Durability of Cross-Neutralizing Antibodies 5.5 Months After Bivalent Coronavirus Disease 2019 Vaccine Booster

Abstract

We analyzed neutralizing antibodies in samples from ancestral + BA.1 and ancestral + BA.4/5 boosted individuals, collected around 5.5 months after booster. Titers of neutralizing antibodies generally decreased compared to a time point early after the bivalent booster immunization. This was more pronounced for individuals without infection history and for recently emerged Omicron variants.

We analyzed durability of neutralizing antibody profiles in individuals after bivalent COVID-19 booster immunization (ancestral + BA.1 or ancestral + BA.4/5). Titers decreased around 5.5 months after boost compared to an earlier time point and were more stable in individuals with hybrid immunity.

In late 2022, updated ancestral + BA.1 and ancestral + BA.4/5 bivalent coronavirus disease 2019 (COVID-19) vaccines were approved. We and others previously showed good neutralization of the parental BA.1, BA.2, and BA.5 Omicron variants early after the bivalent booster vaccine, whereas recently emerged Omicron lineages (eg, BA.2.75, BQ, and XBB) further escaped immunity, albeit to varying extents [1–5]. However, little is known about the durability of neutralizing antibodies beyond 3 months after the bivalent COVID-19 boost, particularly when comparing BA.1 and BA.4/5 bivalently boosted individuals and corresponding hybrid-immune populations with a history of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection prior to bivalent boost.

MATERIALS AND METHODS

Ethic Statement

The local ethics committee of the Medical University of Innsbruck approved this study (ethics committee numbers 1064/2021 and 1059/2022).

Plasma Samples

Plasma samples were obtained from 39 individuals who had received 3 doses of monovalent ancestral COVID-19 vaccine, followed by either ancestral + Omicron BA.1 (n = 16) or ancestral + Omicron BA.4/5 (n = 23) bivalent boost. Patient characteristics are specified in more detail in Supplementary Table 1. On average, blood samples were collected 5.6 months (standard deviation [SD], 0.62 months] after the bivalent boost. Additionally, the majority of study participants [n = 34] had plasma samples analyzed from an earlier time point after bivalent boost [mean, 1.5 months [SD, 0.60 months]) as previously described [1]. Characteristics of individuals with multiple plasma samples are described in Supplementary Table 2.

Serological Testing

To identify individuals with hybrid immunity, plasma samples were analyzed in a commercially available SARS-CoV-2 nucleocapsid (N)–specific enzyme-linked immunosorbent assay (ELISA) (Elecsys, Anti-SARS-CoV-2, reference number 09203095, Roche) as instructed by the manufacturer using a cobas e411 fully automated analyzer (Roche). Samples with a cutoff index ≥1 were considered positive.

Focus-Forming Neutralization Assay

Heat-inactivated plasma samples were tested for neutralization of replication-competent SARS-CoV-2 isolates (Supplementary Table 3) with a focus-forming assay as previously described [6]. In brief, plasma dilutions (1:16 to 1:16 384) were incubated with live SARS-CoV-2 before subconfluent Vero cells overexpressing TMPRSS2 and ACE2 were infected for 2 hours and fixed further 8 hours later. Subsequently, an immunofluorescence staining was performed (SARS-CoV-2 convalescent serum as primary antibody followed by goat antihuman Alexa Fluor Plus 488–conjugated secondary antibody; ThermoFisher Scientific, catalog number 48276) and infected cells were counted using an ImmunoSpot S6 Ultra-V reader and FluoroSpot software (CTL Europe GmbH, Bonn, Germany). Continuous 50% neutralization titers were calculated by nonlinear regression using GraphPad Prism software version 9.0.1 (GraphPad, La Jolla, California). Titers ≤1:1 were set to 1:1 and titers ≥1:16 384 were set to 1:16 384. Neutralization titers <1:16 were considered negative.

Antigenic Cartography

A single exposure map as described previously has been used as base plane to construct antibody landscapes [1, 6]. Geometric mean titer antibody landscapes visualized neutralization profiles in a third dimension and were generated as described elsewhere [1, 6]. The height of antibody landscapes reflects the mean neutralization capacity against underlying virus variant, with increasing height indicating enhanced neutralization.

