Patrick C. Seed
Escherichia coli is an important cause of intraintestinal and extraintestinal infections. Intraintestinal infections present as different diarrheal illnesses. Extraintestinal infections include disease of the urinary tract (see Chapter 553 ) and bloodstream (Chapters 129 , 202 , and 203 ). Intraintestinal pathogenic E. coli , also called enteric E. coli , produce diarrheal diseases. E. coli causing extra- and intraintestinal infections are highly specialized with unique genetic attributes that encode different sets of virulence factors and genetic programs. Extraintestinal pathogenic E. coli increasingly harbor multidrug resistances, including transferrable plasmids resulting in extended-spectrum β-lactamase (ESBL) production. This results in resistance to penicillins, cephalosporins, and aztreonam. Carbapenemase-bearing E. coli have also emerged, often in combination with multi–antibiotic class resistance, resulting in highly drug-resistant strains.
Escherichia coli species are members of the Enterobacteriaceae family. They are facultative anaerobic, gram-negative bacilli that usually ferment lactose. Most fecal E. coli organisms are commensal, are ubiquitous among humans starting in the 1st mo of life, and do not cause diarrhea. Six major groups of diarrheagenic E. coli pathotypes have been characterized on the basis of clinical, biochemical, and molecular-genetic criteria: enterotoxigenic E. coli (ETEC); enteroinvasive E. coli (EIEC); enteropathogenic E. coli (EPEC); Shiga toxin–producing E. coli (STEC), also known as enterohemorrhagic E. coli (EHEC) or verotoxin-producing E. coli (VTEC); enteroaggregative E. coli (EAEC or EggEC); and diffusely adherent E. coli (DAEC).
E. coli strains can also be categorized by their serogroup, where O refers to the lipopolysaccharide (LPS) O-antigen or serotype and H refers to the flagellar antigen, for example, E. coli O157:H7. However, because each pathotype contains many serotypes (e.g., 117 ETEC serotypes have been identified), and some serotypes can belong to more than 1 pathotype (e.g., O26:H11 can be either EPEC or EHEC, depending on which specific virulence genes are present), serotyping frequently does not provide definitive identification of pathotypes.
Because E. coli are normal fecal flora, pathogenicity is defined by demonstration of virulence characteristics and association of those traits with illness (Table 227.1 ). The mechanism by which E. coli produces diarrhea typically involves adherence of organisms to a glycoprotein or glycolipid receptor on a target intestinal cell, followed by production of a factor that injures or disturbs the function of intestinal cells. The genes for virulence properties and antibiotic resistance are often carried on transferable plasmids, pathogenicity islands, or bacteriophages. In the developing world, the various diarrheagenic E. coli strains cause frequent infections in the 1st years of life; diarrheagenic E. coli as a group are responsible for 30-40% of all diarrhea cases in children worldwide. They occur with increased frequency during the warm months in temperate climates and during rainy season months in tropical climates. Most diarrheagenic E. coli strains (except STEC) require a large inoculum of organisms to induce disease, thus necessitating exposure to grossly contaminated ingestible materials. Infection is most likely when food-handling or sewage-disposal practices are suboptimal. The diarrheagenic E. coli pathotypes are also important in North America and Europe, although their epidemiology is less well defined in these areas than in the developing world. In North America, the various diarrheagenic E. coli strains may cause as much as 30% of infectious diarrhea in children <5 yr old.
