Search in the Abstract Database

Search Abstracts 2015

* = Presenting author

P482 Infliximab population pharmacokinetic modelling in patients with Inflammatory Bowel Disease: estimation of individual pharmacokinetic parameters and trough levels prediction

J. Guardiola*1, L. Rodriguez-Alonso1, E. Santacana2, N. Padullés2, H. Colom3, A. Padullés2, A. Ruiz-Cerulla1, C. Arajol1, S. Cobo2, J. Bas4, J. Climent4, F. Morandeira4, F. Rodriguez-Moranta1

1Hospital Universitario de Bellvitge, Gastroenterology, L´Hospitalet de Llobregat, Spain, 2Hospital Universitario de Bellvitge, Pharmacy, L´Hospitalet de Llobregat, Spain, 3Barcelona University, Pharmacokinetics. School of Pharmacy, Barcelona, Spain, 4Hospital Universitario de Bellvitge, Immunology, L´Hospitalet de Llobregat, Spain

Background

Infliximab (IFX) trough levels (TLI) exhibit a high interinviduals variability in inflammatory bowel disease (IBD) patients. This variability is relevant because there is a relationship between TLI and clinical efficacy of IFX in both Crohn Disease (CD) and ulcerative colitis (UC). Two-compartment population pharmacokinetic models (PopPK) using data from pivotal trials have been developed in CD and UC. The main objective of the study was to estimate individual pharmacokinetic (PK) parameters and predict TLI

Methods

Patient data were collected prospectively from patients treated with IFX between July 2013 to March 2014. Blood samples were collected before intravenous infusion at steady state. TLI and antibodies toward infliximab (ATI) were measured by enzyme-linked immunosorbent assay (ELISA) (Promonitor®). Covariates included in final popPK models, C reactive protein (CRP) and smoking were recorded. Individual PK parameters were estimated on the basis of previously developed popPK models and TLI, Cpeak and AUC were individually predicted using the Bayesian posthoc option. Pharmacokinetic analysis was carried out using Nonmem® 7.2

Results

64 patients (39 CD 25 UC). 132 TLI and ATI were measured. 132 TLI and ATI were measured. Mean TLI was 317 mg/L (54%<3 mg/L); of the 113 samples taken in patients who were negative for ATI, 4.4% had undetectable TLI, whereas of the 19 samples taken in patients positive for ATI, 84.2% had undetectable TLI. Mean predicted TLI (IPRED) was: 2.92 mg/L. Mean estimated peak levels and AUC were 115,53 mg/L and 26447.41 mg/h/L, respectively.The comparison between IPRED and observed TLI values using the PopPK model for CD led to a non-significant overall mean relative bias of -3.99% and an acceptable precision of 18.96%. IPRED correlated with observed TLI (coefficient 0.985, p<0.0001). No correlation was found between IPRED and observed TLI using the PopPK model for UC. Individual (Bayesian estimates) PK parameters (mean) were: central clearance (Cl) 7.68 ml/kg/day, volume of distribution (central) (Vd) 50.61 ml/kg, half-life (t1/2): 11.41 days

Conclusion

PopPK model for CD proposed by Fasanmade AA et al (1) adequatly predicts TLI and estimates individual PK parameters in both CD and UC patients. The model may be useful to individualize doses according to patient factors influencing IFX PK

1. Fasanmade AA et al. 2011.

2. Fasanmade AA, et al. 2009