key: cord-0862104-oxkd78kf
authors: Dugas, M.; Grote-Westrick, T.; Merle, U.; Fontenay, M.; Kremer, A. E.; Vollenberg, R.; Lorentzen, E.; Tiwari-Heckler, S.; Duchemin, J.; Ellouze, S.; Vetter, M.; Fuerst, J.; Brix, T.; Denkinger, C. M.; Mueller-Tidow, C.; Schmidt, H.; Tepasse, P.-R.; Kuehn, J.
title: Lack of antibodies against seasonal coronavirus OC43 nucleocapsid protein identifies patients at risk of critical COVID-19
date: 2020-12-07
journal: nan
DOI: 10.1101/2020.12.07.20245241
sha: 35c19b3d8baf07324164638ce16a1874df6d354e
doc_id: 862104
cord_uid: oxkd78kf

The vast majority of COVID-19 patients experience a mild disease. However, a minority suffers from severe disease with substantial morbidity and mortality. We report results from a non-interventional validation study comprising 248 patients (132 males, 116 females) with confirmed SARS-CoV-2 infections from three tertiary care referral centers in Germany and France. Overall median age was 60 years. The ICU group comprised more males, whereas the outpatient group contained a higher percentage of females. For each patient, the serum or plasma sample obtained closest after symptom onset was examined. Patients with critical disease had significantly lower levels of anti-HCoV OC43 nucleocapsid protein (NP)-specific antibodies compared to other COVID-19 patients (p=0.025). OC43 negative inpatients had an increased risk of critical disease (adjusted odds ratio 2.81 [95% CI 1.10 - 7.87]), higher than the risk by increased age or BMI, and lower than the risk by male sex. Frequency of critical disease in COVID-19 inpatients was significantly different according to OC43 status (p=0.009): 23 of 32 (72%) OC43 negative, 46 of 92 (50%) OC43 below or with cutoff, 5 of 18 (28%) OC43 positive inpatients required ICU therapy, respectively. Our results indicate that prior infections with seasonal human coronaviruses can protect against a severe course of COVID-19. Therefore, anti-OC43 antibodies should be measured for COVID-19 inpatients and considered as part of the risk assessment for each patient. Hence, we expect individuals tested negative for anti-OC43 antibodies to particularly benefit from vaccination against SARS-CoV-2, especially with other risk factors prevailing.

Approximately 10 to 20 percent of COVID-19 patients require hospital treatment and about a quarter of those need to be treated in intensive care units (ICU) [1] . In contrast, the majority of COVID-19 patients experience mild symptoms only. Known important risk factors are age, male sex, high body mass index and pre-existing comorbidities [2] . However, severe or fatal COVID-19 occurs also in seemingly healthy individuals without obvious risk factors. COVID-19 disease heterogeneity is still not well understood and complicates patient care.

There are several reports that cross-protection against SARS-CoV-2 might contribute to this phenomenon; for example, Grifoni [3] , Le Bert [4] and Mateus [5] have described T cell responses to SARS-CoV-2 in unexposed human individuals. In a recent manuscript, Henss [6] have stated that "disease severity seemed to correlate with low NL63-neutralizing activities, suggesting the possibility of cross-reactive protection". In a survey of 1,186 convalescent patients with mild COVID-19 [7] , contact to small children has been reported frequently (30.1% of participants). Potentially, childhood-related infections might modify disease severity of COVID-19 and thereby contribute to the low incidence of severe infections in small children [8] . Infections with human coronaviruses (HCoV) NL63, 229E, HKU1 and OC43 frequently occur in children and are a typical cause of common cold during the cold season.

A pilot study with 60 patients at University Hospital Münster reported less severe course of COVID-19 in patients with elevated levels of antibodies against seasonal human coronaviruses (sHCoV) OC43 and HKU1 [9; preprint under review], which like SARS-CoV-2 belong to the genus betacoronavirus of the subfamily orthocoronaviridae. To validate the findings from the pilot study, a study with an independent patient cohort was conducted according to the same protocol.

All patients in the pilot and the validation study showed SARS-CoV-2 infection as confirmed by RT-qPCR. For each patient, the serum or plasma sample obtained closest to onset of symptoms was examined. Critical disease was defined by invasive ventilation or ECMO therapy in intensive care units; severe disease by oxygen supplementation; moderate disease by hospitalization for other reasons without oxygen treatment; mild disease by limited clinical symptoms (fever, cough, diarrhea, myalgia, anosmia/ageusia) that were managed in an outpatient setting.

