key: cord-0309735-fup55b95
authors: Toh, Z. Q.; Anderson, J.; Mazarakis, N.; Neeland, M.; Higgins, R. A.; Rautenbacher, K.; Dohle, K.; Nguyen, J.; Overmars, I.; Donato, C.; Sarkar, S.; Clifford, V.; Daley, A.; Nicholson, S.; Mordant, F. L.; Subbarao, K.; Burgner, D.; Curtis, N.; Bines, J. E.; McNab, S.; Steer, A. C.; Mulholland, K.; Tosif, S.; Crawford, N. W.; Pellicci, D. G.; Do, L. A. H.; Licciardi, P.
title: Reduced seroconversion in children compared to adults with mild COVID-19
date: 2021-10-18
journal: nan
DOI: 10.1101/2021.10.17.21265121
sha: f98623e705abceb308246dcd04be4348c6d4b9c9
doc_id: 309735
cord_uid: fup55b95

Importance: The immune response in children with SARS-CoV-2 infection is not well understood. Objective: To compare seroconversion in children and adults with non-hospitalized (mild) SARS-CoV-2 infection and to understand the factors that influence this. Design: Participants were part of a household cohort study of SARS-CoV-2 infection. Weekly nasopharyngeal/throat swabs and blood samples were collected during the acute and convalescent period following PCR diagnosis for analysis. Setting: Participants were recruited at the Royal Childrens Hospital, Melbourne, Australia between May and October 2020. Participants: Those who had a SARS-CoV-2 PCR-positive nasal/throat swab. Main outcomes and measures: SARS-CoV-2 antibody and cellular responses in children and adults. Seroconversion was defined by seropositivity in all three serological assays. Results: Among 108 SARS-CoV-2 PCR-positive participants, 57 were children (median age: 4, IQR 2-10) and 51 were adults (median age: 37, IQR 34-45). Using three established serological assays, a lower proportion of children seroconverted compared with adults [20/54 69 (37.0%) vs 32/42 (76.2%); (p<0.001)]. This was not related to viral load, which was similar in children and adults [mean Ct 28.58 (SD: 6.83) vs 24.14 (SD: 8.47)]. Age and sex also did not influence seroconversion or the magnitude of antibody response within children or adults. Notably, in adults (but not children) symptomatic adults had three-fold higher antibody levels than asymptomatic adults (median 227.5 IU/mL, IQR 133.7-521.6 vs median 75.3 IU/mL, IQR 36.9-113.6). Evidence of cellular immunity was observed in adults who seroconverted but not in children who seroconverted. Conclusion and Relevance: In this non-hospitalized cohort with mild COVID-19, children were less likely to seroconvert than adults despite similar viral loads. This has implications for future protection following COVID-19 infection in children and for interpretation of serosurveys that involve children. Further research to understand why children are less likely to seroconvert and develop symptoms following SARS-CoV-2 infection, and comparison with vaccine responses may be of clinical and scientific importance.

Since the start of the COVID-19 pandemic, most children with COVID-19 have been 87 either asymptomatic or present with mild illness, and very few require hospitalization 1-3 . 88 However, COVID-19 cases in children are increasing in 2021 due to the emergent of SARS- 

Immunity to SARS-CoV-2 induced through natural infection is likely to be mediated 95 by a combination of humoral and cellular immunity [9] [10] [11] . Some studies comparing children and 96 adults have revealed distinct immune profiles [12] [13] [14] [15] , which have been associated with less severe 97 outcomes in children compared with adults. 98 The correlates of protection against SARS-CoV-2 have not been identified, although 99 neutralizing antibodies are increasingly recognized as the primary mediator of protection [16] [17] [18] . 100 The majority of adults (>90%) infected with SARS-CoV-2 mount an antibody response 19,20 , 101 which can persist for at least 12 months 21 . Seropositive recovered adults are estimated to have 102 up to 89% protection from reinfection against the same strain 22,23 . In contrast, the proportion 103 of children infected with SARS-CoV-2 who seroconvert is unknown, particularly among those 104 with asymptomatic or mild infection. 105 Characterization of the immune response following natural infection is important to 106 better understand factors that may be related to future protection. In this study, we compared 107 seroconversion and cellular immunity in children and adults following infection with the 108 ancestral (Wuhan) strain of SARS-CoV-2 and investigated the factors associated with this 109 response in a household cohort study in Melbourne, Australia.

