March 2009: Volume 35, Number 3
Successful Diagnosis and Management of E. coli 0157:H7
By William M. Janda, PhD, DABMM
One of several thrombotic microangiopathies, hemolytic-uremic syndrome (HUS) is characterized by a triad of thrombocytopenia, microangiophilic hemolytic anemia, and acute renal failure that occur simultaneously. HUS arises predominantly in children and represents the most common cause of acute renal failure in the pediatric age group. But more than 90% of cases of HUS are associated with infections of enterohemorrhagic E. coli strains, predominantly serotype 0157:H7, which cause bloody diarrhea. Other causes of diarrhea-associated HUS include gastrointestinal infections with enteric pathogens, primarily Shigella dysenteriae.
The remaining cases of HUS are termed atypical or non-diarrheal HUS. Atypical HUS has been associated with other bacteria, such as invasive Streptococcus pneumoniae infection. Atypical HUS also may occur following a viral infection or after treatment with certain antineoplastic agents (e.g., mitomycin, bleomycin) and anti-rejection medications administered post-transplant (e.g., tacrolimus, everolimus, cyclosporine) (1–3). Some cases of atypical HUS have been reported in the clinical settings of malignancy and pregnancy (4).
The most common type of microangiopathy with renal failure in adult patients, atypical HUS has a poorer prognosis than diarrhea-associated HUS. In addition, researchers have found that atypical HUS is associated with mutations in genes that code for proteins involved in regulation of the complement alternative pathway and in protection of host cells from complement activation (5).
Recently, the European Pediatric Study Group for HUS published a new classification system for HUS and related medical conditions, including thrombotic thrombocytopenic purpura (Table 1, below)(6). This review will focus on typical or diarrhea-associated HUS and the lab tests used in evaluation of symptomatic pateints.
Classification of Hemolytic-Uremic Syndrome
Level 1: Etiology Known
Level 2: Etiology Unknown
1 i. Infection induced
a. Shiga and shiga-like toxin producing bacteria (enterohemorrhagic E. coli, Shigella dysenteriae type 1)
b. Streptococcus pneumonia
1 ii. Disorders of complement regulation
1 iii. ADAMTS13 deficiency
1 iv. Defective cobalamine metabolism
1 v. Quinine-induced
2 i. HIV infection
2 ii. Malignancy, chemotherapy, ionizing radiation
2 iii. Calcineurin inhibitors and transplantation
2 iv. Pregnancy, HELLP syndrome, oral contraceptives
2 v. Systemic lupus erythematosus, anti-phospholipid antibody syndrome
2 vi. Glomerulonephropathy
2 vii. Familial included in Level 1
2 viii. Unclassified
This classification scheme from the European Pediatric Study Group for HUS includes one level for those cases of HUS where the etiology is known, while level 2 includes those conditions where a specific cause for HUS has not been delineated.
Source: Reference 6.
E. coli O157:H7: An Overview
E. coli O157:H7 was first recognized as a human pathogen in 1982 when the strain was linked to the report of two outbreaks of bloody diarrhea that occurred in Oregon and Michigan following the ingestion of contaminated hamburgers (7). Researchers recovered the strain from stool specimens of affected patients and from undercooked beef and identified Shiga-like toxins (Stxs) from E. coli serotype O157:H7. Several names for E. coli that produce these toxins have evolved, including enterohemorrhagic E. coli (EHEC), Shiga-like toxin-producing E. coli (STEC), and verotoxin-producing E. coli (VTEC) based on the organism’s cytotoxic effects on green monkey kidney (Vero) cells in culture.
Since these initial reports, outbreaks of gastroenteritis caused by the O157:H7 serotype have occurred in schools, day care centers, nursing homes, and entire communities. The sources of the organism in these outbreaks have been traced to ground beef, roast beef, salami, freshly prepared unpasteurized apple cider, and lake or untreated municipal water supplies. Researchers have also documented transmission resulting from direct contact with animals in petting zoos (8). Many additional serotypes of E. coli other than O157:H7 have subsequently been documented to produce Stxs and to cause diarrheal illness, hemorrhagic colitis, and HUS.
