Immune responses to SARS-CoV-2 infection and vaccination in children

Abstract

During the three years since SARS-CoV-2 infections were first described a wealth of information has been gathered about viral variants and their changing properties, the disease presentations they elicit and how the many vaccines developed in record time protect from COVID-19 severe disease in different populations. A general theme throughout the pandemic has been the observation that children and young people in general fare well, with mild symptoms during acute infection and full recovery thereafter. It has also become clear that this is not universally true, as some children develop severe COVID-19 hypoxic pneumonia and even succumb to the infection, while another group of children develop a rare but serious multisystem inflammatory syndrome (MIS-C) and some other children experience prolonged illness following acute infection, post-COVID. Here I will discuss some of the findings made to explain these diverse disease manifestations in children and young people infected by SARS-CoV-2. I will also discuss the vaccines developed at record speed and their efficacy in protecting children from disease.

1. Predominantly mild COVID-19 infection in children and young people

Already early during the first wave of SARS-CoV-2 transmission in Wuhan, China a milder disease presentation in children and young people was noted among hospitalized patients with the new illness, COVID-19[1]. As the epidemic spread and became a pandemic it was apparent that children are infected at similar rates as adults, but with much lower rates of hospitalization[2], [3]. With increasing levels of population immunity, from prior infection and vaccination rates of hospitalization have decreased across all ages. During the Omicron peak wave in the US (December 2021-Februrary 2022) hospitalization rates for kids aged 5-11 years was higher (2.8/100,000) than (1.2/100,000) during the preceding Delta-dominated wave (June -December 2021) and 90% of hospitalized children were unvaccinated[4].

There have been many theories brought forward to explain the milder COVID-19 in children as compared to adults[5] such as better overall health, reduced expression of the viral entry receptor ACE2, recent exposures to related common cold coronaviruses (CoV), and viral interferences in the mucosal membranes from other common viruses in children[6]. Several reports have shown differences in the local mucosal immune system of children as compared to adults characterized by primed and more potent type-I IFN responses upon infection[7], [8], [9]. Mucosal responses to SARS-CoV-2 infections in adults are enriched for neutrophil activation pathways when compared to children[10].

The systemic immune response in children infected by SARS-CoV-2 is also characterized by strong IFN-I responses but less symptoms as compared to adults[11], [12]. Intriguingly, the strong innate response of young children was mediated by a distinct blood monocyte population with a particularly activated antiviral response (Interferon-stimulated genes, ISGs) that is not reported in adults[12]. The antibody response in children kicks off a bit slower as compared to adults[13] but persist longer[14] and even continuous to evolve in previously infected young children in a manner not typically seen in adults[12]. The SARS-CoV-2 specific T cell response in children differs in specificity and is enriched for T memory stem cells when compared to T cell responses of their infected mothers[15].

2. Rare but serious cases of severe COVID-19 in children

Early and local mucosal IFN-I responses following SARS-CoV-2 infection is the strongest determinant of subsequent disease severity[16]. This is illustrated by inborn errors of immunity involving viral sensors, signalling mediators and transcription factors triggering interferon-induced genes as well as type-I IFN receptor defects which are all associated with severe acute COVID-19, even among young people[17], [18], [19]. The many reported genetic causes underlying severe COVID-19 pneumonia have been described in detail previously[17]. Some of these defects are also associated with severe Influenza infections in children, such as IRF7 loss of function[20].

A special condition associated with severe COVID-19 is Autoimmune polyglandular syndrome type 1 (APS-1), also called autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED). This immunodeficiency is caused by inborn errors in the AIRE gene[21] leading to impaired central T cell tolerance and autoantibody production, including anti-cytokine antibodies. Most patients with APS-1 carry high-affinity autoantibodies neutralizing IFN-I and other cytokines such as IL-17A, IL-17F[22] as well as tissue specific antigens[23] explaining the clinical presentation with combined autoimmunity as well as infections such as chronic mucocutaneous candidiasis. Reports of severe COVID-19 in children and adults with APS-1[24], [25], [26] emerged already in the spring of 2020. Although most APS-1 patients develop severe COVID-19 pneumonia[25], clinical presentation is not uniform and for example six young female APS-1 patients in Germany had IFN-I autoantibodies but did not develop severe COVID-19 upon SARS-CoV-2 infection[27]. In patients with neutralizing autoantibodies to type-I IFN plasma exchange and/or exogenous treatment with pegylated IFN-I have proved to be a viable treatment[26], [28] even before the development of vaccines and antiviral medications.

