Kinetics and Durability of Antibody and T-Cell Responses to SARS-CoV-2 in Children

Abstract

Background

The kinetics and durability of T-cell responses to SARS-CoV-2 in children are not well characterized. We studied a cohort of children aged 6 months to 20 years with COVID-19 in whom peripheral blood mononuclear cells and sera were archived at approximately 1, 6, and 12 months after symptom onset.

Methods

We compared antibody responses (n = 85) and T-cell responses (n = 30) to nucleocapsid (N) and spike (S) glycoprotein over time across 4 age strata: 6 months to 5 years and 5–9, 10–14, and 15–20 years.

Results

N-specific antibody responses declined over time, becoming undetectable in 26 (81%) of 32 children by approximately 1 year postinfection. Functional breadth of anti-N CD4+ T-cell responses also declined over time and were positively correlated with N-antibody responses (Pearson r = .31, P = .008). CD4+ T-cell responses to S displayed greater functional breadth than N in unvaccinated children and, with neutralization titers, were stable over time and similar across age strata. Functional profiles of CD4+ T-cell responses against S were not significantly modulated by vaccination.

Conclusions

Our data reveal durable age-independent T-cell immunity to SARS-CoV-2 structural proteins in children over time following COVID-19 infection as well as S-antibody responses in comparison with declining antibody responses to N.

SARS-CoV-2 causes COVID-19, which was the leading cause of death by an infectious disease in 2022 [1]. Children aged <18 years generally experience milder disease as compared with adults [2, 3]. Despite this, some studies show that children have similar viral loads in the upper respiratory tract when compared with adults [4–6]. As of June 2023, 13% of children <5 years old in the United States had received at least 1 dose of a COVID-19 vaccine, despite recommendations for all children ≥6 months old to be vaccinated [7]. Unvaccinated children may contribute to ongoing household and community transmission [8].

Several lines of evidence demonstrate that neutralizing antibodies are immune correlates of protection against SARS-CoV-2 infection in adults and are boosted by vaccination [9–11]. However, a coordinated cellular immune response is critical for support of humoral immunity, resolution of disease, and clearance of infection [12]. Depletion of CD8+ T cells in rhesus macaques led to increased viral titers in the setting of rechallenge after resolution of natural infection with SARS-CoV-2 [13]. In mice, adoptive transfer of SARS-CoV-2–primed T cells contributed to a reduction in pulmonary viral burden [14]. Human studies show an association between T-cell phenotypes and improved clinical outcomes after COVID-19 [15].

Despite this evidence, the development of SARS-CoV-2–induced cellular immunity in children following infection is not well characterized. Some reports indicate that children may have more robust innate immune defenses to SARS-CoV-2 when compared with adults, leading to a reduction in the ensuing adaptive immune response [16–18]. However, other studies note that there is no difference in innate immune response [19] and that T-cell responses in children are less robust than adults [20]. Several studies comparing children with adults following infection or vaccination show significantly reduced CD4+ and CD8+ T-cell responses in children after stimulation with SARS-CoV-2 proteins [19–21]. In contrast, a recent study reports nearly 2-fold increases in IFN-γ responses to spike (S) protein in children aged 3 to 11 years as compared with adults [22].

A limitation of many studies is their cross-sectional design and lack of longitudinal analysis of CD4+ T-cell responses across age strata.

This longitudinal cohort study, initiated in the beginning of the COVID-19 pandemic, recruited children as young as 6 months to characterize age-dependent kinetics of cellular and humoral immunity largely in the absence of vaccination. We also assessed immunity following vaccination in a subset of these children. These data can inform our understanding of the adaptive immune response to SARS-CoV-2 across the pediatric age spectrum.

METHODS

Participants

Children and young adults up to 20 years of age, including those with underlying medical conditions, were recruited from the Seattle area between April 2020 and May 2021 via social media, announcements in clinics and testing sites, and recruitment from symptomatic households or hospitalized patients. Children were included in the study if they had a positive SARS-CoV-2 polymerase chain reaction (PCR) test result, had at least 1 household contact with a positive PCR test result, or were antinucleocapsid IgG positive on clinical testing or during evaluation for multisystem inflammatory syndrome in children. Children with this syndrome were excluded from participation. Participants were considered infected and eligible for the study if they had a positive PCR test result prior to enrollment or a positive SARS-CoV-2 antibody test result at enrollment. Because PCR testing was not widely available for children at the time that this study was initiated, the date of presumed infection onset was defined as the first of the following, in order of priority: (1) date of symptom onset, (2) date of positive PCR test result if no history of acute COVID-19 symptoms, or (3) date of positive household member PCR test result if no PCR test for the child was available. No participants were vaccinated against COVID-19 at the time of enrollment. Parental interview and electronic medical record extraction were used to collect demographics, illness history, viral testing results, intensive care unit admission, and underlying medical conditions. COVID-19 vaccination data were extracted from the Washington State Immunization Information System or the medical record for those who lived outside of Washington. Study data were managed in REDCap.

