Neutralizing antibodies after three doses of the BNT162b2 vaccine, breakthrough infection, and symptoms during the Omicron-predominant wave

Abstract

Objectives

To investigate the role of immunogenicity after the third vaccine dose against Omicron infection and COVID-19-compatible symptoms of infection.

Methods

First, we examined vaccine effectiveness (VE) of the third dose against the second dose during the Omicron wave among the staff at a tertiary hospital in Tokyo. In a case-control study of third vaccine recipients, we compared the preinfection live-virus neutralizing antibodies (NAb) against Omicron between breakthrough cases and their controls who had close contact with patients with COVID-19. Among these cases, we examined the association between NAb levels and the number of COVID-19-compatible symptoms.

Results

Among the 1456 participants for VE analysis, 60 breakthrough infections occurred during the Omicron wave. The third dose VE for infection was 54.6%. Among the third dose recipients, NAb levels against Omicron did not differ between the cases (n = 22) and controls (n = 21). Among the cases, those who experienced COVID-19-compatible symptoms had lower NAb levels against Omicron than those who did not.

Conclusion

The third vaccine dose was effective in decreasing the risk of SARS-CoV-2 infection during Omicron wave compared with the second dose. Among third dose recipients, higher preinfection NAb levels may not be associated with a lower risk of Omicron infection. Contrarily, they may be associated with fewer symptoms of infection.

Introduction

During the ongoing COVID-19 pandemic, the waning of the second vaccine-induced immunogenicity over time [1,2] and the emergence of variants of concerns, such as Delta and Omicron, have led many countries to adapt the booster (third) vaccine campaign. Observational studies have shown that a third dose of existing messenger RNA (mRNA) vaccines is still effective against the infection with Delta and Omicron variants and hospitalization [3], [4], [5]; however, the vaccine effectiveness (VE) of the third dose against Omicron infection is lower than that against Delta [3,4]. The levels of vaccine-induced neutralizing antibodies (NAbs) against Omicron were much lower than those against Delta and former SARS-CoV-2 strains, even after the third dose [6,7], and such a low level of variant-specific immunogenicity may contribute to poor protection against Omicron infection.

Epidemiological evidence on the association between preinfection humoral immunity and the risk of breakthrough infection during the Wuhan (wild-type) Alpha- and Delta-dominant waves is inconsistent [8]. Several studies have shown an inverse relationship between vaccine-induced antibody levels and infection risk, whereas others have suggested that infection occurs irrespective of the antibody levels. Data regarding the association between vaccine-induced immunogenicity and infection risk during the Omicron epidemic are limited [9], which has increased transmissibility and high immune evasion. VE studies have reported a higher ratio of asymptomatic patients to overall infected patients in vaccinated patients with COVID-19 than in unvaccinated patients [10,11], suggesting a suppressive role of the vaccine against symptoms of infection. It is of interest to determine whether preinfection NAb levels are a determinant of symptomatic episodes during breakthrough infection.

The aim of the current study was to investigate the role of immunogenicity after the third vaccine dose against Omicron infection and COVID-19-compatible symptoms. First, we examined the VE of the third dose relative to the second dose during the Omicron dominant wave among the staff of a tertiary hospital in Tokyo. In a case-control study nested in a cohort of third vaccine recipients, we compared the preinfection live-virus NAb levels against Omicron after the third dose between breakthrough cases and their controls who were in close contact with patients with COVID-19. In addition, we examined the association between preinfection NAb and the number of COVID-19-compatible symptoms experienced during the Omicron wave.

Methods

Study setting

In the National Center for Global Health and Medicine (NCGM), Japan, a repeat serological study was launched in July 2020 to monitor the spread of SARS-CoV-2 infection among the staff during the COVID-19 pandemic. The details of the study have been reported elsewhere [12]. As of March 2022, we have completed five surveys, each of which measured the levels of anti-SARS-CoV-2 nucleocapsid and spike (from the second survey onward) protein antibodies for all participants using both the Abbott and Roche assays; the serum samples were stored at -80°C, and information on COVID-19-related factors (vaccination, occupational infection risk, infection-prevention practices, history of close contact with patients with COVID-19, COVID-19-compatible symptoms, and so on) were collected using questionnaires. Self-reported vaccination status was validated using the information provided by the NCGM Labor Office, if any. We identified COVID-19 cases among study participants from the records of patients with COVID-19 maintained by the NCGM Hospital Infection-Prevention and Control Unit, which provided information on the date of diagnosis, diagnostic procedure, possible route of infection (close contact), symptoms, hospitalization, return to work for all cases, and virus strain and cycle threshold values for those diagnosed at the NCGM. Written informed consent was obtained from all participants. This study was approved by the NCGM ethics committee (approval number: NCGM-G-003598).

