Treatment of vascular graft infection by in situ replacement with cryopreserved aortic allografts: An experimental study Christoph Knosalla, MD, Olivier Goëau-Brissonnière, MD, PhD, Véronique Leflon, MD, Patrick Bruneval, MD, Michel Eugène, MD, PhD, Jean-Claude Pechère, MD, Fabien Koskas, MD, Marie-Hélène Nicolas, MD, Jean-Pascal Leschi, MD, Jean Gerota, MD, and Edouard Kieffer, MD, Paris, France, and Geneva, Switzerland Purpose: The purposes of this study were to prove the efficacy of cryopreserved aortic allografts to treat an established vascular graft infection by in situ replacement in an animal model and to evaluate the role of the antibiotics normally used to decontaminate the allografts. Methods: Twenty-three dogs underwent infrarenal aortic replacement with a gelatinsealed knitted polyester graft contaminated in vitro by Staphylococcus epidermidis RP-62. One week later, the 18 surviving animals underwent reoperation for graft removal and were randomized into three groups for in situ replacement: group I (control, n = 6) received a new gelatin-sealed graft; group II (n = 6) received a non–antibiotic-treated cryopreserved allograft; and group III (n = 6) received an antibiotic-treated cryopreserved allograft. Control grafts and allografts were removed 4 weeks after the initial intervention for quantitative bacteriologic analysis and histologic analysis. Bacteriologic results were expressed as colony-forming units per square centimeter of graft material. Qualitative bacteriologic analysis was also obtained from perigraft fluid and tissue. Results: All of the initially implanted grafts and all of the control grafts (group I) were infected at the time of removal. In group II, three out of six allografts were not totally incorporated, whereas in group III incorporation was always complete, with a significantly decreased inflammatory reaction. All of the antibiotic-treated allografts were sterile, whereas three untreated allografts grew bacteria. Conclusions: In this model, cryopreserved aortic allografts were more resistant to reinfection than synthetic grafts after in situ replacement of an infected prosthetic graft. However, the antibiotic loading of the cryopreserved aortic allograft appears to be essential to obtain optimal therapeutic effects. (J Vasc Surg 1998;27:689-98.)
Prosthetic graft infection remains one of the most dreaded complications of reconstructive vascular surgery, especially when the aorta is involved. Despite optimal surgical technique and appropriate systemic antibiotic prophylaxis, the incidence of graft infection persists to range between 1.0% and
6.0%.1-3 Although continuous improvement in results has been achieved over the past decade, prosthetic graft infection is still associated with high mortality and amputation rates.4-6 Dissatisfaction with extraanatomic techniques has led to the development of approaches of in situ replacement by
From the Department of Vascular Surgery, Pitié-Salpêtrière University Hospital, Paris (Drs. Knosalla, Koskas, and Kieffer); the Department of Vascular Surgery (Drs. Goëau-Brissonnière and Leschi) and the Department of Microbiology (Drs. Leflon and Nicolas), Ambroise-Paré University Hospital, Boulogne-Billancourt, France; the Department of Pathology and INSERM U 430, Broussais University Hospital, Paris (Dr. Bruneval); the NMR Laboratory, Physiology Department, Saint-Louis University Hospital, Paris (Dr. Eugène); the Department of Genetics and Microbiology, Faculty of Medicine, Geneva (Dr. Pechère); and the Tissue Bank, ETS, AP-HP, Saint-Louis University Hospital, Paris (Dr. Gerota).
Supported by grants AOA94029 and AOM95146 from The French “Ministère de la Santé,” Assistance Publique-Hôpitaux de Paris, by grant RFG3 from Etablissement Français des Greffes, and by grants from Université Paris V, Faculté de Médecine Paris-Ouest, France. Reprint requests: Olivier Goëau-Brissonnière, MD, PhD, Service de Chirurgie Vasculaire, Hôpital Ambroise Paré, 9 Avenue Charles de Gaulle, 92104 Boulogne-Billancourt, France. Copyright © 1998 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/98/$5.00 + 0 24/1/89392
