Chapter 32: Vibrio, Campylobacter, and Helicobacter

This group of curved Gram-negative rods includes Vibrio cholerae, the cause of cholera, one of the first proven infectious diseases, along with Helicobacter pylori and Campylobacter jejuni, newcomers incriminated as pathogens late in the 20th century (Table 32–1). The peptic ulcer disease now known to be caused by H pylori had been long accepted to be due to stress and disturbed gastric acid secretion. Campylobacter jejuni is one of the most common causes of diarrhea in virtually every country of the world. Cholera has undergone a resurgence in recent decades, spreading from its historic Asiatic locale to the Americas, including the coastline of the United States.

Vibrios are curved, Gram-negative rods (Figure 32–1) commonly found in saltwater. Cells may be linked end to end, forming S shapes and spirals. They are highly motile with a single polar flagellum, non–spore-forming, and oxidase-positive, and they can grow under aerobic or anaerobic conditions. The cell envelope structure is similar to that of other Gram-negative bacteria. Vibrio cholerae is the prototype cause of a water-loss diarrhea called cholera. Other species causing diarrhea, wound infections, and, rarely, systemic infection are listed in Table 32–2.

BACTERIOLOGY

Vibrio cholerae has a low tolerance for acid, but grows under alkaline (pH 8.0-9.5) conditions that inhibit many other Gram-negative bacteria. It is distinguished from other vibrios by biochemical reactions, lipopolysaccharide (LPS) O antigenic structure, and production of cholera toxin (CT). There are over 200 O antigen serotypes, only two of which (O1 and O139) cause cholera. Vibrio cholerae biogroup El Tor, an O1 variant, is a biotype of the classic strain. The O139 strains phenotypically resemble O1 El Tor strains but also produce a polysaccharide capsule. Vibrio cholerae possess long filamentous pili that form bundles on the bacterial surface and belong to a family of pili whose chemical structure is similar to those of the gonococcus and a number of other bacterial pathogens. All strains capable of causing cholera produce a colonizing factor known as the toxin-coregulated pilus (TCP) because its expression is regulated together with CT. In aquatic environments V cholerae produces polysaccharide biofilms, which contain carbohydrate moieties mediating cell–cell adhesion and attachment to surfaces.

The structure and mechanism of action of CT have been studied extensively (Figure 32–2). CT is an A–B type ADP-ribosylating toxin. Its molecule is an aggregate of multiple polypeptide chains organized into two toxic subunits (A1, A2) and five binding (B) units. The B units bind to a GM1-ganglioside receptor found on the surface of many types of cells. Once bound, the A1 subunit is released from the toxin molecule by reduction of the disulfide bond that binds it to the A2 subunit, and it enters the cell by translocation. In the cell, it exerts its effect on the membrane-associated adenylate cyclase system at the basolateral membrane surface. The target of the toxic A1 subunit is a guanine nucleotide (G) protein, Gsα, which regulates activation of the adenylate cyclase system. CT catalyzes the ADP ribosylation (Figure 21–16) of the G protein, rendering it unable to dissociate from the active adenylate cyclase complex. This causes persistent activation of intracellular adenylate cyclase, which in turn stimulates the conversion of adenosine triphosphate to cyclic adenosine 3′,5′-monophosphate (cAMP). The net effect is excessive accumulation of cAMP at the cell membrane, which causes hypersecretion of chloride, potassium, bicarbonate, and associated water molecules out of the cell. Strains of V cholerae other than the two epidemic serotypes may or may not produce CT.

CHOLERA

Epidemic cholera is spread primarily by contaminated water under conditions of poor sanitation, particularly where sewage treatment is absent or defective. Even though convalescent human carriage is brief, if the numerous vibrios purged from the intestines of those infected with cholera are able to reach the primary water supply, the conditions for spread are established. The short incubation period (2 days) ensures that organisms ingested by others quickly enter the epidemic cycle. Even so, modern travel makes imported cases of cholera possible. For instance, one man developed diarrhea in Florida after eating ceviche (marinated uncooked fish) just before departure from an airport in Ecuador.

