Influenza A(H3N2) Antibody Responses to Standard-Dose Versus Enhanced Influenza Vaccine Immunogenicity in Older Adults and Prior Season's Vaccine Status

Abstract

Background

Annual influenza vaccination is recommended for older adults but repeated vaccination with standard-dose influenza vaccine has been linked to reduced immunogenicity and effectiveness, especially against A(H3N2) viruses.

Methods

Community-dwelling Hong Kong adults aged 65–82 years were randomly allocated to receive 2017–2018 standard-dose quadrivalent, MF59-adjuvanted trivalent, high-dose trivalent, and recombinant-HA quadrivalent vaccination. Antibody response to unchanged A(H3N2) vaccine antigen was compared among participants with and without self-reported prior year (2016–2017) standard-dose vaccination.

Results

Mean fold rise (MFR) in antibody titers from day 0 to day 30 by hemagglutination inhibition and virus microneutralization assays were lower among 2017–2018 standard-dose and enhanced vaccine recipients with (range, 1.7–3.0) versus without (range, 4.3–14.3) prior 2016–2017 vaccination. MFR was significantly reduced by about one-half to four-fifths for previously vaccinated recipients of standard-dose and all 3 enhanced vaccines (β range, .21–.48). Among prior-year vaccinated older adults, enhanced vaccines induced higher 1.43 to 2.39-fold geometric mean titers and 1.28 to 1.74-fold MFR versus standard-dose vaccine by microneutralization assay.

Conclusions

In the context of unchanged A(H3N2) vaccine strain, prior-year vaccination was associated with reduced antibody response among both standard-dose and enhanced influenza vaccine recipients. Enhanced vaccines improved antibody response among older adults with prior-year standard-dose vaccination.

During 2 consecutive years with unchanged A(H3N2) vaccine strain, reduced antibody response was observed in older adults who received prior-year vaccination compared to those who did not, regardless of receiving standard-dose or enhanced influenza vaccines in the current year.

Annual influenza vaccination is recommended as the mainstay of prevention for individuals at increased risk of severe illness, including older adults [1]. Trivalent or quadrivalent inactivated influenza vaccines are most widely used, typically containing 15 µg hemagglutinin (HA) of each vaccine strain. However, the protection provided by inactivated influenza vaccines can be reduced in older adults compared to younger adults [2, 3]. To improve vaccine-induced protection, a number of enhanced influenza vaccine technologies have been developed, with the aim of improving immunogenicity and effectiveness [4–9]. In 2022, the Advisory Committee for Immunization Practices in the United States issued a preferential recommendation for older adults to receive one of three enhanced vaccines: the MF59 adjuvanted vaccine (FluAD), the high-dose vaccine (FluZone High Dose) with 60 µg HA per vaccine strain, or the recombinant-HA vaccine (FluBlok) containing 45 µg HA per strain [10].

Reduced antibody responses and reduced vaccine effectiveness (VE) associated with repeated vaccination have sometimes been observed in standard-dose inactivated vaccine recipients, particularly for A(H3N2) viruses [11]. Our previous study in older adults in Hong Kong found that vaccination in the previous two years correlated with reduced immune responses to current-year vaccination [12]. Repeated vaccination effects have also been reported in other studies [11, 13–15], with a number of hypotheses about the underlying biological mechanisms [16–19]. According to the antigenic distance hypothesis, vaccine protection could be compromised when the current vaccine strain is significantly different from the circulating strains and resembles the previous vaccine strain [16]. Because A(H3N2) is the most antigenically diverse strain among influenza viruses, mismatches between vaccine composition and circulating strains occur more frequently, which might contribute to reduced VE against A(H3N2) [20]. Thus, solutions are needed to improve A(H3N2) protection following repeated vaccination, particularly when the vaccine strains remain unchanged for two consecutive years. The aim of this study is to evaluate whether the reduced antibody response to A(H3N2) associated with repeated vaccination with the same vaccine strains could be overcome with receipt of enhanced vaccines rather than standard-dose influenza vaccine in older adults.

