The Regulatory Dossier and What It Must Prove

Part of Regulatory Affairs and the Jurisdiction Map: What Regulatory Affairs Owns.

About 6 minutes. Free, in your language, no account needed.

Why this matters

A clinical trial application is not a form. It is an argument, and like any argument it can be sound or unsound independently of how neatly it is presented.

The argument is this: there is enough evidence that this product may be given to these people, under these conditions, with these safeguards, that the risk is justified by what will be learned. Every document in the dossier exists to support one clause of that sentence. A reviewer who cannot find the support for a clause will ask for it, and the clock stops while you answer.

Understanding the dossier as an argument rather than a checklist is what separates a submission that is authorized in one round from one that takes three.

The core components

Names and structures vary, but the substance is remarkably consistent across jurisdictions.

The protocol. What will be done, to whom, and why. It carries the eligibility criteria, the schedule of assessments, the endpoints, the stopping rules, and the statistical plan. It is the document everything else is measured against.

The Investigator's Brochure. Everything known about the product so far: chemistry, non-clinical pharmacology and toxicology, and any human experience to date. This is the evidence that the first clause of the argument rests on. In the United States, where the application is an Investigational New Drug submission, this sits alongside chemistry, manufacturing and controls data and the non-clinical study reports.

The informed consent documents. What a participant will be told, in the language they speak, at a reading level they can use. Ethics committees read this more closely than any other document, and it is the single most common cause of a conditional approval.

The investigational medicinal product dossier. How the product is made, what is in it, how it is labeled, and how its quality is controlled. A comparator and a placebo need this too, which teams routinely forget.

Site and investigator documentation. Who will conduct the trial, their qualifications, and the suitability of their facilities. In the United States this is formalized in Form FDA 1572, on which an investigator makes specific commitments and signs them.

Insurance, indemnity and the trial agreements. Who bears the risk if a participant is harmed. Several jurisdictions will not authorize without evidence of cover that meets a national minimum.

What "sufficient" means, and why it varies

The same dossier is not sufficient everywhere, and the differences are not arbitrary.

A country whose authority has limited review capacity may lean on the decisions of others, authorizing more readily where a stringent authority has already reviewed the same product. A country with a history of trials conducted badly within its borders may require more, including local safety data, a local sponsor representative, or evidence that the product will be made available after the trial ends.

The World Health Organization's Global Benchmarking Tool gives a vocabulary for this. An authority assessed at maturity level 3 has a stable, well functioning regulatory system; level 4 indicates one operating at an advanced level. CRLN's register records these with their scope and date, because the scope matters: an authority benchmarked at maturity level 3 for vaccines has not been benchmarked for medicines, and quoting the level without its scope overstates what was assessed.

In practice

A sponsor submits the same molecule to four countries in the same month.

In the first, a European Union member state, the submission goes through the shared portal with the full Part I package, and a single coordinated assessment covers seven member states at once. In the second, the United States, the material becomes an Investigational New Drug application, and the sponsor may proceed after thirty days unless the agency places the study on clinical hold. That default is genuinely unusual: in most of the world, silence means nothing has happened, while here silence means proceed.

In the third, an authority with more limited capacity, the reviewer asks for the assessment report from one of the first two and grants authorization substantially on the strength of it.

In the fourth, the authority asks two questions nobody anticipated: what happens to participants who benefit from the product when the trial ends, and who in this country is legally answerable if something goes wrong. Neither question is in the dossier template. Both are reasonable. Answering them takes five weeks because the answers require decisions no one at the sponsor has authority to make alone.

The regulatory lead's contribution was not the filing. It was knowing in advance that the fourth country asks those two questions, and putting them in front of the sponsor in month one.

What this looks like when it goes wrong

The dossier is complete and the argument is unsound.

The protocol's eligibility criteria permit enrolling participants with a degree of renal impairment that the Investigator's Brochure's toxicology section does not support. Each document is correct on its own terms. Nobody read them against each other.

A good reviewer catches this, which costs the sponsor a review cycle and some credibility. A reviewer under pressure may not, which is far worse, because the trial then proceeds with an eligibility criterion that its own safety data does not justify.

The check that prevents this is not a completeness check. Completeness would pass. It is reading the protocol's risk claims against the evidence the dossier actually contains, and that is regulatory affairs' job precisely because nobody else reads all of these documents.

Key takeaways