Analysis of Urinary Escherichia coli Isolates for Ability To Produce Shiga Toxin

Abstract

The frequency of Shiga toxin (Stx)-producing Escherichia coli (STEC) in the urinary tract, which can precipitate the hemolytic-uremic syndrome, is unknown. We tested 597 urinary E. coli isolates by Stx immunoassay and found no STEC. The routine screening of urinary E. coli for the ability to produce Stx is not warranted.

Shiga toxin (Stx)-associated hemolytic-uremic syndrome (HUS) most often follows gastrointestinal infection caused by Escherichia coli O157:H7 (25) or other Stx-producing E. coli (STEC) organisms. However, HUS also can follow urinary tract infection (UTI) caused by non-O157:H7 STEC (3, 15, 18, 21, 24). The possibility has been raised that antibiotic therapy for such infections can precipitate microangiopathic sequelae (15), as occurs when antibiotics are administered to children suffering from diarrhea caused by E. coli O157:H7 (28).

The frequency with which urinary isolates produce Stx is unknown. Beutin et al. were the first to describe a non-O157:H7 STEC urinary isolate, and the patient from whom this organism was recovered did not develop HUS (1). Wilson et al. found no STEC among 81 sorbitol-nonfermenting E. coli urinary tract isolates and concluded that the routine screening of urinary E. coli for the ability to produce Stx was not warranted (27). Thorpe and Acheson questioned this conclusion because the selection of sorbitol-nonfermenting E. coli isolates for study excluded from analysis most non-O157:H7 STEC isolates, which ferment sorbitol when plated on sorbitol MacConkey agar plates (26).

In this study, we examined a diverse collection of E. coli isolates from patients with UTIs to determine what proportion of the bacteria produce Stx, as detected by a commercial immunoassay, regardless of the isolates' abilities to ferment sorbitol.

Bacteria studied.

Table 1 provides the origins of the bacteria studied. To our knowledge, none of the patients from whom these organisms were isolated developed HUS. E. coli O103:H2 (24) and E. coli O157:H7 strain 86-24 (23) were used as positive controls for Stx1 and -2 production, respectively. Strain CFTO73, an E. coli O6:K2:H1 isolate from a patient with pyelonephritis which does not produce Stx, was used as a negative control (16).

TABLE 1.

Sources and descriptions of E. coli clinical isolates tested for Stx production

Clinical syndrome(s)	Population	Geographic locale	Isolate contributor	Isolate no.	Reference(s) or source
Urosepsis	Adults	Boston, Mass.; Long Beach, Calif.; Nairobi, Kenya	J. N. Maslow	74	5, 8, 13, 14
Urosepsis	Adults	Seattle, Wash.	J. R. Johnson	56	4, 6, 9, 11, 12
Cystitis, pyelonephritis	Children	Denver, Colo.	C. E. Johnson	60	Unpublished data
Uncomplicated pyelonephritis	Women	Multicenter	Bayer, Inc.	145	22
Uncomplicated cystitis	Women	Minneapolis, Minn.	J. R. Johnson	78	Unpublished data
Urosepsis	Veterans	Minneapolis, Minn.	J. R. Johnson	27	M. A. Khan et al.a
Uncomplicated cystitis	Women	Seattle, Wash.	A. E. Stapleton	61	7, 10, 20
Cystitis, pyelonephritis	Miscellaneous	Sweden	ECORb	10	17
Cystitis in patients of sexually transmitted disease clinics	Men and     women	Seattle, Wash.	A. E. Stapleton	86	Unpublished data; 19

^aM. A. Khan, T. O'Bryan, N. Kaster, F. Lebahn, and J. R. Johnson, Abstr. 37th Annu. Meet. Infect. Dis. Soc. Am. abstr. 176, 1999.

^bECOR, E. coli reference collection.

Stx EIA.

Bacteria were stored in Luria-Bertani broth with 20% glycerol at −80°C following collection. They were then subcultured onto semisolid agar, inoculated into MacConkey broth (Northeast Laboratory, Waterville, Maine), and grown overnight at 35°C. Stx was assayed in the broth cultures with a commercial enzyme immunoassay (EIA) (Premier EHEC [enterohemorrhagic E. coli]; Meridian Biosciences, Inc., Cincinnati, Ohio) in accordance with the manufacturer's instructions. The assay was read visually at its conclusion.

Findings and conclusions.

None (0%) of the 597 E. coli UTI isolates tested positive in the Stx EIA (one-sided 95% CI is 0 to 0.5%). As expected, the positive controls reacted in the EIA.

These data indicate that STEC rarely causes UTIs in North America. Therefore, the routine use of Stx detection assays to identify STEC, in the hopes of averting the precipitation of HUS with antibiotics (15), does not appear to be justified. These data corroborate the experiences of our respective institutions that suggest that HUS following UTIs is rare. Specifically, we are aware of a total of only two such cases at our institutions since 1990. These data also extend the findings of Wilson et al. (27), who studied only sorbitol-nonfermenting E. coli urinary isolates. Certainly, for patients in whom UTI-associated or nondiarrheal HUS occurs, urinary and fecal STEC isolates should be assayed with a commercially available EIA or other appropriate diagnostic algorithms (2). However, on the basis of currently available data, concerns that HUS can, in rare cases, follow UTI should not discourage the administration of appropriate antimicrobial therapy to patients with UTIs, pending the performance of an Stx assay.