P218 Over 30% of active flares in inflammatory bowel disease patients are associated with gastrointestinal infectious agents

Abstract

Background: GI infections (e.g. Clostridium difficile [CDI]) can worsen inflammatory bowel disease (IBD) activity and outcomes. By conventional methods, only 10% of symptomatic IBD patients (Figure presented) have detectable GI infections. Using the BioFire FilmArray GI PCR panel, a stool test capable of detecting 22 enteropathogenic organisms, we investigated the prevalence of GI infections in symptomatic flares of IBD vs. inactive IBD and healthy controls with the aim to determine the prevalence and impact of detected infectious agents in IBD patients. Method(s): We collected 5 patient cohorts: Active Crohn's disease (CD, n = 112), Inactive CD (n = 53), Active ulcerative colitis (UC, n = 128), Inactive UC (n = 39), and Healthy student Controls (HC, n = 52). The prevalence of positive stool tests was compared in patients with active inflammation (defined by biomarkers, endoscopy, and imaging) vs. those without inflammation. In active IBD patients medication use, and disease course were compared between those with positive and negative tests. Result(s): Overall, 31.25% of active IBD subjects had an infectious agent detected in their stool. The prevalence of infectious agents by group (Figure 1) was 33.9% in CD-active, 3.8% in CD-inactive, 28.9% in UC-active, 12.8% in UC-inactive (12.8%) and 13.4% in HC. Both CD-active and UC-active had significantly higher prevalence rates than inactive IBD. Prevalence increased to 50% in active IBD patients with acute onset (<7 days) of symptoms. Prevalence was 9-fold higher in active vs. inactive CD (p = 0.0001) and 2-fold higher in active vs. inactive UC (p = 0.04). Despite limited statistical power for individual medications, infectious agents were significantly associated with use of vedolizumab, odds ratio of 3.9 (95% CI: 1.7-9.2), and tacrolimus (OR 31.6, 4.0-247.2), but not with anti-TNFs, ustekinumab, or thiopurines. C. difficle was the most common infection detected, followed by coliforms and norovirus (Figure 2). Conclusion(s): Active CD and UC have a higher rate of detectable infectious agents than patients with quiescent disease, suggesting a significant proportion of symptomatic IBD flares could be due to intestinal infection. Few medications are associated with increased risk of a detectable infectious agent, but vedolizumab and tacrolimus are associated with increased risk. These infections (other than C. difficile infection [CDI]) appear to be mostly transient, and treatable with fluids and supportive care. Reflexive use of steroids for symptomatic flares could be harmful in the many IBD flares associated with an intestinal infectious agent.