Pre-Vaccine Antibodies May Predict Blunted COVID-19 Vaccine Response

Pre-existing antibodies to common microbes may help predict which individuals are likely to mount weaker antibody responses to vaccination, according to findings from a national longitudinal study published in Cell Press Blue.

The study included 1644 immunosuppressed patients and 2445 healthy individuals who were evaluated before and after COVID-19 vaccination. Investigators measured humoral immune responses to 185 antigens, including 3 SARS-CoV-2 antigens, 157 common microbial pathogen antigens, and 25 autoantigens.¹

Although blunted COVID-19 vaccine responses were more common among patients with solid organ transplant, multiple myeloma, autoimmune disease, inflammatory bowel disease, and HIV, responses were heterogeneous within every cohort. Approximately 5% to 6% of healthy individuals also mounted weak responses, according to the investigators.¹

“What our study found is that certain biomarkers, when analyzed with AI, can predict who is likely to respond well to a vaccine, even before they receive it. This suggests that some people may be more immune-ready than others,” lead author Joshua LaBaer, executive director, Biodesign Institute, Arizona State University, said in a press release.²

What are sentinel antibodies?

The investigators identified pre-existing antibodies to common microbes—including Staphylococcus aureus, respiratory syncytial virus (RSV), and human respirovirus 3—as “sentinel antibodies” that were associated with post-vaccination antibody responses. These antibodies may reflect baseline humoral immune competence, or the immune system’s broader readiness to generate antibody responses after new exposures.¹

The study found that most non-SARS-CoV-2 antibodies were stable over time. Among 114 seroprevalent antibodies evaluated longitudinally, 103, or 90.4%, did not change meaningfully over 18 months, with fewer than 2% of participants experiencing a change greater than 4-fold. The authors stated that these stable antibody profiles may reliably retain information about an individual’s prior infection history.

Which patients had weaker vaccine responses?

After initial vaccination, weak anti-receptor-binding domain responses were seen in 71.6% of solid organ transplant recipients, 48.5% of patients with multiple myeloma, 43.0% of patients with inflammatory bowel disease, and 39.0% of those with autoimmune disease, compared with 6.2% of healthy individuals. Weaker responses also were observed in 17.6% of patients with solid organ malignancy and 9.8% of patients with HIV, although the HIV difference was not statistically significant.

After booster vaccination, most healthy individuals and many immunosuppressed participants mounted high responses. However, weak or absent booster responses remained more common in certain cohorts, including 37.9% of solid organ transplant recipients and 17.5% of participants with autoimmune disease.

The findings suggest that broad diagnostic categories alone may be insufficient for predicting vaccine response. The authors noted that many immunosuppressed participants had robust responses, while a subset of healthy individuals had blunted responses.

How did the AI model perform?

Using global antimicrobial antibody profiles, the researchers developed a deep-learning model to stratify individuals by likelihood of a blunted vaccine response. The universal cohort-agnostic model included 98 non-SARS-CoV-2 and non-HIV seroprevalent antibodies, demographic information, and cohort labels. In repeated 5-fold cross-validation, the model had a mean area under the receiver operating characteristic curve of 0.766 and sensitivity of 59.4% at 80% specificity across held-out post-booster samples.

Predictive performance was strongest in the healthy cohort, with a mean cross-validation AUC of 0.803 and sensitivity of 73.4% at 80% specificity. Performance was lower in some immunosuppressed cohorts, particularly autoimmune disease and solid organ transplant groups.

The most informative antibodies for distinguishing high and low vaccine responders included antibodies to S aureus IsdB and RSV nucleoprotein in the healthy cohort, consistent with the study’s broader finding that common antimicrobial antibody profiles may serve as markers of humoral immune readiness.

Why is this relevant to primary care?

For primary care clinicians, the study adds to ongoing questions about how to identify patients who may not respond adequately to vaccination. Immunosuppression remains an important clinical risk factor, but these findings suggest that immune response may vary substantially even among patients with the same broad diagnosis.

The concept of sentinel antibodies could eventually support more individualized vaccination strategies, particularly for patients with immunosuppression, chronic inflammatory disease, malignancy, transplant history, or unexpectedly weak responses despite being otherwise healthy. However, this remains a research framework, not a clinical test ready for routine practice.

The findings also reinforce the importance of counseling patients that vaccine response is not uniform. Diagnosis, medication exposure, prior infection history, baseline antibody patterns, and broader immune function may all influence response.

What are the limitations?

The study had several limitations. Although more than 8000 unique samples from more than 4000 participants were analyzed, sample collection was uneven across clinical groups, and some solid organ transplant recipients were vaccinated before baseline samples could be obtained.

The antigen panel was limited to 185 antigens selected from prior knowledge and protein array data, which may have excluded other important differentiating antigens. The authors also noted that cross-reactivity could make some antibody responses difficult to attribute to one specific microbe.

Importantly, the study focused on antibody levels rather than functional immunity. The authors stated that future work should determine whether the antibody-level predictions correlate with infection risk or other clinical outcomes, and noted that socioeconomic status, lifestyle, and other demographic factors were not analyzed.

Reference

1. Song L, Park JG, Qiu J, et al. Pre-vaccine sentinel antibodies predict blunted vaccine responses. Cell Press Blue. 2026;1:100088. doi:10.1016/j.cpblue.2026.100088
2. Arizona State University. Why immune responses to vaccines vary from person to person. News release. Published August 20, 2026. Accessed August 25, 2026. https://news.asu.edu/20260820-science-and-technology-why-immune-responses-vaccines-vary-person-person