16 Campylobacteriosis

Campylobacter jejuni and C. coli are frequent causes of bacterial enteritis in industrialized countries and are a major cause of childhood illness in the developing world. Although deaths due to campylobacteriosis are rare, the morbidity and public health and economic burden is high because of its very high incidence. Campylobacters normally inhabit the intestinal tract of wild birds and domestic animals. Poultry is a major source of campylobacter infection and a large proportion of retail chicken meat is contaminated. Other meats are contaminated to a lesser degree. Human infection is mostly sporadic and outbreaks are uncommon. Infections arise from the consumption of raw or inadequately cooked meat or from other foods contaminated during production or preparation. Contaminated water and raw milk can also act as vehicles of campylobacter infection and have given rise to significant outbreaks. The most effective means of controlling human campylobacteriosis would be the implementation of measures to reduce the contamination of food producing animals during slaughter and processing. Public health education regarding the principles of hygiene and safe food handling are also important.

Cases of campylobacteriosis were probably first observed as early as 1886 by Theodore Escherich who reported non-cultivable spiral bacteria in diarrhoeic stool specimens from kittens and human infants. However, it was not until 1906 that McFadyean and Stockman isolated ‘microaerobic vibrio-like bacteria’, which were later named ‘Vibrio fetus’, from an aborted sheep foetus. Subsequently ‘Vibrio jejuni’ was isolated from calves with ‘winter dysentery’ and ‘Vibrio coli’ from pigs suffering from swine dysentery, although it is now known they are part of the normal microbial flora in these animals. These ‘vibrio-like’ bacteria were eventually assigned to a new genus Campylobacter that originally consisted of Campylobacter fetus (the type species), C. coli, C. jejuni and C. sputorum.

Campylobacter were first implicated as agents of human enteritis by Levy in 1938 during an outbreak of milk-borne diarrhoea in two adjacent USA prisons. At this time campylobacters could not be grown from human faeces, but several cases developed bacteraemia which led to the culture of a ‘vibrio’ that was most likely C. jejuni. In the 1950s King made the connection between ‘microaerobic vibrios’ isolated from cases of bacteraemia in humans and isolates of ‘V. jejuni’ from poultry. She also observed that the human isolates from blood culture were all from patients with symptoms of acute diarrhoea. This work was followed up by Butzler who used filtration-based isolation techniques borrowed from veterinary microbiology to demonstrate that campylobacters were frequently present in the faeces of children with diarrhoea in the Congo (Butzler et al.1973). Butzler’s work was expanded upon by Skirrow working in the UK, who developed an antibiotic-containing selective culture medium for C. jejuni and C. coli (Skirrow 1977). The paper describing Skirrow’s work was a major turning point: it described the clinical entity, suggested appropriate antibiotic therapy and reported an easy culture method which resulted in a high isolation rate from diarrhoeal cases. It further demonstrated that there was a serological response to the infection, and confirmed the previously suspected link between disease in humans and the presence of campylobacter in poultry. Subsequently, throughout the late 1970s and 1980s, laboratories began to culture for campylobacter and the isolation rate increased dramatically as cultivation techniques were refined. In the UK the number of campylobacter isolates exceeded those of salmonella in 1981 and continued to rise reaching a peak of 58, 236 cases in 2000. In the UK and elsewhere the number of isolates reached a plateau in the late 1990s and since the mid 2000s isolate numbers have begun to decline. Nevertheless, campylobacteriosis remains the major cause of bacterial enteritis and a major public health burden worldwide.

The genus Campylobacter together with the genera Arcobacter and Sulfurospirillum constitute the family Campylobacteraceae within the epsilon subdivision of the proteobacteria. This subdivision also contains the closely related genus Helicobacter, which includes H. pylori a major human pathogen associated with gastritis and stomach ulcers that was originally assigned to the genus Campylobacter. At present the Campylobacter genus comprises 21 species and 8 subspecies, most of which are inhabitants of the gastrointestinal tracts of man and animals.

Campylobacter are the major cause of acute bacterial enteritis in humans worldwide. The vast majority of infections are caused by C. jejuni and C. coli, which typically account for 90% and 10%, respectively, of cases. C. jejuni comprises two subspecies: doylei and jejuni. Subspecies doylei was originally isolated from the stomach of a patient with gastritis and subsequently from diarrhoeal stools from children living in poor social conditions. It is distinct from subspecies jejuni phenotypically and its clinical significance is uncertain. Subspecies jejuni accounts for nearly all the isolates of the species and is referred to as C. jejuni throughout this chapter.

Other Campylobacter species have the potential to be zoonotic pathogens but occur comparatively rarely and account for less than 0.5% of cases of campylobacter enteritis in humans in industrialized countries (Lawson et al. 1999): C. upsaliensis is associated with gastroenteritis in cats and dogs and is occasionally reported in cases of gastroenteritis from humans, particularly in children. C. lari was first isolated from gulls and subsequently from other animals and has been occasionally isolated from humans with gastroenteritis; C. hyointestinalis, originally isolated from pigs is also occasionally isolated from cases of diarrhoea in humans. C. fetus is usually found in the intestinal and genital tracts of cattle and sheep and accounts for less than 0.05% of campylobacter isolates from human faeces. Nevertheless, after C. jejuni, it is the second most common campylobacter associated with bacteraemia accounting for ∼25% of isolates from blood culture. Most human infection is caused by C. fetus subspecies fetus and cases are typically elderly patients who often have an underlying condition or immunosuppression. The remainder of this chapter will focus on C. jejuni and C. coli.

