The messenger RNA (mRNA)-based coronavirus disease 2019 vaccines approved for use in children <5 years of age have different antigen doses and administration schedules that could affect vaccine immunogenicity and effectiveness. We sought to compare the strength and breadth of serum binding and neutralizing antibodies to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) elicited by monovalent mRNA-based coronavirus disease 2019 vaccines in young children.
We conducted a prospective cohort study of children 6 months to 4 years of age who completed primary series vaccination with monovalent mRNA-1273 or BNT162b2 vaccines. Serum was collected 1 month after primary vaccine series completion for the measurement of SARS-CoV-2-specific humoral immune responses, including antibody binding responses to Spike proteins from an ancestral strain (D614G) and major variants of SARS-CoV-2 and antibody neutralizing activity against D614G and Omicron subvariants (BA.1, BA.4/5).
Of 75 participants, 40 (53%) received mRNA-1273 and 35 (47%) received BNT162b2. Children receiving either primary vaccine series developed robust and broad SARS-CoV-2-specific binding and neutralizing antibodies, including to Omicron subvariants. Children with a previous history of SARS-CoV-2 infection developed significantly higher antibody binding responses and neutralization titers to Omicron subvariants, which is consistent with the occurrence of identified infections during the circulation of Omicron subvariants in the region.
Monovalent mRNA-1273 and BNT162b2 elicited similar antibody responses 1 month after vaccination in young children. In addition, previous infection significantly enhanced the strength of antibody responses to Omicron subvariants. The authors of future studies should evaluate incorporation of these vaccines into the standard childhood immunization schedule.
What’s Known on This Subject:
The messenger RNA-based COVID-19 vaccines approved for use in children <5 years of age have different antigen doses and administration schedules that could affect vaccine immunogenicity and effectiveness; however, there have not been any direct comparisons of these vaccines.
What This Study Adds:
In a cross-sectional analysis of 75 children 6 months to 4 years of age, we observed no differences in the breadth of severe acute respiratory syndrome coronavirus 2-specific antibody binding or neutralizing activity 1 month after the completion of primary series vaccination with monovalent COVID-19 vaccines.
As of December 2023, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused >3.5 million infections and 800 deaths among children <5 years of age in the United States.1 In June 2022, the US Food and Drug Administration issued emergency use authorizations for 2 vaccines, messenger RNA (mRNA)-1273, which was developed by Moderna, and BNT162b2, which was developed by Pfizer-BioNTech, for the prevention of coronavirus disease 2019 (COVID-19) in young children.2 Both products are mRNA-based vaccines that encode the viral Spike protein of an ancestral SARS-CoV-2 strain; however, the antigen doses and administration schedules for these vaccines differ substantially for this age group. The mRNA-1273 primary series for children in this age category consists of 2 25-µg doses separated by 4 to 8 weeks.2 –4 In contrast, the BNT162b2 primary series consists of 3 3-µg doses, with the second dose administered 3 to 8 weeks after the first dose and the third dose administered at least 8 weeks after the second dose.2 ,5 ,6
The issuance of the emergency use authorizations for the use of mRNA-based COVID-19 vaccines in young children was based on safety, immunogenicity, and efficacy data from large randomized, controlled clinical trials.4 ,6 A phase 2 to 3 trial of mRNA-1273 in 6403 children 6 months to 5 years of age revealed that 2 25-μg doses generated noninferior neutralizing antibody responses to an ancestral SARS-CoV-2 strain compared with young adults who received 2 100-μg doses; the estimated efficacy in preventing symptomatic COVID-19 was 51% among children 6 to 23 months of age and 37% among children 2 to 5 years of age.4 The phase 2 to 3 trial of BNT162b2 in 4526 children 6 months to 4 years of age revealed that 3 3-μg doses generated noninferior neutralizing antibody responses to an ancestral SARS-CoV-2 strain compared with individuals 16 to 25 years of age who had received 2 30-μg doses; the estimated efficacy in preventing symptomatic COVID-19 was 76% in children 6 to 23 months of age and 72% among children 2 to 4 years of age, although these estimates were based on only 34 infections.6 Because of the widespread circulation of Omicron variants, Moderna and Pfizer-BioNTech subsequently developed bivalent formulations that included mRNA encoding the Omicron BA.4/5 Spike and, more recently, a new monovalent vaccine containing mRNA encoding the Omicron XBB.1.5 Spike.7 –9 The overall dosages and administration schedule of the updated vaccine formulations have remained identical to those of the original monovalent vaccines in this age group.
