Associations Among Mucosal and Transmural Healing and Fecal Level of Calprotectin in Children With Crohn's Disease

BACKGROUND & AIMS: Bowel healing is an important goal of therapy for patients with Crohn's disease (CD). Although there have been many studies of mucosal healing, transmural healing (ie, in the bowel wall) has not been investigated in children. We analyzed data from the ImageKids study to determine associations among mucosal, transmural healing and levels of calprotectin and C-reactive protein in children with CD. METHODS: We collected data from a multi-center study designed to develop 2 magnetic resonance enterography (MRE)-based measures for children with CD (6-18 years old). In our analysis of 151 children (mean age, 14.2 +/- 2.4 years), all patients underwent MRE and a complete ileocolonoscopic evaluation; fecal levels of calprotectin and blood levels of C-reactive protein were measured. Mucosal healing was defined as simple endoscopic severity index in CD score below 3, transmural healing as an MRE visual analogue score below 20 mm, and deep healing as a combination of transmural and mucosal healing. RESULTS: We identified mucosal healing with transmural inflammation in 9 children (6%), transmural healing with mucosal inflammation in 38 children (25%), deep healing in 21 children (14%), and mucosal and transmural inflammation in 83 children (55%). The median level of calprotectin was lowest in children with deep healing (mean level, 10 mu g/g; interquartile range, 10-190 mu g/g), followed by children with either transmural or mucosal inflammation, and highest in childrenwithmucosal and transmural inflammation (810 mu g/g; interquartile range, 539-1737 mu g/g) (P <.001). Fecal level of calprotectin identified children with deep healing with an area under the receiver operating characteristic curve value of 0.93 (95% CI, 0.890.98); level of C-reactive protein identified children with deep healing with an area under the receiver operating characteristic curve value of 0.81 (95% CI, 0.71-0.9). A calprotectin cutoff value of 100 mg/g identified children with deep healing with 71% sensitivity and 92% specificity; a cutoff value of 300 mu g/g identified children with mucosal healing with 80% sensitivity and 81% specificity. CONCLUSIONS: In a prospective study of children with CD, we found that one-third have healing in only the mucosa or the bowel wall (not both). Levels of fecal calprotectin below 300 mu g/identify children with mucosal healing, but a lower cutoff value (below 100 mu g/g) is needed to identify children with deep healing.