Abstract
Introduction
Patellar resurfacing affects patellofemoral (PF) kinematics, contact mechanics, and loading on the patellar bone. Patients with total knee arthroplasty (TKA) often exhibit adaptations in movement patterns that may be linked to quadriceps deficiency and the mechanics of the reconstructed knee [1]. Previous comparisons of PF kinematics between dome and anatomic resurfacing have revealed differences in patellar sagittal plane flexion [2], but further investigation of PF joint mechanics is required to understand how these differences influence performance. The purpose of this study was to compare PF mechanics between medialized dome and medialized anatomic implants using subject-specific computational models.
Methods
A high-speed stereo radiography (HSSR) system was used to capture 3D sub-mm measurement of bone and implant motion [3]. HSSR images were collected for 10 TKA patients with Attune® (DePuy Synthes, Warsaw, IN) posterior-stabilized, rotating-platform components, 5 with medialized dome and 5 with medialized anatomic patellar components (3M/7F, 62.5±6.6 years, 2.2±0.6 years post-surgery, BMI: 26.2±3.5 kg/m2), performing two activities of daily living: knee extension and lunge (Figure 1). Relative motions were tracked using Autoscoper (Brown University, Providence, RI) for implant geometries obtained from the manufacturer. A statistical shape model was used to predict the patella and track motions [4].
Subject-specific finite element models of the experiment were developed for all subjects and activities [5]. The model included implant components, patella, quadriceps, patellar tendon, and medial and lateral PF ligaments (Figure 2a). While tibiofemoral kinematics were prescribed based on experimental data, the PF joint was unconstrained. A constant 1000N quadriceps load was distributed among four muscle groups. Soft tissue attachments and pre-strain in PF ligaments were calibrated to match experimental kinematics [5]. Model outputs included PF kinematics, patellar and contact force ratios, patellar tendon angle, and moment arm.
Results and Discussion
Load-bearing activities presented larger variations in PF kinematics and mechanics between dome and anatomic subjects. Consistent with previous findings [2], patients with medialized anatomic geometry achieved greater patellar flexion than those with the medialized dome during lunge (16±3° from 40–100° knee flexion), and demonstrated PF kinematics closer to that of the natural knee (Figure 2b) [6]. Dome subjects experienced greater PF internal-external range-of-motion compared to the anatomic subjects (8±5°).
Model PF kinematics closely replicated the experiment with average root-mean-square differences of flexion-extension<5°, internal-external<3°, and medial-lateral<2 mm. Dome subjects demonstrated larger contact force ratios than anatomic, but presented smaller patellar force ratios in deep flexion (–60°). Smaller PF flexion angles in dome implants likely influenced the increase in contact force ratio, which may increase implant forces and decrease quadriceps efficiency. In contrast, the flexed position of the anatomic design distributed load to the patellar tendon at deeper flexion angles, which may improve extensor function (Figure 3). The current modeling framework can inform implant designers on the impact of articular geometry on quadriceps efficiency.
