The association of neonatal SARS-CoV-2 anti-spike protein receptor-binding domain antibodies at delivery with infant SARS-CoV-2 infection under the age of 6 months: a prospective cohort study.

LAH designed the study and prepared the protocol. LOA, EE, and RD participated in patient enrolment and data collection. LAH, MO, and RC coordinated the project and supervised fieldwork. Data analysis was perfomed by GC, LAH, and LOA. LAH and LOA wrote the first draft of the manuscript. All the authors were involved in the interpretation of the findings and revision of the manuscript. The corresponding author has full access to all data in the study. All the authors approved the final version of the manuscript.

This prospective observational study was conducted at the Rabin Medical Center, Israel. Mothers were enrolled from September 2021 to mid-February 2022, before transfer to the delivery room, and followed at the Maternity and Infant Ward. The study was approved by the institutional ethics committee of the Rabin Medical Center (RMC-0275-21). Written informed consent was obtained from the participating mothers. Inclusion criteria included maternal recipient of at least one dose of the BNT162b2 mRNA COVID-19 vaccine or prior microbiologically confirmed natural SARS-CoV-2 infection. Exclusion criteria included (a) preterm birth at <34 weeks of gestation and (b) congenital or acquired immune deficiency. Since 1 February 2021, pregnant women were eligible for free-of-charge BNT162b2 mRNA COVID-19 vaccination during all trimesters of pregnancy, by the Israeli Ministry of Health. A booster dose for pregnant women was available since 12 August 2021. During the study period, the SARS-CoV-2 PCR testing policy was permissive, with free and available testing for all ages. After delivery, cord blood sera were collected for quantitative measurement of IgG antibodies against the spike receptor-binding domain of the S1 subunit of the spike protein of SARS-CoV-2. Data were collected from mothers' and infants' electronic records and by questionnaires filled by the participating mothers at the time of enrolment. Follow-up virtual interviews were made during the 6-month period, for the collection of data on infant growth, breastfeeding, and microbiologically confirmed SARS-CoV-2 infection, at two pre-defined intervals, 90 + 30 days and 180 + 30 days. Indications for SARS-CoV-2 PCR testing were clinical symptoms or exposure to an infected individual. Primary outcomes were neonatal anti-spike antibody levels at delivery and infant microbiologically confirmed SARS-CoV-2 infection at the age of <6 months. The timing of maternal vaccination was categorized according to the last vaccine dose, as booster dosing was not administered at fixed intervals from the second vaccine dose. A positive SARS-CoV-2 PCR test was defined according to the Israeli Ministry of Health. Testing methods included the Real-Time Fluorescent PCR kit (BGI) and the SARS-COV-2 PCR kit (Seegene). Anti-spike receptor-binding domain IgG antibodies were measured using the SARS-CoV-2 IgG II Quant assay (Abbott); a result of ≥50 AU/mL was defined as positive. Continuous parameters were compared between groups, using the 2-tailed independent t test or one-way analysis of variance and the Mann-Whitney U test for parametric and non-parametric variables, respectively. Fisher's exact test was used to compare categorical variables between groups. Linear regression and polynomial interpolation smoothing were used to assess the association between the time elapsed since the last vaccine dose and antibody level. Linear regression was used to model the relationship between pre-defined independent variables selected on the basis of clinical sense and the literature and level of antibodies (dependent variable). Collinearity was tested by calculating the variance of inflation factor; variance of inflation factor of <5.0 indicated the absence of collinearity. Univariate logistic regression was used to assess the association between antibody titres (independent variable) and infant infection (dependent variable). The Wald test was used for CI calculation. Models' goodness-of-fit was indicated by R-squared. Results are presented based on the full data set. There were no missing data. Interactions were systematically searched for. Data were analysed using IBM SPSS Statistics for Windows, version 27.0 (IBM Corp. Released 2020 and R Core Team, 2016, Armonk, New York), R: A Language and Environment for Statistical Computing (Vienna, Austria). p < 0.05 was considered statistically significant.

