FDA Peptide List Vote: What Synthesis Teams Must Document Now

What the July 2026 Vote Did — And What It Left Open

The Pharmacy Compounding Advisory Committee met on July 23 and 24, 2026 to consider seven peptides nominated for inclusion on the 503A bulk drug substances list — the affirmative list that lets state-licensed physicians and pharmacists compound using a bulk substance that has no applicable USP monograph and is not a component of an approved drug product. Under 21 CFR 216.23(c), FDA evaluates four things: physical and chemical characterization of the substance, safety issues raised by its use in compounded products, evidence of effectiveness or its absence, and historical use in compounded drugs.

FDA’s career staff released briefing documents on June 29, 2026 recommending against adding any of the seven. The committee voted the other way on six.

Vote tallies as reported by RAPS and other outlets covering the July 2026 PCAC peptide vote.

What FDA has done: convened the committee, published briefing documents, and received an advisory recommendation on six substances.

What FDA has not done: added any peptide to the 503A bulks list, authorized any compounding, or approved any product. Adding a substance to the list requires notice-and-comment rulemaking, which had not concluded. The regulatory status of all seven peptides was unchanged by the votes.

Two further facts keep this from being a closed story. FDA has announced a follow-on advisory committee meeting before the end of February 2027 to review additional peptides, reportedly including GHK-Cu, melanotan II, cathelicidin (LL-37), dihexa acetate, and pegylated mechano growth factor. And in July 2026 FDA withdrew its 2021 guidance on highly purified synthetic peptide drug products that reference rDNA-origin listed drugs, saying it no longer reflected current scientific thinking, while stating it intends to reissue it. The framework is moving, not settling.

The practical consequence for a synthesis organization: the specific FDA peptide list outcome matters far less than the evidentiary standard being applied to peptide quality data. That standard is now visible in public documents, and it is the same standard a program will face whether the eventual question is a compounding petition, an investigational new drug application, or a supplier qualification audit.

Why the Disagreement Was Mostly About Paperwork

Read FDA’s briefing materials and a pattern emerges that is easy to miss in coverage focused on the vote tallies. The agency’s objections to characterization were not arguments that the peptides could not be characterized. They were findings that the submitted packages did not characterize them.

The analysis of FDA’s briefing documents published by Bass, Berry & Sims summarizes the core finding: meaningful quality standards cannot be developed unless a substance’s chemical identity is well established. FDA staff reported that literature and commercial sources often describe substances under the same common name despite differences in amino acid sequence, or without clarifying whether the material is a free base, a salt, or an ester — and noted that different salt and ester forms of the same active moiety can behave differently in physicochemical, toxicological, and pharmacokinetic terms.

The BPC-157 case is the sharpest illustration. FDA reported encountering multiple substances marketed under that single common name containing different active moieties, and observed that the common name does not follow established chemical nomenclature conventions. There may be, in the agency’s stated view, no reliable basis for assuming a given name always identifies the same structure and form.

Three further deficiencies recurred across the set:

• Critical quality attribute testing was thin. Nominations and public information frequently lacked data on peptide impurities, aggregates, microbial quality, and bacterial endotoxins — all relevant for injectable dosage forms.

• Method detail was insufficient. FDA stated it could not evaluate how identity, renhet, and impurities were measured from the materials provided. A percentage purity figure without a named chromatographic method and the supporting chromatograms is not an assessable claim.

• Impurity work was not comparative. What FDA wanted was a full impurity profile — identification, quantification, and control across batches and sources — rather than isolated residual values.

Immunogenicity risk threaded through the safety discussion, particularly for injected peptides where impurities and aggregates are not adequately controlled.

None of those four findings requires a novel scientific method. Each of them requires records that a synthesis organization either keeps or does not.

Domain 1 — Identity: Can Your Records Prove Which Molecule You Made?

Identity is the domain FDA treated as foundational in its evaluation of 503A bulks list peptides, and the standard it applied is stricter than most certificates of analysis imply. A single mass confirmation for the intact molecule establishes molecular weight. It does not establish sequence, it does not distinguish an isomer, and it does not tell a reviewer whether the material is a free base or an acetate salt.

