Serology confirms SARS-CoV-2 infection in PCR-negative children presenting with Paediatric Inflammatory Multi-System Syndrome

. Children hospitalised for symptoms consistent with PIMS -TS between 28 April and 8 May 2020, and who were PCR-nega tive for SARS-CoV-2, were tested for antibodies to viral spike glycoprotein using an ELISA test.

. Eight patients (age range 7-14 years, 63% male) fulfilled case-definition for PIMS-TS during the study period. Six of the eight patients required admission to in tensive care. All patients exhibited signific ant IgG and IgA responses to viral spike glycoprotein. Further assessment showed that the IgG isotypes detected in children with PIMS-TS were of the IgG1 and IgG3 subcla sses, a distribution similar to that observed in samples from hospitalised adult COVID-19 patients. In contrast, IgG2 and IgG4 were not detected in children or adults. IgM was not detected in child ren, which contrasts with adult hospitalised adult COVID-19 patients of whom all had positive IgM responses. Conc lus ions. Strong IgG antibody responses can be detected in PCR-negative children with PIMS-TS. The low detection rate of IgM in these patients is consi stent with infection having occurred weeks previously and tha t the syndrome onset occurs well a fter the control of SARS -CoV-2 viral load. This implies that the disease is largely immune-mediated. Lastly, this indicates that serology can be an appropriate diagnostic tool in selec t patient groups. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 7, 2020. ; https://doi.org/10.1101/2020.06.05.20123117doi: medRxiv preprint 3

In adults, SARS-CoV-2 virus causes respiratory infections characterised by a markedly elevated fatality rate, similar to those observed during pandemic influenza outbreaks. Those at ri sk of severe disease or death include the elderly, certain ethnicities and those with underlying co-morbidities such a s cardiova scular disease or obesity(1). In contrast, there is a low rate of symptomatology associated with infec tion in children and a substantially lower risk of death(2). Nevertheless, in recent weeks reports have appeared describing rare pre sentations of a novel multisystem inflammatory syndrome with overlapping features of Kawasaki disease and toxic shock syndrome in children (Pae diatric Inflammatory Multisystem Syndrome temporally associated with SARS-CoV-2 pandemic (PIMS-TS)), associa ted with S ARS-CoV-2 infecti on(3). Diagnosis is complicated by the inconsi stent detection of virus in these patients. Thus, PIMS-TS may be due to the virus or could be incidental to improved surveillance resul ting from the pandemic. Serological tests for anti-viral antibodie s have not been useful to da te in the immediate diagnosis of active COVID-19 infection, which relies on viral detection by PCR in conjunction w ith clinical presentation. This is largely due to the 7-14 day lag between infection and the development of specific antibodies. In primary infections, adaptive immunity develops with slower kinetics than on subse quent exposure. For antibody responses, IgM responses develop first, before eventually waning and IgG response s dominating therea fter. Thus, high levels of Ig G in the absence of IgM are typically suggestive of infection weeks or even months previously. Below, we present findings demonstrating that children with PIMS-TS , who are PCR-nega tive for SARS-CoV-2, can present with very high levels of Ig G antibody to the virus. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 7, 2020. ; https://doi.org/10.1101/2020.06.05.20123117doi: medRxiv preprint 4

s and Methods Ethic s stat e ment. The patients' samples were either tested as part of routine diagnostics on in house COVID-19 antibody ELISAs run by the UKAS accredited Clinical Immunology Service at the Universi ty of Birmingham or used for assay development. The ethical approval for this work and the use of these samples wa s provided by the awarding bodies of the University of Birmingham Research Ethics Committee, the South Birmingham Research Ethics Committee and the National Research Ethics Service Committee West Midlands. All a pprovals are overseen by the United Kingdom National Health S ervice and this is therefore a NHS Health Research Authority approved study. All patients and/or their parents/legal guardians provided signed informed consent to inclusion of de-identified data in this report. Patient c ohor t a nd s a mples . We used a case definition consistent with Royal College of Paediatrics