Equity, Data, and Representation in Real-World Clinical Trials and Virtual Populations

Session Description

This session aims to create a space for informed discussion about how inequity in clinical trials impacts drug development/utilization and which tools can be used to reduce inequities. Despite ongoing efforts to enhance diversity in clinical trials, patients with specific ethnic/racial backgrounds, gender, age, and disabilities are still under-represented in clinical trials due to systemic barriers and complexities surrounding multiple stakeholders. The lack of clinical data in these under-represented populations subsequently raises challenges in understanding and predicting drug safety and efficacy in those populations.

Virtual populations or sets of simulated individuals designed to capture observed variability between subjects with shared characteristics (e.g., gender or age) have increasingly played a role in regulatory decisions (e.g., FDA Modernization Act 2.0). Both clinical and preclinical data can inform these virtual populations, in which datasets are integrated into a coherent framework for treatment assessment and hypothesis testing. As the use of virtual populations become more commonplace, reflecting on the scope of diversity and representation within the modeled populations is critical. We must be cognizant of how the currently limited representation in clinical trials impacts model development and consider novel methods to construct virtual populations more representative of the broader, diverse population.

The session begins with a moderator-provided introduction to health disparities and health equity in clinical trials and the impact inequity has on drug utilization and patient outcomes. Subsequent expert presentations discuss regulatory considerations for diversity in clinical trials and ongoing work/techniques to overcome these inequities. Dr. Liu will present clinical pharmacology and regulatory considerations related to equity and representation in clinical trials. Dr. Woodhead will discuss the development of a virtual population of post-menopausal women, and its application for the prediction of drug-induced liver injury. Dr. Ramanathan will discuss applications of artificial intelligence methods to assess drug dosing with special considerations for age, race/ethnicity, and disease state. The session concludes with a guided panel discussion with all the speakers.

Learning Objectives

1. Upon completion, participants will be able to describe issues of representation in clinical trials, barriers to clinical trial participation for certain populations, and impact on real-world patient care
2. Upon completion, participants will be able to summarize the types of virtual populations currently in use and relate how artificial intelligence methods may be used to generate or inform more diverse virtual populations
3. Upon completion, participants will be able to discuss potential next steps that may be taken to help address these inequities in future clinical trials and virtual populations

Session Schedule

11:00 AM - 11:09 AM MT
Introduction and Audience Polling
Kyunghee Yang, PhD and Daniel Gonzalez, PharmD, PhD

11:09 AM - 11:21 AM MT
Diversity, Equity and Clinical Pharmacology
Qi Liu, PhD

11:21 AM - 11:33 AM MT
Construction of a Simulated Population of Post-Menopausal Women for the Prediction of Drug-Induced Liver Injury
Jeffrey L. Woodhead, PhD

11:33 AM - 11:45 AM MT
Generative Adversarial Networks for Modeling the Race/Ethnicity, Age, Disease-State Dependence of Physiological Determinants of Drug Dosing
Murali Ramanathan, PhD

11:45 AM - 12:00 PM MT
Q&A and Closing Remarks