Guideline for Preclinical Studies of Bone Infections in Large Animals Based on a Systematic Review of 316 Non-Rodent Models

Background: In recent years, animal models of bone infections have been used with increased frequency in order to evaluate novel diagnostic and anti-infective technologies, like antibacterial coating of bone implants or local antibiotic carrier products. Therefore, it is highly relevant to evaluate the scientific quality of existing bone infection models. Methods: We conducted a systematic review of 316 studies of large non-rodent animal models of bone infection (254 rabbit, 16 pig, 23 dog, 11 goat, and 12 sheep) and extracted data on study design, methodological quality, and postmortem evaluation of infection with respect to reporting and quantification of pathology and microbiology. Results: The review demonstrated a substantial lack of study-design information, which hampers reproducibility and continuation of the established work. Furthermore, the methodological study quality was found to be low, as the definition of infection, randomization, power analysis, and blinding were only seldomly reported. The use of histology increased in recent years, but a semi-quantitative scoring of the lesions was often missing, i.e. no objective quantification of outcome. Most of the studies focused on whether the inoculated bacteria were present within the bone tissue post mortem or not. However, very often the bacterial burden was not quantified. In many of the models, different antimicrobial interventions were examined and, although antimicrobial effects were commonly described, a lack of complete sterile outcome was observed in many models. On the basis of the systematic review, we established a study template providing a guideline for the standard reporting of animal models of bone infections, including details related to the animal, pathogen, infected animal, and postmortem analysis that are of crucial importance for validation of results and reproducibility. Conclusions: As the aim of many bone infection models is to examine the effect of an intervention, the guideline emphasizes the importance of objective quantification of outcome, e.g., blinded quantitative scoring of histological findings and quantification of bacterial burden within tissue and on inserted implants. Less than 5% of the analyzed studies adhered completely to the ideal form presented in the study template.