ABSTRACT

Objective.

To determine the prevalence and antimicrobial resistance of Escherichia coli and Salmonella spp. in animal feed samples collected between 2018 and 2021 in Colombia.

Methods.

This was a laboratory-based cross-sectional study using routine data from the program for inspection, surveillance, and control of animal feed at the Colombian Agriculture Institute. Samples of animal feed for swine, poultry, canine, feline, leporine, piscine, and equine species were processed for detection of E. coli and Salmonella spp. using enrichment and selective culture methods. Isolates were tested for antimicrobial susceptibility using an automated microdilution method.

Results.

Of 1 748 animal feed samples analyzed, 83 (4.7%) were positive for E. coli and 66 (3.8%) for Salmonella spp. The presence of E. coli and Salmonella spp. was highest in feed for poultry (6.4% and 5.5%) and swine (6.1% and 4.3%). Antimicrobial resistance testing was performed in 27 (33%) E. coli isolates and 26 (39%) Salmonella isolates. Among E. coli, resistance was most frequently observed to ampicillin (44.5%) followed by cefazolin (33.3%), ciprofloxacin (29.6%), ampicillin/sulbactam (26%), and ceftriaxone (11.1%). The highest resistance levels in Salmonella spp. isolates were against cefazolin (7.7%) and piperacillin/tazobactam (7.7%).

Conclusions.

This is the first study from Colombia reporting on the prevalence and antimicrobial resistance of E. coli and Salmonella spp. in animal feed samples. Its results establish a baseline over a wide geographical distribution in Colombia. It highlights the need to integrate antimicrobial resistance surveillance in animal feed due to the emergence of resistant bacteria in this important stage of the supply chain.

Keywords
Drug resistance, microbial; hazard analysis and critical control points; animal feed; one health; operations research; Colombia

RESUMEN

Objetivo.

Determinar la prevalencia y resistencia a los antimicrobianos de Escherichia coli y Salmonella spp. en muestras de piensos para animales tomadas entre el 2018 y el 2021 en Colombia.

Métodos.

Se trata de un estudio transversal realizado en el laboratorio a partir de los datos regulares del programa de inspección, vigilancia y control de alimentos para animales del Instituto Colombiano Agropecuario. Se procesaron muestras de alimentos utilizados en la cría de cerdos, aves de corral, cánidos, félidos, lepóridos, peces y equinos con el fin de detectar E. coli y Salmonella spp. por medio de métodos de enriquecimiento y cultivo selectivo. Se analizó la sensibilidad a los antimicrobianos de las cepas aisladas mediante microdilución automatizada.

Resultados.

De 1748 muestras de alimentos analizadas, 83 (4,7%) resultaron positivas para E. coli y 66 (3,8%) para Salmonella spp. La presencia de E. coli y Salmonella spp. fue mayor en los alimentos para aves de corral (6,4% y 5,5%) y cerdos (6,1% y 4,3%). Se realizaron pruebas de resistencia a los antimicrobianos en 27 (33%) cepas de E. coli y 26 (39%) de Salmonella. En las cepas de E. coli, se observó una mayor resistencia a la ampicilina (44,5%), seguida de la resistencia a la cefazolina (33,3%), la ciprofloxacina (29,6%), la ampicilina/sulbactam (26%) y la ceftriaxona (11,1%). En el caso de las cepas de Salmonella spp., los niveles de resistencia más elevados fueron para la cefazolina (7,7%) y piperacilina/tazobactam (7,7%).

Conclusiones.

Este es el primer estudio realizado en Colombia en el que se informa sobre la prevalencia y la resistencia a los antimicrobianos de E. coli y Salmonella spp. en muestras de alimentos para animales. Sus resultados establecen una línea de base para una zona geográfica mucho mayor dentro de Colombia. Se subraya la necesidad de integrar la vigilancia de la resistencia a los antimicrobianos en los alimentos para animales debido a la aparición de bacterias resistentes en esta importante etapa de la cadena de suministro.

Palabras clave
Farmacorresistencia microbiana; análisis de peligros y puntos de control críticos; alimentación animal; salud única; investigación operativa; Colombia

RESUMO

Objetivo.

Determinar a prevalência e a resistência a antimicrobianos de Escherichia coli e Salmonela spp. em amostras de ração animal coletadas entre 2018 e 2021 na Colômbia.

Métodos.

Estudo transversal de base laboratorial, usando dados de rotina do programa de inspeção, vigilância e controle de ração animal do Instituto Colombiano de Agricultura. Amostras de ração animal para as espécies suína, avícola, canina, felina, leporina, piscina e equina foram processadas para detecção de E. coli e Salmonella spp., usando métodos de enriquecimento e cultura seletiva. Os isolados foram testados quanto à suscetibilidade a antimicrobianos usando um método automatizado de microdiluição.

Resultados.

