Phase I trial design for drug combinations with Bayesian model averaging

Ick Hoon Jin, Lin Huo, Guosheng Yin, Ying Yuan

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Various statistical models have been proposed for two-dimensional dose finding in drug-combination trials. However, it is often a dilemma to decide which model to use when conducting a particular drug-combination trial. We make a comprehensive comparison of four dose-finding methods, and for fairness, we apply the same dose-finding algorithm under the four model structures. Through extensive simulation studies, we compare the operating characteristics of these methods in various practical scenarios. The results show that different models may lead to different design properties and that no single model performs uniformly better in all scenarios. As a result, we propose using Bayesian model averaging to overcome the arbitrariness of the model specification and enhance the robustness of the design. We assign a discrete probability mass to each model as the prior model probability and then estimate the toxicity probabilities of combined doses in the Bayesian model averaging framework. During the trial, we adaptively allocated each new cohort of patients to the most appropriate dose combination by comparing the posterior estimates of the toxicity probabilities with the prespecified toxicity target. The simulation results demonstrate that the Bayesian model averaging approach is robust under various scenarios.

Original languageEnglish
Pages (from-to)108-119
Number of pages12
JournalPharmaceutical Statistics
Volume14
Issue number2
DOIs
Publication statusPublished - 2015 Mar 1

Fingerprint

Phase I Trial
Bayesian Model Averaging
Drug Combinations
Dose Finding
Drugs
Toxicity
Scenarios
Dose
Statistical Models
Model
Model Specification
Operating Characteristics
Dilemma
Probability Model
Fairness
Estimate
Statistical Model
Assign
Design
Simulation Study

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Pharmacology
  • Pharmacology (medical)

Cite this

Jin, Ick Hoon ; Huo, Lin ; Yin, Guosheng ; Yuan, Ying. / Phase I trial design for drug combinations with Bayesian model averaging. In: Pharmaceutical Statistics. 2015 ; Vol. 14, No. 2. pp. 108-119.
@article{3e22b32f97f4478da10f532f0ffd263a,
title = "Phase I trial design for drug combinations with Bayesian model averaging",
abstract = "Various statistical models have been proposed for two-dimensional dose finding in drug-combination trials. However, it is often a dilemma to decide which model to use when conducting a particular drug-combination trial. We make a comprehensive comparison of four dose-finding methods, and for fairness, we apply the same dose-finding algorithm under the four model structures. Through extensive simulation studies, we compare the operating characteristics of these methods in various practical scenarios. The results show that different models may lead to different design properties and that no single model performs uniformly better in all scenarios. As a result, we propose using Bayesian model averaging to overcome the arbitrariness of the model specification and enhance the robustness of the design. We assign a discrete probability mass to each model as the prior model probability and then estimate the toxicity probabilities of combined doses in the Bayesian model averaging framework. During the trial, we adaptively allocated each new cohort of patients to the most appropriate dose combination by comparing the posterior estimates of the toxicity probabilities with the prespecified toxicity target. The simulation results demonstrate that the Bayesian model averaging approach is robust under various scenarios.",
author = "Jin, {Ick Hoon} and Lin Huo and Guosheng Yin and Ying Yuan",
year = "2015",
month = "3",
day = "1",
doi = "10.1002/pst.1668",
language = "English",
volume = "14",
pages = "108--119",
journal = "Pharmaceutical Statistics",
issn = "1539-1604",
publisher = "John Wiley and Sons Ltd",
number = "2",

}

Phase I trial design for drug combinations with Bayesian model averaging. / Jin, Ick Hoon; Huo, Lin; Yin, Guosheng; Yuan, Ying.

In: Pharmaceutical Statistics, Vol. 14, No. 2, 01.03.2015, p. 108-119.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Phase I trial design for drug combinations with Bayesian model averaging

AU - Jin, Ick Hoon

AU - Huo, Lin

AU - Yin, Guosheng

AU - Yuan, Ying

PY - 2015/3/1

Y1 - 2015/3/1

N2 - Various statistical models have been proposed for two-dimensional dose finding in drug-combination trials. However, it is often a dilemma to decide which model to use when conducting a particular drug-combination trial. We make a comprehensive comparison of four dose-finding methods, and for fairness, we apply the same dose-finding algorithm under the four model structures. Through extensive simulation studies, we compare the operating characteristics of these methods in various practical scenarios. The results show that different models may lead to different design properties and that no single model performs uniformly better in all scenarios. As a result, we propose using Bayesian model averaging to overcome the arbitrariness of the model specification and enhance the robustness of the design. We assign a discrete probability mass to each model as the prior model probability and then estimate the toxicity probabilities of combined doses in the Bayesian model averaging framework. During the trial, we adaptively allocated each new cohort of patients to the most appropriate dose combination by comparing the posterior estimates of the toxicity probabilities with the prespecified toxicity target. The simulation results demonstrate that the Bayesian model averaging approach is robust under various scenarios.

AB - Various statistical models have been proposed for two-dimensional dose finding in drug-combination trials. However, it is often a dilemma to decide which model to use when conducting a particular drug-combination trial. We make a comprehensive comparison of four dose-finding methods, and for fairness, we apply the same dose-finding algorithm under the four model structures. Through extensive simulation studies, we compare the operating characteristics of these methods in various practical scenarios. The results show that different models may lead to different design properties and that no single model performs uniformly better in all scenarios. As a result, we propose using Bayesian model averaging to overcome the arbitrariness of the model specification and enhance the robustness of the design. We assign a discrete probability mass to each model as the prior model probability and then estimate the toxicity probabilities of combined doses in the Bayesian model averaging framework. During the trial, we adaptively allocated each new cohort of patients to the most appropriate dose combination by comparing the posterior estimates of the toxicity probabilities with the prespecified toxicity target. The simulation results demonstrate that the Bayesian model averaging approach is robust under various scenarios.

UR - http://www.scopus.com/inward/record.url?scp=84925625817&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84925625817&partnerID=8YFLogxK

U2 - 10.1002/pst.1668

DO - 10.1002/pst.1668

M3 - Article

C2 - 25641851

AN - SCOPUS:84925625817

VL - 14

SP - 108

EP - 119

JO - Pharmaceutical Statistics

JF - Pharmaceutical Statistics

SN - 1539-1604

IS - 2

ER -