BackgroundChildren have a lower rate of COVID-19, potentially related to cross-protective immunity conferred by seasonal coronaviruses (HCoVs). We tested if prior infections with seasonal coronaviruses impacted SARS-CoV-2 infections and related Multisystem Inflammatory Syndrome (MIS). MethodsThis cross-sectional observational study in Paris hospitals enrolled 739 pauci or asymptomatic children (HOS group) plus 36 children with suspected MIS (MIS group). Prevalence, antigen specificity and neutralizing capability of SARS-CoV-2 antibodies were tested. Antibody frequency and titres against Nucleocapsid (N) and Spike (S) of the four seasonal coronaviruses (NL63, HKU1, 229E, OC43) were measured in a subset of seropositive patients (54 SARS-CoV-2 (HOS-P subgroup) and 15 MIS (MIS-P subgroup)), and in 118 matched SARS-CoV-2 seronegative patients (CTL subgroup). FindingsSARS-CoV-2 mean prevalence rate in HOSP children was 11.7% from April 1 to June 1. Neutralizing antibodies were found in 55{middle dot}6% of seropositive children, and their relative frequency increased with time (up to 100 % by mid-May). A majority of MIS children (25/36) were SARS-CoV-2 seropositive, of which all tested (n=15) had neutralizing antibodies. On average, seropositive MIS children had higher N and S1 SARS-CoV-2 titres as compared to HOS children. Patients from HOS-P, MIS-P, and CTL subgroups had a similar prevalence of antibodies against the four seasonal HCoVs (66{middle dot}9 -100%). The level of anti-SARS-CoV-2 antibodies was not significantly different in children who had prior seasonal coronavirus infection. InterpretationPrior infection with HCoVs does not prevent SARS-CoV-2 infection and related MIS in children. Children develop neutralizing antibodies after SARS-CoV-2 infection. Evidence before this studyChildren seem to be less likely affected by SARS-CoV-2 infection and clinical course of COVID-19 is less severe than in adults. As those asymptomatic or mildly symptomatic children are underdiagnosed and their viral loads are comparable to those of adults, they may act as an asymptomatic reservoir for the spread of the virus. One explanation of the difference between the adult and the pediatric infectious profile might be that infection with seasonal human coronaviruses, which is very frequent from a very young age, could lead to cross protective immunity. We searched in PubMed, MedRxiv and BioRxiv for publications from inception to June 15, 2020, using the terms "COVID-19, SARS-CoV-2, children, serology, Kawasaki, Corona Virus". Added value of this studySARS-CoV-2 mean prevalence rate was 11.7% from April 1 to June 1 and neutralizing antibodies were found in 55% of the tested seropositive children. Among patients with a Multisystem Inflammatory Syndrome, Kawasaki-like disease, 70% were SARS-CoV-2 seropositive and had neutralizing antibodies. COVID-19 and MIS attack rates, and anti-SARS-CoV-2 antibodies titres were not significantly impacted by prior seasonal coronavirus infection. Implications of all the available evidencePrior infection by seasonal coronaviruses does not prevent SARS-CoV-2 infection and associated Multisystem Inflammatory Syndrome in children As antibodies against seasonal coronaviruses are very frequent and as these viruses circulate efficiently in human populations every winter, our results question to what extent the concept of herd immunity based on circulating antibodies can be applied to seasonal coronaviruses and possibly SARS-CoV-2.
