Chair: Yeh-Fong Chen, PhD (FDA)
Vice-Chair: Qian Tang, PhD (University of Iowa)
Instructors:
Johan Vebeeck, PhD (Data Science Institute, UHasselt, Belgium)
Xiaoyu Cai, PhD (Insmed Inc, USA)
Course Description:
When assessing the effects of a treatment in clinical trials, it is recommended by the International Conference on Harmonization to select a single meaningful endpoint. However, in many situations, several clinical meaningful endpoints are simultaneously of interest. Yet, standard methods of analysis for multiple endpoints are limited in a number of ways, not the least in the number and type of endpoints that can be combined. For instance, composite endpoints of the type ‘time-to-first event’, which have gained traction in clinical fields with a low incidence of events (e.g., cardiovascular studies), cannot combine time-to-event endpoints with continuous or categorical endpoints. Additionally, they ignore the relative importance of different clinical events, and by focusing on the first events, they disregard all other subsequent events in time. Last but not least, standard methods for multiple endpoints typically have poor small sample properties.
The Generalized Pairwise Comparisons (GPC) methods were introduced a decade ago to overcome all of these limitations. It compares each patient in one treatment group with each patient in the other group (i.e., pairwise comparisons) based on predefined clinical priorities and thresholds of clinical relevance. GPC methods form a very flexible class of non-parametric techniques that has gained much traction over recent years. Among other advantages, GPC methods can prioritize the endpoints by clinical severity, boost trial power, combine different types of endpoints, allow for matched designs, benefit-risk assessments and allow for thresholds of clinical relevance when performing the pairwise comparisons. These elements pave the way for clinically relevant and patient-centric design of clinical trials. Theoretical developments also point to the good small sample properties of the methodology. Lastly, the method is intuitively appealing and helps enhance communication between all the stakeholders involved in trials.
In this course, we provide an introduction to the flexible framework of GPC, focusing on practical solutions for the design and analysis of clinical trials with examples in cardiology, oncology and rare diseases, followed by a Q&A for the remainder of the session. Interested participants are encouraged to practice the GPC methodology on provided datasets with the aid of user-friendly software.
Learning objectives
The aim of this course is:
- to learn about the practical limitations of combining multiple endpoints in standard methods, and to explore the history, development, and recent advances in Generalized Pairwise Comparisons (GPC),
- to argue that GPC is a suitable method to combine any number and type of efficacy and safety endpoints, even in small sample trials, in a manner that is clinically relevant and can account for the voice of patients,
- to understand the potential advantages and disadvantages of designing a clinical trial with a GPC primary analysis, and the appropriate way to interpret the finding of the analysis results,
- to illustrate GCP with thorough literature review, technical details, numerical simulation, and recent case studies, and provide hands-on practice to the appropriate design and analysis of clinical trials with GPC with the aid of user-friendly software
The course is aimed at statisticians, clinicians and trialists from academia, industry and regulatory agencies with knowledge on clinical trials who want to learn more about the GPC statistical methodology for multiple endpoints of potentially different importance and data types in large and small sample trials.
Prerequisite: | Clinical Trials, Benefit and Efficacy Assessment |
Instructor:
Johan Vebeeck, PhD
Assistant Professor
Data Science Institute
UHasselt, Belgium
Prof. Dr. Johan Vebeeck holds a master’s degree in Biotechnology (1998-2000) from the University of Ghent. After taking several positions in the pharmaceutical industry, including Medical Advisor and Expert medical trainer, he obtained a master’s degree in Biostatistics (2014-2018) from the University of Hasselt, during which he received the Quetelet prize for an outstanding master thesis in 2018. He obtained his PhD in Biostatistics at the University of Hasselt in 2022 on the topic of prioritized endpoints and COVID-19 epidemiology and continues his research focusing on statistical methods to analyze multivariate efficacy, safety and patient-reported outcomes, in particular in small sample trials and benefit-risk assessments, through several H2020 funded projects (has authored multiple research papers and recently co-authored a book (https://doi.org/10.1201/9781003390855) on Generalized Pairwise Comparisons (GPC). Verbeeck has authored multiple research papers and recently co-authored a book (https://doi.org/10.1201/9781003390855) on Generalized Pairwise Comparisons (GPC). He acts as a consultant for clinical trials in several clinical areas, is a lecturer in statistics at Hasselt University, is a frequent speaker at international conferences, and has organized several short-courses.
Instructor:
Xiaoyu Cai, PhD
Associate Director of Biostatistics
Insmed Inc., USA
Dr. Xiaoyu Cai is an Associate Director of Biostatistics at Insmed, Inc. Prior to joining Insmed, she served as a Senior Mathematical Statistician in the Office of Biostatistics at the U.S. Food and Drug Administration’s Center for Drug Evaluation and Research. She received her Ph.D. in Statistics from The George Washington University. Dr. Cai has extensive experience in statistical review and research for clinical trials across multiple therapeutic areas, including hematology and inflammatory diseases. Her research interests include Bayesian methods for small clinical trials, real-world evidence and real-world data, rare disease trial design, missing data imputation, adaptive designs, biosimilar and bioequivalence testing, and product quality assessment.
