Thymalin

What is actually in the vial

Thymalin is not a molecule. The state register entry filed for the product in Kazakhstan describes one vial as 10 mg of dried purified extract of the thymus of cattle, with 20 mg of glycine as excipient, presented as a lyophilised mass for reconstitution and intramuscular injection. That entry describes the raw material as polypeptide fractions from the thymus of young cattle, and states that the source herd is free of viral, prion, bacterial and mycoplasma disease. No donor age is given. Morozov and Khavinson's 1997 account in the International Journal of Immunopharmacology describes the source material as calf thymus and the method as mild acid extraction. The 2003 Khavinson and Morozov paper adds that the manufacturing process is intended to degrade macromolecular protein while retaining peptides below 10 kDa. No sequence and no single molecule defines the material.

Identity databases record that absence in an awkward way. FDA's Global Substance Registration System returns no record for the name. PubChem holds no compound record for the extract, because a compound record requires a structure and the extract has none. It does hold five substance records under the name and CAS 79621-14-0, deposited by a toxicogenomics database, ChemIDplus, the EPA's DSSTox inventory (DTXSID501052972) and two suppliers. One of those, SID 126522450, the Comparative Toxicogenomics Database record for MeSH concept C032805, carries an explicit structure link to CID 3085284, whose title is Nonathymulin: the defined nonapeptide pGlu-Ala-Lys-Ser-Gln-Gly-Gly-Ser-Asn, C33H54N12O15, 858.9 g/mol, CAS 63958-90-7, better known as thymulin. Thymalin also appears in that record's synonym list. Thymulin is a different substance, isolated by a different group in a different decade, and both the synonym list and the curated structure link carry its formula, weight and sequence onto the extract.

NLM keeps the two apart. MeSH holds thymalin as a supplementary concept with registry number 79621-14-0, mapped to Thymus Hormones and to the pharmacologic action Adjuvants, Immunologic. A neighbouring concept, thymarin, carries the note that its registry number refers to a compound with unknown molecular formula and that it may be the same as thymalin. Both entries describe an undefined mixture. A further ambiguity sits inside the primary literature: several papers from Khlystova and colleagues report immunomorphological detection of thymalin in human epidermis and fetal reticuloepithelium, so the same word names a manufactured preparation and a putative tissue antigen, and the retrieved abstracts do not state what the antibody was raised against.

The dipeptides said to be inside it

One component has been located analytically. Morozov and Khavinson reported in 1997 that L-Glu-L-Trp had been isolated from Thymalin by reversed-phase high performance liquid chromatography, and that a separate pharmaceutical, Thymogen, was then built on the synthesised dipeptide. The same paper describes Vilon differently: a novel immunomodulatory dipeptide was synthesised and given that name. Vilon is Lys-Glu. The distinction matters because later papers from the institute treat both dipeptides as constituents of the extract, when the 1997 account describes one as isolated from it and the other as designed.

Linkova and colleagues took that framing into a 2023 paper in the International Journal of Molecular Sciences, describing KE and EW as active substances of the Thymalin drug. Most of it is molecular docking against double-stranded DNA and database mining across promoter, pathway and protein-interaction repositories. The measurement is narrow and separates by agent and by condition. In mononuclear cells from four donors at 1 mg/mL, TNF-alpha under lipopolysaccharide stimulation fell 6-fold with KE, 5-fold with EW and 2.2-fold with the extract; unstimulated, it fell 4.9-fold with KE and 11.4-fold with EW, with no significant effect of the extract. IL-1 beta under stimulation fell only with the extract, 2.7-fold. IL-6 under stimulation fell only with KE, 2.5-fold. In unstimulated cells the extract increased IL-6 release about fourfold. Fold changes span 2.1 to 11.4; the statistics footnote gives n = 4 to 8, Mann-Whitney U test.

Avolio and colleagues ran a wider comparison in 2022 in the same journal, testing five preparations from the same institute, Thymalin among them, on the human monocytic THP-1 line. All five inhibited LPS-stimulated TNF and IL-6 expression in terminally differentiated cells and reduced adhesion of treated THP-1 cells to an activated endothelial monolayer. The concentrations were chosen, in the authors' phrasing, as those known for being effective on recipient cells in culture, and are not given in the abstract. Khavinson is a co-author on both papers.

