Comparison of Small-Intestinal Submucosa and Expanded Polytetrafluoroethylene as a Vascular Conduit in the Presence of Gram-Positive Contamination

Abstract

Objective:

As a vascular conduit, expanded polytetrafluoroethylene (ePTFE) is susceptible to graft infection with Gram-positive organisms. Biomaterials, such as porcine small-intestinal submucosa (SIS), have been successfully used clinically as tissue substitutes outside the vascular arena.

Summary Background Data:

In the present study, we compared a small-diameter conduit of SIS to ePTFE in the presence of Gram-positive contamination to evaluate infection resistance, incorporation and remodeling, morphometry, graft patency, and neointimal hyperplasia (NH) development.

Methods:

Adult male mongrel pigs were randomized to receive either SIS or ePTFE (3-cm length, 6-mm diameter) and further randomized to 1 of 3 groups: Control (no graft inoculation), Staphylococcus aureus, or mucin-producing S epidermidis (each graft inoculation with 10⁸ colonies/mL). Pressure measurements were obtained proximal and distal to the graft to create the iliac/aorta pressure ratio. Morphometric analysis of the neointima and histopathologic examinations was performed. Other outcomes included weekly WBC counts, graft incorporation, and quantitative culture of explanted grafts.

Results:

Eighteen animals were randomized. All grafts were patent throughout the 6-week study period. Infected SIS grafts had less NH and little change in their iliac/aorta indices compared with infected ePTFE grafts. Quantitative cultures at euthanasia demonstrated no growth in either SIS group compared with 1.7 × 10⁴ colonies for ePTFE S aureus and 6 × 10² for ePTFE S epi (each P < 0.001). All SIS grafts were incorporated. Histology demonstrated remodeling into host artery with smooth muscle and capillary ingrowth in all SIS groups. Scanning electron micrography illustrated smooth and complete endothelialization of all SIS grafts.

Conclusions:

Compared with ePTFE, SIS induces host tissue remodeling, exhibits a decreased neointimal response to infection, and is resistant to bacterial colonization. SIS may provide a superior alternative to ePTFE as a vascular conduit for peripheral vascular surgery.

Infected prosthetic grafts can result in graft failure and significant morbidity. The authors compared porcine small-intestinal submucosa (SIS), a cell-free collagen matrix, with expanded polytetrafluoroethylene (ePTFE) as a vascular conduit. Grafts were contaminated with Staphylococcus aureus and S epidermidis. SIS grafts were associated with less neointimal hyperplasia, complete incorporation, and no residual organisms compared with ePTFE.

Despite intensive research efforts, the ideal vascular graft has yet to be developed. Mechanical stability, biocompatibility, nonthrombogenicity, infection resistance, and availability are desired characteristics of a vascular bypass graft. Refinements in conduit fabrication have led to improvements in mechanical stability and patency rates of large-diameter vascular grafts in recent years; however, the use of synthetic conduits for small-diameter (<6 mm) arterial reconstruction is complicated by poor long-term patency rates ranging from 40% to 60% at 5 years.^1–3 Synthetic conduits are also susceptible to infection, which is a devastating complication of vascular reconstruction. Individualized treatment modalities, including debridement and evacuation of surrounding infection with muscle-flap coverage, long-term antibiotic therapy, and partial graft excision with in situ prosthetic reconstruction, have been described with varying degrees of success.^4–6 The most reliable and comprehensive approach to eradicate infection and restore distal perfusion involves total graft excision with extra-anatomic bypass.

ePTFE (expanded polytetrafluoroethylene) is a commonly used synthetic conduit. However, ePTFE is susceptible to infection since its microporous structure provides a “safe haven” for bacteria to survive and proliferate. Staphylococcus and mucin-producing Staphylococcus are the most common infecting organisms and result in chronic infection, lack of graft incorporation, pseudoaneurysm formation, leading to potential graft failure and major morbidity. In addition, an excessive neointimal response due to a persistent inflammatory process and a compliance mismatch with the native artery are significant causes of late graft failure. Thus, a readily available conduit that is resistant to bacterial colonization, mechanically stable, and elicits a favorable host response upon implantation would be superior for elective vascular reconstruction.

