Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2025 Sep 11.
Published in final edited form as: Vaccine. 2025 Jul 5;62:127461. doi: 10.1016/j.vaccine.2025.127461

Humoral, T cell and immune gene expression responses to SARS-CoV-2 vaccination in a small group of children with previous MIS-C

Timothy F Spracklen a,b,#, Jonathan Day a,#, Hamza Van Der Ross a,#, Claire Butters a,#, Ntombi Benede c,d, Avril Walters c,d, Rubina Bunjun c, Thandeka Moyo-Gwete e,f, Mashudu Madzivhandila e,f, Simon C Mendelsohn c,g, Thomas J Scriba g, Muki Shey c,h,i, Wendy A Burgers c,d,I,j, Penny L Moore e,f,k, Liesl J Zühlke a,b,l, Roanne S Keeton c,d, Kate Webb a,m,#
PMCID: PMC7618102  EMSID: EMS208396  PMID: 40618623

Abstract

The effects of SARS-CoV-2 vaccination in children with previous multisystem inflammatory syndrome (MIS-C) are not well understood. In this study, we aimed to assess immune responses to SARS-CoV-2 vaccination in children over the age of 12 years with previous MIS-C and compare them to healthy children. Three children with previous MIS-C and four healthy children received two doses of the BNT162b2 vaccine. Blood was collected before the first dose, one week after the first dose, one week after the second dose and three weeks after the second dose. All participants had detectable SARS-CoV-2 spike IgG before vaccination. Spike binding and neutralising antibody activity increased after the first vaccine dose with no differences between children with a history of MIS-C and healthy children. Serum inflammatory cytokines and whole blood immune gene profiles did not resemble acute MIS-C and there were no differences in these between the two groups at any timepoint. All participants gained a robust SARS-CoV-2-specific T cell response by three weeks after the second dose. A transient increase in SARS-CoV-2-specific CD4 T cells expressing TCR Vβ21.3, a non-specific T cell subset previously found to be enriched in patients with MIS-C, was demonstrated in two of the children with previous MIS-C but none of the healthy children. Together, these data demonstrate that vaccination is effective at boosting SARS-CoV-2-specific immune responses in a small group of children with previous MIS-C, and that it does not induce inflammatory cytokine or gene expression responses resembling acute MIS-C. Although larger-scale studies are needed to confirm these findings, the present evidence supports SARS-CoV-2 vaccination in children with previous MIS-C.

Introduction

Multisystem inflammatory syndrome in children (MIS-C) is a severe immune hyperactivation syndrome associated with SARS-CoV-2 infection in children. MIS-C presents in children two to six weeks after exposure to SARS-CoV-2 with elevated temperatures, rash, conjunctivitis and variable organ involvement including myocarditis and shock.

The exact immune sequence that causes MIS-C is not fully understood, although it is characterised by markedly raised inflammatory markers, a unique inflammatory gene expression signature and significantly elevated inflammatory cytokines [1–8]. Children with acute MIS-C have evidence of a maturing immunoglobulin (Ig)G response to SARS-CoV-2, but no differences in the titres or function of SARS-CoV-2 antibodies compared to children exposed to SARS-CoV-2 who do not develop MIS-C, arguing against an antibody-mediated pathogenesis [1, 9]. An over-expression of non-specific T cells expressing TCR Vβ21.3 has also been reported in patients with MIS-C, which may indicate a non-specific superantigen response [10]. Recent evidence suggests that cross-reactivity between SARS-CoV-2 and human SNX8 protein may underly development of MIS-C [11].

In population studies, SARS-CoV-2 vaccination has been shown to be protective against MIS-C [12–15]. In addition to immunological benefits of SARS-CoV-2 vaccination, there is evidence that this vaccination is safe and well tolerated in children with previous MIS-C [16–20]. The Centers for Disease Control and Prevention (CDC) therefore recommends vaccination of children with previous MIS-C after full clinical recovery and at least 90 days after diagnosis of MIS-C [21]. There are currently no in-depth data describing the inflammatory responses to re-exposure to SARS-CoV-2 antigen through vaccination in children with previous MIS-C. In this study, we aimed to assess the immunogenicity and the inflammatory cytokine, T cell and gene expression responses of SARS-CoV-2 vaccination in a small group of children with previous MIS-C and healthy children.

Methods

Patient cohort

Children with previous MIS-C between the ages of 12 and 18 years were offered vaccination on the State vaccination programme, which was only offering vaccination to children over 12 at the time. Other inclusion criteria included no prior SARS-CoV-2 vaccination and previous enrolment in an MIS-C study carried out from June 2020 to March 2022 at the Red Cross War Memorial Children’s Hospital in Cape Town, South Africa [1, 22]. All patients fulfilled the World Health Organisation criteria for MIS-C [23] at the time of their diagnosis and were healthy at the start of this study. Controls were healthy unvaccinated children enrolled in the abovementioned MIS-C study or referred by other participants. Previous COVID-19 was not recorded for the healthy control children as prior SARS-CoV-2 exposure was not a criterion for exclusion. No children had any underlying chronic infectious or immune diseases.

Ethical approval was obtained from the University of Cape Town Human Research Ethics Committee (HREC 599/2020) and informed consent was collected from all participants. Unvaccinated patients with acute MIS-C (n = 30) and contemporaneous unvaccinated healthy controls (n = 74) were included as comparator groups in some of the assays below (serum cytokine and chemokine analysis and gene expression analysis).

Vaccination schedule and sampling

Children received two doses of the Pfizer-BioNTech COMIRNATY® SARS-CoV-2 mRNA vaccine (BNT162b2) three weeks apart. Longitudinal serum, plasma, whole blood and PAXgene® blood RNA samples were collected before the first dose, one week after the first dose, one week after the second dose and three weeks after the second dose.

