Search

Background The bactericidal activity of an antimicrobial drug is generally assessed by in vitro bacterial time-kill experiments which do not include any components of the immune system, even though the innate immunity, the primary host defence, is probably able to kill a large proportion of pathogenic bacteria in immunocompetent patients. We developed an in vitro tripartite model to investigate the joint action of C57Bl/6 murine bone-marrow-derived macrophages and cephalexin on the killing of Staphylococcus aureus. Results By assessing the bactericidal effects on four bacterial inoculum sizes, we showed that macrophages can cooperate with cephalexin on inoculum sizes lower than 106 CFU/mL and conversely, protect S. aureus from cephalexin killing activity at the highest inoculum size. Cell analysis by flow cytometry revealed that macrophages were rapidly overwhelmed when exposed to large inoculums. Increasing the initial inoculum size from 105 to 107 CFU/mL increased macrophage death and decreased their ability to kill bacteria from six hours after exposure to bacteria. The addition of cephalexin at 16-fold MIC to 105 and 106 CFU/mL inoculums allowed the macrophages to survive and to maintain their bactericidal activity as if they were exposed to a small bacterial inoculum. However, with the highest inoculum size of 107 CFU/mL, the final bacterial counts in the supernatant were higher with macrophages plus cephalexin than with cephalexin alone. Conclusions These results suggest that if the bacterial population at the infectious site is low, as potentially encountered in the early stage of infection or at the end of an antimicrobial treatment, the observed cooperation between macrophages and cephalexin could facilitate its control.

Background Marbofloxacin (MBX), a fluoroquinolone (FQ), is considered as a critical antibiotic requiring antimicrobial susceptibility testing (AST) for prudent use. No clinical breakpoint (CBP) currently exists to interpret the results of such tests in horses. Objectives To compute PK/PD cut-offs (PK/PDCO ) that is one of the three minimum inhibitory concentrations (MICs) considered establishing a CBP for antimicrobial susceptibility test interpretation. Study design A meta-analysis conducted by combining five sets of previously published pharmacokinetic data, obtained in clinical and nonclinical settings. Methods Horses (n = 131) received MBX intravenously at doses of either 2 or 10 mg/kg BW. They were richly sampled (five or six samples per horse). A population model was built to generate a virtual population of 5000 MBX disposition curves by Monte Carlo simulations (MCS) over 24 hours. The selected PK/PD index was the ratio of Area Under the free plasma concentration-time Curve divided by the MIC (fAUC/MIC). The PK/PDCO , which is the highest MIC for which 90% of horses can achieve an a priori selected critical value for the numerical value of the PK/PD index, was established for Gram-positive and Gram-negative bacteria for a dose of 2 mg/kg. Results The PK/PDCO of MBX in horses was 0.125 mg/L for Gram-positive pathogens and 0.0625 mg/L for Gram-negative pathogens. MBX MICs determined by broth microdilution for 54 Escherichia coli and 189 Streptococcus equi isolates are reported. Main limitation No clinical data are taken into account in the determination of a PK/PDco . Conclusion The computed PK/PDco predicts that MBX may be efficacious in horses to treat infections associated with Enterobacteriaceae but unlikely to those involving Staphylococcus aureus or Streptococcus equi.

ObjectivesBacteria developing resistance compromise the efficacy of antibiotics or bacteriophages (phages). We tested the association of these two antibacterials to circumvent resistance.MethodsWith the Hollow Fiber Infection Model (HFIM), we mimicked the concentration profile of ciprofloxacin in the lungs of patients treated orally for Pseudomonas aeruginosa infections and independently, mimicked a single inhaled administration of phages (one or two phages).ResultsEach treatment selects for antibiotic-or phage-resistant clones in less than 30 h. By contrast, no bacteria were recovered from the HFIM at 72 h when ciprofloxacin was started 4 h post-phage administration, even when increasing the initial bacterial concentration by a 1000 fold.ConclusionThe combination of phages with antibiotics used according to clinical regimens prevents the growth of resistant clones, providing opportunities to downscale the use of multiple antibiotics.

