Wear Simulation of Knee Implants

Introduction to Wear Simulation

There are now many different bearings and surgical approaches for total knee replacement (TKR), manufactured from an increasing variety of materials (eg, varied degrees of cross-linkage in polyethylene). There is a need for rigorous in vitro simulator studies to compare the effects of bearing design and material on the function and wear performance of the TKR. Historically, simple configuration tribological tests, such as pin on plate or pin on disk, provided a platform for defining the wear properties of a material under defined cyclic loading and motion patterns. However, the outcomes of these studies do not always correlate with the clinical performance of the same material. A major challenge in the laboratory is to produce simulations that provide clinically relevant data and predictions of future clinical performance. A further consideration is the role that computational predictions can play in a combined experimental and computational approach to assessing the preclinical performance of knee replacements. Computational modeling can support the in vitro experimental design and selection of appropriate test conditions and provide additional insight and understanding that cannot be gleaned from experimental studies alone, such as the contact stresses and cross-shear, providing that the computational models have been adequately experimentally validated.

Numerous factors affect the in vivo performance of a TKR, including the bearing design, material, lubrication, component position, patient loading and activity, and joint kinematics. Therefore, the performance of knee replacement bearings should be assessed in a physiologic simulator designed to replicate the in vivo conditions as closely as possible so that interactions among all these variables may also be assessed. This is challenging, given the range of conditions to which knee replacements are subjected in vivo. Current studies have predominantly focused on simulating standard gait, whereas more recently a wider envelope of conditions is beginning to be more widely considered in line with the SAFER approach. The SAFER approach accounts for variations in surgical delivery, variations in kinematics, variations in the patient population, and degradation of the biomaterials technology, as well as combinations of all these different conditions. Testing under such a wide portfolio of stratified conditions cannot be achieved using experimental simulation alone because too many experimental simulations are needed. A new computational modeling approach has been recently developed that can be combined with the experimental approach. The computational model uses a new complex wear law for polyethylene that accounts for a range of input variables, including variations in wear area, contact stress, cross-shear, and sliding distance. This information has been incorporated into finite element analysis wear prediction models for the knee, which have been experimentally validated under specific walking conditions. Computational models must always be experimentally validated for points at the extreme range of conditions, so that the computational models fall within an experimentally validated envelope.

Several different knee wear simulators have been developed both academically and commercially, and there are significant differences in the function and control of the simulators. Most current simulators use six degrees of freedom, four of which are actively controlled—axial load, anterior-posterior motion, internal-external rotation, and flexion-extension ( Fig. 23.1 ). The remaining two degrees of freedom, abduction-adduction and medial-lateral displacement, tend to be free to move passively or, in some simulators, are fixed. Two hypotheses for the control of motion in the anterior-posterior and internal-external rotational directions have been proposed, and both are defined in standards: force control (International Organization for Standardization [ISO] 14243-1) and displacement control (ISO 14243-3).

The stability of a knee replacement in vivo is determined by a combination of the geometry of the implant and the natural soft tissues surrounding the joint. The relative contribution varies in different patients. Force-controlled simulator studies often use spring elements to represent the motion constraints that are created by the soft tissues in vivo. The rationale for using force control suggests that it generates the most representative kinematics and provides a better indication of the mechanical behavior of a specific design. A simple linear spring cannot represent the complex force displacement relationships found in soft tissues. The geometry of the implant also dictates its resistance to motion, and therefore the tensioning of the springs may not highlight the effect of geometric differences between bearings when testing under force control. Under displacement control, the actual movements are directly controlled and provide a more repeatable set of motion conditions and wear; the machine force is limited, which prevents the knee replacement from experiencing excessive kinematics and ensures that the implant is not driven outside the geometric constraints.

