BPC-157, TB-500 and GHK-Cu: what has been tested in people?

Updated 10 Oct 2026

I became interested in recovery compounds while returning to heavier training and reporting discomfort at the front of my left shoulder. That creates a useful research question, but it does not diagnose a tendon injury or show that an experimental drug is appropriate for me. These studies need to earn their relevance through their populations and outcomes.

The names do not describe interchangeable treatments

FDA’s 2026 TB-500 assessment explicitly distinguishes the acetylated fragment from full-length thymosin β4 and from the unacetylated sequence. Product names in a clinic, registry or paper should therefore be checked before treating their results as one evidence base. FDA TB-500 briefing, pages 24–26

BPC-157: favorable reports with weak causal leverage

The most directly relevant published musculoskeletal report is a retrospective knee-pain series. Seventeen patients were identified and 16 contacted: 12 had received BPC-157 alone and four BPC-157 plus material described as TB4. Eleven of 12 in the first group and three of four in the combination group reported improvement. There was no randomized comparator, standardized functional assessment or imaging demonstration of repaired tissue. Pain relief is interesting, but the design cannot separate treatment effect from expectation, natural change, other care or selective reporting. It also cannot establish that the combination adds benefit. Lee and Padgett, 2021

A separate uncontrolled report in 12 women with interstitial cystitis described marked symptom improvement following a local bladder procedure. It concerns a different disease, tissue and route. It is not a second trial of musculoskeletal recovery. Lee and colleagues, 2024

A 2025 intravenous pilot monitored two older adults, both previously exposed to BPC-157, over a short period. No reported side effects or measured abnormalities appeared in the tested endpoints. This is useful feasibility information; it cannot estimate uncommon adverse effects, establish long-term safety or demonstrate injury recovery. The authors’ broad safety conclusion goes beyond what two short observations can settle. Lee and Burgess, 2025

The older human record is often omitted

It would be inaccurate to say the entire human record consists only of those three papers. FDA’s July 2026 review also identifies older rectal-administration conference abstracts, including a 53-person randomized ulcerative-colitis study and healthy-volunteer work. The colitis comparison had an uncertain between-group result and incomplete endpoint and statistical reporting. These abstracts broaden the history of human exposure while providing insufficient evidence for a confident treatment conclusion. The healthy-volunteer abstracts should not be counted as separate large confirmatory trials merely because the same development work was presented more than once. FDA BPC-157 briefing, pages 26 and 39–41

Full-length thymosin β4 has human trials; TB-500 still needs its own

A double-blind randomized phase 2 study enrolled 73 people with venous ulcers and tested topical full-length thymosin β4. It reported an acceptable short-term safety profile and a potential healing signal at one concentration. Venous-ulcer care, local topical exposure and a 43-residue molecule differ substantially from systemic fragment injections for a painful shoulder. This study is a reason to investigate that formulation further, not to describe TB-500 as clinically validated. Guarnera and colleagues, 2010

FDA’s 2026 fragment assessment found no human administration studies of the nominated TB-500 substances in its reviewed evidence. It also explains why fragment and full-length data cannot be merged. This absence does not prove ineffectiveness or toxicity; it leaves the requested recovery benefit unestablished. FDA TB-500 briefing, pages 28–29

GHK-Cu: cosmetic evidence is real but narrow

In a randomized study after carbon-dioxide laser resurfacing, 13 participants completed follow-up. GHK-Cu skin-care regimens did not improve the objective redness or wrinkle/skin-quality assessments relative to comparison care, although a patient-reported skin-quality measure favored GHK-Cu. This is a mixed result, not an unequivocal regenerative effect. Miller and colleagues, 2006

A small 2026 split-face study reported improved eyebrow hair measures with a topical GHK-Cu serum in 18 participants. Hair-count and diameter differences were nominal week-12 findings across several outcomes and time points; the report did not describe correction for those between-side comparisons. Its cosmetic endpoint, short follow-up and small sample should remain visible. Even if replicated, eyebrow improvement would not establish systemic tendon healing. Original 2026 eyebrow report

FDA continues to note limited human safety information for injectable GHK-Cu, including potential immunogenicity related to aggregation and impurities. This is a route-specific concern, not a declaration that all topical cosmetic use is equivalent to injection. FDA’s compound-specific safety information

Registrations are a watchlist, not results

As inspected on October 10, 2026, NCT02637284, an older oral BPC-157 phase 1 registration, has unknown status and no posted results. A newer University of Arkansas rotator-cuff-repair pilot, NCT07803250, is not yet recruiting and has no results. Its record contains inconsistent planned participant counts: 30 in the structured enrollment field and 20 in its summary. Neither supplies evidence of successful healing today. Older registration, Rotator-cuff registration

Registry contents also require scrutiny. The TB-500 entry NCT07487363 explicitly calls itself fictional despite displaying a recruiting status. I do not count it as a human experiment. A separate BPC-157 hamstring record lists the same sponsor and no posted results. Its listing does not independently verify that the study was conducted, and I do not count it as completed clinical evidence. These observations come from the registry’s own submitted fields, not from assuming an NCT number certifies the study. Self-described fictional TB-500 record, Hamstring record

The regulatory record changed in 2026

FDA considered BPC-157 and TB-500, among other nominated substances, at its July 2026 Pharmacy Compounding Advisory Committee meeting. Consideration by an advisory committee, a compounding category and FDA drug approval are different statuses. The current 503A nomination document lists non-injectable GHK-Cu under evaluation; this does not confer approval on an injectable recovery product. Old summaries that put every compound under the same current category should not be carried forward without checking the date and route. FDA July meeting materials, FDA nomination categories, updated May 2026

The useful research gap is a well-characterized product tested against a credible comparator with functional recovery, structural outcomes and adequate safety follow-up. My immediate shoulder question still belongs with assessment and sensible loading, alongside the recovery page and nutrition guide. The current evidence gives reasons for further trials; it does not supply a demonstrated shortcut for my shoulder.

Back to investigations