively charged members of the clotting cascade and platelets. However, despite these positive actual and theoretical qualities, chitosan proved not to be an inert substance in vivo, as evidenced by an intense inflammatory response not seen in current metallic stents and stent grafts. This foreign body reac-
FIGURE 5. Estimated percent degradation in templates over study period. CURRENT SURGERY • Volume 58/Number 1 • January/February 2001
tion within the vessel lumen and its surrounding thrombosis negates 2 of the theoretical advantages of the material. In vitro, chitosan was degraded uniformly by enzymatic degradation. In vivo, however, the exorbitant cellular response to the material and resultant occlusion of the vessel lumen argues against uniform purely enzymatic degradation of the template. The very type of reaction (foreign body) further disproves that chitosan templates (in their current form) are biologically inert in vivo. Thrombosis and intimal hyperplasia are complications well established in endovascular literature. In fact, much of stent research is focused on the modification of metallic stents to decrease the extent of intimal hyperplasia induced to include irradiation, coating with inert, noncharged polymers as well as modification of stent geometry.7,8 The degree of intimal hyperplasia observed in this model is not different from those seen in a canine femoral artery study performed with metallic stents.9 Unlike metallic stents, chitosan’s multiple-side chemical bonding sites allow for potential structural modification of the molecule to hide the antigentic and inflammatory foci that elicit the neointimal response. This concept argues for the rationale for the development of stents to serve only as a temporary scaffold for the repair of an injured vessel. Work by Sala and associates describes metallic stents within undiseased canine femoral vessels and intimal hyperplasia at the proximal and distal edges of the stent.9 This reaction alone argues for the need for development of degradable stents for repair of vascular trauma. The presence of a metallic foreign body within the lumen of the vessel is the most likely origin of this reaction, which in turn leads to early restenosis, or thrombosis, of the vessel. The very nature of these templates allows for possible modifications to help reduce the intensity of the reaction observed and to the final development of the ideal temporary conduit to allow an injured vessel to heal. Finally, the template’s mechanical qualities lack the memory required to anchor itself to the vessel wall. Independent anchoring and elastic properties remain a major issue in the further 79
manipulation of this biomaterial and will leave further study dependent on commercial stents to anchor new prototypes unless its properties can be modified. In the event these qualities cannot be achieved, pursuit of modified chitosan as the graft material over a metallic stent in a dedicated stent graft format may also hold promise for future research, especially if side branch modifications can be undertaken to reduce the inflammatory and neointimal response to the material.
4. Gutowska A. Resorbable templates for vascular repair. Pa-
per presented at: Sixth World Biomaterials Congress; 2000; Kamuela, Hawaii. 5. Beekley A. Porcine model for study of absorbable vascular
templates. Paper presented at: United States Military Vascular Conference; December 3-5, 1999; Washington, DC. 6. Palmaz J. Intravascular stents: tissue-stent interactions and
design considerations. Am J Roentgenol. 1993;160:613-618.
REFERENCES
7. Williams D. Radiation vascular therapy: a novel approach
1. Slepian M. Polymeric endoluminal paving/sealing: a biode-
to preventing restenosis. Am J Cardiol. 1998;81:18E-20E.
gradeable alternative to intracoronary stenting. Circulation. 1988;78:II-409.
8. Rogers C, Edelman E. Endovascular stent design dictates
2. Dichek D, Neville RF, Zwiebel JA, et al. Seeding of intra-
vascular stents with genetically engineered endothelial cells. Circulation. 1989;80:1347-1353. 3. Irie T, Furui S, Yamauchi T, et al. Relocatable Gianturco
expandable metallic stents. Radiology. 1991;178:575-578.
80
experimental restenois and thrombosis. Circulation. 1995; 91:2995-3001. 9. Sala T, Taylor B, Suggs WD, et al. Reaction to injury fol-
lowing balloon angioplasty and intravascular stent placement in the canine femoral artery. Am Surg. 1994;60:353357.
