FDA real-time clinical trials: the new regulatory data infrastructure layer

TL;DR

The FDA is piloting real-time clinical trials where safety signals flow to reviewers as they happen, not months later. A third-party AI platform — Paradigm Health — now sits between trial sites and FDA reviewers as a regulatory-grade data conduit. Two named sponsors, AstraZeneca and Amgen, are already in the program. A new category of infrastructure has emerged, and it has real implications for how sponsors design trials, workflows, and vendor relationships.

The FDA wants to see safety signals as they happen, not months later. For its first real-time clinical trials, the agency's review pipeline now runs through a third-party AI platform.

That's the structural shift hiding in yesterday's announcement on real-time clinical trials.

AstraZeneca's Phase 2 TRAVERSE in treatment-naïve mantle cell lymphoma has already sent validated signals to the FDA through Paradigm Health's platform. Amgen's Phase 1b STREAM-SCLC in small cell lung carcinoma is the second proof-of-concept, with final site selection still in progress. Pilot program launches this summer. RFI comments close May 29.

The architecture tells a different story

The headline reads as an FDA speed story. The architecture tells a different one. A new infrastructure layer just got inserted between sponsors and the agency.

The traditional flow has been site to sponsor, sponsor analyzes, sponsor submits, FDA reviews. Weeks of delay at every handoff, plus the hard hiatus between phases that costs the most time and capital in early development.

Real-time trials collapse that. To make it work, the FDA needed a regulatory-grade data conduit running between trial sites and reviewers. Audit trails. Patient privacy guarantees. FDA-defined signal logic running at the source. Paradigm Health has been building that infrastructure with the agency since early 2026.

Three implications worth tracking

A new category just emerged: real-time regulatory data infrastructure. Three implications worth tracking.

Continuous trials become technically feasible. If endpoints flow continuously, the wall between phases becomes optional. Adaptive designs that have been theoretical for years now have an operational backbone.

Sponsor workflows need redesigning. When FDA reviewers see predefined safety or efficacy signals as they emerge rather than at scheduled milestones, internal review cadences, escalation decisions, and amendment processes all change.

Choosing a real-time data partner becomes a regulatory matter. For real-time review, the vendor's architecture becomes part of the regulatory path.

The FDA gave a strong signal here. Two named sponsors, validated technical proof, public RFI, named CIO Jeremy Walsh on the record. This is built to scale.

Comments close May 29. For anyone in clinical development, the RFI is worth reading this week.

Key takeaways

• The FDA's real-time clinical trial pilot inserts a third-party AI platform — Paradigm Health — as a regulatory-grade data conduit between trial sites and agency reviewers.
• AstraZeneca's TRAVERSE and Amgen's STREAM-SCLC are the two named proof-of-concept programs; the pilot launches this summer.
• The traditional site → sponsor → submission → FDA review chain, with its weeks of delay at every handoff, is what real-time trials are designed to collapse.
• Continuous endpoint flow makes the wall between trial phases optional, giving adaptive designs an operational backbone they've lacked.
• When the FDA sees signals as they emerge rather than at scheduled milestones, sponsor internal review cadences, escalation decisions, and amendment processes all need to change.
• For real-time review, the choice of data vendor is no longer just a technical decision — the vendor's architecture becomes part of the regulatory path.
• RFI comments close May 29; for anyone in clinical development, it's worth reading this week.