Antimicrobial resistance of Salmonella, Escherichia coli and Campylobacter from pigs on-farm in Alberta and Saskatchewan Canada

Abstract

This cross-sectional study described antimicrobial resistance (AMR) in healthy pigs in 20 Alberta and Saskatchewan herds. All herds used antimicrobials; the daily probability of antimicrobial exposure was 0.8 for nursery pigs and 0.3 for grow-finish pigs. Salmonella spp. (n = 468) were isolated from nursery, grow-finish pigs and sows while Escherichia coli (n = 1439) and Campylobacter spp. (n = 405) were isolated from grow-finish pigs. Fifty-nine percent of the Salmonella were pansusceptible. Isolates from sows were more likely to be pansusceptible than those from other production phases, while Salmonella from nursery pigs were more likely to be multiresistant. All Salmonella and E. coli were susceptible to ceftriaxone and ciprofloxacin, drugs critically important to human medicine, while one E. coli was resistant to ceftiofur. Resistance was most common to tetracycline and sulfamethoxazole (Salmonella, 35% and 27%; E. coli, 68% and 46%). Although often considered an ‘indicator’ organism, E. coli AMR was a poor sentinel for Salmonella AMR at the herd-level. Antimicrobial resistance genes, described in 151 E. coli, were associated in two sets: aadA1 / sul1 / tetA and blaTEM / strA – strB / sul2 / tetB. Associations between genes consistently matched associations between phenotypes suggesting phenotype data may be useful for predicting co-selection. Demonstrating dose-response relationships between various antimicrobial exposures and resistance phenotypes in E. coli reiterated the importance of co-selection. Significant predictors included exposures in other production phases and to unrelated drugs. Four E. coli resistance-phenotypes were associated with macrolide exposure; the most commonly used antimicrobial class in study herds. Additionally, 70% of the Campylobacter were resistant to a macrolide and this resistance was associated with macrolide exposure in nursery pigs. Study herds did not use quinolones. Despite this, 15% of Campylobacter were resistant to a quinolone. Both Campylobacter and E. coli AMR clustered within herds, indicating on-farm interventions could mitigate AMR in pigs.This study described AMR in enteric bacteria from healthy pigs. Identifying dose-response relationships between antimicrobial resistances and exposures to unrelated drugs, and exposures of pigs in different production phases, emphasize the importance of judicious antimicrobial use in pig production.