Long-Term Antibody Response to SARS-CoV-2 in Children

Abstract

Almost 2 years into the pandemic and with vaccination of children significantly lagging behind adults, long-term pediatric humoral immune responses to SARS-CoV-2 are understudied. The C19.CHILD Hamburg (COVID-19 Child Health Investigation of Latent Disease) Study is a prospective cohort study designed to identify and follow up children and their household contacts infected in the early 2020 first wave of SARS-CoV-2. We screened 6113 children < 18 years by nasopharyngeal swab-PCR in a low-incidence setting after general lockdown, from May 11 to June 30, 2020. A total of 4657 participants underwent antibody testing. Positive tests were followed up by repeated PCR and serological testing of all household contacts over 6 months. In total, the study identified 67 seropositive children (1.44%); the median time after infection at first presentation was 83 days post-symptom onset (PSO). Follow-up of household contacts showed less than 100% seroprevalence in most families, with higher seroprevalence in families with adult index cases compared to pediatric index cases (OR 1.79, P = 0.047). Most importantly, children showed sustained seroconversion up to 9 months PSO, and serum antibody concentrations persistently surpassed adult levels (ratio serum IgG spike children vs. adults 90 days PSO 1.75, P < 0.001; 180 days 1.38, P = 0.01; 270 days 1.54, P = 0.001). In a low-incidence setting, SARS-CoV-2 infection and humoral immune response present distinct patterns in children including higher antibody levels, and lower seroprevalence in families with pediatric index cases. Children show long-term SARS-CoV-2 antibody responses. These findings are relevant to novel variants with increased disease burden in children, as well as for the planning of age-appropriate vaccination strategies.

Keywords: Ancestral; COVID-19; Longitudinal; Pediatric; Serology.

Fig. 1

Seroprevalence among household contacts. A Seroprevalence within individual families recalled for follow-up. The seroprevalence, calculated by the seropositive household contacts divided by all household contacts, is shown on the x-axis. Families are depicted by their respective ID and arranged by seroprevalence from top to bottom. At the end of each bar, the number of family members participating in the study is shown. Families are colored according to the age group of the respective index case (child < 18 years or adult > 18 years), where applicable. Seven families were not able to recall the index case. For three families (ID: 28, 29, 73), at least one family member refused blood draw. B Seroprevalence within families by age group of the index case. Families with an adult index case had a higher seroprevalence compared to families with a pediatric index case (mean seroprevalence 73.8 vs. 48.9%). To model seroprevalence with the type of index case, a logistic regression model was applied. Results are shown in the lower panel (odds ratio 1.79, 95% CI 1.01–3.19, P = 0.047)

Fig. 2

Longitudinal follow-up of serum antibody concentrations of SARS-CoV-2 anti-spike IgG in adults and children. Observation times post-symptom onset are indicated on the x-axis and are calculated per family. The y-axis is showing predicted conditional mean values (bold curve) of the serum SARS-CoV-2 anti-spike IgG concentration as determined by the zero-inflated tweedie mixed model. The shaded areas flanking the curves indicate 95% confidence intervals. Predicted conditional means are dependent on zero-inflation. Table in the top right section showing estimated mean ratio of predicted conditional mean values children vs. adults at 90, 180, and 270 days post-symptom onset, along with 95% confidence intervals and corresponding P values. Color coding for adults and children is indicated