Two ways to bottle a gland

Start with the organ, because both of these compounds are really arguments about one organ. The thymus is a small gland behind the breastbone whose job is finishing school for T-cells: immature lymphocytes arrive from the marrow, and the thymic environment teaches them what counts as self, kills the ones that get it wrong, and licenses the rest. Its peculiarity is that it does this work early and then leaves. From about puberty the thymus involutes — its working tissue is steadily replaced by fat, so that output of fresh, naive T-cells falls decade by decade. Much of what we call immune ageing is that shutdown.

So there is an obvious, decades-old idea: if the thymus is what winds down, give the body back whatever the thymus was making. Two research programmes took that idea and arrived at almost opposite objects. One kept the whole extract and never fully opened it. The other opened it all the way, to a single sequence. The gap between them is the real subject here — not which molecule is stronger, but what it means to know what a drug is.

Thymalin (Russian Timalin) is a polypeptide fraction boiled out of the thymus of calves, developed in Leningrad in the 1970s by Vladimir Khavinson and Vyacheslav Morozov — the same St. Petersburg school behind the cartilage tripeptide Cartalax and the pineal peptide Epithalamin. It has been in clinical use across the former Soviet Union for roughly four decades as an immunomodulator, given by injection for infections, burns, and the immune depression of age. And it is, chemically, still a complex: a heterogeneous mixture of peptides rather than one named molecule with one structure and one dose you can put on a certificate of analysis.

The school did not leave it entirely uncharacterised. Over the following decades its programme proposed that the interesting activity in these tissue extracts distils down to very short peptides — and for the thymus line it named a dipeptide, Glu–Trp (marketed separately as thymogen), as a candidate active. But it matters that thymalin the product is not that dipeptide: it is the parent extract, sold as the extract, and the mechanistic story attached to it is the same sweeping bioregulator hypothesis the group applies to all its preparations — that short peptides enter the cell nucleus and tune gene expression directly. That idea is genuinely interesting and covered in full elsewhere on this site; what matters for thymalin specifically is that its clinical reputation rests on one headline number.

Read the study the way you would any other. It is open, not double-blind; its results come overwhelmingly from a single institute and its collaborators; the endpoint that produced the headline is all-cause mortality in a modest cohort followed with the assignment known. None of that makes the number false. It makes it unaudited — a large effect from one group, on a product that is itself a mixture, never reproduced at that scale by an independent lab under blinded conditions. That is exactly the profile a careful field files under “interesting, unconfirmed,” not under “established.”

Now the other route. In the West the thymus was worked on reductively. In 1977 Allan Goldstein’s group took a crude thymic preparation —thymosin fraction 5, a soup not unlike thymalin in spirit — and kept purifying until a single peptide fell out: thymosin α1, a 28-amino-acid, N-terminally acetylated sequence that turns out to be the front end of a larger precursor protein, prothymosin α. It is fully defined: one sequence, one mass near 3,108 daltons, made now by chemical synthesis rather than pulled from tissue, so every vial is the same molecule.

Being defined is what let it become a drug. As thymalfasin, thymosin α1 is marketed as Zadaxin and approved in more than thirty countries, principally as an adjuvant in chronic hepatitis B and C, with a four-decade trail of clinical trials and a low adverse-event burden. It has been studied in sepsis — where the problem is precisely a collapsed, exhausted T-cell compartment — and was used during COVID-19 in China on the same rationale. Whatever one concludes about effect size, this is a molecule that has been through the machinery: named structure, defined dose, registered indication, published randomised trials.

And it has a mechanism you can point at. Thymosin α1 acts largely as an agonist at Toll-like receptor 9 (with reported activity at TLR2 and TLR4) on dendritic cells and other immune cells — the pattern-recognition receptors that read “danger” signals. Through them it drives the IRF3 and NF-κB pathways, pushing dendritic-cell maturation and a T-helper-1 program: more IL-2 and interferon-γ, restored T-cell maturation, sharper natural-killer activity. That is a defined surface receptor and a traceable signal — the thing the bioregulator hypothesis, for all its elegance, does not yet have.

Lay the two side by side and the contrast is almost clinical. Both begin at the same gland and, in effect, the same crude extract. One programme stopped there and defended the mixture; the other kept cutting until it held a single sequence. Everything downstream — whether you get a drug label, a named receptor, an independent trial — follows from that one fork.

• Thymalin is a mixture; thymosin α1 is a molecule. One is a polypeptide fraction of calf thymus with no single defined structure or standardised dose; the other is a synthesised 28-residue sequence identical vial to vial. You can write thymosin α1 on a certificate of analysis. You cannot do the same for an extract in the same way.
• One has a receptor, the other has a hypothesis. Thymosin α1's effects trace to TLR9 on dendritic cells and a Th1 program you can measure. Thymalin's are attributed to the broad 'short peptide enters the nucleus and tunes genes' model that remains largely single-source and unreplicated.
• The evidence lives at different grades. Thymosin α1 carries decades of randomised, published hepatitis-B and sepsis trials and a registered indication in dozens of countries. Thymalin's signature result is an open, single-institute mortality study — striking, but not the kind of evidence that earns a drug label anywhere it would have to be audited.
• Neither result is dismissed here. The point isn't that the extract is fake and the peptide is proven. It's that a defined molecule can be interrogated — dose, receptor, blinded trial — and a mixture largely has to be taken on trust. Bullish on the science in both; sceptical, on the page, in proportion to what each can show.

It is tempting to read this as West-beats-East, and that would be the wrong lesson. The Leningrad school’s instinct — that a whole tissue encodes recoverable instructions, and that the smallest active fragment might carry them — is the same instinct that produced thymosin α1, just stopped one step earlier. What separates the two is not nation or era but how far each was willing to run the purification, and therefore how much of each can be checked. The defined peptide got a receptor and a label because someone finished distilling it. The extract kept its most extraordinary claim and its black box together.