The 266-person trial, read from its tables

The geroprotection claim rests on one report published twice: in Russian in Advances in Gerontology in 2002 and in English in Neuroendocrinology Letters in 2003, with the same abstract. Two centres contributed. In St Petersburg, 94 women aged 66 to 94 living in a War Veterans Home were allocated by stratification randomisation into four arms — placebo, Thymalin, Epithalamin, or both — injected intramuscularly and observed for six years. The paper states the trial was double blind. A separate group of ten women and ten men received both preparations across the same period.

Table 2 carries the mortality figures, and the denominators are small. Over six years of observation, 81.8 per cent of the 22-person placebo group had died, against 41.7 per cent of 24 on Thymalin, 45.8 per cent of 24 on Epithalamin, and 33.3 per cent of 24 on both. Baseline ages across those arms ran from 80.2 to 82.1 years. The 4.1-fold figure that circulates belongs to a fifth column: the separate 20-person group given both preparations over six years, whose mortality was 20.0 per cent against the same 81.8 per cent control. It is a combination result in twenty people, not a Thymalin result.

Kiev's arm was larger and its control group behaved differently. Institute of Gerontology researchers enrolled 152 people, 71 men and 81 women, mostly with ischaemic heart disease, again allocated by stratification randomisation against placebo, treated across three years and observed for five more. Table 5 reports eight-year mortality of 13.6 per cent in 48 controls, 6.6 per cent in 58 on Thymalin and 8.5 per cent in 46 on Epithalamin. Baseline ages sat near 70. Tables 3, 4 and 5 each carry a bracketed source attribution to the paper's references 33 and 34, so the Kiev figures are reprinted from that institute's own prior publications, which were not retrieved. Two control groups in one paper recorded 81.8 per cent mortality over six years, in institutionalised women of baseline age 80.2 years, and 13.6 per cent over eight years, in outpatients of baseline age 69.3 years. The paper offers no reconciliation of the two figures.

COVID-19: one trial, and one paper that disagrees with itself

Khavinson, Kuznik and colleagues published a single-centre open-label trial in Stem Cell Reviews and Reports in 2021. Ninety-two patients hospitalised between April and July 2020 were allocated by envelope randomisation, 42 to the extract added to the standard Russian Ministry of Health regimen and 50 to that regimen alone. Mean ages were 59.8 and 61.7 years. No deaths occurred in either group. In the treated arm IL-6 fell 5.5-fold against 1.41-fold in control, C-reactive protein 9.7-fold against 1.7-fold, and D-dimer 5.7-fold against 2.1-fold; CD3 lymphocytes rose 63.8 per cent and CD4 helpers 88.9 per cent from baseline.

A second report, from Kuznik and colleagues in Advances in Gerontology in 2022, compared the extract against tocilizumab in severe disease and contradicts itself on the primary number. The English abstract gives hospital mortality of 40.9, 28.4 and 20.6 per cent for standard therapy, tocilizumab and Thymalin. The Russian abstract on the same PubMed record gives 40.9, 28.8 and 16.2 per cent. Neither version states group sizes, and no allocation method is described in either. The difference between 20.6 and 16.2 per cent is not a rounding artefact, and no correction notice is attached to the record.

The mechanistic work cited for those clinical papers consists of one 2020 report from the same laboratory. Khavinson and colleagues reported that the extract reduced CD44 and CD117 expression two- to threefold in human haematopoietic stem cells and raised CD28 expression 6.8-fold (PMID 33237528). The PubMed Central deposit for that paper carries the abstract and reference list only; the methods section was not retrievable, so the cell source, the concentration applied and the number of replicates are not established from the published record available here.

Animal and cell work, including where it went nowhere

Zhukova and colleagues at the Rostov Research Institute of Oncology, a group with no author overlap with the St Petersburg institute, transplanted sarcoma 45 into outbred male albino rats and reported tumour growth arrest or regression in more than half of the treated animals, with growth suppressed by 78 per cent in the remainder, alongside raised lymphoproliferative activity in the thymus lobules. Group sizes are not stated in the abstract. The authors attributed the result to doses below the therapeutic range, modulated across the course, which is a claim about scheduling that the abstract does not quantify.

An older experiment recorded a null. Potapov and colleagues injected mice bearing P388 leukaemia at 0.66 mg/kg and found that repeated injection held peripheral large granular lymphocyte counts at control levels, but did not increase the life span of the leukaemic animals. That paper sits beside a 1987 report from Amosova and colleagues describing potentiation of cytostatic drugs, tumour growth inhibition and increased life span in inbred and outbred mice, with no sample sizes and no numbers in the abstract at all.

