Protection of Omicron Bivalent Vaccine, Previous Infection, and Their Induced Neutralizing Antibodies Against Symptomatic Infection With Omicron XBB.1.16 and EG.5.1

Abstract Background Data are limited on the protective role of the Omicron BA bivalent vaccine, previous infection, and their induced neutralizing antibodies against Omicron XBB.1.16 and EG.5.1 infection. Methods We conducted a nested case-control analysis among tertiary hospital staff in Tokyo who had received ≥3 doses of COVID-19 vaccines and donated blood samples in June 2023 (1 month before the Omicron XBB.1.16 and EG.5.1 wave). We identified 206 symptomatic cases between June and September 2023 and selected their controls with 1:1 propensity score matching. We examined the association of vaccination, previous infection, and preinfection live virus neutralizing antibody titers against Omicron XBB.1.16 and EG.5.1 with the risk of COVID-19 infection. Results Previous infection during the Omicron BA- or XBB-dominant phase was associated with a significantly lower infection risk during the XBB.1.16 and EG.5.1–dominant phase than infection-naive status, with 70% and 100% protection, respectively, whereas Omicron BA bivalent vaccination showed no association. Preinfection neutralizing titers against XBB.1.16 and EG.5.1 were 39% (95% CI, 8%–60%) and 28% (95% CI, 8%–44%) lower in cases than matched controls. Neutralizing activity against XBB.1.16 and EG.5.1 was somewhat detectable in the sera of individuals with previous infection but barely detectable in those who were infection naive and received the Omicron bivalent vaccine. Conclusions In the era when the Omicron XBB vaccine was unavailable, the Omicron BA bivalent vaccine did not confer the neutralizing activity and protection against Omicron XBB.1.16 and EG.5.1 symptomatic infection. The previous infection afforded neutralizing titers and protection against symptomatic infection with these variants.

As of 2024, four years after the initial outbreak, the COVID-19 pandemic is still ongoing due to the persistent mutation cycle of SARS-CoV-2.In late 2020, clinical trials showed that COVID-19 vaccination was highly effective in lowering the risk of SARS-CoV-2 infection and severe outcomes [1,2].In late 2021, COVID-19 cases rapidly increased due to the Omicron BA subvariants among the vaccinated population.In 2022, an updated Omicron BA bivalent vaccine lowered the risk of Omicron BA infections [3].In early to mid-2023, the Omicron XBB subvariants (XBB.1.5,XBB.1.16,and EG.5), with multiple spike protein mutations as compared with earlier BA subvariants [4][5][6], dominated worldwide.Until the Omicron XBB vaccine became available in September 2023, people had to rely on immunity acquired by existing vaccines or prior infections against Omicron XBB subvariants.
In immunologic studies, Omicron bivalent BA vaccines and Omicron BA infection elicited neutralizing activity against XBB.1.5,XBB.1.16,and EG.5 subvariants, albeit to a limited extent [7][8][9].In epidemiologic studies, while the Omicron BA bivalent vaccine or previous Omicron BA infection was reported to confer moderate protection against the Omicron XBB.1.5infection [3,10,11], the evidence regarding the protection against Omicron XBB.1.16and EG.5 infection is limited.In a cohort study of 51 017 US health care workers, previous Omicron infection, not the Omicron bivalent vaccine, was associated with a lower risk of subsequent infection when Omicron XBB.1.16and EG.5 subvariants were dominant [12].Quantitative association between vaccine-or infection-acquired neutralizing activity against Omicron XBB.1.16and EG.5 and the risk of infection with these variants remains elusive.
In June 2023, a month before the Omicron XBB.1.16and EG.5.1 epidemic in Japan (July-September 2023), we performed a serologic survey among the staff of the National Center for Global Health and Medicine (NCGM), Tokyo, and stored blood samples.This situation prompted us to investigate whether the Omicron bivalent vaccine and previous infection could confer protection against Omicron XBB.1.16and EG.5.1 infection and whether its induced neutralizing antibody titers could correlate with infection protection.
Here, we examined the protection of the Omicron bivalent vaccine and previous infection against Omicron XBB.1.16and EG.5.1 infection and compared the live virus and preinfection neutralizing antibody titers between infected cases and controls in a nested case-control study of recipients with ≥3 doses of COVID-19 historical monovalent or Omicron BA bivalent vaccines.

