A Prospective Follow-Up Study on Transmission of Campylobacter from Poultry to Abattoir Workers

Introduction

C ampylobacter infection is the most commonly reported zoonosis in the European Union and the most frequent cause of bacterial gastroenteritis in the Western world, with symptoms ranging from mild watery diarrhea to bloody stools, fever, and abdominal pain (EFSA/ECDC, 2013). Possible postinfectious complications include the severe neurological disorder Guillain-Barré syndrome, which has been associated with strains that have the ability to incorporate sialic acids in the lipooligosaccharide (LOS) of the cell wall (Godschalk et al., 2004). Such strains have also been proposed to be associated with increased invasivity and more severe enteritis, but the results are conflicting (Mortensen et al., 2009; Louwen et al., 2012; Ellström et al., 2013a, b). As shown by case–control studies and epidemiological surveys using molecular typing methods, poultry meat is an important source of campylobacteriosis (Schönberg-Norio et al., 2004; Jore et al., 2010). Genotypes of human isolates have been shown to overlap with those of poultry origin by 30–60% in comparative studies (Lindmark et al., 2004; de Haan et al., 2010). However, it is not clear whether chicken may also carry Campylobacter isolates that are less pathogenic for humans or if all chicken Campylobacter isolates can be transmitted to humans and cause disease. The prevalence of Campylobacter among chicken flocks in Sweden shows a considerable seasonal variation, with a peak of 25–40% positive flocks during the summer months, June to September, and only 5–10% during the rest of the year (Jore et al., 2010). The incidence of domestic human infections essentially displays the same seasonal variation (Jore et al., 2010). We studied the transmission of Campylobacter from chicken flocks to the personnel at two abattoirs during the peak season for Campylobacter infections in Sweden, with emphasis on the characteristics of the abattoir workers and the Campylobacter isolates.

Materials and Methods

Results

In total, 28 workers at 2 poultry abattoirs in Sweden were included; 17 women (mean age 33.9 years) and 11 men (mean age, 32.2 years). Twelve participants had been employed ≤1 year (median, 1.3 months) at the beginning of the study (Table 1). All human stool cultures were Campylobacter negative at the beginning of the study in June, but follow-up samples revealed positive Campylobacter stool cultures from a total of seven abattoir workers: five of them in September (Table 1). Importantly, no symptoms were reported by any of the Campylobacter-positive participants at the time of positive stool culture. Four of the Campylobacter-positive participants had been employed within a year from the beginning of the study. These individuals had not reported any previous employment at a poultry abattoir or any previously documented Campylobacter infection. Two of the five participants with earlier confirmed Campylobacter infection also became culture positive during the study.

Five of the seven human Campylobacter isolates were identified as C. jejuni and one as C. lari. One isolate was lost as it could not be resuscitated from the frozen stock. The five human C. jejuni isolates were further identified by MLST to belong to four different clonal complexes: ST-48CC, ST-658CC, ST-677CC, and ST-45CC (two isolates) (Fig. 1). Four of the human isolates—three C. jejuni and one C. lari—had a SmaI pattern identical to that of chicken isolates from flocks slaughtered within 1 month (Fig. 1). For two positive abattoir workers, no chicken isolates with matching PFGE patterns were found at the corresponding abattoir even within 2 months of infection.

OPEN IN VIEWER

FIG. 1.

SmaI restriction patterns after pulsed-field gel electrophoresis (PFGE) for the six human Campylobacter isolates and the corresponding chicken isolates with identical patterns. The column “Isolation date” denotes in which month each positive human sample was obtained as well as how many days earlier the corresponding chicken isolates with matching PFGE patterns were obtained. *Isolates from different flocks of the same breeder. LOS class, lipooligosaccharide locus class (range from A to S); MLST, multilocus sequence typing; ST, sequence type (specific combination of sequence data from seven housekeeping genes in the C. jejuni genome); CC, clonal complex (related STs with at least 4 identical alleles); 18df, the isolate harbored the open reading frame orf18 and thus had any of the LOS locus classes D, F, I, J, K, N, S, or Q which all are unable to incorporate sialic acids; ND, not determined.

