Jump to content
14:00 - 18:00

Pre-Conference Workshop 5 – Using R for Early-Phase Dose Finding Studies

Workshop Presenters

Daniel Sabanés Bové, Roche, Basel, Switzerland
Ying Kuen Cheung, Columbia University, New York, NY
Philip Pallmann, Lancaster University, Lancaster, UK
Graham Wheeler, Cancer Research UK & UCL Cancer Trials Centre, University College London, UK & MRC

Description of Workshop

In this workshop we demonstrate the utility of the open access software R for designing and conducting early-phase dose finding studies. First we give an overview of CRM designs for single or multiple drugs, binary or time-to-event toxicity endpoints, and seamless phase I/II designs. The main focus of the session will be on practical examples and hands-on exercises. We illustrate all methods with data from real clinical trials and make full use of R and its add-on packages ‘dfcrm’, ‘bcrm’, and ‘crmPack’. We also demonstrate a ‘Shiny’ web application with a graphical user interface to compare the operating characteristics of 3+3 and CRM-type methods, and another one to design and carry out a CRM study without any knowledge of R.

Goals of session:

  • Familiarise delegates with the wide range of functionality that R offers for early-phase dose finding.
  • Demonstrate the use of R functions and tools for basic single-agent CRM studies as well as various extensions.
  • Enable delegates to use R and its add-on packages to explore CRM design options for dose escalation studies and thereby build confidence and trust in model-based methods.

Participants should bring a laptop with the latest version of R (www.R-project.org) installed.

Target Audience
Statisticians and clinical investigators interested or involved in the planning, conduct, and evaluation of early-phase dose escalation studies. In-depth knowledge of statistical programming is not required for this course, but some basic familiarity with R may be helpful.