Anti-SARS-CoV-2 Immunoglobulins in Human Milk after Coronavirus Disease or Vaccination—Time Frame and Duration of Detection in Human Milk and Factors That Affect Their Titers: A Systematic Review

Human milk (HM) of mothers infected with or vaccinated against SARS-CoV-2 contains specific immunoglobulins, which may protect their offspring against infection or severe disease. The time frame and duration after infection or vaccination, during which these immunoglobulins are detected in HM, as well as the major factors that influence their levels, have not been fully elucidated. This systematic review aimed to collect the existing literature and describe the immune response, specifically regarding the immunoglobulins in HM after COVID-19 disease or vaccination in non-immune women. We conducted a systematic search of PubMed and Scopus databases to identify studies published up until 19 March 2023. In total, 975 articles were screened, and out of which 75 were identified as being relevant and were finally included in this review. Infection by SARS-CoV-2 virus primarily induces an IgA immune response in HM, while vaccination predominantly elevates IgG levels. These immunoglobulins give HM a neutralizing capacity against SARS-CoV-2, highlighting the importance of breastfeeding during the pandemic. The mode of immune acquisition (infection or vaccination) and immunoglobulin levels in maternal serum are factors that seem to influence immunoglobulin levels in HM. Further studies are required to determine the impact of other factors, such as infection severity, lactation period, parity, maternal age and BMI on immunoglobulin level in HM.


Introduction
The global COVID-19 pandemic has resulted in more than 6.85 million deaths worldwide with about 0.1% of incidents occurring in neonates and children under 5 years [1,2]. The widespread availability of vaccines has played a crucial role in controlling transmission rates and reducing morbidity and mortality. The Center for Disease Control and Prevention (CDC) recommends vaccination for individuals aged 6 months and older, including neonates and non-vaccinated infants [3].
Younger or unvaccinated infants are defenseless against SARS-CoV-2 virus. Breastfeeding could be a protective factor against severe infection for these infants. Human milk (HM) contains various bioactive nutrients, such as immunoglobulins that block the penetration of microorganisms into the endothelium [4]. Initial concerns regarding the safety of breastfeeding during maternal infection led to previous recommendations for infected women to avoid breastfeeding. However, since June 23, 2020, the World Health Organization (WHO) strongly recommends breastfeeding, as the benefits outweigh the

Materials and Methods
In order to perform this systematic review, we followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines' (PRISMA) recommendation (presented as Supplementary Material) [14]. The systematic review was not registered in Prospero. We searched the Pubmed and Scopus databases from 1 December 2019 to 19 March 2023. We searched the existing literature only written in the English language.
Additionally, we searched all of the references of the relevant studies and of previous corresponding systematic reviews in order to confirm the study saturation.

Study Eligibility Criteria
All selected studies examined the immunological response in the HM of pregnant or lactating women after COVID-19 disease or vaccination. After the duplicates were deleted, two investigators (M.D. and R.S.) independently checked the titles and abstracts of the retrieved papers, and consequently studied the full texts to decide which of them were eligible for the review. Any disagreements between the two researchers were analyzed and resolved by a third researcher (Z.I.).
Studies included in the present review met the following eligibility criteria: (1) women vaccinated against or infected by SARS-CoV-2 virus during pregnancy or lactating period were the study population, (2) there was no history of previously confirmed COVID-19 disease or vaccination, (3) current COVID-19 disease was confirmed via a PCR positive test, serology test or other laboratory method and (4) data on HM-specific immunoglobulins against SARS-CoV-2 were described in the studies that were included in the review as well. Irrelevant or non-original studies, case reports or studies with indecisive data were excluded from this review, as well as studies in any language other than English.

Data Extraction
The 2 researchers (M.D. and R.S.) separately studied the eligible studies and extracted useful data in an electronical database (Microsoft Excel). Complete information included the name of the first author, date of publication, country of origin, duration and population of the study, diagnosis method or vaccine type, time of HM collection, studied immunoglobulins and main outcomes or results of the study. Any disagreement between the two researchers was analyzed and resolved by a third researcher (Z.I.). Finally, the selected studies were examined again by another investigator (N.I) to check for eligibility and for duplication, and the extracted data were checked for their accuracy.

Results
A total of 975 articles were initially retrieved. After excluding duplicates and articles with titles and abstracts not related to our review object, 122 studies were selected for their full text to be studied. In these studies, the full text was comprehensively studied and 75 studies finally met the eligibility criteria. The searching and selection processes are depicted in a PRISMA graph ( Figure 1).

Discussion
Breastfeeding is the best dietary choice for infants [4]. HM contains specific immunoglobulins that are produced in response to exposure of the mother to pathogens. These immunoglobulins maintain their structural integrity in the infant's stomach, bind to intestinal mucus and prevent pathogens from entering the bloodstream [89][90][91]. The presence of specific immunoglobulins against SARS-CoV-2 in the HM of infected or vaccinated mothers could potentially shield neonates and infants from future infections or even from severe disease.

Post Infection Immune Response
This systematic review affirms that specific immunoglobulins against SARS-CoV-2 virus were detected in the HM of women who were infected during pregnancy or lactation. These immunoglobulins were mainly IgA, and specifically secretory IgA antibodies, and less IgM and IgG were found [17,20,25,28,32,35,36,43,82], which is compatible with the known proportion of antibody isotypes in human milk [92]. They also primarily (80%) targeted the RBD domain of the S1 subunit [17].

