- Email: [email protected]
Graft preserving methods for managing aortofemoral prosthetic graft infection
Eur J Vasc Endovasc Surg 14 (Supplement A), 38-42 (1997)
Graft Preserving Methods for Managing Aortofemoral Prosthetic Graft Infection K. D. Calligaro ~1 and R J, Veith 2
1Section of Vascular Surgery, Pennsylvania Hospital~Thomas Jefferson Medical College, Philadelphia, PA and 2Division of Vascular Surgery, Montefiore Medical Center~Albert Einstein College of Medicine, New York, NY, U.S.A.
Introduction
The traditional approach to aortofemoral graft infection has included total graft excision and extraanatomic bypasses but has been associated with mortality rates of 25-75% and amputation rates of 1025%. 1-5 Recent reports have documented an improved mortality rate using this traditional approach. 6 A more controversial strategy to treat infected aortofemoral prosthetic graft infections includes selective complete and partial graft preservation. When used appropriately, this method may be associated with lower mortality and limb-loss rates. 7-11 Aortofemoral prosthetic graft preservation should only be attempted when strict criteria are fulfilled, and by vascular surgeons who fully understand the essential adjuncts of this management.
cannot tolerate such prolonged, stressful operations. Third, abdominal exploration after previous placement of an aortic graft may result in injury to the bowel or ureter when dense adhesions are encountered. Fourth, most patients require revascularisation procedures if part or all of the graft is excised, especially if the original aortic operation was performed for limb salvage or aneurysmal disease. If the original operation was performed for claudication, many patients will maintain lower limb viability after graft excision when the proximal anastomosis was performed in an end-toside fashion to the aorta. Revascularisation procedures often pose challenging problems, especially the route of a secondary bypass to avoid the infected wounds. The revascularisation procedure itself may be timeconsuming and contribute to the morbidity associated with graft excision.
Advantages of Prosthetic Graft Preservation
Disadvantages of Prosthetic Graft Preservation
When appropriate indications exist, there are several advantages of preserving all or part of an infected aortofemoral graft. First, totally excising an infected aortic graft that is well-incorporated in surrounding scar tissue can be technically challenging and associated with significant blood loss, especially if inadvertent injury to surrounding vascular structures
There are definite risks and disadvantages associated with complete or partial graft preservation. First, anastomotic haemorrhage can occur if a prosthetic graft is left in an infected wound. However, subsequent haemorrhage may also occur if the artery is oversewn or repaired using a vein patch angioplasty. 12 Arterial ligation represents the most secure closure of a femoral artery, and a double-running layer of monofilament, non-absorbable suture is probably the best closure for an aortic stump, especially when buttressed with omentum. 7'8 Second, recurrent infection may develop weeks to years after the onset of the initial infection if prosthetic material remains. 9'12 Third, sepsis can occur if the graft has been seeded with bacteria and is not completely removed.
Occurs.
Second, total excision of an aortofemoral graft is time-consuming, requires lengthy periods of general anaesthesia and poses considerable cardiac and pulmonary risks. These critically ill patients frequently * Please address all correspondence to: K. D. Calligaro, 700 Spruce St., Suite 101, Philadelphia, PA 19106, U.S.A.
1078-5884/97/SA0038+05 $12.00/0 © 1997 W.B. Saunders Company Ltd.
Graft Preservation
However, if graft preservation is attempted only when patients fulfill well-defined criteria, and our subsequent strategy is closely followed, complete or partial graft preservation will be successful long-term in the majority of patients. 7-12
Complete Graft Preservation
Indications Preservation of an entire aortobifemoral prosthetic graft is only recommended when (1) the infection is confined to the groin, (2) the graft is patent, (3) the anastomoses are intact, and (4) the patient is not septic. 7'8 If any of these criteria are not fulfilled, we believe that all or part of the graft must be removed. This strategy, therefore, should only be used if there are no clinical findings of intra-abdominal graft infection including systemic sepsis, abdominal tenderness or an abdominal pulsatile mass. Radiological studies of the intra-abdominal portion of the graft, including a computed tomography (CT) scan, magnetic resonance imaging (MRI) study, or indium-labelled white blood cell scan, must be negative before preservation of the graft is attempted. CT findings of graft infection include fluid or air around the stem or contralateral limb of the graft. These findings may be normal up to 6 weeks following aortic graft implantation. 13 In addition, if pseudomonas is one of the organisms responsible for the graft infection, graft preservation is less likely to be successful. 14'15Pseudomonas produces toxins that can result in arterial disruption and sepsis, and the organism is particularly resistant to antibiotics and polymorphonuclear cells. 14'15We have also documented that complete graft preservation is more likely to be successful for early graft infections, namely those presenting less than 2 months after graft implantation, than for late graft infections. ~6 This finding may be due to actual bacterial ingrowth into the prosthetic graft rather than simple involvement of the surrounding soft tissue, but this theory is not proven.
