Characterization of SARS-CoV-2 RNA, Antibodies, and Neutralizing Capacity in Milk Produced by Women with COVID-19

Abstract

Whether mother-to-infant SARS-CoV-2 transmission can occur during breastfeeding and, if so, whether the benefits of breastfeeding outweigh this risk during maternal COVID-19 illness remain important questions. Using RT-qPCR, we did not detect SARS-CoV-2 RNA in any milk sample (n = 37) collected from 18 women following COVID-19 diagnosis. Although we detected evidence of viral RNA on 8 out of 70 breast skin swabs, only one was considered a conclusive positive result. In contrast, 76% of the milk samples collected from women with COVID-19 contained SARS-CoV-2-specific IgA, and 80% had SARS-CoV-2-specific IgG. In addition, 62% of the milk samples were able to neutralize SARS-CoV-2 infectivity in vitro, whereas milk samples collected prior to the COVID-19 pandemic were unable to do so. Taken together, our data do not support mother-to-infant transmission of SARS-CoV-2 via milk. Importantly, milk produced by infected mothers is a beneficial source of anti-SARS-CoV-2 IgA and IgG and neutralizes SARS-CoV-2 activity. These results support recommendations to continue breastfeeding during mild-to-moderate maternal COVID-19 illness.IMPORTANCE Results from prior studies assaying human milk for the presence of SARS-CoV-2, the causative virus of COVID-19, have suggested milk may act as a potential vehicle for mother-to-child transmission. Most previous studies are limited because they followed only a few participants, were cross-sectional, and/or failed to report how milk was collected and/or analyzed. As such, considerable uncertainty remains regarding whether human milk is capable of transmitting SARS-CoV-2 from mother to child. Here, we report that repeated milk samples collected from 18 women following COVID-19 diagnosis did not contain SARS-CoV-2 RNA; however, risk of transmission via breast skin should be further evaluated. Importantly, we found that milk produced by infected mothers is a source of anti-SARS-CoV-2 IgA and IgG and neutralizes SARS-CoV-2 activity. These results support recommendations to continue breastfeeding during mild-to-moderate maternal COVID-19 illness as milk likely provides specific immunologic benefits to infants.

Keywords: COVID-19; SARS-CoV-2; antibodies; breastfeeding; breastmilk; human milk; neutralizing capacity.

FIG 1

Overview of 18 participants’ COVID-19 signs and symptoms (A) and COVID-19 diagnostic testing and sampling (B). In panel A, counts for individual signs and symptoms are presented in parentheses. In panel B, the time “0” represents the most recent positive COVID-19 test at enrollment. Participants E, I, N, O, Q, and R had additional COVID-19 tests performed; all were positive except for N’s and O’s second test and Q’s third test (false-negative result), which were negative (open squares). Participants L, M, and N were asymptomatic during the study period, and participants F and P developed symptoms after first testing positive.

FIG 2

Milk antibody concentrations. Panel A shows IgA and IgG to coronavirus antigens in milk produced by COVID-19-infected (red) and healthy, prepandemic (blue) women. Antibody concentrations were measured using ELISA specific to the RBD and S2 domains of the spike and nucleocapsid (N) proteins of SARS-CoV-2 and S proteins from human coronaviruses 229E and OC43. The horizontal solid line in the IgA panel denotes the limit of antigen-specific binding that is defined as mean + 2× standard deviation of RBD-specific IgA signal in prepandemic controls. The dotted horizontal line denotes the ELISA limit of quantification. P values for the difference between infected and prepandemic controls are from the Wilcoxon signed-rank test. Milk from 44 mothers (34/10, infected/controls) was used for IgA testing, and milk from 37 mothers (27/10, infected/controls) was used for IgG testing. Panel B shows correlations between IgA concentrations to SARS-CoV-2 antigens RBD, S2, and N (x axis) and IgA concentrations to S proteins from 229E and OC43 (y axis) in milk produced by COVID-19-infected (red, n = 34) and healthy, prepandemic (blue, n = 10) women. A linear model was fitted to log-transformed IgA concentrations to give r-squared and P values as indicated. Model prediction (red or blue line as defined above) is shown with the 95% confidence interval (gray shading).

FIG 3

Correlation of milk IgA concentrations with microneutralization titers in milk produced by women with COVID-19 (n = 34). Panel A shows the correlation between concentrations of IgA specific to SARS-CoV-2 RBD and microneutralization (MN) titers. Kendall rank correlation τ_p and associated P value are as shown. A linear model was fitted to log-transformed IgA concentrations and MN titers and is shown (black line) with 95% confidence interval (gray shading) for visualization purposes only. Time courses of anti-RBD IgA concentrations and MN titers are shown in panels B and C, respectively. Each color and corresponding point or set of connected points represent one participant. Time of collection is indicated on the x axis as days from the appearance of first symptoms or, for asymptomatic cases, from the time of diagnosis. Antibody concentration is indicated on the y axis of panel B, and MN titer is indicated on the y axis of panel C. Asymptomatic participants are indicated using a dotted line connecting points. The horizontal dashed line in panel B denotes the limit of antigen-specific binding as defined in Fig. 2A.