689
690 Knosalla et al.
prosthetic,7-9 autogenous,10-12 or allograft13 material to avoid complications related to aortic stump disease and subsequent infection and occlusion of extraanatomic bypass grafts.14 In situ replacement with so-called “fresh” aortic allografts preserved at 4° C has been used by Kieffer and associates13 in 43 patients to treat aortic graft infections with very encouraging early and midterm results. Experimental and clinical data suggest that fresh arterial allografts are more resistant to bacterial contamination than prosthetic grafts.13,15,16 However, conservation at 4° C is a less-thanoptimal method. The limited duration of storage precludes effective tissue banking, matching of donor/recipient blood or tissue compatibility, and a sufficient prevention of viral transmission before transplantation. Cryopreservation might be the optimal technique to solve these problems. To date, the clinical experience with cryopreserved aortic allografts to treat prosthetic graft infections or mycotic aneurysms of the aorta is still limited to small numbers of patients.17-19 Nevertheless, the excellent results obtained in cardiac surgery with cryopreserved allografts in managing infections of the aortic valve and the ascending aorta20,21 in larger numbers of patients suggest that cryopreserved allografts are also a suitable replacement graft material for prosthetic graft infection in the abdominal aortic region. The purposes of this study were to prove the efficacy of cryopreserved aortic allografts to treat an established vascular graft infection by in situ replacement in an animal model and to evaluate the role of the antibiotics normally used for decontamination of the allograft during the cryopreservation process. MATERIAL AND METHODS Graft material. Standard, commercially available gelatin-sealed knitted polyester prostheses (Gelsoft, Vascutek Limited, Inchinnan, Scotland) were used in this study. All grafts were 6 mm in diameter. Aortic allografts. The descending thoracic and abdominal aortas of four dogs were harvested under sterile conditions and were cut into 6 cm segments, which were stored according to the study groups. Preservation of allografts. Immediately after excision, 12 aortic allograft segments were stored at 4° C in a sterile preservation fluid containing polyethylene glycol (Microval, Braun Medical S.A., Boulogne, France) and were transported in isothermic boxes to the tissue bank of Saint-Louis Hospital in Paris. After a maximal period of 4 hours, the allografts were further prepared inside a first-grade laminar flow room. The allografts were transferred for 48 hours in a con-
JOURNAL OF VASCULAR SURGERY April 1998
servation medium (Plasmagel, Bellon, Neuilly sur Seine, France) that was or was not, according to the study group, supplemented with antibiotics (vancomycin 0.3 mg/ml, lincomycin 0.125 mg/ml) normally used for the decontamination of vascular grafts in the tissue bank. To submit the allografts to the cryopreservation process, the grafts were placed for 120 minutes in a cryopreservation medium consisting of Ringer’s solution, polyethylene glycol 20.000 d, 30 g/L, and 12.5% dimethylsulfoxide (B. Braun, Boulogne, France).22 The grafts were then cryopreserved at a computer-controlled rate in a freezer that lowered the temperature 1° C per minute down to a temperature of –40° C and 5° C per minute down to –130° C. The increase in temperature during the crystallization process was minimized by a close control of the rate during this period. The specimens were then placed in the vapor phase of a liquid nitrogen freezer. Before implantation the allografts were prepared in the operating room. The allograft was thawed in a water bath at 42° C and was then rinsed in successive baths of saline solution that contained decreasing concentrations of dimethylsulfoxide (8%, 4%, 2%, NaCl 0.9%). No fractures were observed. Before cryopreservation and before implantation, specimens of all allografts were submitted to bacteriologic analysis to rule out a contamination during the preservation process. Bacterial strain. Staphylococcus epidermidis RP62 was used in this study. This clinical isolate provided by the Department of Genetics and Microbiology of the Faculty of Medicine of Geneva has been used in previous experiments.23 This organism demonstrated slime production when incubated in trypticase soy broth with 1% dextrose for 18 hours at 37° C. The strain was susceptible to vancomycin with a minimal inhibitory concentration of 2.0 mg/L and was resistant to lincomycin (minimal inhibitory concentration, >1024 mg/L). According to the diffusion disk agar method with the recommendations of the French Committee of Antibiogram, the strain was susceptible to tetracycline, chloramphenicol, pristinamycin, rifampin, vancomycin, teicoplanin, fusidic acid, pefloxacin, and fosfomycin and was resistant to penicillin G, methicillin (heterologous), gentamicin, tobramycin, kanamycin, erythromycin, lincomycin, and trimethoprim, associated or not with sulfamids.24 When identified with an Api-Staph system (Apisystem, La Balme-les Grottes, France), S. epidermidis RP-62 fermented glucose, fructose, maltose, lactose, and was positive for nitrate reductase, Voges-Proskauer, saccharose, arginine test, and urea reactions. For inoculum preparation, S. epidermidis RP-62
JOURNAL OF VASCULAR SURGERY Volume 27, Number 4
