OP0244 A PRECLINICAL TESTING TOOL: THE IN VITRO 3D FRACTURE GAP MODEL

Background: Approximately 10% of fractures lead to significant fracture healing disorders, with a tendency to further increase due to the aging population. Of note, especially immunosuppressed patients with ongoing inflammation show difficulties in the correct course of fracture healing leading to fracture healing disorders. Most notably, invading immune cells and secreted cytokines are considered to provide an inflammatory microenvironment within the fracture gap, primarily during the initial phase of fracture healing. Current research has the focus on small animal models, facing the problem of translation towards the human system. In order to improve the therapy of fracture healing disorders, we have developed a human cell-based in vitro model to mimic the initial phase of fracture healing adequately. This model will be used for the development of new therapeutic strategies. Objectives: Our aim is to develop an in vitro 3D fracture gap model (FG model) which mimics the in vivo situation in order to provide a reliable preclinical test system for fracture healing disorders. Methods: To assemble our FG model, we co-cultivated coagulated peripheral blood and primary human mesenchymal stromal cells (MSCs) mimicking the fracture hematoma (FH model) together with a scaffold-free bone-like construct mimicking the bony part of the fracture gap for 48 h under hypoxic conditions (n=3), in order to reflect the in vivo situation after fracture most adequately. To analyze the impact of the bone-like construct on the in vitro FH model with regard to its osteogenic induction capacity, we cultivated the fracture gap models in either medium with or without osteogenic supplements. To analyze the impact of Deferoxamine (DFO, known to foster fracture healing) on the FG model, we further treated our FG models with either 250 µmol DFO or left them untreated. After incubation and subsequent preparation of the fracture hematomas, we evaluated gene expression of osteogenic ( RUNX2 , SPP1 ), angiogenic ( VEGF , IL8 ), inflammatory markers ( IL6 , IL8 ) and markers for the adaptation towards hypoxia ( LDHA , PGK1 ) as well as secretion of cytokines/chemokines using quantitative PCR and multiplex suspension assay, respectively. Results: We found via histology that both the fracture hematoma model and the bone-like construct had close contact during the incubation, allowing the cells to interact with each other through direct cell-cell contact, signal molecules or metabolites. Additionally, we could show that the bone-like constructs induced the upregulation of osteogenic markers ( RUNX2, SPP1 ) within the FH models irrespective of the supplementation of osteogenic supplements. Furthermore, we observed an upregulation of hypoxia-related, angiogenic and osteogenic markers ( RUNX2 , SPP1 ) under the influence of DFO, and the downregulation of inflammatory markers ( IL6 , IL8 ) as compared to the untreated control. The latter was also confirmed on protein level (e.g. IL-6 and IL-8). Within the bone-like constructs, we observed an upregulation of angiogenic markers (RNA-expression of VEGF , IL8 ), even more pronounced under the treatment of DFO. Conclusion: In summary, our findings demonstrate that our established in vitro FG model provides all osteogenic cues to induce the initial bone healing process, which could be enhanced by the fracture-healing promoting substance DFO. Therefore, we conclude that our model is indeed able to mimic correctly the human fracture gap situation and is therefore suitable to study the influence and efficacy of potential therapeutics for the treatment of bone healing disorders in immunosuppressed patients with ongoing inflammation. Disclosure of Interests: Moritz Pfeiffenberger: None declared, Alexandra Damerau: None declared, Paula Hoff: None declared, Annemarie Lang: None declared, Frank Buttgereit Grant/research support from: Amgen, BMS, Celgene, Generic Assays, GSK, Hexal, Horizon, Lilly, medac, Mundipharma, Novartis, Pfizer, Roche, and Sanofi., Timo Gaber: None declared