Why so much peptide evidence comes from animals, and how to read it

In the literature on BPC-157, TB-500 and epithalon, controlled human trials are scarce ^1,2,3. That pattern does not show that the compounds are ineffective; it means efficacy and safety in humans remain unestablished, which changes how claims should be read ^2,3. A systematic review of BPC-157 in orthopaedic sports medicine screened 544 articles dated 1993 to 2024 and included 36 studies, of which 35 were preclinical and one was clinical ¹. A later PRISMA review of human evidence found just three studies with 30 participants in total, judged them to be of poor methodological quality, and rated the certainty of evidence very low ².

What the human data actually look like

The human reports for BPC-157 are small and uncontrolled ^2,4. One retrospective chart review of intra-articular use for knee pain contacted 16 of 17 patients by phone, used no specific tools to measure function or quality of life, and described itself as a small study needing future work ⁴. A pilot of intravenous administration enrolled two participants ⁵. The orthopaedic review concluded that no clinical safety data were found, and the later review noted that the evidence is largely limited to animal models and that BPC-157 lacks regulatory approval for human use ^1,2.

For epithalon, a 2026 narrative review described human evidence as limited to small physiological or biomarker studies and an older controlled study in retinitis pigmentosa, found no adequately powered randomised trial establishing therapeutic efficacy, and called pharmacokinetic, dose-ranging and long-term safety data inadequate ³. Thymosin beta-4, the parent protein of TB-500, has some human data: a nine-patient phase 2 trial of an eye-drop formulation reported improvements in dry-eye signs and symptoms ⁶. The compound called TB-500 itself was described in a 2012 analytical paper as a veterinary preparation containing an acetylated synthetic version of the LKKTETQ fragment of thymosin beta-4, and that paper focused on detecting its use in horses rather than on efficacy ⁷. Much of the broader mechanistic case rests on cell and mouse work, such as a 2004 report that thymosin beta-4 promoted cardiomyocyte migration and survival in culture and improved cardiac function after coronary ligation in mice ⁸.

Why animal and cell findings travel poorly

Preclinical findings are informative, but on their own they are unreliable predictors of human effects ⁹. A systematic review comparing animal experiments with clinical trials found discordance in several cases: corticosteroids showed improvement in animal models of head injury but not in clinical trials, and tirilazad reduced infarct volume in animal stroke models yet was associated with worse outcomes in patients ⁹. The authors suggested that discordance may reflect bias or failure of animal models to mimic clinical disease ⁹. Publication bias compounds the problem: an analysis of 16 systematic reviews of animal stroke studies suggested it could account for around one-third of reported efficacy ¹⁰. Reviewers of preclinical guidelines identified recurring recommendations such as power calculations, randomised allocation and characterisation of the disease model, which are the features worth looking for in any animal study ¹¹. In one analysis of 101 basic-science articles that claimed therapeutic promise, 27 led to a published randomised trial and five technologies were licensed for clinical use ¹².

A practical checklist for interpreting claims

• Identify the model first (cells, rodents, larger animals or humans), because findings in one model do not transfer automatically to the next ^9,12.
• Check whether the human data are controlled, blinded and adequately powered, or are case series and chart reviews ^2,4.
• Look for independent replication and sample size, because promising basic-science findings often do not progress to controlled human trials, and publication bias can inflate reported animal efficacy ^10,12.
• Separate the molecule studied from the product described: thymosin beta-4, its LKKTETQ fragment and the veterinary TB-500 preparation are not the same material ^7,13.
• Treat safety statements from samples of a few people as limited data rather than evidence that harm is absent ^1,5.

Regulatory context

The WADA 2026 Prohibited List names BPC-157 among non-approved substances, and lists thymosin-beta-4 and its derivatives, for example TB-500, among prohibited peptide hormones, growth factors and mimetics ¹⁴. FDA warning letters to peptide sellers have stated that "research use only" and "not for human consumption" wording does not change the assessment when evidence shows the products are intended as drugs for human use ¹⁵. Our legal status page covers the wider picture, and the evidence page explains how we grade it.

Key points

• For BPC-157, 35 of 36 studies in one systematic review were preclinical, and a later human-evidence review found three small, low-quality studies ^1,2.
• Epithalon has small human physiological and biomarker studies but no adequately powered efficacy trial ³.
• Animal findings sometimes disagree with human trials, and publication bias may inflate reported animal efficacy ^9,10.
• Always ask which model, which molecule and which design sits behind a claim ^7,11.