Ethics

Informed consent was obtained from parents or participants if aged >18 years, and assent was obtained from all participants aged >7 years. This study proposal was reviewed and approved by the Seattle Children's Hospital institutional review board. This activity was also reviewed by the Centers for Disease Control and Prevention and conducted consistent with its policy and applicable federal law (eg, 45 CFR part 46, 21 CFR part 56; 42 USC §241[d]; 5 USC §552a; 44 USC §3501 et seq).

SARS-CoV-2 IgG Assays

All enrollment serum samples (Supplementary Methods) were tested for antibodies against SARS-CoV-2 nucleocapsid (N) protein by a chemiluminescent microparticle immunoassay with the Abbott Architect system between June 2020 and August 2022 and the Abbott Alinity system thereafter (Abbott). The N index was assessed, and values >1.4 were considered a positive result based on manufacturer's instructions. Enrollment samples with sufficient volume were tested for anti-S IgG antibodies with the SARS-CoV-2 IgG II Abbott Alinity chemiluminescent assay. Anti-S IgG antibodies were measured as arbitrary units per milliliter (AU/mL), which were transformed to binding antibody units per milliliter (BAU/mL) by the conversion factor 1 BAU/mL = 0.142 AU/mL [23, 24], for which values ≥50 AU/mL were considered positive based on the manufacturer's instructions [25, 26]. All follow-up samples from children who were infected were tested for anti-N IgG, and all follow-up samples with sufficient volume were tested for anti-S IgG as described previously.

Neutralization Assay

Neutralization assays were performed on available samples from patients who contributed >1 blood sample by utilizing SARS-CoV-2 D614G S-pseudotyped lentiviral particles based on Wuhan-Hu-1 at the University of Washington Virology Lab as previously described [27]. Details are provided in the Supplementary Methods. Results are reported as 50% neutralizing dilution titers.

Flow Cytometry Analysis

Intracellular cytokine staining was performed as previously described following stimulation of peripheral blood mononuclear cells with S or N peptide pools [28]. Details are provided in the Supplementary Methods. The gating tree is shown in Supplementary Figure 6. Results were reported as T-cell responses to SARS-CoV-2 S or N peptide pools with the background signal subtracted (DMSO). Combinatorial polyfunctionality analysis of antigen-specific T cells (COMPASS) was performed to identify subsets of T cells for which there is a high probability of antigen-specific activation and cytokine response following stimulation (n = 30) [28–30]. Functional subsets identified by COMPASS were added to the gating set for further visualization and analysis, the frequencies of which are expressed as a percentage of total CD4+ T cells. Additionally, functionality scores, a summary measure of COMPASS-defined responses [29], were calculated for each stimulated sample.

Statistics

Statistics are detailed in the Supplementary Methods.

RESULTS

A Longitudinal Pediatric Convalescent COVID-19 Cohort

We enrolled 85 children and young adults with presumed SARS-CoV-2 infection between April 2020 and May 2021 (Table 1). The median age at enrollment was 10.7 years (IQR, 5.4–15.4); 34 participants (40%) were female at birth. Altogether, 64 (75%) were positive for SARS-CoV-2 based on quantitative reverse transcription PCR testing and SARS CoV-2 IgG anti-N detection in their enrollment blood sample; 10 (12%) were positive for SARS-CoV-2 via quantitative reverse transcription PCR but had neither SARS-CoV-2 anti-S or anti-N IgG detected in their enrollment blood sample; and 11 (13%) did not have positive PCR results but had SARS-CoV-2 anti-N antibody detected in their enrollment blood sample (Supplementary Figure 1).

Of 85 children with documented SARS-COV-2 infection, 77 (91%) were symptomatic at enrollment; 33 (39%) were treated as inpatients, including 8 in the intensive care unit. Fourteen children (16%) had at least 1 underlying medical condition, and 4 were mildly to moderately immunocompromised according to their prescribed drug regimens (Supplementary Table 1). Of the 85 children, 55 (65%) completed their enrollment and 6-month blood draw (blood collected at 2 time points), and 36 (42%) completed their 6- and 12-month blood draw (blood collected at all 3 time points; Supplementary Figure 1). Altogether, 35% had no follow-up beyond enrollment, and no child had a 12-month blood draw without a 6-month blood draw. Blood samples were obtained a median 31 (IQR, 18–47), 186 (IQR, 178–198), and 371 (IQR, 363–384) days after the presumed infection date.

