Immune Evasion of SARS-CoV-2 Omicron Subvariants XBB.1.5, XBB.1.16 and EG.5.1 in a Cohort of Older Adults after ChAdOx1-S Vaccination and BA.4/5 Bivalent Booster

The recently emerged SARS-CoV-2 Omicron sublineages, including the BA.2-derived XBB.1.5 (Kraken), XBB.1.16 (Arcturus), and EG.5.1 (Eris), have accumulated several spike mutations that may increase immune escape, affecting vaccine effectiveness. Older adults are an understudied group at significantly increased risk of severe COVID-19. Here we report the neutralizing activities of 177 sera samples from 59 older adults, aged 62–97 years, 1 and 4 months after vaccination with a 4th dose of ChAdOx1-S (Oxford/AstraZeneca) and 3 months after a 5th dose of Comirnaty Bivalent Original/Omicron BA.4/BA.5 vaccine (Pfizer-BioNTech). The ChAdOx1-S vaccination-induced antibodies neutralized efficiently the ancestral D614G and BA.4/5 variants, but to a much lesser extent the XBB.1.5, XBB.1.16, and EG.5.1 variants. The results showed similar neutralization titers between XBB.1.16 and EG.5.1 and were lower compared to XBB.1.5. Sera from the same individuals boosted with the bivalent mRNA vaccine contained higher neutralizing antibody titers, providing a better cross-protection against Omicron XBB.1.5, XBB.1.16 and EG.5.1 variants. Previous history of infection during the epidemiological waves of BA.1/BA.2 and BA.4/BA.5, poorly enhanced neutralization activity of serum samples against XBBs and EG.5.1 variants. Our data highlight the continued immune evasion of recent Omicron subvariants and support the booster administration of BA.4/5 bivalent vaccine, as a continuous strategy of updating future vaccine booster doses to match newly emerged SARS-CoV-2 variants.

Older adults (aged > 60 years) are a high-risk group vulnerable to severe disease and death after SARS-CoV-2 infection [9], thus it has been demonstrated that there is a less robust immune response after COVID-19 vaccination in this priority group [10,11].In that way, since SARS-CoV-2 spike mutations deeply contribute to immune escape and transmissibility, a major goal of vaccine design is the evaluation of an optimal protection for the elderly.Furthermore, the kinetics of neutralizing activity in sera elicited by vaccination and the capacity of boosters to enhance cross-protection against new SARS-CoV-2 variants in elderly cohorts are of great interest.With the emergence and rapid dispersion of Omicron BA.4/5 worldwide [1], it was shown that individuals vaccinated with four doses of prototype mRNA vaccine did not produce a strong immune response against these Omicron sublineages, highlighting the need for bivalent vaccines development targeting the prototype and the BA.4/5 spike protein [12,13].Because the recently emerged Omicron subvariants have accumulated additional spike mutations, it is important to access the vaccine-elicited neutralization against these new sublineages.Herein we investigated the escape of XBB.1.5,XBB.1.16,and EG.5.1 Omicron sublineages from neutralizing antibodies in serum samples collected from a cohort of 59 volunteers aged between 62 and 97 years (median 73 years) after 1 and 4 months of 4th dose vaccination with ChAdOx1-S vaccine and after 3 months post-5th dose with Comirnaty Bivalent Original/Omicron BA.4/BA.5 vaccine.  1 Omicron sublineages' spike region.Amino acid mutations and deletions (Δ) are indicated in reference to the Wuhan-1 (614D) strain.Only distinct mutations between these variants were represented.NTD, N-terminal domain of the spike glycoprotein; RBD, receptor binding domain of the spike glycoprotein.Mutations were determined by the Outbreak.infogenomic reports website [14].

Ethical Statement
We conducted a prospective cohort study of older adults (age > 60 years) receiving the pandemic COVID-19 vaccine between November 2022 and June 2023 at Dom Pedro II Geriatric and Convalescent Hospital, Irmandade da Santa Casa de Misericórdia de São Paulo, in São Paulo, Brazil.All subjects or a legally responsible person provided written informed consent before inclusion in the study, which was approved by the institutional ethics committee from the University of São Paulo and Dom Pedro II Geriatric and Convalescent Hospital (CEPSH-ICB/USP; CAAE: 66951221.2.0000.5467and 69479523.2.0000.5467).The inclusion criteria were willingness to attend scheduled blood sampling visits and being vaccinated with all COVID-19 doses.The exclusion criteria were a history of anaphylaxis or hypersensitivity to vaccines and no vaccination record.The study was conducted in compliance with all International Council for Harmonisation Good Clinical Practice guidelines and the ethical principles of the Declaration of Helsinki.All SARS-CoV-2 experiments were performed in a Biosafety Level 3 (BSL-3) laboratory in the Institute of Biomedical Sciences (ICB) at the University of São Paulo (USP), São Paulo, Brazil.No diagnosis or treatment was involved.

