OBJECTIVE: To evaluate the isolation rate of O157 and non-O157 verocytotoxin-producing Escherichia coli (VTEC) strains, to study the occurrence of additional virulence factors and to correlate these with clinical symptoms. METHODS: Over more than 5 years, 17 296 unduplicated fecal samples submitted for routine culture were screened for VTEC by a single PCR detecting VT1, VT2 and its variants. Verocytotoxin B subunit genotypes of the isolates obtained by testing individual colonies in positive samples were determined by a polymerase chain reaction---restriction fragment length polymorphism (PCR---RFLP) technique, the eaeA gene and the 60-MDa virulence plasmid by PCR, and the hemolytic phenotype by using CaCl2-washed blood agar. RESULTS: Verocytotoxin genes were found in 1.02% of the samples. Non-O157 VTEC strains were isolated in 0.66% and O157 in 0.17%. Overall, VTEC was less frequently isolated than Campylobacter and Salmonella but more frequently than Yersinia and Shigella. All cases except two siblings were epidemiologically unrelated. Cases of hemolytic uremic syndrome (HUS) were only observed in association with serogroup O157, which seems to be more pathogenic than the non-O157 strains. Among non-O157 VTEC strains, eaeA-positive strains are more frequently associated with clinical symptoms than are eaeA-negative strains. Other virulence factors correlate less closely with the presence of symptoms. CONCLUSIONS: VTEC is the third bacterial intestinal pathogen in our study population. All stool samples from patients with diarrhea should be screened for the most frequent serogroup, O157, or, if this is not possible, at least those from patients with bloody diarrhea. Non-O157 VTEC strains, especially if they are eaeA positive, are also associated with diarrhea, more often non-bloody. PCR or the new commercially available immunoassays could be used in selected cases, e.g. in patients suffering from HUS and in cases of outbreaks.
OBJECTIVE: To evaluate the isolation rate of O157 and non-O157 verocytotoxin-producing Escherichia coli (VTEC) strains, to study the occurrence of additional virulence factors and to correlate these with clinical symptoms. METHODS: Over more than 5 years, 17 296 unduplicated fecal samples submitted for routine culture were screened for VTEC by a single PCR detecting VT1, VT2 and its variants. Verocytotoxin B subunit genotypes of the isolates obtained by testing individual colonies in positive samples were determined by a polymerase chain reaction---restriction fragment length polymorphism (PCR---RFLP) technique, the eaeA gene and the 60-MDa virulence plasmid by PCR, and the hemolytic phenotype by using CaCl2-washed blood agar. RESULTS: Verocytotoxin genes were found in 1.02% of the samples. Non-O157 VTEC strains were isolated in 0.66% and O157 in 0.17%. Overall, VTEC was less frequently isolated than Campylobacter and Salmonella but more frequently than Yersinia and Shigella. All cases except two siblings were epidemiologically unrelated. Cases of hemolytic uremic syndrome (HUS) were only observed in association with serogroup O157, which seems to be more pathogenic than the non-O157 strains. Among non-O157 VTEC strains, eaeA-positive strains are more frequently associated with clinical symptoms than are eaeA-negative strains. Other virulence factors correlate less closely with the presence of symptoms. CONCLUSIONS: VTEC is the third bacterial intestinal pathogen in our study population. All stool samples from patients with diarrhea should be screened for the most frequent serogroup, O157, or, if this is not possible, at least those from patients with bloody diarrhea. Non-O157 VTEC strains, especially if they are eaeA positive, are also associated with diarrhea, more often non-bloody. PCR or the new commercially available immunoassays could be used in selected cases, e.g. in patients suffering from HUS and in cases of outbreaks.