Table 227.1
Clinical Characteristics, Pathogenesis, and Diagnosis of Diarrheagenic E. coli
| PATHOGEN | POPULATIONS AT RISK | CHARACTERISTICS OF DIARRHEA | MAIN VIRULENCE FACTORS | DIAGNOSIS | |||
|---|---|---|---|---|---|---|---|
| Watery | Bloody | Duration | Adherence Factors | Toxins | |||
| ETEC | >1 yr old and travelers | +++ | — | Acute | Colonization factor antigens (CFs or CFAs); ECP |
Heat-labile enterotoxin (LT) Heat-stable enterotoxin (ST) |
Detection of enterotoxins (LT and ST) by enzyme immunoassays or PCR (lt, st ) |
| EIEC | >1 yr old | + | ++ | Acute | Invasion plasmid antigen (IpaABCD) | Detection of invasion plasmid antigen of Shigella (ipaH ) by PCR | |
| EPEC | <2 yr old | +++ | + | Acute, prolonged or persistent | A/E lesion, intimin/Tir, EspABD, Bfp | EspF, Map, EAST1, SPATEs (EspC ) | Detection of intimin gene (eae ) ± bundle-forming pili (bfp A) by PCR, and absence of Shiga toxins; HEp-2 cells adherence assay (LA, LLA) |
| STEC (EHEC/VTEC) | 6 mo-10 yr and elderly persons | + | +++ | Acute | A/E lesion, intimin/Tir, EspABD | Shiga toxins (Stx1, Stx2, and variants of Stx2) | Detection of Shiga toxins by enzyme immunoassays or PCR (Stx 1, Stx 2); stool culture on MacConkey-sorbitol media to detect E. coli O157. Simultaneous culture for O157 and nonculture assays to detect Shiga toxins |
| EAEC | <2 yr old, HIV-infected patients, and travelers | +++ | + | Acute, prolonged, or persistent | Aggregative adherence fimbriae (AAF) | SPATEs (Pic, Pet), ShET1, EAST1 | Detection of AggR , AA plasmid, and other virulence genes: aap , aa tA , astA , set1A by PCR; HEp-2 cells adherence assay (AA) |
| DAEC | >1 yr old and travelers | ++ | — | Acute | Afa/Dr, AIDA-I | SPATEs (Sat) | Detection of Dr adhesins (daaC or daaD) and Dr-associated genes by PCR; HEp-2 cells adherence assay (DA) |
—, Not present; +, present; ++, common; +++, very common; A/E lesion, attaching and effacing lesion; AA, aggregative adherence; Bfp, bundle-forming pili; DA, diffuse adherence; DAEC, diffusely adherent E. coli; EAEC, enteroaggregative E. coli; EAST1, enteroaggregative heat-stable toxin; ECP, E. coli common pilus; EHEC, enterohemorrhagic E. coli; EIEC, enteroinvasive E. coli; EPEC, enteropathogenic E. coli ; EspABD, E. coli –secreted proteins A, B, and D; ETEC, enterotoxigenic E. coli; LA, localized adherence; LLA, localized-like adherence; PCR, polymerase chain reaction; ShET1, Shigella enterotoxin 1; SPATEs, serine-protease autotransporter of Enterobacteriaceae; STEC, Shiga toxin–producing E. coli ; Tir, translocated intimin receptor; VTEC, verotoxin-producing E. coli.
Many studies have found diarrheagenic E. coli pathotypes in a significant proportion of asymptomatic healthy children living in developing countries. Fecal contamination (human and animal), which is common in the low-resource environments where many young children live, facilitates the transmission of pathogens. Also, with modern, highly sensitive microbiologic methods, small numbers of bacteria can be detected in stool samples. Therefore, it is important to assess the prevalence of various enteropathogens in children with and without diarrhea to interpret results. Excretion of enteropathogens by children without diarrhea may be explained by characteristics of the pathogens (virulence heterogeneity), the host (host susceptibility, age, nutritional status, breastfeeding, immunity), and environmental factors (inoculum size).
ETEC accounts for a sizable fraction of dehydrating infantile diarrhea in the developing world (10–30%) and of traveler's diarrhea (20–60% of cases); ETEC is the most common cause of traveler's diarrhea. In the Global Enteric Multicenter Study (GEMS) conducted across Asia and Africa, heat-stable enterotoxin (ST)–expressing ETEC (with or without coexpression of heat-labile enterotoxin [LT]) was among the most important causes of diarrhea in young children in developing countries and was associated with increased risk of death. The typical signs and symptoms include explosive watery, nonmucoid, nonbloody diarrhea; abdominal pain; nausea; vomiting; and little or no fever. The illness is usually self-limited and resolves in 3-5 days but occasionally lasts >1 wk.
