BPC-157 Is Entering a Phase 1 Trial After Rotator Cuff Repair: What NCT07803250 Will Actually Test

On 3 September 2026 a Phase 1 study of BPC-157 in patients recovering from arthroscopic rotator cuff repair appeared on ClinicalTrials.gov under the identifier NCT07803250.¹ It has not started, it is not the first human study of BPC-157 to be registered, and it commits to measuring one thing: shoulder strength. The interesting part of the record is the distance between what it describes in prose and what it promises to report. Everything below concerns a public registry entry and the published laboratory literature. Condor Research supplies BPC-157 as a research-use-only reference material, and nothing here describes, recommends or implies human use.

What exactly was registered on 3 September 2026?

The official title is Impact of BPC-157 on Recovery From Rotator Cuff Repair Surgery. The lead sponsor is the University of Arkansas, the single listed site is the University of Arkansas for Medical Sciences Medical Center in Little Rock, and the principal investigator is Shiloah Kviatkovsky, PhD, with orthopaedic surgeons Ryan Hill, Paul Inclan and Justin Rabinowitz named in the eligibility criteria as the operating surgeons.¹ The structured fields describe a randomised, parallel-group, quadruple-masked Phase 1 study with an estimated enrolment of 30 participants. Patients receive either BPC-157 or normal saline, self-administered by subcutaneous injection once daily for 90 days after surgery, in addition to the standard postoperative rehabilitation protocol. The record was submitted on 12 August 2026, cleared quality control on 1 September and was posted publicly on 3 September. Estimated start is January 2027, with primary completion and study completion both listed as August 2027. The oversight module marks it as a study of an FDA-regulated drug.

Is this the first registered human study of BPC-157?

No, and the distinction matters because the claim circulates widely. A ClinicalTrials.gov query for BPC-157 on 13 September 2026 returns four interventional records, the earliest posted in December 2015.² What is plausibly new about NCT07803250 is its specific setting: no other registered study addresses recovery after rotator cuff repair. The correct summary is that BPC-157 has been registered in humans before, that almost none of that work has produced published results, and that the human evidence base remains close to empty.

All four records were read directly from the ClinicalTrials.gov API on 13 September 2026. Registration is a declaration of intent. None of these entries reports efficacy results, and a registry record is not evidence that a compound works.

What will the trial actually measure?

One outcome is registered: isometric shoulder abduction strength, measured with a handheld dynamometer by a blinded evaluator, averaged over three trials, at baseline, 12 weeks and five months after surgery.¹ No secondary outcome measures are listed at all. The narrative summary is considerably more ambitious, describing postoperative MRI assessment of tendon healing, patient-reported outcome measures, range of motion and blood sampling for biomarkers of collagen turnover and inflammation. Those measurements may well happen. They are simply not in the part of the record that creates an obligation to report them.

This is not a technicality invented for the occasion. Prospective registration of outcome measures exists so that the endpoints chosen before a study can be compared with those emphasised after it. When a trial describes MRI-assessed tendon healing in its summary but registers only a strength measurement, a reader in 2028 has no registered baseline against which to check how the imaging was analysed or whether it was reported at all. The Hudson Biotech hamstring study is the useful contrast: it registers time to return to unrestricted sport and change in MRI-assessed injury volume, read by blinded central radiology, as co-primary endpoints.³

A registration is a plan. The only thing that changed on 3 September is that someone wrote the plan down where the rest of us can check it later.

What the record leaves out

The public entry states no dose. It describes route, frequency and duration, 90 daily subcutaneous injections, without specifying how much peptide each injection delivers.¹ For a Phase 1 study, whose stated purpose in the summary is to evaluate safety and feasibility, the dose is the central variable. The record also gives no IND number, although it marks the study as involving an FDA-regulated drug, and says nothing about the source, manufacturer, purity or characterisation of the material to be injected. Anyone who has read a peptide certificate of analysis knows that “BPC-157” names a sequence, not a defined pharmaceutical product, and that identity, purity, counterion and impurity profile vary between suppliers. Our guide to reading a certificate of analysis sets out what those documents do and do not establish.

