SV-40: a real contamination, a cancer that never came, and a modern claim that borrows the name
This is the hardest kind of health claim to answer, because it starts out true. Polio vaccine really was contaminated with a monkey virus. The contaminated lots really were left on the market rather than recalled, and a US government witness said so under oath. What did not happen is the cancer epidemic. And the “SV40” currently circulating in posts about COVID-19 vaccines is not that virus at all — it is a short piece of regulatory DNA that happens to carry the same name.
What this page claims, and what it does not
The contamination is not in dispute and is not denied here. It happened, it was extensive, and the official response to it included a decision not to recall. Anyone who tells you this page calls the contamination a myth has not read it.
What this page examines is the second claim — that the contamination caused cancer in the people who received the vaccine — and a third, separate claim about COVID-19 vaccines that uses the same three characters.
This page is written from government documents, court records, and the peer-reviewed literature. It reports the evidence that cuts against the mainstream position as well as the evidence that supports it, including the loose ends. Where we could not verify something, we say so, including about our own sources.
Nothing here is medical advice. No reader should change anything about their own or their child's care on the strength of a web page, including this one.
The short version
- True: polio vaccine made between 1955 and 1963 was contaminated with simian virus 40. Between 10 and 30 per cent of US doses carried it.
- True, and worse: when this was discovered, lots already released were not recalled. They stayed in use until early 1963.
- Not established: that any of this caused human cancer. Four decades of cancer registry data, in populations totalling several million people across four countries, show no excess of the cancers in question.
- Honestly unresolved: the studies assigned exposure by year of birth rather than by individual vaccination record, which the Institute of Medicine said made even the reassuring results too weak to be conclusive. A small effect cannot be excluded. A large one has been.
- A different thing entirely: the “SV40” in current posts about COVID-19 vaccines is a promoter-enhancer sequence — a fragment of regulatory DNA used in laboratory plasmids. It is not the virus, cannot become the virus, and the scientist who reported finding it says so himself.
The sequence of events
The two claims, side by side
Almost every argument about SV-40 collapses into confusion between these two columns. They share three characters and nothing else.
| The 1955–63 contamination | The modern COVID-19 claim | |
|---|---|---|
| What it is | A live virus — whole, infectious simian virus 40 | A DNA sequence — a short promoter-enhancer used to switch genes on |
| Where it came from | Rhesus monkey kidney cells used to grow poliovirus | A standard laboratory plasmid, in use since the 1980s |
| Which products | Polio vaccine, mostly the injected form; also an army adenovirus vaccine | Pfizer–BioNTech COVID-19 vaccine. Not found in Moderna's |
| How much | 10–30 per cent of US doses; an estimated 10–30 million people exposed | Nanogram traces of residual DNA; below regulatory limits by qPCR |
| Can it infect you? | Yes. It was live virus, and it established infection in recipients | No. A regulatory sequence cannot infect anything |
| Can it make SV40 particles? | It was SV40 | No. The gene for large T antigen — the part implicated in cancer — is absent |
| Was it disclosed? | No. Contaminated lots were not recalled and stayed in use to early 1963 | Partly. Regulators knew residual plasmid DNA was present; Pfizer “did not specifically label” this sequence |
| Evidence of cancer | Forty years of registry data in four countries found no excess of the cancers alleged | No demonstrated integration into human DNA — conceded by the researchers making the claim |
| Official position | IARC: Group 3, not classifiable as to carcinogenicity in humans | Health Canada, FDA, EMA, MHRA and TGA: not a safety concern |
| What the finder says | Eddy's hamster tumours were real, and she was kept from publishing for two years | McKernan says he found the sequence, not the virus |
| Has a court found anything? | No. The one trial of the science was lost on causation in 2008 | No. The allegations sit in an unheard amicus brief; certiorari denied 2025 |
If you remember one line
The first column describes a real virus that really got into a real vaccine, and a real failure to recall it. The second describes a piece of regulatory DNA that shares its name. Neither column contains a demonstrated human cancer.
Part one: what actually happened
Polio vaccine, both the injected Salk vaccine and the oral Sabin vaccine, was grown in kidney cells taken from rhesus macaques. A great many of those monkeys were carrying a polyomavirus that nobody had yet identified. The CDC's own account is that between 1955 and 1963 “an estimated 10-30% of polio vaccines administered in the US were contaminated.”
