Antimicrobial Resistance Profiles of Campylobacter from Humans,Retail Chicken Meat,and Cattle Feces

Abstract

We determined the antimicrobial resistance profiles of Campylobacter isolates from cases of sporadic human infection (n = 119), retail chicken meat (n = 105), and cattle feces (n = 105). Ampicillin and tetracycline resistance was highest in human isolates (32% and 29%, respectively) and retail chicken isolates (25% and 25%, respectively), whereas nalidixic acid resistance was highest in cattle fecal isolates (20%). We found that the antimicrobial resistance profiles were more similar in human and chicken meat isolates than those observed when comparing human and cattle fecal isolates. When we analyzed the distribution of minimum inhibitory concentrations for each antibiotic, in each host, the distribution was similar between human and chicken meat isolates, whereas cattle fecal isolates remained highly distinct from the other two hosts. This study suggests that chicken may be a major source of human Campylobacter infection and that the antimicrobial resistances found in the Campylobacter from this source will therefore also be prevalent in clinical isolates.

Introduction

C ase–control studies and attribution studies using molecular typing have identified poultry meat as a major risk factor for Campylobacter infection (Studahl and Andersson, 2000; Sheppard et al., 2009; Strachan et al., 2009); therefore, it is likely that antimicrobial-resistant isolates from poultry meat could enter the human food chain. Although beef is much less frequently implicated in human infection, there is a recognized association with cattle feces (Stanley and Jones, 2003), particularly in the northeast of Scotland (Grampian) (Strachan et al., 2009), where a high proportion of the land is rural; however, the excretion of resistant strains into the environment presents an as-yet-unquantified risk to humans.

Many classes of antibiotics are used therapeutically in poultry and cattle (Veterinary Medicines Directorate, 2007); therefore, the potential for transfer of antimicrobial-resistant Campylobacter isolates between these sources and humans is high.

The emergence of antimicrobial-resistant Campylobacter has complicated treatment of human campylobacteriosis. Resistance is attributed to the clinical use of antimicrobials and their use in veterinary medicine. Since the 1990s, the prevalence of tetracycline-, fluoroquinolone-, and macrolide-resistant strains from humans has dramatically increased in developed and developing countries around the world, and it is recognized that infection with antimicrobial-resistant campylobacters can lead to prolonged gastrointestinal symptoms (Nelson et al., 2004). We report on the antimicrobial resistance of Campylobacter isolates from the aforementioned sources and further comment on the relative importance of these two sources of human infection based upon the resistances observed.

Materials and Methods

Antimicrobial susceptibility testing

Campylobacter jejuni and Campylobacter coli isolates from patients infected in Grampian, Scotland (n = 119); U.K.-produced retail chicken meat (n = 105); and cattle fecal isolates from Grampian (n = 105) in 2006 from our previous studies (Gormley et al., 2008; Sheppard et al., 2009) were assessed for their antimicrobial susceptibilities.

Minimum inhibitory concentrations (MICs) to ampicillin, ciprofloxacin, erythromycin, gentamicin, kanamycin, nalidixic acid, and tetracycline were determined using the agar dilution technique, according to Clinical and Laboratory Standards Institute methods (NCCLS, 2002). Antibiotics were tested in doubling dilutions ranging from 0.125 to 128 μg/mL, and an inoculum of ∼10⁴ CFU was delivered with 2 mm pins using a multipoint inoculator (AQS Manufacturing Ltd.; serial no. 07/AQS/A400/119) onto Mueller–Hinton agar plates supplemented with 5% defibrinated sheep blood (E&O Laboratories Ltd.) and appropriate antibiotic concentration. Plates were incubated microaerophilically at 37°C for 48 hours. The breakpoints for susceptibility and resistance, respectively, were ampicillin, ≤8 and ≥32 μg/mL; ciprofloxacin, ≤1 and ≥4 μg/mL; erythromycin, ≤8 and ≥32 μg/mL; gentamicin, ≤2 and ≥8μg/mL; kanamycin, ≤16 and ≥64 μg/mL; nalidixic acid, ≤16 and ≥64 μg/mL; and tetracycline, 4≤and ≥16μg/mL, as described previously (Luangtongkum et al., 2007). All antimicrobial media were tested using control strains with established MICs for these antibiotics (C. jejuni NCTC 11351 [=ATCC 33560]; C. coli NCTC 11366 [=ATCC 33559]).

