Compound Physician required WADA prohibited Part of Peptides

BPC-157

Synthetic peptide fragment studied for tissue repair. Animal data promising; human evidence sparse.

The story

BPC-157 begins with a hypothesis rather than a molecule. In 1975, Predrag Sikirić, then a second-year medical student at the University of Zagreb School of Medicine, reasoned that because stress reliably damages the stomach and the stomach reliably survives, the organ must produce something that protects it. He gave the hypothetical substance a placeholder name — Substancija Bože Pomozi, roughly "substance God help me" — and spent the next decade trying to find it (STAT, 2026-06-01).

By 1983 he had assembled a team. Their method was to collect human and porcine gastric juice in volume — from gastroenterology clinics, emergency rooms, a hospital in Split that shipped material to Zagreb, and pig slaughterhouses — store it in university refrigerators, fractionate it, and test the fractions in animal models. Sikirić has described slaughterhouse material that occasionally "harbored rats that had been caught and eaten by the pigs." By 1989 the group reported an active 15-amino-acid fragment, GEPPPGKPADDAGLV, which they called BPC-157: the 157th fraction of body protection compound. That same year Sikirić met Sandor Szabo, a Harvard-affiliated gastric-injury researcher, at a Canadian conference; Szabo became a long-running collaborator (STAT, 2026-06-01).

The origin story is also the compound's most durable scientific dispute. The full sequence of the claimed parent protein was never published, which makes the isolation unreproducible. Anna Mapp, president of the American Peptide Society, has noted that genetic material coding for the BPC-157 sequence has not been found in the human genome or gut microbiome; Patricia Brubaker of the University of Toronto has written that the 1994 patent "gives exceedingly little information about the isolation process" and used crude chromatographic methods, so it is impossible to know whether what was isolated was pure (McGill OSS). Szabo himself co-authored a 2017 paper suggesting the team may have misread an amino acid sequence decades earlier, and has since questioned whether the body produces BPC-157 at all. Whether BPC-157 is endogenous remains unresolved and should be treated as contested, not settled.

The development arc is conventional and it ended badly. In 1993 the Croatian pharmaceutical company PLIVA partnered with Sikirić, later with the American firm Parke-Davis, developing the peptide as PL 14736 for inflammatory bowel disease. Animal results published from 1995 onward were positive, and PLIVA carried the program into human trials for ulcerative colitis — the only sustained clinical development BPC-157 has ever received. Complete results were never published. When GSK acquired PLIVA's research institute in 2006, it dropped the program. Sikirić attributes this to corporate politics; Michael Parnham, a former senior PLIVA scientific adviser, has said the compound "barely outperformed" existing drugs and would have been hard to market. Rights and data reverted to Sikirić in 2009, and research since has been overwhelmingly rodent work from his own group (STAT, 2026-06-01).

The route into fitness and longevity culture is well documented. Around 2010, bodybuilders on forums including Bodybuilding.com and ThinkSteroids.com found Sikirić's published papers and began sourcing the peptide from Chinese suppliers; dedicated Reddit peptide communities formed by 2018 (STAT, 2026-02-03). Edwin Lee, a Florida endocrinologist, became a leading clinical advocate from 2016, founding the Clinical Peptide Society and the nonprofit Save Peptides — and is the first author on all three published human studies. Andrew Huberman released a notably cautious episode in April 2024 that repeatedly flagged cancer and tumor risk, then later that year hosted physician Craig Koniver, who described BPC-157 as "awesome, super-safe." A November 2024 MAHA summit in Washington featured peptide advocacy panels; Gary Brecka promoted the compound to a then-FDA commissioner on his podcast in December 2025; HHS Secretary Robert F. Kennedy Jr. has argued publicly for access (STAT, 2026-02-03; CNN, 2025-11-15). That advocacy reached a procedural peak in July 2026, when an FDA advisory committee voted narrowly in the compound's favor (STAT, 2026-07-23).

