Once possible causes for baseplate failure were identified, focus turned toward understanding the factors involved in baseplate fixation. To this end, the following 4 studies were undertaken involving multiple institutions with their ultimate goal of improving patient outcomes in RSA.

The purpose of this study was to determine the compressive force present between the baseplate and the glenoid for two different types of baseplates: (1) a Grammont style with a smooth central post and (2) an RSP style with a 6.5-mm integrated central cancellous-type screw. Both devices were implanted to manufacturer’s specifications including peripheral screws tightened to 60 in./lb. Film-type force transducers were positioned between the baseplates and a bone analog to measure compressive force. Results showed a tenfold increase in compressive force when a central cancellous-type screw was used instead of a smooth central peg (2000 N vs. 200 N, respectively). This study showed the benefit of having a baseplate with a central cancellous-type screw since it provided a tenfold increase in compressive force over a smooth central post secured only by peripheral screws.

The purpose of the study was to determine resistance to shear force for two different types of baseplates: (1) a Grammont style with a smooth central post and (2) an RSP style with a 6.5-mm integrated central cancellous-type screw. Both devices were implanted to manufacturer’s specifications, but without the use of peripheral screws. A shear force of 756 N was applied to the rim of each baseplate at a rate of 150 N/s until the yield strength of the bone analog was reached (Fig. 4.4). Results showed an almost threefold increase in resistance to shear force when a central cancellous-type screw was used instead of a smooth central peg (631 N vs. 269 N, respectively). This study added to the evidence for having a baseplate with a central cancellous-type screw.

The purpose of this study was to quantify baseplate micromotion when 3.5-mm non-locking and 4.5-mm locked peripheral screws were used for two different types of baseplates when implanted in a bone analog: (1) a Grammont style with a smooth central post and (2) an RSP style with a 6.5-mm integrated central cancellous-type screw (Fig. 4.5). Both devices were implanted to manufacturer’s specifications. The Grammont-style baseplate was implanted with two 3.5-mm non-locked and two 4.5-mm locked peripheral screws, while the RSP was implanted using only 3.5-mm non-locked peripheral screws. A shear force of 756 N was applied to the rim of each baseplate at a rate of 150 N/s until the yield strength of the bone analog was reached. Results showed a significantly higher resistance to shear force when a central cancellous-type screw was used instead of a smooth central peg (367 ± 23 µm vs. 310 ± 20 µm, respectively). This study showed the benefit of using a central 6.5-mm compressive cancellous screw compared to a central post, even when using compressive peripheral screws combined with locked peripheral screws.

The purpose for this study was twofold:

This study, as with the previous study, showed the benefit of using 5.0-mm locked screws over 3.5-mm non-locked screws. The information presented in this section helped modify surgical technique by recommending the use of 5.0-mm locked peripheral screws when implanting an RSP baseplate.

Although baseplate failures have been mostly addressed by the use of 5.0-mm locked screws in place of the 3.5-mm non-locking screws, the few failures that do occur remain a concern with the RSP design. Observations by the design surgeon guided new biomechanical studies into how the position of the baseplate might affect the survivability of the baseplate. The following studies aided in the understanding of how altering design parameters and surgical techniques could affect ROM, glenoid forces, and baseplate displacement.

The purpose of this study was to quantify the effects of baseplate tilt (15° superior tilt, 0° neutral tilt, and 15° inferior tilt) on baseplate micromotion and compressive forces seen by the glenoid. An apparatus was developed to simulate 60° of glenohumeral abduction in the scapular plane. Superior and inferior baseplate micromotion was measured via a linear voltage displacement transducer (LVDT), and glenoid forces were measured by transducers that were placed underneath the baseplate (between the baseplate and a bone analog) in the superior and inferior positions (Fig. 4.8). Results showed less overall micromotion and more evenly spread compressive forces when the baseplate was positioned with a 15° inferior tilt. The results from this study correlated well with what was being seen clinically.

The purpose of this study was to compare the effects of inferior tilt (0°, 3°, 6°, 9°, 12° and 15°) on strain between RSP and Grammont-style baseplates. A finite element model was developed to simulate the RSP and Grammont-style devices in a bone analog substrate (Fig. 4.9). Results showed greater baseplate motion in the superior and inferior directions with the Grammont-style device when compared to the RSP design. Additionally, increasing the inferior tilt of the baseplate in either design showed a decrease in overall motion. The results from this study corroborated the evidence found in the previous biomechanical study and ultimately helped to alter clinical practice by advising surgeons to implant the baseplate with up to a 15° inferior tilt.

