Reverse Hemiarthroplasty for Staged Revision to Reverse Shoulder Arthroplasty

    A 65-year-old female patient presents with a painful and dysfunctional right shoulder 14 years after anatomic total shoulder arthroplasty and subsequent glenoid component removal for aseptic loosening. What are the treatment options for revision shoulder arthroplasty in the setting of significant glenoid-sided bone deficiency?


    Stephen T. Gates, MD, and Gerald R. Williams Jr., MD


    Revision reverse shoulder arthroplasty (RSA) for anatomic glenoid component loosening has been shown to have acceptable results, including significant improvements in pain and motion. [1] However, the associated complication profile is greater than that of primary arthroplasty, and a recent study reported a 17% glenoid baseplate mechanical failure rate in a cohort of revision cases. [1] This problem is particularly concerning with increasing erosion of the native glenoid vault.

    Glenoid bone loss is a challenging problem frequently encountered in both primary and revision shoulder arthroplasty. Options for management of glenoid bone loss include bone grafting techniques and augmented glenoid components. For more severe bone loss, bone grafting techniques are favored, although graft resorption and baseplate loosening remain major concerns. [2]

    RELATED: Register for ICJR’s 10th Annual Shoulder Course, October 7-9 in Las Vegas

    Some surgeons, therefore, favor a staged approach for complex reconstructions that involve massive bone loss because it may be a more reliable way to restore native glenoid bone stock. [3] In delaying final implant placement, time zero loading of the baseplate is minimized, allowing for protected fixation and time for the bone graft to heal to the native glenoid. [4,5]

    The following case involves an active 65-year-old woman with persistent pain, dysfunction, and severe glenoid bone loss more than a decade after anatomic total shoulder arthroplasty (TSA). The case highlights the difficulty in managing significant glenoid bone loss, the reconstructive treatment options, and a staged technique for bone grafting with placement of a reverse hemiarthroplasty, which the authors believe optimizes conditions for bone graft incorporation and baseplate fixation.

    Case Presentation

    A 65-year-old, right hand-dominant, retired medical assistant presents to the office with chronic right shoulder pain and dysfunction. She is a non-smoker with a history of hypertension, obstructive sleep apnea, and breast cancer who enjoys dancing and travel.

    She had a right anatomic TSA performed by an outside surgeon 14 years prior to presentation to the authors’ office. Her postoperative course included an open irrigation and debridement with polyethylene glenoid component removal for aseptic loosening 4 years later. She subsequently underwent an arthroscopic irrigation and debridement with tissue cultures 10 years after the initial operation due to persistent pain. Neither of the subsequent operations yielded any clinical or microbiologic evidence of infection.

    At the time of presentation to the office, she reported a Subjective Shoulder Value (SSV) of 30% and a Visual Analogue Score (VAS) for pain ranging between 8/10 and 10/10.

    Physical Examination

    • Well-healed incisions from open and arthroscopic shoulder procedures; no signs of infection
    • Active forward elevation to 80° with 4/5 strength
    • Active/passive external rotation to -10°; active internal rotation to posterolateral buttock
    • Neurovascularly intact, including axillary, musculocutaneous, radial, ulnar, median, long thoracic, and spinal accessory nerve distributions.


    Radiographs on initial presentation to the authors’ office reveal evidence of the patient’s prior anatomic TSA surgery (Figure 1). An anatomic humeral component is in place without evidence of loosening. A significant amount of glenoid bone erosion, with little bone remaining in the glenoid vault, can be seen. The greater tuberosity rests approximately 1 cm medial to the lateral margin of the acromion.

    Figure 1. Radiographs on initial presentation: Anteroposterior (left) scapular-Y (center), and axillary (right) views of the right shoulder.

    As discussed further in the Treatment section below, the patient initially underwent an open explant procedure with revision to cement hemiarthroplasty spacer, primarily to rule out infection as a source of the clinical findings. Postoperative radiographs confirm that the patient has a significant amount of glenoid bone erosion and show interval placement of a cement hemiarthroplasty spacer status post-humeral component removal.

    Figure 2. Postoperative radiographs: Anteroposterior (left) scapular-Y (center), and axillary (right) views of the right shoulder following humeral component explant and cement hemiarthroplasty spacer placement.

    With the removal of the metal component, the shoulder was amendable to further diagnostic evaluation with a computed tomography (CT) scan for ultimate surgical planning (Figure 3). The CT images show significant bony resorption and deficiency of the glenoid, as well as fragmentation of the proximal humerus.

