Prophylactic Antibiotics Prior to Bacteremia Decrease Endovascular Graft Infection in Dogs

Abstract

Endovascular placement of vascular stent grafts in the aorta and peripheral vessels has become a prominent tool in the armamentaria of the vascular surgeon. Despite, several reports of stent graft infection, no current guidelines exist regarding the administration of antibiotics prior to episodes of potential bacterial seeding. We sought to clarify the role of prophylactic antibiotics in preventing stent graft infection after the parenteral administration of Staphylococcus aureus (S. aureus) at various intervals following device placement. A stent graft device was constructed from a 4 mm thin-walled polytetrafluoroethylene (PTFE) graft attached to the outside of a balloon expandable 394-Palmaz stent (Johnson and Johnson Interventional Systems, Warren, NJ). It was then inserted into the common iliac artery through an 1 IF peal-away sheath placed in the femoral artery. Sixty grafts were placed into 30 dogs. There were 5 groups of equal number (groups A-E). In group A, six dogs received intravenous injection of 3 cc X 104 CFU (colony forming units), biotype 31375 5. aureus, 1 day after stent graft implantation. An identically treated group B received antibiotic prophylaxis (1 gm cefazolin 30 minutes prior to bacterial challenge). Group C received bacterial injection 7 days after graft implantation with no antibiotic prophylaxis. Group D received bacterial injection 7 days after graft implantation with antibiotic prophylaxis. A control group E received no antibiotics and was not infected. All infected animals were sacrificed 7 days following bacterial challenge and the stent graft complex cultured. One half of the control group was sacrificed at 7 days and the other half at 14 days. The overall stent graft patency was 90%. Four of the six graft occlusions occurred in group A. Eleven of 12 (92%) dogs cultured S. aureus (biotype 31375) from the explanted stent graft complex. Two localized perforations occurred at the site of the infected complex. In group B, C, and D, no explanted graft complex cultured S. aureus. One graft occluded in group C and D. No stent graft in the control (group E) cultured S. aureus. A stent graft infection model can be consistently produced. In the canine model, the stent graft is more susceptible to infection in the early postoperative period and becomes less susceptible to bacterial seeding at one week after implantation. The authors recommend the use of prophylactic antibiotics in the prevention of endovascular graft infections in the early postoperative period during times when bacterial seeding may occur. They postulate that pseudointima formation during graft incorporation into the vessel wall may be responsible for the resistance to infection.