In S. aureus biofilms, bacteria are embedded in a matrix of extracellular polymeric substances (EPS) and are highly tolerant to antimicrobial drugs. We thus sought to identify non-antibiotic substances with broad-spectrum activity able to destroy the EPS matrix and enhance the effect of antibiotics on embedded biofilm bacteria. Among eight substances tested, subtilisin A (0.01 U/mL) and calcium gluconate (CaG, Ca2+ 1.25 mmol/L) significantly reduced the biomass of biofilms formed by at least 21/24 S. aureus isolates. Confocal laser scanning microscopy confirmed that they both eliminated nearly all the proteins and PNAG from the matrix. By contrast, antibiotics alone had nearly no effect on biofilm biomass and the selected one (oxytetracycline-OTC) could only slightly reduce biofilm bacteria. The combination of OTC with CaG or subtilisin A led to an additive reduction (average of 2 log10 CFU/mL) of embedded biofilm bacteria on the isolates susceptible to OTC (MBC 10 CFU/mL for 20–25% of the isolates. Further studies are now required to better understand the factors that cause the biofilm produced by specific isolates (20–25%) to be susceptible to the combinations.

International audience

At the request of French Regulatory Authorities, a new formulation of Levothyrox® was licensed in France in 2017, with the objective of avoiding the stability deficiencies of an existing licensed formulation. Before launching the new formulation, an average bioequivalence (ABE) trial was conducted, having enrolled 204 subjects and selected for interpretation a narrow a priori bioequivalence range of 0.90–1.11. Bioequivalence was concluded. In a previous publication, we questioned the ability of an ABE trial to guarantee the switchability within patients of the new and old levothyroxine formulations. It was suggested that the two formulations should be compared using the conceptual framework of individual bioequivalence. The present paper is a response to those claiming that, despite the fact that ABE analysis does not formally address the switchability of the two formulations, future patients will nevertheless be fully protected. The basis for this claim is that the ABE study was established in a large trial and analyzed using a stringent a priori acceptance interval of equivalence. These claims are questionable, because the use of a very large number of subjects nullifies the implicit precautionary intention of the European guideline when, for a Narrow Therapeutic Index drug, it recommends shortening the a priori acceptance interval from 0.80–1.25 to 0.90–1.11.

OBJECTIVE: To compare preparation time, ease of application, and elimination of skin contamination of 3 skin preparation methods for asepsis. STUDY DESIGN: Experimental study. ANIMALS: Healthy dogs (n = 6) with no clinical signs of skin disease. METHODS: Three sites on each dog were randomly allocated to 1 of 3 preparation protocols for asepsis: (1) 5 scrubbings with chlorhexidine gluconate and rinsing (CHXG), (2) washing with mild soap prior to 3 rubbings with hydroalcoholic solution (soap‐HAR), or (3) 3 rubbings with hydroalcoholic solution (HAR). The duration of each method of skin preparation was recorded. A Count‐Tact agar plate was placed in the center of each site before, immediately after, 1 hour after, and 3 hours after antiseptic application. Plates were cultured, and colony forming units (CFU) were counted. RESULTS: Skin preparation lasted an average of 375 seconds for CHXG, 240 seconds for soap‐HAR, and 190 seconds for HAR (P = .00049). Nine CFU (median) were cultured from the skin prior to preparation, with no difference between sites on any animal or for any method. Colony forming units were not detected at any time on any site in any dog after antiseptic application. CONCLUSION: Rubbing with hydroalcoholic (HA) solution was as effective as CHXG and prevented bacterial growth for at least 3 hours under these experimental conditions. Rubbing with hydroalcoholic solution was also faster and easier to perform. CLINICAL SIGNIFICANCE: Because there is currently no known resistance to HA solution, preparation of the surgical site with HAR should be considered to prevent the emergence of bacterial resistance to chlorhexidine as well as potential cross‐resistances to antibiotics. Transfer to clinical animals requires additional investigation.

