key: cord-0751734-go0q1yv9
authors: Zhou, Z.-h.; Dharmarajan, S.; Lehtimaki, M.; Kirshner, S. L.; Kozlowski, S.
title: Early Antibody Responses Associated with Survival in COVID19 Patients
date: 2021-02-26
journal: nan
DOI: 10.1101/2021.02.21.21252168
sha: 697a67d056250f8c937b2137bc76610a4683fdba
doc_id: 751734
cord_uid: go0q1yv9

Neutralizing antibodies to the SARS CoV-2 spike proteins have been issued Emergency Use Authorizations and are a likely mechanism of vaccines to prevent COVID-19. However, benefit of treatment with monoclonal antibodies has only been observed in clinical trials in outpatients with mild to moderate COVID-19 but not in patients who are hospitalized and/or have advanced disease. To address this observation, we evaluated the timing of anti SARS-CoV-2 antibody production in hospitalized patients with the use of a highly sensitive multiplexed bead-based immunoassay allowing for early detection of antibodies to SARS-CoV-2. We found that significantly lower levels of antibodies to the SARS-CoV-2 spike protein in the first week after symptom onset were associated with patients who expired as compared to patients who were discharged. We also developed a model, based on antibody level trajectory, to predict COVID 19 outcome that is compatible with greater antibody benefit earlier in COVID 19 disease.

SARS-COV-2 has led to more than 100 million cases of COVID-19 globally with high morbidity 43 and mortality. There were over 485,000 deaths in the United States as of February 2021 (1). 44 Neutralizing antibodies to the SARS COV-2 spike protein are candidates for therapeutics (2). 45 There have been three Emergency Use Authorizations issued for such antibodies and 46 combinations (3-5), with many others in the development pipeline. Induction of neutralizing 47 antibodies in immunized people is also likely to be a mechanism of vaccines to prevent COVID-48 19 (6, 7).

COVID-19 has an inflammatory phase associated with Acute Respiratory Distress Syndrome and 50 severe disease (8). This phase has macrophage and monocyte activation, cytokine release, may 51 involve viral transmission that is not dependent on the ACE2 viral receptor (9) and may be 52 enhanced by binding of SARS-CoV-2 specific antibodies to cells via Fc receptors (10). A host with 53 activated immune and endothelial cells may also be more sensitive to antibody-virus immune 54 complex-associated inflammation. 55 Some studies evaluating neutralizing antibody treatment in patients who are hospitalized 56 and/or have advanced disease have been stopped due to lack of benefit (11, 12). Current EUAs 57 for neutralizing antibody are limited to outpatients with mild to moderate disease who are at 58 high risk for progressing to severe disease (3-5). In contrast to findings of clinical benefit in 59 outpatients early in disease progression with mild to moderate COVID-19, some studies in 60 hospitalized patients have correlated high levels of neutralizing antibody and "early" 61 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 26, 2021. ; https://doi.org/10.1101/2021.02.21.21252168 doi: medRxiv preprint 4 seroconversion (8-16 days after onset as defined by these studies) with severity of disease for 62 both SARS-Cov-1 (13) and SARS-CoV-2 (14). 63 To address this apparent gap, we evaluated the timing of anti SARS-CoV-2 antibody generation 64 in patients who survived and in patients who expired with the use of a highly sensitive 65 multiplexed bead-based immunoassay method (15) allowing for earlier detection (within days 66 of symptom onset) of antibody to SARS-CoV-2. We also developed a model that was predictive 67 of outcome based on the trajectory of antibody levels over time. 68 These results are compatible with a model of COVID-19 disease with an early viral replication 69 phase, wherein antibody may be more beneficial, and a late inflammatory pathology phase, 70 where there is no current evidence of benefit from virus specific antibody. This disease model is 71 supported by convalescent plasma therapy that suggests better outcomes with early treatment 72 (16, 17) and no benefit with late treatment (18) . This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

