Innovative Methods for the Design and Analysis of Phase II Clinical Trials

Drug development has become increasingly time-consuming, costly, and risky in recent years. There is significant potential for improving clinical trial designs, particularly for phase II trials, which play a critical role in the drug development process. Innovative methods are especially necessary for addressing key challenges in phase II trials in terms of dose-ranging study, patient population selection, and decentralized clinical trials (DCTs). This dissertation presents a comprehensive set of methodologies that address these critical issues with three projects. The first project introduces a Bayesian adaptive dose-ranging design that integrates both efficacy and toxicity data to evaluate each dose comprehensively. The design incorporates a multi-stage, adaptive allocation process that assigns more patients to beneficial treatment arms while dropping those that are overly toxic or ineffective. By employing a utility function to balance the risk-benefit tradeoff, the model selects the minimum effective dose (MED) and the maximum utility dose (MUD). Simulation studies demonstrate that this design provides robust power and dose selection accuracy while ensuring that patients are allocated to beneficial arms, making it a highly effective tool for real-world clinical trials. The second project focuses on a seamless phase II/III trial design using Bayesian adaptive methods. This approach combines the probability of success (PoS) with the expected net present value (eNPV) to guide the selection of target patient groups. The method allows for data-driven interim decisions on whether to continue advancing a drug for specific populations, specifically as first-line (FL) or later-line (LL) patients, by balancing clinical success probabilities with financial considerations. Simulation studies show that this method offers a structured and quantitative approach to patient group selection, increasing the likelihood of advancing drugs with the promising potential for success and financial benefits. The third project explores the challenges of decentralized clinical trials (DCTs), which use both remote and onsite data collection to address the limitations of traditional site-based trials, such as recruitment difficulties and geographical barriers. To deal with the increased variability and potential bias in remote data collection, this project proposes novel methods for sample size planning and adaptive cost-optimized sample size re-estimation for DCTs. Simulations demonstrate that these methods ensure that DCTs maintain their efficiency, scalability, and statistical power while accounting for the heterogeneity between remote and onsite data. As a conclusion, these projects in this dissertation focus on innovative Bayesian adaptive designs and methodologies tailored to the modern landscape of decentralized clinical trials. By improving dose selection and patient allocation, patient group targeting, and the modeling of DCTs data, the research contributions pave the way for more efficient, safer, and more robust trials that are more patient-centric for drug developments.