Most of the contamination was in the injected vaccine, and the reason is a detail worth keeping: SV40 is relatively resistant to formaldehyde, the agent used to kill the poliovirus. Some live SV40 survived a process that had already killed the thing it was meant to kill. Oral vaccine delivered the virus to the gut, where it largely failed to establish itself — recipients of the injected vaccine developed antibodies to SV40, recipients of the oral vaccine mostly did not.
It was found in 1960, twice over. Ben Sweet and Maurice Hilleman at Merck isolated and named it, describing it in their own paper as a common and essentially ubiquitous contaminant of rhesus monkey kidney cultures. Independently, Bernice Eddy at the National Institutes of Health found that extracts of monkey kidney cells produced tumours when injected into newborn hamsters. She announced her findings at a conference in 1960 and was not permitted to publish them for two years.
The part that is hardest to defend, and is true
When the contamination was discovered, the contaminated vaccine already in distribution was not recalled. New lots had to be certified SV40-free; old lots stayed on the shelf and were used until early 1963.
This is not an allegation. It is testimony. At a Congressional hearing on 13 November 2003, the FDA's Dr Philip Egan confirmed that the agency then responsible did not recall the vaccine, and that an advisory committee had concluded it was better to leave it on the market than to interrupt polio vaccination during an epidemic era. Egan's own words were that many of the lots from 1955 through the early 1960s “was certainly contaminated.”
Whatever one concludes about cancer, that decision was made, it was made deliberately, and it was made without telling the public. A reader who has been told this and finds it confirmed is entitled to ask what else is true. The answer to that question is the rest of this page — but the question is a fair one and we are not going to pretend otherwise.
One allegation that goes further is not established: that a manufacturer knowingly shipped contaminated lots after 1961, or reprocessed and reused vaccine. That was put to witnesses by a congressman at the same hearing. It was not confirmed, and the FDA witness said reprocessing would not have been permitted. It remains an assertion, not a finding.
A number that is almost always used wrongly
You will see the claim that 98 million Americans were given a cancer-causing virus. Two different figures have been welded together.
About 98 million Americans received the injected polio vaccine between 1955 and 1963. Of the doses, 10 to 30 per cent were contaminated. The number of people actually exposed to live SV40 is estimated by the National Cancer Institute at 10 to 30 million.
Ten to thirty million people exposed to a monkey virus without their knowledge is a serious fact that needs no inflation. Stating it as 98 million is a denominator error of exactly the kind documented on our page about the 82 per cent miscarriage claim — and the habit of inflating a real number is one of the better reasons to distrust the source doing it.
Did it cause cancer?
This is the question, and it has been studied for forty years in several countries with national cancer registries. The short answer is that nobody has found the cancers.
Denmark produced the strongest single study, published in 2003: roughly 1.8 million people, about forty years of complete registry follow-up. Overall cancer incidence in the exposed was lower than in the unexposed — relative risks of 0.86 and 0.79. No increase appeared for mesothelioma, bone tumours, brain tumours or non-Hodgkin lymphoma. Denmark matters particularly because its own national serum institute confirmed contamination of its own vaccine from first-hand retained samples.
The United States, in a 1998 registry analysis, found no excess of ependymoma or osteosarcoma. Sweden followed about 700,000 people to the early 1990s, Germany 886,000 over twenty-two years. All null.
The most informative cohort is the smallest. 1,073 people who had been given contaminated vaccine as newborns in 1961–63 — the highest-exposure group that exists — were followed for thirty-five years. There were no deaths from any tumour type linked to SV40.
And the US Army adenovirus cohort, where exposure was much better defined than for polio vaccine, produced odds ratios of 0.81 for brain tumours and 0.97 for non-Hodgkin lymphoma. Its authors concluded that the findings “do not support a role for SV40 in the development of these cancers.”
Mesothelioma deserves its own sentence, because it is the cancer most often named. In US data, mesothelioma rates are not higher in the age groups exposed to contaminated vaccine. They are highest in men over 75 — the group least likely to have received the vaccine and most likely to have handled asbestos. The age curve tracks asbestos, not vaccination.
The three loose ends, which we are not going to hide
A page that reported only the null results would be doing what it accuses others of doing. Three findings point the other way, and all three have explanations — but you should know they exist.
Mesothelioma in the 1998 US analysis. The point estimates were elevated — relative risks of 3.00 and 2.45 — though not statistically significant. Reviewers attributed this to the longer follow-up available in the older birth groups.
