https://orcid.org/0009-0007-4802-4107
Abstract
Studies have linked respiratory syncytial virus (RSV) antibody-mediated phagocytosis and complement deposition to severe RSV infection in humans. This study shows waning of these antibody functions in women of childbearing age in 2020-2021 during the implementation of COVID-19 mitigation measures, in absence of RSV circulation. These functions could be explored as correlates of protection against severe RSV disease.
Respiratory syncytial virus (RSV) is a major public health concern and a leading cause of morbidity and mortality from lower respiratory tract infection (LRTI) among young children [1]. Infants, especially those with comorbidities and who are under 6 months of age, are at greatest risk for severe RSV disease [2, 3]. Before the coronavirus disease 2019 (COVID-19) pandemic, RSV infections were responsible for an estimated 57,000 hospitalizations and 2 million emergency department and outpatient clinic visits each year in the US among children under 5 years old [4]. Understanding the role of RSV immunity in protecting children against RSV disease and how these factors evolve over time could help anticipate impacts on health systems.
T cells are important to control viral replication during an infection [5]. On the other hand, antibodies protect against severe RSV disease through antigen-specific, fragment antigen-binding (Fab)-mediated antibody interactions responsible for virus capture and neutralization [6]. Additionally, non-neutralizing, fragment crystallizable (Fc)-mediated interactions facilitate effector immune responses against RSV, such as phagocytosis of viral particles and prevention of viral entry into host cells through structural hindrance due to complement deposition. Previous studies in non-human primates, and human challenge and vaccine studies have shown that virus-neutralizing IgG and antibody-mediated complement deposition helped prevent symptomatic disease, whereas Fc-mediated antibody effector cellular immunity functions (including phagocytosis) are particularly important for protection against more severe RSV disease [7, 8].
During the COVID-19 pandemic, RSV cases declined significantly in British Columbia (BC), in the context of strict COVID-19 mitigation measures from spring 2020 to spring 2021. At the same time, levels of RSV pre-fusion F IgG and neutralizing titers against RSV A declined significantly in infants and in women of childbearing age in the absence of RSV circulation from April 2020 to June 2021 [9, 10]. These data suggest that the duration of humoral immunity to RSV is short-lived and ongoing viral exposure is required to maintain high antibody levels, as supported also by epidemiological data [11]. In contrast, no decrease in overall RSV-specific CD4 T cell memory response could be detected over this period [10]. This is important because young infants solely rely on the placental transfer of maternally derived RSV antibodies, to be optimally protected against more severe RSV infection in the first 4 to 6 months, until they develop their own acquired T and B cell immunity months-to-years after birth. Therefore, a decrease in RSV antibody levels in women of childbearing age could herald a worse RSV season in their young children, particularly after a prolonged period of lack of RSV cases.
We lack established clinical correlate(s) of protection (COP) against RSV infection in humans [12]. While RSV antibody neutralization may correlate with symptomatic RSV infection rather than severe RSV disease, measuring Fc-mediated antibody functions in women of childbearing age could potentially inform the degree of vulnerability of infants to more severe RSV disease. In this study, we sought to determine whether RSV-specific Fc-mediated antibody functions associated with protection from severe RSV infection in previous studies, namely antibody-mediated phagocytosis and complement deposition, declined after a prolonged period of lack of RSV circulation in Vancouver (Canada), from May 2020 to May 2021.
MATERIALS AND METHODS
Sample Description, Outcome Measures, and Ethical Statement
Analyses were performed on 18 paired serum samples collected from women of childbearing age 18–51 years old at BC Children’s & Women’s Health Centre in BC, Canada, between May and June 2020, and nearly 1 year later, from February to May 2021, as described [10]. All participants provided informed consent for these studies. Antibody-dependent cellular phagocytosis (ADCP), using monocytes as the target cells, antibody-dependent complement deposition (ADCD) and antibody-dependent neutrophil phagocytosis (ADNP), using neutrophils as the target cells, were measured as described (Supplementary Methods, Supplementary Figures 1–3). The study was approved by University of British Columbia Children’s and Women’s Hospital Research Ethics Board (H20-01205, H18-01724).