    Figure 3. Axial (left), coronal (center), and sagittal (right) CT cuts of the right shoulder following humeral component explant and cement hemiarthroplasty spacer placement.


    • Severe glenoid-sided bone deficiency following failed anatomic TSA


    In consideration of the patient’s history of 3 prior ipsilateral shoulder surgeries and her chronic pain and dysfunction, the authors recommended right shoulder open explantation with revision to a cement hemiarthroplasty spacer. The patient gave consent for this procedure.

    The primary purpose of the initial operation was rule out infection. One fluid sample from the joint and 5 soft tissue cultures were sent to the microbiology lab and held for 14 days. A single culture (1/6) returned positive for Cutibacterium acnes on postoperative day 12. The patient received the standard 24 hours of postoperative intravenous antibiotics in the hospital, followed by a prophylactic 14-day course of oral doxycycline. Given that only 1 culture was positive for a low-virulence organism, and not until day 12, there was a low index of suspicion for clinical infection, and no further treatment was pursued.

    An added benefit to the removal of metal included the ability to obtain a CT scan for further surgical planning, as discussed in the Imaging section above. The following options remained for the management of the patient’s current pathology.

    Indefinite Retention of the Antibiotic Cement Spacer

    There are reports in the literature of patients being managed with retention of their antibiotic cement spacer, with acceptable clinical outcomes. [6-8] These are primarily performed, however, in the setting of chronic infection and in older, low-demand patients who may not otherwise tolerate revision operations.


    • No further surgery.
    • Glenoid bone deficiency, which is a difficult problem to manage, becomes irrelevant.


    • Further bony erosion will occur over time on both the humeral and glenoid sides, dampening the ability to perform a successful subsequent revision operation.
    • Lacking a true articulation, the patient will have limited functional use of the shoulder.

    Single-stage Revision to RSA with Metal Glenoid Augmentation

    Single-stage revisions, particularly without a concern for infection, are a common treatment option in revision shoulder arthroplasty. Metal glenoid augments expand the ability to correct for various glenoid wear patterns.


    • Performing a single-stage operation avoids the necessity for a subsequent anesthetic and surgical recovery.


    • Metal glenoid augmentation options remain limited and may not be able to fully restore the joint line, version, or inclination altered by the glenoid bony erosion.
    • A metal-backed glenoid may not be able to effectively incorporate into native bone given the severity of the bony erosion.

    Single-stage Revision to RSA with Autograft versus Allograft

    Bone grafting allows for customized deformity correction in patients with bone loss.


    • Performing a single-stage operation avoids the necessity for a subsequent anesthetic and surgical recovery.
    • Bone grafting allows for larger deformity correction and has the ability to incorporate into the remaining native bone of the glenoid.


    • Autograft (eg, iliac crest bone graft) assumes risks associated with donor site morbidity.
    • Allograft (eg, femoral head allograft) assumes risks associated with the use of cadaveric tissue.
    • Immediate loading of the joint, by means of a semi-constrained reverse prosthesis, could jeopardize the time zero fixation of the glenoid baseplate, particularly in the setting of significant backside glenoid bone deficiency and reliance on bone graft incorporation.

    Reverse Hemiarthroplasty with Autograft versus Allograft as Part of a 2-stage Revision to RSA

    Two-stage revision arthroplasty is commonly considered an option in the setting of periprosthetic infections. Similarly, with associated massive bone loss on either the humeral or glenoid side, staging the operation can allow time for bony incorporation.


    • Bone grafting allows for larger deformity correction and has the ability to incorporate into the remaining native bone of the glenoid.
    • Reverse hemiarthroplasty allows the glenosphere to act as a protector cap to the underlying glenoid baseplate and backside bone graft.
    • By delaying humeral component implantation, there is little to no initial loading of the glenoid baseplate and associated bone graft, preventing shearing forces that could otherwise disrupt bony incorporation in the postoperative period.


    • Autograft (eg, iliac crest bone graft) assumes risks associated with donor site morbidity.
    • Allograft (eg, femoral head allograft) assumes risks associated with the use of cadaveric tissue.
    • The 2-stage nature of this technique necessitates multiple trips to the operating room and, therefore, multiple anesthetics and postoperative recovery periods for the patient.
    • Reverse hemiarthroplasty is unconventional, with limited reports in the literature.

    Each of the options above was discussed in detail with the patient. Ultimately, the severity of glenoid bony erosion was felt to be best managed with bone grafting. Similarly, a 2-stage approach was recommended due to the necessity of graft incorporation and baseplate on-growth for implant stability.