was cultivated on sheep blood agar (SBA) plates (BioMérieux, Marcy l’Etoile, France) in an aerobic atmosphere at 37° C for 18 hours. Then four freshly grown colonies were inoculated into 40 ml of trypticase soy broth with 1% dextrose and were incubated for 18 hours at 37° C. The number of viable organisms present after an 18-hour incubation was expressed as the number of colony forming units (CFU) per milliliter of broth. In vitro graft colonization. Gelatin-sealed knitted polyester grafts were colonized in vitro. They were immersed in an 18-hour incubating broth of S. epidermidis RP-62 for 15 minutes at room temperature. After immersion the grafts were rinsed gently with 20 ml of regular sterile saline solution to remove nonadherent microorganisms. Animal model. After intravenous administration of pentobarbital (30 mg/kg), 23 mongrel dogs weighing 15 to 20 kg were orally intubated and mechanically ventilated. Through a median laparotomy, they underwent replacement of the infrarenal aorta with a gelatin-sealed polyester vascular prosthesis under sterile surgical conditions. The infrarenal aorta was transected, and a 5 cm segment of graft was inserted in an end-to-end fashion with 6-0 Prolene sutures (Ethicon, Neuilly sur Seine, France). Before implantation the graft was colonized in vitro as described above. After implantation, the graft was covered by reapproximating the retroperitoneum, and the abdomen was closed in layers with a standard surgical technique. One week after the initial graft had been implanted, the surviving animals underwent reoperation. The aortic graft was exposed through the previous midline incision with sterile surgical technique. After proximal and distal control of the aorta adjacent to the graft, local signs of infection were noted, including tissue inflammation, perigraft fluid and cavity formation, absence of graft incorporation, anastomotic dehiscence, and false aneurysm formation, which was defined as an anastomotic weakening associated with an eccentric dilatation of the aortic wall. The graft was then completely excised, including anastomotic sites and 5 mm of abdominal aorta proximal and distal to the graft, and the retroperitoneum was debrided. Specimens of the anastomotic sites and perigraft tissues were submitted to histologic studies. The prosthetic graft, the perigraft tissues, and the perigraft fluid were also submitted to bacteriologic analysis with bacterial counts. The animals were then randomly assigned to one of three experimental groups: in group I (control, n = 6) animals received an untreated gelatin-sealed knitted polyester vascular prosthesis; in group II
Knosalla et al.
691
(untreated allografts, n = 6) they received a cryopreserved allograft that was not treated with antibiotics; and in group III (antibiotic-treated allografts, n = 6) they received an allograft treated with antibiotics. In each case the replacement graft, 6 cm in length, was implanted in situ in an end-to-end fashion with 6-0 polypropylene sutures. The graft was again covered with the retroperitoneum, and the abdomen was closed in layers. No adjunct systemic antibiotic therapy was administrated to the animals. Four weeks after the initial intervention all animals were anesthetized and killed to recover the specimen for bacteriologic and histologic analysis. Venous blood was sampled just before death for qualitative blood culture. The grafts were evaluated for graft patency, false aneurysm formation, the degree of graft incorporation, and the presence of perigraft fluid or cavity. The prosthetic grafts or the allografts and the adjacent aorta were totally excised. Specimens of lung, liver, spleen, and kidney were also taken for bacteriologic culture. The grafts were rinsed gently with 20 ml of normal saline solution and were submitted to bacterial counts. Specimens of the perigraft tissues were taken separately for histologic and bacteriologic studies. Specimens of the anastomotic sites were taken for histologic studies. All animals received humane care in compliance with the “Principles of Laboratory Animal Care” and the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 86-23, revised 1985). Bacteriologic studies. All removed polyester grafts and allografts were weighed, and their length and internal diameter were measured. Then each graft was cut into two fragments. Weighing each fragment allowed the determination of its surface. Each fragment was hand-crushed for 2 minutes after addition of 2 ml normal sterile saline solution. The crushing effluent was submitted to adequate successive tenfold dilutions, and 0.1 ml aliquots were cultured onto SBA plates at 37° C for 48 hours. Colony counts were expressed as the number of CFU per cm2 of graft material. Venous blood obtained just before graft explantation was cultivated in a blood culture bottle (Hemoline Performance Diphasique, BioMérieux, Marcy l’Etoile, France) at 37° C for 48 hours. The perigraft tissues were hand-crushed for 2 minutes after addition of 2 ml normal sterile saline solution and were cultured onto SBA plates at 37° C for 48 hours. Organ specimens (lung, spleen, liver, and kidney) and perigraft fluid were cultivated in buffered dextrose broth (Sanofi Diagnostic Pasteur, Marnes-la-Coquette, France) for 48 hours at 37° C. Subcultures were made for each broth using SBA plates incubated for 48 hours at 37° C. When Gram-
JOURNAL OF VASCULAR SURGERY April 1998
692 Knosalla et al.
Table I. Macroscopic evaluation of prosthetic grafts and cryopreserved aortic allografts
Group I Initial grafts Control grafts Group II Initial grafts Untreated allografts Group III Initial grafts AB-treated allografts
Graft patency rate
False aneurysms
Anastomotic disruption
Perigraft inflammation
Graft incorporation
Mean score*
6/6 6/6
6/6 4/6
0/6 1/6
6/6 6/6
0/6 0/6
12/18 11/18 (NS)
5/6 5/6
3/6 0/6 (NS)
0/6 0/6
6/6 3/6 (NS)
0/6 3/6 (NS)
9/18 3/18 (p < 0.04)
6/6 6/6
4/6 0/6 (p < 0.03)
0/6 0/6
6/6 0/6 (p < 0.01)
0/6 6/6 (p < 0.01)
10/18 0/18 (p < 0.002)
*Mean score is false aneurysms + anastomotic disruption + perigraft inflammation. Initial grafts, colonized gelatin-sealed knitted polyester grafts implanted for 7 days; control grafts, in situ replacement with gelatin-sealed knitted polyester grafts; untreated allografts, cryopreserved aortic allografts prepared without antibiotics; AB-treated allografts, cryopreserved aortic allografts prepared with antibiotics; NS, not significant. p values are for allografts versus initial grafts.
