The Promise and Limits of Predictive Analytics in Oncology

In a recent video interview with Applied Clinical Trials, Angela Zubel, chief development officer, Debiopharm, described 2026 as an implementation year for AI and advanced analytics in drug development. She explained that many technologies had moved beyond pilot testing and were ready for broader adoption across clinical operations. Zubel highlighted opportunities to shorten timelines, reduce costs, and improve oversight through real-time monitoring, AI-supported site selection, and predictive analytics for compound prioritization. While acknowledging ongoing limitations in predictive modeling—particularly in oncology—she emphasized the importance of organizational openness to innovation. Sponsors that proactively standardized data, adopted practical AI tools, and experimented responsibly, she noted, were already seeing measurable gains in efficiency and competitiveness.

Editor's note: This transcript is a lightly edited rendering of the original audio/video content. It may contain errors, informal language, or omissions as spoken in the original recording.

ACT: How do you expect AI and advanced analytics to move from experimentation to practical impact in clinical trial operations in 2026?

Zubel: In terms of analytics and AI, faster clinical data monitoring is already achievable. Predictive modeling is improving, but accuracy remains a challenge, particularly in oncology.

The biggest unmet need is predicting which compounds will succeed in the clinic. Current predictive models are improving, especially when forecasting success from phase I to phase II or III. Many tools are already demonstrating reasonably strong accuracy in that area, which helps companies prioritize the most promising compounds.

However, predicting success from preclinical stages to clinical outcomes is much more difficult. First-in-class compounds often lack sufficient historical data, making accurate modeling challenging. While we have scientific rationale and mechanism-of-action insights, translating that into reliable clinical predictions remains complex.

In the future, I hope predictive analytics will become strong enough to reduce reliance on animal models, especially in oncology, where those models often fail to predict patient outcomes. Moving away from less predictive approaches would be an important advancement.