Das 1.748 amostras de ração animal analisadas, 83 (4,7%) foram positivas para E. coli e 66 (3,8%) para Salmonella spp. A presença de E. coli e Salmonella spp. foi maior em rações para aves (6,4% e 5,5%) e suínos (6,1% e 4,3%). O teste de resistência a antimicrobianos foi realizado em 27 (33%) isolados de E. coli e 26 (39%) isolados de Salmonella. Em E. coli, a resistência observada com maior frequência foi à ampicilina (44,5%), seguida da cefazolina (33,3%), ciprofloxacino (29,6%), ampicilina/sulbactam (26%) e ceftriaxona (11,1%). Os maiores níveis de resistência em isolados de Salmonella spp. foram contra cefazolina (7,7%) e piperacilina/tazobactam (7,7%).

Conclusões.

Este é o primeiro estudo da Colômbia a notificar a prevalência e resistência a antimicrobianos de E. coli e Salmonella spp. em amostras de ração animal. Os resultados estabelecem uma linha de base com ampla distribuição geográfica na Colômbia. Destaca-se a necessidade de integrar a vigilância da resistência a antimicrobianos na ração animal, devido ao surgimento de bactérias resistentes nesta importante etapa da cadeia de abastecimento.

Palavras-chave
Resistência microbiana a medicamentos; análise de perigos e pontos críticos de controle; ração animal; saúde única; pesquisa operacional; Colômbia

The safety of animal feed is important not only for animals but also for human health, as transmission of infections from animals to humans is a known phenomenon with consequent health and economic implications (¹1. Institute of Medicine. Improving Food Safety through a One Health Approach: Workshop Summary. Washington, DC: The National Academies Press; 2012., ²2. Sapkota AR, Lefferts LY, McKenzie S, Walker P. What do we feed to food-production animals? A review of animal feed ingredients and their potential impacts on human health. Environ Health Perspect [Internet]. 2007 May [cited 2022 Jun 25];115(5):663–70. https://doi.org/10.1289/ehp.9760.
https://doi.org/10.1289/ehp.9760... ). Therefore, microbial contamination of animal feed or animal by-products is one of the challenges to One Health, which is defined as the collaborative efforts of multiple disciplines, working locally, nationally, and globally, to attain optimal health for people, animals, and our environment (³3. World Health Organization; Food and Agriculture Organization of the United Nations; World Organisation for Animal Health. Taking a Multisectoral, One Health Approach: A Tripartite Guide to Addressing Zoonotic Diseases in Countries. WHO, FAO, and OIE; 2019.).

Animal feed plays a vital role in maintaining the nutrition and well-being of animals. Its demand has increased dramatically as livestock production and pet ownership are on the rise. Animal feed safety comprises biological hazards such as the presence of pathogenic bacteria like Escherichia coli, Salmonella enterica, Staphylococcus aureus, Listeria monocytogenes, Clostridium perfringens, Clostridium botulinum, Aeromonas, and Campylobacter. The most commonly found organisms in animal feed are E. coli and Salmonella spp. (⁴4. Molina-Alvarado A, Granados-Chinchilla F. Inocuidad microbiológica de los alimentos para animales en Costa Rica. Nutr Anim Trop. 2015;9(Suppl. 1):13–31. https://doi.org/10.15517/nat.v9i3.22280.
https://doi.org/10.15517/nat.v9i3.22280... , ⁵5. Heredia N, García S. Animals as sources of food-borne pathogens: A review. Anim Nutr. 2018 Sep 1;4(3):250–5.). E. coli is part of the gastrointestinal microbiota of humans and animals, and in some cases the bacterium can be pathogenic. S. enterica can cause salmonellosis in both humans and animals. Transmission of both is usually through contaminated water or food products or by fecal–oral transmission.

The additional problem associated with these pathogens is that of antimicrobial resistance (AMR) (⁶6. Junod T, López-Martin J, Gädicke P. [Antimicrobial susceptibility of animal and food isolates of Salmonella enterica]. Rev Med Chil. 2013;141(3):298–304.). Infections caused by resistant bacteria pose a challenge to antimicrobial treatment, requiring selection of effective alternatives. The resistant pathogens contaminating animal feed could spread through the food chain, from primary production to consumers and to humans in direct contact with colonized and infected animals (⁷7. Founou LL, Founou RC, Essack SY. Antibiotic resistance in the food chain: A developing country-perspective. Front Microbiol. 2016 Nov 23;7:1881.). This potential hazard may contribute to enhance AMR in human pathogens, possibly increasing the cost and duration of treatment and leading to unfavorable treatment outcomes (⁴4. Molina-Alvarado A, Granados-Chinchilla F. Inocuidad microbiológica de los alimentos para animales en Costa Rica. Nutr Anim Trop. 2015;9(Suppl. 1):13–31. https://doi.org/10.15517/nat.v9i3.22280.
https://doi.org/10.15517/nat.v9i3.22280... ). Consequently, microbiological evaluation is essential to ensure animal feed is not a source of E. coli and Salmonella spp. contamination (⁵5. Heredia N, García S. Animals as sources of food-borne pathogens: A review. Anim Nutr. 2018 Sep 1;4(3):250–5., ⁸8. Lazo Pérez L, Sánchez Álvarez C, Díaz M, Madrigal W, Fernández W, Sotelo JA, et al. Factores de riesgo y vulnerabilidad que influyen negativamente sobre la inocuidad de alimentos balanceados en la fábrica de piensos de la provincia de Villa Clara. Rev Electron Vet. 2010;11(3B):1–7.) or a propagator of AMR.

