BPC-157

Identity, and the part of the name that does not resolve

BPC-157 has travelled under four development codes, and the registries treat all four as one thing. The FDA Global Substance Registration System entry for UNII 8ED8NXK95P collects BEPECIN, PL-10, PLD-116 and PL-14736 under a single record, which settles a recurring ambiguity in the secondary literature: the four codes denote one substance, not a series of related ones. That record classes the substance as a protein and lists the subunit sequence GEPPPGKPADDAGLV. Its CAS number and PubChem identifier match the one PubChem compound record for BPC-157: CID 9941957, CAS 137525-51-0, molecular formula C62H98N16O22, molecular weight 1419.5, InChIKey HEEWEZGQMLZMFE-RKGINYAYSA-N.

A second PubChem record exists and is easy to confuse with the first. CID 78169625 carries the identical molecular formula and molecular weight, but a different InChIKey, HEEWEZGQMLZMFE-UHFFFAOYSA-N, and its synonym list gives the sequence in DL form at every chiral centre rather than L. It is the stereochemically undefined twin of the L-peptide. An identity check that matches on formula or mass alone cannot separate the two, and a document quoting C62H98N16O22 at 1419.5 is consistent with either record.

Everything above is verifiable in a registry. The origin story attached to the name is not. Sikirić and colleagues, writing in the Journal of Physiology (Paris) in 1993, described a new gastric juice peptide of relative molecular mass 40,000, named BPC, as recently isolated, and presented the fifteen-residue fragment as the portion thought to carry the activity. That abstract names no isolation paper. UniProtKB queried for "body protection compound", for "BPC-157" and for the string GEPPPGKPADDAGLV returned zero entries on all three searches, and no mass-spectrometric identification of this sequence in human gastric juice was located in PubMed.

The shape of the literature

A PubMed search for BPC-157 returned 226 records on 17 August 2026. Restricting the same search to records carrying Sikirić as an author returned 175. Roughly three quarters of the indexed literature on this compound comes from one laboratory at the University of Zagreb, and a large share of that is review rather than primary data — the same experiments recur across successive narrative summaries in Pharmaceuticals, Biomedicines and Current Pharmaceutical Design. Counting records substantially overstates how much independent measurement exists behind the compound.

Vasireddi and colleagues put a number on the primary base. Their 2025 systematic review in HSS Journal screened 544 articles published between 1993 and 2024, removed duplicates, and included 36 studies: 35 preclinical and one clinical. They graded the entire body as level IV and level V evidence and recorded that no clinical safety data were found. Gwyer and colleagues had reached a similar position in Cell and Tissue Research in 2019, observing that across two decades only a handful of research groups had performed in-depth work on the peptide.

Groups outside Zagreb do exist, and they carry disproportionate weight because they report their methods differently. Chang and colleagues in Taiwan produced the tendon-fibroblast work. He, Xu and colleagues at Air Force Medical University in Xi'an produced the pharmacokinetics and the formal toxicology. Two Turkish groups published rat surgical and ischaemia-reperfusion models in 2026. A Polish group measured acetylcholinesterase inhibition. Where those groups report sample sizes, they state them explicitly, which is not uniformly true of the Zagreb series.

What the soft-tissue experiments measured

The founding musculoskeletal result is Staresinić and colleagues, 2003, in the Journal of Orthopaedic Research. Rats had the right Achilles tendon transected 5 mm proximal to the calcaneal insertion; the peptide was given intraperitoneally once daily at 10 µg, 10 ng or 10 pg per kilogram, first at 30 minutes after surgery, with biomechanical, functional, microscopic and macroscopic assessment at days 1, 4, 7, 10 and 14. Treated groups showed higher load to failure, higher Young's modulus and higher Achilles functional index values than saline controls. The retrieved report does not state group sizes.

Cerovečki and colleagues extended that design to the medial collateral ligament in 2010, following transection out to 90 days and comparing intraperitoneal, drinking-water and topical routes, again without stated group sizes in the retrieved report. Novinscak and colleagues had published a gastrocnemius crush model in Surgery Today in 2008 on the same pattern. Chang and colleagues supplied the cell-level work: in 2011, increased outgrowth from rat Achilles tendon explants, increased fibroblast migration and phosphorylation of FAK and paxillin, with no direct effect on proliferation; in 2014, upregulation of the growth hormone receptor in the same cells.

