Aims: To establish a pharmacodynamic model that allows one to predict the haemoglobin (Hb) response to EPO in children as a function of dose and time, and to derive recommendations for initial dosing and subsequent dose adjustment.
Methods: Haemoglobin was monitored in eight children aged 8-15 years with anaemia due to renal failure during treatment with EPO. All patients were free of conditions known to impair the response to EPO. Pretreatment Hb was 4.9-9.0 g dl(-1). The drug was administered once weekly by subcutaneous injection; doses ranged from 1700 to 6800 U week (-1). Hb was monitored for 4-38 months. The Hb-time data were analysed by applying a population pharmacodynamic model proposed for EPO in adult haemodialysis patients. Internal model validation was carried out by using a bootstrap procedure.
Results: The increase of Hb during treatment with EPO was linear until steady state was reached after 103+/-33 days (mean +/- interindividual s.d.). The weekly gain in Hb from the onset of therapy to steady state was 0.0805+/-0.026 g dl(-1) (mean +/- interindividual s.d.) for every 1000 U EPO week (-1); it did not exhibit a dependence on body weight. Estimated mean prediction errors are +/-1.53 g dl(-1) for predictions that are based on the mean population parameters and +/-0.83 g dl(-1) for predictions that take into account the complete individual Hb-time data up to and including steady state.
Conclusions: The model describes the time course of the Hb response to EPO in children with renal anaemia. The required weekly EPO dose should initially be calculated from the individual pretreatment Hb and the desired Hb at steady state by using the mean population estimates of the weekly gain in Hb per dose unit before steady state (beta) and the time needed to reach steady state (tau). A reduction of the initial dose according to body weight is not justified by the available evidence. beta should be re-estimated individually after 6 weeks of treatment and dose should be adjusted accordingly. A final dose adjustment should be made when steady state has been reached based on individual estimates of beta and tau.