110 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted October 18, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 Methods 111 Study cohort 112 Participants were recruited as part of a household cohort study at The Royal Children's 113 Hospital, Melbourne, Australia between May and October 2020. Infection with SARS-CoV-2 114 was confirmed by PCR of nasopharyngeal/oropharyngeal (NP) swabs. Participants were non-115 hospitalized patients who were asymptomatic or had mild symptoms of 116 headaches, nausea, fever, cough, sore throat, malaise and/or muscle aches 139 We used a modified two-step ELISA based on the Mount Sinai Laboratory method 140 previously described 25, 26 . Briefly, 96-well high-binding plates were coated with receptor 141 binding domain (RBD) or S1 (Sino Biological, China) antigen diluted in PBS at 2 µg/mL.

Serum samples were first screened with RBD antigen, and potential seropositive samples were 143 then confirmed with S1 antigen. Goat anti-human IgG-(1:10,000) horseradish peroxidase 144 (HRP) conjugated secondary antibody was used, and the plates were developed using 3.3', This qualitative commercial assay detects total antibodies (including IgG and IgM) to 159 SARS-CoV-2 RBD antigen. The assays were done according to the manufacturer's 160 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted October 18, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 instructions. Data was reported as a ratio of the absorbance over the kit control cut-off; 161 seropositive is defined as ratio ≥ 1.0. 

For innate cell populations (baseline sample), 100 µL of whole blood was aliquoted for 176 flow cytometry analysis. Whole blood was lysed with 1 mL of red cell lysis buffer for 10 min 177 at room temperature. Cells were washed with 1mL PBS and centrifuged at 350 x g for 5 min.

Following two more washes, cells were resuspended in PBS for viability staining using near perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in Two commercial and one in-house assay were used to measure antibody responses in 211 children and adults at acute (median day 7-12, IQR: 4-13) and convalescence timepoints 212 (median day 41, IQR: 31-49). There was a significant increase in antibody levels between acute 213 and convalescence for adults but not in children for all three assays (Fig. 1A) . Results from all 214 three assays were highly concordant, with 96/108 (88.9%) samples positive by all three assays 215 (Fig. 1B, Supplementary Fig. 4) , and 94-97% agreement between the assays (Supplementary (Fig 2C) .

Individuals were more likely to be seropositive with higher viral loads and longer viral 232 clearance time (based on those with multiple swabs collected), but there were no differences in 233 these parameters between children and adults who were seronegative or seropositive (Fig 2D-234 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted October 18, 2021. ; https://doi.org/10.1101/2021.10.17.21265121 doi: medRxiv preprint E). Interestingly, a Ct value of less than 26 was associated with seroconversion in 80% (12/15) 235 and 91% (10/11) children and adults, respectively. The proportion of children and adults who 236 were seropositive were similar when stratified by sex (Fig. 2F) . A similar age was observed 237 between seronegative and seropositive children as well as between seronegative and 238 seropositive adults (Fig 2G) .

When examining the relationship between symptomatic infection and antibody 240 response, a higher proportion of seronegative adults were asymptomatic compared to 241 seropositive adults (4/10, 40% vs. 2/32, 6.3%; p=0.02) (Fig. 2H) . Symptomatic adults on 242 average had three times more antibodies than asymptomatic adults (median 227.5 IU/mL, IQR 243 133.7-521.6 vs. median 75.3 IU/mL, IQR 36.9-113.6) and higher viral load (not statistically 244 significant) than asymptomatic adults, although the number of adults who were asymptomatic 245 and seropositive was small (Fig. 2I-J) . In contrast, a higher proportion of seropositive children 246 were asymptomatic compared to seronegative children (although not statistically significant) 247 (Fig. 2H) , and similar levels of antibodies and viral load were observed in children regardless 248 of whether they had any symptoms (Fig 2I-J) . Notably, viral load correlated with antibody T effector memory (TEM) cells compared with uninfected adults. These differences were also 255 observed between seropositive and seronegative adults but were not statistically significant 256 (Fig. 3) . There were no differences in IgG+ memory B cells, CD4+ TEM or CD8+ TEM cells in 257 children, however seropositive and seronegative children showed higher levels of transitional 258 B cells compared with uninfected children (Fig. 3) . No other differences were observed for any 259 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted October 18, 2021. ; https://doi.org/10.1101/2021.10.17.21265121 doi: medRxiv preprint of the other cell populations examined in children or adults ( Supplementary Fig. 6 ). We also 260 compared innate responses during the acute phase in children and adults. We found no 261 differences in innate immune responses for both children and adults based on serostatus, 262 although the number of samples available for this analysis was small ( Supplementary Fig. 7) . perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted October 18, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 to be associated with seroconversion in a previous study 40 . Interestingly, asymptomatic 285 infection was associated with lower seropositivity and antibody levels in adults but not in 286 children, consistent with previous studies in adults 35,41 and children 8 . This suggests that the host 287 humoral response to SARS-CoV-2 infection in children is different to adults despite similar 288 viral loads and circulating virus variant exposure.