METHODS

Participants

Community-dwelling older adults aged between 65 and 82 years were recruited in Hong Kong from June 2017 to January 2018. Participants were excluded if they (1) had already received northern hemisphere 2017–2018 formulation of influenza vaccine before recruitment; (2) had a clinical diagnosis of, or showed symptoms of, any cognitive disorders including dementia; (3) reported any contraindications for influenza vaccination; or (4) reported any contraindications for intramuscular injections (eg, anticoagulant medication). Additional details of participant recruitment and laboratory methods were published previously [4].

After the screening process, we collected relevant information from eligible older adults through interview, including age, sex, underlying medical conditions, and self-reported influenza vaccination history in the prior 5 years. Participants were randomly assigned to receive a standard-dose quadrivalent influenza vaccine (0.5 mL; FluQuadri; Sanofi Pasteur) or 1 of 3 enhanced influenza vaccines, which were developed to elicit stronger immune responses than the standard-dose influenza vaccine. These vaccines include MF59-adjuvanted trivalent influenza vaccine (0.5 mL; FluAD; Seqirus), high-dose trivalent influenza vaccine (0.5 mL; Fluzone High-Dose; Sanofi Pasteur) and recombinant-HA quadrivalent influenza vaccine (0.5 mL; FluBlok; Sanofi Pasteur). All vaccines used in this study were formulated according to recommendations for the northern hemisphere in 2017–2018 for which only the A(H1N1) component was changed from the prior 2016–2017 season's formulation. The quadrivalent standard-dose and recombinant-HA influenza vaccines included the vaccine strains of A/Michigan/45/2015(H1N1)-like virus (clade 6B.1), A/Hong Kong/4801/2014(H3N2)-like virus (clade 3C.2a), B/Brisbane/60/2008-like virus (Victoria lineage; clade 1A), and B/Phuket/3073/2013-like virus (Yamagata lineage; clade 3). The trivalent MF59-adjuvanted and high-dose influenza vaccines did not include antigen from the B/Phuket/3073/2013-like virus. Of note, those vaccinated in 2016–2017 or earlier received standard-dose influenza vaccine, which was the only influenza vaccine available to older adults in Hong Kong before 2017–2018. All participants signed informed consent before any study-related procedures took place. The study protocol was approved by the Institutional Review Board of the University of Hong Kong.

Laboratory Methods

Paired blood samples were collected immediately before vaccination (day 0) and approximately 30 days after vaccination (day 30). Sera were separated within 24 hours and stored at −80°C before testing. A subset of 800 participants (200 per vaccine group) with paired samples available were selected for analysis of antibody responses against the vaccine strains. This subset consisted of all participants who provided additional blood samples for cell-mediated immunity and a random sample from the remainder of the participants with available serum samples at day 0 and day 30.

Hemagglutination inhibition (HAI) and virus microneutralization (MN) assays for A(H3N2) viruses were applied to evaluate antibody responses. For the HAI assay, the sera were treated with receptor-destroying enzyme and inactivated in a 56°C water bath before being tested. Serum, 50μL, was serially diluted with 25μL phosphate buffered saline in V bottom 96-well plates, after which 25μL antigen containing 4 HA units of egg-propagated A/Hong Kong/4801/2014(H3N2) was added, and the mixtures were incubated at room temperature for 1 hour. This was followed by adding 50μL 0.5% turkey red blood cells to each well. After incubation for 30 minutes at room temperature, the HAI titer was defined as the reciprocal of the highest 2-fold serum dilution that inhibited HA completely.

We also applied MN assays to test the antibody titer against A/Hong Kong/4801/2014(H3N2) grown in Madin-Darby canine kidney cells. The serum was inactivated in a 56°C water bath, starting with a dilution of 1:10 and followed by serial 2-fold dilutions. Equal volumes of 200 TCID₅₀ (50% tissue culture infectious dose) of cell-propagated antigen were mixed with serum during the dilution. After incubation for 1 hour in a 37°C humidified incubator, a 35 µL virus–serum mixture was added to Madin-Darby canine kidney cell monolayers in 96-well microtiter plates in quadruplicate and incubated at 37°C for 1 hour. Then we discarded the supernatant, washed, and added 150 µL of culture medium (0% minimal essential medium with Tosyl phenylalanyl chloromethyl ketone trypsin) to each well. After incubation in a 37°C and 5% CO₂ humidified incubator for 3 days, the cytopathic effect was recorded as the reciprocal of the highest serum dilution that provided at least 50% inhibition of virus infectivity.

