December 1996
HAIFA, Israel — A strain of Escherichia coli has been associated with encephalopathy in two children in Israel.
This is the first time that the bacterium has been linked to encephalopathy, although central nervous system (CSN) involvement has been associated with Salmonella, Campylobacter, and Shigella. The mechanisms by which these pathogens cause neurological symptoms, however, are not known.
Shigella and E. coli are similar genetically and clinically, but enteroinvasive E. coli infection had not previously been associated with neurological symptoms.
The serogroups of E. coli related to enteroinvasive infection include O28, O29, O112, O124, O136, O143, O144, O152, O164, O167 and O173. (Strain O147 has been linked to foodborne illnesses.) Because stool cultures do not routinely screen for these serogroups, identifying enteroinvasive E. coli is difficult and infections may go undiagnosed.
![[bar]](/art/gradred.gif){kind=small}
The case report of the two children was described by Moshe Ephros, MD, and his colleagues from the Carmel Medical Center, Haifa, and the Children's Medical Center, Petach, Israel. Both cases involved enteroinvasive E. coli O144:NM.
The first case involved a 6-year-old boy who had previously been healthy. He had a one-day history of fever, watery diarrhea and vomiting, and he was confused and disoriented. He was given parenteral fluids, and 24 hours after hospital admission he felt better and was oriented. Enteroinvasive E. coli O144:NM was isolated from his stool, but no other bacterial, viral or parasitic pathogen was identified. Blood and cerebrospinal fluid cultures were negative.
The second case was in an 8-year-old girl. She also had been healthy and presented with a one-day history of fever and dysenteric diarrhea. She had become increasingly apathetic and was stuporous at the time of admission. She did not respond to IV glucose.
Stool cultures revealed polymorphonuclear leukocytes and erythrocytes, leading to a presumptive diagnosis of encephalitic shigellosis. IV fluids and ceftriaxone (50 mg/kg/day) were administered. The girl's fever and diarrhea improved within 48 hours, and her mental state returned to normal. Blood cultures were negative. E. coli O144:NM was the only pathogen isolated in her stool. Two family members also had diarrhea with encephalopathy, but stool samples were not obtained.
One year later, the girl returned to the hospital with the same symptoms. S. sonnei was isolated and, in the absence of other findings, was the only explanation for her encephalopathy. Again, her family members had concurrent diarrhea, but none had neurological symptoms.
"This strongly suggests that host factors probably play an important role in the pathogenesis of the encephalopathy associated with invasive diarrheal disease and probably act in concert with certain bacterial traits," the researchers wrote.
In addition to these two cases, 24 enteroinvasive E. coli O144 isolates were identified in stool samples taken between May and August 1993. None were associated with neurological symptoms.
Eight of the 24 cases of nonencephalopathic enteroinvasive E. coli O144 were examined and compared with the two cases of enteroinvasive E. coli encephalopathy. Of the eight patients, of which two were children, three had dysenteric diarrhea and five had watery stools; two had a fever and one adult required parenteral fluids.
All of the enteroinvasive E. coli O144 isolates studied were susceptible to chloramphenicol (Chloromycetin, Parke-Davis), tetracycline, ampicillin, amoxicillin-clavulanate (Augmentin, SmithKline Beecham), trimethoprim-sulfamethoxazole, ciprofloxacin (Cipro, Bayer Pharmaceutical), nalidixic acid (NegGram, Sanofi Winthrop) and ceftriaxone (Rocephin, Roche Laboratories).
E. coli O144 is apparently becoming more frequent in Israel. An average of 36 E. coli isolates were identified each year between 1989 and 1992; of these, three were identified as O144:NM. In contrast, 31 E. coli strains were isolated in 1993, of which 26 were O144. In June 1993 alone, 17 of 26 isolated strains were identified as enteroinvasive E. coli O144.
These cases were all sporadic and none were associated geographically or epidemiologically.
"The cases of infection described here indicate that enteroinvasive E. coli can be associated with CNS complications," the researchers concluded. "The pathogenesis of encephalopathy caused by enteroinvasive E. coli specifically and other invasive enteric pathogens in general is not well understood."
Although enteroinvasive E. coli had not previously been associated with neurological symptoms, the researchers noted that such an association is not surprising given the clinical and genetic similarities with Shigella. Both pathogens contain a 140-MDa plasmid that encodes certain bacterial proteins that are associated with invasiveness.
Studies of people with shigellosis have indicated that shiga toxin is neurotoxic. Although shiga toxin is produced by Shigella dysenteriae, it not produced by Shigella flexneri and Shigella sonnei, which have also been associated with neurologic symptoms.
"The closely related shiga-like toxins produced by E. coli and other cytotoxins may play a role," Ephros and his colleagues stated, "mainly by causing endothelial cell damage in the brain. ... Severe brain edema has been documented in a patient with Shigella associated encephalopathy. Those investigators suggested that brain edema during shigellosis is often undiagnosed and may be related to hyponatremia. Although direct CNS involvement with Shigella species is possible, it is rare and cannot explain the majority of cases of Shigella associated encephalopathy. The pathogenesis of Salmonella and Campylobacter associated encephalopathy has not been elucidated."
You can express your
views on this article, or other relevant themes, in the
Infectious Diseases in Children
Specialty Forums.