Analysis and comparison of SARS-CoV-2 variant antibodies and neutralizing activity for 6 months after a booster mRNA vaccine in a healthcare worker population

In the context of recurrent surges of SARS-CoV-2 infections, a detailed characterization of antibody persistence over a 6-month period following vaccine booster dose is necessary to crafting effective public health policies on repeat vaccination. To characterize the SARS-CoV-2 antibody profile of a healthcare worker population over a 6-month period following mRNA vaccination and booster dose. 323 healthcare workers at an academic medical center in Orange County, California who had completed primary vaccination and booster dose against SARS-CoV-2 were recruited for the study. A total of 690 blood specimens over a 6-month period were collected via finger-stick blood and analyzed for the presence of antibodies against 9 SARS-CoV-2 antigens using a coronavirus antigen microarray. The primary outcome of this study was the average SARS-CoV-2 antibody level as measured using a novel coronavirus antigen microarray. Additional outcomes measured include levels of antibodies specific to SARS-CoV-2 variants including Delta, Omicron BA.1, and BA.2. We also measured SARS-CoV-2 neutralization capacity for a subset of the population to confirm correlation with antibody levels. Although antibodies against SARS-CoV-2 wane throughout the 6-month period following a booster dose, antibody levels remain higher than pre-boost levels. However, a booster dose of vaccine generates approximately 3-fold lower antibody reactivity against Omicron variants BA.1 and BA.2 as compared to the original Wuhan strain. Despite waning antibody levels, neutralization activity against the original Wuhan strain is maintained throughout the 6-month period. In the context of recurrent surges of SARS-CoV-2 infections despite vaccination with booster doses, our data indicate that breakthrough infections are likely driven by novel variants with different antibody specificity and not by time since last dose of vaccination, indicating that development of vaccinations specific to these novel variants is necessary to prevent future surges of SARS-CoV-2 infections.


Introduction
Since the initial 2019 outbreak of the novel beta coronavirus SARS-CoV-2, rapid international spread of the COVID-19 disease has resulted in a global pandemic. In efforts to contain the spread and severity of COVID-19, the FDA approved the emergency distribution of mRNA vaccines BNT162b and mRNA1273 in December of 2020. Both vaccines provide high rates of protective e cacy of up to 95% against the targeted virus strain following two doses administered at least 3-4 weeks apart 1,2 . There has been a rapid global increase in SARS-CoV-2 cases since then, mainly due to the high infectivity and antibody escape mutations of the new Omicron variants, as well as waning immunity from the BNT162b and mRNA1273 vaccines. 3 The FDA has since approved the administration of a third or "booster" dose of mRNA vaccines and has approved a 4th dose in high-risk patients only with more widespread administration currently under consideration. A booster vaccine dose has previously been effective at protecting against severe COVID-19-related outcomes 4 and has been shown to substantially increase neutralizing antibodies 5,6 . The neutralizing ability of the antibodies has also differed among SARS-CoV-2 variants 7 . Therefore, it is of great importance to elucidate the effectiveness of the booster vaccine in maintaining a persistent antibody response in a population consistently exposed to novel variants of SARS-CoV-2.
Here, we seek to analyze the initial rise and waning of SARS-CoV-2 antibody responses induced by the third-dose mRNA vaccine booster in a healthcare worker population over a 6-month period using a coronavirus antigen microarray, with direct comparison of antibodies against multiple variants of concern. We are not aware of any prior studies that have measured SARS-CoV-2 antibody levels after booster vaccination over a 6-month period against multiple variants of concern. Binding antibodies against SARS-CoV-2 antigens have been shown to correlate strongly with neutralizing antibodies, which are a critical component of clinical immunity. 8- 11 We con rm the correlation of measured antibody responses with SARS-CoV-2 neutralizing capacity for a subset of 30 healthcare workers utilizing an FDAauthorized neutralization assay. 12,13 Methods

