Search in the Abstract Database

Abstracts Search 2019

P161 Faecal calprotectin identifies microscopic inflammation in ulcerative colitis patients with complete endoscopic healing: a post-hoc analysis of the MOMENTUM trial

T. W. Stevens*1, K. Gecse1, K. Barrett2, J. R. Turner3, G. de Hertogh4, D. T. Rubin5, G. R. D’Haens1

1Amsterdam University Medical Centres, Department of Gastroenterology and Hepatology, Amsterdam, The Netherlands, 2Shire, Basingstoke, UK, 3Brigham and Women's Hospital and Harvard Medical School, Department of Pathology, Boston, USA, 4KU Leuven, Leuven, Belgium, 5University of Chicago Medicine, Inflammatory Bowel Disease Centre, Chicago, IL, USA

Background

Histological inflammation is associated with clinical relapse in ulcerative colitis (UC). Faecal calprotectin (FC) is a marker of mucosal inflammation. The aims were to assess (i) diagnostic accuracy of FC for histological inflammation and (ii) develop a prediction model for histological remission at 1 year.

Methods

The phase IV MOMENTUM trial (ClinicalTrials.gov Identifier: NCT01124149) evaluated the efficacy of multi-matrix mesalamine in mild-to-moderate UC. In this post-hoc analysis, endoscopic and histological outcomes were assessed at Week 8 (W8) (N = 604) and 52 (W52) (N = 355). Endoscopic healing and complete endoscopic healing were defined as endoscopy score ≤1 and 0, respectively. The Geboes histopathology index was transformed to an ordinal score (range 0–13). To evaluate the correlation between FC and histology, parameters related to chronic inflammation (Geboes < 2B.1) were scored as 0. Histological remission was defined as a Geboes score < 2B.1 (absence of neutrophils in epithelium and lamina propria) resulting in a drop in the ordinal score from >0 to 0. Receiver-operating characteristic (ROC) curves were used to determine diagnostic accuracy and optimal FC cut-off values. Multi-variable logistic regression was performed using predefined predictors.

Results

Median FC values were lower in patients achieving predefined outcomes compared with patients who did not (Figure 1). Interobserver agreement between both pathologists (GDH and JHT) was moderate (κ = 0.6, 95% CI 0.33–0.87). Area under the ROC curve (AUC) value for endoscopic healing and histological remission (HR) were 0.77 and 0.76 at W8 and 0.79 and 0.80 at W52, respectively. Optimal cut-off value for HR was 75 µg/g (sens 0.65; spec 0.77; PPV 71%; NPV 71%) at W8 and 99 µg/g (sens 0.77; spec 0.75; PPV 84%; NPV 66%) at W52. In the subpopulation with endoscopy score 0, median FC remained lower in patients with HR compared with ongoing microscopic inflammation at W8 (30 vs. 140 µg/g, AUC 0.72) and W52 (21.5 vs. 134.5 µg/g, AUC 0.71). Optimal FC cut-off value was 73 µg/g at W8 and 76 µg/g at W52. Final prediction model for W52 HR comprised endoscopic score (W8) (OR 0.52, 95% CI 0.32–0.82), FC concentration (W8) (OR 1.00, 95% CI 1.00–1.00), and histological activity at baseline (OR 0.92, 95% CI 0.86–0.98) and W8 (OR 0.89, 95% CI 0.81–0.97).

Conclusion

Even in the presence of complete endoscopic healing, FC may discriminate patients with microscopic inflammation from patients in HR. Optimal cut-off lies between 75 and 100 µg/g.

Figure 1