A mouse model of non-union in a femur critical-size defect using locking plate fixation

Abstract Background Mouse fracture models are powerful tools for investigating in vivo bone healing in response to specific genetic alterations or cellular and protein-based osteogenic therapies. However, there are a paucity of murine models that accurately and humanely recapitulate human fracture fixation, with the modelling of non-union posing a particularly difficult challenge. The objective of this study was to develop a translationally relevant non-union fracture model in mice. Methods Twenty-six mice received a unilateral 1.75 mm femoral bony defect, while four control animals underwent a 0.25 mm osteotomy. Internal fixation of bony defects was performed using titanium locking plates and screws, and mice underwent weekly radiographic assessments postoperatively. Of the 1.75 mm osteotomy group, two animals were sacrificed at 5 days, four at 2 weeks, seven at 6 weeks, and thirteen at 12 weeks postoperatively. Histologic analysis was performed with Safranin O staining and immunohistochemistry for collagen types I and II. Nonunion was evaluated in the 6 and 12 week animals only ( n = 20) to allow sufficient time for potential bony healing. Results All control animals with 0.25 mm osteotomy demonstrated bony union of the femur by 2 weeks. In contrast, among the experimental group with 1.75 mm defects, 65% (13 of 20) of animals overall exhibited non-union. When stratified by postoperative healing time, 71% (5 of 7) of animals followed for 6 weeks showed non-union of the femur, whereas 62% (8 of 13) of the animals followed for 12 weeks showed non-union. Conclusions This paper describes a translationally-relevant mouse model of long bone non-union that can be used to recapitulate the key clinical challenges and biology of human fractures. This valuable preclinical investigation technique can subsequently be used to study the effects of genome alterations as well as novel osteogenic scaffolds, drugs, and biologic therapies.