One independent replication of the tendon result states its numbers in full and does not reproduce it cleanly. Biçer and colleagues, in Joint Diseases and Related Surgery in 2026, randomised 32 male Sprague-Dawley rats into four groups of eight — control, BPC-157 at 10 µg/kg/day, TB-500 at 60 µg/kg/day, and both — after standardised Achilles transection and repair, dosing intraperitoneally for four weeks. Maximum load to failure was higher in the BPC-157 and TB-500 groups than in controls but reached significance only for TB-500. Total Bonar and Movin histological scores were significantly lower than control in the TB-500 and combination groups; the BPC-157 group's scores were numerically lower without reaching significance.

Pharmacokinetics and preclinical toxicology

He and colleagues published the first formal ADME study in Frontiers in Pharmacology in 2022, in Sprague-Dawley rats and beagle dogs at six animals per group. After single intravenous administration the elimination half-life was 15.2 minutes in rats and 5.27 minutes in dogs. Intramuscular dosing at 20, 100 and 500 µg/kg in rats and 6, 30 and 150 µg/kg in dogs gave linear kinetics, with mean absolute bioavailability of approximately 14 to 19 per cent in rats and 45 to 51 per cent in dogs. Parent compound was undetectable four hours after dosing.

Distribution in the same study placed the highest total radioactivity in kidney at one hour, then liver, stomach wall, spleen and thymus, all above plasma. Brain sat below plasma at one hour and remained below plasma at 24 hours, in the lowest band alongside body fat. Tritium-labelled material was excreted principally in urine and bile after rapid hydrolysis to small fragments and free amino acids. Total radioactivity counts metabolites as well as parent compound, so these figures do not describe the intact peptide — but they are the only distribution dataset that exists.

Formal toxicology comes from Xu and colleagues in Regulatory Toxicology and Pharmacology, 2020: single-dose and repeated-dose work in mice, rats, rabbits and dogs, with local tolerance, genotoxicity and embryo-fetal studies alongside. The single-dose study identified no test-article-related effects. Repeated dosing in dogs produced no abnormal change against solvent control other than a fall in creatinine at 2 mg/kg but not at lower doses, which reversed after two weeks of withdrawal. Irritation from local administration was recorded as mild. Group sizes are not given in the abstract.

These two datasets sit awkwardly together. A compound cleared from plasma in minutes is credited with differences measured at 14, 28 and 42 days in the surgical models, and no mechanism bridging the two has been demonstrated. The 2026 review by Mateescu and colleagues in Pharmaceutics names this pharmacokinetic-pharmacodynamic disconnect as the central unresolved problem, alongside the absence of any characterised pharmaceutical-grade formulation, permeability data or plasma protein binding figure.

The human record

Three human reports exist. All come from one private clinic in Florida, all appeared in Alternative Therapies in Health and Medicine, and none had a control arm. Lee and Padgett reviewed charts from 2019 to 2020 covering 17 patients given intra-articular injections for knee pain of mixed cause; 16 were reached by telephone and asked to rate their pain retrospectively, with no standardised instrument used. Twelve had received BPC-157 alone, of whom 11 reported significant improvement; across both the peptide-alone and combination groups, 14 of 16 reported relief.

Lee, Walker and Ayadi injected 12 women who had not responded to pentosan polysulfate with a total of 10 mg around the area of bladder inflammation during a single cystoscopy, and recorded Global Response Assessment scores of five out of five in all twelve. Pharmacokinetic data in people come from two subjects: Lee and Burgess infused 10 mg intravenously on day one and 20 mg on day two into a 58-year-old man and a 68-year-old woman, both of whom had received intravenous BPC-157 before the study, reporting no measurable change in cardiac, hepatic, renal, thyroid or glucose biomarkers and a plasma half-life under 30 minutes.