E. coli O157:H7 Associated Diarrheal Illnesses and HUS
A spectrum of clinical syndromes ranging from asymptomatic infection to hemorrhagic colitis with HUS may result following ingestion of E. coli O157:H7. In asymptomatic infections, the organism may be isolated from non-diarrheal stool specimens. An investigation of a Canadian outbreak of gastroenteritis found that 31% of 62 children who had been exposed to the organism source were asymptomatic. Ten of these 19 asymptomatic children had serologic evidence of prior infection (9). Such data suggests that the strain can also cause gastroenteritis with non-bloody diarrhea. It appears that individuals with non-bloody diarrhea are also less likely to progress to HUS; however, most infections with E. coli O157:H7 progress to hemorrhagic colitis.
With an incubation period ranging from 1 to 8 days, hemorrhagic colitis symptoms include severe abdominal cramping, chills, low-grade fever, right lower quadrant pain, and the passage of watery diarrhea that becomes grossly bloody, with mucus and fecal leukocytes. Upper gastrointestinal symptoms such as nausea and vomiting may occur early in the clinical course. Colonic biopsies demonstrate edema and erythema of the colonic mucosa, erosive mucosal lesions, capillary thrombi, and hemorrhage. The most notable abnormalities are found predominantly in the cecal, ascending, and transverse portions of the large intestine (10). Patients usually will show a peripheral leukocytosis with a left shift, while the hematocrit, erythrocyte sedimentation rate, electrolytes, liver function tests, and prothrombin times remain relatively normal or only moderately elevated. The acute diarrheal illness usually lasts from 3 to 8 days and persists in children for a longer period than in adults.
Most patients with diarrhea-associated HUS have the prodrome of gastroenteritis, with bloody diarrhea being the most common presenting symptom. They may also develop microangiopathic hemolytic anemia, thrombocytopenia, and renal failure then supervene. Neurologic complications are not a part of the classical triad seen with HUS, but central nervous system abnormalities such as lethargy and seizure activity occur in about 30% of patients (11). In studies related to outbreaks of E. coli O157:H7, the rate of progression from hemorrhagic colitis to HUS varies from as low as 3% to more than 30%, but researchers estimated that the rate of progression is in the range of 5% to 8% (12).
Indicators associated with development of HUS include leukocytosis on presentation, high fevers, and very young or advanced age. Among adults, HUS may progress to thrombotic thrombocytopenic purpura. These patients may also develop fever and neurologic symptoms along with the classic triad of HUS-associated symptoms. Tragically, 5% to 10% of children die from the infection, and mortality is also higher in elderly patients. In the latter population, infections are often related to co-morbid conditions, such as pneumonia, pulmonary edema, and mycocardial infarction.
HUS Pathology and Treatment
The histopathology of E. coli O157:H7-associated HUS suggests that disease pathogenesis involves direct vascular endothelial damage related to the systemic absorption of Stxs from the gastrointestinal lumen. Researchers also believe that inflammation of the colonic mucosa promotes absorption of the Stxs from the intestinal tract.
Studies in animals have shown that Stxs produce thrombotic, microangiophilic lesions that are histologically similar to those seen in patients with HUS. This damage activates coagulative events that cause the formation of intravascular thrombi. Ischemia caused by the platelet/fibrin thrombi in vascular tissues of the colon leads to hemorrhagic colitis. HUS develops when these platelet-fibrin thrombi form in the kidneys. The direct role of Stxs in the development of vascular endothelial damage, local intravascular coagulation, and fibrin deposition seen in HUS is also supported by the finding of elevated titers of neutralizing antibodies directed against Stxs in patients convalescing from hemorrhagic colitis-associated HUS.
