Katharina Och (1), Jörg T. Bittenbring (2), Thorsten Lehr (1)
1) Clinical Pharmacy, Saarland University, 66123 Saarbruecken, Germany, (2) Internal Medicine I, Saarland University Hospital, 66421 Homburg, Germany
Objectives: Development of a pharmacokinetic/pharmacodynamic (PK/PD) model of Cyclosporine A (CyA) in patients after hematopoietic stem cell transplantation (HSCT) with regard to its nephrotoxic potential.
Methods: Data derived from therapeutic drug monitoring (TDM) at the department of Internal Medicine I at the Saarland University Hospital (UKS) was used for model building. All subjects received HSCT due to different underlying malignancies with a majority of acute myeloid leukemia (AML; 72%). CyA was administered once daily as a 24-hour infusion from the day prior to HSCT (day -1). Later on, patients received CyA as an oral administration twice daily. Plasma concentrations of CyA and serum creatinine were measured once daily. For the pharmacokinetic model of CyA, an evaluation of different models from literature was done. Whereas the pharmacodynamics were built by linking the plasma concentration to creatinine clearance as marker for glomerular filtration rate (GFR). Data were analyzed by non-linear-mixed-effects modelling (NLME) using the software NONMEM® (version 7.3.0) and the graphical interface Pirana (version 2.9.7). Moreover, statistical evaluation was performed within the software R (version 3.4.1) and its graphical interface RStudio (version 1.0.143).
Results: The dataset included 32 subjects with 862 observations presenting plasma concentrations of CyA and 1046 observations of creatinine clearance. Pharmacokinetics were described properly by a two-compartmental model with lag time [1]. Its stochastical submodel consisted of a proportional error model and interindividual variability on all used parameters (clearance, central volume of distribution, peripheral volume of distribution, intercompartmental clearance, rate of absorption and bioavailability). In order to describe changes in GFR it was necessary to treat GFR as a formation product with synthesis rate and degradation rate. A turnover model with zero order synthesis rate (ksyn) and first order degradation rate (kdeg) fitted best. Here, ksyn (0.09 (ml/min) *h-1) was independent of kdeg (0.0015 h-1). The concentration of CyA had a significant impact on the inhibition of GFR synthesis rate. This influence was described very well as an Emax model with great Hill coefficient (Hill = 40) representing a steep concentration-effect relationship. The stochastical submodel included an interindividual variability on ksyn and an additive error as residual error. Different attributes (e.g. age, gender, underlying disease, conditioning regime) were tested as covariates but none of them had a significant impact.
Conclusions: The pharmacokinetic/pharmacodynamic model represented the quantitative relationship between Cyclosporine A exposure and its nephrotoxicity. Considering the steep concentration-effect relationship an on-off switch of the nephrotoxic effect of CyA can be assumed. While the model fits the data further research has to be done in order to improve and validate the model in a lager cohort. A valid PK/PD model could be used as a tool to investigate the nephrotoxic potential of CyA or even as a clinician’s tool to monitor the nephrotoxic risk in patients under treatment with CyA.
References:
[1] Okada A, Ushigome H, Kanamori M, Morikochi A, Kasai H, Kosaka T, Kokuhu T, Nishimura A, Shibata N, Fukushima K, Yoshimura N, Sugioka N. Population pharmacokinetics of cyclosporine A in japanese renal transplant patients: comprehensive analysis in a single center. Eur J Clin Pharmacol (2017) 73: 1111-1119.
Reference: PAGE 27 (2018) Abstr 8615 [www.page-meeting.org/?abstract=8615]
Poster: Drug/Disease Modelling - Oncology