Abstract
This study was aimed at assessing the antimicrobial resistance (AMR) of Campylobacter isolates from broilers and turkeys reared in industrial farms in Northern Italy, given the public health concern represented by resistant campylobacters in food-producing animals and the paucity of data about this topic in our country. Thirty-six Campylobacter jejuni and 24 Campylobacter coli isolated from broilers and 68 C. jejuni and 32 C. coli from turkeys were tested by disk diffusion for their susceptibility to apramycin, gentamicin, streptomycin, cephalothin, cefotaxime, ceftiofur, cefuroxime, ampicillin, amoxicillin+clavulanic acid, nalidixic acid, flumequine, enrofloxacin, ciprofloxacin, erythromycin, tilmicosin, tylosin, tiamulin, clindamycin, tetracycline, sulfamethoxazole+trimethoprim, chloramphenicol. Depending on the drug, breakpoints provided by Comité de l'antibiogramme de la Société Française de Microbiologie, Clinical and Laboratory Standards Institute, and the manufacturer were followed. All broiler strains and 92% turkey strains were multidrug resistant. Very high resistance rates were detected for quinolones, tetracycline, and sulfamethoxazole+trimethoprim, ranging from 65% to 100% in broilers and from 74% to 96% in turkeys. Prevalence of resistance was observed also against ampicillin (97% in broilers, 88% in turkeys) and at least three cephalosporins (93–100% in broilers, 100% in turkeys). Conversely, no isolates showed resistance to chloramphenicol and tiamulin. Susceptibility prevailed for amoxicillin+clavulanic acid and aminoglycosides in both poultry species, and for macrolides and clindamycin among turkey strains and among C. jejuni from broilers, whereas most C. coli strains from broilers (87.5%) were resistant. Other differences between C. jejuni and C. coli were observed markedly in broiler isolates, with the overall predominance of resistance in C. coli compared to C. jejuni. This study provides updates and novel data on the AMR of broiler and turkey campylobacters in Italy, revealing the occurrence of high resistance to several antimicrobials, especially key drugs for the treatment of human campylobacteriosis, representing a potential risk for public health.
Introduction
T
The majority of human Campylobacter infections are usually self-limiting and do not require specific treatment other than fluid and electrolytes replenishment. 41 However, in patients with severe, prolonged, or systemic infections, or belonging to high-risk groups, antimicrobial treatment may be recommended or at least prudent. 41 Currently, macrolides are the drugs of choice when a therapeutic intervention is required. Fluoroquinolones are also recommended as the first-line therapy since they are the drugs of choice for empirical treatment of undiagnosed diarrheal conditions, whereas tetracyclines are considered the second-line treatment.1,2,41 In serious cases of Campylobacter infection, that is, bacteraemia and other systemic infections, the intravenous administration of aminoglycosides may be considered the treatment of choice. 1
Antimicrobial resistance (AMR) in Campylobacter spp. is emerging worldwide and currently represents a serious public health concern, as recognized by the World Health Organization. 41 Indeed, an increasing number of Campylobacter strains resistant to several drugs are now being isolated from human samples, as well as from animals and foodstuffs. 42 AMR in the animal reservoir has serious implications for humans because animals are recognized vehicles through which exposure can occur. Moreover, the contamination of food-producing animals with antimicrobial-resistant strains may lead to the dissemination of AMR through the food chain. 15 The emergence of AMR in Campylobacter seems to be strictly correlated to the use of antimicrobials in veterinary medicine. 2 Particularly, the poultry reservoir played a leading role in the emergence of fluoroquinolone resistance in Campylobacter spp. 49
Very few investigations on the occurrence of AMR in Campylobacter isolated from broilers have been carried out in Italy,13,45,47 and the only study about turkeys 5 is limited to fluoroquinolone resistance. Therefore, the aim of this study was to determine the antimicrobial susceptibility profiles of C. jejuni and C. coli isolated from commercial broilers and meat turkeys reared in Northern Italy.