  1. 1975 Sikirić, a medical student at Zagreb, hypothesizes an endogenous gastric protective substance
  2. 1983 Research team formed; multi-year collection of human and porcine gastric juice begins
  3. 1989 15-amino-acid fragment identified and named BPC-157; Sikirić meets Sandor Szabo
  4. 1993 PLIVA (Croatia) partners with Sikirić; development as PL 14736 for IBD begins, later with Parke-Davis
  5. 1994 Patent filed; isolation process described in minimal detail, later a core criticism
  6. 2003 Staresinic et al. publish accelerated healing of transected rat Achilles tendon — the founding musculoskeletal paper
  7. 2004 Radeljak et al. report BPC-157 inhibits growth and VEGF signalling in a human melanoma cell line — later the most-cited "it's not carcinogenic" datum
  8. 2006 GSK acquires PLIVA's research institute and drops BPC-157; PL 14736 human results never fully published
  9. ~2010 Bodybuilding forums adopt BPC-157; gray-market sourcing from Chinese suppliers begins
  10. 2015–2016 Phase I safety/PK trial NCT02637284 — registered under the name PharmaCotherapia d.o.o., n=42 planned — registered, then abandoned with no published results
  11. 2022-01-01 Added to the WADA Prohibited List under S0, Non-Approved Substances; prohibited at all times
  12. 2021–2025 The entire published human literature appears: three uncontrolled pilot studies from one Florida clinic
  13. 2025 Published dispute in Pharmaceuticals: Józwiak et al. review → Sikirić et al. comment → Józwiak et al. reply
  14. 2026-07-23 FDA Pharmacy Compounding Advisory Committee votes 8–6 to recommend BPC-157 for compounding, limited to ulcerative colitis; non-binding

What it does

A synthetic peptide fragment studied for tissue repair. Animal data is promising; human evidence is genuinely thin and the supplier market is largely unregulated.

A synthetic fragment of body protection compound, hypothesized to support angiogenesis and tissue repair.

  • tissue repair
  • injury recovery

What the evidence shows

After thirty-seven years there has never been a randomized controlled trial of BPC-157, and none is known to exist: the entire published human record is three uncontrolled pilot studies totalling thirty people (n=2, n=12, n=16), all from one Florida clinic, all sharing one first author, all in one journal. The 2025 systematic review that applied inclusion criteria found 36 studies, of which 35 were animal or in vitro — the base is preclinical-dominant in shape, not merely in volume, so there is no controlled human result to weigh. The limitation that matters most is that the cancer question is open in both directions: no in vivo tumour study exists in any species, the reassurance rests on a single unreplicated 2004 cell-line experiment by the originating group, and the concern is mechanistic inference from the compound's own documented VEGFR2 and Akt–eNOS signalling.

Overall evidence strength: Very low certainty

  • Lee E, Padgett B (2021) — intra-articular knee pain (2021) Observational

    Most patients said their knee felt better afterward — 11 of 12 in the BPC-157-only group — but they were asked to remember how bad the pain had been 6–12 months earlier, and nobody was measured against a placebo.

    Source PMID 34324435

  • Lee E, Walker C, Ayadi B (2024) — interstitial cystitis (2024) Case series (uncontrolled)

    Ten of twelve women reported their symptoms completely gone after a single set of injections into the bladder wall, and the other two reported about 80% improvement — an implausibly uniform result for an uncontrolled study of a condition with a large placebo response.

    Source PMID 39325560

  • Lee E, Burgess K (2025) — intravenous safety pilot (2025) Case series (uncontrolled)

    In two people, single IV infusions produced no measurable change in cardiac, liver, kidney, thyroid, or glucose markers and no reported side effects — which tells you almost nothing about safety at a population level.

    Source PMID 40131143

  • Staresinic M et al. (2003) — rat Achilles tendon transection (2003) Preclinical / mechanistic

    Rats whose Achilles tendons had been cut healed faster and stronger with BPC-157 than without, and isolated tendon cells grew faster in the dish — this single paper is the origin of essentially the entire "tendon healing peptide" claim.

    Source PMID 14554208

  • Vasireddi N et al. (2025) — systematic review (2025) Review / meta-analysis

    Someone finally counted, and the answer is that after thirty years there is one human study in orthopaedics and no human safety data at all; the authors advise clinicians and athletes to be cautious.