Campylobacter are small, non-spore forming, Gram-negative bacilli typically 0.5 to 5.0μm long and 0.2 to 0.9μm wide, with a spiral curve and tapering ends. Cells usually possess a polar flagellum at one or both ends and are highly motile. On Gram stain campylobacter cells can appear curved, S-shaped or spiral and may become coccoid in old cultures or on prolonged exposure to air. Campylobacter are microaerobic and are unable to tolerate oxygen at atmospheric levels. For optimum growth they require an atmosphere with an oxygen concentration of between 3 to 5% and a carbon dioxide concentration of 5 to 15%. The addition of hydrogen at 3 to 5% may improve growth of some strains and is a requirement for some species. Campylobacter grow well at 37°C, but certain species, such as C. jejuni and C. coli, have an optimum growth temperature of 42°C and are often termed ‘thermophilic campylobacter’. Campylobacter are susceptible to pasteurization, chlorination, and other forms of disinfection. Although they are generally regarded as fastidious bacteria they are more resistant to stress and adverse conditions than previously thought. They exhibit a heat shock response and are able to survive and respire at temperatures as low as 4°C though viability is rapidly lost. The coccoid form of the bacteria has been associated with the so called ‘viable non-culturable’ (VNC) state, which may represent a dormant form of campylobacter capable of survival in hostile environments. However, the VNC concept remains controversial and difficult to evaluate.

The genome of C. jejuni NCTC 11 168 was sequenced in 2000 (Parkhill et al. 2000): it is 1.64 million nucleotide bases long and encodes 1,643 genes with an average G+C ratio of 30.6%. The genome is relatively small compared with other bacteria (for example Salmonella enterica has a genome of 4.68 million bases) and unlike most other sequenced bacterial genomes there is an almost complete absence of repetitive DNA such as insertion sequences and transposons. It had been estimated that approximately 94.3% of the genome encodes for proteins, a much higher proportion than that found in other bacteria. A significant proportion of this small, compact genome is given over to encoding carbohydrate surface structures (8% in NCTC 11168). These include the lipooligosaccharide (LOS) locus; the capsule locus; and the O-linked and N-linked glycosylation systems (Gundogdu et al. 2007) and presumably these play a major role in interactions between the cell and its host/environment. An interesting feature of the C. jejuni genome is the presence of intragenic homopolymeric tracts, which take the form of variable mononucleotide runs of 8 or more bases. These are thought to provide a slipped strand miss-pairing mechanism for phase variation of key genes such as the capsule, LOS and flagella glycosylation loci leading to a potentially high degree of diversity in these immunogenic structures.

Many campylobacter strains are capable of natural transformation: that is they are naturally able to incorporate homologous DNA from other strains without any specific genetic exchange mechanism. This leads to a high level of genetic exchange and campylobacter exhibit a remarkable degree of genomic plasticity. This has major consequences for the epidemiological study of these bacteria as they exhibit a ‘weakly clonal’ population structure. Here the normal clonal expansion of a strain, where genetic information is passed vertically from one generation to the next, is confused by the horizontal exchange of genetic material with other, often distantly related strains.

C. jejuni possesses a polysaccharide capsule, the exact function of which is unclear. It may contribute to pathogenicity, serum resistance, and survival in the environment. The capsular polysaccharide is the principal component of the Penner serotyping scheme (although epitopes from the LOS and flagella are also involved to some extent), which had long been assumed to be based on membrane-associated lipopolysaccharide (LPS) (Karlyshev et al. 2000). The genes encoding the capsule consist of a highly variable central region flanked by relatively conserved regions. Between different strains there appears to be a great deal of variation by means of insertion and deletion of genes in the central region (Karlyshev et al. 2005). The outer surface of the campylobacter cell membrane lipid bilayer is composed of LOS rather than the LPS found in other bacteria. Campylobacter LOS consists of two covalently linked domains: lipid A, a hydrophobic anchor, and a non-repeating core oligosaccharide, consisting of an inner and outer core region. The repeating O-chain of polysaccharide that characterizes LPS is absent. The genes encoding the biosynthesis of the outer core region vary in number and content between different strains. Campylobacter are capable of producing glycoproteins. Until recently it was presumed that only eukaryotic organisms were able to do this. C. jejuni possesses two distinct glycosylation pathways: an O-linked glycosylation pathway responsible of modification of the flagella and an N-linked general glycosylation system. The O-linked system is highly variable with multiple copies of some genes present in the locus suggesting that it has potential for generating structural diversity in the flagellin protein. In contrast the N-linked glycosylation is highly conserved and in C. jejuni it has been shown to modify upwards of thirty proteins. The role that these N-linked glycoproteins play in the pathogenesis of campylobacter infection is unclear, but disruption of the locus results in cells with a reduced capacity to invade and colonize.

