Maximum Likelihood Methods for Optimal Treatment Schedule-Finding in Clinical Trials

• Activity Number: 66
• Type: Contributed
• Date/Time: Sunday, July 29, 2007 : 4:00 PM to 5:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Changying Liu*+ and Thomas M. Braun
• Companies: University of Michigan and University of Michigan
• Keywords: Phase I trial ; Maximum tolerate schedule ; change-point ; boundary parameter ; maximum likelihood ; large sample property

Abstract

Braun, Yuan, and Thall (2005) constructed a Bayesian paradigm for Phase I trial designs that allows for the evaluation and comparison of several treatment schedules, each consisting of a sequence of administration times. The goal of their design was to determine a maximum tolerated schedule (MTS). Our work focuses upon a maximum-likelihood alternative to their work. We propose using maximum likelihood to estimate the parameters of the triangular hazard used by Braun, Yuan, and Thall (2005) in a single administration setting. We describe how to derive estimators for the change-point and boundary parameters of the triangular hazard model and discuss their large sample properties. We also explain the challenge in extending the results from a single administration setting to a multiple schedules setting, with each schedule consisting of multiple administrations.