Fluoroquinolone-resistant Campylobacter spp. is recognized as an emerging public health problem [1]. Fluoroquinolone resistance in Campylobacter spp. is increasing in many countries throughout the world [Reference Engberg, Aarestrup, Taylor, Gerner-Smidt and Nachamkin2–Reference Velázquez, Jiménez, Chomón and Villa5]. In several countries, an association between the usage of fluoroquinolones in food animals and the occurrence of fluoroquinolone-resistant Campylobacter from humans has been documented [Reference Engberg, Aarestrup, Taylor, Gerner-Smidt and Nachamkin2]. Fluoroquinolones are commonly used in poultry medicine both for therapeutic and prophylactic purposes. Poultry meat is considered as the most important risk factor for campylobacteriosis in humans [Reference Smith, Besser, Hedberg, Leano, Bender and Wicklund4, Reference Friedman, Neimann, Wegener, Tauxe, Nachamkin and Blaser6, 7]. Thus, an extended use of fluoroquinolones in poultry production thereby increases the risk of transmission of fluoroquinolone-resistant Campylobacter spp. to humans. Antimicrobial agents are normally not indicated for treatment of campylobacteriosis in humans. However, severe cases may require treatment. An increased resistance to fluoroquinolones reduces the possibilities to treat severe infections in humans, which can have fatal consequences [1, Reference Helms, Vastrup, Gerner-Smidt and Molbak8].
Norway has a long tradition for a restrictive antimicrobial policy in animal husbandry, and fluoroquinolones were never licensed for use in poultry. In 2000, the Norwegian government established an action plan against antimicrobial resistance, which resulted in the establishment of monitoring programmes for antimicrobial-resistant bacteria from both humans (NORM) and animals and food (NORM-VET). There is close cooperation between these programmes and a joint report (NORM/NORM-VET) is published annually [9] to describe the current situation and to evaluate trends. Monitoring of antimicrobial resistance in zoonotic bacteria such as Salmonella spp. and Campylobacter spp. is included in both programmes.
Campylobacteriosis is currently the most frequently reported cause of bacterial gastroenteritis in Norway. Most cases are sporadic and close to half of the reported cases acquire the infection in Norway. Of the 2192 cases of human campylobacteriosis recorded in Norway in 2002 (incidence rate 48·5/100 000), 52% were reported as acquired abroad [Reference Hofshagen, Aavitsland and Kruse10]. A case-control study conducted in Norway during 1999–2000 identified consumption of untreated drinking water, consumption of poultry meat purchased raw, consumption of barbecued meat, and professional contact with animals as significant risk factors in regard to campylobacteriosis [Reference Kapperud, Espeland and Wahl11]. To reduce the human exposure to Campylobacter through Norwegian broiler meat products, an action plan against Campylobacter in Norwegian broilers was implemented in the spring of 2001 [9]. The present study assesses and compares the prevalence of resistance in C. jejuni isolates from imported and indigenous human cases and from Norwegian broilers using data from NORM/NORM-VET for the period 2001–2003.
The isolates of C. jejuni in broilers originate from the Norwegian action plan against Campylobacter in Norwegian broilers. All broiler flocks slaughtered before 50 days of age are tested for the presence of Campylobacter spp. and 100 samples of broiler meat products from retail level are tested monthly. Annually, one isolate per positive farm as well as one isolate from each batch of positive broiler meat products are submitted for susceptibility testing in NORM-VET. In the period 2001–2003, a total of 413 broiler isolates, 343 from cloacal samples and 70 from broiler meat products, were tested for antimicrobial susceptibility. The human isolates included were obtained from clinical specimens as part of the NORM programme. Each of the five regional NORM laboratories submit isolates from the first five patients with campylobacteriosis for susceptibility testing to the National Reference Laboratory for Enteropathogenic Bacteria at the Norwegian Institute of Public Health each month. In the present study, only isolates of C. jejuni where the source of infection was known (indigenous or abroad) were included. This resulted in 567 human isolates, 184 from patients infected in Norway and 383 from patients infected abroad. This skewed sample, twice as many isolates from imported cases from as indigenous cases, can be explained by the sampling strategy since indigenous campylobacteriosis is subject to seasonal fluctuation with an increased number of cases during the summer season and only few infections during the rest of the year [Reference Sandberg12].
Broiler isolates were isolated and identified according to the Nordic Committee on Food Analyses (NMKL) method number 119 with minor modifications, and human isolates according to conventional methods described in standard reference literature. Broiler isolates were tested at the National Veterinary Institute. Minimum inhibitory concentration (MIC) values for ampicillin, erythromycin, gentamicin, nalidixic acid and enrofloxacin were obtained using the VetMIC™ panels produced at Department of Antibiotics, National Veterinary Institute, Sweden. This is a broth dilution method which is based upon standard methods described by NCCLS and adapted for Campylobacter spp. MIC was recorded as the lowest concentration of the antimicrobial that inhibits bacterial growth. Human isolates were tested for susceptibility to doxycycline, erythromycin, gentamicin, nalidixic acid and enrofloxacin using E tests (AB Biodisk, Solna, Sweden). At the present stage, there are no internationally accepted criteria for interpretation of susceptibility in Campylobacter spp. and subsequently no break-points. The isolates were classified as resistant or susceptible applying microbiological cut-off values. Microbiological cut-off values are determined on the basis of the distribution of MICs for each antimicrobial and bacterial species. The population that clearly departs from the normal genetically unchanged population (‘wild-type’) are defined as resistant (‘non-wild-type’).
