Introduction
Campylobacter is the most common enteric bacterial pathogen in humans in the United States of America (USA) causing an estimated 20 cases for every 100,000 persons each year [1] and is the leading cause of foodborne bacterial infections in the USA [Reference Sher2] and worldwide [Reference Silva3]. The majority of infections (estimated 90%) are caused by Campylobacter jejuni (C. jejuni), and only 5–10% are attributed to Campylobacter coli (C. coli) [1]. In addition to enteric diseases, C. jejuni has been linked to several post-infection complications such as irritable bowel syndrome, Guillain–Barré syndrome, and reactive arthritis [Reference Whitehouse4]. Previous studies described the consumption of contaminated poultry products as a main source of campylobacteriosis [Reference Ford5, Reference Ricke6].
Campylobacter has become resistant to clinically important antimicrobials in human medicine and is therefore listed as a high-priority antimicrobial-resistant pathogen [Reference Tang7]. Recent studies from the USA [Reference Ford5], Switzerland [Reference Ghielmetti8], European Union (EU) [9], and South America [Reference Portes10] reported a high level of resistance in human Campylobacter isolates to tetracycline and ciprofloxacin. Macrolide resistance was reported worldwide; however, currently only a low level of resistance exists, which is an encouraging finding as macrolides (e.g. erythromycin) are the first drug of choice when treating campylobacteriosis.
Chickens and turkeys are important sources of antimicrobial-resistant Campylobacter [Reference Shrestha11, Reference Tedersoo12] as fluoroquinolone-and tetracycline-resistant Campylobacter isolates have been identified previously at chicken and turkey farms, slaughter plants, and retail stores in North America [Reference Shrestha11, Reference Agunos13–Reference Varga15] and worldwide [9, Reference Abraham16].
Antimicrobials have been used effectively for decades to treat, control, and prevent bacterial infectious diseases on poultry farms in North America [Reference Shrestha11, Reference Agunos13–Reference Varga15]; however, the use of antimicrobials has the highest impact on the emergence of antimicrobial resistance (AMR) [Reference Shrestha11, Reference Agunos14]. The US poultry sectors implemented antimicrobial use (AMU) reduction strategies and gradually eliminated the preventive use of medically important antimicrobials to contain the emergence of AMR. There are national [17], and global [18] initiatives to reduce the emergence and dissemination of antimicrobial-resistant commensal and pathogenic bacteria at the human–animal–environment interface. To limit the emergence of AMR in the USA, AMU as growth promoters in food-producing animals was prohibited, and the use of all clinically important antimicrobials in feed and water without the supervision of a licensed veterinarian was banned [19]. Similar bans have also been in effect in other regions, including Canada [Reference Shrestha11] and Denmark [20].
Evaluating AMR monitoring programmes to assess the current AMR patterns in foodborne pathogens to detect emerging AMR patterns and trends and assess the effectiveness of AMU policy changes over time is fundamental.
Considering all the issues presented above, publicly available data from the National Antimicrobial Resistance Monitoring System of Enteric Bacteria (NARMS) were evaluated to compare the prevalence of AMR in C. jejuni and C. coli isolated from the caecal content of chickens and turkeys at slaughter plants across the USA between 2013 and 2019. The provided information could assist public health and animal health authorities in developing effective antimicrobial stewardship programmes.
Methods
Study design
This study analysed publicly available AMR monitoring data collected by the NARMS, comprising data on AMR in Campylobacter spp. isolated from caecal samples obtained from chickens and turkeys at the United States Department of Agriculture (USDA) and Food Safety Inspection Service (FSIS)-regulated poultry slaughter establishments across the USA from 2013 to 2021.
Slaughter establishments were selected randomly by staff at FSIS considering their slaughter volume, type, and location. Once the number of samples per plant was established, caecal products were collected by pooling five samples of turkeys and chickens each into one sample [21].
