Epitalon (Epithalon)

Identity, and the extract behind it

Epitalon is Ala-Glu-Asp-Gly: four residues, free amino and carboxyl termini, nothing appended. PubChem holds it as CID 219042 with the molecular formula C14H22N4O9 and a molecular weight of 390.35 g/mol. The FDA Global Substance Registration System lists UNII O65P17785G and gives 307297-39-8 as the primary CAS registry number; the acetate salt, which the agency treats as a separate substance, carries 307297-40-1. A second number, 64082-79-7, circulates on catalogue pages and resolves to the same PubChem record, but it is not the registry number the FDA record treats as primary.

The peptide was not isolated from anything. It was assembled. Epithalamin is a low-molecular-weight peptide extract prepared from cattle pineal glands, and epitalon was constructed from that extract's amino acid composition, as both the FDA evaluation and the 2025 review by Araj and colleagues describe. In 2017 Khavinson and colleagues reported detecting AEDG within the tetrapeptide fraction of a pineal polypeptide complex by selective reaction monitoring, with tetrapeptides making up 22.1% of the complex by their measurement. That single mass-spectrometry identification is the entire basis for calling epitalon a naturally occurring peptide.

Naming is where the record frays. The FDA review notes that epitalon is a common name rather than a USAN or INN, that multiple salts and derivatives are sold commercially under it, and that both nomination packages submitted to the agency contained certificates of analysis whose titles named one substance while the molecular weight and formula on the same document described another. The agency concluded that neither the free base nor the acetate is well characterised for injectable use, citing absent or inadequate data on peptide-related impurities, aggregates, bioburden and bacterial endotoxins.

The telomerase result, and what it reported

Two short papers carry the claim that made this compound. In 2003, in Bulletin of Experimental Biology and Medicine, Khavinson, Bondarev and Butyugov reported that adding the peptide to telomerase-negative human fetal fibroblast culture induced expression of the telomerase catalytic subunit, enzymatic activity of the enzyme, and telomere elongation. The abstract gives no concentration, no sample size and no effect size. A year later the same group reported that primary pulmonary fibroblasts from a 24-week fetus, which stopped dividing at passage 34 in control, reached passage 44 under treatment and were still dividing.

A later study from the same laboratory measured telomeres in cells taken from named people rather than one fetal line. Khavinson and colleagues incubated phytohaemagglutinin-stimulated blood lymphocytes from five men aged 18 to 22 and six aged 49 to 54, and measured relative telomere length by fluorescence in situ hybridisation. Length changed significantly in seven of eleven donors. It rose in five, by 18, 41, 55, 76 and 156 percent, and fell in two, by 15 and 37 percent. The authors described the pattern as a tendency toward the group mean rather than a uniform increase.

Independent replication arrived in 2025, from a group at Brunel University London who quantified each step of the pathway separately. Messenger RNA for hTERT rose in every line tested, twelvefold in one breast cancer line. Telomerase enzyme activity did not rise significantly in either cancer line. Telomeres lengthened there anyway, by alternative lengthening of telomeres: C-circle activity rose roughly tenfold in one line and threefold in the other, corroborated by promyelocytic leukaemia body counts. In the two normal human lines the pattern inverted, with telomerase activity up four-fold and 26-fold and almost no ALT signal.

Two things belong with that paper. Biogerontology published a correction in November 2025 stating that the wrong figures had appeared as Figures 1, 2 and 3, and supplying replacements. Those are every data figure in the article. PubMed types the notice as a published erratum; there is no retraction and no expression of concern on it, and a check of publication types across the papers cited on this page returned no integrity flags anywhere else. The peptide itself was bought from a commercial research-peptide supplier rather than synthesised and characterised in house.

