Orthopedic infections remain a major clinical problem, with catastrophic costs to healthcare systems (Castaneda et al., 2016). Infections associated with orthopedic implants continue to be life-threatening and devastating complications after orthopedic surgery. Staphylococcus aureus (S. aureus) and coagulase-negative staphylococci (CNS) are commonly found bacteria in implant-related infections. Among them, methicillin-resistant S. aureus (MRSA) is known to be the most difficult to treat organism. Previous studies have reported a 72% cure rate for methicillin-sensitive S. aureus (MSSA) infections, whereas the cure rate dropped to 57% for MRSA (Teterycz et al., 2010).

The formation of biofilms on implants is one of the major reasons for failure in cases of implant-associated infection (Masters et al., 2019; Saeed et al., 2019). Biofilms act as a physical barrier and protect bacteria from attack by the host’s immune cells or antimicrobial agents (Zhang et al., 2020) Bacteria on the surface of implants are hidden within the biofilm, preventing host immune cells and antibiotics from attacking them (Costerton, 2005; Flemming and Wingender, 2010; Masters et al., 2019). In addition, the bacteria present in the biofilm acquire pathogenicity and antimicrobial resistance through their ability to exchange information with each other through the quorum sensing mechanism (Rutherford and Bassler, 2012). Biofilm formation occurs in several stages. These stages, which have been observed in both in vitro and in vivo biofilms, begin with initial attachment, followed by proliferation, biofilm formation, and spread of the mature biofilm (Costerton et al., 1987; O’Toole et al., 2000; Nishitani et al., 2015). In the very early stages of infection, when the biofilm is immature, it may be possible to deal with an implant-associated infection by the administration of standard systemic antibiotics. However, once the biofilm matures, it is difficult to eradicate bacteria within the biofilm using standard antimicrobial treatment (Tomizawa et al., 2021). A high-dose, local antibiotic regimen may have the potential to treat mature biofilms, but effective antibiotics and their effective concentrations are not known.

The efficacy of antibiotics against bacteria is commonly represented by the use of the minimum inhibitory concentration (MIC). MIC is defined as “the minimum concentration of an antibiotic to inhibit bacterial growth” (Li et al., 2017). MIC is generally ascertained by exposing defined amounts of bacteria to increasing concentrations of an antibiotic. However, the methods to derive the measurement are performed on planktonic bacteria in a growth medium. Most bacteria are in biofilm form in orthopedic infections and not in planktonic form (Stoodley et al., 2011; Li et al., 2017). Currently available antibiotic treatments are based on the MIC against the planktonic form of bacteria, and these concentrations are less effective against bacteria in the biofilm, which needs to be up to 1000 times higher. (Verderosa et al., 2019) Thus, the usefulness of MIC for orthopedic implant-associated infections may be limited.

The concept of the minimum biofilm eradication concentration (MBEC) has been proposed as an antibiotic concentration to eradicate bacteria in biofilms (Evans and Holmes, 1987; Girard et al., 2010; Macià et al., 2014). Although MBEC seemed more useful in clinical practice, there are still questions regarding the most appropriate methodologies to measure MBEC. In most current MBEC assays, biofilms are formed on abiotic surfaces, such as polystyrene microtiter well plates, which likely have a different affinity for the attachment of bacteria from the materials used in orthopedic implants, such as titanium alloy or stainless steel (Ceri et al., 2001). Many studies have investigated the effects of antibiotics on biofilms in vitro or ex vivo, but no study has investigated the real in vivo biofilms formed in host animals (Fujimura et al., 2008; Rose and Poppens, 2009; Zimmerli and Sendi, 2019). Compared to in vitro biofilms, in vivo biofilms are more complex structures composed of bacterial polysaccharides, proteins, DNA, and host structures such as fibrin, fibrinogen, and collagen (Høiby et al., 2010; Nishitani et al., 2015).

To overcome these issues, this study measured MBEC using several related methodologies, First, MBEC was measured using an in vitro biofilm formed on a stainless-steel screw. Furthermore, an ambitious attempt was made to evaluate MBEC on in vivo biofilms with orthopedic implants using immature and mature in vivo biofilms formed on stainless steel screw implants in a rodent infection model.