Recent studies have shown the presence of E. coli and Salmonella spp. in animal feed and have further documented their AMR profiles (⁹9. de Mesquita Souza Saraiva M, Lim K, do Monte DFM, Givisiez PEN, Alves LBR, de Freitas Neto OC, et al. Antimicrobial resistance in the globalized food chain: a One Health perspective applied to the poultry industry. Braz J Microbiol [Internet]. 2022 [cited 2022 May 24];53(1):465–86. https://doi.org/10.1007/s42770-021-00635-8.
https://doi.org/10.1007/s42770-021-00635... ). In a study from Kenya, 58% and 28% of poultry feed samples were contaminated with E. coli and Salmonella spp., respectively. The highest resistance was against ampicillin, at 62% for E. coli isolates and 41% for Salmonella spp. (¹⁰10. Ngai DG, Nyamache AK, Ombori O. Prevalence and antimicrobial resistance profiles of Salmonella species and Escherichia coli isolates from poultry feeds in Ruiru Sub-County, Kenya. BMC Res Notes. 2021;14(1):41.). Another study from the United States of America on animal feed also reported the presence of both organisms, where the frequency of bacteria resistant to two or more antimicrobials was 39.3% for E. coli and 22.9% for Salmonella spp. (¹¹11. Ge B, Domesle KJ, Gaines SA, Lam C, Bodeis Jones SM, Yang Q, et al. Prevalence and Antimicrobial Susceptibility of Indicator Organisms Escherichia coli and Enterococcus spp. Isolated from U.S. Animal Food, 2005-2011. Microorganisms. 2020 Jul 15;8(7):1048. https://doi.org/10.3390/microorganisms8071048.
https://doi.org/10.3390/microorganisms80... ). Furthermore, AMR to different groups of antimicrobials, including first-line antimicrobials, has been reported in Salmonella spp. from animal feed (⁶6. Junod T, López-Martin J, Gädicke P. [Antimicrobial susceptibility of animal and food isolates of Salmonella enterica]. Rev Med Chil. 2013;141(3):298–304.). There is no previously published information about the prevalence of AMR in E. coli and Salmonella in animal feed samples in Colombia or in Latin America.

The Colombian Agriculture Institute (ICA) is the institution responsible for inspection, surveillance, and risk-based control of companies that produce and import animal feed in Colombia. Samples of animal feed for several species including swine, poultry, canine, feline, leporine, piscine, and equine species are tested. To address the safety of animal feed in Colombia and understand the AMR profiles of contaminating pathogens, this study aimed to determine the number (and proportion) of samples of animal feed contaminated with E. coli and Salmonella spp. and to describe their AMR characteristics.

MATERIALS AND METHODS

This was a laboratory-based cross-sectional study using routinely collected data from the inspection, surveillance, and control of animal feed program carried out by the National Laboratory of Livestock Supplies (LANIP) within ICA.¹1The dataset used in this article can be made available on request to the corresponding author. The samples analyzed were obtained between January 2018 and November 2021 and covered 27 of the 32 Colombian administrative departments (provinces) (¹²12. Instituto Colombiano Agropecuario. Resolución 061252 del 03 de febrero de 2020. Bogotá: ICA; 2020., ¹³13. Instituto Colombiano Agropecuario. Procedimiento Interno de toma de muestras de alimentos balanceados, suplementos alimenticios y sales mineralizadas para animales. Bogotá: ICA; 2019.). The other five departments were excluded from the sampling because animal production in these provinces is very low, and they do not have factories for the production of animal feed and self-consumption.

Setting

In Colombia, approximately 800 companies produce and market animal feed, of which 400 distribute products to be consumed at the national level and the other 400 import products into the country. In addition, 65 farms produce their own feed products, and they are considered “self-consumption” companies.

ICA works at the national level to ensure the safety of agricultural production in the Colombian countryside. LANIP within ICA verifies the quality of supplies for livestock and the safety of products of animal origin in their primary production phase, to prevent risks to animal and human health. At LANIP, the control of animal feed is carried out following the national sampling plan elaborated and executed by the Technical Directorate of Security and Livestock Supplies of ICA (¹³13. Instituto Colombiano Agropecuario. Procedimiento Interno de toma de muestras de alimentos balanceados, suplementos alimenticios y sales mineralizadas para animales. Bogotá: ICA; 2019.).