Laboratory testing
At the USDA FSIS Eastern Laboratory, standard microbiological methods were used to isolate Campylobacter strains from chicken and turkey samples [22]. Briefly, samples were enriched with buffered peptone water (BPW) and incubated at 42 ± 1.0°C for 29–31 h in a sealed, microaerobic environment. Subsequently, 30 μl from each well or test tube was streaked onto modified charcoal–cefoperazone–deoxycholate agar (mCCDA) plates and incubated at 42 ± 1.0°C for 22–24 h. Typical colonies from mCCDA were then streaked onto trypticase soy agar with 5% sheep blood agar (SBA) plates and incubated at 42 ± 1°C for 24–48 h. Confirmation of Campylobacter was done by examining the SBA plates and re-streaking if necessary for purity. The Bruker Matrix-Assisted Laser Desorption Ionization (MALDI) Biotyper was utilized to confirm the selected colonies. The latter were further tested for their antimicrobial susceptibility and speciated using whole-genome sequencing.
Antimicrobial susceptibility was determined using the Sensititre broth microdilution method using the CMV Campylobacter Selective Agar (CAMPY) plates. The following antimicrobials were tested: gentamicin, clindamycin, azithromycin, erythromycin, ciprofloxacin, nalidixic acid, and tetracycline. The interpretive guidelines for susceptibility testing and the categorization of resistant isolates were based on the minimum inhibitory concentration (MIC) values and breakpoints determined by the Clinical and Laboratory Standards Institute (CLSI) for C. coli and C. jejuni [23]. Supplementary Table 1 lists the breakpoints for both C. coli and C. jejuni. The AMR rate was categorized as rare (<0.1% of isolates), very low (0.1% to 1.0% of isolates), low (1.01%–10.0% of isolates), moderate (10.01%–20.0% of isolates), high (20.01%–50.0% of isolates), and very high (>50.0% of isolates) [24].
Statistical analyses
STATA Intercooled software (Version 18, Stata Corporation, College Station, TX) and R software (Version 4.1.2 (2021-2111-01)) (R Core Team, 2020), within the RStudio platform (Version 1.4.1106 © 2009–2021 RStudio, PBC), were used for statistical analysis. The proportion of resistance to each antimicrobial was calculated by dividing the number of resistant isolates by the total number of isolates tested. For each proportion, the exact binomial 95% confidence intervals (CI) was calculated using the Clopper–Pearson methodology.
Antimicrobial resistance pattern analysis
To analyse co-resistance and multidrug resistance patterns among antimicrobials and the clustering of resistant isolates, single-linkage dendrograms were created using Ward’s hierarchical clustering method, with Euclidean distances. Dendrograms were visualized in heatmaps using the heatmap.2 package in R software and the ggplot and RColorBrewer libraries.
To illustrate the pairwise and total correlations between AMRs, chord diagrams were created by using the chorddiag and devtools R-packages.
Evaluating differences among campylobacter species, poultry type, and years
To determine differences in AMR between poultry species (chicken versus turkey), Campylobacter species (C. coli versus C. jejuni), and years (2013–2021), a multivariable logistic regression model for each antimicrobial was constructed. The dependent binary variable represented the resistance status (resistant = 1/susceptible = 0) of an antimicrobial, while the independent variables included poultry species (comparing turkeys to chickens), Campylobacter species (comparing C. jejuni to C. coli), and the study period (using the year 2013 to which all other years were compared). Statistically significant associations were signified by a p-value of ≤0.05 on the Wald χ2 test. For all model outcomes, odds ratios (ORs), 95% CIs, and p-values were presented. An OR less than 1 indicated a protective effect, while a value >1 signified that the variable had a positive effect on the dependent variable. For each model outcome, predicted probabilities were calculated and displayed graphically.
Results
Prevalence of antimicrobial resistance in C. coli and C. jejuni isolates of chickens and turkeys
A total of 1,899 Campylobacter isolates (1,031 C. coli and 868 C. jejuni) of chickens and 798 Campylobacter isolates (673 C. coli and 123 C. jejuni) of turkeys detected between 2013 and 2021 were included in this study.
In both C. jejuni and C. coli isolates derived from chicken caecal samples, there was a high prevalence of resistance to tetracycline (42–45%), moderate resistance to ciprofloxacin and nalidixic acid (17–22%), and low resistance to clindamycin and azithromycin (1–6%). On the other hand, in C. coli isolates, low resistance rates were observed for erythromycin and gentamicin (5–6%), while C. jejuni isolates showed very low resistance (0.2–1.0%) to the same antimicrobials (Table 1).