Study population
This study was approved by the institutional review board (IRB) of the University of California Irvine (UCI) prior to initiation of the study. Widespread mRNA vaccination of healthcare workers (HCWs) at UC Irvine Health began in December 2020, administering over 16,000 doses of mRNA1273 (Moderna Inc.) or the BNT162b (P zer Inc. and BioNTech Inc.) vaccines within the rst 4 months. In September of 2021, the FDA approved and UCI secured and administered booster shots to all HCWs who wished to be additionally vaccinated. All HCWs working at the UCI Medical Center, located in Orange County CA, who participated in our previous study 14 were invited to receive serological testing by providing serum blood samples via a ngerstick directly before vaccination, 1-2 weeks after vaccination, 2 months, 4 months, and 6 months after booster vaccination. All blood samples were brought to the Institute for Clinical and Translational Science Core Laboratory at the UCI Medical Center. Serum samples were centrifuged using the Eppendorf 5415R and spun at 3000xg for 5 minutes. Serum was quickly transferred into a clean sterile tube and frozen at -80°C until analyzed for IgGs. Reports of their serological test results were returned within 4 weeks of receiving the test. At each assessment, demographic and work-related characteristics, testing frequency, exposure risk, and symptom history were collected via surveys administered prior to serum sample collection. Longitudinal participation was encouraged through an aggressive email campaign as well as ensuring that participants received a report of their antibody levels, but not every subject participated at every time point.
Coronavirus Antigen Microarray 953 independent nger stick blood serum samples were collected over the 6-month period for analysis. This analysis was restricted to 690 samples to adhere to the pre-speci ed guidelines (Table 1). Specimens were probed and analyzed on a coronavirus antigen microarray (CoVAM) for IgG and IgM antibodies against 37 antigens from SARS-CoV-2, other coronaviruses, and other respiratory viruses using a coronavirus antigen microarray. The CoVAM contained 10 SARS-CoV-2 antigens including nucleocapsid protein (NP) and several varying fragments of the spike (S) protein, as well as 4 SARS, 3 MERS, 12 Common CoV, and 8 in uenza antigens. A full list of antigens used in the assay can be found in Supplementary  This model was found to be 93% sensitive and 98% speci c in correctly classifying 91 PCR-positive cases and 88 pre-pandemic negative control. 15 The model was then used to generate a weighted composite measure of IgG reactivity across antigens, with weights corresponding to each antigen's relative importance in the model. This composite IgG reactivity measure was scaled up to represent the weighted mean uorescence intensity (MFI) of all antigens assayed in the CoVAM. Here, we utilized a model containing all SARS-CoV-2 antigens as above with the exception of NP, as this antigen was used to classify prior exposure to SARS-CoV-2 in a subgroup analysis.

Statistical Analysis
In order to characterize SARS-CoV-2 antibody response over time, we t a third-degree polynomial model of the average IgG reactivity measure using all available data from n = 323 HCWs. Due to the variability in the timing of the tests across individuals, we report the model-estimated average IgG reactivity means and standard error of the mean (SEM) error bars at pre-boost, 1-2 weeks, 2 months, 4 months and 6 months post-boost dose, and compared the changes over time. We then explored differences in long-term antibody response by individual characteristics hypothesized to in uence the magnitude and durability of the vaccine-induced antibody response

Neutralization Assay
In order to accurately assess neutralization capacity of SARS-CoV-2 antibodies, we utilized a FDAauthorized research use only surrogate virus neutralization test provided by GeneScript and carried out through a local commercial lab. Samples and controls are pre-incubated with HRP bound RBD to allow for interaction and binding of neutralization antibodies to RBD-HRP. The mixture is subsequently added to a pre-coated plate of hACE2 protein. Unbound RBD-HRP or RBD-HRP bound to non-neutralizing antibody will be captured by the plate. Neutralization antibodies complexed to RBD-HRP remains in the supernatant and is removed through washout steps. The reaction is quenched and samples are read at 450 nm in a microtiter plate reader. Absorbance of the sample is inversely dependent on the titer of the neutralizing antibodies, numbers are shown as a percentage of controls. Sera of 30 individuals who participated at every timepoint were tested using this pseudoneutralization assay to assess neutralizing capacity for binding of spike RBD fragments from the original Wuhan strain to human ACE2 receptors.

Data Availability
The data that support the ndings of this study are available from the corresponding author upon reasonable request.

