Severe glenoid bone loss in patients with osteoarthritis with intact rotator cuff is associated with posterior glenoid bone loss and posterior humeral subluxation. Management of severe glenoid bone loss during shoulder arthroplasty is controversial and technically challenging and options range from humeral hemiarthroplasty, anatomic shoulder replacement with glenoid bone grafting or augmented glenoid component implantation, to reverse replacement with reaming to correct version or structural bone grafting or metallic augmentation of the bone deficiency.

Shoulder replacement with severe glenoid bone loss is technically challenging and characterised by higher rates of complications and revisions. Hemiarthroplasty has limited benefit for pain relief and function especially if eccentric glenoid wear exists. Bone loss with >15 degrees of retroversion likely requires version correction include bone-grafting, augmented glenoid components, or reverse total shoulder replacement. Asymmetric reaming may improve version but is limited to 15 degrees of version correction in order to preserve subchondral bone and glenoid bone vault depth. Bone-grafting of glenoid wear and defects has had mixed results with graft-related complications, periprosthetic radiolucent lines, and glenoid component failure of fixation. Implantation of an augmented wedge or step polyethylene glenoid component improves joint version while preserving subchondral bone, but is technically demanding and with minimal short term clinical follow-up. A Mayo study demonstrated roughly 50% of patients with posteriorly augmented polyethylene had radiolucent lines and 1/3 had posterior subluxation. Another wedge polyethylene design had 66% with bone ingrowth around polyethylene fins at 3 years. Long term outcomes are unknown for these new wedge augmented glenoid components. Reverse shoulder arthroplasty avoids many risks of anatomic replacement glenoid component fixation and stability but is associated with a high complication rate (15%) including neurologic and baseplate loosening and often requires structural bone grafting behind the baseplate with suboptimal outcomes or metallic augmented baseplates with limited evidence and short term outcomes. Reverse replacement with baseplate bone grafting or metal augmentation is technically challenging due to limited native glenoid bone stock available for baseplate component ingrowth and long term fixation. Failure to correct glenoid superior inclination and restore neutral version within 10 degrees increases the risks of reverse baseplate failure of fixation, pull out, and failure of reverse replacement. Reverse baseplate failure rates in patients with severe glenoid bone loss and concomitant glenoid bone grafting range from 5–11%. The minimum native glenoid bony contact with the baseplate is unknown but likely is approximately 1cm of native bone contacting a central ingrowth post and a minority (∼15–25%) of native glenoid contacting the backside of the baseplate. Failure to correct posterior bone loss can lead to retroversion of the baseplate, reduced external rotation, posterior scapular notching, and posteromedial polyethylene wear. In summary, shoulder replacement with severe glenoid bone loss is technically challenging and characterised by higher rates of complication and revision.

Reverse shoulder arthroplasty is becoming a frequent treatment of choice for patients with shoulder disorders. Complication rates after reverse shoulder arthroplasty may be three-fold that of conventional total shoulder arthroplasty especially in high risk patient populations and diagnoses like revision arthroplasty, fracture sequelae, and severe glenoid bone loss. Complications include component malposition, stiffness, neurological injury, infection, dislocation or instability, acromial or scapular spine fractures, scapular notching, and loosening of implants. Recognition of preoperative risk factors and appropriate 3D planning are essential in optimizing patient outcome and intraoperative success. Failure of reverse shoulder arthroplasty is a significant challenge requiring appropriate diagnosis of the failure mode.

The most common neurological injuries involve the brachial plexus and the axillary nerve due to traction, manipulation of the arm, aberrant retractor placement, or relative lengthening of the arm. Intraoperative fractures are relatively uncommon but include the greater tuberosity, acromion, and glenoid. Tuberosity fracture can be repaired intraoperatively with suture techniques, glenoid fractures may be insignificant rim fractures or jeopardise baseplate fixation and require abandoning RSA until glenoid fracture ORIF heals and then a second stage RSA. Periprosthetic infection after RSA ranges from 1 to 10% and may be higher in revision cases and frequently is Propionibacterium acnes and Staphylococcus epidermidis. Dislocation was one of the most common complications after RSA approximately 5% but with increased surgeon experience and prosthetic design, dislocation rates are approaching 1–2%. An anterosuperior deltoid splitting approach has been associated with increased stability as well as subscapularis repair after RSA. Scapular notching is the most common complication after RSA. Notching may be caused by direct mechanical impingement of the humerosocket polyethylene on the scapular neck and from osteolysis from polyethylene wear. Sirveaux classified scapular notching based on the defect size as it erodes behind the baseplate towards the central post. Acromial fractures are infrequent but more common is severely eroded acromions from CTA, with osteoporosis, with excessive lengthening, and with superior baseplate screws that penetrate the scapular spine and create a stress riser. Nonoperative care is the mainstay of acromial and scapular spine fractures.

Recognizing preoperative risk factors and understanding component positioning and design is essential to maximizing successful outcomes.

Durable humeral component fixation in shoulder arthroplasty is necessary to prevent painful aseptic loosening and resultant humeral bone loss. Causes of humeral component loosening include stem design and material, stem length and geometry, ingrowth vs. ongrowth surfaces, quality of bone available for fixation, glenoid polyethylene debris osteolysis, exclusion of articular particulate debris, joint stability, rotator cuff function, and patient activity levels. Fixation of the humeral component may be achieved by cement fixation either partial or complete and press-fit fixation. During the past two decades, uncemented humeral fixation has become more popular, especially with short stems and stemless press fit designs. Cemented humeral component fixation risks difficult and complicated revision surgery, stress shielding of the tuberosities and humeral shaft periprosthetic fractures at the junction of the stiff cemented stem and the remaining humeral shaft. Press fit fixation may minimise these cemented risks but has potential for stem loosening. A randomised clinical trial of 161 patients with cemented vs. press fit anatomic total shoulder replacements found that cemented fixation of the humeral component provided better quality of life, strength, and range of motion than uncemented fixation but longer operative times. Another study found increased humeral osteolysis (43%) associated with glenoid component loosening and polyethylene wear, while stress shielding was seen with well-fixed press fit humeral components. During reverse replacement the biomechanical forces are different on the humeral stem. Stem loosening during reverse replacement may have different factors than anatomic replacement. A systemic review of 41 reverse arthroplasty clinical studies compared the functional outcomes and complications of cemented and uncemented stems in approximately 1800 patients. There was no difference in the risk of stem loosening or revision between cemented and uncemented stems. Uncemented stems have at least equivalent clinical and radiographic outcomes compared with cemented stems during reverse shoulder arthroplasty. Durable humeral component fixation in shoulder arthroplasty is associated with fully cemented stems or well ingrown components that exclude potential synovial debris that may cause osteolysis.