This was a single-center prospective cohort study conducted at Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico in Milan, Italy, between May 23, 2021 and October 30, 2021.
The study protocol was approved by the ethics committee of the sponsoring center (Comitato Etico Milano Area B–Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico Milano, protocol number 2315) and the ethics committee of the Istituto Nazionale per the Malattie Infettive Lazzaro Spallanzani of Rome (appointed National Ethics Committee for the evaluation of clinical trials and medical devices for the treatment of patients affected by COVID-19. Accreditation number 267). Written informed consent was obtained from nursing mothers before they received the first dose of SARS-CoV-2 vaccine, and all procedures were in accordance with the Declaration of Helsinki34.
We recruited lactating women willing to receive the SARS-CoV-2 vaccine while breastfeeding, and determined to continue breastfeeding for the next four months, exclusively or partially, with at least 50% of the intake. total daily milk reported as breast milk (based on average intake by weight and self-reported amount of formula fed). All women received 2 doses of BNT162b2 anti-SARS-CoV-2 vaccine (Pfizer, Inc., and BioNTech) while breastfeeding, 21 days apart. Women and their infants provided biological samples at the following time points (Fig. 1): the day of the first dose of vaccine (before administration, T0), the day of the second dose of vaccine (before administration, T1) and at age 30. ± 1 (T2), 60 ± 1 (T3) and 90 ± 1 (T4) days after the second dose of vaccine. Women reporting a proven prior episode of SARS-CoV-2 infection were excluded from enrollment, as well as those who tested positive for anti-nucleocapsid (anti-N) antibodies at the time of enrollment. In addition, at each time point, maternal serological screening for anti-N antibodies was performed. Neonatal serological screening for anti-N was performed at T4, to exclude a previous neonatal infection. In case of neonatal anti-N positivity, neonatal samples were excluded from data analysis.
The following maternal data were collected at the time of registration and, where applicable, at each time of the study: age of the mother at the time of vaccination, ethnic origin, mode of delivery, type of breastfeeding at each moment (exclusive or partial), number of daily feedings at each moment. . The following post-vaccination maternal symptoms were recorded: injection site pain, malaise, fever > 38°C, headache, myalgia, nausea, rash, adenitis, diarrhoea. In addition, we recorded the following neonatal data: sex, gestational age (GA) at delivery, premature birth weight, age at maternal vaccination. Data from mothers and their newborns were collected prospectively using an electronic database.
At each study visit, we collected maternal serum, breast milk, maternal saliva, infant feces, and an infant buccal mucosa swab. The infant’s blood for anti-N serological screening was only collected at T4. Maternal and infant blood was collected by venipuncture into serum separator tubes, and serum was obtained after centrifugation at 3000 × g. for 10 min at 4°C. Sera were aliquoted into cryogenic vials and stored at -80°C until further analysis. Breast milk was expressed manually by nursing mothers in the presence of a midwife, after cleaning the nipples, and collected in containers provided by the study. The milk was immediately aliquoted into several cryovials and frozen at -80°C until analysis. Maternal saliva was collected fresh at the time of the study visit, aliquoted, and frozen at -80°C. Infant feces were collected from nursing mothers up to 24 h before scheduled study visits in Sterile containers provided by the study, frozen at -20°C, then transferred to -80°C until further analysis. Infant oral mucosa swabs were taken by gently swirling a cotton swab (Copan Diagnostics) over the entire surface of the left and right inner cheek, for 5 s each, and over the pharyngeal/tonsil area (5 s) . The swab was then transferred to a tube containing Universal Transfer Medium (UTM®) and frozen at -80°C until analysis. Buccal swabs were taken at least 2 h after the previous feeding, in order to avoid direct contamination by breast milk.
Sample preparation and antibody detection
For antibody quantification, serum, saliva and breast milk were thawed at room temperature (RT) and centrifuged for 5 min at 5000× g. All samples were then diluted in phosphate buffered saline (PBS) with 1% skimmed human milk powder (Sigma Aldrich). Serum was diluted 1:1000, breast milk (devoid of fat layer and cells) and maternal saliva 1:5. Buccal swabs were also thawed at room temperature, then vortexed in their transport medium for 30 s, and the transport medium was diluted 1/2 in the same 1% PBS milk. Feces were thawed at room temperature, suspended in PBS with 0.1 mg/ml soybean trypsin inhibitor (Sigma Aldrich) at a concentration of 100 mg/ml, and homogenized by shaking at 30 Hz. for 20 s in a tissue lyser (Qiagen), after addition of 1.4 mM ceramic beads (MP Biomedicals). They were then centrifuged at 50× g for 10 min to remove debris, and the supernatant was diluted 1/5 in 1% PBS milk.
