Reading CoAs: A Provider's Guide to Peptide Quality

A Certificate of Analysis usually arrives as a one page PDF, and most of the time it gets filed without being read. That is a problem. For a compounded peptide, the CoA is the only document standing between a prescriber's signature and whatever is actually in the vial. It is also, under federal law, a condition of the compounding exemption itself, not a courtesy from the supplier.

This article is educational. It is not medical, pharmaceutical or legal advice, and it does not recommend any product, supplier or dose.

What the law actually requires

Section 503A of the Federal Food, Drug, and Cosmetic Act sets three conditions on the bulk drug substances used in compounding. The substance must comply with an applicable USP or NF monograph if one exists, or be a component of an FDA approved drug, or appear on FDA's 503A bulks list. It must be manufactured by an establishment registered under section 510, including a registered foreign establishment. And it must be "accompanied by valid certificates of analysis for each bulk drug substance." Those are clauses (i), (ii) and (iii) of 21 U.S.C. 353a(b)(1)(A). Outsourcing facilities compounding under section 503B carry the same certificate and registration conditions.

Read those three clauses together and the point becomes clear. The CoA is not evidence that a product is approved or safe. It is evidence that a specific lot of a specific substance, made by an identifiable registered manufacturer, was tested against stated specifications. Nothing more, and nothing less.

Start with provenance, not with the numbers

Before reading a single result, confirm where the document came from.

Who issued it

The CoA should name the manufacturing establishment, not only the distributor or repackager. FDA's guidance to compounders on knowing their bulks and excipients suppliers, current as of 4 February 2026, asks repackagers to clearly identify the original API manufacturer to their customers, precisely because a repackager's own registration does not satisfy the section 510 condition that attaches to the manufacturer. That same page cites a real failure: Darmerica shipped API labeled quinacrine dihydrochloride to compounders nationwide, and subsequent testing identified the material as artemisinin.

That the manufacturer is registered, and that the lot matches

Registration is searchable through FDA's establishment registration data. It is also a low bar: it confirms that an establishment filed registration and listing information, not that it passed an inspection or that this particular lot meets any standard. Then check the obvious thing that gets skipped surprisingly often. A CoA for a different lot number is not a CoA for the material in front of you.

That it is not a research document

Material labeled "research use only" is not eligible for compounding for human use, whatever the purity figure says. In a warning letter dated 1 May 2026 to Harbin Jixianglong Biotech, following an inspection of 3 to 7 November 2025, FDA rejected exactly that move. The firm proposed marking documents "only for R&D use purpose," and FDA found the response inadequate given the quantities actually shipped.

Purity is not content

This is the single most common misreading of a peptide CoA, and it is worth slowing down for.

Chromatographic purity is normally reported as HPLC area percent. It tells you what share of the peptide related material detected by the method is the target sequence. It says nothing about water, counterion or inorganic salt, because those do not absorb at the detection wavelength.

Net peptide content tells you what share of the powder in the vial is peptide at all. Lyophilised peptides purified by reversed phase HPLC are typically isolated as salts of the ion pairing acid, and they retain bound water. A lot can be 99 percent pure by HPLC and still be well under 90 percent peptide by mass, with both figures entirely honest. The review by McCarthy and colleagues in Pharmaceutical Research (2023), "Reference Standards to Support Quality of Synthetic Peptide Therapeutics", describes the mass balance approach USP uses for this: measure peptide related impurities, residual solvents, acetic acid and trifluoroacetic acid content, residue on ignition and water, then account for the remainder.

If a CoA gives you a purity figure and no net peptide content, no water determination and no counterion, you have one number out of four. Ask for the rest.

Interrogate the method, not just the result

A purity figure is only as good as the method that produced it. For HPLC, a usable CoA states the column, the detection wavelength and ideally includes the chromatogram. For identity, mass spectrometry confirming the monoisotopic mass is the baseline, and MS/MS fragmentation confirming the sequence is better. Amino acid analysis is what distinguishes leucine from isoleucine when a sequence contains both.

This is not a theoretical concern. In the Harbin Jixianglong letter, FDA found that the firm shipped tirzepatide API before completing validation of its HPLC assay and related substances method and its bacterial endotoxin test, and that the validation work it did have showed "poor resolution, overlapping peaks, and absence of structural identification." A clean looking percentage can sit on top of a method that cannot separate the impurities it is supposed to be measuring.