Statistical analysis

Standardized phenotypic distances (d ₂) (adapted from the original method of Manly, 1985) between isolates from humans, chicken meat, and cattle feces were calculated, based upon frequencies of resistance to each antibiotic, generating a number between 0 and 1, where 1 represents complete similarity of populations. A significant distance or “difference” was obtained using randomization tests (10,000 iterations). Briefly, the data from the three populations were randomized in Excel using Poptools (www.cse.csiro.au/poptools/), the analysis was rerun, and the difference in diversity was calculated, repeating the process 10,000 times. The posterior distribution of diversity differences obtained was then compared with the diversity difference calculated for the two original collections to obtain the level of significance (p-value). The difference in MIC distribution in each host was also calculated using the same randomization tests and was represented graphically using cumulative frequency graphs.

Results and Discussion

From 329 Campylobacter isolates, 221 (67%) were susceptible to all 7 antibiotics tested, 57 (17%) were susceptible to 1 antibiotic tested, 27 (8%) to 2 antibiotics tested, 17 (5%) to 3 antibiotics tested, and 6 (2%) to 4 antibiotics tested, and 1 isolate (0.3%) was resistant to 6 antibiotics. Twenty-four (7%) isolates were resistant to three or more classes of antibiotics. The highest resistance prevalence in human and chicken meat isolates was observed with ampicillin and tetracycline (Table 1), while nalidixic acid resistance was highest in cattle fecal isolates.

Table 1.

Prevalence of Antimicrobial Resistance in Campylobacter Human, Chicken Meat, and Cattle Fecal Isolates

	Human, n (%)			Chicken, n (%)			Cattle, n (%)
	S	I	R	S	I	R	S	I	R
Ampicillin	68 (57.1)	13 (10.9)	38 (31.9)	67 (63.8)	12 (11.4)	26 (24.8)	96 (91.4)	1 (1.0)	8 (7.6)
Gentamicin	110 (92.4)	11 (9.2)	3 (2.5)	91 (86.7)	13 (12.4)	4 (3.8)	100 (95.2)	0 (0)	5 (4.8)
Kanamycin	112 (94.1)	5 (4.2)	6 (5.0)	93 (88.6)	2 (1.9)	11 (10.5)	101 (96.2)	0 (0)	4 (3.8)
Tetracycline	85 (71.4)	0 (0)	34 (28.6)	79 (75.2)	0 (0)	26 (24.8)	94 (89.5)	0 (0)	11 (10.5)
Nalidixic acid	98 (82.4)	14 (11.8)	7 (5.9)	84 (80.0)	8 (7.6)	13 (12.4)	83 (79.0)	1 (1.0)	21 (20.0)
Ciprofloxacin	96 (80.7)	6 (5.0)	17 (14.3)	88 (83.8)	4 (3.8)	13 (12.4)	96 (91.4)	7 (6.7)	2 (1.9)
Erythromycin	113 (95.0)	4 (3.4)	2 (1.7)	100 (95.2)	3 (2.9)	2 (1.9)	105 (100.0)	0 (0)	0 (0)

S, susceptible; I, intermediate resistance; R, resistance to the specified antimicrobial.

The difference in resistance frequencies between human and chicken meat Campylobacter isolates was low (phenotypic distance d ₂ = 0.0082, p > 0.05) In contrast, the difference between human and cattle fecal isolates was greater (d ₂ = 0.0136, p < 0.05) as was that between cattle fecal and chicken meat isolates (d ₂ = 0.0361, p < 0.05). In addition, the MICs determined for each antibiotic revealed that Campylobacter isolates from humans and chicken meat were similar to each other for six of the antimicrobials (p > 0.002) but were significantly different for gentamicin (p < 0.001) (Fig. 1). The MIC distributions for cattle fecal isolates were distinct from Campylobacter isolates from the other two sources for all antibiotics (p < 0.001) except gentamicin (p > 0.002).

