Campylobacter jejuni and Related Species

Revised July 1, 2020

Microbiology

Campylobacter organisms are motile, non–spore-forming, comma-shaped, gram-negative rods. Originally isolated from aborted sheep fetuses in 1909, these and similar organisms were considered subspecies of Vibrio fetus. However, because these organisms did not ferment carbohydrates and differed in their guanine plus cytosine (G + C) DNA content from true members of the genus Vibrio, a new genus, Campylobacter, was created. Fourteen species have been recognized within the genus; however, in recent years, taxonomic studies have indicated that splitting the genus is more appropriate. The genus Arcobacter has been created, which now includes Arcobacter butzleri and Arcobacter skirrowi. Helicobacter cinaedi and Helicobacter fennelliae had been named Campylobacter cinaedi and Campylobacter fennelliae when first discovered. Although transfer to the genus Helicobacter is more appropriate on taxonomic grounds, because these two species cause intestinal rather than gastric illnesses they are discussed in this chapter. Helicobacter pylori, previously named Campylobacter pylori, is discussed in Chapter 217 . It is clear that new members of Campylobacter and related genera are being identified with regularity and that many of these will be found to be human pathogens.

Table 216.1 lists the Campylobacter and related species most commonly associated with human disease and indicates the differentiating characteristics. Certain species, such as Campylobacter nitrofigilis and Arcobacter cryaerophilus, have not yet been associated with human illness. In contrast, the “nitrate-negative” campylobacters are associated with diarrheal illnesses, but the appropriate nomenclature for the organisms has not been determined. Two types of illnesses are directly associated with Campylobacter spp.: enteric and extraintestinal. For each of these illnesses, one Campylobacter species predominates and other species are less commonly present. The prototype for enteric infection is C. jejuni; for extraintestinal infection it is Campylobacter fetus ( Table 216.2 ). Because the organisms causing enteric and extraintestinal illnesses are generally the same, they are considered together in the following discussion.

Campylobacters and related organisms grow best in an atmosphere containing 5% to 10% oxygen and are thus considered microaerophilic. Although most of these organisms will not grow under aerobic or anaerobic conditions, C. jejuni can grow in candle jars, which permits isolation when the optimal atmosphere cannot be achieved. All campylobacters grow at 37°C; however, C. jejuni grows best at 42°C. Because C. jejuni is the most common enteric pathogen of humans, many laboratories have used incubation at 42°C for optimal isolation; however, use of this temperature will not permit detection of infections by many of the related species. In particular, Campylobacter upsaliensis may be missed.

Campylobacter spp. multiply more slowly than do the usual bacteria of the enteric flora and therefore cannot be isolated from fecal specimens unless selective techniques are used. The most common isolation methods use blood-based, antibiotic-containing media. Three such media—Skirrow medium, Butzler agar, and Campy-BAP medium—or variations of these have been in wide use. The last two media contain cephalothin, which inhibits C. fetus and several other Campylobacter subspecies, but are best suited for isolating C. jejuni. Several enrichment broths have been developed, but because ill humans usually excrete 10 ⁶ to 10 ⁹ C. jejuni colony-forming units per gram of stool, enrichment usually is not necessary. Blood-free media can also be used. Owing to their small size (0.3–0.6 µm in diameter) and motility, Campylobacter spp. and related organisms pass through 0.45- or 0.65-µm filters that retard the usual enteric flora. Filtration methods permit isolation without use of antibiotic-containing media. It is now clear that use of filtration techniques and nonselective rich media such as chocolate agar, with incubation of plates at 37°C, improves stool culture yields of both C. jejuni and the “atypical” enteric Campylobacter spp. The development of filtration techniques represents a significant advance over the use of selective media, and such techniques are now recommended for primary isolation of Campylobacter spp. from fecal specimens or swabs.

