P125 Histological activity as a predictor of clinical outcome in ulcerative colitis: a 1-year prospective study

Abstract Background The role of histological activity in clinical management of ulcerative colitis (UC) is under investigation. Primary aim was, in a prospective study, to assess the role of histological activity as predictor of clinical relapse in a cohort of UC patients (patients) undergoing colonoscopy and followed-up for 1 year. Secondary aim was to assess the correlation between clinical, endoscopic and histological activity scores. Methods From February 2016 to February 2017 consecutive UC patients with clinical indication for colonoscopy were enrolled and clinically followed-up for 1 year. Inclusion criteria: (1) UC diagnosis; (2) Age > 18, ≤ 80 years; (3) regular follow-up; (4) indication for colonoscopy. During colonoscopy ≥2 biopsies was taken from ≥1 macroscopically involved and, possibly, from ≥1 uninvolved area. The day of colonoscopy clinical activity was assessed by the Mayo partial score, endoscopic activity by the Mayo endoscopic score, histological activity by the Geboes Simplified Score (GSS). Scores blindly assessed by three investigators. Statistical analysis: data expressed as mean [range], Spearman’s correlation coefficients, Cox hazards regression model used for univariate and multivariate analyses to identify predictors of clinical relapse at 1 year (HR[95% CI]). Results UC cohort included 77 UC patients. Characteristics of these 77 UC patients: 43 (55.8%) males, age 51 [24–80]; UC duration 14.7 [1–48] years. UC extent included n (%): 33 (42.8%) pancolitis, 24 (31.2%) left-sided, 20 (26%) proctitis. The day of colonoscopy, UC was clinically active in 15 (19.4%), inactive in 62 (80.6%) patients. Endoscopic activity was observed in 39 (50.6%) patients, histological activity (GSS≥ 3.1) in 37(48%) patients. Moderate correlations were observed between clinical and endoscopic scores (r = 0.439;p < 0.0001) clinical and histological scores (r = 0.32;p = 0.0045), endoscopic and histological scores (r = 0.653;p < 0.0001). During the clinical follow-up at 1 year, UC clinical relapse occurred in 24 (31%) patients, while 53 (69%) patients maintained clinical remission. At baseline colonoscopy, 11/24 (46%) UC patients were clinically active, 15/24 (63%) showed endoscopic activity and 16/24 (67%) patients histological activity. Univariate analysis identified clinical activity (HR 4.82 [2.15–10.82]; p < 0.001) and histological activity (HR 2.599 [1.11–6.08]; p < 0.027) as significant predictive factors for clinical relapse at 1 year. Multivariate model confirmed histological activity as predictive marker of clinical relapse (HR 2.44 [1.04–5.75]; p < 0.041). Conclusion Histological activity provided independent information for clinical relapse in a cohort of UC patients prospectively followed up for 1 year. Histological activity had a significant correlation with the endoscopic and clinical activity scores.