Susceptibility of ESBL Escherichia coli and Klebsiella pneumoniae to fosfomycin in the Netherlands and comparison of several testing methods including Etest, MIC test strip, Vitek2, Phoenix and disc diffusion.

Objectives: Fosfomycin susceptibility testing is complicated and prone to error. Before using fosfomycin widely in patients with serious infections, acquisition of WT distribution data and reliable susceptibility testing methods are crucial. In this study, the performance of five methods for fosfomycin testing in the routine laboratory against the reference method was evaluated.
Methods: Ten laboratories collected up to 100 ESBL-producing isolates each (80 Escherichia coli and 20 Klebsiella pneumoniae). Isolates were tested using Etest, MIC test strip (MTS), Vitek2, Phoenix and disc diffusion. Agar dilution was performed as the reference method in a central laboratory. Epidemiological cut-off values (ECOFFs) were determined for each species and susceptibility and error rates were calculated.
Results: In total, 775 E. coli and 201 K. pneumoniae isolates were tested by agar dilution. The ECOFF was 2 mg/L for E. coli and 64 mg/L for K. pneumoniae. Susceptibility rates based on the EUCAST breakpoint of ≤32 mg/L were 95.9% for E. coli and 87.6% for K. pneumoniae. Despite high categorical agreement rates for all methods, notably in E. coli, none of the alternative antimicrobial susceptibility testing methods performed satisfactorily. Due to poor detection of resistant isolates, very high error rates of 23.3% (Etest), 18.5% (MTS), 18.8% (Vitek2), 12.5% (Phoenix) and 12.9% (disc diffusion) for E. coli and 22.7% (Etest and MTS), 16.0% (Vitek2) and 12% (Phoenix) for K. pneumoniae were found. None of the methods adequately differentiated between WT and non-WT populations. Conclusions: Overall, it was concluded that none of the test methods is suitable as an alternative to agar dilution in the routine laboratory.