E. coli O157:H7 has been associated with typical HUS for some time and is the agent recovered from most of the cases in the U.S. In a retrospective study of HUS cases spanning a 10-year period in Minnesota, E. coli O157:H7 was recovered from 13 (50%) of 26 patients (13). This same study also documented an increasing incidence of HUS in Minnesota, which was due to increases in the number of cases occurring in younger children. A 1987 prospective study performed in the Pacific Northwest found that 58% of the patients with HUS had prior E. coli O157:H7 infections, and another study conducted in British Columbia, Canada found E. coli O157:H7 infection in 9 (64%) of 14 children with HUS (14, 15). While most outbreaks of hemorrhagic colitis and HUS are caused by STEC strains belonging to the O157:H7 serogroup, about 50% of sporadic cases are caused by strains that belong to non-O157:H7 serotypes. These other serotypes include groups O26, O39, O103:H2, O111:H8, O113, O118, O121, O128, and O157.
Treatment modalities for hemorrhagic colitis and HUS associated with E. coli O157:H7 are controversial (16). While some studies have suggested that treatment of hemorrhagic colitis with antimicrobial agents may substantially increase the duration of hemorrhagic colitis and the risk for progression to HUS, other investigations have not demonstrated this (17, 18). In vitro and in vivo studies have shown that certain antimicrobial agents that act by interfering with microbial DNA replication (e.g., fluoroquinolones, trimethoprim) may
increase the amounts of Stxs produced by some E. coli strains, thereby increasing the toxin concentrations in the gastrointestinal tract and augmenting the associated morbidity and mortality. In the absence of additional clinical data supporting the administration of antimicrobial therapy, the Infectious Diseases Society of America currently recommends that patients with E. coli O157:H7 disease should not be treated with anti-infective agents and should be closely followed for the development of renal dysfunction indicative of HUS (17).
Shiga-like Toxins and Other Virulence Factors
Strains of E. coli O157:H7 and several other E. coli serotypes are able to produce Stxs that are related to the Shiga toxin produced by Shigella dysenteriae. Two such toxins, known at Stx-1 and Stx-2, have modes of action simlar to the Shiga toxin of Shigellae. Stx-1 is identical to the shigella toxin, and Stx-2 shows 56% homology with Stx-1. In addition, novel Stxs from E. coli strains recovered from animals have been described: Stx-2d, Stx-2e (associated with disease in piglets), and Stx-2f (associated with avian infections). Toxigenic isolates from human infections usually produce Stx-2 and an Stx-2 variant called Stx-2c. Recently, researchers have described a Stx variant, Stx-2dactivatable, that is activated by enzymes present in intestinal mucus and is associated with HUS cases.
The structural genes for Stxs reside on a bacteriophage, and E. coli strains that are lysogenized by this bacteriophage are able to produce the toxins. Stxs are composed of five binding (B) subunits and one active (A) subunit. Subunit B of the toxins bind to globotriaosyl-ceramide (Gb-3), and the active subunit (A) enzymatically cleaves a specific N-glycoside bond in the 28S RNA of the 60S ribosomal subunit, preventing elongation factor 1-dependent aminoacyl-tRNA binding to the ribsome (19). As a result, protein synthesis in the affected cells is inhibited.
The Stx proteins bind with these globotriaosyl-ceramide receptors present in the gastrointestinal tract, the central nervous system, and the cortical region of the kidneys, resulting in damage to and swelling of the endothelial cells of the glomeruli. This results in detachment of the basement membranes, decreased capillary perfusion to the kidney, and the deposition of platelet/fibrin thrombi, leading to the development of renal insufficiency. The Stx proteins also stimulate production and release of inflammatory cytokines by many cell types, including endothelial cells. These cytokines may function to up-regulate the production of additional Stx receptors, thereby promoting additional attachment and systemic absorption of toxin molecules.
