ESB clinical biomechanics award 2008: Complete data of total knee replacement loading for level walking and stair climbing measured in vivo with a follow-up of 6–10 months

Abstract

Background

Detailed information about the loading of the knee joint is required for various investigations in total knee replacement. Up to now, gait analysis plus analytical musculo-skeletal models were used to calculate the forces and moments acting in the knee joint. Currently, all experimental and numerical pre-clinical tests rely on these indirect measurements which have limitations. The validation of these methods requires in vivo data; therefore, the purpose of this study was to provide in vivo loading data of the knee joint.

Methods

A custom-made telemetric tibial tray was used to measure the three forces and three moments acting in the implant. This prosthesis was implanted into two subjects and measurements were obtained for a follow-up of 6 and 10 months, respectively.

Subjects performed level walking and going up and down stairs using a self-selected comfortable speed. The subjects’ activities were captured simultaneously with the load data on a digital video tape. Customized software enabled the display of all information in one video sequence.

Findings

The highest mean values of the peak load components from the two subjects were as follows: during level walking the forces were 276 %BW (percent body weight) in axial direction, 21 %BW (medio-lateral), and 29 %BW (antero-posterior). The moments were 1.8 %BW*m in the sagittal plane, 4.3 %BW*m (frontal plane) and 1.0 %BW*m (transversal plane). During stair climbing the axial force increased to 306 %BW, while the shear forces changed only slightly. The sagittal plane moment increased to 2.4 %BW*m, while the frontal and transversal plane moments decreased slightly. Stair descending produced the highest forces of 352 %BW (axial), 35 %BW (medio-lateral), and 36 %BW (antero-posterior). The sagittal and frontal plane moments increased to 2.8 %BW*m and 4.6 %BW*m, respectively, while the transversal plane moment changed only slightly.

Interpretation

Using the data obtained, mechanical simulators can be programmed according to realistic load profiles. Furthermore, musculo-skeletal models can be validated, which until now often lacked the ability to predict properly the non-sagittal load values, e.g. varus–valgus and internal–external moments.

Introduction

Total knee replacement (TKR) is a widely used surgical procedure to relieve pain from arthritic joints and to restore their range of motion. The number of TKR surgeries is increasing steadily, as populations becomes older and the age at the time of primary TKR is decreasing. Therefore the survival rates of TKR are still an issue and the longevity of the implant should be maximized. One of the major reasons for revision is wear of the ultra high molecular weight polyethylene (UHMWPE) inserts (Kellett et al., 2004, Sharkey et al., 2002, Vessely et al., 2006), due to the geometry of the articulating surfaces and the high forces and moments acting in the knee joint. Therefore the development of new or modified materials for the articulating components is the subject of intensive research on TKR. For example, highly cross linked UHMWPE was developed or vitamin E was added in many TKR systems to decrease the wear rate (Muratoglu et al., 2007, Teramura et al., 2007). However, prior to market release, these components are required to be tested pre-clinically, using multi-axial mechanical simulators, Finite-Element-Analysis (FEA) or cadaveric testing (Barink et al., 2005, Knight et al., 2007). Unfortunately, results from such investigations have been found to show up differently, because the input variables for load definition and mechanical test conditions vary considerably.

The loading profiles used in these pre-clinical investigations were obtained indirectly from gait data and inverse dynamic analysis. Some of these studies were performed decades ago (Morrison, 1970, Seireg and Arvikar, 1975), and show significant differences in their results, yet serve as the current gold standard. Another disadvantage of these studies is the frequent limitation to two-dimensional (2D) models. In reality the loads acting in the knee joint consist of three force and three moment components. More recently, advanced three-dimensional (3D) modelling provided more information on the loading of TKR (Taylor et al., 2004). However, the complete information of all six load components is rarely available from one single study (Costigan et al., 2002).

To overcome this lack of loading data, several studies to measure the in vitro and in vivo loading conditions of the knee joint have been performed. Some of them have used tibial baseplates instrumented with force transducers to measure loads in cadaveric studies or intraoperatively (Crottet et al., 2007, Crottet et al., 2005, Jeffcote et al., 2007, Kaufman et al., 1996, Nicholls et al., 2007, Singerman et al., 1999, Singerman et al., 1994). However, these implants were not suited for in vivo measurements.

The first TKR for in vivo measurements was an instrumented massive femoral replacement, measuring the forces and moments in its stem more than 20 cm away from the joint line (Taylor et al., 1998). Subsequently, an additional numerical model was necessary to transfer the measured loads to the joint line level. Furthermore, the prosthesis was a hinge joint used for tumor patients with muscular deficiencies. These measurements delivered the first realistic results on the magnitude of loading in a TKR. However, due to the patient group and implant design these results are not representative for today’s TKR designs.

Some years ago, an instrumented tibial baseplate was implanted into one subject allowing the measurement of the axial force component (D’Lima et al., 2005, Mundermann et al., 2008). More recently, load measurements using a six degree of freedom (DOF) prosthesis in up to three subjects with a follow-up of up to twelve months were also reported (D’Lima et al., 2007, D’Lima et al., 2008, Varadarajan et al., 2008). However, not all load components were reported. Especially the rotation moment around the tibial axis was found to be almost lacking. The flexion–extension and varus–valgus moments were only reported for chair-rise and squat activities, but not for walking and stair climbing, which are indeed the most frequent activities.

Therefore, the goal of this study was to obtain accurate in vivo knee contact loads, which are three forces and three moments, during the most frequent strenuous activities, namely walking and going up and down stairs. Here, we report the data of the first two subjects with instrumented knee implants with follow-ups of 6 and 10 months.