BPC-157 Research Uses: A Mechanism-First, Skeptical Review From the Independent Track

Executive summary

I left a master’s program because I got tired of writing elegant narratives around fragile datasets. BPC-157 is exactly the kind of topic that rewards that kind of cynicism: the preclinical literature reports broad, often dramatic healing effects across tissues (gut, tendon/ligament, muscle, nerve, bone, vascular/organ injury models), frequently at very low doses, with comparatively little independent replication and minimal human evidence. [1]

Mechanistically, there is credible cell- and pathway-level work supporting pro-repair signaling—especially vascular/endothelial biology (VEGFR2 internalization and downstream Akt–eNOS signaling; NO-dependent vasomotor effects; endothelial migration and tube formation). [2] A notable new 2026 paper proposes a specific intracellular protein-interaction axis (BPC157–FBXO22–BACH1) to explain proangiogenic effects, which—if independently replicated—would be one of the more concrete mechanistic anchors in this space. [3]

Human data remain thin and methodologically weak: a retrospective knee-pain chart review (single injection; no validated pain instrument), a 12-person chart review in interstitial cystitis (procedural injections during cystoscopy), and a 2-person IV-infusion pilot safety report. [4] A Phase I safety/pharmacokinetics record exists (NCT02637284), but results are not publicly posted in the registry interface available here; some reviews report additional details (including claimed dosing and “no quantifiable plasma/urine levels”), yet those details are not independently verifiable from the primary registry record in this environment and should be treated as unconfirmed. [5]

Safety and regulation are the practical choke points: BPC-157 is not an FDA-approved drug, appears on FDA’s “Category 2” list of bulk substances that may present significant safety risks for compounding, and is prohibited in sport under WADA’s “S0 (Non-Approved Substances)” framing. [6] Anti-doping laboratories have published methods relevant to detecting BPC-157. [7]

For researchers, the highest-value next steps are not “yet another rat tendon figure,” but replication-grade experiments: preregistered protocols, pharmacokinetics/pharmacodynamics (PK/PD) linkage, multi-lab validation, and explicit mechanistic falsification (e.g., VEGFR2 internalization dependence, NO-dependence, FBXO22/BACH1 dependence). [8]

Provenance, definitions, and my stance

BPC-157 is typically described as a synthetic “stable gastric pentadecapeptide” (15 amino acids). One widely cited sequence is GEPPPGKPADDAGLV with molecular weight ~1419. [9]

A recurring claim in the wound-healing literature is that BPC-157 is unusually stable in human gastric conditions and retains activity across multiple administration routes in animals, including per-oral exposure (often via drinking water). [10] That “oral activity” claim matters because it implies either atypical peptide stability, atypical absorption, unusually potent downstream triggering, or (the skeptical interpretation) an experimental artifact that hasn’t been pressure-tested across enough independent labs. [11]

My stance throughout: treat BPC-157 as a hypothesis generator with an unusually positive preclinical track record, not as an established therapeutic. The literature is valuable, but it is not yet “translation-grade” in the way researchers usually mean that term. [12]

Pharmacology and mechanism of action

What is solid versus speculative

Relatively solid (supported by independent groups and standard assays):

• Pro-angiogenic signaling via VEGFR2 internalization → Akt–eNOS activation, shown using CAM assay, endothelial tube formation, hindlimb ischemia readouts (laser Doppler), and endothelial-cell experiments; importantly, this work reports VEGFR2 upregulation but not VEGF-A upregulation, and shows sensitivity to an endocytosis inhibitor (dynasore). [13]
• NO-dependent vascular effects and endothelial signaling consistent with Src–Caveolin-1–eNOS pathway activation, plus shifts in eNOS/Cav-1 binding in endothelial cells. [14]
• Cell migration / cytoskeletal adhesion signaling in tendon fibroblasts, including activation (phosphorylation) of FAK and paxillin. [15]

Plausible but not yet “nailed down”:

