Respiratory Syncytial Virus Humoral Antibody Responses in Older Adults After Vaccination or Infection

Abstract

Background

Immunity to respiratory syncytial virus (RSV) is short lived following infection. We determined if recently licensed RSV preF vaccines induce better immune responses than infection.

Methods

Serum preF binding and neutralizing antibody to RSV A and B was measured in older adults at baseline and 30 days after RSV infection or vaccination with 2 licensed RSV preF protein vaccines.

Results

Vaccination induced higher serum preF binding antibody and neutralizing responses than RSV infection (A2 RSV, 3306 vs 1254, P < .0001; B1 RSV, 5153 vs 2186, P < .0001).

Conclusions

Vaccination with preF protein vaccines induce better humoral immune responses than RSV infection in older adult.

RSV preF vaccines, adults, humoral antibody response, RSV infection

Respiratory syncytial virus (RSV) is a relatively frequent cause of serious respiratory illness in older adults and those with preexisting medical conditions [1, 2]. The burden of RSV in adults is estimated to account for 60 000–180 000 hospitalizations and 10 000–14 000 deaths in adults in the United States per year. Two protein-based vaccines and a messenger RNA (mRNA) vaccine, all based on the prefusion F protein (preF) structure, have recently been licensed in the United States for use in adults [3–5]. Initial clinical trials demonstrated similar high efficacy in persons 60 years of age and older (82.4%–85.7%) against lower respiratory tract disease for each of the vaccines. Postlicensure analyses have demonstrated high vaccine effectiveness (VE) to prevent physician office and emergency department visits and hospitalizations for the 2 protein-based vaccines during the 2023–2024 winter season in the United States [6, 7].

Preliminary immunogenicity data in phase 1/2 trials showed geometric mean fold-rises (GMFR) in serum neutralizing antibody to both group A and B RSV ranging from 9.5 to 12.0 and 8.5 to 9.2, respectively [8, 9]. It has long been recognized that protection from RSV infection is incomplete and relatively short lived, with repeat infections occurring throughout life [1, 10]. It has been suggested that an optimal RSV vaccine should induce a better immune response than natural infection [11]. Thus, we compared the serum preF binding and neutralizing antibody responses in older adults vaccinated with the protein-based preF vaccines to responses following natural RSV infection.

METHODS

Study Populations

Archived serum samples from 82 prospectively enrolled community-dwelling adults ≥60 years of age in 2 prior studies, completed in 1999–2003 and 2005–2009, of natural RSV infection were analyzed [1, 12]. For both studies RSV infection was documented by reverse transcription-polymerase chain reaction (RT-PCR) assay from a respiratory sample. The vaccine cohort consisted of 146 adults ≥60 years of age who resided in the community (n = 76) or in long-term care facilities (LTCF; n = 70) enrolled in an RSV vaccine immunogenicity study [13]. Before (day 0) and approximately 30 days after RSV infection (day 30) sera or pre- and postvaccination sera were available for all subjects. All studies were approved by the University of Rochester Research Subjects Review Board and included consent for future use.

The vaccinated community subjects were randomly administered 1 of the 2 protein-based RSV vaccines, either a monovalent preF vaccine containing 120 µg of preF from a group A RSV strain with ASO1_E adjuvant (Arexvy, GSK) or a bivalent preF vaccine containing 60 µg each of preF from group A and B RSV strains (Abrysvo, Pfizer, Inc) [3, 4]. Residents at one LTCF were administered the monovalent vaccine and residents at the other LTCF received the bivalent vaccine based on the institution's decision. The primary exclusion criteria for the RSV-infected population was the presence of an immunocompromising condition or medication. Exclusions for the vaccine cohort included immunosuppressive conditions, any routine vaccination within a 14-day window before or after RSV vaccination, previous or subsequent receipt an RSV vaccine outside the study, receipt of blood/plasma products or immunoglobulin within 60 days before RSV vaccination, or RSV infection within 2 months prior to enrollment. In a previous publication, antibody responses to RSV preF vaccines in LTCF residents were demonstrated to be equivalent to community-dwelling older adults and thus the 2 groups were combined [13].