In the pilot study, 60 patients were analyzed in the context of the Coronaplasma Project (local ethics committee approval: AZ 2020-220-f-S) and COVID-19 biomarker study (ethics committee approval: AZ 2020-210-f-S) at the University Hospital Münster. This pilot study has enabled us to identify an association between antibody levels against sHCoVs and disease severity of COVID-19. Figure 1 presents basic demographic information on the validation cohort comprising 248 COVID-19 patients (132 males, 116 females) treated at tertiary care referral centers in Heidelberg, Paris and Erlangen. Inclusion criteria were confirmed COVID-19 diagnosis and age 18 years or above. Only patients with limited therapeutic options due to severe preexisting conditions (terminal illness) were excluded. Overall median age was 60 years (range 18 -96 years). Median age in the patient group with critical disease (ICU therapy) was 63, with moderate/severe disease (non-ICU) 61 and with mild disease (outpatient) 53 years, respectively. There was a significant age difference between these groups (p=0.017); COVID-19 outpatients were on average younger than inpatients. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 7, 2020. ;  Overall median body mass index (BMI) was 26.0 (range 16.9 -45.3), in the ICU group 27.2, in the non-ICU group 26.0 and in the outpatient group 25.3. We observed a significant sex difference by type of therapy (p<0.001): While the ICU group comprised more males (54 males, 20 females), the outpatient group included more females (43 males, 63 females).

[ Figure 1 ]

The Heidelberg University Hospital contributed 153 COVID-19 patients to the validation cohort (75 males, 78 females). This study was approved by the local ethics committee (ethics committee approval: S-148/2020) and informed consent was obtained from study participants. Enrollment started in March 2020. Median age for patients from Heidelberg was 60 years (range 19-90 years). Forty patients were diagnosed with critical disease, 20 with moderate or severe disease and 93 with mild disease, respectively. Median BMI was 26 (range 17 -45). Overall, 51 patients showed comorbidities: diabetes mellitus (11), arterial hypertension (12) , heart disease (9), hypothyroidism (6), kidney disease (4) and other comorbidities (9) .

The second external validation cohort included 49 patients from Assistance Publique-Hôpitaux de Paris enrolled between March and April 2020 in Cochin hospital. All patients were informed and expressed their non-opposition to the study. This non-interventional study was approved by the local ethics committee (AAA-2020-08023). Median age for patients from Paris was 53 years (range 29-80 years). The cohort comprised 22 patients with critical disease, 14 with moderate or severe disease and 13 with mild disease, respectively. Median BMI was 26 (range 18 -40). Nineteen patients had comorbidities: diabetes mellitus (8), arterial hypertension or cardiovascular disease (6), respiratory diseases (3) and other comorbidities (2) .

The third external validation cohort included 46 patients from University Hospital Erlangen enrolled between March and June 2020. This study was approved by the local ethics committee (ethics committee approval: 174_20B) and informed consent was collected from study participants. Median age for patients from Erlangen was 65 years (range 18-96 years). Twelve patients were diagnosed with critical disease, 34 with moderate or severe disease. Overall, 35 patients showed comorbidities.

Antibody levels against sHCoVs and SARS-CoV-2 were determined with the immunostrip assay recomLine SARS-CoV-2 IgG from Mikrogen GmbH, Neuried, Germany. With respect to sHCoVs, this assay detects IgG antibodies directed against the nucleocapsid protein (NP) of HCoV 229E, NL63, OC43 and HKU1. Of note, this lab test was designed by the manufacturer for high specificity regarding NP-antibodies and does not pick up antibodies against S protein of sHCoVs. For SARS-CoV-2, the assay allows the detection of NP-specific and spike protein (S)-specific antibodies directed against the S1 subunit and the receptor binding domain (RBD). As in the pilot study, analyses were performed at the Institute of Virology/Department of Clinical Virology of the University Hospital Münster according to the manufacturer's guidelines. To test precision and reliability, internal negative and positive control sera with known antibody reactivity against sHCoVs and SARS-CoV-2, respectively, were included and analyzed as given below. Immunoblot results were analyzed as described in the pilot study [9; preprint under review] . In summary, antibody levels were visually determined as ordinal values using the cutoff band of immunstrips as internal reference. Individual coronavirus-specific bands were rated on an ordinal scale as non-detectable (-), below cutoff (+/-), with cutoff intensity (+), above cutoff (++), and very strong intensity (+++), see also supplemental figure 1. Relative antibody levels were quantitatively determined with ImageJ (version 153, 64bit-Version for windows) [10] using the signal intensity of the cutoff band as internal reference (ratio HCoV-specific band to cutoff band). Standardized photographs from immunostrip assays were used for this . CC-BY 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 7, 2020. ;  analysis. Laboratory analyses were performed blinded regarding patient outcome. Quantitative properties of densitometric immunoblot data were validated with a dilution series (supplemental figure 2 ). In addition, S1-specific IgG antibodies were analyzed by ELISA (Euroimmun, Lübeck, Germany) according to the manufacturer´s instructions.

