Mechanical Properties of Human Bone-Implant Interface Tissue in Aseptically Loose Hip Implants.

The main cause of failure in total hip replacement is aseptic loosening which is associated with the formation of a periprosthetic fibrous (interface) tissue. Despite important applications for finite element modeling of loose implants, the mechanical properties of the bone implant interface tissue have never been measured in humans. In this study, we performed unconfined compression tests to characterize the mechanical properties of the interface tissue and to determine the parameters of various hyperelastic material models which were fitted to the measurements. Human interface tissues were retrieved during 21 elective revision surgeries from aseptically loosened cemented (N=10) and uncemented hip implants (N=11). Specimens were tested at a fixed deformation rate of 0.1 mm/min up to a maximum force of 10 N. Elastic moduli for low and high strain regions of the stress strain curves were determined. Interface tissue from aseptically loose cemented prostheses shows higher elastic moduli (mean=1.85 MPa, 95% C.I.=1.76-1.95 MPa) in the high strain region as compared to that of the interface tissue from the cementless group (mean=1.65 MPa, 95% C.I.=1.43-1.88 MPa). The 5-terms Mooney Rivlin model (W = C-1[I-1 - 3] + C-2[I-2 - 3] + C-3[I-1 - 3][I-2 - 3] + C-4[I-1 - 3](2) + C-5[I-2 - 3](2)) described the stress strain behavior the best. Large variations in the mechanical behavior were observed both between specimens from the same patient as between those of different patients. The material model parameters were therefore estimated for the mean data as well as for the curves with the highest and lowest strain at the maximum load. The model parameters found for the mean data were C-1=-0.0074 MPa, C-2=0.0019 MPa, C-3=0 MPa, C-4=0.0032 MPa and C-5=0 MPa in the cemented group and C-1=0.0137 MPa, C-2=0.0069 MPa, C-3=0.0026 MPa, C-4=-0.0094 MPa and C-5=0 MPa in the cementless group. The results of this study can be used in finite element computer. (C) 2014 Elsevier Ltd. All rights reserved.