Advances in tissue engineering technology have resulted in the clinical application of a variety of biomaterials. Porcine small-intestinal submucosa (SIS) is an acellular collagen matrix, which provides a scaffold for site-specific host tissue remodeling.^7,8 It has been successfully used in herniorrhaphy, chronic wound healing, bladder reconstruction, dural repair, and as a tendon and ligament substitute.^9–14 It has previously been shown that, when used as a vascular patch, SIS is resistant to infection following fecal contamination and is remodeled into host artery.¹⁵ In the present study, we constructed a small-diameter conduit of SIS and compared it to ePTFE in the presence of Gram-positive contamination to evaluate infection resistance, incorporation and remodeling, morphometry, graft patency, and the neointimal response.

METHODS

Male mongrel pigs weighing between 15 and 25 kg were used after acclimation in the Animal Care Facility at the University of Tennessee Health Science Center. The project was approved after review by the Animal Care and Utilization committee of the University of Tennessee Health Science Center. All care was in compliance with the Principles of Laboratory Animal Care and Guide for the Care and Use of Laboratory Animals (National Institutes of Health, Bethesda, MD).

Animals were randomized into 1 of 3 groups: control, Staphylococcus aureus contamination, and mucin-producing S epidermidis contamination. The animals were further randomized to receive either ePTFE or SIS (Fig. 1).

FIGURE 1. Experimental design.

Animals underwent induction of anesthesia with an intramuscular injection of ketamine and xylazine. General anesthesia was maintained with isoflurane. The animals were monitored continuously with ECG and pulse oximetry. All animals were administered cefazolin (500 mg) prior to incision and at closure. After a midline incision, the left common iliac artery was circumferentially dissected and isolated with vascular clamps. Five minutes prior to clamp placement, a heparin bolus (110 units [u]/kg) was administered. Heparin (55 u/kg) was administered every 30 minutes until clamp removal. In the control group, animals underwent interposition grafting of the left common iliac artery with either a 3-cm segment of 6-mm-inner-diameter ePTFE (Gore-Tex; W.L. Gore and Associates) or SIS (Surgisis; Cook Surgical, Bloomington, IN) graft. The SIS graft was constructed by suturing a sheet of commercially available SIS over a sterile pipette after hydration in sterile saline for 10 minutes. The longitudinal suture line was created with a running 6-0 Prolene suture. After completion of the anastomosis, an arterial pressure was transduced in the aorta just proximal to the graft and in the iliac distal to the graft. The iliac systolic pressure was divided by the aortic systolic pressure to calculate the iliac/aorta index. This value was used to determine graft patency and to objectively measure changes at euthanasia. Blood was obtained for WBC count and blood culture. In the control groups, the graft was inoculated with 1 mL of sterile saline prior to closure of the peritoneum over the graft. In the contamination groups, animals underwent the same surgical procedure. Prior to closure of the peritoneum, animals underwent inoculation with either 1 mL of S aureus (ATCC 25923, average concentration 3.75 × 10⁸ CFU/mL) or mucin producing S epidermidis (ATCC 35983, average concentration 2.7 × 10⁸ CFU/mL) directly onto the graft.

After abdominal closure, animals were returned to their cages and allowed free access to food and water. All animals were administered daily aspirin (325 mg/5 mL) until euthanasia and cephalexin (500 mg/5 mL) twice daily for 5 days postoperatively. Animals were sedated with ketamine and xylazine weekly for measurement of WBC counts. Animals were inspected daily for pulse quality, limb viability, and general condition. Temperature measurements were obtained and recorded every third day.

Animals recovered for a period of 42 days. After this recovery period, animals were weighed, anesthetized, and underwent laparotomy under sterile conditions. The graft was inspected for the degree of incorporation and surrounding inflammatory response. Arterial pressures were transduced proximal and distal to the graft. Blood was drawn for WBC count and culture.

The graft was excised and segments of the proximal anastomosis, midgraft, and distal artery were sent for culture. The segments were placed in 2 mL of trypticase soy broth (TSB) and sonicated. Portions (0.1 mL) were then plated onto TSB with 5% sheep blood agar plates, and incubated in 5% CO₂ for up to 3 days at 35°C. Growth was identified as either S aureus or S epidermidis by Gram stain, coagulase test, and API-STAPH identification system (Biomerieux, France).