Statistical Analysis
Log-transformed ADCP, ADCD, and ADNP levels were compared between 2020 and 2021 by paired t-test. In order to confirm that the assays used in this study tested independent functions of RSV antibodies, correlations between fold-changes and levels of ADCP, ADCD, and ADNP, and previously published [10] pre-fusion RSV F protein IgG levels and RSV A live-virus microneutralization outcomes (expressed antibody titer needed to achieve 95% viral neutralization, or NT95) from the same individuals in 2021 versus 2020 were estimated using Pearson correlation, where statistical significance levels were adjusted for multiple comparisons using the Bonferroni method (n = 10 comparisons, adjusted p < .005). In order to explore relationships between functions of RSV antibodies, unsupervised hierarchical clustering was performed. Results for 2020 and 2021 were independently displayed to enable visualization of the relationship between the years and clusters of functions and levels of RSV antibodies. R version 4.1.2 was used for all analyses. Figure 1 was produced with GraphPad Prism v10.0.2.
Function of RSV-specific antibodies in paired samples of women aged 18–51 years collected in 2020 (early-pandemic) and 2021 (nearly 1 year after the pandemic) (A–C). (A) Monocyte antibody-mediated cellular phagocytosis (ADCP); (B) antibody-mediated complement deposition (ADCD); (C) antibody-mediated neutrophil phagocytosis (ADNP). Abbreviations: Pre-F IgG, pre-fusion RSV F protein IgG levels; NT95 A, neutralization titer 95 of RSV strain A; ADCP, antibody-mediated cellular phagocytosis; ADCD, antibody-mediated complement deposition; ADNP, antibody-mediated neutrophil phagocytosis.
RESULTS
The median age of the women (n = 18) was 37 years (interquartile range [IQR]: 28–41), with a median of 277 days (IQR: 256–301) between samples collected in 2020 versus 2021.
Both ADCP (geometric mean phagocytosis score: 228.7 [95% CI: 83.8–624.0] vs 685.5 [95% CI: 338.6–1,388.0]; p = .020) and ADCD (geometric mean: 1.26 [95% CI: 0.3–5.1] vs 62.8 [95% CI: 45.3–87.2]; p < .001) scores significantly decreased in 2021 compared with 2020 (Figure 1A and 1B). However, no statistically significant difference in ADNP scores was detected between the 2021 and the 2020 samples (geometric mean phagocytosis score: 2,741 [95% CI: 2,430–4-3,091] vs 2,778 [95% CI: 2,458–3,141]; p = .870) (Figure 1C).
The correlation between fold-changes and levels of functions of RSV antibodies measured in this study (ADCP, ADCD, ADNP) and those we recently published (pre-fusion RSV F protein IgG levels, NT95 RSV A) [10] between 2021 versus 2020 was low to moderate and not statistically indicating that these functions decreased independently of each other and of pre-fusion RSV F protein IgG levels (Supplementary Figures 4 and 5). The samples collected in 2020 and 2021 showed a unique cluster that displayed lower ADCP and a unique subset of subjects with lower ADCD using unsupervised hierarchical clustering (Supplementary Figure 6).
DISCUSSION
This study shows a decline in RSV Fc-mediated antibody functions in women of childbearing age residing in Vancouver metropolitan area, in the context of a lack of viral circulation during the peak implementation of COVID-19 NPI, from May 2020 to May 2021. We posit that measuring Fc-mediated functions in this specific population, including antibody-mediated phagocytosis and complement deposition, could inform about potential immune correlates associated with severe RSV LRTI in young infants [13]. Examining RSV antibody functions in paired serum samples allowed to minimize the individual variability expected for these outcomes in humans, and maximize detection of the effect of a lack of viral exposure on these antibody functions over time. Moreover, assaying RSV antibody functions across multiple immunological domains strengthens the evidence supporting a broader loss of RSV immunity during the strict period of implementation of NPI, especially in immunologically naïve infants who lack acquired RSV T cell memory at this young age. This provides potential COP that could be tested in future studies.