Cholera is endemic in the Indian subcontinent and Africa. Over the last two centuries, its spread beyond this historic locale to other parts of Asia, Indonesia, and Europe has been described in eight great pandemics, each lasting 5 to 25 years. The current pandemic has brought cholera to the Western Hemisphere for the first time since 1911. Sporadic cases of cholera in the United States first appeared in the early 1970s and were traced to inadequately cooked crabs and shrimp caught off the Gulf Coast of Louisiana and Texas. In 1991, Latin America was hit with epidemic cholera with cases reported from 21 countries from Peru to northern Mexico. In Peru alone, over 500 000 cases of cholera and 4500 deaths occurred in 2 years. A massive epidemic of cholera followed the devastating earthquake of 2010 in Haiti. The disease is now endemic, claiming thousands of lives every year. Virulent V cholerae now lurks in coastal waters throughout the hemisphere and in the drinking water of locales with poor sanitation.

The dominant strain of the 20th century was the El Tor biotype, first isolated from Mecca pilgrims at the El Tor quarantine camp in 1905. This strain survives slightly longer in nature and is more likely to produce subclinical cases of cholera, both of which facilitated its spread. In 1992, the first cases of cholera due to a serotype other than O1 were detected in India and Bangladesh. The new serotype (O139 Bengal) is fully virulent with the additional threat of enhanced ability to produce disease in persons whose immunity is due to exposure to the old serotype. Genomic analysis of the clonal Haitian epidemic strains showed them to be nearly identical to variant V cholera El Tor O1 strains isolated from Bangladesh in 2002 and 2008, likely introduced into Haiti by asymptomatic UN peacekeepers from South Asia, and quite different from earlier Peruvian isolates. These realities illustrate the potential for global spread of cholera and the challenges for the vaccine strategies designed to prevent it.

The epidemic potential of V cholerae depends on its ability to survive in both aquatic environments and human hosts. In the environment, it persists in a dormant state in association with shellfish and plankton by attaching to their chitinous exoskeleton in the biofilms formed in the dormant seawater state but not during infection. This dual life is facilitated by a surface protein able to bind a constituent of chitin as well as glycoproteins and lipids on the intestinal epithelium. Satellite tracking has linked periodic climate changes (warming seawater), plankton blooms, and cholera epidemics along the coast of South America. Otherwise, the organism is fragile, surviving only a few days in the environment outside its human or crustacean hosts.

To produce cholera, V cholerae must reach the small intestine, swim to the intestinal crypts, multiply, and produce virulence factors. In healthy people, ingestion of large numbers of bacteria is required to offset the acid barrier of the stomach. Colonization of the entire intestinal tract from the jejunum to the colon by V cholerae requires adherence to the epithelial surface by the abovementioned protein and surface pili. Bacteria recently passed from cholera cases are hyperinfectious by virtue of chemotactic motility facilitating colonization of the small intestine. The outstanding feature of V cholerae pathogenicity is the ability of virulent strains to secrete CT, which is responsible for the disease cholera. The water and electrolyte shift from the cell to the intestinal lumen is the fundamental cause of the watery diarrhea of cholera. Non-O1, non-O139 strains have been sporadically isolated from cases of gastroenteritis but do not produce CT, and thus not the disease cholera.

Fluid Loss

The fluid loss that results from the adenylate cyclase stimulation of cells depends on the balance between the amount of bacterial growth, toxin production, fluid secretion, and fluid absorption in the entire gastrointestinal tract. The outpouring of fluid and electrolytes is greatest in the small intestine, where the secretory capacity is high and absorptive capacity low. The diarrheal fluid can amount to many liters per day, with approximately the same NaCl content as plasma, but also significant potassium and bicarbonate. The result is dehydration (isotonic fluid loss), hypokalemia (potassium loss), and metabolic acidosis (bicarbonate loss). The intestinal mucosa remains unaltered except for some hyperemia, because V cholerae does not invade or otherwise injure the enterocyte.

Genetic Regulation of Virulence

The expression of the multiple virulence factors of V cholerae is controlled in a coordinated 2-component systems involving environmental sensors and as many as 20 chromosomal genes divided between a pathogenicity island (PAI) containing CT and one containing TCP. The chief regulator is a transmembrane protein (ToxR) that “senses” environmental changes in pH, osmolarity, and temperature, which convert it to an active form. In the active state, ToxR can directly turn on CT genes as well as activate transcription of a second regulatory protein, ToxT. ToxT, whose natural effector may be bile, then activates transcription of virulence genes in both PAIs, including TCP and CT. Another set of environmental sensors switch V cholerae from free-swimming forms to the sessile, biofilm-forming state associated with environmental persistence in crustaceans. Quorum-sensing systems deploy expression of these virulence genes at a time when a critical mass of V cholerae is present to sustain it.

