Abstract
Purpose: Glenoid component loosening is a common reason for failed total shoulder arthroplasty. Multiple factors have been suggested as causes for component loosening including asymmetric loading of the glenoid prosthesis by the humeral head (rocking horse phenomenon). A novel technique was employed to measure in-vitro strain in the subchondral bone adjacent to a cemented all polyethylene pegged glenoid prosthesis. The purpose of the study was to develop and validate a testing protocol to investigate load transfer in the polyethylene glenoid implant and bone construct.
Method: Eight polyethylene components were implanted using standard cementing techniques in eight cadaveric specimens. Loading was performed with a pneumatic actuator capable of applying loads at various angles. A dynamic 10 N/s force was applied for a total of 15 seconds producing a maximum force of 150N at angles of 0, 10, 20, 30, 40 and 50o. Strain gauges were placed around the implant 1mm proximal to the bone-cement interface at the four quadrants. The humeral head was simulated with a custom steel ball with a non-conforming diameter in relation to the prosthesis that is typical in total shoulder arthroplasty.
Results: During pure compressive loading, tension was observed in the superior and inferior quadrants of the glenoid. Superior and inferior loading caused increasing same side (ipsilateral) tension, occurring from 0 to 30o and 0 to 20o, respectively. Compression was recorded superiorly when loading was applied at 40o and 50o in the superior direction while contralateral tension was recorded in the inferior gauges. Strain measurements were less consistent in the anterior and posterior glenoid quadrants and varied between tension and compression.
Conclusion: Tension measurements in the ipsilateral direction at lower angles were unexpected. This observation differs from the previous assumption that applied loads at relatively perpendicular angles to the implant should dissipate as compression. Tension at the bone cement interface is unfavorable. The identification of tension in some quadrants of the implant in this study, therefore, may have revealed a mechanism of implant loosening. Our data support the previously described rocking horse phenomena and also illustrate a new umbrella type effect of polyethylene flexure, which causes the periphery of the glenoid implant to flex upwards superiorly and inferiorly. These findings have the potential to influence future designs of total shoulder arthroplasty perhaps leading to increased implant survival.
Correspondence should be addressed to: COA, 4150 Ste. Catherine St. West Suite 360, Westmount, QC H3Z 2Y5, Canada. Email: meetings@canorth.org