Materials and Methods
Bacterial strains
A collection of 160 C. jejuni and C. coli was analyzed for the antimicrobial susceptibility. Of these, 36 C. jejuni and 24 C. coli were isolated from live broilers (n=60 strains), and 68 C. jejuni and 32 C. coli from live meat turkeys (n=100 strains). Each strain was isolated from different birds commercially reared in four broiler and three turkey farms throughout Northern Italy between 2009 and 2010.20,21
Antimicrobial susceptibility determination
Susceptibility testing was carried out by the disk diffusion method using Mueller–Hinton agar (OXOID, Basingstoke, United Kingdom) supplemented with 5% defibrinated sheep blood (OXOID). Isolates were tested for their susceptibility to 20 antimicrobial drugs belonging to nine classes: aminoglycosides (apramycin [APR], 15 μg; gentamicin [GEN], 10 μg; streptomycin [STR], 10 μg), cephalosporins (cephalothin [CEPH], 30 μg; cefotaxime [CEFO], 30 μg; ceftiofur [CEFT], 30 μg; cefuroxime [CEFU], 30 μg), penicillines (ampicillin [AMP], 10 μg; amoxicillin+clavulanic acid [AMCL], 30 μg), quinolones (nalidixic acid [NAL], 30 μg; flumequine [FLU], 30 μg; enrofloxacin [ENR], 5 μg; ciprofloxacin [CIP], 5 μg), macrolides (erythromycin [ERY], 15 μg; tilmicosin [TLM], 15 μg; tylosin [TYL], 30 μg), lincosamides (clindamycin [CLI], 2 μg), tetracyclines (tetracycline [TET], 30 μg), potentiated sulphonamides (sulfamethoxazole+trimethoprim [SLT], 25 μg), phenicols (chloramphenicol [CAF], 30 μg). Turkey isolates were also tested for their susceptibility to pleuromutilins (tiamulin [TAM], 30 μg). All antimicrobial-impregnated disks were purchased from OXOID, except for TLM, TYL, and TAM, which were obtained from Bio-Rad (Marnes la Coquette, France), Mast Diagnostic Ltd. (Merseyside, United Kingdom), and Abtek Biologicals Ltd. (Liverpool, United Kingdom), respectively.
Inocula were prepared from overnight growth on tryptic soy agar (OXOID) supplemented with 5% defibrinated sheep blood (OXOID) by suspension in sterile saline (0.9% w/v of NaCl), to obtain a turbidity equivalent to that of a 0.5 McFarland standard. Bacterial suspension was streaked onto Mueller-Hinton agar (OXOID) supplemented with 5% defibrinated sheep blood (OXOID) and then incubated at 41.5°C for 24 hours under microaerobic conditions. Results concerning AMP, AMCL, CEPH, CEFO, NAL, CIP, ERY, GEN, TET, and CAF were evaluated in accordance with interpretive criteria provided by the Comité de l'antibiogramme de la Société Française de Microbiologie, 12 while zone diameter breakpoints for APR, FLU, TLM, TYL, and TAM were provided by the manufacturer. Susceptibility categorization for the other drugs tested was carried out using breakpoints established by the Clinical and Laboratory Standards Institute for enteric bacteria in the family Enterobacteriaceae,9,10 as reported in literature, 36 since no breakpoints specific for Campylobacter spp. are currently available for those drugs. C. jejuni ATCC 33560 was used as quality control microorganism in antimicrobial susceptibility determination. Multidrug resistance was defined as resistance to three or more antimicrobial classes and strains were considered resistant to an antimicrobial class if they were resistant to at least one antimicrobial drug within each class.
Statistical analysis
Statistical analysis was performed using the software PASW Statistics version 18.0 (SPSS, Inc., Chicago, IL). Chi-square (χ2) and Fisher's exact two-tailed tests were used to compare differences in the AMR between C. jejuni and C. coli isolates and between strains isolated from broilers and from turkeys. Differences were considered significant at values of p≤0.05.
Results
Antimicrobial resistance of broiler isolates
The number of isolates showing susceptibility, intermediate susceptibility, and resistance to each antimicrobial drug are presented in Table 1. All 60 broiler isolates were resistant to quinolones, both of the first and the second generation, with 100% of strains showing resistance to NAL, FLU, and CIP, and 70% to ENR. Although to a lesser extent, the predominance of resistance was also detected to SLT (72%) and TET (65%). Among β-lactams, resistance rates were very high for AMP (97%), CEPH, and CEFU (100%), and CEFT (93%). In contrast, 63.3% of isolates showed full susceptibility to AMCL and 51.7% to CEFO. Among aminoglycosides, all isolates were susceptible to APR, 97% to GEN, and 82% to STR. Moreover, all strains were susceptible to CAF. Regarding macrolides and lincosamides, results of susceptibility tests differed considerably between Campylobacter species. Indeed, whereas more than 80% of C. jejuni isolates were susceptible to CLI (83%) and macrolides (ERY, 83%; TYL, 86%; TLM, 89%), 87.5% of C. coli strains were resistant to the same drugs. These differences were statistically significant (p<0.0001). Differences between C. jejuni and C. coli were observed also in the susceptibility to STR (p<0.0001), TET (p=0.0002), AMCL (p=0.0005), and CEFO (p=0.013), with an overall predominance of resistance among C. coli isolates compared to C. jejuni strains.
