Abstract
Introduction
Previous studies have shown that third body damage to the femoral head in metal-on-polyethylene hip replacement bearings can lead to accelerated wear of the polyethylene liners. The resulting damage patterns observed on retrieved metal heads are typically scratches and scrapes. The damage created in vitro must represent the third body damage that occurs clinically. A computational model was developed to predict the acceleration of wear of polyethylene articulating against in vitro damaged femoral heads. This involved using a damage registry from retrieval femoral heads to develop standardized templates of femoral head scratches statistically representative of retrieval damage
The aim of this study was to determine the wear rates of polyethylene liners articulating against retrievals and artificially damaged metal heads for the purpose of validating a computational wear prediction model; and to develop and validate an in vitro standardised femoral head damage protocol for pre-clinical testing of hip replacements.
Materials and Methods
Twenty nine, 32mm diameter, metal-on-moderately cross-linked polyethylene bearings (MarathonTM) inserted into Ti-6Al-4V shells (Pinnacle®) were tested in this study. All products were manufactured by DePuy Synthes, Warsaw, Indiana, USA. Following a retrieval study seven different damage patterns were defined, and these were applied to the femoral heads using a four-degree-of-freedom CNC milling machine (Figure 1). The ProSim 10-station pneumatic hip joint simulator (Simulation Solutions, UK) was used for experimental wear simulation using standard gait cycles and testing each experimental group for 3 million cycles. The acetabular cups were inclined at 35° on the simulator (equivalent to 45° in vivo). The wear volumes were determined using a microbalance (Mettler-Toledo XP205, Switzerland) at one million cycle intervals. Statistical analysis used was one way ANOVA followed by a post hoc analysis with significance taken at p<0.05.
Results
Different damage patterns accelerated the wear of polyethylene at different rates (Figure 2). The moderately scratched and severely scratched heads caused a 2 fold (p<0.01) and 5.5 fold (p<0.01) increase when compared to the wear rate of the undamaged head group. However, the scraped damage caused a lower increase than the scratched heads, with a 1.4 fold (p=0.2) increase for the moderately scraped heads and 2.6 fold (p<0.01) increase for the severely scraped heads. The moderate hybrid and severe hybrid groups resulted in a similar increase to the scraped heads with 1.8 fold (p<0.01) increase with the moderate hybrid and 3 fold (p<0.01) increase with the severe hybrid. The wear of polyethylene against the mild hybrid and retrieved heads was not significantly different (p= 0.9) to the wear against undamaged heads.
Discussion
A standardised protocol for generating in vitro damage representative of clinically occurring damage on femoral heads for preclinical testing purposes is needed. The wear rates of polyethylene liners articulating against the retrieval heads were similar to those articulating against the undamaged femoral heads. This study has shown the variations in wear rate of polyethylene bearing under different damage patterns generated in vitro. The wear prediction computational model predict similar trends of the wear acceleration reported in the experimental study.