ETEC causes few or no structural alterations in the gut mucosa. Diarrhea follows colonization of the small intestine and elaboration of enterotoxins. ETEC strains secrete an LT and/or an ST. LT, a large molecule consisting of 5 receptor-binding subunits and 1 enzymatically active subunit, is structurally, functionally, and neutralizing antibody cross-reactive with cholera toxin produced by Vibrio cholerae . LT stimulates adenylate cyclase, resulting in increased cyclic adenosine monophosphate. ST is a small molecule not related to cholera toxin. ST stimulates guanylate cyclase, resulting in increased cyclic guanosine monophosphate. Each toxin induces ion and water secretion into the intestinal lumen, resulting in profuse watery diarrhea. The genes for these toxins are encoded on plasmids.
Colonization of the intestine requires fimbrial colonization factor antigens (CFAs) , which promote adhesion to the intestinal epithelium. Over 25 CFA types exist and can be expressed alone or in combinations. Prevalent colonization factors include CFA/I, CS1-CS7, CS14, and CS17. However, CFAs have not been detected on all ETEC strains. Although 30–50% of ETEC isolates have no characterized CFA by phenotypic screening, novel CFAs continue to be identified. CFAs are highly immunogenic. However, the multiple CFAs and their allelic variants have made the definition of immunity and development of useful vaccines difficult. A large proportion of strains produce a type IV pilus called longus , which functions as a colonization factor and is found among several other gram-negative bacterial pathogens. ETEC strains also have the common pilus, produced by commensal and pathogenic E. coli strains. Among the nonfimbrial adhesions, TibA is a potent bacterial adhesin that mediates bacterial attachment and invasion of cells. For many years, the O serogroup was used to distinguish pathogenic from commensal E. coli . Because the pathogenic E. coli are now defined and classified by using probes or primers for specific virulence genes, determining the O serogroup has become less important. Of the >180 E. coli serogroups, only a relatively small number typically are ETEC. The most common O groups are O6, O8, O128, and O153, and based on some large retrospective studies, these serogroups account for only half the ETEC strains.
Clinically, EIEC infections present either with watery diarrhea or a dysentery syndrome with blood, mucus, and leukocytes in the stools, as well as fever, systemic toxicity, crampy abdominal pain, tenesmus, and urgency. The illness resembles bacillary dysentery because EIEC shares virulence genes with Shigella spp. Sequencing of multiple housekeeping genes indicates that EIEC is more related to Shigella than to noninvasive E. coli . EIEC diarrhea occurs mostly in outbreaks; however, endemic disease occurs in developing countries. In some areas of the developing world as many as 5% of sporadic diarrhea episodes and 20% of bloody diarrhea cases are caused by EIEC (see Chapter 226 ).
EIEC disease resembles shigellosis . EIEC cause colonic lesions with ulcerations, hemorrhage, mucosal and submucosal edema, and infiltration by polymorphonuclear leukocytes (PMNs). EIEC strains behave like Shigella in their capacity to invade gut epithelium and produce a dysentery-like illness. The invasive process involves initial entry into cells, intracellular multiplication, intracellular and intercellular spread, and host cell death. All bacterial genes necessary for entry into the host cell are clustered within a 30-kb region of a large virulence plasmid; these genes are closely related to those found on the invasion plasmid of Shigella spp. This region carries genes encoding the entry-mediating proteins, including proteins that form a needle-like injection apparatus called type III secretion, required for secreting the invasins (IpaA-D and IpgD). The Ipas are the primary effector proteins of epithelial cell invasion. The type III secretion apparatus is a system triggered by contact with host cells; bacteria use it to transport proteins into the host cell plasma membrane and inject toxins into the host cell cytoplasm.