Sixty-seven mothers were enrolled. Nine of them did not meet the study eligibility criteria. Of the 58 mothers included ( Fig. 1 , Table 1 ), the median maternal age at the time of delivery was 34 years (interquartile range, 30–37 years). Documented dates of vaccination and natural infection were obtained for all 58 mothers. Fifty-two of them received BNT162b2 mRNA COVID-19 vaccination. Twelve mothers had recovered from a mild symptomatic infection (Table S1). Fig. 1 Study flow chart. Fig. 1 Table 1 Demographic and clinical characteristics of 58 participating mother-infant dyads Table 1 Characteristic n (%) or median (IQR) Maternal age (y) 34 (30–37) Background systemic illnessa 14 (24) High-risk pregnancy 22 (38) Pharmacological treatment during pregnancyb 18 (31) Gestational age at delivery (wk) 39 (38–40) Intrapartum complicationsc 9 (15) Maternal prior SARS-CoV-2 natural infection 12 (21) Maternal COVID-19 severityd Asymptomatic 4 (33) Mild 8 (67) Moderate 0 Severe 0 Interval from maternal SARS-CoV-2 infection to delivery (d) 114 (46–378) Maternal mRNA COVID-19 vaccine doses 0 6 (10) 1 2 (3) 2 15 (26) 3 35 (60) Mothers fully vaccinatede 51 (82) Interval from the last vaccine dose to delivery (days) 97 (62.25–177.50) Neonatal sex, male 32 (53) Neonatal birth weight (g) 3284 (2626–4220) Neonatal SARS-CoV-2 anti-spike IgG titres, log AU/mL 3.77 (3.19–4.36) Infant nutrition until the age of 6 mo, n  = 57 Exclusive breastfeeding 30 (58) Formula 11 (19) Combination 16 (28) Continuous variables are expressed as medians (IQR); categorical variables are expressed as numbers (percentage). IQR, interquartile range; mRNA, messenger RNA. a Hypothyroidism (4), diabetes mellitus (3), Crohn's disease (1), endometriosis (1), Sjogren's syndrome (1), asthma (1) idiopathic thrombocytopenia purpura (1), antiphospholipid syndrome (1), and epilepsy (1). b Metformin (6), insulin (4), aspirin (2), levothyroxine sodium (4), doxylamine + pyridoxine (1), levetiracetam (1), ursodeoxycholic acid (1), salbutamol (1), and hydroxychloroquine sulfate (1). c Vacuum extraction (5) and caesarean delivery (3) due to fetal bradycardia, meconium-stained amniotic fluid (1). d Defined according to the National Institutes of Health criteria. e A fully vaccinated mother was defined as a mother who received at least two doses of mRNA vaccine or prior COVID-19 followed by one vaccine dose. Study flow chart. Demographic and clinical characteristics of 58 participating mother-infant dyads Continuous variables are expressed as medians (IQR); categorical variables are expressed as numbers (percentage). IQR, interquartile range; mRNA, messenger RNA. a Hypothyroidism (4), diabetes mellitus (3), Crohn's disease (1), endometriosis (1), Sjogren's syndrome (1), asthma (1) idiopathic thrombocytopenia purpura (1), antiphospholipid syndrome (1), and epilepsy (1). b Metformin (6), insulin (4), aspirin (2), levothyroxine sodium (4), doxylamine + pyridoxine (1), levetiracetam (1), ursodeoxycholic acid (1), salbutamol (1), and hydroxychloroquine sulfate (1). c Vacuum extraction (5) and caesarean delivery (3) due to fetal bradycardia, meconium-stained amniotic fluid (1). d Defined according to the National Institutes of Health criteria. e A fully vaccinated mother was defined as a mother who received at least two doses of mRNA vaccine or prior COVID-19 followed by one vaccine dose. One maternal-infant dyad was lost to the 6-month follow-up and was excluded from the analysis of infection. Of the 57 infants with a 6-month follow-up, 12 (21%) tested positive for SARS-CoV-2, at a median age of 2 months (interquartile range, 2–3 months). Ten infants were diagnosed during circulation of the omicron variant and two during the delta-predominant period. Most infants (7, 58%) had mild symptomatic infections, 5 (42%) were asymptomatic, and none exhibited moderate or severe disease or required hospitalization. Fifty-seven (98%) infants were positive for anti-spike IgG at