When the Retraction of a Spike Protein/Nattokinase Study Itself Becomes the Evidence
A small case study in how scientific gatekeeping is actually done
In December 2025, an Elsevier journal retracted a five-patient case study describing a treatment protocol for post-vaccination and post-COVID syndromes. The stated reasons appear nowhere in the publisher's own retraction policy — and one of them is contradicted by the FDA, the CDC, and Elsevier's own vaccine journal.
A retraction notice is supposed to be the boring part. The drama lives in the original paper — the bold claim, the contested method, the data that did or did not survive scrutiny. The retraction is just the journal closing the door. A line of bureaucratic prose. A footnote.
Sometimes the footnote is the story.
In June 2025, the journal IDCases — an Elsevier case-report journal — published a five-patient case series by a clinical team in São Paulo, Brazil. The paper described a treatment protocol for what the authors called Post-Spike Syndrome: a constellation of chronic symptoms following either SARS-CoV-2 infection or mRNA vaccination. The protocol had four components. One of them was nattokinase — a serine protease enzyme produced by Bacillus subtilis during the fermentation of soybeans into the traditional Japanese food natto.
In December 2025, IDCases retracted the paper.
I want to walk through the retraction notice itself, because what is says — and what it cannot say — tells you more about the current state of medical gatekeeping than any case series ever could.
What the retraction notice actually says
The notice gives three reasons for retraction. They are worth quoting precisely.
First: the paper’s “dissemination could promote the premature adoption of unvalidated treatment protocols.”
Second: “there is no empirical evidence supporting a link between the described pathophysiology and vaccines.”
Third: “the potential for harm, particularly through the promotion of unproven medications without sufficient evidence, outweighs the benefits of continued publication.”
That is the entire substantive content of the notice. No allegation of data fabrication. No allegation of plagiarism. No allegation of ethical violation, no allegation of irreproducible results, no allegation of methodological errors that invalidate the conclusions. The standard grounds on which scientific papers are retracted — the grounds the Committee on Publication Ethics actually recognizes — are absent.
What we have instead is an editorial body asserting that a peer-reviewed paper, once accepted by its own peer reviewers, must be removed from the scientific record because (a) other clinicians might read it and try the protocol, and (b) the underlying pathophysiology the authors hypothesized about does not exist.
The first reason, taken seriously, would gut the entire genre of case reports. IDCases publishes case reports. That is the journal’s name. Case reports are, by definition, descriptions of clinical observations that have not yet been validated in randomized controlled trials. If “potential for premature adoption” were a valid retraction criterion, the journal would have to retract most of its archive.
The second reason is the one that should make every working scientist stop and read it twice.
The grounds the retraction notice cites do not appear in Elsevier’s own published retraction policy. That policy enumerates fourteen specific bases for retracting an article — fabrication, falsification, plagiarism, undisclosed conflicts, compromised peer review, citation manipulation, and so on. Every one of them concerns the integrity of the work itself or the conduct of its authors. None of them concerns the downstream effect of publication on clinical behavior. The category that does address health risk in Elsevier’s policy is not retraction at all; it is “Article Removal,” reserved for cases where the work, if acted upon, might cause serious harm. The journal did not invoke removal. It invoked retraction, citing reasons the retraction policy does not contemplate.
There is also a provision in the same policy designed for the situation the editors describe — “Editor’s Notes,” which the policy explicitly authorizes when editors wish to alert readers to elements requiring caution without retracting the work. That instrument was not used either. Instead, the editors reached past the policy’s calibrated tools for editorial concern and used the strongest available action, on grounds the policy does not list.
“No empirical evidence”
The retraction notice asserts, as a settled scientific fact, that there is no empirical evidence linking the syndrome the authors described to mRNA vaccines.
This claim is not contestable. It is wrong.
In September 2022, the journal Vaccine — a respected, peer-reviewed publication owned by the same parent company as IDCases — published Fraiman et al., a reanalysis of the Pfizer and Moderna phase III clinical trials. Using the Brighton Collaboration’s WHO-endorsed list of adverse events of special interest, the authors found that Pfizer and Moderna mRNA COVID-19 vaccines were associated with an excess risk of serious adverse events of special interest of 10.1 and 15.1 per 10,000 vaccinated. The paper has not been retracted. It has not been refuted by a competing reanalysis of the same trial data, because the underlying participant-level data has not been released.