SARS-CoV-2 antibody enzyme-linked immunosorbent assay (ELISA)

Serum from all participants at all four timepoints was assessed to measure anti-SARS-CoV-2 spike IgG antibodies. Briefly, 2 µg/ml of SARS-CoV-2 spike protein was added to each well in a 96-well plate to coat the plate and was incubated overnight at 4°C. After washing, plates were blocked by incubating in blocking buffer (5% skimmed milk, 0.05% Tween 20 and 1X PBS). Samples were diluted to 1:100 and added to plates after they had been washed. The secondary antibody (anti-IgG) was then added at a dilution of 1:3000 in blocking buffer. TMB substrate was added and after 5 minutes, the reaction was stopped using 1M H2SO4. The absorbance was measured at a wavelength of 450 nm.

SARS-CoV-2 antibody neutralisation assay

Serum from all participants at all four timepoints was assessed for neutralisation activity. Briefly, HEK293T cells were co-transfected with SARS-CoV-2 spike (D614G) plasmid and a firefly luciferase lentivirus backbone plasmid to create SARS-CoV-2 pseudotyped lentiviruses. Serum samples were heat-inactivated and incubated with the SARS-CoV-2 pseudotyped lentiviruses for 1 hour at 37°C and 5% CO2. HEK293T cells over-expressing ACE-2 were added and incubated for 72 hours. Luminescence was measured and ID50 values calculated.

Serum cytokine and chemokine analysis

Serum from all participants at all four timepoints was assessed. A customised MILLIPLEX® kit was used to determine the serum concentration of interleukin (IL)-1 receptor antagonist (RA), IL-1β, IL-10, IL-6, IL-27, monocyte chemoattractant protein (MCP)-1, interferon gamma-induced protein (IP)-10 and tumour necrosis factor (TNF) using a Bio-Plex platform as per manufacturer’s instructions. Briefly, antibody immobilised beads were added to wells containing standards, controls and samples for 2 hours at room temperature. After washing, the detection antibody was added and incubated for 1 hour, whereafter Streptavidin-Phycoerythrin was added for 30 minutes in the dark. The plate was washed and sheath fluid was added to each well, after which the plate was read using a Luminex plate reader. Fold changes were calculated using the geometric means of all participants at the different timepoints.

RNA extraction and gene expression analysis

Gene expression analysis of 96 individual transcripts was measured from RNA derived from whole blood, using qRT-PCR as previously described [24]. Transcripts included housekeeping genes and other genes with broad relevance to inflammation, immunopathology, immune regulation and type I interferon responses. RNA samples were available for each participant and at each timepoint. Nine healthy control individuals were also included on the chip. Data were combined with our previously described pre-treatment acute MIS-C (n = 30) and healthy (n = 54) gene expression data after adjusting for batch effect in R v4.3.1 with the batchtma package and using the quantreg method. The data were used to assess changes in expression of our previously described 29-gene signature for MIS-C [24], although two of these genes, MMP9 and TLR8, were removed from the analysis due to high failure rates of the probes (67.9% and 42.9% failure rates, respectively).

T cell flow cytometry

The whole blood assay sample processing used for this study was adapted from a whole blood intracellular cytokine detection assay designed to detect TB-specific T cells in children [25]. Briefly, whole blood was collected in sodium heparin tubes and stimulated with 1 µg/ml SARS-CoV-2 spike peptide pools of 15mer peptides overlapping by 11 amino acids (Miltenyi Biotec) in the presence of costimulatory antibodies against CD28 (clone 28.2) and CD49d (clone L25) (1 μg/mL each; BD Biosciences) and Brefeldin A (10 μg/mL, Sigma-Aldrich). Unstimulated blood was incubated with costimulatory antibodies, Brefeldin A and an equimolar amount of DMSO as a background control. After 24 hours, blood was treated with EDTA (2 mM) for 15 minutes followed by red blood cell lysis and cell fixation using FACS lysing solution (BD Biosciences) for 10 minutes. Cells were then cryopreserved in freezing media (90% fetal bovine serum and 10% DMSO) and stored at −80°C until batched analysis.

Fixed, cryopreserved cells were thawed, washed and surface stained with CD4 ECD (SFCI12T4D11; Beckman Coulter), CD8 BV510 (RPA-8; BioLegend) and TCR Vβ21.3 FITC (REA894; Miltenyi Biotec). Cells were then permeabilised and intracellularly stained with CD3 BV650 (OKT3; BioLegend) and IFN-γ AlexaFluor®700 (B27; BioLegend). Samples were acquired on a BD LSR-II Fortessa flow cytometer (BD Biosciences) using Diva software (v9) and analysed using FlowJo (v10). Cells were gated on singlets, lymphocytes and CD3 cells. Results are expressed as the frequency of CD4 or CD8 T cells expressing IFN-γ or the frequency of SARS-CoV-2-specific CD4 T cells expressing TCR Vβ21.3. Cytokine responses are presented as background subtracted values (from the frequency of cytokine produced by unstimulated cells).

Statistical analysis

Analyses of cytokines and gene expression were conducted using R. Pairwise differences were assessed for each timepoint using two-tailed Mann Whitney U tests, corrected for multiple comparisons using the Holm-Bonferroni method. The pheatmap package was used to perform hierarchical clustering of the gene expression data with centering and scaling of variables. Prior to principal component analysis (PCA), the missForest package was used for random-forest-based imputation of missing delta Ct values with 100 trees in each forest [26]. PCA was performed using the base R prcomp function with zero centering and scaling of variables and visualised with the ggbiplot package.

Statistical analysis of T cell responses was conducted in GraphPad Prism. Due to the small sample size, statistical analysis was limited to a nonparametric Mann Whitney U test to detect differences in SARS-CoV-2-specific CD4 T cell responses between healthy and MIS-C participants at the final timepoint at which all participants had SARS-CoV-2-specific CD4 T cell responses, and a paired Wilcoxon signed rank analysis of the total TCR Vβ21.3 expressing CD4 and CD8 T cells at the third-last timepoint compared to the pre-vaccination timepoint.