Despite the similarity of the antigens contained in mRNA-1273 and BNT162b2, the differences in dosage and administration schedule in children <5 years of age could affect the magnitude and breadth of vaccine-elicited immune responses and, in turn, their effectiveness in preventing COVID-19. To date, there has been no direct comparison of the immunogenicity of these vaccines in this age category. Several studies in adults have suggested that BNT162b2 may be less effective at preventing symptomatic infections and severe disease;10 –13 however, the authors of recent studies in young children have not identified a consistent difference between the 2 vaccines.14 ,15 Notably, vaccine effectiveness estimates have been impacted by differences in the timing of vaccine approvals with respect to the circulation of different variants. Thus, further studies are needed to understand how antigen dose and dosage timing influence vaccine responses in this age group.
In this study, we measured serum SARS-CoV-2-specific humoral immune responses among 75 children 6 months to 4 years of age ∼1 month after completion of the primary series of monovalent mRNA-1273 or BNT162b2. We compared binding of antibodies to the Spike proteins of an ancestral SARS-CoV-2 strain and 13 variants of concern by vaccine product and measured serum neutralizing activity to an ancestral SARS-CoV-2 strain and to Omicron subvariants BA.1 and BA.4/BA.5. Our findings provide a head-to-head comparison of the immunogenicity of these 2 mRNA-based COVID-19 vaccines in young children, including the impact of different dosing strategies on the breadth and functional activity of vaccine-elicited antibodies.
Methods
Study Population and Procedures
The Duke Biospecimens from RespirAtory Virus-Exposed Kids study is a prospective cohort study of children, adolescents, and young adults with SARS-CoV-2 infection or exposure or who have recently been vaccinated for COVID-19.16 ,17 For the analyses presented herein, we recruited children 6 months to 4 years of age who completed the primary vaccine series with monovalent mRNA-1273 or BNT162b between July 1, 2022 and December 5, 2022. Sociodemographic and clinical information were collected through a caregiver questionnaire and review of medical records. Study visits were conducted ∼1 month after completion of the primary vaccine series, during which whole blood was collected from participants via venipuncture. The Serum was isolated from whole blood via centrifugation and frozen to −80°C before to immunologic analyses. We identified previous SARS-CoV-2 infections through medical record review and caregiver report. A positive result from polymerase chain reaction (PCR) or antigen testing of a respiratory sample was considered to be confirmation of previous infection. Additionally, serum from all participants was tested for SARS-CoV-2 nucleocapsid antibodies using the Platelia SARS-CoV-2 Total Antibody assay (BioRad, Hercules, CA). Participants were classified as having previous SARS-CoV-2 infection if they had a history of PCR- or antigen-confirmed infection or detectable serum nucleocapsid antibodies.
Measurement of SARS-CoV-2-Specific Binding Antibodies
We measured immunoglobulin G-binding (IgG-binding) antibodies specific to SARS-CoV-2 Spike proteins in participant sera using 2 binding antibody multiplex assay (BAMA) panels that included SARS-CoV-2 Spike proteins from an ancestral SARS-CoV-2 strain (D614G) and the following variants: Alpha, Beta, Gamma, Delta, Delta Plus, Eta, Iota, Kappa, Lambda, and Omicron subvariants B.1.1.529, BA.1.1, BA.2, and BA.4/5 (see Supplemental Table 4 for reagent details). Binding antibody assays were performed as previously described.16
SARS-CoV-2 Pseudovirus Neutralization Assays
SARS-CoV-2 neutralization was measured by using pseudotyped viruses with Spike proteins from D614G, Omicron BA.1, and Omicron BA.4/5. Pseudovirion production and neutralization assays were performed as previously described (see Supplemental Table 4 for reagent details).16 ,18 Luminescence was measured with the Synergy HTX Microplate Reader (BioTek, Winooski, VT). Neutralization titers were reported as the serum dilution at which relative luminescence units (RLUs) were reduced by 50% (50% inhibitory dilution [ID50]) compared with virus control wells after the subtraction of background RLUs.
Statistical Analysis
Study population characteristics were described by using medians and interquartile ranges (IQR) for continuous variables and frequencies and percentages for categorical variables. The balance between vaccine groups in all categorical covariates was assessed by using Fisher’s exact tests. Pairwise correlation tests were run across all binding and neutralization assays. Antibody binding and neutralization data were compared between vaccine products and stratified by previous history of SARS-CoV-2 infection and age category (6–23 or 24–59 months) using 2-way Wilcoxon rank-sum tests. The results were further validated by using generalized linear models with the log-transformed values for antibody assays as dependent variables and the vaccine product, previous SARS-CoV-2 infection, and age category as independent variables to test for possible interactions between independent variables. A false discovery rate threshold of 0.1 was applied by using the Benjamini-Hochberg method to account for multiple testing. Study data were managed by using REDCap electronic data capture tools hosted at Duke University. Analyses were conducted with R version 4.2.319 and packages lme4,20 corrplot,21 beeswarm,22 and ggplot2.23
Study Approval
This study was approved by the Duke University Health System Institutional Review Board (Pro00106150). Informed consent was provided by a legal guardian for all study participants using an electronic consent document.