Nonspecific defenses such as gastric acidity, gut motility, and intestinal mucus are important in preventing colonization with V cholerae. For example, in persons who lack gastric acidity (gastrectomy or achlorhydria from malnutrition), the attack rate of clinical cholera is higher. The immune state has been most strongly associated with sIgA directed against O-antigen LPS, CT (B subunit), and TCP. The precise protective mechanisms remain to be established.

CHOLERA: CLINICAL ASPECTS

Typical cholera has a rapid onset, beginning with abdominal fullness and discomfort, rushes of peristalsis, and loose stools. Vomiting may also occur. The stools quickly become watery, voluminous, almost odorless, and contain mucus flecks, giving it an appearance called rice-water stools. Neither white blood cells nor blood are in the stools, and the patient is afebrile. Clinical features of cholera result from the extensive fluid loss and electrolyte imbalance, which can lead to extreme dehydration, hypotension, and death within hours if untreated. No other disease produces dehydration as rapidly as cholera.

The initial suspicion of cholera depends on recognition of the typical clinical features in an appropriate epidemiologic setting. A bacteriologic diagnosis is accomplished by isolation of V cholerae from the stool. The organism grows on common clinical laboratory media such as blood agar and MacConkey agar, but its isolation is enhanced by a selective medium (thiosulfate–citrate–bile salt–sucrose agar). Once isolated, the organism is readily identified by biochemical reactions. Outside cholera-endemic areas, the selective medium is not routinely used for stool cultures, so clinical laboratories must be alerted to the suspicion of cholera.

The outcome of cholera depends on balancing the diarrheal fluid and ionic losses with adequate fluid and electrolyte replacement. This is accomplished by oral and/or intravenous administration of solutions of glucose with near physiologic concentrations of sodium and chloride and higher than physiologic concentrations of potassium and bicarbonate. Exact formulas are available as dried packets to which a given volume of water is added. Oral replacement, particularly if begun early, is sufficient for all but the most severe cases and has substantially reduced the mortality from cholera. Antimicrobial therapy plays a secondary role to fluid replacement by shortening the duration of diarrhea and magnitude of fluid loss. A single dose of azithromycin provides optimal antimicrobial therapy, but doxycycline, a fluoroquinolone, or trimethoprim-sulfamethoxazole are also effective agents.

Epidemic cholera, a disease of poor sanitation, does not persist where treatment and disposal of human waste are adequate. Because good sanitary conditions do not exist in much of the world, secondary local measures such as boiling and chlorination of water during epidemics are required. Cholera associated with ingestion of crabs and shrimp can be prevented by adequate cooking (10 minutes) and avoidance of recontamination from containers and surfaces. Vaccines prepared from whole cells, lipopolysaccharide, and CT B subunit have been disappointing, providing protection that is not longlasting. Current interest includes live attenuated vaccine strains because of their potential to stimulate the local sIgA immune response.

Species of Vibrio other than V cholerae may still produce disease, but are uncommon and typically restricted to seacoast locales. Vibrio parahaemolyticus produces a diarrheal illness after ingestion of raw or inadequately cooked seafood due to the production of a pair of its own enterotoxins. For virulence V vulnificus stands out because it can produce a rapidly progressive cellulitis in wounds sustained in seawater as well as a fatal bacteremic infection after ingestion of raw seafood. The latter has been common enough in Florida to threaten the local oyster trade. Cases were also seen in the area devastated by hurricane Katrina. Vibrio vulnificus is also a spectacular scavenger of host iron stores and produces particularly fulminant disease in persons with iron-overload states (eg, thalassemia, hemochromatosis). Features of these and other less common vibrios are shown in Table 32–2.

Campylobacters are motile, curved, oxidase-positive, Gram-negative rods similar in morphology to vibrios. The cells have polar flagella and are often attached at their ends giving pairs “S” shapes or a “seagull” appearance. More than a dozen Campylobacter species have been associated with human disease. Of these, C jejuni is by far the most common and is discussed here as the prototype for intestinal disease. The features of other species are summarized in Table 32–2.

