Systemic and local antibiotic prophylaxis in the prevention of prosthetic vascular graft infection: an experimental study.

AIM: to determine if local, in addition to systemic antibiotic prophylaxis (compared to that provided by systemic prophylaxis alone) provides additional benefit in terms of reducing graft infection. METHODS: gelatin-sealed Dacron grafts were interposed in the infrarenal aorta of 36 mongrels and inoculated with 1 ml of a S. aureus suspension. Group 1 (control group) received no prophylaxis and were inoculated with 1 ml containing 10(9)cfu/ml. Group 2 (n=6) received systemic prophylaxis (1 g cephamandole) and were inoculated with 10(5) cfu/ml (n=3) or 10(7) cfu/ml (n=3). Group 3 received systemic prophylaxis (1 g cephamandole) and were inoculated with 109 cfu/ml. Group 4 received systemic prophylaxis (2 g cephamandole) and were inoculated with 10(9)cfu/ml. In group 5 and 6 grafts were soaked in a rifampicin solution before use and inoculated with 10(9) cfu/ml. Group 5 received no systemic prophylaxis and group 6 received systemic prophylaxis (1 g cephamandole). Grafts were harvested at 2 weeks, and peritonitis, perigraft abscess, anastomotic disruption and graft occlusion recorded. Swabs were taken of the graft, the perigraft tissues and the peritoneal fluid. Graft segments were incubated in broth medium. RESULTS: inoculation with 10(9) cfu/ml ensured graft infection. Systemic or local prophylaxis alone failed to prevent graft infection. Only systemic and local antibiotic prophylaxis provided significant better results than no prophylaxis at all (p<0.01) and local prophylaxis alone (p<0.05). However, total "graft sterility" was not achieved as bacteriologic analysis of the graft segments showed low bacterial counts (<10 bacteria/graft) in 5 of 6 grafts. CONCLUSION: local and systemic prophylaxis provided more protection as demonstrated by the significant decrease in the incidence of "overt" graft infection. Total "graft sterility" cannot be expected in the case of an overwhelming bacterial challenge.