Journal of Antimicrobial Chemotherapy (2007) 60, 715– 723 doi:10.1093/jac/dkm300 Advance Access publication 18 August 2007

Macrolide-resistant Campylobacter: the meat of the matter Aimee E. Belanger and Thomas R. Shryock* Elanco Animal Health, A Division of Eli Lilly and Company, 2001 West Main Street, GL21, PO Box 708, Greenfield, IN 46140, USA The use of macrolide antibiotics in food animals has the potential to select for macrolide-resistant strains of resident bacterial flora. This may include the animal pathogens that are the intended targets of macrolide antibiotic intervention and Campylobacter, common inhabitants of the intestinal tract of food animals that are zoonotic pathogens in man. Such Campylobacter strains are not only resistant to the macrolide antibiotics used in food animals, e.g. tylosin, tilmicosin and tulathromycin, but to the macrolide antibiotics used in human medicine, e.g. erythromycin, azithromycin and clarithromycin, as well. Retail meat is a possible source of Campylobacter and persons consuming the meat derived from macrolide-treated food animals could acquire infections due to macrolide-resistant strains of this organism. Erythromycin is sometimes used to treat human cases of campylobacteriosis and those infected with animal-derived macrolide-resistant Campylobacter may not respond to treatment. The actual risk to human health from the use of macrolide antibiotics in food animals has been difficult to determine because of a lack of information about the macrolide-resistant Campylobacter found on the farm and in the clinic. Recently, however, a plethora of new information has become available on this topic. This review discusses what is currently known about the selection of macrolide-resistant Campylobacter in food animals, the prevalence of macrolide-resistant Campylobacter on retail meat, the prevalence of animal-derived macrolide-resistant Campylobacter in the clinic and the human health consequences associated with macrolide-resistant Campylobacter infection. This work will emphasize the comprehensive body of data generated in Denmark and the US as part of government-sponsored research studies over the last 10 years. These scientific findings may allow informed decisions to be made in the future about how macrolide antibiotics should be used in food animals while still safeguarding human health. Keywords: erythromycin, tylosin, pork, chicken, beef

Campylobacteriosis The inadvertent ingestion of Campylobacter, most commonly Campylobacter jejuni or Campylobacter coli, may lead to campylobacteriosis, one of the most prevalent forms of bacterial gastroenteritis worldwide.1 Infections may be acquired by individuals both domestically and abroad; and in some countries, the number of cases of campylobacteriosis exceeds that of salmonellosis.2,3 Campylobacteriosis may occur as a household or community outbreak or, more commonly, as a sporadic infection.1 The hallmark of campylobacteriosis is diarrhoea that lasts 3 –5 days but other symptoms may include fever, headache, abdominal pain, myalgia, vomiting or bloody stools.4 The illness is usually selflimiting and complications are rare.4,5 Antibiotics play an important but limited role in the treatment of campylobacteriosis. The

vast majority of individuals will never seek medical care and, thus, will not be treated with antibiotics. Regimens of fluoroquinolone or macrolide antibiotics are often administered to individuals with severe symptoms or invasive disease.4,6 Campylobacteriosis is considered to be a ‘new’ disease in that it was not clinically characterized until 1977.7 Since then, considerable efforts have been made to identify potential sources of campylobacteriosis in man. Campylobacter spp. are ubiquitous and known reservoirs include: ground and surface fresh water, salt water, wild and domestic animals, unpasteurized milk, sand, soil, sewage and people.1,8,9 Disease outbreaks have been linked to water, unpasteurized milk and a variety of food stuffs. Risk factors for sporadic infection include exposure to poultry products, the consumption of undercooked chicken, the consumption of unpasteurized milk, exposure to certain kinds of

.....................................................................................................................................................................................................................................................................................................................................................................................................................................

*Corresponding author. Tel: þ1-317-277-5087; Fax: þ1-317-651-6075; E-mail: [email protected] .....................................................................................................................................................................................................................................................................................................................................................................................................................................

715 # The Author 2007. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: [email protected] The online version of this article has been published under an open access model. Users are entitled to use, reproduce, disseminate, or display the open access version of this article for non-commercial purposes provided that: the original authorship is properly and fully attributed; the Journal and Oxford University Press are attributed as the original place of publication with the correct citation details given; if an article is subsequently reproduced or disseminated not in its entirety but only in part or as a derivative work this must be clearly indicated. For commercial re-use, please contact [email protected]

Review water, contact with animals, foreign travel, contact with infected humans and the consumption of various other foods.1,10 – 15

Macrolide-resistant Campylobacter in food animals A number of Campylobacter species may be isolated from food animals16 but, given the importance of C. jejuni and C. coli in disease, antibiotic resistance studies in food animals focus almost exclusively on these two. The predominant Campylobacter species in cattle and chickens is C. jejuni and in swine C. coli.2,17 Erythromycin and other macrolide antibiotics bind to the 50S subunit of the bacterial ribosome and physically prevent the elongation of nascent polypeptide chains.18,19 The 50S macrolide-binding site is composed of portions of the 23S rRNA subunit, ribosomal protein L4 and ribosomal protein L22. Bacteria may acquire macrolide resistance when the genes encoding these components sustain mutations that alter the drug binding site. Among the bacteria that rely on this particular macrolide resistance mechanism, substitutions in the 23S rRNA gene at nucleotide positions that correspond to nucleotide positions 2058 and 2059 in the E. coli 23S rRNA gene are very common.20 In Campylobacter, the nucleotides in question are 2074 and 2075 and mutations at these positions play a prominent role in the development of high-level macrolide resistance in clinical isolates (MIC 128 mg/L).21 The majority of such isolates (78– 100%) harbour an A2075G mutation; of the remainder, A2074C is most common.22 – 24 In bacteria with more than one 23S rRNA gene, the level of macrolide resistance is dependent on the number of mutant alleles.25 Campylobacter contains three 23S rRNA genes and there is some indication that the number of mutant alleles influences the MIC.26,27 Most clinical isolates of Campylobacter with high-level macrolide resistance contain three copies of the mutated allele.21 Mutations in the genes encoding the L4 and L22 ribosomal proteins that affect the binding of macrolide antibiotics to the 50S ribosome subunit and, thus, confer resistance have been identified in some bacteria.24 The critical alterations affect amino acid position 63 –74 of the L4 protein and the C-terminal b-hairpin motif of the L12 protein.24 Some clinical isolates of macrolide-resistant Campylobacter with 23S rRNA mutations have also been shown to contain mutations in the genes encoding the L4 and L22 proteins but the significance of this finding is unknown.22,24 Many of the identified mutations did not fall within the macrolide-binding regions of the two proteins and those that did were present in both resistant and susceptible strains.22,24 The intrinsic resistance of Campylobacter to dyes, detergents and some antibiotics, including erythromycin, is mediated in part by the CmeABC efflux pump.21 With specific regard to erythromycin, an increase in efflux pump activity due to the overexpression of cmeABC 21 or another yet to be defined CmeABC-related mechanism27 confers low-level resistance to Campylobacter (MIC 8 –16 mg/L).21 Increased efflux in tandem with 23S rRNA mutations also leads to increased MICs in some highly-resistant Campylobacter isolates.21 The role of Campylobacter efflux mutants in erythromycin treatment failures in humans remains to be determined. New breakpoints established by the CLSI (susceptible 8 mg/L; resistant 32 mg/L) suggest that the infections caused by efflux mutants with lowlevel erythromycin resistance can still be successfully treated.28

