P375 Histological activity predicts the risk of colectomy after biological treatment in biological-naive patients with Ulcerative Colitis

Abstract

Abstract Background Treatment failure with biological agents is common in patients with ulcerative colitis (UC), but the impact of baseline histological disease activity is unknown. We aimed to investigate the predictive role of histopathological findings for biological treatment failure in UC patients. Methods This is a sub-study of the Danish IBD Biobank Project. Adult bio-naïve patients with UC (n=129) were followed prospectively while undergoing biological treatment. Histological activity was assessed using the Nancy Index and the Geboes score in intestinal biopsies sampled before treatment. Endoscopic activity was assessed using the Mayo Endoscopic Subscore, and clinical activity was measured using the Simple Clinical Colitis Activity Index (SCCAI). Associations between histological activity and treatment outcomes were assessed in multivariable models which incorporated the following fixed covariates: age, gender, disease duration, body mass index >30, disease extent, the SCCAI, and the Mayo Endoscopic Subscore. Correlations between histological, endoscopic, clinical, and biochemical activity were assessed using Spearman’s rank correlation. Results Colectomy was needed in 17.1% (n=22) of UC patients during follow-up (Table 1). Histological activity at baseline measured using the Nancy Index was the only significant predictor of the risk of colectomy during follow-up (Nancy 2 vs. 4: HR 0.14, 95% CI 0.02-0.84, p=0.03; Nancy 3 vs. 4: HR 0.11, 95% CI 0.02-0.53, p=0.01) beside baseline haemoglobin concentration (HR 0.32, 95% CI 0.16-0.64, p<0.01. Figure 1). The risk of colectomy was also significantly predicted by a Geboes grade above 5.0 at baseline (HR 2.72, 95% CI 1.14-6.50, p=0.02). In contrast, endoscopic and clinical activity did not predict the risk of colectomy. At baseline, the Nancy Index and the Geboes score were moderately correlated to endoscopic activity, and weakly, but significantly correlated to biochemical markers haemoglobin, albumin, and C-reactive protein. Meanwhile, poor correlations were found between histological activity and clinical activity. Conclusion The risk of colectomy during biological treatment in bio-naïve UC patients was significantly predicted by histological disease activity, but not endoscopic or clinical disease activity. Our study suggests that histological assessment before initiating biological treatment should be encouraged in UC patients. Funding This study is funded by a restricted grant from Takeda A/S, Louis-Hansen Fonden, and a public fund hosted by Hvidovre Hospital.