    Source PMID 40756949

  • McGuire FP et al. (2025) — narrative/scoping review (2025) Review / meta-analysis

    The mechanism story is genuinely coherent — BPC-157 acts on VEGFR2 and the Akt–eNOS axis, promoting new blood vessels in poorly vascularized tissue like tendon — but the authors conclude it "should be considered investigational" until real trials exist.

    Source PMID 40789979

Common misconceptions

“BPC-157 has been shown not to cause cancer — it actually inhibits tumors.”

No study, in any species, has evaluated whether BPC-157 promotes or inhibits tumor development in a living animal with cancer. The "it inhibits tumors" claim traces to a single 2004 cell-line experiment in which BPC-157 inhibited growth and VEGF signalling in cultured human melanoma cells (DOI 10.1097/00008390-200408000-00050) — one dish of cells, one cell line, never replicated. Józwiak et al. state the anti-tumor assertion lacks supporting in vivo evidence and rests on "a single melanoma cell-line experiment from 2004 (unreplicated)" (DOI 10.3390/ph18101451). The concern in the other direction is mechanistic and real: BPC-157's documented mode of action runs through VEGFR2, Akt–eNOS, ERK1/2 and KRAS — the same pro-angiogenic and pro-growth pathways tumors exploit (McGuire et al. 2025). Sikirić's group argues in print that oncologic risk is "entirely excluded" and points to cornea experiments where the peptide opposed rather than caused neovascularization (DOI 10.3390/ph18101450); Józwiak et al. reply that this contradicts the same group's own characterization of VEGF/eNOS activation as the peptide's key mechanism (DOI 10.3390/ph18101451). Honest answer: The cancer question is open, not resolved in either direction. Anyone claiming BPC-157 is proven safe for cancer risk, and anyone claiming it is proven to cause cancer, is going beyond the evidence. Patricia Brubaker's framing is the useful one: "not every compound that is produced by the body is 'good'" — EGF supports infant gut development and aggressively stimulates breast cancer cell growth in adults (McGill OSS).

“It's a natural peptide your body already makes, so it's inherently safe.”

Whether BPC-157 is endogenous is disputed by serious people. The parent protein's full sequence was never published; the coding sequence has not been located in the human genome or gut microbiome (Anna Mapp, American Peptide Society); Sandor Szabo — Sikirić's own longtime collaborator — co-authored a 2017 paper suggesting the original sequence may have been misinterpreted and now questions whether the body makes it (STAT, 2026-06-01). Separately, "natural" would not establish safety even if true.

“There are hundreds of studies on BPC-157.”

There are roughly 231 PubMed-indexed records mentioning BPC 157, and the count is real — but Józwiak et al. note that over 80% of records under "BPC 157" on Google Scholar/PubMed trace to Sikirić's group, which is the definition of an unreplicated literature (DOI 10.3390/ph18101451). The systematic review that actually applied inclusion criteria to the musculoskeletal literature found 36 studies, of which one was human (DOI 10.1177/15563316251355551).

“Human trials showed it's safe.”

The published human safety data consists of two people, observed for three days, in an open-label pilot with no control group (PMID 40131143). The 2025 systematic review states plainly that "no study assessed the safety or adverse events of BPC-157 in humans" and that no animal study looked past six weeks (DOI 10.1177/15563316251355551). The Phase I trial that would have generated real safety data, NCT02637284, was registered for 42 participants and abandoned without published results (NCT02637284 — T2 registry record: existence and status only; self-asserted, unverified).

“The FDA banned it because it works and threatens pharma.”

The historical record is that a pharmaceutical company did develop it and dropped it. PLIVA carried PL 14736 into Phase II for ulcerative colitis; GSK ended the program on acquiring PLIVA's institute in 2006. Sikirić attributes this to corporate politics; Michael Parnham, a former senior PLIVA scientific adviser, says the compound "barely outperformed" existing drugs (STAT, 2026-06-01). Christopher Milne of the Tufts Center for the Study of Drug Development has noted a different modern barrier: widespread gray-market availability makes running a randomized trial logistically difficult (STAT, 2026-02-03).

“What you buy is what's on the label.”