delivery. Mean log anti-spike IgG concentrations in cord blood sera were 3.77 ± 0.83 AU/mL. Among infants of vaccinated mothers, neonatal anti-spike IgG titres increased with increasing numbers of vaccine doses: mean 1.53 ± 0.0 AU/mL following one vaccine dose, 3.35 ± 0.65 AU/mL following two vaccine doses, and 4.05 ± 0.53AU/mL following the booster dose, p < 0.001. Similarly, among infants of recovered mothers, the mean log antibody titres were 2.49 ± 0.69 AU/mL with no recipient of any vaccine, 4.90 ± 0.0 AU/mL following one vaccine dose, 4.50 ± 0.22 AU/mL following two vaccine doses, and 4.74 ± 0.0 AU/mL following the booster dose; p < 0.001. Among infants of fully vaccinated mothers without prior infection ( n  = 45), neonatal mean log anti-spike IgG titres decreased with increased time since the last vaccination; 4.27 ± 0.6 AU/mL following administration of the last vaccine dose at ≥30 weeks gestation, 4.04 ± 0.7 AU/mL at 27–29 weeks gestation, 3.81 ± 0.6 AU/mL at <27 weeks gestation, and 3.03 ± 0.5 AU/mL at administration before pregnancy; p < 0.001. Linear regression showed an overall linear decrease in titres of 0.004 log AU/mL ([95% CI, 0.006, 0.003], p < 0.001) per follow-up day elapsing since vaccination. Thus, during 24 weeks of pregnancy, infant log titres were anticipated to decrease by 0.672 AU/mL. A spline curve model showed decreasing titres throughout pregnancy until delivery when the last dose was administered at ≤30 weeks of gestation (50 days before delivery; Fig. 2 ). However, when the last maternal vaccination dose was administered at ≥31 weeks of gestation, an opposite trend of increasing titres with time until delivery was shown (R 2  = 0.5). Fig. 2 Smooth spline curve for anti-spike antibody level as a function of the time since the last vaccine dose. The model shows the relation between neonatal SARS-CoV-2 anti-spike antibody levels and the time since the last maternal COVID-19 mRNA vaccination, in fully vaccinated mothers with a maximal follow-up of 350 days ( n  = 44, R 2  = 0.5). The solid line denotes the fitted curve, and the grey area represents 95% CI. In the administration of the last vaccine dose at ≤30 weeks of gestation, the model shows a decrease in antibody level with the time elapsed since the last vaccine dose. An opposite trend of increasing titres is apparent when the last vaccine dose was administered at ≥31 weeks of gestation. Fig. 2 Smooth spline curve for anti-spike antibody level as a function of the time since the last vaccine dose. The model shows the relation between neonatal SARS-CoV-2 anti-spike antibody levels and the time since the last maternal COVID-19 mRNA vaccination, in fully vaccinated mothers with a maximal follow-up of 350 days ( n  = 44, R 2  = 0.5). The solid line denotes the fitted curve, and the grey area represents 95% CI. In the administration of the last vaccine dose at ≤30 weeks of gestation, the model shows a decrease in antibody level with the time elapsed since the last vaccine dose. An opposite trend of increasing titres is apparent when the last vaccine dose was administered at ≥31 weeks of gestation. The mean log anti-spike IgG level was significantly lower among infants who were positive versus negative for SARS-CoV-2 infection during the 6-month follow-up period (3.41 ± 0.74 AU/mL [ n =  12] vs. 3.87 ± 0.84 AU/mL [ n =  45], p 0.036) ( Fig. 3 ). Fig. 3 Anti-spike antibody levels according to infant SARS-CoV-2 infection until the age of 6 months. Box plots for anti-spike antibody levels (AU/mL), measured over the entire study cohort ( n  = 57), according to infant SARS-CoV-2 infection during 6 months of follow-up. The black line corresponds to the group median. The box plots indicate values between the 25 and 75 percentiles. The dotted line represents the optimal diagnostic cut-off of 4.07 AU/mL (11 750). Fig. 3 Anti-spike antibody levels according to infant SARS-CoV-2 infection until the age of 6 months. Box plots for anti-spike antibody