B Fantin1, J Poujade2, N Grégoire3, F Chau4, A Roujansky2, N Kieffer5, M Berleur2, W Couet3, P Nordmann6. 1. IAME, INSERM UMR 1137, F-75018, Paris, France; Université Paris Diderot, Sorbonne Paris Cité, F-75018, Paris, France; AP-HP, Groupe Hospitalier Paris Nord Val de Seine, Service de médecine interne, F-92210, Clichy, France. Electronic address: bruno.fantin@aphp.fr. 2. IAME, INSERM UMR 1137, F-75018, Paris, France. 3. University of Poitiers, School of Medicine and Pharmacy, INSERM UMR1070, France. 4. IAME, INSERM UMR 1137, F-75018, Paris, France; Université Paris Diderot, Sorbonne Paris Cité, F-75018, Paris, France. 5. Emerging Antibiotic Resistance Unit, Medical and Molecular Microbiology, Department of Medicine, INSERM European Unit (LEA, IAME), University of Fribourg, Switzerland. 6. IAME, INSERM UMR 1137, F-75018, Paris, France; Emerging Antibiotic Resistance Unit, Medical and Molecular Microbiology, Department of Medicine, INSERM European Unit (LEA, IAME), University of Fribourg, Switzerland; University Hospital and University of Lausanne, Lausanne, Switzerland.
Abstract
OBJECTIVES: Colistin often remains the last resort antibiotic active against carbapenemase-producing Enterobacteriaceae. However, while in vitro inoculum effect has been reported, therapeutic relevance of this phenomenon remains questioned. METHODS: Ten E. coli strains were used that included the wild-type CFT073 and its transconjugant CFT073-MCR-1 and eight susceptible clinical isolates. Mice with peritonitis were treated for 24 h with colistin sulfate. Bacterial loads were determined in peritoneal fluid (PF) and spleen and colistin-resistant mutants were detected. RESULTS: MICs of colistin against the eight susceptible clinical strains and CFT073 ranged from 0.125 to 0.5 mg/L with an inoculum of 105 CFU/mL and from 2 to 4 mg/L with a 107 CFU/mL inoculum; 5/9 strains with an MIC of 4 mg/L were considered resistant according to EUCAST breakpoint (resistance, > 2 mg/L). When the bacterial load of wild-type CFT073 inoculated in mice increased from 107 to 108 CFU: i) mean log10 CFU reduction generated by colistin in PF and spleen decreased from 5.8/mL and 3.1/g, respectively, (p < 0.01) to 0.9/mL and 0.8/g, respectively (NS); ii) mice survival rate decreased from 15/15 (100%) to 6/15 (40%) (p = 0.017); and iii) proportion of mice with selection of colistin-resistant mutants increased from 4/15 to 15/15 (p < 0.01). These results were comparable to those obtained when peritonitis was produced with a 107 CFU bacterial load of E. coli CFT073 expressing mcr-1, for which the mean log10 CFU reductions were 3.5/mL and 0.6/g in PF and spleen, respectively (NS), and survival rate was 8/15 (53%) (p < 0.01 versus survival of mice infected with wild-type CFT073). CONCLUSIONS: Phenotypic colistin resistance in wild-type E. coli due to an increase in inoculum size had a therapeutic impact in mice with peritonitis that was comparable to that observed when the mcr-1 gene was expressed.
OBJECTIVES: Colistin often remains the last resort antibiotic active against carbapenemase-producing Enterobacteriaceae. However, while in vitro inoculum effect has been reported, therapeutic relevance of this phenomenon remains questioned. METHODS: Ten E. coli strains were used that included the wild-type CFT073 and its transconjugant CFT073-MCR-1 and eight susceptible clinical isolates. Mice with peritonitis were treated for 24 h with colistin sulfate. Bacterial loads were determined in peritoneal fluid (PF) and spleen and colistin-resistant mutants were detected. RESULTS: MICs of colistin against the eight susceptible clinical strains and CFT073 ranged from 0.125 to 0.5 mg/L with an inoculum of 105 CFU/mL and from 2 to 4 mg/L with a 107 CFU/mL inoculum; 5/9 strains with an MIC of 4 mg/L were considered resistant according to EUCAST breakpoint (resistance, > 2 mg/L). When the bacterial load of wild-type CFT073 inoculated in mice increased from 107 to 108 CFU: i) mean log10 CFU reduction generated by colistin in PF and spleen decreased from 5.8/mL and 3.1/g, respectively, (p < 0.01) to 0.9/mL and 0.8/g, respectively (NS); ii) mice survival rate decreased from 15/15 (100%) to 6/15 (40%) (p = 0.017); and iii) proportion of mice with selection of colistin-resistant mutants increased from 4/15 to 15/15 (p < 0.01). These results were comparable to those obtained when peritonitis was produced with a 107 CFU bacterial load of E. coli CFT073 expressing mcr-1, for which the mean log10 CFU reductions were 3.5/mL and 0.6/g in PF and spleen, respectively (NS), and survival rate was 8/15 (53%) (p < 0.01 versus survival of miceinfected with wild-type CFT073). CONCLUSIONS: Phenotypic colistin resistance in wild-type E. coli due to an increase in inoculum size had a therapeutic impact in mice with peritonitis that was comparable to that observed when the mcr-1 gene was expressed.