Table 1. Prevalence of antimicrobial resistance in Campylobacter coli (n = 1704) and Campylobacter jejuni (n = 991) isolates recovered from caecal samples of chickens and turkeys at the slaughterhouse level across the United States, 2013–2019
a AZI, azithromycin; CIP, ciprofloxacin; CLI, clindamycin; ERY, erythromycin; GEN, gentamicin; NAL, nalidixic acid; TET, tetracycline.
b Number and percentage of isolates resistant to the antimicrobial.
c CI, exact binomial 95% confidence interval for the percentage of isolates resistant to the antimicrobial.
In both C. jejuni and C. coli strains isolated from turkey caecal samples, a very high resistance rate to tetracycline (63–69% of isolates) and a high resistance rate to ciprofloxacin and nalidixic acid (24–39%) were observed. Conversely, in C. coli isolates from turkeys, moderate resistance rates were observed for gentamicin, clindamycin, azithromycin, and erythromycin (12–19%), while C. jejuni isolates exhibited a low resistance rate (2–8%) to these antimicrobials (Table 1).
Evaluating antimicrobial resistance patterns and clustering in C. coli and C. jejuni isolates of chickens
Hierarchical clustering dendrograms were constructed to evaluate the co-resistance patterns of the examined antimicrobials (columns) within bacterial isolates (rows).
The cluster analysis of AMR in C. coli and C. jejuni isolates from chickens is represented in Figure 1a,b, respectively.
Figure 1. Clustering dendrogram (heatmap) of antimicrobial resistance in (a) Campylobacter coli and (b) Campylobacter jejuni isolated from chicken caecal samples at the slaughterhouse level across the United States, 2013–2021. AZI, azithromycin; CIP, ciprofloxacin; CLI, clindamycin; ERY, erythromycin; GEN, gentamicin; NAL, nalidixic acid; TET, tetracycline. Red colour: resistant; blue colour: susceptible.
In the columns of both heatmaps (Figure 1a,b), a primary cluster (marked with a plus sign (+)) was identified that signified resistance to tetracycline. The second cluster (++) indicated co-resistance to ciprofloxacin and nalidixic acid. The third cluster (+++) included co-resistance to gentamicin, erythromycin, azithromycin, and clindamycin. The rows of both heatmaps showed a cluster of isolates that were susceptible to all tested antimicrobials (*), another cluster of isolates that were susceptible to all antimicrobials except tetracycline (**), and a group (***) displaying resistance to most of the tested antimicrobials.
Correlations among antimicrobial resistance
The pairwise and total correlations among resistance to the examined antimicrobials in the Campylobacter isolates from chickens are illustrated in Figure 2. C. coli isolates from chickens (Figure 2a) exhibited strong total correlations for azithromycin (3.39) and erythromycin (3.38). The total correlations of other antimicrobials included clindamycin (3.31), ciprofloxacin (2.56), nalidixic acid (2.56), gentamicin (1.73), and tetracycline (1.63).
Figure 2. A chord diagram illustrating the pairwise and total correlations among antimicrobial resistance patterns in (a) Campylobacter coli and (b) Campylobacter jejuni isolated from chickens. AZI, azithromycin; CIP, ciprofloxacin; CLI, clindamycin; ERY, erythromycin; GEN, gentamicin; NAL, nalidixic acid; TET, tetracycline. Each antimicrobial agent is represented as a coloured segment, with the segment’s width reflecting the overall correlation of that antimicrobial agent. The network illustrates the connections and correlation strengths between different antimicrobial agents based on their resistance patterns. The thickness of the colour-coded chords in the diagram signifies the strength of the correlation between the resistance patterns of the respective antimicrobial agents.
In C. coli isolates from chickens, strong positive pairwise correlations were found between ciprofloxacin and nalidixic acid (ρ = 0.99), azithromycin and erythromycin (ρ = 0.99), azithromycin and clindamycin (ρ = 0.89), and erythromycin and clindamycin (ρ = 0.89).
Among C. jejuni from chickens (Figure 2b), the highest total correlations were found for erythromycin (2.69) and azithromycin (2.65), similar to the C. coli isolates. The total correlations for other antimicrobials were clindamycin (2.44), nalidixic acid (2.36), ciprofloxacin (2.35), gentamicin (1.84), and tetracycline (1.71) (Figure 2b).
