The aim of this study was to characterize a new model of chronic osteomyelitis with clinically relevant features. A segmental defect of critical size was surgically created in the rat femur, stabilized with a polyacetyl plate and Kirschner wires, and contaminated with bacteria. The animals were allowed to recover while the contamination progressed to a chronic infection. At a later point in time, the defect was surgically debrided without removing the implant. Further treatments of interest, such as antibiotic therapy or application of an osteogenic agent, could be introduced at this time. To implement this model, an initial experiment was performed to determine the bacterial inoculum and time from contamination that would reliably result in an infected defect without causing excessive bone damage by the time debridement surgery was performed. The number of recovered bacteria, degree of radiographic bony lysis, and torsional stiffness of the defect fixation were measured in 192 rats as a function of 4 inocula of Staphylococcus aureus (lo', lo4, 10' or lo6 CFUs) and 4 times from contamination (1,2, 3 or 4 weeks). A lo4 CFU inoculum over 2 weeks was found to consistently create an infection without severe lysis and loss of fixation stability. Based on these values, a second experiment was performed in 96 rats to characterize the debrided defect over time (2,4,8 and 12 weeks after debridement), with and without 4 weeks of the antibiotic ceftriaxone, in terms of the same outcome variables. Infection was persistent in all animals in spite of debridement and antibiotic therapy. Antibiotic therapy did not reduce the degree of bony lysis. Compared with animals not given antibiotic, bacterial counts significantly decreased during the period of antibiotic therapy, but then rebounded to significantly higher levels at 12 weeks. This model allows us to perform further studies on differing regimens of antibiotic therapy and their relationship to surgical dtbridement, and on the efficacy of osteogenic agents in the presence of infection.

In a model simulating a comminuted fracture (gap), this study found that screws should be placed as close to the fracture site as possible to minimize plate strain. In an anatomically reduced two-part fracture model (no gap), widely spaced screws or those placed away from the fracture resulted in lower strains.

Although these data reveal that stacked VCP create a superior construct with respect to cyclic fatigue, surgeons must decide whether this is a clinical advantage on a case-by-case basis.