P225 Day of admission results predict failure of first-line treatment in acute ulcerative colitis

• Published in: Journal of Crohn's and colitis 2019-01, Vol.13 (Supplement_1), p.S209-S209
• Main Authors: Grant, R, Lynch, R, Bouri, S, Elosua González, A, Manship, T, Jagger, F, Shivakumar, M, Satsangi, J, Ho, G-T, Lees, C, Plevris, N, Tozer, P, Hart, A, Arnott, I
• Format: Article
• Language: English
• ISSN: 1873-9946; 1876-4479
• DOI: 10.1093/ecco-jcc/jjy222.349

Abstract Background Intravenous (IV) steroids remain the standard first-line treatment for patients admitted with acute ulcerative colitis (UC). However, 30% of patients fail to respond and require second-line therapies and/or surgery. The purpose of this study was to determine whether Day 1 parameters could identify a group at high risk of failing first-line therapies. Methods All admissions for acute UC (ICD-10 K51) to hospitals in NHS Lothian (4 sites) and St Mark’s Hospital, Harrow from 1/11/11 to 31/10/16 were obtained from the regional coding departments. Case record review was performed. Response to IV steroids was defined as discharge from hospital with no further acute medical or surgical treatment. Non-response was defined as need to escalate to ciclosporin, infliximab, other acute therapy, or to have surgery. The following parameters were recorded for the first 10 days post admission: haemoglobin (Hb), platelet count, CRP, albumin, stool frequency and faecal calprotectin. Each patient was later attributed a score based on CRP (≤50 mg/dl = 0; >50 mg/dl = 1), albumin (≥30 g/l = 0; < 30 g/l = 1) and platelets (≤400 × 109/l = 0; >400 × 109/l = 1). Results In total, 592 admissions with acute UC were identified; 391/592 (66%) responded to steroids, 201/592 (34%) patients were non-responders. 44 (22%) non-responders received infliximab as second-line therapy, 108 (54%) cyclosporine, and 4 (2%) other. Eighty-three (41%) non-responders required surgery; 7 (8%) had infliximab prior to surgery; 35 (42%) cyclosporine; 12 (14%) went straight to surgery. Insufficient data were available regarding 33 patients. On univariate analysis, albumin (p =