Statistical Analysis

The statistical analysis was designed to examine the association of prior-year vaccination on antibody response to A(H3N2) viruses with standard-dose and enhanced influenza vaccines in older adults, and the potential benefit of enhanced vaccines over standard-dose vaccine stratified by prior-year vaccination history. χ² test, Fisher exact test, or Kruskal-Wallis rank sum test was applied to compare the baseline characteristic between prior-year vaccinated and not vaccinated groups. Values and corresponding 95% confidence intervals (CIs) of geometric mean titers (GMTs) at day 0, GMTs at day 30, mean fold rise (MFR) from day 0 to day 30, the proportion with at least a 4-fold rise in titer, and the percentages with postvaccination titers ≥40, ≥ 80, and ≥160 were calculated in each vaccine type group stratified by 2016–2017 vaccination status. Using the corresponding value in participants without prior-year vaccination as reference, we estimated ratios and 95% CIs of GMTs at day 0, GMTs at day 30, and MFR in participants with prior-year vaccination in each 2017–2018 vaccine type group. Stratified by the prior-year vaccination status, we calculated the ratios and 95% CIs of GMTs at day 30 and MFR, respectively, in each enhanced vaccine group compared to the corresponding value in the standard-dose vaccine group. The difference in antibody response between prior-year vaccinated and not vaccinated groups was estimated by log-linear regression models adjusted for sex and age. The ratios and 95% CIs between 2017–2018 standard-dose and enhanced vaccines were calculated. Statistical significance was defined by P value <.05 or 95% CIs excluding 1 for the ratios. All statistical analyses were performed using R version 1.4.1717 (R Foundation for Statistical Computing).

RESULTS

A total of 1861 participants were enrolled and vaccinated with either a standard-dose or an enhanced 2017–2018 influenza vaccine according to the randomization scheme. Out of these, 1826 provided paired samples for serology tests. A subset of 800 older adults (315 men and 485 women) were selected, and their paired sera were tested by HAI and MN assays against egg- and cell-propagated A(H3N2) antigen, respectively. Among these 800 participants, 542 (67.8%) were vaccinated in 2016–2017, and 258 (32.3%) were not vaccinated. Participants with prior-year vaccination in 2016–2017 were significantly different from those without prior-year vaccination with respect to sex, age, hypertension, and self-report influenza vaccination history prior to 2016–2017 (Table 1). In particular, those vaccinated versus not vaccinated in 2016–2017 were more often female (65.5% vs 50.4%, respectively), older (mean 72 vs 68 years, respectively) and with hypertension, the most common underlying condition overall (53.7% vs 39.8%, respectively). The self-reported vaccination rates in each season from 2012–2013 to 2015–2016 were also higher among those with (range per season, 63.1%–85.6%) than without (range per season, 6.2%–12.8%) 2016–2017 vaccination. Most (70.5%) of those unvaccinated in 2016–2017 had not been vaccinated in any of the 5 seasons from 2012–2013 through 2016–2017.

Table 1.

