Objectives We evaluated early and late results of allograft replacement to treat infrarenal aortic graft infection in a large number of patients and compared the results in patients who received fresh allografts versus patients who received cryopreserved allografts. Methods From 1988 to 2002 we operated on 179 consecutive patients (mean age, 64.6 ± 9.0 years; 88.8% men). One hundred twenty-five patients (69.8%) had primary graft infections, and 54 patients (30.2%) had secondary aortoenteric fistulas (AEFs). Fresh allografts were used in 111 patients (62.0%) until 1996, and cryopreserved allografts were used in 68 patients (38.0%) thereafter. Results Early postoperative mortality was 20.1% (36 patients), including four (2.2%) allograft-related deaths from rupture of the allograft (recurrent AEF, n = 3), all in patients with fresh allografts. Thirty-two deaths were not allograft related. Significant risk factors for early mortality were septic shock ( P < .001), presence of AEF ( P = .04), emergency operation ( P = .003), emergency allograft replacement ( P = .0075), surgical complication ( P = .003) or medical complication ( P < .0001), and need for repeat operation ( P = .04). There were five (2.8%) nonlethal allograft complications (rupture, n = 2; thromboses, which were successfully treated at repeat operation, n = 2; and amputation, n = 1), all in patients with fresh allografts. Four patients (2.2%) were lost to follow-up. Mean follow-up was 46.0 ± 42.1 months (range, 1-148 months). Late mortality was 25.9% (37 patients). There were three (2.1%) allograft-related late deaths from rupture of the allograft, at 9, 10, and 27 months, respectively, all in patients with fresh allografts. Actuarial survival was 73.2% ± 6.8% at 1 year, 55.0% ± 8.8% at 5 years, and 49.4% ± 9.6% at 7 years. Late nonlethal aortic events occurred in 10 patients (7.2%; occlusion, n = 4; dilatation < 4 cm, n = 5; aneurysm, n = 1), at a mean of 28.3 ± 28.2 months, all but two in patients with fresh allografts. The only significant risk factor for late aortic events was use of an allograft obtained from the descending thoracic aorta ( P = .03). Actuarial freedom from late aortic events was 96.6% ± 3.4% at 1 year, 89.3% ± 6.6% at 3 years, and 89.3% ± 6.6% at 5 years. There were 63 late, mostly occlusive, iliofemoral events, which occurred at a mean of 34.9 ± 33.7 months in 38 patients (26.6%), 28 of whom (73.7%) had received fresh allografts. The only significant risk factor for late iliofemoral events was use of fresh allografts versus cryopreserved allografts ( P = .03). Actuarial freedom from late iliofemoral events was 84.6% ± 7.0% at 1 year, 72.5% ± 9.0% at 3 years, and 66.4% ± 10.2% at 5 years. Conclusions Early and long-term results of allograft replacement are at least similar to those of other methods to manage infrarenal aortic graft infections. Rare specific complications include early or late allograft rupture and late aortic dilatation. The more frequent late iliofemoral complications may be easily managed through the groin. These complications are significantly reduced by using cryopreserved allografts rather than fresh allografts and by not using allografts obtained from the descending thoracic aorta.
Read full abstract