ABSTRACT
Objective: To determine the potential longer-term effects of
maternal antenatal respiratory syncytial virus (RSV) vaccination, we
examined the association between cord-blood
RSV-neutralizing antibodies (RSV-NA) and RSV infections in the first
2-years of life,
RSV-NA at 3-years, and respiratory health to age 5-years.
Methods: Two community-based Australian birth cohorts were
combined. For children with at least one atopic parent, paired serum
RSV-NA levels were compared in cord-blood and at age 3-years. Weekly
nasal swabs were collected in one cohort and during acute respiratory
infections (ARI) in the other. Wheeze history up to age 5-years and
physician-diagnosed asthma at 5-years was collected by parent report.
Results: In 264 children,
each log10increase of cord-blood RSV-NA level was associated with 37% decreased
risk (adjusted incidence-rate-ratio (aIRR) 0.63; 95% confidence
interval (CI): 0.40–1.01) of RSV-ARI and 49% decreased risk (aIRR
0.51; 95%CI: 0.25–1.02) of RSV acute lower respiratory infections
(ALRI) at 12–24 months of age. However, higher cord-blood
RSV-NA was associated with increased risk of all-cause ALRI (aIRR 1.29;
95%CI: 0.99–1.69), wheeze-associated ALRI (aIRR 1.75; 95%CI:
1.08–2.82) and severe ALRI (aIRR 2.76; 95%CI: 1.63–4.70) at age
6–<12 months. Cord-blood RSV-NA was not associated with
RSV-ARI in the first 6-months, RSV-NA levels at 3-years, or wheeze or
asthma at 5-years.
Conclusions: Higher
levels of cord-blood RSV-NA did not protect against RSV infections
during the first 6-months-of-life, time-to-first RSV-ARI, or wheeze or
asthma in the first
5-years of life. Additional strategies to control RSV-related illness in
childhood are needed.
INTRODUCTION Respiratory syncytial virus (RSV) is the most common respiratory virus
identified in young children with acute lower respiratory infections
(ALRI)1. It imposes a substantial global health burden
in children aged <5-years with an estimated 33.1 million
episodes of RSV-ALRI, 3.2 million hospital admissions, and almost 60,000
inpatient deaths globally in 20151. In Australia, most
RSV-associated hospitalizations are in children aged
<6-months, with an estimated hospital admission rate for this
age group of 2,224 per 100,000 child years2. Severe
RSV infections in early life are associated with recurrent wheezing
episodes3 and
asthma in later childhood4. However, debate exists
over whether this relationship is causal5,6. Since the
burden of severe RSV-ALRI requiring hospitalization is greatest during
the first few months of life, before sufficient time has passed to allow
active immunization to induce a protective response, any protection of
newborns during this vulnerable period must come from another source.
A potential prevention strategy is maternal antenatal vaccination, which
seeks to induce high-levels of RSV-neutralizing antibodies (NA) that
cross the placenta and protect young infants7.
Maternal vaccination is widely recommended for preventing other
infectious diseases in infants, including tetanus, influenza, and
pertussis8. In a
recent phase III placebo-controlled randomized controlled trial (RCT) of
4,500 pregnant women in 11 countries, there was a 44.4% (95%
confidence interval (CI):
19.6%–61.5%) reduction in
RSV-related hospitalization due to ALRI in babies of vaccinated
mothers9. However, the study did not meet the
prespecified primary endpoint of decreasing RSV-associated, medically
significant ALRI in the first
90-days of life (vaccine efficacy 39.4%; 97.5%CI: -1.0–63.7).
Nevertheless, the concept of maternal vaccination is supported by RSV
monoclonal antibody administration to preterm infants reducing severe
RSV-ALRI10,11 and observational studies linking high
cord-blood NA to a lower risk of severe RSV infections resulting in
hospitalization12,13.
One concern with maternal vaccination is potential interference with
developing B-cell mediated protective immunity against RSV
infection14. Maternal vaccination may suppress the
infant’s immune response, with even low maternal antibody transference
inhibiting infant RSV-specific B-cell and antibody
responses15,16. However, unlike with monoclonal
antibodies17,18, few studies have evaluated the impact
of maternal RSV antibodies upon respiratory infection or lung health
beyond early infancy13,19,20.