Sampling

The study comprised samples of animal feed destined for consumption by swine, poultry, canine, feline, leporine, and equine species. The samples originated from 27 of the 32 provinces of Colombia. The type of animal feed was classified according to its physical appearance as either pellet or flour. Animal feed in flour form originated only from self-consumption companies, as defined above.

Samples were collected under the ICA inspection, surveillance, and control of animal feed program according to the guidelines in the ICA resolution 61252 of 2020 (¹²12. Instituto Colombiano Agropecuario. Resolución 061252 del 03 de febrero de 2020. Bogotá: ICA; 2020., ¹⁴14. Instituto Colombiano Agropecuario. Resolución 1056 de 1996. Por la cual se dictan disposiciones sobre el control técnico los Insumos Pecuarios y se derogan las Resoluciones No. 710 de 1981, 2218 de 1980 y 444 de 1993. Bogotá: ICA; 1996.). Samples were collected in situ by ICA staff under aseptic conditions from complete packaged animal feeds received from production companies, chain stores, agricultural distributors, and small stores. One sample consisted of a total of 500 g drawn from 4–16 subsamples proportionally combined. These subsamples correspond to the number of units selected per batch according to the guidelines. The final sample was packaged in a sterile bag and after labeling, shipped to LANIP for analyses.

Upon arrival at the laboratory, only samples that fulfilled the ICA requirements were included. The requirements included samples with a size of 500 g. The samples were analyzed in the food microbiology laboratory of LANIP in Cundinamarca, Colombia. Samples were stored at room temperature (20–30 °C) until they were analyzed.

Isolation of bacteria

Twenty-five grams of each sample was pre-enriched in 225 mL of buffered peptone water and left incubating for 24 hours at 37 °C for Salmonella spp. and E. coli. For Salmonella isolation, 0.1 mL of pre-enriched buffered peptone water was transferred to Rappaport broth and 1 mL to selenite broth. The Rappaport broth was incubated at 42 °C and selenite broth at 37 °C. Both broths were incubated for 24 hours. Then, 10 µL of Rappaport broth was streaked on XLD agar and 10 µL of selenite broth on Hektoen agar. Presumptive Salmonella spp. colonies were identified by their H₂S production (black colonies) in XLD agar and Hektoen agar. This method was based on the ISO 6579:2002 Salmonella method (¹⁵15. International Organization for Standardization. Microbiology of food and animal feeding stuffs - Horizontal method for the detection of Salmonella spp. ISO 6579:2002. Geneva: ISO; 2002. Available from: https://www.iso.org/standard/29315.html.
https://www.iso.org/standard/29315.html... ).

For E. coli isolation, samples were processed according to ICONTEC-NTC 5652 (¹⁶16. Colombian Institute of Technical Standards and Certification. Microbiología de alimentos y alimentos para animales. Métodos horizontales para la detección y enumeración de enterobacterias. Parte 1: Detección y enumeración mediante la técnica de NMP con pre-enriquecimiento. NTC 5652:2009. Bogotá: Icontec; 2009.). One mL was transferred to BRILA broth for E. coli and incubated at 42 °C. Then, 10 µL was streaked on EMB agar and MacConkey agar. Typical colonies had a metallic shine on EMB, and typical lactose-positive colonies were pink on MacConkey agar.

Antimicrobial susceptibility testing

Investigation for AMR in E. coli and Salmonella spp. isolates was performed as an addition to the routine surveillance in animal feed described above. Isolates were tested for antimicrobial susceptibility using the Becton Dickinson Phoenix™ Automated Microbiology System. Dilutions of presumptive E. coli or Salmonella spp. colonies were made in Phoenix™ ID Broth (Becton Dickinson) at a concentration of 0.45–0.5 on the MacFarland scale according to the manufacturer’s instructions. The bacterial suspensions were added to the panel for identification of microorganisms and antimicrobial susceptibility NMIC/ID 406 for Gram-negative bacteria (Becton Dickinson). The panel determined minimal inhibitory concentrations (MICs) by broth microdilution of the following antimicrobials: amikacin, ampicillin, ampicillin/sulbactam, cefazolin, cefepime, cefoxitin, ceftazidime, ceftriaxone, ciprofloxacin, ertapenem, gentamicin, imipenem, meropenem, piperacillin/tazobactam, tigecycline, and trimethoprim/sulfamethoxazole. The panel was incubated in the Phoenix™ according to manufacturer instructions for 24 hours at 35 °C. Resistance breakpoints were evaluated according to the guidelines of the Clinical and Laboratory Standards Institute (¹⁷17. Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Susceptibility Testing. 32nd edition. Berwyn, PA: CLSI; 2022.). For isolates of porcine origin, resistance to the third-generation cephalosporins ceftazidime and ceftriaxone was evaluated with the CLSI guidelines for bacteria of veterinary origin VET01S-Ed.5 (¹⁸18. Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Bacteria Disk and Dilution Susceptibility Tests for Bacteria Isolated from Animals. 5th edition. CLSI supplement VET01S. Berwyn, PA: CLSI; 2020.).