Study Setting
A repeat serologic study was conducted at the NCGM in Japan in July 2020 to monitor the spread of SARS-CoV-2 infection among staff during the COVID-19 epidemic.The details of this study have been reported elsewhere [13][14][15].In summary, we have completed 8 serosurveys as of June 2023, where we measured anti-SARS-CoV-2 nucleocapsid (N) protein antibodies (all serosurveys) and spike protein antibodies (from the second serosurvey onward) for all participants using Abbott and Roche assays, stored serum samples at −80 °C, and collected information on COVID-19-related factors via a questionnaire (vaccination, occupational infection risk, infection prevention practices, behavioral factors, etc).The self-reported vaccination status was validated by objective information from the NCGM Labor Office.Written informed consent was obtained from all participants.This study was approved by the NCGM Ethics Committee (NCGM-G-003598).

Case-Control Selection
We conducted a nested case-control study among the staff who participated in the eighth survey in June 2023 and had received ≥3 doses of the mRNA COVID-19 vaccine manufactured by Pfizer or Moderna (any of the patterns of historical monovalent vaccine, Omicron BA.1 and wild type bivalent vaccine, and Omicron BA.4/5 and wild type bivalent vaccine; Supplementary Figure 1).Of the 2569 participants, 2409 received ≥3 doses of the mRNA COVID-19 vaccines and donated blood samples.Of those, we excluded 16 participants who lacked information on covariates: body mass index (n = 10), alcohol drinking status (n = 2), living arrangement status (n = 5), adherence to infection prevention practice (n = 3), and infection risk behaviors (n = 2).We further excluded 19 participants with insufficient volume of serum sample (<100 μL), leaving 2374 participants as the base population.
We followed the participants for COVID-19 incidence using the COVID-19 patient records documented by the NCGM Hospital Infection Prevention and Control Unit.Per the NCGM rule, staff should undergo polymerase chain reaction or antigen test for COVID-19 when they have COVID-19-compatible symptoms; if testing positive, they must report the results to the NCGM Hospital Infection Prevention and Control Unit.During the follow-up (June-September 2023), we identified 217 patients with COVID-19.We defined cases as symptomatic SARS-CoV-2 infection.Participants infected after additional vaccination during follow-up were considered cases if the infection occurred within 14 days after the vaccination, assuming they were not sufficiently immunized with the additional booster until then.After 11 asymptomatic patients were excluded, 206 were included as cases (Supplementary Figure 1).We selected a control for each case using propensity score matching.The details of the case-control matching algorithm are described in Supplementary Text 1.We randomly selected 50 matched pairs and measured live virus neutralizing antibody titers to compare neutralizing antibodies between the groups.

Antibody Testing
We measured neutralizing activity against wild type, Omicron XBB.1.16,and Omicron EG.5.1 in the sera of patients and controls by quantifying the serum-mediated suppression of the cytopathic effect of each SARS-CoV-2 strain in HeLa hACE2-TMPRSS2 cells [16,17].The details of the measurement methods are described in Supplementary Text 2.
We assessed anti-SARS-CoV-2 antibodies in all participants at baseline and retrieved data for the case-control pairs.We quantitatively measured the levels of antibodies against the receptorbinding domain (RBD) of the SARS-CoV-2 spike protein using the AdviseDx SARS-CoV-2 IgG II assay (Abbott; ie, anti-RBD immunoglobulin G [IgG]) and Elecsys Anti-SARS-CoV-2 S (Roche; ie, anti-RBD total).We chose the 2 anti-RBD assays with Abbott and Roche because of their established reputation and frequent application in the research field, making it easier to compare with other studies.We also qualitatively measured antibodies against the SARS-CoV-2 N protein using the SARS-CoV-2 IgG assay (Abbott) and Elecsys Anti-SARS-CoV-2 (Roche).

Previous Infection Status at Baseline
Previous infection was defined as a self-reported history of COVID-19 at baseline (confirmed against in-house COVID-19 registry) or anti-N seropositivity with either assay (Roche ≥1.0 cutoff index or Abbott ≥1.40 signal to cutoff) at any of the first through eighth surveys (July 2020-June 2023).We defined participants with no history of COVID-19 but seropositivity on N antibodies as undiagnosed infection [18].We defined phases of previous infection referring to the timing of the last diagnosis: pre-Omicron (predominated by wild type, Alpha, and Delta strains; February 2020-December 2021), Omicron BA (dominated by BA.1, BA.2, and BA.5 subvariants; January 2022-March 2024), and Omicron XBB (primarily XBB.1.5 and XBB.1.16subvariants; April 2024-June 2024).

