John E. Moore Molecular Epidemiology Research Unit, Northern Ireland Public Health Laboratory, Department of Bacteriology, Belfast City Hospital, Belfast, Northern Ireland, BT9 7AD.
Telephone: (01232) 263554 Facsimile: (01232) 321084
E-mail: jemoore@niphl.dnet.co.uk
Members of the genus Campylobacter are now recognised as the most common
cause of acute bacterial enteritis in the British Isles, with Campylobacter
jejuni being the dominant species in human infections. This review aims to
outline the various reservoirs of the organism and the possible modes of
transmission of the organism, that may ultimately result in human infection.
A. Introduction
The past three decades have witnessed the rise of Campylobacter enteritis
in human from virtual obscurity to notoriety, with present isolation rates
superseding those of other enteric pathogens such as Salmonella spp. and Shigella
spp. (Skirrow 1991). Reports from both developing nations including Mexico
(Marquez-Davila et al. 1989), Brazil (Borges et al. 1989), Bangladesh (Blaser et
al. 1980a), Vietnam (MEgraud et al. 1989, 1991) and Thailand (Varavithya et al.
1990) as well as from western countries including Scotland (Brunton and Heggie
1977), Italy (Crotti et al. 1989), the Netherlands (Severin 1978) and the USA (Nachamkin
1993) have shown Campylobacter isolations to be equal to, and sometimes
higher than those of any other enteric pathogen. Campylobacters belonging to the
thermophilic group including C. jejuni, C. coli, C. lari and C.
upsaliensis are now the most frequently isolated pathogens in many part of
the world. In addition this upward trend in isolations both on the British
mainland and in Northern Ireland is showing no sign of lessening.
Unlike the salmonella and other enteric pathogens, the majority (ca. 99%) of
clinical reports concerning Campylobacter are sporadic and Campylobacter
enteritis outbreaks are rare (Cowden 1992). The lack of well developed typing
schemes has hindered the epidemiological investigations seeking natural
reservoirs of the organism and modes of transmission from these sources to man.
Only about 15% of clinical isolates are identified to species level thus making
epidemiological investigations extremely difficult to perform (J. Cowden,
personal communication).
B. Potential sources of human infection
Although campylobacters are not completely new to applied bacteriology, they
have evaded traditional techniques used for the isolation of pure cultures,
apart from single isolations that were free from competing organisms. Until the
development of a selective medium by Skirrow (1977), these organisms were known
mainly by veterinarians as animal pathogens, which were responsible for a wide
variety of disorders in cattle, sheep and pigs (Bokkenheuser and Mosenthal
1981). Since the development of more sophisticated isolation techniques (Dekeyser
et al. 1972; Humphrey 1989), the true disease potential of these organisms has
become apparent and today campylobacteriosis is regarded as a zoonoses, which is
capable of being transmitted to man by a wide range of domestic animals
(Prescott and Munroe 1982).
Various researchers have classified Campylobacter enteritis as a
zoonosis, as animals constitute the main reservoir of these organisms (Prescott
and Munroe 1982). Campylobacters appear to exist as commensals in many wild and
domestic animals (Blaser 1982; Blaser et al. 1983a, 19836), thus presenting a
risk to food safety, due to contamination of carcasses and offal's at slaughter,
as well as foodstuffs due to cross-contamination from raw meat and faecal
material. Smibert (1978) discovered a high frequency of C. jejuni in the
intestinal flora of healthy young cattle, sheep and swine and suggested that
this organism may form a part of the normal faecal flora of the immature animal.
Alternatively it has been suggested that positive carriage of the organism in
asymptomatic animals may be the result of immunity established during infection
at an early postnatal stage (Prescott and Munroe 1982; Park et al. 1991). When
asymptomatic animals are slaughtered, some carcasses may be contaminated through
spillage of intestinal contents onto the offal's, as well as from faecal
contamination from the hides or feathers (Gill and Harris 1982).
The natural habitat of most Campylobacter spp, is the intestine of
warm-blooded animals, including pigs. Although the enteropathogenic
campylobacters have been shown to cause disease in a wide variety of animals and
man, certain species have been shown to have a preferred niche. C. jejuni
have been shown to be most prevalent in poultry, whilst pigs appear to be the
preferred niche of C. coli. Seagulls are mainly infected with C. lari
and cats and dogs have been shown to harbour C. upsaliensis and C.
helveticus. Hence the ecological niche occupied by the Campylobacter
may be of significant importance in relation to the epidemiology of the disease.