Additionally, biochemical screening of carbapenemase production in Salmonella spp. isolates was performed with Carba NP test, which is based on the detection of the hydrolysis of the β-lactam ring of a carbapenem (imipenem) by carbapenemase enzymes produced by the bacteria, changing the color of the medium from red to yellow through decreases in the pH of the medium (¹⁹19. Poirel L, Nordmann P. Rapidec Carba NP Test for Rapid Detection of Carbapenemase Producers. J Clin Microbiol. 2015 Sep;53(9):3003–8.).

Data management and analysis

Metadata were extracted from the electronic database that was used for data collection during the surveillance activities of ICA. Data were summarized as counts and proportions using EpiData analysis software (version 2.2.2.181, EpiData Association, Odense, Denmark).

Ethical considerations

Ethics approval was obtained from the Institutional Review Committee of Secretaría Distrital de Salud in Colombia (SDSCTI2021009), The Union Ethics Advisory Group of the International Union Against Tuberculosis and Lung Disease (EAG28/21), and the Ethics Review Committee of the Pan American Health Organization (PAHOERC.478.01).

RESULTS

A total of 1 748 animal feed samples were analyzed at LANIP-ICA from January 2018 to November 2021. The number of samples tested across the years was similar, though fewer samples were analyzed in 2020 (n = 230). Of all destined animal species, feed for poultry (n = 565) and swine (n = 610) contributed over 65% of samples (Table 1).

Overall, 83 (4.7%) samples were positive for E. coli and 66 (3.8%) for Salmonella spp. The prevalence of E. coli was highest in 2018 (9%) and lowest in 2019 (1.2%). Similarly, the prevalence of Salmonella spp. was highest in 2018 (7.1%) and lowest in 2021 (0.7%). The prevalence of E. coli and Salmonella spp. was highest in feed for poultry (6.4% and 5.5%, respectively) followed by feed for swine (6.1% and 4.3%, respectively).

Among the two types of feed (flour or pellet), flour from self-consumption companies showed the highest prevalence of E. coli (7.8%) and Salmonella spp. (5.3%) (Table 1).

Cundinamarca (n = 281), Antioquia (n = 272), Valle del Cauca (n = 260), and Santander (n = 128) were the provinces that contributed the highest number of samples. The presence of E. coli in these provinces ranged from 3.9% (Cundinamarca) to 6.9% (Valle del Cauca) while the presence of Salmonella ranged from 0.8% (Santander) to 5.5% (Antioquia). The samples from Arauca, Cauca, Cesar, Guainía, and Guaviare provinces did not show E. coli or Salmonella (Table 2).

Of 83 E. coli isolates, 27 (33%) were subjected to antimicrobial susceptibility testing. Resistance to ampicillin (44.5%) was most frequently observed, followed by cefazolin (33.3%), ciprofloxacin (29.6%), and ampicillin/sulbactam (26%) (Table 3). Among E. coli isolates, 14 (40.7 %) were susceptible to all antimicrobials tested.

Of 66 Salmonella spp. isolates, 26 (39%) were subjected to antimicrobial susceptibility testing. Resistance to cefazolin (7.7%) and piperacillin/tazobactam (7.7%) was most frequently observed. Unlike E. coli, where no isolates showed intermediate resistance to any antimicrobial agent, intermediate levels of resistance were observed among Salmonella spp. for imipenem (27%), ciprofloxacin (7.7%), cefepime (7.7%), and meropenem (4.3%) (Table 4). Among Salmonella spp. isolates, 11 (42.3%) were susceptible to all antimicrobials tested.

Six of seven Salmonella spp. isolates with intermediate susceptibility to imipenem and meropenem yielded negative results for carbapenemase activity using the Carba NP test, indicating that this enzymatic mechanism was not involved. One of these showed indeterminate results. Further molecular-based studies are required to characterize this isolate.

DISCUSSION

This is the first study from Colombia reporting on the prevalence and AMR pattern of E. coli and Salmonella spp. in animal feed samples. We found that 4.7% samples were positive for E. coli and 3.8% for Salmonella spp. Among E. coli isolates, resistance was observed for beta-lactam drugs and fluoroquinolones. Among Salmonella isolates, resistance was observed to some beta-lactam drugs. Intermediate levels of resistance to fourth-generation cephalosporins, carbapenems, and fluoroquinolones were observed.