Demographical data, type of treatment and length of stay were extracted from the study databases at Heidelberg and Paris. Descriptive statistics and statistical tests were performed with R (version 3.6.1). Ordinal and numerical values were analyzed with exact Wilcoxon test, Kruskal-Wallis test, Chi-squared test, exact Fisher test, Spearman correlation test and binary logistic regression. 95% confidence intervals for odds ratios were calculated with R-package epitools. A two-sided p-value of 0.05 was considered significant.

Presence of NP-specific IgG antibodies against sHCoVs NP-specific IgG antibodies were detected for OC43 in 14%, for HKU1 in 31%, for NL63 in 41% and for NP229E in 26% of patients (n=248), respectively. No such antibodies were measured for OC43 in 21%, for HKU1 in 17%, for NL63 in 13% and for NP229E in 20% of cases (n=248). Antibody level below or with cutoff was found for OC43 in 65%, for HKU1 in 52%, for NL63 in 46% and for NP229E in 54% of patients (n=248).

We assessed sHCoV-specific IgG antibody levels in relation to symptom onset (days after start of COVID-19 symptoms) for the validation cohort. Figure 2 presents results from this analysis: We observed no systematic effect of the time point of sample collection on relative IgG antibody levels against HCoV OC43 and HKU1 NP. However, regarding HCoV NL63 and 229E, antibody levels were significantly correlated with time since symptom onset. Median time point of sample collection was 8 days for inpatients and 5 days for outpatients (p<0.0001).

[ Figure 2 ] Risk of critical COVID-19 for inpatients in relation to HCoV OC43 serostatus Table 1 presents detailed patient counts for the validation cohort regarding HCoV OC43 serostatus and type of treatment. Frequency of critical disease in COVID-19 inpatients was significantly different according to OC43 status (p=0.009). OC43 negative inpatients required ICU therapy more frequently; thus the key result from the pilot study was validated.

[ Table 1 ] Figure 3 presents levels of sHCoV-specific IgG antibody levels on an ordinal scale for inpatients (ICU group compared to non-ICU group; without outpatients). The proportion of patients without sHCoV antibodies ("-") was increased for critical disease. This pattern was most pronounced for [ Figure 3 ] is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 7, 2020. ; https://doi.org/10.1101/2020.12.07.20245241 doi: medRxiv preprint OC43 serostatus as independent risk factor Sex, age and BMI are known risk factors for COVID-19 severity. Thus, the association between HCoV OC43 antibody levels and these risk factors was analyzed (supplemental figure 3) . There was no significant difference in HCoV OC43 antibody levels between males and females in the validation cohort (p=0.96). There was no significant association between HCoV OC43 antibody levels and age (p=0.13) as or BMI (p=0.93). Therefore, HCoV OC43 antibody levels may provide independent information in addition to known risk factors.

We performed multivariate analysis of our 142 inpatients to compare the effect size of OC43 with established risk factors age, sex and BMI. In binary logistic regression, absence of HCoV OC43-specific antibodies ( Length of stay and sHCoV-specific IgG antibody levels Figure 4 presents length of stay (LoS) of hospitalised patients in relation to relative IgG antibody levels against sHCoV. Of note, a combination of high sHCoV antibody levels and long LoS occurred infrequently. Median length of stay for HCoV OC43-negative inpatients was 15 days, for OC43-positive inpatients 9.5 days (p=0.056).

