P112 Validation of the simplified Geboes score for ulcerative colitis
A. Jauregui-Amezaga*1, Y. Das2, T. Lobatón3, B. Lemmens2, T. Bessissow4, M. Ferrante1, G. Van Assche1, R. Bisschops1, K. Geboes2, G. De Hertogh2, S. Vermeire1
1University Hospitals Leuven, Gastroenterology, Leuven, Belgium, 2University Hospitals Leuven, Pathology, Leuven, Belgium, 3Hospital Universitari Germans Trias i Pujol, Gastroenterology, Badalona, Spain, 4Mc Gill University Health Centre, Gastroenterology, Montreal, Canada
A Simplified Geboes Score (SGS) that includes the presence of basal plasmacytosis, eosinophils, and neutrophils in lamina propria (LP), neutrophils in epithelium, and epithelial injury has been proposed for histological assessment of ulcerative colitis (UC). The aim of this study was to validate the SGS prospectively.
UC patients treated at our tertiary referral centre were prospectively included from January 2011 to March 2014. Patients were followed-up until a) Eendoscopic reassessment because of clinical symptoms, change of medication, or screening for dysplasia or b) end of the follow-up period in October 2015. Endoscopic activity was defined as Mayo Endoscopic Sub-score ≥1. Histological activity was assessed with the Original Geboes Score (OGS) and the SGS by 2 trained readers (readers A and B) and an expert gastrointestinal pathologist (reader C) blinded to clinical data.
Included were 92 UC (72% males, mean age 47 years). Amongst them, 46 (50%) presented with Mayo 0, 18 (20%) with Mayo 1, 11 (12%) with Mayo 2, and 17 (18%) with Mayo 3. Overall, 260 slides were scored. Histological activity defined as a presence of neutrophils in the biopsy (SGS ≥ 2B.1) was present in 52/92 patients (57%): 39/46 (85%) of patients with Mayo ≥1 and 13/46 (28%) of patients with Mayo 0 at inclusion (p < 0.001). At the end of follow-up, 67/92 (73%) patients underwent endoscopic control, and 38 of them (57%) showed endoscopic activity: persistence of active endoscopic disease in 26/67 (39%) and relapse of previously endoscopically inactive UC in 12/67 (18%). Basal plasmacytosis (OR 4.7 [IC 95% 1.6–14.2], p = 0.006), eosinophils in LP (OR 4.3 [IC 95% 1.4–13.7, p = 0.01), neutrophils in LP (OR 7.1 [2.4–21.3], p < 0.001), neutrophils in epithelium (OR 7.4 [IC 95% 2.5–21.8], p < 0.001), and epithelial injury (OR 6.8 (IC 95% 2.3–20.3, p = 0.001) at the baseline endoscopy were identified as predictors of endoscopic activity in univariate analysis, and neutrophils in epithelium (OR 7.4 [IC 95% 2.5–21.8], p < 0.001) as an independent predictor of endoscopic activity in multivariate analysis. Only neutrophils in LP (OR 7 [IC 95% 1.3–38], p = 0.02) and epithelium (OR 12 [IC 95% 1.2–124], p = 0.04) were predictors of relapse in patients with Mayo 0. Inter-observer agreement was better for the SGS than for the OGS (kappa 0.7 vs 0.6 for readers A-C and kappa 0.7 vs 0.5 for readers B-C).
The SGS is a useful and valid histological score for the assessment of UC. All its components (basal plasmacytosis, eosinophils, and neutrophils) have been identified as predictors of endoscopic activity. Only neutrophils have been associated with relapse in patients with Mayo 0. The inter-observer agreement is better with the SGS that with the OGS.