AsapSCIENCE Took On BPC-157. We Checked Every Claim.

A 10.8-million-subscriber science channel called BPC-157 'the deadly drug everyone is taking.' We pulled every source. Most of it checks out — and the parts that don't are more interesting than the parts that do.

On June 30, 2026, AsapSCIENCE — a Canadian science channel with about 10.8 million subscribers, no clinic, no storefront, and nothing to sell in this category — published a nine-and-a-half-minute video titled “The deadly drug everyone is taking.” Its subject was BPC-157, the most talked-about healing peptide in the world. Within a day the video was syndicated by RealClearScience. Millions of viewers heard four numbers and phrases they had no way to check: one lab, lead, cancer, $328 million.

So we checked. We pulled the video’s transcript, traced every load-bearing claim to its source, and graded each one against the primary literature. Here is the honest summary before the detail: this is the most accurate mainstream take on BPC-157 we have seen — and it still gets several important things wrong, one of them backwards. Our corrections are not a defense of BPC-157. The same article that corrects the video carries the video’s strongest charges at full strength, because they happen to be true. Watch the video with our scorecard.

Where they’re right — and it’s the important part

The evidence charge is correct, and it matches the peer-reviewed reviews. The video says most BPC-157 research comes from one lab, that the research is in animals, and that no proper human trials exist. All three are what the literature itself says:

• Józwiak and colleagues, reviewing the field in Pharmaceuticals in 2025, searched PubMed and retrieved more than 190 BPC-157 articles — over 80% of them listing Predrag Sikirić or Sven Seiwerth, of the University of Zagreb group, as first or senior author. An independent 2026 count found 227 indexed papers, 175 of them (77%) sharing Sikirić as an author. The Zagreb group has answered its critics in the same journal; the argument is a live, in-print scientific dispute, not a YouTube invention.
• Vasireddi and colleagues’ 2025 systematic review in the HSS Journal screened 544 articles and found 36 that met inclusion — 35 preclinical and exactly one clinical study: a retrospective, uncontrolled series of 12 knee-pain patients, 7 of whom reported relief beyond six months. No control group, no effect size. The review’s conclusion: no clinical safety data were found.
• The entire published human record is three small uncontrolled reports totaling roughly 30 people — that knee series, a two-person IV safety pilot, and a 12-person bladder-pain series, all from private clinical practice (McGuire et al., 2025). There is exactly one registered human trial, NCT02637284, begun in 2015 — and, as we detail below, abandoned without results.

“Not properly tested in humans” is not a critic’s slogan. It is the state of the record.

The contamination problem is real and documented at scale — bigger, in aggregate, than the video’s single anecdote (which we could not verify; see below). A Belgian public-health laboratory analysis published in Talanta in 2018 (Janvier, Vanhee et al.) tested the ten most frequently encountered falsified peptide drugs seized from illegal internet pharmacies: purity as low as 5–75% for some compounds, 30% of samples with incorrect amino-acid sequences, two-thirds below 95% purity, inorganic arsenic at up to ten times the limit for injectable drugs, and lead detected above the toxicity threshold in at least one sample. Finnrick Analytics, an independent testing platform, reports that across more than 7,000 gray-market samples, roughly one in three fail basic expectations — mislabeled, under- or over-filled, or contaminated (statement by its CEO, February 2026; carried by New York Magazine and The Guardian). A 2026 re-analysis of 6,441 consumer-market samples (Mendias & Awan) found failure rates of 41.6% to 71.1%, depending on the standard applied.

Its two social-science citations are accurate to the digit. The video’s six false-belief percentages — 25, 25, 28, 31, 32, 32 — match the 2026 Edelman Trust Barometer Special Report: Trust and Health (16,009 respondents, 16 countries) exactly. Its “thousand accounts” study is Nickel, Moynihan, Gram et al., JAMA Network Open 2025: 982 posts promoting five medical tests to 194 million combined followers, 87.1% mentioning benefits against 14.7% mentioning harms, and 68% of account holders holding financial interests in what they promoted. Faithful citation like this is rare in the genre, and worth saying so.

And its best observation is genuinely original. When more than 80% of a literature is one lab, an AI chatbot summarizing “the research” is summarizing that lab. AI answers inherit the concentration problem. We have built a whole page around the idea, because it generalizes far beyond BPC-157: The One-Lab Problem.

The one-lab problem — including the part the video got a decade out of date

The video says Sikirić “discovered” BPC-157 in 1993 and “currently holds the patent.” The 1993 date is the landmark paper (Sikirić et al., Journal de Physiologie, Paris) — but the Zagreb group had spent the 1980s isolating the parent compound from gastric juice, and the patent’s priority date is September 12, 1989. And the patent claim is wrong as stated: U.S. Patent 5,288,708, the molecule patent, expired around 2011. Croatian pharma company Pliva had licensed the compound in 1993 and developed it as PL-14736 for ulcerative colitis; GSK acquired Pliva’s research arm in 2006 and dropped the program; the rights reverted around 2009.