To determine whether any specific LOS locus class was over-represented among the isolates only causing asymptomatic infections, the LOS locus class was determined for the five human C. jejuni isolates and the corresponding chicken isolates with matching PFGE patterns. The isolates belonged to four different LOS locus classes and hence there was no over-representation of a specific LOS locus class (Fig. 1). In this study, none of the host characteristics collected in the questionnaires was significantly associated with positive stool cultures.

Discussion

Poultry is an important source of human Campylobacter infections. During this prospective follow-up study on 28 poultry abattoir workers, 7 participants became stool culture– positive for Campylobacter by routine diagnostic methods. Four of the human isolates had PFGE SmaI restriction patterns identical to isolates from chickens slaughtered at the corresponding abattoir within a month of human infection, providing evidence for transmission of Campylobacter from chicken to the personnel at the abattoirs. Immunity developed after years of employment at chicken abattoir might explain the asymptomatic infections of the long-term abattoir workers. However, it is noteworthy that four of the Campylobacter-positive asymptomatic individuals had most probably only had very limited earlier exposure to Campylobacter: short duration of employment at any poultry abattoir, employment before the study mostly during the season with only few contaminated chicken flocks, and no previously verified Campylobacter infection.

Campylobacter infection has been reported previously among poultry abattoir workers (Christenson et al., 1983; Cawthraw et al., 2000; de Perio et al., 2013). However, in all these studies asymptomatic Campylobacter infection has been exclusively reported among experienced workers, whereas all newly employed Campylobacter-positive staff members have reported symptoms of the infection. In a study of a Campylobacter outbreak at an abattoir in Sweden, 17 of 24 newly employed workers were infected together with 20 of 70 experienced workers. Only five asymptomatic Campylobacter-positive individuals were detected; all of them were among the experienced staff members (Christenson et al., 1983). In another study, eight Campylobacter culture-positive abattoir workers were identified and one of them, the only short-term worker, developed symptoms (Cawthraw et al., 2000). Furthermore, in a register-based study of Campylobacter-positive symptomatic poultry-processing workers at an abattoir in the United States between 2008 and 2011, 83% of the infected individuals had worked at the plant for <1 month before the infection (de Perio et al., 2013). Together, these studies indicate that continuous bacterial exposure may protect to some extent the experienced workers from symptomatic Campylobacter infection. However, our findings in the present prospective study, although with a low number of participants, show that asymptomatic infection may also occur among employees with only limited earlier exposure to Campylobacter and thus could further suggest that all Campylobacter isolates of chicken origin might not be equally virulent and cause symptomatic infection in humans. The C. jejuni isolates from the newly employed participants in this study belonged to clonal complexes previously found among symptomatic humans, and it is possible that host factors may explain the lack of symptoms even among these individuals. On the other hand, virulence determinants and phenotypes have been shown to vary even within clonal complexes (Revez et al., 2011).

One suggested virulence trait in the pathogenesis of Guillain-Barré syndrome is the ability of some C. jejuni strains to incorporate sialic acids in the LOS, resulting in structures mimicking human gangliosides (Godschalk et al., 2004). Such strains also displayed increased invasion of epithelial cells in vitro, but results are conflicting regarding their ability to cause more severe symptoms in patients (Louwen et al., 2012; Ellström et al., 2013a). Screening for LOS locus class of the C. jejuni isolates in the present study did not reveal any obvious association between a certain LOS locus class and the asymptomatic infections among the abattoir workers. However, the numbers are too small to draw any further conclusions.

Conclusions

This prospective follow-up study further supports the role of chicken as the source of human Campylobacter infection but suggests that asymptomatic Campylobacter infection might occur even in individuals with only limited earlier exposure to Campylobacter.