ArticlePDF AvailableLiterature Review

Global Epidemiology of Campylobacter Infection

July 2015
Clinical Microbiology Reviews 28(3):687-720

DOI:10.1128/CMR.00006-15

Source
PubMed

Authors:

Nadeem O Kaakoush

UNSW Sydney

Natalia Castaño-Rodríguez

UNSW Sydney

Hazel Mitchell

UNSW Sydney

Si Ming Man

Australian National University

Campylobacter jejuni infection is one of the most widespread infectious diseases of the last century. The incidence and prevalence of campylobacteriosis have increased in both developed and developing countries over the last 10 years. The dramatic increase in North America, Europe, and Australia is alarming, and data from parts of Africa, Asia, and the Middle East indicate that campylobacteriosis is endemic in these areas, especially in children. In addition to C. jejuni, there is increasing recognition of the clinical importance of emerging Campylobacter species, including Campylobacter concisus and Campylobacter ureolyticus. Poultry is a major reservoir and source of transmission of campylobacteriosis to humans. Other risk factors include consumption of animal products and water, contact with animals, and international travel. Strategic implementation of multifaceted biocontrol measures to reduce the transmission of this group of pathogens is paramount for public health. Overall, campylobacteriosis is still one of the most important infectious diseases that is likely to challenge global health in the years to come. This review provides a comprehensive overview of the global epidemiology, transmission, and clinical relevance of Campylobacter infection. Copyright © 2015, American Society for Microbiology. All Rights Reserved.

Environmental reservoirs, routes of transmission, and clinical manifestations associated with Campylobacter species. Campylobacter species can be

…

ncidence and prevalence of campylobacteriosis (C. jejuni/C. coli). The latest information on the global epidemiology of campylobacteriosis from the literature is shown, including data from the United Kingdom (47), Denmark (11), Germany (49), Norway (424), Poland (25, 50, 425), the Netherlands (51), Israel (67), China (60, 61), Japan (26), India (63-65), Australia (69), New Zealand (73), Madagascar (78), Malawi (77), Kenya (79, 426), Guatemala (41), Peru (427), Mexico (428), the United States (10 sites within The Food-Borne Diseases Active Surveillance Network) (34), and Canada (37-39). B.C., British Columbia. (Map adapted from an image from Wikimedia Commons [http://commons.wikimedia.org/wiki/File:A_large_blank_world_map_with_oceans_marked_in _blue.PNG].)

…

Timeline of published campylobacteriosis outbreaks since 2007 (381, 382, 429–447).

…

Figures - uploaded by Si Ming Man

Content may be subject to copyright.

Content uploaded by Si Ming Man

Content may be subject to copyright.

Global Epidemiology of Campylobacter Infection

Nadeem O. Kaakoush,

Natalia Castaño-Rodríguez,

Hazel M. Mitchell,

Si Ming Man

a,b

School of Biotechnology and Biomolecular Sciences, The University of New South Wales, Sydney, NSW, Australia

; Department of Immunology, St. Jude Children’s

Research Hospital, Memphis, Tennessee, USA

SUMMARY ..................................................................................................................................................688

INTRODUCTION ............................................................................................................................................688

GASTROENTERITIS..........................................................................................................................................688

Epidemiology.............................................................................................................................................690

North and Central America ............................................................................................................................691

South America .........................................................................................................................................692

Europe .................................................................................................................................................692

Asia and the Middle East...............................................................................................................................693

Oceania ................................................................................................................................................694

Africa...................................................................................................................................................694

OTHER GASTROINTESTINAL MANIFESTATIONS ...........................................................................................................694

Inﬂammatory Bowel Diseases ............................................................................................................................695

Esophageal Diseases .....................................................................................................................................696

Periodontal Diseases .....................................................................................................................................696

Functional Gastrointestinal Disorders ....................................................................................................................696

Colorectal Cancer.........................................................................................................................................697

Celiac Disease ............................................................................................................................................697

Cholecystitis ..............................................................................................................................................697

EXTRAGASTROINTESTINAL MANIFESTATIONS ............................................................................................................698

Guillain-Barré Syndrome..................................................................................................................................698

Miller Fisher Syndrome ...................................................................................................................................698

Bacteremia and Septicemia ..............................................................................................................................698

Cardiovascular Complications ............................................................................................................................699

Meningitis ................................................................................................................................................699

Extraoral Abscesses.......................................................................................................................................699

Reactive Arthritis .........................................................................................................................................699

Complications of the Reproductive System ..............................................................................................................700

CLINICAL MICROBIOLOGY .................................................................................................................................

700

Isolation Methodologies in Clinical Settings..............................................................................................................700

Laboratory Diagnosis .....................................................................................................................................700

Biochemical identiﬁcation .............................................................................................................................700

Molecular identiﬁcation................................................................................................................................700

Antibiotic Therapies ......................................................................................................................................701

RISK FACTORS, TRANSMISSION, AND ENVIRONMENTAL RESERVOIRS ...................................................................................701

Poultry....................................................................................................................................................702

Domesticated Animals ...................................................................................................................................702

Wild Animals .............................................................................................................................................703

Water .....................................................................................................................................................703

Other Sources ............................................................................................................................................703

IMPACT OF ANTIBIOTIC USAGE IN ANIMALS ON CAMPYLOBACTER RESISTANCE ........................................................................704

(continued)

Published 10 June 2015

Citation Kaakoush NO, Castaño-Rodríguez N, Mitchell HM, Man SM. 10 June 2015.

Global epidemiology of Campylobacter infection. Clin Microbiol Rev

doi:10.1128/CMR.00006-15.

Address correspondence to Si Ming Man, SiMing.Man@StJude.org.

N.O.K. and N.C.-R. contributed equally to this article.

doi:10.1128/CMR.00006-15

crossmark

July 2015 Volume 28 Number 3 cmr.asm.org 687Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

CONTROLLING THE SPREAD OF CAMPYLOBACTERIOSIS .................................................................................................704

Reducing Campylobacter Transmission in Chickens ......................................................................................................705

Bacteriocins ............................................................................................................................................705

Bacteriophages ........................................................................................................................................705

Probiotics ..............................................................................................................................................706

Vaccination ............................................................................................................................................706

Strategies aimed at the processing level...............................................................................................................706

CONCLUSIONS .............................................................................................................................................707

ACKNOWLEDGMENTS......................................................................................................................................707

REFERENCES ................................................................................................................................................707

AUTHOR BIOS ..............................................................................................................................................720

SUMMARY

Campylobacter jejuni infection is one of the most widespread in-

fectious diseases of the last century. The incidence and prevalence

of campylobacteriosis have increased in both developed and de-

veloping countries over the last 10 years. The dramatic increase in

North America, Europe, and Australia is alarming, and data from

parts of Africa, Asia, and the Middle East indicate that campylo-

bacteriosis is endemic in these areas, especially in children. In

addition to C. jejuni, there is increasing recognition of the clinical

importance of emerging Campylobacter species, including Cam-

pylobacter concisus and Campylobacter ureolyticus. Poultry is a ma-

jor reservoir and source of transmission of campylobacteriosis to

humans. Other risk factors include consumption of animal prod-

ucts and water, contact with animals, and international travel.

Strategic implementation of multifaceted biocontrol measures to

reduce the transmission of this group of pathogens is paramount

for public health. Overall, campylobacteriosis is still one of the

most important infectious diseases that is likely to challenge global

health in the years to come. This review provides a comprehensive

overview of the global epidemiology, transmission, and clinical

relevance of Campylobacter infection.

INTRODUCTION

Campylobacter species are Gram-negative spiral, rod-shaped, or

curved bacteria with a single polar ﬂagellum, bipolar ﬂagella,

or no ﬂagellum, depending on the species (1). Campylobacter spe-

cies are non-spore-forming, are approximately 0.2 to 0.8 by 0.5 to

5␮m, and are chemoorganotrophs which obtain their energy

sources from amino acids or tricarboxylic acid cycle intermediates

(2). Most Campylobacter species grow under microaerobic condi-

tions and have a respiratory type of metabolism; however, several

species (Campylobacter concisus,Campylobacter curvus,Campylo-

bacter rectus,Campylobacter mucosalis,Campylobacter showae,

Campylobacter gracilis, and, to a certain extent, Campylobacter hy-

ointestinalis) require hydrogen or formate as an electron donor for

microaerobic growth. In addition, certain species prefer anaerobic

conditions for growth.

The Campylobacter genus was established in 1963 following the

renaming of Vibrio fetus to Campylobacter fetus, forming the type

species of this genus (3). The Campylobacter genus belongs to the

family Campylobacteraceae, the order Campylobacterales, the class

Epsilonproteobacteria, and the phylum Proteobacteria. Since its

ﬁrst description, the genus has grown to include several important

human and animal pathogens that are primarily classiﬁed through

phylogenetic means. The genus Campylobacter consists of 26 spe-

cies, 2 provisional species, and 9 subspecies (as of December

2014).

Campylobacter jejuni is a major cause of gastroenteritis world-

wide. Moreover, C. jejuni infection may lead to autoimmune con-

ditions known as Guillain-Barré syndrome (GBS) and Miller

Fisher syndrome. Many Campylobacter species are known patho-

gens in humans and animals (1). In humans, Campylobacter spe-

cies have been associated with a range of gastrointestinal condi-

tions, including inﬂammatory bowel diseases (IBD), Barrett’s

esophagus, and colorectal cancer (Fig. 1)(

1). They have also been

reported to be involved in extragastrointestinal manifestations,

including bacteremia, lung infections, brain abscesses, meningitis,

and reactive arthritis, in individual cases and small cohorts of

patients (1). A full list of the clinical manifestations associated

with Campylobacter infection is presented in Table 1. The precise

role of Campylobacter species in the development of these clinical

conditions is largely unknown. In this review, we describe the

latest global epidemiological landscape of C. jejuni and other

Campylobacter species in gastroenteritis and other diseases. We

also discuss the modes of transmission and biocontrol methodol-

ogies to prevent transmission of campylobacteriosis.

GASTROENTERITIS

C. jejuni and C. coli are established causes of diarrhea in humans.

A human experimental infection study revealed that the rate of

colonization increased with increasing doses of C. jejuni, whereas

the development of illness did not (4). Infection with a dose as low

as 800 CFU resulted in diarrhea in some volunteers (4). However,

it has been speculated further that the dose of C. jejuni required for

the development of campylobacteriosis can be as low as 360 CFU

(5). Mathematical modeling suggested that an intermediate dose

of 9⫻10

CFU/ml has the highest ratio of illness to infection (6).

In contrast, an association between dose and occurrence of disease

was observed in humans experimentally infected with C. jejuni

strain 81-176. In addition, exposure to C. jejuni strain 81-176 of-

fered only short-term protection (7). This can be reconciled by the

fact that the severity of disease, dose-response relationship, and

illness/infection ratio are dependent, at least in part, on the strain

used. These strain-speciﬁc differences were clearly observed when

experimental infection of naive individuals with C. jejuni strain

CG8421 failed to offer protection against a second bout of cam-

pylobacteriosis upon rechallenge with the same strain (8). Inter-

estingly, an immunocompetent adult experimentally infected

with C. jejuni experienced recrudescence of the infection at the

conclusion of antibiotic therapy (9), suggesting that the incidence

of recurrent infection may be underestimated.

Patients with C. jejuni or C. coli infection experience acute

watery or bloody diarrhea, fever, weight loss, and cramps that last,

on average, 6 days (1). Gastroenteritis induced by C. coli is clini-

Kaakoush et al.

