2380. Fecal Collinsella Abundance is Negatively Associated with Toxin A/B Production in Cancer Patients with Clostridioides difficile

BACKGROUND: The detection of C. difficile (CDI) by nucleic acid amplification test (NAAT) with negative toxin enzyme immunoassay (EIA-) is difficult to interpret in cancer patients. Markers that differentiate true infection from colonization, and are associated with clinical outcomes are needed. We hypothesized that the microbiome composition and inflammatory fecal markers in EIA- patients differed from those who are EIA+ and were associated with disease severity and recurrence. METHODS: We studied the fecal microbiome composition (16s rRNA, V3) of 147 cancer patients with CDI diagnosed by a two-step testing algorithm. Clinical data, CDI bacterial quantity (BQ) by qPCR and markers of intestinal inflammation (calprotectin, lactoferrin, IL-1β and IL-8) were analyzed. Data were stratified according to cancer type [hematologic (H) n = 49, solid tumor (ST) n = 66, or stem cell transplant (SCT) n = 32]. RESULTS: Demographic characteristics and symptoms were similar between the three groups. At baseline, species diversity by Shannon index was similar in all three groups regardless of EIA detection and did not correlate with clinical presentation, response to therapy or recurrence. Microbiome composition did not correlate with inflammatory response except in H in whom a higher diversity correlated with increased IL-8 (P = 0.021) and calprotectin (P = 0.01) levels. At the genus level across all strata and when compared with EIA- cases, EIA+ cases presented with a higher abundance of Peptoclostridium (P = 0.0008) which correlated with CDI BQ qPCR (log of BQ/mg 2.38 ± 1.49 vs 0.92 ± 1.28, P < 0.001). In contrast, EIA- cases had a higher abundance of Collinsella (P = 0.001). SCT patients carried fewer Peptoclostridium when compared with other groups, whereas all three patient groups carried similar amounts of Collinsella. The relative abundance of Peptoclostridium and Collinsella was not associated with response to therapy, or fecal markers of inflammation. Principal component analysis did not demonstrate differences between the three groups studied. CONCLUSION: In this study, the presence of Collinsella, a known butyrate and bile salt hydrolase producer, was associated with the lack of CDI toxin A/B production. Loss of Collinsella may represent a novel risk factor for active CDI. [Image: see text] [Image: see text] [Image: see text] [Image: see text] DISCLOSURES: All authors: No reported disclosures.