A related aim of the study was to investigate the clinical utility of standard of care (SOC) antibiotics on in vivo biofilms. To that end, this study investigated the in vitro and in vivo MBEC of gentamicin (GM), cefazolin (CZ), and vancomycin (VA). A selection of these initial assays was then assessed in combination with a systemic dose of either VA or rifampicin (RF) to reflect the usual clinical practice of accompanying local antibiotic therapy with systemic antibiotics.

The purposes of this study were to establish an MBEC assay for orthopedic implants, with a special focus on using in vivo biofilms formed in rodents, and to investigate the utility of standard of care (SOC) antibiotics on in vivo biofilms.

Two S. aureus strains were used in this study. One was UAMS-1, which is a widely used methicillin-sensitive S. aureus (MSSA) strain isolated from an osteomyelitis patient (Gillaspy et al., 1995), and the other was USA300LAC, which is one of the most prevalent and virulent methicillin-resistant S. aureus (MRSA) strains (Kourbatova et al., 2005). All strains were cultured in a tryptic soy broth (TSB) medium. MIC of two strains was determined by measuring the lowest concentration of antibiotics which prevents visible growth of S. aureus on the TSB agar plate.

In this study, the potential of SOC antibiotics to eradicate biofilm bacteria was evaluated. Notably, in vivo biofilms, which are far more resistant to antibiotics than in vitro biofilms, were formed on the implant in the rodent femur and utilized in this study. As expected, the in vivo MBEC was higher than the in vitro MBEC on stainless steel implants. Not surprisingly, although in vitro MBEC ranged from 64–1024 μg/mL, the in vivo MBEC was as high as 512–4096 μg/mL. When the implant was incubated for the same duration (3 days), bacteria in the in vivo biofilm were more resistant to antibiotics than in the in vitro biofilm. SEM analyses showed that remnants of the biofilm were firmly attached to the implant, and far higher concentrations were needed to eradicate bacteria from the residual biofilm. Among the three antibiotics, only GM had the potential to eradicate USA300LAC on day 14 biofilms in 60% of samples at 1024 μg/mL. This high concentration might be possible with a radical local administration, but cytotoxicity is a concern. A combination of systemic doses of second antibiotics was attempted, and the concentration of GM to obtain 60% effectivity was reduced to 256 μg/mL with RF 1.5 μg/mL in day 14 biofilms of USA300LAC.

We also expected that the MBEC would increase as the biofilm matured; thus, MBEC for day 14 biofilms was expected to be much higher than MBEC for day 3 biofilms, because biofilm maturation is reported at 14 days post-infection (Nishitani et al., 2015). Although this trend was confirmed in most cases, within the in vivo biofilms detailed in Table 2 , GM used on USA300LAC showed the same MBEC₆₀ on days 3 and 14, and CZ used on UAMS-1 had a lower MBEC₁₀₀ on day 14 than it did on day 3. It was speculated that mature biofilms were firmly attached to the implant surface and difficult to remove in these cases. Although VSVM is an established method for removing biofilms, all the biofilms cannot be removed, and some biofilm remains on the implant (Rosa et al., 2019). In the present study, the presence of residual biofilm was confirmed using SEM after VSVM. Therefore, if antimicrobial agents do not penetrate deep into the biofilm, surviving bacteria may multiply again in the residual biofilm. To this end, the implants were cultured again for 24 h after VSVM to examine biofilm persistence. It is noteworthy that persistent biofilms on the implant surface needed a much higher concentration to eradicate the bacteria inside. This can be demonstrated by the finding that VA and CZ were almost ineffective at experimental concentrations of up to 4096 μg/mL in this study. GM was the most effective, but its MBEC₁₀₀ on the residual biofilm was also very high at 2048 μg/mL for day 3 and 4096 μg/mL for day 14 biofilms. Even with the local administration of antibiotics, it seems difficult to reach these high concentrations in actual clinical practice. Therefore, a combination of antibiotics was considered, as a reduction of MBEC was reported by combining multiple antibiotics in previous studies (Deresinski, 2009) (Dall et al., 2018).