BACTERIOLOGY: CAMPYLOBACTER JEJUNI

Before 1973, C jejuni was not recognized as a cause of human disease. Not until selective methods for its isolation were developed was it recognized as one of the most common causes of infectious diarrhea. Like other campylobacters, C jejuni grows well only on enriched media under microaerophilic conditions. That is, it requires oxygen at reduced tension (5%-10%), presumably because of the vulnerability of some of its enzyme systems to superoxides. Growth usually requires 2 to 4 days, sometimes as much as 1 week. Campylobacter jejuni has the structural components found in other Gram-negative bacteria (eg, outer membrane, LPS and LOS). The cells are actively motile through the action of a polar flagellum. In contrast to the vibrios, C jejuni does not break down carbohydrates, but uses amino acids and metabolic intermediates for energy. It is one of a number of pathogens that produce a membrane-bound protein called cytolethal distending toxin (CDT). CDT has an A/B toxin structure in which the A subunit is able to cause cell cycle arrest.

CAMPYLOBACTER ENTERITIS

EPIDEMIOLOGY

It is humbling to consider how a pathogen as common as C jejuni could have been missed for decades. Rates of campylobacteriosis vary widely around the world but at 4% to 30% of diarrheal stools, it is the leading cause of gastrointestinal infection in developed countries. Over 2 million cases occur each year in the United States at a rate roughly double that of Salmonella, the second most common bacterial enteric pathogen. This high rate of disease is facilitated by the low infecting dose of C jejuni—only a few hundred cells.

The primary reservoir is in animals, and the bacteria are transmitted to humans by ingestion of contaminated food or by direct contact with pets. Campylobacters are commonly found in the normal gastrointestinal and genitourinary flora of warm-blooded animals, including sheep, cattle, chickens, wild birds, and many others. Domestic animals such as dogs may also carry the organisms and probably play a significant role in transmission to humans. The most common source of human infection is undercooked poultry, but outbreaks have been caused by contaminated rural water supplies and unpasteurized milk often consumed as a “natural” food. Sometimes a direct association can be made as with a household pet, particularly a new puppy just brought home from a kennel.

PATHOGENESIS

Infection is established by oral ingestion, followed by colonization of the intestinal mucosa. Adherence to enterocytes is facilitated by action of the flagellum followed by entrance into cells in endocytotic vacuoles. Once inside, they move in association with the cell's microtubule structure, rather than the actin microfilaments associated with some other invasive bacteria. Candidate injury mechanisms include the cytotoxic CDT and the action of lipooligosaccharides (LOS) released in outer membrane vesicles. The intestinal pathology is that of an invasive pathogen with acute inflammation, crypt abscesses, and occasional seeding of the bloodstream.

There is an association between C jejuni infection and Guillain-Barré syndrome, an acute demyelinating neuropathy that is frequently preceded by an infection. Although C jejuni is not the only antecedent to this syndrome, it is the most common of identifiable causes. Up to 40% of patients have culture or serologic evidence of Campylobacter infection at the time the neurologic symptoms occur. The mechanism is a type II hypersensitivity involving antibody elicited by epitopes in the C jejuni outer membrane LOS that cross-react with host peripheral nerve myelin gangliosides. These antiganglioside antibodies are found in the serum of patients with Guillain-Barré syndrome motor neuropathies. This molecular mimicry is similar to the mechanism of group A streptococcal rheumatic fever.

IMMUNITY

Acquired immunity after natural infection with C jejuni has been demonstrated in volunteer studies, but the mechanisms involved are unknown. Secretory and serum IgA are formed in the weeks after infection but decline thereafter. The high rate of Campylobacter infection in patients with AIDS suggests the importance of cellular immune mechanisms.

CAMPYLOBACTEROSIS: CLINICAL ASPECTS

MANIFESTATIONS AND DIAGNOSIS

The illness typically begins 1 to 7 days after ingestion, with fever and lower abdominal pain that may be severe enough to mimic acute appendicitis. These are followed within hours by dysenteric stools that usually contain blood and pus. The illness is typically self-limiting after 3 to 5 days but may last 1 to 2 weeks. The diagnosis is confirmed by isolation of the organism from the stool. This requires a special medium made selective for Campylobacter by inclusion of antimicrobials that inhibit the normal facultative flora of the bowel. Plates must be incubated in a microaerophilic atmosphere, which can now be conveniently generated in a sealed jar by hydration of commercial packs similar to those used for anaerobes.

TREATMENT

Since less than 50% of patients clearly benefit from antimicrobial therapy, cases of Campylobacter infection are usually not treated unless the disease is severe or prolonged (lasting longer than 1 week). Campylobacter jejuni is typically susceptible to macrolides and fluoroquinolones but resistant to β-lactams. Erythromycin or azithromycin is considered the treatment of choice but must be given early for maximal effect. Fluoroquinolones are also effective, but resistance is becoming more common, especially in patients with HIV infection who have difficulty clearing the organism despite treatment.