Results
Participant Characteristics
Of 75 study participants, 40 (53%) received mRNA-1273 and 35 (47%) received BNT162b2 (Table 1). Among participants receiving mRNA-1273, the median (IQR) age was 25 (13, 44) months, 45% were 6 to 23 months of age, and 40% were female. Among participants receiving BNT162b2, median (IQR) age was 24 (13–33) months, 49% were 6 to 23 months of age, and 46% were female. A single participant who received mRNA-1273 had a chronic medical condition (valvular heart disease). A total of 24 participants reported a history of previous SARS-CoV-2 infection, including 10 (25%) participants who received mRNA-1273 and 14 (40%) who received BNT162b2. Nucleocapsid antibodies were detected in an additional 2 (5%) mRNA-1273 recipients and 3 (9%) BNT162b2 recipients.
Demographic and Clinical Features of the Study Population by Vaccine Product
| Characteristics | mRNA-1273 (n = 40) | BNT162b2 (n = 35) | P | ||
|---|---|---|---|---|---|
| Median (IQR) age, mo | 25 | (13, 44) | 24 | (13, 33) | .41 |
| Age category | .94 | ||||
| 6–23 mo | 18 | (45%) | 17 | (49%) | |
| 24–59 mo | 22 | (55%) | 18 | (51%) | |
| Female sex | 16 | (40%) | 16 | (46%) | .79 |
| Median (IQR) d between vaccine series completion and serum collection | 34 | (31–39) | 34 | (31–37) | .99 |
| Previous SARS-CoV-2 infection | 12 | (30%) | 17 | (49%) | .15 |
| Characteristics | mRNA-1273 (n = 40) | BNT162b2 (n = 35) | P | ||
|---|---|---|---|---|---|
| Median (IQR) age, mo | 25 | (13, 44) | 24 | (13, 33) | .41 |
| Age category | .94 | ||||
| 6–23 mo | 18 | (45%) | 17 | (49%) | |
| 24–59 mo | 22 | (55%) | 18 | (51%) | |
| Female sex | 16 | (40%) | 16 | (46%) | .79 |
| Median (IQR) d between vaccine series completion and serum collection | 34 | (31–39) | 34 | (31–37) | .99 |
| Previous SARS-CoV-2 infection | 12 | (30%) | 17 | (49%) | .15 |
Immunogenicity of mRNA-1273 and BNT162b2 Vaccines
First, we directly compared the immunogenicity of mRNA-1273 and BNT162b2 among the study population. All participants had detectable binding antibodies against the Spike proteins for D614G and all 13 variants tested, regardless of the vaccine product received. In addition, we identified no significant differences in serum levels of SARS-CoV-2 spike binding antibodies (Fig 1A and B; Supplemental Fig 4). Similarly, neutralizing antibody titers to D614G and Omicron BA.1 and BA.4/5 did not differ by vaccine type (Fig 1C). These results indicate that primary series vaccination with mRNA-1273 or BNT162b2 elicits similar binding and neutralizing antibody responses in children <5 years of age.
Immunogenicity of mRNA-1273 and BNT162b2 vaccines in children <5 years of age. SARS-CoV-2-specific antibody binding and neutralization were compared by vaccine product (mRNA-1273 recipients: n = 40; BNT162b2 recipients: n = 35). A, The top heatmap displays the geometric means of SARS-CoV-2 variant-specific binding antibody responses as measured by BAMA by vaccine product. Antibody binding is expressed as mean fluorescence intensity (MFI) of serum antibodies specific to SARS-CoV-2 Spike proteins for each variant. The bottom heatmap displays the geometric means of variant-specific antibody neutralization titers (ID50) measured using a pseudovirus neutralization assay by vaccine product. B, Antibody binding of Spike proteins from D614G, Delta, and Omicron variants are expressed as mean fluorescence intensity (MFI) and shown by vaccine product. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. C, Antibody-mediated neutralization of D614G, Omicron BA.1, and Omicron BA.4/5 pseudoviruses, reported as ID50, for each vaccine product. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. No comparisons were significant by Wilcoxon rank-sum tests after correction for multiple testing.