In highly-resistant strains with both ribosomal alterations and increased efflux, it is the 23S rRNA mutations that are responsible for the clinical resistance to erythromycin; the chemical inhibition of efflux in such strains decreases the MIC but does not restore susceptibility.21 The ribosomal binding site of macrolide antibiotics overlaps those of lincosamide and streptogramin B antibiotics and modification of the macrolide antibiotic binding site may lead to the MLSB phenotype, the simultaneous expression of macrolide, lincosamide and streptogramin B antibiotic resistance.18 In Campylobacter, the mutations conferring macrolide resistance also confer resistance to lincosamide antibiotics but not to streptogramin B antibiotics; this genus is intrinsically resistant to this class of antibiotics.29,30 As this cross-resistance predicts, the use of lincomycin in food animals co-selects for macrolide-resistant Campylobacter.31 Thus, the prevalence of macrolide resistance among the Campylobacter isolated from the intestinal tracts of food animals today is the result of the use of both macrolide and lincosamide antibiotics. Generally speaking, the use of macrolide and lincosamide antibiotics in food animals may select for macrolide-resistant Campylobacter, but the extent to which this selection occurs depends on the animal species and macrolide antibiotic use in question. In a recent US study, the prevalence of macrolide resistance among the Campylobacter spp. isolated from pigs given tylosin at 40 g/ton of feed during the nursery and finishing stages was 77% while that of pigs reared under antibiotic-free conditions (no antibiotics after 3 weeks of age) was 34.5%.32 In feedlot cattle administered subtherapeutic levels of tylosin for long periods of times, ,1% of the C. jejuni isolated were found to be macrolide resistant.33 In chickens, the administration of low levels of macrolide and lincosamide antibiotics throughout the life of the bird favours the selection of macrolide-resistant Campylobacter,27,31,34,35 but the use of macrolide antibiotics for disease treatment, at higher concentrations of active agent for short periods of time, does not.27 Actual macrolide use patterns in the field have been tracked most closely in Denmark as part of a government-sponsored initiative. Macrolide and lincosamide antibiotics are used to treat enteric and respiratory disease in Danish cattle, swine and broiler chickens but swine receive the bulk of the doses, most often as weaners. In 2005, some 26 million head of swine in Denmark received nearly 65 million doses of macrolide antibiotics and 25 million doses of lincosamide antibiotics or .99% of all macrolide and lincosamide antibiotic doses administered to the three major food animal species.36 Complexities unique to US animal production have precluded government monitoring of antibiotic use patterns, but such data may be obtained from commercially-available marketing research. Approximately half of the 18– 20 million US feedlot cattle on hand in any given quarter of 2005 received macrolide antibiotics, the majority for the reduction in the incidence of liver abscesses (J. Kohler, dmrkynetec, St Louis, MO, USA, personal communication). Some 23– 44% of the 181 million pigs that passed through each of four weight classes in the US in 2005 received macrolide and lincosamide antibiotics for one or more of the following: treatment or control of respiratory or enteric disease or for increased feed efficiency. Pigs at the weaner to 50 lb weight class received the largest percentage of these antibiotics because of age-specific issues with respiratory disease (J. Kohler, dmrkynetec, St Louis, MO, USA, personal

716

Review communication). In 2005, 123.3 million (1.4%) of the estimated 8.9 billion broiler chickens in the US received lincosamide antibiotics for the control of enteric disease (G. Rennier, Rennier & Associates, Inc., Columbia, MO, USA, personal communication). That same year, 201 million macrolide and lincosamide treatments were administered to broiler chickens for the treatment of enteric or skin disease (G. Rennier, Rennier & Associates, Inc., Columbia, MO, USA, personal communication). Assuming that one treatment is equivalent to one bird, some 2.3% of broiler chickens were treated with macrolide and lincosamide antibiotics in 2005. There is a general perception that large amounts of macrolide antibiotics are currently used in broiler chickens for growth promotion purposes but this class of antibiotics has actually not been used for this purpose by the those surveyed since 2001 (G. Rennier, Rennier & Associates, Inc., Columbia, MO, USA, personal communication). The prevalence of macrolide resistance among the Campylobacter isolated from food animals has been monitored in Denmark and the US for nearly a decade. Although absolute numbers vary, similar trends are seen in Denmark and the US: the prevalence of erythromycin resistance is low in the C. jejuni isolated from cattle and chickens and higher in the C. coli isolated from swine and chickens (Tables 1 and 2). The most notable difference between the two countries is that the prevalence of macrolide resistance in the C. coli of Danish swine has declined over time presumably because of discontinued use of macrolide antibiotics as growth promoters in 1998 (Table 1). This same impact is not evident for the C. jejuni isolated from Danish cattle and chicken: macrolide resistance has been maintained at low levels in this Campylobacter species throughout the sampling period (Table 1).

Macrolide-resistant Campylobacter on retail meat Meat is contaminated with Campylobacter when faecal matter is inadvertently released from the intestinal tract of a slaughtered animal during processing. The likelihood of this occurring is not the same for all food animal species as intestinal Campylobacter loads and processing methods vary. Chickens are most likely to become contaminated with Campylobacter in the slaughter plant. The concentration of Campylobacter in the intestinal

Table 1. Prevalence of erythromycin resistance in the Campylobacter isolated from food animals, retail meat and humans in Denmarka Percentage of erythromycin-resistant C. jejuni

Year cattle chickens 1997 1998 1999 2000 2001 2002 2003 2004 2005

4 0 0 0 8 0 0 0 2

Percentage of erythromycin-resistant C. coli

chicken chicken meatb humansc swine chickens meatd

1 2 0 5 0 0 0 1 2

3 5 0 — 2/0 0/0 0/0 0/3 1/2

2 0 0 3.8 (2/7) 5.7 (7/0) 0.9 (0/5) 1.9 (1/7) 5 (5/8) 0 (0/0)