No single technique closes that gap. The frameworks converge on requiring orthogonal methods — separate techniques that could each independently fail, and that rest on different physical principles. The EMA synthetic-peptide guideline, for instance, expects identity to be supported by at least two orthogonal methods.

The last item is where identity documentation and process documentation meet. A sequence made by solid-phase synthesis with a particular coupling strategy has a predictable failure pattern — truncations, deletions, epimerization at specific residues. A record that names the route makes those risks assessable. A record that lists only a final mass does not.

Domain 2 — Purity: Is Your Number Reproducible by Someone Else?

A purity percentage is the most quoted and least informative figure on a typical peptide CoA. It answers one question — what fraction of the ultraviolet-absorbing material in the sample corresponds to the main peak — and it is silent on several others that determine whether the number means anything.

Method purity is not the same as peptide content. Water, counterions, residual solvents, and non-UV-absorbing contaminants sit outside what a chromatographic purity figure captures. A sample reported at 98% purity by area can contain meaningfully less peptide by mass once those components are accounted for. Reporting both, and labelling which is which, is the difference between a measurement and a marketing number.

Lot-to-lot consistency deserves separate mention because it is the claim most often asserted and least often evidenced in custom peptide synthesis documentation. Consistency is not a property of any single batch. It is a property of a data set — which means the evidence is a trend across lots against fixed criteria, not a certificate for the lot in hand. If a supplier cannot produce that trend, “lot-to-lot consistency” in a specification is a statement of intent rather than a demonstrated control.

Domain 3 — Analytical Methods: Could a Reviewer Rebuild Your Method?

FDA’s stated complaint was not that the analytical results were wrong. It was that the materials did not provide enough method-specific information to judge whether the results were reliable. That distinction matters, because it means the fix is descriptive rather than experimental.

For each method used to support a specification, a reviewer needs enough detail to understand what was measured and to form a view on whether the method can do what it claims. What a method must demonstrate before its numbers are usable is set out in ICH Q2(R2) on the validation of analytical procedures, and the chromatographic parameters an implied method must state are defined in USP <621> Chromatography.

A useful test: hand the method section to a competent analyst who has never worked on the molecule and ask them to reproduce the run. If they cannot, the section is not finished. This is a lower bar than validation and it catches the majority of real gaps.

Domain 4 — Peptide Impurity Profiles: What the FDA Peptide List Process Expects

This is where the July 2026 record is most specific and where threshold numeracy matters most.

Synthetic peptides sit in an unusual regulatory position. The EMA synthetic-peptide guideline states that synthetic peptides are fully or partially excluded from the scope of ICH Q3A and Q3B, ICH Q6A and Q6B, and ICH M7, and instead applies peptide-specific impurity thresholds anchored to the European Pharmacopoeia general monograph. FDA’s synthetic peptide guidance has treated ICH Q3A and Q3B as a general framework with peptide-specific modifications. Either way, the small-molecule impurity table does not apply unchanged, and a program should be able to state which framework it is following and why.

Reported tier details vary across secondary summaries of the July 2026 drafts, and the 2021 guidance was withdrawn with a reissue pending — so treat any single number here as framework-specific and verify the current document before you rely on it in a submission. The structural point is stable even though the numbers may shift: peptide-related impurities are governed by their own thresholds, and they are low.

What that means for the peptide impurity profile a program should hold:

The commercial argument for doing this work before anyone asks is straightforward. Impurity characterization is cheapest at the point when the chemistry is still being adjusted. Once a specification is fixed and a program has committed to a supplier, an unidentified impurity at 0.6% becomes a deviation, a retest, and possibly a reformulation of the control strategy. The peptide impurity profile FDA expects under current guidance is not new in kind; the July 2026 record simply made it unusually visible.

Domain 5 — Sourcing: Can You Walk a Batch Back to Its Amino Acids?

This is the domain that received the least public attention and carries the most unexamined risk. Compounding-list coverage focused on the peptide. FDA’s characterization questions, and the general expectations for peptide starting materials, reach all the way back to the amino acid derivatives, resins, and solvents that went into the synthesis.

