Q. Chalret du Rieu (1, 2), S. Fouliard (2), E. Chatelut (1), M. Chenel (2)
(1) Institut Claudius Regaud, Toulouse, France; (2) Clinical Pharmacokinetics, Institut de Recherches Internationales Servier, Suresnes, France
Objectives: To develop a semi-mechanistic PKPD model of thrombocytopenia, taking into account differences within patients, treated with abexinostat for either solid tumors or various types of lymphoma, notably the effect of disease progression on circulating platelet levels in patients with liquid tumors.
Methods: 925 platelet samples from the first 3 cycles of abexinostat treatment coming from 95 patients (i.e. 49 [465 platelet samples] and 46 [460 platelet samples] suffering from solid tumor and lymphoma, respectively) included in 4 different phase I clinical studies, were analyzed. A sequential PKPD modelling approach [1] was performed in NONMEM 7.2, with FOCEI. Individual Empirical Bayesian Estimates of PK parameters (from a previous validated PK model) obtained using the POSTHOC method were input into a PKPD model of platelet dynamics, based on original Friberg et al's one [2;3]. Improvements concerning disease progression were tested on BASE parameter (platelet count at inclusion). The Likelihood-Ratio-Test allowed the discrimination between hierarchical models. Model evaluation was based on standard goodness-of-fit plots, parameter precisions of estimation and Normalized Prediction Distribution Error (NPDE) [4;5].
Results: The original semi-mechanistic myelosuppression structural model was refined [2;3]. A feedback mechanism was added in the mean maturation time in order to quicken the production of circulating platelets by bone marrow in cases of thrombocytopenia and vice versa. Each types of patients were distinguished by estimating two different BASE parameters. A slight decrease in the platelet count over the time in lymphoma patients was modelled by adding an Imax disease progression effect on their BASE parameter, mimicking a pathological effect. Model evaluation by individual fit analysis, goodness of fit plots and NPDE vs TIME graphs showed that this model could fully describe and predict available platelets count in both populations.
Conclusion: A semi-mechanistic PKPD model able to predict toxicity in patients treated by abexinostat, particularly over the time by taking into account physiological or pathological characteristics of the bone marrow according to the patients' disease, was described. Consequently with such a model including disease progression, an optimal administration schedule could be determined particularly in lymphoma patients, for who previous predictions were not as good as for patients with solid tumors.
References:
[1] Zhang L, Beal SL, Sheiner LB. Simultaneous vs. sequential analysis for population PK/PD data I: best-case performance. J Pharmacokinet Pharmacodyn 2003 Dec;30(6):387-404.
[2] Friberg LE, Freijs A, Sandstrom M, Karlsson MO. Semiphysiological model for the time course of leukocytes after varying schedules of 5-fluorouracil in rats. J Pharmacol Exp Ther 2000 Nov;295(2):734-40.
[3] Friberg LE, Henningsson A, Maas H, Nguyen L, Karlsson MO. Model of chemotherapy-induced myelosuppression with parameter consistency across drugs. J Clin Oncol 2002 Dec 15;20(24):4713-21.
[4] Brendel K, Comets E, Laffont C, Laveille C, Mentre F. Metrics for external model evaluation with an application to the population pharmacokinetics of gliclazide. Pharm Res 2006 Sep;23(9):2036-49.
[5] Brendel K, Comets E, Laffont C, Mentre F. Evaluation of different tests based on observations for external model evaluation of population analyses. J Pharmacokinet Pharmacodyn 2010 Feb;37(1):49-65.
Reference: PAGE 22 () Abstr 2827 [www.page-meeting.org/?abstract=2827]
Poster: Oncology