The door swings both ways

A molecule’s biology is climate: its sequence, its receptor, its half-life don’t change from one quarter to the next. What you can actually get is weather — it moves, sometimes fast, and almost nobody charts it. For most of this field the weather is set by two forces: who is allowed to compound a molecule, and where the raw powder comes from. In 2026 both moved at once, and here is the part worth slowing down for: they moved in opposite directions.

One door swung shut and another cracked open in the same year. The approved blockbusters — semaglutide, tirzepatide — were walled off from the compounding that the shortage years had made routine. Meanwhile a set of research peptides that had been pushed out of compounding reach got a federal committee’s nod to come partway back in. Reading the availability map now means reading both doors — and not mistaking a vote for a rule. Bullish on the science, sceptical on the paperwork.

For two years, a quirk of drug-shortage law put the biggest molecules in medicine within reach of compounders. When a drug is on the FDA shortage list, pharmacies may compound a version of it; semaglutide and tirzepatide were both in shortage as demand outran supply, and a large grey-market compounding industry grew in the gap. That window is now shut. Tirzepatide came off the shortage list in December 2024 and semaglutide in February 2025, and the enforcement grace periods for both the 503A pharmacies and 503B outsourcing facilities expired through spring 2025.

Then the FDA pressed further. In April 2026 it proposed excluding semaglutide, tirzepatide and liraglutide from the 503B bulks list outright — the step that would close even the outsourcing-facility route, on the reasoning that there is no clinical need to compound these from bulk when the approved product is available. Narrow patient-specific exceptions survive, but the direction is unambiguous: the approved GLP-1 drugs are being pushed out of compounding and back behind the pharmacy counter. For a buyer, the effect this has already had on the availability layer is real and subtractive — the easy compounded supply is going away.

The opposite motion runs through a different set of molecules. In 2023 the FDA placed several popular research peptides — BPC-157, KPV, MOTS-c, epitalon among them — into its Category 2, the bin for substances it judged to raise significant safety concerns, which in effect barred them from compounding. Compounders and prescribers pushed back; a lawsuit over the opacity of that process settled in 2024; and the settlement forced the agency to actually convene its advisory committee and hear the nominations.

That hearing was the 23–24 July 2026 meeting of the Pharmacy Compounding Advisory Committee, and it did not go the way the FDA’s own scientists asked. The committee voted — by a narrow 8–6, against the recommendation of FDA review staff — to recommend adding six peptides (BPC-157, KPV, TB-500, MOTS-c, epitalon and semax) to the 503A bulk-substances list; a seventh, emideltide, was rejected (docket FDA-2025-N-6895). Read quickly, that sounds like six peptides just became legal to compound. Read correctly, almost nothing happened yet — and the difference is the whole point.

A Pharmacy Compounding Advisory Committee recommendation is advice. It is one step in a longer machine: the committee votes, the FDA weighs that vote against its own review, and any actual change to the 503A list happens through notice-and-comment rulemaking that is slow, public, and free to land differently from the committee. Between the July vote and a final rule sits an indefinite gap in which the legal status of these peptides has not moved at all. The four that were in Category 2 still may not be compounded; the two that were never categorized remain research-use-only material.

This is the same discipline the “first human trial” headlines demand: a committee that votes 8–6 on access grounds, with — as the record shows — no new safety data in the room, is expressing a preference, not delivering a verdict. “FDA panel backs BPC-157” is a true sentence that most readers will complete incorrectly in their heads. What the vote actually changed on the availability layer, as of this writing, is nothing. What it changed about the odds is real, and worth watching — which is exactly why availability is tracked as living data rather than announced once.

While the US argues over who may compound what, the source of the powder is changing under its own regulator. Most finished, lyophilized peptide still traces back to a concentrated set of Chinese manufacturing lines, and China’s 2026 implementing regulations of its Drug Administration Law push that supply chain toward real traceability and tighter oversight. It is the upstream counterpart to the compounding fight: the origin of the material is acquiring a paper trail it did not have before.

So all three layers of the availability map moved in a single year. The approved drugs were pushed out of compounding; a slate of research peptides was pushed, by one vote, toward the edge of coming back in; and the manufacturing source began to professionalize. None of these is finished. A proposed rule can be revised, a recommendation can die in rulemaking, a traceability regime takes years to bind. The map is being redrawn in pencil, in three places at once.

The through-line of this whole series is that reachability is information the spec sheets omit, and that it changes while the biology stays still. 2026 is the year that point became impossible to miss: the peptide world is being sorted, in public, into two legal regimes pulling apart — the approved incretins herded toward pharmacy-only dispensing, the research peptides contested at the 503A line, and the whole supply chain audited from the source. A reference can publish that as data without becoming a store. What it cannot do is pretend the door has stopped moving, because in 2026 it is moving both ways at once.

• Availability moved in two opposite directions in 2026. The approved GLP-1 compounding door closed; the research-peptide 503A door cracked open. Same year, opposite motions.
• GLP-1 compounding is winding down. Tirzepatide (Dec 2024) and semaglutide (Feb 2025) came off shortage; 503A/503B grace periods expired in spring 2025; April 2026 brought a proposal to drop all three from the 503B bulks list.
• The July 2026 PCAC vote was a recommendation, not a rule. 8–6, against FDA staff, to add six peptides to the 503A bulk list (emideltide rejected). Rulemaking hasn't concluded; nothing reachable changed yet.
• A vote, a proposed rule and a delisting are three different tenses. A forecast, a strong unfired signal, and a fact that already bit. Reading all three in the present tense is the field's most common error.
• The upstream is tightening in parallel. China's 2026 drug-law reforms push the manufacturing source toward traceability — the origin of the powder getting a paper trail while the US fights over compounding.