Population Pharmacokinetics and Pharmacodynamics of Hydroxyurea in Sickle Cell Anemia Patients, In Silico Comparison of Two Dosing Regimens
Ines Paule (1), Hind Sassi (2), Anoosha Habibi (3), Dora Bachir (3), Frédéric Galactéros (3), Pascal Girard (1), Anne Hulin (2), Michel Tod (1, 4)
(1) EA3738 CTO, Faculté de Médecine Lyon-Sud, Université Lyon 1, Oullins, France; (2) Laboratoire de toxicologie and (3) Centre de référence pour les syndromes drépanocytaires majeurs, AP-HP, CHU HMAC, Université Paris 12, Créteil, France; (4) Hôpital Croix-Rousse, Hospices Civils de Lyon, Lyon, France.
Objectives: The antineoplastic agent hydroxyurea (HU) is the first approved pharmacological treatment of sickle cell anemia (SCA). It stimulates the production of fetal hemoglobin (HbF) in place of the hemoglobin S that causes SCA and thus reduces the rate and severity of painful attacks and was shown to possibly increase survival time. This study aimed to develop population PK and PD models for HU and to compare two dosing regimens via a simulation of a clinical trial.
Methods: The population PK and PD models were built using NONMEM VI (FOCEI method) based on two datasets of SCA adult patients receiving 500-2000 mg of HU once daily (sparsely sampled PK and PD data of 36 patients and densely sampled PK data of 16 patients). PK and PD models were built sequentially. The datasets included 10 demographic and biological covariates. The PD data included HbF%, Hb, reticulocytes, PMN and platelets. Models were evaluated by VPC, NPC, NPDE. In the clinical trial simulation, the two dosing regimens were 1000 mg daily 7 days a week and 1000 mg daily 5 days a week; the duration was 6 months. The results were compared graphically by representing the HbF% kinetics and its variability.
Results: The PK profiles were described by a bicompartmental model (with first-order absorption and elimination). The typical value (and interindividual CV %) of apparent Cl was 0.95 L/h (46%), the apparent Vcentral was 3.5 L (62%), the apparent Vperif. was fixed to 50 L, and the Ka was 2 h-1 (180%). None of covariates was significant in the PK model. The relationship between mean weekly concentrations of HU and HbF% was described by an indirect effect (inhibition of elimination) model (without significant covariates). The typical value (and interindividual CV %) of Kin was 0.13 %/day (0%), Kout - 0.04 day-1 (134%), the mean concentration needed for 50% of Emax (CM50) - 11 mg/L (155%). PMN and platelet profiles were normal and therefore were not modelled. The simulated HbF% profiles in both arms were comparable.
Conclusions: The PK of HU and its effect on HbF% show high variability. The results suggested that 1000 mg daily doses may reach drug effect saturation. Dosage of HU should be adjusted individually by HbF monitoring.