Strategy Several aspects of graft preservation are critical to achieve a successful outcome. First, appropriate intravenous antibiotics are administered for at least 6 weeks and based on wound culture and sensitivity results. Occasionally we add oral antibiotics for another 6 weeks if the infecting organism is particularly virulent and susceptible to oral agents.
39
Second, repeated and aggressive operative wound debridement of all infected tissue is essential. This radical debridement includes excision of any exudate adhering to the prosthetic graft or anastomosis. The first debridement should be done in the operating room, where adequate lighting and equipment is available in case surgical misadventure with unexpected bleeding occurs. All necrotic and infected tissue must be excised until fresh, bleeding subcutaneous tissue and muscle is evident. Inadequate debridement probably represents the most common cause of failure of attempted complete graft preservation. Third, methods to achieve wound closure after adequate debridement include the following techniques. We use an antibiotic or dilute povidone-iodine solution to moisten gauze dressings packed into the open wound three times a day. Initially these dressing changes should be performed by the surgical housestaff in the intensive care unit. If complete wound healing is attempted by delayed secondary intention, this protocol continues until the graft is covered by granulation tissue. At this point the patient can be transferred to a regular hospital bed and nursing personnel can be allowed to perform the dressing changes. Alternatively, a muscle flap can be placed to cover the graft after all infected tissue has been excised and granulation tissue has formed. During the last 20 years we have used muscle flaps infrequently when complete preservation of aortobifemoral grafts was attempted, and have not found a difference in terms of successful graft preservation compared to secondary intention wound healing. 17However, we recently have been favouring placement of muscle flaps in low-risk patients who can tolerate another major operation. We believe that the gracilis muscle is best for small groin wounds and the rectus abdominis muscle best for large groin wounds.
Subtotal Graft Preservation
Indications Situations exist when complete preservation of an infected aortobifemoral prosthetic graft is not indicated but preservation of most or part of the graft represents a simpler and improved management compared to total graft excision. The advantages of partial graft preservation is that excision of the intra-abdominal segment of the graft with its associated blood loss and other complications may be avoided in appropriate situations. In addition, revascularisation of only one lower extremity may be required if the stem and Eur J VascEndovascSurg Vol 14 SupplementA, December1997
40
K.D. Calligaro and F, J. Veith
uninvolved limb of the graft are left intact. Finally, performance of a secondary bypass to save the threatened limb is often simplified because the remaining part of the uninfected, ipsilateral limb of the graft may be used as an inflow source in appropriate circumstances.
cluded limb of the graft is performed, a proximal cuff of the graft limb is oversewn or ligated, the remaining distal segment of the graft limb is excised, and an axillofemoral bypass may be performed to revascularise the ischaemic leg. This approach should only rarely be considered for patients with an infected, occluded limb of an aortobifemoral graft for fear that the proximal segment of the involved limb of the graft is involved with the infectious process.
Infected, disrupted femoral anastomosis of an aortobifemoral graft A patient who presents with infection in the groin involving an aortobifemoral graft manifested by bleeding or an infected false aneurysm may be treated by preservation of most of the remaining part of the graft in appropriate cases. As previously mentioned, only if further clinical and radiological evaluation reveals that the infection is confined to the distal graft limb in the groin can excision of this infected segment of the graft be carried out and the remaining, uninfected graft preserved. 7-9 Graft incorporation and absence of fluid around the proximal limb of the graft at the time of surgical exploration is confirmatory evidence that the infection has not extended proximally and only involves the distal segment of the graft in the groin. We would not recommend complete graft preservation for patients presenting with a disrupted anastomosis, because they will almost always develop recurrent haemorrhage. ~20versewing a disrupted anastomosis or placing a new prosthetic interposition graft in the infected wound is fraught with hazards, because the arterial wall is involved with the infectious process and predisposed to recurrent rupture.