Remarkably, when analysing 987 patients with life-threatening COVID-19 pneumonia Bastard and colleagues found that neutralizing autoantibodies to IFN-I were present in 101 of these and were able to prevent the antiviral functions of IFN-I[29]. These patients did not have undiagnosed APS-1, Lupus or other known causes of autoantibodies to type-I IFN and these individuals did not have any history of severe viral infections. The presence of such neutralizing autoantibodies to type I IFN were rare (4/1227) among healthy individuals[29]. In a follow-up study of > 34,000 individuals from the general population the incidence of autoantibodies to type I IFN was found to increase with age, be higher in males than females and account for ~20% of fatal COVID-19 cases overall[30].

3. Multisystem inflammatory syndrome, MIS-C

Another notable disease presentation reported early during the pandemic was a Kawasaki-like illness seemingly associated with SARS-CoV-2 infection in children[31]. Kawasaki disease is a postinfectious vasculitis syndrome of unknown cause triggered by various viral infections resulting in clear seasonal incidence patterns. There are also genetic associations resulting in 10-30x more cases of Kawasaki disease in Northeast Asia including Japan where the first case was reported by pediatrician Tomisaku Kawasaki in 1961[32]. Around 80% of cases of Kawasaki disease occurs in children <4 years of age although all ages can be affected and the disease is now the most common cause of acquired heart failure as rheumatic fevers caused by streptococcal infections are increasingly rare[33].

The Kawasaki-like disease following SARS-CoV-2 infection now termed Multisystem-inflammatory disease associated with COVID-19 (MIS-C), or pediatric inflammatory multisystem syndrome temporally associated with SARS-CoV-2 (PIMS-TS)[34] in the United Kingdom, differs from classic Kawasaki disease in multiple ways. MIS-C occurs in children of all ages[35] and even in some adults[36] with young children being most Kawasaki-like while older children and adults more frequently have myocarditis and cardiogenic shock at presentation. Direct comparisons between clinical signs and laboratory measurements in children with MIS-C, Kawasaki disease prior to the emergence of SARS-Cov-2 show that MIS-C is characterized by more lymphopenia, coagulopathy and capillary leak which are generally not seen to the same extent in Kawasaki disease patients[34]. Another direct comparison of the systems-level immunological state in children with MIS-C and Kawasaki disease revealed distinct cytokine profiles with Kawasaki as an IL-17A driven inflammatory disease and MIS-C more associated with the Neutrophil chemoattractant CXCL8 and innate cell activation[37]. A subgroup of Kawasaki disease patients who presented with cardiogenic shock (Kawasaki shock) resembled MIS-C patients with respect to clinical presentation and laboratory features[34]. Patients with an unrelated, but clinically overlapping condition, Toxic shock syndrome, caused by polyclonal T cell activation via a bacterial superantigen was even more similar to MIS-C[34].

Based on the clinical overlap between toxic shock syndrome and MIS-C, Cheng et al. made a key discovery when performing structural modelling of the SARS-CoV-2 spike protein revealing a possible superantigen motif with structural similarity to the staphylococcal enterotoxin B causing toxic shock syndrome[38]. Indirect evidence for SARS-CoV-2 superantigen-mediated T cell activation has also come from multiple groups in the form of T cell receptor repertoire skewing and enrichment of T cells carrying the Vβ21 (TRBV11-2) chain in combination with variable Vα chains[39], [40], [41], [42], [43]. No follow-up analyses have been performed confirming gaps in the T cell repertoire following MIS-C as would be expected following vigorous superantigen-mediated T cell activation and it has also not been possible to verify superantigen-activity of soluble SARS-CoV-2 spike protein due to difficulties in producing this highly reactive protein in its native form. Additional studies of the full and cleaved Spike protein structure and functional properties are warranted.

