Messenger ribonucleic acid (mRNA) vaccines saved humanity from COVID-19, but there were people who were particularly prone to infection even after receiving the same shot. A Korea-Japan joint research team has developed a way to identify in advance vaccinated people who are vulnerable to COVID-19 infection.
A team led by Park Hyeong-gi, a professor in the Department of Convergence Biomedical Engineering at Pusan National University, said in the international journal Science Translational Medicine on the 17th (local time), "Based on the different patterns in how immunity is maintained among COVID-19 vaccine recipients, we discovered a method to single out people who are susceptible to infection."
This study was led by Professor Park Hyeong-gi when he was in the Department of Life Sciences at Nagoya University in Japan. Park moved to Pusan National University in March this year and is researching optimal health policies and clinical trial designs to effectively respond to infectious diseases such as COVID-19.
COVID-19 has passed the pandemic stage and has now shifted to endemic status, but the virus that causes COVID-19 continues to threaten humanity by evolving with new variants. The World Health Organization (WHO), in line with the endemic transition of COVID-19, has recommended since 2023 a "selective vaccination" strategy focused on high-risk groups, instead of the initial strategy of vaccinating the entire population, including medium- and low-risk groups. Amid these changes, deciding who should be prioritized for additional shots (booster shots) is becoming increasingly important.
The researchers conducted a study on 2,526 Japanese people who received mRNA vaccines from U.S. companies Pfizer or Moderna between April 2021 and November 2022. These vaccines contain mRNA, the genetic information for making the coronavirus spike protein. Once inside the body, they produce the virus's spike and trigger an immune response that induces antibodies.
The team collected participants' blood four to five times to track changes in immune responses by examining how many antibodies were present. At that time in Japan, the Alpha variant and the Omicron BA.1 and BA.2 variants were circulating.
Using artificial intelligence (AI) that processed the long-term follow-up data with mathematical modeling and machine learning, the researchers classified participants into three types: "sustained," which maintained high antibody levels after vaccination; "vulnerable," which had a low antibody response from the start; and "rapid-decline," which initially had a strong antibody response that quickly decreased.
According to the study, even after additional shots, about half of the participants did not change from their original type. In particular, the rapid-decline group became infected much sooner after vaccination than the other groups. The researchers also found that these individuals did not produce initial defensive antibodies that block viral entry. Initial defensive antibodies reside in the mucous membranes of the respiratory tract, such as the nose, mouth, and throat, and block the coronavirus as soon as it enters the body. The rapid-decline group had a weaker barrier, allowing the virus to penetrate more easily.
Until now, eligibility for additional shots has been determined only by criteria such as age and underlying conditions. The researchers said that by looking at individual antibody changes after vaccination, vulnerable groups can be identified more accurately.
Park said, "In addition to the existing criteria focused on older adults and the immunocompromised, we should also consider an 'immune marker' of vaccine response patterns," adding, "This study provides new scientific evidence for setting priorities for future booster vaccinations."
The team plans to further determine whether immune response patterns persist even after additional vaccinations and to clarify the mechanism by which antibody decline leads to actual infection risk.
Park said, "Building on these findings, we are developing simple markers, algorithms, and decision-support tools that can be applied immediately in patient care and vaccine development," adding, "At the same time, we are using simulations based on mathematical models to evaluate which vaccination strategies maintain population immunity most efficiently." He explained that they are extending this principle to other viruses, diseases, and vaccines beyond COVID-19.
References
Science Translational Medicine (2025), DOI: https://doi.org/10.1126/scitranslmed.adv4214