0 mentions of a dose appear anywhere in the public registry record for NCT07803250, as read on 13 September 2026.

Three further inconsistencies sit inside the same record. The structured enrolment field says 30 participants while the narrative summary says twenty patients will be enrolled. The structured masking field says quadruple, covering participant, care provider, investigator and outcomes assessor, while the summary describes a double-blind trial. The control arm is typed as a sham comparator and its intervention is named “Sham (No Treatment)”, yet both the arm description and the summary make clear that it is a daily saline injection, which is a placebo. None of this affects whether BPC-157 works. All of it affects how carefully the record was assembled, and registry records are increasingly read as evidence by people who will never see the protocol.

What does the animal literature actually support for tendon?

The preclinical case for BPC-157 in tendon rests mainly on rat Achilles models and on cell work derived from them. Staresinic and colleagues reported accelerated healing of transected rat Achilles tendon in 2003.⁷ Chang and colleagues reported in 2011 that BPC-157 accelerated outgrowth from tendon explants and improved fibroblast survival under hydrogen peroxide stress and cell migration, while having no direct effect on proliferation in an MTT assay,⁶ and in 2014 that it increased growth hormone receptor expression in tendon fibroblasts.¹⁰ The proposed mechanisms across the wider literature involve VEGFR2 signalling, nitric oxide production through the Akt-eNOS axis and ERK1/2 activation.⁹

The most recent tendon study is more sobering. Biçer and colleagues published a four-arm rat experiment in 2026: 32 male Sprague-Dawley rats underwent standardised Achilles transection and repair and received control, BPC-157 at 10 µg/kg/day, TB-500 at 60 µg/kg/day, or both, intraperitoneally for four weeks.⁸ BPC-157 produced numerically lower histological degeneration scores and numerically higher load to failure, but did not reach statistical significance on the total Bonar or Movin scores. TB-500 did, on both histology and biomechanics. Combining the two produced no additive effect. Eight animals per group is small, and the authors present the work as exploratory, but it is the closest thing to a controlled tendon comparison published recently, and it does not show BPC-157 outperforming its most common comparator. We looked at how these two compounds differ in our BPC-157 and TB-500 comparison.

One absence is worth stating plainly: a PubMed search for BPC-157 together with rotator cuff returns no records. The rotator cuff is not the Achilles tendon. It is a tendon-to-bone enthesis repaired under different mechanical conditions, with a healing failure mode that is specific to that interface. The trial is therefore not confirming an established animal model in humans. It is testing a hypothesis extrapolated across tissues.

How much human evidence exists today?

A systematic review published in HSS Journal in 2025 searched PubMed, Cochrane and Embase from database inception to June 2024 and included 36 studies spanning 1993 to 2024.⁵ Of those, the review identified a single human study. That study is a retrospective chart review from a private clinic in Orlando covering 2019 to 2020, in which 17 patients had received intra-articular BPC-157 alone or with thymosin beta-4, 16 were reached by telephone, and pain relief was assessed by asking them, with no validated instrument and no control group.¹¹ Whatever else can be said about it, it is not evidence of efficacy, and its own authors describe it as small and call for future studies.

That is the state of the human record in September 2026: one uncontrolled retrospective case series, one Phase 1 safety study from 2015 that has never reported, one Phase 2 study recruiting in Shenzhen, one supplement study registered after it finished, and now one Phase 1 plan in Arkansas. A narrative review in Current Reviews in Musculoskeletal Medicine in 2025 reached the same conclusion from the opposite direction, pairing a detailed mechanistic account with explicit safety and regulatory concerns about a compound whose availability has run far ahead of its evidence.⁹

Why rotator cuff repair is a hard place to detect an effect

Structural failure after rotator cuff repair is common enough to be worth studying and rare enough to be hard to move. A meta-analysis restricted to level 1 and 2 studies put imaging-confirmed retear at roughly 15% to 21% depending on the follow-up window.¹² Retear risk rises with age, tear size and fatty infiltration. NCT07803250 restricts enrolment to full-thickness posterosuperior tears under 2.5 cm in anteroposterior dimension, and excludes patients with diabetes, autoimmune disease, previous repair, and hormone or testosterone replacement within twelve months.¹ That is a deliberately clean population, and it is also the population with the best baseline prognosis, which leaves less room for an intervention to demonstrate a difference.