688 cmr.asm.org July 2015 Volume 28 Number 3Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

cally indistinguishable from that by C. jejuni. The onset of symp-

toms usually occurs 24 to 72 h following ingestion and may take

longer to develop in those infected with a low dose. The peak of

illness can last 24 to 48 h and may include abdominal pain that

mimics appendicitis (10). Polymorphonuclear leukocytes and

blood (gross or microscopic) can be observed in the stool, and

diffuse inﬂammatory colitis is present in colonic biopsy specimens

from infected patients (10). While infection with C. jejuni or C.

coli can occur in patients of all ages, a recent study from Denmark

showed that infection is more prevalent in toddlers (1 to 4 years)

and young adults (15 to 24 years) than in other age groups (11). A

recent comparison of the characteristics of patients infected with

C. jejuni or C. coli indicated that slightly older patients (34.6 years

versus 27.5 years) and those who traveled abroad were at a greater

risk of being infected with C. coli than with C. jejuni (12). Studies

have also shown that infections with C. jejuni and C. coli are more

common during the summer months (11,13). Although C. coli is

less prevalent than C. jejuni in many geographic regions, C. coli

infections can contribute as many as 25% of all gastroenteritis

cases caused by Campylobacter species (14–18).

C. concisus,Campylobacter ureolyticus,Campylobacter upsa-

liensis, and Campylobacter lari are known as “emerging Campylo-

bacter species,” a term used to describe their underappreciated

roles in human and animal diseases. Emerging Campylobacter

species are likely to contribute to the etiology of gastroenteritis,

especially in cases which have no known association with other

established pathogens (1,19–21). Furthermore, many diagnostic

laboratories fail to detect emerging Campylobacter species owing

to a lack of the specialized cultivation techniques required to cul-

ture these organisms, including the use of microaerobic or anaer-

obic conditions enriched with hydrogen (1). Indeed, introduction

of hydrogen as part of routine microaerobic culture for stool sam-

ples at a University Hospital in Bern, Germany, resulted in a sig-

niﬁcant increase in the rate of isolation of C. concisus (22). As a

consequence, the incidence of C. concisus detection rose from 0.03

to 1.92%.

Patients infected with C. concisus and certain Campylobacter

species other than C. jejuni and C. coli generally experience milder

symptoms, with fewer individuals reporting fever, chills, weight

loss, and mucus and blood in their stools than those infected with

C. jejuni and C. coli (1,23). In general, the milder severity of

symptoms has been found to correlate with the low levels of fecal

calprotectin in those infected with C. concisus (median, 53 mg/kg

of feces; interquartile range, 20 to 169 mg/kg). For comparison,

fecal calprotectin levels are higher (median, 631 mg/kg; interquar-

tile range, 221 to 1,274 mg/kg) in those infected with C. jejuni or C.

coli (24). Symptoms associated with C. concisus infection tend to

be more persistent than those of C. jejuni or C. coli infection, with

FIG 1 Environmental reservoirs, routes of transmission, and clinical manifestations associated with Campylobacter species. Campylobacter species can be

transmitted to humans through consumption of undercooked or contaminated food or via contact with animals. Tap, bore, and pond waters are also sourcesof

Campylobacter species. Person-to-person transmission (fecal-oral or via fomites) can occur. Ingestion of a sufﬁcient dose of organisms via the oral-gastric route

may lead to one or more gastrointestinal and/or extragastrointestinal manifestations; the outcome is dependent on the species or strains of Campylobacter

involved in the infection. Abbreviations: IBD, inﬂammatory bowel diseases; IBS, irritable bowel syndrome. Question marks indicate conditions for which a role

for Campylobacter is implicated but not certain.

Epidemiology of Campylobacter Infection

July 2015 Volume 28 Number 3 cmr.asm.org 689Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

80% of patients reporting diarrhea that lasted 14 days or more,

whereas only 32% of those infected with C. jejuni or C. coli re-

ported prolonged diarrhea (23). Similar to the case proﬁle for C.

concisus infection, infection with C. fetus is more common than

infection with C. jejuni and C. coli in older patients (68.4 years

versus 28.6 years) (12).

Epidemiology

There is evidence to suggest that there has been a rise in the global

incidence of campylobacteriosis in the past decade. The numbers

of cases of campylobacteriosis have increased in North America,

Europe, and Australia. Although epidemiological data from Af-

rica, Asia, and the Middle East are still incomplete, these data

indicate that Campylobacter infection is endemic in these regions.

Differences in the incidence and number of cases reported from

different countries or regions within the same country may vary

substantially (Fig. 2)(

25,26). It is likely that these variations arise,

in part, from differences in the sensitivity of detection methodol-

ogies and the area, population, and scope of the case proﬁle

studied, as well as differences in the standard and stringency of

biocontrol protocols, surveillance bias, food practices, and avail-

ability of natural reservoirs for Campylobacter species in these re-

gions. Furthermore, the reported cases of C. jejuni and C. coli

infections are likely to represent only the tip of the iceberg owing

to underreporting (27).

An additional factor that has been hypothesized to inﬂuence the

prevalence of Campylobacter infections is population-level immu-

nity (28). Population-level immunity refers to the host immune

response against an infection within a population that can provide

protection against transmission of an infection and/or disease for

unprotected individuals. At the population level, this can have

impacts on the epidemiology and risk assessment of campylobac-

teriosis (28). In developing countries where Campylobacter is en-

demic, infection is usually limited to children, with illness/infec-

tion ratios decreasing with age, suggesting that exposure in early

life might lead to the development of protective immunity (13).

This might reﬂect why asymptomatic Campylobacter infections

are common in developing nations, which could also have an im-

pact on the transmission of Campylobacter infections in these re-

gions due to asymptomatic excretion (28). Asymptomatic excre-

tion is also found in developed countries, with a number of studies

showing that a majority of shedders are asymptomatic (29,30).

Outbreaks caused by Campylobacter species are not uncom-

mon. The Centers for Disease Control and Prevention deﬁnes a

foodborne disease outbreak as the occurrence of more cases than

expected in a particular area or among a speciﬁc group of people

during a speciﬁc period, usually with a common cause. Since 2007,

the numbers of individuals reported in published outbreaks of

campylobacteriosis have ranged from 10 to more than 100 and

have correlated with the type of event and environmental source

of infection (Fig. 3). The most common reported sources of Cam-

pylobacter responsible for outbreaks are consumption of poultry

products or water (Fig. 3). Between 1992 and 2009, 143 outbreaks

were reported in England and Wales, United Kingdom. Of these,

114 were due to contaminated food or water, 2 to animal contact,

and 22 to an unknown mode of transmission (31). According to

records from the Centers for Disease Control and Prevention,

there were 4,936 Campylobacter outbreaks in the United States

between 1999 and 2008 (32). Between 2009 and 2010, 56 con-

ﬁrmed and 13 suspected outbreaks were reported to the U.S. Na-

tional Outbreak Reporting System, among which 1,550 illnesses

TABLE 1 Species within the genus Campylobacter and their clinical relevance to humans (as of December 2014)

Campylobacter species

Clinical manifestations

C. coli Established pathogen in gastroenteritis; also found in blood, meningitis, and acute cholecystitis

C. concisus Emerging pathogen associated with gastroenteritis and IBD (Crohn’s disease and ulcerative colitis); also found in Barrett’s

esophagitis, blood, and brain abscess

C. curvus Found in gastroenteritis, ulcerative colitis, Barrett’s esophagitis, blood, liver, and bronchial abscesses

C. fetus

Associated with bacteremia; also found in gastroenteritis, brain abscesses, epidural abscess aspirate, cerebrospinal ﬂuid, cellulitis,

endocarditis, mycotic aneurysm of the abdominal aorta, and peritonitis

C. gracilis Potential periodontal pathogen; also found in IBD, head and neck infection, and brain abscess

C. hominis Found in blood and IBD (possibly a commensal in the intestine)

C. helveticus Found in gastroenteritis

C. hyointestinalis Found in gastroenteritis and blood

C. insulaenigrae Found in gastroenteritis and blood

C. jejuni Established pathogen in gastroenteritis and possible predisposing agent in IBD, postinfectious IBS, and celiac disease; infection may

result in sequelae in the forms of Guillain-Barré syndrome, Miller Fisher syndrome, Bell’s palsy (unilateral facial paralysis), and

reactive arthritis; found in IBD, blood, myocarditis, meningitis, acute cholecystitis, urinary tract infection, and acute febrile

illnesses associated with leukopenia or thrombocytopenia

C. lari Associated with gastroenteritis; also found in blood

C. mucosalis Found in gastroenteritis

C. rectus Putative periodontal pathogen; also found in gastroenteritis, IBD, vertebral abscess, blood, necrotizing soft tissue infection, and pus

C. showae Found in IBD, intraorbital abscess, and blood

C. sputorum Found in gastroenteritis, axillary abscess, and blood

C. upsaliensis Emerging pathogen in gastroenteritis; also found in breast abscess, blood, and placenta

C. ureolyticus Associated with gastroenteritis and IBD; also found in oral, perianal, and soft tissue abscesses, soft tissue or bone infections, and

ulcers or gangrenous lesions of the lower limb

No disease association in humans has been reported for C. avium,C. canadensis,C. corcagiensis,C. cuniculorum,C. lanienae,C. lari subsp. concheus,C. peloridis,C. subantarcticus,

C. troglodytis,C. volucris,“Campylobacter sp. Dolphin DP,” and “Campylobacter sp. Prairie Dog” (as of December 2014).

Includes C. fetus subsp. fetus,C. fetus subsp. venerealis, and C. fetus subsp. testudinum.

The table was updated from the work of Man (1).

Kaakoush et al.

690 cmr.asm.org July 2015 Volume 28 Number 3Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

and 52 hospitalized cases were recorded (33). Unfortunately, out-

break data from developing nations are severely lacking. A full list

of published campylobacteriosis outbreaks since 2007 is shown in

Fig. 3. Below, we summarize the epidemiological data from differ-

ent regions of the world.

North and Central America. In the United States, the annual

number of campylobacteriosis cases, based on 10 years of out-

break data (1998 to 2008), was estimated to be 845,024 cases,

resulting in 8,463 hospitalizations and 76 deaths (32). The U.S.

Food-Borne Diseases Active Surveillance Network (1996 to 2012)

reported an annual incidence of 14.3 per 100,000 population for

Campylobacter infection (34). Batz and colleagues estimated the

annual costs of campylobacteriosis to be $1.7 billion in the United

States (35). Importantly, there was a 14% increase in the incidence

of campylobacteriosis in 2012 compared to the 2006 –2008 period,

whereas the incidences of Cryptosporidium,Listeria,Salmonella,

Shigella, Shiga-toxigenic Escherichia coli (STEC) O157, and Yer-

sinia infections decreased over the same period (34). Analysis of

seven states in the United States within the Food-Borne Diseases

Active Surveillance Network revealed that of nine common food-

borne pathogens, Campylobacter was the leading cause of travel-

associated gastroenteritis from 2004 to 2009, accounting for

41.7% of cases (3,445 of 8,270 cases reported), followed by Salmo-

nella (36.7%) and Shigella (13.0%) (36). The same study also

found that Campylobacter species were the second most prevalent

pathogens in non-travel-associated gastroenteritis, behind Salmo-

nella species, accounting for 26.5% of the 14,782 cases reported

between 2004 and 2009 (36).

In Quebec, Canada, 28,521 cases of campylobacteriosis were

reported between 1996 and 2006, which yielded an estimated an-

nual incidence of 35.2 cases per 100,000 persons (37). A higher

incidence of campylobacteriosis (49.69 cases per 100,000 people)

was reported from 1990 to 2004 in the Waterloo region of On-

tario, Canada (38). In southwestern Alberta, Canada, 36.9% and

5.4% of the patients with diarrhea reported from 31 May to 31

October 2005 were positive for C. jejuni and C. coli, respectively

(14). With a rate among the highest in the country, the province of

British Columbia had an annual average of 38.0 cases per 100,000

people during 2005 and 2009 (39).

In Mexico, C. jejuni was the most common cause of acute gas-

troenteritis in infants and preschoolers in 2006 and 2007 (isolated

in 15.7% of 5,459 cases) (40). Campylobacteriosis is also very

FIG 2 Incidence and prevalence of campylobacteriosis (C. jejuni/C. coli). The latest information on the global epidemiology of campylobacteriosis from the

literature is shown, including data from the United Kingdom (47), Denmark (11), Germany (49), Norway (424), Poland (25,50,425), the Netherlands (51), Israel

(67), China (60,61), Japan (26), India (63–65), Australia (69), New Zealand (73), Madagascar (78), Malawi (77), Kenya (79,426), Guatemala (41), Peru (427),

Mexico (428), the United States (10 sites within The Food-Borne Diseases Active Surveillance Network) (34), and Canada (37–39). B.C., British Columbia. (Map

adapted from an image from Wikimedia Commons [http://commons.wikimedia.org/wiki/File:A_large_blank_world_map_with_oceans_marked_in

_blue.PNG].)