Statistical Analysis
We used conditional logistic regression while accounting for the matched design to examine the association of vaccination status (doses of any COVID-19 vaccines and dose of Omicron BA vaccines) and previous infection status with SARS-CoV-2 infection risk.We used a generalized estimating equation (GEE) with group assignment (case or control) and a robust variance estimator to compare the interval from the last vaccination or COVID-19 diagnosis to baseline blood sampling.To examine the difference in preinfection antibody levels between cases and controls, we compared the log-transformed titers of neutralizing antibodies (wild type, Omicron XBB.1.16,and Omicron EG.5.1) and anti-RBD antibodies (IgG and total) between matched pairs using a GEE model with group assignment and a robust variance estimator.Then, we back-transformed and presented these values as geometric mean titers (GMTs) with 95% CIs.For a sensitivity analysis, we repeated the GEE analysis by restricting matched pairs to infection-naive pairs at baseline (ie, case and matched controls had no history of COVID-19 and were negative on anti-N assays).We used the Kruskal-Wallis test to compare the neutralizing titers across vaccination status (historical monovalent vaccine only or the historical monovalent plus Omicron bivalent vaccines) and previous infection status (infection naive or previously infected).
To examine the difference in neutralizing titers across the timing of previous infection, we used a linear regression model while adjusting age, sex, a history of Omicron bivalent vaccination, and the interval between the last vaccination and blood sampling.For the analyses of neutralizing antibody titers, values below the limit of detection (50% neutralization titer [NT 50 ] <40) were given the limit of detection value.Statistical analyses were performed with Stata version 18.0 (StataCorp), and graphics were generated with Prism version 9 (GraphPad).All P values were 2-sided, and statistical significance was set at P < .05.

Baseline Characteristics Before and After Propensity Score Matching
We ascertained 206 symptomatic breakthrough infection cases during the follow-up in the before-matching cohort, with an incidence rate of 12.7 per 10 000 person-days.Cases were younger and more likely to be female and nurses than the controls in the before-matching cohort (Table 1).After propensity matching with a 1:1 ratio, the 206 matched pairs were well balanced regarding all baseline characteristics.
Previous infection at the Omicron BA or XBB phase, but not the pre-Omicron phase, was significantly associated with a lower risk of SARS-CoV-2 infection during the follow-up (Table 2).When compared with infection-naive status, the ORs (95% CI) of previous infection at the pre-Omicron and Omicron BA waves against infection were 0.20 (.03-1.21) and 0.30 (.18-.50), respectively.No SARS-CoV-2 infection occurred in the group of previous infection at the Omicron XBB Comorbid diseases were defined as cancer, cardiovascular disease, diabetes, hypertension, chronic kidney disease, or lung disease.c Immunosuppression was defined as having an immunosuppressive disease or using steroids (except topical or inhaled), immunosuppressants, or anticancer drugs.d Tobacco products include conventional cigarettes and heated tobacco products.e School-age children include those in nurseries, kindergartens, elementary to high school, and university and those with disabilities.f Infection prevention practice score was calculated per the total score of adherences to avoiding the 3Cs, hand washing, wearing a mask, and social distancing, as well as not touching the face, nose, or mouth, by assigning 2 points to always, 1 to often, and 0 to others (seldom and not at all).
wave (OR, 0.00).Undiagnosed infection was also associated with a lower risk of infection than infection-naive status (OR, 0.27; 95% CI, .13-.55).Among the participants with a history of COVID-19, the interval between the last diagnosed infection and baseline blood sampling was statistically longer in cases than controls, with a mean difference of 81 days (95% CI, 25-137).
The sensitivity analyses restricted to infection-naive matched pairs indicated that the difference in preinfection anti-RBD and neutralizing antibody titers between cases and controls was attenuated and no longer statistically significant.No infectionnaive cases detected neutralization against Omicron XBB.1.16and EG.5.1.