The distal anastomosis was excised and placed in 10% buffered formalin solution. Specimens were allowed to fix for 72 hours. The specimens were sectioned, embedded in paraffin, and cut into 6-μm sections. Each section was stained with hematoxylin and eosin (H&E stain), trichrome, and Gram stain. Grafts were analyzed for remodeling characteristics and the degree of inflammatory response. NIH image 1.55 software (National Institutes of Health public domain software) was used to morphometrically analyze the neointimal response by calculating the neointimal area and area of the graft lumen. These values were then used to determine the neointima/lumen ratio as an objective measurement of the neointimal response.

Statistical analyses were performed by 1-way and repeated-measures analysis of variance within and between treatment groups and controls. Data are expressed as the mean. A value of P < 0.05 was considered to be statistically significant.

RESULTS

A total of 18 animals were randomized: 6 each to the control, S aureus contamination, and mucin producing S epidermidis contamination groups. Three animals in each group received ePTFE grafts and 3 received SIS grafts. All animals survived until necropsy at 6 weeks.

The animals were inspected daily for general condition, pulse quality, and limb viability. All animals maintained a palpable pulse until euthanasia, and there were no instances of limb ischemia. There was no significant difference in the temperature curve among the groups.

WBC counts in the contaminated ePTFE groups peaked at 2 weeks, with a gradual return to baseline in the S aureus group at 4 weeks. Animals in the ePTFE S epidermidis contamination group maintained a leukocytosis until euthanasia, which reached statistical significance compared with controls at 4 and 6 weeks. WBC counts in control and contaminated SIS groups varied little from baseline (Fig. 2).

FIGURE 2. Serial white blood cell counts in all study groups (PC, PSA, PSE = ePTFE control, ePTFE S aureus, ePTFE S epidermis, respectively; SC, SSA, SSE = SIS control, SIS S aureus, SIS S epidermis, respectively).

On gross inspection, control ePTFE grafts elicited an encapsulating foreign-body reaction. There was no anastomotic pseudoaneurysm development. Contaminated ePTFE grafts were poorly incorporated. Grafts contaminated with mucin-producing S epidermidis had a surrounding biofilm layer, which appeared to serve as a barrier to surrounding tissue ingrowth. S aureus–contaminated grafts had patchy areas of adherent surrounding tissue. There was no gross purulence or pseudoaneurysm development in either group. SIS grafts were all well incorporated and appeared to elicit direct tissue ingrowth compared with the encapsulating response observed in the ePTFE controls. There was no difference in the gross appearance of controls and contaminated SIS grafts.

Contaminated ePTFE grafts had a greater reduction in the iliac/aorta index at euthanasia compared with contaminated SIS grafts. The difference was greatest in the S aureus contamination group (−0.12 versus 0.006; P = NS) (Table 1). These measurements correlated with the degree of neointimal change measured at euthanasia, which was objectively measured as the neointimal/lumenal area ratio. Both ePTFE contamination groups had average ratios greater than 1.4. Although not statistically significant, these values were greater in comparison to ratios of less than 1.0 is both SIS contamination groups (Fig. 3). Scanning electron micrography was consistent with these findings, demonstrating more neonintimal hyperplasia in the contaminated ePTFE specimens (Fig. 4).

TABLE 1. Illiac/Aorta Ratios for All Groups

FIGURE 3. Neointima/lumen ratios for all study groups (PC, PSA, PSE = ePTFE control, ePTFE S aureus, ePTFE S epidermis, respectively; SC, SSA, SSE = SIS control, SIS S aureus, SIS S epidermis, respectively).

FIGURE 4. Scanning electron micrographs demonstrating excessive neointimal hyperplasia in contaminated ePTFE and tissue incorporation SIS.

Histopathologic analysis revealed more pronounced inflammatory cell reaction in the contaminated ePTFE grafts compared with controls. The contaminated grafts were largely unincorporated compared with controls. Gram stains revealed Gram-positive cocci residing within the interstices of the contaminated ePTFE grafts (Fig. 5). These findings were in contrast to the appearance of SIS grafts. All SIS grafts were well incorporated, with evidence of host tissue remodeling and transmural capillary ingrowth (Fig. 6). Few inflammatory cells were observed on H&E stain, and none of the gram stains demonstrated evidence of bacterial colonization.

