2D (2 Dimensional) TEQR design for Determining the optimal Dose for safety and efficacy

Area of Science:

• * Oncology drug development
• * Clinical trial design
• * Biostatistics

Background:

• * Traditional dose-finding designs often focus solely on toxicity, potentially missing optimal efficacy.
• * Advanced designs like Eff-Tox, OBD, STEIN, and TEPI incorporate both efficacy and toxicity but can be complex.
• * Frequentist and Bayesian approaches exist for dose selection, each with distinct methodologies and complexities.

Purpose of the Study:

• * To introduce and evaluate a new frequentist optimal dose selection design: the 2D TEQR design.
• * To compare the performance of the 2D TEQR design against established methods like TEPI, STEIN, Eff-Tox, and OBD Isotonic designs.
• * To assess the simplicity and accuracy of the 2D TEQR design in identifying optimal oncology drug doses.

Main Methods:

• * Proposed a novel frequentist design, the 2D TEQR (toxicity equivalence range), for optimal dose selection.
• * Based the 2D TEQR design on empirical toxicity and efficacy rates, requiring no specialized computations.
• * Conducted comparative performance analyses against TEPI, STEIN, Eff-Tox, and OBD Isotonic designs using simulated data.

Main Results:

• * The 2D TEQR design is simpler to implement than TEPI, Eff-Tox, STEIN, and OBD designs.
• * At equivalent sample and cohort sizes, the 2D TEQR design showed lower accuracy than Bayesian TEPI and STEIN designs.
• * Increasing cohort size moderately improved the 2D TEQR design's optimal dose selection accuracy.
• * The 2D TEQR design demonstrated comparable or superior accuracy to Eff-Tox and OBD Isotonic designs, except when a distinct peak in response rates favored OBD Isotonic.

Conclusions:

• * The 2D TEQR design provides a more accessible frequentist alternative for optimal oncology drug dose selection.
• * While initial accuracy may be lower than some Bayesian methods, its performance can be enhanced through adjusted cohort sizes.
• * The 2D TEQR design offers a practical balance between implementation simplicity and effective dose selection compared to existing complex designs.