Immunogenicity of mRNA-1273 and BNT162b2 vaccines in children <5 years of age. SARS-CoV-2-specific antibody binding and neutralization were compared by vaccine product (mRNA-1273 recipients: n = 40; BNT162b2 recipients: n = 35). A, The top heatmap displays the geometric means of SARS-CoV-2 variant-specific binding antibody responses as measured by BAMA by vaccine product. Antibody binding is expressed as mean fluorescence intensity (MFI) of serum antibodies specific to SARS-CoV-2 Spike proteins for each variant. The bottom heatmap displays the geometric means of variant-specific antibody neutralization titers (ID50) measured using a pseudovirus neutralization assay by vaccine product. B, Antibody binding of Spike proteins from D614G, Delta, and Omicron variants are expressed as mean fluorescence intensity (MFI) and shown by vaccine product. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. C, Antibody-mediated neutralization of D614G, Omicron BA.1, and Omicron BA.4/5 pseudoviruses, reported as ID50, for each vaccine product. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. No comparisons were significant by Wilcoxon rank-sum tests after correction for multiple testing.
Impact of Previous SARS-CoV-2 Infection and Age on Vaccine-Elicited Antibody Responses
A total of 24 participants reported a previous SARS-CoV-2 infection based on PCR or antigen testing. These infections occurred a median (IQR) of 165 (84–226) days before serum collection, with no difference in the timing of previous infection by vaccine group (mRNA-1273: 142 [37–204] days vs BNT162b2: 165 [149–228] days; P = .41). Given that many of these previous infections occurred when Omicron subvariants were dominant in the region, we hypothesized that participants with a history of previous infection would exhibit stronger antibody binding and neutralization responses to Omicron subvariants compared with participants without a history of previous infection. Children in both vaccine groups with a history of previous infection tended to have greater antibody binding to Omicron BA.1.1, BA.2, and BA.4/5 and greater serum neutralization against D614G, Omicron BA.1, and BA.4/5 compared with children without known previous infection (Fig 2, Supplemental Fig 5, Supplemental Fig 6). Differences in neutralization activity by history of previous infection were observed in both vaccine groups (Wilcoxon rank-sum tests; D614G: mRNA-1273, Padj = .01, BNT262b2: Padj = .04; Omicron BA.1: mRNA-1273 and BNT262b2, Padj < .0001, BNT262b2: Padj < .0001; Omicron BA.4/5: mRNA-1273 and BNT262b2, Padj < .0001; Fig 2C). Antibody binding responses to Omicron BA.1.1 (Padj = .001), BA.2 (Padj < .0001), and BA.4/5 (Padj < .0001) were higher in mRNA-1273 recipients with a history of previous infection compared with those without previous infection, whereas only antibody binding to Omicron BA.4/5 (Padj = .002) differed by previous infection status in BNT262b2 recipients (Fig 2A and B). In analyses limited to children with a history of previous infection, antibody binding responses to Omicron BA.2 (Padj = .003) and neutralization activity against D614G (Padj = .04) were slightly higher in mRNA-1273 recipients than in BNT262b2 recipients. In contrast, no significant differences in antibody binding or neutralization activity by vaccine type were observed among children without a history of previous infection (Fig 2A and B).
Immunogenicity of mRNA-1273 and BNT162b2 vaccines in children <5 years of age by history of before SARS-CoV-2 infection. SARS-CoV-2-specific antibody binding and neutralization activity were compared by vaccine product and previous SARS-CoV-2 infection history (mRNA-1273 recipients without previous infection: n = 28; mRNA-1273 recipients with previous infection: n = 12; BNT162b2 recipients without previous infection: n = 18; BNT162b2 recipients with previous infection: n = 17). A, The top heatmap displays the geometric means of SARS-CoV-2 variant-specific antibody by BAMA within each vaccine product by previous infection status. Antibody binding is expressed as mean fluorescence intensity of serum antibodies specific to SARS-CoV-2 Spike proteins for each variant. The bottom heatmap displays the geometric means of antibody neutralization titers (ID50) by vaccine product for each variant evaluated. B, Antibody binding of Spike proteins from D614G, Delta, and Omicron variants are expressed as mean fluorescence intensity (MFI) for each vaccine product by previous nfection status. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. C, Antibody-mediated neutralization of D614G, Omicron BA.1, and Omicron BA.4/5 pseudoviruses, reported as ID50, for each vaccine product by previous infection status. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. Statistical comparisons were made by using Wilcoxon rank-sum tests with P values adjusted for multiple comparisons (*Padj < .01; **Padj < .001; ***Padj < .0001).