ECRI and ISMP reported in April 2026 that testing of gray-market peptide products found purity ranging from 5% to 75%, with arsenic and lead contamination exceeding toxicity thresholds (ECRI/ISMP, 2026). The ECRI/ISMP white paper is an expert-body publication — alongside T1 for product-quality and market-conduct facts like these, and never a basis for a clinical efficacy or safety figure; it is used in this file only for the former. The 2025 systematic review cites the broader finding that between 12% and 58% of ergo-nutritional supplements are contaminated with other, often unsafe substances (DOI 10.1177/15563316251355551). Note that the ECRI purity and contamination figures describe the gray peptide market generally; we have not found them attributed to BPC-157 specifically, and we do not attribute them to it here. ---

Who it's for, who should skip

Experimental only. Regulatory status is gray_zone — removed from Category 2 in April 2026 and PCAC-recommended, but compounding remains unlawful pending rulemaking. Any use belongs in a physician-supervised program with third-party purity testing, not self-sourcing.

Running it

No dosing protocol is published for this compound — there is no lawful supervised channel for the use this guide covers. What follows is monitoring, expectations, and safety framing only.

Form
injectable

What to expect

  • Onset (days) — No controlled human data. In the interstitial cystitis pilot, participants reported symptom resolution after a single procedure, assessed at follow-up rather than serially (PMID 39325560) (Confidence: Trial-reported but uncontrolled)
  • Onset (weeks) — Widely reported in forum and clinic settings as "1–2 weeks for soft-tissue pain," with no trial support of any kind (Confidence: Anecdotal, low confidence)
  • Duration of effect — Lee & Padgett report patients recalling relief lasting more than 6 months after a single intra-articular injection; recall was retrospective at 6–12 months, with no objective endpoint (PMID 34324435) (Confidence: Trial-reported, uncontrolled, recall-based)
  • Plateau — Not studied in any species. No dose-response or duration-response experiment exists — published animal work uses single fixed doses (DOI 10.3390/ph18101451) (Confidence: No data)
  • Post-cessation — Not studied. Plasma half-life is under 30 minutes in rats and dogs, so systemic exposure clears quickly, but nothing documents what happens to any claimed benefit afterward (DOI 10.3390/ph18101451) (Confidence: No data)
  • Long-term — No study in any species assessed adverse events beyond six weeks (DOI 10.1177/15563316251355551) (Confidence: No data)
  • First real answer expected — No credible pending readout identified. No randomized controlled trial of BPC-157 is known to exist; a registry record asserting one was withdrawn from this dossier on 2026-08-29 as unverifiable. Nothing on the published record points to a forthcoming controlled human efficacy answer (Confidence: No data)

Response signs

Likely working

  • Reduction in localized joint pain following direct intra-articular injection, reported at 6–12 month recall — Trial-reported, uncontrolled, retrospective recall, no validated instrument · PMID 34324435 Trial-reported
  • Resolution of bladder pain / urgency symptoms after intravesical injection in refractory interstitial cystitis — Trial-reported, uncontrolled, open-label, subjective endpoint · PMID 39325560 Trial-reported
  • Faster return to activity after soft-tissue injury — Anecdotal, low confidence — the dominant claim in circulation, with no human trial support · DOI 10.1177/15563316251355551 Anecdotal
  • Reduced GI discomfort / "gut healing" — Anecdotal, low confidence — inherited from the discontinued IBD program, unpublished results · STAT Anecdotal

Adverse — seek review

  • Severe or escalating anxiety; anhedonia (loss of pleasure/motivation) — Anecdotal, low confidence — documented as user reports in reporting on Reddit communities, never in a clinical study · STAT, 2026-02-03 Anecdotal
  • Intense itching — Anecdotal, low confidence — same source class · STAT, 2026-02-03 Anecdotal
  • Injection-site reaction, particularly spreading, warm, or persisting — Anecdotal, but note FDA's stated concern that compounded BPC-157 "may pose risk for immunogenicity for certain routes of administration" — an immune reaction is the mechanistically plausible version of this sign · DOI 10.1177/15563316251355551; FDA Anecdotal
  • Any systemic reaction — fever, rash, malaise — after a dose — Inference, not documented for BPC-157. Flagged because gray-market product purity has been measured as low as 5%, meaning most of what is in a vial may not be the labeled peptide · ECRI/ISMP, 2026 Inference

