P-108 Infliximab-Associated Suppression of Cytokine Release by Mucosal Explants Predicts Success of Anti-TNF Therapy in Patients with Inflammatory Bowel Disease

Abstract

Background: Only a minority of patients commenced on anti-TNF agents achieve and maintain clinical remission, even with adequate serum drug levels and in the absence of anti-drug antibodies. The variable responsiveness of immune and other cell types to anti-TNF agents may partly explain this phenomenon. A means of identifying individuals with "responsive" mucosal immunology prior to selecting a biological agent could enhance remission rates, reduce costs and avoid safety concerns from ineffective use of anti-TNFs. Methods: We developed an in vitro assay to test responsiveness of patients' biopsies to infliximab as determined by release of pro-inflammatory cytokines. Mucosal biopsies from 45 patients with IBD (24 males/21 females; 15% less than 16 years old, 42% between 17 and 40 years old; 29 with Crohn's disease/16 with ulcerative colitis) were cultured ex-vivo in the presence or absence of infliximab. Levels of 16 cytokines were measured in culture supernatants by multiplex ELISA. Patients were characterized as being "mucosal cytokine responders (MCRs)" if 3 or more cytokines were inhibited in the presence of infliximab. Cytokine responsiveness was correlated with clinical response to anti-TNF therapy in a subgroup of 28 patients with a follow-up of 6 to 54 weeks. Patients were categorized as clinical "non-responders" to anti-TNF treatment based on the following criteria (1) need for hospitalization or surgical intervention for IBD-related problems; (2) dose or frequency escalation of anti-TNF treatment or switching to another anti- TNF regimen; (3) switching to, or adding, another drug class. Results: Amongst all mucosal samples cultured with infliximab, 44% of patients were categorized as mucosal cytokine responders (MCRs). In the cohort with clinical follow-up, 39% of patients were classified as clinical responders. Clinical responders had a greater number of cytokines inhibited in the assay compared to nonresponders (6.0 +/- 1.02 versus 1.29 +/- 0.63 respectively; P = 0.0012 by Mann-Whitney test). IL-13, IL-17A, IL-5 and IL-7 were inhibited in 70% to 90% of clinical responders versus 10% in non-responders (P < 0.001). Clinical response rates to infliximab were 83.3% for MCRs (10 out of 12) versus 6.25% for cytokine non-responders (1 out of 16) (P = 0.0014). Based on a Receiver Operating Characteristic (ROC) analysis, the mucosal cytokine assay had 91% sensitivity and 88% specificity for predicting clinical response to anti-TNF (area under the curve 0.87). Conclusions: These findings suggest that clinical responders to anti-TNF agents can be identified with high accuracy by testing mucosal biopsies in vitro in the presence of infliximab. If these results are replicated in a larger cohort, this mucosal cytokine assay could potentially be used to select patients for either starting or continuing anti-TNF therapy. Furthermore, in the future, it can be adapted to evaluate in parallel different candidate drugs and guide the delivery of personalized treatments in IBD