Role of inoculum and mutant frequency on fosfomycin MIC discrepancies by agar dilution and broth microdilution methods in Enterobacteriaceae.

OBJECTIVES: Fosfomycin is re-evaluated as a treatment of multidrug-resistant Enterobacteriaceae infections. However, MIC differences have been described among the different susceptibility testing. The aim was to study the role of the different inoculum size used in agar dilution with respect to broth microdilution, according to CLSI, in the fosfomycin MIC discrepancies. METHODS: Fosfomycin MICs were determined using agar dilution (reference) and broth microdilution in 220 Escherichia coli (n=81) and Klebsiella pneumoniae (n=139) clinical isolates. Fosfomycin mutant frequencies were determined in 21 E. coli (MIC=1mg/L) and 21 K. pneumoniae (MIC=16mg/L). The emergence of resistant subpopulations of five E. coli strains (MIC=1mg/L) was monitored over the time by microdilution assay using 0, 4 and 8 mg/L of fosfomycin, and eight different inocula (5×105-3.91×103 CFU/well, 1 : 2 dilutions). RESULTS: For E. coli, 86.4% of categorical agreement (CA), 9.1% very major errors (VME), 3.3% major errors (ME) and 9.9% minor errors (mE) were found. For K. pneumoniae, CA was 51.1%, VME 15.7%, ME 28.4% and mE 25.2%. Essential agreement (±1-log2) was observed in 55.45%. By microdilution, 35.9% of the MICs showed discrepancies of ≥2 dilutions. Initial inoculum used was 5.63 times higher in the microdilution method, in range with CLSI methodology for both techniques. Fosfomycin mutant frequencies were 6.05×10-5 (4×MIC) to 5.59×10-7 (256×MIC) for E. coli, and 1.49×10-4 (4×MIC) to 1.58×10-5 (16×MIC) for K. pneumoniae. Resistant subpopulations arose mainly after 8 h of incubation with inocula >3.13×104 CFU/well. CONCLUSIONS: The higher inoculum used in the microdilution method enriched the initial inoculum with resistant subpopulations and could partially explain the fosfomycin MIC discrepancies with respect to the agar dilution method.

Molecular insights into fosfomycin resistance in Escherichia coli.

Objectives: Fosfomycin activity in Escherichia coli depends on several genes of unknown importance for fosfomycin resistance. The objective was to characterize the role of uhpT , glpT , cyaA and ptsI genes in fosfomycin resistance in E. coli. Methods: WT E. coli BW25113 and null mutants, Δ uhpT , Δ glpT , Δ cyaA , Δ ptsI , Δ glpT-uhpT , Δ glpT-cyaA , Δ glpT-ptsI , Δ uhpT-cyaA , Δ uhpT-ptsI and Δ ptsI-cyaA , were studied. Susceptibility to fosfomycin was tested using CLSI guidelines. Fosfomycin mutant frequencies were determined at concentrations of 64 and 256 mg/L. Fosfomycin in vitro activity was tested using time-kill assays at concentrations of 64 and 307 mg/L (human C max ). Results: Fosfomycin MICs were: WT E. coli BW25113 (2 mg/L), Δ glpT (2 mg/L), Δ uhpT (64 mg/L), Δ cyaA (8 mg/L), Δ ptsI (2 mg/L), Δ glpT-uhpT (256 mg/L), Δ glpT-cyaA (8 mg/L), Δ glpT-ptsI (2 mg/L), Δ uhpT-cyaA (512 mg/L), Δ uhpT-ptsI (64 mg/L) and Δ ptsI-cyaA (32 mg/L). In the mutant frequency assays, no mutants were recovered from BW25113. Mutants appeared in Δ glpT , Δ uhpT , Δ cyaA and Δ ptsI at 64 mg/L and in Δ uhpT and Δ cyaA at 256 mg/L. Δ glpT-ptsI , but not Δ glpT-cyaA , Δ uhpT-cyaA or Δ uhpT-ptsI , increased the mutant frequency compared with the highest frequency found in each single mutant. In time-kill assays, all mutants regrew at 64 mg/L. Initial bacterial reductions of 2-4 log 10 cfu/mL were observed for all strains, except for Δ uhpT-ptsI , Δ glpT-uhpT and Δ uhpT-cyaA . Only Δ glpT and Δ ptsI mutants were cleared using 307 mg/L. Conclusions: Fosfomycin MIC may not be a good efficacy predictor, as highly resistant mutants may appear, depending on other pre-existing mutations with no impact on MIC.