The aim
of this study was to use data from two prospective community-based
Australian birth cohorts to examine the association between
maternally-derived RSV-NA levels in cord-blood and (i) RSV infections
and ALRI in the first 2-years of life, (ii) RSV-NA levels at age
3-years, and (iii) respiratory health to 5-years of age.METHODS
Study populations Data from two prospective community-based Australian birth cohorts, the
Childhood Asthma Study (CAS)21 and Observational
Research in Childhood Infectious Diseases (ORChID)
project22, were combined. Both studies collected
cord-blood from participants and are described in detail
elsewhere3,21-24. In brief,
based in Perth, Western Australia,
the CAS was a prospective birth cohort study of 263 infants at high-risk
of atopy (at least one parent with a history of physician-diagnosed
eczema, hayfever, or asthma)3,21. Participants were
born between July 1996 and July 1999. Parents completed daily symptom
diaries for the first 5-years of life. Whenever any respiratory illness
occurred, the research team was contacted, and nasopharyngeal aspirates
were collected in the child’s home within 48-hours of symptom onset and
returned immediately to the laboratory for processing and storage at
-80°C. A history of wheeze and asthma was collected annually. Blood was
collected at 3-years of age.
The ORChID study progressively enrolled 158 unselected healthy, term
newborn infants in Brisbane, Queensland, between September 2010 and
October 201222. Participants were followed from birth
until their second birthday.
Parents completed a daily
respiratory symptom diary and
collected anterior nasal swabs weekly using a single swab for both
nostrils. Specimens were mailed to the laboratory, taking a median
3-days (interquartile range
2–4 days) to reach the laboratory where they were stored at -80°C.
ORChID participants were subsequently invited to have an annual clinical
and respiratory assessment review until 7-years of age as part of the
Early Life Lung Function (ELLF) study25. Blood was
collected as part of the 3-year assessment. To more closely align with
the CAS cohort, a high-risk sub-group of the ORChID cohort was
identified. Children were included if there was a history of parental
asthma or eczema. At enrollment, participants from both the ORChID and
CAS cohorts had socio-demographic and clinical information recorded,
while breastfeeding and childcare attendance were collected
progressively.
The King Edward Memorial and Princess Margaret Hospital Ethics
Committees approved the CAS study (RE95-17.9 and ECO2-53.9). The Royal
Brisbane and Women’s Hospital (HREC/10/QRBW125) Human Research Ethics
Committee (HREC) approved the ORChID study. The Children’s Health
Queensland (HREC/10/QRCH/16 and HREC/12/QRCH/23) and The University of
Queensland (2010000820 and 2014000212)
HRECs approved both the ORChID and
ELLF studies. Informed parental consent was obtained for all
participants in each study.
Acute respiratory infections and asthma categorizationAcute respiratory infections (ARI) were categorized hierarchically as
either ALRI or upper respiratory infections (URI)26.
Similar clinical definitions were used in both cohorts (Supplementary
Table 1). An ALRI was any combination of rattly breathing, moist cough,
shortness of breath, wheeze, or physician-diagnosed pneumonia. If wheeze
was present, an ALRI was sub-categorized as a wheeze-associated ALRI
(wARLI), while an ALRI with temperature > 38°C was
classified as a severe ALRI (sALRI).
A URI included at least one of the
following: nasal discharge/congestion, dry cough, or physician-diagnosed
acute otitis media. Symptomatic RSV episodes in the ORChID study were
identified if symptoms were first detected in the week prior to, or were
present in the week after, the first virus
detection27. Only symptomatic episodes were identified
in the CAS study. Asthma was defined by physician diagnosis of asthma
and wheeze present in the previous year.
Laboratory tests
(i) Analysis of nasal swabs and nasopharyngeal aspiratesValidated real-time polymerase chain reaction (PCR) assays were used to
batch-test stored respiratory specimens for RSV21,22.
(ii) RSV NA assays Virus neutralization in serum from cord-blood and blood collected from
3-year-old participants was assessed against recombinant green
fluorescence protein (GFP)-expressing RSV (rgRSV) derived from D53,
strain A2 that was provided by Prof Mark Peeples28.
