Discussion of the article by Trippa, Rosner, and Müller on Bayesian enrichment strategies for randomized discontinuation trials

Biometrics. 2012 Mar;68(1):212-4; discussion 224-5. doi: 10.1111/j.1541-0420.2011.01624.x. Epub 2011 Jun 29.

Abstract

We congratulate Trippa, Rosner, and Müller (2011, Biometrics, in press) on an intriguing and timely article. The randomized discontinuation design (RDD) has only recently been used in cancer clinical trials, and methodological understanding on how to best design such studies is limited. The authors’ approach to optimize RDD designs based on prior information and utility considerations is an important step forward. A noteworthy element is their use of a semimechanistic model to describe tumor growth. Mathematical models have provided considerable insight on the complex process of tumor evolution (Preziosi, 2003, Cancer Modelling and Simulation). Utilizing this knowledge should lead to better design, analysis, and decisions.

Publication types

  • Comment

MeSH terms

  • Bayes Theorem*
  • Biometry / methods*
  • Data Interpretation, Statistical*
  • Humans
  • Models, Statistical*
  • Neoplasms / epidemiology*
  • Randomized Controlled Trials as Topic / methods*
  • Randomized Controlled Trials as Topic / statistics & numerical data*