The prevalence rates of E. coli and Salmonella spp. reported in our study are lower than those reported by the U.S. Food and Drug Administration (FDA) animal food surveillance program: 12.5% and 12% for E. coli and Salmonella spp., respectively (¹¹11. Ge B, Domesle KJ, Gaines SA, Lam C, Bodeis Jones SM, Yang Q, et al. Prevalence and Antimicrobial Susceptibility of Indicator Organisms Escherichia coli and Enterococcus spp. Isolated from U.S. Animal Food, 2005-2011. Microorganisms. 2020 Jul 15;8(7):1048. https://doi.org/10.3390/microorganisms8071048.
https://doi.org/10.3390/microorganisms80... , ²⁰20. Li X, Bethune LA, Jia Y, Lovell RA, Proescholdt TA, Benz SA, et al. Surveillance of Salmonella prevalence in animal feeds and characterization of the Salmonella isolates by serotyping and antimicrobial susceptibility. Foodborne Pathog Dis. 2012;9(8):692–8.). Also, our estimate seems much lower than estimates reported in other studies, such as 39.3% (E. coli) and 22.9% (Salmonella spp.) reported in 2012 in processing plants in the United States of America (²¹21. Ge B, LaFon PC, Carter PJ, McDermott SD, Abbott J, Glenn A, et al. Retrospective analysis of Salmonella, Campylobacter, Escherichia coli, and Enterococcus in animal feed ingredients. Foodborne Pathog Dis [Internet]. 2013 Aug [cited 2021 Aug 12];10(8):684–91. https://doi.org/10.1089/fpd.2012.1470.
https://doi.org/10.1089/fpd.2012.1470... ), and 58% (E. coli) and 28% (Salmonella spp.) reported in a study from Kenya (¹⁰10. Ngai DG, Nyamache AK, Ombori O. Prevalence and antimicrobial resistance profiles of Salmonella species and Escherichia coli isolates from poultry feeds in Ruiru Sub-County, Kenya. BMC Res Notes. 2021;14(1):41.).

In general, in this study we observed a lower prevalence of E. coli and Salmonella spp. across the years. Interestingly, fewer samples tested positive for Salmonella spp. in 2021 compared to the earlier years, even though there was a decrease in the number of samples tested in 2020 due to the COVID-19 pandemic. These results could have been influenced by the fact that, in addition to the monitoring activities, ICA supports producers with training in implementation of good manufacturing practices (GMP). This could have been complemented by the issuing of resolutions such as 61252 of 2020, in which the regulations of inspection, surveillance, control, and implementation of GMP were reinforced with the purpose of guaranteeing the safety of animal feed (¹⁴14. Instituto Colombiano Agropecuario. Resolución 1056 de 1996. Por la cual se dictan disposiciones sobre el control técnico los Insumos Pecuarios y se derogan las Resoluciones No. 710 de 1981, 2218 de 1980 y 444 de 1993. Bogotá: ICA; 1996.). Another factor could be the emerging willingness of Colombian companies to export their products, thus being subject to more stringent quality control. All these factors might have contributed to improved food safety and a lower prevalence of microbial contamination. However, despite the low prevalence of E. coli and Salmonella spp. in animal feed, this is still above the desired levels, as the Colombian regulation mandates that E. coli and Salmonella spp. should be absent in any 25 g of sample (²²22. Instituto Colombiano Agropecuario. Directivas técnicas de alimentos para animales y sales mineralizadas [Internet]. Bogotá: ICA; 1999. Available from: https://www.ica.gov.co/getdoc/7d27ee5e-cfe4-47a2-868e-7c53f4e49473/directivastecnicasalimentosanimales.aspx.
https://www.ica.gov.co/getdoc/7d27ee5e-c... ). This regulation mandating the absence of Salmonella spp. in a sample is reflected in other guidelines, such as those of the European Food Safety Authority, which emphasize that eliminating Salmonella spp. in animal feed before it gets to the farm will contribute to reducing the presence of Salmonella spp. in food (²³23. European Food Safety Authority. Microbiological risk assessment in feedingstuffs for food-producing animals - Scientific Opinion of the Panel on Biological Hazards. EFSA J. 2008;6(7):1–84.).

The highest number of isolates for both Salmonella spp. and E. coli was found in the flour type of feed. The production of this type of feed is destined for self-consumption. Generally, its production is not done according to commercial standards and is less well-regulated compared to other types of feed. Production is carried out through mechanical homogenization and high temperatures are not usually used. Additional reasons for the contamination of this type of feed could include failures in the cleaning of production lines, poor rodent control practices, storage on the floor and not in food stowage, and poor packaging and storage conditions (²⁴24. Munoz LR, Pacheco WJ, Hauck R, Macklin KS. Evaluation of commercially manufactured animal feeds to determine presence of Salmonella, Escherichia coli, and Clostridium perfringens. J Appl Poult Res. 2021;30(2):100142.).

