Assessment of virulence traits in Cutibacterium acnes from shoulder periprosthetic joint infections

Cutibacterium acnes is the most common pathogen in shoulder periprosthetic joint infection (PJI) but remains difficult to distinguish from contamination due to its low virulence and presence as a skin commensal. Diagnostic ambiguity and variable antibiotic resistance further complicate management.

We analyzed 87 C acnes isolates from patients undergoing evaluation for shoulder PJI across 2 institutions. Isolates were obtained from intraoperative cultures and stored for subsequent analysis. Hemolysis, aerotolerance, and growth rate were assessed in vitro. Antimicrobial susceptibility to 5 antibiotics was determined using minimum inhibitory concentration testing, with interpretive breakpoints defined by both the Clinical and Laboratory Standards Institute (CLSI) and European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines. Infection status was classified per the 2018 International Consensus Meeting criteria into definite, probable, possible, or unlikely categories. Associations between phenotypic traits and infection classification were assessed using univariate and multivariable statistical models.

Only 12.6% (n = 11) of cases met criteria for definite infection, while the majority (48.3%) were classified as probable. Hemolysis and aerotolerance were observed in 51.7% and 54.0% of isolates, respectively, and were strongly associated with one another (P < .001); however, neither phenotype was associated with infection classification in univariate or multivariable analysis. Clindamycin resistance was observed in a minority of isolates but varied by standard, with 12.6% classified as resistant by EUCAST and 6.9% as non-susceptible by CLSI. No isolates were resistant to vancomycin; doxycycline and rifampin minimum inhibitory concentrations were uniformly low.

Neither hemolysis nor aerotolerance reliably distinguished definite PJI from nondefinite cases, limiting their diagnostic utility in this cohort. Clindamycin resistance was uncommon, yet classification differed based on the interpretive criteria used (CLSI vs. EUCAST), highlighting potential inconsistencies in clinical reporting. Moreover, the predominance of "probable" infections illustrates a critical limitation of current consensus criteria, which may both underestimate true infection burden in indolent presentations and overestimate infection due to reliance on positive cultures in the absence of clear clinical findings. More precise diagnostic frameworks may help address current classification challenges and better support clinical decision-making in shoulder arthroplasty.