Comparison of ImmunoCard Toxins A&B and the New Semiautomated Vidas Clostridium difficile Toxin A&B Tests for Diagnosis of C. difficile Infection

We recently showed the ImmunoCard Toxins A&B test (ImmunoCard) (Meridian) to be as specific as and more sensitive than two widely used immunochromatographic tests to diagnose Clostridium difficile infection (CDI) (1). The Vidas C. difficile Toxin A&B test (Vidas-AB) (bioMérieux) is a new enzyme-linked fluorescent immunoassay (ELFA) that detects toxins A and B. As it is automated, it reduces the workload compared to traditional enzyme-linked immunosorbent assays. Our objectives were to evaluate the performance of Vidas-AB and to compare it with that of ImmunoCard for the diagnosis of CDI in stool specimens.

(This study was presented at the 48th Interscience Conference on Antimicrobial Agents and Chemotherapy [ICAAC], Washington, DC, 25 to 28 October 2008.)

The combination of the direct cytotoxicity assay from stool specimens and the cytotoxicity assay from isolates was considered the gold standard, so a true positive result was defined as positive by direct cytotoxicity assay or negative by direct cytotoxicity assay and positive by cytotoxic culture (1-3). ImmunoCard and Vidas-AB were performed according to the manufacturers' instructions. Due to the relatively long test time of Vidas-AB, a one-test-per-sample approach was adopted, and the equivocal values obtained with Vidas-AB were considered negative results.

During the study period, a total of 487 stool specimens from 412 patients were tested. Toxigenic C. difficile was detected using the gold standard in 62 stool specimens (12.7%) from 53 patients. The sensitivity of the direct cytotoxicity assay was 69.4%. The greater specificity of Vidas-AB compared with that of ImmunoCard was statistically significant (P = 0.019), although there were no statistically significant differences between sensitivities and predictive values (Table 1). Vidas-AB gave 13 equivocal results (2.7% of all results). These were for 2 specimens with toxigenic C. difficile and 11 specimens with negative results. ImmunoCard was the quickest technique: the median time to test 5 stool specimens was 24 min and 10 s using ImmunoCard, whereas Vidas-AB took 80 min and 25 s (P < 0.0001; two-tailed Wilcoxon test for paired samples). The hands-on times for both methods were very similar.

To our knowledge, two groups have published evaluations of Vidas-AB (4-6, 8). In the most complete evaluation of Vidas-AB in the three papers published by Shin et al., the sensitivity (63.3%) and specificity (96.7%) values (compared with toxigenic culture and PCR results for tcdA and tcdB as standards) were similar to those obtained in our study (4-6). Terhes et al. (8) compared Vidas-AB with the BD GeneOhm Cdiff assay, a real-time PCR test based on the detection of the tcdB gene of C. difficile, and achieved sensitivity and specificity values for Vidas-AB of 34.5% and 98.5%, respectively. Only two studies have evaluated ImmunoCard using toxigenic culture with or without the cytotoxicity assay as the gold standard (1, 7). In the previous evaluation of ImmunoCard performed in our laboratory (1), the sensitivity (66.7%) and specificity (95.1%) were similar to those obtained in the present study. In their comparison of several enzyme immunoassays and a real-time PCR for the detection of toxigenic C. difficile isolates, Sloan et al. (7) reported a lower sensitivity value (48.0%) and a higher specificity value (99%) for ImmunoCard than we did. These differences could be explained in part by the variability of their estimates due to the relatively low number of specimens tested (200).

In conclusion, our findings show that Vidas-AB was as sensitive as and more specific than ImmunoCard for the rapid diagnosis of CDI in clinical specimens. The use of Vidas-AB instead of ImmunoCard could increase the positive predictive value in the diagnosis of CDI, especially in populations with a low prevalence of the disease. However, the relatively low sensitivity and the limited positive predictive values shown by these tests and other enzyme immunoassays suggest that procedures such as glutamate dehydrogenase detection, toxigenic culture, and molecular detection of the toxin B gene are necessary to achieve an accurate diagnosis of CDI.

• Copyright © 2010 American Society for Microbiology