76. Validation of Systemic Inflammatory Mediators as Biomarkers for Severity and Adverse Outcomes in Clostridium difficile Infection

Abstract

Background Clostridium difficile infection (CDI) can result in severe disease and death. We are currently unable to identify patients at risk for developing adverse outcomes. We previously showed multiple inflammatory mediators were associated with severity and adverse outcomes. Here, we set out to validate these findings in patients and a murine model of CDI.MethodsCDI was diagnosed by the clinical microbiology laboratory. Sera were collected ≤48 hours after diagnosis from pilot (October 2010–November 2012) and validation (January–September 2016) cohorts. Inflammatory mediators were measured with a custom multiplex assay. IDSA severity was defined as serum creatinine >1.5-fold above baseline or white blood cell count >15,000 cells/mL. The 30-day outcomes were all-cause mortality and disease-related complications (DRCs): ICU admission, colectomy, or death attributed to CDI. We sought to validate our patient findings in a murine model of CDI: 67 antibiotic-treated mice were infected with 630 g (37 mice), a low virulence strain, or VPI 10463 (30 mice), a highly virulent strain. Host responses were assessed with a murine version of the multiplex panel. Unadjusted and adjusted models were built using logistic and L1 regression, respectively.ResultsThe pilot cohort had 156 CDI cases; 63 (40%) with IDSA severity. The inflammatory response in IDSA severe cases was distinct based on redundancy analysis of all measured analytes (P = 0.01). In unadjusted analysis, IL-2R, IL-6, and procalcitonin associated with severity (P < 0.001, P = 0.003, and P = 0.003, respectively). The same findings were seen in the validation cohort of 272 cases (Figure 1). Unadjusted analyses revealed several predictors of severity and outcomes (Table 1). Adjusted models performed well (Figure 2) with AUCs of 0.74 [0.67–0.81] (IDSA severity), 0.89 [0.83–0.95] (death), and 0.84 [0.74–0.95] (DRCs). Application of each model to the mouse cohort for high vs. low virulence infections revealed AUCs of 0.59 [0.44–0.74], 0.96 [0.90–1.0], and 0.89 [0.81–0.97] (Figure 3).ConclusionIn both humans and a murine CDI model, a panel of biomarkers from sera associated with severe CDI and predicted adverse outcomes. Our results support the possibility of a serum-based biomarker panel to inform medical decision-making for patients with CDI. Disclosures All Authors: No reported Disclosures.