The diarrhoeal illness caused by campylobacter is clinically indistinguishable from other forms of gastroenteritis and laboratory diagnosis is required to determine the nature of infection. Campylobacter can be readily cultured from faecal samples, which should be promptly delivered to the laboratory and examined within three days. Campylobacter are cultured on a selective media, such as modified Charcoal Cefoperazone Desoxycholate Agar (mCCDA) and incubated in micro-aerobic conditions at 42°C or 37°C for 48 hours. Campylobacter are identified by their colonial morphology, a positive oxidase test, and their appearance on Gram stain. For many laboratories this level of identification is deemed sufficient and isolates are reported as ‘Campylobacter species’ with no attempt made to distinguish between species. C. jejuni can be readily identified by a positive hippurate hydrolysis test. Definitive identification of C. coli is more problematic, but most hippurate-negative, catalase-positive; indoxyl acetate-positive cultures isolated on selective media will be C. coli. Unfortunately, no single phenotypic test is completely infallible: hippurate-negative C. jejuni are known and many uncommon campylobacter and campylobacter-like species are phenotypically similar to C. coli.

Techniques for the detection and enumeration of campylobacter from food have been defined by the International Organization for Standardization and are described in the documents ISO 10272: Horizontal method for detection and enumeration of Campylobacter species Parts 1 and 2. Briefly, these describe techniques for isolation and identification of Campylobacter species; recovery of low numbers using an enrichment broth; and enumeration by dilution and plate count. These standardized techniques have been used to establish baseline levels of campylobacter contamination of chicken carcases in the European Union.

Accurate and timely identification of Campylobacter species can be achieved using PCR assays. Many assays have been developed, but there has been no systematic review of their accuracy and reliability. Although generally reliable PCR assays can be susceptible to the genomic plasticity of campylobacter and the exchange of genes between closely related species such as C. jejuni and C. coli is not uncommon. Genes that are under evolutionary or selective pressure, such as those involved in antimicrobial resistance, should be avoided as targets for PCR assays. PCR assays may also be applied to the direct detection of campylobacter from faecal samples and achieves a similar level of sensitivity to routine culture with the advantages of speed, amenability for automation, and definitive identification (Lawson et al. 1999). PCR is not widely used in the routine clinical laboratories, where these advantages are outweighed by cost. However, in the food industry, where rapid results are at a premium, the latest generation of real-time PCR identification assays are becoming more widely used.

The typing of C. jejuni and C. coli for epidemiological purposes has proven problematic. The genetic diversity and weakly clonal population structure of campylobacter does not lend itself to conventional typing approaches. Initial attempts at typing schemes were phenotypically-based and included serotyping, phage typing, and biotyping. Of these the original serotyping schemes developed by Lior and Penner were probably the most widely used, primarily in reference laboratories. However, the performance of these schemes was poor as common serotypes were widely distributed and an unacceptably large proportion of isolates were untypable.

Theoretically DNA-based molecular typing techniques provide 100% typability and many methodologies have been developed for campylobacter including: ribotyping, Restriction Fragment Length Polymorphism (RFLP), Amplified Fragment Length Polymorphism (AFLP) and Random Amplification of Polymorphic DNA (RAPD) analysis. However, these techniques are not widely used, though they may be of use in local outbreak investigations. The most widely used molecular techniques are Pulsed Field Gel Electrophoresis (PFGE) and Multi Locus Sequence Typing (MLST).

PFGE involves the digestion of chromosomal DNA using bacterial endonucleases, which produces characteristic restriction fragment patterns when run out on an agarose gel. Many variations of PFGE exist for campylobacter, using different restriction enzymes and electrophoresis conditions and despite attempts at standardization by CampyNet in Europe and PulseNet in the USA comparisons between results from different centres are difficult. PFGE is vulnerable to genetic rearrangement which occurs relatively frequently in campylobacter and may result in major changes in PFGE profile between closely related strains.

An MLST-based typing scheme was first developed for C. jejuni in 2001 (Dingle et al. 2001) and subsequently schemes have been published for other Campylobacter species. The technique involves sequencing seven house-keeping genes (loci) and assigning numbers for each new sequence for each allele. The seven allelic numbers from each locus combined together for a sequence type (ST) and related STs that differ by a few base pairs grouped together in clonal complexes (CC). For C. jejuni basic MLST provides a comparable degree of resolution to PFGE. Greater resolution can be provided if additional, more variable loci are sequenced (Dingle et al. 2008). MLST sequence data are unambiguous and easily comparable with an internet-based database. MLST data provides a measure of the degree of relatedness between two strains rather than simply determining if they are the same or different and can thus be adapted to population genetics studies. As the cost of DNA sequencing has decreased over the last decade, so MLST-based typing schemes have become more widely used. For the first time truly global comparisons of campylobacter isolates from different temporal, geographic and ecological niches can be made. However, the high degree of genetic variability and weakly clonal population structure of C. jejuni and C. coli limit the usefulness of any typing scheme and MLST is no exception. MLST-based studies of C. jejuni typically show half as many genotypes as there are strains in the study with relatively few common types accounting for a significant proportion of the isolates (6 ST account for 63% of the isolates) with a ‘long tail’ of uncommon STs (Dingle et al. 2001).

