Concepts on the Glenoid Fixation in Reverse Shoulder Prosthesis

Fig. 1

Glenoid morphology according to Walch’s classification [5]

Assessing the preoperative glenoid morphology and correcting it are fundamental for the surgical procedure since posterior glenoid erosion is not only a risk factor for glenoid loosening but also it results in worse function outcomes and pain [6–8]. In this situation, the theoretical goal of shoulder surgeons performing an arthroplasty is to restore native joint biomechanics by improving glenoid version and maintaining or restoring the glenohumeral joint line.

Biomechanics of Glenoid Retroversion and Posterior Bone Loss

Shoulder biomechanics is inevitably altered by glenoid bone loss, resulting in humeral head displacement and eccentric stresses placed on the glenoid component leading to polyethylene wear, component loosening, or instability. Posterior glenoid bone loss results in altered net humeral joint reaction forces which pass outside the effective glenoid arc, creating joint instability (Fig. 2). Bryce et al. [9] studying the relationship between glenoid wear and humeral head subluxation in their cadaveric biomechanical model demonstrated that subluxation is steadily present already with 2.5° of glenoid retroversion. The degree of bone loss and glenoid retroversion directly influence both glenohumeral forces and humeral head displacement; every 4° increase in retroversion resulted in a 2° shift of joint reactive forces away from the glenoid midline [8]. This condition causes approximately 0.5 mm of posterior humeral head displacement for every corresponding degree of glenoid retroversion. Altered joint forces cause humeral head subluxation which can lead to eccentric loading of the glenoid component, a mechanism described as “rocking-horse,” and associated to high tensile forces across the glenoid component-bone interface [9, 10].

Fig. 2

Principles of biomechanic of the shoulder with no glenoid defects

Farron et al. [11] using three-dimensional finite element analysis stated that retroversion of 20° created a posterior contact point on the glenoid, increasing stresses within the cement mantle and glenoid bone by 326 % and 162 %, respectively. Retroversion of just 10° resulted in an increase in micromotion at the bone-cement interface by >700 %, and they concluded that retroversion beyond this point should be corrected. Placing the glenoid implant in 15° of retroversion (in cadaveric shoulders), Shapiro et al. [12] evaluated the effects of a glenoid component version on joint biomechanics. This procedure significantly decreased the glenohumeral contact area, increased contact pressures, and decreased inferior and posterior glenohumeral forces.

Treatment Options

Eccentric Reaming

Eccentric reaming is a common procedure performed prior to component insertion with the aim to improve excessive glenoid retroversion. An excessive reaming can reduce the subchondral bone available for implant support, medialize the joint line, and allow cortical perforation of the polyethylene implant.

Walch et al. [13] found that motorized reaming was significantly associated with glenoid loosening for both subsidence and posterior tilt and so suggested that subchondral bone be preserved to provide sufficient osseous support to withstand the stresses experienced by the glenoid implant.

Many studies have been performed with the aim to define the limits of eccentric reaming in order to minimize the removal of subchondral bone while maximizing version correction. Attempting to correct 15° of retroversion, Gillespie et al. [14] found implant peg penetration or inadequate bone support in four of eight cadaveric specimens studied. Correcting even 10° of version, a significant decrease in anteroposterior glenoid diameter was found. Clavert et al. [15] reamed to neutral version five cadaveric scapulae in which they have previously created posterior glenoid defects and placed a pegged glenoid component. The result was one peg perforation in all five specimens and one fracture of the anterior glenoid rim leading the authors to conclude that if version exceeds 15°, the surgeon should consider alternatives to reaming the anterior aspect of the glenoid, such as posterior deficiency bone grafting.

Computer software has allowed investigators to simulate the effect that reaming has on glenoid component implantation.

Iannotti et al. [16], using a three-dimensional surgical simulator, compared ideal versus actual retroversion correction and came to the conclusion that retroversion of >19° would have been associated to peg perforation if ideal component placement had been performed.

Nowak et al. [17] considered a version <12° as optimal to implant a standard glenoid component, while version of >18° resulted in peg penetration. However, it is important to note that glenoid perforation after a short-term follow-up period is not correlated to adverse clinical effects or radiographic findings, lacking the literature of long-term follow-up studies.

In summary, an eccentric reaming is restricted by the available bone stock and should be limited to mild defects with no more than 10–15° of glenoid retroversion; an excessive reaming should be avoided to reduce the risk of loss of subchondral bone support, cortical perforation, and consequent implant loosening.

Glenoid Bone Grafts

When posterior glenoid bone loss is too excessive, bone grafting is a valid method to improve version, re-establish the joint line, and restore glenoid bone deficiency with the potential for biologic incorporation.