Results

Vaccination cohort

Of the 66 patients with a history of MIS-C who were previously enrolled, 15 children were over the age of 12 years. Of these, 10 were successfully contacted and 3 agreed to be enrolled in the study. Of the 110 healthy controls who were previously enrolled, 20 were over the age of 12 years. Two consented to be enrolled, and an additional two children over 12 years were referred by one of the participants. These seven children were vaccinated between July and September 2022 (Figure 1). All participants were vaccinated between the ages of 12 and 13 years (Supplementary Table S1).

Figure 1.

Figure 1

Open in a new tab

Study design for vaccination of children with previous MIS-C and healthy controls. Participants M1, M2 and M3 were children with previous MIS-C, while H1, H2, H3 and H4 were healthy participants.

Of the three previous MIS-C patients, one was diagnosed with MIS-C during the first epidemiological SARS-CoV-2 wave in South Africa, while the other two were diagnosed in the second SARS-CoV-2 wave dominated by the Beta variant (Supplementary Table S2). Both patients recruited in the second wave required multiple intensive care unit (ICU) stays during their hospital admission. All patients clinically recovered and were discharged within 10-11 days. The clinical courses of their diseases are summarised in Supplementary Table S2. All patients presented with tachycardia, and all had fever, lung disease and gastrointestinal involvement at some time during their admission. All patients had full resolution of all symptoms on follow up and were clinically well at time of vaccination.

Vaccine-induced SARS-CoV-2 spike-antibody titres are not affected by a history of MIS-C

Humoral responses to SARS-CoV-2 vaccination were determined by measuring SARS-CoV-2 spike IgG antibody levels and neutralising abilities in all participants at all timepoints. Before vaccination, all seven participants had positive (OD450 > 0.4) SARS-CoV-2 spike-antibody ELISA values (Figure 2A), indicating prior exposure to SARS-CoV-2. Following the first vaccine dose, there was a noticeable increase (1.26-fold) in the SARS-CoV-2 spike-antibody responses in both groups, with no difference between the two groups at any timepoint. The responses remained the same after the first dose, with no detectable boost after the second dose and no appreciable waning in antibody titres (1.26-fold difference between pre-vaccine titres and the final timepoint). After the second dose, the assay reached saturation, and the data were inconclusive as to whether the second dose increased antibody titres or showed waning by six weeks.

Figure 2. Anti-SARS-CoV-2 antibodies in serum of vaccine recipients.

Figure 2

Open in a new tab

(A) Anti-spike IgG antibody levels measured at a dilution of 1:100 in the serum of participants with a history of MIS-C (red) and healthy participants (blue) before receiving a dose of the SARS-CoV-2 vaccine, one week after the first dose of the vaccine, one week after the second dose of the vaccine and three weeks after the second dose of the vaccine. (B) Neutralisation ID50 values of anti-spike protein antibodies in the participants across the vaccination schedule. Note that some individual values may be very similar: for individual datapoints please refer to Supplementary Table S3.

The fully titrated neutralising ability of the SARS-CoV-2 spike-antibodies increased 37-fold in both cohorts following the first vaccine dose with no difference between children with a history of MIS-C and healthy children at any of the four timepoints (Figure 2B). There was no evidence of waning neutralising ability, but there was a 99-fold increase in neutralising ability following the second SARS-CoV-2 vaccine dose and a 106-fold increase at the final timepoint.

Cytokine production was not increased with vaccination in children with previous MIS-C

At all four timepoints, there was no difference in the serum concentration of IL-1β, IL-1RA, IL-6, IL-27, IP-10, MCP-1 or TNF between children with a history of MIS-C and healthy children (Figure 3). At the pre-vaccine and one week post first dose sampling timepoints, children with a history of MIS-C had IL-10 serum concentrations that appeared higher compared to healthy controls (median concentration in children with previous MIS-C: 6.6 pg/ml; in healthy controls: 2.21 pg/ml; p = 0.057). However, these IL-10 concentrations were much lower than during acute MIS-C (median: 131.24 pg/ml). They were comparable to IL-10 serum concentrations from samples taken two weeks after MIS-C discharge (median: 5.43 pg/ml) and were within the normal range for the total cohort of healthy children (Supplementary Figure S1). Similarly, IL-1RA appeared slightly raised in children with a history of MIS-C at the second timepoint (median concentration in children with previous MIS-C: 13.53 pg/ml; in healthy controls: 8.19 pg/ml), although this was not significant (p = 0.057) and was lower than the acute phase of disease (median: 493.9 pg/ml). The serum concentrations of all measured cytokines were not affected by vaccination.

Figure 3. Serum cytokine concentration in participants who have a history of MIS-C (red) and healthy participants (blue) throughout SARS-CoV-2 vaccination schedule.

Figure 3

Open in a new tab

Gene expression profiling reveals no differences in vaccine responses between children with previous MIS-C and healthy children

Hierarchical clustering of gene transcripts, supervised by type and visit, did not indicate any clear distinctions in gene expression between the groups (Figure 4). The most noticeable differences occurred after the first dose, with CASP1 (p = 0.057) and IL27 (p = 0.057) up-regulated in children with previous MIS-C compared to vaccinated healthy children and TUBGCP6 (p = 0.057) down-regulated in these patients at this timepoint (Supplementary Figure S2). Expression of these transcripts was not significantly different when compared to 74 unvaccinated healthy children. There were no significant differences in gene expression between the groups at baseline or at the final follow-up three weeks after the second dose.

Figure 4.

Figure 4

Open in a new tab

Hierarchical clustering of expression of gene transcripts in whole blood. Samples are grouped by type (vaccinated MIS-C patients and vaccinated healthy controls) and longitudinally (four vaccine visits corresponding to pre-vaccination, first dose, second dose and final follow-up). Clustering is supervised by timepoint and group.

