Methods: C. acnes collected between 2015 and 2017 were correlated with the presence or absence of infection in a detailed retrospective chart review. To determine the C. acnes genotype, bacterial genomic DNA isolated from a single patient isolate served as template in a six locus multiplex touchdown PCR assay using organism specific primers targeting six genes (16S rRNA, ATPase, sodA, Fic toxin, aspD and recA). Isolates were classified as a contaminant in the absence of multiple positive cultures from an anatomic site and without corresponding clinical, laboratory and histopathologic correlates of infection. The assignment of a diagnosis of prosthetic joint infection (PJI) conformed to the definition recommended by the IDSA Clinical Practice Guidelines of PJI.
Results: Of the C. acnes recovered from 94 patients, 14 (14.9%) were from patients with shoulder implants of which shoulder PJI was present in 10 individuals (10.6% of the total). The remaining 84 (89.4%) isolates were retrieved from a variety of tissue and fluid samples of which the majority (65.5%) were deemed as contaminants. Overall, phylogroups IA1, IB and II predominated (79.8%). Although a similar genetic profile was present in all of the shoulder isolates, no phylogroup association was detected with PJI (p<0.72). No genetic difference was present in the lineage of strains not causing PJI compared to those responsible for PJI (p<0.25).
Conclusion: Our results mirror those from a previous investigation using a less robust four gene MLST PCR based scheme that showed a lack of a phylogenetic association with shoulder PJI. Our results are a reflection of the phylogroup composition of the circulating C. acnes sequence types in our community.
M. Desjardins, None
C. Lee, None