International audience; In France, more than 2.5 million patients are currently treated with levothyroxine, mainly as the marketed product Levothyrox ®. In March 2017, at the request of French authorities, a new formulation of Levothyrox ® was licensed, with the objective of avoiding stability deficiencies of the old formulation. Before launching this new formulation, an average bioequivalence trial, based on European Union recommended guidelines, was performed. The implicit rationale was the assumption that the two products, being bioequivalent, would also be switchable, allowing substitution of the new for the old formulation, thus avoiding the need for individual calibration of the dosage regimen of thyroxine, using the thyroid-stimulating hormone level as the endpoint, as required for a new patient on initiating treatment. Despite the fact that both formulations were shown to be bioequivalent, adverse drug reactions were reported in several thousands of patients after taking the new formulation. In this opinion paper, we report that more than 50% of healthy volunteers enrolled in a successful regulatory average bioequivalence trial were actually outside the a priori bioequivalence range. Therefore, we question the ability of an average bioequivalence trial to guarantee the switchability within patients of the new and old levothyroxine formulations. We further propose an analysis of this problem using the conceptual framework of individual bioequivalence. This involves investigating the bioavailability of the two formulations within a subject, by comparing not only the population means (as established by average bioequivalence) but also by assessing two variance terms, namely the within-subject variance and the variance estimating subject-by-formulation interaction. A higher within individual variability for the new formulation would lead to reconsideration of the appropriateness of the new formulation. Alternatively, a possible subject-by-formulation interaction would allow a judgement on the ability, or not, of doctors to manage patients effectively during transition from the old to the new formulation.

The increase of multidrug-resistant (MDR) bacteria has renewed interest in old antibiotics, such as minocycline, that can be active against various MDR Gram-negative pathogens. The elimination of minocycline by both kidneys and liver makes it suitable for impaired renal function patients. However, the drawback is the possible elimination of a high amount of drug in the intestines, with potential impact on the digestive microbiota during treatment. This study aimed to predict the potential activity of minocycline against Enterobacterales in the gut after parenteral administration, by combining in vivo and in vitro studies. Total minocycline concentrations were determined by UPLC-UV in the plasma and intestinal content of piglets following intravenous administration. In parallel, the in vitro activity of minocycline was assessed against two Escherichia coli strains in sterilized intestinal contents, and compared to activity in a standard broth. We found that minocycline concentrations were 6–39 times higher in intestinal contents than plasma. Furthermore, minocycline was 5- to 245-fold less active in large intestine content than in a standard broth. Using this PK-PD approach, we propose a preclinical pig model describing the link between systemic and gut exposure to minocycline, and exploring its activity against intestinal Enterobacterales by taking into account the impact of intestinal contents.

International audience; Pharmacokinetic/pharmacodynamic (PK/PD) modelling is the initial step in the semi-mechanistic approach for optimizing dosage regimens for systemically acting antimicrobial drugs (AMDs). Numerical values of PK/PD indices are used to predict dose and dosing interval on a rational basis followed by confirmation in clinical trials. The value of PK/PD indices lies in their universal applicability amongst animal species. Two PK/PD indices are routinely used in veterinary medicine, the ratio of the area under the curve of the free drug plasma concentration to the minimum inhibitory concentration (MIC) (fAUC/MIC) and the time that free plasma concentration exceeds the MIC over the dosing interval (fT > MIC). The basic concepts of PK/PD modelling of AMDs were established some 20 years ago. Earlier studies have been reviewed previously and are not reconsidered in this review. This review describes and provides a critical appraisal of more recent, advanced PK/PD approaches, with particular reference to their application in veterinary medicine. Also discussed are some hypotheses and new areas for future developments.First,a brief overview of PK/PD principles is presented as the basis for then reviewing more advanced mechanistic considerations on the precise nature of selected indices. Then, several new approaches to selecting PK/PD indices and establishing their numerical values are reviewed, including (a) the modelling of time-kill curves and (b) the use of population PK investigations. PK/PD indices can be used for dose determination, and they are required to establish clinical breakpoints for antimicrobial susceptibility testing. A particular consideration is given to the precise nature of MIC, because it is pivotal in establishing PK/PD indices, explaining that it is not a "pharmacodynamic parameter" in the usual sense of this term.

