Total shoulder arthroplasty has become the surgical treatment of choice for most severe arthritis affecting the glenohumeral joint. With increasing numbers of primary arthroplasties being carried out, more revision shoulder arthroplasty will be necessary in the future.
This chapter outlines the different causes of failure in primary arthroplasties and presents a comprehensive review of the necessary workup and the technical and surgical procedures required and an overview of results in each type of revision cohort group.
Revision shoulder arthroplasty is technically demanding and requires a thorough knowledge of previous primary surgery, including the prosthetic equipment and implants utilized.
In all cases, infection, especially Propionibacterium acnes, must be considered and ruled out.
Careful workup includes complete evaluation of soft tissue function, bone loss, and neurologic status.
All necessary implant removal equipment should be available, as well as primary and allograft bone supplementation.
In some cases, reverse shoulder arthroplasty is the best option. The surgeon doing revisions should be adept at this technology and have it available.
Be prepared for all aspects of bone and soft tissue deficiency.
Proper workup, including imaging, electromyography (EMG), and aspirations, should be done when indicated.
Failed shoulder arthroplasty is often multifactorial.
Know the implant system being revised and have all necessary equipment.
Have bone graft and allograft available if needed.
Obtain consent for reverse shoulder arthroplasty.
During the last 30 years, shoulder arthroplasty for the treatment of glenohumeral arthritis, rotator cuff tear arthropathy, and proximal humerus fractures has become routine and proven itself as a viable treatment option. Projected survival rates of unconstrained total shoulder arthroplasty from Cofield’s survivorship analysis are 96% at 2 years, 92% at 5 years, and 88% at 10 years. These excellent results have lead to an increase in the number of shoulder arthroplasties performed each year from approximately 10,000 in 1990 to more than 24,000 in 2002.
As the number of primary shoulder arthroplasties has increased, so has the incidence of complications requiring a revision procedure. In a 1996 review by Cofield, 123 of 1183 patients undergoing shoulder arthroplasties had complications. The reported incidence of primary shoulder arthroplasties requiring revision is 5% to 10%. Excellent outcomes can be expected for most patients undergoing revision shoulder arthroplasty as evaluated at the time of early and mid-term follow-up.
Even though total shoulder arthroplasty is a technically demanding procedure and challenging to the most experienced surgeon, surgeons who perform only one or two arthroplasties per year make up to 85% of cases. Patient complication rates, lengths of hospital stays, and readmission rates have all been shown to be lower when shoulder arthroplasties are performed by high-volume surgeons in high-volume hospitals.
Arthroplasty failure can be categorized based on component wear, loosening, soft tissue deficiencies, osseous deficiencies, periprosthetic fracture, and infection. Often the indication for revision surgery is multifactorial. Successful revision surgery requires recognition and treatment of all the failure mechanisms in each individual patient. It is the purpose of this chapter to review the complications of shoulder arthroplasty and provide a comprehensive and logical approach to revision surgery.
INDICATIONS AND CONTRAINDICATIONS
Pain relief remains the primary goal in revision surgery. Revision outcomes are often determined by the surgeon’s ability to recognize and adequately address component wear, loosening, soft tissue and osseous deficiencies, fractures, and infection. Severe bony and soft tissue deformities often limit the restoration of full function and range of motion; therefore, patients need to have a clear understanding of the goals, expected outcomes, and potential complications of a revision procedure.
Revision shoulder arthroplasty should only be considered after a complete history and physical examination as well as appropriate laboratory and imaging studies are performed. The best candidates for revision are patients with good initial outcomes after primary shoulder arthroplasty who suddenly deteriorate and progressively worsen over time.
Contraindications to revision arthroplasty include medical comorbidities limiting patients’ physiologic reserves and capabilities to handle the stress of revision surgery. Also, emotional, psychological, or neurologic disturbances that prevent patients from complying with postoperative rehabilitation programs are contraindicated. Lastly, alcoholism is a relative contraindication to surgery.
Detailed history begins with inquiry into the primary arthroplasty indications, preexisting shoulder pathology, and previous surgical history. The operative report is then reviewed for prosthetic type, size, and mode of fixation. The patient’s postoperative rehabilitation course and motivation, which can often gauge the success of subsequent operations, is also elicited. Lastly, any recent trauma must be documented.
A sudden onset of clicking or crepitus with motion may indicate glenoid loosening or a fractured component. Night pain and weakness are highly suggestive of rotator cuff or soft tissue pathology. A sudden loss of motion that had been regained after the index arthroplasty is an indication of component or soft tissue dysfunction requiring thorough evaluation. Prior to any revision surgery, infection must be ruled out as the cause of implant failure. A history of fever, chills, night sweats, incision erythema, or persistent drainage suggests periprosthetic infection and requires further laboratory and radiologic workup.