The registered endpoint compounds the problem. Isometric abduction strength at twelve weeks after a rotator cuff repair is driven substantially by pain, by the rehabilitation protocol and by how much a given patient has been allowed to load the repair. With twenty to thirty participants split between two arms, a study of this size can establish whether the protocol is feasible and whether anything alarming happens. It cannot establish efficacy, and the summary is honest about this, describing the work as a pilot intended to inform the design of a larger trial.

What would make the eventual results worth reading

Four things, all of them within the investigators’ control. First, disclosing the dose, the supplier and the analytical characterisation of the material used, since a result is uninterpretable without knowing what was injected. Second, updating the record with the secondary outcomes the summary already describes, before the first patient is enrolled rather than after the data are in hand. Third, reporting the MRI analysis in full regardless of which way it comes out. Fourth, posting results at all: Phase 1 drug studies generally fall outside the results-reporting requirement that applies to later-phase trials, which makes voluntary posting a meaningful signal rather than a legal minimum. The 2015 record shows what happens otherwise. A decade later, nobody outside that company knows what its 42 healthy volunteers experienced.

What is established, and what is not

Established: BPC-157 accelerates healing in several rodent injury models, including transected Achilles tendon, and has plausible mechanistic accounts involving angiogenesis and fibroblast behaviour.⁵⁷⁹ Also established: four interventional human studies are registered, and a study specifically addressing rotator cuff repair now exists on paper.¹

Not established: any effect on human tendon healing, any effect on the rotator cuff in any species, the dose that would be required, or the safety profile of 90 consecutive days of subcutaneous administration. Unresolved: whether BPC-157 outperforms TB-500 or placebo on hard tendon endpoints, given that the most recent controlled rat comparison found its advantage did not reach significance.⁸ The finding that would change this assessment is simple to describe: a registered, adequately powered trial reporting a prespecified structural endpoint, with the dose and the material disclosed.

What this changes, and what it does not

Nothing about the evidence changed on 3 September 2026. No patient has been dosed, no data exist, and the study is not scheduled to begin until January 2027. What changed is narrower and still worth noting: a hypothesis that has circulated since the 1990s on the strength of rat experiments has been written down as a protocol at a university hospital, with a blinded assessor, a randomisation scheme and a completion date. That is how a claim stops being a claim. It is also, on the current record, a plan with no stated dose, one registered endpoint and a twenty-versus-thirty discrepancy in its own summary, which is why the useful thing to do in 2027 is to read the results rather than the registration.

Readers who want the background rather than the news will find it in our BPC-157 explainer, our review of the 2026 FDA compounding vote, and our note on the legal position in Europe. For the analytical side, see what HPLC and mass spectrometry can establish about a BPC-157 sample.

How this was checked. Registry data were read directly from the ClinicalTrials.gov API on 13 September 2026 for NCT07803250, NCT02637284, NCT07437547 and NCT07752381, rather than from secondary coverage. Literature was searched on PubMed on the same date, including the query BPC-157 AND rotator cuff, which returned no records. Where the registry record contradicts itself, both readings are reported here rather than reconciled. Version 1.0, first published 13 September 2026; this page will be updated if the registration changes or the study reports.

Condor Research supplies BPC-157 as a characterised reference material for laboratory research use only: not for human or veterinary use, not for diagnostic or therapeutic application, and not for any food or cosmetic purpose. Specifications and handling data are on the BPC-157 product page. Nothing in this article, and nothing in the registration of NCT07803250, should be read as evidence of clinical efficacy or as a basis for any use beyond controlled laboratory research.

Condor Research · Scientific desk
Atrio Sciences s.r.o., IČO 57 669 651, Nitra (SK) · info@condorresearch.com