To limit the spread of bacterial diseases in sheep fattening houses, antibiotics are often administered collectively. Collective treatments can be delivered by drinking water but data on the drug&rsquo, s solubility in water or on plasma exposure of the animals are lacking. We first assessed the solubility of products containing sulfadimethoxine (SDM), associated or not with trimethoprim (TMP), in different waters. We then compared in lambs the SDM and TMP pharmacokinetic profiles after individual intravenous (IV) and oral administrations of SDM-TMP in experimental settings (n = 8) and after a collective treatment by drinking water with SDM-TMP or SDM alone in a sheep fattening house (n = 100 for each treatment). The individual water consumption during the collective treatments was also monitored to characterize the ingestion variability. We showed that TMP had a short terminal half-life and very low oral bioavailability, demonstrating that it would be unable to potentiate SDM by oral route. Conversely, SDM had a long terminal half-life of 18 h and excellent oral bioavailability. However, delivery by drinking water resulted in a very high interindividual variability of SDM plasma concentrations, meaning that although disease spread could be controlled at the group level, some individuals would inevitably be under- or over-exposed to the antibiotic.

The use of chlordecone (CLD), a chlorinated polycyclic pesticide used in the French West Indies banana fields between 1972 and 1993, resulted in a long-term pollution of agricultural areas. It has been observed that this persistent organic pollutant (POP) can transfer from contaminated soils to food chain. Indeed, CLD is considered almost fully absorbed after involuntary ingestion of contaminated soil by outdoor reared animals. The aim of this study was to model toxicokinetics (TKs) of CLD in growing pigs using both non-compartmental and nonlinear mixed-effects approaches (NLME). In this study, CLD dissolved in cremophor was intravenously administrated to 7 Creole growing pigs and 7 Large White growing pigs (1 mg kg−1 body weight). Blood samples were collected from time t = 0 to time t = 84 days. CLD concentrations in serum were measured by GCMS/MS. Data obtained were modeled using Monolix (2019R). Results demonstrated that a bicompartmental model best described CLD kinetics in serum. The influence of covariates (breed, initial weight and average daily gain) was simultaneously evaluated and showed that average daily gain is the main covariate explaining inter-individual TKs parameters variability. Body clearance was of 76.7 mL kg−1 d−1 and steady-state volume of distribution was of 6 L kg−1. This modeling approach constitutes the first application of NLME to study CLD TKs in farm animals and will be further used for rearing management practices in contaminated areas.

International audience

International audience

Sheep are commonly used as animal models for human biomedical research, but descriptions of their use for studying the pharmacokinetics of carbapenem antimicrobials, such as ertapenem, are unavailable. Ertapenem is a critical antimicrobial for human infections, and the description of the pharmacokinetics of this drug is of value for research using ovine as models for human diseases, such as urinary tract infections (UTI). There are currently no ovine models for comparative biomedical research of UTI. The objective of this study was to report the pharmacokinetics of ertapenem in sheep after single and multiple dosing. In addition, we explored the effects of an immunomodulatory drug (Zelnate) on the pharmacokinetics of ertapenem in sheep. Eight healthy ewes (weight, 64.4 ± 7.7 kg) were used in an ovine bacterial cystitis model of human cystitis with Pseudomonas aeruginosa. After disease confirmation, each ewe received 1 g of ertapenem intravenously once every 24 h for 5 administrations. Blood was collected intensively (14 samples) during 24 h after the first and last administration. After multiple-dose administration, the volume of distribution was 84.5 mL/kg, clearance was 116.3 mL/h/kg, T1/2(()λ(z)) was 1.1 h, and the extraction ratio was 0.02. No significant differences in pharmacokinetic parameters or time points were found between groups treated with the immunostimulant and controls or after the 1st or 5th administration of ertapenem. No accumulation was noted from previous administration. Our ovine pharmacokinetic findings can be used to evaluate therapeutic strategies for ertapenem use (varying drug dosing schedules and combinations with other antimicrobials or immune modulators) in the context of UTI.

