Replacement, refinement, and reduction: necessity of standardization and computational models for long bone fracture repair in animals.

In the field of fracture healing it is essential to know the impacts of new materials. Fracture healing of long bones is studied in various animal models and extrapolated for use in humans, although there are differences between the micro- and macrostructure of human versus animal bone. Unfortunately, recommended standardized models for fracture repair studies do not exist. Many different study designs with various animal models are used. Concerning the general principles of replacement, refinement and reduction in animal experiments (three "Rs"), a standardization would be desirable to facilitate better comparisons between different studies. In addition, standardized methods allow better prediction of bone healing properties and implant requirements with computational models. In this review, the principles of bone fracture healing and differences between osteotomy and artificial fracture models as well as influences of fixation devices are summarized. Fundamental considerations regarding animal model choice are discussed, as it is very important to know the limitations of the chosen model. In addition, a compendium of common animal models is assembled with special focus on rats, rabbits, and sheep as most common fracture models. Fracture healing simulation is a basic tool in reducing the number of experimental animals, so its progress is also presented here. In particular, simulation of different animal models is presented. In conclusion, a standardized fracture model is of utmost importance for the best adaption of simulation to experimental setups and comparison between different studies. One of the basic goals should be to reach a consensus for standardized fracture models.