Immunogenicity of mRNA-1273 and BNT162b2 vaccines in children <5 years of age by history of before SARS-CoV-2 infection. SARS-CoV-2-specific antibody binding and neutralization activity were compared by vaccine product and previous SARS-CoV-2 infection history (mRNA-1273 recipients without previous infection: n = 28; mRNA-1273 recipients with previous infection: n = 12; BNT162b2 recipients without previous infection: n = 18; BNT162b2 recipients with previous infection: n = 17). A, The top heatmap displays the geometric means of SARS-CoV-2 variant-specific antibody by BAMA within each vaccine product by previous infection status. Antibody binding is expressed as mean fluorescence intensity of serum antibodies specific to SARS-CoV-2 Spike proteins for each variant. The bottom heatmap displays the geometric means of antibody neutralization titers (ID50) by vaccine product for each variant evaluated. B, Antibody binding of Spike proteins from D614G, Delta, and Omicron variants are expressed as mean fluorescence intensity (MFI) for each vaccine product by previous nfection status. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. C, Antibody-mediated neutralization of D614G, Omicron BA.1, and Omicron BA.4/5 pseudoviruses, reported as ID50, for each vaccine product by previous infection status. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. Statistical comparisons were made by using Wilcoxon rank-sum tests with P values adjusted for multiple comparisons (*Padj < .01; **Padj < .001; ***Padj < .0001).
Early childhood is a period of rapid immune development, leading us to hypothesize that the immunogenicity of mRNA-1273 and BNT262b2 could differ among young children based on age. We therefore compared vaccine-elicited antibody responses in children 6 to 23 months of age to those of children 24 to 59 months of age. Serum binding antibody levels did not differ by age category (Fig 3A and B, Supplemental Fig 7), and serum neutralizing activity was similar among children 6 to 23 and 24 to 59 months of age (Fig 3C).
Immunogenicity of mRNA-1273 and BNT162b2 vaccines in young children by age category. SARS-CoV-2-specific antibody binding and neutralization were compared by vaccine product and age category (mRNA-1273 recipients 6-23 months of age: n = 18; mRNA-1273 recipients 24–59 months of age: n = 22; BNT162b2 recipients 6–23 months of age: n = 17; BNT162b2 recipients 24–59 months of age: n = 18). A, The top heatmap displays the geometric means of SARS-CoV-2 variant-specific antibody by BAMA within each vaccine product by age category. Antibody binding is expressed as mean fluorescence intensity (MFI) of serum antibodies specific to SARS-CoV-2 Spike proteins for each variant. The bottom heatmap displays the geometric means of antibody neutralization titers (ID50) by vaccine product for each variant evaluated. B, Antibody binding of Spike proteins from D614G, Delta, and Omicron variants are expressed as mean fluorescence intensity (MFI) for each vaccine product by age category. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. C, Antibody-mediated neutralization of D614G, Omicron BA.1, and Omicron BA.4/5 pseudoviruses, reported as ID50, for each vaccine product by age category. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. No comparisons were significant by Wilcoxon rank-sum tests after correction for multiple testing.
Immunogenicity of mRNA-1273 and BNT162b2 vaccines in young children by age category. SARS-CoV-2-specific antibody binding and neutralization were compared by vaccine product and age category (mRNA-1273 recipients 6-23 months of age: n = 18; mRNA-1273 recipients 24–59 months of age: n = 22; BNT162b2 recipients 6–23 months of age: n = 17; BNT162b2 recipients 24–59 months of age: n = 18). A, The top heatmap displays the geometric means of SARS-CoV-2 variant-specific antibody by BAMA within each vaccine product by age category. Antibody binding is expressed as mean fluorescence intensity (MFI) of serum antibodies specific to SARS-CoV-2 Spike proteins for each variant. The bottom heatmap displays the geometric means of antibody neutralization titers (ID50) by vaccine product for each variant evaluated. B, Antibody binding of Spike proteins from D614G, Delta, and Omicron variants are expressed as mean fluorescence intensity (MFI) for each vaccine product by age category. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. C, Antibody-mediated neutralization of D614G, Omicron BA.1, and Omicron BA.4/5 pseudoviruses, reported as ID50, for each vaccine product by age category. Each point represents data from an individual participant. The red line indicates the median, and the top and bottom of the box indicate the first and third quartiles. No comparisons were significant by Wilcoxon rank-sum tests after correction for multiple testing.