International audience

"Authors'Reply to Coste et al"; International audience

Of the antimicrobials sold to treat food-producing animals in Europe in 2016, 90.1 per cent were designated for group treatment.1 Group treatment, administered through drinking water or feed, enables farmers to easily treat a large number of animals as soon as clinical signs of disease are observed in the group. This practice – referred to as metaphylaxis or control treatment – has the advantage of quickly treating many animals at the same time in order to slow the spread of disease. It has been shown that metaphylactic treatments are associated with higher cure rates than curative treatments administered when animals are already exhibiting clinical signs of disease.2,3 However, the European guidelines for the prudent use of antimicrobials in veterinary medicine4 highlight treatment via feed and drinking water as a major concern in the development of antimicrobial resistance. To help limit the emergence and spread of antimicrobial resistance, it is necessary to control the concentration of antimicrobials in feed and drinking water, the quantity delivered to the animals and the quantity ingested by each animal. In a study summarised on p 405 of this issue of Vet Record , Vandael and colleagues investigate the use of medicated feed and drinking water on Belgian pig farms and discuss the hurdles that have to be overcome to meet these three criteria.5 Vandael and colleagues found discrepancies in the preparation of medicated feed and drinking water. The concentration incorporated into feed seems to be reasonably well controlled, with 68.2 per cent of the pig farms surveyed purchased premixed medicated feed from a manufacturer. However, the use of …

Sheep are commonly used as animal models for human biomedical research, but descriptions of their use for studying the pharmacokinetics of carbapenem antimicrobials, such as ertapenem, are unavailable. Ertapenem is a critical antimicrobial for human infections, and the description of the pharmacokinetics of this drug is of value for research using ovine as models for human diseases, such as urinary tract infections (UTI). There are currently no ovine models for comparative biomedical research of UTI. The objective of this study was to report the pharmacokinetics of ertapenem in sheep after single and multiple dosing. In addition, we explored the effects of an immunomodulatory drug (Zelnate) on the pharmacokinetics of ertapenem in sheep. Eight healthy ewes (weight, 64.4 ± 7.7 kg) were used in an ovine bacterial cystitis model of human cystitis with Pseudomonas aeruginosa. After disease confirmation, each ewe received 1 g of ertapenem intravenously once every 24 h for 5 administrations. Blood was collected intensively (14 samples) during 24 h after the first and last administration. After multiple-dose administration, the volume of distribution was 84.5 mL/kg, clearance was 116.3 mL/h/kg, T1/2(λz) was 1.1 h, and the extraction ratio was 0.02. No significant differences in pharmacokinetic parameters or time points were found between groups treated with the immunostimulant and controls or after the 1st or 5th administration of ertapenem. No accumulation was noted from previous administration. Our ovine pharmacokinetic findings can be used to evaluate therapeutic strategies for ertapenem use (varying drug dosing schedules and combinations with other antimicrobials or immune modulators) in the context of UTI.