Laboratory Methods

Enzyme Immunoassay

Serum immunoglobulin G (IgG) titers to the RSV prefusion protein of the A2 strain (preFA) and the B1 strain (preFB) were determined by an end-point enzyme immunoassay as previously described [13]. Briefly, serial 2-fold serum dilutions were incubated in preF-coated plates and developed with alkaline phosphatase conjugated anti-human IgG followed by substrate. Mean titers for each cohort were expressed as geometric mean titers (GMTs) with 95% confidence intervals (CI).

Microneutralization Assay

Serum neutralizing titers were determined using an established microneutralization assay for RSV A2 (MNA2) and B1 (MNB1) strains [12]. Briefly, nine 2-fold serum dilutions in DMEM/5% FCS were incubated with approximately 75 plaque-forming units of RSV for 30 minutes followed by addition of 1.5 × 10⁴ HEp-2 cells in 96-well plates. After 3 days’ growth at 37°C in 5% CO₂, RSV antigen production is quantified by enzyme immunoassay using an RSV-specific murine monoclonal antibody followed by substrate. The neutralization titer is defined as the serum dilution resulting in a 50% reduction in color development compared to positive control wells without serum. Inter- and intra-assay standardization was performed using a panel of high, medium, and low neutralizing sera and a control serum. The mean neutralizing titers for each cohort was converted to GMTs with 95% CI. Neutralization titers for the A2 strain and B1 strain can be converted to the World Health Organization international RSV standard in units/mL by multiplying the GMT by 3.086 and 1.01, respectively.

RESULTS

Study Populations

There were 82 participants in the RSV infection community group, 76 in the vaccinated community group, and 70 in the LTCF group. Demographic data for each group are shown in Table 1. The vaccinated community group and the RSV-infected group had similar mean ages, 72.6 and 73.8 years, respectively, while the vaccinated LTCF group was significantly older at 79.6 years (P < .001). All groups were predominately White, non-Hispanic, and female, although men predominated in the RSV-infected group. The vaccinated LTCF subjects had significantly more underlying medical conditions than the vaccinated community and infected community groups (3.2 vs 0.87 and 2.3, respectively; P < .001 for both comparisons). The vaccinated LTCF group had a higher prevalence of coronary artery disease, congestive heart failure, chronic kidney disease, and neurologic conditions than the other groups.

Table 1.

Clinical Characteristics of Study Populations by Group

Characteristic	Vaccinated Community (n = 76)	Vaccinated LTCF (n = 70)	Community RSV Infected (n = 82)	P Value
Age, mean (SD)	72.6 (6.3)	79.6 (9.7)	73.8 (7.8)	<.001a
Female, n (%)	45 (56)	41 (59)	37 (45)	NS
Race, n (%)
White	72 (95)	69 (90)	79 (96)	NS
Black/African American	0 (0)	7 (10)	3 (4)	.005b
Asian	4 (5)	0 (0)	0 (0)	NS
Hispanic, n (%)	1 (1)	4 (6)	2 (2)	NS
Medical history, n (%)
Diabetes mellitus	15 (20)	22 (31)	14 (17)	NS
Coronary artery disease	6 (8)	25 (36)	19 (23)	<.001b
Congestive heart failure	2 (3)	19 (27)	4 (5)	< .001a
Chronic obstructive pulmonary disease	5 (7)	13 (19)	21 (26)	NS
Asthma	5 (7)	5 (7)	6 (7)	NS
Hypertension	27 (36)	56 (80)	38 (46)	<.001b
Chronic kidney disease	2 (3)	21 (30)	2 (2)	<.001a
Cerebrovascular accident	0 (0)	22 (31)	2 (2)	<.0001a
Dementia	0 (0)	32 (46)	0 (0)	<.0001a
Cancer	4 (5)	12 (17)	6 (7)	.03c
No. medical conditions, mean (SD)	0.87 (0.88)	3.2 (1.6)	2.3 (1.1)	<.001b
Infecting RSV group A/B/unknown	NA	NA	37 A/34 B/11 unknown	NA

Abbreviations: LTCF, long-term care facility; NA, not applicable; NS, not significant; RSV, respiratory syncytial virus.