• A unifying “cytoprotection mediator” framework and broad pleiotropy (GI protection → organ protection, vascular protection) that may be real but is often argued in reviews more than proven by falsifiable experiments. [16]
• Claims that BPC-157 “controls” angiogenesis in a context-dependent way (pro-repair without tumorigenic risk), which is possible, but current evidence is insufficient to declare oncologic risk “excluded.” [17]

Notable “new mechanism” worth watching (2026):

A 2026 study proposes that intracellular BPC157 binds the E3 ubiquitin ligase adaptor FBXO22 via the peptide’s proline at position 3, suppressing ubiquitination of transcription factor BACH1, stabilizing BACH1 and enhancing endothelial proliferation/tube formation—framed as a BPC157–FBXO22–BACH1 axis. If this holds up, it is a rare example of a specific molecular interaction story rather than a pathway collage. [3]

A mechanistic map that makes testable predictions

flowchart TD
  A[BPC-157 exposure] --> B[Endothelium / vascular niche]
  B --> C[VEGFR2 upregulation + internalization]
  C --> D[Akt activation]
  D --> E[eNOS phosphorylation]
  E --> F[NO production]
  F --> G[Vasodilation + improved perfusion]
  G --> H[Enhanced repair environment]
  A --> I[Tendon fibroblasts]
  I --> J[FAK / paxillin activation]
  J --> K[Migration + spreading]
  K --> H
  A --> L[Proposed intracellular axis]
  L --> M[FBXO22 interaction]
  M --> N[BACH1 stabilization]
  N --> O[Endothelial proliferation + tube formation]
  O --> H

This diagram is not decorative: each node is experimentally falsifiable with selective inhibitors, gene knockdown/knockout, and route-specific PK/PD mapping. [18]

The oncology/angiogenesis tension

BPC-157 is often described as pro-angiogenic in repair contexts. [19] “Pro-angiogenic” does not automatically mean “pro-cancer,” but in oncology, sustained angiogenic signaling is a known risk channel; this is exactly why claims about “safe angiogenesis” require direct tumor-model evidence rather than rhetorical reassurance. [20]

Counterpoint: there is at least in vitro evidence (human melanoma cell line) suggesting BPC-157 can inhibit cell growth and blunt VEGF signaling via MAPK/ERK pathway modulation. [21] The problem is not that these data are “bad,” but that they are not enough: you can stack pro-angiogenic endothelium data next to anti-proliferative tumor-line data and still not know what happens in vivo in immunocompetent tumor-bearing animals under clinically relevant exposure durations. [22]

Preclinical evidence by tissue and system

This section is organized by tissue/system, emphasizing what models were used, what endpoints were measured, and where limitations cluster.

Gut and gastrointestinal repair

The GI literature is dense and historically central: BPC-157 was developed and discussed as an anti-ulcer / cytoprotective peptide, with multiple rodent injury models reporting improved mucosal integrity and wound repair. [23]

Representative findings include gastric ulcer protection in rats (one example: a World Journal of Gastroenterology paper reporting protective effects in a rat gastric ulcer model). [24]

GI surgical repair models exist as well: in a rat ileoileal anastomosis model, BPC-157 is reported to increase anastomosis breaking strength and promote healing. [25]

Internal/external fistula models, while not purely “gut mucosa,” are relevant to GI wound biology. A colocutaneous fistula model reports improved closure and function, with explicit interaction experiments involving the NO system (L-NAME aggravation; L-arginine partial rescue; BPC-157 counteraction). [26]

Main limitation: many GI findings are scattered across related “cytoprotection” frameworks and frequently come from overlapping author groups; large, blinded, multi-center replication sets are lacking. [27]

Tendon and ligament healing

This is the most “popular” domain in public discourse, but the real interest is methodological: tendon/ligament models have clear biomechanics endpoints, which are harder to fake than subjective symptom reports.

