TB-500 and thymosin beta-4: what has actually been tested in people

Two different things get sold under the name TB-500, and the evidence belongs to only one of them. Thymosin beta-4, a 43-amino-acid protein your cells make anyway, has been through a handful of small randomised trials, mostly as eye drops, with mixed results. TB-500 proper is a seven-amino-acid fragment of it, and as of September 2026 nobody has published a single human study of that fragment.

What it is

Thymosin beta-4 (Tβ4) is one of the most abundant small proteins in mammalian cells. Its best-known job is holding on to actin monomers, the building blocks of the scaffolding that lets a cell change shape and crawl [1]. That is why it turns up in research on wound healing: cells have to move into a wound before they can repair it.

In 2003, a team at the US National Institutes of Health cut Tβ4 into pieces and tested which parts drove blood-vessel growth. A seven-amino-acid stretch at positions 17 to 23, LKKTETQ, the actin-binding motif, did almost everything the full protein did in their cell-migration and chick-artery sprouting assays, at similar concentrations; pieces lacking that motif were inactive [1]. TB-500 is a synthetic version of that stretch with an acetyl group added to the front (Ac-LKKTETQ). It first surfaced as a veterinary product, and Hong Kong's racing laboratory published a test to catch its use in horses in 2012 [2].

In practice the name is used loosely, and some products sold as TB-500 are described by their sellers as full-length Tβ4. The distinction matters because the fragment lacks the other 36 amino acids, and anything the protein does through those regions can't be assumed to carry over. Everything under "How it is thought to work" below comes from cells and animals and should be read as theory.

The human evidence: full-length Tβ4

All of the published human trials used full-length synthetic Tβ4, and most came from one development programme, run first by RegeneRx and later by ReGenTree, which calls the eye-drop form RGN-259.

Safety in healthy volunteers. The first human study gave intravenous Tβ4 to 40 healthy volunteers in four cohorts of 10, each including placebo, at single doses of 42, 140, 420 or 1,260 mg, then the same dose daily for 14 days [3]. Adverse events were infrequent and mild or moderate, with no dose-limiting toxicity. Blood levels rose in proportion to the dose, and the half-life lengthened at higher doses. The stated aim was a drug for heart attacks, and these are still the only published pharmacokinetic data for Tβ4 in people. Note the doses: tens to hundreds of milligrams, given into a vein under trial monitoring.

Skin ulcers. A Phase 2 dose-escalation trial at eight European sites randomised 73 people with venous leg ulcers to topical Tβ4 or placebo [7]. Safety matched placebo. On efficacy, the authors could only say the 0.03% dose "may have the potential" to speed healing, with about a quarter of patients fully healed by three months. That is a hint, not a result.

Dry eye. This is where the most data sit, and it's a story of shrinking effects as trials got bigger.

The two large Phase 3 dry-eye trials have results on the ClinicalTrials.gov registry but, as far as we can find, no peer-reviewed publication. On the numbers posted, the drops did no better than placebo on the main outcomes [8,9]. The 72-person Phase 2 trial had already hinted at this: it used a controlled chamber that dries the eyes on purpose, and neither of its prespecified primary endpoints moved, although discomfort inside the chamber fell by 27% on Tβ4 (p = 0.0244) [5]. Small positive trials followed by flat large ones is a common pattern in drug development, and it is a good reason to be wary of any peptide whose case rests on small studies alone.

Neurotrophic keratopathy. This rare condition, in which nerve damage stops the cornea from healing, produced the most encouraging result. In the SEER-1 Phase 3 trial, 6 of 10 people on 0.1% RGN-259 had complete healing of their corneal defect at four weeks versus 1 of 8 on placebo [6]. Because the trial stopped at 18 people, that difference just missed statistical significance (p = 0.0656). Some secondary measures, including comfort and healing at day 43, did reach significance. The sponsor's employees helped design the study, a fact the paper declares.

The human evidence: the TB-500 fragment

There isn't any. When the FDA reviewed TB-500 for its July 2026 compounding advisory meeting, its briefing noted no human data at all for the fragment, according to one law-firm summary of the documents [14]. The first registered human trial of Ac-LKKTETQ, a US Phase 1/2 study in 80 adults with stable heart disease whose primary outcome is adverse events, began recruiting in February 2026 [10].

How it is thought to work: animal and lab evidence

The animal results for full-length Tβ4 are strong and come from several independent labs, which is more than can be said for many recovery peptides.

• In rats with full-thickness skin wounds, Tβ4 applied to the skin or injected increased re-epithelialisation by 42% over saline at day 4 and 61% at day 7 [11].
• In mice after a heart attack, Tβ4 improved heart-cell migration and survival and cardiac function, in a Nature paper that drew much of the later interest [12].
• In the NIH cell work, the LKKTETQ fragment matched the full protein on endothelial migration, which is the main scientific basis for selling the fragment [1]. That finding covers blood-vessel assays in a dish at around 50 nM. It says nothing about whether an injected fragment survives long enough in the body to reach an injury, and no one has published that data.

A 2026 rat Achilles study that tested TB-500 alongside BPC-157 found TB-500 improved load to failure and tissue scores after four weeks; we cover it in our piece on combining the two.

Safety and side effects reported

For full-length Tβ4, the human safety record is reasonable but narrow: IV doses up to 1,260 mg a day for two weeks in 40 volunteers [3], and topical or eye-drop use in the published trials above, with adverse events similar to placebo where reported [4,5,6,7]. There is no long-term safety data, and none at all for repeated injection of the fragment in people.

Two issues are often raised. Tβ4 encourages cell migration and blood-vessel growth, so the same theoretical worry about feeding an existing tumour that hangs over BPC-157 applies here; no human study has tested it. And products bought outside trials are unregulated, so you can't be sure which molecule, the fragment or the protein, is in the vial.

Legal and regulatory status

United States. Neither Tβ4 nor TB-500 is FDA-approved. On 23 July 2026, the FDA's Pharmacy Compounding Advisory Committee voted 8 to 6, with one abstention, to recommend allowing 503A pharmacies to compound TB-500 for wound healing [13,14]. The FDA's own briefing had proposed not adding it [14], and advisory votes don't bind the agency [13].

United Kingdom. Not a licensed medicine in any form.

Sport. WADA lists "thymosin-β4 and its derivatives e.g. TB-500" under S2 growth factors, prohibited at all times [15].

What we still don’t know

• Whether the TB-500 fragment does anything in people. The first trial [10] has not reported.
• Whether RGN-259 will hold up in neurotrophic keratopathy in a larger trial than SEER-1.
• Why the dry-eye results faded from small Phase 2 trials to flat Phase 3 results.
• Whether injected Tβ4 or its fragment helps tendon, muscle or joint injuries in humans, which is what most people buy it for. No trial has tested this.
• Long-term safety of any form taken for months.

What readers report

First-hand accounts, shared with permission. They are self-reported, not evidence, and sit alongside the research above rather than in place of it.

Educational content only — not medical advice. Many peptides discussed on HPR are not approved for human use. Talk to a qualified clinician before making any decision about your health.

Free live webinar · Thursday 1 October · 7pm UK time

Join the next webinar: Introduction to Peptides

with Mitch O’Callaghan, ISSCA Peptide Therapist · Founder, Proper Protocols

Save my free placeHPR×Proper Protocols