EIEC encompasses a small number of serogroups (O28ac, O29, O112ac, O124, O136, O143, O144, O152, O159, O164, O167, and some untypeable strains). These serogroups have LPS antigens related to Shigella LPS, and as with shigellae, are nonmotile (they lack H or flagellar antigens) and are usually non–lactose fermenting.
EPEC causes acute, prolonged, and persistent diarrhea, primarily in children <2 yr old in developing countries, where the organism may account for 20% of infant diarrhea. In developed countries, EPEC cause occasional daycare center and pediatric ward outbreaks. Profuse watery, nonbloody diarrhea with mucus, vomiting, and low-grade fever are common symptoms. Prolonged diarrhea (>7 days) and persistent diarrhea (>14 days) can lead to malnutrition , a potentially mortality-associated outcome of EPEC infection in infants in the developing world. Studies show that breastfeeding is protective against diarrhea caused by EPEC.
EPEC colonization causes blunting of intestinal villi, local inflammatory changes, and sloughing of superficial mucosal cells; EPEC-induced lesions extend from the duodenum through the colon. EPEC induces a characteristic attaching and effacing histopathologic lesion, which is defined by the intimate attachment of bacteria to the epithelial surface and effacement of host cell microvilli. Factors responsible for the attaching and effacing lesion formation are encoded by the locus of enterocyte effacement (LEE), a pathogenicity island with genes for a type III secretion system, the translocated intimin receptor (Tir) and intimin, and multiple effector proteins such as the E. coli –secreted proteins (EspA-B-D). Some strains adhere to the host intestinal epithelium in a pattern known as localized adherence , a trait that is mediated in part by the type IV bundle-forming pilus (Bfp) encoded by a plasmid (the EAF plasmid). After initial contact, proteins are translocated through filamentous appendages forming a physical bridge between the bacteria and the host cell; bacterial effectors (EspB, EspD, Tir) are translocated through these conduits. Tir moves to the surface of host cells, where it is bound by a bacterial outer membrane protein intimin (encoded by the eae gene). Intimin-Tir binding triggers polymerization of actin and other cytoskeletal components at the site of attachment. These cytoskeletal changes result in intimate bacterial attachment to the host cell, enterocyte effacement, and pedestal formation.
Other LEE-encoded effectors include Map, EspF, EspG, EspH, and SepZ. Various other effector proteins are encoded outside the LEE and secreted by the type III secretion system (the non–LEE-encoded proteins, or Nle). The contribution of these putative effectors (NleA/EspI, NleB, NleC, NleD, etc.) to virulence is still under investigation. The presence and expression of virulence genes vary among EPEC strains.
The eae (intimin) and bfp A genes are useful for identifying EPEC and for subdividing this group of bacteria into typical and atypical strains. E. coli strains that are eae + /bfp A+ are classified as “typical” EPEC; most of these strains belong to common O:H serotypes. E. coli strains that are eae + /bfp A− are classified as “atypical” EPEC. Typical EPEC has been considered for many years to be a leading cause of infantile diarrhea in developing countries and was considered rare in industrialized countries. However, current data suggest that atypical EPEC are more prevalent than typical EPEC in both developed and developing countries, even in persistent diarrhea cases. Determining which of these heterogeneous strains are true pathogens remains a work in progress. In the GEMS, typical EPEC was the main pathogen associated with increased risk of mortality, particularly in infants in Africa.
The classic EPEC serogroups include strains of 12 O serogroups: O26, O55, O86, O111, O114, O119, O125, O126, O127, O128, O142, and O158. However, various E. coli strains defined as EPEC based on the presence of the intimin gene belong to nonclassic EPEC serogroups, especially the atypical strains.