Infected, occluded limb of an aortobifemoral graft A rare indication for partial preservation of an infected aortobifemoral graft is if a patient presents with a groin infection involving an occluded limb of the graft and the remainder of the graft is patent. Generally if a patient presents with infection involving one occluded limb of an aortobifemoral graft, bacteria have seeded the thrombus in the graft and extend to the intraabdominal stem of the graft. In these instances, the entire graft must be removed. However, if a patient represents a prohibitive medical risk for total graft excision, preserving the stem and contralateral limb of the graft is an option if preoperative and intraoperative findings show that these segments of the graft are free of infection. Through a retroperitoneal approach, thrombectomy of the proximal segment of the ocEur J VascEndovasc Surg Vol 14 Supplement A, December1997
Infection involving entire aortobifemoral graft with intact femoral anastomoses One final consideration for partial preservation of an infected aortobifemoral graft is when almost all of the graft must be excised (i.e. intra-abdominal graft infection) but the femoral anastomoses are intact. In these cases, a 3-4 mm cuff of the original prosthetic graft on the underlying femoral artery can be oversewn with a double running layer of prolene suture and the remaining part of the graft excised. 9 This technique may preserve patency of the common femoral artery and avoid the need to ligate or oversew the artery or place an autologous tissue patch on the artery to prevent stenosis. We have been disappointed by a high rate of vein patch rupture when this has been attempted. 12 Patency of the common femoral artery may be important to provide retrograde flow to the iliac artery or antegrade flow to the deep or superficial femoral artery after a revascularisation procedure is performed. As an example, if excision of an aortobifemoral graft is required, the common iliac arteries are occluded, and the external and internal iliac arteries are patent, maintaining pelvic perfusion is often essential to prevent rectal and buttock ischaemia. Maintaining patency of the common femoral artery is important after an axillary-to-superficial or -deep femoral artery bypass to provide retrograde flow to the external iliac artery and then to the internal iliac artery to prevent pelvic ischaemia. Oversewing the common femoral artery might result in arterial stenosis with eventual occlusion, while ligation of the artery would obviously prevent retrograde flow to the internal iliac artery in this example.
Strategy In the previously defined appropriate clinical settings, the stem and contralateral limb of the graft can be preserved and an extra-anatomic bypass performed to the threatened limb to avoid a major amputation.
Graft Preservation
41
Patients who present with a stable infected pseudo- an aortobifemoral graft and an intact anastomosis, aneurysm or with an occluded, infected limb of the oversewing a small remnant of the graft is acceptable. graft can usually undergo a secondary revascularisation procedure before undergoing graft excision. The groin wound is packed with an antibiotic Total Aortic Graft Excision soaked dressing and covered with an adhesive dressing. The entire operative field is then prepared and the groin is again covered with a sterile adhesive We agree with others that total aortic graft excision is dressing. If clinical, radiological, and intraoperative mandatory when a patient presents with systemic findings show that the proximal limb of the graft is sepsis or infection of an intracavitary portion of the not involved, we prefer to use this segment of the graft. 1~ In addition, patients presenting with bilateral graft as an inflow source approached through a re- infected groin pseudoaneurysms will generally require troperitoneal incision. The graft limb is divided, the total graft excision because this type of presentation distal segment oversewn and surrounding soft tissue strongly suggests that the entire graft is infected. is buttressed over it. A new prosthetic or vein graft is Finally, as previously mentioned, most patients prethen anastomosed end-to-end to the proximal limb of senting with infection involving an occluded limb of the original graft, tunnelled lateral to the involved an aortobifemoral graft require total graft excision groin wound under the inguinal ligament and brought because the proximal limb, and therefore stem, of the to a patent, uninfected outflow artery. We have used graft are involved with the infection. Replacement of an infected aortobifemoral prosthe superficial and deep femoral and popliteal arteries thetic graft with a new prosthetic graft placed in the as suitable distal arteries to maintain limb salvage in infected field has been reported to treat aortic graftthese circumstances. 18'19 enteric fistulas, infected thoracoabdominal aortic grafts If the ipsilateral, proximal limb of the graft is not and when the causative organism is coagulase-nega suitable inflow conduit, other choices include the ative Staphylococcus epidermidis3 ° Other options to treat axillary artery or the iliac artery using an obturator infected aortobifemoral graft infections following total bypass, assuming these vessels do not have inflow graft excision include replacing the infected prosthetic stenoses. We prefer to tunnel an axillary graft subgraft with a composite autologous vein graft or a cutaneously lateral to the infected wound in the groin, cadaver aorta. 