Osteochondral defect repair with a collagen/collagen-magnesium-hydroxyapatite (Col/Col-Mg-HAp) scaffold has demonstrated good clinical results. However, subchondral bone repair remained suboptimal, potentially leading to damage to the regenerated overlying neocartilage. This study aimed to improve the bone repair potential of this scaffold by incorporating newly developed strontium (Sr) ion enriched amorphous calcium phosphate (Sr-ACP) granules (100–150 μm). Sr concentration of Sr-ACP was determined with ICP-MS at 2.49 ± 0.04 wt%. Then 30 wt% ACP or Sr-ACP granules were integrated into the scaffold prototypes. The ACP or Sr-ACP granules were well embedded and distributed in the collagen matrix demonstrated by micro-CT and scanning electron microscopy/energy dispersive x-ray spectrometry. Good cytocompatibility of ACP/Sr-ACP granules and ACP/Sr-ACP enriched scaffolds was confirmed with in vitro cytotoxicity assays. An overall promising early tissue response and good biocompatibility of ACP and Sr-ACP enriched scaffolds were demonstrated in a subcutaneous mouse model. In a goat osteochondral defect model, significantly more bone was observed at 6 months with the treatment of Sr-ACP enriched scaffolds compared to scaffold-only, in particular in the weight-bearing femoral condyle subchondral bone defect. Overall, the incorporation of osteogenic Sr-ACP granules in Col/Col-Mg-HAp scaffolds showed to be a feasible and promising strategy to improve subchondral bone repair.

In recent decades, an increasing number of tissue engineered bone grafts have been developed. However, expensive and laborious screenings in vivo are necessary to assess the safety and efficacy of their formulations. Rodents are the first choice for initial in vivo screens but their size limits the dimensions and number of the bone grafts that can be tested in orthotopic locations. Here, we report the development of a refined murine subcutaneous model for semi-orthotopic bone formation that allows the testing of up to four grafts per mouse one order of magnitude greater in volume than currently possible in mice. Crucially, these defects are also "critical size" and unable to heal within the timeframe of the study without intervention. The model is based on four bovine bone implants, ring-shaped, where the bone healing potential of distinct grafts can be evaluated in vivo. In this study we demonstrate that promotion and prevention of ossification can be assessed in our model. For this, we used a semi-automatic algorithm for longitudinal micro-CT image registration followed by histological analyses. Taken together, our data supports that this model is suitable as a platform for the real-time screening of bone formation, and provides the possibility to study bone resorption, osseointegration and vascularisation.