There are several immunological hypotheses as to why children might be less likely to 290 seroconvert. First, antibody profiles (antibody isotypes and subclasses) 12, 14, [42] [43] [44] and memory B 291 cell populations have been reported to be different between children and adults, but this has 292 mostly been related to disease severity 45,46 . We 25 and others 12, 20, 42, 43, 47 have previously reported 293 similar SARS-CoV-2 specific IgG antibody levels between children and adults. We however 294 did not measure IgG subclasses (i.e. IgG1 and IgG3) which have also been associated with 295 COVID-19 severity as well as age effects 14,44 . In our study, we observed a decrease in IgG perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted October 18, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 Children are thought to have a more robust innate and/or mucosal immune response to 309 SARS-CoV-2 than adults 13, 15, [53] [54] [55] . This could explain why children in our study did not appear 310 to trigger the adaptive immune system as well as adults. However, our analysis of innate 311 immune responses in children by serostatus did not reveal any differences, likely due to the 312 small sample size. More efficient innate immunity may also suggest a shorter viral clearance 313 time in seronegative children, but this was not observed in our study. We previously showed 314 that the appearance of mucosal SARS-CoV-2 antibody levels in children was associated with 

It is important to note that our findings are based on the ancestral 'Wuhan' SAR-CoV-330 2 virus that was circulating in 2020. The relevance of our findings to current epidemiology 331 where COVID-19 cases in children have been rising due to the SARS-CoV-2 Delta variant 57 332 is unclear. Whether lower seroconversion rates in children are also observed following 333 All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted October 18, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 infection with Delta variant is unknown and warrants further research. This variant has been 334 associated with 1000x higher viral load compared with the Wuhan strain so a higher 335 seroconversion rate in children might be expected 58 .

The strengths of our study includes the use of three independent serological assays to 337 examine antibody response against SARS-CoV-2, including a subset of samples that correlated 338 positively with neutralizing antibody assay; these assays have also previously been shown to Acknowledgments 350 We thank the study participants and families for their involvement in this study. We also 351 acknowledge the MCRI Biobanking service for their help in processing the samples. Funding perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted October 18, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 (median days, IQR) stratified by serostatus (seronegative children, N=20, seropositive children, 

Seropositivity was defined as seropositive by all three assays. Pearson's correlation analysis 537 was used to examine association. Ct value: cycle threshold.

All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted October 18, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 

Observations of the global epidemiology of COVID-19 from 363 the prepandemic period using web-based surveillance: a cross-sectional analysis

Coronavirus Infections in Children Including COVID-19: An 366 Overview of the Epidemiology, Clinical Features, Diagnosis, Treatment and Prevention 367 Options in Children

Why is COVID-19 less severe in children? A review of the proposed 369 mechanisms underlying the age-related difference in severity of SARS-CoV-2 infections

Covid-19: Cases in children rise sharply in US as doctors call for vaccine approval

If Young Children's Risk of SARS-CoV-2 Infection Is Similar to That of Adults, Can 374 Children Also Contribute to Household Transmission? JAMA Pediatr

Delayed production of neutralizing antibodies 376 correlates with fatal COVID-19

Disease severity dictates SARS-CoV-2-specific neutralizing 378 antibody responses in COVID-19

Asymptomatic or mild symptomatic SARS-CoV-2 380 infection elicits durable neutralizing antibody responses in children and adolescents. JCI 381 Insight

SARS-CoV-2-specific T cell immunity in cases of 383 COVID-19 and SARS, and uninfected controls

Immunity to SARS-CoV-2 induced by 385 infection or vaccination

SARS-CoV-2 mRNA vaccines induce broad CD4+ T 387 cell responses that recognize SARS-CoV-2 variants and HCoV-NL63

Distinct antibody responses to SARS-CoV-2 in children 390 and adults across the COVID-19 clinical spectrum

Immune responses to SARS-CoV-2 infection in 392 hospitalized pediatric and adult patients

Systems serology detects functionally distinct 394 coronavirus antibody features in children and elderly

