Web-EWOC: Interactive Web Tool for Designing and Conducting Dose-Finding Trials in Cancer

• Activity Number: 404
• Type: Contributed
• Date/Time: Tuesday, July 31, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Dror Berel*+ and Andre Rogatko and Shao-Chi Huang and Haibin Wang and Galen Cook-Wiens and Mourad Tighiouart
• Companies: Cedars Sinai Medical Center and Cedars Sinai Medical Center and Cedars Sinai Medical Center and Cedars Sinai Medical Center and Cedars Sinai Medical Center and Cedars Sinai Medical Center
• Keywords: EWOC ; Phase I Clinical Trials ; Bayesian ; R ; BUGS ; cancer

Abstract

EWOC is a Bayesian method for selecting dose levels in Cancer Phase I Clinical Trials while controlling the probability of exceeding the maximum tolerated dose (MTD). We implement a novel two-tier distributed Web application, running R code on web server. Using an adaptive dose escalation scheme, based on a parametric quantal response model, EWOC produces a sequence of doses which converge to the MTD, subject to constraint that the predicted probability of assigning doses higher than the MTD is equal to a specified value. Extended priors for the MTD, and probability of DLT at initial dose are implemented. MCMC methods are run through the BUGS project. It enables the designated end-users to choose an optimal dose in a cancer phase I clinical trial. R and BUGS provides an excellent programming environment for implementing Bayesian methods, supporting both advanced statistical methods and advanced computational abilities. It also simplifies the development and modification of the algorithm for future improvements. Combining it with a simple user friendly interface such as a web based application, allows better bench-to-bedside translational statistical tools.