The organism: Clostridium difficile are spore forming, gram-positive rods that grow under anaerobic conditions. The organisms are part of bowel flora in 3% of general population, and most strains do not normally cause disease. Most patients develop symptoms after antibiotic therapy, which alters the flora and allows for outgrowth of the organisms. C. difficile produces two toxins, which are part of a pathogenicity island on the bacterial chromosome (absent in non-pathogenic strains). Toxin A is an enterotoxin thought to cause the clinical symptoms. It is a large, 308 kDa protein, with a series of repeating units near the carboxyl terminus which recognize galactose containing receptors, likely responsible for the granulocyte chemo-attractant property of the molecule. Toxin B is a cytotoxin; there is some evidence suggesting that it acts synergistically with Toxin A by potentiating the tissue damage caused by Toxin A.

Laboratory Diagnosis: Since C. difficile is part of normal flora, the laboratory analysis is focused on detecting toxin-producing organisms. The assay most commonly used at JHH is a sandwich ELISA test (for toxins A and B) performed on stool samples. In this assay, the primary polyclonal antibodies to Toxins A and B are immobilized on the surface of the plate. The detecting antibodies (conjugated to horseradish peroxidase) are added to the plate wells along with the stool sample prepared in diluent. In this reaction, a "sandwich" is formed between the immobilized primary antibodies which bind the toxins, and the detecting antibodies, which bind the toxin bound to the immobilized primary antibodies. The excess antibodies are washed, leaving only the antibody-antigen-antibody complexes on the plate, which are detected by a colorimetric method utilizing horseradish peroxidase enzymatic activity.

Another, older assay for detection of C. difficile relies on the cytotoxic properties of Toxin B. In this assay, mammalian tissue culture cells are exposed to a filtered (to eliminate live bacteria) stool sample. The presence of cytopathic effect, similar in morphology to that observed in virus detection, is suggestive of the presence of the toxin. To assure specificity, a neutralizing antibody to Toxin B is added in parallel: if the cytopathic effect is not seen in presence of the antibody, it may be safely concluded that the effect is due to the toxin and not other agents contained in the stool sample. Since Toxin A and B are usually co-expressed, measuring only Toxin B correlates well with clinical disease.

In the case presented here, the ELISA assay was repeatedly negative, whereas the tissue culture assay was positive. This was due to "hook" effect, which can be seen in sandwich ELISA assays in the presence of high concentrations of antigen. The concentration of Toxin A and B in this patients sample was so high, that it was sufficient to saturate the binding of the immobilized primary AND the detecting secondary antibodies independently, and no "sandwich" was formed. Therefore, in the washing step, the secondary detecting antibody, bound to the Toxins, was lost and the assay was read as false negative. Due to high clinical suspicion of the ID team, the tissue culture assay was performed which revealed cytopathic effect. The ELISA assay was repeated with diluted stool samples (which decreased the antigen concentration to allow sandwich formation) with positive readout. This was the first episode of "hook" effect observed in this assay at JHH. The manufacturer was contacted, and indicated that "hook" effect had not been previously reported for this assay. A prospective study to examine the incidence of "hook" effect in the C. difficile ELISA negative specimens will be performed by the virology lab staff at JHH.