Tailor-made drug treatment for children: creation of an infrastructure for data-sharing and population PK-PD modeling
Ibrahim Ince (1,2), Saskia N. de Wildt (1,6), Dick Tibboel (1), John N. van den Anker (1,4), Daan J. Crommelin (5), Meindert Danhof (2), Catherijne A.J. Knibbe (1,2,3)
(1) ErasmusMC Sophia Children’s Hospital, Departments of Paediatric Surgery and Paediatrics, Rotterdam, The Netherlands, (2) Leiden/Amsterdam Center For Drug Research, Department of Pharmacology, Leiden, The Netherlands, (3) St. Antonius Hospital, Department of Clinical Pharmacy, Nieuwegein, The Netherlands, (4) Children’s National Medical Center, George Washington University, Washington, DC, USA. (5) Top Institute Pharma, Leiden, The Netherlands, (6) The Hospital for Sick Children, Division of Clinical Pharmacology and Toxicology Toronto, Ontario
Objectives: Despite profound differences in response between children (of different ages) and adults, drugs are used in children in an empirical manner. While about 70% of the prescribed drugs are unlicensed or off-label, this has important implications for the efficacy and safety of pharmacotherapy in these vulnerable patients. When developing dosing schemes for drugs in children, non-linear mixed effects modeling (population approach) is important. It allows the usage of sparse data and implementation of co-variates, and distinguishes between drug-specific and system specific causes of variability in response, to develop evidence-based individualized dosing regimens for (classes of) drugs. Therefore, in the Netherlands, a multidisciplinary research platform for the design of individualized dosing regimens in children has been established, sponsored by the Dutch Top Institute Pharma. Partners in this platform are 4 academic institutions and 6 pharmaceutical industries.
Methods: Modeling and simulation using non-linear mixed effects modeling for the development of rational dosing schemes in children involves: 1) optimization of trial design based on preliminary data, 2) development, including internal validation of population PK-PD models using sparse data, 3) external validation of the proposed models using independent data and 4) prospective clinical evaluation of the model-based individualized dosing regimens.
Results: To date, in the platform an infrastructure for sharing anonimized data in a secure environment, restricted access to predetermined investigators, extensive data checks and strict data management rules, has been created, allowing to implement the four steps in the development of dosing schemes for children. PK-PD model development has recently started.
Conclusions: The proposed infrastructure will allow to share datasets for model building and external validation for not only PK-PD model based age-specific dosing guidelines for specific drugs, but also a general dosing paradigm for drugs with similar disposition and/or effect.