In 1983, a pair of Australian microbiologists (Warren and Marshall) suggested that gastritis and peptic ulcers were infectious diseases, contradicting long-held beliefs concerning their epidemiology, pathogenesis, and treatment. In the same year, the 10th edition of Harrison's Principles of Internal Medicine described peptic ulcers as due to an unfavorable balance between gastric acid–pepsin secretion and gastric or duodenal mucosal resistance. Underlying causes cited included genetic and lifestyle (smoking) as well as psychologic factors (anxiety, stress). Treatment with bismuth salts, antacids, and inhibitors of acid secretion gave relief but not cure. Relapsing patients (50%-80%) were subjected to surgical treatments (vagotomy, partial gastrectomy), which had their own set of complications (reflux, afferent loop syndrome, dumping syndrome). All of this was logical and supported by clinical observations and research studies, but was simply incorrect. The bacteria now called Helicobacter had been observed but dismissed because they were so common and its urease was once considered a secretory product of the stomach itself. The Nobel Prize-winning studies that stimulated the reversal of this dogma have led to cures using antibacterial agents and new ideas linking Helicobacter infection to cancer. This experience has also left us with a sense that we can never be smug about what we “know” in medicine.

BACTERIOLOGY: HELICOBACTER PYLORI

Helicobacter pylori has morphologic and growth similarities to the campylobacters, with which they were originally classified. The cells are slender, curved rods with polar flagella. The cell wall structure is typical of other Gram-negative bacteria. Growth requires a microaerophilic atmosphere and is slow (3–5 days). The cells are rapidly motile due to the action of multiple polar flagella.

A number of unique bacteriologic features have been found in H pylori. The most distinctive is a urease whose action allows the organism to persist in low pH environments by the generation of ammonia. The urease is produced in amounts so great (6% of bacterial protein) that its action can be demonstrated within minutes of placing H pylori in the presence of urea. Another secreted protein called the vacuolating cytotoxin (VacA) causes apoptosis in eukaryotic cells it enters generating multiple large cytoplasmic vacuoles (Figure 32–3). The vacuoles are felt to be generated by the toxin's formation of channels in lysosomal and endosomal membranes. Another protein, CagA, induces changes in multiple cellular proteins and has a strong association with virulence. Both VacA and CagA are delivered to cells by injection secretion systems (type IV). The genes for CagA and the components of its secretion system are located in a large PAI.

HELICOBACTER GASTRITIS

EPIDEMIOLOGY

Infection with H pylori causes what is perhaps the most prevalent disease in the world. The organism is found in the stomachs of 30% to 50% of adults in developed countries, and it is almost universal in developing countries. The exact mode of transmission is not known, but is presumed to be person to person by the fecal–oral route or by contact with gastric secretions in some way. Colonization increases progressively with age, and children are believed to be the major amplifiers of H pylori in human populations. A declining prevalence in developed countries may be due to decreased transmission because of less crowding and frequent exposure to antimicrobial agents.

Once established, the same strain persists for years, decades, even for life. Molecular epidemiologic analysis indicates the strains themselves have strong linkages to ethnic origins that can be traced back to the earliest known patterns of human migration. Helicobacter pylori has been called an “accidental tourist,” which was established in the stomachs of humans thousands of years ago and remained bound to the original population as it dispersed from continent to continent.

Helicobacter pylori is the most common precursor of gastritis, gastric ulcer, and duodenal ulcer cases which are not due to drugs. In addition Helicobacter gastritis caused by Cag⁺ strains is acknowledged to be an antecedent of gastric adenocarcinoma, one of the most common causes of cancer death in the world. It is also linked to a gastric mucosa-associated lymphoid tissue (MALT) lymphoma, which is less common but shows the striking property of regressing with antimicrobial therapy. Helicobacter pylori gained the dubious distinction of being the first bacterium declared a class I carcinogen by the World Health Organization.

Helicobacter pylori is exclusive to humans, but other species have been found in the stomachs of a wide range of animals, where they are also associated with gastritis. It is difficult to imagine the old “stress ulcer” theories surviving the discovery of a cheetah with Helicobacter gastritis. Speculation that domestic animals may serve as a reservoir for human infection has not been confirmed.