Neutralizing Antibody Titers Across Statuses of Omicron Bivalent Vaccination and Previous Infection
There were no substantial differences in preinfection neutralizing antibody titers against wild type, Omicron XBB.1.16,and Omicron EG.5.1 among individuals irrespective of their history of Omicron bivalent vaccination (Figure 3).Neutralization against Omicron XBB.1.16and EG.5.1 was not detected (<40 NT 50 ) in all sera from individuals who were infection naive and had never received the Omicron bivalent vaccine.Among those who were infection naive with a history of Omicron bivalent vaccine, only 21% and 4% had detectable neutralizing titers against XBB.1.16and EG.5.1, respectively.
Participants previously infected had higher preinfection neutralizing antibody titers against wild type, Omicron XBB.1.16,and Omicron EG.5.1 than those who were infection naive (Figure 3).Irrespective of the previous infection phases, wild type neutralizing titers were higher in those with previous infection than those who were infection naive (Figure 4).Those infected during the Omicron XBB periods had the highest neutralizing antibody titers against Omicron XBB.1.16and EG.5.1 as compared with those infected in the Omicron BA period or earlier.Those infected in the Omicron BA period had statistically higher neutralizing titers against Omicron XBB.1.16than those who were infection naive but not in titers against Omicron EG.5.1.

DISCUSSION
From June to September 2023, when Omicron XBB.1.16and EG.5.1 were predominantly circulating in Japan, previous infection during the Omicron BA and XBB waves was associated with 70% and 100% lower risk of subsequent symptomatic SARS-CoV-2 infection, respectively, while the Omicron BA bivalent vaccination was not associated with the risk of infection in a cohort of health care workers with ≥3 doses of vaccination.
The preinfection neutralizing antibody titers against Omicron XBB.1.16and EG.5.1 were lower in infected cases than in matched controls.Although the evidence is scarce regarding the protection of previous infections and the bivalent vaccine against Omicron XBB.1.16and EG.5.1 infection, our findings were similar to a study of US health care workers [12].In that study, previous Omicron infection was associated with a 60% lower risk of subsequent infection during the dominant waves of Omicron BA.4/5, BQ, or XBB subvariants, whereas the Omicron BA bivalent vaccination was not associated with the risk of infection during the dominant wave of the XBB subvariant.Regarding the protection of previous infection, our study has some strengths over the previous one.We classified previous Omicron infection by the subvariant-specific dominant phase (BA or XBB).In addition, we followed subsequent infections restricted to the dominant phase of XBB.1.16and EG.5.1.These gaps allowed us to estimate the Omicron BA-and XBB-specific protection against subsequent infections during the XBB.1.16-andEG.5.1-specific waves, and we found that previous infection during the Omicron XBB wave had superior protection against infection during the XBB.1.16-andEG.5.1dominant phase to those during the Omicron BA wave (100% vs 70% protection).
In spite of much lower preinfection neutralizing titers against XBB.1.16and EG.5.1 than those against wild type, we found that higher neutralizing antibody titers against Omicron XBB.1.16and EG.5.1 were associated with a lower risk of SARS-CoV-2 infection when these variants were dominant.Our findings suggest that variant-specific neutralizing antibody titers could correlate with protection against infection with its variant, even within the low titers range.We also found that the prevalence of those with hybrid immunity (vaccination and previous infection) was lower in cases than controls (25% vs 60%) and that individuals with hybrid immunity had higher neutralizing titers against XBB.1.16and EG.5.1 than those who were infection naive.Similarly, previous studies reported that neutralizing titers against XBB.1.16and EG.history of infections than those without [8,9,19].These results confirm that the evidence that hybrid immunity confers better protective humoral immunity than vaccination alone, which has been recognized for the risk of infection with Omicron BA or earlier variants [20,21], can be extended to the risk of XBB.1.16and EG.5.1 infection.This study had several strengths.We rigorously matched cases and controls using a propensity score estimated by several factors potentially associated with SARS-CoV-2 infection risk, including occupational SARS-CoV-2 exposure risk, living arrangements, comorbidities, infection prevention practices, and high-infection risk behaviors.Blood samples for antibody testing were obtained before infection (1 month before the Omicron XBB.1.16and EG.5 epidemic onset).Previous SARS-CoV-2 infection was determined according to the history of COVID-19 diagnosis and results of anti-SARS-CoV-2 N assays, allowing us to identify undiagnosed infections.We measured the neutralizing antibody titers using live viruses.However, limitations should be acknowledged.We defined cases as patients with physician-diagnosed symptomatic COVID-19.Since we did not conduct active surveillance to detect SARS-CoV-2 infection during the follow-up period, the present results may not apply to asymptomatic SARS-CoV-2 infection.Data on virus strain was not available for the present cases; yet, the cases were most likely due to the Omicron XBB variant (including XBB.1.16and its descendent EG.5), which accounted for >90% of sequenced COVID-19 samples in Japan during the follow-up (June-September 2023; Figure 1).