Study Cohort and Sample Collection
Individuals included in this study have received five doses of the Coronavirus Disease 2019 (COVID-19) vaccine, following the regimen of two initial doses of ChAdOx1-S (Oxford/AstraZeneca) 4 weeks apart, followed by a booster with BNT162b2 (Pfizer-BioNTech) after 7 months post-primary vaccination.A fourth dose (second booster) of ChAdOx1-S was administered 11 months after the 3rd dose.Finally, five months after the 3rd dose, the volunteers were boosted with the 5th dose (third booster) of Comirnaty Bivalent Original/Omicron BA.4/BA.5 vaccine (Pfizer (New York, NY, USA)-BioNTech (Mainz, Germany)).Full cohort and demographic information are provided in Supplemental Tables S1 and S2.Blood samples were taken at three timepoints: 1 (median 36 days) and 4 months (median 126 days) post-4th dose with ChAdOx1-S vaccine, and after 3 months (median 91 days) post-5th dose with Comirnaty Bivalent Original/Omicron BA.4/BA.5 vaccine (Table S2).The serum was obtained by centrifuging the whole blood for 10 min at 4 • C and 2000 r.p.m. and stored at −20 • C until further use.

Cytopathic Effect-Based Virus Neutralization Test (CPE-VNT)
The CPE-VNT was performed following the protocol described by Wendel et al., 2020 [17].Briefly, 5 × 104 cells/mL of Vero cells (ATCC CCL-81) were seeded 24 h before the infection in a 96-well plate.Serum samples were initially inactivated for 30 min at 56 • C. We used 8 dilutions (two-fold) of each serum (1:20 to 1:2560).Subsequently, the serum was mixed vol/vol with 10 3 TCID 50 /mL of each SARS-CoV-2 variant and pre-incubated at 37 • C for 1 h to allow virus neutralization.Then, the serum plus virus mixture was transferred onto the confluent cell monolayer and incubated for 3 days at 37 • C, under 5% CO 2 .After 72 h, the plates were analyzed directly with the automated microscopy EVOS™ M5000 Imaging System (#AMF5000, Invitrogen TM , Waltham, MA, USA).In each assay, a convalescent serum sample was used as a positive control [18] and as a negative pre-pandemic serum sample (collected in 2017) [19].The sample dilution resulting in 50% of virus neutralization (NT 50 ) was calculated using the method of Spearman and Karber [20,21].Each sample was carried out in triplicate for the NT 50 calculation.Samples with no neutralization and observed CPE from the first dilution (1:20) were considered to have VNT 100 titers equal to 10, which is equivalent to half of the limit of detection (LOD).

Statistical Analysis and Reproducibility
Categorical variables were summarized as number (n) and percentage (%) and continuous variables as the geometric mean with 95% confidence intervals (95%CI) or median with interquartile ranges (IQRs).For plot purposes, the geometric mean titer (GMT) calculations, statistical analysis, and sera with undetectable (NT 50 < 18) nAbs titers were assigned an NT 50 of 14 (representing triplicates with a VNT 100 titer equal to 10).Comparison between neutralization titers was performed using a Wilcoxon matched-pairs signed-rank test or a nonparametric Mann-Whitney test, when applicable, using GraphPad Prism version 10.0.3 for macOS (GraphPad Software, Boston, MA, USA, www.graphpad.com).Absolute p values are provided in Tables S4 and S5.p < 0.05 was considered statistically significant.For the infection history analyses, two epidemiological Omicron wave periods in the São Paulo state, Brazil, were defined as follows: 1st Omicron wave by BA.1/BA.2(from January 2022 to May 2022) and 2nd Omicron wave by BA.4/BA.5 subvariants (from June 2022 to January 2023), as previously described by Hojo-Souza et al., 2023 [22] and based on the Fiocruz Genomics Network (https://www.genomahcov.fiocruz.br/dashboard-en/,accessed on 20 October 2023).Images were assembled using Adobe Illustrator, version 28.0.No statistical method was used to predetermine the sample size since the samples were collected based on the availability of volunteers in the Dom Pedro II Geriatric and Convalescent Hospital.No data were excluded from the analyses.

