Peptide Retention Samples: The Material That Makes a COA Re-Testable

Peptide retention samples are the physical material a manufacturer holds back from a batch so that the same batch can be analysed again months or years later. Without retained material, a certificate of analysis is a claim that can never be independently re-checked, because everything the certificate describes has already been shipped or consumed. Regulated pharmaceutical manufacturing treats retention as a formal, auditable obligation with defined quantities and defined storage periods, and the research peptide market almost universally does not.

Every analytical argument about purity eventually runs into the same physical constraint. A purity number describes one aliquot, drawn at one moment, analysed by one laboratory, under one method. If a question about that batch surfaces eighteen months later, the only way to answer it with evidence rather than assertion is to test material from the batch itself. That requires someone to have deliberately set material aside and stored it correctly. Regulators worked this out decades ago and wrote the requirement into the three governing frameworks that cover active ingredients and finished products.

Reference sample and retention sample are two different things

EU GMP Annex 19 splits retained material into two categories that serve different purposes, and the distinction matters more than the shared vocabulary suggests. A reference sample is material from a batch of starting material, packaging material or finished product, stored specifically so that it can be analysed if a question arises during the shelf life of that batch. A retention sample is a fully packaged unit, stored for identification rather than analysis, so that presentation, labelling, batch number and expiry date can be inspected as they were released. Annex 19 notes that for finished products the two are often presented identically as fully packaged units, in which case they may be treated as interchangeable.

That split explains something that confuses buyers reading quality documentation for the first time. One retained specimen answers analytical questions and a different one answers identity and labelling questions. A vial held in a freezer proves what the material was. A sealed, labelled, fully assembled unit proves what the buyer was actually sent. Both records are needed, and neither is produced by the certificate itself.

The storage periods are longer than most people assume

ICH Q7 section 11.71 sets the requirement for active pharmaceutical ingredients. Reserve samples of each batch should be retained for one year after the expiry date assigned by the manufacturer, or for three years after distribution of the batch, whichever is longer. For an ingredient carrying a retest date rather than an expiry date, which is the normal situation for a lyophilised peptide powder, similar reserve samples should be retained for three years after the batch is completely distributed. That last clause is the one worth reading twice, because it means the clock does not start when the batch is made. It starts when the last unit leaves the building.

Annex 19 section 3.1 requires reference and retention samples from each batch of finished product to be retained for at least one year after the expiry date, with the reference sample held in its finished primary packaging or in packaging of the same material as the marketed primary container. Section 3.2 extends the obligation upstream. Samples of starting materials, other than solvents, gases and process water, are to be retained for at least two years after release of the finished product, shortened only where the documented stability period of the material is shorter.

The United States codifies the same idea in 21 CFR 211.170. Subsection (a) requires an appropriately identified reserve sample representative of each lot in each shipment of each active ingredient, retained for one year after the expiration date of the last lot of drug product containing that ingredient. Subsection (b) applies the equivalent obligation to the finished product and adds a monitoring duty that has no analogue in a certificate: reserve samples from representative lots must be examined visually at least once a year for evidence of deterioration, and any evidence found must be investigated under 21 CFR 211.192.

Key Research Findings

• ICH Q7 section 11.71 requires reserve samples of each API batch to be held for one year after the assigned expiry date or three years after distribution, whichever is longer; for retest-dated APIs, three years after the batch is completely distributed.
• ICH Q7 section 11.72 requires sufficient retained quantity to conduct at least two full compendial analyses, or two full specification analyses where no pharmacopoeial monograph exists, in packaging equivalent to or more protective than the marketed system.
• ICH Q7 section 11.70 states explicitly that the purpose of reserve samples is potential future evaluation of batch quality and not future stability testing.
• 21 CFR 211.170(a) requires a reserve sample of at least twice the quantity necessary for all tests needed to determine specification compliance, excluding sterility and pyrogen testing.
• EU GMP Annex 19 section 4.1 requires the reference sample to be large enough to carry out the full analytical controls on at least two separate occasions.
• EU GMP Annex 19 section 3.2 requires starting material samples to be retained at least two years after release of the finished product.
• The International Testing Agency concluded the London 2012 re-analysis programme on 5 December 2022 after re-testing 2,727 stored samples, producing 73 sanctioned anti-doping rule violations, 31 withdrawn medals and 46 reallocated medals across weightlifting (22), athletics (18), wrestling (4) and canoe (2).
• Ashkani et al., International Journal of Peptide Research and Therapeutics, 2024, volume 30, article 42: lyophilised six-peptide and twelve-peptide melanoma vaccine mixtures retained stability, purity and amino acid sequence identity for up to five years at minus 80 degrees Celsius, and 17 of 18 peptides remained stable for up to three months at 4 degrees Celsius or ambient temperature, the single exception showing an increase in an oxidised methionine residue.