PATHOGENESIS

To persist in the hostile environs of the stomach, H pylori uses many mechanisms to adhere to the gastric mucosa and survive the acid milieu of the stomach (Figure 32–4). Motility provided by the flagella allows the organisms to swim to the less acidic locale beneath the gastric mucus, where the urease further creates a more neutral microenvironment by ammonia production. Urease production is regulated in response to changes in the gastric acidity such as rises to a pH as high as 6.0 following the buffering effect of meals. At the mucosa, adherence is mediated by multiple outer membrane proteins which bind to the surface of gastric epithelial cells and certain erythrocyte antigens (Lewis b).

Helicobacter pylori colonization is almost always accompanied by a cellular infiltrate ranging from minimal mononuclear infiltration of the lamina propria to extensive inflammation with neutrophils, lymphocytes, and microabscess formation. Both gastritis and duodenal ulcers are most strongly associated with colonization of the antrum area of the stomach. The inflammation may be due to toxic effects of the urease or the VacA transported into the gastric epithelial cells by the secretion system. Inside the cell, VacA causes vacuolization of the endosomal compartment and has other effects including altered T-cell function. The CagA protein is injected into the gastric epithelial cell by the secretion system, where it triggers multiple enzymatic reactions including those that cause reorganization of the actin cytoskeleton and stimulation of cytokines. Variations in the genes contained in the PAI generate a mixed population of H pylori cells particularly in relation to the multiple properties of CagA. Added together urease, CagA, and VacA provide ample explanation for the gastritis that is universal in H pylori infection. This prolonged and aggressive inflammatory response could lead to epithelial cell death and ulcers. The progression from gastritis to ulcer remains to be explained although a duodenal ulcer-promoting gene has been identified.

That decades of inflammation and assault by the virulence factors just described could cause metaplasia, and eventually cancer seems logical, but the specific mechanisms of carcinogenesis have only recently been explored. CagA, for example, has been shown to trigger a cascade of interactions leading to growth-promoting oncogenic signals. The gastric lymphomas may represent neoplastic transformation of B-lymphocyte clones proliferating in response to chronic antigenic stimulation. The discovery that H pylori colonization may affect hormones involved in glucose homeostasis has led to other hypotheses involving type 2 diabetes.

IMMUNITY

There is obviously little evidence of natural immunity in an infection that typically lasts for decades. The immunosuppressive effect of virulence factors such as VacA may be responsible in combination with yet to be discovered mechanisms.

HELICOBACTER DISEASE: CLINICAL ASPECTS

MANIFESTATIONS

Primary infection with H pylori is either silent or causes an illness with nausea and upper abdominal pain lasting up to 2 weeks. Years later, the findings of gastritis and peptic ulcer disease include nausea, anorexia, vomiting, epigastric pain, and even less specific symptoms such as belching. Many patients are asymptomatic for decades, even up to perforation of an ulcer. Perforation can lead to extensive bleeding and peritonitis due to the leakage of gastric contents into the peritoneal cavity.

DIAGNOSIS

The most sensitive means of diagnosis is endoscopic examination, with biopsy and culture of the gastric mucosa. The H pylori urease is so potent that its activity can be directly demonstrated in biopsies in less than an hour. Noninvasive methods include serology and a urea breath test. For the breath test, the patient ingests ¹³C- or ¹⁴C-labeled urea, from which the urease in the stomach produces products that appear as labeled CO₂ in the breath. A number of methods for the detection of antibody directed against H pylori are now available. Because IgG or IgA remains elevated as long as the infection persists, these tests are valuable both for screening and for evaluation of therapy. The advantage of direct detection of the organism is that culture is the most sensitive indicator of cure following therapy.

TREATMENT AND PREVENTION

Helicobacter pylori is susceptible to a wide variety of antimicrobial agents. Bismuth salts (eg, Pepto-Bismol), which in the past were believed to act by coating the stomach, also have antimicrobial activity. Cure rates approaching 90% have been achieved with various combinations of bismuth salts and/or a protein pump inhibitor plus two antibiotics. Clarithromycin plus either amoxicillin or metronidazole and metronidazole plus tetracycline have been effective. Relapse rates are low, particularly when acid secretion is also controlled with the use of a proton pump inhibitor. These combination regimens must be continued for at least 2 weeks and may be difficult for some patients to tolerate. Prevention of H pylori disease awaits further understanding of transmission and immune mechanisms. Prophylactic treatment of asymptomatic persons colonized with H pylori is not yet recommended.