FIG. 1.

Minimum inhibitory concentration (MIC) cumulative frequency graphs showing distribution of MICs for ampicillin (a), gentamicin (b), kanamycin (c), tetracycline (d), nalidixic acid (e), ciprofloxacin (f), and erythromycin (g) in all hosts. In each graph, the degree of similarity between populations (as represented by a high p-value in phenotypic distance tests) is indicated by line overlapping or a small distance between lines at the middle value (0.5). Graph b (Gentamicin) is the only comparison where all populations showed similar MIC distributions (p < 0.05).

This study has identified that the antimicrobial resistance profiles of Campylobacter isolates from chicken meat and cattle feces are distinct from each other and this is likely due to differences in their antimicrobial exposures. Comparison of these profiles to those of clinical isolates identified that human isolates were much more similar to the chicken meat than to the cattle fecal isolates, suggesting that chicken meat is a much more important source of human infection. In addition, it highlights that the transfer of antimicrobial-resistant strains from animal reservoirs to humans is possible.

Footnotes

Acknowledgments

We would like to thank the University of Aberdeen and The Food Standards Agency-Scotland for funding. We also thank Anne Thomson and Emma Spragg (University of Aberdeen) for their valuable contribution to laboratory work.

Disclosure Statement

No competing financial interests exist.

References

Gormley

, Macrae

, Forbes

, Ogden

, Dallas

, Strachan

. Has retail chicken played a role in the decline of human campylobacteriosis? Appl Environ Microbiol, 2008; 74:383–390.

Luangtongkum

, Morishita

, El-Tayeb

, Ison

, Zhang

. Comparison of antimicrobial susceptibility testing of Campylobacter spp. by the agar dilution and the agar disk diffusion methods. J Clin Microbiol, 2007; 45:590–594.

Manly

BFG

. The Statistics of Natural Selection on Animal Populations. New York: Chapman and Hall, 1985.

[NCCLS] National Committee for Clinical Laboratory Standards. Performance Standards for Antimicrobial Disk and Dilution Susceptibility Tests from Bacteria Isolated from Animals. Approved Standard M31-A2. Wayne, PA: NCCLS, 2002.

Nelson

, Smith

, Vugia

, Rabatsky-Ehr

, Segler

, Kassenborg

, Zansky

, Joyce

, Marano

, Hoekstra

, Angulo

. Prolonged diarrhea due to ciprofloxacin-resistant campylobacter infection. J Infect Dis, 2004; 190:1150–1157.

Sheppard

, Dallas

, Strachan

, Macrae

, McCarthy

, Wilson

, Gormley

, Falush

, Ogden

, Maiden

, Forbes

. Campylobacter genotyping to determine the source of human infection. Clin Infect Dis, 2009; 48:1072–1078.

Stanley

, Jones

. Cattle and sheep farms as reservoirs of Campylobacter. J Appl Microbiol, 2003; 94,Suppl:104S–113S.

Strachan

, Gormley

, Rotariu

, Ogden

, Miller

, Dunn

, Sheppard

, Dallas

, Reid

, Howie

, Maiden

, Forbes

. Attribution of Campylobacter infections in northeast Scotland to specific sources by use of multilocus sequence typing. J Infect Dis, 2009; 199:1205–1208.

Studahl

, Andersson

. Risk factors for indigenous campylobacter infection: a Swedish case-control study. Epidemiol Infect, 2000; 125:269–275.

10.

Veterinary Medicines Directorate. Sales of Antimicrobial Products Authorised for Use as Veterinary Medicines, Antiprotozoals, Antifungals, Growth Promoters and Coccidiostats, in the UK in 2006. 2007. http://www.vmd.gov.uk/Publications/Antibiotic/salesanti06.pdf.