positive cocci were isolated, they were identified with genus-specific Apisystem (Apisystem, La Balme-les-Grottes, France), and their susceptibility to antibiotics was determined using the agar disk antibiotic diffusion method. Histologic studies. During explantation of the prosthetic grafts and allografts, specimens were taken from the perigraft tissues and the proximal and distal aortic anastomotic sites. After fixation in 10% buffered formaldehyde solution, the specimens were embedded in paraffin, sectioned, and stained with hematoxylin-eosin. Histologic parameters of inflammation assessed in this study were the density of cellular infiltration, the amount of polymorphonuclear cells, and the absence or presence of suppurative necrosis. Aortitis was defined as an inflammatory cellular infiltration of the aortic media with polymorphonuclear cells. Statistical analysis. The results of the macroscopic evaluation (false aneurysms, anastomotic disruption, perigraft inflammation; Table I) and the results of the microscopic evaluation of perigraft tissues (Table II) and anastomotic sites (Table III) were summed up and expressed as a mean score in each experimental group for initial grafts and replacement grafts, respectively. The data from the study were analyzed with χ2 analysis and, when appropriate, the one-tailed Fisher’s exact test for small groups or Mann-Whitney-Wilcoxon test. Statistical significance was assigned when p values were less than 0.05. RESULTS The median number of viable organisms present in the inoculum used for the in vitro colonization of the prosthetic grafts before their implantation in the
in vivo experiments was 8.5 × 108 CFU/ml (range, 106 to 4 × 109) without significant difference between the experimental groups. Five animals died before reoperation and in situ replacement. Three had an anastomotic disruption 5 or 6 days after operation, and two died from sepsis on postoperative days three and four, respectively. Eighteen animals were thus included in the study. Pathologic findings. At the time of retrieval, all 18 initially implanted grafts and the six prosthetic replacement grafts of the control group (group I) revealed anatomic signs of infection (Table I). All of these grafts except one in group II were patent. The retroperitoneum was involved with acute inflammation of these 24 prosthetic grafts and adjacent structures. All of the prosthetic grafts were not incorporated and surrounded by a cavity that extended to the anastomoses and contained a suppurative exudate. Seventeen grossly infected pseudoaneurysms were found. In the control group, two animals died before scheduled, one of sepsis and the other of anastomotic disruption. All of the animals that were treated with allograft replacement were healthy at the time of death. Four weeks after the initial intervention all of the antibiotic-treated allografts (group III) and three of the six untreated allografts (group II) were perfectly incorporated without any sign of inflammation or pseudoaneurysm formation. Three of the untreated allografts were not completely incorporated and revealed macroscopic signs of perigraft inflammation, but without perigraft fluid. On microscopic examination, perigraft tissues of all of the initial grafts showed evidence of an inflammatory infiltration with lymphocytes and macrophages (Fig. 1). Polymorphonuclear cells were present in 72% of the samples, and suppura-
JOURNAL OF VASCULAR SURGERY Volume 27, Number 4
Knosalla et al.
693
Fig. 1. Perigraft tissues of a dog 7 days after implantation of an infected gelatin-sealed knitted polyester graft. Hematoxylin-eosin stain at original magnification of ×80 shows inflammatory infiltration in the soft tissues with polymorphonuclear cells.
