Vascular Reconstruction with Cryopreserved Human Allografts in the Setting of Infection: a Single-center Experience with Mid-term Follow Up

Brian G. Peterson, MD, William H. Pearce, MD, Mark K. Eskandari, MD, Melina R. Kibbe, MD, Diana Eastridge, RN and Mark D. Morasch, MD
Northwestern University, Division of Vascular Surgery, Chicago, IL

Objective: Vascular reconstruction in the setting of primary arterial or prosthetic graft infection remains a challenging clinical problem. While replacement or bypass with autogenous tissue is considered optimal, this may not be possible in all patients. Cryopreserved allografts may serve as acceptable alternatives when reconstruction with autogenous tissue or extra-anatomic/in situ prosthetic is not practical.
Methods: From March 1999 to March 2006, 38 cadaveric allografts (14 aortoiliac, 1 femoral vein, and 23 superficial femoral artery segments) were used in 34 patients (mean age 65, range 33-81; 20 male, 14 female) with abdominal or lower limb infections. Indications for treatment included infected aortic graft (n=10), infected femoral artery pseudoaneurysm (n=7), mycotic aortic aneurysm (n=5), infected aortic stent-graft (n=3), infected bare metal stent (n=4), aortoenteric fistula (n=2), and intra-abdominal abscess or wound infection (n=7). Wide local debridement of infected tissue was followed by allograft interposition (n=22), bypass (n=5), or extra-anatomic bypass (n=11). Annual outpatient follow-up was routine, and CT scans were obtained annually in patients who had aortic reconstruction. Mean follow-up was 9 months (0-47 months).
Results: The 30-day mortality rate was 6%. One patient died intraoperatively during removal of an infected aortic endograft, and another died from a perioperative MI following removal of an infected iliac stent and femoral-femoral bypass. The organism responsible for infection was isolated in 29 cases (76%), and the pathogens included staph species (48%), mixed organisms (28%), E. coli, Serratia, enterococcus, enterobacter, candida, and propionibacterium (3.5% each). Three patients required re-exploration for hemorrhage related to the allograft. In addition, there has been one amputation, one aortoenteric fistula, and 2 graft thromboses, all seen at 2 months follow-up. The remaining grafts were patent, without evidence of aneurysmal change, thrombosis, or recurrent infection. There have been 6 late deaths (18%) unrelated to surgical intervention seen on average at 11 months follow-up (range 3-39). Average cost of the homograft was $5,450 (range $2,550 - 11,100).
Conclusion: This single-center experience suggests that in the setting of infection, cryopreserved human allografts can be used for vascular reconstruction. In mid-term follow-up, allografts appear to be resistant to subsequent re-infection, thrombosis, or aneurysmal dilatation. In patients without available autogenous conduit, especially when expedient reconstruction is required, homografts may serve as a viable alternative to the use of alternative autogenous tissue or in situ prosthetic replacement.