Chair: Huan Wang, PhD (FDA)
Vice Chair: Yeh-Fong Chen, PhD (FDA)
Instructors:
James Travis, PhD (FDA)
Huan Wang, PhD (FDA)
Fei Wu, PhD (FDA)
Course Description:
When a new drug is approved for adult patients, the next step in general is to commence studies in pediatric patients to comply with the Pediatric Research Equity Act (PREA). However, it is often not feasible to conduct well-controlled pediatric studies with statistically significant and persuasive findings because pediatric populations often face unique enrollment and feasibility challenges in the design and execution of clinical trials. As a result, randomized studies with conventional frequentist approaches may not be possible. Bayesian borrowing is often proposed as an option as it offers a promising approach to enhance trial efficiency and increase the likelihood of success. Single arm studies and hybrids designs with both concurrent and external control have also been proposed to alleviate the feasibility issues in some settings.
In this course, we will provide a comprehensive overview of statistical approaches for pediatric clinical trials and provide a background regarding the regulatory considerations and common practices for pediatric studies in terms of efficacy and safety assessments. We will also discuss the use of frequentist approaches with illustration from real clinical trials as well as give an overview of designs utilizing Bayesian borrowing techniques that leverage adult and older pediatric study data. The course is divided into three parts:
- Background Introduction about Pediatric Clinical Trials and Frequentist Methods for Small-Size Trials: This part will begin with an introduction to the statistical challenges specific to pediatric trials. It will then proceed to a discussion of the traditional frequentist methods and their limitations when analyzing pediatric studies.
- Introduction to Bayesian Analysis: This part will provide a foundational overview of Bayesian analysis, including its core principles, how it can be used to incorporate prior information into clinical trials, and an introduction to the concept of the effective sample size (ESS).
- Advanced Bayesian Borrowing Techniques: This part will cover Bayesian borrowing methods, such as Bayesian dynamic borrowing models, to incorporate information from historical trials. Specific topics will include borrowing the overall treatment effect from historical trials, borrowing information separately from treatment and control arms, and connecting the mixing weight of the robust mixture prior with the effective sample size.
Through case studies and numerical examples, the course will demonstrate the practical application of these methods, showing how Bayesian designs leveraging adult study data can enhance the interpretability and reliability of pediatric trial outcomes while addressing the ethical and practical constraints of conducting studies in children.
This course is ideal for statisticians, clinical trial designers, and regulatory professionals interested in the application of frequentist or Bayesian methods to enhance the design and analysis of pediatric trials.
Instructors:
James Travis, PhD
Master Mathematical Statistician
Division of Biometrics II, Office of Biostatistics
Center for Drug Evaluation and Research, FDA
James Travis is a master mathematical statistician in the Division of Biometrics II in the US FDA Center for Drug Evaluation and Research. He is the technical lead for the pediatric and maternal health scientists and has been supporting pediatric drug development since joining the Pediatric Review Committee in 2017. He has worked extensively on applications in the complex and innovative trial design program and has led the development of a Bayesian education program for the Office of Biostatistics.
Huan Wang, Ph.D.
Office of Biostatistics
U.S. Food and Drug Administration (FDA)
Dr. Huan Wang joined the FDA in 2021, where he focuses on statistical reviews for drug approval applications related to non-malignant hematologic diseases. Prior to this role, he earned his Ph.D. in Biostatistics from the George Washington University, where he conducted research on developing machine learning-based prognostic systems for cancer patients. His current projects involve developing effect sizes for survival analysis, statistical modeling for externally controlled trials, and Bayesian borrowing for sample size calculation in pediatric and rare disease studies.
Fei Wu, Ph.D.
Mathematical Statistician
Division of Biometrics IX, Office of Biostatistics, CDER
U.S. Food and Drug Administration (FDA)
Dr. Fei Wu is a mathematical statistician in the Division of Biometrics IX, Office of Biostatistics, CDER, FDA, where his work focuses on the review of hematology drug applications. He earned his PhD in Statistics from the University of Iowa. His expertise and research interests span a wide range of statistical topics, including time series analysis, high-dimensional statistics, causal inference, and Bayesian methods.
