Guiding Clinical Trial Design for a Rare Disease Using Natural History Data and Bayesian Disease Progression Modeling (306411)Scott Berry, Berry Consultants
Leslie Biesecker, National Human Genome Research Institute, National Institutes of Health
Melanie Quintana, Berry Consultants
Julie Sapp, National Human Genome Research Institute, National Institutes of Health
*Barbara Wendelberger, Berry Consultants
Keywords: Disease progression modeling, clinical trials, trial design, Bayesian, modeling, rare diseases, natural history data
Rare diseases present challenges to clinical trial design, which may include small, heterogenous patient populations, insufficient understanding of disease etiology, and poorly developed study endpoints. Proteus syndrome is a rare condition resulting from a genetic mutation that manifests in the overgrowth of skin, bone, and other tissues. Here, we investigate the design of a single arm clinical trial in Proteus syndrome that allows us to understand disease progression, simulate virtual patients in virtual trials, and investigate a Bayesian primary analysis model across a range of treatment effects. This Bayesian disease progression model (DPM) provides flexibility that enables estimation of both the control rate of disease progression from natural history data and the rate of disease progression relative to the natural history rate of progression from treated subjects in the single arm trial. The DPM encapsulates both clinical disease progression and statistical disease modeling and provides an intuitive framework for the explanation and interpretation of results for regulatory authorities, effectively providing solutions for researching a challenging rare disease.