Rodent and insect lifespan, as published

Anisimov's group ran the survival experiments, and the shape of the result matters more than its direction. In female Swiss-derived SHR mice, 54 per group, injected subcutaneously on five consecutive days each month from three months of age until natural death, mean life span did not change. The tail did: the last decile of survivors lived 13.3% longer, and maximum life span rose 12.3%. Chromosome aberrations in bone marrow fell 17.1%. Total spontaneous tumour incidence was unchanged, though leukaemia was recorded six-fold less often than in control.

Other strains produced smaller or differently shaped effects. In FVB/N HER-2/neu transgenic mice the reported figures were 13.5% for mean and 13.9% for maximum lifetime, with survival of tumour-free animals up 34.2% and lung metastases reduced by a factor of 1.6. Fifty female CBA mice per group, on the same monthly schedule, showed a 5.3% rise in mean survival. In Drosophila melanogaster, exposure confined to the egg-to-larva stage raised imago life span 11 to 16 percent, and the increase did not scale with concentration across four orders of magnitude, a dose-independence the authors recorded and did not account for.

Light regime changed the answer entirely. Vinogradova and colleagues found no life-span effect in female rats kept under a standard day-night cycle; under the natural light of north-western Russia and under constant light, maximum life span rose by 95 and 24 days respectively. Tumour development was inhibited only in the natural-light group. One discrepancy sits inside this literature and is not resolved anywhere in it: the 2002 report on HER-2/neu mice states a subcutaneous dose of 1 mg five times weekly, while two other papers on the same strain from the same programme state 1 microgram on five consecutive days each month.

Melatonin: the finding, and the experiment that went the other way

The melatonin work is the most coherent part of the file, and the part where a direct test disagreed. Goncharova and colleagues gave old and young female rhesus monkeys a daily intramuscular injection for ten days. In the old animals evening serum melatonin rose more than four-fold and evening cortisol fell about 30%, with no significant change in morning values; in young animals melatonin did not move. The FDA summary of that study notes the investigators tested no dose-response relationship and recorded neither sleep behaviour nor electroencephalographic activity.

Djeridane, Khavinson, Anisimov and Touitou then perifused pineal glands taken from young and old male Wistar rats and applied the tetrapeptide directly across three concentrations spanning two orders of magnitude. Melatonin secretion did not change at either age. Isoproterenol-stimulated melatonin release was likewise unmodified. Their conclusion, published with Khavinson as co-author, was that the tetrapeptide does not appear to participate in the control of melatonin secretion by the rat pineal gland in vitro. Whatever produces the monkey result, this experiment argues it is not direct action on the gland.

One human study bears on the same question. Ivko and colleagues enrolled 75 healthy women aged 40 to 50, most of them night-shift workers, stratified them by urinary 6-sulfatoxymelatonin, and randomised the 40 with low excretion to a sublingual spray or a saline placebo for 20 days. Excretion rose 1.7-fold in the treated group and did not move on placebo; expression of Clock, Csnk1e and Cry2 shifted. A 2025 review reports the same result as 1.6-fold. The publication did not specify blinding, measured no sleep outcome, and reported a psycho-emotional improvement without describing how it was assessed.

What the regulatory record contains

The one regulatory footprint predates the current interest. On 2 September 2010 the FDA granted epitalon an orphan drug designation for the treatment of retinitis pigmentosa, sponsored by an Australian biotechnology company. The Office of Orphan Products Development record lists the status as designated, then withdrawn or revoked, and records no approval for the orphan indication. Designation is a rare-disease development incentive granted on a scientific rationale and a population estimate. It is not a finding of efficacy, and this one no longer stands.

A 64-page FDA evaluation of epitalon was published in 2026, prompted by two nominations to add the substance to the section 503A list that compounding pharmacies may draw on. Both nominations were withdrawn; the agency continued on its own initiative and assessed the substance for insomnia. Its findings, in its own terms: no pharmacokinetic data were found for either form; a FAERS search through 8 December 2025 retrieved no adverse event reports; the food and supplement complaint system through 5 December 2025 retrieved no cases; and no clinical study assessing safety in humans was identified.

