29 What Can Be Done When a Reverse Shoulder Arthroplasty Fails?
Reverse total shoulder arthroplasty (RTSA) is a terrific new tool, but failures invariably occur due to infection, instability, baseplate loosening, mechanical implant failure, humeral loosening, acromial base fractures, periprosthetic fractures, and neurovascular issues. Although some failures can be managed with minimal intervention, others lack good treatment plans. Treatment is dependent on local and host factors. The results of patients treated for a failed RTSA vary depending on the condition of the host, the number of previous operations, and the amount of host bone and deltoid muscle. If the failed RTSA is revised in a good host with good glenoid bone and deltoid, the results can be almost comparable to those of a primary RTSA. However, patients with a failed RTSA represent a challenging group, and the surgeon is often presented with poor hosts. Some failures can be salvaged with a revision RTSA. Others are treated with salvage procedures such as antibiotic hemiarthroplasty spacers, hemiarthroplasty, or bipolar arthroplasty with the anticipation of lower functional results but reasonable pain relief. Even patients with severe issues can be treated with resection arthroplasty with fair pain relief though poor function. In the most impaired hosts, nonoperative treatment may be a better course. This chapter addresses modes of RTSA failure and suggests various treatment approaches as well as some limited outcome results of these approaches.
Reverse total shoulder arthroplasty (RTSA) is a terrific new tool in the shoulder arthroplasty surgeon’s armamentarium. The implant allows management of multiple complex entities, with predictable outcomes, that previously were not treatable or treatable with low outcome expectations. However, with this advancement in technology and subsequent increased use, failures invariably occur. These failures can sometimes be readily managed with minimal intervention, but sometimes they do not have a good treatment option.
Treatment of failed RTSA is dependent on two main factors: local factors (the surgical environment) and host factors. Local factors include the amount of glenoid or humeral bone loss or the state of the deltoid. Host factors include diabetes and heart disease, and take into account the fragility of the host as well as social factors. When treating a patient with a failed RTSA, both sets of factors need to be kept in mind. This chapter will address common modes of RTSA failure and suggest various treatment approaches as well as some limited outcome results of these interventions.
29.2 Causes of Reverse Total Shoulder Arthroplasty Failure
RTSA failures can occur due to multiple causes, including the following: infection, instability, baseplate loosening, mechanical implant failure, humeral loosening, acromial base fractures, periprosthetic fractures, and neurovascular issues. Each of these failure modes will be discussed individually.
Infection is the most common cause of RTSA failure, reported to occur in 1 to 3% of primary implants and more so in revision surgery. 1 , 2 , 3 , 4 It is likely that the actual rate of infection is higher, given that it is frequently recognized late, if at all, due to infection with indolent organisms (e.g., Propionibacterium acnes). The vast majority of the RTSA infections treated at our institution are late infections occurring well after 3 months. It has been our approach to treat all RTSA revisions suspected to harbor infections as a two-stage procedure. At the first stage, an aggressive debridement, removing all foreign material including old sutures and cement, is performed. A temporary reconstruction is performed using an antibiotic-impregnated hemiarthroplasty (► Fig. 29.1). The antibiotic hemiarthroplasty obtains rotational control using a napkin ring of high-dose antibiotic-impregnated polymethylmethacrylate. The patient is then treated with 6 weeks of intervenous antibiotics dictated by cultures (usually Vancomycin or Oxacillin). At any time after 6 weeks, the second stage can be performed. Many of our frail hosts who have undergone multiple operations opt to never undergo a second-stage reimplantation. Care should be exercised in patients with poorly controlled diabetes, as our only second-stage reimplantation failures for recurrent infection have better occurred in patients with brittle diabetes with high A1Cs. Other factors that play a role in the decision to undergo the second stage or not are local issues: whether there is adequate glenoid bone stock to support the glenosphere baseplate, and the status of the multiply operated deltoid and humerus. Patients who opt to forgo the second-stage reconstruction often do surprisingly well.