Two testicular cancer deaths in the 1,073 newborn-exposed cohort, against 0.05 expected. The authors flagged it themselves, attributed it to multiple comparisons, and recommended further study. It is the only positive signal in the highest-exposure group ever assembled.
A 2004 study of children whose mothers were vaccinated during pregnancy found neural and blood cancers about 2.5 times more common. But the tumour types did not match those associated with SV40, the single ependymoma occurred in the child of an unvaccinated mother, and few of the mothers had SV40 antibodies at all.
None of these establishes anything. All three are the sort of thing that turns up when you run many comparisons across large datasets. We record them because the alternative is for a reader to find them elsewhere and conclude, correctly, that we left them out.
What the Institute of Medicine actually said — and why it cuts both ways
In October 2002 the Institute of Medicine's Immunization Safety Review Committee reported on exactly this question. Its conclusion is quoted constantly, usually by only one side at a time:
The finding
“The evidence is inadequate to accept or reject a causal relationship between SV40-containing polio vaccines and cancer.”
Advocates quote this as an admission that the danger is unresolved. Defenders quote it as a finding of no evidence of harm. Both are reading half of it. “Inadequate to accept or reject” means the committee did not think the question had been answered in either direction — and its reasoning is the interesting part.
The committee's criticism of the reassuring studies was that they assigned exposure by year of birth rather than by any individual's actual vaccination record — and since only 10 to 30 per cent of doses were contaminated, most people counted as “exposed” were not. That kind of error systematically pushes results toward finding nothing. The committee was unusual in saying so, and it recommended against further studies of that design until the measurement problem was solved.
The honest statement of where this leaves things: a large effect has been excluded; a small one has not. That is less satisfying than either side's version, and it is what the evidence supports.
On biological plausibility the committee graded the evidence in three steps — strong that SV40 can transform cells, moderate that it could lead to cancer in humans under natural conditions, moderate that vaccine exposure led to human infection. We have not been able to open the report's own text to quote those gradings directly, and we flag that below rather than present second-hand wording as a quotation.
The detective story that settled the laboratory question
Through the 1990s and 2000s, laboratories using PCR reported finding SV40 DNA in mesotheliomas and other tumours. If the virus were sitting in the tumours, that would be powerful. Three things undid it.
SV40 sequences turned up in people who could not have been exposed — those born after 1963, and children. They also turned up in normal tissue from healthy people. Whatever was being detected was not tracking vaccination.
The laboratories could not agree. Detection rates for the same tumour types ranged from about 3 per cent to about 90 per cent depending on who was doing the testing. In 2001 the National Cancer Institute organised a blinded study: nine laboratories, masked mesothelioma samples, each lab using its own method. None of the specimens was consistently positive.
And then somebody found the source of the signal. In 2004 a Memorial Sloan Kettering team examined 71 mesotheliomas with several different sets of PCR primers. The primers that produced positives — in 56 to 62 per cent of cases, matching the earlier positive studies — all targeted one particular region of SV40. That region is present in common laboratory plasmids. Primers aimed outside it produced almost nothing. And the sequences recovered carried a deletion found in the plasmids but not in the real virus.
The decisive finding was quieter: all 71 tumours were negative for T-antigen transcripts. If functional viral DNA were present and doing anything, the corresponding RNA should have been there. It was not. The senior author's summary was that the assays commonly used to detect SV40 in human tumours are uniquely prone to false positives.
The minority view, stated fairly
A real scientific minority position exists, held by established investigators publishing in peer-reviewed journals — principally Michele Carbone, Adi Gazdar and Janet Butel, whose current statement of the case appeared in 2020.
Their argument is not foolish. Laser microdissection work reported SV40 DNA localised to mesothelioma cells rather than surrounding tissue, which a plasmid artefact should not do. They propose that SV40 persists in a form that expresses its transforming protein without destroying the cell, and that it may act as a co-carcinogen alongside asbestos rather than as an independent cause — which would be much harder to detect epidemiologically.
Their own concession is the fairest summary available: SV40 “may contribute to the development of mesotheliomas in humans” but “epidemiological evidence to support this hypothesis is lacking.” They also acknowledge that no cancer epidemic followed the contaminated vaccines.
Two things a reader should weigh. Much of the supporting work comes from these authors' own laboratories. And Adi Gazdar was the plaintiff's sole causation expert in the one case that actually went to trial — where the court found he had “failed to conduct accurate, complete and scientifically reliable testing,” could not exclude contamination in his own samples, and conceded that the epidemiological evidence did not support causation.