S, susceptible; I, intermediate; R, resistant.
All Campylobacter isolates showed multidrug resistance: C. jejuni isolates were resistant to 3–7 antimicrobial classes, whereas C. coli strains to 5–8 classes. The distribution of multidrug resistance among isolates is reported in Table 2. The most common multidrug resistance patterns in C. jejuni included four classes of antimicrobials (i.e., cephalosporins, quinolones, penicillines, and potentiated sulfonamides; pattern CQPS) and five classes of antimicrobials (i.e., cephalosporins, quinolones, penicillines, potentiated sulfonamides, and tetracyclines; pattern CQPST). Each of these patterns was detected in 30.5% of isolates. Also in C. coli, two multidrug resistance patterns prevailed, each one shown by 25% of isolates. These patterns included seven classes of antimicrobials (i.e., cephalosporins, quinolones, lincosamides, macrolides, penicillines, potentiated sulfonamides, and tetracyclines; pattern CQLMPST) and eight classes of antimicrobials (i.e., aminoglycosides, cephalosporins, quinolones, lincosamides, macrolides, penicillines, potentiated sulfonamides, and tetracyclines; pattern ACQLMPST). For both species, all multidrug resistance patterns included cephalosporins and quinolones.
A, aminoglycosides; C, cephalosporins; L, lincosamides; M, macrolides; P, penicillines; Q, quinolones; S, potentiated sulfonamides; T, tetracyclines.
AMR of turkey isolates
Results of antimicrobial susceptibility testing are shown in Table 3. All 100 turkey isolates were susceptible to CAF, TAM, APR, STR, and 98% to GEN. A high occurrence of susceptibility was observed also to AMCL (77%), CLI (89%), and macrolides, with 90% of isolates susceptible to ERY, 89% to TYL, and 87% to TLM. Conversely, isolates showed high resistance to SLT (96%), AMP (88%), TET (80%), and both first- and second-generation quinolones (NAL and ENR, 74%; CIP, 87%; FLU, 88%). Resistance prevailed also among cephalosporins, with all strains resistant to CEPH, CEFT, and CEFU, and 87% to CEFO. C. jejuni and C. coli isolates showed similar AMR rates. The only statistically significant differences between the two Campylobacter species were observed for CLI (p=0.044), AMCL (p=0.003), and SLT (p=0.009). A high susceptibility rate to CLI was detected both in C. jejuni and in C. coli, but it prevailed among C. jejuni isolates (93% vs. 81% in C. coli). Similarly, the prevalence of susceptibility to AMCL was shown by both species, but markedly by C. jejuni (87%) compared to C. coli (56%). On the contrary, all C. jejuni isolates versus 88% of C. coli strains were resistant to SLT.
S, susceptible; I, intermediate; R, resistant.
Among 100 isolates tested for antimicrobial susceptibility, 92% showed resistance to multiple antimicrobial classes, ranging from 3 to 7 among C. coli strains and from 4 to 7 among C. jejuni isolates. The occurrence of multidrug resistance among isolates is shown in Table 4. The most common (72.8%) multidrug resistance pattern included five antimicrobial classes: cephalosporins, quinolones, penicillines, potentiated sulphonamides, and tetracyclines (pattern CQPST). This pattern prevailed among isolates belonging to both Campylobacter species.
A, aminoglycosides; C, cephalosporins; L, lincosamides; M, macrolides; P, penicillines; Q, quinolones; S, potentiated sulfonamides; T, tetracyclines.
Differences in resistance phenotypes between broiler and turkey strains
C. jejuni strains isolated from turkeys showed higher resistance rates to CEFO (p<0.001), CEFT (p=0.039), SLT (p<0.001), TET (p<0.001), and ENR (p<0.001) than broiler isolates. Conversely, broiler isolates were more resistant to NAL (p<0.001) and FLU (p=0.048) than turkey strains. No statistically significant differences (p>0.05) between broiler and turkey isolates were detected for the other antimicrobial drugs tested.
Regarding C. coli, statistically significant differences (p<0.001) were identified for CEFO, with higher resistance in turkey strains, and STR, CLI, and all macrolides, with broiler isolates being more resistant.
Discussion
This study was conducted to investigate on the AMR of C. jejuni and C. coli isolates from commercial broiler and turkey farms in Northern Italy. The AMR of isolates was assessed by the agar disk diffusion method, which has been proven to be suitable for screening purposes and particularly useful when testing several antimicrobials to obtain the qualitative data. 36 Antimicrobial drugs were selected to represent several classes, including those commonly used in both human and poultry therapies.