Demographic and Baseline Characteristics of 800 Participants by 2016–2017 Influenza Vaccination Status

Characteristics	2016–2017 Self-Reported Influenza Vaccination Status
	Not Vaccinated (n = 258)	Vaccinated (n = 542)	P Value
Female sex, No. (%)	130 (50.4)	355 (65.5)	<.001a
Age, y, median (range)	68 (66–71)	72 (68–77)	<.001b
Underlying condition, No. (%)
Hypertension	102 (39.8)	291 (53.7)	.001a
Asthma	4 (1.6)	20 (3.7)	.151a
Coronary artery disease	6 (2.3)	17 (3.1)	.678a
COPD	0 (0)	3 (0.6)	.555c
Congestive heart failure	0 (0)	1 (0.2)	1.000c
Any heart or lung condition	116 (45.3)	318 (58.7)	<.001a
Self-report influenza vaccination history, No. (%)
2015–2016 NH	33 (12.8)	464 (85.6)	<.001a
2014–2015 NH	21 (8.1)	402 (74.2)	<.001a
2013–2014 NH	16 (6.2)	364 (67.2)	<.001a
2012–2013 NH	18 (7.0)	342 (63.1)	<.001a
No influenza vaccination in the previous 5 y	182 (70.5)	0 (0)	<.001a
Vaccine received in 2017–2018, No. (%)
Standard-dose QIV	64 (24.8)	136 (25.1)	.794a
MF59-adjuvanted TIV	70 (27.1)	139 (25.6	…
High-dose TIV	61 (23.6)	137 (25.3)	…
Recombinant-HA QIV	63 (24.4)	130 (24.0)	…

Abbreviations: COPD, chronic obstructive pulmonary disease; NH, northern hemisphere formulation; QIV, quadrivalent influenza vaccine; TIV, trivalent influenza vaccine.

^a P value estimated by χ² test.

^b P value estimated by Kruskal-Wallis rank sum test.

^c P value estimated by Fisher exact test.

In all four 2017–2018 vaccine groups, the prevaccination GMTs by both HAI and MN assays were significantly higher in recipients with versus without prior-year vaccination (β range, 1.57–2.95; Table 2). However, in all four 2017–2018 vaccine groups, GMTs at day 30 by HAI assay were significantly lower in recipients with versus without prior-year vaccination (β range, .50–.61) and by MN assay for the 2017–2018 recombinant-HA vaccine group (β = .54) but not the other 3 egg-based vaccine groups. MFR from day 0 to day 30 by both HAI and MN assays, was significantly lower among recipients with (range, 1.7–3.0) versus without (range, 4.3–14.3) prior-year vaccination in all four 2017–2018 vaccine groups (β range, .21–.48) (Figure 1 and Table 2). Similar associations can be seen in other indicators of antibody response, including percentages of recipients achieving at least a 4-fold rise with day 30 antibody titer ≥40 and meeting various GMT thresholds after 2017–2018 vaccination (Supplementary Table 1).

Table 2.

Comparison of Antibody Responses to 2017–2018 A(H3N2) Vaccine Strain Among Participants With 2016–2017 Influenza Vaccination and Without 2016–2017 Influenza Vaccination

	Influenza Vaccine Received in 2017–2018
	Standard Dose QIV n = 200		MF59 Adjuvanted TIV n = 200		High Dose TIV n = 200		Recombinant-HA QIV n = 200
	β	(95% CI)	β	(95% CI)	β	(95% CI)	β	(95% CI)
HAI assay, A/Hong Kong/4801/2014 (H3N2) egg-propagated antigen
Geometric mean titer at day 0	2.03***	(1.39–2.99)	2.00***	(1.34–2.99)	2.77***	(1.89–4.05)	2.95***	(2.01–4.33)
Geometric mean titer at day 30	.50***	(.35–.70)	.52***	(.38–.72)	.60**	(.42–.84)	.61**	(.43–.86)
Mean fold rise from day 0 to 30	.25***	(.17–.35)	.26***	(.18–.38)	.22***	(.15–.32)	.21***	(.14–.30)
MN assay, A/Hong Kong/4801/2014 (H3N2) cell-propagated antigen
Geometric mean titer at day 0	1.57*	(1.03–2.39)	2.28***	(1.44–3.62)	1.85**	(1.20–2.86)	2.14**	(1.37–3.34)
Geometric mean titer at day 30	.72	(.47–1.11)	.99	(.64–1.55)	.89	(.56–1.40)	.54**	(.37–.79)
Mean fold rise from day 0 to 30	.46***	(.34–.62)	.44***	(.32–.60)	.48***	(.34–0.67)	.25***	(.17–.36)

β estimated by log-linear regression adjusted for age and sex, with corresponding value of group without 2016–2017 influenza vaccination as reference. * P < .05; ** P < .01; *** P < .001.

