Abstract 2389: Empowering early drug development in neoadjuvant non-small cell lung cancer studies using an innovative proof-of-concept platform trial design

Abstract

Abstract Several platform trial designs have recently been proposed in oncology. However, developing an efficient one in the neoadjuvant non-small cell lung cancer (NSCLC) is not easy due to the difficulties in patient recruitment and limited relevant data. These obstacles hinder the inclusion of a control arm, thereby losing the ability to make an informed Go/No-Go decision, and to assess prognostic and predictive biomarkers. To overcome these challenges, we have developed an innovative and efficient proof-of-concept platform trial design that aims to empower drug development in early stage of neoadjuvant NSCLC. The proposed platform trial incorporates available experimental agents, allowing for new agents to be added when available. Within the master protocol structure, a small control arm is included, only about 20% of experimental arm size, and is shared by all experimental arms for the analysis purpose. Consequently, the resulting direct comparison with the control arm enables an informed Go/No-Go decision-making, while the total sample size and budget impacts are minimized. The flexibility of Go/No-Go decision-making is further enhanced in a 3-outcome decision-making framework where a consider zone is introduced for physicians to examine the totality of data. Two innovative statistical techniques are incorporated: Bayesian dynamic borrowing for current control, and response-adaptive randomization (RAR). The former helps reduce the sample size, and the latter aims to optimize patient allocation and improve the ability to identify promising agents. Simulation results demonstrate that the direct comparison sharply reduces the false positive rate below 10%, compared to a single-arm design setting. The effective sample size of the control arm is statistically augmented through Bayesian dynamic borrowing. The patient allocation is guided based on the agent’s efficacy data with a well-calibrated RAR. The promising agent receives meaningfully more patients than if a fixed-randomization is used, while the variability in patient allocation is decreased. Compared to a traditional design, our innovative design boosts the probability of identifying promising agent by up to 2-fold. To conclude, this research develops an innovative platform trial design to improve the success of drug development in early stage NSCLC. The master protocol structure, inclusion of a small control arm, and dynamic borrowing, increase the plausibility of conducting a trial in the neoadjuvant population. Advanced statistical features optimize the patient allocation and enhance the ability to identify promising drugs. Lastly, the proposed design can be easily applied to other oncology indications with appropriate modifications. Citation Format: Zhe Chen, Phillip A. Dennis, Ji Lin. Empowering early drug development in neoadjuvant non-small cell lung cancer studies using an innovative proof-of-concept platform trial design [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2389.