Most isolates were resistant to a variety of antimicrobial agents, and a high occurrence of multidrug resistance was observed. With the exception of penicillines, this finding may be linked to the extensive use of antimicrobial drugs in poultry production. This hypothesis is supported by the evidence that none of our isolates was resistant to chloramphenicol, which has not been used in poultry farming for over 15 years in the EU. Furthermore, the majority of isolates were susceptible to aminoglycosides, rarely administered to poultry. Susceptibility to chloramphenicol and aminoglycosides among Campylobacter strains is widely documented in the literature.19,26,27,36,37,46 Also the full susceptibility to tiamulin was expected since pleuromutilins are not frequently administered to poultry due to their lethal interactions with ionophore anticoccidials. 31
Among β-lactam antimicrobials, penicillines are widely used in the poultry industry, whereas cephalosporins are not currently approved for use in poultry in Italy. It has been reported that most C. coli and C. jejuni strains are intrinsically resistant to these antimicrobials due to the expression of β-lactamases.32,51 Indeed, most of our isolates were susceptible to the association of amoxicillin and clavulanic acid, whereas the resistance shown by a few strains may be due to a class D β-lactamase OXA-61, recently discovered in Campylobacter spp. 3 The high resistance rates to cephalosporins observed in this study may instead be due to low-affinity penicillin binding proteins or low permeability mechanisms.32,51 Although the majority of C. jejuni and C. coli strains have been reported to be resistant to cephalosporins, a moderate susceptibility to certain drugs within this antimicrobial class, including cefotaxime, has been observed in human as well as in chicken isolates.2,25,51 Consistently with this evidence, in our study, all or almost all strains were resistant to three cephalosporins, whereas a smaller amount of strains showed resistance to cefotaxime.
In the present study, very high resistance rates to fluoroquinolones were detected, as previously reported for broiler-derived campylobacters in many European countries and throughout most of the world, in particular where these antimicrobials are administered to poultry.7,18,46 In contrast to broilers, very few reports on the occurrence of AMR in commercial turkeys, especially at farm level, can be found in the literature.19,26,37,39,43 Comparing our results with those from other Italian studies,5,13,45,47 we detected higher resistance rates both in broiler and in turkey Campylobacter isolates. This high fluoroquinolone resistance might be associated to the administration of these antimicrobials (mainly enrofloxacin) to broilers and turkeys. It is well known that the use of fluoroquinolones rapidly leads to the emergence of resistant strains by selecting pre-existing spontaneous resistant mutants, which are able to persist on the farms for several productive cycles in the absence of antimicrobial selective pressure.35,38 Therefore, fluoroquinolone resistance of our isolates did not necessarily develop during the production cycle, but flocks could have been colonized by campylobacters already resistant to these antimicrobials.
Similar to fluoroquinolones, resistance rates to tetracyclines detected in broilers were higher than those observed in other Italian studies,13,45,47 whereas to our knowledge, no data on tetracycline resistance in campylobacters isolated from Italian turkeys have ever been published. Few articles about tetracycline resistance of campylobacters isolated from turkeys can also be found in the international literature, except for a survey carried out in North Carolina, 26 which detected the same resistance level observed in our study, and a German study 19 reporting a lower resistance rate. Although resistance to tetracyclines in Campylobacter spp. from broilers is variable among European countries, it is overall high in the EU, 18 as well as in other Countries worldwide,6,7,50 in accordance to our findings. The widespread use of tetracyclines for over 50 years as a growth promoter has increased the occurrence of resistance in microbial organisms. 8 However, tetracycline resistance we detected not necessarily was due to antimicrobial selective pressure. This consideration is supported by the detection of a high level of tetracycline resistance in campylobacters from organic poultry farms, where antimicrobials are not allowed. 37 Instead, monitored flocks should have been colonized by a Campylobacter population provided with genetic determinants of resistance maintained in the absence of antimicrobial selective pressure. Rapid transmission of the tet(O) gene, encoding resistance to tetracyclines, via natural horizontal transfer has been documented among C. jejuni strains in the intestinal environment of chickens. 4
