PWE-260 Optimal C reactive protein cut-off point for predicting hospitalisation in patients with moderately active Crohn's disease

Abstract

Aim: To identify high risk patients among patients with moderate Crohn's disease (CD), we explored the association between C-reactive protein (CRP) concentration and hospitalization risk for patients with moderately active CD and identified the optimal CRP cutoff point as a marker to predict CD-related hospitalization. CRP is a well-studied and commonly used laboratory marker of inflammation in CD [1]. The relationship between CRP and hospitalization risk given the same Crohn's Disease Activity Impairment (CDAI) score in patients with moderate CD has not been studied. Material(s) and Method(s): Data from CHARM, a 56-week, randomized, placebo-controlled trial of adalimumab maintenance therapy, were analyzed. All patients received adalimumab during a 4-week, open-label induction period; patients were then randomized to adalimumab or placebo for a 52-week double-blind period. For this analysis, only patients who were randomized to placebo at Week 4 and had moderate CD, defined as CDAI +/-300 at Week 4, were analyzed. A Cox model was applied to analyze the association between Week-4 CRP concentration and the probability of having a CD-related hospitalization during the 52-week double-blind period. Week-4 CDAI score, Week-4 steroid use, age, sex, weight, body mass index, and prior anti-tumor necrosis factor use were also adjusted in the model. Patients were censored if they switched to open-label adalimumab or dropped out. A receiver operating characteristic (ROC) curve was used to identify the optimal CRP cutoff point to best predict the 52-week CD-related hospitalization rate. Result(s): The analysis population included 214 patients randomized to placebo with Week-4 CDAI+/-300. An elevated Week-4 CRP concentration was associated with a greater chance of CD-related hospitalization (hazard ratio = 1.24; p = 0.002). The ROC curve identified a CRP concentration of 1.41 mg/dL as the dichotomizing point (area under the curve = 0.68; sensitivity = 0.58; specificity = 0.80). Risk of CD-related hospitalization during the double-blind period was 3.4 times greater for patients with CRP concentrations +/-1.41 mg/dL at Week 4 vs. patients with a CRP concentrations <1.41 mg/dL (p = 0.015), with control for CDAI and other covariates. Conclusion(s): Early CRP concentration represents a moderate to good marker to predict CD-related hospitalization for patients with moderately active CD given the same CDAI score. CRP concentration of 1.41 mg/dL was the optimal cutoff point for predicting long-term CD-related hospitalization.