[ Figure 4 ]

According to visual determination of band intensities, sHCoV-specific IgG antibodies were less frequently detected in COVID-19 inpatients with critical disease compared to all other COVID-19 patients. Supplemental figure 4 presents a comparison of critical disease (ICU group) with non-critical disease (non-ICU and outpatient groups). This pattern was most pronounced for HCoV OC43 (OC43 antibodies negative: odds ratio for critical disease 2. 16 Figure 5 shows relative sHCoV IgG antibody levels determined densitometrically. COVID-19 patients with critical disease (ICU group) presented with lower sHCoV antibody levels than patients with moderate/severe disease (non-ICU group). For HCoV OC43 and HKU1, median antibody levels were lower for outpatients compared to the non-ICU group, but higher in comparison to the ICU group. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 7, 2020. ; https://doi.org/10.1101/2020.12.07.20245241 doi: medRxiv preprint SARS-CoV-2 antibody levels Supplemental figure 6 presents results from SARS-CoV-2 IgG antibody measurements in the full validation cohort at first encounter. In general, patients with critical disease had higher SARS-CoV-2 IgG antibody levels compared to moderate/severe inpatients and outpatients. Additionally, IgG seroreactivity against SARS-CoV-2 S1 protein was determined by ELISA (Euroimmun). S1-specific antibody levels detected by ELISA closely matched results of the immunoblot assay (see supplemental Figure 7 ). There was no evidence for cross-reaction between NPOC43 and S1SARS2 on the antibody level (see supplemental Figure 8 ).

Identification of vulnerable individuals is a key priority in the current stage of the pandemic to guide protective measures and to design vaccination strategies. Recently, we reported an association of antibody levels against OC43 with mild course of COVID-19 in a pilot study comprising 60 patients [9; preprint under review] . To corroborate our observations, we conducted a non-interventional validation study with 248 patients from three major tertiary referral centers. Thus, the key finding from the pilot study was confirmed: Patients with critical COVID-19 disease showed significantly lower levels of anti-HCoV OC43 NP-specific antibodies compared to COVID-19 patients suffering from mild to severe symptoms.

The validation study provided further insights into the characteristics of HCoV antibody levels during COVID-19: First, levels of antibodies against HCoV OC43 and HKU1 NP remain relatively stable during SARS-CoV-2 infection. At the very beginning of the pandemic, Wang et al. reported on increasing HCoV OC43 antibody levels during COVID-19 for a small number of patients [11] . However, we used a different laboratory test, which is specific for NP of HCoV OC43 and was designed to minimize cross-reactivity. In our setting IgG antibodies were measured, as these values can be considered as biomarkers for HCoV infection in the past. Second, absence of anti-HCoV OC43 antibodies is associated with increased length of hospitalization for COVID-19. In our validation cohort, median LoS was elevated by 5.5 days. HCoV OC43 antibody levels were not significantly associated with sex, age or BMI. Accordingly, a lack of HCoV OC43 antibodies can be considered an independent risk factor. Third, HCoV OC43 antibody levels are lower for patients with critical disease compared to COVID-19 patients showing less severe symptoms. In the outpatient group we observed a relatively high variation regarding sHCoV antibody levels. An adjusted odds ratio of 2.30 for HCoV OC43 seronegative patients regarding critical disease might look small at first sight. However, in our validation cohort the risk of HCoV OC43 seronegative patients was higher than the risk by increased age or BMI, and lower than the risk by male sex. It is widely accepted that individuals of high age should be vaccinated with priority. If absence of HCoV OC43-specific antibodies conveys more risk than high age, HCoV OC43 seronegative individuals will particularly profit from vaccination and should also be vaccinated with priority. Fourth, the analysis of COVID-19 inpatients yielded an increased risk for critical disease regarding HCoV OC43 seronegative patients. Of note, the HCoV OC43 related risk was clearly higher than the risk associated with high age or elevated BMI regarding inpatients. A recent study based on data from electronic medical records in the Boston area [12] has reported an odds ratio of 0.1 regarding ICU care for SARS-CoV-2 patients with antibodies for endemic coronaviruses, corresponding well to our findings. For this reason, we propose to determine HCoV OC43-specific IgG antibody levels upon admission of COVID-19 patients to the hospital for individual risk assessment. Fifth, COVID-19 outpatients differed in several characteristics as compared to inpatients: in our validation cohort, outpatients had more favorable risk profiles with respect to age, sex and BMI and are likely to cope with a novel pandemic virus. Prior HCoV OC43 infections may be less important in this group compared to inpatients, which is reflected in the lower OC43 antibody level of outpatients compared to non-ICU inpatients (supplemental figure 5) . is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 7, 2020. ;  This study has important limitations: it was a non-interventional, observational study, and association does not necessarily imply causation. We analyzed a patient cohort from major tertiary care referral centers: This could explain why the overall proportion of critical COVID-19 cases is relatively high. Clearly, a prospective, randomized trial would provide more robust evidence. However, in the current pandemic decisions about vaccination priority and therapy options for COVID-19 inpatients must be taken now, based on available evidence. We observed a very consistent pattern in the data from pilot and validation study: HCoV OC43 is the seasonal coronavirus most similar to SARS-CoV-2 [13] and the association measures are stronger for HCoV OC43 than for all other HCoVs. The length of stay data could be interpreted as a dose-response relationship: The more HCoV OC43 antibodies, the shorter the hospital stay, which hints towards causality.