FIGURE 5. Gram stain of ePTFE demonstrating organisms within the interstices of the graft.

FIGURE 6. Trichrome staining demonstrating new capillary (arrow) and collagen ingrowth in contaminated SIS graft.

Quantitative culture data at euthanasia revealed the presence of the inoculating organism in all contaminated ePTFE grafts, while none of the SIS grafts had positive cultures. All of the control grafts were culture negative. In the mucin-producing S epidermidis group, the average inoculum was 2.7 × 10⁸ CFU/mL. At explantation, ePTFE grafts had an average quantitative culture value of 613 CFU/mL compared with 0 CFU/mL for SIS grafts (P < 0.001) (Table 2). In the S aureus contamination group, the average inoculum was 3.75 × 10⁸ CFU/mL. At explantation, ePTFE grafts had an average quantitative culture value of 17,153 CFU/mL compared with 0 CFU/mL for SIS grafts (P < 0.001).

TABLE 2. Quantitative Cultures of Grafts

DISCUSSION

The search for the ideal vascular graft remains elusive. An increased life expectancy and the prevalence of atherosclerotic vascular disease make the development of a durable, readily available, and cost-effective vascular substitute a significant health care priority. The emergence of endovascular technology has had an impact on the treatment of vascular disease by offering a less invasive approach and reduced initial morbidity; however, anatomic restrictions are a limiting factor in the application of endovascular technology.¹ Surgical bypass grafting remains the most reliable and flexible treatment of occlusive vascular disease.

The ideal vascular graft should be mechanically stable, infection resistant, nonthrombogenic, biocompatible, readily available, and cost efficient. Currently available materials, including Dacron, ePTFE, autogenous saphenous vein, heterografts (bovine carotid artery), and homografts (glutaraldehyde-treated human umbilical cord vein), have their advantages and disadvantages; however, none meet the requirements of the ideal vascular graft. Poor long-term patency, aneurysmal degeneration, size discrepancy, limited availability, and vulnerability to bacterial colonization are limiting features when using substitute conduits for vascular reconstruction.

Infection remains a significant problem in prosthetic vascular reconstruction. The microporous structure of prosthetic grafts provides an ideal environment for bacteria to grow and proliferate. Although infection rates are low (1%–8%),4–6,16–19 the consequences are often disastrous. An aggressive approach requiring explantation, debridement of surrounding tissues, and extraanatomic bypass is usually required to clear the infection and restore perfusion. S aureus and mucin producing S epidermidis are the 2 commonly isolated organisms from infected vascular grafts and result in a lack of tissue incorporation, anastomotic pseudoaneurysm, excessive neointimal hyperplasia, and potential graft failure.

Autogenous saphenous vein has long been considered the gold standard of vascular substitutes. Biocompatibility and superior long-term patency rates have made it the first choice for arterial reconstruction when primary repair is not feasible. There are several disadvantages to the use of autogenous saphenous vein. It is of limited availability and its use requires a separate surgical procedure, which prolongs operative time. In up to 30% of the population, it is not an option secondary to either previous use or involvement with a disease process.²⁰ Although the incidence of infection is 4-fold lower than prosthetic grafts, experimental and clinical evidence has shown that vein is more prone to infective dissolution and subsequent exsanguinations.^21,22

Limited availability and a propensity of vein grafts to disrupt when infected have led some clinicians to support the use of synthetic conduits in the face of contamination.^21–26 Three properties of synthetic materials have been identified which have the greatest influence on bacterial adherence: porosity, surface tension, and electronegativity.²⁷ ePTFE is the most commonly used synthetic conduit and has the most favorable properties in terms of its microporous structure, electronegativity, and relatively high surface tension to minimize bacterial adherence.²⁷ Initial enthusiasm for its use in the face of contamination has been tempered by studies that have identified significant problems. The interstices within the graft matrix provide a safe haven for bacteria to survive and proliferate. When infected, poor tissue incorporation, pseudoaneurysm formation, and thrombosis result in ultimate graft failure. In the present study, a lack of tissue incorporation and an excessive neointimal response were observed in the contaminated ePTFE groups. Organisms were also present within the interstices of the ePTFE grafts.

