Chemotherapy-induced anemia is a frequent complication in cancer treatment. The aim was to develop a pharmacokinetic (PK) model that describes the time-dependent decline of epoetin alfa (rHuEPO) clearance following thrice (t.i.w.) or once (q.w.) weekly subcutaneous injections in cancer patients using a population PK approach. Serum concentrations of rHuEPO were available retrospectively from a phase I study. A one-compartment model with first-order elimination described rHuEPO PK. Sequential zero- and first-order (k(a)) processes with duration (t(lag)) characterized the absorption. Population PK analysis was performed using NONMEM. The influence of several covariates was tested. Model evaluation was performed using visual predictive check. Precision of parameter estimates was assessed by standard errors and confidence intervals determined by bootstrap analysis. Apparent clearance (CL/F) and volume of distribution (V(c)/F) were 25.6% lower and 29.2% higher for q.w. than t.i.w. groups. RHuEPO was absorbed for 10% during 24.6 h through the zero-order process. Following which 90% of the dose was absorbed through the first-order process with k(a) of 0.033 h⁻¹. The most significant covariates were the time-dependent decrease of CL/F with an increase in body weight, a decrease in reticulocyte count, a decrease in hemoglobin baseline, an increase in total number of chemotherapy cycles, and platinum-containing chemotherapy. AGE served as an important covariate on FRAC and k(a). Visual predictive check showed no deviation from observed values. The PK model adequately predicted rHuEPO concentration-profiles in all individuals. Relevant covariates were identified and incorporated into the model.
Copyright © 2010 John Wiley & Sons, Ltd.