Data were analysed using a program developed by the World Health Organization for analysis of antimicrobial resistance data [13]. The results revealed that the prevalence of resistance among C. jejuni isolates from Norwegian broilers is low (Table). There was no significant difference in resistance prevalences between the isolates from broiler meat products and cloacal samples. A total of 91·2% of the isolates from broilers were susceptible to all antimicrobials included in the test panel. None of the isolates were resistant to more than two antimicrobials, and only 1·2% of were fluoroquinolone resistant. Compared to the prevalence of resistance among poultry in other countries this figure 1 is low [Reference Wilson14]. The low prevalence of fluoroquinolone resistance in isolates from broilers most probably reflects the restrictive use of antimicrobials in Norwegian broiler production. Antimicrobials (except coccidiostats) are rarely used, and only for therapeutic purposes. If used, amoxicillin and tetracycline are the drugs of choice. No quinolone preparations are licensed for use in broilers in Norway. However, several fluoroquinolones are approved for use in broilers in the EU and may, therefore, also be used in broiler production in Norway if specifically applied for. The resistance situation among Campylobacter spp. from Norwegian broilers probably reflects the resistance situation among Campylobacter spp. in the Norwegian environment [Reference Hofshagen, Aavitsland and Kruse10, Reference Kapperud, Espeland and Wahl11] as broilers typically are infected from environmental sources, including untreated surface water.
Percentage of isolates of Campylobacter jejuni from various sources and years classified as resistant to various antimicrobials
* Values above the cut-off value indicate resistance.
† Oxytetracycline for broilers.
‡ Enrofloxacin for broilers in 2002 and 2003.
The data show that resistance was significantly more prevalent among the C. jejuni isolates derived from patients infected abroad (23·0% susceptible to all antimicrobials included) than patients infected in Norway (85·3% susceptible to all antimicrobials included). These discrepancies are explained by the widespread prevalence among isolates acquired abroad of resistance to ciprofloxacin/nalidixic acid (67·4/68·1% vs. 6·5/9·2%) and to tetracycline (53·8% vs. 8·2%). The resistance prevalences for domestically acquired human cases correlate quite well with the data for Norwegian broilers, although resistance to quinolones, particularly nalidixic acid, was more prevalent among the human isolates. The discrepancy between broiler isolates and indigenous human isolates might be explained by sporadic cases of secondary human infection with imported Campylobacter strains and infection with Campylobacter spp. from imported poultry. The official import of chicken is relatively limited, but there is some import of other types of poultry products such as ducks or pheasants from the EU (data provided from The Norwegian Food Safety Authority).
The resistance prevalences were stable and low throughout the study period both for isolates from humans infected in Norway as well as for isolates from broilers, whereas the resistance prevalences in isolates from humans infected abroad increased slightly. This is illustrated by the trend for fluoroquinolone resistance during the same period (Fig.). In a study from Northern Ireland, isolates from human cases showed a greater similarity with indigenous broiler isolates than with isolates from imported chicken [Reference Wilson14]. However, the source of infection for the human cases was not presented.
Proportion (%) of isolates of Campylobacter jejuni from various sources and years resistant to fluoroquinolones. □, 2001; 
, 2002; ▪, 2003.
In many countries, Campylobacter spp. has become the most frequently reported cause of bacterial gastrointestinal illness [Reference Friedman, Neimann, Wegener, Tauxe, Nachamkin and Blaser6, 7, 15]. Some countries report that many cases of campylobacteriosis are associated with foreign travel [15]. The humans infected abroad in this study had to a large extent been visiting Southern European countries and Asia. Recent studies have shown that the prevalence of fluoroquinolone resistance among isolates of C. jejuni from humans infected when travelling to Southern European countries is high [Reference Engberg, Aarestrup, Taylor, Gerner-Smidt and Nachamkin2]. The prevalence of fluoroquinolone resistance in Campylobacter is also high in Thailand [Reference Hoge, Gambel, Srijan, Pitarangsi and Echeverria16, Reference Bodhiadatta, Vithayasai, Eimpokalarp, Pitaringsi, Serichantelergs and Isenbarger17].
The present study documents that a limited use of fluoroquinolones in food animal production within a country is associated with a low prevalence of fluoroquinolone resistance in indigenous human C. jejuni isolates. Thus, the restricted use of fluoroquinolones in food-producing animals is recommended in order to keep such antimicrobials efficient for the treatment of severe cases of campylobacteriosis in humans.
