Chair/Organizer: Brad Carlin, PhD (PhaseV Trials)
Abstract:
The implementation of Bayesian adaptive trials in clinical research presents unique practical challenges, particularly when integrating these methodologies into software products. Commercial packages can be somewhat constraining and inflexible, and often fail to incorporate modern tools for design optimization and causal inference. By contrast, the writing of bespoke computer code for each new design is often inefficient and unacceptably work-intensive. In this session, we will hear from 3 speakers who have extensive experience in both the methodological and computational aspects of this tradeoff. Speaker 1 (Fogel) will provide an overview of recent advances in computational power to manage the complexity of the Bayesian approach, contrasting it with frequentist approaches. He will then describe a new platform whose application programming interface (API) facilitates effective communication among diverse stakeholders, and presents complex Bayesian results in an intuitive manner. Speaker 2 (Wathen) will offer an in-depth exploration of the unique challenges of adaptive platform designs, and discuss practical solutions for navigating this intricate landscape using OCTOPUS, an R package for simulation of platform trials. This talk will emphasize the critical importance of simulating the exact platform trial one plans to conduct, and of accounting for the addition and removal of new treatments over time, as well as other sources of variation that may impact the performance of the platform. Speaker 3 (Pryluk) will introduce a novel ensemble estimation approach that leverages causal machine learning methods to enhance the detection and assessment of heterogeneity in adaptive trials. The framework uses conformal prediction to assess uncertainty in its ML estimates for finite samples, facilitating a more nuanced understanding of how different patient subgroups respond to treatments. Finally, the discussant (Thompson) will summarize the presentations, offer a regulatory perspective on the commercial and open-source approaches, give her view on the future of AI in regulatory science, and suggest areas for future work. Time will also be reserved for questions from the floor.
Speaker: Yaniv Fogel, M.Sc. (PhaseV Trials)
Title: To Bayes or Not to Bayes: Practical Challenges and Considerations in Supporting the Design of Adaptive Clinical Trials
Abstract: Clinical research faces distinct operational challenges when putting Bayesian adaptive trials into practice. Through real-world examples, this presentation will explore these obstacles and demonstrate how a new software solution can address them, highlighting recent computational improvements and the advantages of an API-first approach.
We will discuss the computational complexity of Bayesian methodologies compared to frequentist methods. The Bayesian paradigm, necessitating extensive probabilistic computations and iterative simulations, imposes substantial computational requirements. Recent technological advancements in processing capabilities have, however, facilitated novel approaches to managing these computational costs. We shall demonstrate how our new platform capitalizes on these technological progressions to render Bayesian adaptive designs more suited for practical use.
In addition, we will address the complexities inherent in eliciting and integrating contributions from diverse stakeholder groups, while emphasizing the critical nature of transparent communication protocols. The implementation of Bayesian adaptive trials requires collaborative engagement between clinical personnel, statistical experts, and regulatory authorities, each contributing distinct domain expertise. We shall demonstrate how our platform’s API-centric architecture facilitates efficacious communication and renders complex Bayesian analytics more accessible and comprehensible. This methodological approach ensures stakeholders can make clear and meaningful contributions to the trial design.
Through examination of these operational considerations, we present both the pragmatic challenges and potential advantages inherent in Bayesian adaptive trial designs within the context of clinical research. We also demonstrate how our software platform addresses the challenges through enhanced trial design and execution protocols, while maintaining equilibrium between computational demands and communicative effectiveness.
Speaker: Kyle Wathen, PhD (Cytel)
Title: Navigating the Complexities of Adaptive Platform Trials: Design and Simulation
Abstract: As clinical trial design evolves to encompass innovative methodologies like Bayesian adaptive designs and master protocols, the complexity and variety of design options can be overwhelming. This presentation will delve into the intricacies of designing and simulating adaptive platform trials, highlighting the pivotal role of statisticians in this new landscape.
Building on a platform trial currently under design, this talk will illustrate how the role of statisticians has evolved from traditional data analysts to key contributors in the trial design process. As design complexity increases, extensive simulations become a necessity, opening up opportunities for statisticians to guide the design process.
This presentation will provide a brief overview of adaptive platform trials, including their terminology, potential risks and benefits.
The presentation will demonstrate how simulations can be readily conducted using OCTOPUS from within RStudio, combined with trial-specific additions, to accommodate various options explored for Bayesian analysis. It will present several visualizations used to illustrate trade-offs and evaluate the frequentist operating characteristics of a Bayesian design.
The focus will be on how intuitive visuals and advanced simulation tools can guide the team and stakeholders to understand the trial processes and expected outcomes under a wide range of scenarios. This approach facilitates informed decision-making, and optimizes trial design to potentially save time, resources, and most importantly, enhance patient outcomes.
Speaker: Raviv Pryluk, PhD (PhaseV Trials)
Title: Causal Machine Learning for Estimating Uncertainty and Detecting Subgroups in Adaptive Clinical Trials
Abstract: In clinical trials, accurate assessment of heterogeneity is essential for personalized medicine and optimized treatments. This presentation introduces an innovative, combined approach that utilizes causal machine-learning (ML) methods to better identify and evaluate heterogeneity within clinical trials.
Causal ML excels at identifying and interpreting the relationships between variables, enabling a more nuanced understanding of how different patient subgroups respond to treatments. However, selecting the optimal causal ML method is challenging due to frequent disagreements among algorithms, each offering unique advantages and limitations. In addition, many of the causal ML methods still lack well-understood uncertainty estimates, especially in finite samples that are common in clinical trials.
In this talk, we present a method that addresses these challenges by combining different algorithms through both Bayesian and non-Bayesian approaches, while accounting for model uncertainty. This integration leverages the advantages of the various multiple models, resulting in consistent performance across various types of data.
We will present simulation studies demonstrating our approach in clinical trial settings, showing its ability to detect insights into patient heterogeneity and treatment outcomes. This work marks an important advance in applying machine learning to improve clinical trial design and personalized medicine, offering researchers and clinicians a new tool that is both practical and effective.
Discussant: Laura Thompson, PhD (FDA CDER)
