Supplemental Methods and Results – Rat open fracture model

Materials & Methods

The use of Brown Norway rats was conducted in accordance with a pre-approved IACUC protocol at the University of Alabama Medical School. Pain was minimized by providing 1 mg/mL acetaminophen continuously in the drinking water beginning three days prior to surgery and continuing until the end of the study. The procedure used was a modification of the Bonnarens and Einhorn drop weight apparatus(1).

On the modified drop weight apparatus, a 500-g weight is lifted to 25 cm and dropped to create a diaphysealcomminuted femoral fracture. Following injury, constant compression was applied at a pressure of 20 psi for 5 minutes. The fracture was opened through a lateral thigh incision, and an 8-mm segment of the comminuted bone was then removed. The periosteum on the two bone ends was then cauterized circumferentially for a distance of 2 mm on either side of the fracture. The fracture was then stabilized with an intramedullary fully threaded 1.6-mm diameter K-wire. Care was taken to maintain the normal femoral length during this step by manually preserving the fracture gap.

The animals were infected with 1.0 x 108 or 1.0 x 109 CFU of each A. baumannii strain. The wound was closed in a layered fashion. Post-operatively all animals were permitted to be full weight-bearing and fed a standard maintenance diet. Buprenorphine (0.01 mg/kg) was administered subcutaneously for analgesia intra-operatively and every 8 hours for five days post-operatively. Any animals showing severe clinical symptoms of sepsis (porphyrin staining, agonal breathing, and/or lethargy) were sacrificed. All surviving animals were sacrificed on post-op Day 28. The femurs were harvested and dissected free of soft tissue, K-wires were removed, and manual testing of fracture stability was performed on all specimens by grasping the femur and manually applying a bending force.

In order to confirm osteomyelitis, femurs were snap frozen with dry ice and pulverized under sterile conditions. The resulting bone powder was suspended in 20 mL of 0.9% NaCl, then 5 mL of the suspension was serially diluted at a ratio of 1:10 and 100 μL of each dilution was streaked onto phenylelthyl alcohol agar or MacConkey’s agar plates. The plates were incubated at 37°C for 24 hours, and the organisms were identified and counted. The data were presented as the mean value of three plates.

Results

AB5075 displayed a high degree of lethality in other animal models, and because this strain was originally isolated from an infected tibia, we hypothesized that AB5075 could be successfully employed in a rat open-fracture model of infection, to mimic the high energy and crush trauma experienced by wounded military personnel.

The model relies on a dropped weight to induce a fracture crush injury, and a Kirschner wire (K-wire) to link the two resected bone fragments together before bacteria are inoculated into the medullary canal at the fracture site. Each Brown Norway rat received an inoculum of 1.0 x 108 or 1.0 x 109 CFU of AB5075 via pipette delivery. The rats were monitored daily for clinical signs of infection over the course of 28 days. On Day 28 the rats were euthanized and bone tissue was ground and plated to enumerate CFU/g. For rats infected with 1.0 x 108 CFU, the median log CFU/g was 4.47. Rat femurs inoculated with 1.0 x 109 CFU displayed median log CFU/g of 4.37 AB5075-infected femurs. Additionally, Day 28 radiographs revealed typical signs of osteomyelitis in the infected femurs with either periosteal reaction or lytic bone changes or both (Fig. S2). Loss of gap height, a result of bone resorption and K-wire implant failure due to the infection, was also observed in the AB5075-infected rats (Fig. S2).

Reference

1.Stewart RL, Cox JT, Volgas D, Stannard J, Duffy L, Waites KB, Chu TM. 2010. The use of a biodegradable, load-bearing scaffold as a carrier for antibiotics in an infected open fracture model. Journal of orthopaedic trauma 24:587-591.