Epidemiology of Campylobacter Infection

July 2015 Volume 28 Number 3 cmr.asm.org 691Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

common in children in Guatemala, with incidence rates of 185.5

to 1,288.8 per 100,000 children (41). One study estimated that the

incidence of campylobacteriosis in Barbados in 2000 was 5.4 per

100,000 inhabitants—an incidence which had doubled in 2002

(42). However, there is no further report describing the incidence

of human campylobacteriosis in this region. Overall, C. jejuni is a

major pathogen in the United States and Canada, but its precise

impact in regions of Central America is less clear.

South America. In 2011, Fernández reviewed the available data

on the prevalence of C. jejuni and C. coli in South America (43).

The prevalences of C. jejuni ranged from 4.6 to 30.1% of diarrheic

patients in Argentina (three studies), while those of C. coli were 0

to 1.4%, with C. coli being responsible for a third of the Campylo-

bacter-related gastroenteritis cases in one of these studies. How-

ever, none of these studies included a control group. In Bolivia, the

prevalences of C. jejuni ranged from 4.4 to 10.5%; however, of the

two studies cited, one included a control group in which C. jejuni

was detected in 9.6% of the controls. Similarly, two studies in

Brazil found that the levels of detection of C. jejuni were similar

between patients (5.8 to 9.6%) and controls (4.9 to 7.2%), while

those of C. coli ranged from 2.2 to 6.0% for patients and from 1.2

to 2.0% for controls. Detection levels of C. jejuni and C. coli in

gastroenteritis patients ranged from 0 to 14.1% in Chile, 0 to

14.4% in Colombia, 0 to 23.0% in Ecuador, 0.6 to 18.4% in Para-

guay, 0 to 23.0% in Peru, 0 to 14.3% in Uruguay, and 0 to 13.0% in

Venezuela (43).

In 2013, Collado and colleagues reported the detection of C.

jejuni and emerging Campylobacter species in patients with gas-

troenteritis from southern Chile (44). In this study, fecal samples

were collected from participants over the period from November

2010 to March 2012, and the presence of Campylobacter species

was detected by PCR. Of the 140 patients, 11.4% tested positive for

C. concisus DNA, compared to only 3.4% of the 116 healthy con-

trols (P⬍0.05) (44). Notably, the prevalence of C. concisus DNA

in gastroenteritis patients was found to be similar to that of C.

jejuni (10.7%). In Peru, a study that included 150 pediatric stool

samples from the Etiology, Risk Factors and Interactions of En-

teric Infections and Malnutrition and the Consequences for Child

Health and Development (MAL-ED) cohort study detected C.

jejuni/C. coli in 41.3% of the children with gastroenteritis and

18.7% of the controls (P⫽0.007) (45). In contrast, the difference

in the prevalences of other Campylobacter species, including C.

hyointestinalis subsp. lawsonii,C. troglodytis, and C. upsaliensis,in

children with gastroenteritis (33%) and in controls (24%) was not

statistically signiﬁcant (45). Taken together, the data from South

America indicate that Campylobacter species contribute to the eti-

ology of gastroenteritis, but the contribution of campylobacterio-

sis to this disease, relative to the contributions of other major

pathogens, is unclear.

Europe. The most current evaluation of the epidemiology of

campylobacteriosis in 27 European Union (EU) state members

indicated the incidences of Campylobacter infections to range

from 29.9 to 13,500 per 100,000 population in 2009 (with the

lowest incidences in Finland and Sweden and the highest in Bul-

FIG 3 Timeline of published campylobacteriosis outbreaks since 2007 (381,382,429–447).

Kaakoush et al.

692 cmr.asm.org July 2015 Volume 28 Number 3Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

garia) (46). Overall, this equated to 9.2 million cases, compared to

6.2 million cases of salmonellosis, in 2009 (46). A United King-

dom-wide study conducted over the period from April 2008 to

August 2009 identiﬁed Campylobacter species as the most com-

mon bacterial pathogens in cases of gastroenteritis (47). In this

study, the reported rate of campylobacteriosis was 9.3 cases per

1,000 person-years in the community, with an estimated total of

500,000 cases and 80,000 general practitioner consultations across

the United Kingdom annually (47). Investigation of the preva-

lence of pathogen-induced diarrhea in the United Kingdom in

2008 to 2009 revealed that the prevalence of Campylobacter species

had not decreased compared to that observed 15 years prior (21).

In contrast, the prevalences of enteroaggregative E. coli,Salmo-

nella, and Yersinia enterocolitica had all decreased over the same

15-year period (21).

Similarly, in Germany, the prevalence of campylobacteriosis in

2011 was similar to the data from 2001; in contrast, over the same

period, the prevalence of salmonellosis had decreased (48). In

2011, there were 70,560 reported cases of campylobacteriosis, a

prevalence higher than that reported for Salmonella,Shigella,Yer-

sinia, and Listeria infections (48). According to data from Hesse,

Germany, the annual incidences of campylobacteriosis between

2005 and 2011 ranged from 53.4 to 81.4 cases per 100,000 persons

(49). While in Poland the estimated incidence in 2012 was re-

ported to be 1.12 cases per 100,000 inhabitants, it is likely that

campylobacteriosis is underdiagnosed and underreported in this

region (25,50). Interestingly, a study from the Netherlands pre-

dicted that in 2060 the incidence of campylobacteriosis in this

region will be similar to that in 2011, at 51 per 100,000 population

(51). The same model estimated that salmonellosis would account

for only 12 cases per 100,000 persons in 2060. However, it is im-

portant that these estimates were calculated using age-speciﬁc de-

mographic forecasts for 10-year periods between 2020 and 2060,

without having considered other factors, such as social clustering,

health care systems, and food processing and preparation. Alto-

gether, the estimated cost of illness in the Netherlands due to

campylobacteriosis was €21 million per year (52). The high levels

of campylobacteriosis across Europe may be reﬂected in the con-

tinuous increase in the number of C. jejuni-related GBS cases in

Paris between 1996 and 2007 (mean annual increment, 7%; P⫽

0.007) (53).

Epidemiological data from Europe over the last 3 years have

transformed our understanding of the clinical importance of

emerging Campylobacter species in health and disease. Data col-

lected from the Netherlands between March and April 2011 re-

vealed that 71.4% of 493 gastroenteritis cases were PCR positive

for Campylobacter DNA, among which 20 were C. jejuni-associ-

ated cases (4.1%). In addition, a further subset of samples was

sequenced, which allowed the identiﬁcation of other Campylobac-

ter species, including C. concisus (4.1%), C. concisus or C. curvus

(0.8%), C. ureolyticus (0.6%), C. gracilis (0.6%), C. showae or C.

rectus (0.4%), C. upsaliensis (0.4%), C. hominis (0.2%), and C.

sputorum (0.2%) (54). These results suggest that, in the Nether-

lands, the prevalence of C. concisus in gastroenteritis is similar to

that of C. jejuni. A similar ﬁnding based on a culture-dependent

approach has been reported for Denmark, where the prevalences

of C. jejuni and C. concisus in adults and children were compara-

ble, with the annual incidence of C. concisus infection reported to

be 35 cases per 100,000 inhabitants in 2009 and 2010 (11,55). This

is in line with results from Portugal, where Campylobacter species

were detected in 31.9% of diarrheic fecal samples, with C. jejuni

and C. concisus being the most prevalent species (13.7% and 8.0%,

respectively) (56). Furthermore, based on PCR analysis, the prev-

alences of C. ureolyticus in gastroenteritis cases in Ireland in the

period from 2009 to 2012 ranged from 1.15 to 1.30% (57,58).

Furthermore, molecular screening for seven members of the Cam-

pylobacter genus by using PCR revealed the overall prevalence of

Campylobacter species in patients with gastroenteritis from south-

ern Ireland to be 4.7%, with C. jejuni being the predominant spe-

cies, accounting for 66% of all Campylobacter species detected,

followed by C. ureolyticus (22.3% of all Campylobacter species de-

tected), C. coli (6.7%), C. fetus (2.1%), C. hyointestinalis (1.3%), C.

upsaliensis (1.1%), and C. lari (0.5%) (59). Overall, there is com-

pelling evidence from Europe to suggest that in addition to C.

jejuni, emerging Campylobacter species contribute to the etiology

of gastroenteritis in this region.

Asia and the Middle East. Epidemiological data on campylo-

bacteriosis in Asia and the Middle East are limited. Investigation

of the etiology of gastroenteritis in three hospitals in Yangzhou,

China, between July 2005 and December 2006 showed that 4.84%

of 3,061 patients with diarrhea were PCR positive for C. jejuni,

with the highest prevalence being detected in those younger than 7

years of age (60). Between 2005 and 2009, 14.9% (142/950 pa-

tients) of patients with gastroenteritis in a hospital in Beijing,

China, were reported to be positive for Campylobacter species (127

with C. jejuni and 15 with C. coli)(

61). Based on detection rates of

Campylobacter in raw chicken and the consumption trend of

chicken products in China from 2007 to 2010, Wang and col-

leagues predicted that 1.6% of the urban and 0.37% of the rural

population are affected by campylobacteriosis every year (62). In

the Miyagi Prefecture of Japan, Kubota and colleagues estimated

that, from 2005 to 2006, the numbers of acute gastroenteritis ep-

isodes associated with Campylobacter,Salmonella, and Vibrio

parahaemolyticus infections were 1,512, 209, and 100 per 100,000

population per year, respectively (26), suggesting that Campylo-

bacter species are responsible for the majority of bacterial gastro-

enteritis cases in this region. The unusually high incidence of cam-

pylobacteriosis in this region may be attributed to a range of

factors, such as unexpected outbreaks during the time frame ex-

amined and/or the methodologies used to estimate the incidence

rate. Nevertheless, the high incidence of campylobacteriosis in

certain regions of Asia highlights the need for further active sur-

veillance of food safety.

In India, recent data from an infectious disease hospital in

Kolkata reported that, in the period from January 2008 to Decem-

ber 2010, 7.0% (222/3,186 patients) of hospitalized patients with

gastroenteritis were culture positive for Campylobacter species,

with 70% of the isolates identiﬁed as C. jejuni (63). Based on

real-time PCR analysis, Sinha and colleagues reported 16.2%

(11/68 samples) of diarrheic stool samples from patients in the

same region to be positive for Campylobacter species (64). Cam-

pylobacteriosis has also been reported to be most prevalent in

children under the age of 5 years in this region (isolated in 10% of

cases, compared to 3.7% for other age groups; P⬍0.001) (63). In

Vellore, South India, between January 2003 and May 2006, 4.5%

of 349 children under the age of 5 years with diarrhea were positive

by PCR for C. jejuni or C. coli (65). In addition, in a prospective

case-control study conducted between 1 December 2007 and 3

March 2011 to identify the etiology of diarrhea in children aged 0

to 59 months, C. jejuni was reported to be signiﬁcantly associated

Epidemiology of Campylobacter Infection

July 2015 Volume 28 Number 3 cmr.asm.org 693Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

with moderate to severe diarrhea in children from Kolkata, India,

Mirzapur, Bangladesh, and Karachi, Pakistan (66). However, a

further study that examined 144 Bangladeshi children failed to

identify a signiﬁcant association between Campylobacter species,

including C. jejuni/C. coli,C. troglodytis,C. hyointestinalis subsp.

lawsonii,C. concisus, and C. upsaliensis, and diarrhea (45).