In April 2025, the U.S. Food and Drug Administration ordered Pfizer/BioNTech and Moderna to expand the warning labels on their mRNA COVID-19 vaccines. The expanded warning identified males aged 16 to 25 as the highest-risk group for vaccine-associated myocarditis and pericarditis, with an estimated incidence of approximately 27 cases per million doses in males 12 through 24 years of age. The FDA cited a longitudinal observational study showing persistence of abnormal cardiac MRI findings five months after vaccination.
The U.S. Centers for Disease Control and Prevention states, on its own website, that evidence from multiple vaccine safety monitoring systems in the United States and around the globe supports a causal association between mRNA COVID-19 vaccines and myocarditis and pericarditis.
Causal association. Those are the CDC’s words.
So when the editors of IDCases assert in a retraction notice that there is no empirical evidence linking the described pathophysiology to vaccines, they are asserting something that the FDA’s regulatory record, the CDC’s own clinical guidance, and the peer-reviewed literature in their parent company’s own vaccine journal all directly contradict.
This is not a difference of interpretation. It is a public-facing scientific journal making, in its formal retraction record, a categorical claim that does not survive five minutes of fact-checking against documents that are either co-published by its own corporate parent or hosted on the CDC’s own clinical-considerations page.
Why the case is sharper than it looks
I want to be precise about what I am and am not saying.
The retracted paper had real methodological problems. Five patients. A four-drug protocol. No control group. The cases selected from a much larger pool the authors had treated. I would not, in my own work, build a clinical claim on top of that paper. Nobody should.
But the retraction notice does not cite those methodological problems as the reason for retraction. It cites two reasons that fail on their own terms — one that would invalidate the journal’s entire genre, and one that asserts a contestable scientific claim as settled fact — on grounds that do not appear in the publisher’s own retraction policy.
And here is where the gatekeeping mechanism becomes visible.
Three years before the retracted case series was published, a Japanese pharmacology group at Josai University published a paper in the MDPI journal Molecules (Tanikawa et al., 2022). Using purified nattokinase and standard biochemical methods — Western blots, dose-response curves, time courses, heat-denaturation controls, serine-protease-inhibitor controls — they demonstrated that nattokinase enzymatically degrades the SARS-CoV-2 spike protein in a dose- and time-dependent manner, on both isolated cell lysates and the surface of transfected cells, at concentrations that produced no cytotoxicity.
That paper has not been retracted. It is solid biochemistry. The finding is real.
The mechanism is publishable. The clinical exploration of the mechanism is not.
This is what gatekeeping looks like in 2025. Not the suppression of the underlying biochemistry — that has become too well-established and too widely replicated to suppress. What gets suppressed is the bridge: the clinician who notices that patients with a particular post-exposure syndrome respond to a protocol that includes the enzyme that degrades the protein implicated in that syndrome, and writes it up as a hypothesis-generating case series in the appropriate genre journal.
The biochemistry is allowed to exist as a curiosity. The clinical translation is what cannot be allowed to exist as part of the literature.
What this case actually demonstrates
The lesson here is not that the retracted paper was good. It was a weak case series, and weak case series get published all the time without retraction. The lesson is what gets enforced and what does not.
A retraction notice is a public document. It is read by other journal editors, by clinicians evaluating evidence, by the search algorithms that index PubMed, by the AI systems that increasingly mediate medical information. When that notice contains a categorical scientific claim that is demonstrably false against the journal’s own corporate record, the FDA’s regulatory record, and the CDC’s clinical guidance — when it cites grounds for retraction that do not appear in the publisher’s own enumerated retraction policy — and when the only remaining justification would, if applied uniformly, invalidate the journal’s entire genre — the retraction has stopped being an act of scientific self-correction.
It has become a statement about which questions clinicians are permitted to ask and which observations they are permitted to publish.
That is gatekeeping. Not the dramatic kind, with denunciations and committees and public spectacle. The quiet kind. A footnote. A boring line of bureaucratic prose.
The kind that becomes visible only when you read the footnote carefully and notice what it says.
A question that does not appear in the retraction notice
There is one more thing the retraction notice does not say, and it is the thing I keep returning to.