tive necrosis was found in 17% of them (Table II). In the samples of perigraft tissue of the untreated allografts, inflammatory infiltration was found in four of six samples, with presence of polymorphonuclear cells in three samples. In these three cases the grafts were macroscopically unincorporated. In the group that received antibiotic-treated allografts (group III), perigraft inflammatory infiltration was found in two of six specimens, with polymorphonuclear cells in one of these two specimens. The other had no signs of inflammation (Fig. 2). After 3 weeks, the amount of suppurative necrosis did not change significantly in the experimental groups when compared with the results of the initial grafts (p = NS). The adventitia of the aortic anastomotic sites of the initial grafts was infiltrated with inflammatory cells in 83% of cases (Fig. 3). In this infiltrate, polymorphonuclear cells were found in 61% of the specimens (Table III). In groups II and III, 3 weeks after allograft replacement, the incidence of anastomotic inflammatory infiltration was 67% (8 of 12 in each group), and polymorphonuclear cells were present in 37.5% (3 of 8 in each group) of the explants. The other explanted allografts had no evidence of inflammation (Fig. 4). Bacteriologic cultures. The results of bacteriologic cultures are shown in Table IV. At the time of reoperation, all of the initially implanted prosthetic
grafts were infected with S. epidermidis RP-62. There was no significant difference (p = NS) in the density of bacteria measured in the initial grafts of the different study groups. All of the prosthetic replacement grafts (group I) were infected at the time of retrieval. In group II, the study strain was found in one allograft (Table IV). In two other animals of group II, the allografts were not incorporated and grew other bacteria. One grew group C β-hemolytic streptococci (3.5 × 104 CFU/cm2) and the other grew S. chromogenes (3.34 × 105 CFU/cm2). None of the antibiotic-treated allografts of group III was infected, whereas all of the prosthetic replacement grafts (group I) grew S. epidermidis RP-62 (p < 0.01 when compared with group III). No concomitant growth of other bacteria was observed in group III. The perigraft fluid and tissues taken during reoperation and at the time of death in the control group grew S. epidermidis RP-62 in 50% and 25% of cases, respectively, whereas bacteriologic cultures of the perigraft tissues surrounding the allografts always had negative results. In these experiments, all organ and blood samples were sterile. DISCUSSION This study demonstrates the efficacy of in situ replacement by cryopreserved aortic allografts to treat an established S. epidermidis vascular graft
694 Knosalla et al.
JOURNAL OF VASCULAR SURGERY April 1998
Fig. 2. Perigraft tissues of a dog 21 days after implantation of an antibiotic-treated allograft. Hematoxylin-eosin stain at original magnification of ×80 demonstrates noninflammatory fibrotic tissue.
Fig. 3. Perianastomotic tissues of a dog 7 days after implantation of an infected gelatin-sealed knitted polyester graft. Hematoxylin-eosin stain at original magnification of ×80 demonstrates suppurative necrosis in perianastomotic fibrosis (*) adjacent to the media (M), which shows calcification.
infection. The model of infrarenal aortic S. epidermidis polyester graft infection that we used mimics the clinical situation of late graft infection, resulting in a perigraft inflammatory process and anastomotic false aneurysm formation.23,25,26 This model proba-
bly differs from the clinical setting only in regard of the degree of contamination and the required time before the infection is established.23,26 The study strain was a clinical isolate of S. epidermidis, which now causes an increasing number of
JOURNAL OF VASCULAR SURGERY Volume 27, Number 4
Knosalla et al.
695
Fig. 4. Perianastomotic tissues of a dog 21 days after implantation of an antibiotic-treated allograft. Hematoxylin-eosin stain at original magnification of ×80 demonstrates noninflammatory adventitial and perianastomotic fibrosis (*) adjacent to the media (M).
Table II. Microscopic evaluation of the perigraft tissues of prosthetic grafts and cryopreserved aortic allografts Inflammatory infiltration Group I Initial grafts Control grafts Group II Initial grafts Untreated allografts Group III Initial grafts AB-treated allografts
Polymorphonuclear cells
Suppurative necrosis
Mean score*
6/6 6/6
4/6 5/6 (NS)
2/6 2/6
12/18 13/18 (NS)
6/6 4/6
4/6 3/6 (NS)
0/6 2/6 (NS)
10/18 9/18 (NS)
6/6 2/6 (p < 0.03)
5/6 1/6 (p < 0.05)
1/6 1/6
12/18 4/18 (p < 0.01)
*Mean score is inflammatory infiltration + polymorphonuclear cells + suppurative necrosis. Initial grafts, colonized gelatin-sealed knitted polyester grafts implanted for 7 days; control grafts, in situ replacement with gelatin-sealed knitted polyester grafts; untreated allografts, cryopreserved aortic allografts prepared without antibiotics; AB-treated allografts, cryopreserved aortic allografts prepared with antibiotics; NS, not significant. p values are for allografts versus initial grafts.
vascular prosthetic graft infections.8 These organisms produce an extracellular matrix of glycocalyx that facilitates bacterial adherence and colonization of the graft material and protects the bacteria from host defenses and antibiotic penetration. The applied technique of in vitro colonization enables a high reproducibility in the quantitative colonization of prosthetic grafts23 with a constant vascular graft infection after 1 week of implantation. As demonstrated in the animals of group II, untreated cryopreserved aortic allografts are more
resistant to reinfection than prosthetic grafts. Three out of six prosthetic graft infections could be successfully treated by in situ replacement with a cryopreserved aortic allograft within a 3-week period of time without any local or systemic antibiotic treatment. These allografts were macroscopically perfectly incorporated with negative bacterial cultures and decreased histologic signs of inflammation when compared with the initial grafts. This is remarkable because the replacement allografts were implanted into a field of acute infection of perigraft tissues and
