Elsevier

Journal of Biomechanics

Volume 44, Issue 6, 7 April 2011, Pages 995-1000
Journal of Biomechanics

In vivo kinematics of two-component total ankle arthroplasty during non-weightbearing and weightbearing dorsiflexion/plantarflexion

https://doi.org/10.1016/j.jbiomech.2011.02.078Get rights and content

Abstract

Relatively high rates of loosening and implant failure have been reported after total ankle arthroplasty, especially in first and second generation implants. Abnormal kinematics and incongruency of the articular surface may cause increased loads applied to the implant with concomitant polyethylene wear, resulting in loosening and implant failure. The purpose of this study was to measure three-dimensional kinematics of two-component total ankle arthroplasty during non-weightbearing and weightbearing activities, and to investigate incongruency of the articular surfaces during these activities. Forty-seven patients with a mean age of 71 years were enrolled. Radiographs were taken at non-weightbearing maximal dorsiflexion and plantarflexion, and weightbearing maximal dorsiflexion, plantarflexion, and neutral position. 3D–2D model-image registration was performed using the radiographs and the three-dimensional implant models, and three-dimensional joint angles were determined. The implanted ankles showed 18.1±8.6° (mean±standard deviation) of plantarflexion, 0.1±0.7° of inversion, 1.2±2.0° of internal rotation, and 0.8±0.6 mm of posterior translation of the talar component in the non-weightbearing activity, and 17.8±7.5° of plantarflexion, 0.4±0.5° of inversion, 1.8±2.0° of internal rotation, and 0.7±0.5 mm of posterior translation in the weightbearing activity. There were no significant differences between the non-weightbearing and weightbearing kinematics except for the plantarflexion angle. Incongruency of the articular surface occurred in more than 75% of the ankles. Our observations will provide useful data against which kinematics of other implant designs, such as three-component total ankle arthroplasty, can be compared.

Introduction

Total ankle arthroplasty is a treatment of choice for end-stage arthritis of the ankle. After disappointing failure rates of the first generation total ankle arthroplasties in the 1970s (Saltzman, 1999, Gougoulias et al., 2009, Bonasia et al., 2010), the second generation implants were developed with improved designs and fixation methods. These implants have been an increasingly popular alternative to ankle arthrodesis in the last decade (Saltzman et al., 2000). The second generation total ankle arthroplasties include two categories (Lewis, 2004): the two-component type consisting of a talar component and a tibial component with the polyethylene insert fixed with the tibial component, and the three-component type consisting of a mobile bearing insert between the components. Although encouraging short- and mid-term clinical outcomes and low revision rates have been reported using these implants, survivorship still is not comparable to total knee and hip arthroplasties, primarily due to implant loosening and subsidence (Michael et al., 2008).

A better understanding of total ankle arthroplasty kinematics is critical to improve operative techniques, prosthetic designs, and clinical outcomes. For example, kinematic analyses provide quantitative in vivo information to determine if an implant operates in accordance with its design objectives (Leardini et al., 2004). Moreover, abnormal kinematics and incongruency of the articular surface may cause increased contact pressure applied to the implant as shown in biomechanical in-vitro studies (Tochigi et al., 2005, Espinosa et al., 2010, Fukuda et al., 2010). This may result in pathologically increased polyethylene wear leading to component loosening and implant failure (Wood et al., 2009). Several studies have reported in vivo three-dimensional kinematics during the stance phase of walking (Conti et al., 2006, Leszko et al., 2008) and ankle dorsiflexion–plantarflexion (Komistek et al., 2000), but the numbers of the patients were relatively small, and incongruency of the articular surface was not fully investigated.

Additionally, understanding the difference between non-weightbearing kinematics and weightbearing kinematics is clinically important because recreating weightbearing activities is difficult during total ankle arthroplasty surgery, and therefore surgeons need to estimate weightbearing kinematics from the intraoperative non-weightbearing condition. Yamaguchi et al. (2009) reported weightbearing ankle kinematics are significantly different from the non-weightbearing kinematics in vivo, and this difference occurs possibly because the ankle kinematics are mainly determined by the tension of the surrounding ligaments in non-weightbearing activities (Leardini, 2001), while in weightbearing activities they are regulated by the articular surface geometry (Tochigi et al., 2006).

