This study investigated humoral and cellular immunity against SARS-CoV-2 in children and adolescents with natural or hybrid immunity.
In most children having experienced mild or even asymptomatic SARS-CoV-2 infections NCP antibodies persist for nearly one year after infection, although titer height is highly variable. Several studies in adults show a decline in NCP antibody titers beginning 4–6 months after the onset of symptoms with insignificant titers ten months after infection [16–19], a finding in line with one other pediatric cohort [20]. A small reduction in median titer of NCP antibodies was observed over the course of three months in our cohort, confirming previous findings at least in part. In some children eight years or younger we could not detect NCP antibodies, a phenomenon we did not see in older children; otherwise, the NCP antibody titers did not correlate with the age of the participants. As expected, NCP antibody titers were not affected by vaccination.
S antibodies were detectable in all children, with titers being 50-fold higher in vaccinated children. S antibodies could also be detected in younger non-vaccinated children who were tested negative for NCP antibodies as well as for Interferon-γ production in T cells. In asymptomatic or mild wild-type infections in adults seronegativity is described in 1–9% of patients [21–24]. We could not confirm these results, but this may be due to the relatively small sample size.
The increasing S and NCP antibody titers observed in some non-vaccinated children may be attributed to boosting by clinically inapparent SARS-CoV-2 reinfections.
All participants with a hybrid immunity had neutralizing antibodies in contrast to only 50% of those with a natural immunity following an infection in the Omicron era. Moreover, we found significantly increased neutralization capacities in adolescents with a hybrid immunity, which is not surprising as this pattern is already described in adults [25–27] and resembles a heterologous prime-boost-boost regimen [28]. Whether this higher neutralization capacity in adolescents is an effect of age or vaccination cannot be answered due to the small sample size and the different vaccination rates in the different age groups. In another small study non-vaccinated convalescent children had lower mean nAB titers than non-vaccinated convalescent adults. Adults with hybrid immunity had much higher titers than adults with a natural immunity, making effects of both vaccination and age possible [29].
We did not elicit from our data the sequence of infection and vaccination, but at least in adults this is not relevant for the resulting immune response [27] [30].
The QFN-Test was validated [31] in patients infected with SARS-CoV-2 variant B.1.1.7 (Alpha) and its usability was also shown in the Omicron era [29]. More than half of the non-vaccinated convalescent children had a non-reactive interferon-γ release assay, but some exhibited a robust cellular immune response. This finding is somewhat contradictory to the results of other studies, but data on this immune reaction following an Omicron infection are sparse. In one small study group convalescent unvaccinated pediatric and adult participants were described as totally IGRA negative [29] 10 months after infection whereas in another small cohort of unvaccinated adult following infection with Omicron these were mostly positive for T-cell-mediated immunity while being seronegative [6].
Vaccinated adolescents with a hybrid immunity following an Omicron infection had a higher probability of having a reactive IGRA. Again, if this effect is only caused by vaccination or is also dependent on the age of the probands cannot be deduced from this study. An at least partial vaccination effect is probable, as a significantly higher percentage of reactive IGRAs in adults with a hybrid immunity compared to those with a natural immunity is described [29] [32].
Then again, an age dependency with a lower frequency of SARS-CoV-2 specific T cells in children than in adults was also seen in small pediatric study groups infected with the ancestral strain utilizing flow cytometry to detect antigen-specific T cells [20, 33, 34]. In contrast to our study, specific memory T cells were detectable in these cohorts one year after a mild or asymptomatic SARS-CoV-2 infection. The difference to our results is probably attributable to a different method used in these studies to trace memory T cells, as flow cytometry-based assays are shown to have a greater sensitivity in detecting SARS-CoV-2 specific T cell responses [35] [36] than IGRAs. Furthermore, IGRA levels are known to decrease over time [37]. The phenomenon of an age-dependent diminished sensitivity is also described in interferon-γ release assays for the detection of Mycobacterium tuberculosis in children [38] [39]. This raises the question whether younger children are less likely to have SARS-CoV-2 specific memory T cells following infection or whether the IGRA test has weaknesses in monitoring cellular immunity in younger children. As whole blood is used for the QuantiFERON interferon-γ release assay, containing other immune factors such as memory T cells, this phenomenon may be attributable to the fact that younger children have a higher percentage of naïve T cells in peripheral blood [34] therefore interfering with bystander activation of memory CD8+ T cells by cytokine stimulation [40], which is an important aspect of immune responses to pathogens [41]. If this constraint may be overcome by adapting cut-off values in children without impairing specificity needs further studies.
A small influence of the severity of prior SARS-CoV-2 infections cannot be ruled out, as 38% of the younger children with asymptomatic infection had a reactive IGRA versus 48% of the younger children after a symptomatic SARS-CoV-2 infection. But again, the study is not sufficiently powered to control so many variables simultaneously.
Ten months after acquiring natural or hybrid immunity the different compartments of immune memory show different reactions with complex relationships between different aspects of immune memory. Nucleocapsid antibodies were, irrespective of titer, not qualitatively and quantitatively predictive of memory T cells [22] determined by interferon-γ release as described before. The presence of nAB correlates with cellular immunity, as shown in one previous study [42]. Detectable Spike protein antibodies were conjoined with interferon-γ release by T cells when these titers were extremely high; titers of this magnitude were only found in adolescents with a hybrid immunity.
Our findings support the concept of vaccination of convalescent children to further enhance the immune response. Whether this boosted immunological memory results in fewer or less symptomatic SARS-CoV-2 infections with other VOCs is a question not answered by this study.