Why the required quantity is always at least double

Both frameworks converge on the same arithmetic and for the same reason. ICH Q7 section 11.72 requires enough retained material to conduct at least two full compendial analyses, or two full specification analyses where no pharmacopoeial monograph applies, and requires storage in a packaging system equivalent to or more protective than the marketed one. Annex 19 section 4.1 requires the reference sample to permit the full analytical controls on at least two occasions, and adds that unopened packs should be used for each set of controls where necessary.

Two, not one. A single analysis consumes the evidence. If a laboratory runs the retained material once and the result is ambiguous, contested or itself suspect, there is nothing left to arbitrate with. Retaining a second full analysis worth of material is what converts a stored specimen into a genuinely disputable record. This is the physical precondition behind the entire logic of an out-of-specification investigation, where the ability to test again under controlled conditions is the difference between an investigation and an opinion.

Annex 19 section 4.4 closes a gap that is easy to overlook. It requires that all necessary analytical materials and equipment remain available, or readily obtainable, so that every test in the specification can still be performed until one year after the expiry of the last batch manufactured. Retaining material is not sufficient on its own. The capability to test it has to survive as long as the material does, which means reference standards, columns, and validated methods all have to be kept alive alongside the vials.

The EU rewrote this guidance in June 2026

On 23 June 2026 the European Commission issued decision C(2026) 4135 final, replacing the 2006 version of Annex 19, with a deadline for coming into operation of three months from the date of publication. The revision was recommended jointly by the GMP and GDP Inspectors Working Group and the PIC/S Committee, and its substantive addition is a new section 9 covering parallel imported, parallel distributed and parallel traded products.

The new section does something quietly significant. It states that reference samples of the re-packaged product are not required, but that a retention sample of the re-packaged finished product must be taken for each re-packaging operation and retained for at least one year after the expiry date, representing both primary and secondary packaging. Where a physical retention sample cannot reasonably be kept and the competent authority has agreed in advance, section 9.5 permits a photographic or digital sample instead, provided it is a complete record, permits investigation equivalent to a physical sample, allows traceability to the batch packaging record, captures batch number and expiry date, shows that safety features were applied, and complies with Annex 11 principles for record integrity.

The concession is narrow and heavily conditioned, and the conditions are the interesting part. Even when regulators allow an image to replace a physical specimen, they allow it only for the identification function, never for the analytical one. Nothing in the revision permits a photograph to substitute for a reference sample, because a photograph cannot be injected into a chromatograph.

What retained material makes possible that a certificate cannot

The clearest demonstration of retained specimen value does not come from pharmaceutical manufacturing at all. It comes from anti-doping, where storing samples for later re-testing is a deliberate deterrence strategy. The International Testing Agency concluded the London 2012 Olympic re-analysis programme on 5 December 2022, having re-tested 2,727 stored samples selected by risk assessment. The programme produced 73 sanctioned anti-doping rule violations, the withdrawal of 31 medals and the reallocation of 46 medals across weightlifting, athletics, wrestling and canoe. Most of those findings came from a long-term metabolite method that simply did not exist when the samples were first analysed in 2012.

That is the whole argument in one example. The original analysis was competent, complete and negative. A decade later, better methods applied to preserved material produced 73 different answers. The World Anti-Doping Code sets a ten year statute of limitations for exactly this reason, and the IOC subsequently committed a fund of up to five million US dollars and established a Centralised Long-Term Storage Facility in 2020 to keep the option open for future Games. The certificate was never the evidence. The frozen aliquot was.