At study initiation in April 2020, no COVID-19 vaccines were available; however, vaccines became available for children aged >5 years during the study period. Nineteen (22%) children received at least 1 vaccine dose prior to the last blood draw (Table 1). Patients were predominantly vaccinated with mRNA vaccines manufactured by Pfizer and Moderna (BNT162b2 and mRNA1273, respectively), and 1 participant aged >18 years received Janssen Ad26.COV2.S (Supplementary Table 1).

Decay of Antibody Responses Targeting N Over 12 Months

Serum samples were assessed for antibody responses to SARS-CoV-2 S and N proteins (Figure 1A). Of 85 samples collected at enrollment, 74 (87%) were positive for N antibodies but these decayed rapidly: specifically, 62% and 81% of those with follow-up reverted to seronegative by their respective 6- and 12-month blood draws (Figure 1B). During acute infection, there was a significant association between N index and age among outpatients (Supplementary Figure 2A). Outpatient children <10 years old tended to exhibit a higher N index than older children (Supplementary Figure 2B). When stratified by hospitalization status, inpatient children <5 years old produced significantly lower anti-N IgG titers than their outpatient counterparts (Supplementary Figure 2C). There were no differences in mean N index values across age strata at 6 to 8 months after the presumed infection date (Figure 1C). Five participants, all aged >10 years, were observed to have measurable anti-N IgG antibodies at their 12-month blood draw after having negative test results 6 months earlier, suggesting that these children may have been exposed to SARS-CoV-2 a second time (Figure 1B). Of these 5 children, none had underlying medical conditions, and 4 had received at least 1 dose of a COVID-19 vaccine prior to their last blood draw (Supplementary Table 1).

Decay of T-Cell Responses Targeting N Over 12 Months

Archived peripheral blood mononuclear cells (n = 27 at enrollment [approximately 2-month time point], n = 27 at 6-month time point, and n = 28 at 12-month time point) were assessed for T-cell responses to N (Figure 1A). We determined the magnitude and functionality of N-specific CD4+ T cells with intracellular cytokine staining and COMPASS [29]. COMPASS identified 8 N-specific CD4+ T-cell subsets (Supplementary Figure 3A–C). The proportion of N-specific CD154+ IFN-γ+ IL-2+ TNF-α+ T cells was significantly decreased at 12 months as compared with the acute period (Figure 2A). Functionality scores were significantly decreased in the 5- to 9-year and 15- to 20-year age strata by 12 months of follow-up, but there were no significant differences detected between age strata at either time point (Supplementary Figure 3D and 3E). N-specific functionality scores were positively associated with the level of anti-N antibodies (r = 0.31, P = .008; Figure 2B). Inpatients had a wide range of functionality scores at the 0- to 2-month blood draw and trended higher than outpatients (Figure 2C). Taken together, these data reveal that the functional breadth of CD4+ T-cell responses to N were low by 12 months of follow-up, with no significant differences across age strata.

Age Correlates With Anti-S Antibody in Outpatients

Among outpatients, increasing age was negatively correlated with anti-S antibodies (Figure 3A). Children <5 years old produced significantly higher anti-S antibody titers than participants >10 years old; however, differences in neutralizing titers were not significant after correcting for multiple comparisons (Figure 3B). Inpatient children <5 years old produced significantly fewer anti-S IgG and neutralizing antibody titers than outpatient children. Interestingly, the opposite trend was observed for older participants aged 15 to 20 years: inpatients produced significantly higher titers of anti-S IgG, but there was no difference in neutralizing titers (Figure 3C).

Vaccination Enhances Anti-S Binding Antibody, Neutralizing Antibody, and Polyfunctional CD4+ T-Cell Responses to S

Of the 30 children in the T-cell response analysis, 15 (50%) received at least 1 dose of COVID-19 vaccine during the study (5–9 years, n = 4; 10–14 years, n = 5; 15–20 years, n = 6; Table 1, Supplementary Table 1). Anti-S antibodies and neutralizing titers were significantly increased following vaccination (P < .001 and P = .002, respectively; Figure 4A and 4B). Anti-S and neutralizing antibody titers trended upward or remained stable over 1 year (Supplementary Figure 4A) and were significantly higher in vaccinated participants than unvaccinated participants (P < .001 and P =.005; Supplementary Figure 4B). Conversely, there was little change in S-specific functionality scores following 1 or 2 vaccine doses across age strata, and as expected, N-specific functionality scores were not enhanced following vaccination (Figure 4C). The frequency of polyfunctional T cells postvaccination was not significantly changed when compared with prevaccination visits (Figure 4D). The magnitude of functional subsets did not differ by age (Supplementary Figure 4C and 4D).

