AI and real-world data integration: Transforming clinical trials and drug development - panel at DIA Global

DIA-panel-How-are-AI-and-real-world-data-speeding-up-drug-development.jpg
© Getty Images (Getty Images/iStockphoto)

During a panel discussion held at DIA Global in San Diego this week, which was moderated by Peter Wahl, vice president and global head of scientific affairs at CorEvitas, industry experts delved into the integration of real-world data (RWD) and advanced technologies to accelerate drug development.

The panel featured Ariel Berger, executive director of integrated solutions and real-world evidence at Evidera; John Van Hoy, executive director of data science and advanced analytics at the PPD clinical research business of Thermo Fisher Scientific; and Gino Pirri, vice president of product and technology, also with the PPD clinical research business. The conversation highlighted how AI and real-world data are transforming the landscape of clinical trials.

Disease models and trial simulators

Berger emphasized the importance of disease and trial simulators in optimizing clinical trials. “With a large amount of deep and broad global data these days, we can easily build a natural history of disease that we can convert into a disease model,” he explained. By overlaying trial parameters such as efficacy and safety, and varying these parameters, the optimal set of values for trial execution can be identified. This process, Berger noted, can be completed “before you lock the protocol, before you begin recruiting—all the expensive and time-consuming things about trial development.”

AI in drug discovery

Van Hoy highlighted AI's role in drug discovery, referencing a recent article from The New York Times. “Everyone's focused, especially in the AI space, on drug development, specifically in finding novel targets or developing stable molecules,” he said. He also mentioned the potential of AI in label expansion, where an existing drug is found to be effective for another disease, thereby providing a great return on investment. “It's almost like a two-for-one, so it's a great ROI (return on investment),” he added.

DIA-2024-panel.jpg

Patient-centric clinical studies

Pirri spoke about the significance of understanding the patient voice in clinical studies, particularly through social networks. “We own a platform called Health Unlocked, a public health network with over a million subscribers,” Pirri said. By leveraging data from social media, researchers can gain insights into patient experiences and preferences, which can lead to more effective and patient-centric clinical trials.

Optimizing trial design with AI

Discussing AI's role in optimizing trial design, Van Hoy pointed out that AI can help identify subpopulations of responders, potentially reducing the number of patients needed for a trial while maintaining its power. “If you can better target the responders, you could recruit fewer patients and still get the same sort of study power,” he noted. Van Hoy also introduced a product called Study Gauge, which uses decision science and modeling to compare different protocol design concepts, providing empirical data on patient willingness to participate.

Digital technologies and regulatory guidance

Pirri emphasized the importance of digital technologies in expanding evidence capture for efficacy and safety. He noted the FDA's recent guidance on using digital tools, highlighting the need for validation and fit-for-use. “It's not just about giving patients a bunch of devices and applications; we need to think about consumer-grade experiences to ensure the expected data capture,” he said.

AI and data management

Wahl asked about PPD's approach to AI and data management. Berger described PPD's investments in infrastructure to pull data from various sources and apply advanced analytics. “We're creating harmonization among functions, ensuring forecasts are as sophisticated and updated as possible,” he explained. Berger also mentioned partnerships with AI-focused companies to leverage disease area models for trial optimization.

AI strategies for operational efficiency

Pirri outlined three AI strategies aimed at reducing cycle times: fine-tuning language models to specific data, infusing AI into workflows to assist experts, and leveraging generative AI to connect data with systems for insightful responses. “These strategies are set to significantly reduce the time and effort required in the study start-up phase and throughout clinical trials,” he said.

The panel discussion underscored the transformative potential of AI and real-world data in drug development. From optimizing trial design to enhancing patient-centric studies, the integration of advanced technologies promises to expedite clinical trials and improve outcomes. As Ariel Berger aptly put it: “Technology is a tool. We have a lot of very cool tools we play with every day, but they shouldn't overwhelm your process and objectives.”