Periprosthetic Shoulder Infections Are Less Elusive to Diagnostic Arthroscopy

Establishing the diagnosis of deep infection as a complication of shoulder arthroplasty continues to be a challenge. Preoperative evaluation to identify deep infection after arthroplasty involving a lower extremity (hip or knee replacement) relies mainly on radiographic findings as well as various analyses of peripheral blood (CBC [complete blood-cell count], ESR [erythrocyte sedimentation rate], and CRP [C-reactive protein]) and of synovial fluid (cell count, differential cell count, Gram stain, culture, and various other investigational markers). The sensitivity, specificity, and predictive values of these investigations have turned out to be suboptimal when used after shoulder arthroplasty. This is believed to be due partly to the higher prevalence of infection with low-virulence bacteria, primarily Propionibacterium acnes. To further complicate matters, the clinical importance of intraoperative cultures that are unexpectedly positive for P. acnes is not fully understood.

Selected patients presenting with a failed shoulder arthroplasty may be suspected of having a deep infection despite negative preoperative investigations. When the index of suspicion is high, the safest approach may be staged revision surgery, with the goal of the first stage being to remove the failed prosthesis and obtain tissue samples for culture. However, this approach has a substantial number of disadvantages, including increased morbidity, prolonged recovery time, and cost. Repeated surgical violation of the subscapularis to obtain the necessary exposure likely increases the risk of subscapularis failure. Multiple surgical procedures also increase the risk of permanent stiffness. Finally, the decision to remove well-fixed components in the setting of a possible infection needs to be balanced against the potential for bone loss that potentially compromises future reimplantation.

Diagnostic arthroscopy is a very appealing tool in the evaluation of the painful, possibly infected shoulder arthroplasty. It allows tissue samples to be obtained with substantially less morbidity than open surgery. Component fixation and soft-tissue integrity (especially as it relates to the subscapularis and rotator cuff) can be assessed as well. Finally, a shorter operative time and shorter hospital stay both translate into cost savings compared with open resection. The main drawback of diagnostic arthroscopy is that it can potentially result in three surgical procedures if the arthroplasty is found to be infected and is treated with a two-stage reimplantation.

The study by Dilisio et al. highlights the role of diagnostic arthroscopy to identify deep infection complicating anatomic (non-reverse) shoulder arthroplasty. In this study, the authors report on the results of diagnostic arthroscopy performed in nineteen shoulders (sixteen primary and three revision arthroplasties; fourteen total shoulder arthroplasties and five hemiarthroplasties, one with biologic glenoid resurfacing) out of a total of 350 painful anatomic shoulder arthroplasties evaluated during a five-year period. Thus, the authors performed this diagnostic procedure in approximately 5% of their patients a mean of three years (range, 0.7 to 7.7 years) after the index shoulder arthroplasty. None of these patients had concomitant elevation of CBC, ESR, and CRP values. Preoperative aspiration had been performed in fourteen shoulders, with only one being positive for P. acnes.