The physical examination begins with a thorough evaluation of the cervical spine, thorax, and abdomen to rule out potential sources of referred shoulder pain. Examination of the shoulder and upper extremity is then performed to assess atrophy, active and passive range of motion, strength, and stability. Rotator cuff pathology usually produces atrophy, weakness, and signs of impingement. Soft tissue contractures or an oversized humeral head often result in stiffness. Stability testing includes anterior, posterior, and inferior translation of the humeral head on the glenoid. Clicking during range of motion may represent a loose or fractured glenoid component. Scarring of the long head of the biceps tendon has recently been reported as a cause of restricted motion and stiffness. The long head of the biceps tendon is evaluated with provocative tests such as Speed’s and Yergason’s.
The importance of a careful neurovascular examination cannot be overemphasized. Lynch et al. reviewed 417 total shoulder arthroplasties and reported a 4% incidence of neurologic complications. Following discovery of abnormal neurovascular exam findings, electromyography (EMG) and nerve conduction studies are indicated.
Radiographic evaluation begins with standard anteroposterior (AP), axillary, and outlet views. Plain films are used to assess humeral height, head size, and version. A humeral component set too high or an oversized humeral head often overstuffs the joint, causing stiffness and early glenoid wear. Plain radiographs are also used to identify sites of impingement such as bone spurs off the acromion and acromioclavicular joint. In the setting of trauma, plain films are used to rule out periprosthetic fractures. Glenoid bone loss, deformity, and component version are assessed using the axillary view.
Occasionally, bilateral humeral scanograms are used as an adjunct to help assess humeral height and can often differentiate between implant malposition and subsidence secondary to muscular dysfunction. In addition, computed tomography (CT) with implant subtraction may be utilized to evaluate osseous deficiencies and magnetic resonance imaging (MRI) can be used to evaluate soft tissue and osseous integrity ( Fig. 25-1 ). Recently, we have begun utilizing ultrasound to assess soft tissue integrity in cases in which we suspect rotator cuff failure, including subscapularis rupture.
When assessing for component loosening, recent studies have shown that extensive radiolucency usually precedes clinical symptoms. Component loosening is suggested by radiolucency of 2 mm or more surrounding the implant and is confirmed when subsequent radiographs indicate a shift in component position. Fluoroscopy has also been used successfully to evaluate implant loosening.
Routine laboratory tests include complete white blood cell count (WBC), erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP) in order to rule out infection. If infection is suspected, radionuclear studies such as three-phase bone scan and indium-labeled white blood cell scan are also performed. In addition, joint aspiration can be used to confirm the diagnosis and identify the offending organism for directed antimicrobial therapy. Unfortunately, in the setting of infection by an indolent organism, laboratory tests, nuclear studies, and joint aspiration are not highly accurate. If infection is suspected but the diagnosis is uncertain, preparation must be taken to perform a two-stage revision based on analysis of intraoperative tissue specimens.
The successful use of diagnostic arthroscopy in the evaluation and treatment of failed shoulder arthroplasty has been well established. Hersch and Dines reported on the use of diagnostic arthroscopy to diagnose glenoid loosening, remove loose bodies, and treat impingement syndrome and rotator cuff tears ( Fig. 25-2 A ). Freedman et al. also described arthroscopic treatment of shoulder arthroplasty impingement. Loose glenoid components can also be removed arthroscopically. O’Driscoll et al. reported the removal of loose glenoid components in five patients with contraindications to glenoid reimplantation. These glenoid components were removed after sectioning the components into quarters using small osteotomes. This allowed the four smaller pieces to pass through an arthroscopic portal. Tuckman and Dines recently reported on successful arthroscopic debridement or tenodesis of the biceps tendon in symptomatic shoulder arthroplasty patients.
Arthroscopy following shoulder arthroplasty can be technically challenging secondary to the “mirror effect” created by the surface of the humeral head. Directing the arthroscope away from the humeral component helps compensate for this effect. Precise manipulation of arthroscopic instruments helps avoid damaging the humeral and glenoid components. Gross glenoid stability is best tested using a probe placed through the anterior portal.
SOFT TISSUE DEFICIENCIES
Rotator Cuff Tears
An intact rotator cuff after shoulder arthroplasty is necessary for optimal function and maximal range of motion. Rotator cuff tear following shoulder arthroplasty has a reported incidence of 2% to 2.7%. The development of a tear postoperatively may be asymptomatic; however, it often results in pain, instability, and decreased function. In preparation for revision surgery due to rotator cuff tear, MRI is utilized to determine tear size, tissue quality, and additional soft tissue pathology. Unfortunately, even with use of limited pulse sequence parameters, soft tissues can be difficult to visualize with the shoulder prosthesis in place. In these cases, diagnostic arthroscopy may help establish an accurate diagnosis. Revision surgery is indicated based on the patient’s clinical symptoms and physical examination. Patients with minimal pain and loss of function often benefit from gentle physical therapy. For isolated rotator cuff tears, arthroscopic or mini-open techniques are utilized; however, if complex soft tissue pathology exists, the rotator cuff and soft tissues are best addressed through the standard deltopectoral approach ( Fig. 25-2 B ).
