key: cord-0689094-9ldxtqn1
authors: Ren, Lili; Fan, Guohui; Wu, Wenjuan; Guo, Li; Wang, Yeming; Li, Xia; Wang, Conghui; Gu, Xiaoying; Li, Caihong; Wang, Ying; Wang, Geng; Zhou, Fei; Liu, Zhibo; Ge, Qing; Zhang, Yi; Li, Hui; Zhang, Lulu; Xu, Jiuyang; Wang, Chen; Wang, Jianwei; Cao, Bin
title: Antibody Responses and Clinical Outcomes in Adults Hospitalized with Severe COVID-19: A Post hoc Analysis of LOTUS China Trial
date: 2020-08-25
journal: Clin Infect Dis
DOI: 10.1093/cid/ciaa1247
sha: 7480bc3787f8c92f6486505dbaf4cbd23acf0e8e
doc_id: 689094
cord_uid: 9ldxtqn1

BACKGROUND: The characteristics of neutralizing antibodies (NAbs) and antibody against major antigen proteins related to clinical outcomes in severe COVID-19 patients were still less known METHODS: The neutralizing antibodies (NAbs) and antibodies targeting nucleocapsid (N), spike protein (S), and the receptor-binding domain (RBD) in longitudinal plasma samples from the LOTUS China trial were measured by microneutralization assay and ELISA. Viral load was determined by real-time RT-PCR. A total of 576 plasma and 576 throat swabs were collected from 191 COVID-19 patients. Antibody titers related to adverse outcome and clinical improvement were analysed. Multivariable adjusted generalized linear mixed model for random effects were developed RESULTS: After day 28 post symptoms onset, the rate of antibody positivity reached 100% for RBD-IgM, 97.8% for S-IgM, 100% for N-IgG, 100% for RBD-IgG, 91.1% for N-IgM and 91.1% for NAbs. The NAbs titers increased over time in both survivors and non-survivors and correlated to IgG antibodies against N, S and RBD, while its presence showed no statistical correlation with death. N-IgG (slope -2.11, 95% CI -3.04 to -1.18, p&0.0001), S-IgG (slope -2.44, 95% CI -3.35 to -1.54, p&0.0001) and RBD-IgG (slope -1.43, 95% CI -1.98 to -0.88, p&0.0001) were negatively correlated with viral load. S-IgG titers were lower in non-survivors than survivors (p=0.020) at week 4 after symptoms onset CONCLUSIONS: IgM, IgG against N, S and RBD and NAbs developed in most severe COVID-19 patients, and do not correlate clearly with clinical outcomes. The levels of IgG antibodies against N, S and RBD were related to viral clearance

During the past several months, coronavirus disease has engulfed the world and caused millions of casualties [1] . As of August 2020, only two drugs have demonstrated minor benefits in randomized trials, and no vaccine is yet available. Understanding the humoral immune response has been important for developing diagnostic tests, and measuring antibodies may have a role in disease prognosis [2] [3] [4] [5] [6] . The dynamics of neutralizing antibodies (NAbs) are essential to the development of vaccines and antibody therapeutics.

Studies from SARS-CoV and MERS-CoV have demonstrated that many fragments in spike (S) proteins, S1-N terminal domain (NTD), receptor binding domain (RBD), and S2 can be used as targets to develop vaccines as they can elicit robust NAbs responses [7] . According to recent studies, convalescent plasma and antibodies isolated from recovered COVID-19 patients may be promising on prevention and treatment of COVID-19 [8] [9] [10] [11] . Long et al reported that total Immunoglobin (Ig) G and IgM against nucleocapsid (N) proteins and a peptide (LQPELDSFKEELDKYFKNHTSPDVD) from the S protein were detected in all enrolled patients by day 19 after illness onset [12] and the IgG levels in a high proportion of asymptomatic convalescent individuals decreased within 2-3 months after infection [13] .

Robbiani et al [14] found that plasma neutralizing activity was low in most convalescent (at an average of 39 days after the onset of symptoms) individuals among 149 COVID-19

patients. Fourati et al reported anti-S-IgA and -IgG antibody titers were higher in survivors than non-survivors at day-28 post ICU admission [15] . These studies have described the dynamic changes of antibodies to each protein in COVID-19 patients, and measuring the NAbs response during acute infection has been highlighted as a research priority [16] .

However, the dynamics of the NAb and humoral immune response and the association with clinical outcome in COVID-19 remains poorly understood. To address these issues, we longitudinally assessed 191 hospitalized COVID-19 patients from the LOTUS China trial and A c c e p t e d M a n u s c r i p t 6 characterized dynamics changes of antibody responses to explore the associations with viral clearance and clinical outcomes in severe COVID-19 patients.

