Abstract
INTRODUCTION
Knee simulators are being used to evaluate wear. The current international standards have been developed from clinical investigations of the normal knee [1, 2] or from a single TKA patient [3, 4]. However, the forces and motions in a TKA patient differ from a normal knee and, furthermore, the resulting kinematic outcomes after TKA will depend on the design of the device [5]. Consequently, these standard tests may not recreate in-vivo conditions; therefore, the goal of this study was to perform a novel wear simulation using design-specific inputs that have been derived from fluoroscopic images of a deep knee bend.
METHODS
A wear simulation was developed using fluoroscopic data from a pool of eighteen TKA patients performing a deep knee bend. All patients had a Sigma CR Fixed Bearing implant (DePuy) and were well functioning (Knee Society Score > 90). A single patient was selected that represented the typical motions, which was characterized by early rollback followed by anterior motion with an overall modest internal tibial rotation (Figure 1). The relative motion between the femoral and tibial components was transformed to match the coordinate system of an AMTI knee wear simulator [6] and a compressive load input was derived using inverse dynamics [7]. The resulting force and motions (Figure 2) were then applied in a wear simulation with 5 MRad crosslinked and remelted polyethylene for 3 Mcyc at 1 Hz. Components were carefully positioned and each joint (n=3) was tested in 25% bovine calf serum (Hyclone Laboratories), which was recirculated at 37±2°C [3]. Serum was supplemented with sodium azide and EDTA. Wear was quantified gravimetrically every 0.5 Mcyc using a digital balance (XP250, Mettler-Toledo) with load soak compensation.
RESULTS
The knee simulator was able to recreate the in-vivo input kinematics. The femoral low point location revealed good agreement between in-vivo and in-vitro conditions and the overall pattern of the motion from full extension to maximum knee flexion was replicated (Figure 3). The measured wear from these inputs was very low (0.7 ± 0.2 mg/Mcyc).
DISCUSSION
We have performed a device-specific wear simulation for a deep knee bend. Surprisingly, the wear associated with this activity was very low. It is possible that abnormal kinematics, including paradoxical anterior slide and reverse rotation, would generate higher wear. The deviations the between in-vivo and in-vitro kinematics (Figure 3) are likely due to a size mismatch across the transformation process. In a previous study [7] we recreated the in-vivo motions with better fidelity (RMS error = 0.6mm) using size matched components. Further work is needed to improve the transformation technique for different sized components. Also, similar approaches will be used in future investigations to study the effect of abnormal kinematics as well as other designs including rotating platform and cruciate substituting devices.