and Child Health guidel ines and pa tients were identified based on fulfilling the case definition for PIMS-TS described in Table 1. All were admitted to hospital betw een 28 th Apr il-8 th May 2020. Tests for SARS-CoV-2 infection by PCR gave negative results. All patients received standard supportive care that included empirical antibiotic s, re spiratory and cardiovascular support as indicated. Patients received intravenous immunoglobulin and/or steroids if they fulfilled either full or atypical Kawasaki disease criteria . EL ISA to SARS-CoV-2 spike g lycoprotein. Antibodies to near-full-length trimeric viral spike glycoprotein(4, 5), we re detected by ELISA. High-binding plates (Greiner Bio -One) were coated with spike glycoprotein (1 μ g/ml) and blocked with with Stabilcoat solution (Sigma Aldrich) before test serum was added at 1:40 or 1:50 and diluted 5-fold down the plate. HRP-labelled mouse monoclonal anti-human IgG, IgA IgM, IgG 1-4 secondary antibodies, generated at the University of Birmingham (available from Abingdon Health Ltd), were added individually or combined and HRP activity detec ted using TMB core (Bio-rad). . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 7, 2020. ; https://doi.org/10.1101/2020.06.05.20123117doi: medRxiv preprint 5

Patient presentation. Eight patients, 5 (63%) male and median age 9 (range 7-14) years, fulfilled case definition for PIMS -TS and had SARS-CoV-2 serology tested. Pati ents were of mixed ethnicity. What wa s unusual was the cluste ring of case s and in light of recent reports, the potential that this was a SARS-CoV-2 presentation was considered. 7 patients had overlapping features of hyper- inflammation with either typical or atypical Kawasaki di sea se, and one patient had overlapping features of hype r-inflammation and toxic shock syndrome. All pati ents had fever a nd at lea st one gastrointestinal symptom (abdominal pain, vomiting and diarrhoea) and 75% had a ra sh. Hyperinflammation was supported by presence of fever and the median (IQR) CRP was 188 (136- 255) mg/L and ferritin was 1325 (819-2121) μ g /L in this cohort of children. 63% of patients had impaired myocardial function on echocardiography. 75% required admission to pa ediatric intensive care predominantly for cardiovascular support due to hypotension. All patients improved with supportive therapy that included immunomodulation with immunoglobulins and/or steroids and were discharged from PICU, remaining hospital inpa tients. Antibody dete c t i on t o viral spike g lyc op r otein. Sera from these children were tested against viral spike glycoprotein, the major immunodominant antigen and compared to pre-2019 sera and plasma from adul ts with severe COVI D-19 infections admitted to the intensive treatment unit (ITU). A screening test to detect IgG, IgA and IgM was performed at a single dilution of 1:40 of sample, which demonstrated that all children had antibody against the SARS -CoV-2 spike glycoprotein (Fig. 1A). Since antibody isotypes can reflec t recent infection (IgM), or more historic infections (IgG and IgA), we examined individual antibody isotypes. In children, IgM levels were simila r to pre-2019 sera; in contrast, spike glycoprotein-specific IgM levels were high in adult ITU COVID-19 patients. (Fig . 1B). IgA and IgG were more similar in childre n and adult COVID-19 patients. Assessment of Ig G isotypes, which informs on the effector function of the antibodies, revealed IgG1 and IgG3 were the predominant isotypes pre sent in these children and in adults (Fig. 1C), with IgG2 and IgG4 similar to negative controls in all but one child, who had a weak IgG4 response (data not shown). Therefore, children with Kawa saki-like inflammatory syndrome who are negative by PCR can have high IgG1, IgG3 and IgA antibody levels to SARS -CoV-2 in the absence of maintained IgM lev els. