Taper connections of modular hip prostheses are at risk of fretting and corrosion, which can result in reduced implant survival. The purpose of this study was to identify the minimum torque required to initiate a removal of the passivation layer at the taper interface as a function of assembly force and axial load. Titanium stems and cobalt-chromium heads were assembled with peak impaction forces of 4.5 kN or 6.0 kN and then mounted on a materials testing machine whilst immersed in Ringer's solution. The stems were subjected to a static axial load (1 kN or 3 kN) along the taper axis. After a period of equilibration, a torque ramp from 0 to 15 Nm was manually applied and the galvanic potential was continuously recorded. Prostheses assembled with a force of 6 kN required a significantly higher torque to start a removal of the passivation layer compared to those assembled with 4.5 kN (7.23±0.55 Nm vs. 3.92±0.97 Nm, p=0.029). No influence of the axial load on the fretting behaviour was found (p=0.486). The torque levels, which were demonstrated to initiate surface damage under either assembly force, can be readily reached during activities of daily living. The damage will be intensified in situations of large weight and high activity of the patient or malpositioning of the prosthesis.

The patellofemoral joint is a common site of pain and failure following total knee arthroplasty. A contributory factor may be adverse patellofemoral biomechanics. Cadaveric investigations are commonly used to assess the biomechanics of the joint, but are associated with high inter-specimen variability and often cannot be carried out at physiological levels of loading. This study aimed to evaluate the suitability of a novel knee simulator for investigating patellofemoral joint biomechanics. This simulator specifically facilitated the extended assessment of patellofemoral joint biomechanics under physiological levels of loading. The simulator allowed the knee to move in 6 degrees of freedom under quadriceps actuation and included a simulation of the action of the hamstrings. Prostheses were implanted on synthetic bones and key soft tissues were modelled with a synthetic analogue. In order to evaluate the physiological relevance and repeatability of the simulator, measurements were made of the quadriceps force and the force, contact area and pressure within the patellofemoral joint using load cells, pressure-sensitive film, and a flexible pressure sensor. The results were in agreement with those previously reported in the literature, confirming that the simulator is able to provide a realistic physiological loading situation. Under physiological loading, average standard deviations of force and area measurements were substantially lower and comparable to those reported in previous cadaveric studies, respectively. The simulator replicates the physiological environment and has been demonstrated to allow the initial investigation of factors affecting patellofemoral biomechanics following total knee arthroplasty.

No abstract