We congratulate Hsu et al. for their important contribution to the existing knowledge on the treatment options for failed shoulder arthroplasty with suspected infection. The goal of their study was to answer the highly relevant question of whether single-stage revision shoulder arthroplasty performed because of pain, stiffness, or component loosening in patients without any obvious clinical signs of infection is a feasible treatment option even if Propionibacterium is detected in cultures taken from intraoperative periprosthetic specimens.
Staphylococci used to be the predominant cause for periprosthetic shoulder infections in the early decades of shoulder arthroplasty, as shown in a prospective collection of >2,000 cases of total shoulder arthroplasties over 33 years from the Mayo Clinic1. However, Propionibacterium has caught up1 or has even taken over the lead in recent years2. Instead of an actual shift of the bacterial spectrum, the improvement of microbiological techniques enabling the identification of low-grade infections may have contributed to this change. This hypothesis is supported by the fact that the infections with staphylococci remained steady over time, but the clinically suspected infections with negative cultures decreased1.
Propionibacterium infections after shoulder arthroplasty represent a major diagnostic challenge due to the lack of traditional clinical signs of infection and the limited reliability of routine laboratory examinations, as well as anaerobic growth requirements and slow replication of this pathogen requiring up to 14 days of cultivation3. Despite recent improvement of microbiological techniques, the still-limited sensitivity and specificity of diagnostic tests lead to uncertainty during treatment decisions. For example, the prevalence of unexpected positive cultures in revision shoulder arthroplasty is approximately 15%, with P. acnes being the predominant cause. Although a persistent infection can be demonstrated in only 10% of these cases, 25% of the unexpected positive cultures do not develop any clinical relevance, and the relevance of the remainder of the unexpected positive cultures remains unclear4. No matter how these numbers are interpreted, they clearly show the uncertainty with which surgeons have to deal when cultures of intraoperatively taken tissue samples during revision arthroplasty return positive. Furthermore, in a recent study, Mook et al. showed that, in 18% of primary open shoulder surgical procedures without any sign of infection, at least 1 positive pericapsular tissue culture after prolonged culture incubation can be found, mostly associated with P. acnes5. Therefore, it is still unclear whether Propionibacterium-positive cultures always demonstrate an infection or whether the bacterium can be a commensal of deeper tissue layers. An interesting fact is that, in the same study by Mook et al., a control group using sterile sponges was established and yielded mostly Propionibacterium-positive cultures in 13% of the cases5. Therefore, the possibility of culture contamination and false-positive results should always be kept in mind.
Because of the lack of obvious clinical signs of infection, the chance for false-positive culture results, and the possibility of non-pathogenic Propionibacterium growth in deep-tissue layers, the identification of true low-grade infections remains a challenge and often can only be approached by educated guessing without ever getting to know the actual truth. Therefore, the distinction of the group with low-grade infection (≥2 positive cultures) from the non-infected group (0 or 1 positive culture) in the study by Hsu et al. can also be argued with and probably allocates patients to one group or the other when they would be better kept in the large gray area in between. Nonetheless, single-stage revision seems indeed to be a valid choice when performing revision shoulder arthroplasty in these patients without any obvious clinical signs of infection. When doing so, it is important to actively search for low-grade infections, relying not only on microbiological examinations but also on clinical signs (arthrofibrosis, loosening of the implant, ossifications), cell count in joint aspirate (>2,000 leukocytes/μL or >70% granulocytes), and histopathological examination of periprosthetic tissue. If the infection is diagnosed or suspected, the patients should be treated with antibiotics as a septic revision. As Propionibacterium shows broad antimicrobial susceptibility, eradication of infection is highly successful with single-stage exchange of the prosthesis and a postsurgical treatment with biofilm-active antimicrobials including rifampin. Future research may provide new ways to prevent colonization of the revision prosthesis (e.g., by patient vaccination against Propionibacterium or anti-biofilm implant coating), to improve Propionibacterium detection (e.g., by molecular tests or pathogen-specific markers), and to treat biofilms (e.g., by local antibiotic-delivery systems or bacteriophage therapy).
↵* Disclosure: There was no source of external funding for this work. On the Disclosure of Potential Conflicts of Interest forms, which are provided with the online version of the article, one or more of the authors checked “yes” to indicate that the author had a relevant financial relationship in the biomedical arena outside the submitted work.
- Copyright © 2016 by The Journal of Bone and Joint Surgery, Incorporated