(WindNight/Shutterstock)
Severe COVID-19 raises lung cancer risk by 24%, study of 76 million Americans finds
In A Nutshell
- People hospitalized with severe COVID-19 had roughly a 24 percent higher risk of developing lung cancer compared to those never infected, even after accounting for smoking and other known risk factors.
- Animal experiments show severe respiratory infections leave durable immune changes in the lungs that accelerate tumor growth long after the virus is gone, an effect not seen after mild infections.
- Vaccines appeared to prevent the lung changes linked to increased tumor growth in animal models, suggesting protection may extend beyond the acute illness itself.
- A two-drug combination reduced tumor burden and restored cancer-killing immune cell activity in previously infected animals, pointing toward a potential treatment strategy for high-risk survivors.
Surviving a severe case of COVID-19 or influenza feels like crossing a finish line. Fever breaks, oxygen levels stabilize, and eventually life returns to normal. A major new study published in the journal Cell suggests that for people who were hospitalized with either virus, recovery may not be the clean ending it appears to be. Deep inside the lung tissue, the infection may leave behind a biological imprint that silently raises the risk of cancer later on.
Researchers at the University of Virginia, along with colleagues at the University of Pittsburgh, Mayo Clinic, and Cedars-Sinai Medical Center, analyzed de-identified health records from nearly 76 million American adults. People hospitalized with severe COVID-19 had roughly a 24 percent higher risk of subsequently developing lung cancer compared to people who had never been infected. That elevated risk held up after researchers accounted for smoking history, age, sex, and a range of pre-existing conditions. Severe COVID-19 was independently associated with higher lung cancer risk. People with mild or moderate COVID-19, by contrast, showed a modest decrease in lung cancer diagnoses, a contrast that makes the severity distinction difficult to dismiss.
The human data is correlational, meaning it cannot prove that infection directly causes cancer, and the observation window is still relatively short. But the association persisted after accounting for every major known risk factor the researchers could measure.
What the team set out to understand was not just whether the risk was real, but why. Their answer points to something the lungs carry long after the virus is gone.
How Severe COVID-19 and Flu Raise Lung Cancer Risk
To trace the mechanism, researchers infected mice with either SARS-CoV-2 or influenza A, allowed them to fully recover, and then introduced lung tumor cells. Previously infected animals consistently developed larger, faster-growing tumors and died sooner than mice that had never been infected. The pro-tumor effect from influenza was still detectable four months after the initial infection.
What the infection left behind was not the virus itself. Using a technique that maps which regions of DNA are switched on or off, the team found that severe respiratory infections produced durable changes in multiple types of lung cells, including immune cells, structural cells, and the cells lining the airways. Those changes caused the lungs to keep overproducing a protein called G-CSF long after recovery. G-CSF is a recruiting signal for neutrophils, white blood cells that normally arrive during infection, do their job, and leave.
In the post-viral lung, a specific subset of neutrophils accumulated in large numbers and stayed. Rather than protecting the host, they suppressed anti-tumor immune responses and made it easier for tumor cells to grow. At the same time, the immune system’s dedicated cancer killers, called CD8 T cells, were functionally weakened. Fewer were producing the proteins needed to destroy tumor cells.
Cell, DOI: 10.1016/j.cell.2026.02.013)
The Biological Scar Behind the Lung Cancer Link
A central question was whether this altered immune state was a local lung phenomenon or a body-wide change. Researchers answered it by surgically joining the circulatory systems of two mice so they shared a bloodstream. When a previously infected mouse was paired with a healthy one, the abnormal neutrophil buildup in the infected animal’s lung tissue persisted. Healthy immune cells flowing in from the uninfected partner did not fix it. The lung had been changed on its own terms, independent of the rest of the body.
When researchers directly gave G-CSF to healthy, never-infected mice, those animals developed the same abnormal neutrophil buildup and showed faster tumor growth. Blocking the G-CSF signal in previously infected mice reduced the problematic neutrophils and slowed cancer progression. The protein driving all of this was not a remnant of the virus. It was being produced by a lung that the infection had durably rewired.
How COVID and Flu Vaccines May Reduce Lung Cancer Risk
Mice vaccinated against SARS-CoV-2 with an mRNA spike vaccine before virus exposure were fully protected from serious illness. When those vaccinated animals were later challenged with tumor cells, they showed substantially lower tumor burden than unvaccinated animals that had experienced full infection. A parallel experiment using an inactivated influenza vaccine produced the same result.
