High prevalence of toxin A-negative toxin B-positive Clostridium difficile in hospitalized patients with gastrointestinal disease.

The incidence of Clostridium difficile toxin A-negative toxin B-positive in hospitalized patients with severe gastrointestinal disease was evaluated. Of 530 stool specimens tested in parallel by two immunoassay tests, Tox A and Tox A/B (TechLab, Inc., Blacksburg, VA), 422 produced negative results on both tests. One hundred eight specimens (20.4%) tested positive by Tox A/B assay, and only 47 of them were also positive by Tox A. The 61 specimens with discrepant results were confirmed to be positive for toxin B by tissue culture cytotoxicity assay. Furthermore, 3 of the 422 specimens that were negative by enzyme immunoassay tested positive by cytotoxicity assay. The sensitivity and specificity of the Tox A/B test were 95.3% and 100%, respectively, with negative and positive predictive values of 99.3% and 100%, respectively, and a correlation rate of 99.4%. The high prevalence (56.5%) of specimens from symptomatic patients having detectable toxin B, but undetectable toxin A emphasizes the importance of using diagnostic tests that include toxin B. Furthermore, our data support the potential role of toxin B in the pathogenesis of toxigenic Clostridium difficile.