Before almost every agency meeting, someone on the regulatory team is asked the same set of questions. What has FDA raised on this programme? Which of the hold issues did we answer, and in which sequence? Is anything from the last meeting still open? How did the protocol get from version 1.0 to where it is now?

All of those answers are in the filed sequences. That is the frustrating part. They exist, in the agency’s letters and in your responses, and yet in most companies they are reconstructed by hand every time — from a viewer, a folder of PDFs and a spreadsheet somebody keeps on the side.

Why the spreadsheet is never quite right

A tracking spreadsheet is a summary, and every summary is a judgement. Someone decided how to paraphrase the agency’s request. Someone decided that a response “addressed” it. Someone forgot to update it after the last information request arrived by email.

The trouble shows up at exactly the wrong moment: in a briefing document that quotes the agency slightly wrongly, or in a meeting where nobody can say for certain whether a request from two years ago was ever formally answered.

What teams actually need is not a better summary. It is the agency’s own words, side by side with the response that answered them, read directly from what was filed.

Reading the history from the record

DnXT Reviewer now builds a Clinical Development History for any application, directly from its filed sequences. Each time it is opened it shows:

  • A timeline of sequences — the submission type, the date, and what each one carried: protocols and amendments, consent forms, investigator’s brochures, meeting materials, responses, safety reports and study reports.
  • Every protocol version, with the filed redline or summary of changes attached to the version it explains.
  • Clinical holds — placed, continued and removed — in the order of the letters.
  • What the agency asked, item by item, in the agency’s own words.
  • Where it was answered — the response document, its heading and page, and the answer text.

Word for word, or not at all

Agency letters come in several shapes, and each states its requests differently. Hold letters list issues under their basis in 21 CFR 312.42(b), followed by what must be done to resolve the hold. Information-request letters number their requests after a standard opening. Meeting letters often add a list of further requests by discipline, outside the question-and-answer sections. And some information requests arrive as an email chain that the sponsor files as it was received.

Each of these is read as written. Nothing is paraphrased, and no AI summarises the letter. The full wording is always one click away.

Responses are linked by an equally strict rule. Sponsors usually restate the agency’s request before answering it — in quotation marks, in bold italic, or as a numbered block followed by a “Sponsor Response” label. A response is tied to a request only when the agency’s wording appears in it verbatim. When the answer follows under a response label, the item is marked Answered; when the wording is merely cited, for example in a regulatory history section, it is marked Quoted.

Anything that does not match exactly is left unlinked rather than guessed. That is deliberate. A tool that confidently links the wrong response to a hold issue is worse than no tool at all.

“Not quoted” is not “unresolved”

One more distinction matters. If no filed response restates a request, the history says exactly that. It does not label the request as open or unresolved, because it cannot know that — the answer may have been given in a meeting, by telephone, or in a document that paraphrased the request. Presenting it as an open issue would be a claim the record does not support.

Every fact on the page also shows where it came from: read directly from a filed document, or worked out from a document’s title and its place in the submission. A reader always knows which is which.

Who uses it

  • Regulatory leads preparing for a meeting see the whole development story in one place instead of opening thirty sequences.
  • Writers find the agency’s exact wording, and the response that addressed it, ready to cite.
  • New team members learn a programme’s history from the record rather than from whoever has been there longest.
  • Due diligence and in-licensing teams read an acquired programme’s regulatory history from its filings, not from a summary someone else wrote.

The principle underneath

Your filed submissions are the most authoritative account of your programme that exists. The work is not to summarise them better, but to make them readable — so that the answer to “what did the agency ask?” is the agency’s own sentence, and the answer to “did we respond?” is a page number.

DnXT builds eCTD publishing, submission planning, document management and dossier review software for regulatory operations teams. Book a demo to see your own application’s agency requests and responses, read straight from the filed sequences.