P410 Segmental endoscopic scoring improves outcome prediction in acute severe colitis
S. Raimundo Fernandes*, P. Santos, C. Moura, P. Marques da Costa, J. Rita Carvalho, C. Baldaia, A. Valente, A. Rita Gonçalves, P. Moura Santos, L. Correia, J. Velosa
Centro Hospitalar Lisboa Norte, Gastrenterology, Lisbon, Portugal
Up to 25% of patients with ulcerative colitis will be admitted with acute severe colitis (ASC). Although the majority will respond to intravenous steroids, a third will require rescue therapy or surgery. Endoscopy is mandatory in ASC, but its role in predicting important outcomes has seldom been evaluated.
The electronic records of patients admitted to our centre between January 1993 and October 2015 with ASC, according to the Truelove–Witts criteria, were reviewed. For each patient, the endoscopic Mayo (Mayo) and Ulcerative Colitis Endoscopic Index of Severity (UCEIS) were determined. In addition, we produced a segmental Mayo (sMayo) and UCEIS (sUCEIS) score by adding the values of the rectum and sigmoid evaluation. The area under the receiver operating characteristics curve (AUC) was used to assess the accuracy of each scoring system in predicting intensive care unit (ICU) admission, need for rescue therapy, colectomy, and death.
In total, 108 patients, 60 (55.6%) male, mean age 37.6 ± 15.3 years, were enrolled in this study. All patients received intravenous steroids. Further, 44 patients (40.7%) were admitted in the ICU, and 4 patients (3.7%) died. In addition, 35 patients (34.3%) required rescue therapy (26 with infliximab and 9 with cyclosporin), and 38 patients (33.3%) underwent colectomy during the index admission or during the first year of follow-up. Median (range) endoscopic scores were Mayo 3 (1–3), sMayo 6 (1–6), UCEIS 5 (2–8), and sUCEIS 10 (4–15). The UCEIS and sUCEIS both accurately predicted ICU admission (area under the curve [AUC] 0.678 and 0.664, p < 0.004) and the need for colectomy (area under the curve [AUC] 0.715 and 0.744, p < 0.001). The sUCEIS was the only predictor of the need for rescue therapy or surgery (AUC 0.664, p = 0.004) and of any adverse outcome (AUC 0.638, p = 0.021). Patients requiring colectomy had significantly higher UCEIS (5 [4–8] vs 4 [2–8]) and sUCEIS scores (10 [7–15] vs 8 [4–14]), p < 0.001. In logistic regression analysis, the UCEIS (OR 2.0 CI95% [1.37–2.95]) and the sUCEIS (OR 1.5 CI 95% [1.21–1.85]) were the only independent predictors of colectomy. A cut-off of UCEIS ≥ 7 or sUCEIS ≥ 12 identified 75.0% and 72.2% of patients requiring rescue therapy or colectomy (p < 0.001).
Endoscopic severity can predict worse outcome in patients with ASC. The UCEIS score outperformed the Mayo score in all important outcomes. Segmental endoscopic scoring can improve outcome prediction, especially surgery, in patients with ASC.