From our data we cannot deduce evidence for cross-reactive antibodies between sHCoVs and SARS-CoV-2. The pattern for SARS-CoV-2 antibodies in our cohort is concordant with the literature [14] . We hypothesize that prior exposure to HCoV OC43 virus facilitates T-cell based immune responses to SARS CoV-2. While previous work supports this hypothesis [3, 4, 5, 15] , further research is needed to gain deeper insights into the underlying immunological processes. According to a recent simulation study [16] , crossimmunity between HCoV OC43 and SARS-CoV-2 could affect transmission dynamics of SARS-CoV-2.

Our results provide evidence that prior infections with sHCoVs, specifically HCoV OC43, can protect against a severe course of COVID-19. Therefore, anti-HCoV OC43 antibodies should be determined in COVID-19 inpatients and considered as part of the risk assessment for individual patients. Hence, we expect individuals tested negative for anti-HCoV OC43 antibodies to particularly benefit from vaccination against SARS-CoV-2, especially with other risk factors prevailing. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 7, 2020. ; https://doi.org/10.1101/2020.12.07.20245241 doi: medRxiv preprint

The ICU group comprised more male, the outpatient group more female individuals (p<0.001).

. CC-BY 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 7, 2020. ; https://doi.org/10.1101/2020.12.07.20245241 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 7, 2020. ; https://doi.org/10.1101/2020.12.07.20245241 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted December 7, 2020. ; https://doi.org/10.1101/2020.12.07.20245241 doi: medRxiv preprint

Characteristics of and Important Lessons From the Coronavirus Disease 2019 (COVID-19) Outbreak in China: Summary of a Report of 72 314 Cases From the Chinese Center for Disease Control and Prevention

Covid-19: risk factors for severe disease and death

Targets of T Cell Responses to SARS-CoV-2 Coronavirus in Humans with COVID-19 Disease and Unexposed Individuals

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

Selective and cross-reactive SARS-CoV-2 T cell epitopes in unexposed humans

Analysis of humoral immune responses in SARS-CoV-2 infected patients

Association of contact to small children with mild course of COVID-19

Systematic review of COVID-19 in children shows milder cases and a better prognosis than adults

Less severe course of COVID-19 is associated with elevated levels of antibodies against seasonal human coronaviruses OC43 and HKU1 (HCoV OC43, HCoV HKU1)

Kinetics of viral load and antibody response in relation to COVID-19 severity

Recent endemic coronavirus infection is associated with less severe COVID-19

A metagenomic viral discovery approach identifies potential zoonotic and novel mammalian viruses in Neoromicia bats within South Africa

Antibody responses to SARS-CoV-2 in patients with differing severities of coronavirus disease 2019

Covid-19: Do many people have pre-existing immunity?

Projecting the transmission dynamics of SARS-CoV-2 through the postpandemic period

Supported by grants from BMBF (HiGHmed 01ZZ1802V, Use Case Infection Control; 01KI20152 RECOVER trial), Bavarian State Ministry for Sciences and Art (TP-10 and TP-11 to AEK), and National research network for University Medicine (NUM to AEK and MD). We are grateful to Petra Klöters-Plachky, Jutta Mohr, Alina Bauer and Amandine Houvert for excellent technical assistance.