The most recent data from the Middle East show Campylobac-

ter species to be a major and increasing cause of gastroenteritis in

this region. For example, the annual incidence of campylobacte-

riosis in Israel increased from 31.04 cases per 100,000 population

in 1999 to 90.99 cases per 100,000 population in 2010, with chil-

dren under the age of 2 years having the highest incidence (356.12

cases per 100,000 population) (67). Consistently, a study con-

ducted between 2007 and 2009 showed that of 99 hospitalized

children with gastroenteritis reported in a hospital in Nahariya,

Israel, 61% were positive for Campylobacter species, followed by

Shigella (24%) and Salmonella (16%) (68). It is difﬁcult to accu-

rately assess the burden of Campylobacter infections in Asia owing

to insufﬁcient epidemiological data. There is nevertheless a trend

indicating a possible rise in the incidence of campylobacteriosis in

the Middle East.

Oceania. Campylobacteriosis is the most commonly notiﬁed

foodborne infection in Australia, with 16,968 notiﬁed cases (112.3

cases per 100,000 cases of notiﬁed foodborne infection) in 2010

(69–71). In 2010, the prevalence of Campylobacter infections in-

creased 6% compared with the data from 2008 and 2009 (69–71).

Similarly, an increase in salmonellosis notiﬁcations was observed,

with 11,992 notiﬁcations (53.7 cases per 100,000 cases) in 2010,

compared to 9,533 notiﬁcations (43.6 cases per 100,000 cases) in

2009 and 8,310 notiﬁcations (39 cases per 100,000 cases) in 2008

(69–71). In contrast, Gibney and colleagues reported Campylo-

bacter to be the second leading cause of acute gastroenteritis in

Australia in 2010, after norovirus, with the highest disability-ad-

justed life-year burden (72).

In New Zealand, over the period from 2002 to 2006, the inci-

dence of campylobacteriosis was reported to be 353.8 cases per

100,000 population. However, in 2008, this high incidence

dropped substantially, to 161.5 cases per 100,000 population, due

to successful intervention strategies within the poultry sector in

this region (discussed further below) (73). Epidemiological data

available from 364 patients with campylobacteriosis from New

Zealand, obtained through telephone and postal questionnaires,

showed that 47% of the cases were due to consumption of con-

taminated food, 27.7% from direct contact with animals, 6.9%

from overseas travel, 3.3% from consumption of contaminated

water, and 11% from an unknown mode of transmission (74). A

study from New Zealand suggested that emerging Campylobacter

species are not associated with gastroenteritis cases (75). Among

the fecal samples collected in that study, between 2007 and 2009,

C. concisus (healthy controls, 53.1%; patients, 46.9%), C. ureolyti-

cus (healthy controls, 24.5%; patients, 10.9%), C. hominis (healthy

controls, 16.3%; patients, 8.6%), and C. gracilis (healthy controls,

6.1%; patients, 14.1%) were detected in samples from both pa-

tients and healthy controls, with similar frequencies (75). How-

ever, the authors concluded that given the level of genetic diversity

within these species, in particular C. concisus, the possibility that

they may play a role in disease cannot be ruled out.

Only one study has investigated the prevalence of Campylobac-

ter species in the regions of Oceania other than Australia and New

Zealand. Howard and colleagues reported the isolation of Campy-

lobacter species from 143/1,167 (12%) gastroenteritis cases, com-

pared to 20/660 (3%) controls, among children admitted to the

Goroka Base Hospital, Papua New Guinea, between October 1985

and March 1990 (76). More data are required to elucidate the

epidemiological landscape of campylobacteriosis in most Oceania

regions.

Africa. Data from a limited number of countries in Africa have

indicated that Campylobacter infection is most prevalent in the

pediatric population. A 10-year study (1997–2007) from Blantyre,

Malawi, Africa, found that C. jejuni and C. coli were detected in

21% (415/1,941 children) of hospitalized children with diarrhea

by real-time PCR, with C. jejuni accounting for 85% of all campy-

lobacteriosis cases (77). Between 1997 and 1999, nondiarrheic

children were also examined, and 14% were PCR positive for C.

jejuni and C. coli (77). Although this prevalence was signiﬁcantly

lower than that for children with diarrhea within the same period

(28%; P⬍0.001) (77), these observations indicate that C. jejuni

and C. coli are endemic in this pediatric population. These ﬁnd-

ings are supported by a study conducted in Moramanga, Mada-

gascar, where the rate of Campylobacter isolation from diarrheic

samples was reported to be 8.9% (41/459 samples), while that in

nondiarrheic samples was 9.4% (278/2,965 samples) (78). In Ke-

nya, samples collected from May 2011 to May 2012 at a hospital in

Kisii for the detection of a range of enteric pathogens showed 5.8%

(9/156 samples) of samples from patients with diarrhea to be cul-

ture positive for Campylobacter species, which was signiﬁcantly

higher than the incidence of 0.6% (1/156 samples) in the controls

(P⫽0.02) (79). A further study which described 138 Tanzanian

children and the use of diverse detection techniques, including

culture, enzyme immunoassay, and PCR, reported the detection

of C. jejuni/C. coli in 34.8% of gastroenteritis cases and 30.4%

of controls and other, non-jejuni/coli Campylobacter species in

47.8% of gastroenteritis cases and 42.0% of controls (45). These

differences, however, did not reach statistical signiﬁcance (45).

Pioneering work conducted by Lastovica and colleagues in South

Africa, who used culture methods optimal for the isolation of

most Campylobacter species, unveiled a more complete and real-

istic epidemiological landscape of C. jejuni and emerging Campy-

lobacter species. From 2005 to 2009, 5,443 strains of Campylobac-

ter species were isolated from stools of children with diarrhea at

the Red Cross Children’s Hospital in Cape Town. Of these, 40%

were C. jejuni (32.3% C. jejuni subsp. jejuni and 7.7% C. jejuni

subsp. doylei), while the second most prevalent organism was C.

concisus (24.6%) (80). From 1990 to 2009, Lastovica cultivated

more than 2,000 clinical isolates of C. concisus (81), a valuable

collection which could be characterized further by more extensive

genomic sequence analyses. Overall, it is not unreasonable to con-

clude that C. jejuni and other Campylobacter species are endemic

to children in most surveyed regions of Africa.

OTHER GASTROINTESTINAL MANIFESTATIONS

While gastroenteritis is a major clinical condition resulting

from Campylobacter infection, these organisms have also been

associated with a range of other serious conditions within the

gastrointestinal tract, including IBD, esophageal diseases, perio-

dontitis, functional gastrointestinal disorders, celiac disease, cho-

lecystitis, and colon cancer. Here we review the epidemiology and

impact of Campylobacter infection in these gastrointestinal dis-

eases.

Kaakoush et al.

694 cmr.asm.org July 2015 Volume 28 Number 3Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

Inﬂammatory Bowel Diseases

IBD are chronic inﬂammatory conditions of the gastrointestinal

tract which include Crohn’s disease (CD) and ulcerative colitis

(UC). The phenotype in patients with CD is characterized by

transmural lesions that may occur in any site along the gastroin-

testinal tract, while patients with UC are affected by continuous

submucosal inﬂammation restricted to the colon. Despite exten-

sive research, the etiology of IBD has yet to be elucidated; however,

the general hypothesis is that they are complex diseases in which a

dysregulated immune response that leads to chronic inﬂamma-

tion arises as a result of a dysregulated gastrointestinal microbial

ecology, host genetic factors, and a disruption of the gastrointes-

tinal epithelium triggered by environmental factors (82).

The role of Campylobacter species in IBD has been investigated

for the past 3 decades. C. jejuni was the initial focus of research

(83–85), but it was not until 2009 that Gradel and colleagues pro-

vided evidence that indicated an association between C. jejuni in-

fection and an increased risk of IBD (86). Furthermore, recent

studies investigating the role of other emerging Campylobacter

species in IBD have provided solid evidence that demonstrates an

association between C. concisus and these gastrointestinal disor-

ders (81,87–94).

The association between emerging Campylobacter species (C.

concisus,C. showae,C. hominis,C. gracilis,C. rectus, and C. ureo-

lyticus) and CD was ﬁrst described by the Mitchell group in 2009

(87). For a cohort of newly diagnosed pediatric CD patients, 82%

of intestinal biopsy specimens were found to be positive for Cam-

pylobacter DNA by PCR, compared to 23% of control samples

(87). Only the prevalence of C. concisus DNA was found to be

signiﬁcantly higher in patients with CD (51%) than in controls

(2%) (P⬍0.0001) (87). Consistent with this, a study by Tankovic

and colleagues found that C. concisus was present in 21% (4/19

patients) of IBD patients but only 9% (1/11 controls) of controls

(88). In 2010, Man and colleagues further reported that 65% of

fecal samples from patients with CD were positive for C. concisus,

compared to 33% of samples from healthy controls and 37% of

samples from non-IBD controls, and the differences were statisti-

cally signiﬁcant (P⫽0.03 and P⫽0.008, respectively) (89). Fur-

ther analyses to investigate the fecal microbiota in a subset of these

patients by using pyrosequencing techniques detected C. concisus

in two CD samples but not in controls, which indicates that C.

concisus DNA was present in sufﬁcient quantity to be detected by

less sensitive approaches (90). The increased prevalence of C. con-

cisus DNA in CD patients compared to controls appears to be

speciﬁc to the intestinal tract, because no difference in prevalence

of C. concisus DNA was found for saliva samples from IBD patients

(100%; 13 CD patients and 5 UC patients) and healthy controls

(97%; 57/59 controls) (95). This raises the possibility that the oral

cavity may be a natural reservoir for C. concisus.

In a study to investigate whether speciﬁc microorganisms were

selectively transported to the lymph nodes of CD patients, O’Brien

and colleagues used high-throughput sequencing and reported

Campylobacteraceae DNA to be present in three CD patients (96).

In line with this, a study by Kovach and colleagues identiﬁed 37

immunoreactive proteins of C. concisus, detected using sera col-

lected from 10 C. concisus-positive children with CD (97). Of these

proteins, ﬂagellin B, the ATP synthase F1 ␣subunit, and outer

membrane protein 18 were consistently recognized by all CD pa-

tients (97).

Similarly, the prevalence of C. concisus was reported to be higher

in patients with UC (91–94). For example, Mahendran and col-

leagues found a higher prevalence of C. concisus, not only in co-

lonic biopsy specimens from adult CD patients (53%; 8/15 pa-

tients) but also in those from UC patients (31%; 4/13 patients),

than in controls (18%; 6/33 individuals) (P⬍0.05) (91). Two

further studies, conducted in Scotland, showed an increased prev-

alence of C. concisus DNA in both adults and children presenting

with UC (92,93). The ﬁrst study isolated C. concisus from three

children with IBD (two with CD and one with UC) but not from

any of the controls; however, based on PCR, the prevalences of C.

concisus were not signiﬁcantly different between patients and con-

trols (92). In contrast, the second study detected a signiﬁcantly

higher prevalence of C. concisus DNA (33.3%; 23/69 samples) in

intestinal biopsy specimens from adult UC patients than in con-

trols (10.8%; 7/65 individuals) (P⫽0.0019) (93). More recently,

Rajilic-Stojanovic and colleagues examined the fecal microbiota

of 15 UC patients during remission and 15 controls, using a highly

reproducible phylogenetic microarray assay that can detect and

quantify more than 1,000 intestinal bacteria in a wide dynamic

range. This showed the levels of Campylobacter and other patho-

gens (Fusobacterium,Peptostreptococcus, and Helicobacter)tobe

increased in the fecal samples from UC patients compared with

those in controls (P⫽0.0004) (94). The reason for the increased

level of Campylobacter species during the remission stage of UC is

unclear, and the identity of the Campylobacter species is unknown,

as only the genus information was provided (94).

Despite solid evidence supporting an association between C.

concisus and IBD, the observation that C. concisus is detected in the

intestines of one-third of cohorts without IBD raises the possibil-

ity that C. concisus may simply be present as a result of dysbiosis

and intestinal inﬂammation (1,98). Although causality between

C. concisus and IBD has not yet been established, recent studies

have focused on identifying speciﬁc genetic variants of C. concisus

or genomospecies that may be associated with disease. C. concisus

is a genetically heterogeneous species which is deﬁned by 2 to 4

genetically variable genomospecies (99–105). A recent study by

our group addressed this issue by determining the levels of C.

concisus exotoxin 9/DnaI, a putative virulence factor postulated to

be associated with increased survival in the cell, in patients with

CD (106–108). This showed exotoxin 9/DnaI levels to be signiﬁ-

cantly higher in fecal samples from CD patients [48.8 ⫾20.7 pg

(g feces)

⫺1

] than in controls [4.3 ⫾1.1 pg (g feces)

⫺1

](P⫽0.037).

