The patient was in his otherwise healthy state of being until 3 days prior when he developed non-bloody diarrhea. On the morning of presentation the stool had become bloody. The patient was afebrile, had some reduced intake of food, but drinking fine. Most notably, he periodically stops walking and bends over as if he is in pain. This happened 3-4 times the previous day and these episodes tended to last for about one minute, after which the boy would continue to play. He does not attend daycare and his immunizations are up to date.
Stool culture was sent and a predominant organism was an oxidase negative, lactose-fermenting, Gram-negative rod (Figure 1). The organism was non-sorbitol fermenting based on growth on Sorbitol-MacConkey agar, and grew as mauve colonies on E. coli 0157 screening agar. MUG testing was negative.
The isolate was E. coli O157. Isolates of E. coli O157 commonly produce shiga toxins (sxt1 and sxt2) which are responsible for diarrhea, hemorrhagic colitis, and most famously hemolytic-uremic syndrome (HUS). Typical illness starts with non-bloody diarrhea which becomes bloody after 2-3 days due to onset of hemorrhagic colitis. Often severe abdominal pain and low grade fever are present as well. HUS is a serious complication of E. coli O157 infection which results in acute renal dysfunction. HUS most often occurs in children < 5 years of age, of which 15% of those with laboratory confirmation of E. coli O157 developing this complication, compared to 6% in the general population. It is possible for other E. coli to produce shiga toxins, with 1% of HUS is caused by non-E. coli O157 infection.
It is recommended that all patients with suspected HUS should have stool cultured on selective and differential media for detection of E. coli O157 and direct shiga toxin detection should be performed to identify non-E. coli O157 isolates that are producing toxin. E. coli 0157 isolates look exactly the same as non-E. coli O157 normal fecal flora on 5% sheep blood, chocolate, and MacConkey agars. All E. coli ferment lactose on MacConkey agar (Figure 1A). E. coli O157 can be differentiated from other E. coli strains by growth on Sorbitol-MacConkey (SMAC) agar; E. coli O157 is a non-sorbitol fermenter while most other E. coli will ferment sorbitol (Figure 1B). Chromagenic agar for E. coli O157 is another option to screen stool specimens for E. coli O157. E. coli O157 grow mauve colored colonies on this particular agar (BBL CHROMagar O157 , Becton Dickinson) (Figure 1C). A summary of this data can be found in Table 1.
Growth of organisms suspicious for E. coli O157 on any media requires confirmation prior to reporting. Biochemical confirmation tests include E. coli O157 antiserum or latex agglutination and 4-methylumbelliferyl-beta-D-glucuronide (MUG) testing. For latex agglutination or antisera testing, it is essential to test the isolate of interest with the E. coli O157-specific reagent as well as a non-specific control to exclude non-specific binding. Unlike most E. coli strains, E. coli-O157 does not express beta-glucuronidase and is therefore MUG test negative (Table 1).
Table 1. Characteristics of E. coli O157 in comparison to other E. coli strains
|Test||Non-E. coli O157||E. coli O157|
|Appearance on MacConkey agar||Lactose fermenter||Lactose fermenter|
|Appearance on Sorbitol-MacConky agar||Sorbitol fermenter||Non-sorbitol fermenter|
For direct detection of shiga toxin, there are several commercially available immunoassays available for detection of shiga toxin protein. New on the market are multiplex gastrointestinal panels that can be used for molecular based detection of shiga toxin genes sxt-1 and sxt-2 among a host of other agents of gastrointestional disease.
E. coli O157 is spread via fecal oral route. It can be acquired directly from person to person or indirectly through food and water sources contaminated with fecal matter from infected humans and animals. Classic scenarios are undercooked ground beef, leafy greens, unpasteurized milk and juice, petting zoos, and contaminated drinking water. The incubation period prior to symptoms is 3-4 days (range 1-8 days).
Treatment for E. coli O157 is largely supportive consisting of fluids to prevent dehydration. The role of antibiotics is controversial with some studies suggesting antibiotics increase the risk of developing HUS while others found no association between the their use and increased HUS.
Following our patient’s stool culture result for E. coli O157, he was recalled to the Emergency Department for evaluation. He was still having diarrhea and vomiting, but it was reduced compared to the previous day. The patient was given fluids and sent home without antibiotic treatment and via phone conversation with his mother, his symptoms resolved a few days later.
-Erin McElvania TeKippe, PhD, D(ABMM), is the Director of Clinical Microbiology at Children’s Medical Center in Dallas Texas and an Assistant Professor of Pathology and Pediatrics at University of Texas Southwestern Medical Center.