Complex Innovative trials using adaptive, Bayesian and other novel designs can increase efficiency and reduce costs. They enhance the likelihood of identifying a true clinical benefit and accelerate patient access to good health technologies. Evolving technologies, novel statistical methods, increases in computational power, and the evolving Regulatory Framework enable and encourage Innovative Development approaches.
Sample Size Reassessment (SSR) and Adaptive Population Enrichment
Population enrichment helps to identify high responsive groups and detect treatment effect with smaller sample sizes. Population enrichment can also play a rescue role in the sense that failed molecules from one study may succeed in a subgroup, instead of in the whole population of patients. Our experience has shown that detailed trial simulations, a focus on the control of type I error and minimization of operational bias are key to regulatory success in trial design. We are experienced using different types of alpha-spending approaches for interim analyses, to suit different needs, and have developed multiplicity schemes where more than one endpoint is formally assessed within a clinical trial, including complex multiple endpoint scenarios. We also have experience in applying adaptive rules to achieve more flexible design scenarios, including multiple adaptations in confirmatory trials, for example, mixing enrichment with SSR.
We have expertise in the planning of confirmatory oncology trials where there are correlated endpoints of Objective Response Rate, Progression-Free Survival and Overall Survival. Optimizing the frequency, timing and objectives for interim and final analyses is critical to clinical development planning and to business needs.
Bayesian Methods and Historic / External Data in Confirmatory Designs
Bayesian and Hybrid confirmatory designs using historical / external data have been mainly applied in the development of medicines for rare diseases and medical devices. In this space the use of Bayesian methodology has been successful, both in the incorporation of external data (historical control arm, metadata from the literature) as informative priors and for designing adaptive trials and in early phase decision making.
In the last few years, the industry has seen a rapid rate of adoption in biomarkers and how they can be used to improve biomedical interventions. Trial investigators have been showing interest in biomarker-guided trials such as basket trials and umbrella trials, developed under the master protocol framework. As a result, we have been seeing a rapid rate of adoption of these innovative trial methods.