The capacity for natural transformation that confounds convention typing approaches can be used to provide a degree of source attribution. Campylobacter of chicken origin are likely to have acquired genes that are current among chicken campylobacter strains and it has been estimated that in an average C. jejuni isolate 86 of its 1643 genes will have been acquired from the gene pool of the host campylobacter population (McCarthy et al. 2007). This effect is apparent in the loci used for MLST and it has been found that certain alleles have a host species signature. Although this approach doesn’t work on some common widely distributed STs, it is able to distinguish between strains of chicken, ruminant, pig origins, and wild bird origin.

Campylobacter enteritis is primarily a human disease and C. jejuni and C. coli are considered to be part of the commensal microbial flora of food producing animals such as chickens, cattle, sheep, and pigs. Nevertheless, campylobacter do elicit an immune response in the animals that they colonize and have been associated with gastroenteritis in ruminants and pigs and ‘vibrionic hepatitis’ in poultry. The impact of these conditions is difficult to assess as campylobacter are so widespread in food producing animals that they are readily found in both sick and health animals. Wild birds have high carriage rates of campylobacter but are usually colonized with strains not commonly found in humans. Companion animals may act as reservoirs of campylobacter infection and there is evidence of enteric illness particularly in young animals.

Humans are not natural hosts of C. jejuni and C. coli and infection is usually transitory. Infections may be asymptomatic, and there is considerable variation in the severity of illness. Humans are not a significant reservoir of infection and person-to-person spread is uncommon. Not all members of the genus Campylobacter are pathogenic to humans and some such as C. hominis (Lawson et al. 2001) seem to be part of the normal human intestinal microbial flora.

The infectious dose of C. jejuni has been estimated at as few as 500 cells and the typical incubation period is between 3 to 5 days but may range from 1 to 10 days. Upon ingestion campylobacters pass through the acid environment of the stomach where their survival will be influenced by the buffering capacity of the food with which they are consumed. Initially colonization occurs in the jejunum and upper ileum and then spreads to rest of ileum and colon. Campylobacters cross the intestinal mucosa, adhere to epithelial surface and enter epithelial cells. Infected mucosa show acute inflammatory infiltration accompanied by fluid secretion which varies in extent dependent on the degree of host response and the extent of epithelial damage. Cell adherence and invasiveness are essential in campylobacter induced diarrhoea. The mechanisms involved are poorly understood, but are dependent on flagella-mediated motility and may involve other factors such as the capsule. Several host cell adhesion factors have been identified in C. jejuni and though their exact roles remain unclear, attachment to the host cells is required for subsequent invasion. Once internalized, C. jejuni cells are primarily confined to membrane bound vacuoles, termed C. jejuni containing vacuoles, in which they are able to survive intercellularly. The role of toxins in campylobacter pathogenesis is not fully understood, but it is now clear that, contrary to some early reports, C. jejuni does not produce a cholera-like endotoxin. Most strains of C. jejuni produce a cytolethal distending toxin, which has been shown to cause cell cycle arrest and apoptotic cell death.

Infected individuals typically show a rapid immune response with circulating antibodies detectable 6 to 7 days after the onset of illness. Specific IgA antibodies are secreted in the intestines and these give protection against the infecting strain and to some degree against other strains that share or have similar epitopes in immunogenic structures such as the capsule, flagella, and LOS.

The symptoms of the diarrhoeal illness caused by C. jejuni and C. coli vary from asymptomatic to severe. The first signs are abdominal pain and diarrhoea or in some cases a prodomal fever with headache, flu-like illness, and myalgia. Cases with this prodome usually have a more severe illness. The abdominal pain can be severe so that it may be mistaken for appendicitis. In England 82% of patients admitted to hospital with ‘food poisoning’ have a campylobacter infection (Adak et al. 2002). Nausea is a commonly reported symptom but vomiting is rare. The diarrhoea is profuse and may be watery or bloody (Gillespie et al. 2006). After three to four days the diarrhoea will ease and most patients recover spontaneous within a week. Recovery is usually complete but relapse is possible. In some cases excretion of campylobacter may continue for several weeks after symptoms abate, but person to person spread is uncommon, and long-term carriage does not occur except in immunosuppressed individuals. In children the illness is usually less severe than that in adults, but bloody diarrhoea, especially in children under one year of age, and vomiting, are more commonly reported. Complications are rare but include bacteraemia, intestinal haemorrhage, toxic mega colon and haemolytic uraemic syndrome. Bacteraemia occurs more frequently in the elderly and reported rates vary from between 1 to 8 cases per 1,000. Deaths due to campylobacter are uncommon and usually associated with underlying illness.

Campylobacter infection has been associated with the development of Irritable Bowel Syndrome (IBS), although a causal link between the two is still contentious. It is thought that a combination of the damage to the mucosa and disruption of the normal flora lead to prolonged bowel dysfunction. It has been estimated that as much as 25% of post infectious IBS could be attributable to campylobacter infection (Neal et al. 1997).