Demographics of the patients in the study are provided in subsets by outcome in Table 1. The   79 numbers are small but a notable difference in outcome by sex is observed. There were 4 to 16 This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Given the amount of missing data in daily antibody levels, we tested for differences across 97 expired and discharged groups using antibody levels at week one, week two and after week 98 two. For patients with more than one sample within a week, we used the median of the 99 available antibody levels. Figure Table 2 ). The mean increase from week one was also found to be significantly higher in the 114 expired group at week two for both S1 and S2 spike proteins and at after week 2 for S1 spike 115 protein (Table 2 ). These findings suggest that later increases in antibody response are unlikely 116 to mitigate the negative effects of an initial slow response. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 26, 2021. ; https://doi.org/10.1101/2021.02.21.21252168 doi: medRxiv preprint 10 --0.943) for spike protein domains S1 and S2 respectively (Figure 3) , indicating that the patient-134 specific trajectories are highly predictive of the eventual outcome. 135 Results for the linear mixed model and the joint model for other antibody isotypes, IgM and 136 IgA, are less striking as compared to IgG and are available in the Supplementary Materials. 137 There is also a sensitivity analysis excluding two patients in the Supplementary Materials. We observed that lower levels of early antibody responses to several spike protein antigens 141 correlated with poor outcomes. In addition, a joint model (Model 2) was developed, based on 142 antibody trajectory, that was predictive of poor outcomes. This appears to differ with prior 143 studies of SARS-CoV-1 (13) and SARS-CoV-2 (14) that noted a negative correlation with early 144 antibody peak levels. These studies consider early seroconversion as <16 days or 8-14 days. 145 When the actual timing of the results is compared, these studies are not in conflict with our 146 data regarding early seroconversion at < 7 days from symptom onset. Although early antibody 147 detection is seen in other studies (19) many assay formats have limited sensitivity (20) and may 148 not detect early antibody, at low levels. Bead-based immunoassays can be highly sensitive (15, 149 21) allowing for early antibody detection. In addition, the bead-based assay described here had This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Detection for anti-CoV2 antibodies in heat-inactivated serum or plasma samples was done 244 using 96 well U-shaped plates for high-throughput runs. Serum/plasma samples were serially 245 diluted 1:100, 1:1,000, 1;10,000 and 1:100,000 for detection of IgG and IgA, and 1:20, 1:80, 246 1:320, 1:1,280 for detection of IgM. For every 96-well plate, ~ 2.4 x 10 6 beads containing targets 247 and control beads were added to 2 mL serum enhancement buffer (supplied in the BD human 248 CBA kits). Target beads and control beads were mixed in total of 9 mL buffer (2 mL serum 249 treatment buffer, 2 mL beads capture buffer, 5 mL sample diluent). Then 80µL beads and 80 µL This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. For patients who expired, the earliest sample was at Day 0 and the latest sample was at Day 30, 279 with a median of 6 samples per patient. For patients who were discharged, the earliest sample 280 was at day 0 and the latest sample was at Day 42, with a median of 6 samples per patient. 281 As there were limited sequential sample days for any given patient, we tested for differences 282 across expired and discharged groups using antibody levels at week one, week two and after 283 week two. For patients with more than one sample within a week, we used the median of the 284 available antibody levels. Density and box plots were used to display the distribution of 285 antibody levels at week one, two and after week two, along with results from a t-test 286 comparing the means of the distributions for different proteins. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. We also fit a linear mixed model (Model 1), accounting for patient-level correlation, to the data 288 to model the antibody response using time, expired / discharged status and the interaction 289 between time and expired / discharged status as predictors. Specifically, the mean antibody 290 levels at week 1, week 2 and after week 2 were estimated for the expired group and the 291 discharged group. The estimated antibody titer value differences between expired group and 292 the discharged group were calculated and underwent statistical test. The mean increase 293 antibody titer of week 2 and after week 2 from week one value was also calculated for each 294 patient group to compare antibody response. 295 To determine whether the antibody response trajectory itself can be predictive of the eventual 296 outcome, we used a joint model (Model 2) to relate the patient-specific antibody response to 297 the eventual outcome. Specifically, we first fit a linear mixed model to the data to model the This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 313 We would like to thank Mat Soukup and Yong Ma for critical review of statistical analysis. We This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 26, 2021. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 26, 2021. ; https://doi.org/10.1101/2021.02.21.21252168 doi: medRxiv preprint