That expiry is not a footnote — it is the most plausible explanation for the missing human trials. A 2026 review in Pharmaceutics calls the absence of exclusivity “a significant disincentive for industry-sponsored Phase II and III development.” Nobody funds a pivotal trial for a molecule anyone can sell.

But the video’s instinct — that the discoverer has a financial stake — is right, and the documented version is stronger than the one it told. Reporting by STAT News and Undark (February 2026) found that Sikirić part-owns PharmaCotherapia, the company that sponsored the only registered human trial (NCT02637284), and is CEO of Diagen, which holds a live patent on a stable BPC-157 salt formulation that it advertises for sale. The reporters found none of these interests disclosed in the papers they reviewed. And that one registered trial? It began in 2015 with about 42 healthy volunteers; its registry record was never updated after December 2015, and no results were ever posted. Per STAT and Undark, data was submitted and then withdrawn before outside review.

The video undersold its own best material. The truth is not “he holds the patent.” The truth is: the patent expired, the trial vanished, and the stakes were never disclosed.

Where they overstate — the graded errors

1. The cancer segment is built on the wrong family. The video introduces trefoil factor peptides — a family defined by a distinctive three-disulfide-bond “trefoil” fold — presents evidence that they contribute to cancer-cell movement and metastasis, and then pivots to BPC-157’s history as though it were one of them. BPC-157 is not a trefoil peptide. It is a linear 15-amino-acid fragment of a different gastric protein, with no trefoil domain. The metastasis evidence belongs chiefly to TFF3 — and even there the story is two-directional: TFF3 is genuinely overexpressed in several cancers and promotes migration and invasion in studies, yet in the largest recent breast-cancer analysis (2026), high TFF3 expression was an independent protective factor for survival. Evidence about TFF3 does not transfer to BPC-157, and it does not point only one way even about TFF3. We built the honest version as its own page: Cancer and Peptides: What Is Actually Known.

2. The free-radicals claim appears inverted. The video says BPC-157 “has been found to increase free radicals.” The published literature — including the Zagreb group’s own reviews and independent organ-protection studies — frames BPC-157 as a free-radical scavenger that normalizes oxidative-stress markers. Those antioxidant claims are themselves single-lab animal claims and prove nothing about humans. But a fact-check has to note the direction of the arrow: as stated, the video’s sentence contradicts the literature it draws on.

3. The $328 million is not what the video says it is. “U.S. imports of unregulated injectable chemicals from China doubled in 2025” is the video’s most quotable statistic and its least accurate sentence. The source is The New York Times (January 3, 2026), citing customs data: imports of hormone and peptide compounds from China rose from $164 million in the first three quarters of 2024 to $328 million in the first three quarters of 2025. That is a broad customs category — it includes bulk ingredients feeding legal demand, such as GLP-1 active ingredients for compounding pharmacies, alongside gray-market supply. It is a dollar value for nine months, not a full year, not units, and not separable into legal and illegal trade. The doubling happened; what doubled is not what the video says doubled. The sharper gray-market measure is the money trail: Chainalysis traced $32 million in cryptocurrency to gray-market peptide vendor wallets in the first quarter of 2026 alone — a 159% rise on the prior quarter, against roughly $1 million per quarter through 2024. We keep both numbers, with their definitions attached, on one page.

Two smaller stretches, graded on the video page: the video’s lead anecdote — a named buyer, a “trusted source,” a BPC-157 product, a lead finding — could not be matched to any publication; the documented lead and arsenic findings are in other falsified peptides (the 2018 Belgian study’s product list does not include BPC-157), and BPC-157’s own documented failure modes in the Finnrick dataset are wrong quantity and endotoxin, not heavy metals. And the video’s bridge from eating-disorder trends to peptide use is speculation wearing a citation’s clothes: the trend numbers are real (Liu et al., 2025, Global Burden of Disease data), but that study measures anorexia and bulimia, not muscle dysmorphia, and attributes no share of the trend to peptides.

“Deadly,” calibrated

The video’s title promises death. Its description warns peptides “could cause cancer, and maybe result in early death.” So: what does the documented harm record actually contain? This is the table we wish every piece on this topic — seller hype and scare headline alike — had to publish next to its adjective.

