BNT162b2 mRNA vaccine elicited antibody response in blood and milk of breastfeeding women

Following the rapid rollout of the BNT162b2 mRNA COVID-19 vaccine in Israel beginning in December 2020, lactating women became eligible for vaccination despite their exclusion from initial clinical trials. While prior studies in women with COVID-19 showed the presence of virus-specific antibodies in breastmilk, limited evidence existed on vaccine-induced antibody responses in lactating women, their temporal dynamics in breastmilk versus blood, and the neutralization capacity of these antibodies. This study aimed to characterize the kinetics and magnitude of SARS-CoV-2 spike- and RBD-specific IgG and IgA responses in paired serum and breastmilk samples from lactating women after two vaccine doses, to assess the neutralization capacity of breastmilk antibodies, and to compare vaccine-elicited breastmilk antibody levels to those induced by routine TDaP vaccination. The purpose was to inform public health recommendations regarding vaccination during lactation and the potential for passive neonatal protection via breastfeeding.

Prior evidence indicated robust and specific secretory IgA against SARS-CoV-2 in human milk from women with natural infection, and detectable neutralizing capacity in a substantial proportion of such samples. Two early reports described antibody responses in breastmilk following mRNA vaccination (Pfizer-BioNTech and Moderna), but key gaps remained regarding the neutralization capacity of vaccine-elicited breastmilk antibodies, the synchronization and dynamics of responses between blood and milk, and how these responses compare to antibodies elicited by other maternal vaccines such as TDaP. These gaps motivated a focused longitudinal analysis in vaccinated lactating women.

Design: Prospective cohort study of 10 lactating healthcare providers receiving two doses of BNT162b2 (21-day interval). Controls: Pre-pandemic serum and breastmilk from 10 healthy women. Ethics: Informed consent obtained; IRB approvals 0002269-4, 0002757-1 (Tel Aviv University) and 1088-20-TLV (Tel Aviv Sourasky Medical Center). Sampling: Paired breastmilk and blood collected at four time points relative to vaccination: 1D7 (7 days post–first dose), 1D14 (14 days post–first dose), 2D7 (7 days post–second dose), and 2D14 (14 days post–second dose). Plasma was isolated by density-gradient centrifugation; breastmilk aqueous phase obtained by centrifugation and removal of lipid layer; all stored at −20 °C. ELISA for endpoint titers: Half-area 96-well plates coated overnight at 4 °C with 2 µg/ml recombinant SARS-CoV-2 spike or RBD proteins. Blocking with 3% skim milk in PBS. Serial threefold dilutions: serum starting 1:100; breastmilk 1:11. Detection with HRP-conjugated anti-human IgG or IgA, development with TMB, read at 450 nm. Endpoint titer defined as highest dilution with OD450 exceeding background by 3 SD. TT-specific (tetanus toxoid) IgG/IgA in breastmilk measured for comparison. IgG:IgA molar ratio: Spike-coated ELISA with included standards (protein G for IgG, protein M for IgA). Pure IgG or IgA standards (starting 60 µg/ml, twofold dilutions) used to generate 4-parameter logistic curves. Sample molarities interpolated and adjusted by dilution factors (serum 1:100; breastmilk 1:6) to calculate IgG:IgA ratios in serum and breastmilk across time points. Neutralization assays: SARS-CoV-2 spike-bearing pseudovirus produced by co-transfection (PCMV-deltaR8.2, pLenti-GFP, PCDNA3.1-SAC19) in Expi293F cells; supernatant concentrated and filtered. HEK-293 cells stably expressing hACE2 were infected with pseudovirus pre-incubated with serial dilutions of filtered breastmilk (starting 1:1) or with purified IgG/IgA (50 µg/ml). After infection, GFP-positive cells were quantified by IncuCyte imaging and percent inhibition calculated; ID50 determined by nonlinear regression (inhibitor vs response). Breastmilk was pre-tested for lack of cell toxicity. Purification of breastmilk antibodies: From 2D7 samples, IgG purified with protein G affinity chromatography; flow-through subjected to peptide M affinity chromatography to purify IgA. Eluted antibodies neutralized, dialyzed, quantified, and purity checked by SDS-PAGE. Neutralization tested at 50 µg/ml for each isotype; inhibition normalized per 5 µg antibody for comparison to controls. Statistics: Mann–Whitney U-tests for independent groups with Bonferroni correction (alpha 0.0125 where applicable); Wilcoxon matched-pairs signed-rank for paired comparisons; two-tailed P values. Analyses performed with GraphPad Prism 9.0.2.