Strong positive pairwise correlations between ciprofloxacin and nalidixic acid (ρ = 0.99), azithromycin and erythromycin (ρ = 0.75), erythromycin and clindamycin (ρ = 0.61), and azithromycin and clindamycin (ρ = 0.59) were identified in C. jejuni isolates from chickens.
Evaluating antimicrobial resistance patterns and clustering in C. coli and C. jejuni isolates of turkeys
The cluster analysis of AMR in C. coli and C. jejuni from turkeys is illustrated in Figure 2a,b, respectively.
Distinct clusters were observed in the columns of both heatmaps (Figure 2a,b). The first cluster, identified by the symbol (+), signified resistance only to tetracycline and the second cluster (++) indicated resistance to ciprofloxacin and nalidixic acid, while the third cluster (+++) included isolates that showed resistance to erythromycin, azithromycin, clindamycin, and gentamicin.
Both Campylobacter species isolates illustrated in the rows of the heatmaps (Figure 3a,b) exhibited identical clustering patterns. The primary cluster (*) contained isolates that showed susceptibility to all tested antimicrobials, and the second cluster (**) contained isolates susceptible to all antimicrobials tested except for tetracycline. The third cluster (***) comprised isolates resistant to ciprofloxacin, nalidixic acid, and tetracycline, while the fourth cluster (****) comprised isolates resistant to all tested antimicrobials, thereby signifying a multidrug-resistant (MDR) group.
Figure 3. Clustering dendrogram (heatmap) of antimicrobial resistance in (a) Campylobacter coli and (b) Campylobacter jejuni isolated from turkey caecal samples at the slaughterhouse level across the United States, 2013–2021. AZI, azithromycin; CIP, ciprofloxacin; CLI, clindamycin; ERY, erythromycin; GEN, gentamicin; NAL, nalidixic acid; TET, tetracycline. Red colour: resistant; blue colour: susceptible.
Correlations among antimicrobial resistance
The pairwise and total correlations among resistance to the examined antimicrobials in the Campylobacter isolates from turkeys are illustrated in Figure 4.
Figure 4. A chord diagram illustrating the pairwise and total correlations among antimicrobial resistance patterns in (a) Campylobacter coli and (b) Campylobacter jejuni isolated from turkeys. AZI, azithromycin; CIP, ciprofloxacin; CLI, clindamycin; ERY, erythromycin; GEN, gentamicin; NAL, nalidixic acid; TET, tetracycline. Each antimicrobial agent is depicted as a coloured segment, with the segment’s width reflecting the overall correlation of that antimicrobial agent. The network illustrates the connections and correlation strengths between different antimicrobial agents based on their resistance patterns. The thickness of the colour-coded chords signifies the strength of the correlation between the resistance patterns of the respective antimicrobial agents. The blurring effect observed in (b) is attributed to the negative correlation between erythromycin and tetracycline.
In C. coli isolates (Figure 4a), the highest total correlations were detected for azithromycin (4.09) and erythromycin (4.09). Other antimicrobials showed the following total correlations: clindamycin (3.96), ciprofloxacin (3.51), nalidixic acid (3.51), gentamicin (3.24), and tetracycline (2.51) (Figure 4a). Strong pairwise correlations were detected between ciprofloxacin and nalidixic acid (ρ = 1), azithromycin and erythromycin (ρ = 0.99), azithromycin and clindamycin (ρ = 0.85), and erythromycin and clindamycin (ρ = 0.86).
In C. jejuni isolates (Figure 4b), ciprofloxacin (2.93) and nalidixic acid (2.93) showed the highest total correlations. Other antimicrobials had the following total correlations: azithromycin (2.71), erythromycin (2.67), gentamicin (2.48), clindamycin (2.28), and tetracycline (1.78) (Figure 4b). All pairwise correlations were positive, except erythromycin and tetracycline (ρ = −0.04). Strong positive pairwise correlations were identified between ciprofloxacin and nalidixic acid (ρ = 1), and azithromycin and erythromycin (ρ = 0.81).