Based on these ﬁndings, it is possible that IBD patients are colo-

nized by strains harboring speciﬁc virulence factors (106). Fur-

thermore, our group also identiﬁed the zonula occludens toxin

(Zot) gene within the genomes of some C. concisus strains (109),

and we showed that the levels of Zot gene DNA in patients with

moderate to severe CD were increased compared to those for mild

CD [for mild CD, 1.6 ⫾0.7 pg (g feces)

⫺1

; and for moderate/

severe CD, 4.4 ⫾1.0 pg (g feces)

⫺1

](P⫽0.059) (110). Based on

these and other ﬁndings on the pathogenicity and immunogenic-

ity of C. concisus (108,111,112), we hypothesized that C. concisus

strains can be subdivided into the following two pathotypes which

differ from nonpathogenic strains: (i) adherent and invasive C.

concisus (AICC), which possesses a superior ability to survive in-

tracellularly within host cells (potentially involving exotoxin

9/DnaI and other virulence factors); and (ii) toxigenic C. concisus

(AToCC), which produces Zot, with the potential to target tight

Epidemiology of Campylobacter Infection

July 2015 Volume 28 Number 3 cmr.asm.org 695Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

junctions of host cells (98). Further characterization of the prev-

alence of these strains in IBD is required.

In addition to C. concisus, Mukhopadhya and colleagues found

21.7% (15/69 samples) of samples from UC patients and 3.1%

(2/65 samples) of samples from controls (P⫽0.0013) to be PCR

positive for C. ureolyticus (93). Overall, further investigations are

required to establish a causative role; nevertheless, these ﬁndings

collectively indicate an important contribution of C. concisus to

the pathogenesis of IBD.

Esophageal Diseases

Esophageal diseases include gastroesophageal reﬂux disease

(GERD), Barrett’s esophagus (BE), and esophageal adenocarci-

noma. GERD is a chronic disorder in which mucosal damage to

the esophagus occurs due to stomach acid or, occasionally, stom-

ach content, which may contain bile, ﬂowing back into the esoph-

agus, which over time increases the risk of BE. In turn, BE is a

preneoplastic condition deﬁned by the replacement of normal

squamous mucosa by metaplastic columnar mucosa in the distal

esophagus. This event increases the predisposition to the develop-

ment of esophageal adenocarcinoma.

Early studies have reported the bacterial composition to differ

in individuals with a healthy esophagus, GERD, and BE. The bac-

terial communities detected in these sites are primarily character-

ized by members of four phyla: the Actinobacteria,Bacteroidetes,

Firmicutes, and Proteobacteria (113–118). Recent studies have

demonstrated that Campylobacter species, and C. concisus in par-

ticular, are among the dominant species present in patients with

GERD and BE (113,119). For example, in a study by Macfarlane

and colleagues which examined the presence of aerobic, mi-

croaerobic, and anaerobic microorganisms in esophageal aspi-

rates and mucosal samples from patients with BE, 57% of patients

were reported to be colonized by Campylobacter species, the ma-

jority of which were C. concisus (113). In agreement with these

ﬁndings, Blackett and colleagues reported Campylobacter species,

almost exclusively C. concisus, to be increased in patients with

GERD and BE, but not in esophageal adenocarcinoma patients,

compared with healthy controls (119). This ﬁnding suggests a

possible association between C. concisus colonization and reﬂux

into the esophagus. Furthermore, the authors showed a strong

correlation between C. concisus colonization and production of

interleukin-18 (IL-18) (119), a cytokine that stimulates both in-

nate and adaptive immune responses and has been widely associ-

ated with carcinogenesis (120).

Periodontal Diseases

C. rectus,C. gracilis,C. showae, and C. concisus have been identiﬁed

as potential oral pathogens, while other Campylobacter species,

including C. curvus,C. sputorum, and C. ureolyticus, have been

isolated from the oral cavity; however, it remains unclear if they

are linked to periodontal disease (1,121–129). Gingivitis is a pre-

ventable and reversible clinical condition that includes erythema,

edema, bleeding, sensitivity, tenderness, and enlargement. Perio-

dontitis is a more severe condition characterized by a loss of clin-

ical attachment level, reduction in bone level, and, ultimately,

tooth loss. These oral inﬂammatory conditions are induced by

bioﬁlms that accumulate in the gingival margin and are reported

to be initiated in periodontal tissue by a number of bacterial spe-

cies, including C. rectus (121).

Anumber of studies have shown C. rectus to be associated with

higher levels of clinical attachment loss, bleeding on probing of

the sampled site, and probing depth (121,130–133). Furthermore,

the abundance of C. rectus has been reported to be elevated signif-

icantly in patients with chronic gingivitis and moderate periodon-

titis but not in severe periodontitis patients, suggesting that this

organism is associated with the early stages of periodontitis (130,

134). C. gracilis has been isolated from the oral cavities of individ-

uals presenting with dental caries (135–137), and its reported co-

aggregation with Actinomyces species has led to the suggestion that

these Gram-negative obligate anaerobic rods contribute to the

development of bioﬁlms, dental plaque, and root caries (138).

Evidence that C. showae and C. concisus may also play a role in

periodontal disease has been reported in a number of studies. For

example, an increased prevalence and abundance of both species

have been observed at active periodontal disease sites (121,128,

129,139–141). Furthermore, the ﬁndings that substantially in-

creased levels of C. concisus are observed at periodontal sites with

a more severe gingival bleeding index and that the presence of a

systemic humoral immune response against C. concisus can be

observed in patients with periodontal disease support the view

that this species is an oral pathogen (142–144). In contrast, one

study observed reduced levels of C. showae in plaque from white-

spot or dentin lesions of patients with periodontal disease com-

pared with the levels in healthy subjects (135), while another re-

ported that C. concisus was associated with an increase in tooth

attachment in a patient with periodontitis (145).

It has been speculated that oral colonization by Campylobacter

species may lead to other pathological consequences in the body.

For example, a study by Ercan and colleagues suggests that the

presence of C. rectus in the oral cavity of pregnant women with

periodontal disease may lead to adverse pregnancy outcomes, in-

cluding preterm birth and low birth weight (146). Furthermore,

this observation was reported to be more pronounced in those

with generalized periodontitis and a high bleeding index. Whether

this phenomenon relates to the ability of bacteria and their prod-

ucts to diffuse more readily when vascular permeability increases

in gingival tissues during pregnancy remains to be determined. A

small number of studies have investigated a possible etiological

association between oral Campylobacter species and IBD. It is in-

teresting that periodontal disease and IBD share some common

clinical features and are both associated with an unusual microbi-

ota. For example, levels of C. gracilis have been reported to be

signiﬁcantly higher at periodontitis sites of patients with CD than

at those of patients with UC or than the levels in healthy controls

(147). In addition, elevated levels of C. concisus can be found in

periodontal lesions of IBD patients, and the oral cavity in these

patients may be colonized by speciﬁc orally afﬁliated and enteri-

cally invasive C. concisus strains (148,149).

Functional Gastrointestinal Disorders

C. jejuni and other Campylobacter species are associated with the

development of foodborne gastroenteritis-associated sequelae, in-

cluding postinfectious functional gastrointestinal disorders

(PFGD). Two PFGD have received the most attention: irritable

bowel syndrome (IBS) (150–156) and functional dyspepsia (FD)

(156–161). IBS is deﬁned by recurrent abdominal pain or discom-

fort during at least 3 days/month in the last 3 months, associated

with an alteration of bowel habits (diarrhea, constipation, or

both), in accordance with the Rome III classiﬁcation system (162).

FD is characterized by persistent or recurrent symptoms (pain or

Kaakoush et al.

696 cmr.asm.org July 2015 Volume 28 Number 3Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

discomfort centered in the upper abdomen) in the last 3 months,

in the absence of organic disease (including on upper endoscopy).

The postinfectious forms of these disorders develop de novo de-

spite clearance of the causative agent.

The mechanisms underlying postinfectious IBS are poorly un-

derstood but might include persistent changes in the gut micro-

biota as well as in mucosal immunocytes, enterochromafﬁn cells,

mast cells, and enteric nerves (163). In addition, host factors, in-

cluding female gender, depression, hypochondriasis, smoking,

adverse life events in the preceding 3 months, and treatment with

antibiotics, are risk factors for the development of postinfectious

IBS (163).

The percentages of individuals presenting with gastroenteritis

who develop postinfectious IBS range from 3.7% to 36% (163);

however, studies exclusively investigating C. jejuni-associated

postinfectious IBS showed percentages ranging from 9.0 to 13.8%

(152,164). Furthermore, long-term follow-up studies have re-

vealed that C. jejuni-associated postinfectious IBS symptoms can

persist for up to 10 years after the infectious event (165,166).

Current evidence suggests that both bacterial and host factors play

a crucial role in the predisposition to C. jejuni-associated postin-

fectious IBS. These include increased cytotoxic virulence of the

Campylobacter strain and increased transcellular bacterial trans-

location, a reduced absorptive capacity of the gut, and increased

mucosal permeability in the host during acute gastroenteritis

(167). Experimental evidence that Campylobacter toxins are im-

portant determinants for the development of chronic gastrointes-

tinal symptoms following acute gastroenteritis comes from a

study by Thornley and colleagues, who observed that Campylo-

bacter strains associated with postinfectious IBS were more toxi-

genic to both HEp-2 and African green monkey kidney epithelial

(Vero) cells (164). Further studies have shown that C. jejuni infec-

tion is the strongest risk factor for postinfectious IBS compared to

Salmonella and Epstein-Barr virus infections (150,151).

More recently, other Campylobacter species have been identi-

ﬁed to play a role in postinfectious IBS. Nielsen and colleagues

reported patients infected with C. jejuni,C. coli, and C. concisus to

be more likely to develop IBS symptoms at 6 months postinfection

(23). In a follow-up study, they assessed the risk of postinfectious

IBS associated with C. concisus and found that patients with gas-

troenteritis associated with C. concisus carried a 25% risk of devel-

oping IBS (168).

Similar to the case for IBS, a number of studies have provided

evidence for an association between Campylobacter infection and

a risk of postinfectious FD. A meta-analysis of 19 studies found

that following infections with several pathogens, including C. je-

juni,Salmonella spp., Escherichia coli O157, Giardia lamblia, and

norovirus, the prevalences of postinfectious FD were 9.6 and

30.5% in adults and children, respectively (157). Consistent with

these ﬁndings, Ford and colleagues reported the odds ratio (OR)

of postinfectious FD to be 2.30 (95% conﬁdence interval [CI], 1.63

to 3.26) in a cohort study following a waterborne outbreak of

infections with Campylobacter species and E. coli O157 (158). In

addition, a recent study by Porter and colleagues reported the

relative risk of Campylobacter-associated FD among active-duty

U.S. military personnel with acute gastroenteritis from 1998 to

2009 to be 2.0 (95% CI, 1.3 to 3.0) (159). Of particular interest,

Campylobacter species and E. coli O157 can be identiﬁed in blood

tests and stool cultures from postinfectious FD patients (157).

Overall, there is good evidence to suggest a link between Campy-

lobacter infection and IBS or FD.

Colorectal Cancer

Increasing evidence indicates that dysbiosis of the gut microbiota

contributes to the development of colorectal cancer. Currently,

due to a lack of epidemiological studies, evidence supporting a

role for Campylobacter species in colorectal cancer is very limited.