Interactions of Age, Previous Infection, and Vaccine Product on SARS-CoV-2-Specific Antibodies
Although we did not identify differences in antibody responses by vaccine group, we observed significant differences by previous infection status. Additionally, when evaluating the distribution of antibody responses by age group, we observed a consistent trend of higher binding and neutralizing antibody responses among mRNA-1273 recipients in the 24- to 59-month age group (Fig 3A and B, Supplemental Fig 6). We therefore used generalized linear models to evaluate the extent to which vaccine group, previous SARS-CoV-2 infection, and age category were independently associated with SARS-CoV-2 binding (Table 2; Supplemental Table 5) and neutralizing antibody responses (Table 3). These analyses confirmed that previous infection was associated with higher levels of binding antibodies to Omicron BA.1.1 (Padj < .0001), BA.2 (Padj < .0001), and BA.4/5 (Padj < .0001; Table 2) and higher serum neutralization activity against D614G (Padj < .0001) and Omicron BA.1 (Padj < .0001) and BA.4/5 (Padj < .0001; Table 3). Additionally, receipt of mRNA-1273 was associated with greater neutralization activity against D614G (Padj = .02) than BNT262b2, but not against Omicron BA.1 or BA.4/5. We did not observe any differences in SARS-CoV-2 binding antibody responses by vaccine product. No evidence of interaction was found between vaccine group, previous SARS-CoV-2 infection, and age category.
Generalized Linear Models Investigating Factors That Influence Antibody Binding to Selected SARS-CoV-2 Variants After Vaccination
| Exposure | Adjusted Fold Change in MFI (95% Confidence Interval) | Padj | |
|---|---|---|---|
| Antibody binding to D614G | |||
| Previous infection vs no previous infection | 0.89 | (0.73–1.08) | .54 |
| Receipt of mRNA-1273 vs BNT162b2 | 0.85 | (0.71–1.02) | .54 |
| Age 24–59 mo vs 6–23 mo | 1.13 | (0.94–1.36) | .54 |
| Antibody binding to Omicron BA.1 | |||
| Previous infection vs no previous infection | 0.89 | (0.68–1.15) | .61 |
| Receipt of mRNA-1273 vs BNT162b2 | 0.79 | (0.61–1.02) | .54 |
| Age 24–59 mo vs 6–23 mo | 1.26 | (0.98–1.62) | .54 |
| Antibody binding to Omicron BA.1.1 | |||
| Previous infection vs no previous infection | 1.80 | (1.45–2.25) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 1.37 | (1.10–1.70) | .19 |
| Age 24–59 mo vs 6–23 mo | 1.12 | (0.91–1.39) | .55 |
| Antibody binding to Omicron BA.2 | |||
| Previous infection vs no previous infection | 1.48 | (1.28–1.70) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 1.22 | (1.06–1.40) | .19 |
| Age 24–59 mo vs 6–23 mo | 0.99 | (0.87–1.14) | .97 |
| Antibody binding to Omicron BA.4/5 | |||
| Previous infection vs no previous infection | 2.16 | (1.81–2.58) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 1.23 | (1.04–1.47) | .34 |
| Age 24–59 mo vs 6–23 mo | 1.00 | (0.84–1.19) | .99 |
| Exposure | Adjusted Fold Change in MFI (95% Confidence Interval) | Padj | |
|---|---|---|---|
| Antibody binding to D614G | |||
| Previous infection vs no previous infection | 0.89 | (0.73–1.08) | .54 |
| Receipt of mRNA-1273 vs BNT162b2 | 0.85 | (0.71–1.02) | .54 |
| Age 24–59 mo vs 6–23 mo | 1.13 | (0.94–1.36) | .54 |
| Antibody binding to Omicron BA.1 | |||
| Previous infection vs no previous infection | 0.89 | (0.68–1.15) | .61 |
| Receipt of mRNA-1273 vs BNT162b2 | 0.79 | (0.61–1.02) | .54 |
| Age 24–59 mo vs 6–23 mo | 1.26 | (0.98–1.62) | .54 |
| Antibody binding to Omicron BA.1.1 | |||
| Previous infection vs no previous infection | 1.80 | (1.45–2.25) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 1.37 | (1.10–1.70) | .19 |
| Age 24–59 mo vs 6–23 mo | 1.12 | (0.91–1.39) | .55 |
| Antibody binding to Omicron BA.2 | |||
| Previous infection vs no previous infection | 1.48 | (1.28–1.70) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 1.22 | (1.06–1.40) | .19 |
| Age 24–59 mo vs 6–23 mo | 0.99 | (0.87–1.14) | .97 |
| Antibody binding to Omicron BA.4/5 | |||
| Previous infection vs no previous infection | 2.16 | (1.81–2.58) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 1.23 | (1.04–1.47) | .34 |
| Age 24–59 mo vs 6–23 mo | 1.00 | (0.84–1.19) | .99 |
MFI, mean fluorescence intensity.
