Skip to content
PeptideHormone

The packet and the peptide

In July 2026 the Pharmacy Compounding Advisory Committee recommended Epitalon free base for the 503A bulks list, 7–4 with one abstention, against FDA review staff. The write-ups of that day settled on a tidy line: the nomination failed to capture the record, so a good substance was judged on a bad packet. The line is half the story. A nomination, a literature, and a vote are three different objects, and Epitalon’s day mixed them. The packet named insomnia, longevity, circadian rhythm and cellular ageing, then gave the agency a reviewable question only for insomnia — and the one human study in the briefing was sublingual, twenty days, in seventy-five women, and measured a urinary melatonin metabolite rather than sleep. The literature that commenters said was missing is real and mostly one St. Petersburg network: a fibroblast telomerase assay, a mouse study that moved maximum lifespan and not the mean, a non-randomized mortality report. Completing the packet would not have turned any of those into an insomnia trial or into a carcinogenicity study. The vote applied the four-factor bulks test, a lower bar than drug approval, and it did not audit the science. A committee can find the factors close enough for a patient-specific prescription while the molecule stays on the cell-culture rung. Those are different verdicts. Bullish on reading the meeting; sceptical on treating the tally as the evidence.

12 min read · reviewed November 2026

Three objects, one afternoon

On 23–24 July 2026 the Pharmacy Compounding Advisory Committee spent part of a two-day meeting on Epitalon, the pineal tetrapeptide Ala-Glu-Asp-Gly. The committee recommended adding the free base to the 503A bulk-substances list, 7–4 with one abstention. FDA review staff had recommended against it. Accounts of the day, including a widely circulated meeting report, landed on a sentence a supporter had offered from the floor: this was a nomination-quality problem, not a substance-quality problem.

That sentence explains why the slides felt smaller than the comment file. It does not explain what anyone decided. A nomination is a packet. A literature is a set of experiments. A vote is a recommendation under a statute. Epitalon’s afternoon treated them as interchangeable, and the tidy line preserves the confusion. The useful reading separates them.

The real split

Filling in the packet would have given the committee a fuller file. It would not have converted a melatonin-metabolite study into an insomnia trial, or a theoretical telomerase concern into either a clearance or a conviction. The vote can be reasonable under the bulks standard while the science stays unaudited.

What the packet actually put in front of FDA

Nominations for Epitalon named several proposed uses: insomnia, circadian disorders, longevity, age-related disease, stress response, cellular ageing. FDA’s briefing evaluated insomnia. The agency’s reason, stated in the meeting, was procedural and specific: the nomination did not include enough information to evaluate the other uses. A use you name and do not document is not a use the statute asks the committee to grade.

The human study the briefing did discuss is easy to mis-file as the missing insomnia trial. Seventy-five women received Epitalon under the tongue for twenty days. The measured endpoint was a urinary metabolite of melatonin, which rose. During questioning, FDA confirmed that the paper also reported improvement in psycho-emotional and physical condition in about 83 percent of the Epitalon group and 25 percent of the placebo group, and said the publication did not explain how those outcomes were measured. The route was sublingual. The nomination’s product was subcutaneous. A study can be clinical and still answer a different question from the one on the form.

Commenters were right about a process fact. Translated material on human exposure, melatonin regulation, and ageing markers had been submitted, and FDA said characterization data had received a preliminary review. Members were asked to vote on a public file the agency’s slides had not fully absorbed. That is a real defect in how the day was run. It is also a defect with a boundary: an unread paper does not become stronger because it arrived late. The telomerase result in human fibroblasts, the mouse lifespan paper, and the 266-person mortality report were already public, already graded on this site, and already limited in the ways a fuller slide deck would still have had to say.

Three columns that do not answer each other
What the packet asked
  • Insomnia, subcutaneous
  • A defined bulk substance
  • A dose and a course
  • Safety at that exposure
What the literature has
  • A melatonin metabolite, sublingual
  • Telomerase, in a dish
  • Max lifespan, in mice
  • One research network
What 7–4 decided
  • Recommend the free base
  • Under the four factors
  • Against FDA staff
  • Not a listing, not an audit

The literature answers questions the form did not ask

The record commenters wanted read into the meeting is the St. Petersburg pineal program: epithalamin distilled to a tetrapeptide, a cell-culture experiment in which Epitalon induced telomerase and was associated with longer telomeres in fetal fibroblasts, mouse work in which mean lifespan did not change and maximum lifespan rose by about twelve percent with fewer of some tumors, and a non-randomized human series on pineal and thymic extracts. Almost all of it comes from one research network. None of the lifespan findings has an independent replication.

Put that record next to the nomination and the mismatch is the insight. The papers ask whether a short peptide can move a melatonin metabolite, a telomere assay, or the tail of a survival curve. A 503A nomination has to support a bulk substance a pharmacy could compound: a defined salt, a route, a dose, a duration, and a safety picture for the exposure that implies. Foreign-language studies of a different preparation, by a different route, on a different endpoint, are evidence of a research tradition. They are not a substitute dossier for the product that was nominated. Calling the gap a failure to “capture the full record” flatters both sides. The record is full of a different subject.

