Show your thinking: the TMF under ICH E6(R3)

ICH E6(R3) replaced the 2016 E6(R2) guideline in stages. The Principles and Annex 1 were adopted in January 2025 and took effect across the EU and UK by July 2025, with Switzerland following in August. The FDA released final guidance in September 2025. Annex 2, covering decentralized trials, pragmatic designs, and real-world data, reached final adoption on June 3, 2026, closing the last open piece of the rewrite.

On paper, this reads as restructuring: new principles, two annexes, a narrower main document. Talk to the people who live inside trial master files, and the picture changes. The TMF stops being a filing obligation and starts being evidence.

Shaun Hastings, Dawn Niccum, Sholeh Ehdaivand, and Nick Hargaden joined me at TMF Week to talk through what ICH E6(R3) is asking of the TMF.

Stop counting documents, start showing your thinking

Shaun Hastings, Director of QA at PHARMExcel, puts the shift plainly: proof of controls, proof of decisions, proof of a risk-based approach to the study. "It's about having a smarter trial master file, one that demonstrates your documents are fit for purpose and meet the risks of the trial," he said.

I still hold to this: "We're shifting away from being archivists and toward being information managers, and that makes the job more interesting, not less." The shift isn't about collecting fewer documents. The point is making sure the content inside those documents reflects what happened during the trial. The distinction matters more than most guidance documents give credit for.

Inspectors used to ask whether everything got filed and whether the file was complete. Under ICH E6(R3), the question changes: is there enough evidence filed to show your thinking. Shaun's version: "Show your thinking."

Ninety eight percent complete means nothing on its own

Dawn Niccum, Executive VP of Quality Assurance at inSeption Group, runs mock inspections for a living, and her sharpest example cuts straight to the point. Ask an organization about oversight, and they'll describe meetings, metrics, governance. Ask where the evidence lives, and the answer often points to email threads or an unvalidated SharePoint site, not the TMF.

"I don't care that your metrics say ninety eight percent completeness," she said. "It's around the quality of records versus the volume of records." For sponsors running trials through 35-40 vendors, showing proportionate oversight across every relationship depends on documenting the right records, not the most records.

Montrium’s Study Oversight and Risk Management solve that together. Instead of reviewing everything or guessing where to focus, Risk Management shows you where quality risk sits across a study. Study Oversight turns that into a targeted review, documented inside eTMF Connect, tied to the study it belongs to.

When a regulator asks how you managed risk-based quality across your studies, the answer is already in the system, not reconstructed after the fact.

Appendix C solves one problem and creates another

Sholeh Ehdaivand, Managing Partner at CRG Advisors, sums up her relationship with Appendix C in two words: love hate. The essentiality criteria give organizations a defensible, trial-specific way to decide what belongs in the TMF. The accompanying tables risk pulling everyone straight back into checklist thinking, the exact habit ICH E6(R3) tries to break.

"We end up creating documents just to collect them, when that's not what R3 is telling us to do," she said. Her fix is change management: pulling regulatory, clinical, data management, and every function with a hand in the TMF into the conversation about what evidence a trial needs.

Nick Hargaden, Senior Director of TMF at Merus, adds a wrinkle worth sitting with: essentiality at the sponsor level and essentiality at the site level aren't the same question. A record created and needed at a site doesn't automatically belong in the sponsor's eTMF too. "I'm voting for a smaller TMF, at least on the sponsor side," he said, pushing back on the instinct to treat the essentiality tables as a shopping list.

Running oversight across a multi-study portfolio raises the same question on every study: can you defend how you scoped it? Risk Management scores your artifacts by dimension so every review has a documented rationale behind it, not a guess. Run the same way on every study, that rationale is what turns a checklist into evidence.

Bring the TMF manager into the room earlier

Critical to quality factors, the tolerance limits sponsors set for a trial before writing the protocol, determine what records get created downstream. Most TMF planning happens in the final weeks before site initiation. Nick's argument: TMF managers belong in the room while those factors get defined, not after.

"By the time you get to selecting and onboarding a CRO, you already have 80% of your index, because you've been having those planning conversations," he said. This shift moves the TMF manager from archivist to something closer to a coach, sitting with each function early enough to help them identify their own evidence before the trial starts generating records.

Dawn adds a caution here. Standardizing critical to quality factors across every study defeats the purpose. "If you're not looking at the individual protocol, the individual indication, the stage you're at, you're missing the boat on critical quality," she said. Consistency across studies sounds efficient. Applied without adjustment for the trial in front of you, consistency becomes the same checklist problem in a different form.

Centralize the management, not the filing

Does ICH E6(R3) favor a centralized TMF function over a clinical-ops-led model? Sholeh's answer uses a daisy. Put the TMF at the center, with every function, clinical operations, regulatory, data management, supporting from the outside. "It's not centralizing the filing, it's centralizing the management," she said. Dawn adds the guardrail: a centralized TMF team becomes the organizer, not the sole owner of every decision about what belongs in the file. The functions still decide what evidence proves their part of the trial happened.

A smarter TMF, not a bigger one

Nick calls this a smarter TMF, a meaningful hub of insight rather than a records dump. Dawn calls this quality by design, applied to the TMF itself, not only the protocol. Sholeh points to understanding what inspectors are truly looking for, and starting the culture shift now. Shaun brings the point back to three words: show your thinking.

If you manage a TMF today, this shift is already showing up in your inbox, your inspection prep, and your conversations with clinical ops. The TMF isn't shrinking under ICH E6(R3), and isn't growing for the sake of growing. The TMF is getting smarter, earlier, and more honest about what evidence a trial needs. This is the opportunity, and the work, ahead of every TMF team this year.