P637 High faecal and serum tumour necrosis factor-α levels are associated with non-response to infliximab in patients with ulcerative colitis

Abstract

Background: Infliximab (IFX) is an effective therapy for patients with ulcerative colitis (UC). However, a substantial number of patients do not respond to treatment (primary non-responders), especially those with severe colitis. This may be explained by a high inflammatory burden, specifically due to high quantities of mucosal TNF-α or increased drug loss from the “leaky gut”. We aimed to assess whether TNF-α levels (faecal or serum) or IFX faecal loss is associated with non-response to IFX. Methods: A prospective cohort of moderate-to-severe UC patients receiving induction therapy with IFX. For each patient, faecal and serum samples were evaluated for pre-treatment TNF-α levels using enzyme-linked immunosorbent assay. Repeated faecal and serum samples have been evaluated for IFX levels, pre-treatment and up to Week 14. Clinical response was assessed at Weeks 0, 2, 6, and 14 after treatment initiation according to Mayo or partial Mayo score. We also recruited as controls patients in remission with maintenance IFX and patients with mild UC treated with only 5-aminosalicylic acid. Results: UC patients (n = 46) were recruited: 28 with moderateto-severe UC, 10 in remission and 8 with mild disease (defined as 18 control patients). Faecal TNF-α was detected among moderateto-severe UC patients compared with controls (33.3% vs. 10% in patients in remission and 0% in mild UC, p = 0.178). Values above the upper limit of the assay (>1000 pg/ml) in faecal or serum samples were detected only in non-responders. IFX was detected in the first faecal sample collected during induction in 82% moderate-to-severe UC patients compared with none of the patients in remission. IFX non-responders (defined as the absence of clinical response at Week 14, or patients needed additional or higher IFX doses because of deteriorating symptoms) had significantly shorter disease duration compared with responders (median disease duration 18 months in non-responders [IQR = 11.5-60 months] vs. 78 months [33-165] in responders, p = 0.01). Conclusions: Extremely high pre-treatment TNF-α level in faecal or serum samples may predict risk for IFX non-response in UC. Nonresponse was associated with higher faecal loss of IFX and shorter disease duration before treatment initiation. Additional studies are warranted to examine whether intensified treatment protocol will lead to an improved outcome in these patients.