Canada: what we know, and what nobody has written down
Connaught Laboratories in Toronto made polio vaccine throughout the relevant period. It supplied all 3,000 litres of poliovirus fluid for the 1954 field trial, opened a dedicated polio building in 1956, received a Sabin seed pool in 1959, shipped three million doses of oral vaccine to Japan in 1961, and saw Canadian oral vaccine licensed in March 1962. It used the same rhesus monkey kidney substrate as everyone else.
An honest gap
We could not find a single source — government, archival or scholarly — stating whether Connaught's own lots were tested for SV40, what proportion were contaminated, or when its substrate was cleared. The University of Toronto's Connaught history archive does not mention SV40. Neither does the Canadian Encyclopedia's account of Canadian polio vaccine.
Contamination of Canadian-produced vaccine is highly likely on first principles, because the cell substrate was the same one that was contaminated everywhere else. But likely is not documented, and this page is not going to assert it. If a reader can point us to the Connaught or Sanofi Pasteur Canada records on this, we will publish what they show.
The only Canadian federal statement we could locate is an archived Health Canada page from roughly 2005–2006, which says Canadian polio vaccines have been SV40-free since the early 1960s and that studies “have not found an increase in cancer rates.” The current Canadian Immunization Guide does not mention SV40 at all.
The lawsuits, and what they actually decided
A table of SV-40 lawsuits circulates online. It is worth going through, because the outcomes are routinely misdescribed in both directions.
| Case | What it was | What was decided, and why |
|---|---|---|
| Gannon v. United States (3d Cir., 8 Sept 2008) | A Federal Tort Claims Act suit — not a Vaccine Injury Compensation Program case, as is often stated. Alleged oral polio vaccine received 1973–76 caused a medulloblastoma. | Judgment for the government, affirmed. This was a loss on the science, not a technicality. The trial court expressly denied the government's motion to exclude the plaintiff's expert, heard his evidence, and found the causation case unproven. |
| Garcia v. Pfizer (5th Cir., 12 Feb 2008) | Alleged oral vaccine received in 1970 caused a meningioma. | Lost on product identification — she could not show which manufacturer supplied her dose. The court never reached the science. |
| Rivard v. American Home Products (N.J. App. Div., 8 Mar 2007) | A child who received oral vaccine, diagnosed with medulloblastoma in 1998, died 1999. | Held that an SV40 claim is a vaccine-related injury and must go to the federal Vaccine Court first, because SV40 was not a contaminant “intentionally added.” |
| Gannon v. American Home Products (N.J. Supreme Court, 15 Aug 2012) | State product-liability claims by the same plaintiff. | Ended on collateral estoppel — the causation question had already been fully litigated and lost in federal court. |
Two corrections to the table that circulates. Gannon v. United States is described in it as a Vaccine Injury Compensation Program case; it was not. And it lists a Moreno decision we could not locate at all under the docket given.
The net record: no plaintiff has ever recovered damages on an SV-40-caused-my-cancer theory in a US court. And the science was tried properly exactly once, in Pennsylvania in 2007, where it lost. SV-40 has never been on the Vaccine Injury Table, and neither has cancer.
Part two: the modern claim is a different thing
Everything above concerns a virus. What is circulating now concerns a sequence, and almost all of the confusion comes from the two sharing a name.
The companion page
What follows is the short form. The modern claim has its own page on this site — “Full of SV40”: how a harmless DNA fragment became a cancer scare — which goes further on the 2025 Ontario paper now under investigation by its publisher, on how Florida’s Surgeon General moved the claim into an official document, and on the pattern common to both: each measured the vaccines against a regulatory standard written for a different kind of product. If you have just been shown a post, start there.
A promoter is a stretch of DNA that sits in front of a gene and tells the cell where to start reading it. An enhancer boosts it. Molecular biologists have used a short piece of SV40 as an off-the-shelf promoter since the 1980s, because it is compact, well characterised and reliable. It is a standard laboratory part, present in enormous numbers of plasmids that have nothing to do with vaccines.
In April 2023 Kevin McKernan and colleagues reported residual plasmid DNA in COVID-19 vaccines, including an SV40 promoter-enhancer sequence in the Pfizer product but not in Moderna's. The finding travelled, and by the time it reached social media it had become “the shots contain cancer-causing monkey virus.”