71 68 36 38 30 32 24 23 20

58 33 25 40 27 — — — —

13 — — — — — — 33/12 0/17

a

Derived from DANMAP 1997– 2005.36,102 – 109 Data reported for domestic product or domestic/imported product. c Data reported for total cases or total cases (domestically acquired/ travel-related). d Data reported for domestic product or domestic/imported product. b

contents of the chicken may be as high as .109 cfu/g37,38 and the accidental spillage of this material during mechanical evisceration may lead to significant contamination levels on the meat. The use of common chill tanks during processing also promotes cross-contamination; even carcasses that were clean after evisceration may become contaminated with the Campylobacter released from the intestinal tract of other processed birds.37,38 The concentration of Campylobacter in the intestinal contents of cattle and swine is much lower than that of chickens, 102 – 105 cfu/g2,38 – 40 and manual evisceration, carcass washes and individual air-chilling of beef and pork carcasses effectively minimize Campylobacter contamination.38,41 Numerous studies have been performed to evaluate the fate of Campylobacter through the various microbiological control

Table 2. Prevalence of erythromycin resistance in the Campylobacter isolated from food animals, retail meat and humans in the USa Percentage of erythromycin-resistant C. jejuni

Percentage of erythromycin-resistant C. coli

Year

cattle

beef

chicken

chicken meat

humans

chicken

chicken meat

swine

pork

humans

1997 1998 1999 2000 2001 2002 2003 2004 2005

— — 0.5 — — — — — —

— — — — — 0 0 0 —

— 3.1 0.2 5.1 3.1 0.6 1.6 — —

— — 6 — — 0 0 0.8 —

3 1 2.4 1.6 1.9 1.8 0.3 — —

— 23.8 14.9 22.7 11.5 18.8 20.2 — —

— — 0 — — 18.9 34 9.2 —

— — — — — — — 59.7 28.4

— — — — — 33 75 33 —

0 37.5 10 8.3 5.9 4 9.1 — —

a

111

Derived from Englen et al.,110 NARMS Annual Retail Meat Report 2002– 2004,50 – 52 USDA-NARMS,

USDA-CAHFSE 2003 –2005.114,115

717

Gupta et al.,112 CDC-NARMS113 and

Review points that are in place along the processing chain.38 Such studies are important for determining what measures should be implemented to control contamination in the slaughter plant but have limited value in assessing human exposure. The levels of Campylobacter contamination on meat in the packing plant does not necessarily reflect those that a consumer may encounter because bacterial load on the meat may decline after additional refrigeration or freezing.38 The likelihood that a person will be exposed to disease-causing quantities of macrolide-resistant Campylobacter on meat is best represented by the microbiological quality of packaged retail meat. The surveys of raw retail meat conducted in Denmark and the US show the same trend with regards to Campylobacter contamination: these bacteria may be isolated from a significant proportion of chicken meat samples but is rarely isolated from beef or pork (Table 3). The prevalence of Campylobacter on raw beef and pork is so low that these bacteria/meat combinations were not even considered in similar Canadian or Australian retail meat surveys.42,43 The prevalence of erythromycin resistance among the C. jejuni isolated from raw retail meat is extremely low and sometimes no macrolide-resistant strains are recovered (Tables 1 and 2). In keeping with animal surveillance studies, erythromycin resistance is most common in the C. coli isolated from pork and chicken meat (Tables 1 and 2). In order to become sick with campylobacteriosis, a person must ingest an infectious dose of Campylobacter and not all retail meat, even that which tests positive in microbiological surveys, contains a sufficient number of organisms to induce illness. Determining the precise amount of Campylobacter organisms that are required for illness has been confounded by acquired immunity. Individuals experience symptomatic disease only when they are exposed to Campylobacter strains that their immune systems do not recognize,44 – 46 which makes it difficult to predict how any one person will react after consuming a given quantity of bacteria. What is known is that the ingestion of as little as 500 organisms is sufficient to cause disease in some people.47 Other scientific data support the assumption that the infectious dose required for becoming sick with campylobacteriosis is at least 500 cfu of Campylobacter. Known vehicles

Table 3. Prevalence of Campylobacter spp. in retail meat samples % Positive Country

Year

beef

pork

chicken

Reference

Denmark Denmark Denmark Denmark Denmark US US US US US US

1997 2002 2003 NDb 2005 1999 1999–2000 ND 2002 2003 2004

– 0.1 0 0 — — 0.5 — 0 0.1 0

1 0.2 0 0.3 — — 1.7 1.3 0.8 0.4 0.3

33 41.7 32.9/50.3a 38.7 19.1/70.5 44 70.7 — 46.8 52.3 60.2

116 53 117 118 55 119 120 121 50 52 51

a

Domestic/imported product. Not defined.

b

for the transmission of Campylobacter, e.g. water, raw milk and chicken rinsate, contain very high concentrations of these bacteria, 105 – 108 cfu/mL48 and the ingestion of even small quantities of these liquids would be sufficient to cause disease. A minimal infectious dose of 500 cfu is also consistent with the finding that the number of Campylobacter organisms that may be transferred in the kitchen via cross-contamination can exceed this amount.49 In the US studies, the actual amount of Campylobacter in a positive meat sample could not be determined because the experimental method employed an enrichment step.50 – 52 However, the methods used in Denmark have permitted some estimation of bacterial loads. In 2002, 0.2% (5) of the 2413 raw minced pork samples was found to contain Campylobacter at levels of 0.4 – 40 cfu/g.53 In this same study, 0.1% (3) of the 3046 raw minced beef samples was found to contain Campylobacter at levels of 0.4 – 4 cfu/g.53 It is important to note that these findings do not actually prove that some pork and beef samples contained as much as 40 and 4 cfu/g, respectively, of Campylobacter. The ranges reported reflect predetermined sensitivity parameters of the assay used and it is possible that the load reported is based on the growth of a random colony on one of the dilution plates in the series. Assuming that the amount of Campylobacter that is detected represents the true amount of infectious organisms that is present, and that a dietary serving of meat is 100 g, it appears that beef and pork, even if eaten raw, rarely, if ever, contain an infectious dose of Campylobacter, macrolide-resistant or otherwise. That chicken is indeed a source for an infectious dose of Campylobacter was more convincing: 42% of the 712 samples were found to be positive and some samples contained .4000 cfu/g of bacteria using the same assay methods as described earlier.53

Macrolide-resistant Campylobacter in humans For unknown reasons, the incidence of campylobacteriosis (reported cases per 100 000 people) may vary widely from country to country: New Zealand, 432.4;54 Australia, 117;42 Denmark, 68;55 US, 12.8.56 In Denmark and the US, the majority of all cases of campylobacteriosis (90 – 100%) are caused by C. jejuni,55,57 but C. coli and other unidentified thermophilic campylobacters play a larger role in disease elsewhere.46,58,59 The prevalence of macrolide resistance among clinical isolates of C. jejuni in Denmark and the US is generally very low (Tables 1 and 2). One of the earliest US surveys of erythromycin resistance among human Campylobacter strains was published in 1984 or 23 years after macrolide antibiotics were introduced for use in food animals.60 The prevalence of resistance was low at that time (4%) and has remained so (Table 2).61,62 The prevalence of erythromycin resistance among C. coli clinical isolates in the US has been ,10% since 2000 (Table 2); Denmark does not report resistance data for this Campylobacter species. A review of worldwide macrolide resistance rates in the Campylobacter of humans and food animals is already available.63 Source attribution studies with Campylobacter have long been hampered by typing methods with limited discriminatory power such as serotyping but this situation has changed with the advent of more sophisticated genetic-based methods. The use of genomotyping has revealed that the majority of C. jejuni strains isolated from humans are most similar to those originating in the

