Background: In Denmark, antimicrobial resistance (AMR) in pigs has been monitored since 1995 by phenotypic approaches using the same indicator bacteria. Emerging methodologies, such as metagenomics, may allow novel surveillance ways. Aim: This study aimed to assess the relevance of indicator bacteria (Escherichia coli and Enterococcus faecalis) for AMR surveillance in pigs, and the utility of metagenomics. Methods: We collated existing data on AMR and antimicrobial use (AMU) from the Danish surveillance programme and performed metagenomics sequencing on caecal samples that had been collected/stored through the programme during 1999-2004 and 2015-2018. We compared phenotypic and metagenomics results regarding AMR, and the correlation of both with AMU. Results: Via the relative abundance of AMR genes, metagenomics allowed to rank these genes as well as the AMRs they contributed to, by their level of occurrence. Across the two study periods, resistance to aminoglycosides, macrolides, tetracycline, and beta-lactams appeared prominent, while resistance to fosfomycin and quinolones appeared low. In 2015-2018 sulfonamide resistance shifted from a low occurrence category to an intermediate one. Resistance to glycopeptides consistently decreased during the entire study period. Outcomes of both phenotypic and metagenomics approaches appeared to positively correlate with AMU. Metagenomics further allowed to identify multiple time-lagged correlations between AMU and AMR, the most evident being that increased macrolide use in sow/piglets or fatteners led to increased macrolide resistance with a lag of 3-6 months. Conclusion: We validated the long-term usefulness of indicator bacteria and showed that metagenomics is a promising approach for AMR surveillance.
This work was supported by the Danish Veterinary and Food Administration, the Novo Nordisk Foundation (NNF16OC0021856: Global Surveillance of Antimicrobial Resistance) and the European Union’s Horizon-2020 through the One-Health European Joint Programme.
- Antimicrobial resistance
- correlation analyses