Patients with an aggressive brain cancer who received a COVID-19 vaccine in the months before tumor surgery lived roughly twice as long as unvaccinated patients, according to an analysis that has not yet been peer-reviewed.
The preprint, posted to medRxiv in July by researchers at the University of Alabama at Birmingham, examined 187 people with glioblastoma who had tumors biopsied or surgically removed between 2022 and 2025.
The distinction that governs how to read this is stated by the researchers themselves. James Markert, the UAB neurosurgeon who led the work, told Science News that his team found an association between vaccination and survival, not a cause-and-effect relationship.
Inside the Numbers from 187 Glioblastoma Patients
Among patients vaccinated within 100 days before biopsy or surgery, half lived 743 days or longer. Median survival among unvaccinated patients was 349 days. For a cancer with no cure and few options, that difference amounts to roughly an additional year.
The comparison depends on how the control group is drawn. The preprint's own unmatched analysis reports 743 days against 318 days for patients who were not vaccinated in that perioperative window, a group that is not identical to patients who were never vaccinated at all. Both figures show the same direction of effect but use different denominators.
The link appeared only with COVID-19 vaccines. Other vaccines the team examined as negative controls showed no comparable association.
Glioblastoma is the most common malignant primary brain tumor in adults and among the most lethal. Standard treatment is surgery followed by radiation and chemotherapy, and median survival is typically measured in months rather than years. That context is why a signal of this size drew attention, and also why it needs to be handled carefully.
The researchers attempted to control for confounders, including the severity of COVID-19 infections and whether vaccinated patients had better access to health care than unvaccinated patients. Markert has acknowledged that accounting for everything is impossible, leaving the mechanism unresolved.
Preprint Status Is the First Thing to Read
A preprint is a manuscript posted publicly before independent experts have reviewed it. Peer review catches errors in statistics, design, and interpretation, and papers routinely change between preprint and publication. Some never publish at all. The medRxiv posting includes the platform's standard notice that the work has not been peer-reviewed and should not be used to guide clinical practice.
This is also a retrospective analysis of records rather than a trial. Nobody was randomly assigned to receive the vaccine or not, so the two groups may differ in ways the researchers could not measure. People who sought out a COVID vaccine may differ systematically from those who did not in health engagement, comorbidities, performance status at diagnosis, or in factors nobody has thought to check.
The sample is also modest. One hundred eighty-seven patients across a single institution's experience is enough to generate a hypothesis and not enough to settle one.
None of that makes the finding uninteresting. It makes it preliminary, which is a different thing.
A Pattern That Has Now Appeared in Three Cancers
The result does not stand alone, and that is the strongest argument for taking it seriously.
Last year, researchers reported in Nature that patients with advanced non-small cell lung cancer or metastatic melanoma who received an mRNA COVID vaccine within 100 days of starting immune checkpoint inhibitor therapy lived substantially longer than unvaccinated patients. That work, from University of Florida and MD Anderson researchers and summarized in a Nature news article, also found no equivalent effect from influenza or pneumococcal vaccines. Laboratory results in the same paper indicated a type I interferon response that increased tumor PD-L1 expression, a plausible mechanism by which a nonspecific immune trigger could sensitize tumors to immunotherapy.
Elias Sayour, a pediatric oncologist at the University of Florida who was not involved in the glioblastoma work, said that evidence is mounting that mRNA COVID-19 vaccines may spur anti-cancer activity.
A separate line of caution applies to that earlier work. An independent reanalysis posted to medRxiv applied methods designed to avoid common biases in this kind of retrospective comparison and found a much smaller difference in lung cancer survival, 6.4 percentage points at three years, with a confidence interval that crossed zero. That is the central problem with observational findings of this type: the size of the effect can shift substantially depending on how the analysis accounts for differences between people who get vaccinated and those who do not.
Two observational findings pointing in the same direction are more persuasive than one, but they are not a substitute for a randomized trial, which is the only design that can separate the vaccine from the patients who choose to receive it.
Nothing Here Changes Treatment Today
No treatment guideline has changed, and nobody involved in this research is recommending that patients act on it.
Delaying glioblastoma surgery is dangerous, and any theoretical benefit from timing a vaccine would be swamped by the harm of a postponed operation. The preprint itself carries an explicit warning against using it to guide clinical practice.
Patients or families who want to discuss COVID vaccination should raise it with their oncology team, as they would with other standard reasons a person undergoing cancer treatment might be advised to stay up to date on vaccines. Immunosuppressed patients have their own vaccine timing considerations that a treating clinician is positioned to weigh.
Two things happen next. The UAB team is working to validate the finding in a separate group of patients, and researchers in this field have called for a trial testing the question directly rather than inferring it from records.
What remains unknown is whether the association survives peer review, whether it holds in other patient groups, and whether any of it reflects the vaccine rather than differences between people who receive it and those who do not. MedicalDaily has reported on related mRNA cancer vaccine research and will report the peer-reviewed publication and any trial results.
Key Questions Answered
What did the analysis find? Among 187 glioblastoma patients, those vaccinated against COVID within 100 days before biopsy or surgery had a median survival of 743 days, compared with 349 days for unvaccinated patients.
Has it been peer-reviewed? No. It is a preprint posted to medRxiv and has not undergone independent expert review. Findings may change or not be published.
Does this prove the vaccine extended survival? No. It is a retrospective association. Patients were not randomly assigned, so unmeasured differences between the groups could explain the result.
Did other vaccines show the same effect? No. Other vaccines examined as negative controls showed no comparable association in this analysis.
Should a patient delay surgery to get vaccinated? No. The preprint states it should not be used to guide clinical practice, and delaying surgery carries real harm. Vaccination questions belong with the treating oncology team.
Is there other evidence pointing in the same direction? Yes, with caveats. A Nature analysis of lung cancer and melanoma patients found longer survival among those vaccinated near the start of immunotherapy, though an independent reanalysis found a much smaller effect.
What happens next? The team is trying to replicate the finding in a different patient group, and researchers have called for a randomized trial to test the question directly.
