Multiple-dosing per cohort in phase I and phase I/II clinical trials
University of Wisconsin--Madison, 2008 - Architecture - 131 pages
A new class of dose-finding designs in phase I/II clinical trials in cancer combining both toxicity and efficacy are delineated in the second major part. Instead of trying to find a single optimal dose, I would like to identify an acceptable region in which doses are both effective and safe, along with the optimal one within the region. For each cohort of three patients, the currently estimated minimum effective dose (MED), most successful and acceptable dose (MSAD) which maximizes the probability of joint efficacy and no toxicity, and maximum tolerated dose (MTD) are each assigned to a patient. Simulation results show that the new designs exhibit good operating characteristics and have competitive performance compared with the existing designs. Confidence sets for the MED and MTD are also established to gain further assurance and to have a flexible manipulation of the acceptable region.
What people are saying - Write a review
We haven't found any reviews in the usual places.
ThreeDoseCohort Designs in Phase I Trials
The DoseFilter Designs Modeling both Toxicity and Efficacy
3 other sections not shown
0.0 Moderate-LMH-CRM 0.0 Scenario 2-sided dose-filter designs 2f.Baves 2f.Baves.indep 2f.MLE.dep 2f.MLE.indep 2f.Bayes 2f.Bayes.dep 2f.Baves.indep 2f.MLE.dep acceptable dose acceptable region assigned assumption of independence Bayes Bayes estimators clinical trials coherence cohorts of patients compromise structure confidence sets de-escalation Dose allocations dose level dose recommendation effective dose efficacy rates estimated MTD Fast-LMH-CRM independent dose-filter design isotonic regression lf.MLE likelihood based likelihood function LMH-CRM designs m-CRM m-CRM(l m-CS m-ND Moderate-LMH-CRM 0.0 modified CRM designs modified Ivanova's monotonicity MTD estimates O'Quigley O'Quigley's observed toxicity optimal dose overall toxicity parameter patients per cohort Percent of dose percent of overlapping percentiles phase I trial posterior density posterior distribution posterior probability prior probability ratio test quantiles recommended dose region and percent sample size distribution Scenarios 1-6 sequential probability ratio Simulation 1000 Slow-LMH-CRM stopping rule Table target three patients toxic response toxicity and efficacy toxicity probabilities toxicity rate true MSAD true MTD unmodified LMH-CRM vague prior