Diversity detected in commensals at host and farm level reveals implications for national antimicrobial resistance surveillance programmes. (13th November 2021)

• Record Type: Journal Article
• Authors: Laird, Tanya J.; Jordan, David; Lee, Zheng Zhou; O'Dea, Mark; Stegger, Marc; Truswell, Alec; Sahibzada, Shafi; Abraham, Rebecca; Abraham, Sam
• Is Part Of: Journal of antimicrobial chemotherapy. Volume 77:Number 2(2022)
• Page Start: 400
• Page End: 408
• Publication Date: 2021-11-13
• Subjects: Anti-infective agents -- Periodicals; Chemotherapy -- Periodicals; 615.58
• DOI: 10.1093/jac/dkab403
• Languages: English
• ISSNs: 0305-7453

Abstract

Abstract: Background: A key component to control of antimicrobial resistance (AMR) is the surveillance of food animals. Currently, national programmes test only limited isolates per animal species per year, an approach tacitly assuming that heterogeneity of AMR across animal populations is negligible. If the latter assumption is incorrect then the risk to humans from AMR in the food chain is underestimated. Objectives: To demonstrate the extent of phenotypic and genetic heterogeneity of Escherichia coli in swine to assess the need for improved protocols for AMR surveillance in food animals. Methods: Eight E. coli isolates were obtained from each of 10 pigs on each of 10 farms. For these 800 isolates, AMR profiles (MIC estimates for six drugs) and PCR-based fingerprinting analysis were performed and used to select a subset ( n = 151) for WGS. Results: Heterogeneity in the phenotypic AMR traits of E. coli was observed in 89% of pigs, with 58% of pigs harbouring three or more distinct phenotypes. Similarly, 94% of pigs harboured two or more distinct PCR-fingerprinting profiles. Farm-level heterogeneity was detected, with ciprofloxacin resistance detected in only 60% of pigs from a single farm. Furthermore, 58 STs were identified, with the dominant STs being ST10, ST101, ST542 and ST641. Conclusions: Phenotypic and genotypic heterogeneity of AMR traits in bacteria from animal populations are real phenomena posing a barrier to correct interpretation of data from AMR surveillance. Evolution towards a more in-depth sampling model is needed to account for heterogeneity and increase the reliability of inferences.