A transected rat Achilles tendon model reports functional index improvements, increased load to failure, and histologic signs of improved collagen/fibroblast organization across a 14-day window after daily intraperitoneal dosing (reported across microgram to picogram dosing levels). [28]

A medial collateral ligament transection model reports reduced instability/contracture and improved biomechanical parameters; dosing approaches include intraperitoneal, topical (cream), and per-oral routes (drinking water). [29]

Mechanistic tendon-cell work supports a migration/adhesion axis: BPC-157 increased tendon-fibroblast migration/spreading and increased phosphorylation of FAK and paxillin. [15]

Main limitations: (a) the very low-dose potency claims are unusual; (b) many experiments bundle multiple endpoints without preregistered primary outcomes; (c) dosage–response relationships and exposure confirmation (plasma/tissue levels) are often underdeveloped. [30]

Muscle repair

Muscle injury models include contusion/crush paradigms and models with impaired healing (e.g., corticosteroid exposure). In a rat gastrocnemius injury model, BPC-157 is reported to accelerate healing and counteract methylprednisolone-aggravated impairment, measured functionally, macroscopically, and histologically. [31]

Because muscle has stronger baseline regenerative capacity than tendon, the most informative experiments are those that quantify function (strength, gait metrics) and compare against established pro-repair controls. Many BPC-157 studies report strong effects, but cross-lab replication and dose confirmation are again the thin points. [32]

Nerve and spinal cord

Peripheral nerve repair: in a rat transected sciatic nerve model, BPC-157 is reported to improve outcomes when delivered intraperitoneally, intragastrically, locally at the anastomosis, or directly into a nerve tube (for non-anastomosed nerve gap conditions). [33]

Spinal cord injury: a rat study reports improved neurologic recovery (e.g., tail function/spasticity) with BPC-157 treatment, with the usual caveat that rodent SCI improvement claims often fail translation without stringent blinding and standardized behavioral testing. [34]

Main limitations: SCI literature is notorious for inflated effect sizes when blinding and allocation concealment are weak; the BPC-157 SCI space should be treated as “interesting, not proven.” [35]

Bone and connective-tissue interfaces

The bone literature is less cleanly represented in my retrieved primary set than tendon/ligament, but multiple reviews cite animal evidence for pseudoarthrosis healing and reduced bone resorption in specific bone-loss models. [36]

One practical bridge: tendon-to-bone and junctional healing. The musculoskeletal literature includes tendon-to-bone and myotendinous junction themes, which are biologically “bone-adjacent” and arguably closer to near-term orthopedic relevance than pure fracture union claims. [37]

Main limitations: for bone proper, the field needs modern micro-CT endpoints, biomechanical testing of union, and clear comparator arms (BMPs, PTH analogs, etc.)—not just histologic impressions. [38]

Cardiovascular and vascular injury biology

There are two distinct “cardio” tracks:

1) Endothelial function / angiogenesis, supported by VEGFR2–Akt–eNOS activation, NO-dependent vasomotor tone changes, and ischemia perfusion recovery in hindlimb models. [39]

2) Whole-animal organ injury models reporting improvements in thrombosis/bleeding contexts and in myocardial injury models.

For myocardial injury, one rat model reports that BPC-157 may counteract isoprenaline-induced myocardial infarction. [40]

Vascular thrombosis/bleeding claims are often presented within a broader cytoprotection/NO-system framework; the most cautious interpretation is that BPC-157 affects vascular integrity signaling and microcirculatory resilience, but the exact boundary condition (dose, duration, target receptor) remains unresolved. [41]

Liver and kidney protection

A historically cited hepatoprotective line of work dates back to early 1990s models involving stress or bile duct/hepatic artery ligation or CCl4 exposure (reported as hepatoprotective in rats). [42]

More recent work includes models framed as ischemia–reperfusion and “distant organ damage,” where BPC-157 is reported to reduce injury across organs including liver and kidney. [43]

Main limitation: “multi-organ protection” can blur mechanisms. Without tissue exposure quantification and target engagement assays, protective findings can represent many things—anti-inflammatory tone shift, microvascular effects, or even confounding due to experimental handling. [44]

Table of representative preclinical studies

The table below is deliberately selective. It does not imply the omitted studies are low quality; it reflects what can be summarized cleanly with accessible primary records and clear endpoints.