We have used the revised version of the TNK Ankle™ (Japan Medical Materials, Osaka, Japan) since 1990 (Fig. 1), and have reported good clinical results (Takakura et al., 2004) comparable to other second generation total ankle arthroplasty designs (Buechel et al., 2004, Claridge and Sagherian, 2009, Wood et al., 2009, Gougoulias et al., 2010). Results have been less favorable in patients with rheumatoid arthritis than in patients with osteoarthritis (Nagashima et al., 2004, Takakura et al., 2004). The TNK ankle is a two-component alumina ceramic prosthesis coated with beads and hydroxyapatite, and a high-density polyethylene insert is fixed on the tibial component. It has cylindrical joint surfaces, and the diameter of the talar surface is slightly smaller than that of the tibial surface, allowing the talar component to slide and rotate in addition to dorsiflexion/plantarflexion. Other features include the medial facet to sustain increased medial loads in ankles with primary varus osteoarthritic deformity, and the use of tissue engineered bone mounted on the upper surface of the tibial component to enhance bonding between the bone and implant (Tohma et al., 2006).

3D–2D model-image registration techniques are widely accepted methods for the accurate measurement of in vivo joint kinematics, (Banks and Hodge, 1996, You et al., 2001, Mahfouz et al., 2003) although few ankle studies have been reported (Yamaguchi et al., 2009). To begin this study, we quantified the accuracy and reproducibility of the 3D–2D model-image registration techniques for the TNK Ankle system. We then measured in vivo kinematics of the TNK Ankles during non-weightbearing and weightbearing dorsiflexion/plantarflexion activities using these techniques. From our clinical experience, we hypothesized: (1) in the non-weightbearing activity, incongruency of the articular surfaces would be observed, and (2) in the weightbearing activity, the loaded articular surfaces would better control motions and reduce the observed surface incongruencies.

Section snippets

3D–2D model-image registration

Three-dimensional implant models were obtained from the manufacturer, and an anatomic coordinate system was embedded in each implant (Fig. 1). For the tibial component, the mediolateral midline of the inferior surface of the implant was defined as the anteroposterior axis, the anteroposterior midline was defined as the mediolateral axis, and the intersection of the axes was the origin. The superoinferior axis was defined as the cross product of the two other axes. For the talar component, a

Accuracy and reproducibility

The average RMS differences between matching metal beads and matching implants were 0.14 mm for sagittal plane translations and 0.48° for all rotations (Table 1). The average interobserver RMS differences were 0.16 mm for sagittal plane translations, and 0.39° for all rotations. The ICCs were higher than 0.92 for all rotations and translations except for medial/lateral translation, in which the ICC was 0.66 (Table 1). The average intraobserver RMS differences were 0.14 mm for sagittal plane

Discussion

The purpose of this study was to measure kinematics of the two-component total ankle arthroplasty during non-weightbearing and weightbearing activities. We hypothesized that incongruency of the joint surface would occur with higher frequency in non-weightbearing activities than in weightbearing activities. At least one type of incongruency occurred in greater than 75% of the ankles. However, we were unable to find significant differences between the non-weightbearing and weightbearing joint

Conflict of interest statement

Three-dimensional CAD models of the TNK Ankle were supplied by Japan Medical Materials, Osaka, Japan. No funds were received in support of this study.

References (33)

  • N. Espinosa et al.

    Misalignment of total ankle components can induce high joint contact pressures

    Journal of Bone and Joint Surgery (American)

    (2010)
  • T. Fukuda et al.

    Impact of talar component rotation on contact pressure after total ankle arthroplasty: a cadaveric study

    Foot and Ankle International

    (2010)
  • N.E. Gougoulias et al.

    History and evolution in total ankle arthroplasty

    British Medical Bulletin

    (2009)
  • N. Gougoulias et al.

    How successful are current ankle replacements?: a systematic review of the literature

    Clinical Orthopaedics and Related Research

    (2010)
  • A. Hayes et al.

    Ankle morphometry on 3D-CT images

    The Iowa Orthopedic Journal

    (2006)
  • R.D. Komistek et al.

    A determination of ankle kinematics using fluoroscopy

    Foot and Ankle International

    (2000)
  • Cited by (0)

    View full text