Vaccines prevent severe illness in many people, but the immune protection they provide can vary significantly from person to person. New research led by Arizona State University offers insights into what might be behind these differences.
Antibodies in the Blood May Reveal Vaccine Responsiveness
The immune system may provide signs of how strongly it will respond even before vaccination. Researchers at ASU and collaborating institutions analyzed blood samples from more than 4,000 people, identifying antibodies that recognized 185 antigens. These immune targets included common viruses and bacteria, as well as targets associated with autoimmune diseases. Artificial intelligence was then used to search for patterns in blood samples collected before and after a COVID-19 vaccination. The analysis revealed antibody signatures that could help distinguish individuals with a strong immune response to the vaccine from those with a weaker response. The findings could ultimately contribute to vaccination strategies that are more precisely tailored to an individual’s immune system.
“Our study found that certain biomarkers, when analyzed with AI, can predict who is likely to respond well to a vaccine even before it is administered. This suggests that some people may be better prepared immunologically than others,” said Joshua LaBaer, who led the study. LaBaer is executive director of the Biodesign Institute at ASU and director of the Virginia G. Piper Center for Personalized Diagnostics. The project also involved researchers from ASU as well as collaborators from medical and research institutions across the United States. The study was published in the latest issue of the journal Cell Press Blue .
Typically, scientists evaluate the immune response after vaccination by measuring whether the immune system has produced antibodies against the intended target molecule. In this study, the researchers approached the problem from the opposite direction. They wanted to know whether immune patterns already present in the blood could provide insight into how someone would respond even before receiving a vaccine. Many factors can influence the immune response to vaccination, including age, gender, genetics, past illnesses, and underlying medical conditions. People with conditions that compromise the immune system often show weaker responses. However, vaccination outcomes can vary widely even among individuals in the same general health category. The researchers used one of the first approaches to examine a comprehensive antibody “fingerprint”—present even before vaccination—as a measure of immune readiness. While some other predictive strategies rely on genetic testing, this method analyzes antibody patterns in the blood, which could potentially make it easier to implement in clinical practice.
Health Status Alone Says Nothing About the Response
To investigate whether these antibody fingerprints might indicate a readiness for vaccination, the team measured immune responses to 185 antigens. The target antigens included SARS-CoV-2, the virus responsible for COVID-19, as well as other common viruses and bacteria and target antigens associated with autoimmune diseases.
In total, the researchers analyzed 8,687 samples from 4,089 participants. The group included healthy volunteers as well as people with conditions or treatments associated with immunosuppression, including HIV, multiple myeloma, malignant solid-organ tumors, autoimmune diseases, inflammatory bowel diseases, and solid-organ transplants.
Several immunosuppressed groups were more likely to have a reduced response to the COVID-19 vaccine. However, simply classifying a person as “immunosuppressed” or “healthy” did not allow for a reliable prediction of the outcome. Some participants with weakened immune systems nevertheless developed strong responses. At the same time, about 5% to 6% of healthy participants showed weak vaccine responses.
“Sentinel” Antibodies Signal Immune Readiness
Certain antibodies that were already present in the blood prior to vaccination stood out particularly in the analysis. Higher concentrations of antibodies against common microbes, includingStaphylococcus aureus, RSV, and human respiratory syncytial virus type 3, were associated with stronger responses to COVID-19 vaccines. The researchers refer to these antibodies as “sentinel” antibodies, as they can serve as a kind of early warning signal for a person’s underlying immune readiness. This does not mean that these antibodies directly support the vaccine. Rather, they could provide clues as to how well certain parts of the immune system can respond to a new stimulus even before vaccination. The antibodies also reflect a person’s individual immune history, which has been shaped by previous exposure to viruses and bacteria.
The team therefore also investigated whether the complete antibody fingerprint might be more informative than individual biomarkers. To do this, a deep-learning model analyzed numerous antibody levels simultaneously and searched for patterns associated with a strong or weak immune response. This approach could be particularly helpful because the immune response is shaped by many factors and an individual’s personal history. The results suggest that it is not a single antibody that is decisive, but rather the interplay of many immune signals. AI could help reveal such complex patterns, which would be difficult to detect using conventional analytical methods. Further studies are still needed to determine whether these signatures can be reliably applied to other vaccines and population groups.
AI Scans Through Millions of Immune Signals
The results highlight a potential advantage of using AI in biomedical research. Machine learning systems can analyze millions of biological data points simultaneously and search for subtle correlations that are difficult to detect using conventional methods. This approach can be particularly helpful when studying the immune system, as numerous factors interact simultaneously. Age, pre-existing conditions, past infections and vaccinations, as well as an individual’s immune history, can influence how the body responds to a vaccine.
In this case, the deep learning model analyzed not only individual antibodies but also the patterns of the entire antibody profile under investigation. This enabled the researchers to recognize that certain combinations of immune signals may reveal more about the subsequent vaccine response than a single biomarker. This suggests that the immune system should be viewed more as an interconnected whole. The AI does not make medical decisions but helps researchers identify correlations in large datasets that might otherwise be easily overlooked.
The research also highlights the potential of newer technologies that allow for the simultaneous measurement of many antibody responses. Instead of merely testing whether a person has antibodies against a single pathogen, researchers can create a much more comprehensive picture of the immune system. This profile reflects the immune system’s numerous encounters with viruses, bacteria, and other immune targets. Such large-scale analyses could help us better understand in the future why people react differently to the same vaccine.
Toward More Personalized Vaccination
However, the application of such analyses in medical practice is still in its infancy. The patterns identified must first be verified in further studies. If the results are confirmed in future studies and can also be applied to other vaccines, this approach could find application far beyond COVID-19. The analysis of sentinel antibodies could help identify, even before vaccination, which people are likely to develop a strong immune response and which may develop a weaker one. This would be particularly useful for older adults or people with medical conditions and treatments that can compromise the immune system. At the same time, these findings could support vaccine research and development by providing a better understanding of the biological conditions that favor successful immunization.
In the long term, such a test could provide physicians with additional information to plan vaccinations more individually. For people with an expectedly weak immune response, for example, closer monitoring of the immune reaction might be advisable. Additional vaccine doses or other protective measures could also be considered in a more targeted manner, provided that future studies show that such strategies actually provide additional benefit. Conversely, people expected to have a strong immune response might not need additional measures. The goal here would not be to replace existing vaccination recommendations outright, but rather to supplement them with information about an individual’s immune status.
Vaccine development could also benefit from a better understanding of these differences. If it is known which characteristics of the immune system are associated with a particularly good vaccine response, researchers could investigate more specifically how to promote an adequate immune response even in people with a weaker baseline immune system. These findings could thus be relevant not only to the question of who responds well to a vaccine, but also to how vaccines can be used or developed in the future so that as many people as possible develop reliable protection.
However, there is still a long way to go before these findings can be applied in medical practice. The current study initially shows that measurable indicators of the subsequent immune response can be found in the blood even before vaccination. Further studies are now needed to clarify how reliable these signatures actually are, whether they also work with other vaccines, and whether concrete medical decisions can be derived from them. If this approach is confirmed, it could represent an important step toward more personalized vaccination medicine in the long term—one that takes into account not only the vaccine but also the individual’s immune biology to a greater extent.