Abstract
Introduction
A large number of total knee arthroplasty (TKA) patients, particularly in Japan, India and the Middle East, exhibit anatomy with substantial proximal tibial torsion. Alignment of the tibial components with the standard anterior-posterior (A-P) axis of the tibia can result in excessive external rotation of the tibial components with respect to femoral component alignment. This in turn influences patellofemoral (PF) mechanics and forces required by the extensor mechanism. The purpose of the current study was to determine if a rotating-platform (RP) TKA design with an anatomic patellar component reduced compromise to the patellar tendon, quadriceps muscles and PF mechanics when compared to a fixed-bearing (FB) design with a standard dome-shaped patellar component.
Methods
A dynamic three-dimensional finite element model of the knee joint was developed and used to simulate a deep knee bend in a patient with excessive external tibial torsion (Figure 1). Detailed description of the model has been previously published [1]. The model included femur, tibia and patellar bones, TKA components, patellar ligament, quadriceps muscles, PF ligaments, and nine primary ligaments spanning the TF joint. The model was virtually implanted with two contemporary TKA designs; a FB design with domed patella, and a RP design with anatomic patella. The FB design was implanted in two different alignment conditions; alignment to the tibial A-P axis, and optimal alignment for bone coverage. Four different loading conditions (varying internal-external (I-E) torque and A-P force) were applied to the model to simulate physiological loads during a deep knee bend. Quadriceps muscle force, patellar tendon force, and PF and TF joint forces were compared between designs.
Results
The RP design demonstrated consistently lower medial-lateral (M-L) force at the PF joint than the FB design, with greater differences between designs in later flexion once the patella was engaged in the sulcus groove; root-mean-square (RMS) differences in M-L force averaged 50 N less in the RP design throughout the flexion cycle, and 70 N less after 45° flexion (Figure 2). The FB design aligned for optimal bone coverage demonstrated 15% higher M-L forces than the FB design aligned with the tibial A-P axis. RMS load required by the quadriceps muscle was 60 N lower with the RP design than the FB design throughout the cycle (Figure 2).
Discussion
Comparing a RP design with an anatomic patellar component and a FB design with a domed patellar component, the RP design demonstrated lower M-L PF joint and soft-tissue extensor mechanism forces. Differences were more pronounced under conditions of high I-E torque where the RP design accommodated large relative TF rotation. Differences in FB alignment resulted in substantially different PF M-L forces; when the FB component was mal-aligned with respect to the tibial A-P axis (and the line-of-action of the patellar tendon) the resulting M-L PF force was increased. The RP design reduced the demands on the extensor mechanism and loads on the PF joint and facilitated better coverage of the resected tibial bone surface.