Evaluating differences among antimicrobial resistance in Campylobacter species, poultry types, and years
The probability of resistance to all tested antimicrobials (except for tetracycline) was significantly reduced during the study period when compared to 2013 (Table 2). The predicted probabilities of AMR across the study period considering poultry species and Campylobacter species are illustrated in Figure 3. Prediction for nalidixic acid was not illustrated as it was identical to ciprofloxacin.
Table 2. Multivariable logistic regression models representing the probability of resistance to antimicrobials among two Campylobacter species (n = 2,695) isolated from chicken and turkey samples at the slaughterhouse level across the United States, 2013 to 2021
a AZI, azithromycin; CIP, ciprofloxacin; CLI, clindamycin; ERY, erythromycin; GEN, gentamicin; NAL, nalidixic acid.
b CI, exact binomial 95% confidence interval.
c Statistically significant at P ≤ 0.05.
d Turkeys versus chicken.
e C. jejuni versus C. coli.
Turkeys had a higher probability of resistance to all examined antimicrobials compared to chickens (Table 2 and Figure 5).
Figure 5. Predicted probabilities of antimicrobial resistance across the study period (2013–2021) considering poultry species (turkeys vs. chickens) and Campylobacter species (Campylobacter jejuni vs. Campylobacter coli). AZI, azithromycin; CIP, ciprofloxacin; CLI, clindamycin; ERY, erythromycin; GEN, gentamicin.
Likewise, the odds of resistance to ciprofloxacin and nalidixic acid were higher in C. jejuni than in C. coli, but C. jejuni isolates showed significantly lower odds of being resistant to all the other antimicrobials when compared to C. coli isolates (Table 2 and Figure 5).
Discussion
This study evaluated publicly available longitudinal surveillance data collected by the NARMS programme on AMR in C. coli and C. jejuni isolates of chickens and turkeys sampled at the slaughterhouse level across the USA over 9 years. We provided evidence-based data on the prevalence, patterns, and differences in AMR between C. jejuni and C. coli isolates and between turkey- and chicken-origin isolates.
A lower probability of resistance to all antimicrobials (except for tetracyclines), particularly macrolides (azithromycin and erythromycin) and quinolones (ciprofloxacin and nalidixic acid), was observed among Campylobacter isolates in the later years of the study (2019, 2020, and 2021). This finding agrees with two recent studies from the USA that showed a decrease in the probability of resistance to antimicrobials in Campylobacter isolated from chicken and turkey samples at retail [Reference Sarkar and Okafor25] and slaughter [Reference Chandra Deb, Jara and Lanzas26]. Additional studies from the USA [Reference Awosile27] and Canada [Reference Shrestha11] described a reduction of AMR in foodborne pathogens in chickens and turkeys. These are encouraging findings and might be related to the policy changes to AMU in poultry and livestock sectors, such as the veterinary oversight of the use of medically important antimicrobials in feed and water and the ban on using antimicrobials as growth promoters [28, 29].
The study results indicated that in both chickens and turkeys, C. coli isolates had higher AMR rates than C. jejuni isolates for most of the antimicrobials examined. This finding agrees with previous studies that described a higher proportion of MDR C. coli isolates from poultry [Reference Shrestha11, Reference Varga15]. In contrast, it was described that C. jejuni isolates have a lower rate of AMR, but they better survive the food processing environment that aids them to enter the food chain and infect humans [Reference Corcionivoschi and Gundogdu30, Reference Hull31].
A higher prevalence of resistance to all examined antimicrobials was observed in the Campylobacter spp. isolates derived from turkeys when compared to isolates derived from chickens. The longer production cycle of turkeys compared to chickens might increase the probability of contracting infections that necessitate antimicrobial treatment. This elevated exposure to antimicrobials might contribute to the higher AMR rates observed in turkey flocks. This finding agrees with a recent study from the USA, which described a higher level of AMR in Salmonella serovars of turkeys compared to isolates obtained from other food animals [Reference Awosile27]. Future investigations are needed to identify factors and underlying reasons behind this issue.