Abbreviations: CI, confidence interval; HAI, hemagglutination inhibition assay; MN, microneutralization assay; QIV, quadrivalent influenza vaccine; TIV, trivalent influenza vaccine.

Figure 1.

Comparisons of changes in influenza A(H3N2) geometric mean titers from day 0 to day 30 by (A) HAI assay using A/Hong Kong/4801/2014 (H3N2) egg-propagated antigen and (B) MN assay using A/Hong Kong/4801/2014 (H3N2) cell-propagated antigen between participants with and without 2016–2017 standard-dose influenza vaccination in each 2017–2018 vaccine group. Values tested significantly different (P < .05) from participants without 2016–2017 influenza vaccination are shown in bold (estimated by log-linear regression adjusted for age and sex). Abbreviations: HAI, hemagglutination inhibition; MN, microneutralization; QIV, quadrivalent influenza vaccine; TIV, trivalent influenza vaccine.

Figure 2 shows the relationships between the distributions of prevaccination antibody titers and postvaccination antibody titers by HAI and MN assays. Recipients of prior-year vaccination clustered more closely around the diagonal of the plot indicating little change from pre- to postvaccination, whereas those without prior-year vaccination showed a greater density toward the upper left corner (ie, more participants with low prevaccination titers and high postvaccination titers). These patterns were similar across all 4 vaccine groups.

Figure 2.

Heatplot of pre- and postvaccination titers by (A) hemagglutination inhibition assay using A/Hong Kong/4801/2014 (H3N2) egg-propagated antigen and (B) microneutralization assay using A/Hong Kong/4801/2014 (H3N2) cell-propagated antigen stratified by the receipt of 2016–2017 standard-dose vaccine in the prior year for each vaccine group. A darker shade indicates a greater density of points in the relevant area of the plot. Antibody titers are interval censored within the 2-fold dilutions used in the assays, hence the use of a heatplot rather than a scatterplot to show these data. Areas on the diagonal represent no change in titer from prevaccination to postvaccination. Areas to the upper left represent low prevaccination titers and high postvaccination titers.

Among those who had received standard-dose vaccine in 2016–2017, the 2017–2018 enhanced vaccine stimulated higher day 30 MN GMTs (range, 1.43 to 2.39-fold higher) and MFR (range, 1.28 to 1.74-fold higher), with similar pattern by HAI assay, compared to 2017–2018 standard-dose vaccine (Figure 3 and Supplementary Table 2). However, this relative benefit of enhanced over standard-dose vaccine was only statistically significant among repeat vaccinees. Among participants without prior-year vaccination, the day 30 GMTs and MFR by HAI and MN assay did not significantly differ between 2017–2018 enhanced and standard-dose vaccine recipients. The only exception among prior-season unvaccinated individuals was for recombinant-HA vaccine, which by MN assay showed 3.02-fold higher day 30 GMT and 2.94-fold higher MFR compared to standard-dose vaccine. Although recombinant-HA vaccine tended to provide the greatest relative advantage overall, differences between 2017–2018 enhanced vaccine groups were not significant, with overlapping CIs. The corresponding ratios and 95% CIs can be found in Supplementary Table 2.

Figure 3.

Ratios of day 30 geometric mean titers and mean fold rise by (A) hemagglutination inhibition assay using A/Hong Kong/4801/2014 (H3N2) egg-propagated antigen and (B) microneutralization assay using A/Hong Kong/4801/2014 (H3N2) cell-propagated antigen estimated for each enhanced vaccine compared to standard-dose influenza vaccine among participants with and without 2016–2017 influenza vaccination. Open squares represent the point estimates and dashed lines represent the 95% confidence intervals (CIs) in participants without 2016–2017 vaccination. Solid points represent the point estimates and solid lines represent the 95% CIs in participants with 2016–2017 vaccination. The 95% CIs that do not cross 1 (null value) are statistically significant. Values to the right of the null value favor the corresponding 2017–2018 enhanced vaccine over the 2017–2018 standard-dose influenza vaccine.

