A Novel Adaptive Design Approach for Early-Phase Clinical Trials to Optimize Vaccine Dosage.

Abstract

Introduction: Vaccines are a cornerstone of global health, with their efficacy and safety dependent on appropriate dosage determination. Early-phase vaccination trials face significant challenges due to minimal toxicity and nonmonotonic dose response curves, creating a major obstacle in vaccine development. To address this gap, we propose a novel Bayesian phase I/II trial design for dose response curves exhibiting plateau or unimodal patterns to identify the optimal biological dose (OBD), effectively balancing efficacy and toxicity.

Methods: We employ a logistic dose-efficacy design that makes dose-escalation and de-escalation decisions while simultaneously considering both efficacy and safety parameters. Extensive simulation studies evaluate the performance of this design.

Results: Comparative analyses with commonly used vaccine dose-finding designs demonstrate that our method excels in identifying the optimal toxicity-efficacy trade-off, offering both simplicity and accuracy. Sensitivity analyses across various prior settings confirm the robustness and efficiency of our approach. Additionally, our design provides a user-friendly framework for clinicians, with superior operating performance compared to existing designs, particularly in terms of accuracy and robustness.

Discussion: Our innovative Bayesian design represents a significant advancement in addressing the inherent challenges of early-phase vaccination clinical trials, offering improved accuracy and efficacy in vaccine dosage determination.