^aVaccinated LTCF versus vaccinated community and RSV infected groups.

^bVaccinated LTCF versus vaccinated community group.

^cVaccinated LTCF versus vaccinated community.

RSV Antibody Responses to Vaccination or RSV Infection

The pre- and postinfection and pre- and postvaccination serum titers to preFA, preFB, and microneutralization titers to A2 (MNA2) and B1 (MNB1) RSV for the 2 groups are shown in Figure 1. The day 30 GMT for binding antibody to preFA and preFB were significantly higher in vaccinated subjects than RSV-infected subjects (preFA, 128 214 vs 44 285, P < .0001; preFB, 114 087 vs 44 973, P < .0001; Figure 1A). The GMFR to both antigens were also significantly greater in vaccinated subjects (P < .0001 for each). The day 30 GMT of neutralizing titers to RSV A2 and B1 viruses were also significantly greater than in RSV-infected subjects (MNA2, 3306 vs 1254, P < .0001; MNB1, 5153 vs 2186, P < .0001; Figure 1B). The baseline (day 0) titers were not significantly different based on overlap of the 95% CI for each of the comparisons. The MNA2 and MNB1 GMFRs were approximately twice as great for the vaccinated subjects compared to RSV-infected subjects (MNA2, 14.0 vs 6.5, P < .0006; MNB1, 15.7 vs 7.6, P < .004). Because LTCF subjects were significantly older than the other 2 groups, we performed a subanalysis limited to the community groups (Supplementary Figure 1). This confirmed that the vaccinated community group had significantly greater preF binding and neutralizing antibody responses compared to the infected community group (P < .005 for each comparison of peak titer and GMFR).

Figure 1.

Serum geometric mean titers (GMT) with 95% confidence intervals and geometric fold rise (GMFR) of antibody responses to respiratory syncytial virus (RSV) preF vaccination (orange bars) or infection (blue bars). A, Binding antibody to preFA and preFB prior to (day 0) and 1 month after (day 30) vaccination or infection. B, Neutralizing antibody responses to group A RSV (MNA2) and group B RSV (MNB1) at day 0 and day 30 after vaccination and infection.

Although the total dose of preF proteins was equivalent (120 µg) in the 2 vaccine preparations, the bivalent vaccine contained equal proportions of preFA and preFB (60 µg each) while the monovalent contained only preFA with ASO1_E as an adjuvant. Thus, we were interested to determine if there were differences in the response to each vaccine. The baseline (day 0) and day 30 preFA and preFB GMTs and GMFRs were equivalent between the 2 vaccine preparations, with near identical peak titers at day 30 for both antigens (Supplementary Figure 2). The peak GMT and GMFR for MNA2 at day 30 were also equivalent for both vaccines (Supplementary Figure 2). The peak day 30 GMT MNB1 titer for the monovalent vaccine was 28% lower than the peak titer for the bivalent preF vaccine (GMT, 4408 vs 6106, P = .07) and the MNB1 GMFR was also lower for the monovalent vaccine compared to the bivalent vaccine (GMFR, 11.5 vs 22.1, P = .001).

DISCUSSION

The recent development of several new RSV vaccines for use in adults has offered the possibility to effectively reduce the burden of RSV in older and high-risk adults. Because immunity to RSV after natural infection appears to be relatively brief, we sought to compare the humoral immune response after vaccination to that of natural infection in adults ages 60 and above. Notably, we found that serum binding antibody to preF and neutralizing responses to the 2 protein-based vaccines were approximately 2-fold greater 1 month after vaccination compared to natural infection. This is notable because some neutralizing activity in the RSV-infected group could be contributed by antibody responses to the RSV attachment protein (G) that is known to carry neutralizing epitopes and is not a component of the vaccines [14]. It is also possible that this difference is due to immunosuppressive effects of RSV infection [15]. Although a correlate of protection from RSV infection has not been defined, it is quite plausible that higher levels of serum neutralizing or preF binding antibody are causally related to protection as suggested by the success of the vaccine trials and postmarking effectiveness analysis [3–7]. If so, vaccination should provide substantially better protection than natural infection.

