Rapid Ultrasensitive Detection of Clostridiodes difficile Toxins in Stool Samples Using A Single-Molecule Counting Method

Abstract

Background:Clostridiodes difficile infection is considered an urgent antibiotic resistance threat by the CDC, accounting for ∼225,000 hospitalizations, 12,800 deaths, and ∼$1 billion in healthcare costs in the United States in 2017. The presence of the secreted toxins that cause the devastating symptoms of this gastrointestinal infection are diagnostic of C. difficile infection (CDI). However, the rapid testing methods currently used to detect CDI lack accuracy. Enzyme immunoassays are specific but lack sensitivity because they do not detect CDI patients that have low levels of the toxins. Nucleic acid amplification tests (NAATs) are sensitive, but they lack specificity because they detect patients colonized with C. difficile in the dormant spore form that does not produce the toxins. This insufficiency has resulted in the adoption of complex multitest algorithms for C. difficile diagnosis. We present results for a new toxin test that demonstrates both high clinical sensitivity and clinical specificity for C. difficile toxin B on a fully automated benchtop platform. Methods: The detection technology uses nonmagnified digital imaging to count single toxin molecules that tether together target-specific magnetic and fluorescent particles. The 30-minute method includes the use of a dye cushion to eliminate wash steps and the need for time-consuming specimen preparation steps. We determined analytical performance characteristics of the test using negative clinical stool samples spiked with purified toxin. To assess clinical performance, we tested 785 stool samples from 5 clinical sites and compared the results with the cellular cytotoxicity neutralization assay (CCNA). Results: The test’s limit of detection for toxin B was 60 pg/mL. A comparison of the new test to the CCNA reference method gave 98% positive percentage agreement (83 of 85 samples) and 95% negative percentage agreement (667 of 700 samples). Conclusions: The new method demonstrated 96% accuracy compared to the gold standard for C. difficile toxin tests. The results also demonstrate an analytical sensitivity (limit of detection, 60 pg/mL). Thus, the test has the potential to detect CDI patients missed by enzyme immunoassay (EIA) tests due to their low analytical sensitivity. Because the test detects toxins directly, it is expected to have a lower false-positive rate than NAAT methods, which detect patients colonized with the non–toxin-producing spore form. A single accurate test for toxin-producing C. difficile could eliminate the need for multitest algorithms.

Funding: First Light Diagnostics, Inc., provided support for this study.

Disclosures: Donald Straus reports that he is the founder and chief scientific officer of First Light Diagnostics (FLDx) with salary and ownership interest in the form of stocks, stock options, and warrants. Adam Williams reports salary from First Light Diagnostics.