USP General Chapter <1504> addresses quality attributes of starting materials for the chemical synthesis of therapeutic peptides and is intended to be used alongside the peptide drug substance chapter. It names the non-amino-acid-derivative impurity risks these materials carry: residual solvents and reagents left over from their own preparation, elemental impurities, and transmissible spongiform encephalopathy or bovine spongiform encephalopathy risk. ICH Q7 Chapter 7 sets the materials management baseline — each raw material batch recorded with manufacturer name, identity, quantity, supplier, and supplier control number, purchased against agreed specifications, and traceable into the batches where it was used.

Two practical points. First, the manufacturer of a critical material should always be known and documented — not just the distributor. Identity of the actual producer is what makes a supply risk assessable. Second, change control is the part of supplier qualification most often left verbal. A written commitment to notify before a process or site change is what converts a supplier relationship into a managed one. Programs that outsource synthesis support with a peptide CRO services partner inherit that partner’s supply chain, so the same traceability questions apply one layer further up.

Domain 6 — Intended Use: Does Your Specification Match the Route and Dose?

Here the July 2026 record offers the clearest lesson about specifications generally. Every one of the seven peptides was evaluated for a specific proposed use. The committee that reviewed BPC-157 considered ulcerative colitis. Semax was reviewed for cerebral ischemia, migraine, and trigeminal neuralgia. Emideltide was assessed for opioid withdrawal, chronic insomnia, and narcolepsy, and FDA’s stated concern included the absence of safety data for the subcutaneous route in particular.

That is not a detail of the compounding framework. It is the general principle: the specification follows the intended use, the route, and the dose — not the molecule name in isolation.

The EMA synthetic-peptide guideline makes the same point structurally. It applies to synthetic peptides used in a medicinal product, distinguishes research-use-only material from pharmaceutical material, expects the counterion to be identified with its stoichiometry, and expresses peptide content on a counterion-free, anhydrous basis within a mass-balance approach to content and purity. A material that is entirely adequate for in vitro screening can be inadequate for an injectable formulation, and the specification is where that difference lives.

The last row is deliberately uncomfortable. Being able to say “this material is qualified for X and would need Y before it could support Z” is more useful to a reviewer and a partner than a specification broad enough to permit everything and specific enough to prove nothing. It is also the record most often missing, because it requires a decision rather than a measurement. For teams designing peptide variants toward a target profile, the same logic applies one step earlier — read across the protein engineering lessons for peptide optimization to see how design intent constrains the quality attributes worth tracking from the first batch.

Scoring the Package Before Someone Else Does

FDA’s four criteria give a program something better than a checklist: they imply a scoring rule that predicts which parts of a documentation package will be challenged. The criterion that receives the most attention publicly is characterization, but the one that sinks petitions is usually the weakest link rather than the average.

The scoring discipline that matters is not adding up the good rows. It is identifying the one domain where the program would have nothing to say if asked, and closing that gap first. Work on the strong domains is comfortable and mostly adds documentation weight; work on the weak domain is what changes the outcome.

A second practical step is to convert these tables into a vendor questionnaire and answer it as though it were being asked by someone else. Most programs find that the material they can produce readily — chromatograms, mass spectra, CoAs — maps onto the domains they are already strong in, while the domains that depend on deliberate record-keeping, sourcing traceability and intended-use justification, require decisions they have not yet made. That asymmetry is predictable, and it is better found internally than during a review.

None of this depends on how the rulemaking concludes or what the February 2027 committee does with a further set of peptides. The record set described in this peptide documentation checklist is what any peptide program should be able to produce to describe its own material honestly — to a regulator, to a partner conducting due diligence, or to its own quality function during an investigation to which external review is a detail rather than the trigger.

Next steps

If any of the six domains above came back thin, the useful next move is a technical review of the specific gaps rather than a general quality claim. MOL Changes works with development teams on exactly that: walking through an existing documentation package against identity, renhet, methodological, impurity, sourcing, and intended-use expectations, and identifying what a lot-specific data package would need to include to close the gap. If that would help your program, ask for a technical feasibility assessment and we will tell you plainly what your current records support, what they do not, and what it would take to get there.