PK/PD model extension to characterise bone marrow exhaustion in cancer patient making use of a prior paclitaxel PK model
Andrea Henrich (1,2), Markus Joerger (3), Stefanie Kraff (4), Ulrich Jaehde (4), Wilhelm Huisinga (5), Charlotte Kloft (1), Zinnia P. Parra-Guillen (1)
(1) Dept. Clinical Pharmacy & Biochemistry, Institute of Pharmacy, Freie Universitaet Berlin, Germany, (2) and Graduate Research Training program PharMetrX, Germany, (3) Dept. of Oncology & Haematology, Cantonal Hospital, St Gallen, Switzerland, (4) Dept. Clinical Pharmacy, Institute of Pharmacy, Universitaet Bonn, Germany and (5) Institute of Mathematics, Universitaet Potsdam, Germany
Objectives: Due to complex, non-linear pharmacokinetics (PK) including high inter-individual variability and severe toxicity, especially neutropenia (pharmacodynamics (PD)) of paclitaxel (PTX), dose-individualisation and therapeutic drug monitoring are indicated. A population PK/PD model  was externally evaluated using data from the CEPAC-TDM trial . Worsening neutropenia over repeated chemotherapy treatment cycles, we hypothesised this process could be due to bone marrow exhaustion (BME). The aim of this work was to refine our PK model and implement BME to describe neutrophil concentrations in cancer patients over several cycles.
Methods: Patients (n = 183) received PTX (doses adjusted according to a published algorithm ) in combination with a carbo- or cisplatin every 3 weeks for ≤ 6 cycles. PTX plasma concentrations were measured approx. 24 h after PTX administration, while neutrophil concentrations were obtained on day 1 and 15 ± 2 of each cycle. A stepwise analysis was performed: First, the prior information from the published PK model was utilised applying the frequentist approach for re-estimating the PK parameters. Second, to implement BME in a mechanistic way, an additional compartment was added to the Friberg model  accounting for slowly proliferating stem cells, while the proliferation compartment mimicked the rapidly dividing progenitor cells. Therefore, both cell types were replicating with different proliferation rate constants, but were influenced by the same drug effect and feedback mechanism. NONMEM 7.3, PsN 4.4 and Xpose4 4.5.3 were used.
Results: Only Km and bilirubin on Vmax of the saturable elimination of the re-estimated fixed-effects PK parameters, were outside the 90% confidence interval reported for the original PK model. However, the visual predictive check indicated a better PTX prediction than with the original PK model. The optimised PK/PD model was able to describe the observed BME pattern and showed high parameter precision (RSE < 10% for fixed- and < 20% for random-effects parameters). The proliferation rate constant of the progenitor cells was 3.67-fold higher than the one of the stem cells.
Conclusions: Using the frequentist approach previous knowledge from the former rich data were successfully combined with the sparse CEPAC-TDM study data, thus enabling an adequate description of PTX PK. A pathophysiologically plausible PK/PD model was developed to describe the hypothesised bone marrow exhaustion.
 M. Joerger, S. Kraff, A.D.R. Huitema, et al. Evaluation of a pharmacology-driven dosing algorithm of 3-weekly paclitaxel using therapeutic drug monitoring: a pharmacokinetic-pharmacodynamic simulation study. Clin Pharmacokinet (2012) 51(9): 607–17.
 A. Henrich, M. Joerger, W. Huisinga, et al. External evaluation of a PK/PD model describing the time course of paclitaxel and neutropenia in patients with advanced non-small cell lung cancer. PAGE 24 (2015) Abstr. 3460 [www.page-meeting.org/?abstract=3460].
 L.E. Friberg, A. Henningsson, H. Maas, et al. Model of chemotherapy-induced myelosuppression with parameter consistency across drugs. J. Clin. Oncol. (2002) 20(24): 4713–21.