Campylobacter has been associated with a range of post infection rheumatological conditions, (Townes 2010) the most commonly reported being reactive arthritis (ReA). Symptoms usually occur two weeks post infection and manifest as pain and swelling in more than one joint, most typically the knees, ankles, wrists, and lower back. Symptoms may last for up to two months. The estimated incidence of ReA varies from between 1 to 15% and has been particularly associated with individuals with the HLA-B27 genotype.

Guillain-Barré Syndrome (GBS) may occur 1 to 3 weeks after infection with campylobacter and is an acute inflammatory demyelinating polyneuropathy characterized by progressive symmetrical motor weakness ranging from weakness of the extremities to complete paralysis and respiratory problems. GBS is now the major cause of flaccid paralysis worldwide since the near eradication of poliomyelitis. GBS occurs due to molecular mimicry between certain campylobacter LOS epitopes and nerve gangliosides. This causes an autoimmune response resulting in demylation of the nerve and degeneration of the axons and this is initially experienced as a tingling sensation, then as apparent weakness and ultimately flaccid paralysis. Mortality in severe cases of GBS ranges from 2 to 10% reflecting the varying quality of care available in different counties. Full recovery may take between 6 to 12 months but often there is some residual deficit (Nachamkin 2002; Wierzba et al. 2008). Other agents such as Mycoplasma pneumoniae and cytomegalovirus are associated with GBS and the proportion of cases with campylobacter as an antecedent is unclear. Estimates range from 15% in culture-based studies (Tam et al. 2003) to 80% where serology has been used (Van Koningsveld et al. 2000). The occurrence of GBS has been estimated at 1 in every 1,000 cases of campylobacter. Penner serotypes HS:19 and HS:41 are especially associated with GBS (Nachamkin 2002) and though it is now known that the serotype is derived primarily from capsular polysaccharide, the LOS of these strains possess epitopes analogous with human nerve gangliosides. Many campylobacter strains are capable of expressing LOS epitopes that are sufficiently similar to human nerve gangliosides to be able to induce GBS. Antibodies analogous to over 20 different gangliosides have been identified in the sera of cases of GBS.

Milder forms of ReA and GBS may be under reported. In a survey of patients recovering from campylobacteriosis, 37% reported musculoskeletal problem, 11% sensory problems, and 9% general weakness (Zia et al. 2003).

In most cases campylobacteriosis is best treated with rest and regular dehydration to replace lost fluid and electrolytes. The use of antibiotics is rarely indicated but may be effective if given early enough and has been shown to reduce the period of faecal shedding. The most commonly used antibiotics are erythromycin and ciprofloxacin. Resistance rates vary widely between countries but generally erythromycin resistance rates for C. jejuni are fairly low, but higher in C. coli. Ciprofloxacin resistance rates have increased alarmingly especially from travellers returning from countries where fluoroquinolones are licensed for use in animal husbandry. There is some evidence to suggest that infection with resistant campylobacter strains leads to a more severe infection.

Campylobacter is the major cause of bacterial enteritis worldwide and the World Health Organization estimate that 1% of the population of Western Europe are infected with campylobacter each year. The reported incidence varies widely from country to country primarily due to differences in approach to laboratory culture and reporting practices. Incidence is estimated from the number of laboratory confirmed cases and the models used in different countries may vary greatly. Generally, worldwide, the picture has been one of a rise in incidence rates during the 1980s as isolation techniques were refined. There then seems to have been a more gradual increase in incidence throughout the 1990s, which seems to have peaked in most countries in the early 2000s and in some countries a subsequent decline. In the European Community the overall incidence of campylobacteriosis is 51.6 cases per 1 00, 000 persons per year (Janssen et al. 2008).

However, the incidence in individual member states varies from 0 to over 300 per 1 00, 000. In the UK laboratory confirmed cases reached a peak at 57, 674 cases in 2000, declined to 44, 294 cases in 2004 and have since begun to rise once more (Fig. 16.1). The estimated incidence in the UK, based on 2005 data, is 88.5 per 1 00, 000 (EFSA 2006) at the same time the USA incidence was estimated at 12.7 per 1 00, 000 (CDC 2005), whilst the highest incidence occurred in New Zealand with 400 per 1 00, 000 persons (Baker et al. 2007).

In industrialized countries cases of campylobacter typically show a bimodal age distribution with the highest peak in 1 to 4 year olds and a less pronounced second peak in young adults (15 to 24 year olds). The reasons for these peaks remain unclear. Perhaps small children are over represented in samples as they are more likely to have stool samples taken. There is a slight bias amongst cases towards males (1.1 to 1.5 times higher) and this is most evident in young adults.

Recent analysis of longitudinal data in the UK suggests that underlying the gradual rise in reported campylobacter cases is a more complex story of changing incidence amongst different age groups that may reflect changes in society. Over the last twenty years the number of reported cases in the over 50’s has risen by a factor of 3.8, whilst during the same period cases in the under 20’s have decreased by 5% (Fig. 16.1) (Gillespie et al. 2009). There are probably many factors behind these changes not least the fact the UK has an increasing aging population, but also changes in access to health care, and eating habits. The wider use of immunosuppressive therapies may also be an important factor as individuals with an impaired immune system are more prone to campylobacter infection. Interestingly, the rise in campylobacter infection amongst the over 50’s since the 1980s coincides with discovery of Helicobacter pylori as the cause of gastric ulcers and its treatment with a combination of antibiotics and a proton pump inhibitor. Gastric ulcers are seen most frequently amongst the over 50’s and it has been suggested that the use of proton pump inhibitors to reduce stomach acidity during treatment of H. pylori infections effectively removes or reduces a significant barrier to campylobacter infection by effectively lowering the infectious dose in this group (Gillespie et al. 2009).