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 7, 2020. ; https://doi.org/10.1101/2020.06.05.20123117doi: medRxiv preprint 6 Dis c us sion. Recent anecdotal reports of a Kawasaki-like inflammatory syndrome, often without detection of SARS-CoV-2 virus, appea r at odds with the rela tively mild or asymptomatic presentation of SARS- CoV-2 in fection in th e vast majority of children(2, 6). Te sting for SARS-CoV-2 infec tion in eight hospitalised children presenting with PIMS-TS found significant levels of IgG and IgA to thi s virus despite no evidence of current infection by PCR. Although PCR d etection of infection is an imperfect technique, it i s the nearest to a gold standard for determining active infection(7). Not de tecting virus in any of these eight children is unlikely to be due to false-negative PCR results and more consistent with pre sentation after viral clearance. Compatible with this conclusion, all samples had low levels of specific IgM and high levels of IgG, indicating that infection may have occurred weeks or even months previously. As we report in a forthcoming manuscript, after testing hundreds of pre- pandemic control sera we have found minimal cross-reactivity of pre-pandemic sera with the spike glycoprotein, indic ating our a ssay i s highly specific for SARS-CoV-2 and that this virus i s indeed the aetiological trigger. The absence of current infecti on would suggest that symptomatology in these children relates to immune-mediated pathology. Immune-mediated disease has been suggested to contribute to disease severity in adult infections(8). Associated with this is the detection o f IgG1 and IgG3 in these children. These isotypes a re associated with complement ac tivation(9), which has been shown to be enhanced in adult patients(10), otherwise this could simply be a proxy for other immune activity. An alternative mechanism would be antibody-dependent enhancement (ADE) which has been reported with other coronaviru ses(11-13). This is a paradoxical phenomenon in which binding of non- neutralizing antibodies to a virus enhanc es entry into host cells, resulting in more severe disease in secondary infections, but this i s unlikely in these PCR negative children. Other routes to immune- mediated damage may also be important in these individual cases. The SARS-CoV-2 specific antibodies could either induc e a pathogenic res pons e to a self-antigen t hrough molecular mimicry or could simply be marke rs of a ‘hit and run’ virus induced inflamma tory condition. This report does not explain the mechanism behind the antibody pattern but it is an important observation when considering potential complications post-vaccination and warrants further investigation. In summary, PCR-negative patients can present with a severe inflammatory syndrome whose aetiology can only be determined through antibody testing. This is important, as until now, serology has not been useful diagnostically, only for epidemiology. This therefore offers a widening of the value of serology in the identificati on and understanding of infec tions caused by SARS-CoV-2. Indee d, since all pa tients were positive serologically, it may be worth considering amending the defini tion of PIMS-TS so that TS i s not just "temporally associated with SARS-CoV-2 pandemic", but "triggered by SARS-CoV-2 in fection". . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 7, 2020. ; https://doi.org/10.1101/2020.06.05.20123117doi: medRxiv preprint 7 Ack nowledgem ent s. We would like to thank the University of Birmingham Clinical Immunology Service for their invaluable support in sample collec tion and processing. AFC is grateful for funding from The Medic al Research Council and The Institute for Global Innovation, The University of Birmingham for funding. This study was supported by the UK National Institute for Health Re search, Birmingham Biomedical Research Centre s Funding scheme. Dr Barnaby Scholefield is funded by the NIHR Clinician Scientist fellowship programme. The work in Prof. Max Crispin’s laboratory was funded by the International AIDS Vaccine Initiative, Bill and Melinda Gates Founda tion through the Collaboration for AIDS Vaccine Disc overy (OPP1084519 and OP P1115782), the Scripps Consortium for HIV Vaccine Development (CHAVD) (AI144462), and the University of Southampton Coronavirus Response Fund which has over 1000 donors from a round the world. We would like to acknowledge the support of the Birmingham Women’s and Children’s Hospital NHS Foundation trust sta ff and patients, including Drs Fiona Reynolds, Jim Gray, Mitul Patel, Phillip Hurley, Tristan Ramcharan, Habib Ali, Sakeena Samar, Penny Davis, Kathryn Harrison, William Coles, Pam Dawson, Sean Monaghan, Deevena Chinthala, Heather Duncan, Nick Richens and Sanket Sontakke . We thank Ja son McLellan for the expre ssion pla smid for the SARS-CoV-2 glycoprotein. We are gra te ful to Dr Galit Alter, Harvard University for helpful comments. We thank The Binding Site for technical assistance. