Purpose: The purpose of this study was to treat an established prosthetic vascular graft infection by in situ replacement with a rifampin-bonded gelatin-sealed Dacron graft in an animal model.

Methods: The infrarenal aorta of 18 dogs was replaced with a gelatin-sealed graft contaminated in vitro by soaking it in a solution with Staphylococcus epidermidis. One week later, animals were randomized into three groups. In group I (control, (n = 6), the dogs did not undergo repeat operations. The dogs in groups II and III underwent repeat operation. In these animals the infected grafts were removed for bacteriologic analysis and replaced in situ with one of two types of grafts: group II (n = 6) received an untreated, gelatin-sealed graft; group III (n = 6) received a rifampin-bonded, gelatin-sealed graft. Antibiotic bonding was obtained by soaking grafts for 15 minutes in a 60 mg/ml saline solution of rifampin at 37 degrees C. All 18 dogs received no systemic adjunct antibiotic therapy. Control grafts and replacement grafts were removed 4 weeks after the initial implantation for bacteriologic analysis. When harvested, all the grafts were cut into two fragments, and quantitative bacterial cultures were obtained from all the fragments. Results were expressed as colony-forming units (CFU)/cm2 of graft material.

Results: All 18 initially implanted grafts and all the untreated replacement grafts were grossly infected at the time of removal, whereas all the rifampin-bonded replacement grafts had normal incorporation. None of the rifampin-bonded grafts grew bacteria, whereas all the initially implanted and all the untreated replacement grafts were infected (p < 0.01). Bacterial counts from the infected fragments were similar in control grafts (2.6 +/- 1.9 x 10(6) CFU/cm2), in initially implanted grafts of groups II (9 +/- 1.1 x 10(5) CFU/cm2) and III (1.3 +/- 1.5 x 10(6) CFU/cm2), and in untreated replacement grafts of group II (1.7 +/- 2.5 x 10(6) CFU/cm2). Blood culture results and culture results of liver, spleen, kidney, and lung specimens at the time of sacrifice were negative.

Conclusion: This study demonstrates that rifampin-bonded gelatin-sealed Dacron grafts are resistant to infection when used for in situ replacement of an infected graft in the dog.