Völler, S. (1); Boddy, A.V. (2); Boos, J. (3), Kontny, N.E. (1); Krischke, M. (4); Würthwein, G. (4); Hempel G. (1)
(1) Westfälische Wilhelms-Universität, Institut für pharmazeutische und medizinische Chemie, Klinische Pharmazie, Corrensstraße 48, 48149 Münster, Germany, (2) Northern Institute for Cancer Research, Paul O'Gorman Building, Medical School, Framlington Place, Newcastle upon Tyne, NE2 4HH, UK, (3) Universitätsklinikum Münster, Klinik und Poliklinik für Kinderheilkunde, Pädiatrische Hämatologie und Onkologie, Funktionsbereich Pädiatrische Hämatologie und Onkologie, Albert-Schweitzer-Straße 33, 48149 Münster, Germany, (4)Universitätsklinikum Münster, Centre for Clinical Trials, ZKS Münster, Münster, Germany
Objectives: Although almost 60% of children diagnosed with cancer receive anthracyclines as part of their treatment, the knowledge on the pharmacokinetics (PK) of the drug in children is very limited. As doxorubicin was featured on the European Medicines Agency priority list for studies on off-patent paediatric medicinal products, a multicentre, multinational phase II PK study investigating a possible age-dependency in the clearance (CL) of doxorubicin in children with solid tumours and leukaemia was conducted. The data were analysed using a population PK model generated in NONMEM® 7.2.
Methods: Samples from 2 doxorubicin administrations in 101 patients treated according to the tumour-specific national or European therapeutic trial were collected with a particular focus on recruiting children less than 3 years. Data were analysed using NONMEM® 7.2, R, Xpose4 and a predefined stepwise strategy. A large number of covariates including patient characteristics, laboratory values (e.g. bilirubin, serum creatinine, alanine aminotransferase, serum albumin, haematocrit) and 17 single nucleotide polymorphisms were available in the study. Covariate modelling was performed using the stepwise covariate modelling strategy (SCM) integrated in PsN® with the linearization option. Using a linearized SCM was not possible for genetic polymorphisms as most were coded with three stages (wild type, heterozygous, homozygous). Therefore, each genetic polymorphism was tested separately for an effect on the CL of doxorubicin.
Results: A three compartment model was most suitable to characterise the PK of doxorubicin. All parameters of the population model were scaled to body surface area. The inclusion of age on the CL yielded a significant improvement of the model. No other patient-related covariate, including liver function, was found to influence the parameters of the model. Pharmacogenetic variants, including those in transporters and drug metabolizing enzymes, had no influence on pharmacokinetic parameters.
Using the mean estimated CL value for each individual, children less than 3 years had a lower CL (21.1 ± 5.8 l/h/m2) than older children (26.6 ± 6.7 l/h/m2) (p=0.0004), even after correcting for body size.
Conclusion: This study demonstrates an age-dependency in the clearance of doxorubicin in children. The results may be useful for refining dosage regimens in this patient group.
Reference: PAGE 23 (2014) Abstr 3053 [www.page-meeting.org/?abstract=3053]
Poster: Drug/Disease modeling - Oncology