The most useful fact on this page
The researcher who found the sequence says the virus is not there. McKernan has stated that his team found the promoter, not the whole SV40 virus. His own 2025 paper concedes that genomic integration of these DNA fragments “has yet to be demonstrated.”
This matters more than any regulator's denial, because it requires nobody to trust a regulator. The version circulating online is a distortion of the original finding that the original researcher does not endorse.
The reason it cannot become the virus is specific. SV40's cancer-associated activity depends on a protein called large T antigen, and the gene for it is not present. Pfizer told the European regulator in a formal response that the large T antigen is not in the plasmid, and that without it there can be no replication. A short regulatory sequence cannot assemble itself into a virus any more than a light switch can assemble itself into a house.
Regulators converged on near-identical wording. Health Canada, in October 2023: “The presence of the SV40 promoter enhancer sequence is not the same as the presence of the whole virus itself.” The UK's MHRA used almost exactly that sentence in March 2024. The FDA's Peter Marks, writing to Florida's Surgeon General on 14 December 2023: “No SV40 proteins are encoded for or are present in the vaccines,” and integration is “quite implausible.”
Where the critics have a point, and it should be said
Three things in this story are not comfortable for the regulators, and a page that skipped them would not be worth reading.
Pfizer did not flag the sequence. Australia's Therapeutic Goods Administration, answering a Senate question, said the company “did not specifically label the SV40 promoter sequence in the plasmid.” Regulators knew residual plasmid DNA would be present; that is different from knowing which sequences the plasmid contained.
The 10 nanogram limit is softer than it sounds. The World Health Organization document that established it says in terms that the figure “does not take into consideration important factors such as the size of the DNA fragments,” and that product-specific limits should be agreed case by case. Both sides cite this limit as though it were a bright line. It was never designed to be one.
There is a genuine measurement dispute. The headline claim of “hundreds of times over the limit” comes from fluorometry, which is confounded here because RNA interferes with the dyes used to measure DNA. The same papers' own qPCR results come in below the limit. That internal split is the heart of the technical argument, and it is a real argument between people with relevant expertise — not one side inventing things.
Where it currently stands: a Slovak Academy of Sciences team tested fifteen batches by four independent methods in December 2025 and found residual DNA below limits in all of them. The 2025 paper making the opposite case has been under investigation by its publisher since September 2025 — one of its own peer reviewers has said publicly that he twice recommended against publication. It has not been retracted, and anyone describing it as retracted is wrong, just as anyone citing it as settled is wrong.
The Supreme Court case that is not what it is said to be
A table circulating online lists a 2025 US Supreme Court case on SV40 in COVID-19 vaccines, attributed to plaintiffs named Beaudoin and Wilkins. We checked the docket.
The case is real but the description is wrong in every material respect. It is Wilkins v. Herron, No. 25-9, which reached the Supreme Court from Oregon by way of the Ninth Circuit. Beaudoin is not a plaintiff. He is one of four signatories to a friend-of-the-court brief filed on 1 August 2025 — and another signatory is Kevin McKernan, the originator of the scientific claim. The SV40 allegations appear only in that brief.
The Supreme Court declined to hear the case on 6 October 2025. No court examined the SV40 allegations, and a refusal to hear a case decides nothing and endorses nothing.
How to tell the two claims apart
When someone shows you an SV-40 post, three questions settle which claim is being made.
Is it about polio vaccine before 1963, or a COVID-19 vaccine? If the former, the contamination is real and the cancer question is covered above. If the latter, it concerns a DNA sequence, not a virus.
Does it say “virus” or “promoter”? If a post about COVID-19 vaccines says the virus is present, it is wrong on a point the original researcher agrees about.
Is a court said to have found something? No court has ever found that SV-40 caused anyone's cancer. The one trial of the question went the other way, and the Supreme Court case being cited was never heard.
Why this belongs on a Canadian site
Because the true part makes the false part durable. A claim that opens with a verifiable government failure — a contaminated vaccine that was not recalled, and a scientist who was not allowed to publish — earns a hearing that a straightforwardly invented claim would not. Readers who check the first part find it stands up, and reasonably extend their trust to the rest.
The answer is not to minimise the first part. It is to be more accurate about it than the people using it — which includes saying that ten to thirty million exposed is the real figure rather than ninety-eight million, that the lots really were left on the market, and that the epidemiology, while reassuring, has a weakness the Institute of Medicine itself identified.
