P503 Can daily SCCAI measurements over a 6-month period better phenotype the pattern of disease in ulcerative colitis?. (16th January 2018)

• Record Type: Journal Article
• Authors: Walsh, A; Kormilitzin, A; Hinds, C; Sexton, V; Wilson, J; Peters, M; Brain, O; Keshav, S; Uhlig, H; Geddes, J; Goodwin, G; Collins, G; Travis, S
• Is Part Of: Journal of Crohn's and colitis. Volume 12:Number 1(2018:Jan.)Supplement 1
• Page Start: S359
• Page End: S360
• Publication Date: 2018-01-16
• Subjects: Inflammatory bowel diseases -- Periodicals; 616.344005
• DOI: 10.1093/ecco-jcc/jjx180.630
• Languages: English
• ISSNs: 1873-9946

Abstract

Abstract: Background: Access to daily disease activity measurements may allow more accurate phenotyping of the pattern of disease in UC. Clustering patients by the percentage of time spent in each disease category (remission, mild, moderate, severe) may provide a more objective classification of the pattern of disease. Methods: TrueColours UC, a comprehensive real-time web-based programme, collected daily SCCAI results from 66 patients for 6 months. Disease pattern was defined as the proportion of time spent in each SCCAI disease activity category: remission (SCCAI 0–2), mild (SCCAI 3–5), moderate (SCCAI 6–11) and severe (SCCAI 12). Principal component analysis (PCA) and unsupervised clustering was performed. To determine whether clustering provided adequate separation of different disease patterns, each data point was examined with reference to three clinical categories: Remission (>80% time with SCCAI 2 and no time with SCCAI 5), Relapsing (some time spent in remission, but with isolated episodes of SCCAI 5 requiring escalation in medical therapy), and Persistently Active (>80% time spent with SCCAI 3). Results: The unsupervised cluster plot identified five clusters (Figure 1A–E), with most clusters matching the pre-determined categories. The Remission category matched Cluster A, with patients in clinical remission for >80%. Other clusters represented different patterns of active disease. The Relapsing category matched Cluster B, with 11 of 11 patients having an exacerbation requiring escalation of therapy (prednisolone in 7 of 11). The Persistently Active category matched with Group D and Group E. In group D, 3 of 3 patients spent >50% in mild disease activity and >30% in moderate disease activity. In group E, six of six patients spent < 5% in remission and >80% in mild disease activity. In Cluster C, all patients had some active disease, but had components of both relapsing and persistently active categories. Conclusions: Unsupervised clustering from daily symptom scores of SCCAI accurately classified most patients. Most importantly, it separated those in remission from those with active disease. It also enabled sub-classification of those with active disease.