JOURNAL OF VASCULAR SURGERY April 1998
696 Knosalla et al.
Table III. Microscopic evaluation of the anastomotic sites of prosthetic grafts and cryopreserved aortic allografts Perianastomotic inflammatory infiltration Group I Initial grafts Control grafts Group II Initial grafts Untreated allografts Group III Initial grafts AB-treated allografts
Polymorphonuclear cells
Suppurative necrosis
Aortitis
Mean score*
10/12 10/12
7/12 8/12
2/12 4/12
1/12 3/12
20/48 25/48 (NS)
10/12 8/12
7/12 3/12 (NS)
1/12 1/12
0/12 2/12 (NS)
18/48 14/48 (NS)
10/12 8/12
8/12 3/12 (p < 0.05)
3/12 0/12 (NS)
1/12 0/12 (NS)
22/48 11/48 (p < 0.01)
*Mean score is perianastomotic inflammatory infiltration + polymorphonuclear cells + suppurative necrosis + aortitis. Initial grafts, colonized gelatin-sealed knitted polyester grafts implanted for 7 days; control grafts, in situ replacement with gelatin-sealed knitted polyester grafts; untreated allografts, cryopreserved aortic allografts prepared without antibiotics; AB-treated allografts, cryopreserved aortic allografts prepared with antibiotics; NS, not significant. p values are for allografts versus initial grafts.
Table IV. Bacteriologic evaluation of graft infection with S. epidermidis RP-62 No. with positive culture/total examined
Group I Initial grafts Control grafts Group II Initial grafts Untreated allografts Group III Initial grafts AB-treated allografts
Grafts
Fragments
Viable counts (CFU/cm2): median (range)
6/6 6/6
12/12 12/12
7.9 × 104 (6 × 103 to 1 × 106) 9.8 × 104 (1.6 × 104 to 1.1 × 106)
6/6 1/6*
12/12 2/12†
2.03 × 104 (1.94 × 103 to 4.61 × 105) 5.54 × 104
6/6 0/6*
12/12 0/12‡
3.32 × 103 (1.03 × 103 to 2.95 × 104) 0
Initial grafts, colonized gelatin-sealed knitted polyester grafts implanted for 7 days; control grafts, in situ replacement with gelatin-sealed knitted polyester grafts; untreated allografts, cryopreserved aortic allografts prepared without antibiotics; AB-treated allografts, cryopreserved aortic allografts prepared with antibiotics. *p < 0.01 when compared with control grafts and initial grafts. †p < 0.001 when compared with control grafts and initial grafts. ‡p < 0.0001 when compared with control grafts and initial grafts.
anastomotic sites and all of the prosthetic replacement grafts of the control group were infected at the time of death. The mechanisms by which cryopreserved aortic allografts resist infection are not well defined. Different factors can be hypothesized. As was shown by Mitchell and coworkers,27 cryopreservation preserves an amorphous and fibrillar extracellular matrix that is immunologically inert and protects the graft from autolysis. Compared with fresh allografts, the cryopreserved allografts appear to keep their intimal integrity, which results in a reduced surface thrombogenicity28 and reduces the risk of hematogenous bacterial colonization. The active and passive immunologic properties of aortic allografts, like the
presence of MHC class I and II molecules29 with subsequent T-cell activation30 or the expression of endothelial-leukocyte adhesion molecules such as endothelial leucocyte adhesion molecule-1, vascular cellular adhesion molecule-1, and intercellular adhesion molecule-1, which can be induced after cryopreservation,31 may play a role in modulating the host defenses and bacterial adherence. All of these hypothesizes need to be verified by further experimental studies. However, in our experience the resistance of untreated cryopreserved aortic allografts was not complete. Persistent perigraft infection and reinfection of the replacement graft, even with other bacteria present at the time of reoperation in a septic
JOURNAL OF VASCULAR SURGERY Volume 27, Number 4
operation field, remain a risk. Antibiotic loading resulting from the 48-hour decontamination in a medium supplemented with antibiotics performed before cryopreservation of the allografts might then have contributed to the increased resistance to infection of the aortic allografts that we observed in the animals of group III. Previous experimental studies with antibiotic-bonded prosthetic grafts demonstrated that rifampin-bonded gelatin-sealed polyester grafts were highly resistant to early reinfection after in situ replacement of an infected graft.23,32 In the present work, the study strain was highly sensitive to vancomycin, and the concentration of vancomycin in the decontaminating solution was very high. The antibiotic may then have been carried over in the tissues surrounding the allografts after implantation, playing a role in the treatment of graft infection. The antibiotic-treated allografts were perfectly incorporated with negative bacterial cultures and significantly decreased histologic signs of inflammation when compared with the initial grafts (p < 0.01) at the time of death, whereas persistent perigraft infection and reinfection of the replacement allografts were observed in three of the six untreated allografts. The group C β-hemolytic streptococci and the S. chromogenes that grew from two untreated allografts in group II were probably contaminants that colonized the allografts at the time of graft replacement in a septic field. Because these allografts were not antibiotic-treated, the bacteria could grow, leading to reinfection. The 3-week duration of implantation chosen for the replacement grafts in this study may appear short and does not allow the exclusion of the occurrence of a late infection. Nevertheless, negative bacterial cultures and the perfect incorporation of the allografts can be considered as an evidence of their resistance to early reinfection. In this study, the antibiotic treatment of the allografts was performed according to the clinical protocol used for vascular allografts by the tissue bank, using vancomycin and lincomycin to decontaminate the allografts. Nevertheless, the demonstrated role of antibiotics in the resistance of allografts to infection should have important implications for the antibiotic procurement of cryopreserved aortic allografts. The antibiotics used in the cryopreservation process should be chosen not only in regard to the decontamination of the allograft but also in regard to the spectrum of bacteria involved in the infections intended to treat by in situ replacement with cryopreserved aortic allografts. Further experimental studies are necessary to determine the optimal antibiotics for the preparation of aortic allografts.