A variety of biomaterials have been developed for vascular reconstruction.²⁸ The ideal material would have favorable mechanical properties, serve as a scaffold for regeneration of a host vessel, and degrade over time while maintaining prosthetic integrity. Porcine SIS is a material which has significant potential as a vascular substitute. It is an acellular collagen matrix, which serves as a scaffold for site specific host tissue remodeling.⁷ Numerous growth factors, including VEGF, FGF, and TGF-β, have been extracted from the material and are thought to contribute to the remodeling response.^29,30 Its mechanical properties are favorable for use as a vascular conduit. Its suture retention strength and burst pressure are adequate to withstand pulsatile arterial pressures. Its compliance is closer to that of native artery than any available synthetic conduit.³¹ SIS has the unique property of directional porosity. A higher porosity on the abluminal surface allows tissue ingrowth, while a low porosity on the luminal surface prevents graft leakage.^32,33

Preliminary work utilizing SIS as a vascular conduit is encouraging. Problems observed with synthetic materials such as neointimal hyperplasia, infection, and aneurysmal degeneration have not been reported with SIS.^34–40 Badylak et al³⁴ demonstrated short-term patency rates (<18 months) of 92% when used in a canine model. Lantz et al³⁸ showed that, when used as a small-diameter autograft (<4.5-mm ID), SIS has a patency rate of 75%. None of the grafts in this study demonstrated evidence of aneurysmal degeneration or neointimal hyperplasia. These studies demonstrate potential for use of SIS as a conduit regardless of species. The porcine-derived SIS has been well tolerated in humans when used as a soft-tissue patch.

SIS has demonstrated resistance to deliberate bacterial contamination in animal studies.9,15,19,41 When used as a vascular patch in the presence of gross fecal contamination and shock, SIS was resistant to bacterial colonization and demonstrated evidence of host-tissue remodeling into a native blood vessel. SIS maintained mechanical integrity and was superior to ePTFE in terms of infection resistance, pseudoaneurysmal degeneration, and thrombosis.¹⁵ SIS induces an integrative response with neovascularization through transmural capillary ingrowth, allowing the host to respond to the presence of bacteria and prevent colonization. This type of response is much different from an encapsulating foreign-body reaction elicited by synthetic materials, which restricts neovascularization of the graft, allowing bacteria to reside in an environment protected from the host's immune response. Another interesting property of SIS is the presence of naturally occurring antibacterial peptides within the matrix which have demonstrated activity against Gram-positive and Gram-negative organisms in vitro.⁴²

The present study was designed to evaluate the potential clinical utility and infection resistance of the readily available and cost-efficient biomaterial SIS. Our study was the first to use the material in its commercially available form and test it as a small-diameter conduit against ePTFE in the presence of deliberate Gram-positive contamination. Previous work has used SIS harvested and made into a conduit in vivo without being subjected to the subsequent processing steps. ePTFE is a commonly used prosthetic material in vascular reconstruction; therefore, its clinical performance was compared with SIS. SIS grafts demonstrated an integrative host-tissue response with incorporation into the surrounding tissues. This was in contrast to a lack of incorporation observed in the ePTFE contamination groups. When exposed to bacterial contamination, SIS grafts exhibited a healing response comparable to the control group and were resistant to bacterial colonization. All ePTFE grafts were culture positive for the inoculating organism at explantation. A decreased neointimal response in the presence of contamination was observed in the SIS grafts compared with ePTFE, which may be a significant cause of late graft failure. Histologically, SIS grafts demonstrated remodeling into a host blood vessel with smooth-muscle- and endothelial-cell migration and capillary ingrowth. Scanning electron micrography illustrated confluent endothelialization of the SIS grafts.