Visible colonies usually appear on the plating media within 24 to 48 hours, and usually longer for the “atypical” species. Campylobacter spp. can be distinguished from other microorganisms on the basis of several standard criteria and can be distinguished from one another on the basis of biochemical testing. Organisms from young cultures have a typical vibrioid appearance ( Fig. 216.1 ), but after 48 hours of incubation, organisms appear coccoid. The ability to hydrolyze hippurate distinguishes C. jejuni from most other members of the genus, but hippurate-negative C. jejuni isolates also occur. Isolation of organisms from sites without a normal microbiota, such as the bloodstream, is not difficult, although when Campylobacter is the suspected pathogen, incubation of cultures should be extended to 2 weeks. With radiometric detection systems, turbidity of the medium may not be present, and the increase in released radiolabeled substance may be less than usually specified thresholds, reflecting suboptimal conditions for certain of these organisms. State-of-the-art identification to the species level should include polymerase chain reaction studies of 16S recombinant RNA (rRNA) or other targets for comparison with known species. As discussed later (see “ Diagnosis ”), these techniques have been developed for culture confirmation and for typing of strains and may be more sensitive than traditional culture methods.

As with other bacteria whose ecologic niche is the gastrointestinal tract of mammals and avian species, the serotypical diversity of C. jejuni is enormous. More than 90 different serotypes based on somatic (O) antigens and 50 different serotypes based on heat-labile (capsular and flagellar) antigens have been identified ; phase variation of flagellar antigens occurs. O-antigen variation reflects the presence of differing genetic cassettes that contain the enzymes for O-antigen formation. No group somatic or flagellar antigen has been identified; however, several superficial proteins appear to have broad serotypical specificity, a factor that may aid in the development of a broadly specific vaccine.

C. jejuni cannot long withstand drying or freezing temperatures, which are characteristics that limit its transmission. However, C. jejuni survives in milk or other foods or in water kept at 4°C for several weeks. Pasteurization effectively destroys the organism, as does chlorine at concentrations in standard use for water disinfection.

Epidemiology

Campylobacteriosis is a worldwide zoonosis. Campylobacter spp. are commonly found as commensals of the gastrointestinal tract in wild or domesticated cattle, sheep, swine, goats, dogs, cats, rodents, and all varieties of fowl. C. jejuni has a very varied reservoir, but Campylobacter coli and Campylobacter hyointestinalis are most commonly isolated from swine, and C. upsaliensis is most commonly isolated from dogs. C. fetus subsp. fetus has been isolated from sheep, cattle, poultry, reptiles, and swine. Primary acquisition of Campylobacter spp. by animals often occurs early in life and may lead to morbidity or mortality, but in most colonized animals a lifelong carrier state develops. The vast reservoir in animals is probably the ultimate source for most enteric Campylobacter infections in humans. Meats originating from infected animals frequently become contaminated with intestinal contents during the slaughtering process. In particular, commercially raised poultry is nearly always colonized with C. jejuni, slaughterhouse procedures amplify contamination, and chicken and turkey in supermarkets, ready for consumers to take home, frequently are contaminated. Excreta from infected animals may contaminate soil or water. Most infections in humans probably result from consumption of contaminated food and water. Investigations of more than 50 outbreaks indicate that unpasteurized (raw) milk is such a vehicle. Similarly, untreated surface water has been responsible for both endemic and epidemic campylobacteriosis. Backpackers in Wyoming who drank untreated water and developed acute diarrheal illnesses had three times more Campylobacter infections than Giardia infections. Several large outbreaks have been traced to defects in municipal water systems. Undercooked meats, especially poultry, have been associated with infection. Other vehicles include raw clams, raw or undercooked beef, and unpasteurized cheeses and goat's milk. In one US study, children riding in grocery store shopping carts next to raw meat or poultry had higher rates of Campylobacter infections. Nevertheless, consumption of undercooked poultry is estimated to be responsible for 50% to 70% of sporadic Campylobacter infections in developed countries. Increases in the isolation of Campylobacter spp. reflect both improved recognition and increased consumption of poultry in recent years.