21'22 These aspects of management of even if the graft needs to be placed lateral to the aortic graft infections are discussed elsewhere. anterior superior iliac spine. An obturator bypass is preferred if a medial tunnel must be used, which we have rarely found to be the case. The contralateral femoral artery is rarely an option as an inflow source, Acknowledgements because it is difficult to tunnel a cross-over femoral graft to an outflow artery in the threatened limb Supported by grants from the John F. Connelly Foundation, U.S. Health Service (HL 02990-01), the James Hilton Manning and without having the new graft come in contact with Public Emma Austin Manning Foundation, the Anna S. Brown Trust, and the infected wound. If the contralateral femoral artery the New York Institute for Vascular Studies. must be utilised as an inflow vessel, a possible route to avoid the open groin wound is to tunnel the graft across the perineum and approach an outflow artery References medially. We do not have experience with this technique and fear it is prone to infection because of 1 BUNT TJ. Synthetic vascular graft infections. I. Graft infections. difficulty sterilising this area. Surgery 1983; 93: 733-746. After the secondary bypass is completed and these 2 SZILAGYIDE, SMITH RF, ELLIOTTJI~t VRANDECICMP. Infection in arterial reconstruction with synthetic grafts. Ann Surg 1972; 176: incisions are closed and dressed, the infected segment 321-323. of the graft in the groin is removed. By following this 3 YEAGERRA, McCONNELLDB, SASAKITM, VETTORM. Aortic and strategy, ischaemia of the threatened limb is minperipheral prosthetic graft infection: Differential management and causes of mortality. Am J Surg 1985; 150: 36-43. imised. The underlying common femoral artery can 4 LIEKWEG WG t GREENFIELD LJ. Vascular prosthetic infections: then be oversewn or ligated depending on the degree Collected experience and results of treatment. Surgery 1977; 81: of necrosis of the artery. We would not leave an 335--342. 5 LORENTZENJg, NIELSEN OM, ARENDRUPH et al. Vascular graft oversewn patch of prosthetic graft on the femoral infection: an analysis of sixty-two graft infections in 2411 conartery if the anastomosis was disrupted, but if the secutively placed implanted synthetic vascular grafts. Surgery 1985; 98: 81-86. patient presented with an infected, occluded limb of Eur J Vasc Endovasc Surg Vol 14 Supplement A, December 1997
42
K . D . Calligaro and F, J, Veith
6 RJCOTTAJJ, FAGGIOLIGL, STELLAA et al. Total excision and extraanatomic bypass for aortic graft infection. Am J Surg 1991; 162: 145-149. 7 VEITH FJ. Surgery of the infected aortic graft. In: Bergan JJ, Yao JST, eds. Surgery of the Aorta and its Body Branches. New York: Grune and Stratton, 1979: 521-533. 8 CALLIGAROKD, VEITHFJ. Diagnosis and management of infected prosthetic aortic grafts. Surgery 1991; 110: 805-813. 9 CALLIGAROKD, VEITEIFJ, GUPTA SK et al. A modified method for management of prosthetic graft infections involving an anastomosis to the common femoral artery. J Vasc Surg 1990; 11: 485-492. 10 MORRISGE, FRIENDPJ, VASSALLODJ~ FARRINGTONMr LEAPMAN S, QuIcK CRG. Antibiotic irrigation and conservative surgery for major aortic graft infection. J Vasc Surg 1994; 20: 88-95. 11 TomN KD. Aortobifemoral perigraft abscess: treatment by percutaneous catheter drainage. J Vasc Surg 1988; 8: 339-343. 12 SAMSON Rid, VEITH FJ, JANKO GS, GUPTA SK, SCHtER LA. A modified classification and approach to the management of infections involving peripheral arterial prosthetic grafts. J Vasc Surg 1988; 8: 147-153. 13 O'HAKA PJ, BORKOWSKI GP, HERTZER NR, O'DONOVAN PB, BRIGHAMSL, BEVENEG. Natural history of periprosthetic air on computerized axial tomographic examination of the abdomen following abdominal aortic aneurysm repair. J Vasc Surg 1984; 1: 429-433. 14 CALLIGARO KD, VEITH FJ, SCIKWARTZML, SAVARESERP, DELAURENTIS DA. Are gram-negative bacteria a contraindication
Eur J Vasc Endovasc Surg Vol 14 Supplement A, December 1997
to selective preservation of infected prosthetic arerial grafts?
J Vasc Surg 1992; 16: 337-346. 15 GEARY KJ, TOMKIEWICAAM, HARRISON HN et al. Differential effects of a gram-negative and a gram-positive infection on autogenous and prosthetic grafts. J Vasc Surg 1990; 11: 339-347. 16 CALLIGAROKD, VEITH FJ, SCHWARZ M, DOUGHERTYMJ, DELAUI~NTIS DA. Differences between early and late arterial graft infections. J Vasc Surg (in press). 17 CALLIGAROKD, VEITH FJ, SALESCM, SAVARESERP, DOUGHERTY MJ, DELAUI~ENTISDA. A comparison of muscle flaps and delayed secondary intention wound healing in the management of infected lower extremity arterial grafts. Ann Vasc Surg 1994; 8: 31-37. 18 VEITEIF'J,ASCERE, GUI'TASK, WENGERTERKR. Lateral approach to the popliteal artery. J Vasc Surg 1987; 6: 119-123. 19 NUNEZ AA, VEITH FJ, COLLIER P, ASCER E~ WHITE FLO1KESS, GUrTA SK. Direct approaches to the distalportions of the deep femoral artery for limb salvage bypasses. J Vasc Surg 1988; 8: 576-581. 20 BANDYKDF, BERGAMINITM, KINNEY EV, SEABROOKG, TOWNE JB. In situ replacement of vascular prostheses infected by bacterial biofilms. J Vasc Surg 1991; 13: 575-583. 21 CLAGETTGP, BOWER BL, LOPES-VIEGOMA et al. Creation of a neo-aortoiliac system from lower extremity deep and superficial veins. Ann Surg 1993; 218: 239-249. 22 KIEFFER E, BAI.ININI 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: 349-356.