International audience

An easily implementable strategy to reduce treatment failures in severe bacterial infections is to combine already available antibiotics. However, most in vitro combination assays are performed by exposing standard bacterial inocula to constant concentrations of antibiotics over less than 24h, which can be poorly representative of clinical situations. The aim of this study was to assess the ability of static and dynamic in vitro Time-Kill Studies (TKS) to identify the potential benefits of an antibiotic combination (here, amikacin and vancomycin) on two different inoculum sizes of two S. aureus strains. In the static TKS (sTKS), performed by exposing both strains over 24h to constant antibiotic concentrations, the activity of the two drugs combined was not significantly different the better drug used alone. However, the dynamic TKS (dTKS) performed over 5 days by exposing one strain to fluctuating concentrations representative of those observed in patients showed that, with the large inoculum, the activities of the drugs, used alone or in combination, significantly differed over time. Vancomycin did not kill bacteria, amikacin led to bacterial regrowth whereas the combination progressively decreased the bacterial load. Thus, dTKS revealed an enhanced effect of the combination on a large inoculum not observed in sTKS. The discrepancy between the sTKS and dTKS results highlights that the assessment of the efficacy of a combination for severe infections associated with a high bacterial load could be demanding. These situations probably require the implementation of dynamic assays over the entire expected treatment duration rather than the sole static assays performed with steady drug concentrations over 24h.

International audience; Given that: (1) the worldwide consumption of antimicrobial drugs (AMDs) used in food-producing animals will increase over the coming decades; (2) the prudent use of AMDs will not suffice to stem the rise in human antimicrobial resistance (AMR) of animal origin; (3) alternatives to AMD use are not available or not implementable, there is an urgent need to develop novel AMDs for food-producing animals. This is not for animal health reasons, but to break the link between human and animal resistomes. In this review we establish the feasibility of developing for veterinary medicine new AMDs, termed "green antibiotics," having minimal ecological impact on the animal commensal and environmental microbiomes. We first explain why animal and human commensal microbiota comprise a "turnstile" exchange, between the human and animal resistomes. We then outline the ideal physico-chemical, pharmacokinetic, and pharmacodynamic properties of a veterinary green antibiotic and conclude that they can be developed through a rational screening of currently used AMD classes. The ideal drug will be hydrophilic, of relatively low potency, slow clearance and small volume of distribution. It should be eliminated principally by the kidney as inactive metabolite(s). For oral administration, bioavailability can be enhanced by developing lipophilic pro-drugs. For parenteral administration, slow-release formulations of existing eco-friendly AMDs with a short elimination half-life can be developed. These new eco-friendly veterinary AMDs can be developed from currently used drug classes to provide alternative agents to those currently used in veterinary medicine and mitigate animal contributions to the human AMR problem.

Resistance in canine pathogenic staphylococci is necessitating re-evaluation of the current antimicrobial treatments especially for biofilm-associated infections. Long, repeated treatments are often required to control such infections due to the tolerance of bacteria within the biofilm. To comply with the goal of better antibiotic stewardship in veterinary medicine, the efficacies of the available drugs need to be directly assessed on bacterial biofilms. We compared the activities of amoxicillin, cefalexin, clindamycin, doxycycline and marbofloxacin on in vitro biofilms of Staphylococcus pseudintermedius and Staphylococcus aureus. Exposure of biofilms for 15 hours to maximum concentrations of the antibiotics achievable in canine plasma only reduced biofilm bacteria by 0.5 to 2.0 log10 CFU, compared to the control, except for marbofloxacin which reduced S. aureus biofilms by 5.4 log10 CFU. Two-antibiotic combinations did not improve, and even decreased, bacterial killing. In comparison, 5 min-exposure to 2 % chlorhexidine reduced biofilms of the 2 tested strains by 4 log10 CFU. Our results showed that S. pseudintermedius biofilm, unlike S. aureus biofilm, was highly tolerant to all the drugs tested, consistent with the treatment failures observed in practice. Under our conditions, the use of topical chlorhexidine would probably be the best currently available strategy to reduce S. pseudintermedius biofilm.