Study for the effectiveness of the third vaccine dose relative to the second dose

To examine the VE of the third versus second vaccine doses during the Omicron BA.1 dominant wave, we analyzed the data of NCGM staff who participated in the fifth survey (March 2022), for whom detailed information on vaccine status was available. The outcome was polymerase chain reaction- or antigen-confirmed COVID-19, which was reported in the in-house registry. In Japan, the Omicron BA.1-2 dominant wave began in January 2022. Thus, we started the follow-up on January 1, 2022 (baseline) and censored the date of COVID-19 diagnosis or the attendance date to the fifth survey (March 1-9, 2022), whichever came first.

Of the 1578 participants, we excluded those who received less than two vaccine doses (n = 30) at baseline (i.e., January 1, 2022), received non-mRNA-based COVID-19 vaccines (n = 49), had a history of COVID-19 before the start of the follow-up (n = 21), and tested positive on anti-SARS-CoV-2 nucleocapsid protein assays (positive with Abbott and/or Roche) in any previous survey (i.e., first to fourth surveys; n = 41), leaving 1456 who were included in the analysis.

A nested case-control study among third vaccine dose recipients

We conducted a case-control study among the staff who participated in the fourth survey in December 2021, after receiving the third dose (Figure 1 ). Of the 948 participants, 224 received the third vaccine of BNT162b2 at least 7 days before the fourth survey and donated a blood sample. Among these participants, those with a history of COVID-19 (n = 3) and those who tested positive with the anti-SARS-CoV-2 nucleocapsid protein antibody assay (Abbott and/or Roche) in any previous survey (n = 3) were excluded. The remaining 218 infection-naïve participants formed the basis of this case-control study and attended the fifth survey (follow-up) conducted in March 2022.

Figure 1.

Flowchart for the case-control selection.

In this cohort, 22 breakthrough infections were identified during the follow-up. Of these, 10 cases were ascertained using an in-house registry: nine tested positive by polymerase chain reaction testing and one by antigen test. The remaining 12 patients were identified using anti-SARS-CoV-2 nucleocapsid protein assays (Abbott and/or Roche) during the follow-up survey. Five of ten cases confirmed using the in-house registry were sequenced and confirmed to be infected with the Omicron BA.1 variant. Of the remaining 196 participants without evidence of COVID-19, 21 participants reported that they had close contact with patients with COVID-19 during the follow-up and served as controls.

The NAb against Wuhan, Omicron BA.1, and BA.2 after third vaccine and the preinfection sera was determined by quantifying serum-mediated suppression of the cytopathic effect of each SARS-CoV-2 strain in VeroE6TMPRSS2 cells [13,14]. Details of the measurement methods are described in Supplemental Text 1.

We assessed the anti-SARS-CoV-2 antibodies for all participants at the baseline and follow-up and retrieved the data for the case-control subsets. We quantitatively measured the levels of antibodies against the receptor-binding domain of the SARS-CoV-2 spike protein using the AdviseDx SARS-CoV-2 immunoglobulin (Ig)G II assay (Abbott) and Elecsys® Anti-SARS-CoV-2 S RUO (Roche) (including IgG).

We also qualitatively measured antibodies against the SARS-CoV-2 nucleocapsid protein using the SARS-CoV-2 IgG assay (Abbott) and Elecsys® Anti-SARS-CoV-2 RUO (Roche) and used these data to exclude those with possible infection before the baseline survey and to identify those who experienced breakthrough infection during the follow-up period.