Vero cells were seeded into 96 well, optical bottom/black wall cell
culture plates (Thermo Fisher Scientific, Swedesboro, NJ, USA) at a
density of 4x104 cells per well in 100µl of Opti-MEM
(Gibco, Thermo Fisher Scientific, Swedesboro, NJ, USA) with 3% fetal
calf serum and incubated overnight at 37°C in 5% CO2.
The following day, human plasma samples were titrated by
4-fold dilutions in Opti-MEM within a round bottom 96 well plate. Plasma
dilutions were then mixed with 2 x 103 plaque-forming
units/mL of rgRSV and incubated for 1-hour at room temperature before
addition to Vero cells. Following 6-days incubation at 37°C in 5%
CO2, GFP fluorescence intensity was measured on a
CLARIOstar Microplate Reader (BMG LABTECH, Melbourne, VIC, Australia).
Fluorescence was graphed against plasma dilution and a three-parameter
dose-response curve was fitted by nonlinear regression using GraphPad
Prism v7.0.0 (GraphPad Software, San Diego, CA, USA).
OutcomesThe childhood respiratory outcomes available for assessment were
RSV-related ARI and ALRI, and all-cause ALRI, wALRI and sALRI in the
first 2-years of life; serum RSV-NA levels at age 3-years; wheeze
history up to 5-years of age and physician-diagnosed asthma by age
5-years.
AnalysisSummary statistics are presented as frequency (percentage) for
categorical variables. Serum RSV-NA half-maximal inhibitory
concentration (IC50) values were normalized by
log10 transformation. Associations between cord-blood
RSV-NA level and rate of RSV infections in the first 2-years of life
were presented as incidence rate ratios (IRR) using mixed-effects
Poisson regression with study (CAS/ORChID) included as a random-effect.
Models were offset by the natural logrithim of each child’s time-at-risk
of infection. First, univariable models were constructed, then models
were adjusted for the potentially confounding variables season of birth
and presence of an older child in the household at birth. Secondly,
time-to-event analyses were conducted using Cox proportional hazards
regression models. Sensitivity analysis was conducted using all ORChID
children who provided cord-blood to evaluate whether there were
differences between symptomatic and asymptomatic RSV infections.
Thirdly, the association between cord-blood RSV-NA level and serum
RSV-NA levels at age 3-years was analyzed using linear regression.
Finally, the association between cord-blood RSV-NA level and risk of
wheeze at ages 1 to 5-years and asthma at 5-years of age was analyzed
using both univariable and multivariable mixed-effects logistic
regression models. Study was included as a random effect. Multivariable
models were adjusted for season of birth and presence of older child in
the household at birth. Data were analyzed using Stata v13 (StataCorp,
College Station, TX, USA).
RESULTS
Participants Cord-blood RSV-NA data were available for 109/158 (69%) ORChID and
214/236 (91%) CAS children. Fifty children from the ORChID cohort had
at least one atopic parent and were combined with the 214 children from
the CAS cohort to form the high-risk group (Figure). Characteristics of
high-risk children by each study cohort are described in Table 1. All
had RSV-NA detected in cord-blood (log10IC50 values ranged from 1.57 to 4.19) and except for
paternal atopy, their characteristics were similar across the study
cohorts.
