23 Full factorial design of experiment elucidates interaction of mechanical parameters driving in vitro cartilage formation in a bioreactor
Background: Traumatic articular cartilage injuries lead to cartilage degeneration and increase the risk of post-traumatic osteoarthritis. Since adult cartilage exhibits a low regenerative capacity, therapies for its repair are urgently needed. We investigated how mechanical stimulation drives stem cells to form cartilage to replace the damaged tissue. This is conducted in an in vitro bioreactor that simulates the mechanical environment that the cells are usually exposed to. A full factorial design of experiment is applied to explore the interaction between different mechanical parameters. The aim is to establish a loading protocol that optimizes production of proteins important for cartilage formation.
Material and methods: Human stem cells were encapsulated in a fibrin-poly(ester-urethane) scaffold and subsequently subjected to joint-mimicking mechanical load within a bioreactor. According to the full factorial design, different combinations of two factors, namely type of indenter and loading protocol, were chosen to elucidate the factors’ main and interaction effects on biomarker secretion. Two types of indenters were used: a cylinder and a ball. The loading protocols consisted of different combinations of shear and compression. Culture medium was collected and investigated using biochemical assays. The analysis was performed using a linear mixed model.
Results: A significant interaction between type of indenter and loading protocol was observed for various biomarkers. Using a cylindrical indenter and shear and compression on the highest settings resulted in the largest activation of growth factors driving cartilage formation without leading to increased cell death.
Conclusion: The full factorial experimental design represents a viable methodology to investigate interaction between loading parameters. We were able to determine loading parameters that favor in vitro cartilage formation. Additionally, by using the right statistical modelling approach, we were able to demonstrate significance that would have otherwise been hidden. Insights gained from this study could be used for rehabilitation protocols in clinics.