Adaptive Borrowing of Adult Data for Pediatric Trials: Collaborative Research at the Intersection of Pharmacometrics and Statistics

• Activity Number: 607 - Effective Application of Modeling, Simulation and Knowledge Sharing in Drug Development
• Type: Topic Contributed
• Date/Time: Thursday, August 1, 2019 : 8:30 AM to 10:20 AM
• Sponsor: Statistics and Pharmacometrics Interest Group
• Author(s): Chyi-Hung Hsu*
• Companies: Janssen, R&D
• Keywords: Pediatrics; Bayesian Statistics; Commensurate Prior; Borrowing

Abstract

Pediatric drug development is challenging in many aspects, particularly, a slow and difficult recruitment may contribute to the failure of a pediatric program. To increase the efficiency of pediatric drug development, the extrapolation approach, based on adult data and other information, was proposed by the US Food and Drug Administration (FDA) in 1990’s. The impact of this approach was evaluated by Dunne, et al (2011), and they concluded that extrapolating did streamline pediatric drug development and help increase the number of approvals for pediatric use. In practice, the appropriate level of extrapolation, e.g. full or partial, is often determined by the similarity in disease progression, and in response to intervention. In addition, the decision on extrapolation level usually needs to be made and agreed upon in a very early stage, sometimes even before the collection of pediatric data. In this presentation, a Bayesian alternative of fitting population pharmacokinetic models is investigated, of which the level of extrapolation and information borrowing is based on the concordance of adult and pediatric data.