Knosalla et al.
697
Although late deterioration may be expected with arterial allografts stored at 4° C,13 excellent long-term results have been reported by heart surgeons with cryopreserved allograft replacement for management of infections that involve the ascending aorta.19,33 Future modifications in the techniques of allograft preparation and cryopreservation should allow the reduction of antigenicity and have better long-term results. In conclusion, cryopreserved aortic allografts are a suitable material for arterial reconstruction in grossly infected fields. In this animal model, even without adjunct antibiotic therapy, aortic allografts were capable of bringing a prosthetic graft infection into complete remission within a 3week period of time. However, the antibiotic procurement of cryopreserved aortic allograft appears to be essential to obtain optimal therapeutic effects. We are indebted to Thierry Marchix, BS, and Roger Gouezo, BS, for their technical assistance in cryopreservation of the allografts; to Virginie Metral, MD, for her technical assistance in bacteriologic preparations; and to Mrs. Martine Douheret for her technical assistance in histologic processing. We also thank Sulzer-Vascutek Ltd. for providing the prosthetic grafts used in this study. REFERENCES 1. Bunt TJ. Synthetic vascular graft infections. I. Graft infections. Surgery 1983;93:733-46. 2. Lorentzen JE, Nielsen OM, Arendrup H, Kimose HH, Bille S, Andersen J, et al. Vascular graft infection: an analysis of sixty-two graft infections in 2411 consecutively implanted synthetic vascular grafts. Surgery 1985;98:81-6. 3. O’Hara PJ, Hertzer N, Beven EG, Krajewski LP. Surgical management of infected abdominal aortic grafts: review of 25-year experience. J Vasc Surg 1986;3:725-31. 4. Reilly LM, Stoney RJ, Goldstone J, Ehrenfeld WK. Improved management of aortic graft infection: the influence of operation sequence and staging. J Vasc Surg 1987;5:421-31. 5. Schmitt DD, Seabrook GR, Bandyk DF, Towne JB. Graft excision and extra-anatomic revascularization: the treatment of choice for septic aortic prosthesis. J Cardiovasc Surg 1990;31:327-32. 6. Yeager RA, Moneta GL, Taylor LM, Harris EJ, McConnel DB, Porter JM. Improving surgical and limb salvage in patients with aortic graft infection. Am J Surg 1990; 159:466-9. 7. Walker WE, Cooley DA, Duncan JM, Hallman GL, Ott DA, Reul GJ. The management of aortoduodenal fistula by in situ replacement of the infected abdominal aortic graft. Ann Surg 1987;205:727-32. 8. Bandyk DF, Bergamini TM, Kinney EV, Seabrook GR, Towne JB. In situ replacement of vascular prostheses infected by bacterial biofilms. J Vasc Surg 1991;13:575-83. 9. Torsello G, Sandmann W, Ghert A, Jungblut R. In situ replacement of infected vascular prosthesis with rifampicinsoaked vascular grafts: early results. J Vasc Surg 1993; 17:768-73.
