Shiga Toxin Antigen Detection Should Not Replace Sorbitol MacConkey Agar Screening of Stool Specimens

The interesting paper by Gavin et al., describing their experience using an enzyme immunoassay (EIA) to identify stools containing Shiga toxin (Stx)-producing Escherichia coli (STEC) (3), adds to the growing literature that non-O157:H7 STEC are overlooked pathogens. However, as physicians, clinical and public health microbiologists, and disease control epidemiologists, we are concerned by the authors' use of an EIA to screen stools and, only if this test is positive, to then seek E. coli O157:H7 with sorbitol MacConkey agar culture (SMAC).

The promoted approach has several pitfalls. First, in a study which several of us coauthored, EIA was less sensitive than SMAC screening in detecting E. coli O157:H7 (5). We suspect that in vitro Stx production is not consistently above the level of detection by EIAs. Particularly troubling is that E. coli O157:H7, the STEC with the strongest and most enduring association with epidemics and severe illnesses, might be overlooked unless EIAs are complemented by parallel, and not sequential, culture. Second, while there is value in detecting infections caused by non-O157:H7 STEC, these organisms have lesser risks of precipitating the hemolytic-uremic syndrome (HUS) than do E. coli O157:H7, which are optimally identified using SMAC screening in parallel to, and not sequential with, the EIA. Providers need to know, as early in illness as possible, if a patient infected with an STEC has a high (i.e., those infected with E. coli O157:H7) or a considerably lower (i.e., those infected with non-O157:H7 STEC) likelihood of developing HUS (1). Even a 1-day delay in finding the agent producing the signal in the EIA can have considerable implications for clinical care (7). Third, because E. coli O157:H7 has a well-established association with outbreaks from common sources or exposures, it is critical that infecting organisms be forwarded to public health laboratories urgently, so that they can be genotyped (2, 6). Identifying the sources of epidemics of this high-profile pathogen could be slowed, or made impossible, by laboratory algorithms that employ SMAC only after an EIA on a broth culture produces a signal that suggests the presence of an STEC. Such delayed isolation of E. coli O157:H7 is inappropriate, considering its ability to cause outbreaks and life-threatening human illnesses.

We applaud the attempts by laboratories to find unusual and overlooked causes of human enteric infection, such as non-O157:H7 STEC; in fact, we encourage Stx antigen detection to accomplish this goal and note that there are circumstances where this methodology is more sensitive than culture alone for detecting the presence of E. coli O157:H7 (4). Unfortunately, however, a diminishing proportion of stools submitted for culture in the United States undergo SMAC screening (8). While Stx antigen detection by EIA should, intuitively, always identify E. coli O157:H7, there are insufficient data to justify eliminating SMAC screening in favor of EIA detection of Stx, and we are concerned that algorithms as proposed by Gavin et al. might lead to missed or tardy diagnoses, delay disease control interventions, and adversely affect clinical care and public health.