Treating a patient with a positive culture for Propionibacterium can feel like a leap of faith. These slow-growing, aerotolerant anaerobes are ubiquitous. They are the most common cause of periprosthetic shoulder infections and were present in 75% of the revision shoulder arthroplasties performed by Dr. Matsen, an author of this study. P. acnes has even been grown on culture of specimens from a high percentage of shoulders without prior surgery, indicating potential culture contamination or the presence of the bacteria as a commensal organism (present but causing no pathological changes)1,2. This leaves open the possibility that Propionibacterium living in the sebaceous glands of the deep dermis is not adequately addressed with routine skin preparation and therefore, as the skin is incised, these oily glands ooze Propionibacterium into the surgical wound3.
Propionibacterium also seems to act differently compared with any other bacteria commonly encountered as a cause of periprosthetic joint infection. Unlike other postsurgical infections, which have a predilection for patients with multiple comorbidities, Propionibacterium infections most commonly affect young males. Also, these infections lack the traditional signs of suppurative infections—i.e., rubor, dolor, calor, and tumor—and there is no commercially available, reliable, preoperative or intraoperative test that can detect P. acnes. Measurements of the erythrocyte sedimentation rate, C-reactive protein level, white blood-cell count (WBC), and WBC differential as well as aspiration and intraoperative frozen sections have all been shown to be unreliable. Only long-hold cultures (grown for 3 weeks under aerobic and anaerobic conditions in this study) can be used to test for the organism, but the false-positive rate of these cultures has been questioned1,4. Without a confirmatory test or any other clinical or laboratory signs, it is an open question whether an individual positive test represents a contaminant, a commensal organism, or a pathogen.
It is easy to doubt the existence of something that is reportedly so ubiquitous but cannot be seen or identified with a clinical examination or a wide range of diagnostic tests. We are tempted to shout “The emperor has no clothes!” This, however, would be a mistake. P. acnes is a real cause of shoulder infection, even if certain patients have positive cultures from a contaminant (caused by either incising the sebaceous glands or a break in the sterility of the sample) or have bacteria present as a commensal organism.
A high percentage (43%) of specimens from culture-positive shoulders in this study showed no growth. It is unclear whether this may be a product of sampling technique/error or whether, for example, a finding that 1 of 5 cultures is positive for broth only may represent a “false-positive” result whereas 5 of 5 cultures being positive for heavy growth may indicate a true infection. Without a confirmatory test, the threshold remains elusive and this differentiation remains purely speculative.
In this debate over what represents a true infection versus a false-positive result, Ahsan et al. stay above the fray while presenting Dr. Matsen’s impressive experience with revision shoulder arthroplasty. Instead of trying to classify which Propionibacterium cultures represent true infections, they propose a straightforward, semiquantitative scheme for quantifying bacterial load in individual cultures and, by combining those results, for the culture specimens overall. Interestingly, they observed no bimodal distribution in these scores—thus indicating no useful threshold for defining a true infection. However, perhaps the greatest impact of this study is to define a strategy that can be used for future research: taking a minimum of 5 samples with separate instruments for each sample and placing them directly into the culture tube, performing long-hold cultures under aerobic and anaerobic conditions, and using the semiquantitative grading scale to characterize the overall bacterial burden.
While the semiquantitative nature of the score is limited by the lack of standardization in the amount of tissue in each sample, this concept is clearly a step in the right direction. Until such time as a confirmatory test for P. acnes becomes widely available, the long-hold cultures remain the only standard. A strategy such as the one described in this study to glean information from these cultures that is more nuanced than a simple binary “culture-positive versus culture-negative” classification may one day provide an understanding of how to distinguish a true infection from a false-positive result.
↵* The author indicated that no external funding was received for any aspect of this work. On the Disclosure of Potential Conflicts of Interest form, which is provided with the online version of the article, the author checked “yes” to indicate that the author had a relevant financial relationship in the biomedical arena outside the submitted work.
- Copyright © 2017 by The Journal of Bone and Joint Surgery, Incorporated