However, a recent study by Warren and colleagues, investigating

metatranscriptome data obtained from colorectal cancer and con-

trol tissues, demonstrated that Campylobacter species, predomi-

nantly C. showae, coaggregate with Fusobacterium and Leptotrichia

species (169). This ﬁnding is of particular interest because previ-

ous studies have shown that Fusobacterium species are overrepre-

sented in colorectal tumors compared to control specimens (170,

171). As part of their study, Warren and colleagues isolated a novel

C. showae strain (CC57C) from colorectal cancer tissue, which

they showed harbored a number of potential virulence genes, in-

cluding a VirB10/D4 type IV secretion system. Furthermore, in

vitro assays showed that this strain aggregates with another tumor

strain of Fusobacterium nucleatum (CC53) (169). Based on these

ﬁndings, Warren and colleagues raised the possibility that a

Gram-negative anaerobic bacterial population comprising Cam-

pylobacter and Fusobacterium might be associated with colorectal

cancer (169). Consistent with this, a study by Wu and colleagues

(172) which used culture-independent pyrosequencing and re-

verse transcription-quantitative PCR (RT-qPCR) reported a spe-

ciﬁc microbial proﬁle, characterized by signiﬁcant increases in

Bacteroides,Enterococcaceae,Fusobacterium, and Campylobacter

species, to be associated with colorectal cancer. Although these

studies provide an indication that Campylobacter species are pres-

ent in patients with colorectal cancer, further studies will be re-

quired to determine if any relationship between Campylobacter

and the development of colorectal cancer exists.

Celiac Disease

Celiac disease is a digestive disorder in which the immune system

reacts abnormally to gluten, resulting in damage to the lining of

the small intestine. Celiac disease is estimated to affect approxi-

mately 1% of people worldwide (173,174), and the incidence of

this disease has increased up to 5-fold in some countries, including

the United States (175). In 2007, a case study described for the ﬁrst

time the development of celiac disease in a young healthy woman

following infection with C. jejuni (176). More recently, Riddle and

colleagues reviewed the U.S. Department of Defense medical en-

counter database to identify whether there is a risk of developing

celiac disease following foodborne infection (177). They found

that the rates of celiac disease were similar in those with prior

diarrhea caused by bacteria (0.07 per 100,000 person-years) and

matched controls (0.04 per 100,000 person-years). However, per-

sons diagnosed with campylobacteriosis had a 3.5-fold higher rate

of celiac disease (0.15 per 100,000 person-years) than unexposed

individuals (177). Furthermore, no patients exposed to any of the

other gastrointestinal pathogens, including Salmonella (nonty-

phoidal), Shigella, and Y. enterocolitica, developed celiac disease

(177). To date, there are insufﬁcient epidemiological data to con-

clude whether campylobacteriosis is associated with celiac disease.

Cholecystitis

Cholecystitis refers to inﬂammation of the gallbladder that usually

arises when the cystic duct is blocked by gallstones, leading to the

Epidemiology of Campylobacter Infection

July 2015 Volume 28 Number 3 cmr.asm.org 697Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

accumulation of bile within the gallbladder. C. jejuni has been

implicated in the development of cholecystitis; however, this is

considered rare given that only 15 cases have been described in the

literature over the last 30 years (178). One possible reason for this

is that the standard conditions used for culture of bacteria from

bile samples do not generally favor the growth of Campylobacter

species; thus, some cases of Campylobacter-associated cholecysti-

tis may have been overlooked (178).

EXTRAGASTROINTESTINAL MANIFESTATIONS

In addition to gastrointestinal infection, Campylobacter species

also cause a range of clinical manifestations in other parts of the

body, as either a local isolated infection, a systemic manifestation

after an episode of enteritis, or a postinfectious immune disorder.

These manifestations include Guillain-Barré syndrome, Miller

Fisher syndrome, brain abscesses and meningitis, bacteremia, sep-

sis, endocarditis and myocarditis, reactive arthritis, and clinical

manifestations that result in complications in the reproductive

tract. The clinical importance and epidemiology of these extragas-

trointestinal manifestations as a result of Campylobacter infection

are discussed in the following sections.

Guillain-Barré Syndrome

GBS was ﬁrst reported by Landry in 1859 (179,180); however, it

was not until 1916 that the French neurologists Guillain, Barré,

and Strohl ﬁrst described the clinical features of GBS (181). GBS is

a neurologic condition characterized by a progressive symmetrical

weakness in the limbs, with or without hyporeﬂexia, which can

also affect respiratory and cranial nerve-innervated muscles (180).

The two main subtypes of GBS are acute motor axonal neuropathy

(AMAN) and acute inﬂammatory demyelinating polyneuropathy

(AIDP), with each subtype displaying a distinct immunopatho-

genesis and response to treatment (182). AMAN is an axonal sub-

type that progresses more rapidly and is considered the major

subtype (30 to 65% of patients) in Asia and Central and South

America (182,183), while AIDP is more prevalent in Europe and

North America (182). The annual incidence of GBS is approxi-

mately 1.2 to 2.3 cases per 100,000 persons; however, the incidence

increases with patient age and male gender (180).

GBS is considered a postinfectious disease, and the major trig-

ger of this disease is C. jejuni infection, with a direct correlation

between annual rates of GBS and campylobacteriosis being re-

ported. For example, following the implementation of stricter hy-

giene measures on poultry meat in New Zealand, decreased rates

of GBS were observed, which correlated with a fall in the number

of campylobacteriosis cases (184). Furthermore, outbreaks of GBS

have been associated with outbreaks of C. jejuni infection (185). A

recent systematic review reported the proportion of Campylobac-

ter cases resulting in GBS to be 0.07% (95% CI, 0.03% to 0.15%)

(186). Moreover, evidence exists to suggest that the number of C.

jejuni-related GBS cases is increasing in some countries (53).

The underlying mechanism of the nerve damage associated

with GBS is reported to be due to cross-reactivity between anti-

bodies produced in response to C. jejuni lipooligosaccharide

(LOS) and human gangliosides, such as the GM

ganglioside

(187). For example, C. jejuni strains expressing ␣2,3-sialylated

GD1a/GM1a- and ␣2,8-sialylated GD1c-mimic LOS structures

have been shown to interact with sialoadhesin and sialic acid-

binding immunoglobulin-like lectin-7 (Siglec-7), respectively

(188,189). T-cell responses are also important in GBS, with LOS

that was ␣2,8-sialylated being shown to induce Th1 immune re-

sponses, while LOS containing ␣2,3-linked sialic acid induces Th2

responses (188). Evidence of this Th1/Th2 polarized immune re-

sponse following C. jejuni infection comes from a study by Malik

and colleagues, who showed that IL-10-deﬁcient mice infected

with colitogenic C. jejuni had upregulated Th1/17 but not Th2

responses, while GBS-associated C. jejuni enhanced Th2 responses

but blunted Th1/17 responses (190). Thus, the GBS-associated C.

jejuni strains may protect against colitis but instead promote au-

toimmunity. In addition to sialylated LOS structures in C. jejuni,

two capsule biosynthesis genes (cj1421c and cj1428c) have been

shown to have higher conservation rates among strains isolated from

GBS patients than among strains from enteritis patients (191). Fur-

thermore, a gene encoding a glucosyltransferase (cj1135) was re-

ported to be more conserved in enteritis strains whose LOS did not

mimic gangliosides, suggesting that this gene may function to silence

the neuropathogenesis of C. jejuni LOS (191).

Miller Fisher Syndrome

Miller Fisher syndrome is a clinical variant of GBS that was dis-

covered in 1956 by Charles Miller Fisher. This condition is deﬁned

by acute-onset ophthalmoparesis, areﬂexia, and ataxia, which

arise from the development of anti-GQ1b antibodies following

exposure to LOS from certain bacteria (192,193). Several patho-

gens have been linked to the molecular mimicry that leads to the

development of this condition. Of these, C. jejuni is the most fre-

quently identiﬁed one (194). Siglec-7 has been shown to exclu-

sively bind to C. jejuni strains that express terminal disialylated

ganglioside mimics. C. jejuni binding to Siglec-7 is reported to

correlate with the presence of anti-GQ1b antibodies and oculo-

motor weakness in patients, suggesting that this may be a potential

trigger of Miller Fisher syndrome (195).

In some cases, patients are negative for antibodies against GQ1b

(196,197), indicating that antibodies against gangliosides other

than GQ1b may be involved in the development of disease. For

example, Oyazato and colleagues reported a recent atypical per-

sistent case of Miller Fisher syndrome following Campylobacter

enterocolitis where the patient had anti-GA1 antibody in his se-

rum, but not anti-GQ1b and anti-GT1a (198). Further work is

required to investigate the etiology and epidemiology of Campy-

lobacter-associated Miller Fisher syndrome and the ability of Cam-

pylobacter species other than C. jejuni to trigger this disease.

Bacteremia and Septicemia

One of the most common extragastrointestinal manifestations of

Campylobacter species is bacteremia, which is predominantly as-

sociated with C. jejuni,C. coli, and C. fetus infections (1). At least

10 different Campylobacter species have been documented in bac-

teremia cases, but bacteremia cases associated with C. lari,C. in-

sulaenigrae, and C. upsaliensis infections are rare (1,80,199,200).

Campylobacter-associated bacteremia cases are often underre-

ported (201). Most cases occur in elderly or immunocompro-

mised patients with one or more concurrent pathologies, includ-

ing liver cirrhosis or neoplasia; among these patients, 10 to 15%

die within 30 days of disease diagnosis (202–204). A recent study

conducted in a Danish population reported the estimated inci-

dence of C. jejuni-,C. coli-,C. fetus-, and C. lari-associated bacte-

remia to be 2.9 cases per 1,000,000 person-years, with a peak in-

cidence in patients older than 80 years (205). In contrast, a 10-year

nationwide study in Finland concluded that C. jejuni and C. coli

Kaakoush et al.

698 cmr.asm.org July 2015 Volume 28 Number 3Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

bacteremias affect predominately younger individuals without

major underlying diseases (206). The disease generally results

from a single gastroenteritis complication in children or as recur-

rent episodes in immunocompromised children without gastro-

intestinal symptoms (207). A number of Campylobacter species,

including C. jejuni,C. coli,C. fetus, and C. upsaliensis, have also

been associated with sepsis in both immunocompetent and im-

munocompromised children and adults (208–210). Furthermore,

anumber of cases of C. fetus-associated neonatal sepsis have been

reported (discussed below) (211).

Cardiovascular Complications

Campylobacter species, mainly C. jejuni and C. fetus, have been

detected in association with a wide spectrum of cardiovascular

complications, including endocarditis, myocarditis, pericarditis,

myopericarditis (pericarditis with concurrent myocardial in-

volvement), atrial ﬁbrillation, and aortitis with aortic dissection.

Myo(peri)carditis (refers to either myocarditis, pericarditis, or

myopericarditis) associated with bacterial enteritis is a rare but

serious condition in immunocompetent individuals. These con-

ditions can lead to arrhythmia, dilated cardiomyopathy, conges-

tive heart failure, and sudden cardiac death. Salmonella and Shi-

gella are the main gastrointestinal pathogens linked to myo(peri)

carditis, but the increasing incidence of campylobacteriosis

worldwide over the last 10 years has drawn attention to Campylo-

bacter-associated myo(peri)carditis. Interestingly, Becker and col-

leagues recently calculated the annual incidence rates of myocar-

ditis, using data obtained from 6,204 individuals with stool

cultures positive for Campylobacter and 62,040 matched controls,

and they found the incidence rate to be 16.1 cases (95% CI, 2.3 to

114.4 cases) per 100,000 person-years in the population with

Campylobacter-positive stool samples, compared to only 1.6 cases

(95% CI, 0.2 to 11.4 cases) per 100,000 person-years in the control

population (212). Notably, the same authors did not ﬁnd a signif-

icant difference in the number of myopericarditis cases between

the Campylobacter-infected and control populations (212). How-

ever, because of the rarity of these events (only two cases of myo-

carditis and two cases of pericarditis were found in the entire study

sample), it is important to highlight the low statistical precision of

this study.