Human evidence and clinical-trial record

Published human studies

Intra-articular knee pain (retrospective chart review):
A small clinic-based chart review reported intra-articular BPC-157 injections for heterogeneous knee pain. The paper reports a BPC-157 dose of 2 cc or 4 mg (2000 mcg/mL) for BPC-157 alone, with some patients receiving BPC-157 plus thymosin beta-4 at varying doses; injections were performed without ultrasound guidance. [49] Outcomes were subjective (“improvement” categories) and the cohort lacked standardized diagnostic workup and validated pain instruments, limiting interpretability. [50]

Interstitial cystitis (retrospective chart review):
A 12-woman series reports cystoscopy plus injections around the inflammatory bladder area during a single procedure, using a total of 10 mg BPC-157; outcomes were captured via the Global Response Assessment questionnaire and follow-up cystoscopy. [51] The study is not randomized, and placebo effects in pain syndromes are large; still, it is at least a documented human exposure report. [52]

IV infusion safety (pilot study):
A 2-participant pilot reports 10 mg BPC-157 in 250 cc normal saline infused over 1 hour, followed by 20 mg in 250 cc over 1 hour on a subsequent day, with repeated labs and vital signs; the authors report no clinically meaningful biomarker changes across measured panels. [53] This is safety-relevant but far too small for any robust adverse-event inference. [54]

Bottom line: these papers document human use, not human efficacy, and they should not be used as dosing standards. Human dosing standards are not established. [55]

ClinicalTrials.gov record and the gap between registration and knowledge

A Phase I record exists for “PCO-02 – Safety and Pharmacokinetics Trial” (NCT02637284). [56] The registry interface available here does not present structured protocol/outcome details or results; the “results” view is effectively a placeholder (“No Study Results Posted”). [57]

Secondary sources (including a 2025 PMC review by the original research group) describe the trial as involving 42 volunteers and claim oral dosing regimens (1/3/6/9 mg daily for up to two weeks) with “safe/well-tolerated” conclusions and no quantifiable BPC-157 in plasma/urine. Those details may be accurate, but because they are not verifiable here as primary registry results, I treat them as unconfirmed until independently posted or published. [58]

Dosing, formulations, and routes in studies

What the literature actually uses

Across animal and cell studies, BPC-157 is often reported as administered without carriers (saline solution, drinking water, or neutral creams for topical application), with frequent emphasis that results are consistent across routes. [59]

Common preclinical routes include:

• Intraperitoneal (i.p.) daily dosing in musculoskeletal injury models (tendon; ligament; muscle; nerve). [60]
• Per-oral exposure (often via drinking water) in some ligament and osteoarthritis-type studies. [61]
• Local application at the injury site, including topical cream formulations for ligament or muscle injury models. [62]
• In vitro concentrations spanning ng/mL to µg/mL in endothelial/tendon cell studies. [63]

Human studies (again: not standards) used: - Intra-articular: 4 mg (2 cc; 2000 mcg/mL) BPC-157 for knee pain in the BPC-157-only group. [49]
- Procedural bladder injections: total 10 mg around inflamed bladder area during cystoscopy. [51]
- IV infusion: 10 mg then 20 mg, each in 250 cc saline over 1 hour (pilot). [53]

Pharmacokinetics and metabolism

PK/ADME data remain a weak spot relative to the breadth of efficacy claims, but at least one dedicated PK paper reports:

• Rapid systemic clearance: reported plasma half-life < 30 minutes in rats after IV dosing. [64]
• Dose-proportional exposure in tested ranges, and bioavailability estimates for IM dosing (reported at ~14–19% in rats and ~45–51% in beagle dogs). [64]
• Excretion patterns: major recovery via urine and bile in rats, and metabolism into smaller peptide fragments and amino acids. [64]

