Abstract

To evaluate the ProSpecTrM Shiga-Like Toxins I and II Microplate Assay (Alexon-Trend, Ramsey MN, USA) for detecting enterohemorrhagic Escherichia coli (EHEC) shigatoxins, 3 sets of samples were tested with this ELISA and the standard verocell cytotoxicity assay in comparison. First, 47 fecal samples from 25 patients with positive isolation of verotoxinogenic Escherichia coli were cultured overnight in trypticase soy broth to improve shigatoxin detection. All (100%) were subsequently cytotoxicity positive. The ELISA detected 38/47 (80.9%), including each patient's fïrst sample (in case of follow-up). When stool samples were tested directly, only 16/45 (35.6%) were cytotoxicity positive and 13/45 (28.9%) were positive by ELISA. Second, of 60 clinical strains of verotoxinogenic Escherichia coli maintained in pure culture, 59 (98.3%) were both verocytotoxic and ELISA positive, 1 remained equivocal in both tests. Finally, of 422 stools from enteritic children, both the cytotoxicity assay and the ELISA detected the 2 EHEC isolation-positive samples. While true positives were always reproducible in this stady, 12/14 results initially ELISA positive in the prospective study proved to be false positives, resulting in a positive predictive value of only 14.3% in this low incidence population. Overall, ELISA sensitivity, specifÏcity, positive and negative predictive values in stool cultures were calculated as 81.6%, 97.1%, 76.9%, and 97.8%, respectively. In conclusion, the ProSpecTrM shigatoxin ELISA proved to be a sensitive screening test for both stools and EHEC isolates. Due to the low positive predictive value in a low incidence population, however, repeat analysis of reactive results is strongly recommended. Prior cultural enrichment of fecal samples is crucial for both cytotoxi-city and ELISA.

DOI: Clin. Lab. 1999;45:13-21