Dose-finding Designs for Phase II Clinical Trials Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 20, 2019
  • Xiao, Changfu
    • Affiliation: Gillings School of Global Public Health, Department of Biostatistics
  • Most existing dose-finding designs have been proposed for Phase I oncology trials where the main outcome is toxicity, and dose escalation is guided by ethical considerations. Most of these designs have been developed assuming that dose-response curve is strictly increasing and the outcome is binary. The main outcome in Phase II non-oncology trials is efficacy, which is often continuous. A bivariate outcome combining efficacy and safety is also constantly considered in Phase II non-oncology trials. The goal of this work is to investigate the suitability of Phase I oncology designs for Phase II non-oncology trials and to develop dose-finding designs that better address the needs of Phase II non-oncology trials. Specifically, the first paper investigates which of the several known dose-finding methods is most suitable when dose response curve plateaus. Some of the designs tend to spread the allocation among the doses on the plateau, others, like the continual reassessment method and the t-statistic design, concentrate allocation at one of the doses with the t-statistic design selecting the lowest dose on the plateau more frequently. The second paper examines the optimal allocation for estimating the minimum effective and peak doses in a dose-ranging trial when the set of dose level is fixed and isotonic regression is used as a method of estimation. We propose fully sequential strategy for subject assignment. The proposed strategy includes adaptive randomization procedure to balance the allocation to placebo and active doses. We also consider estimation in presence of covariates, where randomization procedure also balances the allocation with respect to a set of known covariates. Simulations show that the new adaptive strategy is superior to equal allocation. The third paper investigates a Bayesian adaptive two-stage design to efficiently estimate the minimum effective dose or the maximum dose in a dose-finding trial where some monotonicity assumptions regarding dose-response relationship can be made. The new design allocates subjects in stage 2 according to the posterior distribution of the location of the target dose. Simulations show that the proposed two-stage design is superior to equal allocation and to two-stage strategy where only one dose is left in stage 2.
Date of publication
Resource type
Rights statement
  • In Copyright
  • "... in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Biostatistics"
  • Ivanova, Anastasia
Place of publication
  • Chapel Hill, NC
  • Open access

This work has no parents.