When comparing to acute MIS-C, we observed a significant distinction between unvaccinated acute MIS-C and unvaccinated healthy controls (Supplementary Figure S3). Gene expression from participants who received the SARS-CoV-2 vaccine clustered almost entirely with the healthy control participants at all timepoints regardless of previous MIS-C (Supplementary Figure S3). Further, PCA of all transcripts revealed separate clusters of acute MIS-C and healthy controls, with vaccinated participants clustering with healthy individuals (Supplementary Figure S4).

T cell Vβ21.3 expression and SARS-CoV-2 specific responses

T cell analysis was performed on all participants’ samples at all timepoints, except for one participant (M2) for whom T cell analysis could only be performed at the final two timepoints (Figure 5). Prior to vaccination only 1/4 (25%) of the healthy children and no children with a history of MIS-C had a detectable SARS-CoV-2-specific CD4 T cell response (Figure 5B). After the first SARS-CoV-2 vaccination, 4/4 healthy children had a detectable SARS-CoV-2-specific CD4 T cell IFN-γ response, which was maintained at one week post second vaccination and three weeks after the second dose (Figure 5B). Of the children with a history of MIS-C, one child (M3) gained a T cell response following the first vaccination which was boosted one week post the second vaccination (median: 0.013% vs 0.055% after the first and second vaccination respectively). The second child with a history of MIS-C (M1) gained a response after two vaccination doses at the final timepoint, while the third child for which we had no pre-vaccination data (M2), had a SARS-CoV-2-specific CD4 T cell IFN-γ response at both the one-week post second vaccination and final timepoints. There were no significant differences in the magnitude of the SARS-CoV-2-specific CD4 T cell IFN-γ responses between the healthy children and those with a history of MIS-C, with all children gaining robust CD4 T cell responses by three weeks after the second dose (Figure 5B). Fewer participants had SARS-CoV-2-specific CD8 T cell IFN-γ responses (Figure 5E and F). One of the healthy children had a detectable CD8 T cell IFN-γ response after the first vaccination and two had a response after the second vaccination, while only one child with a history of MIS-C developed a SARS-CoV-2-specific CD8 T cell IFN-γ response. This was detectable at one week after the second vaccination and the final timepoint (Figure 5E and F).

Figure 5. SARS-CoV-2-specific CD4 T cell responses in vaccine recipients.

Figure 5

Open in a new tab

(A) Representative flow cytometry plots showing IFN-γ production from whole blood stimulated with spike peptides or the unstimulated control. (B) The frequency of SARS-CoV-2-specific CD4+IFN-γ+ T cells in participants with a history of MIS-C (red), or healthy participants (blue) before receiving a dose of the SARS-CoV-2 vaccine, one week after the first dose of the vaccine, one week after the second dose of the vaccine and three weeks after the second dose of the vaccine. (C) Overlay flow plot examples of TCR Vβ21.3 profile in SARS-CoV-2-specific CD4 T cells in a healthy (left) or MIS-C (right) child. (D) The percentage of SARS-CoV-2-specific CD4 T cells expressing TCR Vβ21.3 across the vaccination schedule. A minimum of 15 CD4+IFN-γ+ events were required for phenotyping. (E) Representative CD4 T cell flow cytometry plots showing IFN-γ production. (F) The frequency of SARS-CoV-2-specific CD8+IFN-γ+ T cells. One of the participants who previously had MIS-C only had samples available from after the second dose. Note that some individual values may be very similar: for individual datapoints please refer to Supplementary Table S3.

We next examined CD4 and CD8 T cell expression of TCR Vβ21.3, due to its reported expansion in patients with MIS-C [10]. There was no difference in the frequency of total CD4 or CD8 T cells expressing TCR Vβ21.3 between the groups at any time timepoint (Supplementary Figure S5). There was also no significant change in the frequency of CD4 or CD8 Vβ21.3+ T cells between the pre-vaccination timepoint and a week after the second dose, either for children with previous MIS-C or for healthy children. We next characterised the SARS-CoV-2-specific CD4 T cells and there was a transient increase in the proportion of SARS-CoV-2-specific CD4 T cells expressing TCR Vβ21.3 in children with a history of MIS-C relative to vaccinated healthy children. In participant M3, 38.9% of SARS-CoV-2-specific T cells expressed TCR Vβ21.3 at the one-week timepoint which decreased to 17.5% and 9.5% at the one-week post second vaccine and six-week timepoints respectively. Although there was a lack of pre-vaccination data, participant M2 also had high proportion of TCR Vβ21.3 expression (43.8%) one week after the second vaccine dose, decreasing to 29.7% at the six-week timepoint. In contrast, TCR Vβ21.3 expression on the SARS-CoV-2-specific T cells in healthy children remained below 10% at all timepoints (Figure 5C and D).

Discussion

MIS-C is a poorly understood acute immune response to SARS-CoV-2 exposure. Although SARS-CoV-2 vaccination has been shown to be safe in children with previous MIS-C [16–20], the effect of antigen re-exposure on cytokine expression, gene expression and T cell responses is not known. In this study we confirm that SARS-CoV-2 vaccination in children with previous MIS-C results in robust SARS-CoV-2-specific antibody and T cell responses, with comparable viral neutralisation capacity to vaccinated children with no history of MIS-C. These immune responses were observed following the first SARS-CoV-2 vaccine dose and were maintained through the second dose and to the end of the follow-up period, three weeks after the second dose. SARS-CoV-2 vaccination did not induce inflammatory cytokine or gene expression responses and did not result in an increased proportion of CD4 or CD8 TCR Vβ21.3+ T cells in children with previous MIS-C compared to those without. Intriguingly however, this small series showed a transient increase in the proportion of SARS-CoV-2-specific T cells expressing TCR Vβ21.3 in children with a history of MIS-C relative to vaccinated healthy children.