Applied to peptide supply, the parallel is direct. Analytical capability improves. A batch certified in 2024 against the methods of 2024 cannot be re-examined with the methods of 2028 unless someone kept the material. This is also what separates a genuine quality claim from an unfalsifiable one, and it sits directly alongside the problem of confirming that a certificate resolves to the laboratory that issued it. A verifiable certificate and a retained specimen answer different halves of the same question.

Why research peptide supply sits outside this system

None of the three frameworks above binds a research chemical supplier. ICH Q7 applies to APIs intended for medicinal products. Annex 19 applies to holders of manufacturing authorisations in the EEA. 21 CFR part 211 applies to finished pharmaceuticals. Material sold for laboratory research falls outside all three, which means no legal obligation exists to retain anything, in any quantity, for any period.

The practical consequence is that in most of this market a batch becomes untestable the moment it sells out. When a researcher a year into an experimental programme sees results drift and wants to know whether the material changed, there is usually nothing left to test. This is the same structural problem that makes two lots reading 99 percent genuinely different materials, except worse, because comparability at least assumes both lots still exist somewhere.

The physics are not the obstacle. Ashkani and colleagues, publishing in the International Journal of Peptide Research and Therapeutics in 2024, reported that lyophilised six-peptide and twelve-peptide melanoma vaccine mixtures retained stability, purity and amino acid sequence identity for up to five years at minus 80 degrees Celsius, and that 17 of 18 individual peptides remained stable for up to three months at 4 degrees Celsius or ambient temperature, with the only observed change being an increase in an oxidised methionine residue. Lyophilised peptide material stored cold is entirely capable of surviving a multi-year retention period in analysable condition. What is missing is the decision to keep it, the freezer space to hold it, the traceability records Annex 19 section 2.4 requires, and the willingness to be tested against it later.

What a research buyer can actually ask for

The useful questions are narrow and answerable. Ask whether material from each batch is retained after the batch sells out, and for how long after final distribution rather than after manufacture. Ask how much is retained, measured against the quantity a full re-analysis would consume, since anything under roughly double is a single-shot answer. Ask under what conditions it is stored, because retention at ambient temperature for a compound requiring frozen storage preserves a specimen of degradation rather than of the batch. Ask whether the retained material is traceable to a specific lot number by a record a third party could audit.

Ask, finally, what happens if a buyer disputes a result. Annex 19 section 10 requires that when a manufacturer closes down, reference and retention samples are transferred to an authorised storage site so that unexpired batches on the market remain supported by retrievable evidence. No research supplier is bound by that, and the fragility of the supply base makes it worth asking who would hold the material if the supplier ceased trading. The batch certificates published by Maple Research Labs document what a laboratory measured on a specific lot. Understanding what retention adds is understanding what a certificate structurally cannot do on its own.

Limitations and open questions

Three caveats belong on this analysis. First, the regulatory citations describe obligations attaching to medicinal product manufacture, and importing them wholesale into a research supply context is an analogy, not a legal requirement. They are useful as a benchmark for what a defensible retention practice looks like, not as a compliance standard anyone in this market is failing to meet.

Second, ICH Q7 section 11.70 draws a boundary that is frequently misread. It states that reserve samples exist for potential future evaluation of batch quality and explicitly not for future stability testing. A retained sample tells you what the batch was, evaluated under conditions chosen to preserve it. It does not tell you what happened to the units that shipped, which travelled a different thermal history in a different container.

Third, the anti-doping comparison has a real limit. Doping control samples are collected under a formal chain of custody with a witnessed split into A and B specimens, which is a far stronger evidentiary structure than a manufacturer voluntarily setting aside a few vials. The 2,727 sample re-analysis programme demonstrates the value of preserved material, not that any voluntary retention scheme would carry equivalent evidentiary weight.

Retained material matters most when the release document itself carries no numbers. The difference between a certificate of conformance and a certificate of analysis is whether measured results accompany the release statement or the supplier is attesting to conformity without reporting them.

Research use statement

All materials discussed are supplied strictly for laboratory research. For research purposes only. Not for human consumption. Not for diagnostic or therapeutic use.