Infection After Total Shoulder Arthroplasty

Pathogens 

The most common pathogens associated with shoulder arthroplasty infections are Gram-positive bacteria. The most common organisms are Staphylococcus aureus, Staphylococcus epidermidis, and Propionibacterium acnes.5 The infecting organism has clinical and prognostic significance regarding both treatment and success. S. aureus is a Gram-positive bacterium that often requires treatment with beta lactamase–resistant penicillin such as nafcillin. S. aureus, because of the increasing prevalence of methicillin-resistant strains, frequently requires treatment with vancomycin. While S. aureus is often antibiotic resistant, it does not usually produce a significant glycocalyx. This lack of a significant glycocalyx allows improved antibiotic penetrance and can allow successful treatment of infection despite implant retention. While infection with S. aureus is the most common, it also has a good prognosis for eradication of the infection.

Staphylococcus epidermidis is a Gram-positive, coagulase negative cocci and is highly antibiotic resistant. S. epidermidis, unlike S. aureus, is particularly adept in producing a robust glycocalyx. The formation of this glycocalyx makes S. epidermidis especially difficult to eradicate in infections around implants.6, 7 The presence of glycocalyx-producing bacteria significantly reduces the efficacy of parenterally administered antibiotics.6, 7 For this reason we recommend that prosthetic infections involving S. epidermidis be treated with implant explantation.8 Propionibacterium acnes has become a well-recognized pathogen in the infection of shoulder arthroplasty. Propionibacterium is an anaerobic Gram-positive rod that is part of the normal skin flora. P. acnes remains very antibiotic susceptible despite the emergence of some tetracycline-resistant strains.

Infection Diagnosis 

Making the diagnosis of infection in the face of a total joint arthroplasty can be challenging and is critical for prompt initiation of treatment. The risk factors that are associated with increased infection rates include revision surgery, systemic steroid use, immunosuppression, multiple steroid injections, and radiation therapy.5 The gold standard for diagnosis of infection is two positive cultures obtained at different time intervals that exhibit the same organism. Unfortunately, this scenario is not common and diagnosis is frequently made with the combination of a positive culture along with clinical suspicion. The use of a positive culture alone to dictate treatment should be undertaken with prudence because of the significant implications of a false-positive result. When taking intraoperative specimens it is important to take a negative control to facilitate interpretation of a positive culture. A negative control can be performed by opening a culture set in the operating room, waving it in the air, and sending it to the laboratory. In this way, the interpretation of a positive culture can be compared with a negative control, thus avoiding a potentially long, protracted, and possibly unnecessary treatment course.

Early signs of clinical infection include pain and swelling surrounding the shoulder. Erythema and purulent drainage often manifest as later signs of infection. In contrast, constitutional signs, including fever and malaise, are often rare presentations of upper extremity infections. Laboratory markers for infection include the use of C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and white blood cell count (WBC). These laboratory results are nonspecific but are useful in contributing to the overall clinical picture and monitoring the response to treatment. The hip arthroplasty literature has support for using ESR and CRP in following treatment, with a normal ESR and CRP being 100% specific for ruling out infection.9 CRP will normalize to less than 3.0 after 3 weeks while ESR may remain elevated for as long as 6 months.10 Unlike ESR and CRP, the leukocyte count has been shown to be unreliable in the detection of infection, with elevated counts in only 15 to 20% of patients.11, 12

The use of aspiration can be an important piece of armamentarium in the diagnosis of total joint infections, but it must be used prudently. The use of preoperative aspirations is particularly helpful and provides a control sample to intraoperative cultures that are subsequently obtained. A consistent organism between two different time points adds significant reliability to the diagnosis of infection and can help to guide antibiotic therapy. While aspiration can be a helpful adjunct, there can be a high false-negative rate that has been reported to be as high as 50% in the series by Coste and colleagues.1, 5

Radiographic signs of infection include radiolucency surrounding the implant–cement or cement–bone interface. While a radiolucent line is not an uncommon occurrence after shoulder arthroplasty, serial radiographs can show a progression of radiolucency. Loosening of the implant, including subsidence or change in implant positioning, is another radiographic marker for infection. While aseptic loosening of an implant is possible, it is uncommon in the early postoperative period. The use of adjunctive radiologic modalities such as technetium and indium-111 scans can contribute to diagnosis. Indium-111 scans have been shown to be more sensitive than technetium scans and have the added ability to differentiate between septic and aseptic loosening.13

Infection Treatment 

Treatment of the infected total arthroplasty of the upper extremity includes both operative and nonoperative treatments. Nonoperative treatment, including the use of suppressive antibiotics, is primarily reserved for patients who are too sick to undergo an operative procedure and who are infected with an organism with low virulence. Operative treatment includes irrigation and debridement along with retention of the prosthesis, as well as single- and two-stage reimplantations.

