P218 Linked colour imaging predicts the risk of the clinical relapse in ulcerative colitis

Abstract

Abstract Background Ulcerative colitis (UC) is a chronic inflammatory bowel disease with unknown aetiology. Recently, mucosal healing (MH) has emerged as an important therapeutic endpoint in UC. Linked colour imaging (LCI) is a novel endoscopic system that enhances the colour differences of the gastrointestinal mucosa. We aimed to evaluate the correlation between the endoscopic observation using LCI and the clinical relapse rate in UC patients. Methods We retrospectively analysed UC patients who underwent a total colonoscopy between August 2016 and May 2019 at our facility with Mayo endoscopic score (MES) of 0 or 1. We assessed the most relevant endoscopic findings, including erythema, vascular pattern, and friability using white-light imaging (WLI) and LCI observation. We analysed the correlation of clinical relapse rate, which means requiring additional treatment for UC, during the 2-year follow-up period with endoscopic assessment using WLI and LCI. We also evaluated the utility of LCI-observation method for predicting the prognosis in UC patients with clinical remission. Results Forty-seven patients were analysed, and clinical relapse was observed in 10 patients (21%) during the 2-year. Univariate analysis revealed a significant difference in decreased vascular pattern and friability between the clinical relapse and remission groups; multivariate analysis showed decreased vascular pattern observed on LCI as the only variable, which remained significantly associated with clinical relapse (p <0.03). We defined LCI-mucosal score (MS) as LCI-MS 0, distinct vascular pattern and no friability; LCI-MS 1, either indistinct purple/magenta vascular pattern or friability; and LCI-MS 2, both indistinct purple/magenta vascular pattern or friability on LCI observation. The relapse rate in patients with LCI-MS 0 was 10% (3/30), which was significantly lower than in patients with LCI-MS 1 or 2 (36 %; 6/16). Conclusion LCI findings have the diagnostic implications in predicting the risk of clinical relapse in UC during the 2-year follow-up period.