745 Developing Precision Medicine in IBD: Peripheral Blood Eosinophilia Identifies Individuals With Worse Outcomes Following Initiation of Anti-TNF Biologic Agents

Abstract

INTRODUCTION: Peripheral blood eosinophilia (PBE) has been recognized as a biomarker, identifying a subgroup of inflammatory bowel disease (IBD) patients with worse clinical outcomes. With more severe disease, anti-TNF therapy is often employed, with varied rates of success. The aim of this study was to determine whether the presence of pre-treatment PBE identifies IBD patients at risk for diminished anti-TNF treatment response. METHODS: We performed a registry analysis of IBD patients starting anti-TNF therapy who were enrolled in a consented, prospective, natural history IBD cohort at a tertiary center. Metadata was collected in the 2 years preceding and 2 years following anti-TNF initiation. The presence of PBE before and after anti-TNF initiation was recorded and treatment response was assessed for those with/without PBE using disease activity scores [Harvey Bradshaw Index (HBI) and Ulcerative Colitis Activity Index (UCAI)], quality of life measures [Short Inflammatory Bowel Disease Questionnaire (SIBDQ)], inflammatory markers (ESR, CRP), steroid use (prednisone or methylprednisolone), IBD-related hospitalizations, and IBD-related surgeries. Outcomes were assessed via linear mixed-effects and Cox proportional-hazards models for Crohn's disease (CD) and ulcerative colitis (UC). RESULTS: A total of 580 IBD patients (CD 430, UC 150) were included with 111 (19%) demonstrating pre-treatment PBE (CD 66, UC 45). Table 1 summarizes demographic and clinical data. Following anti-TNF therapy, CD patients with pre-treatment PBE were more likely to have an IBD-related hospitalization [adjusted hazard ratio (AHR) 1.88, 95% CI 1.22-2.90, P = 0.004] and require anal fistula repair (AHR 3.88, 95% CI 1.09-13.8, P = 0.04) than CD patients without PBE. UC patients with pre-treatment PBE were more likely to have an elevated CRP following anti-TNF therapy (OR 2.77, 95% CI 1.50-5.10, P = 0.001) with a non-significant trend towards increased IBD-related admissions and surgeries. CD patients with pre-treatment PBE were less likely to require steroids following anti-TNF therapy than CD patients without PBE (AHR 0.40, 95% CI 0.18-0.86, P = 0.019). There was no significant difference in HBI, UCAI, or SIBDQ scores in patients with/without PBE undergoing anti-TNF therapy. CONCLUSION: Pre-treatment PBE identifies IBD patients at risk for worse outcomes following anti-TNF therapy, particularly in CD. Further studies are warranted to determine if alternate mechanism of action biologics will function optimally in the PBE IBD population.