Level of maternal antibodies against respiratory syncytial virus (RSV) nucleoprotein at birth and risk of RSV very severe lower respiratory tract infection

Abstract Background The nucleoprotein (N protein) of respiratory syncytial virus (RSV) is a candidate antigen for new RSV vaccine development. The aim of the present study was to investigate the association between maternal antibody titers against the RSV N protein at birth and the newborns' risk of developing very severe lower respiratory tract infection (VS‐LRTI). Methods In this single‐center prospective cohort study, 578 infants born during the RSV epidemic season in France were included. Among these, 36 were hospitalized for RSV VS‐LRTI. A generalized linear model was used to test the occurrence of a VS‐LRTI in function of sex, mode of delivery, parity of the mother, type of pregnancy, date of birth in relation to the peak of the epidemic, and antibody titer against N protein. Results All cord blood samples had detectable antibodies against N protein. The mean titers were significantly lower in newborns with risk factors for RSV severe LRTI (preterm infants, birth before the peak epidemic, multiparous mother). There was no association between antibody titer against the N protein and a protection against VS‐LRTI. Conclusions Further studies are needed to support the hypothesis that transfer of maternal antibodies against the RSV N protein can provide a significant immune protection early in infancy and that N protein candidate vaccine may be a suitable target for maternal vaccine.

was no association between antibody titer against the N protein and a protection against VS-LRTI.
Conclusions: Further studies are needed to support the hypothesis that transfer of maternal antibodies against the RSV N protein can provide a significant immune protection early in infancy and that N protein candidate vaccine may be a suitable target for maternal vaccine.  1 It is recognized to cause substantial mortality in low-and middle-income countries, 2 and major lower respiratory tract infection burden in high-income countries 3,4 ; the highest rate of hospitalization is in infants aged less than 3 months. 5 Despite this, preventive measures are to date limited to the monthly injection of RSV-specific neutralizing antibodies for children at a high risk of severe complications, the cost of which may be prohibitive. 6 There are, however, vaccines currently in development that have the potential to be a more affordable option to reduce the worldwide RSV burden. 6 Most candidate vaccines target the fusion protein, which is an envelope protein highly conserved between different subtypes or RSV, stabilized in its prefusion form. 7,8 However, RSV nucleoprotein (N protein), which is implicated in nucleocapsid-RNA complex formation that allows RSV replication and transcription, 9 is one of the most conserved viral proteins among strains 10 and is also considered as a potential target for candidate vaccines inducing T CD8+ response. 11,12 These vaccines may be good candidates for maternal immunization strategy, an approach that appears as safe because of the particular vulnerability of infants and their inability to produce effective antibodies. 13 Such a strategy has already demonstrated its ability to provide a passive humoral protection for infants against tetanus, influenza, and pertussis. 14 Humoral protection is conferred at birth by the transfer of maternal IgG, the titer of which depends on the maternal level of RSV antibodies and the effectiveness of transplacental RSV antibody transfer from mother to infant. 15 Although the protection conferred to infants by neutralizing antibodies induced by natural exposure to RSV of mothers has been suggested, published data are conflicting [16][17][18][19][20] ; furthermore, there are no published data concerning the protection provided by maternal antibodies against the N protein. The aim of the present study was therefore to investigate the association between maternal antibody titers against the RSV N protein at birth and the newborns' risk of developing RSV VS-LRTI early in infancy. According to local protocols, all infants hospitalized with an LRTI diagnosis were tested for RSV on a nasopharyngeal sample. Laboratoryconfirmed RSV infection were those diagnosed by real-time reverse transcriptase (RT)-PCR, as previously described. 25 Clinical records were reviewed to further classify these as VS-LRTI, as defined by the WHO: cough or difficulty breathing, associated with fast breathing or peripheral capillary oxygen saturation (SpO2) < 90%, or inability to feed or unconscious. 13 The peak of the epidemic was defined as December 9, 2019, based on both pediatric emergency admissions for bronchiolitis and RSV detection data in Lyon, in accordance with data from previous years.

| Laboratory procedures
Cord bloods were collected at birth in heparinized tubes, centrifuged, and stored at À20 C until testing. IgG against RSV N protein were measured in 1:100 diluted plasma using an in-house enzyme-linked immunosorbent assay (ELISA) performed according to Roux

| Statistical analysis
Mean and SD were calculated to describe clinical continuous variables. Concerning antibody titers against RSV N, both mean (SD) and median (interquartile range [IQR]) were calculated. Comparisons between mean antibody titers were tested using a two-sided parametric Student t-test or an ANOVA for more than two groups. Normality assumption for the antibody titers was visually inspected before performing the ANOVAs and the Student tests. The birth weight and the gestational age were treated as categorical variables. A generalized linear model (GLM) was used to test the occurrence of a VS-LRTI (with a binomial distribution and a logit link function) in F I G U R E 1 Flowchart. LRTI, lower respiratory tract infection function of sex, mode of delivery, parity of the mother, type of pregnancy, date of birth in relation to the peak of the epidemic (all previous variables are qualitative), and titer antibodies against RSV N (quantitative variable). Preterm infants as well as patients who contracted a VS-LRTI but for whom the result for RT-PCR RSV detection was negative were excluded from the GLM. Statistical analyses were performed using R software version 4.0.3.

| Ethics
This study was approved by the review board of university hospitals of Lyon is registered on ClinicalTrials.gov (NCT04144816) and was declared to national data protection commission   (n = 32) were born to multiparous mothers, and none were from multiple births (Table 1).