Resection arthroplasty is an option that gives patients a high success rate of eradicating the infection. Unfortunately, resection arthroplasty has also been associated with unreliable pain relief, a high complication rate, and poor functional outcomes.5, 14 Sperling and colleagues5, 14 have reported a 50% complication rate after resection arthroplasty, including a fracture in 6 of 21 patients. In the same study, however, they also demonstrated no reoperations due to persistent infection.

Irrigation and debridement with retention of the prosthesis can be considered in an acute infection less than 3 weeks old with absence of prosthetic loosening. Sperling and colleagues5 showed a 50% failure rate with this treatment, resulting in continuing infection. In our experience, cases involving S. epidermidis have a poor success rate using this method and we recommend against implant retention unless medically unfeasible.8

While a single-stage reimplantation can be performed, it is usually reserved for patients with significant medical comorbidities that would preclude a second operative procedure. A two-staged reimplantation is the preferred method for treating an infected arthroplasty of the upper extremity.5, 15, 16 The use of a staged reimplantation of shoulder arthroplasties has been demonstrated to show a low recurrence rate of infection (Fig. 1).14, 16 It has also been shown to have superior functional outcomes to a resection arthroplasty.5, 14 These patients, while faring better than with other treatment options, do not typically attain the same preoperative functional level.15 A thorough debridement, including removal of all cement, is performed. The complete removal of the cement is extremely important and can be facilitated with the use of an ultrasound extraction device. After removal of the cement, a complete synovectomy is necessary. A cement spacer can then be fabricated and used until the delayed reimplantation can be performed. The previous implant is used as a template and a cement spacer with the same dimensions is made. The purpose of the anatomic antibiotic spacer is to provide high local concentrations of antibiotics that are not attainable intravenously. The spacer also acts to prevent contracture formation and to maintain the length of the rotator cuff. If the patient is unable to undergo a second operation, an anatomic spacer can also allow a more functional joint.

View Large Image Download to PowerPoint

Figure 1. (A) X-rays from a patient who presented 2 years after total shoulder arthroplasty with a spontaneous loosening and fragmentation of the glenoid fixation. Subsequent operative exploration revealed gross purulence and an infected joint. (B) The proximal humeral prosthesis was removed, and the canal as well as the glenoid was thoroughly irrigated. After this, an antibiotic cement spacer was molded into the shape of an appropriate-sized hemiarthroplasty. (C) Three months later, after a 6-week course of intravenous antibiotics and a 6-week observation period, a revision reimplantation of a total shoulder was performed. The patient went on to have a painless, quite functional shoulder, achieving an elevation of 130°.

Explantation is followed by 6 weeks of appropriate antibiotics. After the completion of antibiotics, reimplantation is performed utilizing a hemiarthroplasty, total shoulder arthroplasty, or reverse total shoulder (Fig. 2). When multiple debridements in the face of infection have been performed, the rotator cuff is often deficient. In these patients, the advent of the reverse total shoulder arthroplasty can be a good option. Clinical outcome studies are needed to investigate this application.

View Large Image Download to PowerPoint

Figure 2. (A) X-rays from a patient who presented with a purulent infection and proximal humeral osteomyelitis after repair of a massive rotator cuff tear. The proximal humerus was osteotomized and a short-stem antibiotic cement spacer was placed. (B) After a period of 3 months, including 6 weeks of intravenous antibiotics, a reverse total shoulder arthroplasty was performed for this cuff-deficient shoulder. (C) The patient now has restoration of full active elevation and a painless shoulder. (Color version of figure is available online.)

Summary 

Infection after arthroplasty of the upper extremity is a difficult clinical problem to undertake. However, many patients can do well after a staged reconstruction with low recurrence of infection and good functional results. While rapid diagnosis and treatment is important, the avoidance of a false positive is equally imperative. Good clinical judgment along with appropriate ancillary tests can contribute to reliable diagnosis and prudent treatment.