Population Pharmacokinetic Study of Fotemustine
Y Bourhis (1), A Iliadis (2), C Lucas (3), AB Depouilly (3), C Laveille (3), B Tranchand (1)
(1) Centre Léon-Bérard, Lyon, (2) Faculté de Pharmacie, Marseille, (3) I.R.I.S., Paris
The main purpose of this study was to perform the pharmacokinetic modelling of fotemustine (MUPHORANâ) and to improve the pharmacokinetic and variability parameters estimation by introducing covariates into the model.
84 patients with various types of cancer (non Hodgkin lymphoma n=6, renal cell carcinoma n = 5, malignant glioma n=27, hepatocarcinoma n=2, colocarcinoma n=11, malignant melanoma n=13, myeloma n=20) were included in this multicentre study. Each patient was treated with one or two courses of intravenous fotemustine, given as a one hour constant-rate infusion (a total of 109 courses were administered). Depending on the tumour type, patients received either 100 mg/m²/week (conventional dose regimen) or 300-500 mg/m²/day for two consecutive days (high dose regimen).
2 to 13 blood samples per patient were collected at each course, and fotemustine was assayed in plasma by high performance liquid chromatography (limit of quantitation < 20 ng/ml).
The pharmacokinetic analysis was performed using NONMEM version V under VisualNM. The candidate covariates were age ([19;76] years), sex (54 males, 22 females, 8 unknown), weight ([46;95] kg), height ([149;190] cm), creatininemia ([26;112] µmol.l-1), serum urea nitrogen ([0.2;1.4] µmol.l-1). The tumor type was also considered.
We tested 2- and 3-compartment linear models, with additive, multiplicative, and mixed error model, using the First Order method. The inter-occasion variability was also taken into account.
The pharmacokinetics of fotemustine was best described by a 2-compartment linear model (ADVAN3 TRANS4) with a mixed error model. The overall mean clearance and volume of distribution values were estimated at 85.4 l.h-1 and 21.5 l, respectively with an important inter-individual variability (82% and 78% respectively). The introduction of the statistically significant covariates decreased the variability of the clearance from 82 to 44%, and of the volume of distribution from 78 to 46%. The inter-occasion variability was 27% for the clearance and 43% for the volume of distribution.
In this way, the clearance of fotemustine could be expressed as a function of sex, age, creatininemia, serum urea nitrogen, and tumor type. Volume of distribution was expressed as a function of age, serum urea nitrogen and tumor type.
This model will be used to analyse the plasma concentrations of children receiving high doses fotemustine in an on-going phase I/II study and to compare the pharmacokinetic parameters of this paediatric population with those of adults.