Feeds for swine and poultry constituted the largest proportion of the samples, as these are produced in larger quantities than feeds destined for other species. Further, E. coli and Salmonella spp. were isolated more commonly in this type of feed. This finding is in accordance with the findings for the flour type of feed, which is mostly destined for these animals.

Nariño was the province with the highest percentage of bacterial isolates of E. coli (9.8%) and Salmonella spp. (7.3%). The feed production in this province concentrates mostly on food destined for leporine species. We therefore recommend further investigation in the production in this province to identify any specific source of the pathogens during production of feed for leporine species.

We found resistance to ampicillin, ampicillin/sulbactam, cefazolin, and ciprofloxacin in E. coli. The highest level of resistance was observed against ampicillin, followed by cefazolin and ciprofloxacin. High levels of resistance to ampicillin have also been reported in a study from Kenya (¹⁰10. Ngai DG, Nyamache AK, Ombori O. Prevalence and antimicrobial resistance profiles of Salmonella species and Escherichia coli isolates from poultry feeds in Ruiru Sub-County, Kenya. BMC Res Notes. 2021;14(1):41.). However, levels of resistance in Salmonella spp. to other drugs in the Kenyan study differed from ours. For instance, no resistance against ciprofloxacin was observed in that study. Our results also contrast with a previous study from the FDA animal food surveillance program, where the highest levels of resistance were observed for tetracycline (11.2%) and streptomycin (4.6%). In the FDA study, low levels of resistance to most beta-lactam drugs, including ampicillin (2.9%), were observed (¹¹11. Ge B, Domesle KJ, Gaines SA, Lam C, Bodeis Jones SM, Yang Q, et al. Prevalence and Antimicrobial Susceptibility of Indicator Organisms Escherichia coli and Enterococcus spp. Isolated from U.S. Animal Food, 2005-2011. Microorganisms. 2020 Jul 15;8(7):1048. https://doi.org/10.3390/microorganisms8071048.
https://doi.org/10.3390/microorganisms80... ).

The AMR levels for E. coli observed in our study seem to be specific to the situation in Colombia, and could be related to the use of antimicrobials such as ampicillin, ciprofloxacin, and trimethoprim/sulfamethoxazole in animal production in the country (²⁵25. Puentes Martínez AR. Determinación de perfiles de resistencia antimicrobiana para Salmonella spp, Campylobacter spp., Listeria monocytogenes, Escherichia coli, Enterococcus spp en la cadena productiva avícola: abuelas, reproductoras y pollo de engorde. Ibagué: Universidad del Tolima; 2017.). In addition, some companies in Colombia still use non-therapeutic antimicrobials as prophylaxis in animal feed, and this can contribute to the emergence of AMR against these groups. Such AMR patterns, therefore, may limit the use of these antimicrobials in human and animal therapies (²⁶26. Gilchrist MJ, Greko C, Wallinga DB, Beran GW, Riley DG, Thorne PS. The potential role of concentrated animal feeding operations in infectious disease epidemics and antibiotic resistance. Environ Health Perspect. 2007;115(2):313–6.). The resistance against cefazolin and ciprofloxacin is also a matter of concern, as these antimicrobials are in the group of critically important and highly important antimicrobials for human health (²⁷27. World Health Organization. Critically Important Antimicrobials for Human Medicine, 6th revision. Geneva: WHO; 2019.). This has important implications for public health and may lead to calls for more regulated use of these antimicrobials in animals.

Compared with E. coli, lower levels of resistance were found in Salmonella spp. isolates. The highest levels of resistance in Salmonella spp. were against cefazolin and piperacillin/tazobactam. Additionally, intermediate levels of resistance were observed for ciprofloxacin, the fourth-generation cephalosporin cefepime, and the carbapenem drugs imipenem and meropenem. Excluding intermediate resistance levels, most Salmonella spp. isolates were susceptible to all the drugs tested. Similar susceptibility patterns have been reported from Kenya and the United States of America (¹⁰10. Ngai DG, Nyamache AK, Ombori O. Prevalence and antimicrobial resistance profiles of Salmonella species and Escherichia coli isolates from poultry feeds in Ruiru Sub-County, Kenya. BMC Res Notes. 2021;14(1):41., ²⁰20. Li X, Bethune LA, Jia Y, Lovell RA, Proescholdt TA, Benz SA, et al. Surveillance of Salmonella prevalence in animal feeds and characterization of the Salmonella isolates by serotyping and antimicrobial susceptibility. Foodborne Pathog Dis. 2012;9(8):692–8.) where Salmonella from animal feed exhibited low resistance levels to most of the antimicrobials tested. Cefepime, imipenem, and meropenem are critically important antimicrobials for human health, where cefepime is in the group of highest priority (²⁷27. World Health Organization. Critically Important Antimicrobials for Human Medicine, 6th revision. Geneva: WHO; 2019.).