(1) Farm animal reservoirs
Prescott and Munroe (1982) reported enteritis in calves caused by C.
jejuni. In Norway, Rosef et al. (1983) reported a faecal carriage rate of
0.8%, with several serotypes prevailing within a herd at any one time. In
Finland, Honninen and Roevuori (1981) showed the faecal carriage rate in cattle
to be 5.5%, represented by C. jejuni. In Sweden, a similar study by
Svedhem and Kaijser (1981) showed a positive faecal isolation rate of 19%. In an
extensive study by Garcia et al. (1985), C. jejuni was isolated from 50%
of cattle examined.
In 1955, Russell described an important epidemiological link between the
"vibrio" which was isolated from sheep with scour and from similar
clinical cases in man. Duffell and Skirrow (1978) published evidence supporting
the work of Russell, where bacteriological evidence linked campylobacter-related
abortions in ewes with human enteritis. Poultry has been shown to be highly
contaminated with C. jejuni and high isolation rates have been reported,
ranging from 30% (Stern et al. 1985) to 100% (Munroe et al. 1983). Consequently
poultry, especially chicken has been considered a major reservoir of C.
jejuni. The Communicable Disease Centre in the USA showed a strong
correlation implicating chicken as a major vector in human enteric disease and
indicated that chicken accounted for an aetiological factor of 70% of the risks
contributing to the disease (Deming et al. 1987). In the UK, Pearson et al.
(1987) concluded that there was a definite correlation between human infection
and consumption of chicken. In studies, sporadic cases were traced back to a
chicken producer whose birds had the same serotype (C. jejuni Lior 1,
Penner 4), as seen in clinical cases. These workers showed that when this strain
was removed the human disease frequency diminished by 35%. These workers
consequently concluded that
there was a definite correlation between human infection and consumption of
chicken. Park et al. (1991) suggested that the rearing of campylobacter-free
poultry was not economically feasible and that reduction in incidence should be
attempted at the processing stage, through the addition of a solute or by drying
operations, in order to reduce the amount of available water, thus creating a
hurdle for the growth of these organisms.
(11) Pigs
Many workers have demonstrated C. coli to be a pathogen in the
non-immune postnatal pig, where enteritis was induced by inoculating the animals
with pure cultures of the organism (Taylor and Olubunmi 1981; Olubunmi and
Taylor 1982; Ward and Jones 1990). C. coli may become established as a
commensal organism in the intestinal flora of the young pig, whereby it takes a
non-pathological role, after the initial colonization and pathogenic stage,
similar to the role of C. jejuni in poultry (Prescott and BruinMosch
1981; Stern and Meinersmann 1989). Straw (1990) could not demonstrate any loss
in performance with respect to growth rate or feed conversion in pigs with Campylobacter-induced
enteritis. Scarcelli et al. (1991) showed that C. coli was the only Campylobacter
spp. in 22/30 pigs. In addition, these workers found that 15/25 apparently
healthy slaughter pigs yielded Campylobacter. Various researchers have
found faecal contamination rates greater than 50%, including 66% (Sticht-Groh
1982) and 95% (Svedhem et al. 1981). In contrast to this, Prescott and Bruin-Mosch
(1981) reported a carriage rate of only 0.9% in pig faeces. Oosterom (1980)
discovered 60.7% of intestinal contents to be contaminated, whilst carcass
contamination rates varied from 12.5% (Bracewell et al. 1985) to 58% (Stern
1981).
Unlike isolates found in other domestic animals such as cattle, most isolates
from swine have phenotypic characteristics resembling those of C. coli (Skirrow
and Benjamin 1980), although Finlay et al. (1986) isolated both C. coli
and C. jejuni from pigs at 55% and 45% respectively, with an overall
isolation rate of 46.5% from faeces sampled. These workers found that faeces was
the major source of contamination of both the neck and diaphragm muscles and
concluded that faecal contamination may be considered a potential source of
carcass contamination.
Although there is a high contamination level of carcasses with Campylobacter
spp. from intestinal and faecal material, Oosterom et al. (1983) questioned
the epidemiological link between this and porcine-related enteric
campylobacteriosis in humans, in contrast to the obvious links shown between
carriage of C. jejuni in poultry and infection in man.
From birth to slaughter, there are numerous opportunities for the
colonization and cross-contamination of Campylobacter spp. Large litter
sizes may result in certain piglets being deprived of colostrum containing
antibodies and hence become more susceptible to infection from Campylobacter
spp. During the weaning, fattening and finishing stages, animals are in
close contact with each other. Brent (1982) recommended at weaning, each animal
should have at least 0.3m² floor space. Consequently this may allow
infected animals to contaminate ad lib feeding systems with faeces containing
large numbers of organisms and thus help the transmission of the organism from
animal to animal. Therefore young animals which have been intensively reared may
be infected with Campylobacter spp, at the point of slaughter.