CURRENT SURGERY • Volume 58/Number 1 • January/February 2001
Bioabsorption Qualities of ChitosanAbsorbable Vascular Templates Mohamad I. Haque, CPT, MC, USA,* Alec C. Beekley, CPT, MC, USA,* Anna Gutowska, PhD,§ Ruth Ann Reardon, CPT, MC, USA,† Paul Groo, LTC, MC, USA,† Sean P. Murray, MAJ, MC, USA,‡ Charles Andersen, COL, MC, USA,* and Kenneth Azarow, LTC, MC, USA* Departments of *Surgery, †Pathology, and ‡Radiology, Madigan Army Medical Center, Tacoma, Washington and the §United States Pacific Laboratory Northwest/Battelle Corporation, Richland, Washington PURPOSE: The scope of endovascular surgical techniques has
expanded to include the treatment of diseases considered at one time to be amenable only to surgical treatment. The development of the biodegradable template follows as an extension of current permanent stent technology. The goal of our project is to develop and test chitosan as an absorbable template for the vascular system. METHODS: Ultrapure chitosan, heparin sodium salt and ly-
sozyme, and contrast agents MD-76R and Oxilan-350 were used to give radioopaque quality. Prototype chitosan vascular templates were obtained by a dip coating method in which alternate layers of chitosan were coagulated with nonsolvents or heparin. The amount of loaded and released heparin was determined using Azure II colorimetric assay. In vitro enzymatic degradation of templates was evaluated using lysozyme solutions in phosphate buffered saline. Mechanical properties were analyzed using the Dynamic Mechanical Analyzer, DMA-7 (Perkin Elmer, Foster City, Calif.). The microstructure of freeze-dried templates was investigated by field emission scanning electron microscopy (FE SEM) using an LEO 982 electron microscope (Zeiss, Thornwood, NY). In vivo deployment of the templates was undertaken in 10 full-sized pigs (Sus scrofa). After open expose and control of the iliac artery, a closed balloon catheter technique was used to advance and place the balloon catheter and template. The balloon was then expanded, deploying a Palmaz stent with a chitosan template anchored distally. Patency and deployment of the stent-template complex was confirmed by an arteriogram. The animals were sacrificed at 1, 2, 3, 4, and 5 weeks poststent placement, and arterial sections were taken for microscopic
Correspondence: Inquiries to Kenneth Azarow, LTC, MC, USA, Madigan Army Medical Center, MCHJ-SGY, MAMC, Tacoma, WA 98431; fax: (253) 968-0232. Support: This study was supported by a grant from the National Medical Test Bed. This article contains the opinions of the authors only and does not represent the opinion of the United States Department of Defense or the United States Army.
CURRENT SURGERY • Published 2001 by Elsevier Science Inc.
analysis. The amount of chitosan remaining was estimated to determine an in vivo rate of absorption. RESULTS: On hematoxilyn and eosin staining of the section arterial samples, a marked inflammatory response was noted and progressed with duration of in vivo contact. A giant cell foreign body reaction coupled with intense intimal hyperplasia and organized thrombus was also noted and progressed with duration of time in vivo. Also noted was the degradation of the template material with only small remnants of material noted within the giant cell by week 4. Clinically, none of the pigs developed limb ischemia or evidence of thromboembolic events. CONCLUSIONS: In this in vivo study, the chitosan template
proved to be biodegradable but elicited an intense thrombotic and foreign body reaction despite heparin bonding. Further investigation is ongoing as to decreasing the thrombogenic and antigenic qualities of the template materials by either alteration of the base material or addition of bioactive side chains. (Curr Surg 58:77-80) KEY WORDS: chitosan, endovascular surgery, vascular stents
INTRODUCTION Over the last 3 decades, the explosion in the field of endovascular surgery has led to rapid advances in the biomaterials used for intraluminal deployment. The mainstay of this technology has focused on permanent, expandable metallic stents for use virtually anywhere in the vascular tree. This technology has been applied to the therapy of both occlusive and aneurysmal disease of the coronary, carotid, renal, iliac arteries, and, most recently, aortic repair. Since the early 1980s, little work has been done in the area of degradable stents (templates) for use in the repair of arterial or venous injuries. A single study from 1982 described the unsuccessful attempts at vicryl stents in porcine coronary arteries.1 Stent grafts with harvested saphenous vein, polytetrafluoroethelene, or bioengineered endothelial cells have been devel0149-7944/01/$20.00 PII S0149-7944(00)00422-0
77
FIGURE 1. Chitosan: (1,4)-linked 2-amino-2-deoxy-B-D-glucan, negatively charged organic compound.
