Parametric Dose Standardization for Optimizing Two-Agent Combinations in a Phase I-II Trial with Ordinal Outcomes

Area of Science:

• * Oncology
• * Biostatistics
• * Clinical Trial Design

Background:

• * Phase I-II clinical trials require optimizing doses for multiple agents.
• * Solid tumors present challenges due to complex efficacy and toxicity profiles.
• * Ordinal outcomes (four-level) complicate dose-response modeling.

Purpose of the Study:

• * To develop a Bayesian model for jointly optimizing doses of targeted agents and chemotherapy.
• * To address challenges posed by ordinal efficacy and toxicity outcomes.
• * To provide a robust and reliable method for dose pair selection in solid tumor trials.

Main Methods:

• * Employed a Bayesian framework with generalized continuation ratio models for marginal distributions.
• * Utilized a copula to model the bivariate distribution of efficacy and toxicity.
• * Incorporated elicited outcome probabilities for prior construction and numerical utilities for posterior mean criteria.
• * Implemented adaptive randomization to avoid suboptimal dose pair selection.

Main Results:

• * Parametric dose standardization with additive dose effects demonstrated robustness and reliability.
• * The proposed model compared favorably against designs with dose-dose interaction terms.
• * Simulation studies validated the model's effectiveness in dose pair optimization.

Conclusions:

• * The developed Bayesian model and design effectively optimize dual-agent dosing in phase I-II trials.
• * The methodology is suitable for clinical trial settings with similar dose and outcome structures.
• * Offers a reliable alternative to existing models, particularly those including complex interaction terms.