The glenohumeral joint can be classified as an enarthrodial ball and socket joint. It allows polyaxial motion of the arm relative to the scapula. These motions can be described relative to the scapular plane in the following manner: humeral rotation around the sagittal axis is specified as abduction/adduction, rotation around the frontal axis, where the axis is directed in a medial–lateral direction, named flexion/extension, and rotational movement around a longitudinal-humeral axis is referred to as external/internal rotation.

Even though articular surfaces are more or less congruent, glenohumeral joint kinematics do not fully equate enarthrodial kinematics. Exact determination of this kinematics has been subject to numerous studies. However, some controversy exists, and results are difficult to compare due to the variety of methodological approaches. Recently, development of open MRI and three-dimensional biplanar fluoroscopy techniques promise a highly increased accuracy and more realistic representation of the actual in vivo situation. Sahara et al. found an overall 1.1 mm inferior humeral translation during abduction using a vertically open MRI-technique. In the anterior–posterior direction, there was a 2.4 mm anterior translation while abduction from 0° to 90° and a posterior translation of −1.4 mm in abduction angles of 90°–150° (Fig. 2.3) [30]. Gibhart et al. found slightly bigger values with maximum superior–inferior translation of 4.2 ± 2.3 mm and 5.1 ± 1.1 mm for anterior–posterior translation, respectively.

These kinematics are highly dependable on the interaction of the rotator cuff, the deltoid and the scapulothoracic muscles. The anterior (SSC) and posterior (ISP/Teres minor) cuff act as a transverse force couple generating compressive forces. These glenohumeral joint reaction forces reach 45–90 % of total body weight, with the maximum forces seen in full abduction position. As a result, the humeral head is maintained centrally in the glenoid fossa, unaffected by the superior translational forces of the deltoid muscle [31]. The superior cuff (SSP) does also contribute to concavity compression, especially at early abduction [32]. However, with isolated supraspinatus tears, maximum joint compression forces remain unchanged, while tears of the anterior or posterior rotator cuff result in significantly decreased compression forces [33]. This decrease in compression forces leads to the inability to withstand the superior translational forces of the deltoid and superior migration of the humeral head might occur.

Motion of the arm is initiated by both the rotator cuff and the deltoid muscle. The muscle’s capacity to generate a rotational force can be estimated by measurement of their moment arm respective to the particular axis. The deltoid muscle with its broad origin and big muscle mass is an effective abductor, external and internal rotator. When the arm is at the neutral, resting position, the lateral (middle) part of the deltoid has an abduction moment arm of about 16 mm, which is considerably smaller compared to the SSP abduction moment arm in this position (23 mm). However, with increasing abduction, this moment arm increases to 30–35 mm, making him the strongest abductor in this position. The bigger SSP moment arm at the resting position underlines its importance in initiating the abduction motion [34].

Humeral rotation is initiated by the anterior and posterior rotator cuff and deltoid portions. The subscapularis has an internal rotational moment arm of about 23 mm in the resting position making him the strongest internal rotator. Posteriorly, the Teres minor has a 17-mm external rotational moment arm and the infraspinatus 24-mm external rotation moment arm, respectively [35, 36]. The anterior and posterior subregions of the deltoid muscle contribute to rotational motion to a lesser degree in healthy shoulders, as their rotational moment arms are smaller with 4-mm external rotation moment arm for the posterior regions and 5-mm internal rotation moment arm for the anterior regions, respectively [37].

While the measurement of moment arms can give a rough estimation about the muscles capacity to generate rotational forces, other factors such as muscle tension and integrity should not be underestimated.

In RSA, the center of rotation (COR) is shifted medially and inferiorly. This shift affects biomechanical properties of the deltoid muscle and the remaining rotator cuff in various ways: First of all, the deltoid muscle gains tension. This tension is mandatory for prosthetic stability after RSA and does also increase sufficiency of the deltoid when initiating abduction. The total lengthening of the deltoid sums up to about 20 % compared to the normal shoulder [38]. In shoulders with cuff tear arthropathy, where there is a substantial cranialization of the humerus, this effect might even be more pronounced. The medialization of the COR does also affect the deltoids abduction moment arm: While in healthy shoulders, an overall moment arm in respect to the sagittal axis of 16–35 mm can be seen [37, 39], it increases to 50–60 mm following RSA [38, 39]. Additionally, the amount of deltoid muscle fibres positioned lateral to the COR, which therefore have the capacity to create abductional forces increases with medialization of the COR [40]. On the other hand, a decrease of fibres positioned antero- or posteromedial to the COR, which have the potential to create internal or external, rotational forces in healthy shoulders, can be seen.

Analogous to the COR, the humerus with the insertion sites of the rotator cuff is shifted medially and inferiorly in conventional RSA. These changes have a significant influence on the biomechanical properties of the remaining rotator cuff. In patients with cuff tear arthropathy, supra- and infraspinatus seem to be the most commonly involved, whereas teres minor and subscapularis will often remain intact [41, 42]. Even though these muscles are strong external and internal rotators in healthy shoulders and their integrity might be maintained after RSA, humeral rotation often remains compromised or even decreases postoperatively [41, 43, 44]. While the decrease of deltoid fibres with rotational potential after RSA might be one contributing factor to these findings, a change to the biomechanical properties of the remaining rotator cuff has been identified as the other. Following conventional RSA, rotational moment arms of Teres minor show a significant decrease of up to 25 %, whereas the SSC rotational moment arms decrease by up to 36 % [36]. Moreover, muscle tension of both muscles, especially at the resting position of the arm, declines. The overall decrease of muscle length after conventional RSA can be as high as 20 mm for the teres minor and 16 mm for the subscapularis. With increasing humeral abduction, the reduced muscle tension resolves, and the distal muscle segments might even have a minimal increased tension at abduction of 60° or more [36].

One potential way to address these disadvantageous biomechanical properties is to shift the COR laterally. Lateralization can be achieved by implant design or by interposition of an autologous bone block between the glenoid and the baseplate [2, 45]. While with these efforts a lateralization of 7–10 mm in comparison with the conventional RSA is achieved, the COR is still shifted medially compared to normal shoulders. In lateralized RSA, rotational moment arms are preserved and no significant decrease of muscle tension for either SSC or teres minor is observed (Figs. 2.4 and 2.5). Only the very distal regions of the SSC show a slight decrease of muscle tension when the arm is in a 0° abduction position (139 mm vs. 145 mm) [46].