Leveraging Artificial Intelligence for Predictive Enrichment in Drug Trials: A Scoping Review Protocol with a Focus on Botanical Drugs

Clinical trials often fail because treatments do not work equally well for all patients. Predictive enrichment is an approach that aims to improve trial success by identifying patient subgroups who are more likely to benefit from a treatment. Recent advances in artificial intelligence (AI) and machine learning (ML) have made it possible to use complex clinical and real-world data to better understand these differences in treatment response. Botanical drugs, such as herbal and plant-based medicines, present special challenges for predictive enrichment because they contain multiple active components and are used in diverse patient populations. Although AI and ML methods are increasingly applied in drug development, it is unclear how these techniques have been used specifically to support predictive enrichment in botanical drug trials. This scoping review will systematically map existing studies that use AI or ML to identify treatment-responsive subgroups in drug trials, with a particular focus on botanical drugs. We will summarize the types of AI methods used, the data sources involved, and how these models are developed and applied. The expected outcomes include an overview of current research practices, identification of key knowledge gaps, and a practical framework to guide future use of AI-based predictive enrichment in botanical drug development and clinical trial design.