CONCLUSION
In the era when Omicron XBB.1.16and EG.5.1 variants were predominant and the Omicron XBB vaccine was still unavailable in Japan, previous Omicron BA or XBB infection, not Omicron bivalent vaccination, was associated with a lower risk of symptomatic SARS-CoV-2 infection.The preinfection and live virus neutralizing antibody titers against Omicron XBB.1.16and EG.5.1 were lower in infected cases than in their matched controls.Those with a history of Omicron BA bivalent vaccine had barely detectable neutralizing titers against these variants.Our results highlight the importance of infection prevention practices when the circulating variants had high immune evasion from immunity acquired by existing vaccines.

Figure 1 .
Figure 1.Relative frequency of circulating SARS-CoV-2 variants from June to September 2023 in Japan.The distribution of SARS-CoV-2 lineages in Japan during the study period (June-September 2023) was analyzed by using all domestic genome sequences registered in the GISAID EpiCov database (https://gisaid.org).For lineage analysis, the extracted sequences (n = 27 899) were applied to Phylogenetic Assignment of Named Global Outbreak Lineages version 4.3.1 with pangolin-data 1.25.1.

d
GMT ratio for cases to controls, estimated by the generalized estimating equation model.e Analyzed among 50 matched pairs randomly selected from 206 matched pairs.f Analyzed among 57 infection-naive pairs.g Analyzed among 12 infection-naive pairs.Protection Against XBB.1.16/EG.5 Infection • OFID • 5

Figure 2 .
Figure 2.Comparison of the preinfection live virus neutralizing and anti-RBD antibody titers between propensity score-matched cases and controls.A and B, Among 206 cases with breakthrough and 206 matched controls, anti-RBD total antibody titers were measured with Roche reagent, and anti-RBD IgG antibody titers were measured with the Abbott reagent.C-E, In addition, the live virus neutralizing antibody titers against wild type, Omicron XBB.1.16,and Omicron EG.5.1 among the 50 matched pairs were randomly selected from 206 matched pairs.In each panel, the horizontal bars indicate the geometric mean titers, and the I-shaped bars indicate the geometric SD.The limit of detection of the neutralizing assay is 40, as shown by dashed horizontal lines.The frequency (percentage) of the upper limit of detection is denoted above the x-axis.P values were calculated via the generalized estimating equation model.AU, arbitrary units; IgG, immunoglobulin G; NT 50 , 50% neutralizing titer; RBD, receptorbinding domain.

Figure 3 .
Figure 3.Comparison of the preinfection live virus neutralizing antibody titers across the histories of Omicron bivalent vaccine and previous SARS-CoV-2 infection.A-C, Live virus neutralizing antibody titers against wild type, Omicron XBB.1.16,and Omicron EG.5.1 among the 100 samples from 50 matched pairs.In each panel, the horizontal bars indicate the geometric mean titers, and the I-shaped bars indicate the geometric SD.The limit of detection of the neutralizing assay is 40, as shown by dashed horizontal lines.The frequency (percentage) of the upper limit of detection is denoted above the x-axis.P values were calculated via the Kruskal-Wallis test.NT 50 , 50% neutralizing titer; RBD, receptor-binding domain.

Table 1 . Baseline Characteristics Before and After Propensity Score Matching
Data are presented as mean ± SD for continuous variables and percentage for categorical variables.An absolute standardized difference <0.10 indicates a relatively small imbalance.Abbreviation: 3Cs, crowded places, close-contact settings, and confined and enclosed spaces.a Occupational SARS-CoV-2 exposure risk was categorized as low (those not engaged in COVID-19-related work), moderate (those engaged in COVID-19-related work without heavy exposure to SARS-CoV-2), or high (those heavily exposed to SARS-CoV-2).b

Table 2 . Association of Vaccination Status, Previous Infection Status, and Preinfection Antibody Titers With the Risk of Symptomatic SARS-CoV-2 Infection
Analyzed among those with a history of COVID-19 diagnosis (case/control: n = 37/91).
Abbreviations: AU, arbitrary units; GMT, geometric mean titer; NA, not applicable; NT 50 , 50% neutralization titer.a Odds ratio of SARS-CoV-2 infection across exposure groups, estimated by the conditional logistic regression model.b Mean difference between cases and controls, estimated by the generalized estimating equation model.c