Nattokinase is a food-derived enzyme. It has been consumed in fermented soybeans by hundreds of millions of people across multiple centuries. Its toxicological profile is well-characterized and shows no significant safety signals at therapeutic doses. It is not patentable as a molecule. It is not owned by any pharmaceutical company. It is sold as a dietary supplement for a few cents per dose.
The methodologically rigorous biochemistry on its spike-degradation activity has been in the peer-reviewed literature since 2022. The cardiovascular and fibrinolytic literature on it goes back four decades. Whatever its ultimate clinical role turns out to be in the post-exposure syndromes the IDCases paper described, the enzyme itself is not exotic, not novel, and not dangerous.
So consider the asymmetry the retraction notice enforces.
A clinician treating patients with refractory post-COVID and post-vaccination syndromes — patients who are real, whose symptoms are real, and for whom no approved therapy exists — observes that a protocol including a food-derived enzyme appears to help. He writes it up as a hypothesis-generating case series, in the appropriate genre journal, with appropriate methodological caveats. The journal retracts it on the grounds that the observations might cause clinicians to try the protocol before randomized controlled trials have been completed.
But randomized controlled trials of nattokinase for post-exposure syndromes are not being funded. They are not being funded because the molecule is unpatentable, the market is fragmented across dozens of supplement manufacturers, and no pharmaceutical company stands to recoup a hundred-million-dollar trial cost from sales of a fermented soybean enzyme. The institutional infrastructure that would generate the validation the retraction notice demands does not exist for compounds like this, and will not exist, because the financial incentives that fund pivotal trials are oriented around molecules that can be exclusively owned.
The retraction notice’s logic is therefore circular in a way that has real consequences. We retract this clinical observation because it has not been validated in trials. The trials will not happen because the substance cannot be owned. Therefore the observation can never enter the literature. The standard of evidence demanded is the standard the system structurally cannot supply for this category of intervention.
Meanwhile, patients exist now. Some of them are very sick. Some of them are running out of options. The clinicians who treat them are making decisions today, on the basis of whatever evidence is actually available — mechanistic studies, safety profiles, clinical observation, accumulated experience. Removing carefully-described clinical observation from the literature does not protect these patients. It deprives the clinicians treating them of one more piece of information they could weigh.
I am not arguing that nattokinase has been proven effective for post-spike syndromes. It has not by conventional evidentiary standards. Nor am I claiming that it is not profoundly therapeutic. I think the significant body of research that I have indexed on GreenMedInfo.com makes it clear that it is. I am arguing that “we retract clinical observations of unpatentable interventions until pharmaceutical-grade trials validate them” is not a neutral epistemic standard. It is a standard whose effect — whether anyone intends it or not — is to ensure that only interventions backed by ten-figure development capital can become part of the medical literature.
A retraction policy that produces this outcome, applied selectively to a food-derived enzyme being studied in the context of a post-vaccination syndrome the agencies regulating those vaccines have already conceded is real, is not protecting public health.
It is protecting something else.
This piece is part of an ongoing investigation into how post-pandemic chronic illness is being defined, attributed, and treated within the institutional record. For the broader argument that “Long COVID” may be functioning as a diagnostic umbrella that absorbs and obscures vaccine-induced injury — and the molecular and clinical evidence underlying that hypothesis — see Unmasking the “Long COVID” Cover Story for Vaccine Injury (Poisoned, Not Infected, Part 2).
The retraction case examined here is one observable instance of the larger pattern that piece describes: when carefully-described clinical observations point toward a vaccine-related etiology and recommend a non-patentable intervention, the institutional response is removal from the literature rather than further investigation. The two phenomena are connected. The structural logic is the same.
View and comment on the thread on X below.
It is the same with ivermectin.
It is the same with hydrochloroquine.
It is the same with fenbendazole.
It will be the same for anything with no profit potential.
The angle is the same, which tells you exactly where this pressure to unpublish and bury is coming from.
The science is settled. The vast majority of scientists agree with whoever is funding them. – Anon
A study designed not to find harm will never find harm.
We know they are lying. They know they are lying
They know that we know they are lying. We know that they know that we know they are lying. And they still continue to lie.
– Alexander Solzhenitsyn
They refuse to acknowledge God as the creator, and instead, they worship the creation!
Then scam and play god by "saving" you from certain doom.