Infectious diseases, including invasive bacterial infections, represent the most frequent cause of death among infants and young children worldwide.1–6 Immunization programs have been the most successful strategy to decrease the morbidity and mortality caused by infections in children across the world.7–9 However, despite the enormous success of the immunization programs, important knowledge gaps remain that need to be addressed to further optimize the global impact of childhood vaccinations. One aspect that has attracted considerable attention is the variability of individual immune responses.
In this issue of Pediatrics, Chapman et al.10 provide initial evidence of an important factor that may contribute to the differences in immune responses among young children after vaccination. Based on previous studies, mostly conducted in animal models, the authors hypothesized that the composition and diversity of the gut microbiome play a significant role in immune development and that by altering the gut microbiome with antibiotic treatment modifies childhood vaccine immune responses. These hypotheses are bolstered by a recent study that found that adults treated with broad-spectrum antibiotics had reduced gut bacterial load and diversity and a significant reduction in neutralizing antibody responses to the H1N1 influenza strain after vaccination.11 The vaccine response showed a marked inflammatory profile similar to that seen among vaccinated elderly individuals instead of the interferon profile observed in children, young adults, and pregnant women.11–14
This is a retrospective study of a prospectively enrolled cohort of children 6 to 24 months of age initially studied for acute respiratory infections, including acute otitis media. Blood samples were obtained during well-child visits at 6, 9, 12, 15, 18, and 24 months of age, as well as at the onset of acute otitis media. The investigators reviewed the medical records and queried parents about illnesses and antibiotic prescriptions. Study children received the standard vaccines (diphtheria, tetanus, and acellular pertussis [DTaP], haemophilus influenzae type b [Hib], and inactivated polio vaccine [IPV] at 2, 4, 6 and 18 months; pneumococcus conjugate vaccine [PCV] was administered at 2, 4, 6 and 15 months of age). They measured serum antibody titers to several vaccine antigens included in DTaP, Hib, IPV, as well as PCV serotypes 6B, 14, and 23F.10 A total of 560 children met the inclusion criteria. 342 children had 1678 antibiotic courses prescribed and 218 children with no antibiotic prescriptions served as controls. Overall, the investigators measured 11 888 antibody titers to the different vaccine antigens. Not surprisingly, day care attendance was associated with more frequent antibiotic prescriptions (42.7% vs 20.6%, P < .001). The authors also noticed that race and ethnicity, likely reflecting differences in access to care and the social determinants of health, were also associated with the frequency of antibiotic prescriptions.
The study results are remarkable. Children from 9 to 24 months of age who were prescribed antibiotics had a higher frequency of vaccine-induced antibody levels below the level of protection compared with children with no antibiotic prescriptions. With this initial observation, additional analyses were performed to examine the consistency of their findings. Treatment with broad-spectrum antibiotics and a longer duration of therapy were associated with antibody titers that were more frequently below the level of protection. A greater number of antibiotic courses during the first year of life was also associated with lower antibody titers. Together, these observations suggest a dose-dependent effect of the antibiotics in reducing the antibody responses to vaccines The investigators also examined the association of the antibiotics in the second year of life, after vaccine boosters. Consistent with the earlier findings, vaccine-induced antibody titers were reduced according to the number of antibiotic courses that the children had received.
As the authors indicated, the study has inherent limitations because of its retrospective study design and the lack of stool samples, which would have allowed direct assessment of the effect of antibiotics on the composition of the gut microbiome. Other aspects need to be considered, such as whether there was a possible bias by including children with frequent otitis media who might have lower vaccine-induced antibody titers in the study, regardless of antibiotic exposure. In the discussion, the authors summarized their previous studies where they identified otitis-prone children with associated low vaccine responses. But they clearly pointed out that the percentage of children with the phenotype of frequent otitis media and low vaccine antibody titers represented only 10% to 15% of the study population, whereas the percentage of children showing lower vaccine-induced antibody titers was much larger in the present study. Future research is needed to explore whether the findings are due to other sources of confounding; for example, might there be a group of children with lower responses to vaccines and who also develop more frequent viral infections leading to greater exposure to antibiotics. Regardless, it is important to note the high percentage of children in both groups (with or without antibiotic prescriptions) who had antibody titers below the level of protection. Although the authors did not report increased incidence of vaccine-associated infections, this observation clearly deserves future research.
Despite these limitations, the authors deserve major credit for their innovative investigation. Their observations could have major implications for clinical practice and should stimulate larger prospective multicenter studies involving more diverse populations to confirm these findings. Future studies should also include mechanistic analyses that permit understanding the origin of these reduced antibody responses, and eventually design approaches to mitigate their potential clinical impact. Furthermore, the availability of the novel multiomic analytical tools that can be used with small volume samples will facilitate these studies. Comprehensive analysis of early life immunity is a new research frontier where there is much to be learned to improve our immunization programs. It is important to mention the role of the National Institutes of Health in encouraging and supporting these studies.15 Now, it is up to us to optimize study design and to advocate for pediatric institutions and organizations to further support clinical research in this fundamental area of children’s health.
FUNDING: The authors are supported in part by the National Institutes of Health grant U01AI1313.
CONFLICT OF INTEREST DISCLOSURES: Dr Mejias has received research grants from the National Institutes of Health, Janssen and Merck, and fees from Janssen, Sanofi and Merck for participation in advisory boards. Dr Ramilo has received research grants from the Bill & Melinda Gates Foundation, Janssen and the National Institutes of Health; and fees from Adagio, Merck, Lilly, Pfizer, and Sanofi for participation in advisory boards; and fees from Pfizer and Sanofi for educational lectures.
COMPANION PAPER: A companion to this article can be found online at www.pediatrics.org/cgi/doi/10.1542/peds.2021-052061.