Innate cell profiles during the acute and 396 convalescent phase of SARS-CoV-2 infection in children

A correlate of protection for SARS-CoV-2 vaccines is urgently needed

Neutralizing antibody levels are highly predictive of 400 immune protection from symptomatic SARS-CoV-2 infection

Evidence for antibody as a protective 403 correlate for COVID-19 vaccines

Humoral Immune Response to SARS-CoV-2 405 in Iceland

Robust neutralizing antibodies to SARS-CoV-2 infection 407 persist for months

Protective humoral and cellular immune responses to SARS-CoV-2 409 persist up to 1 year after recovery

Antibody Status and Incidence of SARS-CoV-2 411 Infection in Health Care Workers

Incidence of SARS-CoV-2 infection in health care 413 workers from Northern Italy based on antibody status: immune protection from secondary 414 infection-A retrospective observational case-controlled study

A Comparison of Five SARS-CoV-2 Molecular Assays 417 With Clinical Correlations. American journal of clinical pathology

Persistence of SARS-CoV-2-Specific IgG in Children 6 419 Months After Infection

A serological assay to detect SARS-CoV-2 421 seroconversion in humans

Association of Age With SARS-CoV-2 Antibody 423 Response

Seroprevalence of SARS-CoV-2 antibodies in 425 children: a prospective multicentre cohort study

Difference in SARS-CoV-2 Antibody Status Between 427 Patients With Cancer and Health Care Workers During the COVID-19 Pandemic in Japan

SARS-CoV-2 Seroconversion and Viral Clearance in 430 Patients Hospitalized With COVID-19: Viral Load Predicts Antibody Response. Open Forum 431 Infect Dis

Factors associated with anti-SARS-CoV-2 IgG antibody 433 production in patients convalescing from COVID-19

Factors affecting antibody response to SARS-CoV-2 in 435 patients with severe COVID-19

Antibody response to SARS-CoV-2 infection in humans: A 437 systematic review

Virological and immunological features of SARS-439

CoV-2-infected children who develop neutralizing antibodies

Characterization of antibody response in 441 asymptomatic and symptomatic SARS-CoV-2 infection

Predictors of Nonseroconversion after SARS-CoV-2 443 Infection

Multi-center nationwide comparison of seven 445 serology assays reveals a SARS-CoV-2 non-responding seronegative subpopulation

Longitudinal Serology of SARS-CoV-2-448 Infected Individuals in India: A Prospective Cohort Study

Ct threshold values, a proxy for viral load in 453 community SARS-CoV-2 cases, demonstrate wide variation across populations and over 454 time

Clinical and immunological assessment of asymptomatic 456 SARS-CoV-2 infections

Humoral signatures of protective and pathological SARS-458

CoV-2 infection in children

Children develop strong and sustained cross-reactive 460 immune responses against Spike protein following SARS-CoV-2 infection, with enhanced 461 recognition of variants of concern. medRxiv

Serological analysis reveals an imbalanced IgG subclass 463 composition associated with COVID-19 disease severity

The immune system of children: the key to 465 understanding SARS-CoV-2 susceptibility?

Shared B cell memory to coronaviruses and other 467 pathogens varies in human age groups and tissues

Mild SARS-CoV-2 Infections and Neutralizing 469 Antibody Titers

Expansion of plasmablasts and loss of memory B 471 cells in peripheral blood from COVID-19 patients with pneumonia

Cell Subsets as Severity-474 Associated Signatures in COVID-19 Patients

SARS-CoV-2 specific T cell responses are lower in children 476 and increase with age and time after infection

Peripheral CD4+ T cell subsets and antibody response in COVID-478 19 convalescent individuals

The Role of CD4 T Cell Memory in 480

Generating Protective Immunity to Novel and Potentially Pandemic Strains of Influenza

Immune responses to SARS-CoV-2 in three children of 483 parents with symptomatic COVID-19

Pre-activated antiviral innate immunity in the upper 485 airways controls early SARS-CoV-2 infection in children

Natural mucosal barriers and COVID-19 in children

SARS-CoV-2 seropositivity and subsequent infection risk in 489 healthy young adults: a prospective cohort study

Covid-19: Cases in children rise sharply in US as doctors call for vaccine approval

Viral infection and transmission in a 498 large well-traced outbreak caused by the Delta SARS-CoV-2 variant. 2021. Accessed 11th 499

Antibody Levels in Coronavirus Disease 2019 Convalescent Patients

Robust correlations across six SARS-CoV-2 serology 504 assays detecting distinct antibody features