Volume 13 Number 5 May 1991
bodies may thereby be more likely to develop a truly biocompatible vascular prosthesis. Ralph S. Greco, MD Untiersityof Medicine and Dentisty of New Jersey Robert Wood Johnson Medical School New Brunswick, NJ.
REFERENCES 1. Elek SD, Cohen PE. The virulenceof staphylococcus pyogenesfor man: a studyof the problemsof wound infection.Br J Exp Path01 1957;38:573. 2. Margiotta MS, Robertson FM, Greco RS. Selective induction of intercellular adhesion molecule (ICAM-I) expression by human endothelial cells following adherence to vascular grafts using an in vitro model. Surg Forum 1980;41:339-41. 3. Whiffen JD, Gott VL. In vivo adsorption of heparin by graphite-benzalkonium intravascular surfaces. Surg Gynecol Obstet 1965;121:287-90. 4. Harvey RA, Greco RS. The non-covalent bonding of antibiotics to a polytetrafluoroethylene graft. Ann Surg 1981;194: 642-7. 5. GrecoRS, Harvey BA, Henry R, et al. Preventionof graft infection by antibiotic bonding. Surg Forum 198O;XXXI:2930. 6. Greco RS, Harvey RA. The role of antibiotic bonding in the prevention of vascular prosthetic infections. Ann Surg 1982; 195:168-71. Greco RS, Harvey RA, Smilow PC, et al. Prevention of vascular prosthetic infection by a benzalkonium-oxacillin bonded polytetrafluoroethylene graft. Surg Gynecol Obstet 1982;155:28-32. Harvey RA, Greco RS. Antibiotic bonding to polytetralluoroethylene with tridodecylmethylammonium chloride. Surgery 1982;92:504-12. Greco RS, Donen. Al’, Harvey BA. The application of antibiotic bonding to the treatment of established vascular prosthetic infection. Arch Surg 1985;120:71-5. 10. Shue WB, Worosilo SC, Donetz Al’, Trooskin SZ, Harvey RA, Greco RS. Prevention of vascular prosthetic infection with an antibiotic bonded Dacron graft. J VASC SURG 1988;8:600-5.
PATHOGENESIS OF VASCULAR GIUFT INFECTIONS The increasingly routine use of prosthetic vascular grafts has revolutionized our ability to reperfuse and salvagepatients with severeperipheral vasculardisease.It is estimated that more than 60,000 vascular graft prostheses are implanted in the United States yearly.’ Despite improved prosthetic materials and surgical techniques, vascular graft infections remain a serious source of clinical morbidity and deaths. Approximately 2% of patients with vascular grafts will develop a prosthetic infection, the consequencesof which are often grave, that is, limb loss or death.’ To limit the incidence of these infections it is of paramount importance that we have a full understanding of the pathogenesis of graft infections. Only in this way can we hope to make significant inroads into the prevention and eradication of vascular prosthetic infections. Graft infections can be separated into those that occur acutely and those that occur late. Early postoperative graft
Special
cummunicatim
755
infections are often the result of wound infections and usually involve grafts anatomically superficial in location (femoral, axillary, popliteal sites). Late graft infections may occur months to yearsafter vascularreconstructive surgery. Indeed, there may be a long dormant period between the time of contamination and evidence of clinical infection.’ A wide spectrum of organisms has been reported to infect vascular prostheses.It has long been maintained that Staphylococcus aureus is the most common offending pathogen. However, multiple organisms, including gramnegative species, are frequently found.’ More recently, S. epidemtidis is being recognized as a common pathogen, and infections caused by this low virulence organism usually occur late after operation.’ It is well accepted that the introduction of bacteria commonly occurs at the time of implantation of a prosthetic device. Defective barriers, such as breaks in sterile technique, improper handling or sterilization of the graft material, or the development ofwound infections may play a role.’ Infections may also result from bacterial seeding. Of particular importance in the realm of peripheral vasculardiseaseis the presenceof lower extremity ischemic ulcers, which are often infected with polymicrobial flora.’ Any surgical incision involving the groin carries a higher incidence of wound infection and therefore graft infection. This is likely related to contaminated skin as a result of proximity to the perineum, redundant folds of moist skin, superficial location of the graft, and the transection of numerous lymphatics.2~3 The arterial wall itself might rarely be a source of contamination. Ilgenfiitz and Schwartz have respectively demonstrated a 10.4% and 19.6% incidence of positive bacterial cultures of the aortic wall taken at the time of aneurysmectomy. Cultures were more often positive in patients with ruptured or expanding aortic aneurysms.3 Interactions among the local host tissues and the prosthesis provide further insight into the pathogenesis of vasculargraft infections. Somehow local host defensesseem to fail in this environment, thus allowing bacterial propagation. Tissue reactivity in the presence of a foreign body has long been recognized as a potentiator of infection. Dougherty and Simmons* were able to demonstrate that infection could be produced with only 100 colony forming units of S. aureus in the presence of silk suture. The acute inflammatory processwill vary depending on the size and shape of the foreign material, its surface characteristics, and its chemical composition.*Most prosthetic materials commonly used in vascular surgery are fairly inert, but, nonetheless, they do endure some degree of inflammation when exposed to host tissues. The implantation site and mechanical interaction with host tissuesmay also influence the degree of inflammation.’ For example, the inflammation resulting from constant mechanical pulsatile interaction of an aortic graft against the bowel may lead to periprosthetic abscessformation and eventually enteroperivascular or enterovascular fistulas. At the cellular level there is evidence that host defenses are hampered. Many investigators believe that normal
756
Special
Journal of VASCULAR SURGERY
communication
neutrophil function may be diminished in the presence of a foreign materia13” Klock and Bainton’ showed that neutrophils exposed to nylon fibers demonstrated intact chemotaxis but reduced bactericidal ability. Zimmerli et al.* have shown that fluid from Teflon tissue cages implanted subcutaneously had decreased opsonization capacity after 1 hour of experimental infection with S. aweus.’ Neutrophils from even sterile tissue cages showed diminished phagocytic and bactericidal capability when compared to peripheral or exudative neutrophils.“’ These studies render support to the theory that contact between neutrophils and foreign bodies trigger the release of lysosomal enzymes and oxygen free radicals, leading to local tissue damage and metabolic exhaustion of normal neutrophil killing capacity.3 Our own recent in vitro studies suggest that human neutrophils are “primed” in the presence of bioprosthetic materials and have a significant increase in superoxide production when exposed to cell stimulators. Whether or not with time these neutrophils become exhausted and less affective against bacterial invasion remains to be clearly demonstrated. Architectural features of graft materials can play a role in infectivity. Bacteria introduced at the time of implantation may find refuge in the interstices of porous graft surfaces, thereby gaining protection against cellular host defenses.4 Vascular grafts are particularly vulnerable before the completion of pseudointimal formation or in the presence of ulcerations in the pseudointimal lining.4 Bacterial trapping at these denuded sites may act as a harbinger of late prosthetic infections. Factors directly related to the bacterial organisms themselves may assist in infectivity. The production of an extracellular slime substance or glycocalyx by certain bacteria has recently been recognized.4 Much remains to be known about the chemical composition and structure of extracellular slime, but it is felt that bacteria that produce this substance have enhanced ability to adhere to foreign bodies and produce prosthetic related infections4 It is not clear whether the slime is initially produced acting as an adhesive of pathogens to foreign body surfaces or whether the slime is produced secondarily after bacterial adhesion. Factors determining whether a certain strain of bacteria will be a slime producer are also unclear. Christensen et al. ’ ’ have identified two different strains of S. epidwmidisone produces slime (RP62A) and the other does not (RP62A-NA). Other pathogens in addition to S. epidermidis have been found to be slime producers; these include Pseudomonas aeruginosa, Streptococcus vin’dans and S. aureus. 3 There appear to be several mechanisms by which extracellular slime substance functions as a potentiator of infection. Slime may provide a physical barrier against local host defenses and antibiotics. Extracellular slime substance has also been shown to inhibit neutrophil chemotaxis, phagocytosis and oxidative metabolism. Additionally, it has been found to suppress mononuclear cell lymphoproliferative response, the T helper/suppressor cell ratio, natural killer cell cytotoxicity, and immunoglobulin synthesis.3 The slime produced by P. aem&zosa may be