This matters because many preclinical efficacy papers imply sustained repair effects from brief exposure. If those effects are real, the mechanism likely involves early “switch flipping” (gene expression programs, vascular remodeling triggers) rather than prolonged receptor occupancy—another testable prediction. [65]

Safety, toxicity, and regulatory status

Toxicology signal

A 2020 preclinical safety evaluation (Regulatory Toxicology and Pharmacology) reports nonclinical safety testing in animals with “no observed toxicity,” including mention of no systemic anaphylaxis in guinea pigs, no hemolysis, no vascular irritation, negative mutagenicity/genotoxicity signals in the tests performed, and stated margins suggesting a wide safety window relative to proposed human doses for trials. [66]

That is a meaningful data point, but it does not erase the practical safety problems in real-world “research chemical” supply chains (purity, endotoxin, mislabeling, contamination), which is exactly why regulators have flagged compounded products as risky. [67]

Regulatory and anti-doping status

• In the United States, BPC-157 is not FDA-approved for any indication, and FDA has listed BPC-157 on its “Category 2” bulk substances list (substances that may present significant safety risks for compounding). [67]
• BPC-157 is treated as a prohibited substance in sport under World Anti-Doping Agency[68] frameworks, commonly described as falling under the “S0: Non-Approved Substances” category in the Prohibited List. [69]
• U.S. Anti-Doping Agency[70] has public guidance noting BPC-157 is prohibited and highlighting that purported benefits are largely based on animal studies with limited human research and that available “studies” are often inconclusive or unpublished. [71]

Detectability and research ethics around misuse

Anti-doping labs have published analytical work on detection and in vitro metabolism of confiscated peptides including BPC-157, reinforcing that “undetectable” narratives are not credible and that any human-use ecosystem intersects with testing science. [7]

For an institutional researcher, this matters ethically: once a compound is culturally positioned as a performance enhancer, your work can be co-opted. That raises burdens for clear disclaimers, data transparency, and careful discussion of routes/doses in publications. [72]

Research gaps and next-step experimental program

The gap is not “we need more positive animal studies.” The gap is credibility under adversity: independence, reproducibility, and mechanistic falsification.

What a serious next-step program would prioritize

Replication-grade core experiments (multi-lab):

1) Pick two flagship models with hard endpoints:
- Tendon transection with biomechanical testing (load to failure, stiffness, modulus) [28]
- Hindlimb ischemia perfusion recovery + histologic vessel quantification [73]

2) Pre-register: primary endpoint, analysis plan, exclusions, stopping rules. [74]

3) Dose–response reality check: include at least 3 doses plus vehicle, and confirm exposure (LC–MS/MS quant or labeled tracer) so the model is not purely “story-driven.” [75]

Mechanistic falsification experiments (not just “associated with”):

• VEGFR2 internalization dependence: repeat key endothelial outcomes with endocytosis perturbation (dynasore-like design) and measure downstream Akt/eNOS activation. [13]
• NO dependence: use pharmacologic NO scavenging or eNOS inhibition and test whether functional repair endpoints collapse. [76]
• FBXO22/BACH1 axis (2026): attempt CRISPRi/siRNA FBXO22 knockdown in endothelial assays and test whether BPC-157 loses proangiogenic effects, as predicted. [3]

Safety/oncology boundary experiments:

• A minimal credible test would include an immunocompetent mouse tumor model with clinically plausible exposure windows, tracking tumor growth, vascular density, and metastasis signals—designed to detect both acceleration and inhibition. The current discourse includes both pro-angiogenic repair data and anti-proliferative melanoma-cell data; only in vivo tumor biology can reconcile them. [77]

Practical guidance for researchers: study design, endpoints, assays, sample sizes

I’ll keep this pragmatic and conservative. If you can’t afford full mechanistic depth, at least design studies so your results are interpretable.

