Management of Humeral and Glenoid Bone Defects in Reverse Shoulder Arthroplasty

Lisa G. M. Friedman, MD, MA; Grant E. Garrigues, MD


J Am Acad Orthop Surg. 2021;29(17):e846-e859. 

In This Article

Glenoid Bone Loss

Glenoid bone loss is not an uncommon problem and can stem from a variety of etiologies. Primary glenohumeral osteoarthritis can cause glenoid bone loss. The Walch classification takes into account glenoid morphology, including glenoid wear patterns (Figure 4).[23] Rotator cuff arthropathy can also lead to glenoid bone loss, with multiple classifications, including those of Favard[24] and Sirveaux[25] (Figure 5 and Figure 6). Finally, glenoid bone loss is often encountered in revision arthroplasty. This can be caused simply by the implantation, which can compromise glenoid bone stock with reaming, peg holes, keel punch, and thermal necrosis from the polymethyl methacrylate cement. In addition, loosening of the glenoid implant can induce an osteolytic response that further compromises the bone within and surrounding the glenoid vault. The Antuna classification takes into account both the location of wear and its severity (Figure 7).[26]

Figure 4.

Illustration showing the Walch classification of glenoid morphology in primary glenohumeral osteoarthritis.23

Figure 5.

Illustration showing the Favard classification of glenoid bone erosion in the setting of rotator cuff arthropathy.25

Figure 6.

Illustration showing the Sirveaux classification for glenoid notching in reverse total shoulder arthroplasty.25

Figure 7.

Illustration showing the Antuna classification for glenoid wear after glenoid implant removal, taking into account both the location and the severity.26

Treatment Strategies

Glenoid Bone Grafting. There are a variety of grafting options to treat bone loss of the glenoid (Table 3). When available, that is, in the primary arthroplasty setting, using humeral head autograft is an attractive option for glenoid bone loss because it is a readily available local graft. The burr and cut technique uses the humeral head as an autograft to fill a defect that it created on the glenoid. In this technique, which can be used for either reverse or anatomic total shoulder arthroplasty, the size of the humeral autograft is determined based on a preoperative template. The glenoid is prepared to bleeding bone with a burr. The cut humerus graft is secured in place on the glenoid with temporary Kirschner wires or screws. The glenoid and the graft are reamed with the baseplate secured thereafter (Figure 8).[27] In cases in which there may not be humeral head autograft available, such as in the revision setting, iliac crest autograft is an option that can facilitate the reconstruction of the glenoid. In addition, multiple studies have examined a variety of different graft sources. Outcomes across studies have generally been favorable although the incidence of complications and revision surgeries speaks to the difficult surgical problem.

Figure 8.

Photographs showing the burr and cut technique for humeral head autografting of glenoid defects. A, Radiograph determining the humeral head cut using the T-square method. (B) Note the significant glenoid retroversion in this case. C, The glenoid is prepared by burring to a bleeding surface, and multiple 2-mm drill holes are placed to promote healing. D, The burred humeral head is placed in situ with Kirschner wires for provisional fixation. E, The preoperative template shows graft thickness after reaming. F, Postoperative radiograph demonstrating that the template plan has been executed.27

Glenoid Augments. Glenoid bone loss can be addressed by using an augmented glenoid implant, which bolsters the area of the glenoid where there is bone deficiency. This obviates some of the complications seen with bone grafting, such as nonunion, resorption, and symptomatic implant and failure.