Direct contact with infected animals may result in transmission. Household pets, especially young dogs and cats with diarrhea, have been implicated as vectors for campylobacteriosis. In 2017, a large multistate US outbreak of multidrug-resistant C. jejuni infections occurred in puppies and was transmitted to more than 100 people. Because healthy dogs, cats, rodents, and birds may excrete Campylobacter and related organisms, it is not surprising that human infections associated with these animals also have been reported. People with occupational exposure to cattle, sheep, and other farm animals are at increased risk for infection, and laboratory-acquired infections have been reported. C. fetus strains in reptiles and mammals probably diverged 200 million years ago ; however, humans may become infected with reptile strains, possibly due to consuming a food of reptile origin. Most reported strains in the United States and elsewhere have been from peoples of Asian origin, suggesting that some particular contaminated food is involved. The reptile isolates are sufficiently distinct that a new subspecies has been proposed: C. fetus subsp. testudinum.

As with other enteric pathogens, fecal-oral person-to-person transmission of C. jejuni has been reported. People in contact with the excreta of infected individuals who are not feces continent (e.g., infants) are at risk for infection. Infected school-age children rarely may transmit Campylobacter infection. Transmission from infected food handlers who are asymptomatic is at best uncommon. Perinatal transmission from a mother who may not have been symptomatic may be due to exposure in utero, during passage through the birth canal, or during the first days of life. Infection has been associated with blood transfusion from an infected patient. Because of a variety of sexual practices, homosexual men appear to be at increased risk for infection caused by H. cinaedi, H. fennelliae, and other “atypical” Campylobacter spp. Human immunodeficiency virus (HIV)–infected patients are at substantially increased risk for infection. The standardization of serotyping methods and the development of molecular methods for identification and typing of C. jejuni and related organisms should improve our understanding of transmission.

C. jejuni infections occur year-round in the United States and other developed countries but with a sharp peak in summer and early fall. C. fetus infections show the same seasonal variation, but the peak is less marked. The reason for this seasonal variation is unclear. In developed countries, the incidence of infection is higher when air temperatures rise. Flies also have been suggested as a potential source of transmission to humans. In tropical countries, the seasonal variation of C. jejuni infection appears to be influenced by rainfall.

For many years, the incidence of C. jejuni infections continued to rise in the United States and Europe and exceeded rates of Salmonella and Shigella infections combined. Beginning in the mid-1990s, the incidence has waxed and waned but C. jejuni remained one of the most frequently reported pathogens that cause foodborne illness. Improved hygienic practices on farms and especially in poultry slaughterhouses may have contributed to stabilization in disease frequency. Beginning in 2014, the incidence began to increase again. Indeed, in 2017, the most recent year reported by the Centers for Disease Control and Prevention's Foodborne Diseases Active Surveillance Network (FoodNet), the incidence of Campylobacter rose by 10%, climbing to 19.1 per 100,000 population. Over the past decade, the incidence of Campylobacter infections has also increased in Europe and Israel. The reasons for these increases are not obvious, but increasing reliance on culture-independent diagnostic tests (CIDTs) could be playing a role. In the United States, the number of Campylobacter cases diagnosed by CIDTs alone increased by 114% in 2016 when compared with 2013 to 2015. Because CIDT methods are quicker and more readily available, clinicians may order these tests more frequently, making year-to-year comparisons difficult. As a result, changes in Campylobacter incidence over time and determinations of the effectiveness of prevention strategies are becoming increasingly challenging. Nevertheless, Campylobacter infections have always been underdiagnosed vis-à-vis infections with Enterobacteriaceae, due to their slow and fastidious nature in culture, the need for specialized culture conditions, and the importance of laboratory experience in isolating and identifying these organisms. The increasing use of CIDTs is permitting a better assessment of C. jejuni infections, but unless the constituents of the panels are broadened in the future, will fail to identify infections due to the less common species, further underdiagnosing these infections. Other technologic advancements, such as whole-genome sequencing of Campylobacter isolates, may be available in the future for use in outbreak investigations and in further characterizing the epidemiology of these infections.