Design basics (non-negotiable):

• Follow ARRIVE 2.0 reporting expectations (randomization, blinding, sample-size justification, inclusion/exclusion transparency). [78]
• Use PREPARE principles at planning stage to reduce avoidable noise (housing, welfare, perioperative consistency). [79]
• If human studies are contemplated, align with CONSORT expectations for trial transparency. [80]

Endpoints that actually discriminate signal from wishful thinking:

• Tendon/ligament: load to failure, stiffness, modulus; blinded histologic scoring + collagen organization quant (polarized light) [81]
• Muscle: force production (in situ), locomotor metrics, cross-sectional area, centralized nucleation proportion [31]
• Nerve: electrophysiology (NCV), functional indices, axon counts + myelination metrics [33]
• Gut: permeability assay, histology with blinded scoring, inflammatory cytokine panels with correction for multiplicity [82]
• Vascular/ischemia: perfusion (laser Doppler), vessel density, VEGFR2/Akt/eNOS activation panels, NO proxies [39]

Assays to confirm target engagement / mechanistic direction:

• Westerns or phospho-ELISAs for pVEGFR2 / pAkt / peNOS, and co-IP for Cav-1/eNOS binding when relevant. [83]
• LC–MS/MS quantification where possible (even if absolute quant is hard; relative exposure across dose groups is often enough to validate PK plausibility). [75]

Sample sizes (how to avoid nonsense):

• Do an a priori power calculation. If you don’t have variance estimates, run a pilot to estimate SD and operational variability, then power for a single primary endpoint (not 12 co-primaries). [84]
• As a pragmatic starting point for continuous biomechanical endpoints in rodent injury models, many labs land in the n≈8–12 per group band to detect moderate-to-large effects, but you should treat that as a placeholder until you compute power from your own variance and expected minimal clinically important difference (MCID) equivalent for the model. [84]

Recommended experimental protocols table

These are research protocols, not human-use recommendations. The goal is reproducible signal detection with mechanistic leverage.

A timeline of where BPC-157 stands as of early 2026

timeline
  title BPC-157: research and regulatory milestones (selected)
  1993 : Early reports describe BPC-157 as a gastric-derived pentadecapeptide with organ-protective effects
  2003 : Rat Achilles tendon transection model reports large functional/biomechanical improvements
  2010 : Rat MCL transection study reports improved ligament biomechanics across routes
  2015 : Murine wound-healing work reports VEGF/ERK-related changes in skin injury model
  2017 : VEGFR2 internalization → Akt–eNOS mechanism published; anti-doping detection work published
  2020 : Preclinical safety evaluation published; endothelial NO/Src–Cav-1–eNOS signaling explored
  2024-2025 : Small human clinic reports (knee, interstitial cystitis, IV infusion) appear
  2026 : Proposed BPC157–FBXO22–BACH1 axis published; WADA 2026 Prohibited List remains in force

This timeline compresses a sprawling literature into a few “load-bearing” datapoints—useful for deciding where to invest replication effort. [88]

Final research-grade conclusion

If you’re a researcher, BPC-157 is interesting precisely because it sits at the intersection of too many positive outcomes and too little translational confirmation. The best use of your time is not to amplify the hype. It is to find the smallest set of experiments that can kill the hypothesis—or force it to become real through reproducibility, target engagement, and careful boundary testing (especially oncology and thrombosis/bleeding edges). [89]

And if you’re tempted to treat “human case reports” as a substitute for real trials: don’t. The human literature is currently too small and too weak to set dosing norms, establish safety, or claim efficacy. [90]