Graft Preserving Methods for Managing Aortofemoral Prosthetic Graft Infection K. D. Calligaro ~1 and R J, Veith 2
1Section of Vascular Surgery, Pennsylvania Hospital~Thomas Jefferson Medical College, Philadelphia, PA and 2Division of Vascular Surgery, Montefiore Medical Center~Albert Einstein College of Medicine, New York, NY, U.S.A.
Introduction
The traditional approach to aortofemoral graft infection has included total graft excision and extraanatomic bypasses but has been associated with mortality rates of 25-75% and amputation rates of 1025%. 1-5 Recent reports have documented an improved mortality rate using this traditional approach. 6 A more controversial strategy to treat infected aortofemoral prosthetic graft infections includes selective complete and partial graft preservation. When used appropriately, this method may be associated with lower mortality and limb-loss rates. 7-11 Aortofemoral prosthetic graft preservation should only be attempted when strict criteria are fulfilled, and by vascular surgeons who fully understand the essential adjuncts of this management.
cannot tolerate such prolonged, stressful operations. Third, abdominal exploration after previous placement of an aortic graft may result in injury to the bowel or ureter when dense adhesions are encountered. Fourth, most patients require revascularisation procedures if part or all of the graft is excised, especially if the original aortic operation was performed for limb salvage or aneurysmal disease. If the original operation was performed for claudication, many patients will maintain lower limb viability after graft excision when the proximal anastomosis was performed in an end-toside fashion to the aorta. Revascularisation procedures often pose challenging problems, especially the route of a secondary bypass to avoid the infected wounds. The revascularisation procedure itself may be timeconsuming and contribute to the morbidity associated with graft excision.
Advantages of Prosthetic Graft Preservation
Disadvantages of Prosthetic Graft Preservation
When appropriate indications exist, there are several advantages of preserving all or part of an infected aortofemoral graft. First, totally excising an infected aortic graft that is well-incorporated in surrounding scar tissue can be technically challenging and associated with significant blood loss, especially if inadvertent injury to surrounding vascular structures
There are definite risks and disadvantages associated with complete or partial graft preservation. First, anastomotic haemorrhage can occur if a prosthetic graft is left in an infected wound. However, subsequent haemorrhage may also occur if the artery is oversewn or repaired using a vein patch angioplasty. 12 Arterial ligation represents the most secure closure of a femoral artery, and a double-running layer of monofilament, non-absorbable suture is probably the best closure for an aortic stump, especially when buttressed with omentum. 7'8 Second, recurrent infection may develop weeks to years after the onset of the initial infection if prosthetic material remains. 9'12 Third, sepsis can occur if the graft has been seeded with bacteria and is not completely removed.
Occurs.
Second, total excision of an aortofemoral graft is time-consuming, requires lengthy periods of general anaesthesia and poses considerable cardiac and pulmonary risks. These critically ill patients frequently * Please address all correspondence to: K. D. Calligaro, 700 Spruce St., Suite 101, Philadelphia, PA 19106, U.S.A.
1078-5884/97/SA0038+05 $12.00/0 © 1997 W.B. Saunders Company Ltd.
Graft Preservation
However, if graft preservation is attempted only when patients fulfill well-defined criteria, and our subsequent strategy is closely followed, complete or partial graft preservation will be successful long-term in the majority of patients. 7-12
Complete Graft Preservation
Indications Preservation of an entire aortobifemoral prosthetic graft is only recommended when (1) the infection is confined to the groin, (2) the graft is patent, (3) the anastomoses are intact, and (4) the patient is not septic. 7'8 If any of these criteria are not fulfilled, we believe that all or part of the graft must be removed. This strategy, therefore, should only be used if there are no clinical findings of intra-abdominal graft infection including systemic sepsis, abdominal tenderness or an abdominal pulsatile mass. Radiological studies of the intra-abdominal portion of the graft, including a computed tomography (CT) scan, magnetic resonance imaging (MRI) study, or indium-labelled white blood cell scan, must be negative before preservation of the graft is attempted. CT findings of graft infection include fluid or air around the stem or contralateral limb of the graft. These findings may be normal up to 6 weeks following aortic graft implantation. 13 In addition, if pseudomonas is one of the organisms responsible for the graft infection, graft preservation is less likely to be successful. 14'15Pseudomonas produces toxins that can result in arterial disruption and sepsis, and the organism is particularly resistant to antibiotics and polymorphonuclear cells. 14'15We have also documented that complete graft preservation is more likely to be successful for early graft infections, namely those presenting less than 2 months after graft implantation, than for late graft infections. ~6 This finding may be due to actual bacterial ingrowth into the prosthetic graft rather than simple involvement of the surrounding soft tissue, but this theory is not proven.