Our results are of clinical significance. Indeed, high ADNP and ADCD levels correlated with a reduction in viral load in bronchoalveolar lavage and thus protection from LRTI disease in non-human primates (African green monkeys) vaccinated with six different RSV vaccine platforms and later challenged with RSV [7]. In another study, ADCP, ADNP, and ADCD correlated with protection from RSV infection in humans (18–50 years of age) vaccinated with the adenovirus-based Ad26.RSV.preF vector (protection defined as intranasal RSV plaque-forming units area under the curve < 100 in follow-up after RSV challenge) [8]. The aforementioned studies are biologically relevant given that RSV viral load was shown to be associated with more severe RSV disease in children [14], making RSV viral load as a potential measure of severity of RSV infection. However, other studies have shown that high RSV viral loads to be associated with a decreased risk for severe RSV infection because it more rapidly engages the innate immune response [15, 16]. Indeed, higher viral load at time of presentation and a faster viral load decline were both associated with less severe RSV disease in children [17]. It is also well-established that phagocytosis of RSV immune complexes effectively activates target cells and appears necessary for a broad range of effector immune responses [6]. The current study findings corroborate epidemiological data showing increased RSV hospitalizations in young children in BC after COVID-19 NPI measures were gradually relaxed [18]. Our findings are also relevant as these functions, if confirmed by future studies to serve as COP against RSV, could be tested in RSV vaccine studies, to confirm protective immune responses. This is important also given that maternal RSV vaccines have recently been approved and now available for pregnant women in the US.
While we showed that antibody-mediated phagocytosis declined using monocytes as the target cells (in the ADCP assay), this was not shown using neutrophils as the target cells (in the ADNP assay). This finding could be due to differential expression of receptors implicated in the function of target cells and their affinity to RSV F IgG, and the collaboration between Fab-mediated and Fc-mediated functions. These observations highlight the importance of examining multiple immune outcomes to obtain a more complete picture of RSV immunity in humans [12].
A limitation of this study is its small sample size and from a single geographical region, which requires confirmation in other regions. The study was performed on health-care workers who may have a unique RSV exposure and implementation of NPI, thus this cohort may not reflect the general population. An additional limitation is that early-pandemic antibody function may vary depending on the severity of RSV seasons preceding the time point of blood collection in 2020. The lack of correlation between NT95 and pre-fusion RSV F IgG levels in this sub-analysis may be due to low sample size as high correlation was noted between NT95 and pre-fusion IgG F levels in our previous paper [10]. In our study, we used a single dilution of Palivizumab as appositive control but we did not use different concentrations of Palivizumab, which could have provided a reference to relate the different functions of RSV antibodies to known palivizumab concentration with known neutralizing activity and in vivo effectiveness. The study also lacks clinical outcome measures, which precludes analysis of the impact of the decrease in RSV antibody functions on the severity of RSV cases in infants. Although some subjects had an increase in levels of ADCP and ADNP in 2021 compared with 2020, this increase was marginal and was not associated with increase in pre-fusion RSV F IgG levels in those subjects (data not shown). In addition, small changes in levels of immunological outcomes are expected in biological assays and are within the range of expected variation.
Overall, our study sheds light on the longevity of Fc-mediated RSV antibody responses during a prolonged period of lack of viral circulation. These data add to the growing body of evidence supporting that protective RSV antibody immunity is short-lived. These antibody functions should be formally tested in hypothesis-driven studies of COP against RSV. Further studies are needed to understand how these antibody functions are restored after viral re-exposure.
Author contribution. Conceptualization by BA, FR, PML. Methodology—Neutralization of RSV A (FR); ADCP, ADNP and ADCD (Assay set up: BA with input from CM, PML; Laboratory analysis: BA and MPV). Data analysis and visualization by BA, FR. Data interpretation by BA, CM, FR, MV, PML, AM, LG, MG, AS, DM. Writing–original draft by BA. Writing–review and editing and approval of submitted paper by BA, CM, FR, MV, PML, AM, LG, MG, AS, DM.
Financial support. This work was supported by the Government of Canada via its COVID-19 Immunity Task Force (P. M. L.). B. A. was funded by Michael Smith Health Research BC. F. R. was funded by the German Research Foundation (Deutsche Forschungsgemeinschaft) RE 4598/1-2. P. M. L. received grant salary support from the British Columbia Children’s Hospital Foundation.
Potential conflicts of interest. No relevant conflicts of interest to declare.
Acknowledgments
The authors wish to acknowledge BC Children’s Hospital Research Institute Core Technologies and Services, Vancouver, Canada for providing the support to perform the flow cytometry experiments. The authors would like to thank Jeffrey Bone, Biostatistical Lead at BC Children’s Hospital Research Institute for providing input on statistical analysis.
References
Author notes
Bahaa Abu-Raya and Frederic Reicherz are first co-authors.
