Toxin B PCR cycle threshold as a predictor of poor outcome of Clostridium difficile infection: a derivation and validation cohort study.

Abstract

Prediction of patients with poor outcome is necessary in order to plan the proper management of Clostridium difficile infection (CDI); however, clinical criteria are insufficient. In a previous study, we observed that high toxigenic C. difficile cfu stool counts at diagnosis were associated with a poor outcome. Our objective was to investigate the role of the PCR toxin B amplification cycle threshold (Ct) in the prediction of CDI poor outcome and to derive and validate a high-risk prediction rule using this marker. We prospectively included patients with CDI (derivation cohort, January 2013 to June 2014; and validation cohort, December 2014 to May 2015), who were followed for at least 2 months after their last episode/recurrence. All samples were tested with Xpert™ C. difficile. For the derivation cohort (n = 129) toxin B Ct was independently associated with poor outcome (P < 0.001). The receiver operating characteristic (ROC) curve yielded an AUC of 0.816. Using a cut-off of <23.5 cycles for high risk of poor outcome, the diagnostic accuracy was 81.4%, the sensitivity was 46.5% (95% CI 32.5-61.1) and the specificity was 98.8% (95% CI 93.7-99.8). For the validation cohort (n = 170), the diagnostic accuracy was 81.8%, the sensitivity was 88.4% (95% CI 75.5-94.9) and the specificity was 79.5% (95% CI 71.7-85.6). The ROC curve yielded an AUC of 0.857. Low toxin B Ct values from samples collected at the initial moment of diagnosis appears to be a strong marker for poor outcome. This available test may identify, at an early stage, patients who are at higher risk of a poor outcome CDI.