Internal kinetic changes in the knee due to high tibial osteotomy are well-correlated with change in external adduction moment: An osteoarthritic knee model
Introduction
Osteoarthritis (OA) of the knee is most often unicompartmental, more frequently affecting the medial compartment (Baliunas et al., 2002). The tibial plateau articulates with the femoral condyles, creating a joint that bears up to 2.5 times body weight during walking (Mundermann et al., 2008). Given that the line of action of the ground reaction force passes medially to the knee joint throughout stance phase, loading in the medial compartment is significantly higher than in the lateral (Shelburne et al., 2006). A varus knee alignment would tend to increase this load asymmetry, and so varus alignment has been identified as a risk factor for medial compartment OA (Teichtahl et al., 2009).
The external knee adduction moment (EKAM) has been demonstrated to be a valid proxy for medial compartment loading and OA severity (Hunt et al., 2006). However, the correlation between an EKAM and medial compartment load is not perfect, since many of the soft tissues structures spanning the knee joint tend to share the external adduction moment during gait (Zhao et al., 2007). Additionally, studies have shown that patterns of muscle activation in subjects with osteoarthritis are different from healthy subjects in that the augmented muscle activation patterns serve to reduce loads on the knee as subjects attempt to avoid pain (Lewek et al., 2004, Hubley-Kozey et al., 2008). Due to the subject-specific nature of OA and the multiple factors that can affect joint loading, a more direct measure of an in-vivo medial and lateral compartment loading is often sought. Previous studies have used an inverse dynamics approach in two-dimensions (Collins and O’Connor, 1991) and three-dimensions (Shelburne et al., 2006, Kakihana et al., 2005; Glitsch et al., 1997) with an in-vivo kinematic and kinetic gait data from patient populations to estimate the internal kinetics of knee joint soft tissue structures. More complex finite element modeling has also been used to estimate stress distributions within the meniscus, cartilage and ligaments (Papaioannou et al., 2008, Yao et al., 2008, Donahue et al., 2002, Penrose et al., 2002, Li et al., 1999, Perie and Hobatho, 1998). But these models are difficult to validate. Accurate modeling depends on representative geometry of the in-vivo knee. Anatomical geometries have been derived from cadaveric specimens (Morrison et al., 1969; Crowninshield and Brand, 1981) and medical imaging, such as MRI (Hashemi et al., 2008, Yao et al., 2008, Heller et al., 2008, Bei and Fregly, 2004, Penrose et al., 2002, Arnold et al., 2000, DeFrate et al., 2004, Li et al., 1999, Perie and Hobatho, 1998) or CT (Au et al., 2008). But applying subject-specific knee models to a large patient cohort is unfeasible in a clinical setting. The model in this study uses an inverse dynamic, optimization approach with geometry derived from a generic knee that is scaled to the patient size.
An alternative to total knee replacement for treating medial compartment OA is the opening-wedge high tibial osteotomy (Spahn and Wittig, 2002). The HTO creates a wedge in the medial proximal tibia, which is plated medially. This realigns the lower leg, moving the knee joint medially and reducing varus angulation. HTO is thought to shift loading from the medial into the lateral compartment (Agneskirchner et al., 2007). A study of medial and lateral compartment pressures in cadavers under simulated walking gait showed a strong correlation between the medial-to-lateral compartment load ratio and knee alignment, and showed a significant change in the load ratio with a HTO (Agneskirchner et al., 2007). Due to the significant alteration of geometry with the HTO, this procedure is a prime candidate to examine with a model of internal kinetics. Such a model could test whether load is indeed shifted laterally with the HTO.
This study examines the correlation between the external kinetic measure of external knee adduction moment (EKAM) and the internal kinetic measures of medial compartment load (ML) and medial-to-lateral load ratio (MLR) during walking gait and the stability of these changes over time. These measures were examined at three time points: immediately pre-HTO, 6 and 12 months post-HTO. Three hypotheses were tested: (1) that the HTO procedure reduces an EKAM, an ML and an MLR, (2) these measures are not significantly different at 6 and 12 months post-HTO, and (3) the change in the impulse of EKAM due to a HTO is well-correlated with the impulse of an MLR. The change in an EKAM-impulse with HTO was used, since it has been shown to better differentiate OA severity than peak magnitudes of EKAM (Maly et al., 2008, Thorp et al., 2006, Thorp et al., 2007). An in-house computational model estimated internal kinetic measures of an ML and an MLR. The EKAM was determined from the an in-vivo walking gait data using an inverse dynamic approach. A large cohort of 30 knee OA patients, who were undergoing a HTO realignment was examined.
Section snippets
Subjects
Kinematic and kinetic data for 30 HTO patients (mean (SD) at pre-HTO collection: mass 92.17 (13.70) kg; height 174.87 (8.38) cm); during walking gait was retrieved over the database. Patients were randomly selected from the population of unilateral opening-wedge HTOs to correct a varus malalignment (Hunt et al., 2006). Each patient had performed three gait analyses (immediately pre-HTO, 6 and 12 months post-HTO).
Gait analysis data and inverse dynamics
Kinematic data was collected with an eight-camera motion capture system (Hi-Res Eagle
Results
An average curves for an EKAM, an ML and an MLR at each test time point are shown in Fig. 4. An average instance of the first peak of an EKAM across the 30 patients was 23% stance phase.
There was a significant decrease (p<0.01) for all three kinetic measures between pre-HTO and 6 months post-HTO (Table 1). An EKAM dropped by 1.70 (1.37) %BW-ht, an ML dropped by 0.56 (0.52) %BW and an MLR dropped by 1.00 (1.05). There were no significant differences between 6 and 12 months post-HTO for any of
Discussion
The first hypothesis was confirmed with the first peak in an average EKAM reduced significantly due to the HTO. The instance used for comparison was found by observing the peak EKAM from the population average curves (Fig. 4—A, B and C). At this same point in stance phase, the medial compartment load (ML) and the ratio of medial-to-lateral loading were also significantly reduced. These changes were also stable at 12 months post-HTO, which confirmed the second hypothesis, with no significant
Conflict of interest statement
There are no commercial relationships to this work that may lead to conflict of interests. There are no financial or personal relationships with other people or organizations that could inappropriately influence this work.
Acknowledgements
This research was funded by a Natural Science and Engineering Research Council Discovery grant.
References (40)
- et al.
Orientation of tendons in-vivo with active and passive knee muscles
Journal of Biomechanics
(2005) - et al.
The effects of valgus medial opening wedge high tibial osteotomy on articular cartilage pressure of the knee: a biomechanical study
Arthroscopy
(2007) - et al.
Sensitivity of kinematics-based model predictions to optimization criteria in static lifting tasks
Medical Engineering & Physics
(2006) - et al.
A NURBS-based technique for subject-specific construction of knee bone geometry
Computer Methods and Programs in Biomedicine
(2008) - et al.
Increased knee joint loads during walking are present in subjects with knee osteoarthritis
Osteoarthritis and Cartilage
(2002) - et al.
Multibody dynamic simulation of knee contact mechanics
Medical Engineering & Physics
(2004) - et al.
The sensitivity of muscle force predictions to changes in physiologic cross-sectional area
Journal of Biomechanics
(1986) - et al.
A physiologically based criterion of muscle force prediction in locomotion
Journal of Biomechanics
(1981) - et al.
In-vivo tibiofemoral contact analysis using 3D MRI-based knee models
Journal of Biomechanics
(2004) - et al.
Muscle co-activation patterns during walking in those with severe knee osteoarthritis
Clinical Biomechanics (Bristol, Avon)
(2008)