levels (AU/mL), measured over the entire study cohort ( n  = 57), according to infant SARS-CoV-2 infection during 6 months of follow-up. The black line corresponds to the group median. The box plots indicate values between the 25 and 75 percentiles. The dotted line represents the optimal diagnostic cut-off of 4.07 AU/mL (11 750). A receiver operator characteristic curve was plotted to identify the best diagnostic cut-off of antibody level for differentiating COVID-19 positive from negative infants. The area under the receiver operator characteristic curve was 0.70 (95% CI, 0.60, 0.84); the optimal diagnostic performance was attained at a cut-off of 4.07 log AU/mL (11 750). Logistic regression showed a strong association of antibodies at delivery with a decreased risk of infant SARS-CoV-2 infection at the age of <6 months; antibody level of ≥4.07 log AU/mL at delivery was associated with a reduced likelihood for infection, 1 of 26 versus 11 of 32 in infants with antibody level of <4.07 log AU/mL, OR of 0.076 (95% CI, 0.076, 0.64), p 0.018. There was no significant difference between infants who were positive versus negative for SARS-CoV-2 infection in maternal age, breastfeeding, or day care attendance (Table S2). In linear regression analysis, the maternal recipient of a booster dose had the strongest effect on the level of antibodies at delivery. An increasing number of vaccine doses, prior maternal infection, and recipient of the last vaccine dose at ≥31 weeks of gestation were also significantly associated with a higher level of antibodies at delivery ( Table 2 ). Table 2 Multivariate linear regression analysis of factors affecting neonatal anti-spike antibody level Table 2 Variable Unstandardized coefficients 95% CI p B SE COVID-19 vaccination, per dose 0.68 0.135 0.413, 0.956 <0.001 Last vaccine dose at <27 wk gestation 0.430 0.226 −0.024, 0.884 0.063 Last vaccine dose at 27–30 wk gestation 0.666 0.267 0.130, 1.202 0.016 Last vaccine dose at ≥31 wk’ gestation 0.894 0.234 0.424, 1.364 <0.001 Prior maternal natural SARS-CoV-2 infection 1.159 0..135 0.413, 0.956 <0.001 Model was explored with a full model including pre-defined independent variables selected on the basis of clinical sense and the literature (number of vaccine doses, time of last vaccine dose categorized to three periods, and SARS-CoV-2 natural infection status) [ [4] , [5] , [6] , [7] , 16 , 17 ] and level of SARS-CoV-2 anti-spike antibody titres (dependent variable). There were no missing data. No interactions were found. Variance of inflation factors was between 1.872 and 3.510, indicating absence of collinearity (variance of inflation factor, <5.0). Adjusted R 2  = 0.797. Analysis shows that maternal recipient of a booster dose had the strongest effect on the level of antibodies at delivery. Compared with infants of mothers vaccinated with a booster dose during the third trimester of pregnancy, infants of unvaccinated mothers had 2.706 log AU/mL lower antibodies at birth. Multivariate linear regression analysis of factors affecting neonatal anti-spike antibody level Model was explored with a full model including pre-defined independent variables selected on the basis of clinical sense and the literature (number of vaccine doses, time of last vaccine dose categorized to three periods, and SARS-CoV-2 natural infection status) [ [4] , [5] , [6] , [7] , 16 , 17 ] and level of SARS-CoV-2 anti-spike antibody titres (dependent variable). There were no missing data. No interactions were found. Variance of inflation factors was between 1.872 and 3.510, indicating absence of collinearity (variance of inflation factor, <5.0). Adjusted R 2  = 0.797. Analysis shows that maternal recipient of a booster dose had the strongest effect on the level of antibodies at delivery. Compared with infants of mothers vaccinated with a booster dose during the third trimester of pregnancy, infants of unvaccinated mothers had 2.706 log AU/mL lower antibodies at birth.