Generalized Linear Models Investigating Factors That Influence Serum Neutralization of SARS-CoV-2 Variants After Vaccination
| Exposure | Adjusted Fold Change in ID50 (95% Confidence Interval) | Padj | |
|---|---|---|---|
| D614G | |||
| Previous infection vs no previous infection | 2.87 | (1.89–4.36) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 1.96 | (1.30–2.94) | .02 |
| Age 24–59 mo vs 6–23 mo | 1.36 | (0.91–2.03) | .26 |
| Omicron BA.1 | |||
| Previous infection vs no previous infection | 10.84 | (6.61–17.78) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 0.92 | (0.57–1.50) | .21 |
| Age 24–59 mo vs 6–23 mo | 1.37 | (0.85–2.21) | .08 |
| Omicron BA.4/5 | |||
| Previous infection vs no previous infection | 8.83 | (5.69–13.69) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 0.68 | (0.44–1.04) | .50 |
| Age 24–59 mo vs 6–23 mo | 1.67 | (1.09–2.54) | .52 |
| Exposure | Adjusted Fold Change in ID50 (95% Confidence Interval) | Padj | |
|---|---|---|---|
| D614G | |||
| Previous infection vs no previous infection | 2.87 | (1.89–4.36) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 1.96 | (1.30–2.94) | .02 |
| Age 24–59 mo vs 6–23 mo | 1.36 | (0.91–2.03) | .26 |
| Omicron BA.1 | |||
| Previous infection vs no previous infection | 10.84 | (6.61–17.78) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 0.92 | (0.57–1.50) | .21 |
| Age 24–59 mo vs 6–23 mo | 1.37 | (0.85–2.21) | .08 |
| Omicron BA.4/5 | |||
| Previous infection vs no previous infection | 8.83 | (5.69–13.69) | <.0001 |
| Receipt of mRNA-1273 vs BNT162b2 | 0.68 | (0.44–1.04) | .50 |
| Age 24–59 mo vs 6–23 mo | 1.67 | (1.09–2.54) | .52 |
Discussion
We compared the humoral immune responses of children <5 years of age who completed the primary vaccine series with the original monovalent vaccine formulations of mRNA-1273 or BNT162b2. Despite substantial differences in antigen dosage and administration schedule, we found that these vaccines elicited similar levels of serum SARS-CoV-2-specific binding antibodies and neutralization activity ∼1 month after completion of the primary series. In addition, we found that children with a history of previous infection exhibited greater antibody binding and neutralization activity after vaccination regardless of vaccine product, whereas no differences in antibody responses were observed by age group. Moreover, all vaccine recipients developed robust responses to the original ancestral strain targeted by the vaccine and recent Omicron variants, suggesting that both vaccines induce the broad humoral responses that are necessary to protect from severe disease caused by future variants of concern. This direct comparison of 2 mRNA-based vaccines targeting the Spike protein of SARS-CoV-2 provides important insights into how antigen dose and dosing schedule can influence immune responses and additionally highlights the components of the humoral immune response that may correlate with epidemiologic observations of vaccine effectiveness.
Previous studies comparing the immunogenicity of mRNA-1273 and BNT162b2 among adults have revealed that immunization with mRNA-1273 elicits more robust SARS-CoV-2-specific antibody responses and neutralization activity, with the differences in immunogenicity attributed primarily to differences in antigen dosage and dose timing.10 ,11 ,24 –26 In addition, the authors of several studies observed a higher risk of infection and hospitalization after immunization with BNT162b2 compared with mRNA-1273.10 –13 We identified only subtle differences in the humoral immune responses elicited by mRNA-1273 and BNT162b2 among children <5 years of age. These small differences are of questionable significance for vaccine effectiveness; indeed, 2 recent studies revealed that these vaccines provided similar protection against symptomatic SARS-CoV-2 infection and COVID-19-related emergency department or urgent care encounters in this age group.14 ,15
Immune responses elicited by a combination of infection and vaccination, termed “hybrid immunity,” have been found to enhance both the potency and breadth of SARS-CoV-2-specific immune responses.27 –32 Although we observed that all participants developed serum antibodies against each variant tested, those with a history of previous infection exhibited greater antibody-mediated neutralization of D614G and Omicron BA.1 and BA.4/5 and slightly stronger binding antibody responses to Omicron subvariants. Other studies have similarly revealed that individuals with hybrid immunity were more likely to have greater SARS-CoV-2-specific antibody binding and neutralizing activity compared with unvaccinated individuals with a history of previous infection or vaccinated individuals without a history of infection.4 ,33 Taken together, these findings highlight the additional benefit conferred by COVID-19 vaccination even among children with a history of previous SARS-CoV-2 infection.