The clinical outcomes of rotator cuff repair after shoulder arthroplasty remain unpredictable. Although patients often report improvement in pain scores, recurrent rotator cuff tears and limited functional gains continue to be significant problems. Therefore, prevention should be emphasized through diligent soft tissue balancing during primary arthroplasty and supervised postoperative rehabilitation.
Glenohumeral instability, with a reported incidence of 5.2%, is a well-recognized and challenging complication of shoulder arthroplasty. Initial stability following shoulder arthroplasty is provided by the combination of implant size, version, and surrounding soft tissue attachments. Often the cause of glenohumeral arthroplasty instability is multifactorial. The most common causes include subscapularis failure, capsular contracture, incorrect component size, and component malposition. Although much less common, component loosening and nerve injury can also lead to instability. Successful revision for instability requires a thorough preoperative assessment of the shoulder soft tissue integrity, component position, and neurologic status.
Anterior instability is most commonly due to incompetent subscapularis tendon secondary to traumatic rupture or poor repair. Moeckel et al. reported a subscapularis tear in all seven patients requiring reoperation for anterior instability following shoulder arthroplasty. Following thorough history and physical examination, MRI, ultrasound, or arthrogram is used to confirm the diagnosis of incompetent subscapularis tendon. In acute traumatic ruptures, the subscapularis may be repaired primarily through bone holes or with suture anchors. In cases where adequate repair of the subscapularis cannot be achieved, a pectoralis major transfer may be a viable option. In salvage cases, the use of Achilles tendon allograft to reconstruct the static anterior restraint has been used with limited success. An oversized humeral head component or excessive component version may also cause anterior instability; however, this is much less common and more easily addressed at time of revision.
Posterior instability is often associated with excessive retroversion of the glenoid component (more than 20 degrees) or humeral component (greater than 45 degrees). Excessive glenoid retroversion often results from posterior glenoid erosion secondary to long-standing osteoarthritis or following hemiarthroplasty. Failure to recognize and correct the excessive glenoid retroversion during primary arthroplasty increases the risk of posterior instability. Correction for excessive posterior glenoid erosion requires eccentric reaming of the anterior aspect of the glenoid and appropriate adjustment of humeral component version. During revision arthroplasty for posterior instability, posterior capsular distension is a common finding and requires capsular plication for a successful outcome. Capsular plication is typically begun after stem and glenoid revision but prior to placement of the final humeral head. A vertical row of sutures is placed in the posterior capsule after a longitudinal incision is made. The posterior sutures are tightened after impaction of final humeral head component in order to avoid excessive tightening of the capsule.
Inferior instability most commonly results from failure to restore humeral height but also can be caused by deltoid dysfunction. Bilateral humeral scanograms are used to differentiate between implant malposition and subsidence secondary to muscular dysfunction. During revision secondary to implant malposition, restoration of the anatomic humeral length is necessary to reestablish the resting tension of the deltoid and rotator cuff. If appropriate resting tension cannot be restored with component reposition and soft tissue rebalancing, a reverse prosthesis should be considered.
Progressive anterosuperior instability results from rotator cuff and coracoacromial ligament deficiencies. In patients with rotator cuff deficiency and previous coracoacromial ligament resection, the humeral head migrates proximally and articulates eccentrically with the superior portion of the glenoid. The “rocking-horse” effect results in increased stress at the bone-cement interface and accelerates glenoid loosening. Several techniques have been described to reconstruct the superior restraint of the glenohumeral joint. Use of Achilles tendon allograft to reconstruct the coracoacromial arch has recently been reported with moderate success. Patients were satisfied with pain relief and limited functional goals; however, UCLA and L’Insalata questionnaire scores were low. Galatz et al. also reported moderate success utilizing a subcoracoid pectoralis major transfer to treat anterosuperior subluxation in 14 patients with massive rotator cuff insufficiency. Most of these patients reported decreased pain and improved activities of daily living; however, overhead function remained limited.
The reverse total prosthesis has evolved as an additional treatment option for anterosuperior instability and rotator cuff deficiency. The prosthesis design medializes the glenohumeral center of rotation, thereby increasing the lever arm for the deltoid. An intact deltoid is necessary to establish the biomechanical fulcrum. As it contracts, the humeral head is forced into the glenoid and the arm elevates. Several reports have described good functional results using reverse total shoulder prosthesis for rotator cuff arthropathy. Guery et al. recently reported 5-year survival rates, with replacement of the prosthesis and glenoid loosening as end points, of 91% and 84%, respectively. In this series, reverse total shoulders performed for massive rotator cuff tears faired significantly better than those performed for other disorders. Although long-term results have yet to be established, the reverse ball prosthesis is a promising treatment option for anterosuperior instability secondary to rotator cuff and coracoacromial arch deficiency.
Aseptic component loosening is the most common indication to perform revision total shoulder arthroplasty, with glenoid component loosening being more common than humeral stem loosening ( Fig. 25-3 ). Successful treatment of symptomatic glenoid loosening has been reported with both component reimplantation or component resection. The decision is often made at the time of surgery based on the remaining glenoid bone stock. When assessing the remaining glenoid bone stock, it is important to consider not only the quantity of bone remaining but also the quality and orientation.