Based on current estimates, there are probably more than 2 million Campylobacter infections annually in the United States, although there is great geographic variability in incidence rates, even within the United States. Population-based studies show peak incidence in children younger than age 1 year and in people 15 to 29 years old ; however, cases have been reported in patients of all ages. The incidence in males may be higher. The prevalence of infection in healthy people is very low (<1%).

The epidemiology of infection in developing countries is markedly different. C. jejuni is often isolated from healthy people, and the infection is especially common during the first 5 years of life. During the first 2 years of life, most children have numerous Campylobacter infections, but those occurring early in life frequently are symptomatic, whereas later infections are mostly asymptomatic. The source of these frequent infections has not been defined, but preliminary evidence suggests that human-to-human transmission may be more common than in developed countries. The substantial age-related difference in the infection-to-illness ratios in developed and developing countries appears primarily to be due to differences in age- or exposure-related immunity of the populations rather than to differences in the isolates. Areas where Campylobacter infection is most prevalent are associated with growth shortfalls among children. Interestingly, even in developed nations, the incidence of infection in rural areas is higher, similar to patterns observed in developing nations, and possibly associated with more direct contact with animal vectors of the infection. C. jejuni and other Campylobacter spp. are important causes for the acute diarrheal illnesses suffered by travelers.

Update: Updated Incidence for 2018

Pathogenesis and Pathologic Characteristics

The absence of a nonprimate animal model that is closely analogous to human infection makes understanding Campylobacter pathogenesis more difficult. However, it is clear from outbreak investigations and from volunteer studies that not all Campylobacter infections produce illness. Although all factors responsible for this phenomenon are not known, three of the most important appear to be the dose of organisms reaching the small intestine, the virulence of the infecting strain, and the specific immunity of the host to the pathogen ingested. Among exposed people who become ill, the incubation period varies from 1 to 7 days, a characteristic that is probably inversely related to the dose ingested. Although there are reports of illness occurring within 24 hours of exposure, most infections occur 2 to 4 days after exposure. In one study, volunteers became ill after ingesting as few as 500 organisms, but with a dose of less than 10 ⁴ organisms, illness was infrequent. C. jejuni, like Salmonella typhimurium, is susceptible to hydrochloric acid. Taken together, these data suggest that the infectious dose for C. jejuni is similar to that for Salmonella. The acidic milieu of the stomach provides an effective barrier against Campylobacter infection. However, vehicles such as milk, fatty foods, and water that favor passage through the gastric acid barrier may permit some infections to occur at relatively low doses. Similarly, patients who use proton pump inhibitors or histamine type 2 blockers are more susceptible to infection. Also, some Campylobacter spp. appear to be well adapted to survival outside animal hosts and are more resilient to physical stresses ; these strains may be more available to infect humans.

C. jejuni multiplies in human bile a characteristic that aids colonization of the bile-rich upper small intestine early in infection. The sites of tissue injury include the jejunum, ileum, and colon, with similar pathologic features in each. Inspection of affected tissues may reveal a diffuse, bloody, edematous, and exudative enteritis, but pathologic examinations are generally only performed on specimens from patients with the most severe cases. Microscopic examination of rectal biopsy specimens has shown a nonspecific colitis with an inflammatory infiltrate of neutrophils, mononuclear cells, and eosinophils in the lamina propria; degeneration, atrophy, loss of mucus, and crypt abscesses in the epithelial glands; and ulceration of the mucosal epithelium. Rectal biopsy samples with these nonspecific features have been interpreted as showing acute ulcerative colitis or Crohn disease. In other cases, the appearance of the rectal biopsy sample has been similar to that of specimens obtained in Salmonella or Shigella infections. In a series of 124 patients with C. jejuni infection, 18 of the most severely ill patients underwent sigmoidoscopic examination or rectal biopsy; 17 of these procedures showed colonic involvement. Some patients have terminal ileitis as well as colitis. Unspecified host factors are also clearly important; in volunteers, a single strain produced a wide spectrum of clinical manifestations.