Bone grafting is a valid method when there is insufficient bone stock for component fixation or an inability to correct component position with glenoid reaming as it would result in an incorrect glenoid implant in cases of retroversion >15° [18–20]. The aim of bone grafting is that of improving version, re-establishing the joint line, and restoring glenoid bone deficiency with the potential for biologic incorporation. Problems connected to this procedure are nonunion, resorption, or subsidence, in addition to the technical demand of graft placement and fixation [19, 21–27].

Few studies evaluated the clinical and radiological outcomes of reverse shoulder arthroplasty using bone grafting for excessive glenoid retroversion.

Mizuno et al. [28] studied 27 reverse shoulder replacements performed for the treatment of primary glenohumeral osteoarthritis with a biconcave glenoid; retroversion (mean: 32°) and humeral head subluxation (mean: 87 %) were not such as to be corrected by asymmetric reaming. Ten patients required a bone graft if version could not be corrected to within 10° of neutral or when the baseplate surface contact was <80 %. Constant score increased from 31 to 76 points (p < 0.0001). In four (15 %) of 27 patients, a complication occurred, with three patients having neurologic issues and one patient having early glenoid loosening. At the latest follow-up evaluation (mean FU: 44 months), 25 patients (93 %) were either very satisfied or satisfied with their results. No radiolucent lines were observed around the central peg or screws; no recurrence of posterior instability was found. The authors concluded that reverse shoulder arthroplasty offers a viable solution for the treatment of severe static posterior glenohumeral instability and severe glenoid erosion. Wall et al. reviewed the results of reverse total shoulder arthroplasty in 240 patients (mean age: 72 years) according to different surgical indications [29]. Of those patients, 33 underwent reverse total shoulder arthroplasty because of severe posterior glenoid bone loss and posterior humeral head subluxation. The mean Constant score after a mean FU of 38 months passed from 24.7 to 65.1 points and mean shoulder flexion from 77° to 115°. A rapid loosening of both the graft and implant was found, and this patient needed surgery for conversion to a cuff tear arthroplasty.

Our Experience

In our practice, each patient submitted to reverse shoulder prosthesis undergoes preoperative evaluation with standard Rx examination (true AP and axillary view) and CT scan (with 3D reconstructions) in order to obtain a detailed surgical planning. In case of glenoid retroversion <15°, we perform an eccentric reaming in order to restore a correct joint congruity and the right glenoid version. If retroversion is >15°, we utilize a bone grafting using the humeral head bone (Fig. 3a).

Fig. 3

Bone grafting using the humeral head bone. Intraoperative phases (a–d)

Augmented Glenoid Components

In alternative to eccentric reaming and bone grafting, augmented glenoid components were designed.

Clinical and radiological outcomes regarding this technique are controversial. Rice et al. [30] reviewing 14 shoulders treated with an asymmetric wedge-shaped posteriorly augmented glenoid component (mean FU: 60 months) found only two clinical unsatisfactory results. However, more than half of the glenoid components demonstrated radiolucent lines, and one-third demonstrated moderate or severe posterior glenohumeral subluxation, although no revision surgery was performed.

Rice et al. concluded that the contribution of the modified glenoid component to overall correction of glenoid bone wear and humeral subluxation seemed marginal, and use of this implant was discontinued.

In the last years, we have seen the development of a stepped, posteriorly augmented glenoid design that places the component perpendicular to the vector of joint forces and allows for improved biomechanical properties [31–33]. Iannotti et al. [33] compared the resistance to anterior glenoid lift off of four different all-polyethylene augmentation designs, under both compressive and eccentric loads. The stepped glenoid resulted to have lower initial and final lift off values compared with the augmentation designs, although not all reached significance.

Glenoid implant augmentation can improve glenoid version while preventing implant perforation, joint line medialization, and subchondral bone loss. However, more clinical studies are needed. Furthermore, augmented glenoid implantation is technically demanding procedure; a precise creation of a glenoid bone bed to seat the augmented component is essential. High rate of micromotion and the risk of loosening are reported [34].

Lateralization of the Center of Rotation in Reverse Shoulder Prosthesis

The Grammont-style reverse shoulder prosthesis had two biomechanical principles: medialization of the glenohumeral center of rotation together with the lowering of the humerus [35]. These principles reduce torque on the glenoid component and increase the deltoid lever arm, overcoming the weakness or the absence of rotator cuff tendons [36, 37].

Many studies reported problems and complications attributed to this design [38–43]. The scapular notching is the most frequent, ranging from 50 to 96 % in postoperative radiograms [44–47] (Fig. 4). It consists of the inferomedial impingement of the humeral component against the scapular neck during arm adduction and rotation responsible for bone erosion and polyethylene wear. Prosthetic instability is a further complication consequent to humeral medialization because of glenohumeral impingement and the poor soft tissue tension; it has been observed in 3–6 % of cases [36, 38, 48]. Finally, patients submitted to reverse shoulder prosthesis complain of cosmetic concerns related to the loss of their normal shoulder contour [35, 36] (Fig. 5).