The purpose of this study was to compare the effects of varying glenosphere COR on (1) baseplate micromotion in two separate studies, (a) biomechanical and (b) finite element model, and (2) reaction moments from a numerical model. Seven different glenospheres (RSP: 32, 36, and 40 mm neutral and 32 mm −4, 36 mm −4 and 40 mm −4; Grammont: 36 mm) with five different CORs (0 mm [40 mm −4], 1 mm [36 mm—Grammont], 2 mm [36 mm −4], 4 mm [40 mm neutral], 6 mm [36 mm neutral and 36 mm −4], and 10 mm [32 neutral]) were tested.

The results from the analytical and finite models showed good agreement with both showing 30 µm or less variability in baseplate micromotion by varying glenosphere designs. The analytical model showed a general increase in moments at the baseplate as both coefficient of friction and COR offset were increased. These results showed minimal differences in motion between all the designs of glenosphere, as long as good initial glenoid fixation was achieved.

The purpose of this study was to classify glenoid morphology and examine how variances in this morphology can affect the fixation of glenoid components. The morphology of 216 glenoids was divided into normal and abnormal groups according to six anatomic parameters [height, width, version, inclination, distance from coracoid to the glenoid (C-G), and distance from the acromion to the glenoid (A-G)] and 2 surgical parameters (central/standard centerline and spine centerline for central axis device fixation). The abnormal groups were further subdivided by erosion site (posterior, superior, global, and anterior) (Fig. 4.12). Results showed a decrease in the standard centerline distance in abnormal glenoids (19.6 ± 9.1 mm) versus normal glenoids (28.6 ± 4.1 mm, p < 0.0001). Conversely, the spine centerline showed a larger distance in the abnormal glenoids (34.9 ± 17.0 mm). Additionally, peripheral screw placement was reduced by 42 % in abnormal glenoids, with the placement limited to the anterior and inferior quadrants. This study showed the significant effect that abnormal glenoid morphology had on both baseplate and peripheral screw placement, thus altering surgical technique.

The purpose of this study was to (1) quantify anatomic measurements and glenohumeral relationships and determine whether they change linearly based on patient size, and (2) determine how patient shoulder prosthetic size mismatch can affect soft tissue strain and the middle deltoid moment arm. The anatomic measurements and glenohumeral relationships of 92 patients undergoing RSA, without bony deformities, were evaluated using 3D CT reconstructions. The anatomic measurements were as follows: coronal plane head diameter, thickness of humeral head, neck-shaft angle, articular arc angle, superior humeral head to greater tuberosity distance, lateral humeral contact point to greater tuberosity distance, axial plane head diameter, upper arm length, glenoid height, superior glenoid width, and inferior glenoid width (Fig. 4.13). The glenohumeral relationships were as follows: lateral distance to greater tuberosity, lateral distance of greater tuberosity to acromion, superior distance of greater tuberosity to acromion, lateral distance to coronal head center, lateral distance to axial head center, superior subacromial space, and moment arm. Results showed that humeral head size, glenoid width, lateral offset, and moment arm all independently increased linearly (r² ≥ 0.92), but the rate of increase varied (range from 0.59 to 1.92). Glenoid height, coronal humeral head diameter, and gender predicted greater tuberosity position within 1.09 ± 0.84 mm of actual position in 90 % of the studied population. Results also showed that patient shoulder prosthetic size mismatch can have an adverse effect on soft tissue strain and middle deltoid moment arm. Strains increased from 3 to 7 % when a small shoulder was implanted with a small implant versus a large implant, respectively. Moment arms were increased by 35 % when a small shoulder was implanted with a large implant, while the moment arm of a large shoulder with a small implant only increased the moment arm by 25 %. This study showed that differences in shoulder measurements and glenohumeral relationships can help group patients by size, sex, and certain morphological measurements. The clinician should be cognizant of choosing the correct size implants based on patient size and should avoid large deviations in non-anatomic reconstructions.