Despite initial concerns that SARS-CoV-2 vaccination could cause MIS-C recurrence, and case reports of MIS-C onset after vaccination [27–37], larger-scale cohort studies and pharmacovigilance studies have shown this to be a very rare occurrence indeed [15, 38], more likely due to breakthrough infection after SARS-CoV-2 exposure than an immunological response to the vaccine itself. We showed a boost in spike IgG antibodies and neutralisation ability in patients with and without previous MIS-C shortly after vaccination and an average of 18 days after the vaccine course was concluded. These findings are in accordance with studies from other regions in the USA and Europe. A prospective cohort study of 20 Polish children with prior MIS-C demonstrated the immunological benefit of the Pfizer-BioNTech mRNA vaccine, with children mounting effective spike IgG antibodies a median of 23 days after completing the vaccine course [39]. There was also no difference in IgG responses between these children and 34 healthy children vaccinated at the same time [39]. Vaccination boosted spike IgG antibodies for up to three months in a series of seven children with previous MIS-C [40], with similar neutralising ability and cytotoxicity as in a cohort of vaccinated healthy children. What appeared to influence neutralising antibodies more was the infecting SARS-CoV-2 variant, with neutralising titres significantly reduced for the Omicron variant (B.1.1.529) compared to Wild-type [40]. In two small case series totalling five children with previous MIS-C, SARS-CoV-2 vaccination led to elevation of IgG antibodies to levels higher than in acute MIS-C [41, 42].

Through longitudinal profiling of participants’ immunological status, we provide further evidence of the overall tolerance of children to vaccination. As a hyperinflammatory condition, MIS-C is characterised by elevation of numerous inflammatory cytokines and markers [1–8]. In our setting, we previously found elevated IL-10 and IL-6 to be specific to MIS-C compared to other clinically similar acute inflammatory diseases [1]. In the present study, inflammatory cytokine levels were low in both healthy children and those with previous MIS-C at the start of the vaccination programme, with no observed clinical effect of vaccination on the production of any of the tested cytokines. Notably, IL-10 appeared elevated in children with prior MIS-C at two timepoints, including before vaccination, although the observed levels were within the normal range and much lower than during acute MIS-C [1]. Similarly, gene expression profiling revealed no differences between the experimental groups, and transcript-level vaccine responses were virtually indistinguishable between patients with prior MIS-C and healthy children. This is in contrast to the widespread levels of gene dysregulation we have described in acute MIS-C [24].

All of the children, regardless of previous MIS-C, gained a SARS-CoV-2-specific CD4 T cell response during the course of the study. Despite positive anti-spike titres in all participants, it is notable that only one, a healthy child, had detectable T cell responses before vaccination. We have previously shown lack of T cell responses in 17% of seropositive children in our setting [43]. Here, most children mounted a T cell response after the first dose, with one of the children with previous MIS-C only mounting a T cell response after the second vaccination. Nevertheless, at the last timepoint, there was no difference in the magnitude of T cell responses between the two groups. Polyclonal TCR Vβ21.3 T cell expansion is characteristic of MIS-C and not observed in COVID-19 or other similar inflammatory conditions [10, 44], and suggestive of a superantigen response to SARS-CoV-2 in MIS-C. Interestingly, a transient increase of the proportion of SARS-CoV-2-responding CD4 T cells expressing TCR Vβ21.3 was observed in two of the children with previous MIS-C (66.7%) after the first vaccine dose, while this was not observed in any of the healthy children. While a superantigen-like polyclonal TCR Vβ21.3 T cell expansion induced by spike peptide stimulation could possibly contain not only SARS-CoV-2-specific T cells, these IFN-γ producing T cells do represent a subset of activated T cells within which to analyse TCR Vβ21.3 expression. Studies have shown an increased proportion of activated T cells expressing Vβ21.3 during MIS-C disease [10, 45].

This study was limited by the small sample size, which makes it underpowered to detect significant differences between the experimental groups and to draw definitive conclusions. The exclusion of participants younger than 12 years was due to South African regulations at the time of study design which had not yet authorised paediatric vaccinations at ages younger than this. Another limitation is that data concerning safety and adverse events were not collected as this was beyond the scope of the investigation. Nevertheless, our use of in-depth profiling provides some of the first data on inflammatory responses to SARS-CoV-2 vaccination in children with and without previous MIS-C.

A reduction in MIS-C incidence has been observed across the globe and is thought to be due, at least in part, to SARS-CoV-2 vaccination [46]. Nationwide cohort studies have demonstrated that vaccination reduces the risk of MIS-C in children and adolescents [12–14, 47–52]. We still do not fully understand the underlying causes of this disease. These data describing the immune responses upon SARS-CoV-2 vaccination add to the global understanding of this enigmatic disease. Although larger-scale cohort studies and randomised clinical trials are needed to confirm these findings, the present evidence further supports SARS-CoV-2 vaccination in children with previous MIS-C.

