Indeterminate tcdB using a Clostridium difficile PCR assay: a retrospective cohort study

Abstract

Background: C. difficile (CD) real-time polymerase chain reaction (PCR) for toxin B gene (tcdB) is more sensitive, and reduces turnaround time when compared to toxin immunoassay. We noted typical amplification curves with high tcdB cycle thresholds (Ct) and low endpoints (Ept) that are labeled negative by the Xpert(®) C. difficile assay (Cepheid) and undertook this study to determine their significance.

Methods: We defined an indeterminate CD assay result as detection of a typical PCR amplification curve with an Ept >10 that was interpreted as negative by the Xpert(®) assay. Samples with indeterminate Xpert(®) result were collected for 5 months and retested by Xpert(®), cultured for toxigenic CD, and isolates subjected to PCR ribotyping, detection of toxin genes and multilocus variable-number tandem repeat analysis (MLVA) typing. Chart reviews were completed to assess if patients met the Society of Healthcare Epidemiology of America and the Infectious Diseases Society of America CD infection (CDI) clinical case definition. Illness severity was compared with tcdB Ct and culture results.

Results: During the 5-month study period, 48/3620 (1%) of specimens were indeterminate and 387/3620 (11%) were positive. Of the 48 patients with indeterminate results, 39 (81%) met the clinical case definition of CDI, and 7 of these (18%) met criteria for severe CDI. Toxigenic stool cultures were positive for 86% (6/7) of patients with severe CDI, 19% (6/32) of patients with non-severe CDI, and 44% (4/9) of patients who did not meet the clinical case definition of CDI (p = 0.002). Lower tcdB Ct and higher Ept were associated with greater likelihood of toxigenic culture positivity (p = 0.03) and more severe symptoms (p = 0.06). Indeterminate results were not associated with a particular technologist or instrument module, or CD strain type.

Conclusions: A subset of specimens (1%) using the Xpert(®) C. difficile assay have typical amplification curves and are interpreted as negative. At least one-third of these results are associated with positive CD culture. The mechanism of these indeterminate results is not technique-related, equipment-related, or due to particular CD strains. Clinicians should be aware that even PCR testing has the potential to miss CDI cases and further highlights the importance of clinical context when interpreting results.

Figure 1

Ribotyping and multiple loci variable-number tandem-repeat analysis (MLVA) results for C. difficile isolated from stool specimens with indeterminate Xpert® C. difficile assay results. % similarity was calculated using ribotyping and MLVA composite data sets using the unweighted pair group method with arithmetic mean. rel = related.