Loss of response to infliximab is common amongst patients with inflammatory bowel disease and varies according to definition

Abstract

Background: Loss of response (LOR) to infliximab (IFX) is common amongst patients with inflammatory bowel disease (IBD); however, there is heterogeneity as to how LOR is defined. Roda et al. (2016) suggest that LOR may be defined according to recurrence of symptoms, or alternatively only when there is objective evidence of recurrent disease, or only when a change in biologic was instituted.1 Aims: To assess the rate of LOR to IFX amongst patients with IBD treated at a tertiary referral centre based on different definitions of LOR; and to assess how LOR is managed amongst this cohort. Methods: A retrospective audit of 169 patients with IBD treated with infliximab during January to December 2015 was performed. The study was approved by the human research ethics committee at the Canberra Hospital. Results: Seventeen patients were excluded due to inadequate records (e.g. private patients or interstate transfer), leaving 152 patients. One hundred twenty-four (81.6%) patients with Crohn's disease (CD) and 28 (18.4%) with ulcerative colitis (UC) received IFX for a mean duration of 36.3 months. • 116 (76.3%) patients initially underwent combination therapy with an immunomodulator (IM). Of these patients, 93 (80.2%) were given azathioprine, 7 (6.0%) were given 6-mercaptopurine and 16 (13.8%) were given methotrexate. The mean duration of combination therapy was 11.7 months. • Baseline characteristics for CD: ileal (25 patients, 20.2%), colonic (19 patients, 15.3%), ileocolonic (64 patients, 51.6%), perianal modifier (38 patients, 30.6%). Phenotype: inflammatory 37 (30.0%), stricturing 23 (17.7%), fistulising 7 (5.6%, excluding perianal fistulising disease). 24 patients underwent bowel resection for CD in the past (19.4%). • Extent for UC: extensive (12, 42.6%), left sided (10, 35.7%), distal disease (2, 7.1%). • Loss of response was experienced by 35 patients (23.0%). Of these patients, 21 (60.0%) experienced a partial LOR comprising of symptoms prior to each infusion which resolved after each infusion. 26 (74.3%) patients had LOR defined by objective evidence (e.g. CRP, calprotectin, and endoscopy). 4 patients (11.4%) developed LOR requiring a change in biologic. 5 patients (14.3%) developed LOR as defined by symptoms without objective evidence. • As treatment of LOR, 14 patients (40.0%) were given at least one double dose of IFX, 3 patients (8.6%) underwent ‘re-induction’, an IM was reintroduced in 6 patients (17.1%), 5 patients (14.3%) were treated with prednisolone, and 4 underwent a change to an alternative biologic (11.4%). Some patients underwent more than one intervention. • 27 patients with LOR (77.1%) had IFX trough levels and antibodies to IFX measured. The mean trough level was 1.70 μg/ml (therapeutic range 3.0–7.0 μg/ml), and the mean antibody level was 23.2 μg/ml (only tested if levels below 0.14 μg/ml). The trough level was lower in patients with partial LOR compared those with full LOR (1.66 vs. 2.23 μg/ml). Conclusions: Loss of response to infliximab in a tertiary referral centre setting is common; however, the rate of LOR varies depends upon the definition. The authors propose a novel definition of ‘type 2’ LOR referring to symptoms of LOR which resolve after each infusion. This retrospective audit provides a ‘real world’ snapshot of treatment with IFX over 12 months.