718

Review environment; only 44% are apparently derived from livestock.8 Interestingly, the number of campylobacteriosis cases observed in Belgium declined by about that same amount following the withdrawal of chicken from the market due to dioxin contamination.64 Indeed, some human C. jejuni strains have been shown to be related to those derived from chicken and chicken meat by amplified fragment length polymorphism (AFLP).65 Genetic overlap also exists between some human and cattle C. jejuni strains,65 a result that is still compatible with the low Campylobacter contamination rates of beef as unpasteurized milk is a well-known source of Campylobacter. AFLP analysis also suggests that poultry, but not pigs, are a source of the C. coli strains that infect man.66 Campylobacter strains that appear to be human-specific, i.e. not related to those found in food animals, have also been identified by AFLP,65,66 a finding that is consistent with the genomotyping work. An analysis of C. coli strains by multilocus sequence typing (MLST) analysis indicates that few sequence types are shared between strains from humans and pigs and that these shared sequence types are also found in strains from poultry.67 But are all macrolide-resistant Campylobacter strains that infect man derived from food animals? The evidence available thus far suggests that the answer to this question is no. Food animal-independent reservoirs of antibiotic-resistant Campylobacter do exist59 and although campylobacteriosis is not a communicable disease in the more traditional sense, human-to-human transmission of both macrolide-susceptible and macrolide-resistant strains is known to occur.10 – 13,44,68,69

Human health consequences associated with macrolide-resistant Campylobacter Treatment failures Only a small number of people with campylobacteriosis actually receive erythromycin, which may seem counterintuitive given that this antibiotic is the treatment of choice for this disease.6 However, erythromycin is the preferred treatment for diagnosed cases of campylobacteriosis and most diarrhoeic patients that seek a physician’s care are empirically treated with a broadspectrum antibiotic. In a recent Danish study involving 122 patients with known treatment histories, 40 (32.8%) were treated with antibiotics. Of the 40 patients, 33 (82.5%) were treated with a fluoroquinolone antibiotic, 6 (15%) were treated with a macrolide antibiotic and 1 (2.5%) was treated with both.70 In other studies, 25% of the treated campylobacteriosis patients received erythromycin antibiotics.5,71,72 Physician guidelines suggest that erythromycin will shorten the duration of diarrhoea in campylobacteriosis patients if given early enough after the onset of symptoms.11,73,74 Thus, patients experiencing erythromycin treatment failure due to infection with macrolide-resistant Campylobacter might experience extra days of diarrhoea relative to those patients that are successfully treated. A number of studies with antibiotic-treated campylobacteriosis patients have been conducted and yet erythromycin treatment failures have not been described.5,59,70,72 There may be several reasons for this: (i) the researchers may have limited themselves to only defining fluoroquinolone treatment failures; (ii) there were too few erythromycin-treated patients infected with macrolide-resistant Campylobacter to analyse; or (iii) the

treatment outcomes may not have differed between erythromycin-treated patients infected with macrolide-resistant Campylobacter strains and those infected with macrolidesusceptible strains. With regards to the latter, the empirical treatment of campylobacteriosis patients with erythromycin decreases the duration of diarrhoea by 2 days,75 but the treatment of patients with erythromycin after laboratory diagnosis, the more true-to-life scenario, does not ameliorate symptoms.76 – 78 Indeed, a meta-analysis of controlled antibiotic (macrolide as well as fluoroquinolone) clinical treatment studies showed a reduction of just 1.3 days of intestinal symptoms, with early (i.e. empirical) treatment achieving better results, but leading the authors to conclude that a ‘restrictive attitude’ to administration of antibiotics in uncomplicated cases in non-high risk patients should be employed.79 Effective antibiotic treatment is especially important for those patients that are seriously ill. Bacteraemia occurs at an average incidence rate of 1.5 per 1000 cases of campylobacteriosis in the UK80 or 0.2 per 100 000 people in Denmark81 and afflicts otherwise healthy individuals and those with underlying disease such as AIDS.80,82 Of all campylobacteriosis patients, those that are seriously ill are probably the least likely to experience erythromycin treatment failure. Standardized antibiotic susceptibility testing, e.g. methods approved by the CLSI,28 is routinely performed on the Campylobacter of hospitalized patients and if an individual was infected with a macrolide-resistant Campylobacter strain, then erythromycin would not be administered. An inability to use macrolide antibiotics for the treatment of Campylobacter-infected patients with serious disease is not necessarily detrimental. Erythromycin is not a drug of last resort and bacteraemic patients may be successfully treated with other classes of antibiotics.6,81 – 83 An inability to use macrolide antibiotics was considered to further limit treatment options for children because the use of fluoroquinolones was previously contraindicated.84 However, current physician guidelines have considered new evidence on safety and suggest that fluoroquinolones may be cautiously used in children.85,86

Other proposed consequences Some people with campylobacteriosis develop autoimmune disorders, e.g. Guillain– Barre´ Syndrome (GBS) or Reiter’s Syndrome, after a period of convalescence from acute gastrointestinal symptoms.4,87,88 The incidence of GBS, which may be triggered by viral or bacterial infection, is 1.3 cases per 100 000 people.88 At least one-third of all GBS cases are specifically preceded by infection with Campylobacter.88 Reiter’s syndrome develops in anywhere from 0 –16% of campylobacteriosis patients.4,89,90 Host factors play a significant role in determining whether a person develops either GBS or Reiter’s syndrome.4,87,88 Persons that develop Reiter’s syndrome may experience longer bouts of diarrhoea during their initial campylobacteriosis infection90 and the role of antibiotic treatment failure, with its potential for extra days of diarrhoea, in the development of this condition has been questioned.91 It is known, however, that the successful antibiotic treatment of the initial Campylobacter infection, and, presumably, a shorter duration of diarrhoea, has not been shown to prevent the development of Reiter’s syndrome.89

