Surveillance of antimicrobial resistance and foodborne pathogens within agri-food production systems

Zoonotic pathogens common among food-producing animals are known to cause foodborne illness in humans. Exploring the routes of transmission for zoonotic pathogens such as Salmonella from food-producing animals to the human food supply may alleviate this issue. Furthermore, the monitoring of antimicrobial resistance (AMR) in Salmonella isolated from food-producing animals is useful to describe the emergence and spread of resistant bacteria within animal populations. The objectives of my studies were to 1) compare two parallel surveillance systems using retrospective data to determine if each system provides different and useful information to describe the burden of AMR in Salmonella isolated from French poultry flocks; and 2) explore a diverse assortment of peripheral lymph nodes (PLN) from cattle fattened in feedlots from areas associated with high and medium Salmonella prevalence to determine the proportion of lymph node types from which Salmonella could be recovered. For the first study, two surveillance systems established in France for monitoring AMR in Salmonella enterica subsp. enterica isolated from agri-food systems were investigated. The voluntary and event-based Salmonella Network (SN) surveillance system was compared to the mandatory surveillance system regulated by the European Union (hereafter referred to as the EU surveillance system). Parallel monitoring provides an excellent opportunity to gain new knowledge in AMR surveillance through the exploration of trends in diversity and resistance profiles, as well as an understanding of the emergence of new phenotypes using retrospective data collected by event-based and mandatory surveillance systems. Three poultry production systems were investigated in this study: laying hens, broilers and fattening turkeys, which have been continuously monitored through both the SN and EU surveillance systems since 2008. The EU surveillance system includes antimicrobial susceptibility testing against a panel of 15 antimicrobials for a minimum of 170 Salmonella isolates per poultry sector each year. Bacterial isolates (e.g., representing animal health, food, or feed) collected through the SN surveillance system are tested for antimicrobial susceptibility for a similar panel of antimicrobials. This study was performed to determine if the two surveillance systems are redundant or if each system provides different and useful information. Several ecological and epidemiological methods were employed to compare directly the phenotypic diversity observed within each surveillance system. Our results suggest that each system is monitoring different aspects of the bacterial populations within each respective poultry population. Using retrospective data from these parallel surveillance programs, we demonstrate that both systems provide different and useful information that may be useful in the ongoing effort to monitor AMR in French poultry production For the second study, peripheral lymph nodes (n=1,270) were collected in two commercial slaughter establishments during the process in which carcasses are disassembled into smaller parts suitable for packaging and distribution. Peripheral lymph nodes (PLN) were collected from six anatomically distinct regions of the carcass. To provide context to previously reported data, approximately 50 subiliac PLN were also collected. Qualitative and quantitative methods were used to detect and estimate the concentration of Salmonella, respectively. The mean weights of PLN surveyed in this study were 0.5 (n= 193), 2.5 (n= 407), 4.1 (n= 200), 12.7 (n= 174), 13.6 (n= 96), and 14.4 (n= 200) grams among the accessory superficial cervical, sternal, paralumbalis, medial iliac, subiliac, and popliteal lymph nodes, respectively. Salmonella was recovered from 0, 0.5 (95%-CI, 0-1.5), 0.5 (0.3-0.8), 1 (0-2.4), 3.4 (0.7-6.2), 3.9 (0.2-7.6), 12.5 (5.9-19.1) and 26.5% (20.4-32.6) of cranial sternal (establishment 1 only), sternal (establishment 2 only), accessory superficial cervical, paralumbalis, medial iliac, caudal sternal (establishment 1 only), subiliac and popliteal lymph nodes, respectively. Few PLN contained quantifiable levels of Salmonella. Of all Salmonella isolates recovered, 22 distinct subtypes were identified. Based on these data, harborage of Salmonella in peripheral PLN does not appear to be uniform; in particular, Salmonella is most likely to be recovered from specific – namely the larger – PLN.