Tocilizumab in the treatment of rheumatoid arthritis (RA) is a potential candidate for concentration-guided tapering because the standard dose of tocilizumab results in a wide range of serum concentrations, usually above the presumed therapeutic window, and an exposure-response relationship has been described. However, no clinical trials have been published to date on this subject. Therefore, the objective of this study was to assess the feasibility of the tapering of intravenous (iv) tocilizumab with the use of a pharmacokinetic model-based algorithm in RA patients. A randomized controlled trial with a double-blind design and follow-up of 24 weeks was conducted. RA patients who received the standard of tocilizumab for at least the past 24 weeks, which is 8 mg/kg every 4 weeks, were included. Patients with a tocilizumab serum concentration above 5 mg/L at trough were randomized between concentration-guided dose tapering, referred to as therapeutic drug monitoring (TDM), or the standard 8 mg/kg dose. In the TDM group, the tocilizumab dose was tapered with a recently published model-based algorithm to achieve a target concentration of 4-6 mg/L after 20 weeks of dose tapering. Twelve RA patients were included and 10 were randomized between the TDM and standard dose group. The study was feasible regarding the predefined feasibility criteria and patients had a positive attitude toward therapeutic drug monitoring. In the TDM group, the tocilizumab trough concentration within patients decreased on average by 24.5 ± 18.3 mg/L compared with a decrease of 2.8 ± 12 mg/L in the standard dose group. None of the patients in the TDM group reached the drug range of 4-6 mg/L. Instead, tocilizumab concentrations of 1.6 and 1.5 mg/L were found for the 2 patients who completed follow-up on the tapered dose. No differences in RA disease activity were observed between the 2 study groups. This study was the first to show that it is feasible to apply a dose-reduction algorithm based on a pharmacokinetic model in clinical practice. However, the current algorithm needs to be optimized before it can be applied on a larger scale.
Read full abstract