particularly virulent in that it seems to function as an exotoxin with systemic sequelae mimicking sepsis. Mice injected with slime from l? aerzginosa rapidly develop hepatorenal failure and die.3 Evidence exists that certain plasma proteins can influence bacterial adhesion. High molecular weight hydrophobic polymers coated with albumin promote compatibility of the prosthetic material with host tissues. Furthermore, the albumin coat reduces bacterial adhesion to the polymer.3 Other plasma proteins such as collagen and fibronectin may promote cellular adhesion of not only bacteria but other endogenous cells such as endothelial cells.” Our understanding of the factors that promote prosthetic related infections has certainly grown, and with this expanding body of knowledge we can direct our efforts toward the prevention ofgraft infections. The development of inert materials coated with antibiotics and other host compatible substances holds much promise. Further study at the cellular level may provide insight as to the preservation of normal host defense function. The eradication of slime and slime producing pathogens is also a key area to explore. The reduction or even eradication of vascular graft infections is a foreseeable goal, one with far reaching clinical and medical cost related implications. Rekha Cbaudbay, MD Richard L. Simmons, MD University of Pittsburgh School of Medicine Pittsburgh, Pa. REFERENCES
1. Sugarman B, Young EJ.: Infections associated with prosthetic devices: magnitude of the problem. Infect Dis Clin North Am. 1989;2:187-98. 2. Golan JF. Vascular Graft Infections. Infect Dis Clin North Am. 1989;2:247-58. 3. Dougherty SH, Simmons RL. Endogenous factors contributing to prosthetic device infections. Infect Dis Clin North Am. 1989;2:199-209. 4. Ilgenfritz FM, Jordan FT. Microbiologic monitoring of aortic aneurysm wall and contents during aneurysmectomy. Arch Surg 1988;123:504-8. 5. Schwartz JA, Powell TW, Burnham JJ, et al. Culture of abdominal aortic aneurysm contents: an additional series. Arch Surg 1987;122:777-80. 6. James RC, McLeod CJ. Induction of staphylococcal infection in mice with small inocula introduced on sutures. Br J Exp Path01 1961;42:266-77. 7. Howard RJ, Simmons RL. Surgical infectious diseases, 2nd ed, 1988. 8. Klock JC, Bainton DF. Degranulation and abnormal bactericidal function of granulocytes procured by reversible adhesion to nylon wool. Blood 1976;48: 149-61. 9. Zimmerli W, Waldvogel FA, Vaudaux P, Nydegger UE. Pathogenesis of foreign body infection: description and characteristics of an animal model. J Infect Dis 1982;46:487-95. 10. Zimmerli W, Lew I’D, Waldvogel FA. Pathogenesis of foreign body infection: evidence of local granulocyte defect. J Clin Invest 1984;73:1191-200. 11. Christensen GO, Baddour LM, Simpson WA. Phenotypic variation of Staphylococcus epiahnidis slime production in vitro and in vivo. Infect Immun 1987;55:2870-7.
bodies may thereby be more likely to develop a truly biocompatible vascular prosthesis. Ralph S. Greco, MD Untiersityof Medicine and Dentisty of New Jersey Robert Wood Johnson Medical School New Brunswick, NJ.
REFERENCES 1. Elek SD, Cohen PE. The virulenceof staphylococcus pyogenesfor man: a studyof the problemsof wound infection.Br J Exp Path01 1957;38:573. 2. Margiotta MS, Robertson FM, Greco RS. Selective induction of intercellular adhesion molecule (ICAM-I) expression by human endothelial cells following adherence to vascular grafts using an in vitro model. Surg Forum 1980;41:339-41. 3. Whiffen JD, Gott VL. In vivo adsorption of heparin by graphite-benzalkonium intravascular surfaces. Surg Gynecol Obstet 1965;121:287-90. 4. Harvey RA, Greco RS. The non-covalent bonding of antibiotics to a polytetrafluoroethylene graft. Ann Surg 1981;194: 642-7. 5. GrecoRS, Harvey BA, Henry R, et al. Preventionof graft infection by antibiotic bonding. Surg Forum 198O;XXXI:2930. 6. Greco RS, Harvey RA. The role of antibiotic bonding in the prevention of vascular prosthetic infections. Ann Surg 1982; 195:168-71. Greco RS, Harvey RA, Smilow PC, et al. Prevention of vascular prosthetic infection by a benzalkonium-oxacillin bonded polytetrafluoroethylene graft. Surg Gynecol Obstet 1982;155:28-32. Harvey RA, Greco RS. Antibiotic bonding to polytetralluoroethylene with tridodecylmethylammonium chloride. Surgery 1982;92:504-12. Greco RS, Donen. Al’, Harvey BA. The application of antibiotic bonding to the treatment of established vascular prosthetic infection. Arch Surg 1985;120:71-5. 10. Shue WB, Worosilo SC, Donetz Al’, Trooskin SZ, Harvey RA, Greco RS. Prevention of vascular prosthetic infection with an antibiotic bonded Dacron graft. J VASC SURG 1988;8:600-5.
