Escherichia coli 'O' group serology of a haemolytic uraemic syndrome (HUS) epidemic.

This is the first comprehensive serological analysis of a haemolytic uraemic syndrome (HUS) outbreak. A wide range of 'O' group Escherichia coli antibody responses in patients and controls was examined. The study provides a unique insight into the epidemiology of such epidemics, points a way to the most appropriate investigation of these and indicates possible answers to a number of issues related to severity of disease. In order to be able to test for a wide variety of E. coli 'O' antigens, a microagglutination assay was used to examine E. coli 'O' group serological responses of 22 children admitted to hospital with HUS and 14 contemporaneous age-matched controls. A total of 51 'O' serogroup strains were used. These included 'O' groups reported to be associated with cases of HUS, with 6 isolates from patients associated with the Adelaide outbreak (O26, O111, O123 and O157), environmental Verocytotoxigenic/Shiga-toxin producing Escherichia coli (VTEC/STEC) strains and common human commensal strains. Sixteen clinically confirmed HUS cases (72.7%) of 22 seroconverted to 1 or more serogroups of which 11 (50%) seroconverted to O111 (the serogroup isolated from 16 patients). In addition, 11 (50%) and 10 (45.5%) developed antibody to O137 and O145, respectively, although no stool isolates of these serogroups were made. Seventeen (77.3%) of 22 HUS patients had antibody to serogroup O157, with 11 (50%) seroconversions, however, O157:H- was isolated from only 2 of these. Overall, titres ranged from 100 to 6400, some of the highest in 3 patients were against O157, whose faeces yielded only Enterohaemorrhagic E. coli (EHEC) O111, and only 1 developed O111 antibody. Mixed infection was demonstrated serologically by microagglutination (confirmed by Western blot) and was consistent with the findings of multiple serogroups of VTEC found in the mettwurst incriminated as the source, and suggests further strains (not found in the source or in patients' faeces) were probably also involved. In HUS associated with EHEC infection, multiple strain infection may be the rule rather than the exception. A relationship with clinical severity deserves further investigation. Non-O157 EHEC (in addition to O157) should be sought in all future outbreaks of EHEC disease.