Symptoms indicative of COVID-19

To detect COVID-19-compatible symptoms among breakthrough infection cases, we collected information on symptoms using a questionnaire during the follow-up survey. In the questionnaire, we asked participants about the following symptoms since January 1, 2022: (i) common cold-like symptoms (sore throat, cough, runny nose, nasal congestion, fever, and fatigue.) persisting for 4 or more days; (ii) common cold-like symptoms persisting for less than 4 days; (iii) high fever; (iv) severe fatigue; (v) dyspnea; (vi) loss of sense of taste or smell; and (vii) sore throat. In the questionnaire, we highlighted that adverse reactions after vaccination should not be included. For the analysis, the responses to two questions regarding common cold-like symptoms were combined and categorized into three groups: no symptoms, persists for <4 days, and persists for ≥4 days. We counted five specific symptoms: high fever, severe fatigue, dyspnea, taste and smell disorders, and sore throat and classified them into three categories: 0, 1-2, and 3-5. We did not inquire about the timing of the symptom onset during the follow-up. For cases notified using the in-hospital registry, we confirmed that the self-reported symptoms in the follow-up survey were similar to those at COVID-19 diagnosis recorded in the registry.

Statistical analysis

In the study on the VE of the third vaccine dose relative to the second dose, we fitted a Cox proportional hazards model with adjustment for age and sex. We treated vaccination status as a time-dependent covariate. If the participants received the third vaccine dose during the follow-up period, the vaccination status was changed from two to three doses 14 days after the date of receiving the third dose. The estimated hazard ratio with a 95% confidence interval (CI) for vaccination status was used to calculate VE (%) according to the following formula: VE = (1 - adjusted hazard ratio) × 100.

In the nested case-control study among third dose vaccine recipients, we compared the characteristics between cases and controls using the Mann-Whitney U test or Fisher's exact test. To examine the difference in the humoral response after the third vaccination between cases and controls, we compared the log-transformed titers of live-virus NAb (Wuhan, Omicron BA.1, and Omicron BA.2) and antispike antibodies (Abbott and Roche) using a multivariable linear regression model with adjustment for age (continuous) and sex (male or female). We also constructed a multivariate linear regression model to examine the association between preinfection NAb titers after the third vaccination and COVID-19-compatible symptoms during the study period among the cases. To compare the interindividual differences in NAb titers against Wuhan, Omicron BA.1, and Omicron BA.2 between cases and controls, we used a generalized estimating equation with a robust variance estimator. The estimated effects of covariates were back-transformed and presented as ratios of geometric means and geometric mean titers (GMTs) with 95% CIs. For the sensitivity analysis, we compared preinfection NAb and antispike antibody titers between cases and controls, restricting those who experienced close contact at home, where virus transmission would be higher than in other settings.

For the analyses, the values below or above the limit of detection for NAb titers (NT₅₀ <40) and spike antibody titers with the Roche assay (titer >25,000 U/ml) were given the limit of detection value. Statistical analyses were performed with Stata version 17.0 (StataCorp LLC), and the graphics were prepared with GraphPad Prism 9 (GraphPad, Inc.). All P-values were two-sided, and statistical significance was set at P <0.05.

Results

Effectiveness of the third dose relative to the second dose

Of 1456 infection-naïve participants at baseline, 863 and 593 had completed two and three vaccine doses, respectively. Of the two-dose vaccine recipients, 782 received the third dose during the follow-up. The person-days of the two- and three-dose groups were 14,429 and 73,645, respectively. Of the total participants, 60 (4%) SARS-CoV-2 breakthrough infections occurred between January 1 and March 9, 2022 (Table 1 and Figure S1). Regarding the variant types, 33 (55%) were identified as Omicron, and 27 were unknown (unmeasured). The median age of the patients was 39 years (interquartile range [IQR], 28-45 years), and 80% were females. At the beginning of the follow-up, the median intervals since the last vaccination in the two- and three-dose groups were 261 (IQR: 233-267) and 15 (IQR: 15-19) days, respectively. A total of 14 recipients who received two doses were infected at a median of 218 (IQR: 160-255) days after the second vaccination, whereas 46 cases among the three-dose vaccine recipients were infected at a median of 55 (IQR: 43-61) days after the third vaccination. Regarding the symptoms during infection identified from the in-house registry, all patients were asymptomatic (n = 10, 17%) or had mild symptoms (n = 50, 83%). Three patients with mild symptoms (6%) were admitted to a hospital. The proportion of asymptomatic patients of the overall infected patients was higher in patients who received three doses (20%) than in those who received two doses (7%). The infection rates per 10,000 person-days for the two- and three-dose vaccines were 9.7 and 6.3, respectively, and the age- and sex-adjusted VE of the three doses against two doses for infection was 54.6% (95% CI: 14.0-76.0).