[1] [12] [27] [89] https://pubmed.ncbi.nlm.nih.gov/30915550/

https://pubmed.ncbi.nlm.nih.gov/30915550/

[2] [13] [18] [19] [22] [39] [73] [77] [83] https://pubmed.ncbi.nlm.nih.gov/27847966/

https://pubmed.ncbi.nlm.nih.gov/27847966/

[3] [8] https://link.springer.com/article/10.1186/s12964-026-02694-6

https://link.springer.com/article/10.1186/s12964-026-02694-6

[4] [49] [50] [90] https://asipp.org/wp-content/uploads/Intra-Articular-Injection-of-BPC-157-for-Multiple-Types-of-Knee-Pain-2021-Alternative-Therapies-in-Health-and-Medicine.pdf

https://asipp.org/wp-content/uploads/Intra-Articular-Injection-of-BPC-157-for-Multiple-Types-of-Knee-Pain-2021-Alternative-Therapies-in-Health-and-Medicine.pdf

[5] [56] https://clinicaltrials.gov/study/NCT02637284

https://clinicaltrials.gov/study/NCT02637284

[6] [67] https://www.fda.gov/drugs/human-drug-compounding/certain-bulk-drug-substances-use-compounding-may-present-significant-safety-risks

https://www.fda.gov/drugs/human-drug-compounding/certain-bulk-drug-substances-use-compounding-may-present-significant-safety-risks

[7] [72] https://pubmed.ncbi.nlm.nih.gov/28035768/

https://pubmed.ncbi.nlm.nih.gov/28035768/

[9] [28] [30] [59] [60] [81] [85] [88] https://pubmed.ncbi.nlm.nih.gov/14554208/

https://pubmed.ncbi.nlm.nih.gov/14554208/

[10] [16] [23] https://pubmed.ncbi.nlm.nih.gov/21548867/

https://pubmed.ncbi.nlm.nih.gov/21548867/

[11] [44] [64] [65] [75] https://www.frontiersin.org/journals/pharmacology/articles/10.3389/fphar.2022.1026182/full

https://www.frontiersin.org/journals/pharmacology/articles/10.3389/fphar.2022.1026182/full

[14] [47] [68] [76] https://pmc.ncbi.nlm.nih.gov/articles/PMC7555539/

https://pmc.ncbi.nlm.nih.gov/articles/PMC7555539/

[15] [63] https://pubmed.ncbi.nlm.nih.gov/21030672/

https://pubmed.ncbi.nlm.nih.gov/21030672/

[17] [20] https://pubmed.ncbi.nlm.nih.gov/23782145/

https://pubmed.ncbi.nlm.nih.gov/23782145/

[21] https://journals.lww.com/melanomaresearch/fulltext/2004/08000/bpc_157_inhibits_cell_growth_and_vegf_signalling.50.aspx

https://journals.lww.com/melanomaresearch/fulltext/2004/08000/bpc_157_inhibits_cell_growth_and_vegf_signalling.50.aspx

[24] https://www.wjgnet.com/1007-9327/full/v10/i7/1032-T4.htm

https://www.wjgnet.com/1007-9327/full/v10/i7/1032-T4.htm

[25] https://www.researchgate.net/publication/388836839_Protective_Effects_of_BPC_157_on_Liver_Kidney_and_Lung_Distant_Organ_Damage_in_Rats_with_Experimental_Lower-Extremity_Ischemia-Reperfusion_Injury

https://www.researchgate.net/publication/388836839_Protective_Effects_of_BPC_157_on_Liver_Kidney_and_Lung_Distant_Organ_Damage_in_Rats_with_Experimental_Lower-Extremity_Ischemia-Reperfusion_Injury

