Development and evaluation of a population pharmacokinetic-pharmacodynamic model of darbepoetin alfa in patients with nonmyeloid malignancies undergoing multicycle chemotherapy.

Anemia is frequently observed in patients undergoing chemotherapy. Administration of darbepoetin alfa, a recombinant erythropoiesis-stimulating agent that has longer residence time than endogenous erythropoietin, to patients with chemotherapy-induced anemia (CIA) increases mean hemoglobin concentration, reduces risk of red blood cell transfusions, and improves patient-reported outcomes. A pharmacokinetic/pharmacodynamic (PkPd) model was developed using data from patients with nonmyeloid malignancies and CIA who were receiving darbepoetin alfa. A 2-compartment Pk model with linear elimination described the Pk data obtained in 140 CIA patients after intravenous and subcutaneous (s.c.) doses of 2.25 microg/kg every week and s.c. doses of 6.75 microg/kg every 3 weeks. The population typical values of key Pk parameters were clearance, 2010 mL/day; steady-state volume of distribution, 3390 mL; and bioavailability, 44.3%. A modified indirect response model, wherein serum concentrations stimulated the production of hemoglobin through an Emax-type equation, described the hemoglobin levels after s.c. doses of 0.5 microg/kg every week to 15 microg/kg every 3 weeks in 573 CIA patients. The estimated incremental maximum stimulation of hemoglobin production was 43.7% and darbepoetin alfa serum concentration at half-maximal stimulation was 3.68 ng/mL. The impact of covariates (body weight and platinum-containing chemotherapy) on the PkPd response was evaluated based on point and interval estimates of parameters, rather than through stepwise hypothesis testing. The final PkPd model adequately predicted hemoglobin response in a test data set, thereby confirming the predictive capability of the model. Based on simulations, it was not possible to categorize the influence of any covariate as clinically important.