Another important breakthrough in understanding MIS-C pathogenesis came from whole exome/genome sequencing analyses of 558 patients within the COVID Human Genetic effort consortium (https://www.covidhge.com/). In these MIS-C patients, Lee et al. identified patients with deleterious mutations in genes OAS1, OAS2 and RNAseL involved in the OAS-RNAseL pathway involved in RNA sensing and degradation[44]. These mutations led to defective negative regulation of the pathway and excessive inflammatory responses in SARS-CoV-2 infected myeloid cells triggered in a MAVS dependent manner[44]. This was particularly clear in monocytes and provides an explanation for innate inflammation in MIS-C which is also a target of effective therapies used to treat MIS-C such as such recombinant IL1RA (Anakinra) and anti-TNFα as well as steroids and intravenous immunoglobulins[45]. Interestingly patients with mutations in the OAS-RNAseL pathway also displayed signs of superantigen-mediated T cell activation with skewed T cell receptor repertoires suggesting a link between the two features of MIS-C pathogenesis, excessive innate cell activation and superantigen-mediated T cell activation. Additional mutations in related pathways of innate cell activation are likely to explain additional MIS-C cases and further investigation is warranted into candidate variants.

The main unresolved aspect of MIS-C pathogenesis is the delayed presentation of 1-2 months following acute SARS-CoV-2 infection. Superantigen-mediated diseases are likely to activate T cells primarily when superantigen abundance is maximal which should be during peak SARS-CoV-2 viremia. Similarly, excessive innate inflammatory responses triggered by SARS-CoV-2 would be likely to occur during acute infection and but most MIS-C patients have mild to asymptomatic acute SARS-CoV-2 infection. One possible explanation to the delayed presentation is that MIS-C is triggered, not in the respiratory epithelium or circulation where SARS-CoV-2 viruses are found during acute infection, but rather that repeated stimulation within a tissue specific niche from a persistent viral reservoir is required. The most likely such tissue specific niche would be the intestine, since most patients have signs of intestinal inflammation[35], [46], elevated markers of intestinal permeability[47], viral shedding is prolonged in stool from children as compared to adults[48], and viral particles have been found in intestinal biopsies in children months following mild COVID-19[49]. Possibly such a persistent intestinal viral reservoir could trigger T cell activation within the large pool of resident T cells in the intestine and associated lymphoid structures and such T cells engage with highly activated myeloid cells in genetically predisposed children such as the ones with OAS-RNAseL mutations ( Fig. 1). If so, this could explain both the location, timing and rapid resolution of inflammation by immunomodulatory treatments that shut down this self-propagating T-cell/monocyte activation loop (Fig. 1). Rebound inflammation rarely occurs in treated MIS-C patients, despite the fact that viral particles are likely to be present or seen again upon reinfection. More analyses of the tissue specific immune responses in MIS-C in the intestine are needed to test this hypothesis.

Fig. 1.

Exuberant inflammatory responses to viruses in MIS-C. Since MIS-C develops 1-2 months following a mild SARS-CoV-2 infection occurring in the respiratory tract, and viral persistence in the intestine has been reported in MIS-C, one possibility is that the disease is caused by a self-propagating activation loop of tissue-resident myeloid and T-cells and triggered by a persistent virus in the intestine and spreading to a systemic inflammatory syndrome in patients with predisposing variants such as defects in OAS- RNAseL pathway. Figure shows Super-Ag mediated activation of T cell carrying Vb21.3 TCR chain independently of an antigen- presenting cell. The outlined hypothesis behind MIS-C is that two phenomena act synergistically in a tissue-specific manner in the intestine if viral particles are present for sufficient time in individuals also carrying predisposing mutations and the Vb21.3 expressing T cells in their intestine.

The incidence of MIS-C has changed dramatically during the pandemic and is currently low. Vaccines and likely immunity from prior infection protects from MIS-C[50] and there are reasons to believe that novel SARS-CoV-2 variants with mutated spike proteins might be less likely to trigger MIS-C[51].

4. Post-COVID syndrome in children

Post-infectious disorders can develop following multiple infections in children and adults[52]. Postinfectious, multiorgan dysfunction involving cognitive impairment, autonomic dysregulation and coagulopathy have been well described in adults following COVID-19[53]. These affect women of reproductive age more than men and often follows a mild to moderate COVID-19 and can be a progressive disease associated with chronic immune activation[54] and detectable spike protein in circulation many months after initial SARS-CoV-2 infection[55].