719

Review Certain people are more likely to acquire campylobacteriosis or to suffer from a poor disease outcome than others. Predisposing factors include co-morbidity, immunosuppression, age and certain kinds of drug therapy.80,83,92 – 94 Mortality is highest in bacteraemic patients4,80,83,95 and is most likely to occur in the first month following illness.79 Certain people are also more likely to be infected with antibiotic-resistant strains of Campylobacter: those that have received antibiotic treatment prior to laboratory diagnosis59,72 and those with underlying medical conditions.59 In a recent Danish registry-based cohort study, patients infected with macrolide-resistant Campylobacter were found to be .5-fold more likely to experience either invasive illness or death than patients infected with macrolidesusceptible strains.5 To explain these findings, the authors suggest that macrolide-resistant strains of Campylobacter may be more virulent than their macrolide-susceptible counterparts.5 Increased virulence is but one hypothesis among many that could explain the association observed in the study and conclusions cannot be made until these study limitations are addressed: (i) the adverse events experienced by the patients were not shown to be the result of their infection with macrolide-resistant Campylobacter; (ii) the patients infected with macrolide-resistant Campylobacter that experienced adverse events were elderly (67– 91 years of age) and nearly all had serious underlying medical conditions; (iii) the association between having an infection with a macrolide-resistant Campylobacter strain and experiencing an adverse event was not observed until 90 days; (iv) potential confounders such as the exposure of patients to macrolide antibiotics for medical conditions other than their campylobacteriosis, e.g. respiratory illness, were not considered; and (v) patient data were incomplete. There is of yet no scientific evidence to suggest that certain strains of Campylobacter spp. that infect man are indeed more virulent than others. In the past, some have described the infections caused by C. coli as being more severe than those caused by C. jejuni 96 while others have stated just the opposite.30,97 A more recent study suggests that individuals infected with either C. jejuni or C. coli experience similar symptoms.70 Genome sequence analysis does indicate that the repertoire of putative virulence genes varies slightly between C. jejuni and C. coli,98 but the biological significance of these differences is unknown. The C. jejuni strains originating in livestock do not appear to cause more severe disease than do environmental strains8 and although certain serotypes have been linked to GBS,87,88 certain clonal complexes have not.99 With regards specifically to macrolide-resistant Campylobacter, participants in a previous Danish epidemiological study that were infected with macrolide-resistant C. coli were not observed to experience more severe symptoms than those infected with macrolide-susceptible C. coli.5

Future prospects There is a continued need to use antibiotics in food animals for the purposes of health and welfare. At the same time, measures must be taken to ensure that the use of antibiotics in food animals does not adversely impact human health. In recognition of these seemingly competing needs, the World Health Organization and the Organisation Internationale des Epizooties

have issued a number of recommendations for how antimicrobial resistance might be contained within food animals.84,100 One theme found throughout these recommendations is to use only those veterinary antibiotics with acceptable risk profiles, as determined by scientific-based analyses. Thus far, the use of risk analysis methods to determine antibiotic risk profiles has not proved entirely satisfactory because of the uncertainty produced by gaps in scientific understanding.101 The new data discussed here closes some of these data gaps and may lead to increased confidence in the use of risk assessment to make decisions on how best to manage the potential risk to human health from the use of macrolide antibiotics in food animals.

Transparency declarations A. E. B. and T. R. S. are employees of Elanco Animal Health, A Division of Eli Lilly & Co., Greenfield, IN, USA.

References 1. Friedman C, Neimann J, Wegner J et al. Epidemiology of Campylobacter jejuni infections in the United States and other industrialized nations. In: Nachamkin I, Blaser M, eds. Campylobacter. 2nd edn. Washington, DC: ASM Press, 2000; 121–38. 2. Anonymous. Annual Report on Zoonoses in Denmark 2004. http://www.dfvf.dk/Default.aspx?ID=9606 (27 June 2007, date last accessed). 3. The OzFoodNet Working Group. Reported foodborne illness and gastroenteritis in Australia: annual report of the OzFoodNet network. Commun Dis Intell 2004; 29: 164 –90. 4. Skirrow M, Blaser M. Clinical aspects of Campylobacter infection. In: Nachamkin I, Blaser M, eds. Campylobacter. 2nd edn. Washington, DC: ASM Press, 2000; 69–88. 5. Helms M, Simonsen J, Olsen K et al. Adverse health events associated with antimicrobial drug resistance in Campylobacter species: a registry-based cohort study. J Infect Dis 2005; 191: 1050–5. 6. Gilbert D, Moellering R Jr, Eliopoulos G et al. The Sandford Guide to Antimicrobial Therapy. 35th edn. Hyde Park, VT: Antimicrobial Therapy Inc., 2005. 7. Skirrow M. Campylobacter enteritis: a “new” disease. Br Med J 1977; 2: 9–11. 8. Champion O, Gaunt M, Gundogdu O et al. Comparative phylogenomics of the foodborne pathogen Campylobacter jejuni reveals genetic markers predictive of infection source. Proc Natl Acad Sci 2005; 102: 16043–8. 9. Maugeri T, Carbone M, Fera M et al. Distribution of potentially pathogenic bacteria as free living plankton associated in a marine coast. J Appl Microbiol 2004; 97: 354 –61. 10. Ashkenazi S, Danziger Y, Versano Y et al. Treatment of Campylobacter gastroenteritis. Arch Dis Child 1987; 62: 84–5. 11. Butzler J, Goosens H. Campylobacter jejuni infection as a hospital problem: an overview. J Hosp Infect 1988; 11 Suppl A: 374–7. 12. Gaudreau C, Michaud S. Cluster of erythromycin- and ciprofloxacin-resistant Campylobacter jejuni subsp. jejuni from 1999 to 2001 in men who have sex with other men, Quebec, Canada. Clin Infect Dis 2003; 37: 131–6. 13. Goosens H, Henocque G, Kremp L et al. Nosocomial outbreak of Campylobacter jejuni meningitis in newborn infants. Lancet 1986; ii: 146–9.