RESULTS

We collected plasma samples from ancestral + BA.1 bivalent– and ancestral + BA.4/5 bivalent–boosted individuals on average 5.6 months after bivalent booster and characterized their neutralization capacity against a panel of pre-Omicron (D614G, Beta, Delta), parental (BA.1, BA.2, BA.5), and recently emerged Omicron (CH.1.1, BQ.1.1, XBB.1.5.1) variants using a live virus neutralization assay. As a prior SARS-CoV-2 infection significantly influences an individual's neutralization profile, we first performed an anti-N ELISA to identify hybrid-immune individuals (Supplementary Figure 1). Accordingly, study participants were assigned to 1 of 4 cohorts: ancestral + BA.1 bivalent-boosted individuals with or without detectable levels of N antibodies (BA.1 biv./N⁺ and BA.1 biv./N⁻) and ancestral + BA.4/5 bivalent-boosted individuals with or without detectable levels of N antibodies (BA.4/5 biv./N⁺ and BA.4/5 biv./N⁻). The characteristics of study cohorts analyzed are specified in Supplementary Table 1. Although tested negative in the anti-N ELISA, 3 ancestral + BA.4/5 bivalent-boosted participants were assigned to the BA.4/5 biv./N⁺ cohort, due to positive anti-N titers in a previous study approximately 2–3 months after the bivalent booster (referred to as BA.4/5 biv./N⁺ to N⁻ and marked with triangles in Figure 1A) [1]. Participants with detectable N antibodies at both time points or those with only the late blood collection but positive for N at that time point were labeled as BA.4/5 biv./N⁺ only.

Figure 1.

Neutralization profiles against severe acute respiratory syndrome coronavirus 2 variants after ancestral + BA.1 or ancestral + BA.4/5 bivalent boost. Plasma samples from individuals with 3 doses of ancestral vaccine were collected at an average of 5.6 mo after a fourth dose with ancestral + BA.1 (red) or ancestral + BA.4/5 (teal) bivalent booster. An anti-nucleocapsid (N) enzyme-linked immunosorbent assay (ELISA) was performed to identify hybrid-immune (N⁺) individuals and study cohorts were grouped accordingly: ancestral + BA.1 boost and N negative (BA.1 biv./N⁻; n = 10), ancestral + BA.1 boost and N positive (BA.1 biv./N⁺; n = 6), ancestral + BA.4/5 boost and N negative (BA.4/5 biv./N⁻; n = 13), and ancestral + BA.4/5 boost and N positive (BA.4/5 biv./N⁺; n = 10). A, Continuous neutralization titers (50% inhibitory concentration [IC₅₀]) of study participants against indicated variants are shown. Assay negativity (IC₅₀ titer <1:16) is marked in gray. Individual results are illustrated by symbols, and titers for each individual are connected by lines. Bold lines and white circles represent geometric mean titers (GMTs) of study cohorts. Three ancestral + BA.4/5 bivalent-boosted, N ELISA-negative samples were included in the BA.4/5 biv./N⁺ study cohort, due to positive N titers in the past (referred to as BA.4/5 biv./N⁺ to N⁻, open triangle symbols). B, An antigenic map with colored circles indicating virus variants as previously published [1]. C, antibody landscapes. (B) served as a base plane to illustrate neutralization profiles of analyzed study cohorts in 3D by GMT antibody landscapes (C) (Licensed under a Creative Commons Attribution 4.0 International License http://creativecommons.org/licenses/by/4.0/). C, The left panel shows antibody landscapes of BA.1 biv./N⁻, BA.1 biv./N⁺, BA.4/5 biv./N⁻, and BA.4/5 biv./N⁺ study cohorts as indicated. The right panel shows antibody landscapes of the BA.4/5 biv./N⁻ and BA.4/5 biv./N⁺ study cohorts in comparison to the BA.4/5/N⁺ to N⁻ cohort (n = 3; light gray) and the BA.4/5 biv./N⁺ excluding BA.5/N⁺ to N⁻ cohort (referred to as BA.4/5 biv./N⁺ only; n = 7; dark gray).

As shown in Figure 1, around 5.5 months after the bivalent boost, almost all analyzed individuals neutralized pre-Omicron and parental BA.1, BA.2, and BA.5 Omicron variants. Titers were generally higher in hybrid-immune individuals (BA.1 biv./N⁺ and BA.4/5 biv./N⁺) compared to vaccination-only cohorts. This is in line with data 2–3 months after the bivalent boost [1, 2, 4]. Although we observed reduced cross-neutralization of the recent CH.1.1, BQ.1.1, and XBB.1.5.1 variants, the majority of hybrid-immune individuals still had detectable titers. In contrast, only a few samples from the N⁻ study cohorts neutralized these 3 variants. Interestingly, the 3 BA.4/5/N⁺ to N⁻ samples (Figure 1A [right panel; triangle symbols]) also tended to have lower titers against several analyzed variants. As the time point of infection is unknown for most study participants, we can only speculate that loss of N seropositivity and lower neutralizing antibody titers may be due to infections that occurred longer ago. Overall, we did not observe differences in long-term neutralizing antibody responses between the 2 different booster regimens, but differences were mainly caused by the N antibody serostatus (ie, past infection history).