The design and protocol of LOTUS China (Lopinavir Trial for Suppression of SARS-Cov-2 in China) study has been reported previously [17] . 

The patients' clinical information and laboratory findings were retrieved from LOTUS China trial. A total of 576 blood and 576 throat swabs samples were taken at days 1, 5, 10, 14, 21 and 28 after recruitment, until death or discharge, whichever came first. Viral RNA load in throat swabs and blood antibodies over time were detected. Viral RNA concentrations were tested at Teddy Clinical Research Laboratory (Tigermed--DiAn Joint Venture) by targeting to envelope (E) gene [17] . The blood samples were tested by enzyme-linked immunosorbent assay (ELISA) and microneutralization assay to detect IgM, IgG and NAbs titers.

The IgM and IgG antibody titers against nucleocapsid (N), S, and RBD for plasma samples were evaluated by using ELISA assays. The N protein was expressed by our group as reported previously [18] . The full-length ectodomain of S protein and RBD protein with a purity of ≥ 90% were used for coated proteins (Cat: 40589-V08B1 and 40592-V08H, respectively, Sino Biological, Beijing, China).

The purified N (10 ng/well), S (10 ng/well), and RBD (10ng/well) proteins were coated in 96-well plates, diluted in coating buffer (0.05 mol/L carbonate/bicarbonate, pH 9.6) at 4°C overnight and blocked with 3% bovine serum albumin. A c c e p t e d M a n u s c r i p t 9

Clinical improvement was defined as a decline of two categories on the modified sevencategory ordinal scale of clinical status, or hospital discharge. The modified seven-category ordinal scale of clinical status was defined as reported [17] .

Data were expressed as number (proportion), median (interquartile range [IQR]) or mean ± SD where appropriate. The titer of NAbs was log2 transformed for its skewed distribution. To compare the temporal changes of antibody titers or response rates, the time periods were divided by 7 days post symptoms onset (PSO) to sampling. A locally weighted scatterplot smoothing method was used to portray the trend of titer changes and antibody response rates as time. Generalized linear mixed model for random effects was conducted in univariate and multivariable analyses, which adjusted for age, gender and days PSO to sampling, to evaluate the correlations between antibody titers and viral load. We also compared the differences on antibody response rates and median days PSO to its first positivity between survivors and non-survivors, and patients with and without day 28 clinical improvement after recruitment of each antibody. To minish type I error in multiple comparison, FDR (False Discovery Rate) adjusted p values were estimated.

All statistical analyses were performed by SAS 9.4 (SAS Institute Inc.), unless otherwise indicated. A two-sided alpha of < 0.05 was considered statistically significant.

A c c e p t e d M a n u s c r i p t 10 

A total of 576 plasma samples were collected from the 191 patients. The antibody positive rates of all samples were 70.8% for NAbs, 93.1% for N-IgM, 92.9% for N-IgG, 95.7% for RBD-IgM, 91.8% for RBD-IgG, 94.4% for S-IgM and 98.0% for S-IgG (Table S1 ). The median time to antibody positivity PSO were 14, 14, 13 and 17 days for RBD-IgG, N-IgG, S-IgG and NAbs, respectively. All antibody specificities tested appeared rapidly after symptoms onset, with seroconversions plateauing at day 15, and by day 28 reaching 91.1%

for N-IgM, 97.8% for S-IgM, 100% for RBD-IgM, 100% for N-IgG, 98.9% for S-IgG, 100%

for RBD-IgG, and 91.1% for NAbs ( Figure 1A , Table S1 ). The positive rate of each antibody showed no significant difference in all the subjects, the survivors and non-survivors in weeks after symptoms onset (Table S1). No significant difference of antibody responses in improved or non-improved patients by day 28 after recruitment (Table S2) .

