Long-Term Outcome of Patients with Ulcerative Colitis and Primary Non-response to Infliximab

Background and Aims: We studied the long-term outcome of patients with ulcerative colitis [UC] and primary non response [PNR] to infliximab and searched for predictors of colectomy in these patients. Methods: This retrospective, multi-centre study included UC patients from three European referral centres, with PNR to infliximab defined as a lack of clinical improvement after the induction therapy, leading to drug discontinuation. Relapse, for patients who continued on biologicals after PNR to infliximab, was defined as drug discontinuation for PNR, loss of response, or serious adverse event. Serum infliximab concentrations at Weeks 2 and 6 were evaluated using an enzyme-linked immunosorbent assay [ELISA] developed in house. Results: The study population consisted of 99 anti-tumour necrosis factor [TNF]-naive patients with UC and PNR to infliximab. At the end of follow-up (median: 3.2 [interquartile range 1-6.3] years), 55 [55.6%] of these patients underwent colectomy. Multiple Cox regression analysis identified acute severe UC (hazard ratio [HR]: 24; 95% confidence interval [CI]: 2.5-231; p = 0.006], baseline C-reactive protein [CRP] > 5 mg/l [HR: 11; 95% CI: 2.1-58.8; p = 0.005], baseline albumin < 40 g/l [HR: 9.5; 95% CI: 1.3-71.4; p = 0.026], and infliximab concentration at Week 2 < 16.5 mu g/ml [HR: 5.6; 95% CI: 1.1-27.8; p = 0.034] as independent predictors of colectomy. Regarding patients who continued on biologicals after PNR to infliximab, there was a marginally higher cumulative probability for relapse in patients switching to another anti-TNF agent compared with those swapping to vedolizumab [p logrank = 0.08]. Conclusions: About half of UC patients with PNR to infliximab will undergo colectomy. Patients with severe inflammation and low serum infliximab concetrations during the induction phase are at greatest risk.