Dose-Finding Designs for Phase I Clinical Trials in Oncology and Use of Selective Phenotypying to Increase Power of Genetic Association Studies Public Deposited

Downloadable Content

Download PDF
Last Modified
  • October 10, 2018
  • Wang, Yunfei
    • Affiliation: Gillings School of Global Public Health, Department of Biostatistics
  • The goal of phase I clinical trials in oncology is to find a dose for a study that has acceptable toxicity or adverse effect associated with a pre-specified probability in patients experiencing DLT (dose limiting toxicity) for a drug. We propose a dose-finding design for Phase I oncology trials where each new patient is assigned to the dose most likely to be the target dose given observed data. The only assumption is that the dose-toxicity curve is non-decreasing. This method is especially beneficial when it is desirable to enroll patients into a study as soon as they present for the trial. To prevent assignments to doses with limited toxicity information in fast accruing trials we propose assigning temporary fractional toxicities to patients still in follow-up. The goal of a Phase I clinical trial in oncology is to find a dose with acceptable dose limiting toxicity rate. Often when a cytostatic drug is investigated or when the maximum tolerated dose is defined using a toxicity score, the main endpoint in a Phase I trial is continuous. We propose a new method to use in a dose-finding trial with continuous endpoints. The new method performs on par with other methods and provides more flexibility in assigning patients to doses in the course of the trial when the rate of accrual is fast relative to the follow-up time. Blood-based biomarkers and other quantitative measures can provide valuable insights into disease etiology and are often used as intermediate outcomes for identifying risk factors associated with disease. Genome-wide association studies (GWAS) between quantitative traits and single-nucleotide polymorphisms (SNPs) are routinely performed on large samples from population-based cohorts. Replication studies are an important step in controlling the type I error rate of reported GWAS findings. Many potential replication cohorts have existing genome-wide SNP data but have not yet measured the quantitative trait of interest. Measuring these traits can be expensive and time consuming, which can deter studies from pursuing replication. Given the expense and time of measuring these quantitative traits on large samples, it would be desirable to identify a subset of subjects that could be phenotyped to optimize statistical power under fixed sample size constraints. We describe an approach of utilizing existing genotype data to identify an optimal subset of samples to be phenotyped and included in a genetic replication study. Specifically, we have developed a simulated annealing-based algorithm to optimally select samples to be phenotyped conditional on a list of candidate SNPs and available genotype data for those SNPs under a fixed sample size constraint. We demonstrate the increase in power of our approach relative to random sampling using simulations and a real replication study for C-reactive protein levels. Our approach is flexible enough to allow constraints on available genotype counts and differential weighting of SNPs in the power calculations.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Ivanova, Anastasia
  • Doctor of Public Health
Degree granting institution
  • University of North Carolina at Chapel Hill
Graduation year
  • 2014

This work has no parents.