Despite encouraging preliminary studies in animals, much work is warranted prior to the first clinical implant of SIS as a vascular conduit in humans. Further studies need to be performed to characterize the immune response elicited when implanted across species lines. Long-term studies are needed to evaluate the fate of the original SIS material and changes in the mechanical properties of the graft as it remodels. Since the matrix serves as a scaffold for tissue ingrowth, a limitation may exist in the effective length of the graft. Additionally, the effect of contamination by collagenase-producing organisms such as Pseudomonas species needs to be tested. Nonetheless, the results of this study are promising and demonstrate that SIS is a biologically active matrix which is resistant to bacterial colonization, induces host tissue remodeling, and is mechanically stable. SIS may provide a superior alternative to ePTFE as a substitute conduit in peripheral vascular surgery.

Discussions

Dr. David V. Feliciano (Atlanta, Georgia): Drs. Shell, Croce, Fabian, and colleagues have presented a follow-up study on the use of porcine small-intestinal submucosa as an implant in the vascular system. They have used a popular laboratory model of overt contamination of both the SIS product sewn into the shape of a tube graft and of expanded PTFE graft that was popularized by Dr. Wesley Moore of UCLA in 1975 (J Surg Research 18:229, 1975).

When newly implanted grafts in animal models are exposed to overt bacterial loads, particularly if postoperative antibiotics are not being administered, the results are predictable. Saphenous vein grafts undergo dissolution, and synthetic prostheses such as Dacron or expanded PTFE dehisce graft-artery suture lines. When postoperative antibiotics are used, autogenous conduits often survive while the synthetic conduits fester and do not become incorporated, much as the authors have described today.

As stated in this study, the acellular small-intestinal mucosa used as a tube graft had 100% patency, no incidence of infection despite bacterial contamination, and little evidence of neointimal hyperplasia at the distal SIS-artery suture line.

These data are most intriguing, and I am sure that many members of the audience are calling their stockbrokers as we speak to invest in the SIS vascular graft of the future. While they are doing that, I am going to ask some questions.

There is a comment in your abstract that “expanded PTFE is the conduit of choice for peripheral vascular reconstructions.” Now, the authors know that this is not true, and I hope they reword the abstract. But the larger issue is why didn't you compare the SIS to autogenous vein in this model? Wouldn't this have been a more appropriate comparison as PTFE is, in everybody's hands, a secondary choice in elective bypasses for true limb salvage and in trauma when a substitute vascular conduit is needed after arterial resection? Another way to ask this question is, you knew what was going to happen with the PTFE based on previous studies.

Second, a question on the model. Did you do sensitivity testing on the ATCC Staphylococcus aureus and epidermidis to prove that these are not mutated lines with resistance to the cephalexin that was used postoperatively?

Third, is there really enough evidence in your study that the neointimal hyperplasia at the distal PTFE graft-artery suture lines was related to the presence of bacteria? Aren't you just redescribing a well-known phenomenon of PTFE grafts that has led many vascular groups and trauma centers, including our own, to place all patients with PTFE vascular grafts on clopidogrel bisulfate tablets and aspirin postoperatively for many years?

Finally, I know you would agree that this model is not clinically relevant as bacteria are irrigated out of incisions and wounds clinically, and, importantly, human beings are unable to reintimize substitute vascular conduits the way pigs do. What is the next step here? How long a graft can you actually make out of SIS? Will it resist the contraction scarring that often occurs in groin incisions or post-trauma cavities? What will happen to this acellular matrix when neointima does not develop except within a centimeter or 2 of the 2 suture lines?

Dr. Timothy C. Flynn (Gainesville, Florida): As you point out, vascular surgeons continue to seek the Holy Grail of an off-the-shelf vascular conduit, as anyone who has spent the day sewing together 3 pieces of arm vein will attest. Unfortunately, all too often have our hopes been shattered as much-touted solutions have proven ineffective in the long run, particularly when these were based on animal experiments in which vascular conduits seemed to behave in a much more favorable environment.

This paper reports on a well-designed experiment using prepared porcine small-intestine submucosa as an arterial substitute in an infected field compared to the use of PTFE under the same conditions in a porcine model. At 6 weeks, as reported, the SIS grafts were incorporated, showed remodeling and endothelialization, as well as smooth-muscle growth. They were free of bacterial contamination. As one would expect, the opposite was found with PTFE grafts. While a small study, there were only 3 animals in each of the experimental groups; this suggested SIS may have a role in vascular surgery.

