P555 Failure of infliximab treatment in Crohn’s disease is associated with the presence of fibrosis
J. de Bruyn*1, J. Steenkamer2, M. Wildenberg1, S. Meijer3, W. Bemelman4, G. van den Brink1, G. D’Haens1
1Academic Medical Centre, Gastroenterology & Hepatology, Amsterdam, Netherlands, 2Academic Medical Centre, Tytgat Institute for Liver and Intestinal Research, Amsterdam, Netherlands, 3Academic Medical Centre, Pathology, Amsterdam, Netherlands, 4Academic Medical Centre, Surgery, Amsterdam, Netherlands
Intestinal fibrosis in Crohn’s disease (CD) is stimulated by chronic inflammation, and characterised by an increased presence of myofibroblasts and collagen deposition in all layers of the intestinal wall. Infliximab is a highly effective anti-inflammatory treatment in CD. Despite initial good response, a considerable proportion of patients loses response during maintenance treatment. We aimed to investigate if this phenomenon could be explained by the presence of fibrosis.
Between 2005 and 2012, patients with ileocecal CD failing thiopurine treatment were randomised to medical therapy with infliximab or ileocolonic resection. Patients not responding to at least 3 doses of infliximab also underwent subsequent resection. Demographics and preoperative C-reactive protein (CRP) were recorded. Haematoxylin and eosin (H&E) stained resection specimens were scored for severity of inflammation by an experienced IBD pathologist (validated score, range 0–13 points). Moreover, immunohistochemistry was used to stain for collagen I and III, and fibronectin. All 4 layers of the intestinal wall were assessed lamina propria, muscularis mucosa, submucosa, and muscularis propria. Staining intensity was measured as positive stained area divided by the total surface area as indicated by H&E staining.
We examined 20 specimens from patients operated after infliximab failure (median duration of treatment 38 weeks), and 20 specimens from infliximab-naive patients. CRP levels did not differ between treated and non-treated patients (respectively 17 vs 6 mg/L), indicating similar levels of systemic inflammation. Both infliximab-failure and infliximab-naive groups had considerable intestinal inflammation but this did not differ significantly (11 vs 10 points). In the infliximab-failure group, significantly more collagen I deposition was found in the submucosa (respectively 67% vs 58%, p = 0.03); in the other intestinal layers the same trend was observed. The collagen III deposition was more pronounced in the infliximab-failure patients in the submucosa and muscularis propria (60% vs 48%, p = 0.01; 43% vs 32%, p = 0.01). Most importantly, infliximab-failing patients had significant more expression of fibronectin in the lamina propria, muscularis mucosa, and submucosa (22% vs 10%, p = 0.001; 19% vs 7%, p = 0.001; and 29% vs 13%, p = 0.03).
More collagen and fibronectin deposition was observed in patients failing infliximab compared with patients naive to infliximab, predominantly in the deeper layers of the intestine. Therefore, intestinal fibrosis may be a cause for infliximab failure in patients with CD.