RESUMO
In our tertiary care center, the reported susceptibility of E. coli blood isolates to amoxicillin/clavulanic acid exceeded 90% in 2005 and showed a progressive decrease to 50% by 2017. In this study, we investigate whether there is a real increase in resistant E. coli strains or if this apparent decline in reported susceptibility might be attributed to the substitution of CLSI by EUCAST guidelines in 2014. We randomly selected 237 E. coli blood isolates (stored at - 80 °C) from 1985 to 2018 and reassessed their MIC values, applying both the CLSI (fixed ratio of clavulanic acid) and EUCAST guidelines (fixed concentration of clavulanic acid). In parallel, the susceptibility of these isolates was retested by disk diffusion, according to the EUCAST guidelines. Whole genome sequencing was successfully performed on 233 of the 237 isolates. In only 130 of the 237 isolates (55.0%), testing according to the EUCAST and CLSI criteria delivered identical MIC values for amoxicillin/clavulanic acid. In 64 of the 237 isolates (27.0%), the MIC values diverged one dilution; in 38 (16.0%), two dilutions; and in five (2.1%), three dilutions. From these 107 discrepant results, testing according to EUCAST methodology revealed more resistant profiles in 93 E. coli strains (94.1%). Also, phenotypical susceptibility testing according to EUCAST guidelines tends to correlate better with the presence of beta-lactamase genes compared to CLSI testing procedure. This study highlights the low agreement between EUCAST and CLSI methodologies when performing MIC testing of amoxicillin/clavulanic acid. More strains are categorized as resistant when EUCAST guidelines are applied. The low agreement between EUCAST and CLSI was confirmed by WGS, since most of EUCAST resistant/CLSI sensitive isolates harbored beta-lactamase genes.