Objectives. We aimed to measure SARS-CoV-2 serologic responses in children hospitalized with multisystem inflammatory syndrome (MIS-C) compared to COVID-19, Kawasaki Disease (KD) and hospitalized pediatric controls.
Methods. From March 17, 2020 – May 26, 2020, we prospectively identified hospitalized children at Children’s Healthcare of Atlanta with MIS-C (n=10), symptomatic COVID-19 (n=10), KD (n=5), and hospitalized controls (n=4). With IRB approval, we obtained prospective and residual blood samples from these children and measured SARS-CoV-2 spike receptor binding domain (RBD) IgM and IgG, full-length spike IgG, and nucleocapsid protein antibodies by quantitative ELISAs and SARS-CoV-2 neutralizing antibodies by live-virus focus reduction neutralization assays. We statistically compared the log-transformed antibody titers among groups and performed linear regression analyses.
Results. All children with MIS-C had high titers of SARS-CoV-2 RBD IgG antibodies, which correlated with full-length spike IgG antibodies (R2 =0.956, P<0.001), nucleocapsid protein antibodies (R2 =0.846,P<0.001), and neutralizing antibodies (R2 =0.667, P<0.001). Children with MIS-C had significantly higher SARS-CoV-2 RBD IgG antibody titers (geometric mean titer [GMT] 6800, 95% CI 3495-13231) than children with COVID-19 (GMT 626, 95% CI 251-1563, P<0.001), KD (GMT 124, 95% CI 91-170, P<0.001) and hospitalized controls (GMT 85, P<0.001). All children with MIS-C also had detectable RBD IgM antibodies, indicating recent SARS-CoV-2 infection. RBD IgG titers correlated with erythrocyte sedimentation rate (R2 =0.512, P<0.046) and with hospital (R2 =0.548, P=0.014) and ICU lengths of stay (R2 =0.590, P=0.010).
Conclusion. Quantitative SARS-CoV-2 serology may have a role in establishing the diagnosis of MIS-C, distinguishing it from similar clinical entities, and stratifying risk for adverse outcomes.