487. Severity and Clinical Outcomes of Clostridium difficile Infection Based on Toxin B Assay Results

BACKGROUND: Clostridium difficile infection (CDI) remains a major health problem in the United States. The IDSA guidelines recommend using stool toxin assay as part of a multistep algorithm rather than nucleic acid amplification test (NAAT) alone. However, the clinical significance of toxin negative tests remains a subject of debate. We performed a prospective study in our institution to describe clinical outcomes of CDI based on the results of the stool toxin assay. METHODS: Our laboratory utilizes a 2-step algorithm, using glutamate dehydrogenase plus detection of toxin B by enzyme immunoassay (EIA) arbitrated by NAAT for testing stool samples submitted for C. difficile testing. The study was conducted between January and December 2017. Patients diagnosed with CDI based on laboratory results were divided into two groups based on toxin B assay results. Shotgun metagenomics was performed directly on stool specimens using Illumina NextSeq in a subset of patients. Chart reviews were performed to assess clinical outcomes. Our primary outcome was incidence of severe CDI and 30-day mortality. RESULTS: A total of 2,823 samples were submitted to the laboratory for testing for suspected CDI. Three hundred thirty-eight samples in 290 discrete patients were considered positive using the two step algorithm. Whole genome sequencing was performed on samples from 57 patients (Figure 1). Clinical outcome data were available for 53 patients. Thirty percent were on active chemotherapy. Thirty-four patients were toxin B positive (group 1), 19 were toxin B negative (group 2) by EIA. Hospital onset disease was seen in 10 (27%) of patients in group 1 vs. 7 (37%) in group 2 (P = 0.57). Thirty-day mortality was 3% in toxin positive vs. 5% in toxin negative groups (P = 0.67). Severe CDI was seen in 14 (41%) in group 1 vs. 8 (42%) in group 2 (P = 0.94). NAP 1 strain was detected in 10.5% of patients in group 2. Percentage of C. difficile reads on sequencing in fecal samples in group 1 (0.17%) was not significantly different from group 2 (0.24%) (P = 0.70, Figure 2). CONCLUSION: In our cohort, detection of C. difficile toxin in stool samples was not associated with increased severity of disease. Our cohort has a higher prevalence of patients on active chemotherapy than previously studied cohorts. Bioburden of C. difficile was not significantly different in toxin positive and negative disease. [Image: see text] [Image: see text] DISCLOSURES: All authors: No reported disclosures.