The first sampling time differed among patients and the earliest sample was taken at day 5 PSO. The sampling days were scattered over time, which made the trends of antibody titers and response rates of antibodies difficult to describe. Thus, locally weighted scatterplot smoothing method was adopted to fit smoothing curves that may reflect the temporal changes of each antibody. As is shown in Figure 1B , the titers of IgG antibody were all higher than that of IgM antibody and the titer of RBD-IgG was the highest, followed by S-IgG, N-IgG, RBD-IgM, S-IgM and N-IgM. After log2 transformed, the titers of NAbs increased quickly before days 20 PSO (Figure 1B) . The antibody titers continued to increase until days 15-21 for IgM antibodies against N, S, RBD, and days 22-28 for IgG antibodies against N, S, RBD, and NAbs ( Figure 1B, Table S3 ). The titers of each antibody significantly increased over A c c e p t e d M a n u s c r i p t 11 time, and the same trends were also observed in survivors. However, for non-survivors, significant increasing of antibody titers was only found in NAbs (P < 0.0001), S-IgG (P = 0.0310) and RBD-IgG (P= 0.0128) ( Table S3 ). The S-IgG titers showed significantly lower in non-survivors at days 22-28, compared to survivors (P= 0.020) (Figure 2) . The NAbs titers showed related to all the tested IgM and IgG antibodies against N, RBD and S, but the high correlation matrix was showed in RBD-IgM (0.53) and S-IgG (0.59). The tested antibody titers showed no difference in patients with clinical improvement and nonimprovement in weeks PSO ( Figure S1 ).

The increasing titers of all antibodies were significantly associated with lower viral load in throat swab samples (all P< 0.05) in the univariate analysis. However, after being adjusted for age, gender and days from symptoms onset to sampling in the multivariable regression model, only N-IgG (slope -2.11, 95% CI -3.04 to -1.18, P < 0.0001), S-IgG (slope -2.44, 95% CI -3.35 to -1.54, P< 0.0001) and RBD-IgG (slope -1.43, 95% CI -1.98 to -0.88, P< 0.0001) remained significant (Figure 3 ).

In this study, we retrospectively characterized the temporal changes of IgM and IgG antibodies against the major antigen proteins of SARS-CoV-2, especially the NAbs responses among moderate to severe COVID-19 inpatients in LOCUS China trial. The correlations between the positive rates, antibody levels, virologic characteristics, and clinical outcomes were evaluated. We found that antibody responses occurred in most patients, and their titers increased over time PSO, which were comparable to previous reports [4, 12] . The IgG Compared to IgM and IgG antibodies against SARS-CoV [2, 20] , which appeared at two weeks after viral infections, the IgM and IgG antibodies against SARS-CoV-2 appeared from 5 days PSO. The early IgG antibody responses has also been found in previous report [12] , suggesting a rapid and robust host humoral immune responses against SARS-CoV-2.

The IgG levels in COVID-19 patients have been considered related to disease severity [4, 5, 12, 15 ]. S-IgG antibody showed in a lower level in critical cases than those in mild and severe patients [4] , or lower in non-intensive care unit (ICU) patients compared to ICU cases in week 3 PSO [4] . Whether S-IgG levels related to death has still not been well documented [15] . In our study, we noted correlations of antibody titers with clinical improvements. However, when compared to that of survivors, significant lower average S-IgG titers in non-survivors was found, but showed only in the fourth week. The antibody levels related to adverse outcomes in convalescence stage was similar with that in SARS patients [20] , indicating the prediction role of S-IgG levels on adverse outcome.

The binding proteins are the viral major antigen which elicits protective humoral immunity against CoVs, relating to viral-receptor interaction blocking and viral clearance A c c e p t e d M a n u s c r i p t 13 [21, 22] . The negative correlation of S and RBD IgG antibody levels to the viral load, furtherly emphasized the protectivity IgG levels against the progression of COVID-19.

Intensive evaluation of S-IgG levels at different disease stages of COVID-19 would be useful to triage the critical ill patients, and provide evidence for plasma infusions treatment, which has been used to treat COVID-19 patients [6, 23, 24] . We also noticed the negativecorrelation of N-IgG antibody titers with viral load. Whether it was involved in virus clearance by antibody-dependent cell-mediated cytotoxicity (ADCC) warrants further investigation.

A strong positive correlation was found between the S/RBD-IgG levels and NAbs. This is in line with previously findings that a strong correlation between antibodies against binding antigen proteins and SARS-CoV-2 neutralization in COVID-19 patients [25] [26] [27] , and negative-correlation with viral load observed in our study. Interestingly, we found that the N-IgG antibody levels showed positive correlation with NAbs levels. Such findings have been reported previously [3] , however, the mechanisms were still not clear.