STEC causes a broad spectrum of diseases. STEC infections may be asymptomatic. Patients who develop intestinal symptoms can have mild diarrhea or severe hemorrhagic colitis. Abdominal pain with initially watery diarrhea that may become bloody over several days characterizes STEC illness. Infrequent fever differentiates STEC disease from the otherwise similar appearance of shigellosis or EIEC disease. Most persons with STEC recover from the infection without further complication. However, 5–10% of children with STEC hemorrhagic colitis go on within a few days to develop systemic complications such as hemolytic-uremic syndrome (HUS), characterized by acute kidney failure, thrombocytopenia, and microangiopathic hemolytic anemia (see Chapter 538 ). Severe illness occurs most often among children 6 mo to 10 yr old. Young children with STEC-associated bloody diarrhea and neutrophilic leukocytosis in the early course of their diarrhea are at risk for HUS progression. Older individuals can also develop HUS or thrombotic thrombocytopenic purpura.
STEC is transmitted person to person (e.g., in families and daycare centers) as well as by food and water; ingestion of a small number of organisms is sufficient to cause disease with some strains. Poorly cooked hamburger is a common cause of food-borne outbreaks, although many other foods (apple cider, lettuce, spinach, mayonnaise, salami, dry fermented sausage, and unpasteurized dairy products) have also been incriminated in STEC transmission.
STEC affects the colon most severely. These organisms adhere to intestinal cells, and most strains that affect humans produce attaching-effacing lesions such as those seen with EPEC and contain related genes (e.g., intimin, Tir, EspA-D ). Unlike EPEC, STEC produces Shiga toxins (Stx; previously called verotoxins and Shiga- like) as key virulence factors. There are 2 major Shiga toxin families, Stx1 and Stx2, with multiple subtypes identified by letters (e.g., Stx2a, Stx2c). Some STEC produce only Stx1, and others produce only 1 of the variants of Stx2; many STEC have genes for several toxins. Stx1 is essentially identical to Shiga toxin, the protein synthesis–inhibiting exotoxin of Shigella dysenteriae serotype 1. Stx2 and variants of Stx2 are more distantly related to Shiga toxin, although they share conserved sequences.
These ETEC Shiga toxins are composed of a single A subunit noncovalently associated with a pentamer composed of identical B subunits. The B subunits bind to globotriaosylceramide (Gb3 ), a glycosphingolipid receptor on host cells. The A subunit is taken up by endocytosis. The toxin target is the 28S rRNA, which is depurated by the toxin at a specific adenine residue, causing protein synthesis to cease and affected cells to die. These toxins are carried on bacteriophages that are normally inactive (lysogenic) in the bacterial chromosome; when the phages are induced to replicate (e.g., by the stress induced by many antibiotics), they cause lysis of the bacteria and release of large amounts of toxin. Toxin translocation across the intestinal epithelium into the systemic circulation can lead to damage of vascular endothelial cells, resulting in activation of the coagulation cascade, formation of microthrombi, intravascular hemolysis, and ischemia.
The clinical outcome of an STEC infection depends on a strain-specific combination of epithelial attachment and the toxin factors. The Stx2 family of toxins is associated with a higher risk of causing HUS. Strains that make only Stx1 often cause only watery diarrhea and are infrequently associated with HUS.
The most common STEC serotypes are E. coli O157:H7, E. coli O111:NM, and E. coli O26:H11, although several hundred other STEC serotypes have also been described. E. coli O157:H7 is the most virulent serotype and the serotype most frequently associated with HUS; however, other non-O157 serotypes also cause this illness.
EAEC is associated with (1) acute, prolonged and persistent pediatric diarrhea in developing countries, most prominently in children <2 yr old and in malnourished children; (2) acute and persistent diarrhea in HIV-infected adults and children; and (3) acute traveler's diarrhea; EAEC is the 2nd most common cause of traveler's diarrhea after ETEC. Typical EAEC illness is manifested by watery, mucoid, secretory diarrhea with low-grade fever and little or no vomiting. The watery diarrhea can persist for ≥14 days. In some studies, many patients have grossly bloody stools, indicating that EAEC cannot be excluded on stool characteristics. EAEC strains are associated with growth retardation and malnutrition in infants in the developing world.