Enzyme immunoassays (ELISA) for the detection of anti-S, anti-RBD and anti-N IgG and IgA antibodies were performed on Nunc MaxiSorp 96-well flat-bottom plates, in duplicate technique. Briefly, each plate was coated with 2 μg/mL of full-length Spike protein (S), receptor binding domain (RBD) or nucleocapsid (N) proteins (SinoBiological) of SARS-CoV-2 in PBS, and incubated at 4°C overnight.
After incubation, the plates were dried and blocked with 300 μL of PBS with 3% skimmed human milk powder for 1 h at room temperature. Then the block solution was discarded and the plates wiped clean. Fifty μL of diluted samples were added to the wells and the plates were incubated for 2 h at room temperature. Plates were then washed 3 times with wash buffer (0.1% Tween 20 in 1X PBS) and incubated at room temperature for 1 h with a 1:5000 dilution of mouse anti-human IgG conjugate horseradish peroxidase (HRP) (Genescript, clone 12H3C4A6 cat. no. A01855) or a 1:10,000 dilution of goat anti–human IgA alkaline phosphatase (AP) conjugate (Thermo Fisher Scientific, RRID AB_2535561, cat. no. A18784), or a 1:2000 dilution of AP conjugated mouse anti-human IgA1 (Southern Biotech, clone B3506B4, catalog # 9130-04), or a 1:8000 dilution of conjugated mouse anti-human IgA2 HRP (Southern Biotech, clone A9604D2, catalog no. 9140-05). The plates were washed 3 times and 50 μL of undiluted p-nitrophenyl phosphate (PNPP, Thermo Fisher Scientific) or 3,3′,5,5′-tetramethylbenzidine (TMB) solution (Thermo Fisher Scientific) was added to each well; TMB reactions were stopped by adding 25 μL of 0.18 M sulfuric acid per well. Absorbance at 405 (PNPP) or 450 (TMB) nm was recorded with a ClarioStar microplate reader (BMGLabtech). On each plate, two duplicates of a 1:3 serial dilution in 1% PBS milk of the World Health Organization (WHO) International Standard and Reference Group for anti-SARS-CoV-2 antibody (NIBSC, UK) were used as a standard curve, from a concentration of 30 IU/mL. Before quantifying antibody titers in complex biological matrices such as breast milk and fecal supernatants, we tested the performance of anti-IgA and anti-IgG detection antibodies with WHO standard curves diluted in three negative samples (T0) of the matrices of interest, at the same dilution of the samples tested, and compared to a standard curve diluted in 1% PBS milk. The correlations between the optical density (OD) of the standard curve diluted in 1% PBS milk or in the different biological matrices are reported in Fig. 1 additional: r2 were consistently >0.95, with no significant increase in background signal. To calculate antibody titers, absorbance (OD) values of each experimental sample were interpolated with the mean standard curve after correction for absorbance of blank controls (1% milk in PBS).
Descriptive statistics were calculated using GraphPad Prism 9.0.1 (GraphPad Software) and Stata 13.0 (StataCorp LP). Continuous variables are presented as median with interquartile range (IQR), categorical variables as numbers and proportion. To determine differences in antibody concentration between different samples at specific time points, a t-test was applied. To determine longitudinal differences in antibody concentration, a repeated measures ANOVA was applied. In case of missing values, we analyzed the data by fitting a composite that uses a compound symmetry covariance matrix and is fitted using restricted maximum likelihood. In the absence of missing values, this method gives the same P values and multiple comparison tests in the form of repeated measures ANOVA. In the presence of missing values (missing completely at random), the results can be interpreted as a repeated measures ANOVA. Pearson r coefficient was used to assess the correlation between antibody concentrations. Reporting of results follows guidelines on strengthening reporting of observational studies in epidemiology35.
Summary of reports
Further information on the research design can be found in the summary of nature research reports linked to this article.