Having said that, SIS in its current form has been the subject of experimental research since at least the late 1980s. And while it has achieved some use in hernia repairs and as patches for urologic defects, it has not been well studied as a vascular substitute.

Experiments using SIS as a small-diameter vascular conduit in microvascular anastomoses were unsuccessful. There is 1 previous report of using SIS as an aortic substitute in pigs with reported success in a very-short-term follow-up. Tubularized SIS was not effective as a ureteral substitute because of occlusion, while SIS used as a ligament or in a bladder suspension procedure seems to do very well, and, as the authors point out, it almost miraculously assumes the histologic appearance of the site into which it is implanted. Perhaps the authors can make some comment about this property.

Your group has also shown that an arterialized patch of SIS was effective in a field contaminated by Gram-negative organisms. One wonders why there have not been more animal studies, and to my knowledge no human studies, of SIS in the arterial tree. I have several questions.

Previous experience with biologic arterial substitutes resulted in aneurysm generation. While you show smooth-muscle and capillary ingrowth, is there any evidence of other vascular components such as elastin or type 3 collagen? Other reports show that type 1 and type 4 collagen are prevalent in these grafts, but 3 and elastin are probably much more important for long-term integrity.

Can you make any statement on the stability of this conduit with only a 6-week experiment? Obviously, we are expecting this to hold up over decades.

How do you see this as an improvement over the currently available products, such as cyropreserved homograft, which has achieved some utility in graft replacement in infected fields? And how do you see this material being used? Perhaps the best use would be as a short-term substitute in an infected field with a planned reoperation and insertion of a more durable conduit after the infection has cleared.

Dr. Selwyn M. Vickers (Birmingham, Alabama): I enjoyed your paper. I have 2 quick questions. In our use of SIS in hernias, there have been some difficulties in relationship to its breakdown to specific bacteria collagenases. Have you looked at it in reference to other bacteria and its breakdown? Secondly, because although this is acellular collagen, wouldn't it have made more sense to do this in dogs versus pigs, since this is pig mucosa. Thus, would it have been more applicable to humans if you had done the grafts in dogs versus pigs to understand the xenograft response?

Dr. Basil A. Pruitt, Jr. (San Antonio, Texas): Dr. Croce and Dr. Fabian, I rise to compliment you on continuing this work that you brought before this group last year. My first question is whether you can accelerate the maturation of the conduit by adding growth factors? Secondly, does the collagen production represent a possible long-term deleterious effect? If it does, do you need to curtail that by cytokine manipulation or by use of an anticollagen agent? Thirdly, is there some product left in the graft in the process of producing it that has an antibacterial effect?

Dr. William C. Lineaweaver (Jackson, Mississippi): I think one of the principal issues here is to decide whether or not this material acts more like a vein graft or more like the PTFE. I think one of the critical areas to try to answer this question would be the proximal anastomoses. I wonder if you have examined this area closely for any evidence of early pseudoaneurysm and if you have done any biomechanical testing such as tensile strength or breaking strength to see if in effect that proximal anastomosis is really forming a healed wound.

Second, as mentioned before, this material has been looked at in microsurgical grafting, where it potentially could be used for replantation and flap procedures. The small grafts don't seem to work. Your internal diameter I think was 6 mm, and I wonder if you looked at small diameters and if you had any idea about the limits of smallness in terms of diameter this material can tolerate.

Dr. Adrian Barbul (Baltimore, Maryland): I think that the authors make a very convincing case for this model in the infected setting. However, when one looks at the control noninjected groups, it appears that the PTFE performed better in maintaining luminal diameter measurements and the SIS actually seemed to have much more hypertrophic response.

Dr. Dan Shell (Memphis, Tennessee): Dr. Feliciano, you asked why we did not compare autogenous saphenous vein to small-intestinal submucosa in our study. Autogenous saphenous vein would be our first choice to use when direct primary repair of an artery is not feasible. The intent of our study was to investigate the utility of an off-the-shelf, readily available biomaterial and compare it to the most commonly used synthetic conduit, ePTFE, for vascular reconstruction.

