Are rapid immunoassays for in vivo detection of toxin a sufficient for diagnostic purposes of Clostridium difficile-associated diseases?

One hundred and fifty-five stool specimens of patients suspected for Clostr idium difficile-associated diarrhoea, colitis or pseudomembranous colitis ( PMC) were investigated. All patients were pre-treated with antibiotics, suf fered from watery diarrhoea and abdominal pain and were hospitalized in dif ferent hospital units. Units varied from departments of surgery, internal m edicine, intensive care, paediatry, dermatology, orthopaedy to gastroentero logy. Fifty C. difficile strains were isolated from the faecal samples. Clo stridium difficile toxin detection was done directly in the stool samples, and also in cultured C. difficile strains (in vivo and in vitro, respective ly). We observed clear differences between in vivo and in vitro toxin A det ection by using commercial rapid immuno-enzymatic tests: from 25 in vivo to xin A-negative samples, 17 were positive in vitro. This observation suggest s that culturing of C. difficile on selective medium is mandatory for adequ ate toxin detection and necessary for confirming the presence of toxin-prod ucing C. difficile. This is especially important among patients with clinic al symptoms and history of pretreatment with antibiotics and when in vivo t oxin A detection is negative. It was established that toxin gene detection by PCR is optimal and PCR results were concordant with results of other in vitro assays. Genotyping by using AP-PCR and PCR ribotyping showed heteroge neity among the toxigenic C. difficile strains cultured from in viva toxin A-negative stool samples. (C) 2000 Academic Press.