One line in that conclusion runs against the direction of the rodent tumour literature. The agency wrote that epitalon has the potential to be carcinogenic from a mechanistic standpoint, because it has been shown to activate telomerase and lengthen telomeres, and longer telomeres are generally associated with increased cancer risk. It judged the existing mouse studies too limited in dose range, sex and duration to inform genotoxic or carcinogenic potential. The 2025 finding that the peptide switched on alternative lengthening of telomeres specifically in cancer lines sits on the same side of that argument.

Agency reviewers proposed adding neither form to the list. The advisory committee that met on 23 and 24 July 2026 voted the other way on six of the seven peptides before it, epitalon among them by the narrowest margin of the seven. A committee vote is advice rather than a decision, and the FDA had not posted minutes for that meeting when this page was compiled. Elsewhere the position is plain: epitalon appears in neither the European nor the Japanese pharmacopoeia, and the FDA review identified no approved epitalon product in the European Union, the United Kingdom, Canada, Australia or seven named European states.

Registered trials, and what circulates instead

ClinicalTrials.gov returns nothing. Searches on 17 August 2026 for epitalon, epithalon, AEDG and epithalamin, run as general terms and again against the intervention field, each returned zero registered studies. PubMed indexes exactly one publication under the Clinical Trial type: a 2002 report in Neuroendocrinology Letters on retinitis pigmentosa, whose abstract states a positive clinical effect in 90% of cases and gives no sample size, no randomisation and no description of the control. Secondary accounts put that trial at 162 patients; the figure appears nowhere in the abstract.

Most of the human longevity material attributed to epitalon was generated with a different material. Khavinson and Morozov's 2003 report of reduced mortality across 266 elderly people administered Thymalin and Epithalamin, a thymic preparation and the bovine pineal extract, not the synthetic tetrapeptide. Korkushko's twelve-year randomised study in elderly patients with coronary disease, the origin of the 28% mortality difference in circulation, likewise used epithalamine. The extract and the tetrapeptide are different substances, and no mortality or survival endpoint has been published for the tetrapeptide in humans.

Material reaches people regardless. In 2015 Vanhee and colleagues at the Belgian public-health institute reported identifying epitalon by liquid chromatography with tandem mass spectrometry in two illegal pharmaceutical preparations, lyophilised and intended for reconstitution and injection. They noted that the isomeric tetrapeptide Glu-Ala-Asp-Gly presents a counterfeiting risk that ordinary mass measurement cannot resolve, and separated the two by heat treatment. The FDA review recorded finding websites that sell epitalon products while identifying no pharmacy that compounds them.

What is not known

There is no registered human trial of this compound anywhere. ClinicalTrials.gov returns zero studies for epitalon, epithalon, AEDG and epithalamin, searched both as general terms and against the intervention field, so no dose-ranging, pharmacokinetic, safety or efficacy work in people has been registered, let alone reported. Nothing is known about absorption, distribution, metabolism or clearance in any species: FDA's 2026 review searched the standard databases and found no pharmacokinetic data at all, which means the relationship between an administered amount and any tissue concentration is unestablished. Immunogenicity has not been assessed; the agency flagged aggregation and peptide-related impurities as unquantified risks specific to the injectable route that the nominations proposed. The rodent record cannot answer the carcinogenicity question it raises, because the available studies used a single fixed dose, female animals only, and intermittent short exposure. No study has characterised reproductive or developmental effects, chronic administration, or behaviour in non-rodent species beyond two short monkey experiments. Group sizes are missing from a large share of the primary abstracts, and the great majority of the literature originates with one laboratory in St Petersburg, so most findings have never been produced by an unrelated group. Epitalon is not an approved medicine in any jurisdiction, appears in no major pharmacopoeia, and material sold outside a registered trial has not been subject to identity, purity or sterility controls.

Questions