EAEC organisms form a characteristic biofilm on the intestinal mucosa and induce shortening of the villi, hemorrhagic necrosis, and inflammatory responses. The proposed model of pathogenesis of EAEC infection involves 3 phases: adherence to the intestinal mucosa by way of the aggregative adherence fimbriae or related adhesins; enhanced production of mucus; and production of toxins and inflammation that results in damage to the mucosa and intestinal secretion. Diarrhea caused by EAEC is predominantly secretory. The intestinal inflammatory response (elevated fecal lactoferrin, interleukin [IL]-8 and IL-1β) may be related to growth impairment and malnutrition.
EAEC strains are recognized by adherence to HEp-2 cells in an aggregative, stacked-brick pattern, called aggregative adherence (AA). EAEC virulence factors include the AA fimbriae (AAF-I, -II, and -III) that confer the AA phenotype. Some strains produce toxins, including the plasmid-encoded enterotoxin EAST1 (encoded by ast A), a homolog of the ETEC ST; an autotransporter toxin called Pet; other STATE toxins; and the chromosomally encoded enterotoxin ShET1 (encoded by setA and setB ). Other virulence factors include outer membrane and secreted proteins, such as dispersin (aap ), and the dispersin transport complex (aatPABCD). EAEC is a heterogeneous group of E. coli . The original diagnostic criteria (HEp-2 cell adherence pattern) identified many strains that are probably not true pathogens; genetic criteria appear to more reliably identify true pathogens. A transcriptional activator called AggR controls the expression of plasmid-borne and chromosomal virulence factors. Identification of AggR appears to reliably identify illness-associated pathogenic EAEC strains (“typical” EAEC). EAEC agg R-positive strains carrying 1-3 of the genes aap , ast A, and set1A are significantly associated with diarrhea compared with EAEC isolates lacking these genes. Other than the factors AAF and AggR, EAEC strains are genetically diverse and thus display variable virulence. EAEC strains belong to multiple serogroups, including O3, O7, O15, O44, O77, O86, O126, and O127.
Although the status of DAEC strains as true pathogens has been in doubt, multiple studies in both developed and developing countries have associated these organisms with diarrhea, particularly in children after the 1st 1-2 yr of life. DAEC strains isolated from children and adults seem to represent 2 different bacterial populations. Age-dependent susceptibility may explain discrepancies among epidemiologic studies to diarrhea or by the use of inappropriate detection methods. Data suggest that these organisms also cause traveler's diarrhea in adults. DAEC produces acute watery diarrhea that is usually not dysenteric but is often prolonged.
DAEC strains produce diffuse adherence in cultured epithelial cells. They express surface fimbriae (designated F1845) that are responsible for the diffuse adherence phenotype in a prototype strain. These fimbriae are homologous with members of the Afa/Dr family of adhesins, which are identified by hybridization with a specific probe, daaC , common to operons encoding Afa/Dr adhesions. A 2nd putative adhesin associated with the diffuse adherence pattern phenotype is an outer membrane protein, designated AIDA-I. The contribution of other putative effectors (icuA, fimH, afa, agg-3A, pap , astA, shET1) to virulence is still under investigation. The only documented secreted factor associated with DAEC infection is the serine-protease autotransporters of Enterobacteriaceae (SPATE) cytotoxin Sat. Bacteria expressing Afa/Dr adhesins interact with membrane-bound receptors, including decay-accelerating factor (DAF). The structural and functional lesions induced by DAEC include loss of microvilli and a decrease in the expression and enzyme activities of functional brush-border–associated proteins. Afa/Dr DAEC isolates produce a secreted autotransporter toxin that induces marked fluid accumulation in the intestine. DAEC strains typically induce IL-8 production in vitro. Serogroups of DAEC strains are less well defined than those of other diarrheagenic E. coli .