Table 1.

Vaccine effectiveness of three vs second doses of vaccine for COVID-19 infection among 1456 staff of a large referral medical and research institution in Tokyo during the Omicron wave (January 2022 to March 2022).

Vaccination status	Total		COVID-19 cases			Person-days	Incident rate per 10000 person-days	Adjusted VE, % (95% confidence interval)
	No.	Interval from last vaccination to tracking start date, median days (IQR)	No.	Interval from last vaccination to infection, median days (IQR)	Asymptomatic cases, n (%)
Two doses	863	261 (233-267)	14	218 (160-255)	1 (7)	14429	9.7	reference
Three doses	1375	15 (15-19)	46	55 (43-61)	9 (20)	73645	6.3	54.6 (14.0-76.0)

IQR, interquartile range; VE, vaccine effectiveness.

VE was calculated as ([1-hazard ratio]×100), estimated using a time-dependent Cox proportional hazard regression model with adjustment for age and sex.

In the two-dose group (n = 863), 782 received the third dose during the follow-up without infection and moved to the three-dose group.

Characteristics of breakthrough infection in the case-control analysis

Of the 22 patients with breakthrough infections after the third vaccine dose, 86% were female, with a median age of 31.8 (IQR: 26.6-38.4) years, similar to those of the total study population (80% were female, and the median age was 39 [IQR: 28-45] years). All cases were nurses, and 23% were affiliated with the COVID-19-related department (Table 2 ). The majority (91-100%) showed good adherence to infection-prevention practices. In the follow-up survey, 36% reported having spent 30 minutes or more in crowded places, close contact settings, and confined and enclosed spaces (the 3Cs) without a mask, whereas 27% reported having dined in a group of five or more people for more than 1 hour. In the follow-up survey, 23% of respondents reported close contact with patients with COVID-19 in their households, 5% in the community, and 5% in the workplace. These figures were similar between cases and controls, except for those with close contact; close contact in the workplace setting was more frequent in controls than in cases (57% vs 5%).

Table 2.

Characteristics of the participants in the case-control study.

	Cases (n = 22)	Controls (N = 21)	P-value
Age	31.8 (26.6-38.4)	34.7 (29.3-38.7)	0.66
Female	19 (86)	17 (81)	0.70
Job			0.23
Doctors	0 (0)	2 (10)
Nurses	22 (100)	19 (90)
Department
COVID-19-related department	5 (23)	4 (19)	1.00
Non-COVID-19-related department	17 (77)	17 (81)
Adherence to infection-prevention practices during the follow-upa,b
Keeping social distance	21 (95)	20 (95)	1.00
Wearing a mask	22 (100)	21 (100)	-
Not touching eyes, nose, and mouth	20 (91)	20 (95)	1.00
Washing or sanitizing hands	22 (100)	21 (100)	-
High-risk behaviors during the follow-upa
Spending ≥30 minutes in the 3Cs without mask (≥1 time)	8 (36)	5 (24)	0.51
Dinner in a group of ≥5 people for >1 hour (≥1 time)	6 (27)	2 (10)	0.24
Close contact with COVID-19 patients during the follow-upa
Household	5 (23)	5 (24)	1.00
Community	1 (5)	2 (10)	0.61
Workplace	1 (5)	12 (57)	<0.01
Others	0 (0)	2 (10)	0.23
Unknown	15 (68)	0 (0)	<0.01
Interval between the second and third doses	239 (236-242)	239 (238-243)	0.56
Interval between the third dose and blood sampling	10 (9-11)	11 (10-12)	0.09
Interval between the third dose and infectionc	58 (52-62)		-
Interval between blood sampling and infectionc	47 (44-61)		-
Type of variantsc			-
Omicron BA.1	5 (50)
Unknown (unmeasured)	5 (50)

Data are presented as median (interquartile range) for continuous measures and n (%) for categorical measures.