And Canada has a specific gap: nobody appears to have written down what happened with Connaught. That is a question for an archive, and we would like the answer.
A note on how this page is written, and what we could not check
Sources are government documents, court records, national cancer registries and the peer-reviewed literature. Advocacy publications were excluded in both directions.
Three limitations should be on the record rather than buried.
We could not open the Institute of Medicine's 2002 report itself. Its causality conclusion is quoted here because four independent sources — sworn FDA testimony, a CDC conference abstract, a federal appellate opinion and the National Cancer Institute — quote it identically. The committee's graded assessments of biological plausibility are described but not quoted, because we have them only at second hand.
IARC's classification of SV40 as Group 3 is sourced to a peer-reviewed citation of the monograph rather than to the monograph text, which we could not retrieve.
The Connaught question is unanswered, as set out above.
Corrections are welcome and will be posted.
Sources. On the contamination and the response: CDC, “Historical Vaccine Concerns,” last reviewed 31 July 2024; US House of Representatives, Committee on Government Reform, Subcommittee on Human Rights and Wellness, Preventing Another SV40 Tragedy: Are Today's Vaccine Safety Protocols Effective?, 13 November 2003, Serial No. 108-127, for the FDA's testimony on the non-recall; Sweet BH and Hilleman MR, Proceedings of the Society for Experimental Biology and Medicine 1960;105:420–427; Butel JS, Bulletin of the World Health Organization 2000;78(2):195–198; Children's Hospital of Philadelphia Vaccine Education Center, “Vaccine Ingredients: SV40,” reviewed 15 September 2026; Statens Serum Institut, EPI-NEWS No. 8, 21 February 2001. On the epidemiology: National Cancer Institute briefing on SV40 exposure and cancer development, 2003; Strickler HD et al., JAMA 1998;279(4):292–295; Engels EA et al., Journal of the National Cancer Institute 2003;95(7); Carroll-Pankhurst C et al., British Journal of Cancer 2001; Rollison DEM et al., American Journal of Epidemiology 2004;160(4):317–324; Dang-Tan T et al., Oncogene 2004;23:6535–6540. On the laboratory question: Strickler HD et al., Cancer Epidemiology, Biomarkers & Prevention 2001;10(5):523–532; López-Ríos F, Illei PB, Rusch V and Ladanyi M, The Lancet 2004;364(9440):1157–1166. On the IOM: Stratton K, Almario DA and McCormick MC, eds., Immunization Safety Review: SV40 Contamination of Polio Vaccine and Cancer, Institute of Medicine, released 22 October 2002, with the National Academies' release of the same date. On the minority view: Carbone M, Gazdar A and Butel JS, Translational Lung Cancer Research 2020;9(Suppl 1):S47–S59. On IARC: Monographs Volume 104, working group February 2012. On Canada: University of Toronto Connaught history archive; The Canadian Encyclopedia; archived Health Canada immunization page, c. 2005–2006, via Library and Archives Canada. On the litigation: Gannon v. United States, No. 07-3428 (3d Cir. 2008); Garcia v. Pfizer, No. 06-40703 (5th Cir. 2008); Rivard v. American Home Products, No. A-2478-05T3 (N.J. App. Div. 2007); Gannon v. American Cyanamid (N.J. 2012); the Vaccine Injury Table at 42 C.F.R. § 100.3. On the modern claim: Health Canada statements of October 2023 as reported by The Canadian Press; MHRA FOI 24/212, 28 March 2024; FDA, Peter Marks to Joseph Ladapo, 14 December 2023; Therapeutic Goods Administration Senate answers and media release of 18 October 2024; WHO Technical Report Series No. 978, Annex 3 (2013); European Medicines Agency, Comirnaty response to request for supplementary information, 2024; Achs A et al., npj Vaccines 2025;10:259; Speicher DJ, Rose J and McKernan K, Autoimmunity 2025;58(1):2551517, together with its publisher investigation as reported by Retraction Watch, 18 September and 21 November 2025; US Supreme Court docket 25-9, Wilkins v. Herron, certiorari denied 6 October 2025; and fact-checks by Full Fact (7 November 2025), AAP FactCheck (18 July 2025) and FactCheck.org (26 October 2023).
QuackeryWatch is a non-commercial consumer information site operated by Terry Polevoy, MD, FRCPC, a retired paediatrician.