The presence of bacteremia in some patients, the finding of cellular infiltration in biopsy specimens, and the presence of blood in stools from patients with Campylobacter colitis also suggest that tissue invasion occurs. The process of C. jejuni invasion is multifactorial. Some evidence suggests C. jejuni breaches epithelial cell barriers via a paracellular route by disruption of tight junctions. Other evidence exists for basolateral and apical transcellular routes of invasion. Invasion requires microtubule-dependent mechanisms, and subsequently Campylobacter organisms are contained within a compartment that does not fuse with lysosomes but allows release of the organisms to the basolateral side of the epithelial cell and underlying lamina propria. Subsequent detection of the organisms by the host immune system leads to an influx of inflammatory cells and cytokines/chemokines and to tissue destruction, which ultimately produce the disease signs and symptoms. The bacteria's flagellae are also important virulence factors because they promote the motility and chemotaxis needed for C. jejuni to colonize the intestinal tract. The bacterial flagellar export apparatus is involved in the secretion of a number of proteins that affect invasion. Some secreted proteins produce rapid apoptotic death of cell cultures ; others may modulate the immune response to favor bacterial survival. However, the role of these proteins in causing diarrhea is not known. Unlike other enteric pathogens such as Salmonella, Campylobacter flagellins are not recognized by the host pattern recognition receptor Toll-like receptor (TLR) 5 due to differences in specific regions of flagellin genes, and do not elicit production of the proinflammatory cytokine interleukin-8, suggesting that flagellins are involved in evasion of innate immune responses in their reservoir hosts.

Unlike other pathogens such as Escherichia coli , the C. jejuni genome does not encode any conventional known enterotoxins. Some strains express the cytolethal distending toxin (Cdt), named for its effect on mammalian cell lines, but its role in pathogenesis is unclear because it is not always present in strains isolated from patients with diarrhea. However, this toxin may affect cell cycle kinetics and may play a role in suppressing innate immunity by inducing death of macrophages.

A high-molecular-weight plasmid (pVir) encodes proteins involved in secretion and also enhances the invasive capabilities of C. jejuni virulence. It has been identified in some clinical Campylobacter isolates and has been significantly associated with bloody stools, although strains lacking pVir may remain virulent. Several epithelial cell adhesins have been identified in C. jejuni. The superficial antigen (PEB1) that appears to be the major adhesin and is conserved among C. jejuni strains also is a target of the immune response, and may represent a vaccine candidate. Other important adhesins include JIpA, a surface-exposed lipoprotein, and CadF, which mediates adhesion by binding to fibronectin. Acquisition of ferrous and ferric iron in the gut is critical for colonization by C. jejuni, and the molecules involved in the process may be considered as virulence factors and targets for interventions as well.

Campylobacter outer membranes contain lipopolysaccharides (LPSs) with typical endotoxic activity. The structure of the LPS O antigen is highly variable. Many C. jejuni O antigens possess sialic acid–containing structures that are recognized by host immune cells bearing sialoadhesins. The close resemblance of these sialylated structures to those seen in human gangliosides such as GM1, GD1a, GD3, and GT1a and their presence in strains isolated from patients who developed the Guillain-Barré syndrome (GBS) support a role in the pathogenesis of this disorder, via molecular mimicry. Certain capsular genotypes are also associated with the sialylated lipooligosaccharide structures implicated in GBS, raising the possibility that capsule also may contribute to GBS pathogenesis. Sialylated lipooligosaccharides confer resistance to host cationic antimicrobial peptides and proteins and may also be associated with more severe gastrointestinal disease.