Recent work is dental and Ukrainian. Boiko and colleagues drilled a defect in the mandible of 48 mature WAG rats in four groups of twelve, and injected the soft tissue surrounding the defect, alone or with a hydroxyapatite graft. Two 2024 papers from that experiment report accelerated clearance of necrotic tissue, altered osteoblast and osteoclast counts across the stages of repair, and a shift in macrophage phenotype from M1 to M2 with rising T- and B-lymphocyte counts to day 28. Domina's review of Chornobyl cytogenetics notes a different pattern entirely: in irradiated human lymphocytes the extract acted protectively or as a sensitiser depending on concentration and radiation dose.

Where it stands with regulators

This is one of the few compounds in this category that is a registered medicine somewhere. Thymalin, transliterated Timalin, is authorised in Russia as a prescription lyophilisate for intramuscular injection under ATC code L03AX, other immunostimulants. The 2003 paper dates the original approval to USSR Ministry of Health order number 1008 of 10 November 1982 and prints the registration number as 82.1008.8; the 2021 clinical paper prints the current registration certificate as 82/1108/8. The two disagree in the third digit group and both are recorded here. A Kazakh registration under number RK-LS-5 number 018545 ran from 22 December 2011 to 22 December 2016 and has lapsed. That registration file is the source of a blunt admission: under the pharmacokinetics heading, the label reads not studied.

United States status is the opposite. No approved application exists, and in a warning letter dated 7 February 2024 the FDA cited a firm selling the substance online, finding the product to be an unapproved new drug under sections 505(a) and 301(d) of the Federal Food, Drug and Cosmetic Act and a misbranded drug under section 502. Thymalin was not among the seven substances considered at the Pharmacy Compounding Advisory Committee meeting of 23 and 24 July 2026, so it has not been evaluated for the section 503A bulk drug substances list even to the extent that epitalon and semax have.

Registered trials do not exist. Searches of ClinicalTrials.gov on 18 August 2026 for thymalin and timalin, run as general terms and again against the intervention field, each returned zero studies, while a control query on thymus extract returned 423, confirming the endpoint was live. A search of the FDA adverse event reporting system through the openFDA interface returned no matches under either spelling. The published record is also overwhelmingly one-language: PubMed indexes 293 records under the term, of which 254 are typed Russian, 30 English and 8 Ukrainian.

What is not known

No registered trial of this material exists anywhere. ClinicalTrials.gov searches on 18 August 2026 for thymalin and timalin, run as general terms and again against the intervention field, each returned zero studies, so nothing has been registered, prospectively powered or reported under an accountable protocol. Nothing is known about absorption, distribution, metabolism or clearance in any species: the state register file for the licensed product prints its pharmacokinetics section as not studied, and no PubMed record reports a half-life, a plasma concentration or a bioavailability figure. The substance is a fraction of undefined composition, so batch-to-batch equivalence is also unestablished in the public record, and no independent analytical characterisation of what a vial contains was located. The human evidence has two roots and both trace to one institutional lineage: the 266-person geroprotection trial, in which the largest single arm held 58 people and the group carrying the most-quoted result held twenty; and one 92-patient open-label COVID-19 trial in which no participant in either group died, so no mortality inference is available from it. The severe-disease paper that does report mortality states two different values for its own treated arm in its English and Russian abstracts, and states no group sizes. Sample sizes are absent from a large share of the animal abstracts, and the two oncology reports point in opposite directions in different tumour models. Long-term administration, reproductive and developmental effects, carcinogenicity and immunogenicity have not been assessed in any retrievable regulatory review. On the transmissible spongiform encephalopathy risk inherent to bovine-derived injectables, two manufacturer-side statements exist: the 2003 paper reports that composition analysis by electrophoresis, immunoblotting and electron microscopy demonstrated the absence of prions, protooncogenes and viruses, and the Kazakh label states the source herd is free of viral, prion, bacterial and mycoplasma disease. No antibody, method detail or detection limit is given for that work, and no independent or regulatory assessment of the assertion was located. Thymalin is not an approved medicine in the United States, the European Union, the United Kingdom, Canada or Australia, and no identity, purity or sterility standard applies to this material outside the registered pharmacy supply chains of the two countries where it holds or held an authorisation.

Questions