Mobile colistin resistance (MCR), extended-spectrum beta-lactamase (ESBL) and multidrug resistance monitoring in Escherichia coli (commensal and pathogenic) in pig farming: need of harmonized guidelines and clinical breakpoints.

Abstract

Current data on antimicrobial resistance in pig production is essential for the follow-up strategic programs to eventually preserve the effectiveness of last-resort antibiotics for humans. Here, we characterized 106 Escherichia coli recovered in routine diagnosis (2020-2022) from fecal sample pigs, belonging to 74 Spanish industrial farms, affected by diarrhea. The analysis of virulence-gene targets associated with pathotypes of E. coli, determined 64 as pathogenic and 42 as commensal. Antimicrobial susceptibility testing (AST) performed by minimal inhibitory concentration (MIC) assay, was interpreted by applying breakpoints/cut-off values from the different standards EUCAST/TECOFF 2022, CLSI VET ED5:2020, and CASFM VET2020. Comparisons taking EUCAST as reference exhibited moderate to high correlation except for enrofloxacin, neomycin, and florfenicol. Of note, is the lack of clinical breakpoints for antibiotics of common use in veterinary medicine such as cefquinome, marbofloxacin, or florfenicol. AST results determined multidrug resistance (MDR) to ≥3 antimicrobial categories for 78.3% of the collection, without significant differences in commensal vs pathogenic isolates. Plasmid-mediated mobile colistin resistance gene (mcr) was present in 11.3% of 106 isolates, all of them pathogenic. This means a significant decrease compared to our previous data. Furthermore, 21.7% of the 106 E. coli were ESBL-producers, without differences between commensal and pathogenic isolates, and mcr/ESBL genes co-occurred in 3 isolates. Phylogenetic characterization showed a similar population structure (A, B1, C, D, and E), in both commensal and pathogenic E. coli, but with significant differences for B1, C, and E (38.1 vs 20.3%; 19 vs 1.6%; and 7.1 vs 25%, respectively). Additionally, we identified one B2 isolate of clone O4:H5-B2-ST12 (CH13-223), positive for the uropathogenic (UPEC) status, and in silico predicted as human pathogen. We suggest that a diagnosis workflow based on AST, detection of mcr and ESBL genes, and phylogenetic characterization, would be a useful monitoring tool under a "One-Health" perspective.