Computer modeling to support personalized strategies for treatment of compromised conditions in bone healing

Bone has the ability to regenerate itself following a fracture or injury. However, there are situations such as Type 2 Diabetes Mellitus (T2DM) or bone defects, that impairs this ability. Such compromised conditions affect cellular activities necessary for bone healing, impairing the overall process and potentially leading to delayed healing or non-union. However, the causes underlying impaired bone regeneration in such compromised conditions are to this day not clear. Identifying the causes behind impaired bone regeneration will not only allow for unraveling mechanisms behind such impairment but also create a basis for personalized treatment strategies. This thesis aims to develop multiscale computer models to support treatment strategies for compromised conditions in bone healing. A multiscale in silico approach, combining finite element analysis and agent-based computer models, was used to simulate the bone healing process. Finite element techniques investigated mechanical signals within the bone healing region, while agent-based models simulated cellular dynamics within the callus. T2DM bone regeneration was investigated by incorporating reported T2DM-related effects on bone regeneration into a computational model. Furthermore, scaffold design, particularly gyroid architecture, was investigated for its influence on bone healing as a potential treatment strategy. In addition, a computational framework integrating omics data was made to individualize generic computer models of bone regeneration. Validation of the predictions from the computer models was done using in vivo experiment data such as micro-CT, histological, or x-ray images from pre-clinical animal studies. Computer model results revealed that mechanical alterations minimally affect reduced bone regeneration in T2DM. Instead, alterations in mesenchymal stem cell proliferation, migration, and osteoblast differentiation were identified as key factors to impaired bone regeneration in T2DM. Compared to a strut-like scaffold, gyroid scaffolds showed lower mechanical stiffness and higher deformations within the callus region. Bone regeneration simulations found that gyroid scaffolds resulted in slower healing dynamics and reduced bone tissue formation. Moreover, personalized omics-driven computer models of bone regeneration, based on experimental omics data, were able to predict bone healing dynamics of a mouse fracture model. In addition, simulated compromised protein regulation led to reduced bone healing outcome predictions. The project highlights the possibility of using computer models for the simulation of complex bone healing processes, revealing key factors influencing bone regeneration in T2DM and the effects of scaffold design. Such findings are fundamental for designing effective therapeutic strategies targeted at enhancing bone healing. Moreover, the developed personalized computer models offer valuable tool to clinicians in tailoring regenerative therapy to the individual patient's profile.