• RAADfest, Las Vegas, July 13, 2025. Two women received injections at an anti-aging conference booth; both developed anaphylaxis-type reactions, were intubated, spent days in intensive care, and recovered. Investigators could never establish what was injected — the product was gone — and the booth physician was unlicensed in Nevada. Nevada’s pharmacy board fined the parties a total of $35,000. (ProPublica.) Lesson: unidentified product, unlicensed administration.
• FDA Class I recall, October 2025. GenoGenix NAD+ injections were recalled with endotoxin measured at 3,360 EU/mL; three patients were hospitalized. Lesson: sterility and endotoxin — the risks a purity certificate does not measure.
• Products sold as retatrutide. CBS News (October 13, 2026) reported that the FDA had received two reports of deaths following use of products sold as retatrutide — an investigational Eli Lilly drug that has never been approved or legally sold. The products were unverified gray-market items; a report to the FDA is not a causal finding. This is the closest the record comes to the video’s framing — and it concerns a stolen investigational drug, not BPC-157.
• BPC-157 itself. An independent search of the FDA’s adverse-event database found three reports for injectable BPC-157 through December 2025. KFF Health News (October 13, 2026) described entries including a hospitalized 36-year-old woman and a 48-year-old man hospitalized with a “life-threatening” condition, and published first-person accounts of months of fatigue, anxiety and twitching after a nasal BPC-157 product. Those accounts are patient-reported and causally unverified — anecdata, in the video’s own useful term, pointing the other direction. And adverse-event databases under-report in both directions, which has to be said aloud too.
• Enforcement texture. The FDA issued 58 warning letters in a single day (September 9, 2025) over unlawful online sales of unapproved drugs. The owner of Paradigm Peptides was sentenced to 70 months in July 2026 after admitting he forged the lab certificates his customers relied on. The “tested” paper trail itself is falsifiable.

Now the empty cell: no death has been attributed to BPC-157 in any source we could find, and there is no human cancer data in either direction — no attributed cases, and no carcinogenicity studies at all. The documented harms cluster in product quality, product identity, and unlicensed administration. The video’s implied harm — this molecule, via cancer, at population scale — is the one cell in the table with no entries in it yet. “Deadly” is the channel’s franchise format doing marketing. It is the mirror image of the sellers’ “miracle” — and it is the video’s least defensible element.

What changed after the video aired

The video was published June 30, 2026, and its closing advice — wait until doctors can prescribe it — was, on that date, the only defensible advice. The ground has since moved, and fairness requires the update:

• July 23, 2026: the FDA’s Pharmacy Compounding Advisory Committee voted 8–6 (one abstention) to recommend adding BPC-157 to the 503A bulks list — non-binding, and against FDA staff’s stated concerns.
• October 30, 2026: HHS published a draft, non-binding enforcement-discretion policy covering BPC-157 (acetate) and five other peptides for state-licensed 503A compounding pharmacies, under ten conditions, in specified forms, against patient-specific prescriptions. It is not FDA approval, not a bulks-list addition, does not cover research-chemical sales, and does not bind state pharmacy boards. Public comments close November 13, 2026. What the policy actually says

Precision about the law matters as much as precision about the science: selling these products for human use is unlawful, but the molecules are not controlled substances and possessing them is not a crime. “Unapproved and unlawfully marketed, in a gray zone the government is actively renegotiating” is the accurate sentence. Our standing summary lives on the Legal Mandate page.

The Huberman paragraph

The video notes that Andrew Huberman has injected himself with BPC-157. True — first-person, on the record, repeatedly: the L5 deadlift-injury account on the Joe Rogan Experience (#2195, 2024), his own April 2024 episode on peptide therapeutics, and appearances through 2026, including describing keeping it on hand for strains (The Tim Ferriss Show #876, July 2026).

What the video strips out is the context the man himself attaches. His stated route to it was a doctor’s prescription filled by a compounding pharmacy — the regulated channel, not a gray-market website. And he pairs the anecdote with the caveat: “there isn’t any clinical data for BPC-157. It’s all animals” — in 2026 appearances he openly wondered whether saline would have done the same. Anecdotes are data about belief, not about effect. That standard is the video’s own. It applies to heroes too.

The kicker

Here is the standard both sides of this fight keep failing. Sellers sell certainty the data does not contain. Scare headlines claim deaths the record does not contain. In between sits the actual record — one lab, no completed human trials, a third of the gray market failing basic quality checks, an expired patent, a vanished trial, undisclosed stakes, and a federal policy window open right now — which is bad enough, and interesting enough, without either side’s embroidery.

AsapSCIENCE got more of that record right than almost anyone with an audience that size. Where it got things wrong, it got them wrong in ways that matter — a wrong family, a backwards arrow, a statistic stripped of its definition, a title the evidence can’t carry. Checking those things, in public, with the receipts attached, is the entire job of this site.

Explained, not prescribed. If you are considering any of these compounds, the channel that exists for that conversation is a licensed clinician — the directory we are building is where that starts. If you want the law as it stands today, start with the Legal Mandate.

Read next

• The video page, with the full scorecard
• The One-Lab Problem
• Cancer and Peptides: What Is Actually Known
• The $328 Million Question: Peptide Imports from China
• BPC-157 — the compound page

Educational information only — not medical or legal advice. Fact-check prepared November 9, 2026, from the video’s transcript and the sources cited inline.