Although both preF vaccines contain 120 µg of protein, the major differences between the 2 are that the monovalent vaccine includes an adjuvant while the bivalent vaccine delivers equal proportions of preFA and preFB, representing the 2 major RSV groups. As part of our analysis, we assessed differences in humoral immune response based on virus group. Both vaccines induced equivalent binding antibody to preFA and preFB and induced equivalent neutralizing responses to group A RSV. However, we did find the bivalent vaccine induced a nonsignificant but higher peak GMT and a significantly greater GMFR in neutralizing titer to group B RSV compared to the monovalent adjuvanted vaccine. It is interesting that the difference in MNB1 responses was noted despite equivalent levels of binding antibody to preFB. It is possible that cross-reactive binding antibody may not equate to equivalent cross-group neutralizing capacity, as the correlation between preF binding and neutralization responses, although high, did not show perfect agreement in our analysis (r = 0.73, data not shown). Mapping of antibody binding responses to the neutralization epitopes on both preF proteins could be informative. However, it should be recognized that the clinical relevance of this relatively small difference is unproven. Furthermore, we were unable to assess the potential protective differences in T-cell responses between the vaccinated and infected groups.

During the first year of availability of the 2 protein-based vaccines for the 2023–2024 winter season, approximately 24% of eligible US adults received either the adjuvanted monovalent or the bivalent protein-based vaccines. VE using a test-negative case-control design demonstrated VE of 77%, 80%, and 81% against RSV-associated emergency department encounters, hospitalizations, and intensive care admission, respectively [6]. The clinical trial results have not shown significant RSV-group-specific differences in VE although the number of cases of each viral group was small with large 95% CIs, thus making it difficult to discern subtle differences in VE between vaccines against RSV A or RSV B. In addition, the postmarketing effectiveness studies noted above do not allow group-specific effectiveness calculations as most clinical laboratories do not group type RSV by PCR. However, it should be noted that these studies did not find overall significant differences in point estimates of protection between the 2 vaccines, at a time when both viral groups were circulating nationally [6, 7].

Our study had several limitations to our findings and conclusions. The sample size for both cohorts is relatively small, especially when the analysis is separated by vaccine type. Although there were no major demographic differences between the 2 populations there were some differences in underlying disease characteristics that may have affected immune responses. Finally, we did not have serum samples at later time points to determine if the differences translate into longer durability of vaccine protection compared to natural infection.

In conclusion, we found that the humoral immune response to the protein-based preF vaccines were significantly better than that induced by natural infection. Thes finding suggest vaccination may offer improved protection from RSV compared to naturally acquired immunity.

Supplementary Data

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

Notes

Author contributions. A. F. and A. B. designed the study and contributed to analysis and editing of the manuscript. M. P. executed data generation and contributed to the analysis of the data. E. W. designed the study, executed the data generation, and provided analysis of data, and wrote and edited the manuscript.

Financial support. No financial support was received for this work.

Author notes

Potential conflicts of interest. E. W. has had grant support from Merck, Pfizer, and Moderna; and serves on advisory boards to Merck, Pfizer, Moderna, AstraZeneca, and GSK. A. F. has had grant support from AstraZeneca, Pfizer, and Moderna; and serves on advisory panels to Pfizer, GSK, AstraZeneca, and Moderna. A. B. has grant support from Pfizer, Moderna, and Cyanvac; and is a consultant to Moderna, GSK, and Sanofi. M. P. reports no potential conflicts.

All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.