Two-stage shoulder arthroplasty after deep infection of the shoulder caused by arthroplasty, osteosynthesis or other surgical procedures: microbial spectrum, complications and functional outcome after at least 1 year follow-up.

PURPOSE Infection represents probably the most challenging complication in shoulder arthroplasty and all other surgical procedures of the shoulder. A deep infection of the shoulder is often combined with a destruction of the joint and a loss of function. In many cases the arthroplasty is the last resort for these patients to obtain a satisfactory function. The objective of this study was to determine outcome data on microbiota and clinical results of a two-stage shoulder arthroplasty procedure after deep infection of the shoulder. METHODS Twenty-six patients with a deep shoulder infection after arthroplasty, osteosynthesis, or rotator-cuff repair were included, while two patients had an empyema without prior surgery. All underwent initial surgical debridement with implantation of an antibiotic-loaded spacer, followed by postoperative systemic antibiotics. The patients obtained definitive shoulder arthroplasty in a second surgery. None of the patients had to undergo more than two surgeries before the arthroplasty. The follow-up analysis including microbiota results, complication rates, and functional outcomes could be determined after at least 1 year in 60% of patients (n = 16). RESULTS The most frequently detected microorganisms were S. epidermidis (31%, n = 10) and Cutibacterium acnes (19%, n = 6). In 28% (n = 9) of cases multi-drug resistant bacteria were detected and in 35% (n = 8) of cases more than one microorganism was found. The overall revision rate was 42%, 11 of 26 patients, in the first 8 weeks after arthroplasty. Reasons for revision were joint dislocations in 23% (n = 6), 15% (n = 4) postoperative hematomas, one (4%) re-infection and one (4%) periprosthetic fracture. At follow-up after 19.3 ± 5.5 months postoperatively, a mean abduction of 113.4°, anteversion of 122.8° and external rotation of 14° was found, with an average age and gender-adjusted Constant Score of the affected shoulder of 63. The subjects' pain and impairment in normal life measured by a questionnaire with school grades were significantly reduced (p < 0.05) as compared to prior shoulder arthroplasty. CONCLUSION The most relevant bacteria in the study cohort were Staphylococcus subspecies and to a lower extent Cutibacterium acnes. A high number of multi-drug resistant and mixed microbiota spectra were detected, defining a need to adapt therapeutic regimen targeting these microbiotas. The two-stage shoulder arthroplasty after a deep infection resulted in excellent infection control with however high early postoperative rates for joint dislocation (23%) and hematomas (15%). At follow-up a stable joint was achieved with a sufficient functional outcome and satisfying outcome for activities of daily living and pain. CLINICAL TRIAL Trial registration number: DRKS00016927, date of registration: 2019/03/19.