    The patient consented to right shoulder reverse hemiarthroplasty with femoral head allograft as part of a staged conversion to RSA.


    Revision to Reverse Hemiarthroplasty with Structural Femoral Head Allograft

    • The patient was taken to the operating room. General anesthesia was administered, and the patient was placed into a well-padded beach chair position with the head of the bed elevated to approximately 40°. Her right arm was prepped and draped in the standard sterile fashion.
    • The previous anterior shoulder incision was utilized, and full-thickness flaps were developed. Given the revision nature of the surgery, there was no true deltopectoral interval. The tip of the coracoid was identified instead, from which the deltoid and pectoralis major muscles were split distally to the deltoid tuberosity.
    • Dense adhesions deep to the deltoid and pectoralis major muscles were lysed both bluntly and sharply to develop the appropriate planes. The plane between the conjoint tendon and the subscapularis was also developed and the axillary nerve was identified.
    • Fluid was aspirated from the joint, and 5 soft tissue cultures were taken during the procedure and sent for cultures, to be held for 14 days.
    • Beginning within the bicipital groove, the entirety of the remaining anterior soft tissue envelope was reflected subperiosteally and retracted medially. The cement hemi-spacer ball was removed from the joint and the inferior capsule was excised while protecting the axillary nerve.
    • The remainder of the capsule was released and the glenoid was exposed.
    • A cement mold was placed into the glenoid deficiency and allowed to harden before being removed.
    • A burr was used to prepare the surface of the glenoid. A patient-specific guide, made based on a plan from preoperative planning software (Match Point, DJO Surgical; Austin, Texas) (Figure 4), was placed into position and the guide pin for the glenoid baseplate was advanced through the anterior neck of the scapula just medial to the base of the coracoid. The pin was then exchanged for the tap, which was left in place.


    Figure 4. Preoperative planning software screenshots depict the planned trajectory of glenoid guide pin placement and subsequent position of the baseplate after use of backside bone grafting.

    • A femoral head allograft was fashioned on the back table to match the deformity depicted by the cement mold. A center hole was over-drilled through the allograft, which was then passed over the tap and tamped into position. A pin was placed to provide provisional fixation of the graft. The glenoid and graft were reamed over the tap until there was a concentric surface for the baseplate.
    • The 26-mm diameter baseplate was then screwed into position, with final compression achieved after removal of the provision pin. It was noted that the final position of the baseplate sat 15 mm inferior from the most inferior extent of the base of the coracoid and 15 mm posterior from the anterior glenoid rim. The anterior 10% of the baseplate rested on native bone, while the remainder was supported by the structural allograft.
    • Four peripheral locking screws were placed. A 36 neutral glenosphere consisting of 6 millimeters of lateralization was impacted over the taper of the baseplate and its corresponding central locking screw was placed.
    • After copious irrigation, the anterior soft tissue envelop was closed, followed by standard skin closure.
    • Sterile dressings were applied, and the patient woke from anesthesia without complication.

    Postoperative Course

    The operative arm was placed into a simple sling. Immediate postoperative radiographs were taken in the recovery room (Figure 5), and the patient was admitted for an overnight stay, where she received 24 hours of postoperative intravenous antibiotics. She was placed on a 14-day course of prophylactic doxycycline at discharge. Intraoperative cultures were followed for 14 days, and all 6 specimens were negative for any growth.

    Figure 5. Immediate postoperative anteroposterior radiograph of the reverse hemiarthroplasty.

    2 Weeks Postoperatively

    The surgical incision was healing appropriately, and the patient was compliant with wearing the sling. All 6 intraoperative cultures remained negative through 14 days. No notable changes in implant position were seen on radiographs (Figure 6).

    Figure 6. Progressive postoperative radiographs of the reverse hemiarthroplasty – anteroposterior (top) and axillary (bottom) – at 2 weeks, 6 weeks, 3 months, and 6 months postoperatively, respectively.

    6 Weeks Postoperatively

    The surgical incision had healed by 6 weeks postoperatively, with no signs of infection. The patient was doing well, was compliant with use of her sling, and had less pain than she had preoperatively. No notable changes in implant position were seen on radiographs (Figure 6, above).

    The patient was allowed to remove the sling and use her arm for activities below shoulder height, with a 1-pound lifting restriction.