One of the key concerns related to the original monovalent mRNA-based COVID-19 vaccines is their ability to induce immune responses against new viral variants because the majority of mutations in new viral variants occur in the Spike protein.34 The authors of several studies have found that recent Omicron subvariants can escape immune responses induced by the original vaccines or previous infection with an ancestral strain, necessitating the development and deployment of updated vaccine products.35 –40 We observed that participants with a previous history of SARS-CoV-2 infection had higher neutralizing antibody titers against Omicron subvariants compared with participants without previous infection, likely because most of these infections occurred at a time when Omicron subvariants were the dominant circulating strains. In contrast to previous studies in adults,35 ,41 ,42 most participants in our study exhibited some degree of neutralizing activity against Omicron subvariants even in the absence of documented previous infection. Previous studies of young children vaccinated with either mRNA-1273 or BNT162b2 similarly suggested that vaccine recipients had broad humoral responses to SARS-CoV-2 variants. Nziza and colleagues found that young children vaccinated with mRNA-1273 developed a stronger antibody response to an Omicron variant than vaccinated adults,43 whereas Bellusci and colleagues found that children <5 years of age developed more cross-neutralizing antibodies to Omicron BQ.1, BQ.1.1, and XBB.1 variants compared with older children.40 Additional studies are needed to fully characterize the mechanisms underlying age-related differences in vaccine-elicited immune responses to SARS-CoV-2 and other pathogens.
Our study has several strengths and limitations. First, this study represents the first direct comparison of the immunogenicity of mRNA-1273 and BNT162b2 in children <5 years of age. Moreover, we specifically measured antibody responses to a broad panel of SARS-CoV-2 variants, providing information on the breadth of humoral immune responses elicited by these vaccines and relevant to the protection that they are likely to provide against future SARS-CoV-2 variants. In addition, our study provides comparative immunogenicity data that bolsters recent epidemiologic evidence suggesting that mRNA-1273 and BNT162b2 have similar effectiveness in young children. Of note, we evaluated immune responses in the peripheral blood and did not evaluate mucosal immune responses, which may contribute to differences in protection against severe disease.44 –46 All participants received the original monovalent vaccine formulations, and the authors of future studies will need to evaluate immune responses to updated mRNA-based COVID-19 vaccine products. In part due to relatively poor uptake of these vaccines among young children, our study sample size was small, and most participants were without comorbidities, potentially limiting generalizability to other populations. Finally, our study was focused on immune responses 1 month after completion of the primary vaccine series; future studies are needed to evaluate antibody waning and the impact of subsequent booster doses on vaccine-elicited antibody responses.47 ,48
Conclusions
We demonstrate that the original monovalent formulations of mRNA-1273 and BNT162b2 induced similar antibody binding and neutralization responses in children <5 years of age despite differences in antigen doses and administration schedule. In addition, our study reveals that both vaccines elicited broad humoral immune responses to SARS-CoV-2, indicating that vaccination of young children is likely to provide some degree of lasting protection despite viral evolution and the emergence of new variants of concern. These findings add to the growing literature revealing that mRNA-based COVID-19 vaccines are highly immunogenic in children and adolescents.
Acknowledgments
We offer sincere gratitude to the children and families who participated in this research. Plasmids encoding Spike proteins for SARS-CoV-2 variants were provided by David Montefiore (Duke University School of Medicine).
Ms Dalapati and Dr Williams collected data, conducted the initial analyses, developed data visualizations, drafted the initial manuscript; Dr Giorgi conducted the initial analyses and developed data visualizations; Dr Hurst conceptualized and designed the study, collected data, drafted the initial manuscript, and coordinated the project; Ms Herbek, Ms Kosman, Ms Pulido, Ms Aquino, Mr Pfeiffer, Mr Rodriguez, and Drs Chen, Rotta, and Turner collected data; Drs Fouda and Permar conceptualized and designed the study, acquired funding, and supervised the execution of the study; Dr Kelly conceptualized and designed the study, acquired funding, supervised the execution of the study, and collected data; and all authors critically reviewed and revised the manuscript, approved the final manuscript as submitted, and agree to be accountable for all aspects of the work.
All data are available in the main text or the Supplemental Information. The statistical files and script used for data analyses are also publicly available.
FUNDING: Funded by the National Institutes of Health (NIH). This research was supported by an NIH grant to Drs Fouda, Kelly, and Permar (R01-AI161008). Drs Kelly and Hurst were supported by National Institutes of Health Career Development Awards (K23-AI135090 to Dr Kelly and K01-AI173398 to Dr Hurst). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
CONFLICT OF INTEREST DISCLOSURES: Dr Permar serves as a consultant for Merck, Pfizer, Moderna, Hookipa, Dynavax, and GSK in the area of cytomegalovirus vaccines and has led sponsored programs with Merck and Moderna. Dr Kelly reports consulting or advisory board fees from Merck & Co, Inc and Invivyd. All other co-authors declare that they have no competing interests.
Comments