As far as sensitivity testing, we did not perform sensitivity testing. Our study was designed to mimic the clinical scenario utilizing empiric antibiotic therapy. Regarding neointimal hyperplasia, previous studies describing the behavior of ePTFE in the face of infection have noted a decreased neointimal response. One of the results of our study which we found interesting was actually an increased neointimal response in the contaminated ePTFE groups. The increased neointimal response measured at the distal anastomosis in the contaminated groups did not reach statistical significance compared to controls, but it was markedly increased and seemed to correlate with an increased TNF α level measured at euthanasia.

As far as contraction scarring, we did not observe a significant degree of contraction at 42 days postimplantation. Previous studies evaluating small-intestinal submucosa as a growing vascular graft have actually shown growth of the graft concomitant with the blood vessel at up to 1 year postimplantation. The grafts actually showed an increase in length up to 150% of the original postimplantation length.

Our next step will involve using small-intestinal submucosa at varying lengths and lumenal diameters. We also intend to evaluate its durability in the face of contamination with collagenase and elastase producing organisms such as Pseudomonas.

Dr. Flynn asked about aneurysmal degeneration, elastin, and type 3 collagen. Type 3 collagen has been identified within this matrix. Elastin has not. Aneurysmal degeneration has been shown to be a significant problem with biologic grafts such as modified human umbilical vein and negatively charged bovine collagen. This has been thought to be due to calcific degeneration associated with processing steps, specifically, glutaraldehyde crosslinking. The processing of small intestinal submucosa does not involve glutaraldehyde or any crosslinking agents.

As a small-diameter conduit, small intestinal submucosa has been shown to maintain patency and mechanical stability at 180 days postimplantation with no evidence of aneurysmal degeneration.

As far as the stability of the conduit, it has been shown that over a 60-day time course, the elastic modulus and compliance of small intestinal submucosa changes to approach values similar to the native artery. The burst strength of small-intestinal submucosa is much greater than that required to withstand pulsatile arterial pressures.

Your next question was regarding the use of small-intestinal mucosa as a short-term substitute in an infected field. I think that the use of small-intestinal submucosa as a conduit in an infected field to maintain distal perfusion until the infection can be cleared and definitive arterial reconstruction can be performed is a significant potential clinical application. We have shown that it is acutely resistant to infection, maintains mechanical stability, is nonthrombogenic, and could serve a significant clinical role as a bridge to definitive arterial reconstruction.

Dr. Vickers asked about collagenase. One thing we are currently working on is to evaluate the stability of small-intestinal submucosa when exposed to collagenase-producing organisms such as Pseudomonas.

As far as studies in dogs are concerned, dogs are almost prohibitively expensive to perform long-term, statistically significant studies. Xenographic rejection has not been an issue with small-intestinal submucosa. It has been used in over 10,000 human applications with no episodes of rejection. The GAL epitope is responsible for whole-organ xenographic rejection. It has been identified in small quantities within the matrix of small-intestinal submucosa; however, in vitro studies have shown that it elicits an IgG-2-mediated immune response which is a weak activator of complement and thus is not felt to result in rejection response.

Dr. Pruitt asked if anything is left in the graft with antimicrobial properties. There have been substances identified within the matrix of small-intestinal submucosa called cecropins which are antibacterial peptides. They are very small, less than 10 kDa in size. They have been shown to inhibit growth of Gram-positive bacteria, as well as some Gram-negative species in vitro.

Dr. Lineaweaver asked about pseudoaneurysmal degeneration and tensile strength at the proximal anastomosis. We did not observe, and pseudoaneurysmal degeneration has not been observed, in previous studies using small-intestinal submucosa as a vascular conduit. It has been shown, as previously mentioned, that it does approach the biomechanical characteristics of the native artery at time periods of 60 to 90 days postimplantation. As far as its use as a microvascular conduit, a previous study using small-intestinal submucosa in a microvascular model showed uniform failure within 1 hour postimplantation. Covalent heparin bonding was used in an attempt to improve patency rates; however, those grafts failed as well. Future studies are needed to determine the smallest effective lumenal diameter at which the grafts maintain patency and to investigate the effect of lumenal modification with ionic bonding heparin and in vitro endothelialization on small diameter graft patency.