DISCUSSION

This study aimed to determine whether enhanced influenza vaccines for older adults could overcome the impaired antibody response to A(H3N2) associated with repeated influenza vaccination of the same vaccine strains in 2 consecutive years. This is of substantial public health importance given A(H3N2) epidemics, compared to other influenza type or subtype epidemics, are typically associated with greater older adult population-level mortality, who are recommended to receive annual influenza vaccination [21]. In the context of 2017–2018 A(H3N2) vaccine antigen that was unchanged from 2016–2017, we observed reduced MFR to A(H3N2) among prior-year vaccinated older adults following both 2017–2018 standard and enhanced influenza vaccination, compared to those unvaccinated in the prior year. These findings are consistent with the antigenic distance hypothesis, which theorized potentially greater negative interference from prior season's vaccination on current season's vaccine response when identical antigens are used cross-season [16]. Our immunogenicity findings are also consistent with epidemiological observations of reduced 2017–2018 VE among prior recipients of 2016–2017 influenza vaccine in Canada [22], and under similar 2016 and 2017 conditions in Australia [23]. Moreover, ferret experiments recapitulating these conditions showed that animals vaccinated twice consecutively versus current season only with the identical 2016–2017 and 2017–2018 egg-adapted A(H3N2) antigen had reduced vaccine protection and greater viral shedding [24]. It would be valuable to repeat our study and attempt to correlate immunogenicity with epidemiological observations when vaccine components are changed between seasons.

Interestingly, when assessed by MN assay, using cell-propagated A(H3N2) virus, participants not vaccinated the prior year who received recombinant-HA vaccine achieved at least twice higher MFR than recipients of egg-based MF59-adjuvanted or high-dose vaccines (12.6 vs 5.0 vs 6.2, respectively) while such differences were not as prominent by HAI assay using the egg-propagated A(H3N2) vaccine strain. The very high MFR evident by the MN assay among recombinant-HA vaccine recipients may be tied to the cell-based recombinant vaccine manufacturing process that avoids egg-adaptation mutations accrued with egg-based manufacturing. The antibody stimulated by recombinant-HA vaccine (also at higher dose of HA than standard vaccine) may have been a better antigenic match to virus used in MN versus HAI assays. According to the Crick Worldwide Influenza Centre, virtually all A(H3N2) subclade 3C.2a1b (32/33) and 3C.2a2 (24/25) viruses that circulated in 2017–2018 were antigenically similar to the cell-propagated 3C.2a (A/Hong Kong/4801/2014) vaccine strain; whereas, among the same isolates, all 3C.2a1b and all but one 3C.2a2 viruses were antigenically distinct from the egg-adapted 3C.2a (A/Hong Kong/4801/2014) vaccine strain, reinforcing a potential role for egg adaptation mutations in these relative immunogenicity responses by vaccine group [25].

The effect of repeated vaccination cannot be fully explained by the “antibody ceiling effect” [26], which has been raised as a concern for using serological end points in evaluating VE since the 1960s [19]. Apart from being used to explain the negative correlation of the antibody boost against infection with higher levels of preexisting antibody titers or multiple exposures [27, 28], this effect was also found in repeated vaccination [29, 30]. A study following HAI antibody response against A(H1N1) in older adults for up to 75 days observed a ceiling effect of 640, the preexisting titer that was not likely to increase after vaccination [31]. However, our study found that the majority of repeat vaccinees did not have high prevaccination titers and typically achieved 2-fold rises in titer after vaccination, with no obvious ceiling effects playing a role in the reduced response to vaccination (Figure 2). Specifically, we also found that reduced antibody responses associated with repeated vaccination could occur even when the prevaccination antibody titer was low.