The correlation between viral load and NAb levels has been reported responsible for the reduction of the viral load in SARS patients [28] . In our study, the presence of NAbs were detectable around day 6 and peaked plateau about day 21 PSO. However, there was no significant difference in NAb levels between survivors and non-survivors, and NAbs titers showed no relation to viral load. This result suggested that NAbs might not suffice for the viral clearance of SARS-CoV-2 during disease progression. Clinical evidence indicated that the fatal outcome of SARS-CoV-2 infections caused mainly by lung damage and multi-organ failure, and the organ dysfunctions were associated with cytokines storm, rather than uncontrolled replication of SARS-CoV-2 [29] . These data suggested that factors related to adverse disease progresses were complicated. NAbs combined with other factors, such as cytokine storm, cellular immunity, aging, co-occurring disease, etc, would be responsible for A c c e p t e d M a n u s c r i p t 14 clinical outcomes of COVID-19 [30] [31] [32] [33] . Thus the NAbs titers would be not used as unique indicator for vaccination effect evaluation. The similar NAb levels to SARS-CoV-2 in survivor and non-survivor patients were different with patients infected with SARS-CoV, in which NAb titers was significantly lower in the recovered than in the deceased patients [20] .

The mechanism underlying such dichotomy in the two emerged coronavirus is unknown and need to further study.

The NAbs against SARS-CoV-2 were produced in 89.0% (170/191) patients, which was a little higher than Wang et al's report, who found 73.9% patients positive on NAbs [3] . The discrepancy might related to the cohorts recruited. It was surprising to note that NAbs waned from low titers to undetectable within short periods of time in 7 patients, including one nonsurvivor patients and six survivor patients. The undetectable NAb response might because of inefficient T-cell helper response [34] . Both isotype switch of antibody production and the antibody affinity maturation requires T-cell help [34] . Some patients infected with SARS-CoV-2 could recovery from infection without producing virus specific NAbs revealed that other immune responses, such as innate immunity and cellular immunity, might also be critical for recovery. However, whether these patients who did not developing NAbs might be re-infected by SARS-CoV-2 was still unknown. These findings indicated that cellular immune responses should be considered intensively in vaccine design.

There are some limitations in our study. Firstly, the involved subjects were mainly moderate to severe COVID-19 inpatients, lack of mild and asymptomatic cases, which may A c c e p t e d M a n u s c r i p t 

COVID-19) Dashboard

Profile of specific antibodies to the SARS-associated coronavirus

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

Detection of IgM and IgG antibodies in patients with coronavirus disease 2019

Kinetics of SARS-CoV-2 specific IgM and IgG responses in COVID-19 patients

Effect of Convalescent Plasma Therapy on Time to Clinical Improvement in Patients With Severe and Life-threatening COVID-19: A Randomized Clinical Trial

Neutralizing Antibodies against SARS-CoV-2 and Other Human Coronaviruses

Broad neutralization of SARS-related viruses by human monoclonal antibodies

Isolation of potent SARS-CoV-2 neutralizing antibodies and protection from disease in a small animal model

Antibody cocktail to SARS-CoV-2 spike protein prevents rapid mutational escape seen with individual antibodies

Studies in humanized mice and convalescent humans yield a SARS-CoV-2 antibody cocktail

Antibody responses to SARS-CoV-2 in patients with COVID-19

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

Convergent antibody responses to SARS-CoV-2 in convalescent individuals

SARS-CoV-2 viral loads and serum IgA/IgG immune responses in critically ill COVID-19 patients

A Coordinated Global Research Roadmap

A Trial of Lopinavir-Ritonavir in Adults Hospitalized with Severe Covid-19

Profiling Early Humoral Response to Diagnose Novel Coronavirus Disease (COVID-19)

A simple method of estimating fifty percent endpoints

Antibody responses against SARS coronavirus are correlated with disease outcome of infected individuals

Antibody Dependent Enhancement Due to Original Antigenic Sin and the Development of SARS

COVID-19, an emerging coronavirus infection: advances and prospects in designing and developing vaccines, immunotherapeutics, and therapeutics

The convalescent sera option for containing COVID-19

Passive antibody therapy for infectious diseases

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

Serological assays for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)

The receptor binding domain of the viral spike protein is an immunodominant and highly specific target of antibodies in SARS-CoV-2 patients

Neutralizing antibodies in patients with severe acute respiratory syndrome-associated coronavirus infection

Cytokine storm and leukocyte changes in mild versus severe SARS-CoV-2 infection: Review of 3939 COVID-19 patients in China and emerging pathogenesis and therapy concepts

COVID-19: consider cytokine storm syndromes and immunosuppression

Elevated exhaustion levels and reduced functional diversity of T cells in peripheral blood may predict severe progression in COVID-19 patients

Clinical Characteristics of 138 Hospitalized Patients With 2019 Novel Coronavirus-Infected Pneumonia in Wuhan, China

Treatment of 5 Critically Ill Patients With

With Convalescent Plasma

The changing preference of T and B cells for partners as T-dependent antibody responses develop

A c c e p t e d M a n u s c r i p t A c c e p t e d M a n u s c r i p t