Therapeutic Drug Monitoring for Dose Optimization of Infliximab in Patients With Inflammatory Bowel Disease: An Analysis of Canadian Real-World Data.

Abstract

Background: Although it is generally believed that infliximab dose optimization in patients with inflammatory bowel disease with low serum infliximab concentration at trough results in increased treatment persistence, empirical data to support this notion are lacking. This study evaluated the association of infliximab therapeutic drug monitoring (TDM) and TDM-associated dose optimization with persistence in real-world practice. Methods: Data from adults with Crohn's disease (CD) or ulcerative colitis (UC) who participated in a national patient support program (PSP) in Canada were analyzed. Participants who had a first TDM evaluation (with a recorded infliximab trough concentration) in the maintenance phase of treatment were assessed (excluding those who underwent prior dose optimization). Persistence was evaluated using time-dependent Cox proportional hazards models. Results: In the overall population of patients with CD or UC, TDM was not associated with longer persistence (n = 13,203). In patients with no prior dose optimization (n = 2729) who had a serum infliximab concentration of < 3 μg/mL, dose optimization within 9 weeks of TDM was associated with significantly longer persistence (HR: 0.36; 95% CI: 0.26, 0.50 for CD [n = 711] and HR: 0.30, 95% CI: 0.21, 0.43 for UC [n = 501]). Sensitivity analyses yielded similar results when using a threshold concentration of < 5 μg/mL. In an analysis excluding patients who received no further treatment after TDM, the association between dose optimization and longer persistence was not confirmed in patients with CD, and mostly confirmed in patients with UC at a threshold concentration of < 3 μg/mL. Conclusion: TDM-associated dose optimization in patients with UC with low serum infliximab concentrations was associated with longer persistence. This association was not confirmed in patients with CD. This study demonstrated that real-world data from a PSP-generated cohort can be evaluated to inform clinical practice and that this approach may be complementary to other types of cohort studies.