Campylobacter resistance to sulphonamides and trimethoprim has generally received scarce attention. However, in the few published articles regarding these drugs, high levels of resistance have been reported in poultry isolates in different countries, similar to our findings.19,43 Also in previous Italian investigations, very high resistance rates to sulfamethoxazole and trimethoprim have been detected both in C. jejuni and in C. coli isolated from broilers.45,47 Resistance to trimethoprim is widely distributed among C. jejuni and C. coli, and for a long time, it has been considered intrinsic in these bacteria. 53 However, recent studies showed that resistance to trimethoprim may be transferred to Campylobacter spp. through horizontal gene transfer. 22 Indeed, trimethoprim resistance in C. jejuni has been linked to the dfr1 and dfr9 genes usually located on integrons or transposons. 22 Integrons can also account for resistance to sulphonamides, mediated by the sul1 gene located in the 3′-conserved segment of class 1 integrons. 40 Class 1 integrons has been detected in both C. jejuni and C. coli from different sources, including poultry,34,44 and the existence of chromosomally located integrons carrying a dfr1 containing gene cassette has been reported in clinical isolates of C. jejuni. 23 However, in a previous investigation, we found neither class 1 nor class 2 integrons in the same Campylobacter isolates analyzed in the present study. 48
Regarding resistance to macrolides of Campylobacter strains isolated from broilers, we detected considerably higher rates than those observed in other Italian studies45,47 and in other countries.11,18,27,28,46 Although the development of macrolide resistance in Campylobacter spp. is known to be a slow process, 35 it has been reported that the administration of both therapeutic and subtherapeutic doses of tylosin to broilers during their usual lifespan lead to the emergence of resistance to erythromycin. 33 Even if we do not have information about antimicrobial treatments in the farms and flocks under investigation, we hypothesize that macrolide resistance observed in our study could be linked to the administration of tylosin to birds. Indeed, this drug is often administered to broilers early in the production cycle to avoid intestinal bacterial overgrowth (i.e., dysbacteriosis), which occurs frequently in the first weeks of life. Conversely, in turkeys, the susceptibility of the great majority of isolates to macrolides was unexpected. Indeed, this finding is in contrast with results obtained in broilers and with the literature reporting that strains from turkeys are more likely to be resistant to erythromycin than strains from broilers. 37 However, the low resistance to macrolides among turkeys is consistent with results obtained by Lutgen et al. 39 and Gu et al. 26
Resistance rates to clindamycin reflected those observed for macrolides both in broilers and in turkeys, and this finding may be explained by cross-resistance between this antimicrobial class and lincosamides, although it cannot be excluded that resistance could also be linked to the administration of lincosamides to poultry. 52
In the present study, differences in the AMR between C. jejuni and C. coli were observed in broilers, mainly for macrolides. This finding is in accordance with other studies reporting a more frequent occurrence of resistance to this antimicrobial class in C. coli than in C. jejuni.7,24,33,35,39,46 All the other significant differences in AMR between the two Campylobacter species isolated from broilers consisted in the predominance of resistance in C. coli isolates. Consequently, multidrug resistance to a very high number of antimicrobial classes was detected more commonly in this species. These observations are consistent with the suspected higher tendency of C. coli to harbour resistance to multiple antimicrobials compared to C. jejuni. 14 In contrast to broiler isolates, C. jejuni and C. coli from turkeys showed similar AMR profiles for most drugs. This finding suggests that broilers and turkeys could be colonized by different Campylobacter populations possessing distinct AMR phenotypes, as detected for certain drugs by the statistical analysis we performed. This hypothesis is also supported by the fact that in a previous investigation20,21 on the same Campylobacter isolates, fla-typing showed a limited overlapping of genotypes between broilers and turkeys. However, in-depth genetic analysis would be helpful in understanding if the antimicrobial susceptibility of C. jejuni and C. coli could be associated to particular genotypes, which evolved primarily in broilers or in turkeys, as hypothesized by D'lima et al. 14 This could be due to the selection driven by the production practices and the antimicrobial treatments differing between broiler and turkey husbandry.
The present study provides updates on the AMR of broiler campylobacters in Italy, and it is the first article describing AMR in Campylobacter spp. isolated from live turkeys in our country. The high resistance rates detected to several antimicrobials and the high occurrence of multidrug resistance, with all broiler strains and 92% of turkey isolates resistant to at least three antimicrobial classes, should be taken into consideration. Particularly, resistance to fluoroquinolones and macrolides is alarming since they are the drugs of choice for the treatment of campylobacteriosis in human medicine. Also the high resistance to tetracycline should be highlighted since it represents a second-line therapeutic agent in human campylobacteriosis therapy. Our findings are of great concern considering that poultry are the major source of human Campylobacter spp. infection and that antimicrobial-resistant strains can be easily transmitted to humans via the food chain, potentially increasing the burden of campylobacteriosis. Therefore, the assessment and monitoring of AMR of Campylobacter spp. in their main reservoir species is needed to pursue the goal of protecting public health.
Footnotes
Acknowledgments
The present study was supported by the Cariverona Foundation and the University of Padua (Grant Ref. No. CPDA095771/09).
Disclosure Statement
The authors have none to declare about competing financial interests.
