Due to an aging population and advances in medical technology, prosthetic joint placement is increasing in frequency. Prosthetic joint infection (PJI) affects up to 2% of prosthetic joints. Staphylococci are the most common causes of PJI, with Staphylococcus aureus being the most common pathogen overall and Staphylococcus epidermidis being the second most common staphylococcal species involved. Treatment of PJI is problematic due to the limited activity of most antimicrobial agents against microbial biofilms. Rifampin, however, has anti-biofilm activity, and is a cornerstone drug in the management of staphylococcal PJI, when managed with dbridement and implant retention (DAIR). Unfortunately, resistance to rifampin, mediated by single base pair mutations in the target gene, is readily selected and would be expected to compromise management of staphylococcal PJI using a DAIR strategy. Our preliminary data surprisingly suggest, however, that selected rifampin resistance may not necessarily negatively impact future treatment using rifampin-based therapy. Specifically, we have shown that, as expected, rifampin resistance is selected following rifampin monotherapy in a rat model of methicillin-resistant S. aureus foreign body osteomyelitis (FBO). But, surprisingly, rifampin resistance ?disappears? two weeks following cessation of treatment, with only rifampin susceptible S. aureus being detectable. We have shown that rifampin resistance does not confer decreased fitness when bacteria are grown alone in vitro or in vivo; however, rifampin-resistant bacteria are out-competed by rifampin-susceptible bacteria when grown together in vitro and in vivo. And, when treating FBO with two rounds of rifampin monotherapy, resistance is observed at a lower frequency following the second than the first round of rifampin exposure. These observations suggest that rifampin may be a usable agent in biofilm- associated infections in which rifampin resistance has been previously selected. The central hypothesis of our proposed studies is therefore that in experimental FBO, treatment with rifampin can be successful even if rifampin resistance has been previously selected. Our objectives are to develop an experimental FBO model that allows sampling of bacteria in real time from the same animal and to use the model to define the kinetics of selection and disappearance of rifampin resistance. Whereas in traditional osteomyelitis models, microorganisms are analyzed at single time-points (i.e., when animals are sacrificed), the proposed model will allow analysis of bacteria from the same animal over time. We will perform whole genome sequencing on bacteria recovered at multiple time-points during and following one and two rounds of rifampin treatment. This will allow definition of rifampin resistance- and fitness-associated mutations that emerge during and following initial rifampin treatment, and subsequent re-exposure to rifampin. This approach will pinpoint mutations that may occur in rifampin-susceptible bacteria allowing their persistence or conferring upon them a selective advantage over their rifampin-resistant counterparts, even when challenged with rifampin. Finally, we will compare outcomes of FBO treated with either rifampin monotherapy followed by rifampin combination therapy or rifampin combination therapy alone. This will allow us to determine whether rifampin-containing regimens may be an option in the treatment of PJI in which rifampin resistance has been previously selected. Results of the proposed studies will define the kinetics of in vivo selection of rifampin resistance and identify mutations that play a role in out-competition of rifampin-resistant bacteria by those that are rifampin-susceptible. Additionally, our studies will inform clinical treatment options in the scenario of selected rifampin resistance and lay the groundwork for future experiments defining optimal therapy following selection of rifampin resistance.
Rifampin is a key antibiotic in the treatment of staphylococcal prosthetic joint infection; unfortunately, rifampin resistance is easily selected. In our preliminary studies, however, we showed that resistance to rifampin may not be durable and that therefore it may be possible to use rifampin to treat prosthetic joint infection even if rifampin resistance has previously been observed. We propose to develop a novel foreign body osteomyelitis model that will allow monitoring of the kinetics of selection of rifampin resistance; we will use our new model to monitor the microevolution of rifampin resistance using whole genome sequencing. Finally, we will define optimal treatment regimens in the context of selected rifampin resistance.