The majority of reported campylobacter cases are sporadic infections involving individuals or small family groups. Case-control studies have identified several risk factors associated with sporadic campylobacter infection these include the consumption of chicken in various forms; the consumption of barbecued or undercooked meat; the consumption of salad vegetables; drinking unpasteurized milk; drinking untreated water; swimming in natural surface water; and foreign travel (Adak et al. 1995; Frost et al. 2002; Gillespie et al. 2002, 2003, 2006). Interestingly, case-control studies report that handling and cooking of chicken at home have a ‘protective’ effect: i.e. those people who regularly eat and prepare chicken at home have a lower rate of infection presumably from acquired immunity (Adak et al. 1995).

Campylobacter outbreaks (i.e. those affecting more than one household) are rarely recognized. There are a number of reasons why this is so: the bacterium generally survives poorly outside the host and is unable to multiply on food or to grow below 30°C; person to person transmissibility is low; campylobacter have a relatively long incubation period and with a further 48 hours required for culture, often several days will have elapsed between infection and diagnosis, and this reduces the chances of recognition and reporting of outbreaks. Of the 2,374 general outbreaks of infectious intestinal disease reported to the Public Health Laboratory Service between 1995 and 1999 for which an aetiological agent was identified, campylobacter accounted for only 50 (2%), of which food-borne transmission was identified in 35 (Frost et al. 2002). Nevertheless, point source outbreaks of campylobacter may be more common than previously thought. An extended MLST-based longitudinal study has shown evidence of temporal and spatial clustering of specific types within a locale (Dingle et al. 2008). Detailed epidemiological questionnaires distributed to cases of campylobacter infection often reveal knowledge of people with similar symptoms (Gillespie et al. 2003).

When campylobacter outbreaks are successfully investigated they are commonly associated with the consumption of chicken (especially when under cooked); consumption of unpasteurized milk and drinking untreated or contaminated water. However, the investigation of outbreaks can be complicated by the fact that several campylobacter strains may be isolated from epidemiologically linked cases. This may occur when faecal material from several sources contaminates a watercourse or when material from several different animals is combined in a foodstuff: for example the livers from a number of birds combined in chicken liver paté (Forbes et al. 2009).

The incidence of campylobacter infections in humans exhibits a pronounced and consistent seasonal pattern (Nylen et al. 2002). In temperate countries such as the UK, incidence peaks in late spring and early summer then falls away during late summer and autumn. Elsewhere in the world seasonality is variable but the effect seems to be more pronounced with increasing latitude and is more evident in rural than urban areas. The factors behind seasonality are not fully understood, but seem to be associated with climate factors such as temperature, sunlight, rain fall, and humidity. Of these higher temperature seems to be the best correlate (Kovats et al. 2005). Seasonality is also evident in animals. In the UK the prevalence of campylobacter in chickens rises sharply during late spring and early summer to reach a peak in July/August. Interestingly, in a study from Wales the seasonal peak in chickens seemed to occur a few weeks after the peak in human infections (Meldrum et al. 2005). The role of other animals such as flies (Nichols 2005) and migratory birds (Colles et al. 2008) either as sources or vectors of campylobacter has been suggested but the relative importance of these, if any, has yet to be determined.

A combination of the lack of chlorinated drinking water, poor sanitation, and frequent contact with animals leads to hyperendemic infection in many developing countries. The repeated exposure to campylobacter that children experience during the first few years of life results in immunity and older children and adults are generally unaffected by the disease. Nevertheless, campylobacter may contribute significantly to infant diarrhoea and mortality in some circumstances. The often unapparent high prevalence of campylobacter in developing countries may pose a high risk of infection to tourists and other visitors. Campylobacter enteritis is a common form of travellers’ diarrhoea.

The estimate of the economic cost of campylobacteriosis varies enormously between countries. In the UK it has been estimated from data gathered between 1996 and 2000 when laboratory confirmed cases ranged from 43, 978 to 57, 674 (Fig. 16.1) that the average number of actual cases of campylobacter infection was 3 37, 655. From this figure it has been extrapolated that this would have resulted in 160, 788 GP consultations, 15, 918 hospital admissions, 58, 897 bed days spent in hospital, and 80 deaths (Adak et al. 2005). The impact of campylobacter infections should not be underestimated: of causes of food-borne illness in the UK campylobacter is the greatest single cause accounting for 20% of all cases, but in terms of hospital bed days needed it accounts for 58% of the total (Adak et al. 2005). The greater requirement for hospitalization arises from the severity of abdominal pain and sequalae. In terms of financial cost in England in 1995 the average cost of an uncomplicated case of campylobacter enteritis was estimated at £1,315. That would have given a conservative estimate of a total cost £65 million per year. However, if one considers the additional cost of hospitalization the estimate might be as much as £500 million per year (Humphrey et al. 2007).