Some authors have advocated primary exchange arthroplasty, but although we do this occasionally by accident with a revision it has not been our practice. However, the reported results of primary exchange arthroplasty are about the same as the two-stage results. 5 , 6 Based on the results of these two studies and discussions with other senior shoulder surgeons, we are changing our practice to perform primary exchange arthroplasties in immunocompetent hosts. Immunosuppressed hosts and hosts with inadequate local tissues to support a primary exchange will continue to be treated with a two-stage procedure. In the few acute RTSA arthroplasty infections that we have experienced, we have abandoned the early irrigation and poly exchange technique because this has uniformly failed.
Many systems now have a common platform stem, so not all humeral stems need to be removed. The advantages of a convertible platform stem not requiring removal include a lower complication rate, shorter operating time, less blood loss, decreased destruction of the native humerus, and decreased cost. 7 , 8 , 9 However, if the humeral stem needs to be removed for infection or revision purposes, it may be relatively easy or very arduous and destructive. If no infection is present, a cemented stem can be removed using flexible osteotomes proximally to break up the cement bonding to the metaphyseal texturing; then, using the stems extractor or a large bone tamp about the medial calcar, the implant can be extracted. A new stem can then be cemented into the old but stable cement mantle. However, if the implant is infected, all the cement will need to be removed. Unless the cement mantle is loose, removing cement that is well fixed distally requires windowing the humerus. It is our preference to perform a vascularized humeral door if a simple humeral split does not work for cement extraction. This technique leaves the brachialis attached to the door to facilitate rapid bone healing. 10 If an ultrasonic cement removal device is used, we suggest making working times short so as not to cause a thermal injury to the radial nerve due to its close proximity. If there is an infected cemented implant with cement to the elbow, it is our practice to remove the implant and all the loose cement we can without destruction of the humerus. Then the proximal end of the distal remaining cement can be cauterized with short bursts of the ultrasonic cement remover; however, the surgeon must be sure the device is in the humerus and keep the energy bursts short so as not to cause a thermal injury to the radial nerve. We have no data to support this practice.
Instability occurs in 1 to 3% of primary RTSAs and is much higher in revision RTSA. 4 , 11 Instability is often due to humeral contact with bony or noncompliant tissues about the inferior and posterior inferior glenoid, which then act as a shoe horn to dislocate the RTSA. The most common cause is placing the glenosphere too high, resulting in contact between the humeral implant and the native inferior or posterior inferior glenoid. When planning a revision RTSA for instability, a computed tomographic (CT) scan is very helpful for delineating areas of contact as well as looking for acromial base fractures and abscess formation.
Assuming that closed reduction is not successful, the options for treating a dislocated RTSA are revision RTSA; two stage for infection, hemiarthroplasty, or bipolar arthroplasty; and resection arthroplasty. Closed reduction under anesthesia is successful about half the time. 12 In several frail patients who have undergone multiple operations with normal neurologic function, we have opted to leave the dislocations (► Fig. 29.2). Because of their low demand, it is surprising how little pain some of these patients have. If they have profound metal-on-metal contact (this is not usual and feels like sandpaper on metal), then the glenosphere and baseplate should be removed.
When performing a revision for instability, the impinging structures, if present, should be eliminated. If the baseplate is too high, it should be lowered or a lateralized glenosphere, larger glenosphere, or some combination should be used. The goal is to have the humeral component clear the offending structure in adduction or adduction, extension, and external rotation. Displaced greater tuberosity fragments are particularly adept at causing posterior impingement and dislocation and must be at least partially removed. Patients with medial glenoid wear and RTSA instability should have this addressed with bone-grafting procedures (► Fig. 29.3), augmented baseplates (► Fig. 29.4), or lateralized glenospheres that will push the humerus and greater tuberosity lateral to the acromion and restore the very important deltoid wrapping angle (which provides a deltoid compressive vector at low angles of abduction; ► Fig. 29.5). In patients without a great tuberosity and instability, the tuberosity should be reconstructed using allograft strut humerus, thereby restoring the stability by promoting the deltoid wrapping angle (► Fig. 29.6).