Journal

Clinical Need: The Legal Test That Decides Compounded GLP-1s

The whole compounded GLP-1 question now turns on two threshold questions in a statute most patients have never read. Neither of them is about money.

The two doors, and why both are closing

An outsourcing facility cannot compound a drug from a bulk substance unless one of two conditions is met: the substance appears on the 503B Bulks List, or the compounded drug is on the FDA drug shortage list at the time of compounding, distribution and dispensing. That is section 503B(a)(2)(A) of the FD&C Act, and it is the whole architecture.

The shortage door has already shut. FDA determined the tirzepatide shortage resolved in December 2024 and the semaglutide shortage resolved in February 2025. The bulks-list door is the one the 1 May 2026 notice proposes to close permanently, for semaglutide, tirzepatide and liraglutide together.

This is why the current docket matters more than any single enforcement action. Warning letters address what a company said. A bulks-list determination addresses whether the category can lawfully exist at scale at all.

The two threshold questions

FDA reads the phrase “bulk drug substances for which there is a clinical need” to mean the list may include a substance if there is a clinical need for an outsourcing facility to compound the drug product, and the product must be compounded using the bulk substance rather than starting from an approved drug.

Operationally that becomes two threshold questions, asked in this order:

1. Is there a basis to conclude that an attribute of the FDA-approved product makes it medically unsuitable to treat certain patients for a condition FDA has identified for evaluation — and that the product proposed to be compounded is intended to address that attribute?
2. Is there a basis to conclude that the product must be produced from a bulk drug substance rather than from an FDA-approved drug product?

If the answer to either is no, the substance generally does not go on the list. FDA did not answer yes to both for any of the three substances, and so stopped there.

That detail is easy to skim past and it is the most consequential sentence in the notice. FDA never reached the second stage of its own analysis — the part where safety, quality and effectiveness of compounded products would have been weighed. The proposal does not rest on a finding that compounded GLP-1s are dangerous. It rests on a finding that the nominations never established why an approved product could not be used instead.

Three arguments the statute puts out of bounds

The notice is unusually direct about what does not count. Three exclusions are stated on the face of the document:

• Supply problems. FDA does not interpret supply issues, such as backorders, to be within the meaning of clinical need. The statute already handles shortages through the separate shortage-list route, so importing them into the clinical need analysis would collapse two distinct pathways into one.
• Convenience. FDA does not consider convenience in administering a particular compounded drug product — a ready-to-use presentation, for example.
• Cost. FDA does not consider the cost of the compounded drug product as compared with an FDA-approved drug product when assessing clinical need.

Read those together and the shape of the decision becomes predictable. The three things that actually drove the compounded market — it was available when brands were not, it came in easier formats, and it cost a fraction as much — are precisely the three things the test excludes.

Why the affordability argument cannot win on this docket

It is worth being blunt, because a great deal of commentary is not. If your argument for compounded semaglutide is that the approved product costs more than you can pay, that argument is true, it is serious, and it is not addressed to this question. FDA has said in terms that relative cost is not part of the clinical need assessment.

This is not the agency being callous. It follows from what the test is for. Clinical need asks whether the approved medicine is medically unsuitable for identifiable patients. A drug that works but is unaffordable is not medically unsuitable; it is inaccessible, which is a real problem with a different set of instruments — coverage policy, direct-pricing channels, manufacturer programmes, legislation.

The practical consequence for anyone reading a comment campaign, a provider email or a petition asking you to write to FDA about affordability: that comment is unlikely to move the determination, because it does not engage the standard. FDA said in the extension notice that it encourages comments addressing the clinical need standard provided by the statute and the agency’s rationale in the May notice. That is a fairly direct hint about what it considered responsive.

The factors that were never reached

Had both threshold questions been answered yes, FDA would have moved to a balancing stage weighing four factors against each other: the physical and chemical characterisation of the substance; any safety issues raised by its use in compounding; available evidence of effectiveness or lack of effectiveness of a product compounded with it; and current and historical use of the substance in compounded products, including the conditions treated and any references in peer-reviewed literature.

None of that was reached for semaglutide, tirzepatide or liraglutide. Arguments about salt forms, additives, sterility findings and adverse event reports — the material that dominates public debate — would have been weighed at that second stage. FDA noted specifically that commenters’ safety and quality arguments about the nominated substances would be addressed in Part 2 of the analysis, which it did not reach.

So a finalised exclusion would not be a verdict that compounded GLP-1s are unsafe. It would be a verdict that nobody demonstrated why they were needed. Those are different findings and they should not be reported as the same one.

What this means if you are on a compounded product now

Nothing about your prescription changed when the comment period closed on 30 July, and nothing changes until FDA publishes a final determination in the Federal Register. There is no announced timeline for that.

Two distinctions are worth holding clearly. First, this docket concerns 503B outsourcing facilities compounding from bulk substance. Patient-specific compounding under section 503A is a separate regime with separate conditions, and it is not what this notice decides. Second, an exclusion from the bulks list restricts compounding from bulk drug substance — FDA’s own reasoning notes that where it is feasible to compound starting from an approved drug product, there are benefits to doing so.

The practical steps are the ones we set out in our transition guide: find out in writing which pharmacy fills your prescription and under which section, ask whether the product is compounded from bulk substance, and price the approved alternatives now rather than during a disruption.

What would actually change the answer

A comment that identifies a specific patient population for whom a named attribute of a named approved product is medically unsuitable, and explains why the compounded product addresses that attribute, and why it must be made from bulk substance rather than from the approved drug. That is the whole target.

FDA also said it may finalise the proposed determination without change, or finalise a modification reflecting new evidence or analysis regarding clinical need. The door is not nailed shut. But it opens on evidence about medical suitability, not on volume of correspondence.

After a final determination publishes, FDA will no longer consider comments to the docket on that substance. The remaining route is a citizen petition under 21 CFR 10.30.

Frequently asked questions