Campylobacter are adapted to grow in the ruminant and avian gut and at some point the strains that cause campylobacteriosis in humans must have originated in an animal host. Recent MLST-based attribution studies suggest ∼97% of C. jejuni isolates from humans originated from animals farmed for meat and poultry (Sheppard et al. 2009; Wilson et al. 2008). This approach suggests that for C. jejuni isolated from humans between 57 to 78% originated from chicken, 18 to 39% from ruminants (sheep and cattle), ∼1% from pigs and between 3 to 4% from environmental sources such as wild birds and water. For isolates of C. coli from humans 40 to 56% originated from chicken, 42 to 54% from ruminants, and 1 to 6% from pigs (Sheppard et al. 2009). Detailed questionnaire-based follow up studies of cases of campylobacteriosis suggest that a food-borne vehicle was responsible for the infection in about half of the cases, the route of infection in the remainder of cases remains unknown. Table 16.1 illustrates the levels of campylobacter commonly found in food animals and the corresponding (uncooked/unpasteurized) foods that they produce.

The transmissibility of camplylobacter is low and survival of the cells outside the body is short. In the absence of a protective food matrix cells are unlikely to pass through the stomach’s acid barrier unless consumed in large numbers. Simple hygiene measures should be sufficient to protect against direct campylobacter infection. In outbreaks and sporadic cases person to person spread does not seem to occur and secondary cases are uncommon. In the food industry there is a risk of acquiring campylobacter infection associated with the handling colonized animals or carcasses. Newly recruited poultry and abattoir workers often report illness in the first few months of employment, but thereafter do not seem to be troubled, presumably having acquired immunity though repeated exposure.

Over the last two decades the consumption of poultry, primarily chicken, has increased dramatically. In the UK poultry production increased between 1985 and 2005 from 120 to 174 million birds, the increase being primarily due to a near doubling in the number of ‘table birds’ produced (Defra). Likewise in the USA over the same period yearly chicken consumption per capita increased from 53.1 to 87.4 pounds (USDA). This dramatic increase in production and consumption coincides with the increase in confirmed campylobacter cases and poultry consumption has been demonstrated to be a major risk factor in several studies. As can be seen in Table 16.1 poultry are frequently heavily contaminated with campylobacter before and after production. Various types of poultry have been implicated as a vehicle of human campylobacterosis: any type of chicken; poultry and poultry liver; raw or rare chicken; cooked chicken; processed chicken; and chicken prepared by or eaten in a commercial food establishment (ACMSF 2005). In the UK it is thought that contaminated chicken is the single most important cause of foodborne illness being responsible for an estimated 398, 420 cases each year. This is equivalent to 111 cases per million servings (Adak et al. 2005). Further evidence of the role of poultry in human campylobacteriosis came from Belgium in 1999 when poultry meat was withdrawn from sale due to fears that chicken feed had become contaminated with dioxins. During the period of withdrawal there was a concomitant decline of ∼40% of in the number of cases of human campylobacteriosis (Vellinga and Van Loock 2002). It is important to distinguish between chicken as a source of campylobacter and chicken as a vehicle for campylobacter infection. Undoubtedly chicken is a major source of campylobacter as has been shown in MLST-based attribution studies. However, epidemiological evidence suggests that actual consumption of chicken is responsible for only between 20 to 40% of cases (ACMSF 2005). A significant proportion of the campylobacter that originated in chicken must reach humans via other vehicles. Such routes might include: other foodstuffs cross-contaminated during production or in the household; companion animals; contaminated water; or other as yet to be identified routes of infection. Alternatively, the comparatively long incubation period and laboratory isolation required to detect campylobacter might confound epidemiological analysis making it difficult to attribute a vehicle in some cases.

Campylobacter are part of the normal flora of cattle, sheep, and pigs and can readily be isolated from these animals. However, within any given herd carriage rates may vary considerably between animals with some individuals being heavily contaminated and others below the level of detection. In contrast to poultry, the process of converting cattle, sheep, and pigs to beef, mutton, and pork results in a significant reduction in the degree of campylobacter contamination (see Table 16.1). This is primarily because the slaughter process for larger animals is less automated than that for poultry and the practice of roding and bunging, whereby the gastrointestinal tract of the animal is sealed during slaughter and initial processing reduces the chance of faecal contamination of the carcass. Once prepared the carcasses are hung in chillers where a combination of cold and desiccation is effective in reducing campylobacter numbers (Humphrey et al. 1995). Despite the lower levels of contamination, red meats still pose a risk of campylobacter infection. MLST-based source attribution studies suggest that campylobacter originating in cattle and sheep are responsible for a significant proportion of human infections. In contrast, by this approach, campylobacter originating from pigs seem to be a relatively unimportant source of human campylobacteriosis with even C. coli of porcine origin contributing a maximum of 6% of human isolates in the UK. This may be because a significant proportion of pork often undergoes additional curing processes prior to consumption, which might further reduce campylobacter numbers. Epidemiological studies identify undercooked and barbecued red meats as a vehicle for campylobacter infection (Adak et al. 1995).

