09/26/2025: 1:30 PM EDT - 2:45 PM EDT
Parallel
Room: Salon F,Salon G
Early stopping for efficacy or futility, as in group sequential, goldilocks, and promising zone designs, aim to make decisions before the maximum sample size has been reached. These trials can decrease expected sample size without sacrificing power or type I error. Response adaptive randomization, arm dropping, and other randomization manipulation strategies have been researched thoroughly and implemented widely. Adaptive allocation can focus randomization so that more data is collected in on arms of interest. These methods are used broadly and recognized as cutting edge methods that diverge from the classic paradigm of how information is used in a clinical trial.
A frontier that has similar goals as the previously mentioned adaptations, but has not been the focus of as much research as the other two is the use of early endpoint data in adaptive decision making. It's a seemingly obvious statistical observation that a well-designed clinical trial should use all data available to it to make decisions. Despite that, it is common to ignore early data about subjects in a clinical trial, often to the extent that only subjects with complete information are included in statistical models. This session will describe how early visit information can be leveraged through multiple imputation in Bayesian adaptive clinical trials.
The three talks in this session complement each other, creating a natural flow. The first, by Nick Berry, will present general statistical efficiencies available by incorporating early endpoint data into the interim analyses. The second talk, by Tony Jiang, presents an example of a complex innovative design that used longitudinal patient data to improve the quality of its adaptations. Finally, Telba Irony will present on regulatory perspectives of Bayesian designs, and the origin story of the FDA guidance on the use of Bayesian statistics for medical devices.
Adaptive Trials
Longitudinal Data
Bayesian
Multiple Imputation
Organizer
Nicholas Berry, Berry Consultants
Chair
Kert Viele, Berry Consultants
Topic Description
Clinical Trial Design (Innovative/complex design, Master Protocol, multiplicity, multi-regional clinical trial, etc.)
ASA Biopharmaceutical Section Regulatory-Industry Statistics Workshop 2025
Presentations
Complex innovative designs (CID) with adaptive design features can potentially reduce cost, shorten cycle time and increase probability of success in drug development programs. Making Decisions using early data is critical in enabling features of adaptive design such as early futility and response adaptive randomization. In this presentation, a case study accepted by the FDA CID pilot program will be described and especially focusing on using early data for decision-making. Detailed simulation results will be presented to quantify the impact of this approach.
Presenting Author
Xun Jiang, Amgen
In this session, we will discuss lessons learned on the use of Bayesian approaches by presenting examples in which an early endpoint predicts later outcomes at an interim analysis and when strength is borrowed to empower a clinical trial for pediatrics. The FDA guidance on the use of Bayesian statistics for medical devices will clarify the regulatory perspective on these examples.
Presenting Author
Telba Irony, Johnson & Johnson
This talk will focus on the statistical efficiencies gained in adaptive clinical trials that include intermediate endpoint values in the estimation of the final endpoint. The early data augments the complete subjects to increase the effective sample size of the analysis population – this sample size improvement is quantifiable. The increased effective sample size leads to improved precision and MSE of the treatment response estimates at the final endpoint time. This precision benefit increases power over a design that only uses completers. Additionally, incorporating longitudinal data in the analysis models produces interim estimates nearer to the estimates available after complete data is available on the subjects enrolled at the interim. Leveraging early endpoint data, and its relationship to the final endpoint, leads to more accurate predictive probabilities, which results in better decision making at interim analyses.
Presenting Author
Nicholas Berry, Berry Consultants