International audience; The efficacy of an early and low inoculum-adjusted marbofloxacin treatment was evaluated on microbiological and clinical outcomes in calves infected with 4.10(7) CFU of Mannheimia haemolytica A1. Twenty-two calves were included based on their rectal temperature rise in the 10 h after challenge and allocated in four groups, receiving a single intramuscular injection of saline (CON), 2 mg/kg marbofloxacin 2-4 h after inclusion (early treatment, E2), 2 or 10 mg/kg marbofloxacin 35-39 h after inclusion (late treatments, L2, L10). In CON calves, M. haemolytica DNA loads in bronchoalveolar lavages continuously increased from inclusion to day 4, and were associated with persistent respiratory clinical signs and lung lesions. At times of early and late treatments, M. haemolytica loads ranged within 3.5-4 and 5.5-6 log10 DNA copies/mL, respectively. Early 2 mg/kg marbofloxacin treatment led to rapid and total elimination of bacteria in all calves. The late treatments induced a reduction of bacterial loads, but 3 of 6 L2 and 1 of 6 L10 calves were still positive for M. haemolytica at day 4. Except for CON calves, all animals exhibited clinical improvement within 24 h after treatment. However, early 2 mg/kg treatment was more efficacious to prevent pulmonary lesions, as indicated by the reduction of the extension and severity of gross lesions and by the histopathological scores. These results demonstrated for the first time that a reduced antibiotic regimen given at an early stage of the disease and targeting a low bacterial load could be efficacious in a natural bovine model of pneumonia.

The antibacterial activity of some antimicrobials may be under-estimated during in vitro microbiological susceptibility tests, due to their instability under such conditions. The in vitro degradation of seven widely used antimicrobials (amoxicillin, cephalexin monohydrate, cefotaxime sodium salt, ciprofloxacin, erythromycin hydrate, clarithromycin, and doxycycline hyclate) and its effect on MIC and MBC determinations was studied using the broth microdilution method, considered as the gold standard for MIC determinations. In vitro concentrations of antimicrobials, over a 24 h incubation period in the medium tested without bacteria, decreased from 33% for ciprofloxacin to 69% for clarithromycin. For cephalexin, cefotaxime, clarithromycin, and doxycycline which were the most degraded drugs, MIC and MBC values for one strain of E. coli and one strain of S. aureus were compared using the standard method or after ad-hoc drug complementation aiming at maintaining constant drug concentration. Abiotic degradation during the standard method was associated with a significant increase of the MIC (2 antibiotics) and MBC (3 antibiotics). However, the observed discrepancy (less than one twofold dilution), even for the most degraded drug for which the concentration at 24 h was reduced by two thirds, suggests that this would only be clinically significant in special cases such as slow-growing bacteria.

Maintaining quinolone concentrations outside the mutant selection window (MSW) between the MIC and mutant prevention concentration (MPC) was suggested by in vitro and in vivo studies to prevent the selection of resistant mutants. However, selection also may depend on the presence of resistant bacterial mutants at the start of treatment, which is highly dependent on the initial inoculum size. In this study, a mouse thigh bacterial infection model was used to test the influence of different exposures to marbofloxacin on the selection of resistant bacteria after infection with a low (10 5 CFU) or high (10 8 CFU) initial inoculum of Escherichia coli . The inoculum size was shown to influence the exposure to marbofloxacin and the values of pharmacokinetic/pharmacodynamic indices. When the abilities of the indices time within the MSW ( T MSW ), area under the concentration-time curve of 0 to 24 h divided by the MIC, and the maximum concentration of drug in plasma divided by the MIC to predict the selection of resistant bacteria were compared, only T MSW appeared to be a good predictor of the prevention of resistance for values less than 30%. When the T MSW was higher than 34%, the selection of resistant bacteria occurred less often in thighs initially infected with the low inoculum (11/24; 46%) than in those infected with the high inoculum (30/36; 80%), suggesting that the selection of resistant mutants depends on both the T MSW and inoculum size. The relevance of these results merits further investigation to test different strategies of antibiotic therapy depending on the expected bacterial burden at the infectious site.

We demonstrate using an in vitro pharmacodynamic model that the likelihood of selection of Escherichia coli mutants resistant to a fluoroquinolone was increased when the initial size of the bacterial population, exposed to fluoroquinolone concentrations within the mutant selection window, was increased.

National audience