(111) Red meats and offals
Sticht-Groh (1982) found that 59% of healthy slaughtered pigs, washed
intestines and water samples collected from a slaughterhouse and a butcher's
shop were infected with Campylobacter spp. and 24% of isolates were
identified as C. jejuni, Sticht-Groh concluded that this may be a
possible source of infection for man, as intestines are still used as casings
for sausage production in central Europe. In a survey of pig carcasses in
Georgia, USA, Bracewell et al. (1985) found 12.5% were positive for C. coli.
It is interesting to note that positive carcasses for Campylobacter spp.
only harboured C. coli and no other species of Campylobacter. In an
extensive survey of meats from both a slaughterhouse and retail outlets,
Turnbull and Rose (1982) showed 1.6% of samples to be positive for Campylobacter
spp. These workers also found that all strains in the slaughterhouse were C.
coli and suggested that the four strains of C. jejuni isolated from
retail pork samples were from other sources. In addition, they concluded that
there was no apparent correlation between the isolation of Campylobacter
and Salmonella, but did conclude that the high isolation rates of C.
jejuni in chickens are not paralleled in red meats.
Gill and Harris (1982) in a survey of animal carcasses showed that positive
carcasses were contaminated with low numbers of C. jejuni (1-10 cfu
cm-2). However in this study, all swabs taken in butchers' premises over a six
month period failed to grow campylobacters, both by the direct culture and after
enrichment. These workers suggested that failure to isolate campylobacters in
retail butchers' outlets was attributed to the oxygen sensitivity of the
organism and concluded that fresh meat sold to the public was not an important
source of infection for man. Turnbull and Rose (1982) in a survey of meats from
both slaughterhouses and retail outlets showed abattoir meat to be more
frequently contaminated than meat from a retail outlet and suggested that the
lower incidence in retail meat was due to an inability of C. jejuni to
multiply readily in or on meat. Fricker and Park (1989) isolated campylobacters
from retail outlets by using enrichment protocols. These workers concluded that
offals may be a more important source of human enteropathogenic campylobacters
than red meats, as offals are not subjected to drying, to which the
campylobacters are sensitive (Doyle and Roman 1982). Itoh et al. (1980)
implicated pork as the causative agent in an outbreak in a Japanese day care
centre and amongst school children. Pork has also been implicated as the cause
of an outbreak of enteritis amongst school children (Yanagisawa 1980).
Porcine offals have been shown to be a source of C. jejuni and C.
coli. Bolton et al. (1985) reported isolation rates of 2.5 and 11.5% for
abattoir and retail porcine offals respectively. Likewise, Moore and Madden
(1998) demonstrated that porcine liver in Northern Ireland had a prevalence of
approximately six percent. However in contrast to this, Arwana and Promberger
(1991) in a comprehensive study of beef liver, kidney, spleen and muscles failed
to detect Campylobacter spp. from 137 samples at slaughter and 61 samples
taken from butchers' premises. The presence of these organisms on offals may be
attributable to cross-contamination from infected faeces during the slaughtering
and dressing processes (Franco 1988). Rosef (1981) found a high prevalence of C.
jejuni in the gallbladders and bile ducts of clinically healthy slaughter
pigs and concluded that these infected sources may contaminate the liver during
the dressing process. Alternatively the presence of such organisms in offals
could be the result of the invasive capacity of the organism, as members of this
genus have been reported to be invasive (Butzler 1984). In contrast, Breuer
(1986) found that chicken offal was very frequently contaminated with Campylobacter
spp. and showed 85% of chicken livers to be contaminated with C. jejuni
and C. coli. In a recent study of the contamination of poultry,
Khalafalla (1990) reported that 40%, 36% and 30% of liver from chickens, ducks
and turkeys respectively were contaminated with Campylobacter spp.
Various other researchers have considered offals to be a potentially
important source of campylobacters associated with sporadic human infection (Fricker
and Park 1989; Skirrow 1991).
Modes of transmission
The widespread presence of C. jejuni in animals and birds provides an
almost unlimited source of infection for humans (Bokkenheuser and Mosenthal
1981). Transmission vectors are important in the spread of the disease from
infected individuals to new hosts. Campylobacter spp, have been shown to
be transmitted to man in a number of ways.