Influence of History of Infection by SARS-CoV-2 in nAbs Titers
We further investigated if older adults with a history of SARS-CoV-2 infection had higher neutralizing titers against the studied variants at the different time points of sample collection.We observed slightly higher titers in individuals with infection history but these were not statistically significant (p > 0.05) unless against BA.4/5 (GMTs:279 vs. 556, p = 0.0367) and EG.5.1 (GMTs: 31 vs. 49, p = 0.0346) at 4 months post-vaccination with the ChAdOx1-S vaccine (Figures 3A,B and S1).When analyzing the samples with infection history and separating them into two groups, those with a previous infection before the epidemiological wave of BA.4/BA.5 (until June 2022, mainly during the BA.1/BA.2wave) and those during the BA.4/5 wave (from June 2022 to January 2023), it was noticed that the difference against BA.4/5 (GMTs: 279 vs. 823, p = 0.0109) and EG.5.1 (GMTs: 31 vs. 55, p = 0.0328) was related to samples with infection history during the BA.4/5 wave (Figure 3A,B and Table S5).These results suggest that BA.1/BA.2 or BA.4/BA.5 breakthrough infection did not substantially increase the magnitude of nAbs' levels against the tested Omicron sublineages XBB.1.5,XBB.1.16,and EG.5.1.
Vaccines 2024, 12, x FOR PEER REVIEW 9 of 14  S5 for the complete data.

Discussion
The immune response in older adults after the COVID-19 vaccination remains understudied.In this study, we measured the levels of neutralizing antibodies (nAbs) against D614G, BA.4/5, XBB.1.5,XBB.1.16,and EG.5.1 viruses in a Brazilian cohort of older adults after vaccination with a 4th dose of the ChAdOx1-S vaccine and a 5th dose of the bivalent mRNA vaccine.We found higher levels of nAbs after the 5th dose and a substantial immune escape in sera from these individuals against newly emerged SARS-CoV-2 variants.
Vaccination with ChAdOx1 has been demonstrated to protect older adults [23,24].However, with the rapidly evolving of SARS-CoV-2, and mainly several Omicron sublineages (i.e., XBB) that exhibit enhanced fusogenicity and substantial immune evasion in the elderly population [25], the administration of a booster dose with mRNA bivalent vaccines targeting the ancestral and the BA.4/5 spike protein was recommended to generate immunity.Our data on Omicron neutralization following a 4th dose of ChAdOx1-S vaccine mirrors others who showed decreases in neutralization of >35-fold against XBB.1.5,XBB.1.16,and EG.5.1 [26,27].This difference is reduced by less than 10-fold against the same variants after 3 months of a booster shot with Comirnaty Bivalent Original/Omicron BA.4/BA.5 vaccine, as observed by others [27,28].In that way, it is clear from our data that boosting with a bivalent mRNA vaccine generates a much higher overall titer of neutralizing antibodies, enhancing cross-protection against high immune evasive variants such as XBB.1.5,XBB.1.16,and EG.5.1 [29].
Recent studies have demonstrated that vaccinated individuals with SARS-CoV-2 infection history develop a stronger humoral immune response against Omicron subvariants [12,30].However, the impact of the infection subvariant and the time between infection and sample collection might influence these observations.Here we showed that older adults with reported infection mainly during the first Omicron wave (BA.1/BA.2,from January 2022 to May 2022, São Paulo, Brazil) [22,31] did not present higher levels of nAbs against D614G, BA.4/5, XBB.1.5,XBB.1.16,or EG.5.1.Furthermore, individuals infected during or after the second Omicron wave (BA.4/5, from June 2022, São Paulo, Brazil) showed discretely higher titers against only BA.4/5 and EG.5.1 (p < 0.05).In the beginning of the Omicron era, a low cross-neutralization against Omicron from previously non-Omicron infection was proposed [32], reinforcing the importance of vaccination to mitigate Omicron dispersion.In that way, our data is similar to Yang et al., 2023, who observed low levels of neutralizing antibodies against XBB.1.5after BA.5 infection [33].Also, it has been shown that vaccinated individuals with BA.1, BA.2, BA.5, or BF.7 breakthrough infections had a low humoral response against XBB lineages [34,35].
Some limitations of this study warrant consideration.First, we have not examined either the binding antibody titers against spike and/or NP or the cellular immune response.Neutralizing antibodies together with the T-cell immune response protect patients from severe disease and death [36,37] and can provide different insights when analyzed together.Second, we were not able to investigate possible differences in other vaccine regimens since all the participants took the same vaccine doses.Third, while this study, to our knowledge, represents one of the largest cohort studies to date on the antibody response in older adults after COVID-19 vaccination, it still represents a relatively small sample size.Fourth, we were not able to collect blood samples just before the boost with the mRNA bivalent dose; in that way, we were not able to determine the nAbs baseline titers.Finally, we did not perform the sequencing of samples from the participants with infection history, because of sample unavailability.

Irmandade da Santa Casa de Misericórdia de São Paulo, in São Paulo, Brazil. All subjects or a legally responsible person provided written informed consent before inclusion in the study, which was approved by the
19 vaccine between November 2022 and June 2023 at Dom Pedro II Geriatric and Convalescent Hospital,

Table 1 .
Demographic and clinical characteristics of the older adults' cohort (n = 59).See TableS1for the complete data.