Molecular detection of viable bacteria in prosthetic joint infection

Introduction Diagnosis of periprosthetic joint infection (PJI) can be challenging and discrimination between septic and aseptic loosening might be difficult due to the high rate of false negative bacterial culture results particularly when patients have been on long-term antibiotics. Quantitative polymerase chain reaction (qPCR) seems to be a promising technique with higher sensitivity than conventional microbiological methods in the diagnosis of PJI in such cases. However, this technique has proved to have a lower specificity due to, at least in part, detection of DNA from dead bacteria. Propidium monoazide (PMA), a DNA binding reagent, can inhibit the DNA from membrane-compromised cells from being amplified during the PCR. In this study, we have tested the performance of qPCR of clinical samples with and without prior treatment with PMA as a diagnostic tool for PJI diagnosis and compared to the conventional microbiological culture. Methods 208 periprosthetic tissues and/or explanted prosthesis samples were collected from 62 revision episodes from 60 patients undergoing revision arthroplasties due to either PJI or non-infective causes. Tissues were homogenized, and prostheses were sonicated. Homogenates and sonicates were used for aerobic and anaerobic cultures, qPCR, and PMA-qPCR. Our PCR assay included genus-specific primers for staphylococci and enterococci and species-specific primers for Propionibacterium acnes being the commonest PJI causative organisms. Sample analysis was performed blindly without referral to patients’ clinical data which was reviewed later. Results Among the 62 revision episodes, 15 satisfied the Musculoskeletal Infection Society (MSIS) criteria for diagnosing PJI and 46 did not. One patient who was indeterminate as not satisfying the MSIS criteria but with a high clinical suspicion was included in the PJI group for a more conservative data analysis. Two PJI episodes caused by organisms outside our qPCR panel were excluded. Sensitivity of culture, qPCR, and PMA-qPCR were 63%, 71%, and 79% respectively. Their specificities were 96%, 72%, and 89% respectively. Conclusion PMA-qPCR (viability PCR) has the potential to improve the diagnosis of PJI as it increases not only the specificity but also the sensitivity of qPCR with a higher overall accuracy. It could be used as a viability marker to confirm eradication of infection prior to reimplantation in the two-stage treatment protocol of PJI or in post-septic arthritis patients before considering arthroplasty