A mechanism-based population pharmacokinetic model for characterizing time-dependent pharmacokinetics of midostaurin and its metabolites in human subjects

Clin Pharmacokinet. 2008;47(12):807-16. doi: 10.2165/0003088-200847120-00005.


Background and objective: Midostaurin, a novel potent inhibitor of protein kinase C enzyme and class III receptor tyrosine kinases, including Fms-like tyrosine kinase-3 (FLT3) and c-KIT, shows time-dependent pharmacokinetics in human subjects, presumably due to enzyme auto-induction. The purpose of this study was to develop a mechanism-based population pharmacokinetic model to describe the plasma concentration profiles of midostaurin and its metabolites and to characterize the time course of auto-induction.

Subjects and methods: Data from 37 diabetic patients who received oral doses of midostaurin (25 mg twice daily, 50 mg twice daily or 75 mg twice daily) for 28 days were analysed using nonlinear mixed-effects modelling. The structural model included a gut compartment for drug input and central and peripheral compartments for midostaurin, with drug output from the central compartment to either of two compartments for the midostaurin metabolites CGP62221 and CGP52421. Different enzyme induction sub-models were evaluated to account for the observed time-dependent decrease in midostaurin concentrations.

Results: An enzyme turnover model, with CGP62221 formation (CL(1)) being a linear process but CGP52421 formation (CL(2)) being inducible, was found to be most appropriate. In the pre-induced state, CL(1) and CL(2) of midostaurin were determined to be 1.47 L/h and 0.501 L/h, respectively. At the end of 28 days of dosing, CL(2) was increased by 5.2-, 6.6- and 6.9-fold in the 25 mg, 50 mg and 75 mg groups, respectively, resulting in a 2.1- to 2.5-fold increase in total clearance of midostaurin. The final model estimated a mean maximum fold of induction (E(max)) of 8.61 and a concentration producing 50% of the E(max) (EC(50)) of 1700 ng/mL (approximately 2.9 micromol/L) for CGP52421-mediated enzyme induction.

Conclusions: The population pharmacokinetic model that was developed was able to describe the time-dependent pharmacokinetic profiles of midostaurin and its auto-induction mechanism. Thus it may be useful for designing an appropriate dosage regimen for midostaurin. The unique feature of this model included a precursor compartment that was able to capture the time delays of auto-induction. The use of such precursor extension in the model may be applicable to other drugs showing long time delays in enzyme auto-induction.

Publication types

  • Clinical Trial, Phase I
  • Controlled Clinical Trial
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Administration, Oral
  • Aged
  • Algorithms
  • Chromatography, High Pressure Liquid
  • Diabetes Mellitus, Type 1 / blood
  • Diabetes Mellitus, Type 1 / drug therapy
  • Diabetes Mellitus, Type 1 / metabolism
  • Diabetes Mellitus, Type 2 / blood
  • Diabetes Mellitus, Type 2 / drug therapy
  • Diabetes Mellitus, Type 2 / metabolism
  • Dose-Response Relationship, Drug
  • Drug Administration Schedule
  • Enzyme Inhibitors / administration & dosage
  • Enzyme Inhibitors / metabolism
  • Enzyme Inhibitors / pharmacokinetics
  • Female
  • Humans
  • Male
  • Middle Aged
  • Models, Biological*
  • Protein Kinase C / antagonists & inhibitors
  • Staurosporine / administration & dosage
  • Staurosporine / analogs & derivatives*
  • Staurosporine / metabolism
  • Staurosporine / pharmacokinetics
  • Time Factors


  • Enzyme Inhibitors
  • Protein Kinase C
  • Staurosporine
  • midostaurin