Bacteremia can sometimes be detected in patients with Campylobacter infections, whether or not they show signs of systemic illness. Most bacteremias reported to the Centers for Disease Control and Prevention have been due to C. fetus subsp. fetus, whereas C. jejuni is by far the more common pathogen overall. One explanation for the apparently greater tendency of C. fetus to cause bacteremia is that it is usually resistant to the bactericidal activity present in normal human serum. Important risk factors for C. fetus bacteremia and meningitis include immunosuppression and occupational exposure to infant animals that are the reservoir hosts for C. fetus. C. fetus is covered with a surface (S)–layer protein that functions as a capsule. Virtually all human isolates of C. fetus possess an S-layer protein that completely disrupts C3b binding to these organisms. Lack of C3b binding explains both serum and phagocytosis resistance. C. fetus also has the ability to change the major S-layer protein expressed. This results in antigenic variation and is facilitated by recombination among several highly homologous genes encoding full-length proteins. The S-layer protein of C. fetus is the major virulence factor explaining its extraintestinal spread ( Fig. 216.2 ).

Immunology

The adaptive human immune responses responsible for protection and recovery from Campylobacter infection are not completely understood. Immunocompetent individuals who develop C. jejuni diarrheal disease usually resolve the diarrhea and become asymptomatic without antibiotic treatment. Protection from recurrent disease may not develop following a single infection, as has been shown in human experimental infection and reinfection models of C.jejuni ; early antibiotic treatment might thwart the development of a full immune response to the infection. Patients infected with Campylobacter spp. develop specific immunoglobulin (Ig) G, IgM, and IgA antibodies in serum and IgA antibodies in intestinal secretions.

In low-income countries, where C. jejuni infection is hyperendemic, serum IgA levels rise progressively with age, and the decreasing case-to-infection ratio with age suggests acquisition of protective immunity following multiple exposures. Further supporting the importance of humoral immunity are multiple reports of severe and recurrent C. jejuni infection in patients with congenital or acquired hypogammaglobulinemia. In HIV-infected patients as well, failure of C. jejuni infection to respond to antimicrobial therapy has been correlated with failure to produce a humoral response to infection. Overall, patients with immunodeficiency from HIV have a markedly increased incidence of C. jejuni, although the role of cellular immunity is not well understood. The heightened incidence is consistent with an epidemiologic model of wide circulation of the organism in human foods, but most often at low doses to which immunocompetent hosts are not susceptible, but to which immunocompromised hosts are.

Components of the innate immune response have been demonstrated to both limit infection and trigger inflammation. C. jejuni is serum sensitive, allowing for complement-mediated killing as well as opsonization followed by phagocytosis. Phagocytes, intestinal epithelial cells, and many other cells express the key pattern recognition receptors TLRs 2 and 4, which recognize cell wall components and LPS/lipooligosaccharide. Engagement of these receptors triggers proinflammatory signaling. This leads to the production of inflammatory cytokines (e.g., interleukin-6, tumor necrosis factor-α) and chemokines (e.g., interleukin-8) that assist in the recruitment and activation of host cells that kill C. jejuni directly via neutrophilic phagocytosis, for example. Unlike most other enteric pathogens, the flagellin of C. jejuni is not recognized by TLR5, and TLR9 is poorly activated by C. jejuni as well, suggesting these are means of immune evasion. Nonetheless, C. jejuni is susceptible to killing by cationic antimicrobial agents, including α- and β-defensins produced by intestinal epithelial cells and phagocytes.

As discussed earlier, one important virulence factor in C. fetus that is not found in C. jejuni is a proteinaceous paracrystalline array attached to the O antigen of LPS known as an S layer. This renders C. fetus serum resistant, owing to the inability of the complement component C3b to opsonize its surface. Patients with C. fetus infections much more frequently have evidence of impaired immunity, including conditions such as chronic alcoholism, liver disease, old age, diabetes mellitus, and malignancies, permitting an organism that a priori can resist innate immunity to take hold.

Clinical Manifestations