The presence of some strains with reduced susceptibility to carbapenems warranted further investigation. After phenotypic screening testing, the presence of carbapenemases (β-lactamases with the ability to hydrolyze β-lactams including carbapenems) in six of seven isolates was excluded. This means that other mechanisms apart from enzymatic actions are contributing to carbapenem and cephalosporin resistance. For instance, the active expulsion of carbapenems from the periplasmic space (²⁸28. Schweizer HP. Efflux as a mechanism of resistance to antimicrobials in Pseudomonas aeruginosa and related bacteria: unanswered questions. Genet Mol Res. 2003;2(1):48–62.) and porin mutations may be associated with decreased antimicrobial susceptibility, as have been described for other Gram-negative bacilli (²⁹29. Meletis G, Exindari M, Vavatsi N, Sofianou D, Diza E. Mechanisms responsible for the emergence of carbapenem resistance in Pseudomonas aeruginosa. Hippokratia. 2012;16(4):303–7.). Our results indicate that uncommon AMR mechanisms may be circulating in Salmonella isolates from animal feed in Colombia. This provides evidence to support the design and implementation of better and more sophisticated AMR surveillance in animal feed, considering the impact of cross-transmission of strains between humans and animals as stated by the One Health approach.

The study had several strengths. An analysis was conducted on a large number of samples originating from the nationwide surveillance system covering most of the provinces of the country, collected over a four-year span. The results of the study serve as a baseline against which future reports may be compared. Baseline information is crucial to understanding the effectiveness of current policies on animal feed manufacturing and safety. In addition, given that the sample collection and the analysis were part of a national surveillance system, this process is ongoing and will allow an analysis of trends in the variables evaluated in the future. This information will enable close monitoring and allow suggestions for corrective action when needed.

The study had some limitations. First, the small number of positive samples obtained, despite the large sample size, precluded further statistical analysis of factors associated with the presence of the pathogens in animal feed samples. Second, due to resource constraints, only 27 (33%) and 26 (39%) strains of E. coli and Salmonella isolates, respectively, were tested for AMR, reducing the scope of the statistical analysis of the distribution of AMR patterns. Third, assessment of the performance of the surveillance system was outside the scope of this study because we had no access to the specific data system involved; but it would be useful to look into this in the future.

Conclusion

The presence of E. coli and Salmonella spp. in this study was lower than expected when compared with previous reports from other countries. Despite that, the presence of resistance to antimicrobials critically important for human and animal health deserves attention. Our results highlight the need to integrate antimicrobial susceptibility analyses into the current surveillance program for close monitoring of resistance patterns of isolates to antimicrobials. Future research on how to integrate the multiple sectors and disciplines using a One Health approach will help us to understand the origin, dissemination, and impact on public health of the microbiological and AMR contamination identified in this study.

Disclaimer.

Authors hold sole responsibility for the views expressed in the manuscript, which may not necessarily reflect the opinion or policy of the RPSP/PAJPH and/or those of the Pan American Health Organization.

Acknowledgment.

This research protocol was developed through the Structured Operational Research and Training Initiative (SORT IT), a global partnership coordinated by TDR—the UNICEF, United Nations Development Programme, World Bank, and World Health Organization (WHO) Special Program for Research and Training in Tropical Diseases, hosted at WHO. The specific SORT IT program that led to this study protocol included an implementation partnership of TDR and the Pan American Health Organization (PAHO), and the PAHO Colombia and Ecuador country offices; the Ministry of Health and Social Protection, Colombia; Food and Agriculture Organization, Sierra Leone; Sustainable Health Systems, Freetown, Sierra Leone; the Tuberculosis Research and Prevention Center non-governmental organization, Armenia; the International Union Against Tuberculosis and Lung Disease, Paris, France, and South East Asia office, India; Institute of Tropical Medicine, Antwerp, Belgium; Damien Foundation, Belgium; Indian Council of Medical Research–National Institute of Epidemiology; Jawaharlal Institute of Postgraduate Medical Education & Research (JIPMER); GMERS Medical College, Gotri Vadodara, India; Medical College Baroda, Vadodara, India; Sri Manakula Vinayagar Medical College, Puducherry, India; Public Health Ontario, Canada; Quadram Institute Bioscience, Norwich, United Kingdom; Universidade Federal de Ciências de Saúde de Porto Alegre, Brazil; Universidade de Brasília, Brazil; Universidad de Concepción, Chile; Universidad de los Andes, Colombia; Universidad Pontificia Bolivariana, Colombia; Universidad Pedagógica y Tecnológica de Colombia; Central University of Ecuador; California State University, Fullerton, United States of America; and the Autonomous University of Yucatán, Mexico.

REFERENCES

Publication Dates

History