Strategy Several aspects of graft preservation are critical to achieve a successful outcome. First, appropriate intravenous antibiotics are administered for at least 6 weeks and based on wound culture and sensitivity results. Occasionally we add oral antibiotics for another 6 weeks if the infecting organism is particularly virulent and susceptible to oral agents.
39
Second, repeated and aggressive operative wound debridement of all infected tissue is essential. This radical debridement includes excision of any exudate adhering to the prosthetic graft or anastomosis. The first debridement should be done in the operating room, where adequate lighting and equipment is available in case surgical misadventure with unexpected bleeding occurs. All necrotic and infected tissue must be excised until fresh, bleeding subcutaneous tissue and muscle is evident. Inadequate debridement probably represents the most common cause of failure of attempted complete graft preservation. Third, methods to achieve wound closure after adequate debridement include the following techniques. We use an antibiotic or dilute povidone-iodine solution to moisten gauze dressings packed into the open wound three times a day. Initially these dressing changes should be performed by the surgical housestaff in the intensive care unit. If complete wound healing is attempted by delayed secondary intention, this protocol continues until the graft is covered by granulation tissue. At this point the patient can be transferred to a regular hospital bed and nursing personnel can be allowed to perform the dressing changes. Alternatively, a muscle flap can be placed to cover the graft after all infected tissue has been excised and granulation tissue has formed. During the last 20 years we have used muscle flaps infrequently when complete preservation of aortobifemoral grafts was attempted, and have not found a difference in terms of successful graft preservation compared to secondary intention wound healing. 17However, we recently have been favouring placement of muscle flaps in low-risk patients who can tolerate another major operation. We believe that the gracilis muscle is best for small groin wounds and the rectus abdominis muscle best for large groin wounds.
Subtotal Graft Preservation
Indications Situations exist when complete preservation of an infected aortobifemoral prosthetic graft is not indicated but preservation of most or part of the graft represents a simpler and improved management compared to total graft excision. The advantages of partial graft preservation is that excision of the intra-abdominal segment of the graft with its associated blood loss and other complications may be avoided in appropriate situations. In addition, revascularisation of only one lower extremity may be required if the stem and Eur J VascEndovascSurg Vol 14 SupplementA, December1997
40
K.D. Calligaro and F, J. Veith
uninvolved limb of the graft are left intact. Finally, performance of a secondary bypass to save the threatened limb is often simplified because the remaining part of the uninfected, ipsilateral limb of the graft may be used as an inflow source in appropriate circumstances.
cluded limb of the graft is performed, a proximal cuff of the graft limb is oversewn or ligated, the remaining distal segment of the graft limb is excised, and an axillofemoral bypass may be performed to revascularise the ischaemic leg. This approach should only rarely be considered for patients with an infected, occluded limb of an aortobifemoral graft for fear that the proximal segment of the involved limb of the graft is involved with the infectious process.
Infected, disrupted femoral anastomosis of an aortobifemoral graft A patient who presents with infection in the groin involving an aortobifemoral graft manifested by bleeding or an infected false aneurysm may be treated by preservation of most of the remaining part of the graft in appropriate cases. As previously mentioned, only if further clinical and radiological evaluation reveals that the infection is confined to the distal graft limb in the groin can excision of this infected segment of the graft be carried out and the remaining, uninfected graft preserved. 7-9 Graft incorporation and absence of fluid around the proximal limb of the graft at the time of surgical exploration is confirmatory evidence that the infection has not extended proximally and only involves the distal segment of the graft in the groin. We would not recommend complete graft preservation for patients presenting with a disrupted anastomosis, because they will almost always develop recurrent haemorrhage. ~20versewing a disrupted anastomosis or placing a new prosthetic interposition graft in the infected wound is fraught with hazards, because the arterial wall is involved with the infectious process and predisposed to recurrent rupture.