Supplementary Material

Supplementary Figure S1

Funding

This work was supported by the South African Medical Research Council (96863). This work was also supported by the Crick African Network (CAN). This fellowship funded Dr Webb’s salary and study consumables. The CAN receives its funding from the UK’s Global Challenges Research Fund (MR/P028071/1), and by the Francis Crick Institute which receives its core funding from Cancer Research UK (FC1001647), the UK Medical Research Council (FC1001647), and the Wellcome Trust (FC1001647). This work was also funded by a grant from the Wellcome Centre for Infectious Disease Research in Africa (CIDRI) through a rapid COVID grant to Prof Liesl Zühlke and Dr Kate Webb which allowed for the funding of members of the team salaries and study consumables, software and hardware. This research was supported by South African Medical Research Council with funds received from National Treasury. The content and findings reported/illustrated are the sole deduction, view and responsibility of the researcher and do not reflect the official position and sentiments of the SAMRC or National Treasury. The funders did not play a role in the data collection, analysis, or interpretation; trial design; patient recruitment; or any aspect pertinent to the study. Penny Moore is supported by the South African Research Chairs Initiative of the Department of Science and Innovation, South Africa and National Research Foundation of South Africa (NRF 9834), the SA-MRC SHIP programme and the Center for the AIDS Programme of Research in South Africa (CAPRISA). Thandeka Moyo-Gwete is supported by a South African Medical Research Council SIR grant. Wendy Burgers is supported by the EDCTP2 program of the European Union’s Horizon 2020 programme (TMA2016SF-1535-CaTCH-22), the Wellcome Trust (226137/Z/22/Z) and the EU-Africa Concerted Action on SARS-CoV-2 Virus Variant and Immunological Surveillance (COVICIS), funded through the EU’s Horizon Europe Research and Innovation Programme (101046041). Rubina Bunjun is supported by the Fondation Botnar under the EDCTP2 programme (TMA2020CDF-3187-ACROSS) supported by the European Union, and the Carnegie Corporation of New York DEAL3 program. Timothy Spracklen is supported by the DSI/NRF Innovation (PDP) postdoctoral fellowship. Liesl Zühlke is funded by the SAMRC through its Division of Research Capacity Development under the Mid-Career Scientist Program from funding received from the South African National Treasury. The content hereof is the sole responsibility of the authors and do not necessarily represent the official views of the SA-MRC. Liesl Zühlke also receives support from the National Research Foundation of South Africa (NRF), as well as the UK Medical Research Council (MRC) and the UK Department for International Development (DFID) under the MRC/DFID Concordat agreement, via the African Research Leader Award (MR/S005242/1).