Antibody landscapes are a tool to visualize neutralization profiles after complex exposure histories above antigenic maps [7, 8]. We previously described an antigenic map containing 21 SARS-CoV-2 variants (Figure 1B) [1]. On the antigenic map, shorter distances between variants indicate antigenic similarity with good cross-neutralization (eg, D614G to Alpha variant), while variants with no or low level of cross-neutralization are found relatively distant (eg, pre-Omicron to Omicron variants) [6]. Above this antigenic map, the neutralization profiles of the study cohorts were illustrated by antibody landscapes (Figure 1C), where increasing magnitudes indicate enhanced neutralization of the underlying variants. Antibody landscapes of N serostatus–matched cohorts did not differ from each other. This is in line with data collected approximately 2 months after bivalent vaccination in our previous study [1]. The tendency for an increased magnitude in the landscape of the BA.1 biv./N⁺ compared to the BA.4/5 biv./N⁺ cohort might be explained by the low-level neutralization of the 3 BA.4/5 biv./N⁺ to N⁻ individuals in this cohort. To verify this, we excluded the 3 individuals who lost N antibodies during follow-up (BA.4/5 biv./N⁺ to N⁻) from the BA.4/5 biv./N⁺ landscape, which then was referred to as BA.4/5 biv./N⁺ only cohort (Figure 1C [right panel, dark gray]). Compared to the BA.4/5 biv./N⁺ landscape, we indeed found an improved magnitude of the BA.4/5 biv./N⁺ only landscape, equalizing the level of the BA.1 biv./N⁺ landscape. Complementarily, the BA.4/5 biv./N⁺ to N⁻ landscape (Figure 1C [right panel, light gray]) was more similar to that of the BA.4/5 biv./N⁻ cohort. These data suggest that it is becoming increasingly difficult to analyze vaccine regimens in individuals with a well-defined infection history. Antibodies against N should be a good indicator of recent infections but may be less reliable for older infections.

To further investigate the durability of neutralizing antibody responses, we compared neutralization profiles early and later after the bivalent booster for 34 individuals from our study cohorts, for whom we had additional plasma samples approximately 1 to 1.8 months after the bivalent boost (Supplementary Table 2). Over the 4-month follow-up period, we found a considerable decrease in neutralization in N-negative individuals whereas titers were more stable in N-positive individuals (Supplementary Figure 2). In particular, neutralization capacity against the recently emerged CH.1.1 and XBB.1.5.1 variants dropped, while other variants were less affected. These observations agree with other studies [3, 9]. This may also be conceivable for hybrid-immune individuals whose infection occurred relatively long ago, as we have seen similar neutralization profiles in BA.4/5 biv./N⁻ and BA.4/5 biv./N⁺ to N⁻ study cohorts. In contrast, neutralization capacity remained more stable over the follow-up period in hybrid-immune individuals with persistent N titers.

DISCUSSION

One limitation of our study is that we cannot rule out past infections in our N-negative cohorts or multiple infections in the N-positive cohorts. Infection status was determined based on N ELISA and self-reported infections. However, N seroconversion rates might be lower in vaccinated individuals after subsequent infection, or N antibodies may be lost over time [10, 11]. Additionally, infecting variants are unknown, which might influence neutralization profiles. We did not analyze neutralizing antibody titers against the most recent BA.2.86 Omicron variant; however, other studies indicate that neutralization of this variant early after bivalent booster immunization is similar to that against XBB.1.5 [12, 13].

Taken together, these data show that cross-neutralization against currently circulating Omicron variants is limited approximately 6 months after the bivalent boost, particularly in individuals who do not have antibodies against N. Further vaccine updates might be beneficial to overcome the immune escape of current and, potentially, future SARS-CoV-2 variants.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.

Notes

Author contributions. A. R. designed and performed neutralization assays and wrote the manuscript. L. K., D. B., and W. B. performed clinical work. A. N., supervised by D. J. S., performed the antigenic cartography. D. v. L. co-designed the study. J. K. designed and supervised the study. All authors edited and reviewed the manuscript prior to submission.

Acknowledgments. We thank Albert Falch, Maria Huber, and Verena Pittl for their excellent technical support.

Financial support. J. K. was supported by the European Union's Horizon 2020 research and innovation program (grant agreement 101016174) and the Austrian Science Fund (project number P35159-B). J. K., D. J. S., and A. N. were supported by the Centers of Excellence for Influenza Research and Response, National Institute of Allergy and Infectious Diseases, National Institutes of Health (contract number 75N93021C00014 as part of the SAVE program). A. N. was supported by the Gates Cambridge Trust.