Early Response to Corticosteroid and Baseline C-Reactive Protein Predicts Outcomes in Children with Moderate to Severe Ulcerative Colitis

BackgroundInitial response to corticosteroids (CS) is recognized as a strong predictor of outcomes in ulcerative colitis (UC).AimTo compare outcomes of early poor responders (PR) versus good responders (GR) to initial CS at 1, 2, and 3years from diagnosis.MethodsIn this retrospective study, we report longitudinal outcomes of children with moderate-severe UC, initiating oral/IV CS<1month of diagnosis and a minimum follow-up (FU) of 1year. CS resistance (CSR) and CS dependency (CSD) were combined as PR, and those with CS-free remission (CSFR) at 6months were GR.ResultsOf 116 children with UC, 76 (33 males) fulfilled study criteria. Median age at diagnosis was 12years (IQR 12-14), and a median FU was 48months (IQR 27-65). Thirty-five (46%, CSR=10, CSD=25) were PR, and 41 (54%) were GR. Mean relapse (2.39 vs. 1.1, p=0.0009), acute severe UC flare-up (40% vs. 9.7%, p=0.002), and colectomy rates (34.2% vs. 2.4%) were greater in PR versus GR, despite frequent early (<6months) use of azathioprine (74% vs. 27%, p=0.004) and anti-TNFs (43% vs. 2.4%, p=0.0001). Cumulative colectomy at 3years was lowest in those with GR versus CSD and CSR (2.4% vs. 28% and 50% p=0.001). On univariate analysis, CRP>20mg/L at diagnosis, Mayo Clinical Score>1 at 3months, and PR predicted colectomy. On multivariate regression, only baseline CRP>20mg/L predicted colectomy (HR 4.9, p=0.03).ConclusionsBaseline CRP and poor response to initial CS are associated with unfavorable outcomes in children with UC.