Expanded application of in situ replacement for prosthetic graft infection

Purpose: The purpose of this study was to analyze the outcome of an individualized treatment algorithm for prosthetic graft infection, including the application of in situ graft replacement, based on clinical presentation, extent of graft infection, and microbiology. Methods: There was a retrospective review (1991-2000) of 119 patients with 68 aortoiliofemoral or 51 extracavitary (infrainguinal, 19; axillofemoral, 16; femorofemoral, 16) prosthetic graft infections presenting more than 3 months (range, 3-136 months) after implantation/revision. The treatment algorithm consisted of graft excision with or without ex situ bypass grafts for patients presenting with sepsis or graft-enteric erosion, whereas in situ replacement (autogenous vein, rifampin-bonded polyester, polytetrafluoroethylene [PTFE]) was used in patients with less virulent gram-positive graft infection, in particular infections caused by Staphylococcus epidermidis. Outcomes (death, limb loss, recurrent infection) were correlated with treatment type and infecting organism. Results: In situ replacement was used in 52% of aortoiliofemoral (autogenous vein, 10; rifampin-bonded polyester, 6; PTFE, 9) and 80% of extracavitary (autogenous vein, 26; PTFE, 9; rifampin, 6) graft infections. Total graft excision with ex situ bypass was performed in 34 patients, including 21 patients with graft-enteric erosion/fistula, with a 21% operative mortality and 9% amputation rate. In situ graft replacement was used to treat 76 graft infections with a 30-day operative mortality rate of 4% and an amputation rate of 2%. Graft excision alone was performed in nine patients with one 30-day death. Gram-positive cocci were the prevalent infecting organisms of both intracavitary (59% of isolates) and extracavitary (76% of isolates) graft infections. S epidermidis was the infecting organism in 40% of patients, accounting for the expanded application of in situ prosthetic replacement using a rifampin-bonded polyester or PTFE prosthesis. During the mean follow-up interval of 26 months, recurrent graft infection developed in 3% (1 of 34) of patients after conventional treatment, 3% (1 of 36) patients after in situ vein replacement, and 10% (4 of 40) patients after in situ prosthetic graft replacement (P>.05). Failure of in situ replacement procedures was the result of virulent and antibiotic-resistant bacterial strains. Conclusions: In situ replacement was a safe and durable option in most (64%) patients presenting with prosthetic graft infection. In situ replacement with a rifampin-bonded graft was effective for S epidermidis graft infection, but when the entire prosthesis is involved with either a biofilm or invasive perigraft infection, in situ autogenous vein replacement is preferred. Virulent graft infections presenting with sepsis, anastomotic dehiscence, or graft enteric fistula should continue to be treated with total graft excision, and if feasible, staged ex situ bypass graft.