698 Knosalla et al.
10. Clagett GP, Bowers BL, Lopez-Viego MA, Rossi MB, Valentine RJ, Myers SI, et al. Creation of a neo-aortoiliac system from lower extremity deep and superficial veins. Ann Surg 1993;218:239-49. 11. Ehrenfeld WK, Wilbur BG, Olcott CN IV, Stoney RJ. Autogenous tissue reconstruction in management of infected prosthetic grafts. Surgery 1979;85:82-92. 12. Quiñones-Baldrich WJ, Gelabert HA. Autogenous tissue reconstruction in management of aortoiliofemoral graft infection. Ann Vasc Surg 1990;4:223-8. 13. Kieffer E, Bahnini A, Koskas F, Ruotolo C, LeBlevec D, Plissonnier D. In situ allograft replacement of infected infrarenal aortic prosthetic grafts: results in forty-three patients. J Vasc Surg 1993;17:349-56. 14. Bacourt F, Koskas F, French Association for Research in Surgery. Axillobifemoral bypass and aortic exclusion for vascular septic lesions: a multicenter retrospective study of 98 cases. Ann Vasc Surg 1992;6:119-26. 15. Moore WS, Swanson RJ, Campagna G, Bean B. The use of fresh tissue arterial substitutes in infected fields. J Surg Res 1975;18:229-33. 16. Koskas F, Goëau-Brissonnière O, Nicolas MH, Bacourt F, Kieffer E. Arteries from human beings are less infectible by Staphylococcus aureus than polytetrafluoroethylene in an aortic dog model. J Vasc Surg 1996;23:472-6. 17. Mestres CA, Mulet J, Pomar JL. Large-caliber cryopreserved arterial allografts in vascular reconstructive operations: early experience. Ann Thorac Surg 1996;60:S105-7. 18. Knosalla C, Weng Y, Yankah CA, Hofmeister J, Hetzer R. Using aortic allograft material to treat mycotic aneurysms of the thoracic aorta. Ann Thorac Surg 1996; 61: 1146-52. 19. Vogt PR, von Segesser LK, Goffin Y, Niederhäuser U, Genoni M, Künzli A, et al. Eradication of aortic infections with the use of cryopreserved arterial homografts. Ann Thorac Surg 1996;62:640-5. 20. McGriffin DC, Galbraith AJ, McLachlan GJ, Stower RE, Wong ML, Stafford EG, et al. Aortic valve infection: risk factors for death and recurrent endocarditis after aortic valve replacement. J Thorac Cardiovasc Surg 1992;104:511-20. 21. Knosalla C, Siniawski H, Weng Y, Yankah AC, Hetzer R. Diagnosis and surgical treatment of active infective aortic valve endocarditis with associated periannular abscess. Cardiovasc Surg 1995;3(suppl 1):14.
JOURNAL OF VASCULAR SURGERY April 1998
22. Eugène M, M’Bengue A, Bauza G, LeMoyec L, Gouezo R, Gerota J, et al. Method for cryopreserving human arteries: 1H NMR spectroscopy for measuring the kinetics of permeation and the ice forming tendency of cryoprotective agents. Transplant Proc 1996;28:345. 23. Goëau-Brissonnière O, Mercier F, Nicolas MH, Bacourt F, Coggia M, Lebrault C, et al. Treatment of vascular graft infection by in situ replacement with a rifampicin-bonded gelatin-sealed Dacron graft. J Vasc Surg 1994;19:739-44. 24. Comité de l’Antibiogramme de la Société Française de Microbiologie 1994. Communiqué. Path Biol 1994; 42(8):1-8. 25. Bergamini TM, Bandyk DF, Govostis D, Kaebnick HW, Towne JB. Infection of vascular prostheses caused by bacterial biofilms. J Vasc Surg 1988;7:21-30. 26. Martin LF, Harris JM, Fehr DM, Peter AO, Appelbaum PC, Spangler SK, et al. Vascular prosthetic infection with Staphylococcus epidermidis: experimental study of pathogenesis and therapy. J Vasc Surg 1989;9:464-71. 27. Mitchell RN, Jonas RA, Schoen FJ. Structure-function correlations in cryopreserved allograft cardiac valves. Ann Thorac Surg 1995;60:S108-13. 28. Boren CH, Anthony AJ, Moore WS. Maintenance of viable arterial allografts by cryopreservation. Surgery 1977;83: 382-91. 29. Khatib HE, Lupinetti FM. Antigenicity of fresh and cryopreserved rat valve allografts. Transplantation 1990;49:765-7. 30. Fischlein T, Schütz A, Haushofer M, Frey R, Uhlig A, Detter C, et al. Immunologic reaction and viability of cryopreserved homografts. Ann Thorac Surg 1995;60:S122-6. 31. Mulligan MS, Tsai TT, Kneebone JM, Ward PA, Lupinetti FM. Effects of preservation techniques on in vivo expression of adhesion molecules by aortic valve allografts. J Thorac Cardiovasc Surg 1994;107:717-23. 32. Colburn MD, Moore WS, Chvapil M, Gelabert HA, Quiñones-Baldrich WJ. Use of an antibiotic-bonded graft for in situ reconstruction after prosthetic graft infections. J Vasc Surg 1992;16:651-60. 33. Zwischenberger JB, Shalaby TZ, Conti VR. Viable cryopreserved aortic homograft for aortic valve endocarditis and annular abscesses. Ann Thorac Surg 1989;48:365-70. Submitted Dec. 10, 1997; accepted Feb. 4, 1998.