PS50: Leveraging Synthetic Data and Digital Twins in Clinical Development

Synthetic data and digital twins are increasingly influential in clinical development. Synthetic data represents larger, more diverse populations, ensuring privacy and enabling extensive hypothesis generation and testing. Digital twins model virtual copies of entities like patients, utilizing data from multiple sources to create personalized models. These technologies are leveraged throughout the entire drug development process, enabling researchers to accelerate processes, reduce costs, and improve efficiency. By creating a virtual patient population or identical twins, we can enable more comprehensive analyses, optimize experimental designs and make more informed decisions before undertaking expensive clinical studies. Although challenges remain, these innovative technologies that generate privacy-compliant patient data will make a significant impact in drug and medical device development. In this session, we aim to explore the fundamentals and the current landscape of synthetic data and digital twins in the realm of drug and medical device development. The session will feature diverse perspectives from health authorities, academia and/or industry. The speaker from the FDA will provide an overview of synthetic data and digital twins, discussing what the agency observes in sponsor plans and submissions, any regulatory guidance or best practices, and future directions for this technology. The speaker from academia will cover the fundamentals and methodological considerations with synthetic Data, including a statistical perspective and motivating examples. The speaker from industry will discuss the case studies of these technologies in clinical development programs, highlighting key considerations for implementation.