Footnotes

Author contributions

Kate Webb conceived and designed the study. Claire Butters, Timothy F Spracklen, Ntombi Benede, Roanne S Keeton and Kate Webb interpreted the results and prepared the first draft of the manuscript. Jonathan Day and Hamza van der Ross recruited participants and arranged vaccinations. Antibody binding and neutralisation data were generated by Thandeka Moyo-Gwete, Mashudu Madzivhandila and Penny Moore. Cytokine and chemokine data were generated and analysed by Claire Butters and Muki Shey. RNA expression data were generated and analysed by Timothy F Spracklen, Simon C Mendelsohn, Thomas J Scriba and Kate Webb. T cell flow cytometry data were generated and interpreted by Ntombi Benede, Avril Walters, Rubina Bunjun, Wendy A Burgers and Roanne S Keeton. Institutional and intellectual support was provided by Liesl J Zühlke, Roanne S Keeton and Kate Webb. All authors have read and approved the manuscript for submission. All authors attest they meet the ICMJE criteria for authorship.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  • [1].Butters C, Benede N, Moyo-Gwete T, et al. Comparing the immune abnormalities in MIS-C to healthy children and those with inflammatory disease reveals distinct inflammatory cytokine production and a monofunctional T cell response. Clin Immunol. 2024;259:109877. doi: 10.1016/j.clim.2023.109877. [DOI] [PubMed] [Google Scholar]
  • [2].Consiglio CR, Cotugno N, Sardh F, et al. The Immunology of Multisystem Inflammatory Syndrome in Children with COVID-19. Cell. 2020;183:968–981.:e7. doi: 10.1016/j.cell.2020.09.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [3].Gruber CN, Patel RS, Trachtman R, et al. Mapping Systemic Inflammation and Antibody Responses in Multisystem Inflammatory Syndrome in Children (MIS-C) Cell. 2020;183:982–995.:e14. doi: 10.1016/j.cell.2020.09.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [4].Esteve-Sole A, Anton J, Pino-Ramirez RM, et al. Similarities and differences between the immunopathogenesis of COVID-19-related pediatric multisystem inflammatory syndrome and Kawasaki disease. J Clin Invest. 2021;131 doi: 10.1172/JCI144554. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [5].Lee PY, Day-Lewis M, Henderson LA, et al. Distinct clinical and immunological features of SARS-CoV-2-induced multisystem inflammatory syndrome in children. J Clin Invest. 2020;130:5942–5950. doi: 10.1172/JCI141113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [6].Sacco K, Castagnoli R, Vakkilainen S, et al. Immunopathological signatures in multisystem inflammatory syndrome in children and pediatric COVID-19. Nat Med. 2022;28:1050–1062. doi: 10.1038/s41591-022-01724-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [7].Ramaswamy A, Brodsky NN, Sumida TS, et al. Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory syndrome in children. Immunity. 2021;54:1083–1095.:e7. doi: 10.1016/j.immuni.2021.04.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [8].Hoste L, Roels L, Naesens L, et al. TIM3+ TRBV11-2 T cells and IFNγ signature in patrolling monocytes and CD16+ NK cells delineate MIS-C. J Exp Med. 2022;219 doi: 10.1084/jem.20211381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [9].Weisberg SP, Connors TJ, Zhu Y, et al. Distinct antibody responses to SARS-CoV-2 in children and adults across the COVID-19 clinical spectrum. Nat Immunol. 2021;22:25–31. doi: 10.1038/s41590-020-00826-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].Moreews M, Le Gouge K, Khaldi-Plassart S, et al. Polyclonal expansion of TCR Vbeta 21.3(+) CD4(+) and CD8(+) T cells is a hallmark of Multisystem Inflammatory Syndrome in Children. Sci Immunol. 2021;6 doi: 10.1126/sciimmunol.abh1516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [11].Bodansky A, Mettelman RC, Sabatino JJ, et al. Molecular mimicry in multisystem inflammatory syndrome in children. Nature. 2024 doi: 10.1038/s41586-024-07722-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [12].Levy M, Recher M, Hubert H, et al. Multisystem Inflammatory Syndrome in Children by COVID-19 Vaccination Status of Adolescents in France. Jama. 2022;327:281–283. doi: 10.1001/jama.2021.23262. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [13].Zambrano LD, Newhams MM, Olson SM, et al. BNT162b2 mRNA Vaccination Against Coronavirus Disease 2019 is Associated With a Decreased Likelihood of Multisystem Inflammatory Syndrome in Children Aged 5-18 Years-United States, July 2021 - April 2022. Clin Infect Dis. 2023;76:e90–e100. doi: 10.1093/cid/ciac637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [14].Holm M, Espenhain L, Glenthøj J, et al. Risk and Phenotype of Multisystem Inflammatory Syndrome in Vaccinated and Unvaccinated Danish Children Before and During the Omicron Wave. JAMA Pediatr. 2022;176:821–823. doi: 10.1001/jamapediatrics.2022.2206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [15].Ouldali N, Bagheri H, Salvo F, et al. Hyper inflammatory syndrome following COVID-19 mRNA vaccine in children: A national post-authorization pharmacovigilance study. Lancet Reg Health Eur. 2022;17:100393. doi: 10.1016/j.lanepe.2022.100393. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [16].Yousaf AR, Kunkel A, Abrams JY, et al. COVID-19 Vaccine Reactogenicity and Vaccine Attitudes Among Children and Parents/Guardians After Multisystem Inflammatory Syndrome in Children or COVID-19 Hospitalization: September 2021-May 2022. Pediatr Infect Dis J. 2023;42:252–259. doi: 10.1097/INF.0000000000003803. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [17].Curtti T, Vázquez J, Yori S, et al. Immunization with messenger RNA vaccines against COVID-19 in adolescents with a history of multisystem inflammatory syndrome: a case series. Arch Argent Pediatr. 2023;121:e202202757. doi: 10.5546/aap.2022-02757.eng. [DOI] [PubMed] [Google Scholar]
  • [18].Epalza C, Prieto-Tato L, Pino R, et al. Safety and Acceptance of COVID-19 Vaccination After Multisystem Inflammatory Syndrome in Children (MIS-C) in Spain. J Pediatric Infect Dis Soc. 2022;11:471–473. doi: 10.1093/jpids/piac076. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Wisniewski M, Chun A, Volpi S, et al. Outcomes After SARS-CoV-2 Vaccination Among Children With a History of Multisystem Inflammatory Syndrome. JAMA Netw Open. 2022;5:e224750. doi: 10.1001/jamanetworkopen.2022.4750. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [20].Elias MD, Truong DT, Oster ME, et al. Examination of Adverse Reactions After COVID-19 Vaccination Among Patients With a History of Multisystem Inflammatory Syndrome in Children. JAMA Netw Open. 2023;6:e2248987. doi: 10.1001/jamanetworkopen.2022.48987. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [21].Centers for Disease Control and Prevention. Clinical Guidance for COVID-19 Vaccination. 2024. Available: https://www.cdc.gov/vaccines/covid-19/clinical-considerations/interim-considerations-us.html#covid19-vaccination-misc-misa.
  • [22].Butters C, Abraham DR, Stander R, et al. The clinical features and estimated incidence of MIS-C in Cape Town, South Africa. BMC Pediatr. 2022;22:241. doi: 10.1186/s12887-022-03308-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [23].World Health Organisation. Multisystem inflammatory syndrome in children and adolescents with COVID-19. 2020. 08/06/2022 Available: https://www.who.int/publications/i/item/multisystem-inflammatory-syndrome-in-children-and-adolescents-with-covid-19.
  • [24].Spracklen TF, Mendelsohn SC, Butters C, et al. IL27 gene expression distinguishes multisystem inflammatory syndrome in children from febrile illness in a South African cohort. Front Immunol. 2022;13:992022. doi: 10.3389/fimmu.2022.992022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [25].Hanekom WA, Hughes J, Mavinkurve M, et al. Novel application of a whole blood intracellular cytokine detection assay to quantitate specific T-cell frequency in field studies. J Immunol Methods. 2004;291:185–95. doi: 10.1016/j.jim.2004.06.010. [DOI] [PubMed] [Google Scholar]