Abbreviations: 3Cs, crowded places, close contact settings, and confined and enclosed spaces.

^aThe information was collected using a questionnaire at the follow-up survey and represents the context during the follow-up period.

^bIn each question related to infection-prevention practices, participants’ responses were categorized using a 4-point Likert scale: “always”, “often”, “seldom”, and “not at all”, with the first two response options defined as good adherence to infection-prevention practices. These P-values represent the results of comparison between always or often and seldom or not at all.

^cThe denominator is the number of COVID-19 patients ascertained using the in-house registry (n = 10).

As for the 10 cases that were notified from the in-house COVID-19 registry, the median intervals from the third vaccine to the infection and the blood sampling to the infection were 58 (IQR: 52-62) and 47 (IQR: 44-61) days, respectively. Of these, 10% were asymptomatic, and 90% had mild symptoms. None of the patients required hospital admissions. Of the 12 cases identified solely by antibody test, 42% reported some COVID-19-compatible symptoms during the follow-up: sore throat (n = 5), common cold-like symptoms (n = 4), severe fatigue (n = 2), and dyspnea (n = 1).

Neutralizing and antispike antibodies in cases and controls

The median (IQR) intervals between the third dose and blood sampling were 10 (9-11) and 11 (10-12) days for cases and controls, respectively. There were no statistical differences between the cases and controls in the preinfection levels of NAb against Wuhan, Omicron BA.1, or Omicron BA.2 or antispike antibodies (Figure 2 and Supplementary Table S1). The GMT (NT₅₀) of NAb against Wuhan, predicted by the linear regression with adjustment for age and sex, was 733 (95% CI, 552-973) for cases and 613 (95% CI, 407-923) for controls; against Omicron BA.1 was 221 (95% CI, 141-349) for cases and 223 (95% CI, 129-383) for controls; against Omicron BA.2 was 102 (95% CI, 72-143) for cases and 108 (95% CI, 75-155) for controls. The results were the same among individuals who had close contact with family members with COVID-19 (Supplementary Table S2).

Figure 2.

Preinfection neutralizing and spike antibody titers after the third vaccine dose among cases and controls who experienced close contact with COVID-19 patients.

Among the 22 SARS-CoV-2 breakthrough infection cases and the 21 controls who experienced close contact with COVID-19 patients, neutralizing antibody titers against the Wuhan (A), Omicron BA.1 (B), and Omicron BA.2 (C) strains after the third vaccine dose during the preinfection period (median of 10 days since the third vaccination) are shown. Results of comparison of post-vaccination antispike antibody titers measured using the Abbott reagent (D) and those using the Roche reagent (E) in the two groups. In each panel, the horizontal bars indicate the geometric mean titers, and the I-shaped bars indicate geometric standard deviations. The dashed horizontal lines indicate the limits of detection in the present analysis (NT₅₀ <40 in neutralizing antibody assays and >25,000 U/ml in Roche assay).

P-values were calculated using a multivariable linear regression model with adjustment for age and sex.

Abbreviations: AU, arbitrary units; ns, not significant; NT, neutralizing titer.

Cross-reactive NAbs

NAb titers against Omicron BA.1 and Omicron BA.2 were much lower than those against Wuhan (Figure 3 ). Among the cases, the NAb titers against Omicron BA.1 and Omicron BA.2 were 3.3-fold and 7.2-fold lower than those against Wuhan, respectively. Those against Omicron BA.2 were 2.2-fold lower than those against Omicron BA.1. Similar results were obtained in the control group.

Figure 3.

Cross-reactive neutralizing antibody titers after the third vaccine dose among cases (n = 22) or controls (n = 21).

Results of comparison of preinfection neutralizing antibody titers against the Wuhan, Omicron BA.1, and Omicron BA.2 strains among cases or controls. The bars indicate geometric mean titers, and I-shaped bars indicate their geometric standard deviations. The dashed horizontal lines indicate the limits of detection (NT₅₀ titer <40). The fold-change values are estimated ratios of geometric means for antibody titers based on the generalized estimating equation model (ns: not significant; *P <0.05; **P <0.01; ***P <0.001).

Abbreviations: NT₅₀, 50% neutralization titer.

