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

Humeral Bone Loss

There are numerous causes for humeral bone loss in the setting of RTSA. Revision for any reason can result in humeral bone loss, both antecedent to the revision surgery, such as aseptic loosening, septic loosening, osteolysis, and stress shielding, or, alternatively, as a result of extracting primary implants. Oncological resection can also cause bone loss given the need to remove large quantities of affected bone with adequate margins en bloc. In addition, humeral bone loss is often seen in the setting of periprosthetic infection, not only because infection can lead to osteolysis but because treatment also involves the removal of implants and tissue as part of source control.

Humeral bone loss can be characterized using the Proximal Humeral Arthroplasty Revision Osseous inSufficiency classification system (Table 1), which differentiates bone loss based on the segment of bone that is involved (Figure 1, A and B). Although used primarily to describe bone loss, the system is associated with treatment approaches. Type 1 and type 2 are associated with greater tuberosity fixation, type 2 and type 3 are associated with structure humeral bone grafting, and type 3 is associated with proximal humeral or total humerus replacement.[4]

Figure 1.

A, Illustration showing numeric types of the Proximal Humeral Arthroplasty Revision Osseous inSufficiency classification system with (B) subtypes

Treatment Strategies

Allograft Prosthetic Composite. Allograft prosthetic composite (APC) describes the use of arthroplasty to treat glenohumeral pathology in combination with bone allograft to address bone loss. There are multiple advantages in using this treatment approach to address bone loss. Using a proximal humerus allograft reduces the high torsional loads that are present at modular junctions in the setting of bone loss. The presence of the bony allograft helps fill the subdeltoid space and reestablishes the deltoid wrapping effect, thus providing enhanced stability to the prosthetic articulation. Finally, an APC can provide attachment sites for tendons, especially in oncological cases, that can enhance stability, motion, and strength.

However, APCs also come with a clear set of disadvantages. As with any bone grafting procedure, there are complications relating to graft resorption and failure of graft incorporation, which can result in the loss of tuberosity contour, stress shielding, and ultimately fracture. In addition, there is the risk of infection not only from the typical risk of an extensive surgery involving a revision arthroplasty but also the unique risk of donor-to-host transmission of infection from the graft, albeit an uncommon occurrence. Allografts are also a high-cost technology, and the complexity of these cases increases surgical time, which is both costly and can lead to increased morbidity.

When conducting a RTSA with APC, technical considerations are important to consider. In the setting of available bone that is proximal to the deltoid insertion, a step-cut osteotomy can be used, which enhances the chance of bony union by increasing the area of contact between the allograft and the host bone and provides some intrinsic stability. However, the step cut must be perfectly cut to avoid loss of bony contact. Some authors recommend a straight osteotomy and advocate that fixating the graft with a plate allows for more rotational control and compression across the graft-host interface. The senior author recommends a male allograft proximal humerus for female patients and a female allograft proximal femur for male patients to obtain an allograft, even when not specifically size matched that is larger than the host bone (Figure 2). This allows a long flange of the allograft to be cut, placed along the anterior humerus, and affixed with cables for excellent bone apposition, as demonstrated in the associated Video (Figure 3). Frozen allografts are preferred given cost and availability because the cartilage-friendly properties of fresh allografts are not required as that the articular surface will be removed.

Figure 2.

Photograph showing sizing of the humeral allograft such that the graft is larger than the native humerus. Image courtesy of Grant E. Garrigues, MD.

Figure 3.

A, Photograph showing intraoperative fixation of the allograft prosthetic composite using cerclage wires. B, AP postoperative radiograph. C, Scapular Y lateral postoperative radiograph. The circle indicates the shingle method. Images courtesy of Grant E. Garrigues, MD.

Clinical outcomes have demonstrated that APC is an efficacious solution to a challenging problem with Chacon et al,[12] showing a high rate of both clinical success and radiographic incorporation of the allograft. Indeed, Sanchez-Sotelo et al[13] followed 26 patients who underwent RTSA with APC conducted with compression plating. These patients had notable improvement in pain, elevation, and external rotation, with no difference between primary and revision outcomes, and a high revision-free survival rate of 96% at 2 and 5 years was observed. Longer-term outcomes were studied by Cox et al[14] who followed 73 patients who underwent RTSA with APC from 2 years postoperatively up to 10 years postoperatively. Patient-reported outcomes and range of motion markedly improved, but complications were notable, with 14 patients undergoing revision for prosthetic fracture (6), glenosphere dissociation (2), humeral loosening (2), and infection (2). The survival rate dropped off markedly with prolonged follow-up with a reoperation-free survival rate of 88% at 5 years, 78% at 10 years, and 67% after 10 years. Thus, the use of APC in RTSA for humeral bone loss is effective, but it is not without complications and there is concern for long-term survival.

