Graft infection after endovascular abdominal aortic aneurysm repair

Introduction: Although the natural history and management of infected open abdominal aortic aneurysm (AAA) repair is well described, only sporadic case reports have described the fate of patients with infected endografts placed in the abdominal aorta. The present study describes a tertiary referral center's experience with infected endovascular aneurysm repairs (EVARs). Methods: The medical records of 1302 open and endovascular aortic procedures were queried from January 2000 to January 2010. The cases were reviewed for prior aortic procedures, prosthetic implants, and etiology of current open procedure. Demographics, operative details, and perioperative courses were documented. Results: Nine patients (1 woman) with a mean age of 71 years had an EVAR that later required an open procedure for explantation and surgical revision for suspected infection. All grafts were explanted through a midline transperitoneal approach, with a mean time to explant of 33 months. The explanted endografts included 4 Zenith (Cook, Bloomington, Ind), 2 Ancure (Endovascular Technologies, Menlo Park, Calif), 2 Excluders (Gore, Flagstaff, Ariz), and 1 AneuRx (Medtronic, Minneapolis, Minn). Eight of the nine original EVARs were performed at other hospitals; 1 patient had EVAR and open explant at the University of Michigan. All patients had preoperative computed tomography scans, except one who was transferred in extremis with a gastrointestinal hemorrhage. Three patients also had a tagged leukocyte scan, and two had magnetic resonance imaging to further reinforce the suspicion of infection before explantation and bypass planning. Rifampin-soaked Hemashield (Boston Scientific) in situ grafts were used in four patients, with extra-anatomic (axillary-bifemoral) bypass used in the other five. The in situ group had no positive preoperative or postoperative cultures, with the exception of the unstable patient who died the day of surgery. For the other five patients, positive tissue cultures were found for Bacteroides, Escherichia coli, coagulase-negative Staphylococcus, Streptococcus, and Candida. Three patients were found to have aortic-enteric fistula, two of whom died before discharge from the hospital. The remaining seven survived to discharge. Average length of stay was 22 days, with a median follow-up of 11 months. Conclusion: This series of infected EVARs is the largest group of infected AAA endografts reported to date. Because EVAR of AAAs is presently the most common method of repair, development of endograft infection, while rare, can be managed with acceptable mortality rates. Patients presenting with aortic-enteric fistula after EVAR appear to have a more virulent course.