Common / neutral

  • No perceptible subjective effect at all — Consistent with the published record — BPC-157 has no acute perceptual signature; the IV pilot reported no side effects and no biomarker movement in either participant · PMID 40131143
  • Mood disturbance without clear direction — Anecdotal, low confidence — reported inconsistently in both directions · DOI 10.1177/15563316251355551 Anecdotal

Getting it right

  • Monitoring: Establish a baseline before, not after. The only human study that measured anything measured cardiac, hepatic, renal, thyroid and glucose markers before and after exposure (PMID 40131143). Whatever a clinician chooses to track, tracking it from a pre-exposure baseline is what makes any later change interpretable. mitigate side effect
  • Monitoring: Accept that there is no efficacy biomarker. Nothing validated predicts or confirms response. Any monitoring plan here is safety monitoring, not effectiveness monitoring. mitigate side effect
  • Monitoring: Time-box the question. No study in any species assessed adverse events past six weeks (DOI 10.1177/15563316251355551). Open-ended use sits entirely outside the observed window of every study ever conducted. mitigate side effect
  • Monitoring: Age- and history-appropriate cancer surveillance is the reasonable response to an unresolved angiogenesis question — not because a signal has been documented, but because the question has never been asked in a living animal with a tumor. mitigate side effect
  • General safety: Product identity is the dominant risk, ahead of pharmacology. Purity in tested gray-market peptide products ranged from 5% to 75%, with arsenic and lead above toxicity thresholds (ECRI/ISMP, 2026). A certificate of analysis supplied by the seller is not independent verification. mitigate side effect
  • General safety: FDA has stated a specific concern that compounded BPC-157 "may pose risk for immunogenicity for certain routes of administration" (FDA). mitigate side effect
  • General safety: Combination use compounds the unknowns. No stack containing BPC-157 has been studied against its components; adverse events in a multi-compound regimen cannot be attributed. mitigate side effect
  • General safety: Athletes and service members face strict liability. See — this is not a risk that intent or ignorance mitigates. mitigate side effect
  • The physician conversation: What is the specific injury or condition, and has it been imaged or diagnosed? (Tendinopathy, partial tear, and referred pain have different natural histories and different evidence-backed treatments.) mitigate side effect
  • The physician conversation: What conservative options with randomized-trial support have actually been tried to completion? (— eccentric loading and physical therapy have the evidence BPC-157 lacks.) mitigate side effect
  • The physician conversation: What is the personal and family cancer history? (is an open question, and the answer changes how open it should feel.) mitigate side effect
  • The physician conversation: What else is being taken, including anything from a non-pharmacy source? mitigate side effect
  • The physician conversation: What would count as it working, and by when — and what is the plan if that date passes? mitigate side effect
  • The physician conversation: Is there any competitive sport, collegiate, or military testing obligation? (.) mitigate side effect

Safety

  • highly variable product purity across the supplier market
  • anaphylaxis risk with injection of unknown compounds

Stacking & alternatives

  • Complements TB-500 — The canonical pairing, marketed together as the "Wolverine stack." Lee & Padgett's knee series actually included 4 patients who received BPC-157 + TB-500 rather than BPC-157 alone — the only human data on the combination, and it performed slightly worse (3 of 4 improved vs 11 of 12) in a sample far too small to mean anything (PMID 34324435)
  • Alternative to GHK-Cu — Also positioned for tissue repair, with a different mechanism (copper-peptide, ECM remodeling) and a longer cosmetic-science record; competes for the same "recovery peptide" intent
  • Alternative to KPV — Positioned for the gut/anti-inflammatory use case that BPC-157 inherited from the PL 14736 IBD program. Also went to the July 2026 PCAC vote (STAT)

Access & cost

Legally unsettled. Access, where it exists at all, runs through a physician — not a consumer purchase. See the status note above.

No lawful channel to price. As of 2026-08-28 there is no lawful US route to obtain BPC-157 for human use, so there is no legitimate telehealth, compounding-pharmacy, or brand price to report (Holt Law analysis). A gray market for the compound nonetheless exists and has been documented by mainstream reporting, with product reaching buyers through unregulated websites and overseas laboratories at prices set entirely outside any regulated channel (STAT, 2026-02-03).

