Abstract
Implant designs for hip and knee arthroplasty have undergone a continual improvement process, but development of implants for total elbow arthroplasty (TEA) have lagged behind despite the marked mechanical burden placed on these implants. TEA is not as durable with failure rates approaching thirty percent at five years. The Coonrad-Morrey (Zimmer, Warsaw, IN), a linked design, remains the standard-bearer, employing polyethylene bushings through which a metal axle passes. A common failure mode is bushing wear and deformation, causing decreased joint function as the bushing-axle constraint decreases and osteolysis secondary to release of large volumes of wear debris.
Improving upon this poor performance requires determining which factors most influence failure, so that failure can be avoided through design improvements. The approach integrates clinical observations of failed TEAs with implant retrieval analysis, followed by measurements of loads across the elbow for use in stress analyses to assess the performance of previous designs, and, finally, new design approaches to improve performance.
Examination of the clinical failures of more than seventy Coonrad-Morrey TEAs revealed patterns of decreased constraint and stem loosening. Implant retrieval analysis from more than thirty of these cases showed excessive bushing deformation and wear and burnishing of the fixation stems consistent with varus moments across the joint.
To determine loads across the elbow, motion analysis data were collected from eight TEA patients performing various activities of daily living. The kinematic data were input into a computational model to calculate contact forces on the total elbow replacement. The motion that produced the maximum contact force was a feeding motion with the humerus in 90° of abduction. For this motion, the joint reaction forces and moments at the point of maximum contact were determined from a computational model.
We applied these loads to numerical models of the articulating bushings and axle of the Coonrad-Morrey to examine polyethylene strains as measures of damage and wear. Strain patterns in response to the large varus moment applied to the elbow during feeding activities showed extensive plastic deformation in the locations at which deformation and wear damage were observed in our retrieved implants (Fig. 1).
Finally, we examined a new semi-constrained design concept intended to meet two goals: transfer contact loads away from the center of the joint, thus allowing contact to provide a larger internal moment to resist the large external varus moment; and reduce polyethylene strains by utilizing curved contacting surfaces on both the axle and the bushings (Fig. 2). After a sensitivity analysis to determine optimal dimensional choices (e.g., bushing and axle radii), we compared the resulting polyethylene strains between the Coonrad-Morrey and new design at locations that experienced the largest strains (Fig. 3). Substantial decreases were achieved, suggesting far less deformation and wear, which should relate to marked improvements in performance.
Currently, we are incorporating this new design concept, along with alterations in stem design achieved from examination of load transfer at the fixation interfaces based on the same loading conditions, to achieve an implant system intended to improve the performance of TEA.