The highest prevalence of resistance in both C. coli and C. jejuni isolates derived from chickens and turkeys was identified against tetracycline. This finding agrees with previous studies from North America and worldwide [9, Reference Shrestha11, Reference Agunos14, Reference Varga15, Reference Rahimi and Ameri32]. Tetracyclines are commonly used to treat and prevent bacterial poultry diseases [Reference Agunos, Carson and Léger33], and the selection pressure of AMU is a major factor in the selection of tetracycline resistance. In 2018, tetracycline comprised 66% of the total antibiotics sold for administration to livestock and poultry in the USA [17, Reference Hull31]. Also, genetic factors might contribute to the persistence and selection of tetracycline resistance as previous studies have indicated that mobile genetic elements carrying resistance genes can be transferred among various Campylobacter strains [Reference Hull31]. Earlier research [Reference Guernier-Cambert34] identified the plasmid-encoded gene tet(O) as the key determinant of tetracycline resistance in Campylobacter. A recent US study further reported that 13.2% of Campylobacter isolates from food animals carried the tetO gene [Reference Hull35]. Additionally, it has been documented that this gene can undergo horizontal transfer between C. jejuni and C. coli within the gastrointestinal tracts of food animals [Reference Elhadidy36]. Apart from tetracycline use and genetic determinants, other factors contributing to the selection of tetracycline resistance should be further investigated.
Fluoroquinolones (e.g. ciprofloxacin) are the preferred empirical treatment choices for campylobacteriosis in humans [Reference Luangtongkum37], and the increase in fluoroquinolone resistance poses a public health risk. Similar to the findings of this study, high resistance to fluoroquinolones in Campylobacter was also previously reported among isolates in poultry, humans, and environment worldwide [9, Reference Varga15]. The presence of fluoroquinolone-resistant Campylobacter is concerning, and to limit its emergence, the World Health Organization (WHO) included fluoroquinolone-resistant Campylobacter as a high-priority pathogen that requires increased research and development focus to advance new and effective antibiotic treatments [38]. In the USA, since 2005, fluoroquinolones have not been used in water to treat poultry bacterial infections [Reference Zawack39], and the use of fluoroquinolones on turkey [Reference Singer40] and chicken [Reference Singer41] farms is limited, which points to the impact of non-AMU factors that influence the selection of fluoroquinolone resistance. Previous research studies have documented the presence of fluoroquinolone-resistant Campylobacter isolates in poultry in the absence of fluoroquinolone use [Reference Shrestha11, Reference Abraham16]. Moreover, a recent Australian study suggested that the infection source of fluoroquinolone-resistant Campylobacter in poultry might be attributable to humans, wild birds, or pests [Reference Abraham16]. Cattle might also be a source for drug-resistant Campylobacter isolates in poultry as a recent Canadian study using molecular epidemiological methods showed genetic relatedness among cattle, poultry, and human C. jejuni isolates [Reference Teixeira42].
Biosecurity and farm management factors might also impact the prevalence of Campylobacter isolates in poultry flocks [Reference Schweitzer43]. A recent Canadian investigation revealed that in chicken flocks the use of virginiamycin as a feed additive, using traps to control rodents, and the number of birds in a barn increased the prevalence of fluoroquinolone-resistant C. jejuni [Reference Caffrey44].
Macrolides (e.g. azithromycin and erythromycin) are the primary treatment choice for human campylobacteriosis [Reference Aidara-Kane45]. Our results revealed a low prevalence of macrolide-resistant Campylobacter, apart from moderate resistance of C. coli isolates from turkeys, which is consistent with past observations of higher macrolide resistance in C. coli compared to C. jejuni in turkey flocks [Reference Shrestha11]. Macrolide resistance in both C. jejuni and C. coli is associated with point mutations in the 23S ribosomal RNA (rRNA) gene [Reference Zhao46] and can also be conferred by the erm(B) gene [Reference Qin47]. Notably, in chickens, substitutions in the 23S rRNA gene (specifically A2075G or A2074C/G) have been linked to reduced colonization of flocks with C. jejuni [Reference Wieczorek and Osek48]. Furthermore, it has been reported that the erm(B) gene is more frequently detected in C. coli compared to C. jejuni [Reference Liu49]. Further molecular-level investigations are needed to validate our findings.
Campylobacter isolates from both turkey and chicken caecal samples displayed low levels of resistance to gentamicin. Previous US investigations reported that aph(2″)-Ig and aph(2″)-If variants are the most predominant AMR genetic determinants associated with gentamicin resistance among Campylobacter [Reference Whitehouse4, Reference Zhao46]. Historically, gentamicin was prescribed for the prevention of necrotic enteritis. However, the use of gentamicin in hatcheries (in-ovo) in the USA decreased between 2013 and 2019, with no reported usage after 2019 [Reference Singer50]. Our findings align with the mentioned intervention, showing a reduction in gentamicin resistance among Campylobacter isolates during the study period.