Endoprosthetic Reconstruction. Endoprosthetic reconstruction can be used for bone loss of the proximal humerus by using a megaprothesis that is implanted directly into the humeral diaphysis. The steps for implantation are more similar to those for a primary arthroplasty because less shaping and carpentry is required than with an allograft. In addition, there is theoretically a lowered risk of mechanical complications, and because there is no reliance on bony healing as there is for APC, no complications were observed related to nonunion and hardware failure. As a result, it has lower reoperation rates.[15] However, tendon reattachment sites are typically limited to suture holes and perhaps a porous coating that may promote scar formation.

Technical considerations may lead one to favor the use of an endoprosthesis for reconstruction. Endoprosthesis are typically modular, and although this modularity unprotected by bone can be a weak point, it can also be an advantage if later revision is required. This is particularly useful in the oncologic population because many patients are young and revision with stem retainment is an advantage. In addition, there are some systems that allow the proximal humerus megaprosthesis to mate with total elbow prosthesis, allowing for the replacement of the entire humerus when warranted, although the results of this so-called total humerus replacement are sobering because the deltoid insertion is, by definition, compromised.[16,17]

Most endoprosthesis literature comes from oncologic experience, and care must be taken because this patient population may differ significantly from the revision arthroplasty patient these devices are frequently used for today. The use of endoprothesis was originally limited to only hemiarthroplasty and has only recently expanded to more widespread use of RTSA by tumor surgeons. Surgeons who used a hemiarthroplasty endoprosthesis found longevity in the implant, but very modest functional outcomes.[18]

Clinical outcomes have shown that endoprosthesis reconstruction using RTSA technology is successful in at least meeting limited goals, reflective of the difficult nature of the cases in which it is used, although sample sizes across studies have been small. Guven et al[19] found improved range of motion, pain scores, and functional outcomes in the 10 patients they followed for reverse shoulder tumor prosthesis for resection of proximal humerus tumors with two patients undergoing open reduction for instability. Maclean et al[20] followed 8 patients who underwent primary and revision surgery using the Bayley Walker (Stanmore Implants) reversed polarity, linked shoulder replacement for oncological reconstruction. At a minimum follow-up of 36 months, there was a 100% survivorship rate. Giffiths et al[21] followed 42 patients who underwent a massive, fixed-fulcrum endoprosthesis for reconstruction after tumor resection. Patients with primary bone tumors did better than those with malignant lesions, as did those who received constrained liners. Those with sacrificed axillary nerves and skeletal immaturity did worse. Fourteen patients dislocated their prosthesis, with 4 undergoing open reduction and 10 electing to stay dislocated because they were minimally symptomatic.

Comparative Studies. Some studies have compared outcomes for a variety of surgical techniques to treat proximal humerus tumors with outcomes largely dependent on the amount of resection required, rather than the specific surgical approach used. Potter et al[22] followed 49 patients who underwent proximal humerus resection with osteoarticular allograft, APC, or endoprosthetic reconstruction. The 5-year implant survivorship was highest for endoprosthesis reconstruction at 100%, followed by APC at 91%, with osteoarticular allograft at only 56%. Complications were common with 51% of patients experiencing complications across the 3 groups, leading to a 33% revision rate. The average Musculoskeletal Tumor Society functional scores were highest for APC (79%), followed by osteoarticular allograft (71%) and endoprosthetic reconstruction (69%).

Similar results were found in a systematic review by Dubina et al[15] that included 1,227 patients undergoing a variety of procedures after resection of proximal humerus tumors. Osteoarticular allografts resulted in a higher revision rate of 34%, whereas megaprosthesis had the lowest revision of 10%.

Based on these studies, it appears that APC is the best procedure for promoting greater functional scores and range of motion, whereas endoprosthetic reconstruction is the best option for decreasing complications and need for revision surgery. A summary of the advantages and disadvantages of these techniques is demonstrated in Table 2. Retaining the deltoid insertion is an important consideration in choosing a surgical approach. More research is necessary to compare these techniques in broader patient populations and, specifically, in the revision arthroplasty setting.