Two names travel together, so they get treated as one thing with two brand labels. They are not. Pulling the two literature strands apart and reading the abstracts of the reviews that summarise them takes about an hour, and it produces two very different shapes.
BPC-157: a large animal literature and a very small human one
Two reviews published in 2025 cover the orthopaedic literature on BPC-157, one systematic and one narrative. Both describe the same shape. Across muscle, tendon, ligament and bone models, treated animals show better functional, structural and biomechanical outcomes than controls, and the direction holds across injury types and research groups. Tendon rupture models - rat Achilles tendon after transection or compression - are among the most represented models in the dataset.
Then the number that organises everything else. Thirty-five preclinical studies. One clinical study. That is the entire human evidence base the reviewers could find. They state plainly that no human safety data were identified. No side effects in the animal work only reassures within the boundaries of animal work.
A second, narrative review in the same year reaches the same conclusion from the same corpus. The mechanism has never been described as a clean binding at one receptor; it is described as broad-acting. That is the honest rendering of several effects at once, none of them fully mapped.
TB-500: a smaller dataset, built more carefully
Thymosin beta 4, sold under the name TB-500, has a genuinely human study in it, and it is worth reading properly. A phase 1 trial in healthy volunteers in China: randomised, double-blind, single and multiple administration, observed for four weeks. Thirty healthy participants in the multiple-administration arm, given a course over ten days.
The reported result: adverse events were mild to moderate in intensity, there were no treatment-limiting toxicities and no serious adverse events, and there was no obvious accumulation with continued administration.
Read that carefully, because the sentence is easy to over-read. It is a tolerability study in healthy people. It is not an efficacy study, because it was not designed to test any condition. It concerns a recombinantly produced thymosin beta 4, which is not the same object as whatever is in a vial bought without a prescription. That gap is exactly what a third-party test is for, and also its limit: what a peptide purity report establishes is the short version. And the conflict-of-interest statement records that two of the authors are employees of the sponsor. None of that makes the study wrong. It means the study answers one question and leaves the rest open.
The line that decides the argument
Both substances sit on the World Anti-Doping Agency’s prohibited list under category S0, which covers substances not approved for any therapeutic use. That line is not an opinion about evidence quality. It is a definition: if a substance is not approved anywhere for any use, it goes in that category, regardless of how many animal studies exist. Anyone arguing that a large animal literature changes the regulatory status is arguing against a definition.
What the data do not show
They do not show that either substance works in people. They do not show safety over years, or in people with existing conditions, or in combination with anything else. The phase 1 trial says nothing about a person with a tendon problem, because such a person was not in it. And the reviews do not establish that an animal effect size is a human effect size, which is a claim that recurred throughout the musculoskeletal supplement literature and was never sustained by a single head-to-head trial in humans.
The pattern that shows up in both cases is the same: a real scientific literature, honestly described in reviews, sitting underneath a commercial literature that describes it in superlatives. The reviews are the reason this subject can be discussed without the forum vocabulary.
This article summarises published reviews and a phase 1 trial. It is not medical advice, and it does not describe any use for any individual.
