Q&A: How the FDA is approaching AI in clinical trials and drug development

In the last year, pharma incumbents have leaned into partnerships with AI-driven startups to supercharge drug discovery efforts. High-powered models of protein structure and molecule binding can help identify new disease targets, and even help design new drugs. But AI-derived or not, a new drug candidate still has to wind its way through the painstaking clinical trial process. That is why the pharma industry is also leaning into AI as a tool to make its clinical research more efficient — and the Food and Drug Administration is taking steps to clarify its regulatory approach to those trials. 

AI models can predict which kinds of patients will respond the best to a drug, helping to design evidence-based inclusion criteria or optimize doses. AI-derived biomarkers are increasingly proposed as endpoints in human studies. And some groups are supplementing placebo groups with simulations, using models to predict what would happen to participants receiving an experimental drug if they had received the control instead. 

advertisement

Those applications and more appeared in a recent analysis of AI in FDA drug and biologic submissions through 2021. The report was co-authored by Tala Fakhouri, who now co-leads an AI Council that FDA’s Center for Drug Evaluation and Research established in late August.

STAT+ Exclusive Story

STAT+

This article is exclusive to STAT+ subscribers

Unlock this article — and get additional analysis of the technologies disrupting health care — by subscribing to STAT+.

Already have an account? Log in

View All Plans

To read the rest of this story subscribe to STAT+.

Subscribe