PS40: Integrating AI/ML to Advance Drug Development: Innovative Methodologies and Regulatory Perspectives

In recent years, the FDA's Center for Drug Evaluation and Research (CDER) has witnessed a significant increase in regulatory submissions incorporating artificial intelligence (AI) and machine learning (ML) methodologies. Between 2016 and 2021 1, most of these submissions-primarily through Investigational New Drug (IND) applications-focused on therapeutic areas such as oncology (27%), psychiatry (15%), gastroenterology (12%), and neurology (11%). Key objectives of these submissions included disease prognosis, treatment response prediction, pharmacometric modeling, imaging analysis, drug discovery, and endpoint and biomarker assessment. The commonly used AI/ML methodologies were decision tree-based models and deep learning approaches, particularly for analyzing imaging data. This session will feature expert speakers who have made significant advancements in applying AI/ML to drug development and will provide insights into the regulatory perspective on these innovative methodologies. Session Overview: Speaker 1: The first speaker will discuss the application of natural language processing (NLP) to automate outcome adjudication in clinical trials, with a focus on cardiovascular endpoint adjudication. The presentation will illustrate how this approach can enhance the efficiency of clinical trials, observational studies, and quality improvement initiatives while addressing the current limitations of automated adjudication systems. Speaker 2: The second speaker will explore the adoption of AI/ML methodologies in late-stage clinical development 2, emphasizing the critical importance of quality, robustness, and interpretability in regulatory discussions. The talk will also address detecting treatment-effect subgroups and interpreting complex functional relationships between variables and clinical outcomes using AI/ML. Discussants: Two discussants from the FDA and EMA will provide insights into how the inclusion of AI/ML technologies is evolving in regulatory submissions. They will share updates on efforts to develop formal guidance and best practices for reviewing submissions that incorporate AI/ML components. This session aims to deliver a comprehensive overview of the application of AI/ML methods across different phases of clinical trials. It will also address the evolving regulatory landscape, bridging the gap between innovation and regulation to ensure that AI/ML advancements are effectively and responsibly integrated into drug development.