Extrapolation of praziquantel pharmacokinetics to a pediatric population: a cautionary tale

• Published in: Journal of pharmacokinetics and pharmacodynamics Vol. 45; no. 5; pp. 747 - 762
• Main Authors: Bonate, Peter L., Wang, Tianli, Passier, Paul, Bagchus, Wilhelmina, Burt, Howard, Lüpfert, Christian, Abla, Nada, Kovac, Jana, Keiser, Jennifer
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.10.2018; Springer Nature B.V
• Subjects: Bioavailability; Biochemistry; Biomedical and Life Sciences; Biomedical Engineering and Bioengineering; Biomedicine; Children; Data processing; Health care; Original Paper; Pharmacokinetics; Pharmacology/Toxicology; Pharmacy; Praziquantel; Public health; Tablets; Veterinary Medicine/Veterinary Science; Population pharmacokinetics; Root cause analysis; Linear mixed effects models; Oral dispersion tablet; Complex pharmacokinetics; NONMEM
• ISSN: 1567-567X; 1573-8744; 1573-8744
• DOI: 10.1007/s10928-018-9601-1

L-praziquantel (PZQ) pharmacokinetic data were analyzed from two relative bioavailability Phase 1 studies in adult, healthy subjects with two new oral dispersion tablet (ODT) formulations of L-PZQ administered under various combinations of co-administration with food, water, and/or crushing. Linear mixed effects models adequately characterized the noncompartmental estimates of the pharmacokinetic profiles in both studies. Dose, food, and formulation were found to significantly affect L-PZQ exposure in both studies. The model for AUC was then extrapolated to children 2–5 years old accounting for enzyme maturation and weight. The predicted exposures were compared to an external Phase 1 study conducted by the Swiss Tropical and Public Health Institute using a currently marketed formulation (Cesol 600 mg immediate-release tablets) and found to be substantially lower than observed. A root cause analysis was completed to identify the reason for failure of the models. Various scenarios were proposed and tested. Two possible reasons for the failure were identified. One reason was that the model did not account for the reduced hepatic clearance seen in patients compared to the healthy volunteer population used to build the model. The second possible reason was that PZQ absorption appears sensitive to meal composition and the model did not account for differences in meals between a standardized Phase 1 unit and clinical sites in Africa. Further studies are needed to confirm our hypotheses.