SARS-CoV-2 Antibody levels 6 months post-vaccination
A total of 690 tests were analyzed from 323 HCWs throughout a 6-month period following booster vaccine dose (Table 1). Composite IgG antibody levels as measured by mean uorescence signi cantly increased in the 2 weeks following the booster vaccination (mean 23,303 vs 48,384, P < 0.01) (Fig. 1). Antibody levels continued to increase from 2 weeks post-boost until 2 months post-boost (mean 48,384 vs. 55,613, p < 0.01) but decreased signi cantly between months 2 and 4 (mean 55,613 vs. 50,105, p < 0.01) with a non-signi cant trend towards decreasing antibodies between months 4 and 6 (mean 50,105 vs 46,946, p = 0.14) Despite waning, antibody levels at 6 months post booster dose remain signi cantly higher than pre-boost levels.

Antibody Levels By Demographic
The cohort of HCWs were divided into subgroups based on their gender, age, presence of exposure in their community, worker role, presence of comorbidities, and history of PCR positivity. Individuals with a history of PCR positivity after their booster were found to have higher antibody levels at the 4 month timepoint (57,262 vs 40,846, p < 0.01) and at the 6 month timepoint (47,222 vs 42,451, p = 0.04) (Fig. 2a).
Vaccination side effects were assessed through a 45-point subjective scale of severity encompassing 9 side effects with severity from 0 to 5 points. Baseline antibody levels were equivalent between those with no side effects and those with severe side effects (21,639 vs 20,872, p = 0.85), but those with severe systemic side effects had signi cantly higher antibody levels at 2 months after booster vaccination (51,313 vs 69,849, p < 0.01), with a continued non-signi cant trend towards higher antibody levels compared to those without side effects (Fig. 2b). Strati cation by community exposure, healthcare worker role, gender, and age did yield any signi cant differences (Fig. 2c-2f).
Antibody Levels Against Variants Of Concern 30 healthcare workers were retrospectively chosen to undergo further analysis using a more advanced microarray that included antigens against the original Wuhan strain and 3 variants of concern: Delta, Omicron BA.1, and BA.2. This analysis was restricted to analyze only the receptor binding domain (RBD) for unbiased comparison between variants and included a total of 213 samples over the 6-month period, with a minimum of 27 samples per time point (Fig. 3). Prior to booster vaccine dose, antibody levels against the Wuhan strain were not signi cantly different from the Delta strain (mean 3473 vs. 2449, p = 0.24) but were signi cantly lower against Omicron strains BA.1 and BA.2 (mean 3473 vs. 300, p < 0.01 and 3473 vs. 542, p < 0.01). At 2 weeks post booster vaccination, all variants had signi cantly lower antibody levels directed against them as compared to the Wuhan strain (Delta: 1.3 fold-reduction, p < 0.01 Omicron BA.1 6.0 fold-reduction, p < 0.01; BA.2 3.4 fold reduction, p = < 0.01). This trend persistent over the 6-month period.

Persistence Of Neutralizing Antibodies Against Wuhan Strain Despite Waning Antibody Levels
The same 30 healthcare workers who were selected to undergo variant testing were also selected for analysis of neutralization capacity. These 30 individuals expressed a similar trend to the overall cohort, with their average antibody levels initially increasing from pre-boost to 2 months post-boost, but signi cant waning between months 2 to 6. Signal inhibition of neutralizing titers increased from 77% before booster vaccination to 96% immediately after booster vaccination (p < 0.01). Despite waning antibody levels from months 2 to 6, signal inhibition did not wane signi cantly over this time (Fig. 4).