- Antibody kinetics and synchronization: Spike- and RBD-specific IgG and IgA responses in serum and breastmilk were rapid and highly synchronized. No significant increase vs controls at 1D7; significant increases appeared by 1D14. Titers peaked at 2D7 and showed stabilization by 2D14 (no significant increase vs 2D7). - Isotype dynamics: RBD-specific IgA in both serum and breastmilk reached significant levels vs controls by 2D7, slightly delayed relative to IgG. In breastmilk, the fold-change in IgA over 1D7 declined significantly by 2D14, indicating a decreased production rate, whereas IgG remained elevated and peaked at 2D14. - IgG:IgA molar ratios: Serum responses were IgG-dominant with consistently higher IgG:IgA ratios than breastmilk at all time points and little temporal change. Breastmilk responses were IgA-dominant overall, but the IgG:IgA ratio increased significantly at 2D7 and 2D14, suggesting a temporal rise in IgG levels. - Comparison to TDaP: Breastmilk spike- and RBD-specific IgG and IgA endpoint titers at 2D14 did not differ significantly from tetanus toxoid (TT)-specific titers induced by TDaP vaccination during pregnancy in the same participants. - Neutralization capacity (whole breastmilk): All 10 breastmilk samples at 2D7 exhibited pseudovirus inhibition with mean ID50 ≈ 17 (reciprocal dilution; range 3.2–52.2), significantly higher than pre-pandemic controls. - Neutralization capacity (purified antibodies): At 50 µg/ml, 4/10 purified IgG samples from 2D7 exceeded 50% inhibition; none of the purified IgA samples exceeded 50% inhibition at the same mass concentration. Normalized inhibition activity per 5 µg antibody was on average 21-fold higher for IgG and 4-fold higher for IgA from vaccinated women compared to controls. - Overall, neutralization capacity of breastmilk was evident in all participants despite inter-individual variability.

The synchronized rise of spike- and RBD-specific IgG and IgA in serum and breastmilk, peaking seven days after each vaccine dose, indicates coordinated systemic and mucosal humoral responses in lactating women following BNT162b2 vaccination. Breastmilk IgG levels peaked by 2D14, whereas IgA showed a modest decline by that time, consistent with differential isotype kinetics and class-specific regulation. The similarity of vaccine-elicited breastmilk antibody titers to those induced by maternal TDaP suggests a potential for comparable passive protection in infants. Neutralization activity was present in all breastmilk samples, supporting functional relevance of the elicited antibodies. Although purified IgG showed higher per-mass inhibition than IgA in vitro, the higher physiological concentration and dimeric form of secretory IgA in milk, along with continuous daily intake by infants, suggests that IgA likely contributes substantially to mucosal neutralization in the neonatal gut and upper airways. These findings support the plausibility that maternal mRNA vaccination during lactation could confer passive protection to breastfed infants.

This study demonstrates that BNT162b2 vaccination in lactating women induces a rapid, robust, and synchronized antibody response in serum and breastmilk, comprising both IgG and IgA with demonstrable neutralization capacity. Breastmilk antibody levels after two doses were comparable to those elicited by TDaP vaccination, underscoring potential for passive neonatal protection. Future research should extend follow-up to define durability of milk antibodies, investigate functional breadth against variants, quantify dose–response for purified isotypes (e.g., full IC50 curves), and provide real-world evidence of reduced COVID-19 incidence in breastfed infants of vaccinated mothers.

- Neutralization assays for purified breastmilk IgG and IgA were performed at a single concentration (50 µg/ml) due to limited sample volumes, precluding IC50 determination and limiting inferences about comparative potency, particularly for IgA which did not exceed 50% inhibition at that mass concentration. - Small cohort size (n=10) and short follow-up (up to 14 days after the second dose) limit generalizability and assessment of durability. - Inter-individual variability in neutralization activity was observed, and the study did not evaluate clinical protection in infants.