P426 In failure of non-optimised adalimumab (ADA) with therapeutic serum levels, a change in biotherapy class is greater than an intensification of ADA dose in IBD patients

Background Nearly 40% of patients who failed on ADA have therapeutic residual trough levels (pharmacodynamic failure). It is then recommended to change the therapeutic class to a non-anti-TNF biological. In this situation, however, a quarter of patients respond to an ADA dose optimisation. The purpose of this work was to compare these two strategies and identify predictors of response for each strategy. Methods This is a bi-centre, retrospective study based on a clinical and biological database that included patients followed from 2015 to 2016 with an IBD in maintenance treatment with ADA as monotherapy (40 mg/sc/14 days) and in loss of response (CDAI > 220 with faecal calprotectin > 250 µg/g stool for CD and Mayo score > 6 with endoscopic sub-score > 1 for UC). All patients included had residual trough ADA levels > 5 µg/ml. Relapsed patients were treated either with dose optimisation (ADA: 40 mg/sc/7 days) or other biotherapy (ustekinumab or vedolizumab, according to classic regimens and dosages). The primary endpoint was the relapse-free survival rate (defined as clinical and biological or endoscopic activity requiring therapeutic change). The survival curve analysis was performed using the Kaplan–Meier method and the log-rank test. Results In total, 125 patients (mean age: 38 years, 75% CD, sex ratio M/F = 57%, previous infliximab = 42%) were included and 50 were treated with another biological (Vedolizumab: 30 patients, Ustekinumab: 20 patients). The characteristics of the two groups were comparable except for significantly higher levels of CRP and faecal calprotectin at inclusion in the biological change arm. The duration without relapse was significantly longer after the class change (p = 0.0005; HR = 0.39[0.23–0.68]). The relapse-free survival rates at 54 weeks were 35% for the optimised arm and 66% for a class change (p = 0.04). In multivariate analysis, the only long-term response predictor was biotherapy class change vs. ADA dose optimisation (p = 0.042; HR = 0.3[0.02–0.95]). Previous use of IFX or type of IBD were not associated with relapse. For the dose optimisation strategy, in multivariate analysis, only age < 40 years was associated with a progression without relapse (p = 0.007; HR = 0.09[0.01–0.52]). Residual ADA levels were not predictive of response to optimisation (p = 0.73) and no ADA quartile was significantly associated with a more favourable clinical response. For the class change strategy, no relapse predictors were isolated in univariate analysis. Conclusion In patients on ADA who are losing response despite therapeutic rates, the change of class to non-anti-TNF biotherapy allows a significantly greater lasting response than a dose optimisation strategy for ADA.