oped and used in clinical trials; however, these leave permanent material within the lumen of the vessel.2,3 The application for a vascular template with reliable bioabsorption qualities has a particular application in the setting of vascular trauma, especially in blood vessels that are difficult to approach through open surgical techniques. In this setting, endovascular techniques could be used to place a template across the injured vessel and to serve as a bridge for regeneration of the intimal surface while preventing exanguinating hemorrhage. As the vessel heals, in vivo serum enzymes degrade the template material until it is completely degraded or incorporated into the vessel wall . The goal of this project was to study the bioabsorption qualities of a novel template material within the vascular tree of a porcine model. Gutowska describes the development of the chitosan templates.4 Beekley describes the porcine model and endovascular technique for deployment of these templates.5
MATERIALS AND METHODS Ultrapure chitosan was purchased from Carbomer (Westborough, Mass) (Fig. 1). Heparin sodium salt and lysozyme was purchased from Sigma Chemical Company (St. Louis, Mo). Oxilan-350 nonionic x-ray contrast was obtained. The templates were produced by alternate dipping of thin glass rods (2-mm diameter) in an ultrapure chitosan solution and a dilute acetone solution. This was then alternated with a heparin salt solution to produce a heparin-impregnated template. The templates were then exposed to in vitro enzymatic degradation studies in a lysozyme solution. Before phase 2, a series of templates was then created using a sterile technique and soaked in Oxilan-350 until the moments before in vivo deployment.4 Ten adult pigs (Sus scrofa) were brought to a veterinary operating suite with fluoroscopic capability, placed under general anesthesia, systemically heparinized, and surgical exposure of the left external iliac artery was obtained. Through a transverse arteriotomy, sequential dialators were introduced distally until a 10 French sheath could be easily introduced. The balloon catheter was then loaded with a Palmaz (Johnson & Johnson, Warren, NJ) stent (2.5 mm ⫻ 15 mm) and the chitosan template advanced over the stent such that 5 mm of the templatestent complex overlapped, and this complex was then deployed under fluorscopic guidance. Predeployment and postdeployment arteriograms were performed. The animals were then sacrificed at weeks 1, 2, 3, 4, and 5 postdeployment, and the entire stent-template complex was removed en block and preserved in formalyn.5 Sections were then taken 2 to 4 mm distal to the 78
FIGURE 2. Intense foreign body giant cell reaction: eosinophilic chitosan (a) within the hypercellular granuloma and foreign body giant cell at medial aspect of the granuloma (b).
palpable end of the Palmaz stent and analyzed by hematoxylin and eosin staining.
RESULTS Clinically, none of the animals studied displayed evidence of sepsis, acute limb ischemia, or thromboembolic phenomenon. Nine of 10 stent-template arterial sections were recovered and studied microscopically. Study of these templates revealed an intense foreign body reaction (Fig. 2) and severe intimal hyperplasia (Fig. 3). These templates were degraded in a nonuniform manner (Fig. 4, left and right). It appears that the entire resorption took place at the week 5 time frame. The only stent and template not recovered was at week 5. Qualitative analysis of chitosan detected in lumen reflected as percent degradation (Fig. 5). Although the absorption over time was unequal throughout the circumference and length of the template, the rate of total resorption was equivalent between each pair of animals sacrificed at the specific intervals measured.
FIGURE 3. Intimal hyperplasia: proliferative effect of chitosan on intima of vessel (a) in a 5th-week harvest (no intraluminal chitosan noted). CURRENT SURGERY • Volume 58/Number 1 • January/February 2001
FIGURE 4. Organized thrombus: extent of organized thrombosis within the lumen of the vessel with classic evidence for organized thrombosis. (Left) Two weeks in vivo. (Right) Four weeks in vivo.