Infected, occluded limb of an aortobifemoral graft A rare indication for partial preservation of an infected aortobifemoral graft is if a patient presents with a groin infection involving an occluded limb of the graft and the remainder of the graft is patent. Generally if a patient presents with infection involving one occluded limb of an aortobifemoral graft, bacteria have seeded the thrombus in the graft and extend to the intraabdominal stem of the graft. In these instances, the entire graft must be removed. However, if a patient represents a prohibitive medical risk for total graft excision, preserving the stem and contralateral limb of the graft is an option if preoperative and intraoperative findings show that these segments of the graft are free of infection. Through a retroperitoneal approach, thrombectomy of the proximal segment of the ocEur J VascEndovasc Surg Vol 14 Supplement A, December1997
Infection involving entire aortobifemoral graft with intact femoral anastomoses One final consideration for partial preservation of an infected aortobifemoral graft is when almost all of the graft must be excised (i.e. intra-abdominal graft infection) but the femoral anastomoses are intact. In these cases, a 3-4 mm cuff of the original prosthetic graft on the underlying femoral artery can be oversewn with a double running layer of prolene suture and the remaining part of the graft excised. 9 This technique may preserve patency of the common femoral artery and avoid the need to ligate or oversew the artery or place an autologous tissue patch on the artery to prevent stenosis. We have been disappointed by a high rate of vein patch rupture when this has been attempted. 12 Patency of the common femoral artery may be important to provide retrograde flow to the iliac artery or antegrade flow to the deep or superficial femoral artery after a revascularisation procedure is performed. As an example, if excision of an aortobifemoral graft is required, the common iliac arteries are occluded, and the external and internal iliac arteries are patent, maintaining pelvic perfusion is often essential to prevent rectal and buttock ischaemia. Maintaining patency of the common femoral artery is important after an axillary-to-superficial or -deep femoral artery bypass to provide retrograde flow to the external iliac artery and then to the internal iliac artery to prevent pelvic ischaemia. Oversewing the common femoral artery might result in arterial stenosis with eventual occlusion, while ligation of the artery would obviously prevent retrograde flow to the internal iliac artery in this example.
Strategy In the previously defined appropriate clinical settings, the stem and contralateral limb of the graft can be preserved and an extra-anatomic bypass performed to the threatened limb to avoid a major amputation.
Graft Preservation
41
Patients who present with a stable infected pseudo- an aortobifemoral graft and an intact anastomosis, aneurysm or with an occluded, infected limb of the oversewing a small remnant of the graft is acceptable. graft can usually undergo a secondary revascularisation procedure before undergoing graft excision. The groin wound is packed with an antibiotic Total Aortic Graft Excision soaked dressing and covered with an adhesive dressing. The entire operative field is then prepared and the groin is again covered with a sterile adhesive We agree with others that total aortic graft excision is dressing. If clinical, radiological, and intraoperative mandatory when a patient presents with systemic findings show that the proximal limb of the graft is sepsis or infection of an intracavitary portion of the not involved, we prefer to use this segment of the graft. 1~ In addition, patients presenting with bilateral graft as an inflow source approached through a re- infected groin pseudoaneurysms will generally require troperitoneal incision. The graft limb is divided, the total graft excision because this type of presentation distal segment oversewn and surrounding soft tissue strongly suggests that the entire graft is infected. is buttressed over it. A new prosthetic or vein graft is Finally, as previously mentioned, most patients prethen anastomosed end-to-end to the proximal limb of senting with infection involving an occluded limb of the original graft, tunnelled lateral to the involved an aortobifemoral graft require total graft excision groin wound under the inguinal ligament and brought because the proximal limb, and therefore stem, of the to a patent, uninfected outflow artery. We have used graft are involved with the infection. Replacement of an infected aortobifemoral prosthe superficial and deep femoral and popliteal arteries thetic graft with a new prosthetic graft placed in the as suitable distal arteries to maintain limb salvage in infected field has been reported to treat aortic graftthese circumstances. 18'19 enteric fistulas, infected thoracoabdominal aortic grafts If the ipsilateral, proximal limb of the graft is not and when the causative organism is coagulase-nega suitable inflow conduit, other choices include the ative Staphylococcus epidermidis3 ° Other options to treat axillary artery or the iliac artery using an obturator infected aortobifemoral graft infections following total bypass, assuming these vessels do not have inflow graft excision include replacing the infected prosthetic stenoses. We prefer to tunnel an axillary graft subgraft with a composite autologous vein graft or a cutaneously lateral to the infected wound in the groin, cadaver aorta. 