This prospective cohort study showed a strong association between neonatal anti-spike antibody level at delivery with decreased risk of infant SARS-CoV-2 infection at the age of <6 months. Accordingly, a titre of ≥11 750 AU/mL at delivery was associated with lower odds of infant SARS-CoV-2 infection. These findings complement recent descriptions of the effectiveness of maternal COVID-19 vaccination during pregnancy against infant infection and hospitalization during this vulnerable period [ 11 , 12 ] and strengthen the recommendation for maternal vaccination for the added benefit of infant protection [ 18 , 19 ]. Our results can contribute to the decision-making process regarding pregnant women and decrease vaccine hesitancy [ 20 ]. A prospective cohort study showed that at the age of 6 months, 43% of infants of mRNA fully vaccinated mothers during pregnancy had no detectable anti-spike antibody level; this compares with 2% at the age of 2 months [ 21 ]. Similarly, for other transplacental vaccine-elicited antibodies, the protective efficacy is reduced by the age of 6–12 months [ 22 ]. Our findings concur with the correlation reported of COVID-19 vaccination-derived immunity, with antibody titres in vaccinated health care workers [ 23 ]. Our study described several significant factors associated with neonatal antibody levels at delivery, namely, the number of maternal vaccine doses and the time since vaccination. Our results corroborate studies that reported higher antibody levels with increasing vaccine doses. A retrospective cohort study demonstrated that fully vaccinated compared with partially vaccinated women transmitted higher levels of anti-spike antibodies to neonates at the time of delivery and that a third trimester booster dose was associated with the highest umbilical cord antibody level [ 24 ]. However, only approximately 2% of the cohort of that study received a booster dose, and all the boosters were given during the third trimester. Our multivariate analysis showed that the maternal recipient of a booster dose had the strongest impact on the level of antibodies at delivery, even more than prior natural infection or timing of maternal vaccination. Compared with natural infection, maternal vaccination was previously reported to result in greater antibody persistence at the age of 6 months [ 21 ]. When the last vaccine dose was administered at ≤30 weeks of gestation, neonatal titres ‘decreased’ with time since vaccination, after which titres increased. This corroborates a previous report of lower anti-spike antibody levels when self-reported maternal vaccination was initiated at an earlier trimester (pre-pregnancy or first trimester vs. second trimester vs. third trimester) [ 24 ]. However, in that study, the maternal vaccine initiated after 32 weeks of gestation was associated with lower neonatal IgG levels at delivery. This discrepancy from our findings can be explained by the different starting points used to calculate the time from vaccination, either the first or the last dose, to delivery. Initiation of mRNA vaccination at ≥32 weeks of gestation, with the second dose given after 3–4 weeks of interval, at 35–36 weeks, would not yield an adequate interval from vaccination to term delivery, for complete maternal immune response and transplacental antibody transfer. Furthermore, in that study, most of the women vaccinated at ≥32 weeks of gestation received only one vaccine dose and were thereby anticipated to transfer lower antibody levels. This interpretation is in line with a prospective cohort study, in which vaccine administration at ≥32 weeks of gestation (10–50 days before delivery), showed an ‘increase’ in cord blood anti-spike antibodies according to the time elapsed from the maternal second vaccine dose to delivery ( n  = 8, r  = 0.81, p 0.01) [ 25 ]. Our model showed a linear decay in antibody levels throughout pregnancy. This finding concurs with a prospective study that evaluated neonatal titres in response to a two-dose regimen given during the second trimester [ 16 ]. This gradual decrease in neonatal titres since vaccination correlates with the two-fold decrease in vaccine effectiveness against infant hospitalization, described by Halasa et al. [ 13 ] when the second dose of the vaccine was given at ≤20 vs. >20 weeks of gestation (38% vs. 69%). Presently, pregnancy alone is not an indication for a second booster [ 18 , 19 ]. However, because of the decreasing vaccine immunity over time, planning a second booster dose in late pregnancy may be necessary to extend the duration of protection of infants, particularly in the woman who completed the vaccination regimen before or during early pregnancy. According to our