[26] https://www.jstage.jst.go.jp/article/jphs/108/1/108_FP0072161/_pdf

https://www.jstage.jst.go.jp/article/jphs/108/1/108_FP0072161/_pdf

[29] [45] [61] [62] [86] https://pubmed.ncbi.nlm.nih.gov/20225319/

https://pubmed.ncbi.nlm.nih.gov/20225319/

[31] [32] https://pubmed.ncbi.nlm.nih.gov/20190676/

https://pubmed.ncbi.nlm.nih.gov/20190676/

[33] [46] https://pubmed.ncbi.nlm.nih.gov/19903499/

https://pubmed.ncbi.nlm.nih.gov/19903499/

[34] https://link.springer.com/article/10.1186/s13018-019-1242-6

https://link.springer.com/article/10.1186/s13018-019-1242-6

[35] https://pmc.ncbi.nlm.nih.gov/articles/PMC9164058/

https://pmc.ncbi.nlm.nih.gov/articles/PMC9164058/

[36] https://www.mdpi.com/2227-9059/9/11/1547

https://www.mdpi.com/2227-9059/9/11/1547

[37] https://pmc.ncbi.nlm.nih.gov/articles/PMC9687817/

https://pmc.ncbi.nlm.nih.gov/articles/PMC9687817/

[38] [74] [78] https://arriveguidelines.org/sites/arrive/files/documents/ARRIVE%20guidelines%202.0%20-%20English.pdf

https://arriveguidelines.org/sites/arrive/files/documents/ARRIVE%20guidelines%202.0%20-%20English.pdf

[40] https://pubmed.ncbi.nlm.nih.gov/36359218/

https://pubmed.ncbi.nlm.nih.gov/36359218/

[41] https://pubmed.ncbi.nlm.nih.gov/23755725/

https://pubmed.ncbi.nlm.nih.gov/23755725/

[42] [70] https://link.springer.com/article/10.1007/s10787-999-0022-z

https://link.springer.com/article/10.1007/s10787-999-0022-z

[43] https://pmc.ncbi.nlm.nih.gov/articles/PMC11857380/

https://pmc.ncbi.nlm.nih.gov/articles/PMC11857380/

[48] https://pubmed.ncbi.nlm.nih.gov/25995620/

https://pubmed.ncbi.nlm.nih.gov/25995620/

[51] [52] https://www.alternative-therapies.com/oa/pdf/11423.pdf

https://www.alternative-therapies.com/oa/pdf/11423.pdf

[53] [54] https://www.alternative-therapies.com/oa/pdf/11513.pdf

https://www.alternative-therapies.com/oa/pdf/11513.pdf

[55] https://www.opss.org/article/bpc-157-prohibited-peptide-and-unapproved-drug-found-health-and-wellness-products

https://www.opss.org/article/bpc-157-prohibited-peptide-and-unapproved-drug-found-health-and-wellness-products

[57] https://www.clinicaltrials.gov/study/NCT02637284?tab=results

https://www.clinicaltrials.gov/study/NCT02637284?tab=results

[58] https://pmc.ncbi.nlm.nih.gov/articles/PMC12195719/

https://pmc.ncbi.nlm.nih.gov/articles/PMC12195719/

[66] https://www.sciencedirect.com/science/article/abs/pii/S027323002030091X

https://www.sciencedirect.com/science/article/abs/pii/S027323002030091X

[69] https://www.wada-ama.org/sites/default/files/2025-09/2026list_en_final_clean_september_2025.pdf

https://www.wada-ama.org/sites/default/files/2025-09/2026list_en_final_clean_september_2025.pdf

[71] https://www.usada.org/spirit-of-sport/bpc-157-peptide-prohibited/

https://www.usada.org/spirit-of-sport/bpc-157-peptide-prohibited/

[79] https://pubmed.ncbi.nlm.nih.gov/28771074/

https://pubmed.ncbi.nlm.nih.gov/28771074/

[80] https://pmc.ncbi.nlm.nih.gov/articles/PMC2844943/

https://pmc.ncbi.nlm.nih.gov/articles/PMC2844943/

[82] [87] https://pmc.ncbi.nlm.nih.gov/articles/PMC4717094/

https://pmc.ncbi.nlm.nih.gov/articles/PMC4717094/

[84] https://eda.nc3rs.org.uk/experimental-design-group

https://eda.nc3rs.org.uk/experimental-design-group

Mechanistic map 

Preclinical studies

Recommended experimental protocols

Timeline