720

Review 14. Llovo J, Mateo E, Munoz A et al. Molecular typing of Campylobacter jejuni isolates involved in a neonatal outbreak indicates nosocomial transmission. J Clin Microbiol 2003; 41: 3926– 8. 15. Scho¨nberg-Norio D, Takkinen J, Hanninen M-L et al. Swimming and Campylobacter infections. Emerg Infect Dis 2004; 10: 754–60. 16. Vandamme P. Taxonomy of the family Campylobacteraceae. In: Nachamkin I, Blaser M, eds. Campylobacter. 2nd edn. Washington, DC: ASM Press, 2000; 3 – 26. 17. Nielsen E, Engberg J, Madsen M. Distribution of serotypes of Campylobacter jejuni and Campylobacter coli from Danish patients, poultry, cattle and swine. FEMS Immunol Med Microbiol 1997; 19: 47 – 56. 18. Walsh C. Antibiotics that block protein synthesis. Antibiotic Actions, Origins, Resistance. Washington, DC: ASM Press, 2003; 51 – 69. 19. Tenson T, Lovmar M, Ehrenberg M. The mechanism of action of macrolides, lincosamides and streptogramin B reveals the nascent peptide exit path in the ribosome. J Mol Biol 2003; 330: 1005–14. 20. Nakajima Y. Mechanisms of bacterial resistance to macrolide antibiotics. J Infect Chemother 1999; 5: 61– 74. 21. Payot S, Bolla J-M, Corcoran D et al. Mechanisms of fluoroquinolone and macrolide resistance in Campylobacter spp. Microb Infect 2006; 8: 1967–71. 22. Gibreel A, Kos V, Keelan M et al. Macrolide resistance in Campylobacter jejuni and Campylobacter coli : molecular mechanism and stability of the resistance phenotype. Antimicrob Agents Chemother 2005; 49: 2753– 9. 23. Vacher S, Menard A, Bernard E et al. Detection of mutations associated with macrolide resistance in thermophilic Campylobacter spp. by real-time PCR. Microb Drug Res 2005; 11: 40 – 7. 24. Corcoran D, Quinn T, Cotter L et al. An investigation of the molecular mechanisms contributing to high-level resistance in Campylobacter. Int J Antimicrob Agents 2006; 27: 40–5. 25. Tait-Kamradt A, Davies T, Appelbaum P et al. Two new mechanisms of macrolide resistance in clinical strains of Streptococcus pneumoniae from Eastern Europe and North America. Antimicrob Agents Chemother 2000; 44: 3395– 401. 26. Vacher S, Me´nard A, Bernard E et al. PCR-restriction fragment length polymorphism analysis for detection of point mutations associated with macrolide resistance in Campylobacter spp. Antimicrob Agents Chemother 2003; 47: 1125– 8. 27. Lin J, Yan M, Sahin O et al. Effect of macrolide usage on emergence of erythromycin-resistant Campylobacter isolates in chickens. Antimicrob Agents Chemother 2007; 51: 1678– 86. 28. Clinical and Laboratory Standards Institute. Methods for Antimicrobial Dilution and Disk Susceptibility Testing of Infrequently Isolated or Fastidious Bacteria: Approved Guideline M45-A. CLSI, Wayne, PA, USA, 2006. 29. Burridge R, Warren C, Phillips I. Macrolide, lincosamide and streptogramin resistance in Campylobacter jejuni/coli. J Antimicrob Chemother 1986; 17: 315–21. 30. Taylor D, Echeverria P, Pitarangsi C et al. Influence of strain characteristics and immunity on the epidemiology of Campylobacter infections in Thailand. J Clin Microbiol 1988; 26: 863– 8. 31. Luangtongkum T, Morishita T, Ison A et al. Effect of conventional and organic production practices on the prevalence and antimicrobial resistance of Campylobacter spp. in poultry. Appl Environ Microbiol 2006; 72: 3600– 7. 32. Thakur S, Gebreyes W. Prevalence and antimicrobial resistance of Campylobacter in antimicrobial-free and conventional pig production systems. J Food Prot 2005; 68: 2402– 10. 33. Inglis G, McAllister T, Busz H et al. Effects of subtherapeutic administration of antimicrobial agents on beef cattle on the prevalence

of antimicrobial resistance in Campylobacter jejuni and Campylobacter hyointestinalis. Appl Environ Microbiol 2005; 71: 3872–81. 34. Ladely S, Harrison M, Fedorka-Cray P et al. Development of macrolide resistant Campylobacter in broilers administered subtherapeutic levels of tylosin. International Poultry Scientific Forum, Atlanta, GA, 2007. Abstract M57. 35. Berrang M, Ladely S, Meinersmann R et al. Subtherapeutic tylosin phosphate in broiler feed affects Campylobacter on carcasses during processing. Poult Sci 2007; 86: 1229–33. 36. DANMAP 2005. Use of Antimicrobial Agents and Occurrence of Antimicrobial Eesistance in Bacteria From Food Animals, Foods and Humans in Denmark. http://www.danmap.org/ (27 June 2007, date last accessed). 37. Berndtson E, Tivemo M, Engvall A. Distribution and numbers of Campylobacter in newly slaughtered broiler chickens and hens. Int J Food Microbiol 1992; 15: 45 –50. 38. Jacobs-Reitsma W. Campylobacter in the food supply. In: Nachamkin I, Blaser M, eds. Campylobacter. 2nd edn. Washington, DC: ASM Press, 2000; 467–81. 39. Oosterom J, Dekker R, de Wilde G et al. Prevalence of Campylobacter jejuni and Salmonella during pig slaughtering. Vet Q 1985; 7: 31 –4. 40. Stanley K, Wallace J, Currie J et al. The seasonal variation of thermophilic campylobacters in beef cattle, dairy cattle and calves. J Appl Microbiol 1998; 85: 472–80. 41. Pearce R, Wallace F, Call J et al. Prevalence of Campylobacter within a swine slaughter and processing facility. J Food Prot 2003; 66: 1550–6. 42. Australian Government. Pilot Surveillance Program for Antimicrobial Resistance in Bacteria of Animal Origin. http://www.health. gov.au/internet/wcms/publishing.nsf/content/health-pubhlth-strategjetacar-pdf-amrstrategy_affa.htm (27 June 2007, date last accessed). 43. Government of Canada. Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS) 2004. http://www. phac-aspc.gc.ca/cipars-picra/2004_e.html (27 June 2007, date last accessed). 44. Christenson B, Ringer A˚, Blucher C. An outbreak of Campylobacter enteritis among the staff of a poultry abbatoir in Sweden. Scand J Infect Dis 1983; 15: 167–72. 45. Blaser M, Sazie E, Williams JP. The influence of immunity on raw milk-associated Campylobacter infection. J Am Med Assoc 1987; 257: 43 –6. 46. Oberhelman R, Taylor D. Campylobacter infections in developing countries. In: Nachamkin I, Blaser M, eds. Campylobacter. 2nd edn. Washington, DC: ASM Press, 2000; 139–53. 47. Black R, Levine M, Clements M et al. Experimental Campylobacter jejuni infections in humans. J Infect Dis 1988; 157: 472 –9. 48. Yang C, Jiang Y, Huang K et al. Application of real-time PCR for quantitative detection of Campylobacter jejuni in poultry, milk and environmental water. FEMS Immunol Med Microbiol 2003; 38: 265 –71. 49. Luber P, Brynestad S, Topsch D et al. Quantification of Campylobacter species cross-contamination during handling of contaminated fresh chicken parts in chicken in kitchens. Appl Environ Microbiol 2006; 72: 66 –70. 50. National Antimicrobial Resistance Monitoring System. NARMS Retail Meat Annual Report, 2002. http://www.fda.gov/cvm/narms_pg. html (27 June 2007, date last accessed). 51. National Antimicrobial Resistance Monitoring System. NARMS Retail Meat Annual Report, 2004. http://www.fda.gov/cvm/narms_pg. html (27 June 2007, date last accessed).

