Mechano-regulation study in scaffolds for tissue engineering using fluid-structure interaction models

Abstract

The body cannot heal bone fractures surpassing 2 cm, so nowadays, graft surgery is the only available treatment. Nonetheless, this entails serious risks such as immune rejection, second medical interventions, and cross-infection. Therefore, bone tissue engineering has arisen as an alternative approach able to solve this problem. Bone tissue engineering provides temporary mechanical support for bone regeneration through an artificially prepared extracellular matrix (i.e. scaffolds) to allow cell differentiation, proliferation, and migration. To do so, global mechanical load (fluid or structural) is transferred as stimuli to cells through the scaffold architecture. Adequate mechanical characteristics, biomaterials and stimuli promote a proper mesenchymal cell differentiation to bone phenotype. The present study aims to develop an in-silico study of bone tissue differentiation in diverse scaffold designs using fluid-structure interaction models. The relation between scaffold strain deformation and fluid mechanical stimuli developed at the cell microscopic level are analysed. The optimal configuration that leads to cell differentiation in order to restore a bone lesion is chosen. To accomplish this, on the one hand, computational solid mechanics and computational fluid transient states models were implemented for all the scaffolds with steady-state and transient state inputs. On the other hand, fluid-structure interaction models were performed considering four scenarios. Finally, cell differentiation studies considering the octahedral shear strain and fluid shear stress have been compared. It has been found that high porous scaffolds with low transient state compression and velocity resulted in an increment in bone tissue phenotype. Moreover, it has been established that computational models not presenting interaction between solid and fluid phases can lead to overestimating bone tissue differentiation. For this reason, it is concluded that fluid-structure models are capable of mimicking and evaluating both transient state mechanical stimulations closest to reality.