Abstract
Objective. To analyze current experimental models of fracture union failure and bone nonunion formation, assess their pathogenetic validity, advantages, limitations, and potential applicability in preclinical studies. Methods. A scoping literature review was conducted, including a systematic search for publications devoted to experimental models of fracture nonunion and nonunion. The literature search was conducted in PubMed, Scopus, and Web of Science using the following keywords and their combinations: fracture nonunion, delayed healing, animal model, and bone healing. The review included original experimental studies on laboratory animals that described the method for modeling fracture nonunion, the animal species, the mechanism of nonunion formation, and the methods for evaluating the results. Clinical studies, abstracts, reports, and publications without experimental modeling were excluded from the analysis. The selected studies were analyzed based on the model type, pathogenetic mechanism, replication features, and limitations of use. Results.The analysis revealed significant heterogeneity in existing experimental models. Mechanical models based on fixation instability predominantly reproduce hypertrophic nonunion variants and are characterized by high reproducibility. Critical bone defect models are widely used to evaluate biomaterials and regenerative technologies; however, they do not always reflect the pathogenesis of clinical nonunion. Biological models that include periosteal damage, impaired blood supply, or suppression of the osteogenic potential of tissues are more capable of reproducing atrophic forms of consolidation failure. Combined models that integrate mechanical and biological factors demonstrate a higher pathogenetic similarity to clinical nonunion variants but are characterized by greater technical complexity. Systemic and genetic models allow for the study of the influence of metabolic disorders, aging, and molecular mechanisms on reparative osteogenesis. Differences in animal species, observation periods, and nonunion confirmation criteria have been shown to limit the comparability of results between studies. Conclusions. A universal experimental model capable of fully reproducing all clinical variants of fracture healing failure currently does not exist. The choice of model should be determined by the study objective and the proposed mechanism of bone healing impairment. A promising direction is the standardization of existing approaches and the development of combined models that simultaneously consider mechanical, biological, and systemic factors. This could improve the reproducibility of experimental studies and facilitate the translation of preclinical data into clinical practice.