In children the diagnosis of postCOVID syndrome has been controversial with highly variable results across studies and inconsistent case definitions used[56]. Many studies also report persistent symptoms in children without comparing to a control group of children. The CLoCK consortium in the United Kingdom performed a matched cohort study of 1658 children and adolescents (age 11-17 years) testing positive for SARS-CoV-2 infection from January to March 2021 and compared self-reported symptoms in these to 1737 matched children without any positive SARS-CoV-2 test. This study found several symptoms more commonly in SARS-CoV-2 positive children such as abdominal pain, diarrhoea, chest pain and confusion all decreasing in frequency at 3- and 6-months following infection[57]. Similarly, a Danish population study of 35,277 children (0-17 years) with a positive SARS-CoV-2 test and 78,037 control children showed symptom resolution within 5 months following infection, although observation periods in this study was not fully matched[58]. The impression from reported studies is that post-COVID symptoms in children and young people are less frequent as compared to adults, in particular when comparing to adult women of reproductive age. Also, the severity and duration of post-COVID symptoms in children is seemingly lower as compared to adults, despite that fact that individual cases of very severe post-COVID disease have been reported in some teenagers[59]. It is curious that post-COVID in adults is associated with viral persistence in blood[55], while MIS-C in children is associated with viral persistence in the intestine[47], suggesting that tissue specific differences or age-associated differences in immune responses to persistent SARS-CoV-2 viruses may explain these different presentations.

5. A unifying theory to explain diverse disease presentations in children

To explain the predominantly mild SARS-CoV-2 infections seen in children and young people as compared to the elderly, and the seemingly increased propensity for viral persistence leading to prolonged shedding in stool, abdominal symptoms and in rare cases development of MIS-C, I recently proposed a model based on life history theory[60]. According to this model, children with their efficient mucosal IFN-I responses will prioritize limited resources from systemic immune responses in favour of structural growth and induce disease tolerance when deemed safe ( Fig. 2)[48]. This could explain reduced systemic inflammatory responses and limited symptoms during acute infection in children, and the conversely elevated risk of MIS-C development in genetically predisposed children[40], [44]. Also, in adults infected with SARS-CoV-2 is disease severity determined by the timing and robustness of IFN-I responses [61], [62] and in some individuals with milder COVID-19 disease, peripheral innate cells display attenuated inflammatory responses indicating a certain level of disease tolerance [61].

Fig. 2.

An energy allocation theory to explain diverse clinical outcomes in children and adults infected by SARS-CoV-2. Differing investments in growth, reproduction and immune responses among children and adult males and females could explain the likelihood of mounting energy costly systemic inflammatory responses upon infection and conversely the likelihood of viral persistence in the young associated with MIS-C.

Given that females of reproductive age will also invest significantly more of its limited resources into reproduction as compared to males, the same model could also explain stronger systemic inflammatory responses and cytokine storms leading to increased COVID-19 mortality in males vs. females[63] while females, like children, develop milder symptoms but also are at higher risks of viral persistence and associated post-COVID syndromes[48].

6. SARS-CoV-2 vaccine responses in children

The SARS-CoV-2 mRNA vaccines (BNT162b2) and (mRNA-1273) are approved by European Medicines Agency (EMEA) and US Food and Drugs administration (FDA) for use in children from 6 months of age. These vaccines offer excellent protection from severe disease in children[64], [65], [66] and adults[67], [68] and also protect children from MIS-C[50], [69]. At a comparable dose (100μg), children aged 6-11 years elicit higher titre antibody responses to the mRNA-1273 vaccine as compared to adults and with stronger Fc binding capacity[70]. T cell responses to SARS-CoV-2 vaccines in adults are stronger as compared to the responses induced by infection[71], but comparisons between children and adults are lacking. History of prior infection imprints on T cell responses to SARS-CoV-2 vaccination[72], which might have an impact on children as the last population to receive COVID-19 vaccines during the pandemic leading many to have been previously infected prior to vaccination.