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 7, 2020. ; https://doi.org/10.1101/2020.06.05.20123117doi: medRxiv preprint 8 Table 1 Case definition for Paediatric Inflammat ory Multisystem Syndrome temporally associated with SARS-CoV-2 pandemic Any child (16 years) presenting with 1 AND 2 AND 3 below 1. Presenting with • Persistent fever • Inflammation (neutrophili a, elevated CRP and lymphopaenia) • Evidence of single or multi-organ dysfunc tion (shock, cardiac, respiratory, renal, gastrointestinal or neurological disorder) • Additional features such as coagulopa thy. • This may include children fulfilling full or partial criteria for Kawasaki disea se 2. Exclusion of any other microbial cause (waiting for results of these investigations should not delay seeking expert advice) 3. SARS-CoV-2 PCR testing may be positive or negative Based on RCPCH guidance: https://www.rcpch.ac.uk/sites/default/files/2020 -05/COVID-19- Paediatric-multisystem-%20inflammatory%20syndrome-20200501.pdf . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 7, 2020. ; https://doi.org/10.1101/2020.06.05.20123117doi: medRxiv preprint 9

1. Jordan RE, Adab P, Cheng KK. Covid-19: risk factors for severe disease and death. BMJ (Clinical research ed). 2020;368:m1198. 2. Ludvigsson JF. Systematic review of COVID-19 in children shows milder cases and a better prognosis than adults. Acta paediatrica (Oslo, Norway : 1992). 2020;109(6):1088-95. 3. Riphagen S, Gomez X, Gonzalez-Martinez C, Wilkinson N, Theocharis P. Hyperinfla mmatory shock in children during COVID-19 pandemic. . Lancet 2020. 4. Watanabe Y, Allen JD, Wrapp D, McLellan JS, Crispin M. Site-specific glycan analysis of the SARS-CoV-2 spike. Science (New York, NY). 2020. 5. Wrapp D, Wang N, Corbett KS, Goldsmith JA, Hsieh CL, Abiona O, et al. Cryo-EM structure of the 2019-nCoV spike in the prefusion conforma tion. Science (New York, NY ). 2020;367(6483):1260-3. 6. Wu Z, M cGoogan JM. Characterist ics of and Impor tant Less ons From the Coronavirus Disease 2019 (C OVID-19) Outbreak in China: Summary of a Report of 72qiNN314 Cases From the Chinese Center for Disease Control and Prevention. Jama. 2020. 7. Esbin MN, Whitney ON, Chong S, Maurer A, Darzacq X, Tjian R. Overcoming the b ottleneck to widespread testing: A rapid review of nucleic acid testing approaches for COVID-19 detection. RNA (New York, NY). 2020. 8. Kadkhoda K. COVID-19: an Immunopathological View. mSphere. 2020;5(2). 9. Valenzuela NM, Schaub S. The Biology of IgG Subclasse s and Thei r Clinical Relevance to Transplantation . Transplantation. 2018;102(1S Suppl 1):S7-s13. 10. Risitano AM, Mastellos DC, Huber-Lang M, Yancopoulou D, Garlanda C, Ciceri F, et al. Complement as a target in COVID-19? Nature reviews Immunology. 2020:1-2. 11. Jaume M, Yip MS, Cheung CY, Leung HL, Li PH, Kien F, et al . Anti-severe acute re spiratory syndrome coronavirus spike antibodies trigger infection of human immune cells via a pH- and cysteine protease-independent FcγR pathway. Journal of virology. 2011;85(20):10582-97. 12. Kam YW, Kien F, Roberts A, Cheung YC, L amirande EW, Vogel L, et al. Antibodies against trimeric S glycoprotein protec t hamste rs against SARS-CoV challenge despite their capacity to mediate FcgammaRII-dependent entry into B cells in vitro. Vaccine. 2007;25(4):729-40. 13. Wang SF, Tseng SP, Yen CH, Yang JY, Tsa o CH, Shen CW, et al. Antibody-depende nt SARS coronavirus infection is mediated by antibodies against spike proteins. Biochemical and biophysical research communications. 2014;451(2):208-14. Legends Fig ur e 1. Detection of a nt i-S A RS -C oV -2 a ntibody respons es in c hildren with PI M S - T S. Se rological responses were detected against purified near-full-length trimeric SARS-Co V-2 viral spik e glycoprotein by ELISA. A) Absor bance values of individual sera at a s ingle dilution (1:40) from pre- 2019 healthy adult donors (green) or sera from children with PIMS-TS (red), or plasma from adult ITU patie nts (orange ) detected using combined HRP-labelled anti-IgG, IgA and IgM. One symbol represents results for a single serum and the bar shows the median values for each group. B) Absorbance value s for individual sera from pre-2019 negative control donors (green), children with PIMS-TS (red), or plasma from adult ITU patien ts (orange) serially diluted five-fold from 1:50, primary antibodies were detected using HRP-labell ed anti-IgG, IgA or IgM individually, or C) HRP- labelled IgG1 or IgG3. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 7, 2020. ; https://doi.org/10.1101/2020.06.05.20123117doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 7, 2020. ; https://doi.org/10.1101/2020.06.05.20123117doi: medRxiv preprint