Discussion
In summary, we report data on the boosting of both composite IgG levels and neutralizing antibodies through the administration of a third mRNA vaccine dose against SARS-CoV-2. The strengths of our study include a 6-month follow-up period which is longer than prior studies, the comparison of antibody responses against SARS-CoV-2 variants of concern, and inclusion of subgroup analyses with respect to demographics, vaccine side-effects, and breakthrough SARS-CoV-2 infection. The weaknesses of our study include limited power to detect differences in subgroups and limitations of the HCW population, which is a relatively young and healthy population with few individuals who are immunocompromised or over the age of 65.
Both IgG antibody levels and neutralization have been shown to wane over the course of months following vaccination. 14,16 The booster vaccine dose has been shown to signi cantly increase both IgG antibody levels and neutralization compared to the pre-boost levels. 5,17 Similarly, the booster vaccine dose has been shown to reduce risk of SARS-CoV-2 breakthrough infections. 6 However, while previous studies have focused on short-term effect of booster vaccination on SARS-CoV-2 antibody responses 18 , we are not aware of any prior studies that have characterized variant-speci c antibody responses over 6month period following booster vaccination.
Using our CoVAM assay, we observed a signi cant increase in antibody levels following the booster dose of the vaccine. Despite signi cant waning between months 2-6, the antibody levels at 6 months remain higher than pre-boost levels. In the context of the high rate of breakthrough infections prior to 6 months observed in recent surges of SARS-CoV-2 infections, these data indicate that waning of antibody levels over time is not the primary driver of SARS-CoV-2 breakthrough infections.
We strati ed study participants based on gender, age, community exposure risk, worker role, presence of comorbidities, and breakthrough COVID-19 infection based on a history of PCR positivity. We did not nd signi cant differences in antibody levels for subgroups strati ed by gender, age, co-morbidities, selfreported exposure to COVID-19, or clinical role, although the study may have been underpowered to detect small differences within these subgroups. We had a small cohort of individuals who had break-through infections con rmed through PCR positivity, and these individuals showed less waning in antibody response with signi cantly higher antibody levels at 4 and 6 months.
We, among others, hypothesized that the quantity and severity of side effects may be correlated with a more robust immune response, resulting in higher antibody titers. We show that the presence of systemic side effects is associated with higher antibody titers up to 4 months following booster vaccination. Severe side effects are associated with higher antibody levels in a longitudinal fashion throughout the 6months period. These data can reassure patients that systemic side-effects associated with vaccination may represent a more robust immune response to the vaccine dose.
The neutralizing capacity of SARS-CoV-2 antibodies has been shown to increase immediately the vaccine booster dose 19 . While we focused on binding antibody levels in this study, we did con rm correlation between binding antibodies and neutralizing antibodies in a subset of patients using an FDA-authorized pseudoneutralization assay 12 . We report signi cant increase in neutralizing antibodies against the original Wuhan strain following booster vaccine dose with maintenance throughout the 6-month period in this cohort, despite waning antibody levels measured on the CoVAM (Fig. 3). It is important to note that our neutralization data are limited in that our assay was likely saturated, with our values being at 96% inhibition, which may not fully allow us to detect small differences between timepoints. However, based on these data, we hypothesize that the decrease in antibody levels over the course of the 6-month period is not su cient to reduce neutralization capacity and therefore unlikely clinically signi cant.
We acknowledge that others have directly compared the neutralization capacity of vaccinee sera against SARS-CoV-2 variants of concern 19 , however, estimates for the magnitude of antibody escape by the Omicron variant of concern vary widely from as low as 1.5-fold to as high as 32-fold 20 21 22,23 . Using an updated version of the CoVAM, we were able to directly compare antibodies against the spike protein receptor-binding domain antigens of variants of concern elicited by the booster.

Con ict of Interest
The coronavirus antigen microarray is intellectual property of the Regents of the University of California that is licensed for commercialization to Nanommune Inc. (Irvine, CA), a private company for which P. Felgner is the largest shareholder and several co-authors (R. de Assis, A. Jain, R. Nakajima, and S. Khan) also own shares. Nanommune Inc. has a business partnership with Sino Biological Inc. (Beijing, China) which expressed and puri ed the antigens used in this study.

Figure 1
Antibody levels over 6 months. Mean IgG antibody levels measured by mean uorescence intensity (MFI).  side effect pro le, c) exposure in community, d) healthcare worker role, e) gender, and f) age.

Figure 3
Page 16/16 Antibody reactivity against variant RBD antigens. The antibody reactivity as measured by mean uorescence intensity (MFI) against variant-speci c RBD antigens is shown for different variants of concern at different time points after vaccination.

Figure 4
Neutralizing capacity of vaccinee sera. Neutralizing capacity of sera, measured as percent signal inhibition of binding of RBD antigen from the Wuhan strain to human ACE2 receptors compared to negative control sera, is shown for different time points after vaccination.