Stx-2 is the virulence determinant that is most important in human disease and is associated with an increased risk of HUS following infection. While three genes encoding Stx-1 have been identified, more than 12 genes that code for Stx-2 have been described. Some of these Stx-2 genes are found in E. coli strains isolated from animals, while others are associated with clinical disease in humans.
Other virulence factors found in strains of E. coli O157:H7 include pili, termed hemorrhagic coli pili (HCP), which also mediates attachment of organisms to the colonic mucosa (20). Serum specimens from patients with HUS but not from control patients react with HCP antigens, suggesting that these pili are actively expressed in infected patients and play a pivotal role in disease pathogenesis.
Lab Diagnosis of E. coli O157:H7 Infection
Many approaches can be used to recover E. coli O157:H7 from stool specimens submitted to the clinical microbiology lab. Most isolates of E. coli O157:H7 are unable to ferment sorbitol. Consequently, labs use MacConkey agar containing 1% sorbitol (Sorbital MacConkey agar, SMAC) instead of lactose to screen for E. coli O157:H7 in bloody diarrheal stool specimens. Early after infection, culture on SMAC usually yields isolates of sorbitol-negative E. coli O157:H7; however, as the disease progresses, the isolation rate rapidly falls.
Some studies have demonstrated that the direct culture technique is less than ideal, with sensitivities for detection of E. coli O157:H7 in the 50% to 60% range. Some commercial vendors have made modifications of SMAC formulations to enhance the selective recovery of E. coli O157:H7. For example, one formulation (Hardy Diagnostics, Santa Martia, Calif.) supplements SMAC agar with potassium tellurite and cefixime (0.5 µg/mL media). This medium inhibits the growth of non-STEC strains and most of the other non-sorbitol-fermenting species found in stool specimens. SMAC will not detect non-O157:H7 E. coli or strains of E. coli O157:H7 that may be able to metabolize sorbitol.
Rainbow agar O157 (Biolog Inc., Hayward, Calif.) is another selective medium that contains a chromogenic substrate for the enzyme β-glucuronidase. E. coli O157:H7 usually does not produce this enzyme, while most non-O157:H7 E. coli strains do. On Rainbow agar, most E. coli strains grow as purple, blue, or magenta-pink colonies, but O157:H7 strains grow as black, gray, or purple-blue colonies.
Once isolated, the E. coli strains can be serotyped for somatic (O) and flagellar (H) antigens and tested for Stx production. The drawback with this approach is that several sorbitol-negative colonies recovered from SMAC or β-glucuronidase-negative colonies from Rainbow agar O157 need to be subcultured and serotyped to determine if they are the O157:H7. Suspicious isolates from these selective media also must be tested for ability to produce Stxs. Not all sorbitol-negative E. coli isolates are E. coli O157:H7, and several sorbitol-positive serotypes of E. coli are capable of producing Stxs, hemorrhagic colitis, and HUS.
Labs can test for Stx production using various methods. The reference procedure is a Vero cell cytotoxicity assay (21). In this assay, stool specimens or E. coli isolates to be tested are grown in MacConkey broth and, following centrifugation and treatment with antibiotics, aliquots are incubated with Vero cell monolayers. After 2 to 3 days, the monolayers are inspected for cytotoxicity. If cytotoxicity is observed, a portion of the original stool or organism suspension is mixed with Stx-specific polyclonal anti-toxin and incubated. Aliquots of the original suspension and the neutralized specimen are then placed on Vero cell monolayers and incubated. The presence of cytotoxicity in the toxin-treated monolayers and absence of cytotoxicity in monolayers exposed to the neutralized aliquot constitute a positive cytotoxicity assay.