Association between cord-blood RSV-NA level and risk of
RSV-ARIs, including
RSV-ALRIs, and all-cause ALRIs in the first 2-years of life The association between cord-blood RSV-NA levels and ARI episodes are
displayed in
Table 2. Each log10 IC50 increase in
cord-blood RSV-NA level was associated with a 37% decreased risk of
RSV-ARI (adjusted IRR 0.63; 95%CI: 0.40–1.01) and a 49% decreased
risk of RSV-ALRI (aIRR 0.51; 95%CI: 0.25–1.02) in the second-year of
life. In contrast, in the first-year of life higher RSV-NA was
associated with an increased risk of all-cause ALRI (age 0–<6
months; aIRR 1.41; 95%CI: 0.98–2.04; age 6–<12 months; aIRR
1.29; 95%CI 0.99–1.69), which at age 6–<12 months was
significant for wALRI (aIRR 1.75; 95%CI: 1.08–2.82) and sALRI (aIRR
2.76; 95%CI: 1.63–4.70).Time-to-first RSV detectionRSV-NA level was not associated with time-to-first symptomatic RSV
detections (Supplmentary Table 2 and Supplementary Figure 1). With each
log10 IC50 increase in
RSV-NA levels, the hazard ratio (HR) decreased slightly by 11%
(adjusted HR (aHR) 0.89; 95%CI: 0.62–1.29) for RSV-ARI, and by 10%
(aHR 0.90; 95%CI: 0.54–1.51) for RSV-ALRI. To test the sensitivity of
this result, we repeated the analysis using all 109 ORChID children who
provided cord-blood (Supplementary Table 3), and RSV-NA level was not
associated with time-to-first RSV detection (Supplementary Table
4).Association between serum RSV-NA levels in
cord-blood and at age 3-years There was no significant association between paired serum RSV-NA levels
in cord-blood and at 3-years of age in the 192 children for whom paired
data were available (Supplementary Figure 2).
Cord-blood RSV-NA levels and wheezing episodes during the first
5-years of life and asthma at age 5-years No significant associations between cord-blood RSV-NA levels and risk
of wheeze during the first 5-years of life and physician-diagnosed
asthma at age 5-years were observed (Table 3).
DISCUSSION In this combined cohort of community-based children at increased risk
of asthma, there was weak evidence that each log10increase in maternally-derived RSV-NA levels in cord-blood was
associated with a lower risk of both RSV-ARI and RSV-ALRI in the second
year of life. Cord-blood RSV-NA levels were not associated with the
timing of symptomatic RSV infections in the first 2-years of life. In
contrast, each log10 increase in cord-blood RSV-NA level
may be associated with increased risk of all-cause ALRI in the
first-year of life, with this association statistically significant for
wALRI and sALRI episodes in children aged 6–<12 months. There
was no association between RSV-NA cord-blood levels and wheezing in the
first 5-years of life or an asthma diagnosis at age 5-years.
Both hospital and community-based studies that have shown higher
cord-blood RSV antibody levels are associated with older age at
infection, decreased hospitalization, and overall less severe
disease13,29-31. However, this is not a universal
finding and studies in higher-risk populations from Alaska, Kenya, and
Nepal have not shown a protective effect for cord-blood antibodies on
early RSV infection risk or severity32-34. Our study
found that that higher cord-blood RSV-NA levels were associated with a
decreased risk of RSV infections at
12–24 months, but not in the first 6-months when a substantial decline
in maternally-derived antibody levels occurs. A possible explanation for
our failure to identify an association between cord-blood RSV-NA and the
timing of RSV infections during infancy is that our cohorts were
community-based, and there were relatively few cases in the first
6-months of life. In contrast, hospital-based studies reporting high
cord-blood RSV-NA levels protect against RSV disease during infancy are
comprised predominantly of infants experiencing severe infections in the
first-months of life12,30.
Our finding that children at high risk of atopy who had the highest
cord-blood RSV-NA levels were also at increased risk of all-cause ALRI,
and especially wALRI and sALRI, episodes in the first-year of life, but
at lower risk of RSV-ARI and RSV-ALRI in their second-year was
unexpected. It may be that high RSV-NA levels represent maternal
exposure to multiple other viruses, rather than RSV alone, and the child
develops immunity from prior RSV exposure in their second 6-months of
life. While potential confounding factors such as season of birth and
presence of children in the household at birth were included as
covariables in the regression analyses, additional unmeasured
environmental/lifestyle and genetic factors may have influenced the
result. This finding was similar to a Danish birth cohort study, where
they reported that although elevated cord-blood RSV-NA levels protected
infants against hospitalization from RSV infection in the first 6-months
of life (aIRR 0.74; 95%CI: 0.62–0.87), they were also associated with
increased risk of recurrent wheeze (aIRR 1.28; 95%CI:
1.04–1.57) in the same age group35. While serum
antibodies may protect against ALRI, there is emerging evidence that
local mucosal immunity, especially for URI, may also be important. Adult
human RSV challenge trials found that pre-existing nasal IgA levels
correlated with protection, whereas no such relationship was observed
for serum NA levels, despite these being moderately high in the healthy
volunteers36. Furthermore, although the pivotal
studies use of monoclonal RSV antibodies in young infants significantly
reduced RSV-ALRI hospitalizations, they did not affect milder RSV
infections, including URI episodes10, which comprise
the greatest overall community burden of disease27.
