Information Borrowing in Rare Disease and Non-Oncology Clinical Trials
Chair: Qiming Liao (ViiV)
Vice Chair: Freda Cooner (Eli Lilly)
Speaker: Satrajit Roychoudhury (Pfizer)
Title: Use of External Data in to Accelerate Rare Disease Drug
Abstract: Design and analysis of clinical trials for rare and ultra-rare disease pose unique challenges to the practitioners. Meeting conventional power requirements is infeasible for diseases where sample sizes are inherently very small. Moreover, rare diseases represent a highly heterogeneous group of disorders. External data sources such as natural history data and real-world data (RWD) play significant roles in defining and characterizing disease progression, final patient populations, novel biomarkers, genetic relationships, and treatment effects in this context. This talk will focus on applications of different types of external data in rare diseases. Considerations for data quality and limitations when using natural history and RWD will be elaborated. Opportunities will include possible avenues of cross-sector collaboration, standardized and high-quality data collection using new technologies, and more comprehensive evidence generation using quantitative approaches such as disease progression modeling, artificial intelligence, and machine learning. Advanced statistical approaches to integrate natural history data and RWD to further disease understanding and guide more efficient clinical study design and data analysis in drug development in rare diseases will be discussed. Concepts will be further elaborated using real-life examples.
Speaker: Jenny Huang (GSK)
Title: Bayesian Dynamic Borrowing in Pediatric Clinical Trials
Abstract: In drug development program, if a product is being developed for adults and is anticipated for pediatric patients, clinical trials for pediatric population are often needed. Due to various reasons including challenges in enrollment and adherence etc., pediatric clinical trials are usually in small sample size and with no controls. However, large quantity of efficacy and safety data are available in reference populations (adults for the drug, and adults/pediatrics for the similar class of HIV drugs etc.) and Bayesian borrowing approach can be useful to incorporate reference data into pediatric clinical trials to support efficacy and safety profile. In this talk, the method of Bayesian dynamic borrowing and its application to HIV pediatric clinical trials will be discussed.
Speaker: Bradley Hupf (Takeda)
Title: Bayesian semiparametric meta-analytic-predictive prior for historical control borrowing in clinical trials
Abstract: When designing a clinical trial, borrowing historical control information can provide a more efficient approach by reducing the necessary control arm sample size while still yielding increased power. Several Bayesian methods for incorporating historical information via a prior distribution have been proposed, e.g. (modified) power prior, (robust) meta-analytic predictive prior. When utilizing historical control borrowing, the prior parameter(s) must be specified to determine the magnitude of borrowing before the current data is observed. Thus, a flexible prior is needed in case of heterogeneity between historic trials or prior data conflict with the current trial.
To incorporate the ability to selectively borrow historic information, we propose a Bayesian semiparametric meta-analytic-predictive prior. Using a Dirichlet process mixture prior allows for relaxation of parametric assumptions, and lets the model adaptively learn the relationship between the historic and current control data. Additionally, we generalize a method for estimating the prior effective sample size (ESS) for the proposed prior. This gives an intuitive quantification of the amount of information borrowed from historical trials, and aids in tuning the prior to the specific task at hand. We illustrate the effectiveness of the proposed methodology by comparing performance between existing methods in an extensive simulation study and a phase II proof-of-concept trial in ankylosing spondylitis (AS).
In summary, our proposed robustification of the meta-analytic-predictive prior alleviates the need for pre-specifying the amount of borrowing, providing a more flexible and robust method to integrate historical data from multiple study sources in the design and analysis of clinical trials.
Speaker: Hwanhee Hong (Duke)
Title: Comparing Bayesian methods for creating external controls using multiple historical controls in clinical trials
Abstract: Approving new therapies for rare diseases are urgently needed. However, traditional randomized controlled trials can be suboptimal for this purpose with several challenges, including an ethical challenge (e.g., how many patients should be randomized to placebo) and a practical challenge (e.g., how to recruit a sufficiently large number of patients that is required for confirmatory clinical trials.) Recently, using external controls, such as placebo arms of historical trials or real-world data such as electronic health records, in designing clinical trials with patients having rare diseases gains extreme popularity because the external information can complement the information gained from the concurrent trials. However, there is no clear guidance about creating an external control by borrowing information adaptively from multiple sources. In this work, we consider Bayesian methods for adaptive borrowing information such as power prior and meta-analytic prior. We compare the performance of the methods with respect to operating characteristics (type I error, calibrated power, and effective sample size) and estimation accuracy (bias and mean squared error) under various scenarios. We illustrate how these methods perform differently in adaptive borrowing using a real clinical trial example in idiopathic pulmonary fibrosis. This is joint work with Eric Yanchenko (North Carolina State University), Scott Palmer (Duke), and Lisa Wruck (Duke).