All 3 vaccinated children with underlying conditions and longitudinal data demonstrated sustained anti-S antibody titers 6 to 8 months after infection. Of these, antibody titers dropped below the positivity threshold at 12 months in 1 participant. This individual entered inpatient treatment and failed to produce substantial anti-N antibodies (Supplementary Table 1).

These data reveal that while mRNA vaccination significantly boosts antibody responses in children previously infected, it may not lead to significant increases in T-cell breadth or functionality in response to S glycoprotein beyond that established by the prior infection.

Polyfunctional T-Cell Responses to S and Neutralizing Antibodies Remain Stable up to 12 Months Postinfection in the Absence of Vaccination

Finally, we sought to characterize immune responses to S in the absence of vaccination. Among children who did not receive a SARS-CoV-2 vaccine, anti-S antibodies were detected in 57 (95%) of 60 children early after infection, and titers remained relatively stable up to 12 months postinfection (Figure 5A). Neutralizing antibodies were detected in 11 (92%) of 12 children tested 12 months postinfection (Figure 5B). When stratified by age, we found no difference in anti-S antibodies at 6 months postinfection (Figure 5C); however, neutralization titers were significantly higher among children <5 years old when compared with the 15- to 20-year age strata 6 to 8 months following infection (Figure 5D).

Because most participating children aged 5 to 14 years were vaccinated, these age groups were removed for the following analysis, and only unvaccinated children aged <5 years and 15 to 20 years were compared. COMPASS identified 10 S-specific CD4+ T-cell subsets among children in the absence of vaccination (Supplementary Figure 5A and 5B). There were no significant differences in magnitude of functional subsets between participants aged 15 to 20 years and children <5 years (Supplementary Figure 5C). No significant differences in functionality score were observed between age groups at the 0- to 2-month and 12-month blood draws (Figure 6A), and functionality scores overall remained stable over time (Figure 6B). Notably, S-specific functionality scores were significantly higher than N at all time points (Figure 6C). Among unvaccinated children, S-specific functionality scores were not positively correlated with anti-S antibodies or neutralizing titers (Figure 6D). These data reveal that polyfunctional T-cell responses to S glycoprotein may not vary significantly by age or time since SARS-CoV-2 infection in the absence of vaccination.

DISCUSSION

The kinetics and durability of humoral and cell-mediated immunity to SARS-CoV-2 in children have not been well established. By following children ranging from 6 months to 20 years of age, we were able to profile T-cell responses, neutralizing antibodies, and binding antibody responses to SARS-CoV-2 structural proteins over 12 months of follow-up. We demonstrate the stability of anti-S antibodies and T-cell responses in children over time and across age strata, as well as the decay in anti-N antibodies and T-cell responses. We also show that even young children can produce long-lasting immune responses to SARS-CoV-2 that persist in the absence of vaccination.

In this study, most participants were identified by a positive PCR detection test result from a nasal sample, but anti-N IgG antibodies were used as a surrogate of infection for some participants who did not have molecular testing performed. Nearly all children demonstrated seroreversion of anti-N IgG antibody. The Abbott SARS-CoV-2 anti-N chemiluminescent assay used in this study is a standard single antigen–binding sandwich immunoassay, and while a useful tool to assist in diagnosis of recent COVID-19 infection, it is thought to primarily detect low-affinity antibodies, which peak and decline rapidly as the humoral response matures [26, 30]. With the decrease in N-specific T-cell functionality scores, these data suggest that humoral and cellular immunity to SARS-CoV-2 N protein wanes quickly in children. Interestingly, children who were suspected of having been reinfected exhibited N-specific functionality scores on the lower end of the spectrum at the 12-month blood draw. In contrast, we observed that children as young as 6 months developed robust immune memory to SARS-CoV-2 S protein, as demonstrated by the longevity of anti-S CD4+ T-cell responses, neutralizing antibody titers, and anti-S binding antibodies in unvaccinated participants. All children had markedly enhanced CD4+ responses to S as compared with N proteins. It remains to be seen how S-specific CD4+ T cells protect against variants of concern, but some studies show CD4+ T-cell cross-reactivity in children against other coronaviruses and variants of concern [22, 31].