An alternative prevention strategy to maternal vaccination is newborn
monoclonal antibody administration. Palivizumab has been licensed since
1998. However, it has no effect upon RSV-URI and its impact upon RSV
hospitalizations on a population basis is limited by its high cost. This
has restricted its use to a small number of babies deemed to be at
highest risk of severe RSV infections, such as those requiring home
oxygen for chronic neonatal lung disease or with complex congenital
cyanotic heart disease37. The new monoclonal anti-RSV
antibody, Nirsevimab, has more than three-times the half-life of
palivizumab11. Thus one dose can potentially provide
protection for the typical RSV season of 5–6 months. In a phase III
placebo-controlled, RCT involving 1,500 preterm infants from 23
countries, there was a 78.4% (95%CI: 51.9–90.3) reduction in
RSV-associated hospitalizations in infants receiving Nirsevimab compared
to placebo up to 150-days after the dose was
administered11. This strategy is seen as providing
protection during the vulnerable period in the first months of life when
the lower airways are of small caliber, the immune system is immature,
and maternal RSV NA levels are declining. While maternal antenatal RSV
vaccines and long-acting monoclonal RSV antibodies may protect against
severe ALRIs in young infants, which of itself is important, their
overall impact on RSV disease will be limited in older infants and
children and will require additional strategies, such as that provided
by approved RSV vaccines.
While passive immunoprophylaxis can reduce RSV-associated
hospitalization, the preventative effect on subsequent wheeze is still
unclear. A recent systematic review and meta-analysis of observational
studies and RCTs evaluated the association of RSV-ALRI on subsequent
wheezing illness38. Eight immunoprophylaxis studies
contributed to the analysis, and although the odds point estimate of
subsequent wheezing illnesses was higher in those who had not received
RSV monoclonal antibodies (palivizumab or motavizumab), the effect was
weak (odds ratio 1.21; 95%CI: 0.73–1.99). A 2020 World Health
Organization Report reached the same conclusion that there was
insufficient evidence to support a causal relationship between RSV-ALRI
and subsequent wheezing
illnesses5.
The major strength of the
present study is that we combined data from two high-quality Australian
birth cohorts. Although the two studies were conducted by different
research teams, and consequently the study procedures were not
identical, we worked with both research teams to combine the different
data sets and standardize results. The RSV-NA from all samples were
measured in the same laboratory, ensuring consistent results. There are
however several limitations. Firstly, parent-reported symptoms were
captured, and only otitis media and pneumonia were doctor-diagnosed. To
reduce misclassification of symptoms, parents were trained to recognize
the respiratory symptoms itemized in the daily
diary26. Secondly, the diagnosis of asthma relied upon
assessment by a physician and was not supplemented by more objective
measures, such as the forced oscillation technique, to determine airway
resistance and compliance. Nevertheless, and although controversial,
clinical assessment is still the most reliable means of diagnosing
asthma in this age group39. However, a history of
parent-reported wheeze is subject to information bias, including
reporting error with differential misclassification. Thirdly, we were
not able to assess the protective effect of maternal
RSV-NA on RSV-ALRI occurring in the first 3-months due to the small
number (n=6) of cases. Fourthly, CAS captured only symptomatic RSV
episodes and thus asymptomatic RSV infections in this cohort went
undetected. Finally, while serum RSV-NA represents the main immune
correlate of protection, a standardized protective threshold has not
been defined40.
RSV remains an important cause of morbidity and mortality in infants
worldwide. Although RSV vaccines are a global priority, none are
licensed at present. The data from the present study suggest that in
contrast to studies focusing upon hospitalization, high levels of
transplacental RSV-NA that might be achieved by maternal immunization
may not decrease overall RSV infections in the first 6-months of life
and may not protect against wheeze and asthma developing in early
childhood at a community level. Additional strategies to control RSV
disease beyond the first-months of life are needed.