Labs may also choose to use an EIA to detect Stx directly in stool specimens or in overnight broth cultures. The availability of two FDA-cleared assays, ProspecT Shiga Toxin E. coli (STEC) assay (Remel Laboratories, Lenexa. Kan.) and the Premier EHEC EIA (Meridian Bioscience, Inc., Cincinnati, Ohio), has made it possible for labs to easily screen stool specimens for Stx-producing organisms independent of the serotype of the producing E. coli strain. Meridian Bioscience, Inc. also markets the ImmunoCardSTAT! EHEC, a rapid immunochromatographic assay that detects Stx-1 and Stx-2 directly in stool specimens or specimens placed into Cary-Blair transport media.
In the ProspecT assay, polyclonal antibodies directed against Stx-1 and Stx-2 are fixed to microtiter wells and capture toxins in the specimen, and mouse monoclonal antibodies conjugated to horseradish peroxidase detect bound Stx protien. In the Premier assay, monoclonal antibodies against Stx-1 and Stx-2 are present in the microtiter wells for antigen capture, and polyclonal anti-Stx-1 and anti-Stx-2 antibodies tagged with horseradish peroxidase detect the bound toxins.
With these EIAs, labs can test stool specimens or stool from Cary-Blair stool preservative/transport or broth media. With the latter specimen type, stool is inoculated into 5 mL of MacConkey or GN broth and incubated for 16–24 hours at 35–37°C. Aliquots from these enrichment broths are then used for toxin detection by EIA. Inocula for the Premier EIA can also be prepared from an individual colony or several colonies directly from agar media. The sensitivity of the Premier EIA for the detection of Stx-1 and Stx-2 exceeds 90% (22, 23). In general, sensitivities of Stx EIAs are better when testing is performed on growth-enhanced broth specimens.
Stool specimens that yield positive Stx EIA results need to be inoculated onto growth media for recovery of isolated colonies for serotyping. Plating media for Stx EIA-positive specimens should include a SMAC agar or a Rainbow O157 agar plate to detect the sorbitol-negative colonies of E. coli O157:H7. Several latex agglutination tests are available commercially for serotyping of presumptive E. coli O157:H7 isolates and include Dryspot E. coli O157 (Oxoid, Cambridge, U.K. ), the RIM E. coli O157:H7 latex test (Remel, Lenexa, Kan.), the Wellcolex E. coli O157 latex agglutination kit (Remel), and the Wellcolex E. coli O157:H7 kit (Remel).
Outbreaks of E. coli O157:H7 associated HUS are rare but they constitute a deadly threat to children and the elderly. HUS occurs most frequently in the context of a severe gastroenteritis characterized by copious, bloody diarrhea. Labs can contribute to the successful diagnosis and treatment of these illnesses by offering testing protocols that pinpoint the infectious agent in a timely manner.
- Dlott JS, Danielson CF, Blue-Hnidy DE, et al. Drug-induced thrombotic thrombocytopenic purpura/hemolytic uremic syndrome: a concise review. Ther Apher Dial 2004;8:102–111.
- Pratap B, Abraham G, Srinivas CN, et al. Post-renal transplant hemolytic uremic syndrome following combination therapy with tacrolimus and everolimus. Saudi J Kidney Dis Transplant 2007;18:609–612.
- Waters AM, Kerecuk L, Lik D, et al. Hemolytic uremic syndrome associated with invasive pneumococcal disease: the United Kingdom experience. J Pediatr 2007;151:140–144.
- George JN. The association of pregnancy with thrombotic thrombocytopenic purpura-hemolytic uremic syndrome. Curr Opin Hematol 2003;10:229–244.
- Kavanagh D, Richards A, Atkinson J. Complement regulatory genes and hemolytic uremic syndrome. Annu Rev Med 2008;59:293–309.
- Besbas N, Karpman D, Landau D, et al. A classification of hemolytic uremic syndrome and thrombotic thrombocytopenic purpura and related disorders. Kidney Int 2006;70:423–431.
- Riley LW, Remis RS, Helgerson SD, et al. Hemorrhagic colitis associated with a rare Escherichia coli serotype. N Engl J Med 1983;308:681–685.
