Background Non-union is a well-recognised complication after open fractures. Fifty percent of open extremity trauma returning from military operations developed non-union. Aetiology of non-union is multi- factorial, with infection reported as major contributory factor. The aim of this study was to develop an in-vivo model of infected fracture non-union managed with external fixation which allowed real time in-vivo monitoring of infection to evaluate potential therapeutic strategies.

Method Ten male Wistar rats underwent application of external fixator and midshaft tibia osteotomy. Osteotomy sites were inoculated with bioluminescent Staphylococcus aureus Xen36 (infected group; n=6); or phosphate buffer solution (control group; n=4). Animals were monitored for infection with invivo bioluminescent imaging and fracture healing with plain radiographs. Animals were sacrificed at eight weeks. Post-mortem micro-computed tomography (uCT) was used to assess fracture union; in-vivo bioluminescent imaging to assess persistence of Xen36 infection; tissue samples were processed for bacterial colony forming unit counts and histology to assess for fracture healing and infection.

Results Eight animals reached experiment endpoint (infected=5, control=3). All five infected animals demonstrated radiographic non-union on x-ray and uCT. Bioluminescence, at fracture site in infected cohort, peaked at week two and reduced to chronic baseline of 105 photons per second for duration of experiment. At experiment endpoint bioluminescence was confirmed at fracture site and bioluminescent bacteria was cultured from fracture site tissue samples in all of the infected cohort. Two of three control animals demonstrated radiographic non-union, none luminesced, one grew bacteria from tissue samples but was not bioluminescent.

Conclusions This study has developed an infected fracture non-union animal model. Use of bioluminescent bacteria allows for non-invasive and real-time monitoring of infection. This model is more representative of the military casualty than previously reported models and could be used to evaluate therapeutic strategies for prevention and management of infected fracture non-union.