721

Review 52. National Antimicrobial Resistance Monitoring System. NARMS Retail Meat Annual Report, 2003. http://www.fda.gov/cvm/narms_pg. html (27 June 2007, date last accessed). 53. Anonymous. Annual Report on Zoonoses in Denmark 2002. http://www.dfvf.dk/Default.aspx?ID=9606 (27 June 2007, date last accessed). 54. ESR for the Ministry of Health. New Zealand Public Health Surveillance Report, September, 2006. http://www.surv.esr.cri.nz/ PDF_surveillance/NZPHSR/2006/NZPHSRSept2006.pdf (27 June 2007, date last accessed). 55. Anonymous. Annual Report on Zoonoses in Denmark 2005. http://www.dfvf.dk/Default.aspx?ID=9606 (27 June 2007, date last accessed). 56. Centers for Disease Control and Prevention (CDC). 2005 FoodNet Annual Report. http://www.cdc.gov/foodnet/index.htm (27 June 2007, date last accessed). 57. Centers for Disease Control and Prevention (CDC). 2004 FoodNet Annual Report. http://www.cdc.gov/foodnet/index.htm (27 June 2007, date last accessed). 58. Banmali P, Fritschi L, Atrie J. Epidemiology of notified campylobacteriosis in Western Australia. Nepal Med Coll J 2006; 8: 31 –5. 59. Unicomb L, Ferguson J, Stafford J et al. Low-level fluoroquinolone resistance among Campylobacter jejuni isolates in Australia. Clin Infect Dis 2006; 42: 1368– 74. 60. Wang W-L, Reller L, Blaser M. Comparison of antimicrobial susceptibility patterns of Campylobacter jejuni and Campylobacter coli. Antimicrob Agents Chemother 1984; 26: 351– 3. 61. Huang M, Baker C, Banerjee S et al. Accuracy of the E test for determining antimicrobial susceptibilities of staphylococci, enterococci, Campylobacter jejuni, and Gram-negative bacteria resistant to antimicrobial agents. J Clin Microbiol 1992; 30: 3243–8. 62. Liebers D, Baltch A, Smith R et al. Comparative in-vitro activities of A-56268 (Te-031) and erythromycin against 306 clinical isolates. J Antimicrob Chemother 1988; 21: 565– 70. 63. Moore J, Barton M, Blair I et al. The epidemiology of antibiotic resistance in Campylobacter. Microbes Infect 2006; 8: 1955–66. 64. Vellinga A, Van Loock F. The dioxin crisis as experiment to determine poultry-related Campylobacter enteritis. Emerg Infect Dis 2002; 8: 19 –22. 65. Siemer B, Harrington C, Nielsen E et al. Genetic relatedness among Campylobacter jejuni serotyped isolates of diverse origin as determined by numerical analysis of amplified fragment length polymorphism (AFLP) profiles. J Appl Microbiol 2004; 96: 795–802. 66. Siemer B, Nielsen E, On S. Identification and molecular epidemiology of Campylobacter coli isolates from human gastroenteritis, food and animal sources by amplified fragment length polymorphism analysis and Penner serotyping. Appl Environ Microbiol 2005; 71: 1953–8. 67. Litrup E, Torpdahl M, Nielsen E. Multilocus sequence typing performed on Campylobacter coli isolates from humans, broilers, pigs and cattle originating in Denmark. J Appl Microbiol 2007; 103: 210–8. 68. Centers for Disease Control and Prevention (CDC). Outbreak of Campylobacter jejuni infections associated with drinking unpasteurized milk procured through a cow-leasing program - Wisconsin, 2001. Morb Mortal Wkly Rep 2002; 51: 548– 9. 69. Graham C, Whyte R, Gilpin B et al. Outbreak of campylobacteriosis following pre-cooked sausage consumption. Austral N Z J Pub Health 2005; 29: 507– 10. 70. Engberg J, Gerner-Smidt P, Schuetz F et al. Quinolone-resistant Campylobacter infections in Denmark: risk factors and clinical consequences. Emerg Infect Dis 2004; 10: 1056–63. 71. Sanders J, Isenbarger D, Walz S et al. An observational clinicbased study of diarrheal illness in deployed United States military

personnel in Thailand: presentation and outcome of Campylobacter infection. Am J Trop Med Hyg 2002; 67: 533–8. 72. Smith K, Besser J, Hedberg C et al. Quinolone-resistant Campylobacter jejuni infections in Minnesota, 1992–1998. N Engl J Med 1999; 340: 1525–32. 73. DuPont H, The Practice Parameters Committee of the American College of Gastroenterology. Guidelines on acute infectious diarrhea in adults. Am J Gastroenterol 1997; 92: 1962–75. 74. Guerrant R, Van Giler T, Steiner T et al. IDSA Guidelines. Practice guidelines for the management of infectious diarrhea. Clin Infect Dis 2001; 32: 331–51. 75. Salazar-Lindo E, Sack R, Chea-Woo E et al. Early treatment with erythromycin of Campylobacter jejuni-associated dysentery in children. J Pediatr 1986; 109: 355–60. 76. Anders B, Lauer B, Paisley J et al. Double-blind placebo controlled trial of erythromycin treatment of campylobacter enteritis. Lancet 1982; ii: 131–2. 77. Pai C, Gillis F, Tuomanen E et al. Erythromycin treatment of Campylobacter enteritis in children. Am J Dis Chil 1983; 137: 286–8. 78. Williams D, Schorling J, Barrett L et al. Early treatment of Campylobacter jejuni enteritis. Antimicrob Agents Chemother 1989; 33: 248–50. 79. Ternhag A, To¨rner A, Svensson A˚ et al. Mortality following Campylobacter infection: a registry-based linkage study. BMC Infect Dis 2005; 5: 70– 5. 80. Skirrow M, Jones D, Sutcliffe E et al. Campylobacter bacteraemia in England and Wales, 1981–1991. Epidemiol Infect 1993; 110: 567–73. 81. Schønheyder H, Søgaard P, Frederiksen W. A survey of Campylobacter bacteraemia in three Danish counties, 1989 to 1994. Scand J Infect Dis 1995; 27: 145–8. 82. Tee W, Mijch A. Campylobacter jejuni bacteremia in human immunodeficiency virus (HIV)-infected and non-HIV-infected patients: comparison of clinical features and review. Clin Infect Dis 1998; 21: 91 –6. 83. Pigrau C, Bartolome R, Almirante B et al. Bacteraemia due to Campylobacter species: clinical findings and antimicrobial susceptibility patterns. Clin Infect Dis 1997; 25: 1414–20. 84. World Health Organization. Critically Important Antibacterial Agents for Human Medicine for Risk Management Strategies of NonHuman Use: Report of a WHO Working Group Consultation, Canberra, Australia, February 15 –18, 2005. http://www.who.int/foodborne_ disease/resistance/en/ (27 June 2007, date last accessed). 85. Thielman N, Guerrant R. Acute infectious diarrhea. N Eng J Med 2004; 350: 38 –47. 86. White N, Dung N, Vihn H et al. Fluoroquinolone antibiotics in children with multidrug resistant typhoid. Lancet 1996; 348: 547. 87. Hughes R, Cornblath D. Guillain-Barre´ syndrome. Lancet 2005; 366: 1653–66. 88. Nachamkin I. Campylobacter jejuni infection and the association with Guillan-Barre´ syndrome. In: Nachamkin I, Blaser M, eds. Campylobacter. 2nd edn. Washington, DC: ASM Press, 2000; 155–75. 89. Locht H, Krogfelt K. Comparison of rheumatological and gastrointestinal symptoms after infection with Campylobacter jejuni/coli and enterotoxigenic Escherichia coli. Ann Rheum Dis 2002; 61: 448–52. 90. Neal K, Wright J. Reactive arthritis like symptoms following gastro-enteritis - an increased risk with severe Campylobacter infection. Int J Med Microbiol 2001; 291: 110. 91. Nelson J, Smith K, Vugia D et al. Prolonged diarrhea due to ciprofloxacin-resistant Campylobacter infection. J Infect Dis 2004; 190: 1150–7.