| Maternal antibody titers against the RSV N protein
The mean (SD) value of maternal antibody titers against the RSV N protein among newborns was 502 (255) RU/mL. There was no significant difference in mean values according to sex (Figure 2A), type of pregnancy ( Figure 2B), or mode of delivery ( Figure 2C). The mean value was significantly different according to the term of pregnancy (P < 0.001; Figure 2D); it was significantly lower in preterm newborns (i.e., extremely, very, and moderate-to-late combined) than in those born at term (377 vs. 516 RU/mL, P < 0.001), in newborns with a low birth weight than in those with a birth weight ≥2500 g (401 vs. 515 RU/mL, P < 0.001; Figure 2E), in newborns from primiparous mother than those from multiparous mothers (470 vs. 522 RU/mL, P < 0.05; Figure 2F), and in newborns whose birth occurred before the peak of the epidemic than in those born after (458 vs. 542 RU/ mL, P < 0.001; Figure 2G). There was no significant difference in mean antibody titers between infants who experienced a VS-LTRI (irrespective of RSV detection) and those who did not (466 vs. 505 RU/mL, P = 0.37; Figure 2H).

| Prediction of VS-LRTI using clinical variables and serological status
Preterm infants were excluded from the GLM given that a 3-month follow-up was not sufficient to analyze their risk of being admitted to hospital for VS-LRTI (they might have been hospitalized since birth at 3 months of age). In multivariate analysis, maternal multiparity (relative risk, RR: 2.34, 95%CI [1.58; 3.01]) and a date of birth before the peak of the epidemic (RR: 2.84, 95%CI [2.08; 3.59]) were significantly associated with the occurrence of VS-LRTI; there was no significant association between antibody titer and VS-LRTI (Table 2).

| DISCUSSION
In the present study, no association was found between antibody titer against the RSV N protein measured in cord blood and a protection against RSV VS-LRTI. This result differs from the largest cohort studies that analyzed cord blood titers; these studies, conducted in Denmark, Mali, and on American Indian infants, found that a higher concentration of neutralizing antibodies at birth was associated with a lower risk of contracting an RSV LRTI and of being hospitalized for this. 14,15,23 Considering the power of the present study, the sample size is comparable with the study reported by Buchwald et al (587 infants), 18 and greater than that reported by Eick et al (372 infants), 16 which both concluded to a protection conferred by neutralizing antibodies (whereas the study conducted in Alaska [155 infants] and Kenya [90 infants] did not find any association). 17,18 For the present study, owing to the lack of data available regarding antibody titers against RSV N protein, there was no determination of number of subjects required; the data provided herein will therefore help for the power calculation in future studies.
It is of note that the mean antibody titers against RSV N protein at birth were lower in premature infants, those with a low birth weight, born to a primiparous mother, and born before the peak of the epidemic. This profile is not surprising; for instance, it is well known that majority of transplacental transfer of IgG occurs during the third trimester of pregnancy and that preterm newborns have a compromised passive immunity. [27][28][29] Furthermore, multiparous mothers are likely to have been exposed to RSV through household transmission from a co-occupant aged between 2 and 13 years, 30 as well as those delivering at the end of the epidemic season. These factors are also known to be risk factors for severe RSV infection. [31][32][33] The multivariate analysis indicates that this risk is mostly mediated through another pathway that the lack of RSV N antibodies.
The study does have some limitations. The most important is that only patients admitted to the hospital they were born in were considered for the analysis of VS-LRTI; thus, any cases visiting elsewhere may have been missed. However, there are few other pediatric emergency departments in the region; the HFME university hospital is the largest pediatric hospital in the area and the only one with an intensive care unit; thus, it is likely that most cases of RSV VS-LRTI occurring before 3 months of age were captured from either direct admission or transfers from other hospitals. Although VS-LRTI before 3 months of age is a relevant outcome, the stringency of this definition reduces the probability of observing an outcome in some less frequent groups. As a consequence, we were not able to test the impact of prematurity and low birth weight in the multivariable analysis.
Furthermore, detection used ELISA that allows standardization of result but does not quantify neutralizing activity. However, a neutralization assay would not have specifically detect antibodies against N protein that were investigated herein. Finally, the severity of LRTI episodes might have been overestimated; a third of cases (n = 12/36) were classified as VS-LRTI only on the "inability to feed" criterion, and patients hospitalized for LRTI were very likely to have enteral nutrition, sometimes for less than 24 h (data not shown).
However, it is unlikely to have led to a differential selection according to the antibody titer level, and the classification of all episodes of LRTI according to the WHO standardized severity criteria allows comparison with future studies. Finally owing to the lack of data on antibody titers against RSV N protein, we were not able to determinate the number of subjects prior to the study. It is not possible to conclude here to the absence of protection confer by the N-antibody against severe LRTI. Next studies on RSV N antibody level at birth should considered adding others RSV antibody assays and included more patients.
In conclusion, in this study, we did not find any significant association between anti-N antibody titers at birth and VS-LRTI occurrence the first 3 month of life. Further studies are needed to support the hypothesis that transfer of maternal antibodies against the RSV N protein can provide a significant immune protection early in infancy and that N protein candidate vaccine may be a suitable target for maternal vaccine.

ACKNOWLEDGEMENT
We warmly thank the VRS study group in Lyon for their support in this work.