Introduction: Potential clinical advantages for using reverse shoulder prostheses, such as enhanced stability or function, can only be realized if adequate glenoid component fixation is achieved. This study evaluates fixation of uncemented reverse glenoid components during physiologic loading, including radiographic assessment of in vivo component position. The relationships between initial fixation, glenoid component design (offset and screw geometry) and baseplate position were established using in-vitro biomechanical tests.

Methods: Clinical: Twelve patients received Reverse Shoulder Prostheses (RSP, Encore Medical). Six patients had good outcomes (ASES score > 95), whereas the remaining six patients had glenoid loosening. Patient follow-up radiographs were digitized and glenoid base-plate position relative to the scapular spine was measured using a computer-guided goniometer.

Mechanical Tests: RSP glenoid components were inserted in-vitro into synthetic bone foam blocks with material properties similar to human cancellous bone. Baseplates were secured using the RSPs central screw and either four 3.5 mm standard cortical screws in countersunk peripheral holes or four 5.0 mm diameter screws in threaded peripheral holes to fully capture the screw in the baseplate. Glenosphere lateral offset was 27 mm (neutral) or 23 mm (reduced). Angled baseplate positions of 15 superior, 0, and 15 inferior were tested. Loads were applied to the glenoid components through the polyethylene humeral component, consistent with physiologic forces measured at the shoulder joint during activity. Component motion and contact forces at the baseplate-foam interface were measured during cyclic loading using a displacement transducer and force transducers attached to the underside of the glenoid base-plates. Data were analyzed using ANOVA and t-tests.

Results: The mean baseplate-to-scapular spine angle on the clinical radiographs was 84.5 for failed prosthesis, while those that did not fail had a significantly smaller (inferior tilt) mean angle of 73.4 (p< 0.05). Motion and forces at the baseplate-foam interface were lowest with a 15 inferior baseplate position. Peripheral screw type (p< 0.05), but not offset (p> 0.05), significantly affected baseplate motion. Fixation with 5.0 mm captured screws reduced the average baseplate motion by 21% to 32% compared to the 3.5 mm screws.

Discussion: Changing the inclination angle or type of fixation screw affects clinical outcome and the base-plate motion and interface stress. Inferior baseplate tilt resulted in more even force distribution beneath the baseplate, a decreased force magnitude, and lower baseplate motion during physiologic loading. Fixation with 5.0 mm captured screws reduced baseplate motion compared to 3.5 mm screws. Obtaining similar results in vivo partially depends on surgical baseplate and screw placement and the patients glenoid bone stock.