We used the concentration of a systemic administration dose to mimic the condition of the combination of local and systemic therapy. As GM had the lowest MBEC among the three antibiotics, the MBEC of GM was evaluated again in combination with a systemic dose of VA and RF. VA and RF were selected because they have been reported to penetrate deep into biofilms (Jacqueline and Caillon, 2014). The addition of a systemic dose of VA did not reduce the MBEC of GM. On the other hand, RF reduced the in vivo MBEC, especially in the in vivo implant MBEC on day 3. The reason for this may be that the bactericidal effect of VA in biofilms is inferior to that of RF and RF is reported to be more effective than VA in MRSA biofilm infections) (Jørgensen et al., 2016). Some in vivo animal studies have demonstrated the benefits of combined treatment with RF, and these studies have reported that the addition of RF significantly reduces CFU (Stavrakis et al., 2014; Thompson et al., 2017). Previous studies have also reported that sublethal doses of VA induce robust biofilm formation through the promotion of extracellular DNA-dependent release (Hsu et al., 2011; Abdelhady et al., 2014). The 20 μg/mL dose used in this study may be insufficient for bacteria in biofilms, but higher blood concentrations increase the risk of side effects. GM is effective in eradicating biofilms formed by S. aureus, including MRSA, and GM’s MBEC has been reported to be lower than that of VA (Mottola et al., 2016; Metsemakers et al., 2018). In addition, the synergistic effect of GM with other antimicrobial agents on staphylococcal biofilms has been reported previously (Dall et al., 2018). GM has a long history of local administration in the form of antibiotic-loaded cements (Vugt et al., 2019), and is also used for systemic administration by intramedullary antibiotic perfusion (Maruo et al., 2021). GM seemed a good candidate for local antibiotic treatment.

This study had some limitations. First, only one bacterial strain of MSSA and MRSA was investigated. Although these bacteria are representative strains that produce robust biofilms (Nishitani et al., 2015), the effectiveness of antibiotics against other S. aureus strains remains unknown. GM was the most effective among the three SOC antibiotics, and whether GM has similar efficiency against GM-resistant MRSA strains is unknown. Moreover, other bacterial strains that cause implant-associated infections have not yet been investigated. However, since we established a novel method to investigate in vivo MBEC with real implants, it became possible to investigate in vivo MBEC with various clinically acquired bacterial strains from patients with implant-associated infections. Second, three SOC antibiotics were selected, but there are many candidate antibiotics, including other anti-MRSA drugs (e.g., linezolid and daptomycin), that might be more effective. Third, stainless steel was used to grow the biofilm, but the MBEC of other implant materials, such as titanium alloy or cobalt-chrome, may be different. Fourth, antibiotic treatment was administered for only 24 h. Some reports have shown that a longer treatment duration reduced the required MBEC (Castaneda et al., 2016; Post et al., 2017). Although very high concentrations were needed for in vivo implant MBEC in this study, these high concentrations may not be needed when biofilms on implants are exposed to antibiotics for longer durations. This study only showed the concentration required to eradicate bacteria in biofilms on orthopedic implants within 24 h of treatment.

In conclusion, a novel MBEC assay for in vivo biofilms on orthopedic implants was developed. GM was the most effective SOC antibiotic against both MSSA and MRSA biofilms in both the in vitro and in vivo experiments. Although in this study GM + RF only showed its greater effect against MRSA on day 14, early initiation of treatment with antimicrobial intervention is still desirable in clinical practice because the required concentration of antibiotics increases with the maturation of the biofilm.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by Kyoto University Animal Experiment Committee.

KN conceived the ideas for experimental designs, analyzed data, and wrote the manuscript. YO conducted the experiments, analyzed data, and wrote the manuscript. AS and YK acquired findings and advised on interpretations. TK contributed to the development of the animal model. TT and MS conducted the experiments. SM contributed the concept and supervised the project. All authors approved the final version of the manuscript.

This study was supported by Grants-in-Aid for Scientific Research (19K09548, 19K09621, 22K09329).

SM received honoraria for meeting chair and a scholarship donation for orthopaedic research from Diichi-Sankyo, outside of this study. This pharmaceutical company has no contribution or influence on this study.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.