NAbs and symptoms in patients in the case-control study

Among cases identified in the case-control analysis, those who experienced some COVID-19-compatible symptoms during the study period had lower, albeit not statistically significant, preinfection NAb levels against the Omicron BA.1 and BA.2 than those who did not experience any symptoms (Table 3 ). Those who experienced common cold-like symptoms had significantly lower preinfection NAb titers against Omicron BA.1 than those who did not have these symptoms. Compared with those who did not experience common cold-like symptoms, the GMT (95% CIs) ratio of NAb against the Omicron BA.1 for those who had the symptoms for less than 4 days and for 4 or more days was 0.81 (0.04-15.1) and 0.39 (0.20-0.77), respectively (P for trend <0.01). In addition, a larger number of symptoms was associated with lower preinfection NAb titers against Omicron BA.1. Although not significant, similar associations of common cold-like symptoms and the number of symptoms were observed with NAb titers against Omicron BA.2. High fever was significantly associated with lower NAb titers against Omicron BA.2. The loss of taste or smell was associated with lower Omicron BA.1 titers.

Table 3.

Association between preinfection NAb titers after the third vaccine dose and symptoms among breakthrough infection cases.

Symptoms	N	NAb against Wuhan	NAb against Omicron BA.1	NAb against Omicron BA.2
		Ratio of mean (95% CI)	Ratio of mean (95% CI)	Ratio of mean (95% CI)
Any of symptoms
Asymptomatic	8	reference	reference	reference
Symptomatic	14	1.22 (0.72-2.07)	0.54 (0.26-1.12)	0.63 (0.35-1.13)
Common cold-like symptoms
No	9	reference	reference	reference
Persists <4 days	2	2.93 (0.99-8.67)	0.81 (0.04-15.1)	0.93 (0.43-2.02)
Persists ≥4 days	11	1.12 (0.57-2.19)	0.39 (0.20-0.77)	0.56 (0.29-1.08)
P for trend		0.91	<0.01	0.07
Number of specific symptomsa
0	9	reference	reference	reference
1-2	6	1.45 (0.70-3.02)	0.65 (0.20-2.15)	0.82 (0.40-1.66)
3-5	7	0.82 (0.48-1.38)	0.44 (0.20-0.98)	0.54 (0.26-1.11)
P for trend		0.29	0.04	0.09
High fever
No	18	reference	reference	reference
Yes	4	0.62 (0.34-1.13)	0.76 (0.28-2.05)	0.32 (0.18-0.57)
Severe fatigue
No	13	reference	reference	reference
Yes	9	0.77 (0.46-1.28)	0.60 (0.26-1.37)	0.74 (0.37-1.47)
Dyspnea
No	17	reference	reference	reference
Yes	5	0.80 (0.53-1.21)	0.56 (0.23-1.37)	0.71 (0.34-1.48)
Loss of sense of taste or smell
No	20	reference	reference	reference
Yes	2	0.60 (0.41-0.88)	0.27 (0.11-0.69)	1.00 (0.30-3.38)
Sore throat
No	9	reference	reference	reference
Yes	13	1.06 (0.60-1.86)	0.53 (0.25-1.10)	0.65 (0.36-1.16)

Data are shown as the ratio of means estimated using the multivariable linear regression model with adjustment for age and sex.

Abbreviations: CI, confidence interval; NAb, neutralizing antibody.

^aNumber of specific symptoms counted: high fever, severe fatigue, dyspnea, loss of sense of taste or smell, and sore throat.

Discussion

Among the staff of a medical research center in Tokyo, the incidence rate of breakthrough infection with SARS-CoV-2 during the Omicron dominant wave was lower in those who received three doses of the mRNA vaccine (within 3 months after the vaccination) than in those who received only two doses (within 11 months after the vaccination). In a case-control study nested in a cohort of vaccine recipients of all three doses, the preinfection NAb levels against Omicron BA.1 and Omicron BA.2 did not materially differ between breakthrough infection cases and uninfected controls who were in close contact with patients with COVID-19 during the Omicron wave. The patients who experienced COVID-19-compatible symptoms during the Omicron wave had a lower preinfection NAb levels against Omicron than those without symptoms.

