Abstract
Introduction
It is believed that wear of replacement joints vivo in is strongly dependent on input motions (kinematics) and loading. There is difficulty in accurately measuring total disc replacement (TDR) kinematics in vivo. It is therefore desirable to ascertain the sensitivity of implant wear in vitro to perturbations of the standard testing parameters. An anterior-posterior (AP) shear force input is not currently included in the present ISO and ASTM testing standards for lumbar TDRs but is known to exist in in vivo. Other joint-replacement wear tests have shown that the phasing of input motions influences the ‘cross-shear’ process of polyethylene wear. Polyethylene bearing materials do not behave linearly to axial loading changes and so the effect on wear rate is difficult to predict. The study aim was to assess the effects on wear of a ProDisc-L TDR under the following conditions: ISO 18191-1 standard inputs; an additional input AP shear; input kinematics phasing changes; axial loading changes.
Methods
A five active degree of freedom (DOF) spine simulator was used to compare the effects of varying the kinematic and loading input parameters on a ProDisc-L TDR (Synthes Spine). A four DOF standard ISO (ISO18192-1) test was followed by a five DOF test which included the AP shear force. The standard ISO test was repeated on a second simulator (of identical design) but with the phasing of the lateral bend (LB) and flexion extension (FE) motions changed to be in-phase, creating a low cross-shear motion pattern. The standard ISO test was then modified to give half the ISO standard axial loading. All tests conducted were based on the ISO18192-1 standard for lumbar implants with 15 g/l protein lubricant and modified as described. Gravimetric wear measurements were taken every million cycles (mc) in units of milligrams (mg). Six discs were tested to give statistically significant results.
Results
When the fifth DOF AP input force was added, the wear rate showed a non-significant (p=0.78) change in mean wear rate from 12.7 ± 2.1 mg/mc (± standard deviation) to 11.6 ± 1.2 mg/mc. For the repeated test, on the second simulator, changing from standard ISO to in-phase FE-LB conditions (producing a low cross-shear wear pattern) showed a significant mean wear rate fall of 16.1 ± 1.4 mg/mc to 6.0 ± 1.3 mg/mc. The low-load test showed a marginally non-significant (p=0.18) difference in mean wear rate from 16.0 ± 0.8 mg/mc to 15.1 ± 0.8 mg/mc.
Conclusion
When comparing the standard ISO test with the modified five DOF AP input test no significant difference in mean wear rate was observed. Comparing the standard ISO test with the modified in-phase (low cross-shear) test produced a significant 62% reduction in wear rate. Reducing the loading by half did not produce a significant fall in mean wear rate. The wear of lumbar TDRs is strongly dependant on input phasing kinematics and perhaps not so dependent on axial loading and AP shear. This counter-intuitive result is important for in vivo wear performance estimation
