CORR Insights®: Do Early Results of Proximal Humeral Allograft-Prosthetic Composite Reconstructions Persist at 5-year Followup?

Where Are We Now?

The proximal humerus is the most-common site for primary malignant and metastatic disease in the upper extremity. Most patients who have surgery for upper-extremity malignancies undergo a limb-salvage procedure. A well-reconstructed shoulder must be both mobile and stable; these goals often are difficult to achieve, since tumor resections may involve capsule, deltoid, and rotator cuff. The use of allograft-prosthetic composites, which allows the surgeon to reconstruct the proximal humerus and repair any remaining rotator cuff and deltoid, is appealing in this setting, as this may reduce the risk of complications and improve functional outcome [1, 2, 4-8]. Although using an allograft-prosthetic composite provides patients with an excellent functional outcome, complications still arise, in particular glenohumeral instability. Newer allograft-prosthetic composite techniques, such as combing the allograft-prosthetic composites with a reverse total shoulder arthroplasty, may allow for improved function with an acceptable complication profile [8].

In the current study, El Beaino and colleagues [3] retrospectively studied allograft-prosthetic composite reconstructions using a hemiarthroplasty of the proximal humerus following intraarticular resection from their institutional database. Patients had acceptable shoulder function in terms of ROM and Musculoskeletal Tumor Society (MSTS) score, and a low initial complication profile at 1-year postoperation. However, continued followup found that the risk of late (> 12 months) complications was high, with a cumulative incidence of revision of 10.1% 5 years after the index reconstruction. Additionally, patients’ active ROM and MSTS scores had decreased at their 5-year followup. Although this approach to allograft-prosthetic composite reconstructions may provide patients with an acceptable initial functional outcome, it may not be the final answer to the ideal shoulder reconstruction method. Indeed, the solution to proximal humeral reconstruction following tumor resection still eludes us.

Where Do We Need to Go?

Reconstruction of the proximal humerus generally provides patients with a stable platform for elbow and hand function. However, active shoulder motion is limited and subluxation of the implant is common, resulting in poor function [6]. In the current study, the authors noted that the mean active shoulder forward flexion is reasonable at 1-year (101^o), which was superior to previous reports of this technique [1, 4]. However, the risk of revision and MSTS scores appear to be similar at 1-year followup to previous reports [1, 4, 7], and decreased over time [2]. This raises some important questions: (1) Do patients’ active ROM decrease over time because of the repair technique or because the tendons from the allograft fail over time? (2) Although the active ROM is improved compared to other methods of reconstruction, it is not near normal. Do other methods of reconstruction—that do not rely on the rotator cuff—warrant further investigation for these patients? (3) Are MSTS scores low due to the decreased ROM, the superior migration of the prosthesis, or greater tuberosity resorption? (4) Is there a way to prevent the superior migration of the proximal humerus, and if so, how will this impact functional outcome? (5) Allograft resorption and delayed union remains a problem. In the patient cohort a compression plate was utilized later in the series at the site of the humeral osteotomy, will the routine use of a compression plate reduce the rate of delayed union?

How Do We Get There?

Since proximal humeral resections and reconstructions are uncommon, a promising first step would be to establish a retrospective study examining modern reconstruction techniques to determine which technique performs the best. Next, we could prospectively compare the best forms of reconstruction for patients following proximal humeral resection and reconstruction. In order to obtain proper data, we would likely need a multicenter study.

Limb-salvage techniques in the upper extremity have improved over time; however, advances in reconstructive implants and soft-tissue attachments to the implants have lagged, particularly in the upper extremity, and near-normal function continues to elude us. A reconstructive technique that allows for near normal shoulder function with a low complication profile would be a tremendous advancement in our field, and provide patients with not only a stable platform for hand and elbow motion, but also shoulder function.