It is believed that the bearing type may dictate the appropriate test conditions. Because of differences in test conditions, implants studied, and simulator setup, it has been difficult to directly compare the two modes. However, studies involving identical implants tested in the same simulator, under load or displacement control, have been reported and have shown that similar levels of anterior-posterior translation occurred under force and load control. A low-conforming, fixed-bearing study in which rotation and anterior-posterior translation were controlled by force or displacement, also showed similar magnitudes of internal-external rotation, but the phase of the motion within the gait cycle differed significantly. The volumetric wear under load control was significantly higher than under displacement control. Low-conformity bearings and cruciate-retaining implants may undergo excessive displacement under load control testing. Conversely, high-conformity bearings, tested under displacement control, may be driven to excessive displacement, which would not occur physiologically, and therefore experience very high stress. Moderately conforming bearings under both load and displacement control have been shown to generate equivalent wear volumes. There are obvious merits to both test hypotheses, depending on the design philosophy of the implant. Wear simulation studies should be conducted to produce clinically relevant kinematic profiles and wear data, and there is merit in simulation systems that have the flexibility to adopt both approaches to control on the anterior-posterior and rotation motion axes.

Furthermore, it is important to note that the current international standards, regardless of force and displacement control, describe a single set of idealized standard gait conditions for a standard patient in whom the prosthesis has been implanted with perfect surgical technique and under a single activity.

Effect of Input Kinematics on Bearing Wear

Early wear simulator studies of TKR produced very low wear rates, which was related to the input kinematics used for the simulations. It has been shown that the magnitude of anterior-posterior displacement and internal-external rotation has a significant effect on the wear of the polyethylene bearing. A standard displacement-controlled simulator study might use an anterior-posterior displacement of 0 to 10 mm and an internal-external rotation of ±5 degrees to re-create the motion within the natural knee (high kinematics). A 50% reduction in anterior-posterior displacement, while maintaining a ±5-degree rotation, has been shown to result in a twofold reduction in mean wear rate in fixed-bearing knees. Reducing the internal-external rotation to ±2.5 degrees while maintaining the 0- to 5-mm displacement resulted in a fourfold reduction in wear compared with high-kinematics studies. Removal of the internal-external rotation or the anterior-posterior displacement decreased the wear rate by an order of magnitude. A shorter anterior-posterior displacement results in a reduced sliding distance and therefore a reduced polyethylene surface area subjected to wear. However, the significant change in wear resulting from reduced rotation and displacement is caused by a change in the cross-shear on the surface of the bearing. A reduction in rotation reduces the cross-shear, therefore reducing the exposure of the polyethylene to wear in the strain-softened direction. An increase in internal-external rotation results in an increased frictional force transverse to the sliding distance and therefore increases the wear in a fixed-bearing knee.

Lift-off of the femoral condyles from the tibial bearing has been shown in vivo through fluoroscopic studies. One experimental study investigated the effect of lift-off on the wear performance of fixed and mobile TKRs by examining inserts with similar contact geometry. Condylar lift-off was achieved in the study by application of a rotation moment to the abduction-adduction axis to create 1 mm of lateral condyle lift-off during each gait cycle. A significant increase in wear rate was measured for fixed and mobile bearings, and it was notable that the reduced wear rates observed in the mobile knees compared with the fixed knees under standard gait conditions were not seen under lift-off. There was no significant difference between the wear rates of the bearings under lift-off conditions. The uneven loading of the insert during lift-off caused a medial-lateral shift, increasing the cross-shear on the medial condyle and increasing the wear. This study showed lift-off to have a significant impact on the wear rate of the bearing; however, this test had lift-off during every cycle and the clinical translation of this research would depend on the frequency of lift-off in vivo.

Influence of Bearing Design on in Vitro Wear Performance of Knee Replacements

A wide variety of knee replacements are clinically available, which can be subdivided into two groups, fixed-bearing knees and mobile-bearing knees. In addition, designs may vary on the degree of conformity between the ultra–high-molecular-weight polyethylene (UHMWPE) insert and the femoral bearing. Unicompartmental knee replacements (UKRs) are also becoming increasingly popular clinically. In vitro wear simulator testing, under defined loading and motion conditions, can demonstrate the effect of bearing design on the wear performance of a knee replacement.