Severity mattered even in unvaccinated animals. Mice exposed to low doses of either virus, enough to cause only mild disease, did not show a meaningful increase in tumor burden. It was severe infection, the kind involving major lung inflammation, that drove the reprogramming. Vaccines, in these models, appeared to block not just the acute illness but the cancer-related consequences that followed it.
A Drug Combination That Reversed Post-Viral Tumor Growth
On the treatment side, researchers tested two drugs in post-COVID tumor-bearing mice. Reparixin, which blocks neutrophil recruitment, reduced the problematic neutrophils and partially decreased tumor burden. An antibody targeting PD-L1, a protein tumors use to hide from immune attack, produced modest improvement on its own. Together, the two drugs worked considerably better: tumor burden dropped, cancer-killing T cells increased inside tumor tissue, and more of those T cells were actively doing their job.
That combination, cutting off the neutrophil-driven immune suppression while restoring T cell function, suggests a treatment approach that could eventually be tested in people who survived severe COVID-19 or influenza.
For the tens of millions of Americans hospitalized with COVID-19 during the pandemic, and the millions more who have been seriously ill with influenza over the years, this study raises a question medicine has not yet fully confronted: whether severe respiratory viral infection should be considered a potential lung cancer risk factor alongside smoking history and occupational exposures.
The biological mechanism, demonstrated primarily in mice, still needs to be confirmed in humans, and prospective studies are needed before any formal recommendations can be made. But the depth of this work moves that conversation into considerably firmer scientific ground.
Disclaimer: This article is based on findings from a peer-reviewed study combining population-level health record data and animal experiments. The human data are associational and cannot prove that COVID-19 or influenza directly causes lung cancer. The biological mechanism was demonstrated primarily in mice and has not yet been confirmed in humans. Readers should not draw clinical conclusions from this research. Consult a qualified healthcare provider with any concerns about lung cancer risk.
Paper Notes
Limitations
The human data in this study are retrospective, meaning researchers analyzed existing health records rather than following patients prospectively in a controlled trial. As the authors acknowledged, it cannot be ruled out that people who developed severe COVID-19 may have already harbored undetected precancerous changes that contributed both to disease severity and to subsequent cancer diagnosis. The animal models, while mechanistically informative, used specific inbred mouse strains and may not fully capture the complexity of human immune responses. The research focused on the lung microenvironment, and whether infection-imprinted mechanisms operate similarly in other organs remains unknown. While G-CSF and SiglecF-high neutrophils are identified as key drivers, the authors note that other pro-tumor immune populations, including macrophages and pathways involving IL-6 and IL-1, may also contribute. Prospective human studies will be needed before any clinical surveillance recommendations can be formally established.
Funding and Disclosures
This research was supported by multiple grants from the U.S. National Institutes of Health, including awards AI147394, AG069264, AI112844, HL170961, AI176171, and AG090337, all awarded to lead author Jie Sun. Additional funding came from the UVA Pinn Scholar Award, the UVA Comprehensive Cancer Center Collaborative Grant (U01CA224293), the UVA Shannon Fellowship, the UVA Lung TRT Pilot Grant, the American Lung Association Catalyst Grant, and several NIH training grants supporting co-authors. Schematics were created with BioRender.com. The University of Virginia has filed a provisional patent disclosure related to local targeting of G-CSF signaling in lung cancer. No additional competing interests were declared.
Publication Details
Authors: Wei Qian, Xiaoqin Wei, Andrew J. Barros, Xiangyu Ye, Haibo Zhang, Qing Yu, Samuel P. Young, Eric V. Yeatts, Yury Park, Chaofan Li, Sijie Hao, Gislane Almeida-Santos, Jinyi Tang, Harish Narasimhan, Nicole A. Kirk, Valeria Molinary, Ying Li, Li Li, Bimal N. Desai, Peter Chen, Kwon-Sik Park, Anny Xiaobo Zhou, Jeffrey M. Sturek, Wei Chen, In Su Cheon, and Jie Sun. | Journal: Cell, Volume 189, May 14, 2026. | DOI: https://doi.org/10.1016/j.cell.2026.02.013 | Correspondence: [email protected]
