P798 Understanding anti-TNF treatment failure: mechanisms and management of loss of response to anti-TNF therapy, three-year data from the PANTS study

Abstract

PANTS (Personalised Anti-TNF Therapy in Crohn’s disease) is a prospective observational UK-wide study investigating the effectiveness of infliximab (IFX) and adalimumab (ADL) in active luminal Crohn’s disease. Here we report remission rates through three years of follow up; rates of and the factors that predicted loss of response (LOR); and the effect of dose intensification on drug persistence. Remission was defined as CRP of ≤3 mg/L and HBI of ≤4 points (sPCDAI ≤15 in children), without corticosteroid therapy or exit for treatment failure. LOR was defined in patients who initially responded to anti-TNF therapy at week 14 and continued in the PANTS extension, by symptomatic IBD activity that warranted an escalation of corticosteroids, immunomodulatory or anti-TNF therapy, resectional surgery, or exit due to adverse events. The effect of dose intensification at the time of LOR on drug persistence was stratified by optimal anti-TNF drug level (IFX 7, ADL 12) (mg/L) and presence of antibodies (anti-IFX 9, anti-ADL 6) (AU/mL), measured using IDKmonitor® ELISA assays. Between March 2013 and July 2016, 1610 patients were included in PANTS: 358/955 (37%) treated with IFX (221/358 [62%] originator and 137/358 [38%] with biosimilar) and 187/655 (29%) treated with ADL, who had not exited for treatment failure at the end of year 1, opted to enrol in the extension. Overall, 41.8%, 39.4%, and 41.3% of IFX- and 38.2%, 39.7%, and 39.2% of ADL-treated patients were in remission at year 1, 2 and 3, respectively. For both drugs, we observed a dose response association between week 14 drug concentration and remission status at the later timepoints (A). 34.3%, 54.0%, and 60.3% of IFX-, and 31.9%, 47.3%, and 69.0% of ADL-treated patients lost response at year 1, 2 and 3, respectively. Multivariable regression analyses showed that low drug concentration was the major independent risk factor associated with LOR for both drugs (B). In patients who developed immunogenicity, factors associated with subsequent drug clearance were lower week 14 drug level (p<0.01), non-immunomodulator use (p=0.04), and obesity (p<0.01), but not carriage of HLA-DQA1*05 (p<0.34). In the setting of LOR (n = 686 episodes), 42% patients had their anti-TNF dose intensified and 29% exited the study without further action (C). Compared to patients who experienced LOR with optimal drug levels, dose intensification was associated with lower rates of drug persistence in patients with immunogenic-pharmacokinetic and non-immunogenic pharmacokinetic treatment failure in patients treated with IFX, but not ADL (D). Long term remission was observed in about 40% of patients treated with IFX and ADL. Loss of response and non-remission was predicted by low drug levels.