In 2011, a massive outbreak of an unusual O104:H4 strain of diarrheagenic E. coli began in Germany. Eventually, >4,000 individuals were sickened with hemorrhagic colitis; the outbreak involved primarily adults (<100 children were reported affected). More than 800 people developed HUS, and >50 of these individuals died. Genomic analysis suggested the outbreak strain was most closely related to EAEC and had acquired a lambdoid bacteriophage with genes for Shiga toxin Stx2a. It was thus a hybrid pathogen with colonization mechanisms similar to a typical EAEC strain and toxin production typical of an STEC strain. This outbreak strain carries Pic on the chromosome and a pAA-like plasmid encoding AAF, AggR, Pet, ShET1, and dispersin. A 2nd virulence plasmid encodes multiple antibiotic resistances. The high morbidity and mortality associated with this strain may reflect the stronger adherence of EAEC compared with STEC, delivering more Stx to target cells. Alternative terminology for this strain includes enteroaggregative hemorrhagic E. coli and Shiga toxin–producing EAEC . Whether Shiga toxin production in an EAEC background merits separate classification is unclear. Organisms with Shiga toxin genes in an atypical EPEC background were designated as a separate group (referred to as STEC , EHEC , or verotoxin-producing E. coli ) before the relative importance of the various genes was clear. EPEC strains are a heterogeneous group themselves. The important issue is not the nomenclature but rather the concept that virulence genes can move between E. coli , resulting in new variants.
The features of illness are seldom distinctive enough to allow confident diagnosis strictly on clinical observations, and routine laboratory studies such as blood counts rarely prove effective in the diagnosis. Practical, non–DNA-dependent methods for routine diagnosis of diarrheagenic E. coli have been developed primarily for STEC. Serotype O157:H7 is suggested by isolation of an E. coli that fails to ferment sorbitol on MacConkey sorbitol medium; latex agglutination confirms that the organism contains O157 LPS. Other STEC strains can be detected in routine hospital laboratories using commercially available enzyme immunoassay or latex agglutination assays to detect Shiga toxins, although the variable sensitivity of commercial immunoassays has limited their value.
Although some STEC (O157:H7 strains) can be detected in routine microbiology laboratories using selective media and appropriate antisera, the diagnosis of other diarrheagenic E. coli infection is traditionally made based on tissue culture assays (e.g., HEp-2-cells assay for EPEC, EAEC, DAEC) or identification of specific virulence factors of the bacteria by phenotype (e.g., toxins) or genotype. Multiplex, real-time, or conventional polymerase chain reaction (PCR) can be used for presumptive diagnosis of isolated E. coli colonies. The genes commonly used for diagnostic PCR are lt and st for ETEC; IpaH or iaL for EIEC; eae and bfp A for EPEC; eae, Stx1, and Stx2 for STEC; AggR or the AA plasmid for EAEC; and daaC or daaD for DAEC. Commercial assays such as the FilmArray Gastrointestinal Panel and Eurofins Diatherix Panel now detect genetic markers for EPEC, EAEC, ETEC, STEC, and EIEC, among other pathogen genes, directly from a fecal sample in several hours.
Serotyping does not provide definitive identification of pathotypes (except for selected cases such as O157:H7) because each pathotype contains many serotypes and some serotypes can belong to >1 pathotype. Consequently, serotyping should not be used routinely for diarrheagenic E. coli identification in clinical laboratories (e.g., to diagnose EPEC in infantile diarrhea), except during an outbreak investigation.
Other laboratory data are at best nonspecific indicators of etiology. Fecal leukocyte examination of the stool is often positive with EIEC or occasionally positive with other diarrheagenic E. coli . With EIEC and STEC there may be an elevated peripheral blood PMN count with a left shift. Determination of Stx2 blood levels in the early, postbloody diarrhea period may be useful to identify children at risk of HUS; however, this method requires further evaluation. Fecal lactoferrin, IL-8, and IL-1β can be used as inflammatory markers. Electrolyte changes are nonspecific, reflecting only fluid loss.
