The Guideline That Took Four Years to Write

Start with the boundary, because it is the thing most often lost. The document at the centre of this story is a medicines-manufacturing guideline, written for companies that develop and make synthetic peptides as medicinal products in Europe. It is not a regulation of research chemicals, it does not govern what a laboratory supplier may sell, and it says nothing about use. It took effect on 1 June 2026, and it exists because the people who make peptide medicines had spent years working in a gap that the older rule-books never quite covered 1.

On that date a document with a long reference number, EMA/CHMP/CVMP/QWP/367182/2025, became the formal European expectation for how a synthetic peptide must be developed, made, characterised and controlled before a regulator will accept it. The numbers inside the reference tell a small story of their own. The last group is the year of adoption. The earlier ones belong to committees: the human medicines committee, the veterinary one, and the quality working party that did the drafting. Four years of work are folded into that string.

A reference number, a date and a gap in the rules

Regulators write guidelines when a recurring question has no home. For most of the twentieth century the question of what a peptide medicine had to prove about itself was answered by borrowing. A short peptide made by chemical synthesis looked, to a quality assessor, rather like a small molecule: it was built step by step from defined starting materials, it was purified, and it was dried. A long one, or one made in a bacterium or yeast cell, looked more like a protein, with all the questions of folding and host-cell residue that come with biology.

Neither analogy was comfortable. A peptide of thirty residues assembled on a resin is a chemical product with the impurity profile of a biological one: dozens of near-identical relatives, differing by one missing or altered residue, which cannot all be removed and which are hard to tell apart from the intended molecule. The scientific literature spent years describing that problem and proposing how specifications for such products might be framed 24. A regulator reading it had a choice between stretching rules that were written for something else and writing a new one.

Why peptides fell between two sets of rules

The gap widened for a commercial reason. As synthetic manufacturing became cheaper and more reliable, the range of peptide medicines grew, and so did the number of companies preparing the active substance rather than only the finished product. Reviews of the manufacturing field in the late 2010s describe an industry in which solid-phase synthesis, solution-phase methods and hybrid routes all competed, with large differences in how each handled purification, and a growing base of contract manufacturers supplying companies that did not make the material themselves 3.

Every one of those routes leaves its own fingerprint. A chain built on a resin accumulates deletion products and truncated chains. A chain joined from fragments risks different by-products at the joins. A product cut from a resin with strong acid carries the acid's counter-ion with it, and the salt form matters to everything downstream. An assessor comparing two dossiers for what looked like the same peptide could be looking at two materially different mixtures, with no agreed vocabulary for saying how different.

2022: the concept paper

European guideline-making is deliberately slow and deliberately public. It begins with a concept paper, a short document that says in effect: there is a problem, here is why existing text does not solve it, and here is what we propose to write. The agency's own record shows the concept paper for synthetic peptides, reference EMA/CHMP/QWP/735422/2022, first published on 20 September 2022, with a consultation that ran until 20 December 2022 1.

The three-month window is the first point at which outsiders can push back, and it is easy to underestimate how much that matters. Manufacturers, trade bodies, academic groups and national authorities can all say whether the proposed scope is right. A concept paper that is too narrow produces a guideline that misses the products it was meant to catch. One that is too broad pulls in material the authors never considered. The decisions made at that stage, about what counts as a synthetic peptide for the purpose of the document, shaped everything that followed.

2023 to 2024: the draft and the long consultation

The draft followed thirteen months later. The agency's page records the draft guideline, reference EMA/CHMP/CVMP/QWP/387541/2023, as first published on 18 October 2023, with a consultation that ran to 30 April 2024 1. That is more than six months, twice the length of the concept stage, and the extra time is itself information. A draft with real technical content draws real technical comment.

The agency did not publish its overview of comments until 12 May 2026, under reference EMA/66185/2026, which is to say after the final guideline had been issued and shortly before it took effect 1. Readers who want to know who objected to what, and which objections changed the text, therefore have a document to read rather than a rumour to repeat. This account does not summarise that overview, because the point of publishing it is that the primary text can be read directly, and the page itself should be consulted for its contents.

What the final text covers

According to the agency's description, the guideline addresses the manufacturing process, characterisation, specifications and analytical control of synthetic peptides. It also covers conjugation, medicinal product development, development of a synthetic peptide that refers to a biologically produced peptide as its European reference medicine, and clinical trial applications. The trial-application element applies to human medicines only, whereas the guideline as a whole is issued for both human and veterinary use. Matters already handled by the agency's general guidelines on the chemistry of active substances, for human and for veterinary products, are outside its scope 1.

The detailed numerical expectations, such as how an individual impurity is to be reported, identified or qualified, are in the guideline text itself and should be read from the live page rather than from a summary that may age. What can be said safely from the scope description is its shape: it is organised around the questions a quality assessor asks of any active substance, applied to a product whose defining difficulty is that it is a mixture of closely related chains.

The generics question

One item in that scope list carries more weight than it appears to. Several of the best-known peptide medicines were first made in living cells by recombinant methods. When a company wishes to market a chemically synthesised version, the regulator must decide what evidence shows the two are the same enough, given that their impurity profiles differ in kind. The guideline includes development of synthetic peptides that use a biologically produced peptide as the European reference medicine, so the question has a written answer to start from 1.

The United States reached the same problem from its own direction and has issued its own generic-drug guidance for highly purified synthetic peptides that refer to recombinant-origin reference products. The parallel is worth noting and nothing more: the two documents belong to separate systems, and this article does not compare their thresholds. The agency's page does not say how the guideline treats generics beyond listing them as a keyword, so its practical effect on generic filings is a matter for the text.

What it does not cover

A guideline of this kind binds the people who apply for marketing authorisations and clinical trial authorisations. It describes what a dossier for a medicinal product must demonstrate. It is not a statute, it does not create offences, and it does not regulate laboratory reagents sold for research. Nothing on the agency's page addresses research use at all, and a reader should not infer from silence that research materials are either blessed or condemned by it 1.

There is nonetheless a way in which the document matters beyond its legal reach. It sets down in public the vocabulary of what it means to know what is in a peptide: how the manufacturing route is described, how the impurities are identified, how the specification is justified. Anyone reading a certificate of analysis for any peptide, from any source, can use that vocabulary as a yardstick for what a complete characterisation looks like. A certificate is not a medicine and is not judged by this guideline, but the questions are the same ones.

What this means for a researcher reading the literature

The practical lesson is about reading labels and papers with the right frame. A paper that says a peptide was purified to a stated percentage is describing one number from one method. The guideline is a reminder that regulators of peptide medicines ask for a good deal more than a single purity figure: the process, the related substances, the counter-ion and the control strategy as a whole. The research literature will not meet that standard and does not have to, but a reader should know which questions have been answered and which have not.

The second lesson is about the pace of rule-making. Four years from concept paper to effective date is not slowness for its own sake. It is the time a public consultation takes when the people being consulted are industries with strong and conflicting views about cost and risk. The manufacturers of peptide medicines had spent decades working without this text, and the guideline reads best as a field formally acknowledging that it had grown up.

The handbook and definition articles elsewhere in this network take over from here: one explains what a synthetic peptide is, and others set out how analytical confirmation of identity and purity is done in practice. The guideline itself, with its reference number and its dates, is on the agency's own page for anyone who wants the primary source.