A number of C. jejuni-associated myo(peri)carditis cases have

been published since 1980 (213–230). These reports collectively

indicate that patients usually present with symptoms such as tho-

racic pain, with concomitant electrocardiogram changes as well as

increased levels of cardiac enzymes, 3 to 5 days after the onset of

gastroenteritis. These studies also suggest that males are more sus-

ceptible, and microbiological stool cultures and/or serological

analyses in these cases reveal C. jejuni as the only causative agent.

In many cases, blood cultures remain sterile and the outcome is

generally benign.

About 11 cases of C. fetus-associated myo(peri)carditis have

been documented in the English literature. For these cases, blood

cultures are generally positive and the outcome is more severe,

sometimes even leading to death (225,231). This might relate to

the fact that, in contrast to C. jejuni,C. fetus tends to cause a more

severe compromise of the pericardium, which might be due to

colonization of the pericardium following bacteremia and septi-

cemia and might explain the development of nonspeciﬁc symp-

toms, such as fever, malaise, and weight loss (232).

The mechanism by which Campylobacter species cause myo-

(peri)carditis remains unclear. It has been hypothesized that inva-

sion of the cardiac tissue, bacterial exotoxins, circulating immune

complexes, and cytotoxic T cells are involved (225). C. fetus-asso-

ciated myo(peri)carditis may also require expression of bacterial

surface layer proteins to confer evasion of the host immune

system.

Campylobacter-associated endocarditis is an infrequent condi-

tion in which both native and prosthetic valves can be affected.

Previous studies have shown that individuals with Campylobacter-

associated endocarditis have either C. jejuni or C. fetus infection

and that one-third of these patients suffer from concurrent

chronic diseases, including hepatic cirrhosis, connective tissue

disease, tuberculosis, or cancer (233–239). Furthermore, four

cases of atrial ﬁbrillation associated with C. jejuni infection have

been reported in the literature (230,240), and both C. jejuni and C.

fetus infections have been associated with Campylobacter-associ-

ated aortitis (241–244). Overall, the limited current evidence sug-

gesting an association between Campylobacter species and cardio-

vascular complications precludes causal inference.

Meningitis

Both C. jejuni and C. fetus subsp. fetus have been implicated in the

development of meningitis in humans (245–252). Meningitis

caused by C. fetus subsp. fetus has generally been reported for

immunocompromised adults and is rare, with only eight cases

reported from 1983 to 1998 (248). C. jejuni-associated meningitis

is also rare and may affect both healthy and immunocompro-

mised children and adults (246,247,251).

Extraoral Abscesses

Campylobacter species have been reported to be present in several

types of abscesses outside the oral cavity. C. rectus has been asso-

ciated with a chest wall infection (253), a breast abscess (254), and

a vertebral abscess (255), while C. curvus has been associated with

a liver abscess in a patient with complicated ovarian cancer and a

bronchial abscess in a patient with lung cancer (254). C. gracilis

and C. concisus have both been implicated in brain abscesses, while

C. showae was detected in an intraorbital abscess (255). Most ab-

scesses are polymicrobial in nature, making it difﬁcult to assess the

contribution of a speciﬁc Campylobacter species to the clinical

outcome.

Reactive Arthritis

Reactive arthritis is a form of arthritis which most commonly

occurs in patients in their 30s or 40s and develops following gas-

trointestinal or genitourinary infections. This condition can affect

joints, such as knees and ankles, as well as the eyes and the genital,

urinary, and gastrointestinal systems. Symptoms can begin ap-

proximately 1 month following infection and resolve within a

year, although in some patients this condition may persist for up

to 5 years (256). In 2007, a systematic review by Pope and col-

leagues reported the incidence of reactive arthritis associated with

Campylobacter infection to be 1 to 5% (257). Other studies have

estimated the risk of reactive arthritis associated with Campylo-

bacter infection to be 3 to 13%, compared with 0 to 9% for E. coli

O157:H7, 2 to 15% for Salmonella, 1 to 10% for Shigella,and0to

14% for Yersinia (256). Recently, Ajene and colleagues performed

acomprehensive systematic review to identify the global incidence

of reactive arthritis associated with infections by enteric patho-

gens, namely, Campylobacter,Salmonella, and Shigella (258). They

Epidemiology of Campylobacter Infection

July 2015 Volume 28 Number 3 cmr.asm.org 699Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

found 25 articles, among which 14 cohort studies were chosen for

calculation. Among a total of 63,206 patients infected by Campy-

lobacter, 573 developed reactive arthritis, generating an incidence

rate of 9 reactive arthritis cases per 1,000 cases of Campylobacter

infection (258). Furthermore, the incidences of reactive arthritis

resulting from Campylobacter infection ranged from 8 to 16% in

adults and 0 to 6% in children. In comparison, the incidence rate

for both Salmonella and Shigella infections was slightly higher,

with 12 reactive arthritis cases per 1,000 infections (258). In a

more recent meta-analysis assessing the proportion of Campylo-

bacter cases that develop chronic sequelae, Keithlin and colleagues

found that only 2.86% (95% CI, 1.40 to 5.61%) of patients in-

fected with Campylobacter, mainly with C. jejuni and C. coli, de-

veloped reactive arthritis (186), suggesting that the incidence rate

of Campylobacter-associated reactive arthritis is largely dependent

on the geographic regions and cohorts used.

Complications of the Reproductive System

C. jejuni,C. coli,C. fetus subsp. fetus, and C. upsaliensis have been

shown to cause septic abortion and neonatal sepsis in humans and

animals (259,260). These species are generally associated with

abortion in pregnant women following an aggressive bowel infec-

tion that results in sepsis, with the infection eventually transmitted

to the fetus (259). The clinical presentation and outcome of abor-

tion caused by Campylobacter do not differ between species.

Other Campylobacter species, such as C. rectus and C. curvus,

have been associated with premature birth and low birth weight in

pregnant humans or mice (146,261,262). It has been shown that

C. rectus translocates from the oral cavity to the reproductive tract,

leading to inﬂammation that results in preterm birth (146). In

mice, this inﬂammation appears to be mediated by the activation

of Toll-like receptor 4 (TLR4) in placental tissues infected with C.

rectus (262,263). It remains to be seen if similar mechanisms

control the inﬂammation that leads to preterm birth and intra-

uterine growth restriction following Campylobacter infection in

humans.

CLINICAL MICROBIOLOGY

Isolation Methodologies in Clinical Settings

There is no gold standard or common method for the isolation of

all Campylobacter species from clinical samples. Several selective

agar media, using blood-based agar or blood-free agar, have been

used for the isolation of Campylobacter species, particularly ther-

motolerant species (2). However, given the variability in antibiotic

susceptibilities among Campylobacter species, these methods are

effective for only a subset of species.

A more robust method—the Cape Town protocol—requires

ﬁltration of homogenized clinical samples through membrane ﬁl-

ters with a pore size of 0.45 or 0.65 ␮m onto blood agar media

(with or without vancomycin supplementation). The plates are

then incubated at 37°C under microaerobic conditions (⬃5% O

)

enriched with CO

and H

(264). While not an absolute require-

ment, H

enhances the growth of some Campylobacter species.

The Cape Town protocol has been used successfully to isolate a

range of Campylobacter species from fecal, intestinal biopsy, and

saliva samples (264).

Enrichment procedures have been suggested to improve isola-

tion rates from samples with a small starting number of Campy-

lobacter cells (e.g., intestinal biopsy specimens). Enrichment of

homogenized intestinal biopsy specimens in Ham’s F-12 medium

supplemented with vancomycin, followed by incubation at 37°C

for 2 days and in combination with the Cape Town protocol, has

been effective for the isolation of a number of Campylobacter spe-

cies from patients with chronic gastroenteritis and IBD (108).

Other enrichment broths that have been used to successfully iso-

late Campylobacter species include brucella-FBP (a combination

of ferrous sulfate, sodium metabisulﬁte, and sodium pyruvate),

Preston, Doyle and Roman, modiﬁed charcoal cefoperazone de-

oxycholate, Park and Sanders, Bolton, Hunt and Radle, and Hunt

broths (265).

Laboratory Diagnosis

Laboratory diagnosis of Campylobacter infection requires the use

of culture-dependent and/or culture-independent methodolo-

gies. In culture-dependent methodologies, single isolated colonies

can be subjected to a range of conventional biochemical tests to

identify phenotypic traits. Temperature, incubation time, and at-

mospheric conditions, for example, an H

-enriched atmosphere

or growth at 42°C, can also be used to favor the growth of Cam-

pylobacter species during isolation procedures. The biochemical

proﬁle of an unknown organism is then matched to previously

deﬁned characteristics of the Campylobacter genus/species to en-

able identiﬁcation of the organism to the genus or species level. In

culture-independent methodologies, DNA or RNA can be iso-

lated from clinical samples or from a pure culture. A genetic sig-

nature or marker of the organism can then be determined using

sequencing techniques or genus- or species-speciﬁc PCR ampliﬁ-

cation of the gene of interest, for example, the 16S rRNA gene.

Culture-independent tests are increasingly being used for the de-

tection of Campylobacter species, and while in many cases they can

enhance detection sensitivity, it has been proposed that this will

have an impact on public health surveillance given that detailed

analyses of isolates are required to monitor the distribution of

different strains (266).

Biochemical identiﬁcation. Biochemical tests can be used to

differentiate Campylobacter species from related genera and to

identify organisms to the species level (2). The relevant biochem-

ical tests used to differentiate between members of the Campylo-

bacter genus have been reviewed by Lastovica (80). In that article,

a biochemical ﬂowchart to identify Campylobacter species is out-

lined, beginning with growth with or without supplementation of

, followed by an indoxyl acetate test, a hippurate test, growth on

MacConkey agar, an aryl sulfatase test, and production of H

Furthermore, detection of L-alanine aminopeptidase activity can

be employed to differentiate between Campylobacter,Helicobac-

ter, and Arcobacter species and other Gram-negative bacteria (80).

However, the ability to differentiate between C. jejuni and C. coli

would be the most relevant in the clinical setting. The only bio-

chemical test that distinguishes between these two species is the

hippurate hydrolysis test. C. jejuni isolates have the ability to hy-

drolyze hippurate, whereas C. coli isolates yield a negative test

result.

Molecular identiﬁcation. The 16S rRNA gene has been used

extensively for rapid detection and identiﬁcation of many bacte-

rial taxa, including Campylobacter species (267,268). Owing to the

sequence similarity among Campylobacter species, the 16S rRNA

gene sequence cannot be used to differentiate between very closely

related species, such as C. jejuni and C. coli (269). The larger 23S

rRNA gene and the internal transcribed spacer (ITS) region, a

Kaakoush et al.

700 cmr.asm.org July 2015 Volume 28 Number 3Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

region which lies between the 16S and 23S rRNA genes, have also

been used to differentiate between Campylobacter species and

strains (270–272). The 23S rRNA genes contain strain-speciﬁc

intervening sequences, whereas the ITS region is highly variable in

size and sequence composition depending on the species (273,

274). Indeed, a comprehensive analysis revealed the ITS region to

be the most discriminatory region, compared with the 16S and

23S rRNA genes, for species and strain differentiation for the

Campylobacter genus (270). When all three regions (16S rRNA-

ITS-23S rRNA) were combined to create a phylogenetic tree, the

resultant tree had the highest resolution in differentiating between

members of the Campylobacter genus (270).

Several enzyme immunoassays are also available for the de-

tection of C. jejuni and C. coli in clinical samples. A recent study

by Granato and colleagues compared three commercially avail-

able kits with culture-based techniques and found that the three

immunoassays had sensitivities that ranged from 98.5 to 99.3%

and speciﬁcities that ranged from 98.0 to 98.2%, while standard

culture had a sensitivity of 94.1% (275). Furthermore, a number

of real-time assays are also available for the detection of Campy-

lobacter species, some of which are capable of detecting more than

one species at a time, including C. jejuni,C. coli, and C. lari (57).

Interestingly, Javed and colleagues also described an assay to de-

tect C. jejuni and C. coli based on the ability of recombinant re-

ceptor binding proteins from the C. jejuni bacteriophage

NCTC12673 to agglutinate in the presence of these species (276).

