Anatomic studies have moved from simply descriptive [1–5] to pathomechanical explanation of several shoulder disorders and to specific surgical techniques development.
Recently, reversed shoulder prosthesis design has gained popularity in the management of massive cuff tears associated with glenohumeral arthritis, even though results refer short and mid term follow-up [17–20]. The reversed design is, however, cause of concern because of the fixation of its components, specially the glenoid component, as well as the potentially rate of complications such as component loosening [22, 23]. When this study was carried out, Delta III (DePuy International Ltd, Leeds, England) was the unique reversed shoulder prostheses available in Spain. Primary fixation of the glenoid component in Delta III prosthesis relays on a central stem that should be located into the glenoid body, and four screws. Delta III glenoid component present a fixed – angle orientation of the superior and inferior screws (70° between glenoid surface and screw) and a free-angle orientation for the anterior and posterior ones. Superior and inferior screws should be located in divergence, directing the superior one to the base of the coracoid process and the inferior one to the upper posterior column of the scapula. The anterior and the posterior screws should be placed into the body of the glenoid. In addition, the superior and inferior holes of the glenoid component to insert the superior and the inferior screws are positioned in line. It is to be supposed that fail in peg and/or screws location may affect stability of the implant as it has been shown in previous studies . It is also to be supposed that the more the screws run inside the bone, the better fixation will be obtained.
The present study has found two different types of scapulas as far as glenoid surface to upper posterior column of the scapula angle is concerned, and although no attempt has made to measure the 3-D bone coverage of the inferior screw in the different types of scapulas, type I, which is the most frequent (61,43% in the three-dimensional computed tomography scapulas and the 71,30% in the cadaveric group), determines a mean angle of 50°–52°, meaning that if the inferior screw has a prefixed position of 70°, it will be poorly placed into bone because the different orientation of the screw and the lateral border of the scapula determining thus less bony coverage. Type II determines a mean angle of 62°–64°, meaning that the prefixed screw direction better fits in the lateral border of the scapula leading to a more bony coverage of the screw.
Taking into account the coronal plane, this study demonstrates that the center of the coracoid process and the upper posterior column of the scapula are not in line, moreover, the center of the base of the coracoid process is located a mean of 18,25° anterior with regard to the major craneo-caudal glenoid axis and the upper posterior column of the scapula is located 8° posterior to this axis, giving a mean of 10°of difference. In the Delta III glenoid component the holes for the superior and inferior screws are placed in line, that means that if the inferior screw is properly located in the posterior column of the scapula, the superior screw is directed to the posterior part of the base of the coracoid process, giving thus a poor placement into bone.
The inferior part of the glenoid in the anterior face as well as in the posterior can be divided into two grossly different length necks. In the so called "short-length" glenoid neck, the glenoid articular surface is close to the upper posterior column of the scapula and allows inferior screw to reach easily to the posterior column of the scapula. In the so called "long-neck" glenoids, the glenoid articular surface is located far from the upper posterior column of the scapula and determines that if the inferior screw has a prefixed angle it may conduct the screw through the glenoid neck instead of into the upper posterior column of the scapula, giving thus a short bone in through location.
All the anatomical variations described advice for major changes in the metaglene component of the reversed prostheses to improve bone fixation. Inferior and superior screws may have to have a minimum of 10° of free orientation to adapt in the upper part of the posterior column of the scapula and be able to fit both scapular types. The 10° free orientation may also help to better place the superior screw into the base of the coracoid process.
One major cause of concern regarding the glenoid component fixation is the formation of a notch at the inferior pole of the scapula as a result of the contact of the medial part of the humeral component and the glenoid during adduction. Recently, to avoid this complication, the implantation of the glenoid component extending beyond the inferior glenoid rim has been proposed . Several preoperative measures have to be done before deciding to extend beyond the inferior glenoid rim the glenoid component to assess the type of scapula and the length of the inferior glenoid neck. Positioning inferiorly the glenoid component in case of a "long-neck" glenoid may determine the screw run through the glenoid neck instead of into the upper posterior column, and in the same way, in a type I scapula the more inferior the glenoid component is located, the less chance to get the lateral border of the scapula with the inferior screw in an angle-fixed component design. Avoiding scapular notch by extending beyond the inferior glenoid rim the glenoid component positioning requires glenoid component to be modified in order to allow variation in the direction of positioning the inferior screw.
The different scapular morphologies founded in this study advise to individualize screws positioning in the glenoid component to adjust them to the anatomy present in each particular case. Three-dimensional computed tomography of the scapula constitutes an unvalued source when planning surgery with reversed prostheses for better understanding the particular scapular morphology of each individual case and the adjustments to be done to better place glenoid component. Prefixed angle screws leads several times to a decrease of bone coverage, so adjustments have to be done to change direction depending on the type of angle between the glenoid surface and the upper posterior column of the scapula, the different location of the base of the coracoid process and the upper posterior column of the scapula and the length of the neck of the glenoid. Maybe two different implant types of glenoid component should be considered to address different glenoid neck lengths.
Recently Codsy et al have also stressed on the importance of the glenoid vault and the integrity of the subchondral bone to obtain proper fixation of the glenoid component and even though they find in normal glenoids a uniform morphology of the glenoid vault, 5 different sizes are defined to fit an average clinical population.
Is to be believed that bony coverage of the screw may affect stability if the implant although many other parameters are involved in glenoid component stability such as bone quality around screw, orientation of the screw with respect to the forces, etc.
No relationship has been found between the different scapular morphologies and sex or age in the three-dimensional computed tomography group. No correlation has been found between the different types of scapulas as far as glenoid surface and posterior column of the scapula angle is concern and glenoid neck length in anterior or posterior face. No correlation has been found between the length of the neck in the anterior face of the glenoid and the length of the neck in the posterior face.
Kappa studies revealed a moderate to substantial agreement of anterior and posterior neck lengths which means a reasonable level of concordance and reproducibility of these measures, and a level almost perfect in the analysis of the type of angle of glenoid surface and upper posterior column of the scapula.