(I). Animal to animal contact
One factor contributing to these high incidence rates in pigs is the
intense-rearing systems of pig production units and other intensive production
methods employed on modern farms, where rapid spread of infection may occur
through faecal contamination of feed and drinking water. In contrast, Rosef et
al. (1983) did not isolate campylobacters in Norway from populations of wild
moose, reindeer or beavers. Gill and Harris (1982) suggested that
cross-infection may occur in flocks and herds of animals during the
transportation and holding of animals before slaughter. Once established in the
rearing house, spread of C. jejuni to all birds is rapid through
cross-contamination in a variety of ways (Berndtson et al. 1989a, 1989b). A
combination of poor hygiene procedures between batches of broilers,
inappropriate use of fabrication materials, water quality, presence of rodents
and flies and working practices have all been shown to be contributing factors
in the spread of C. jejuni to broilers (Berndtson et al. 1989a, 1989b;
Pearson 1989; Pearson et al. 1989).
(II). Direct animal to man contact
Duffell and Skirrow (1978) reported on the incidence of campylobacter
enteritis in man following handling aborted foetuses in ewes carrying C.
jejuni. Skirrow (1981) estimated that 5% of human infection was acquired
from contact with dogs, as they have been shown to be both symptomatic and
asymptomatic excretors of this organism (Jorgensen 1981). In a recent study (Bossinger
and Dillon 1992), dogs were shown to remain positive excretors of C. jejuni
for up to 28 days after artificial inoculation.
(III). Person to person contact
Blaser et al. (1981) reported on the person to person transmission of C.
jejuni and concluded that the faecal-oral route may be a particularly likely
mode of transmission of C. jejuni amongst faecally incontinent children.
Mawer and Smith (1979) reported on the vertical transmission of infection from a
mother to her neonate. Although there have been isolated reports supporting
person-to-person transmission, person-to-person spread of Campylobacter
is rare (Butzler and Skirrow 1979). Unlike the common spread of other enteric
pathogens such as Shigella spp (Weissman et al. 1975), the faecal-oral
route does not appear to be of any significance in the transmission of the
organism. This is surprising as campylobacters are excreted in large numbers
(106-109 cells per gram) in faeces of infected individuals (Blaser et al. 1980)
and as the infective dose is relatively small (500 cells) (Robinson 1981).
(IV). Cross-contamination
Under normal processing conditions it is very difficult to produce fresh
meat, especially poultry, which is free from campylobacters. The entire
slaughter processing environment has the potential to perpetuate contamination
because of the close contact of the meat with faeces and other contaminated
material.
During poultry processing, campylobacter-free birds are likely to be
contaminated by debris and other contaminated material from positive birds on
the processing line. Although washing with chlorinated chiller water (340 p pm
available chlorine) has helped to reduce the incidence of Salmonella contamination
in oven-ready birds, it has proven insufficient at reducing the rate of
contamination with C. jejuni (Simmons and Gibbs 1979; Luechtefeld et al.
1981).
Berndtson et al. (1991) suggested that the scalding procedure used in opening
the feather follicles for subsequent feather removal may allow cells of C.
jejuni to be trapped in the subcutaneous region and thus escape subsequent
decontamination with chlorinated chiller water, as this process will only
surface sterilization of the carcass of the bird.
DeBoer and Hahne (1990) demonstrated that C. jejuni isolated from
chickens could be easily transferred from raw chicken to cutting-boards, plates
and hands and subsequently to cooked chicken products. Park et al. (1991)
suggested that cross-contamination may occur by storage of cooked chicken in
unwashed cartons previously used for delivery of the raw product.
Coates et al. (1987) suggested the faecal-oral route from contaminated hands
of food handlers may be a means of contamination. Doyle and Schoeni (1986)
suggested that one source of C. jejuni contamination of retail fresh
mushrooms was handling of the product by staff with poor personal hygiene.
However washing and thorough drying of contaminated hands has been shown to be
effective in the elimination of these organisms (Coates et al. 1987). Faecal
shedding of C. jejuni and C. coli by asymptomatic food handlers in
a high risk food production area e.g. pork pie manufacture, is of considerable
risk to food safety in terms of cross contamination. As with other intestinal
pathogens, infection with Campylobacter does not always produce symptoms
(Butzler and Skirrow 1979). Asymptomatic excretors and mild cases can commonly
be found among the close contacts of cases (Butzler 1984). Such cases as these
give subsequent cause for concern in relation to personnel within the food
industry. Although routine screening identifies positive cases (Jones and Harrop
1981). there is the possibility that asymptomatic excretors will continue to
work possibly in a high-risk operation, unaware of the potential dangers.
Butzler and Skirrow (1979) reported that the carriage rates of these pathogens
in humans may last up to six weeks and even up to one year in some cases.
Coupled with this, a reluctance by clinicians to prescribe antibiotic therapy
for uncomplicated cases has made the management of positive food handlers more
difficult. Consequently various managerial guidelines have been proposed to the
food industry in an attempt to minimise cross-contamination of foodborne
pathogens (Anon 1987).
|