- Davies M, Engel J, Griffin D, et al. Outbreaks of Escherichia coli O157:H7 associated with petting zoos. Morbid Mortal Weekly Rep 2005;54:1277–1280.
- Laboratory Center of Disease Control. Outbreak of gastrointestinal disease —Ontario. Canada Diseases Weekly Rep 1987;13:5–8.
- Kelly J, Oryshak A, Wenetsek M, et al. The colonic pathology of Escherichia coli O157:H7 infection. Am J Surg Pathol 1990;14:87–92.
- Cimolai N, Morrison BJ, Carter JE. Risk factors for the central nervous system manifestations of gastroenteritis-associated hemolytic uremic syndrome. Pediatrics 1992;90:616–621.
- Ochoa TJ, Cleary TG. Epidemiology and spectrum of disease of Escherichia coli O157. Curr Opin Infect Dis 2003;16:259–263.
- Martin DL, MacDonald KL, White KE, et al. The epidemiology and clinical aspects of the hemolytic-uremic syndrome in Minnesota. N Engl J Med 1990;323:1161–1167.
- Gransden WR, Damm MA, Anderson JD, et al. Further evidence associating hemolytic-uremic syndrome with infection by Verotoxin-producing Escherichia coli O157:H7. J Infect Dis 1986;154:522–524.
- Niell MA, Tarr PI, Clausen CR, et al. Escherichia coli O157:H7 as the predominant pathogen associated with hemolytic-uremic syndrome: a prospective study in the Pacific Northwest. Pediatrics 1987;80:37–40.
- Orth D, Grif K, Zimmerhackl LB, et al. Prevention and treatment of enterohemorrhagic Escherichia coli infections in humans. Expert Rev Anti-Infect Ther 2008;6:101–108.
- Guerrant RL, Van Gilder T, Steiner TS, et al. Practice guidelines for the management of infectious diarrhea. Clin Infect Dis 2001;32:331–351.
- Safdar N, Said A, Gangnon RE, et al. Risk of hemolytic uremic syndrome after antibiotic treatment of Escherichia coli O157:H7 enteritis: a meta-analysis. JAMA 2002;288:996–1001.
- Nakajima H, Kiyokawa N, Katagiri YU, et al. Kinetic analysis of binding between Shiga toxin and receptor glycolipid Gb3Cer by surface plasmon resonance. J Biol Chem 2001;276:42915–42922.
- Xicohtencatl-Cortes J, Monteiro-Neto V, Ledesma MA, et al. Intestinal adherence associated with type IV pili of enterohemorrhagic Escherichia coli O157:H7. J Clin Invest 2007;117:3519–3529.
- Novicki TJ, Daly JA, Mottice SL, et al. Comparison of sorbitol MacConkey agar and a two-step method which utilizes enzyme-linked immunosorbent assay toxin testing and a chromogenic agar to detect and isolate enterohemorrhagic Escherichia coli. J Clin Microbiol 2000;38:547–551.
- Kehl SC. Role of the laboratory in the diagnosis of enterohemorrhagic Escherichia coli infections. J Clin Microbiol 2002;40:2711–2715.
- Mackenzie AMR, Lebel P, Orrbine E, Rowe PC, et al. Sensitivities and specificities of Premier E. coli 0157 and Premier EHEC enzyme immunoassays for diagnosis of infection with verotoxin (shiga-like toxin)-producing Escherichia coli. J Clin Microbiol 1998;36:1608–1611.
William M. Janda, PhD, is professor of Pathology, Microbiology, and Immunology and the director of the Clinical Microbiology Laboratory at the University of Illinois Medical Center in Chicago. His research interests include the pathogenic Neisseria, fastidious Gram-negative bacilli, and the aerobic and facultative Gram-positive bacilli. Dr. Janda is a co-author of several clinical microbiology textbooks and manual, as well as several self-study courses.