David Ternant, Denis Mulleman, Alexandre Aubourg, Stéphanie Willot, Thierry Lecomte, Hervé Watier, Philippe Goupille, Gilles Paintaud
CNRS UMR 7292
Objectives: Infliximab, an anti-TNF-α antibody, has profoundly modified the treatment of several inflammatory diseases but its pharmacokinetics (PK) is highly variable between patients. This variability influences the clinical response. The objective of this study was to describe infliximab pharmacokinetics using routine therapeutic drug monitoring (TDM) data and to develop a model allowing the prediction of infliximab concentrations using only dose and time since last perfusion.
Methods: Between 2005 and 2010, infliximab concentrations were monitored in 655 patients treated with infliximab in Tours university hospital. Infliximab concentrations associated with positive antibodies toward infliximab were removed from the analysis. To describe infliximab pharmacokinetics, a one-compartment PK model was used. A simplified model for TDM (TDM model) was built to devise the expected infliximab concentrations us ing only dose and time since last infusion. This model described, for each visit, a monoexponential decrease of infliximab concentrations since the last infusion. 2/3 and 1/3 patients were randomly assigned to estimation and validation groups, respectively. In the validation group, infliximab concentrations were predicted using typical PK and TDM model parameter estimates. To validate both PK and TDM models, their respective concentration predictions were compared.
Results: A total of 354 patients were eligible. They were treated for rheumatoid arthritis (RA, n=49), ankylosing spondylitis (AS, n=132), inflammatory bowel disease (IBD, n=112), both AS and IBD (AS-IBD, n=7), Psoriatic rheumatism (PR, n=44) and other diseases (OTH, n=10). Both PK and TDM models described the data satisfactorily and provided estimations of volume of distribution (Vd) and clearance (CL). Using PK analysis, median Vd, CL and half-life (T½) were 6.7 L, 0.3 2 L/day and 14 days, respectively. Using TDM model, apparent volume (Vd*) and clearance (CL*) were 5.7 L and 0.27 L/day, respectively. The two types of PK parameters were similarly influenced by covariates: Vd and Vd* were influenced by disease whereas CL and CL* were influenced by disease, weight and sex. The validation step showed that the two models provided similar concentration predictions.
Conclusions: Infliximab PK is different between SPA, AS and IBD patients. The developed TDM model is precise and accurate and may be used for TDM of infliximab.
Reference: PAGE 21 (2012) Abstr 2321 [www.page-meeting.org/?abstract=2321]
Poster: Other Drug/Disease Modelling