    3 Months Postoperatively

    The patient was doing well and had almost no pain. No notable changes in implant position were seen on radiographs (Figure 6, above). She was advised to continue with her current restrictions.

    6 Months Postoperatively

    Between the 3-month and 6-month postoperative visits, the patient underwent a lumbar laminectomy for chronic lumbar disc disease, which she tolerated well.

    Regarding the right shoulder, she noted an SSV of 50% with VAS pain scores ranging from 1.5 to 4. On examination, she could achieve active forward elevation of 80°, with external rotation to neutral and internal rotation to L5. No notable changes in implant position were seen on radiographs (Figure 6, above).

    At the time of this office visit, the authors discussed further treatment options with the patient, including retaining the reverse hemiarthroplasty indefinitely versus stage 2 conversion to RSA. A potential disadvantage of the former is progressive humeral bony erosion over time, which could preclude a standard humeral component if a conversion were undertaken later.

    Through shared decision-making, the patient consented to undergo stage 2 revision to RSA. This procedure was performed 7 months after the stage 1 reverse hemiarthroplasty.

    At the time of the second stage, the baseplate was found to be well-fixed, with incorporation of the bone graft. The glenosphere was exchanged to a 32 neutral sphere consisting of 10 millimeters of lateralization, and a standard length cementless reverse humeral stem was placed in 30° of retroversion. A standard polyethylene liner yielded a stable construct (Figure 7).

    Figure 7. Anteroposterior postoperative radiograph following stage 2 conversion to reverse total shoulder arthroplasty.


    Management of glenoid-sided bone loss in shoulder arthroplasty is a difficult problem, particularly in the revision setting. When severe glenoid erosion and bone loss necessitate bone grafting reconstruction techniques, use of a staged reverse hemiarthroplasty allows for bone graft incorporation while mitigating the risk of excessive mechanical loading otherwise seen if a single-stage revision is undertaken.

    Author Information

    Stephen T. Gates, MD, is a shoulder & elbow clinical fellow at The Rothman Institute in Philadelphia, Pennsylvania. Gerald R. Williams Jr., MD, is a board-certified orthopaedic surgeon with The Rothman Institute in Philadelphia, Pennsylvania, specializing in the treatment of shoulder and elbow conditions.

    Disclosures: Dr. Gates has no disclosures relevant to this article. Dr. Williams has disclosed that he receives royalties and research support from DePuy Synthes; that he receives royalties from, is a paid consultant for, and is a paid presenter or speaker for DJO; that he is a paid presenter or speaker for Mathys, Ltd.; and that he has stock or stock options in Parvizi Surgical Innovations.

    Shoulder Section Editor, Rothman Institute Grand Rounds

    Daniel E. Davis, MD, MS


    1. Bartels DW, Marigi E, Sperling JW, Sanchez-Sotelo J. Revision reverse shoulder arthroplasty for anatomical glenoid component loosening was not universally successful: a detailed analysis of 127 consecutive shoulders. J Bone Joint Surg Am. 2021.
    2. Scalise JJ, Iannotti JP. Bone grafting severe glenoid defects in revision shoulder arthroplasty. Clin Orthop Relat Res. 2008;466(1):139-145.
    3. Neyton L, Walch G, Nove-Josserand L, Edwards TB. Glenoid corticocancellous bone grafting after glenoid component removal in the treatment of glenoid loosening. J Shoulder Elbow Surg. 2006;15(2):173-179.
    4. Gupta A, Thussbas C, Koch M, Seebauer L. Management of glenoid bone defects with reverse shoulder arthroplasty-surgical technique and clinical outcomes. J Shoulder Elbow Surg. 2018;27(5):853-862.
    5. Neyton L, Boileau P, Nove-Josserand L, Edwards TB, Walch G. Glenoid bone grafting with a reverse design prosthesis. J Shoulder Elbow Surg. 2007;16(3 Suppl):S71-78.
    6. Cronin KJ, Hayes CB, Sajadi KR. Antibiotic cement spacer retention for chronic shoulder infection after minimum 2-year follow-up. J Shoulder Elbow Surg. 2020;29(9):e325-e329.
    7. Levy JC, Triplet J, Everding N. Use of a Functional Antibiotic Spacer in Treating Infected Shoulder Arthroplasty. Orthopedics. 2015;38(6):e512-519.
    8. Mahure SA, Mollon B, Yu S, Kwon YW, Zuckerman JD. Definitive treatment of infected shoulder arthroplasty with a cement spacer. Orthopedics. 2016;39(5):e924-930.