A Pfizer-BioNTech COVID-19 vaccine study has found significant short-term changes in the corneas of patients after two doses, prompting researchers to recommend closer monitoring of people with vulnerable eyes.
The peer-reviewed paper, published in Ophthalmic Epidemiology, examined 128 eyes belonging to 64 patients before vaccination and about 76 days after their second dose of the Pfizer-BioNTech BNT162b2 vaccine.
Researchers recorded an increase in corneal thickness, a reduction in the density of endothelial cells and changes in the size and shape of those cells following vaccination.
The corneal endothelium is a delicate layer of cells lining the inside of the cornea, the transparent front surface of the eye. These cells help keep the cornea clear by pumping out excess fluid.
Unlike many other cells, corneal endothelial cells have little ability to regenerate. When some are lost, neighbouring cells expand to cover the space.
The study found that median endothelial cell density fell from 2,597 to 2,378 cells per square millimetre, a reduction of about 8%.
Median central corneal thickness increased from 528 micrometres before vaccination to 542 micrometres afterwards, an increase of about 2.7%.
Researchers also found that the cells became less uniform in size. Their coefficient of variation increased from 39 to 42, while the proportion retaining a normal six-sided shape declined slightly.
All of the reported changes were statistically significant, according to the paper.
However, the researchers did not report that participants lost vision, developed corneal disease or suffered permanent eye damage during the short follow-up period.
The study also did not establish whether the changes persisted, worsened or returned to normal over time.
“The endothelium should be closely monitored in those with a low endothelial count or who have had a corneal graft,” the researchers concluded.
People who already have reduced endothelial reserves — including some older patients, people with conditions such as Fuchs’ endothelial dystrophy and those who have undergone eye surgery or a corneal transplant — could theoretically be more vulnerable to further cell loss.
If endothelial function deteriorates severely, fluid can accumulate inside the cornea, causing swelling, cloudiness and blurred vision. Advanced corneal decompensation can require transplantation.
The study does not show that the Pfizer vaccine causes those conditions, however. It found changes in measurements associated with endothelial health over a period of roughly two and a half months.
SMALL STUDY WITHOUT UNVACCINATED CONTROL GROUP
The research was a relatively small before-and-after observational study involving 64 people.
Because it did not compare vaccinated participants with a separate unvaccinated control group, it cannot establish with certainty that vaccination caused the changes.
Other factors occurring between the two examinations could potentially have influenced the results. The study also measured both eyes of each participant, meaning the 128 eyes were not 128 independent patients.
The researchers called for longer-term studies involving larger groups to determine whether the findings are temporary, clinically important or reproducible in other populations.
The paper, titled Evaluation of the Effects of mRNA-COVID-19 Vaccines on Corneal Endothelium, was written by Fatma Sumer and Sevgi Subasi and published online in July 2025. It appeared in Volume 32 of Ophthalmic Epidemiology.
The findings do not amount to evidence that Pfizer vaccination commonly causes blindness or major eye damage.
They instead identify measurable short-term changes that the authors said warrant further investigation, particularly among patients who already have low endothelial cell counts or transplanted corneas.
Separately, the U.S. Food and Drug Administration updated the labels of Pfizer and Moderna mRNA vaccines in June 2025 with additional information about the recognised risks of myocarditis and pericarditis. That regulatory action concerned inflammation of the heart and surrounding tissue and was unrelated to the corneal study.
