Draft guidance is where the rules are written, so stop waiting for the final text

The unfinished rulebook

Read this week's records as a set and a pattern emerges that has little to do with any single approval. FDA's biologics centre issued a draft guidance on how individualised genome editing products can be covered by a single BLA, including how product variants can be added after licensure. It also issued a draft on electronic submission of IDE applications. And it held a public meeting on 26 October on proposed recommendations for reauthorising the Biosimilar User Fee Act for fiscal years 2028 to 2032, with a docket open for comments.

None of these is a final rule. Each is a moment when the agency is still listening. My position is that regulatory affairs teams treat these moments as optional reading far too often. The drafts and dockets are where the practical shape of future obligations is decided, and they deserve the same ownership and deadlines as a live submission.

What the drafts actually tell us

The genome editing draft addresses a structural question: whether bespoke therapies with multiple variants must each be licensed separately, or can sit under one BLA with variants added later. Anyone developing in this space has an obvious interest in how that is worked out, and the draft is the agency's invitation to test its reasoning against real programmes.

The IDE draft is more instructive for what it leaves open. It introduces existing resources for electronic IDE submissions, but binding standards, a timetable and exemption criteria are due only when the guidance is finalised. A team that reads only the headline will conclude that electronic IDE submission is settled. A team that reads closely will see that the consequential parts are still to come, and that this is exactly when feedback is most useful.

The BsUFA process works the same way. Fee structures and performance commitments for fiscal years 2028 to 2032 are being shaped now. Agenda and slides from the meeting are posted, and the docket is open. The records do not give a comment deadline, so teams should confirm it on the docket itself.

A cautionary note on thin information

The Commission's publication of a Health Security Committee opinion on mosquito-borne diseases is a reminder from the other direction. The available information does not set out its content, scope or any obligations. Whether it matters to your portfolio is not yet known. The only responsible response is to read the primary document rather than infer from the title. That is the same discipline I am urging for drafts: go to the text, and assign someone to do it.

The strongest counter-argument

The best objection is about resources. Drafts are non-binding, they change, and some never reach final form in the shape first proposed. Small teams already stretched by labelling changes, such as the new boxed warning on Adzynma for neutralising antibodies, or by lot withdrawals like those affecting several immune globulin products, can reasonably argue that live obligations come first. Commenting on a docket produces no visible return, and effort spent tracking a draft that is later rewritten is effort lost.

That is a fair point, and it is why I am not asking for comment on everything. But the argument proves too little. The cost of a short, targeted review is small against the cost of discovering in final guidance that your submission model, variant strategy or electronic workflow was built on assumptions the agency never endorsed. Triage is the answer, not abstention.

Recommendation

Give draft guidance and open dockets a named owner and a calendar. Score each item by how directly it touches your pipeline, then decide whether to read, comment or pilot. Where a draft is directly relevant, as the genome editing draft is for any bespoke-therapy developer, test your own processes against it now and send the agency specific, evidence-based comments. Where it is not, record why you decided to ignore it.

— The VigiEye Regulatory Desk