The cluster analysis of chicken isolates revealed similar AMR clusters in both Campylobacter species. However, C. coli exhibited a higher prevalence of tetracycline resistance, while C. jejuni showed greater resistance to ciprofloxacin and nalidixic acid. Given that C. jejuni is responsible for 80–90% of human campylobacteriosis cases [Reference Corcionivoschi and Gundogdu30, Reference Hull31] and exhibits higher resistance to fluoroquinolones, it is crucial to intensify monitoring and investigations into the emergence of fluoroquinolone resistance, particularly in chickens, a main protein source for humans [51].
The cluster analysis of turkey isolates also indicated identical AMR and MDR clusters in C. coli and C. jejuni. The main distinction was that C. coli isolates displayed higher resistance to all tested antimicrobials, which could be explained by the inherent characteristics of this species, which is known to exhibit greater resistance to multiple antibiotics [Reference Corcionivoschi and Gundogdu30, Reference Hull31].
Here, higher resistance to antimicrobials in C. coli isolates was observed in turkeys compared to chickens, suggesting that the former might play a larger role in the emergence of multidrug resistance in Campylobacter isolates. This finding is supported by a previous study, which suggested that antimicrobial-resistant C. coli might have the potential for better adaptation to the turkey farm environment and a higher tendency to colonize turkeys compared to C. jejuni [Reference Zhang52].
The strongest pairwise correlation coefficients found here were observed between macrolides and quinolone classes in both C. coli and C. jejuni isolates from both poultry species. This may be explained by the cross-resistance within the same class of antimicrobials, facilitated by mobile genetic elements harbouring multiple resistance genes.
In addition to the NARMS programme in the USA, other countries have also integrated AMR surveillance systems to monitor indicator, foodborne, and pathogenic bacteria from poultry. These include the Danish Integrated Antimicrobial Resistance Monitoring and Research Programme (DANMAP), the European Antimicrobial Resistance Surveillance Network in Veterinary Medicine (EARS-Vet), and the Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS) [9, Reference Agunos13, 20, Reference Mader53–56]. These programmes serve to identify emerging AMR trends and evaluate the effectiveness of antimicrobial stewardship strategies.
The present study is not without limitations. The absence of detailed information regarding the exposure history of turkeys and chickens sampled at the slaughter plants, including their on-farm AMU history and biosecurity and farm management factors, limited the ability to make direct links between the selection of AMR in Campylobacter isolates and the impact of various exposure factors.
In conclusion, we have shown a higher prevalence of resistance to most of the examined antimicrobials in C. coli isolates compared to C. jejuni, in both poultry species. Additionally, higher resistance rates were observed in C. coli and C. jejuni isolates obtained from turkeys compared to chickens. Over the study period, there was an overall decrease in the prevalence of resistance to the tested antimicrobials in Campylobacter isolates, particularly in the later years. Molecular epidemiological and on-farm studies are needed to acquire insights and promote understanding of the factors associated with the selection and persistence of antimicrobial-resistant Campylobacter isolates in the poultry production system.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0950268824000359.
Data availability statement
The data are publicly available, and we shared the link to the data set in the reference section.
Acknowledgements
We would like to thank the National Antimicrobial Resistance Monitoring System for Enteric Bacteria (NARMS) for its role in data generation and public accessibility. We would also like to thank the veterinarians and slaughterhouse staff who actively took part in this study for their valuable contributions.
Author contribution
H.R.S. and C.V. conceptualized the study; C.V., I.A., and S.Y. involved in formal analysis; C.V. acquired funding; H.R.S., C.V., I.A., and S.Y. designed the methodology; C.V. administered the project; C.V. supervised the data; C.V., I.A., and S.Y. visualized the data; H.R.S. and M.N.S. wrote the original draft; and H.R.S., C.V, M.N.S., I.A., and S.Y wrote, reviewed, and edited the manuscript.
Funding statement
The study did not receive funding.
Competing interest
The authors declare none.
Open access