During the Omicron BA.1-dominant wave, we observed a 54.6% VE of the third dose of vaccine (within 3 months of vaccination) compared with the second dose (within 11 months of vaccination). This finding is similar to those of previous studies that examined VE approximately 1-2 months after the third dose relative to unvaccinated individuals during the Omicron wave (51.2 and 52.9%) [3,4]. Our results were consistent with our laboratory data from the same cohort [15], where Omicron BA.1-specific NAb was detected in all patients who received the third dose within 1 month but not in all patients who received only the second dose approximately 8 months previously (GMT, 152 vs <40 NT₅₀). However, compared with the VE during the Delta epidemic (88.1 and 92.6%) [3,4], the VE during the Omicron epidemic was considerably low. In serum samples obtained from the third vaccine dose recipients, we found much lower NAb titers against Omicron BA.1 than against Delta (152 vs 1563 NT₅₀) [15]. Thus, the third vaccine dose was effective in decreasing the risk of Omicron infection. However, the magnitude of this effect was much smaller than that for former SARS-CoV-2 strains.

Among the recipients of the third dose who had close contact with patients with COVID-19 during the Omicron wave, we observed no measurable differences in the levels of live-virus NAb against Omicron variants and antispike antibodies between breakthrough infection cases and their controls. This result is consistent with the findings of a study of nursing home residents, which showed no difference in pseudotyped virus NAb titers against Omicron after the third vaccination between patients infected with Omicron and arbitrarily selected controls [9]. These results suggest that among healthy individuals, breakthrough infections with highly immune-evasive Omicron variants may occur independent of the humoral immune response to the third vaccine dose.

In our VE analysis, asymptomatic infection was more frequent among recipients of the third dose (20%) than among those of the second dose (7%), which is consistent with previous data [16,17]. Among the cases in our case-control study of third vaccine recipients, higher preinfection NAb titers against Omicron were associated with fewer symptoms during the Omicron wave. Similarly, in a study of patients infected with Omicron, after vaccination, a higher total antibody level within 7 days of infection (i.e., peri-infection period) was associated with a lower rate of fever, hypoxia, C-reactive protein elevation, and lymphopenia [18]. These findings suggest that preinfection NAb may play a role in alleviating symptoms of SARS-CoV-2 infection.

The current study has several strengths. We measured the NAb titers against Wuhan, Omicron BA.1, and Omicron BA.2 using live viruses. Both cases and controls were derived from a well-designed cohort. The cases were identified through the registry of patients with COVID-19 and a serological survey. This point is important because many patients with Omicron infection, characterized by asymptomatic or mild symptoms [19], might have been left undiagnosed. In addition, we excluded undiagnosed cases from controls using the results of the antibody test at the follow-up.

This study has some limitations. First, we measured the antibody levels at a median of 10 days after the third dose; antibody titers may still be increasing, and we did not measure the peri-infection antibody levels. Second, although we selected controls from among those who had close contact with patients with COVID-19, the major setting of the contact differed between cases (unknown) and controls (workplace). Nonetheless, we confirmed no material change in the results after restricting the cases and controls to those who had close contact at home (Supplementary Table S2). Third, among the cases reported from the in-house registry, viral sequence data were available for only 33 (55%) cases in the VE analysis and 5 (50%) cases in the case-control analysis. Nevertheless, we could reasonably assume that the remaining breakthrough infections were also due to the Omicron variant, which accounted for more than 90% of the sequenced COVID-19 samples in Japan during the follow-up period (January-March 2022) [20]. Fourth, we did not assess the cellular immune response, which is another important mechanism for preventing COVID-19 [21]. Finally, we could not determine whether the COVID-19-compatible symptoms reported in the follow-up survey corresponded to those that occurred during infection.

In conclusion, compared with second dose, the third vaccine dose halved the risk of SARS-CoV-2 infection during the Omicron wave. Among the third dose recipients, preinfection and live-virus NAb titers against Omicron were not materially different between the cases and controls, whereas a higher preinfection NAb titer against Omicron was associated with fewer symptoms. Higher levels of NAb after the third vaccination may not indicate a lower risk of Omicron infection, whereas they may suppress symptomatic episodes of infection.

Declaration of competing interest

Abbott Japan and Roche Diagnostics provided reagents for antispike antibody assays. The authors have no competing interests to declare.