Questions

Does BPC-157 actually work?
For tendon, muscle and ligament injury the honest answer is that it works reliably in rats and has never been tested against placebo in a person. Thirty-five of the 36 studies in the 2025 systematic review were preclinical (DOI 10.1177/15563316251355551). The three published human studies are uncontrolled pilots totaling 30 people, all from one clinic and all sharing a first author (PMIDs 34324435, 39325560, 40131143). No randomized controlled trial of BPC-157 is known to exist. A registry record asserting one was withdrawn from this dossier on 2026-08-29 as unverifiable, and nothing else on the published record points to a controlled answer arriving.
Does BPC-157 cause cancer?
Nobody knows, and both confident answers are wrong. There is no in vivo tumor study. The mechanism runs through VEGFR2 and other pro-growth pathways, which is a legitimate theoretical concern (DOI 10.1007/s12178-025-09990-7), and the counter-evidence is one unreplicated 2004 melanoma cell-line experiment (DOI 10.3390/ph18101451). See.
Oral vs. injectable — which one works?
Neither has been compared to the other in a human trial, and this is the single most misrepresented question about the compound. The claim underpinning oral capsules is Sikirić's assertion that BPC-157 is "stable in human gastric juice" (DOI 10.2174/092986712803414015) — a stability claim, not a bioavailability claim, and gastric stability does not establish that an intact peptide crosses the intestinal wall into circulation. Every human study to date used direct local injection or IV infusion; the 2025 systematic review notes only that cash-pay practices offer both injectable and oral formulations, without evidence comparing them (DOI 10.1177/15563316251355551). Treat oral-vs-injectable claims in either direction as marketing.
Is BPC-157 detectable on a drug test?
Yes. A validated weak-cation-exchange solid-phase extraction method detects BPC-157 in urine at a limit of 0.1 ng/mL, developed after the peptide was found in confiscated vials (PMID 28035768); metabolite profiling by UHPLC-HRMS has since extended detection targets (PMID 37959764). A detection window of at least 72 hours is reported, but only by the BSCG blog (BSCG) — a commercial testing company's own publication, outside T1/T2/T3, weaker than independent journalism, and not sufficient on its own for a claim an athlete or service member would act on. Note the contrast within this answer: the detection limit is carried by two peer-reviewed method papers (T1), while the window — the number that would actually govern a decision — is not. See.
What are the side effects?
Not established. The published human record — 30 people across three uncontrolled studies — reports no adverse events, which is far too small to characterize a safety profile (DOI 10.1177/15563316251355551). Journalists have documented user reports on Reddit of intense itching, severe anxiety, and anhedonia (STAT, 2026-02-03); the 2025 systematic review similarly notes anecdotal online reports of injection-site reactions, anxiety and mood disturbance that "remain undocumented clinically." These are anecdotal, low confidence.
How much of the research is independent?
Very little. Józwiak et al. calculate that over 80% of indexed BPC 157 records trace to Sikirić's group (DOI 10.3390/ph18101451). All three published human studies share a first author, Edwin Lee, who also founded the Clinical Peptide Society and the advocacy nonprofit Save Peptides (STAT, 2026-02-03). Neither fact makes the findings wrong; both are reasons independent replication matters more here than usual.
Why isn't it approved if it's been studied since the 1990s?
It reached Phase II for ulcerative colitis under PLIVA as PL 14736 and was dropped by GSK in 2006; complete results were never published. A 2015 Phase I safety trial was registered and abandoned (NCT02637284 — T2 registry record: existence and status only; self-asserted, unverified). See.
Is what I'd buy actually BPC-157?
Often not, or not only. ECRI and ISMP reported gray-market peptide purity ranging from 5% to 75%, with arsenic and lead above toxicity thresholds (ECRI/ISMP, 2026). Clinicians interviewed on the record describe the market as "the wild, Wild West… you especially don't know about the purity of the products" (MDLinx). ---

Last reviewed 27 August 2026. Regulatory status verified 27 August 2026.