A significant proportion of dairy cattle carry campylobacter and inevitably milk may become contaminated (see Table 16.1). Faecal contamination at the time of milking is thought to be the primary means by which the bacteria get into milk, but they may also be excreted directly from infected udders in cases of campylobacter mastitis. Historically many major outbreaks of campylobacteriosis have been milk-related. These have been due to the consumption of raw or inadequately pasteurized milk and have sometimes involved thousands of people. Milk-related outbreaks are now less common due to legislation and better awareness of the risks associated with the consumption of unpasteurized milk. Also with the decline of smaller farms, milk is more likely to be pasteurized in larger centralized facilities that are well maintained. Small outbreaks have occurred due to the consumption of milk delivered to doorsteps that has been pecked by birds (usually of the crow family) (Southern et al. 1990). Typically this occurred in early summer when birds have young to feed. It is thought that the birds become contaminated whilst gathering insects from cow faeces. This route of infection has diminished with the decline in use of foil topped bottles and door step deliveries.

Campylobacters are capable of survival in unchlorinated water, in planktonic suspension, in biofilms, and in the vacuoles of certain amoeba species. However, it is unlikely that they are able to multiply in water to any great extent and water must therefore be seen as a vehicle for rather than a source of campylobacteriosis. Open water is frequently contaminated with faeces from wild bird and farm animals. In some circumstances it may become more heavily contaminated with slurry run off or untreated human sewage. Sporadic infection due to recreational use of water, either inadvertently through water sports activities or in cases of trekkers drinking streams or lake water has been reported. Often these types of infection are associated with uncommon campylobacter types associated with wild birds. A more significant threat to human health is the risk posed by the contamination of potable water supplies. This might be an important and underappreciated route of infection particularly in rural areas where local unchlorinated sources of potable water might be utilized and where chlorinated mains water is available the greater length of pipe required to connect properties to the mains might make them vulnerable to breakage and contamination. In Sweden an association was found between the incidence of campylobacter in different regions and the average water pipe length per person (Nygard et al. 2004). Epidemiological studies have identified consumption of bottled mineral water as a risk factor for campylobacteriosis (Gillespie et al. 2002). Although campylobacters have been shown experimentally to be capable of long term survival in refrigerated bottled water they have yet to be isolated or detected by PCR-based techniques from commercially available bottled waters. Presumably the positive association between campylobacteriosis and bottled water is due to some other as yet undetermined risk factor to which regular consumers of bottled water are more exposed than the general populace.

Ready-to-eatfoods such as vegetables, salad and fruit have occasionally been implicated as vehicles of campylobacteriosis. These food stuffs may become contaminated either during production (i.e. if irrigated or washed with water contaminated with faeces) or in the home due to cross-contamination from contaminated meat via surfaces, kitchen implements, or unwashed hands. The extent to which this route of infection contributes to overall human illness is difficult to assess.

Companion animals and animals in petting zoos and farms may also act as vehicles for campylobacteriosis. Cats and dogs have their own campylobacter species: C. helveticus and C. upsaliensis and the latter has been implicated as an uncommon cause of illness in humans. MLST analysis of C. jejuni isolates from dogs shows that they originated in chicken and ruminants. Studies suggest that cats and dogs are capable of acting as vehicles for C. jejuni and C. coli and sick animals, particularly puppies and kittens, may pose a risk of infection to humans especially young children.

The widespread distribution of campylobacter amongst food producing animals is the most significant reservoir of infection for humans. Of all of the elements affecting the food chain, control of level of campylobacter contamination in poultry production would be likely to have the biggest impact. Even if complete eradication were not possible it has been estimated a 2 log reduction of campylobacter on poultry carcases would reduce the risk of human infection 30-fold (Rosenquist et al. 2003). Broiler chickens are probably the most numerically significant poultry product and are reared on an industrial scale in high density sheds and slaughtered and prepared for retail in high throughput processing plants. Interventions could be made during production in the form of better animal husbandry, increased biosecurity and greater efforts to avoid cross contamination between flocks. During slaughter, priority possessing of campylobacter-negative or low-level contaminated flocks might be effective. During processing, carcasses identified as highly contaminated could be earmarked for freezing or decontamination with disinfectant spays or washes. The development of a campylobacter vaccine for poultry capable of achieving similar results to the Salmonella Enteritidis vaccine (Fig. 16.1) would have a major impact on the number of cases of human campylobacteriosis. However, at present the prospects for such a vaccine seem quite distant. For a more detailed view of intervention during food production see the review article by (Humphrey et al. 2007).

In industrialized countries the provision of safe drinking water and adequate sanitation are vital preventative measures that are largely taken for granted. They only become apparent when the system fails. Continued public education should be undertaken regarding basic hygiene in particular the importance of hand washing after handling animals and before eating or handling food. Clear instructions on best practice in the kitchen, particularly with regard to the handling, storage and preparation of raw meat should also be available.

National clinical laboratories and epidemiology networks continue to play a major role in identifying and monitoring trends in campylobacter infection in humans. Whilst the usefulness of typing methods for campylobacter is debated, it is generally accepted that typing remains epidemiologically useful in the investigation of outbreaks.