Pharmacokinetic modelling of ifosfamide administered by continuous infusion on 5 days at the dose of 6 G/M-2

Background. We aimed to create a model for Ifosfamide (IFX) pharmacokinetic s for drug monitoring in order to improve protocol dose intensity. Material and Methods. We studied ifosfamide pharmacokinetics in 12 patients aged 8 - 19 years. Sixteen courses were modelled (6 g/m(2) on 5 days). The auto-in duction of ifosfamide was taken into account in the model. Ifosfamide measu rement was performed on serum samples by gas chromatography with thermo-ion ic detection. Two pharmacokinetic models were compared. The following param eters were estimated: volume of distribution (Vd), clearance at the beginni ng of the induction (CLi), clearance extrapolated to infinity (CLf), cleara nce at the end of infusion (CL120), a rate constant (Kc) indicating the cle arance variation with time and the lag rime (Lag) indicating the rime elaps ed between the start of infusion and the start of induction. The Wilcoxon t est was used to investigate possible differences between models. We tested the hypothesis that Boddy's model is an acceptable simplification of Levy's model. Results: Four of six parameters were significantly different betwee n the two models (p = 0.05). The best curve fitting was obtained using the Levy's model which provided the following estimates, Cli = 2.46 +/- 0.94 L. h(-1).m(-2), CLf = 5.22 +/- 1.02 L.h(-1).m(-2), Kc = 0.024 +/- 0.014 h(-1), Vd = 18.84 +/- 5.04 L and Lag = 4.86 +/- 6.61 h. The most important differ ence is found for the distribution volume. Conclusion: Levy's model is more accurate and takes into account the integration of clearance.