  • [26].Stekhoven DJ, Bühlmann P. MissForest--non-parametric missing value imputation for mixed-type data. Bioinformatics. 2012;28:112–8. doi: 10.1093/bioinformatics/btr597. [DOI] [PubMed] [Google Scholar]
  • [27].Schmidt M, Hébert S, Wallukat G, et al. “Multisystem Inflammatory Syndrome in Children”-Like Disease after COVID-19 Vaccination (MIS-V) with Potential Significance of Functional Active Autoantibodies Targeting G-Protein-Coupled Receptors (GPCR-fAAb) for Pathophysiology and Therapy. Children (Basel) 2023;10 doi: 10.3390/children10121836. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [28].Saeed S, Cao J, Xu J, et al. Case Report: A case of multisystem inflammatory syndrome in an 11-year-old female after COVID-19 inactivated vaccine. Front Pediatr. 2023;11:1068301. doi: 10.3389/fped.2023.1068301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [29].Collignon C, Frachette C, Callot D, et al. Two pediatric cases of multisystem inflammatory-like syndrome following COVID-19 vaccination. Arch Pediatr. 2022;29:620–623. doi: 10.1016/j.arcped.2022.08.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [30].Lee JS, Cho KS, Choe YJ. A Suspected Case of Multisystem Inflammatory Disease in Children Following COVID-19 Vaccination: A Case Report and Systematic Literature Review. Pediatr Infect Dis J. 2022;41:e456–e460. doi: 10.1097/INF.0000000000003674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [31].Varghese M, Alsoub H, Koleri J, et al. Multisystem inflammatory syndrome in children (MIS-C) secondary to COVID-19 mRNA vaccination - A case report from Qatar. IDCases. 2022;30:e01606. doi: 10.1016/j.idcr.2022.e01606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [32].Santilli V, Manno EC, Giancotta C, et al. Two Pediatric Cases of Multisystem Inflammatory Syndrome with Overlapping Neurological Involvement Following SARS-CoV-2 Vaccination and Unknown SARS-CoV2 Infection: The Importance of Pre-Vaccination History. Vaccines (Basel) 2022;10 doi: 10.3390/vaccines10071136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [33].Consolini R, Costagliola G, Spada E, et al. Case Report: MIS-C With Prominent Hepatic and Pancreatic Involvement in a Vaccinated Adolescent - A Critical Reasoning. Front Pediatr. 2022;10:896903. doi: 10.3389/fped.2022.896903. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [34].Cole LD, Slate M, Minneman S, et al. Multisystem Inflammatory Syndrome after Breakthrough SARS-CoV-2 Infection in 2 Immunized Adolescents, United States. Emerg Infect Dis. 2022;28:1456–1459. doi: 10.3201/eid2807.220560. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [35].Jain E, Donowitz JR, Aarons E, et al. Multisystem Inflammatory Syndrome in Children after SARS-CoV-2 Vaccination. Emerg Infect Dis. 2022;28:990–993. doi: 10.3201/eid2805.212418. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [36].Poussaint TY, LaRovere KL, Newburger JW, et al. Multisystem Inflammatory-like Syndrome in a Child Following COVID-19 mRNA Vaccination. Vaccines (Basel) 2021;10 doi: 10.3390/vaccines10010043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [37].Yalçinkaya R, Öz FN, Polat M, et al. A Case of Multisystem Inflammatory Syndrome in a 12-Year-old Male After COVID-19 mRNA Vaccine. Pediatr Infect Dis J. 2022;41:e87–e89. doi: 10.1097/INF.0000000000003432. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [38].Yousaf AR, Cortese MM, Taylor AW, et al. Reported cases of multisystem inflammatory syndrome in children aged 12-20 years in the USA who received a COVID-19 vaccine, December, 2020, through August, 2021: a surveillance investigation. Lancet Child Adolesc Health. 2022;6:303–312. doi: 10.1016/S2352-4642(22)00028-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [39].Ludwikowska KM, Popiel A, Matkowska-Kocjan A, et al. COVID-19 mRNA BNT162b2 vaccine immunogenicity among children with a history of paediatric multisystem inflammatory syndrome temporally associated with COVID-19 (PIMS-TS) Vaccine. 2023;41:3317–3327. doi: 10.1016/j.vaccine.2023.04.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [40].Perez MA, Hsiao HM, Chen X, et al. Serologic responses to COVID-19 vaccination in children with history of multisystem inflammatory syndrome (MIS-C) Vaccine. 2023;41:2743–2748. doi: 10.1016/j.vaccine.2023.03.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [41].Chang AJ, Baron S, Hicar MD. Robust humoral immune response after boosting in children with Multisystem Inflammatory Syndrome in Children. IDCases. 2022;29:e01569. doi: 10.1016/j.idcr.2022.e01569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Opoka-Winiarska V, Morawska-Michalska I, Mertowska P, et al. Anti-SARS-CoV-2 Vaccination and PIMS-TS-Friends or Foe? Case Reports and Literature Review. Vaccines (Basel) 2024;12 doi: 10.3390/vaccines12030278. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [43].Benede N, Tincho MB, Walters A, et al. Distinct T cell polyfunctional profile in SARS-CoV-2 seronegative children associated with endemic human coronavirus cross-reactivity. iScience. 2024;27:108728. doi: 10.1016/j.isci.2023.108728. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [44].Porritt RA, Paschold L, Rivas MN, et al. HLA class I-associated expansion of TRBV11-2 T cells in multisystem inflammatory syndrome in children. J Clin Invest. 2021;131 doi: 10.1172/JCI146614. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [45].Goetzke CC, Massoud M, Frischbutter S, et al. TGFβ links EBV to multisystem inflammatory syndrome in children. Nature. 2025 doi: 10.1038/s41586-025-08697-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [46].Buonsenso D, Perramon A, Català M, et al. Multisystem Inflammatory Syndrome in Children in Western Countries? Decreasing Incidence as the Pandemic Progresses?: An Observational Multicenter International Cross-sectional Study. Pediatr Infect Dis J. 2022;41:989–993. doi: 10.1097/INF.0000000000003713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [47].Schwartz N, Ratzon R, Hazan I, et al. Multisystemic inflammatory syndrome in children and the BNT162b2 vaccine: a nationwide cohort study. Eur J Pediatr. 2024 doi: 10.1007/s00431-024-05586-4. [DOI] [PubMed] [Google Scholar]
  • [48].Ali M, Phillips L, Kaelber DC, et al. Characteristics of pediatric COVID-19 infections and the impact of influenza and COVID-19 vaccinations during the first two years of the pandemic. Front Pediatr. 2023;11:1046680. doi: 10.3389/fped.2023.1046680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [49].Cortese MM, Taylor AW, Akinbami LJ, et al. Surveillance for Multisystem Inflammatory Syndrome in US Children Aged 5-11 Years Who Received Pfizer-BioNTech COVID-19 Vaccine, November 2021 through March 2022. J Infect Dis. 2023;228:143–148. doi: 10.1093/infdis/jiad051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [50].Nygaard U, Holm M, Hartling UB, et al. Incidence and clinical phenotype of multisystem inflammatory syndrome in children after infection with the SARS-CoV-2 delta variant by vaccination status: a Danish nationwide prospective cohort study. Lancet Child Adolesc Health. 2022;6:459–465. doi: 10.1016/S2352-4642(22)00100-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [51].Zambrano LD, Newhams MM, Olson SM, et al. Effectiveness of BNT162b2 (Pfizer-BioNTech) mRNA Vaccination Against Multisystem Inflammatory Syndrome in Children Among Persons Aged 12-18 Years - United States, July-December 2021. MMWR Morb Mortal Wkly Rep. 2022;71:52–58. doi: 10.15585/mmwr.mm7102e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [52].Bizjak M, Emeršič N, Zajc Avramovič M, et al. High incidence of multisystem inflammatory syndrome and other autoimmune diseases after SARS-CoV-2 infection compared to COVID-19 vaccination in children and adolescents in south central Europe. Clin Exp Rheumatol. 2023;41:1183–1191. doi: 10.55563/clinexprheumatol/i1l2xn. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Figure S1
EMS208396-supplement-Supplementary_Figure_S1.pdf (1.1MB, pdf)

RESOURCES