DISCUSSION Chitosan as an absorbable vascular template material proved to have many desirable qualities in the in vitro stage of the study. Its biodegradability and saturation quality with heparin are novel qualities to current stent technologies and initially allowed for deployment and confirmation of arterial patency immediately post-template deployment. Furthermore, the negative charge of this organic compound contributes to its antithrombogenic quality. Unlike in commercially expandable metallic stents, well described by Palmaz, these stents carry an obligate positive charge that elicits an irregular clot between the strut indentations and is attributed to early stent thrombosis or restenosis at the edge of the intimal-stent interface.6 Chitosan’s negative charge carries the theoretical advantage of repelling positively charged members of the clotting cascade and platelets. However, despite these positive actual and theoretical qualities, chitosan proved not to be an inert substance in vivo, as evidenced by an intense inflammatory response not seen in current metallic stents and stent grafts. This foreign body reac-
FIGURE 5. Estimated percent degradation in templates over study period. CURRENT SURGERY • Volume 58/Number 1 • January/February 2001
tion within the vessel lumen and its surrounding thrombosis negates 2 of the theoretical advantages of the material. In vitro, chitosan was degraded uniformly by enzymatic degradation. In vivo, however, the exorbitant cellular response to the material and resultant occlusion of the vessel lumen argues against uniform purely enzymatic degradation of the template. The very type of reaction (foreign body) further disproves that chitosan templates (in their current form) are biologically inert in vivo. Thrombosis and intimal hyperplasia are complications well established in endovascular literature. In fact, much of stent research is focused on the modification of metallic stents to decrease the extent of intimal hyperplasia induced to include irradiation, coating with inert, noncharged polymers as well as modification of stent geometry.7,8 The degree of intimal hyperplasia observed in this model is not different from those seen in a canine femoral artery study performed with metallic stents.9 Unlike metallic stents, chitosan’s multiple-side chemical bonding sites allow for potential structural modification of the molecule to hide the antigentic and inflammatory foci that elicit the neointimal response. This concept argues for the rationale for the development of stents to serve only as a temporary scaffold for the repair of an injured vessel. Work by Sala and associates describes metallic stents within undiseased canine femoral vessels and intimal hyperplasia at the proximal and distal edges of the stent.9 This reaction alone argues for the need for development of degradable stents for repair of vascular trauma. The presence of a metallic foreign body within the lumen of the vessel is the most likely origin of this reaction, which in turn leads to early restenosis, or thrombosis, of the vessel. The very nature of these templates allows for possible modifications to help reduce the intensity of the reaction observed and to the final development of the ideal temporary conduit to allow an injured vessel to heal. Finally, the template’s mechanical qualities lack the memory required to anchor itself to the vessel wall. Independent anchoring and elastic properties remain a major issue in the further 79
manipulation of this biomaterial and will leave further study dependent on commercial stents to anchor new prototypes unless its properties can be modified. In the event these qualities cannot be achieved, pursuit of modified chitosan as the graft material over a metallic stent in a dedicated stent graft format may also hold promise for future research, especially if side branch modifications can be undertaken to reduce the inflammatory and neointimal response to the material.
4. Gutowska A. Resorbable templates for vascular repair. Pa-
per presented at: Sixth World Biomaterials Congress; 2000; Kamuela, Hawaii. 5. Beekley A. Porcine model for study of absorbable vascular
templates. Paper presented at: United States Military Vascular Conference; December 3-5, 1999; Washington, DC. 6. Palmaz J. Intravascular stents: tissue-stent interactions and
design considerations. Am J Roentgenol. 1993;160:613-618.
REFERENCES
7. Williams D. Radiation vascular therapy: a novel approach
1. Slepian M. Polymeric endoluminal paving/sealing: a biode-
to preventing restenosis. Am J Cardiol. 1998;81:18E-20E.
gradeable alternative to intracoronary stenting. Circulation. 1988;78:II-409.
8. Rogers C, Edelman E. Endovascular stent design dictates
2. Dichek D, Neville RF, Zwiebel JA, et al. Seeding of intra-
vascular stents with genetically engineered endothelial cells. Circulation. 1989;80:1347-1353. 3. Irie T, Furui S, Yamauchi T, et al. Relocatable Gianturco
expandable metallic stents. Radiology. 1991;178:575-578.
80
experimental restenois and thrombosis. Circulation. 1995; 91:2995-3001. 9. Sala T, Taylor B, Suggs WD, et al. Reaction to injury fol-
lowing balloon angioplasty and intravascular stent placement in the canine femoral artery. Am Surg. 1994;60:353357.
CURRENT SURGERY • Volume 58/Number 1 • January/February 2001