Development of an atrophic nonunion model and comparison to a closed healing fracture in rat femur.

Although most fractures heal, some fail to heal and become nonunions. Many animal models have been developed to study problems of fracture healing. The majority of nonunion models have involved segmental bone defects, but this may not adequately represent the biologic condition in which nonunions clinically develop. The objective of the present study is to develop a nonunion model that better simulates the clinical situation in which there is soft tissue damage including periosteal disruption and to compare this model to a standard closed fracture model utilizing identical fracture stabilization, providing a similar mechanical environment. A total of 96 three month old Long Evans rats were utilized. A 1.25 mm diameter K-wire was inserted into the femur in a retrograde fashion, and a mid-diaphyseal closed transverse fracture was created using a standard three-point bending device. To create a nonunion, 48 of the rats received additional surgery to the fractured femur. The fracture site was exposed and 2 mm of the periosteum was cauterized on each side of the fracture. Fracture healing was evaluated with serial radiographs every two weeks. Animals were maintained for intervals of two, four, six or eight weeks after surgery. Specimens from each time interval were subjected to biomechanical and histological evaluation. None of the cauterized fractures healed throughout the eight weeks experimental duration. The radiographical appearance of nonunion models was atrophic. This investigation showed pronounced differences between the experimental nonunions and standard closed fractures both histologically and biomechanically. In conclusion, we have developed a reproducible atrophic nonunion model in the rat femur that simulates the clinical condition in which there is periosteal disruption but no bone defect.