JHU. COVID 19 Dashboard by the Center for Systems Science and Engineering

Therapeutic 343 Neutralizing Monoclonal Antibodies: Report of a Summit sponsored by Operation Warp Speed 344 and the National Institutes of Health

5. FDA. Casirivimab and imdevimab EUA Letter of Authorization

Efficacy and Safety of the 354 mRNA-1273 SARS-CoV-2 Vaccine

An 356 mRNA Vaccine against SARS-CoV-2 -Preliminary Report

Lessons Learned to Date on COVID-19

Hyperinflammatory Syndrome: Considerations for Interventions to Mitigate SARS-CoV-2 Viral 359 Infection and Detrimental Hyperinflammation

SARS-CoV 2; Possible alternative virus receptors and pathophysiological 361 determinants

Implications of 363 antibody-dependent enhancement of infection for SARS-CoV-2 countermeasures

Lilly Statement on the NIAID Decision to Pause Enrollment in ACTIV-3

Regeneron. REGN-COV2 Independent Data Monitoring Committee Reccomends Holding 369 Enrollment in Hospitalized Patients with High Oxygen Requirements and Continuing Enrollment 370 in Patients with Low or No Oxygen Requirements

Anti-SARS-CoV IgG response in relation to 374 disease severity of severe acute respiratory syndrome

High neutralizing antibody titer in 376 intensive care unit patients with COVID-19

Anaphylaxis Associated with Polyethylene Glycol

It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted

Plasma Therapy to Prevent Severe Covid-19 in Older Adults

Patients with Convalescent Plasma Reveals a Signal of Significantly 383 Decreased Mortality

Analytical performances of a 387 chemiluminescence immunoassay for SARS-CoV-2 IgM/IgG and antibody kinetics

Evaluation of two automated and three rapid lateral flow immunoassays for the detection of 391 anti-SARS-CoV-2 antibodies

Ultrasensitive high-393 resolution profiling of early seroconversion in patients with COVID-19

Persistence and decay 399 of human antibody responses to the receptor binding domain of SARS-CoV-2 spike protein in 400 COVID-19 patients

Distinct Early Serological 402 Signatures Track with SARS-CoV-2 Survival

Monotherapy or in Combination With Etesevimab on Viral Load in Patients With Mild to 405

Moderate COVID-19: A Randomized Clinical Trial

Company ELa. Lilly Statement Regarding NIH's ACTIV-3 Clinical Trial

https://www.lilly.com/news/stories/authors/eli-lilly-and-company. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. transformed) in expired (orange) and discharged (green) groups of patients targeting the 417 nucleocapsid protein (NP) and five spike protein components, i.e., S1+S2 ECD, S2, S1 418 (D614G), S1 and RBD. Dots are titer values for each patient at the corresponding day 419 from onset; lines are smoothed regression fit to the observed data with 95% confidence 420 interval bands. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. week 1 after onset targeting SARS CoV2 NP and five spike protein components, S1+S2 433 ECD, S2, S1 (D614G), S1 and RBD. The expired (orange) and discharged (green) 434 groups were shown with t-test comparison p-values. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Figure S1 , S2, S3 450 IgG Levels at Week 2 and After Week 2 Post-onset Comparisons Figure S4 , S5, Table S1 451 IgG Analysis Using IgG Cut-point MFI Method Figure S6 This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted February 26, 2021. ; https://doi.org/10.1101/2021.02.21.21252168 doi: medRxiv preprint