Registered trials are thin. A ClinicalTrials.gov intervention search returned three records. NCT02637284, a phase 1 safety and pharmacokinetic study of 42 participants sponsored by PharmaCotherapia with a March 2016 completion date, carries status unknown and has posted no results. NCT07752381 tested a peptide gummy product rather than the isolated compound. NCT07437547, a phase 2 placebo-controlled trial of 120 participants with acute hamstring strain sponsored by Hudson Biotech, began recruiting in February 2026 with completion scheduled for February 2028. If it reports, it will be the first controlled human evidence on this compound.

Regulatory position, and one withdrawn paper

United States compounding regulation moved twice in three years. BPC-157 was placed in Category 2 of the section 503A bulk drug substances list — substances that raise significant safety risks — and then taken off it. The FDA category document updated 14 May 2026 lists BPC-157 in none of the three categories. The agency's safety-risks page retains it under bulk drug substances nominated but withdrawn, with the original assessment intact: compounded drugs containing it may pose risk for immunogenicity for certain routes of administration and complexities regarding peptide-related impurities and characterisation of the active ingredient, and the agency identified no or only limited safety-related information for the proposed routes.

Removal followed the nominators withdrawing their nominations. It was a procedural consequence, not a finding about the substance, and evaluation continued regardless. The Pharmacy Compounding Advisory Committee agenda for 23 July 2026 lists BPC-157 free base and BPC-157 acetate among the substances discussed for possible inclusion on the 503A bulks list, with the use evaluated recorded as ulcerative colitis — the indication the originating group has named since 2003. The FDA briefing package for that session, linked from the agenda page as media item 193343, returned HTTP 404 on repeated retrieval attempts on 17 August 2026, while the packages for the KPV, TB-500 and MOTS-c sessions at the same meeting downloaded normally.

Anti-doping status is disputed inside the peer-reviewed literature. Józwiak and colleagues, writing in Pharmaceuticals in 2025, state that the compound was temporarily banned by WADA in 2022 and is not currently listed as banned. The United States Anti-Doping Agency states that it is prohibited under section S0, non-approved substances, which is a catch-all clause covering any pharmacological substance without regulatory approval rather than a named listing. Repeated attempts to retrieve the 2026 Prohibited List PDF from the WADA site returned an empty response body, so the List document itself could not be read for this page. The two positions reconcile only if listed is read narrowly.

A single paper in the corpus carries a formal flag, and its history is unusual. Current Neuropharmacology published a withdrawal notice in January 2025 for Stable Gastric Pentadecapeptide BPC 157 as a Therapy of Severe Electrolyte Disturbances in Rats, by Grubišić and colleagues, stating that the article had been withdrawn at the author's request; PubMed types that record as a Retracted Publication. The same title, by the same authors, reappeared in the same journal in 2026 under a different DOI and carries no flag. A publication-type and correction-link check across every other paper cited on this page returned no retraction, expression of concern or erratum.

What is not known

No controlled human trial of BPC-157 has been completed or published. The entire human record is three uncontrolled reports from one private clinic totalling fewer than thirty people, two of them retrospective or single-arm symptom surveys in conditions with high placebo response rates, and the third a two-subject pharmacokinetic and safety pilot. The one registered phase 1 study never reported. Nothing is therefore established about efficacy for any indication, about dose-response in humans, about inter-individual variability, or about any adverse event less frequent than roughly one in ten. Long-term administration has not been characterised in people at all, and the preclinical toxicology, while it covers four species, has no published group sizes. Subcutaneous pharmacokinetics — the route most discussed outside the literature — have not been measured in any species. Oral bioavailability has not been measured. Intestinal stability, epithelial permeability, hepatic first-pass extraction, plasma protein binding and volume of distribution are all uncharacterised. The mechanism connecting a plasma half-life of minutes to differences observed at six weeks in rodent surgical models is unexplained. Roughly three quarters of the indexed literature comes from a single laboratory, and where independent groups have attempted replication with stated sample sizes, the tendon result did not reach significance. BPC-157 is not an approved medicine in any jurisdiction, and material distributed outside a registered trial has not been subject to identity, purity or sterility controls; the FDA's own assessment identified immunogenicity risk and characterisation complexity as open questions for compounded preparations.

Questions