In the setting of RTSA, Wright et al[28] evaluated 39 patients with minimum 2-year follow-up after augmented baseplate. They found both posterior and superior augments to be safe and efficacious because neither augment dislocated nor lead to revision. Michael et al[29] also studied reverse augments and followed 22 superior, 50 posterior, and 67 posterosuperior augmented baseplates. All groups improved functional and pain scores.[29]

Custom and Patient-specific Technology. Although glenoid augments are an off-the-shelf option to remedy bone loss, a custom glenoid baseplate is a patient-specific approach that uses computer-assisted design and manufacturing to create a glenoid baseplate unique to the patient's degree of bone loss. Bodendorfer et al followed 12 shoulders for a minimum of 2 years that underwent RTSA with a custom system with severe glenoid bone loss medialized to the base of the coracoid. Pain, functional patient-reported outcome scores, and range of motion improved after surgery. No complications or revisions were observed.[30]

Patient-specific instrumentation involves creating surgical instruments based on a patient's unique anatomy and using this technology to implant components in an optimal position. Levy et al[31] used patient-specific glenoid baseplate drill guides for RTSA in 14 cadavers and compared the resultant trajectory with preoperative plans. The translational accuracy averaged 1.2 ± 0.7 mm, whereas the glenoid version accuracy averaged 2.6° ± 1.7°.

Although patient-specific instrumentation is effective, it may not be superior to standard instrumentation. Throckmorton et al[32] studied 70 cadavers with glenohumeral osteoarthritis who were randomized to patient-specific guides or instrumentation for reverse and anatomic total shoulder arthroplasty. No significant difference was observed in accuracy of version or inclination between the two groups.

By contrast, Eraly et al[33] made glenoid defects in 10 cadaveric shoulders and implanted RTSAs using patient-specific positioning guide or standard instrumentation. The patient-specific guide markedly reduced angular deviations from planned glenoid implant position and markedly improved the position of the screws.

Dallalana et al[34] used a patient-specific instrumentation system and implanted 10 RTSAs that were then compared with the preoperative plan. Deviation from the planned version was 1.1° ± 1.1°, and deviation from the planned inclination was 1.6° ± 1.1°. Thus, although this technology is effective, it is a matter of debate to what extent it adds benefit over standard instrumentation.


Navigation, which uses an optical tracking system to determine the location and orientation of the central guide pin, can be helpful in the setting of glenoid bone loss by implanting the baseplate where there is maximal bony support. Navigation has shown promise in allowing surgeons to precisely target glenoid fixation.[35–37] This may be particularly helpful in cases with poor glenoid bone stock in which there is less room for error in fixating the glenosphere in a precise location.

O-arm navigation utilization in RTSA was studied by Sasaki et al[36] in a case-control study. No significant difference was found between the range of error for version in the control and navigated groups, but a significant improvement was observed in the range of error for inclination in the navigated group. In this study, 3 of 15 cases were Walch class B2/3, whereas 4 of 20 controls were Walch class B2/3, suggesting that the use of navigation may be a promising developing technology for cases involving bone loss. However, more research is necessary for this specific indication.

Alternative Central Line

For shoulders with abnormal glenoid morphology, implanting the glenoid baseplate in a typical orientation in reference to the joint surface may result in insufficient bone stock into which to secure the baseplate. An alternative central line aligning with the axis of the junction of the scapular spine and scapular body allows for more robust fixation in those with notable glenoid erosion (Figure 9, A–D). Frankle et al[38] examined the use of an alternative center line in 216 glenoids in primary RTSA. They found that in abnormal glenoids, the spine centerline provided longer bony distance. Abnormal glenoid morphology also reduced the peripheral screw placement area by 42% and limited it to the anterior and inferior quadrants.

Figure 9.

Illustration showing A, anterior and (B) inferior view of the standard centerline and C, anterior and (B) inferior view of the alternative centerline line along the scapular spine.

Similar results were found by Klein et al,[39] in which they compared surgical techniques in normal and abnormal glenoids for 143 shoulders undergoing primary RTSA with minimum 2-year follow-up. In their series, 87 normal glenoids were implanted according to the standard centerline, whereas 56 abnormal glenoids were implanted according to the alternative centerline, of which 22 also had bone grafting. They found that larger glenosphere was used more often in those with glenoid defects. However, there were no significant difference between groups in American Shoulder and Elbow Surgeons scores and no radiographic failure or loosening in either group. In both groups, all outcomes were improved.