The cornerstone of management is appropriate fluid and electrolyte therapy. In general, this therapy should include oral replacement and maintenance with rehydration solutions such as those specified by the World Health Organization. Pedialyte and other readily available oral rehydration solutions are acceptable alternatives. After refeeding, continued supplementation with oral rehydration fluids is appropriate to prevent recurrence of dehydration. Early refeeding (within 6-8 hr of initiating rehydration) with breast milk or infant formula or solid foods should be encouraged. Prolonged withholding of feeding can lead to chronic diarrhea and malnutrition. If the child is malnourished, oral zinc should be given to speed recovery and decrease the risk of future diarrheal episodes.
Specific antimicrobial therapy of diarrheagenic E. coli is improving with accurate, rapid molecular diagnostic panels using direct fecal samples. However, the unpredictability of antibiotic susceptibilities remains problematic. Treatment is complicated by these organisms often being multiply resistant to antibiotics because of their previous exposure to inappropriate antibiotic therapy. Multiple studies in developing countries have found that diarrheagenic E. coli strains typically are resistant to antibiotics such as trimethoprim-sulfamethoxazole (TMP-SMX) and ampicillin (60–70%). Most data come from case series or clinical trials in adults with traveler's diarrhea. ETEC responds to antimicrobial agents such as TMP-SMX when the E. coli strains are susceptible. ETEC cases from traveler's diarrhea trials respond to ciprofloxacin, azithromycin, and rifaximin. However, other than for a child recently returning from travel in the developing world, empirical treatment of severe watery diarrhea with antibiotics is seldom appropriate.
In resource-poor settings where rapid molecular panel tests are not available, EIEC infections may be treated before culture results are finalized because the clinician suspects shigellosis and has begun empirical therapy. If the organisms prove to be susceptible, TMP-SMX is an appropriate choice. Although treatment of EPEC infection with TMP-SMX intravenously or orally for 5 days may be effective in speeding resolution, the lack of a rapid diagnostic test in the resource-poor setting makes treatment decisions difficult. Ciprofloxacin or rifaximin is useful for EAEC traveler's diarrhea, but pediatric data are sparse. Specific therapy for DAEC has not been defined.
The STEC strains represent a particularly difficult therapeutic dilemma; many antibiotics can induce bacterial stress, toxin production, and phage-mediated bacterial lysis with toxin release. Antibiotics should not be given for STEC infection because they can increase the risk of HUS (see Chapter 538 ). In settings with rapid molecular diagnostics, a delay in providing antibiotics is rarely consequential and can allow the clinician to more confidently recommend or exclude antibiotics from the therapeutic plan.
In the developing world, prevention of disease caused by pediatric diarrheagenic E. coli is probably best done by maintaining prolonged breastfeeding, paying careful attention to personal hygiene, and following proper food- and water-handling procedures. People traveling to these places can be best protected by handwashing, consuming only processed water, bottled beverages, breads, fruit juices, fruits that can be peeled, or foods that are served steaming hot.
Prophylactic antibiotic therapy is effective in adult travelers but has not been studied in children and is not recommended. Public health measures, including sewage disposal and food-handling practices, have made pathogens that require large inocula to produce illness relatively uncommon in industrialized countries. Food-borne outbreaks of STEC are a problem for which no adequate solution has been found. During the occasional hospital outbreak of EPEC disease, attention to enteric isolation precautions and cohorting may be critical.
Protective immunity against diarrheagenic E. coli remains an active area of research, and no vaccines are available for clinical use in children. Multiple vaccine candidates based on bacterial toxins and colonization factors have shown promise for prevention of ETEC in adult travelers, but long-term protection with these vaccines has not been optimal, particularly in children.