In addition, we also found that enhanced vaccines were more immunogenic than the standard vaccine in recipients with prior-year vaccination, although the difference was not statistically significant in recipients without prior-year vaccination, except for recombinant-HA vaccine, which still achieved higher MN antibody titer compared to the standard-dose vaccine (Figure 3). The immunogenicity of repeated vaccination with enhanced vaccines has seldom been reported in previous publications, with a few exceptions [29, 32], both of which lack a comparison group with no vaccination in the first year. A randomized trial concluded that high-dose influenza vaccine tended to perform better than standard-dose vaccine in immunogenicity regardless of the vaccine types received in the preceding year [32]. Combined with findings of our study, it suggests that one of the enhanced vaccines’ advantages is to offer improved protection for those with prior vaccination. Whether such relative advantage will persist if enhanced vaccines are repeatedly used annually remains to be seen and is an important area for further research.

This study has several limitations. First, the proportion of recipients without prior-year vaccination was relatively small (32%) among all participants, thus the comparable immunogenicity between the enhanced vaccines and standard-dose vaccine found in recipients without prior-year vaccination could be further verified in larger studies. Second, the vaccination history was self-reported and there may be misclassification bias; thus we only classified the recipients by vaccination status in 2016–2017 rather than based on multiple prior years. Third, we assessed immunogenicity using HAI assay, which is a conventional assay that is furthermore subject to variation based on parameters such as erythrocytes (eg, turkey vs guinea pig) used, antigen (egg- vs cell-propagated) and other conditions (eg, oseltamivir carboxylate to address neuraminidase effects). For that reason, we also assessed antibody response by MN assay, a functional assay detecting antibodies that can prevent infection of mammalian cells in vitro and one for which we used cell-propagated viruses that may compliment but also complicate direct comparison with HAI findings based instead on egg-propagated viruses. Our findings of reduced immunogenicity with repeat vaccination in 2017–2018 correlated well with epidemiological findings elsewhere (eg, Canada, Australia) and with animal model investigations, but we cannot directly translate our antibody findings into estimates of clinical vaccine protection. Finally, the association between immune response to enhanced vaccines and repeated vaccination over multiple seasons and the longitudinal effects on the duration of the antibody should be elucidated in future study, including whether the capacity to overcome prior year's vaccination will be maintained in the context of repeated annual use of enhanced vaccines.

In conclusion, this study found that the reduced antibody response to A(H3N2) vaccine strain related to repeated vaccination occurred for both enhanced and standard vaccines, with the enhanced vaccines more immunogenic than the standard vaccine in prior repeated recipients of standard-dose vaccine. The results of this study provide unique evidence that reduced antibody responses associated with repeated vaccination cannot be fully overcome by adding adjuvant, increasing the dose of HA antigen, or using recombinant-HA instead of egg-propagated HA in the vaccine. Our findings are consistent with the antigenic distance hypothesis, which postulates greater negative interference from prior season's vaccination in the context of unchanged antigen, as per the current evaluation. In the future, the association between antibody response and repeated vaccination should be investigated over multiple seasons, and solutions to improve the immune response to repeated vaccination need to be further explored and developed.

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

Acknowledgments. The authors thank the study staff for assistance in project implementation, including Tin-Kin Chau, Jennifer Ho, Fiona Kee, Janisy Lai, Cecily Leung, Anita Li, Jessamine Luk, Loretta Mak, Yvonne Ng, Angel Wong, Miyuki Wong, Phoebe Wong, and Kitty Yu; Leo Luk, Emily Yau, Chi Tsang, and Kin Chan for laboratory support; and Julie Au for administrative support.

Disclaimer. The sponsor had no role in the data collection and analysis, or the decision to publish, but was involved in the study design and preparation of the manuscript. The contents of this report are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention or the Department of Health and Human Services.

Financial support. This work was supported by the Centers for Disease Control and Prevention (cooperative agreement number IP001064-02); the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health and Human Services (contract number 75N93021C00016); and the Research Grants Council of the Hong Kong Special Administrative Region Government (Theme-based Research Scheme project number T11-712/19-N). B. J. C. is supported by a Research Grants Council Senior Research Fellowship (grant number HKU SRFS2021-7S03).