data, administration of a booster dose during weeks 31–34 of gestation would yield higher transplacental antibody transfer. Our study has several strengths; first, its prospective nature enabled comprehensive and accurate data collection, including the maternal presence of immunodeficiency, documentation of maternal vaccine doses, and perinatal and postnatal outcomes. Second, the study included infants of mothers after pre-pregnancy and booster vaccination, who were not widely evaluated in previous studies. However, as the sample size was small, larger studies are needed to validate our preliminary results regarding a serologic correlate of protection. In addition, differences in protective behaviour between vaccinated and unvaccinated mothers may have had an impact on the risk of infant infection [ 26 ]. Additionally, parturients were not tested for anti-nucleocapsid antibodies; thus, some mothers categorized as ‘vaccinated-only’ may have had prior infection. However, this would not influence the association between antibody level and later infection. In conclusion, our findings suggest that higher neonatal anti-spike antibody levels at delivery are associated with increased infant protection against SARS-CoV-2 infection during the first 6 months of life. This supports routine SARS-CoV-2 maternal vaccination for the benefit of infant protection. As antibody level decreases linearly with time, a second booster dose for pregnant women vaccinated before or during early pregnancy may be considered. Future studies are needed to evaluate the effectiveness of various maternal vaccination strategies to increase the durability of maternally acquired antibodies and achieve protection of infants during the first months of life.

Maternal COVID-19 messenger RNA (mRNA) vaccination is effective against the morbidity of pregnant women [ 1 ]; however, its role in protecting young infants has not been fully elucidated. To date, mRNA COVID-19 vaccination is available from the age of 6 months, leaving infants aged 0–6 months without protection against COVID-19. Although infection at this age is usually mild, the risk of severe illness remains and the rate of hospitalization is not negligible [ 2 , 3 ]. Evidence shows that vaccine-elicited and naturally acquired SARS-CoV-2 functional anti-spike IgG antibodies passively cross the placenta from the mother to the foetus and that maternal and cord blood antibody concentrations correlate positively [ [4] , [5] , [6] ]. Serum anti-spike IgG levels have been shown to correlate with the neutralization of SARS-CoV-2 [ 7 ]. However, placental transfer ratios were reported as relatively lower than other vaccine-elicited antibodies [ [8] , [9] , [10] ]. Preliminary data suggest that maternal COVID-19 vaccination may protect young infants against infection and hospitalization [ 11 , 12 ]. However, a serologic correlate of protection against infant SARS-CoV-2 infection is currently unknown, and whether higher antibody levels result in greater protection during the first months of life remains unanswered. The optimal timing of COVID-19 vaccination during pregnancy, to maximize infant protection, is not known; and it does not necessarily coincide with the optimal timing to protect the mother. As transplacental transfer begins at approximately 17 weeks of gestation [ 13 ] and as a minimal interval of at least 2 weeks is required from vaccination to delivery for maternal antibody production and placental transfer [ 4 ], some authors have recommended maternal COVID-19 vaccination in the second trimester [ 14 , 15 ]. However, others have advocated administration early in the third trimester (first dose at 27–31 weeks) [ 16 ], similar to the recommendation for the pertussis vaccine, at 27–28 weeks [ 17 ]. These recommendations mostly rely on data based on a two-dose vaccine regimen given during the second and third trimesters of pregnancy. However, currently, maternal vaccination is generally earlier, including the administration of a booster dose after the completion of the initial vaccine series. Therefore, the timing of maternal vaccination during gestation and the effect of a booster dose remain to be explored. The main aim of this study was to identify serologic correlates of protection by evaluating the association between neonatal SARS-CoV-2 antibody level at delivery after maternal vaccination or natural infection and infant SARS-CoV-2 infection under the age of 6 months. Our second aim was to describe the predictors of neonatal antibody levels. These data could aid in developing maternal vaccination strategies for maximizing infant protection.

The authors declare that they have no conflicts of interest. No external funding was received.