21'22 These aspects of management of even if the graft needs to be placed lateral to the aortic graft infections are discussed elsewhere. anterior superior iliac spine. An obturator bypass is preferred if a medial tunnel must be used, which we have rarely found to be the case. The contralateral femoral artery is rarely an option as an inflow source, Acknowledgements because it is difficult to tunnel a cross-over femoral graft to an outflow artery in the threatened limb Supported by grants from the John F. Connelly Foundation, U.S. Health Service (HL 02990-01), the James Hilton Manning and without having the new graft come in contact with Public Emma Austin Manning Foundation, the Anna S. Brown Trust, and the infected wound. If the contralateral femoral artery the New York Institute for Vascular Studies. must be utilised as an inflow vessel, a possible route to avoid the open groin wound is to tunnel the graft across the perineum and approach an outflow artery References medially. We do not have experience with this technique and fear it is prone to infection because of 1 BUNT TJ. Synthetic vascular graft infections. I. Graft infections. difficulty sterilising this area. Surgery 1983; 93: 733-746. After the secondary bypass is completed and these 2 SZILAGYIDE, SMITH RF, ELLIOTTJI~t VRANDECICMP. Infection in arterial reconstruction with synthetic grafts. Ann Surg 1972; 176: incisions are closed and dressed, the infected segment 321-323. of the graft in the groin is removed. By following this 3 YEAGERRA, McCONNELLDB, SASAKITM, VETTORM. Aortic and strategy, ischaemia of the threatened limb is minperipheral prosthetic graft infection: Differential management and causes of mortality. Am J Surg 1985; 150: 36-43. imised. The underlying common femoral artery can 4 LIEKWEG WG t GREENFIELD LJ. Vascular prosthetic infections: then be oversewn or ligated depending on the degree Collected experience and results of treatment. Surgery 1977; 81: of necrosis of the artery. We would not leave an 335--342. 5 LORENTZENJg, NIELSEN OM, ARENDRUPH et al. Vascular graft oversewn patch of prosthetic graft on the femoral infection: an analysis of sixty-two graft infections in 2411 conartery if the anastomosis was disrupted, but if the secutively placed implanted synthetic vascular grafts. Surgery 1985; 98: 81-86. patient presented with an infected, occluded limb of Eur J Vasc Endovasc Surg Vol 14 Supplement A, December 1997
42
K . D . Calligaro and F, J, Veith
6 RJCOTTAJJ, FAGGIOLIGL, STELLAA et al. Total excision and extraanatomic bypass for aortic graft infection. Am J Surg 1991; 162: 145-149. 7 VEITH FJ. Surgery of the infected aortic graft. In: Bergan JJ, Yao JST, eds. Surgery of the Aorta and its Body Branches. New York: Grune and Stratton, 1979: 521-533. 8 CALLIGAROKD, VEITHFJ. Diagnosis and management of infected prosthetic aortic grafts. Surgery 1991; 110: 805-813. 9 CALLIGAROKD, VEITEIFJ, GUPTA SK et al. A modified method for management of prosthetic graft infections involving an anastomosis to the common femoral artery. J Vasc Surg 1990; 11: 485-492. 10 MORRISGE, FRIENDPJ, VASSALLODJ~ FARRINGTONMr LEAPMAN S, QuIcK CRG. Antibiotic irrigation and conservative surgery for major aortic graft infection. J Vasc Surg 1994; 20: 88-95. 11 TomN KD. Aortobifemoral perigraft abscess: treatment by percutaneous catheter drainage. J Vasc Surg 1988; 8: 339-343. 12 SAMSON Rid, VEITH FJ, JANKO GS, GUPTA SK, SCHtER LA. A modified classification and approach to the management of infections involving peripheral arterial prosthetic grafts. J Vasc Surg 1988; 8: 147-153. 13 O'HAKA PJ, BORKOWSKI GP, HERTZER NR, O'DONOVAN PB, BRIGHAMSL, BEVENEG. Natural history of periprosthetic air on computerized axial tomographic examination of the abdomen following abdominal aortic aneurysm repair. J Vasc Surg 1984; 1: 429-433. 14 CALLIGARO KD, VEITH FJ, SCIKWARTZML, SAVARESERP, DELAURENTIS DA. Are gram-negative bacteria a contraindication
Eur J Vasc Endovasc Surg Vol 14 Supplement A, December 1997
to selective preservation of infected prosthetic arerial grafts?
J Vasc Surg 1992; 16: 337-346. 15 GEARY KJ, TOMKIEWICAAM, HARRISON HN et al. Differential effects of a gram-negative and a gram-positive infection on autogenous and prosthetic grafts. J Vasc Surg 1990; 11: 339-347. 16 CALLIGAROKD, VEITH FJ, SCHWARZ M, DOUGHERTYMJ, DELAUI~NTIS DA. Differences between early and late arterial graft infections. J Vasc Surg (in press). 17 CALLIGAROKD, VEITH FJ, SALESCM, SAVARESERP, DOUGHERTY MJ, DELAUI~ENTISDA. A comparison of muscle flaps and delayed secondary intention wound healing in the management of infected lower extremity arterial grafts. Ann Vasc Surg 1994; 8: 31-37. 18 VEITEIF'J,ASCERE, GUI'TASK, WENGERTERKR. Lateral approach to the popliteal artery. J Vasc Surg 1987; 6: 119-123. 19 NUNEZ AA, VEITH FJ, COLLIER P, ASCER E~ WHITE FLO1KESS, GUrTA SK. Direct approaches to the distalportions of the deep femoral artery for limb salvage bypasses. J Vasc Surg 1988; 8: 576-581. 20 BANDYKDF, BERGAMINITM, KINNEY EV, SEABROOKG, TOWNE JB. In situ replacement of vascular prostheses infected by bacterial biofilms. J Vasc Surg 1991; 13: 575-583. 21 CLAGETTGP, BOWER BL, LOPES-VIEGOMA et al. Creation of a neo-aortoiliac system from lower extremity deep and superficial veins. Ann Surg 1993; 218: 239-249. 22 KIEFFER E, BAI.ININI 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: 349-356.