PATHOGENESIS OF VASCULAR GIUFT INFECTIONS The increasingly routine use of prosthetic vascular grafts has revolutionized our ability to reperfuse and salvagepatients with severeperipheral vasculardisease.It is estimated that more than 60,000 vascular graft prostheses are implanted in the United States yearly.’ Despite improved prosthetic materials and surgical techniques, vascular graft infections remain a serious source of clinical morbidity and deaths. Approximately 2% of patients with vascular grafts will develop a prosthetic infection, the consequencesof which are often grave, that is, limb loss or death.’ To limit the incidence of these infections it is of paramount importance that we have a full understanding of the pathogenesis of graft infections. Only in this way can we hope to make significant inroads into the prevention and eradication of vascular prosthetic infections. Graft infections can be separated into those that occur acutely and those that occur late. Early postoperative graft
Special
cummunicatim
755
infections are often the result of wound infections and usually involve grafts anatomically superficial in location (femoral, axillary, popliteal sites). Late graft infections may occur months to yearsafter vascularreconstructive surgery. Indeed, there may be a long dormant period between the time of contamination and evidence of clinical infection.’ A wide spectrum of organisms has been reported to infect vascular prostheses.It has long been maintained that Staphylococcus aureus is the most common offending pathogen. However, multiple organisms, including gramnegative species, are frequently found.’ More recently, S. epidemtidis is being recognized as a common pathogen, and infections caused by this low virulence organism usually occur late after operation.’ It is well accepted that the introduction of bacteria commonly occurs at the time of implantation of a prosthetic device. Defective barriers, such as breaks in sterile technique, improper handling or sterilization of the graft material, or the development ofwound infections may play a role.’ Infections may also result from bacterial seeding. Of particular importance in the realm of peripheral vasculardiseaseis the presenceof lower extremity ischemic ulcers, which are often infected with polymicrobial flora.’ Any surgical incision involving the groin carries a higher incidence of wound infection and therefore graft infection. This is likely related to contaminated skin as a result of proximity to the perineum, redundant folds of moist skin, superficial location of the graft, and the transection of numerous lymphatics.2~3 The arterial wall itself might rarely be a source of contamination. Ilgenfiitz and Schwartz have respectively demonstrated a 10.4% and 19.6% incidence of positive bacterial cultures of the aortic wall taken at the time of aneurysmectomy. Cultures were more often positive in patients with ruptured or expanding aortic aneurysms.3 Interactions among the local host tissues and the prosthesis provide further insight into the pathogenesis of vasculargraft infections. Somehow local host defensesseem to fail in this environment, thus allowing bacterial propagation. Tissue reactivity in the presence of a foreign body has long been recognized as a potentiator of infection. Dougherty and Simmons* were able to demonstrate that infection could be produced with only 100 colony forming units of S. aureus in the presence of silk suture. The acute inflammatory processwill vary depending on the size and shape of the foreign material, its surface characteristics, and its chemical composition.*Most prosthetic materials commonly used in vascular surgery are fairly inert, but, nonetheless, they do endure some degree of inflammation when exposed to host tissues. The implantation site and mechanical interaction with host tissuesmay also influence the degree of inflammation.’ For example, the inflammation resulting from constant mechanical pulsatile interaction of an aortic graft against the bowel may lead to periprosthetic abscessformation and eventually enteroperivascular or enterovascular fistulas. At the cellular level there is evidence that host defenses are hampered. Many investigators believe that normal
756
Special
Journal of VASCULAR SURGERY
communication
neutrophil function may be diminished in the presence of a foreign materia13” Klock and Bainton’ showed that neutrophils exposed to nylon fibers demonstrated intact chemotaxis but reduced bactericidal ability. Zimmerli et al.* have shown that fluid from Teflon tissue cages implanted subcutaneously had decreased opsonization capacity after 1 hour of experimental infection with S. aweus.’ Neutrophils from even sterile tissue cages showed diminished phagocytic and bactericidal capability when compared to peripheral or exudative neutrophils.