Radially and axially graded cellular tibial stems for total knee replacement

Abstract

In spite of the high success rate of existing total knee replacement implants, the aseptic loosening of components, particularly the tibial one, is a matter of concern. To control the short- and long-term loosening mechanisms, the tibial stem material shall have diverse stiffness properties in different regions and directions, such that it encourages osseointegration and prevents bone resorption. The previous studies in the literature have failed to provide designs which could satisfy these two requirements concurrently. The purpose of this study was to design heterogeneous and anisotropic porous titanium tibial stems, based on radial and axial functionally graded cellular (FGC) structures, with precisely tailored stiffness properties. A 3D voxel-based finite element model of the bone-implant construct was developed to predict the nodal micromotion at the interface, assuming a frictional contact, and the stress-shielding effect in the periprosthetic bone, assuming a perfectly bonded connection. An exponential FGC material model was assigned to the implant along each of the radial and axial directions, with the gradient parameters considered as design variables. A design of experiment (DOE) -based multi-objective optimization procedure was then employed to maximize the osseointegration potential and minimize the bone resorption index. Equivalent FGC lattices, composed of Cartesian and cylindrical arrangements of gyroid unit cells, were produced for the optimal designs and fabricated by selective laser melting (SLM) technique. In comparison with the solid stem, the radial graded design resulted in a substantial improvement in the bone resorption behavior (19.4%), but a moderate reduction in the osseointegration potential (9.4%). The axial graded design, on the other hand, was associated with a moderate improvement in the bone resorption characteristics (12.0%) and slight improvement in the osseointegration index (2.2%). The dimensional accuracy of the SLM-fabricated specimens was reasonably high (strut thickness: 0.032 to 0.076 mm RMSE, pore size: 0.029 to 0.057 mm RMSE). The radial and axial graded specimens behaved as brittle and ductile materials, respectively, in axial compression tests and failed due to buckling and excessive plastic deformation. The cylindrical lattice arrangement demonstrated a lower elastic modulus (radial graded: 1945 vs. 3053 MPa, axial graded: 779 vs. 2324 MPa) and yield strength (radial graded: 77 vs. 114 MPa, axial graded: 17 vs. 43 MPa) than the Cartesian arrangement. It was concluded that an optimal bi-directional axial and radial gradient structure with a cylindrical lattice arrangement would provide the most favorable biomechanical performance for the tibial stem.