722

Review 92. Neal K, Slack R. Diabetes mellitus, anti-secretory drugs and other risk factors for Campylobacter gastro-enteritis in adults: a casecontrol study. Epidemiol Infect 1997; 119: 307– 11. 93. De Bruyn G. Infectious disease: diarrhea. Clin Evid 1999; 2: 267–73. 94. Neal K, Scott H, Slack R et al. Omeprazole as a risk factor for Campylobacter enteritis: case-control study. BMJ 1996; 312: 414–5. 95. Smith G, Blaser M. Fatalities associated with Campylobacter jejuni infections. J Am Med Assoc 1985; 253: 2873–5. 96. Figura N, Guglielmetti P. Clinical characteristics of Campylobacter jejuni and C. coli enteritis. Lancet 1988; i: 942–3. 97. Popovic-Uroic T, Gmajnicki B, Kalenic S et al. Clinical comparison of Campylobacter jejuni and C. coli diarrhea. Lancet 1988; i: 176–7. 98. Fouts D, Mongodin E, Mandrell R et al. Major structural differences and novel potential virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol 2005; 3: 72– 85. 99. EE Leonard I, Thompkins L, Falkow S et al. Comparison of Campylobacter jejuni isolates implicated in Guillain-Barre´ syndrome and those that cause enteritis by DNA microarray. Infect Immun 2004; 72: 1199–203. 100. Organisation Internationale Epizooties. Guidelines for the Responsible and Prudent Use of Antimicrobial Agents in Veterinary Medicine, OIE Terrestrial Animal Health Code 2006, Part 3, Section 3.9, Chapter 3.9.3. http://www.oie.int/eng/normes/mcode/en_chapitre_3.9. 3.htm) (27 June 2007, date last accessed). 101. World Health Organization, Food and Agriculture Organization of the United Nations, and World Organization for Animal Health. Background Document for the Joint WHO/FAO/OIE Expert Workshop on Non-Human Antimicrobial Usage and Antimicrobial Resistance: Scientific Assessment, Geneva, Switzerland, December 1–5, 2003. http://www.who.int/salmsurv/links/RelWHO/en/ (27 June 2007, date last accessed). 102. DANMAP 1997. Consumption of Antimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria From Food Animals, Food and Humans in Denmark. http://www.danmap.org/ (27 June 2007, date last accessed). 103. DANMAP 1998. Consumption of Antimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria From Food Animals, Food and Humans in Denmark. http://www.danmap.org/ (27 June 2007, date last accessed). 104. DANMAP 1999. Consumption of Antimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria From Food Animals, Food and Humans in Denmark. http://www.danmap.org/ (27 June 2007, date last accessed). 105. DANMAP 2000. Consumption of Antimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria From Food Animals, Foods and Humans in Denmark. http://www.danmap.org/ (27 June 2007, date last accessed). 106. DANMAP 2001. Use of Antimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria From Food Animals, Foods and Humans in Denmark. http://www.danmap.org/ (27 June 2007, date last accessed). 107. DANMAP 2002. Use of Antimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria From Food Animals, Foods and

Humans in Denmark. http://www.danmap.org/ (27 June 2007, date last accessed). 108. DANMAP 2003. Use of Antimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria From Food Animals, Foods and Humans in Denmark. http://www.danmap.org/ (27 June 2007, date last accessed). 109. DANMAP 2004. Use of Antimicrobial Agents and Occurrence of Antimicrobial Resistance in Bacteria From Food Animals, Foods and Humans in Denmark. http://www.danmap.org/ (27 June 2007, date last accessed). 110. Englen M, Fedorka-Cray P, Ladely S et al. Antimicrobial resistance patterns of Campylobacter from feedlot cattle. J Appl Microbiol 2005; 99: 285–91. 111. United States Department of Agriculture – National Antimicrobial Resistance Monitoring System (USDA-NARMS). Campylobacter 2003 Report. http://www.ars.usda.gov/Main/docs. htm?docid=6750 (27 June 2007, date last accessed). 112. Gupta A, Nelson J, Barrett T et al. Antimicrobial resistance among Campylobacter strains, United States, 1997–2001. Emerg Infect Dis 2004; 10: 1102–9. 113. Centers for Disease Control and Prevention (CDC)–National Antimicrobial Resistance Monitoring System: Enteric Bacteria (CDC-NARMS). 2003 Human Isolates, Final Report. http://www.cdc. gov/narms/annual/2003/NARMS2003AnnualReport.pdf (27 June 2007, date last accessed). 114. United States Department of Agriculture - Collaboration in Animal Health and Food Safety Epidemiology (USDA-CAHFSE). CAHFSE Annual Report, July 2003–June 2004. http://www. aphis.usda.gov/cahfse/results/index.htm (27 June 2007, date last accessed). 115. United States Department of Agriculture - Collaboration in Animal Health and Food Safety Epidemiology (USDA-CAHFSE). CAHFSE Annual Report, July 2004–July 2005. http://www.aphis. usda.gov/cahfse/results/index.htm (27 June 2007, date last accessed). 116. Anonymous. Annual Report on Zoonoses in Denmark 1997. http://www.dfvf.dk/Default.aspx?ID=9606 (27 June 2007, date last accessed). 117. Anonymous. Annual Report on Zoonoses in Denmark 2003. http://www.dfvf.dk/Default.aspx?ID=9606 (27 June 2007, date last accessed). 118. Nielsen E, Fussing V, Engberg J et al. Most Campylobacter subtypes from sporadic infections can be found in retail poultry products and food animals. Epidemiol Infect 2005; 134: 759–67. 119. Gupta A, Nelson JM, Barrett TJ et al. Antimicrobial resistance among Campylobacter strains, United States, 1997–2001. Emerg Infect Dis 2004; 10: 1102–9. 120. Zhao C, Ge B, De Villena J et al. Prevalence of Campylobacter spp., Escherichia coli, and Salmonella serovars in retail chicken, turkey, pork, and beef from the Greater Washington, D.C., area. Appl Environ Microbiol 2001; 67: 5431–6. 121. Duffy EA, Belk KE, Sofos JN et al. Extent of microbial contamination in United States pork retail products. J Food Prot 2001; 64: 172 –8.

723