Telomerase is one event with two readings

FDA’s main safety question was the mechanism the brochure treats as the benefit. Telomerase rebuilds chromosome ends. It is also how many cancer cells keep dividing. The agency asked what prolonged or continuous exposure could do. The animal studies in the review did not show tumor promotion; some reported fewer or smaller spontaneous tumors and longer life, at limited doses and limited time. Public comment answered with the pattern of use people actually describe — short courses, then a stop — and with the phrase “theoretical rather than demonstrated.”

Both descriptions can be true, and neither is a study. “Theoretical” means the carcinogenic experiment at the relevant dose and duration has not been done. “Not demonstrated in these animals” means the experiments that were done, at the exposures they used, did not promote tumors. An intermittent course is a hypothesis about exposure. It is not a carcinogenicity design. The disagreement worth having is which uncertainty a 503A factor analysis is allowed to leave with a prescriber. The disagreement the floor kept scoring — benefit wins, or hazard wins — needs data the meeting did not have on either side.

Exposure counts are not a safety study

Several speakers brought retrospective pharmacy numbers. One described more than 14,000 source-identified vials and a single complaint, a cloudy vial, which is a quality observation. Others cited records covering more than 200,000 doses with no reported adverse event and no reported immunogenicity reaction. FDA’s reply was the distinction the phrase “real-world evidence” keeps swallowing: a pile of doses is data, and data become evidence when you can say who received them, for what, at what dose, for how long, from which product, and whether anyone was looking for harm. Absence from a system that does not collect events is not a measured rate of zero.

The same hour produced the opposite quality story, and it is the one the nomination debate usually drops. A domestic manufacturer reported testing research-market Epitalon: samples missing potency, purity under 80 percent in some of them, residual trifluoroacetic acid as high as 14 milligrams per gram. That is an argument about unregulated supply. It is not an argument that the tetrapeptide is toxic, and it is not an argument that listing the substance creates a single clean manufacturer. FDA’s line on that last point was blunt, and it is the structural one: once a substance is on the bulks list, the agency cannot confine it to one firm, one route of synthesis, or one certificate of analysis. A capable analytical method — four amino acids, mass spectrometry, chromatography — shows the molecule can be specified. The statute, after a listing, does not require that every compounder meet the best specification in the room.

The nominations were also inconsistent about which substance they had named, free base or acetate, and the certificates of analysis did not line up cleanly with either. Missing impurity, aggregation, bioburden, endotoxin, and injectable-quality information is a characterization gap. It sits under a different factor from “does the peptide lengthen telomeres.” A meeting that argues only the second factor will always feel, to people holding the first, like the wrong trial.

What 7–4 is allowed to mean

Members who voted yes described the standard they thought they were applying. The bulks list is governed by four factors — chemistry, safety, evidence of effectiveness for the proposed use, and historical use — under section 503A. Several said the room kept sliding toward the burden of an investigational new drug application. Pharmacist members talked about a patient-specific prescription and a named pharmacist, which is the compounding model, as against a product made for an unnamed market. One yes vote said outright that the file in front of the committee and the file in the slides were not the same file. The National Association of Boards of Pharmacy abstained, its usual position of neither supporting nor opposing peptide additions.

The no votes named characterization, the telomerase question, and the absence of a dose and a frequency a prescriber could actually use. One member called it a significant lack of information both to evaluate the product and to know how to prescribe it. That objection is not a demand for Phase 3. It is a demand that “patient-specific” still have a regimen to be specific about.

Hold the tally next to the rest of that July meeting. Six peptides were recommended; a seventh was not; staff opposed the set; nothing on the list changed, because a committee advises and a rule decides. Epitalon’s 7–4 is one cell in that table. Read as a forecast of rulemaking, it is a signal. Read as a verdict that the longevity claim survived contact with evidence, it is the wrong document.

  • The packet. Nominated uses included insomnia and longevity. FDA could review insomnia. The human study in the briefing was sublingual, measured a melatonin metabolite, and did not establish a sleep outcome.
  • The literature. Telomerase in fibroblasts, a mouse maximum-lifespan nudge, a non-randomized mortality series — mostly one research network, answering different questions from a bulks dossier.
  • The hazard. Telomerase activation is the proposed benefit and the theoretical cancer concern. Limited animal studies did not show promotion. They also did not test continuous exposure.
  • The exposure counts. Thousands of vials and hundreds of thousands of doses are real-world data. Without a population, a dose, and active follow-up they are not a safety rate.
  • The vote. 7–4 with one abstention recommended the free base for the 503A list, against staff, under a standard below drug approval. The recommendation is not a listing, and it is not an audit of the science.

Educational reference on a public advisory-committee meeting and the evidence discussed in it, summarized from the meeting record as reported and from the scientific literature, and simplified in places. Not legal advice, not individualized medical advice, and not a prescription. Not approved, and may not be compounded: epitalon. Status checked November 2026. A committee recommendation is not a change in regulatory status. The vote tally is taken from the 23–24 July 2026 webcast as recorded by McDermott Will & Emery. Meeting details of FDA’s briefing and public comment follow contemporaneous accounts of that session; verify them against the official transcript when you need a quotation.