More recently, due to the reduced costs of bacterial genome

sequencing, the genomes of Campylobacter species and strains can

now be sequenced fully, and many of these are completed or

drafted. To date, the complete and draft genomes of over 100

Campylobacter species or strains—predominantly C. jejuni and C.

coli strains—are available in the NCBI database. Genome se-

quencing and assembly have become commonplace in research

labs for the characterization of isolates, and this will likely become

routine practice in diagnostic facilities in the future.

Antibiotic Therapies

Most Campylobacter infections are self-limiting and require no

therapeutic intervention other than supportive therapy, such as

maintenance of hydration and electrolyte balance. However, an-

tibiotics are employed in immunocompromised patients, patients

whose symptoms are severe or persistent, and those with extraint-

estinal infections (277,278).

Ciproﬂoxacin, a ﬂuoroquinolone, is often used for the em-

pirical treatment of gastroenteritis, particularly in travel-re-

lated cases. The major targets of quinolones in bacteria are

DNA gyrase and topoisomerase IV, which are enzymes essen-

tial for DNA replication, transcription, recombination, and re-

pair. However, in the case of campylobacteriosis, macrolides

are the preferred choice of therapy (279). Macrolides bind to

the 23S rRNA nucleotides 2,058 and 2,059 in the 50S ribosomal

subunit, which results in blockage of the translocation step of

protein synthesis, thereby preventing release of tRNA after

peptide bond formation and resulting in the termination of

peptide chain elongation. The reason for the increasing impor-

tance of macrolides for the treatment of Campylobacter infec-

tions is that the rates of ciproﬂoxacin resistance are relatively high

for Campylobacter species due to the use of antibiotics in the poul-

try industry and animal husbandry operations (discussed further

below) and, to a lesser extent, the indiscriminate use of ciproﬂoxa-

cin for treatment of human diseases (280). This is particularly

important in developing countries, where the use of ﬂuoroquino-

lones is usually a suboptimal approach, for the above-mentioned

reasons. For example, a recent randomized double-blind trial

conducted in Thailand that compared macrolide regimens (sin-

gle-dose and 3-day azithromycin treatments) with a ﬂuoroquin-

olone regimen (3-day levoﬂoxacin treatment) for the empirical

management of traveler’s diarrhea, mainly caused by C. jejuni/C.

coli (64%), reported a cure rate of 96% with azithromycin, com-

pared to 71% with levoﬂoxacin (281). Similarly, the rate of micro-

biological eradication was superior with azithromycin-based reg-

imens (96% to 100%) compared to the levoﬂoxacin regimen

(38%) (P⫽0.001) (281). The high efﬁcacy of azithromycin was

supported by a further study comparing a single-dose azithromy-

cin regimen to a 5-day erythromycin treatment in children with

campylobacteriosis, which showed that azithromycin was signiﬁ-

cantly superior to erythromycin in eradicating the pathogen and

accelerating the time to clinical cure (282).

Anumber of macrolides are also used in food animal produc-

tion for both growth promotion and therapeutic reasons (283),

and this has resulted in high levels of macrolide resistance being

reported in some countries (284,285). Consequently, ciproﬂoxa-

cin has been recommended for the treatment of human infections

caused by macrolide-resistant Campylobacter species.

Another problem related to antibiotic resistance in Campylo-

bacter infections is the emergence of multidrug-resistant (MDR)

strains (deﬁned as strains with resistance to three or more antibi-

otics), which have been isolated in many countries throughout the

world. The levels of MDR strains in humans are still relatively low

overall (⬍25%) (286), but an increase in these strains in domes-

ticated animals has raised concerns in relation to human disease

(287). These concerns are well placed given that infections with

Campylobacter species which are resistant to antibiotics have been

associated with a longer duration of illness, an increased risk of

invasive disease and death, and increased health care costs (288–

290). In the case of severe Campylobacter infection in humans,

treatment with aminoglycosides (e.g., gentamicin or kanamycin)

is commonly employed (291). The mechanism of action of amin-

oglycosides involves binding to the 30S ribosomal subunit, where

they interfere with the binding of formylmethionyl-tRNA to the

ribosome, preventing the formation of the initiation complexes

from which protein synthesis proceeds, and the rate of resistance

to these antibiotics remains relatively low for Campylobacter spe-

cies (3%) (286,292).

RISK FACTORS, TRANSMISSION, AND ENVIRONMENTAL

RESERVOIRS

A number of risk factors contribute to the susceptibility of hu-

mans to campylobacteriosis. A recent meta-analysis revealed that

international travel was the most important risk factor for cam-

pylobacteriosis (OR, 4.9; 95% CI, 2.9 to 8.2), followed by con-

sumption of undercooked chicken, environmental exposure, and

direct contact with farm animals (293). In agreement with this,

another study revealed that international travel was a risk factor

for infection, with C. jejuni being one of six organisms that con-

tributed to 70% of all gastrointestinal infections acquired during

overseas travel between 1996 and 2005 (294). Moreover, campy-

lobacteriosis was found to be the main cause of travel-related dis-

ease in Canada from 2005 to 2009, being responsible for 27.6%

(123/446 cases) of cases (295). Similar levels have been reported in

Epidemiology of Campylobacter Infection

July 2015 Volume 28 Number 3 cmr.asm.org 701Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

the United States, with 18% of Campylobacter infections being

estimated to be associated with international travel (296). Domes-

tic travel also plays a role in the transmission of human campylo-

bacteriosis (297). In Europe, the proportion of diarrhea cases

caused by Campylobacter species detected in travelers increased

from 7% in 2008 to 12% in 2010 (298). In England, 17% of Cam-

pylobacter infections were classiﬁed as being associated with travel

(299).

The level of risk for travel-related campylobacteriosis appears

to be associated with the travel destination. A meta-analysis in

2009 showed that the locations with the highest levels of risk are

Southeast Asia (32.4%; 162/500 cases), South Asia (7.8%; 39/499

cases), Africa (4.6%; 54/1,177 cases), and Latin America (2.5%;

51/2,031 cases) (300). More recently, Mughini-Gras and col-

leagues found that Dutch travelers visiting South Asia (OR, 28.9;

95% CI, 2.4 to 265.1), Southeast Asia and China (OR, 27.8; 95%

CI, 4.5 to 170.9), sub-Saharan Africa (OR, 25.4; 95% CI, 2.7 to

310.7), Latin America and the Caribbean (OR, 20.8; 95% CI, 2.0 to

211.6), western Asia (OR, 10.6; 95% CI, 2.8 to 39.9), and northern

Africa (OR, 10.6; 95% CI, 2.3 to 49.0) were at higher risk of devel-

oping campylobacteriosis than those traveling to western Euro-

pean countries. While travel does contribute to the overall burden

of Campylobacter transmission, the transfer or spread of exotic or

antibiotic-resistant strains to previously unexposed populations is

a further concern. In addition, travel-related infections are often

associated with consumption of contaminated meat or water

(300), suggesting that a complex interrelationship exists between

risk factors.

Consumption of contaminated food, particularly poultry

products, unpasteurized milk, and water, is a risk factor for C.

jejuni and C. coli infection (Fig. 1)(

18,301,302). One approach to

estimate the source attribution of campylobacteriosis cases is

through the analysis of outbreak data. As expected, Campylobacter

infections are mainly attributed to poultry products by using these

techniques (303,304). More recently, however, molecular tech-

niques, such as multilocus sequence typing, porA/ﬂaA typing, and

gyrase subunit A typing, have played a more signiﬁcant role in

source attribution (305–310), allowing further insights into the

contributions of the different reservoirs to the burden of campy-

lobacteriosis in humans and the subsequent transfer of Campylo-

bacter species to wildlife through environmental contamination.

Immunodeﬁciency is also a risk factor for campylobacteriosis

(101,311). For example, human immunodeﬁciency virus (HIV)-

infected patients presenting with diarrhea are more frequently

infected with Campylobacter than uninfected individuals with di-

arrhea (311). In addition, the incidence of Campylobacter-related

illness among HIV-infected patients is higher than the incidence

found in the general population (312). Below, we further examine

the prevalence of Campylobacter species in the environment and

discuss how these reservoirs facilitate transmission of Campylo-

bacter to humans.

Poultry

Poultry is recognized as a primary source of food-related trans-

mission of Campylobacter species to humans (313). A major con-

tributing factor is the high carriage rate of Campylobacter within

broiler chickens. Therefore, Campylobacter species are found in

abundance on poultry farms and their surrounding environment,

including the soil, water sources, dust, building surfaces, and the

air (314). In addition to chickens, commercial turkeys and ducks

can also serve as reservoirs of C. jejuni and C. coli (315–317).

Furthermore, poultry is also an important reservoir of other Cam-

pylobacter species, such as C. lari,C. upsaliensis, and C. concisus

(318,319). Domesticated broiler chickens and imported chickens

both contribute to the overall burden of Campylobacter infections

(320).

It has been estimated that 71% of human campylobacteriosis

cases in Switzerland between 2001 and 2012 were attributed to

chickens (321,322). A study investigating the prevalence of Cam-

pylobacter in crop (enlarged part of the digestive tract of birds that

serves as a temporary storage space for food) and cecal samples

from market-age broiler chickens found 62% of crop samples to

be positive, compared to only 4% of cecal samples (P⬍0.001),

suggesting that the crop is an important niche for Campylobacter

and may represent a major source of contamination during pro-

cessing (323). Given that C. jejuni strains survive in chicken feces

for up to 6 days after excretion, chicken feces may also be a poten-

tial source of transmission to the environment or humans when

poultry manure is used as a fertilizer (324).

The UK Food Standards Agency reported preliminary ﬁndings

showing that 72.9% of fresh whole retail chickens surveyed during

2014 to 2015 were infected with Campylobacter, with 18.9% of

these harboring a level of ⬎10,000 CFU/g, which is considered

highly contaminated (325). Data from Canada also support the

ﬁnding that broiler chickens are a major source of Campylobacter

species. There are, however, indications that chicken-associated

Campylobacter infections may be more common in urban dwell-

ers than in rural dwellers (326,327). One study reported the levels

of thermotolerant Campylobacter species to be three times higher

in organic broilers than in conventional broilers (54.2% versus

19.7%) (328), suggesting that the likelihood of purchasing Cam-

pylobacter-contaminated broiler meat is higher for organic

sources than for conventional sources. Furthermore, the relative

risk of becoming ill from Campylobacter on a per-serving basis is

1.7 times higher for consuming organic carcasses than for con-

suming conventional carcasses.

Domesticated Animals

Domesticated animals are another reservoir of Campylobacter

species (313,329). Fresh and frozen meats are frequently contam-

inated with Campylobacter, whereas commercial cooked products

are less affected (1,330). C. jejuni,C. coli, and C. lari accounted for

69, 30, and 1% of the contaminating bacteria in these products,

respectively (330). In addition to consumption of contaminated

meat from domesticated animals, contact with domesticated and

companion animals poses a signiﬁcant risk for the transmission of

Campylobacter species (1,301,302,321,331,332).

Recent studies from Denmark showed that cattle were the at-

tributed source for 16 to 17% of the total cases of campylobacte-

riosis (320). Similarly, in Switzerland, cattle have been estimated

to be responsible for 19.3% of Campylobacter infections, which is

substantially higher than the contribution from pigs (1.2%) (321).

The prevalence of Campylobacter in cattle varies signiﬁcantly

across studies, with rates ranging from 23% to close to 90% (333–

335). The species detected in cattle include C. jejuni,C. coli,C. lari,

and C. lanienae (335–337).

Campylobacter species are also prevalent in pigs and piglets.

These bacteria are capable of colonizing piglets as early as 24 h

after birth, as a result of exposure to contaminated feces (338). The

carriage rates of Campylobacter range from 32.8 to 85.0%, de-

Kaakoush et al.

702 cmr.asm.org July 2015 Volume 28 Number 3Clinical Microbiology Reviews

on June 10, 2015 by ST JUDE CHILDRENS RES HOSPITALhttp://cmr.asm.org/Downloaded from

Global Epidemiology of Campylobacter Infection

Abstract and Figures