The immune systems of children undergo rapid development over time, differing significantly among infants, toddlers, and adolescents [32]. For example, when exposed to the superantigen staphylococcal enterotoxin B, T cells in the peripheral blood of children display less activation and polyfunctionality [33]. Similarly, polyfunctional CD4+ T cells were reported to be markedly reduced in response to upper airway infection in children aged <3 years [34]. In adults, enhanced CD4+ polyfunctionality with an effector memory phenotype was observed following COVID-19 vaccination, typically associated with durable vaccine-induced immunity [35], and was associated with recovery from severe COVID-19 [36]. However, humoral responses in children may be more durable than in adults [37]. Our previous work suggested that children produce lower titers of SARS-CoV-2–specific antibodies soon after infection as compared with adults but that neutralizing titers were similar by 6 months postinfection [38]. Here, we report that among children with mild disease, very young children (<5 years old) produced significantly greater anti-S IgG than children >15 years old and similarly aged children treated inpatient, suggesting that young children may rely heavily on a robust humoral response to protect from severe disease. Notably, we did not observe a difference in T-cell polyfunctional profiles across age strata in our study.

During the early phase of the pandemic, testing was largely limited to hospital settings; thus, our cohort includes a relatively large proportion of patients who were hospitalized. This limited our ability to make generalizable conclusions about the majority of pediatric patients with COVID-19 but provided the opportunity to explore immunologic relationships with disease severity in children. There are few reports detailing the immune kinetics in children across the spectrum of disease, but children with severe COVID-19 may have less robust early IgG responses [39, 40]. In this study, severity of disease in young children was marked by a depressed humoral response during acute infection. Though rare, severe COVID-19 in children is highly associated with underlying conditions [41], but none were noted for most children in this cohort. Patients with underlying medical conditions are part of a vulnerable subset of the pediatric population that would benefit most from vaccination, and their inclusion provides important information showing that development of protective immune responses to SARS-CoV-2 is not completely hindered by immunosuppressive therapies. Pediatric transplant recipients demonstrate delayed and less robust humoral responses when compared with healthy controls, but improved CD4+ T-cell responses and S-specific antibodies can be induced with repeated doses [42, 43].

While our study reveals important insights into CD4+ T cells in children after SARS-CoV-2 infection, we were limited by the overall sample size, especially for comparisons across age strata. It is possible that unidentified reinfections could have influenced data from later time points. Additionally, CD8+ T cells demonstrated minimal increases in cytokine production following stimulation with N or S peptide pools (data not shown). This may reflect the length of the peptides used for stimulation (13- and 17-mers), which are more suited for presentation to CD4+ T cells. Other studies have shown that CD4+ T-cell responses are more prominent than CD8+ T-cell responses [44]. Finally, we were unable to profile antigens beyond N and S, including those that may account for age-associated immune responses due to differences in antigen processing as hypothesized by Cohen et al [20].

Children represent an important link in the chain of transmission of respiratory viruses [45], including SARS-CoV-2 [8]. Our data reveal surprisingly robust T-cell and neutralizing antibody activity against S glycoprotein up to 12 months postinfection even in the absence of vaccination. Consistent with prior studies, in children with prior COVID-19, vaccination boosts the humoral immune response, as demonstrated by the enhanced anti-S binding and neutralizing antibody titers, without having a substantial impact on the breadth or longevity of the cellular response [46]. Overall, our data suggest that infection-induced immunity may provide some level of protection against COVID-19 disease in children across the pediatric age spectrum but that very young children may benefit from repeat vaccinations to protect against severe disease.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.

Notes

Acknowledgments. We thank the children and their families at Seattle Children's Hospital for participation in this study and Research Laboratory Services at Seattle Children's Hospital. We thank Jesse Bloom, PhD, for assistance and advice regarding SARS-CoV-2 neutralization assays. We acknowledge Erik D. Layton for technical assistance with T-cell assays. We also gratefully acknowledge the contributions of all our study participants and their families to this study.

Author contributions. C. S. and J. A. E. designed the study. M. A. F. performed the experiments and analyzed the data. L. K., K. L., and A. A. contributed to curation of clinical data and analysis. A. G., J. A. D., L. G., and A. W. performed neutralization assays. C. M. M., M. B. H., K. P., and T. F. facilitated data collection and provided a critical review of the manuscript. C. S., M. A. F., and J. A. E. wrote the manuscript with contributions from all authors.

Disclaimer. The findings and conclusions of this report are those of the authors and do not represent the official position of the Centers for Disease Control and Prevention.

Financial support. This work was supported by the Centers for Disease Control and Prevention (BAA 75D301-20-R-67897); the Seattle Children’s Hospital Foundation; the US National Institutes of Health (R01-AI142670); and the National Science Foundation (DGE-1762114).

References

Author notes