USP has published chapters aimed at exactly this gap. General Chapter 1503 covers quality attributes of synthetic peptide drug substances, and chapter 1504 covers the starting materials for their chemical synthesis, focusing on protected amino acid derivatives. Neither chapter sets universal impurity limits; those are justified case by case. Worth noting for anyone who learned the older framework: on 28 July 2026 FDA withdrew its May 2021 guidance on ANDAs for certain highly purified synthetic peptide drug products, saying it "no longer reflects FDA's current scientific thinking," and published revised draft product specific guidances the same day covering impurity thresholds, innate immune response testing, higher order structure and biological activity. As of 1 October 2026 a replacement general guidance has not been reissued, so the older impurity thresholds from the 2021 document should not be quoted as current policy.

For anything that will be injected

For sterile preparations, the component CoA should also carry bacterial endotoxin results against a stated limit, microbial or bioburden limits, residual solvents and elemental impurities. Endotoxin is the attribute most often missing, and it is the one that cannot be fixed downstream by sterile filtration.

For registered manufacturers and 503B outsourcing facilities, CGMP at 21 CFR 211.84 sets the rule for relying on a supplier's document at all. At least one specific identity test must be conducted in house on each component, and the reliability of the supplier's analyses must be established through appropriate validation of the supplier's test results at appropriate intervals. State licensed 503A pharmacies are exempt from CGMP under section 503A and are instead governed by USP chapters 795 and 797 through state law, but the underlying logic is the same. A supplier's CoA is the starting point for a quality decision, never the whole of it.

FDA has cited that failure directly. In a warning letter dated 8 December 2025 to Darmerica, LLC, following a 3 to 19 March 2025 inspection, the agency found the firm released APIs for distribution before its quality unit completed review, including before the certificate of analysis had been reviewed and authorised by its third party testing laboratory. The letter also cited requalification of a supplier whose vendor survey answers about inspection history conflicted and went uninvestigated.

Red flags worth an email before an order

• No lot number, or a lot number that does not match the container.
• No manufacturer named, only a distributor or repackager.
• A purity figure with no method, no wavelength and no chromatogram.
• No net peptide content, no water content and no counterion identified.
• No endotoxin result on material destined for an injectable.
• A "research use only" statement anywhere on the document.
• A CoA that is an image or a retyped table rather than an issued, signed document from the testing laboratory.
• Any discrepancy between the CoA and the container label. That is the quinacrine and artemisinin scenario.

The regulatory ground is still moving

Two things are true at once as of 1 October 2026, and conflating them causes real problems.

First, FDA has been reorganising its nominated bulk substances categories. Its page on certain bulk drug substances that may present significant safety risks now also tracks substances previously in category 2 whose nominations were withdrawn, and several peptides moved into that withdrawn group during April and May 2026. Second, the Pharmacy Compounding Advisory Committee met on 23 and 24 July 2026 to discuss BPC-157, KPV, TB-500, MOTS-c, emideltide, Semax and Epitalon related substances for possible inclusion on the 503A bulks list. FDA proposed that BPC-157 free base and BPC-157 acetate not be included.

Neither development makes any peptide eligible to compound. Removal from category 2 is not addition to the 503A bulks list, and addition to that list requires notice and comment rulemaking that has not occurred for these substances. Committee recommendations are advisory, and FDA is not bound by them. Any claim that a peptide is "now legal to compound" because of a 2026 list change should be checked against FDA's own 503A bulks pages on the date you are reading it, not against a summary.

Where PEPTPlus and its pharmacies fit

PEPTPlus is a prescribing and fulfilment platform. It does not manufacture API, it does not test lots, and it does not make any compounded preparation FDA approved. What the platform does is route a prescription to a licensed compounding pharmacy and keep the record of that transaction intact and auditable.

The document review described above belongs to the pharmacy's quality function, and the eligibility question belongs to the prescriber and the pharmacy together. If you prescribe through PEPTPlus and want to see how a partner pharmacy qualifies a bulk supplier, which attributes it requires on an incoming CoA, or what it tests on receipt, ask. A pharmacy that cannot answer those questions in writing is telling you something.

If you would like to walk through how sourcing documentation flows on the platform before you write anything, get in touch and we will take you through it.