A semi-physiological framework to predict changes in pharmacokinetics of cytotoxic drugs in pregnant women
J.M. Janssen (1), J.G.C. van Hasselt (2), K. van Calsteren (3), F. Amant (3,4), J.H. Beijnen (1,5), A.D.R. Huitema (1,6), T.P.C. Dorlo (1)
(1) Department of Pharmacy & Pharmacology, Antoni van Leeuwenhoek/Netherlands Cancer Institute, Amsterdam, The Netherlands, (2) Division of Systems Biomedicine and Pharmacology, Leiden Academic Center for Drug Research, Leiden Academic Center for Drug Research, Leiden University, Leiden, The Netherlands, (3) Centre for Gynecologic Oncology Amsterdam (CGOA), Antoni van Leeuwenhoek/Netherlands Cancer Institute, Amsterdam, The Netherlands and Amsterdam University Medical Center, Amsterdam, The Netherlands, (4) Department of Oncology, Catholic University of Leuven, Leuven, Belgium, (5) Utrecht Institute for Pharmaceutical Sciences (UIPS), Utrecht University, Utrecht, The Netherlands, (6) Department of Clinical Pharmacy, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands.
Objectives:
Physiological changes during pregnancy might have an influence on pharmacokinetics (PK) and hence on the efficacy and toxicity of pharmacological treatment. Recently, it was shown that oncological treatment during pregnancy is safe and recommended.[1] Given the severity of the disease but at the same time high potential impact on both the mother and child, there is a high unmet medical need for adequate and tolerable treatment of this neglected patient population. In order to make adequate dose adjustments, it is important to assess the PK of cytotoxic drugs in pregnant patients. With this work we aimed to develop a methodology enabling the simulation of individual PK profiles of a range of cytotoxic drugs in pregnant patients and subsequently predict adequate and safe dosing regimens.
Methods:
A selection of relevant empirical equations for physiological changes from Abduljalil et al. were implemented in our semi-physiological simulation framework.[2] Firstly, the change in unbound drug concentration as a function of estimated gestational age (EGA) was implemented, using the change in albumin or alpha-1-glycoprotein levels. Since the proportion renally cleared drug for all four studied drugs is eliminated by glomerular filtration (GFR), the change in renal clearance was scaled using the change in GFR during pregnancy. To describe the change in hepatic clearance, a well-stirred liver model was used, in which intrinsic clearance was adjusted for gestational changes in enzyme activity. Reported non-pregnant volumes of distribution were scaled with a previously proposed algorithm using the gestational change in total body water, extracellular water and plasma volume.[3]
Non-linear mixed effects population models that described the PK of doxorubicin, epirubicin, docetaxel and paclitaxel in non-pregnant patients were obtained from literature.[4,5,6,7] These models were integrated with the semi-physiological alterations and drug specific characteristics. Individual concentration-time profiles were simulated in R (R package deSolve) and simulations were visually evaluated using PK data from 26, 16, 9 and 19 pregnant patients that were available for doxorubicin, epirubicin, docetaxel and paclitaxel, respectively.[1,8] Individual model fits were obtained for the observed data, by using the MAXEVAL=0 and POSTHOC options in NONMEM (v7.3).[9] The fit of the semi-physiological model was compared for the fit of the model parameters for the non-pregnant versus the pregnant state.
Results:
Typical parameters for an EGA of 28 weeks were predicted. A typical increase of 15.8%, 14.6% and 29.0% was observed for doxorubicin CL, V1 and V2, respectively and 14.2%, 14.5%, 13.7% and 39.0%, for epirubicin. For docetaxel, typical increases were predicted of 18.1%, 18.0%, 20.5% and 38.7% for CL, V1, V2 and V3, respectively. For paclitaxel, an empirical PK model including saturable distribution to the first peripheral compartment and saturable elimination was used. A typical increase of 19.8%, 15.0% and 38.4% for the maximal elimination rate, V1 and V3 was observed. The simulations showed an adequate prediction of the observed pregnant PK for all four drugs at therapeutic doses. Also, the simulations clearly demonstrated that the use of non-pregnant parameter estimates resulted in an overprediction of the observed concentrations for all four drugs. Comparison of the model fit for the individual predictions based on the semi-physiological pregnant parameter estimates versus non-pregnant parameter estimates showed a significantly improved fit for paclitaxel (ΔOFV –18.0, P=0.0004, χ2-distribution, degrees of freedom (df)=3), epirubicin (ΔOFV –148, P<0.00001, χ2-distribution, df=4) and doxorubicin (ΔOFV –62.2, P<0.00001, χ2-distribution, df=3). For docetaxel, a decrease in OFV of 4.66 points (P=0.324, χ2-distribution, df=4) was observed.
Conclusions:
The semi-physiological framework provided an adequate prediction of the PK for four cytotoxic agents of two distinct drug classes in women over varying stages of gestation. This method may therefore be used for extrapolation purposes to adjust anticancer dosing regimens in pregnant women for drugs for which PK data from pregnant women are unavailable.
References:
[1] Haan J De, Verheecke M, Calsteren K Van, Calster B Van, Shmakov RG, Gziri MM, et al. Oncological management and obstetric and neonatal outcomes for women diagnosed with cancer during pregnancy: a 20-year international cohort study of 1170 patients. Lancet Oncol. 2018;19:337–46.
[2] Abduljalil K, Furness P, Johnson TN, Rostami-Hodjegan A, Soltani H. Anatomical, Physiological and Metabolic Changes with Gestational Age during Normal Pregnancy. Clin Pharmacokinet. 2012;51:365–96.
[3] Gibaldi M, McNamara PJ. Apparent volumes of distribution and drug binding to plasma proteins and tissues. Eur J Clin Pharmacol. 1978 Jul 30;13(5):373-80.
[4] Koolen SLW, Oostendorp RL, Beijnen JH, Schellens JHM, Huitema ADR. Population pharmacokinetics of intravenously and orally administered docetaxel with or without co-administration of ritonavir in patients with advanced cancer. Br J Clin Pharmacol. 2010;69:465–74.
[5] Crombag MBS, Schultink AHMDV, Koolen SLW, Wijngaard S, Joerger M, Schellens JHM, et al. Impact of Older Age on the Exposure of Paclitaxel: a Population Pharmacokinetic Study. Pharm Res. Pharmaceutical Research; 2019;36.
[6] Sandstrom M, Lindman H, Nygren P, Johansson M, Bergh J, Karlsson M. Population analysis of the pharmacokinetics and the haematological toxicity of the fluorouracil-epirubicin-cyclophosphamide regimen in breast cancer patients. Cancer Chemother Pharmacol. 2006;58:143–56.
[7] Joerger M, Huitema ADR, Richel DJ, Dittrich C, Pavlidis N, Briasoulis E, et al. Population Pharmacokinetics and Pharmacodynamics of Doxorubicin and Cyclophosphamide in Breast Cancer Patients A Study by the EORTC-PAMM-NDDG. 2007;46:1051–68.
[8] RC Team (2009) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna.
[9] Bauer RJ. NONMEM USERS GUIDE INTRODUCTION TO NONMEM 7.3.0. ICON Development Solutions, 2014.
