Bisphosphonate-Conjugated Sitafloxacin for Treatment of Staphylococcus aureus Infection Associated with Cortical Bone Screws: Case Series in Sheep Model.

Background/Objectives: Hydroxybisphosphonate-conjugated sitafloxacin (HBCS) was developed to achieve higher antibiotic concentrations within infected bone. Small animal studies supported further development, but the feasibility of HBCS treatment in a more clinically relevant and larger animal model is unknown. Methods: In this study, we present case reports on four sheep, each receiving four MRSA-contaminated tibial screws treated with different regimens of intravenous antibiotics. The first two sheep received two screws contaminated with 103 CFU and two screws contaminated with 105 CFU. Sheep 1 only received vancomycin, starting on day two. Sheep 2 received vancomycin, starting on day 2, but also received 7 doses of HBCS (2 mg/kg/48 h). The protocol for the final two sheep was revised, and both received four screws contaminated with 103 CFU, and vancomycin was started preoperatively. Sheep 3 and 4 received 7 doses (starting on day 6) and 9 doses (starting on day 2) of HBCS (4 mg/kg/48 h), respectively. Bacteriology was performed on three screws per animal. Longitudinal radiography and histology (n = 1 screw) were assessed for signs of osteolysis and reactive bone formation. Electron microscopy (EM) was performed in the first two sheep to evaluate antibiotic-induced bacterial damage. Results: All sheep tolerated HBCS infusion without clinical signs of discomfort. In addition to a high bacterial load (~104 CFU on all screws), Sheep 1 displayed extensive radiographic and histologic evidence of peri-implant osteolysis and reactive bone formation. Despite having a high bacterial load (~104 CFU on all screws), Sheep 2 displayed only mild radiographic and histologic evidence of peri-implant osteolysis and periosteal reactive bone formation. Bacteriology in Sheep 3 and 4 demonstrated near MRSA eradication (<100 CFU on 2 screws). Both sheep displayed no evidence of osteolysis or new bone formation adjacent to the screw head. EM confirmed the presence of bacteria resorbing bone and replicating in biofilm in Sheep 1, while antibiotic-killed bacteria with ruptured septal planes were seen in Sheep 2. Conclusions: This study demonstrates the feasibility of HBCS therapy in a clinically relevant animal model and provides guidance on future efficacy studies, such as the use of an inoculum of 103 CFU per screw, the initiation of antibiotic treatment commencing at the time of surgery, and the usability of antibiotic-killed bacteria within altered glycocalyx observed by TEM as a potential biomarker for HBCS efficacy.