Scientists at St. Jude Children’s Research Hospital have proposed a new method for designing phase II clinical trials and developed software to implement the approach.
The work was recently described in detail in the journal Pharmaceutical Statistics. Corresponding authors are Haitao Pan, PhD, of the St. Jude Department of Biostatistics, and Jianrong Wu, PhD, of the University of Kentucky’s Markey Cancer Center.
“For rare cancers like pediatric cancers, using historical data is a pragmatic approach to speed clinical trials,” said Pan. “The proposed method focuses on a type of phase II study with features of a one-arm study that uses historical control data and a survival endpoint.”
In recent years, drug development in oncology has expanded from mainly randomized controlled trials to subtype-specific clinical trials. The latter often rely on historical controls, Pan said.
“Although the randomized controlled trial is the established gold standard for clinical research, there have been an increasing number of single-arm studies in recent years, particularly in the oncology areas with phase I expansion cohorts and in rare diseases,” said Pan. “Such an approach is better suited for cases where a randomized trial would be unethical or impractical and for smaller trials to give patients faster access to novel therapies.”
The authors present an event-driven approach to Bayesian single-stage and two-stage, single-arm Phase II study designs. Its theoretical and empirical results are intended to help researchers who want to design a study using the Bayesian approach to achieve the operational characteristics often required by regulatory authorities. The software they have developed, called BayesDesign, can be accessed free of charge at https://CRAN.R-project.org/package=BayesDesign.
The study was funded by the National Cancer Institute (CA177558) and ALSAC, St. Jude’s fundraising and awareness raising organization.
Edited by Gary Cramer