“’ These studies render support to the theory that contact between neutrophils and foreign bodies trigger the release of lysosomal enzymes and oxygen free radicals, leading to local tissue damage and metabolic exhaustion of normal neutrophil killing capacity.3 Our own recent in vitro studies suggest that human neutrophils are “primed” in the presence of bioprosthetic materials and have a significant increase in superoxide production when exposed to cell stimulators. Whether or not with time these neutrophils become exhausted and less affective against bacterial invasion remains to be clearly demonstrated. Architectural features of graft materials can play a role in infectivity. Bacteria introduced at the time of implantation may find refuge in the interstices of porous graft surfaces, thereby gaining protection against cellular host defenses.4 Vascular grafts are particularly vulnerable before the completion of pseudointimal formation or in the presence of ulcerations in the pseudointimal lining.4 Bacterial trapping at these denuded sites may act as a harbinger of late prosthetic infections. Factors directly related to the bacterial organisms themselves may assist in infectivity. The production of an extracellular slime substance or glycocalyx by certain bacteria has recently been recognized.4 Much remains to be known about the chemical composition and structure of extracellular slime, but it is felt that bacteria that produce this substance have enhanced ability to adhere to foreign bodies and produce prosthetic related infections4 It is not clear whether the slime is initially produced acting as an adhesive of pathogens to foreign body surfaces or whether the slime is produced secondarily after bacterial adhesion. Factors determining whether a certain strain of bacteria will be a slime producer are also unclear. Christensen et al. ’ ’ have identified two different strains of S. epidwmidisone produces slime (RP62A) and the other does not (RP62A-NA). Other pathogens in addition to S. epidermidis have been found to be slime producers; these include Pseudomonas aeruginosa, Streptococcus vin’dans and S. aureus. 3 There appear to be several mechanisms by which extracellular slime substance functions as a potentiator of infection. Slime may provide a physical barrier against local host defenses and antibiotics. Extracellular slime substance has also been shown to inhibit neutrophil chemotaxis, phagocytosis and oxidative metabolism. Additionally, it has been found to suppress mononuclear cell lymphoproliferative response, the T helper/suppressor cell ratio, natural killer cell cytotoxicity, and immunoglobulin synthesis.3 The slime produced by P. aem&zosa may be
particularly virulent in that it seems to function as an exotoxin with systemic sequelae mimicking sepsis. Mice injected with slime from l? aerzginosa rapidly develop hepatorenal failure and die.3 Evidence exists that certain plasma proteins can influence bacterial adhesion. High molecular weight hydrophobic polymers coated with albumin promote compatibility of the prosthetic material with host tissues. Furthermore, the albumin coat reduces bacterial adhesion to the polymer.3 Other plasma proteins such as collagen and fibronectin may promote cellular adhesion of not only bacteria but other endogenous cells such as endothelial cells.” Our understanding of the factors that promote prosthetic related infections has certainly grown, and with this expanding body of knowledge we can direct our efforts toward the prevention ofgraft infections. The development of inert materials coated with antibiotics and other host compatible substances holds much promise. Further study at the cellular level may provide insight as to the preservation of normal host defense function. The eradication of slime and slime producing pathogens is also a key area to explore. The reduction or even eradication of vascular graft infections is a foreseeable goal, one with far reaching clinical and medical cost related implications. Rekha Cbaudbay, MD Richard L. Simmons, MD University of Pittsburgh School of Medicine Pittsburgh, Pa. REFERENCES
1. Sugarman B, Young EJ.: Infections associated with prosthetic devices: magnitude of the problem. Infect Dis Clin North Am. 1989;2:187-98. 2. Golan JF. Vascular Graft Infections. Infect Dis Clin North Am. 1989;2:247-58. 3. Dougherty SH, Simmons RL. Endogenous factors contributing to prosthetic device infections. Infect Dis Clin North Am. 1989;2:199-209. 4. Ilgenfritz FM, Jordan FT. Microbiologic monitoring of aortic aneurysm wall and contents during aneurysmectomy. Arch Surg 1988;123:504-8. 5. Schwartz JA, Powell TW, Burnham JJ, et al. Culture of abdominal aortic aneurysm contents: an additional series. Arch Surg 1987;122:777-80. 6. James RC, McLeod CJ. Induction of staphylococcal infection in mice with small inocula introduced on sutures. Br J Exp Path01 1961;42:266-77. 7. Howard RJ, Simmons RL. Surgical infectious diseases, 2nd ed, 1988. 8. Klock JC, Bainton DF. Degranulation and abnormal bactericidal function of granulocytes procured by reversible adhesion to nylon wool. Blood 1976;48: 149-61. 9. Zimmerli W, Waldvogel FA, Vaudaux P, Nydegger UE. Pathogenesis of foreign body infection: description and characteristics of an animal model. J Infect Dis 1982;46:487-95. 10. Zimmerli W, Lew I’D, Waldvogel FA. Pathogenesis of foreign body infection: evidence of local granulocyte defect. J Clin Invest 1984;73:1191-200. 11. Christensen GO, Baddour LM, Simpson WA. Phenotypic variation of Staphylococcus epiahnidis slime production in vitro and in vivo. Infect Immun 1987;55:2870-7.
