Mélanie L. Pastor (1), Céline M. Laffont (1), Antonin Schmitt (2), Etienne Chatelut (2), Didier Concordet (1)
(1) Université de Toulouse, INP, ENVT,31076 Toulouse, France. INRA, UMR1330, Toxalim, F-31027 Toulouse, France ; (2) Université de Toulouse, Institut Claudius-Regaud, EA4553, 20, rue du Pont-Saint-Pierre, 31052Toulouse, France.
Objectives: While administration of granulocyte-colony stimulating factor (G-CSF) is often used in patients receiving cytotoxic drugs to reduce high grade neutropenia, we need to increase our understanding of the effect of G-CSF in cancer patients. The aim of this work was to develop a semi-physiological model with a population pharmacokinetic-pharmacodynamic (PK-PD) approach in order to reach this goal.
Methods: Absolute neutrophil counts (ANC) were measured in 375 cancer patients receiving carboplatin over 2 cycles. Among them, 47 had G-CSF administration given as a daily or pegylated formulation. A model was built to describe ANC time course in all patients whatever the treatments they received. This model, developed in NONMEM 7.2, was inspired by the “Friberg’s model” [1] with two main differences: (i) the carboplatin plasma concentration was assumed to induce cell loss from the stem cell compartment by increasing apoptosis; (ii) the feedback mechanism regulating ANC was a function of the free circulating concentration of G-CSF (endogenous and exogenous when applicable), which was modeled using a previous compartmental model from Krzyzanski et al. [2] with linear (renal) and non-linear elimination (receptor binding to neutrophils and internalization). A modification of their model was made in order to improve data fitting, in line with the observations made by Quartino et al. [3]. As in physiological conditions, free G-CSF could stimulate stem cell proliferation and accelerate maturation time, with two different maximum effects and potency parameters.
Results: Our population PK-PD model well described neutrophil time course after carboplatin administration in all patients as assessed by Visual Predictive Checks and computation of the percentage of patients undergoing high grade neutropenia. High grade neutropenia was much more frequent in patients who did not receive G-CSF. Simulations performed with the model shed light on the differences between G-CSF formulations and dosage regimens (time and frequency of administration) and the consequences on their effects.
Conclusions: The population PK-PD model described here allows a good description of neutrophil time course in 375 patients with various tumors treated with carboplatin, whether or not they received G-CSF. This model appears more physiological than the traditional ones and gives new directions to increase our understanding of the effect of G-CSF in cancer patients.
References:
[1] Friberg LE, et al. Model of chemotherapy-induced myelosuppression with parameter consistency across drugs. J Clin Oncol 2002. 20:4713-21.
[2] Krzyzanski W, et al. Population modeling of filgrastim PK-PD in healthy adults following intravenous and subcutaneous administrations. J Clin Pharmacol 2010. 50:101S-112S.
[3] Quartino LQ, et al. An integrated G-CSF-myelosuppression model characterizing the target mediated disposition of endogenous G-CSF in breast cancer patients following chemotherapy. PAGE 20 2011, Abstr 2255.
Reference: PAGE 21 (2012) Abstr 2511 [www.page-meeting.org/?abstract=2511]
Poster: Oncology