The recent outbreaks of Escherichia coli O157:H7 linked to spinach and lettuce remind us yet again how limited our tools are when it comes to treating this infection and its sequelae. Focusing our efforts on prevention is by far the best medicine.
As of Oct. 6, a total of 199 people infected with the outbreak strain of E. coli serotype 0157:H7 had been reported to the Centers for Disease Control and Prevention from 26 states, including 22 cases in children younger than 5 years of age. Of the total group, 51% were hospitalized and 16% developed hemolytic uremic syndrome (HUS). Twenty-nine percent of children younger than 18 years developed HUS, compared with 8% of adults aged 18–59 years and 14% of those aged 60 years and older, confirming the increased risk for HUS in children and the elderly.
There were three deaths, including a 2-year-old child with HUS whose stool sample contained evidence of the outbreak strain confirmed by “DNA fingerprinting.”
About 73,000 infections with E. coli 0157:H7 occur annually in the United States. Such infections are reportable nationally as well as in most states. In most states, HUS is reportable to departments of public health as well. The CDC investigates all reported cases to ascertain whether they are outbreak-associated or isolated. Most are the latter. Half of all cases occur between June and September.
Produce was the source in the recent outbreak, but in the past we've seen disease in children associated with undercooked meat, nonpasteurized milk products, and even water. Petting zoos are a major hazard.
During 2004–2005, a total of 173 cases of E. coli 0157:H7 were reported from outbreaks in Arizona, Florida, and North Carolina. Illnesses primarily affected children who had visited petting zoos at agricultural fairs or festivals. There were 22 cases of HUS, but fortunately no fatalities (MMWR 2005;54:1277–80).
In a study the CDC conducted at a petting zoo, illness was associated with touching or stepping in manure, falling or sitting on the ground, using a pacifier or “sippy” cup, and thumb-sucking. Use of alcohol-based sanitizer was not protective, but reported awareness by the accompanying adults of the risk for disease from contact with livestock was. We should counsel parents about the potential risk and the preventive strategies such as avoidance of manure and of the use of a pacifier or eating while at the petting zoo.
Direct human-to-human contact is a rarer source of E. coli infection, but it's important to keep in mind when we see a child with bloody diarrhea, particularly if that child is in day care.
Unfortunately, we don't have a way to interrupt the progression from colitis to HUS. The role of antibiotics in children with E. coli gastrointestinal infection remains controversial. Epidemiologic data have suggested that antibiotics may increase the risk for HUS, perhaps by increased toxin exposure to the kidneys following bacteriolysis in the gut. A meta-analysis of 26 studies conducted between January 1983 and February 2001 did not show a higher risk of HUS due to antibiotic use. However, the authors concluded that a randomized trial of adequate power is needed to conclusively answer the question (JAMA 2002;288:996–1001).
Until then, the potential benefit of antimicrobial therapy in a specific patient presenting with gastroenteritis must be weighed against the potential risk. Stool cultures should be obtained from any child who presents with bloody diarrhea and abdominal pain. If the child is afebrile and otherwise does not appear ill, supportive care is advised. But of course, a child with gastroenteritis who is hypotensive and appears septic requires urgent intervention that may include antimicrobial therapy.
Although we don't know which children with E. coli-associated diarrhea will progress to HUS, we do know that certain risk factors, such as young age, long duration of diarrhea, elevated leukocyte count, and proteinuria, are predictive (Emerg. Infect. Dis. 2005;11:1955–7). Fortunately, there is usually a lag time of several days to a week between the onset of bloody diarrhea and renal failure. If we see the child soon enough, we can intervene with fluid replacement and close monitoring.
At the time of progression to HUS, stool cultures often are negative. The diagnosis is made clinically, on the basis of renal failure and hemolytic anemia, with or without thrombocytopenia. Treatment is supportive: Dialysis has dramatically reduced HUS mortality from about 21% before 1974 to about 4% today.
Intriguing early work is now being done looking at treating HUS with infusion of the human plasma protein serum amyloid P component (J. Infect. Dis. 2006;193:1120–4) and use of specific neutralizing antibodies directed against the A subunit of the toxin (Clin. Microbiol. Rev. 2004;17:926–41). Clinical use is probably years away, however.