Guidelines Support the Value of Stand-Alone Nucleic Acid Amplification Tests for Clostridioides (Clostridium) difficile Infection

LETTER

The commentary by Sandlund and Wilcox (1) entitled “Ultrasensitive detection of Clostridium difficile toxins reveals suboptimal accuracy of toxin gene cycle thresholds for toxin predictions” discusses the potential role of “ultrasensitive” toxin assays for diagnosis of Clostridioides (Clostridium) difficile infection (CDI) in comparison to the off-label use of nucleic acid amplification tests (NAATs), in which cycle threshold values are used to predict the presence of free toxin in a stool specimen. Unfortunately, the authors fail to discuss on-label testing with NAATs to put toxin testing into perspective. Guidelines from the Infectious Diseases Society of America–Society for Healthcare Epidemiology of America (IDSA-SHEA) (2), the American Society for Microbiology (3), and the European Society for Clinical Microbiology and Infectious Diseases (ESCMID) (4) all make it clear that NAATs play an essential role in the laboratory diagnosis of CDI. The IDSA-SHEA guideline states that if patients meet clinical criteria for CDI, are not on laxatives, and the laboratory does not test formed stool, either a stand-alone NAAT or an algorithm-based approach is acceptable (2). The ASM guideline also endorses a role for stand-alone NAATs for CDI.

The authors’ statement that using NAATs for the diagnosis of CDI leads to overdiagnosis is true only if the above criteria are not met. Their article perpetuates the erroneous belief that NAATs are unique in their potential to lead to overdiagnosis of CDI. In fact, any diagnostic assay for Clostridioides difficile, including ultrasensitive toxin assays, can be positive in asymptomatic carriers as well as in patients with CDI (5). A recent study of an ultrasensitive toxin assay by Pollock et al. “…clearly demonstrated that stool toxin A and B concentration alone cannot distinguish a patient with CDI diagnosed by NAAT or toxin detection from an asymptomatic carrier because concentration distributions on both types of patients overlap substantially” (5).

Studies cited by the authors correlating toxin levels with CDI mortality and severity (6, 7) are likely to be outliers and have been contradicted by other studies, including by a large multicenter study (5, 8–16). Other studies have demonstrated a correlation between cycle threshold values and free toxin levels (17). While there is some correlation between organism burden, free toxin levels, and clinical features of CDI, no specific cutoff value, whether a PCR-based cycle threshold or an ultrasensitive toxin level, can reliably distinguish patients with asymptomatic carriage from those with CDI (5).

We are not advocating the off-label use of NAATs. Rather, we simply note that three recent guidelines support the value of NAATs for diagnosing CDI, while none indicates a role for ultrasensitive toxin tests. We do agree with Sandlund and Wilcox that the diagnosis of CDI must combine clinical presentation with diagnostic testing. However, the key to minimizing CDI overdiagnosis is appropriate patient selection, regardless of the diagnostic method used (18). Standalone NAAT testing continues to be widely used for CDI. NAATs offer the combination of speed, sensitivity, high negative predictive value, and cost effectiveness when used appropriately (19, 20). Whether ultrasensitive toxin testing can truly improve patient outcomes remains to be seen.