Clinical trial designs in rare and ultra rare diseases yield multiple challenges due to restricted sample sizes. In such settings, innovative Bayesian methods can often pay dividends, allowing the sensible incorporation of auxiliary data and other relevant information to bolster that collected by the trial itself. Previous work has demonstrated the usefulness of one-arm trials augmented by the participants’ own natural history data, from which the future course of the disease in the absence of intervention can be predicted. Patient response can then be defined by the degree to which post-intervention observations are inconsistent with the predicted “natural” trajectory. However, such methods can offer no protection against biases arising from the presence of any “placebo effect,” the tendency of some patients to improve merely by being in the trial.
In this webinar, after a very brief review of Bayesian adaptive clinical trial methods, we introduce our Bayesian responder approach to one-arm clinical trials in rare disease modeling, investigating the impact of both static and transient placebo effects. We then go on to describe two-arm versions that incorporate a small concurrent placebo group, but still borrow strength from the natural history data. We also propose more traditional Bayesian changepoint models that specify a parametric functional form for the patient’s post-intervention trajectory, which in turn allow quantification of the treatment benefit in terms of the model parameters, rather than semiparametrically in terms of a response relative to some “null” model. Our results indicate that our two-arm responder and changepoint methods can offer protection against placebo effects, improving power while controlling the trial’s Type I error rate. We offer illustrations in the context of a clinical trial in a particular rare disease, where large patient-to-patient and visit-to-visit heterogeneity can be observed. In such settings, our innovative Bayesian techniques facilitate increased power to detect an effect with respect to more classical methods. We also offer advice regarding computational approaches in these settings, as well as our experience with key regulatory authorities, dialog with whom of course remains crucial in rare disease research.
Senior Advisor, Data Science, PharmaLex
Brad is a statistical researcher, methodologist, consultant, and instructor. Prior to joining PharmaLex, he spent 27 years on the faculty of the Division of Biostatistics at the University of Minnesota School of Public Health, serving as division head for 7 of those years. He has published more than 185 papers in refereed books and journals, and has co-authored three popular textbooks on Bayesian statistical methods and their applications in spatial statistics and adaptive clinical trials. During his spare time, Brad is a health musician and bandleader, providing keyboards and vocals in a variety of venues. He is excited and proud to be a part of the PharmaLex data science and statistics team.
Senior Manager Statistics, PharmaLex
Arnaud holds a bachelor’s degree in Mathematics and a master degree in statistics and econometrics (Université Catholique de Louvain). He is a Senior Manager Statistics and Data Science at PharmaLex and has been at PharmaLex for the past 5 years. He has been primarily involved in various projects covering non-clinical, pre-clinical as well as clinical projects.