Anti-SARS-CoV-2 IgA Immunoglobulins
IgA titers in HM increase one week after infection, and these titers are even higher 2 weeks post COVID-19 disease [40]. Many studies indicate that they remain high in HM and detectable even 2-3 months after infection [27,31,34,35,[38][39][40]42,43,83,85,86]. Pace RM et al. reported that IgA remained positive in 77% of 64 lactating women 2 months after infection [40]. Junker HG et al. found that HM conversion was observed after a median of 15 days and IgA levels peaked after 35 days. After 70 days, however, IgA was detectable only in 33% of HM samples [42]. Conti MG et al. reported that IgA was detectable in all HM samples of 28 lactating women even 2 months after delivery [23]. These women had been infected during pregnancy, which indicates that IgA may persist for a longer period of time. Indeed, other studies confirm their persistence in HM even 5-10 months post COVID-19 disease [34,39,43]. Fox A. et al. reported that all of the 28 tested women in their study had detectable IgA in HM 4-10 months after infection and 43% of them had even higher titers than what they had at 1 month after the infection [43]. The presence of specific IgA against SARS-CoV-2 in HM is optimal for infants. According to a recent cross-sectional study in Brazil, the titers of IgA in the HM of women infected during pregnancy were negatively correlated with the presence of clinical symptoms in their neonates [83].
Various factors influence anti-SARS-CoV-2 IgA levels in HM. The concentration of specific IgA immunoglobulins in HM is positively correlated with the levels of total IgA, IgM and IgG titers in HM [21]. Additionally, levels of specific IgG in maternal serum are also significantly correlated with IgA levels in HM [16]. However, this is not the case for IgA titers in serum. There is no reported association between IgA titers in HM and maternal serum [25,35,47]. This is logical as IgA in HM after natural infection is not of serum origin, but of muscular origin [93].
As for the time from infection, the data are conflicting. Some report a negative correlation with antibody titers [21,23], while others report a positive correlation [37]. Infection induces a humoral response and antibody titers begin to rise. After an unknown period of time, they reach a peak, and then they begin to wane over time [94]. Therefore, contradicting results in the studies may be due to different time-points of sample collection.
Additionally, not all types of antibodies have the same response over time. Bobik T.V. et al. tested sIgA against specific epitopes of SARS-CoV-2 virus (N protein, linear NTD, RBD-SD1 and RBD) and reported that the levels of sIgA against N-protein and against linear NTD and RBD-SD1 were higher in women that were infected during the third trimester compared to women infected during the first and second trimesters. Regarding sIgA against RBD, they found that their levels were similar, independent of the trimester of pregnancy when infection occurred [19]. No correlation between time of infection during pregnancy and anti-RBD antibodies was reported by Szczygioł, P. et al. either [28]. This indicates that this kind of antibody is more stable over time. Interestingly, Wachman EM et al. reported that anti-RBD IgA titers were not stable but significantly higher in women infected during the first or second trimester of pregnancy [84].
The severity of COVID-19 disease is another factor, but its impact on antibody titers has not yet been clarified. Pace, R.M et al. reported higher concentrations of antibodies in the HM of women with symptomatic COVID-19 disease than in the HM of asymptomatic women; yet, the difference was not significantly important [40]. In other relevant studies, no association between the two parameters has been found [28,83]. Probably, the severity of the disease has a positive impact on antibody levels. Many other studies on the general population have reported that antibody titers are higher in people with severe/moderate COVID-19 disease [94]. More studies are required to reach safe conclusions. From the existing data in the literature, no correlation has been found between IgA levels in HM and maternal age or infant gender [35,37].

Anti-SARS-CoV-2 IgM Immunoglobulins
IgM immunoglobulins are the second most abundant antibodies in HM, and yet their titers are significantly lower than IgA titers. In a prospective study, Decenti EC detected anti-SARS-CoV-2 antibodies in only 7.5% of milk samples [26]. Specific IgM against 2 SARS-CoV-2 was mainly detected in samples collected 10-40 days after infection, and after that, their levels declined [20,21]. In a relevant study by Luo QQ et al., a positive correlation between IgM levels in HM and maternal serum was found [24]. The presence of IgM antibodies against infectious diseases in HM can provide passive immunity to infants, while simultaneously hindering the entry and transportation of viruses, such as HIV, to the infant [95,96]. Given these findings, it is plausible that breastfeeding by SARS-CoV-2infected mothers offers postpartum protection to the infant through antibodies, reducing the risk of viral transmission.

Anti-SARS-CoV-2 IgG Immunoglobulins
HM contains low titers of IgG immunoglobulins, and in some studies they were not even detectable [20,24]. Decenti, E. C. et al. studied HM samples from 141 women, and they detected IgG only in 3% of the study population [26]. Bauerl et al. reported low IgG titers in HM, with an increase in these titers from Day 40 to Day 205 after COVID-19 disease [21]. Pullen KM et al. detected IgG in a low concentration, but they did not observe any significant change in the titers over time (the mean time of sample collection was 66 days post infection). They also claimed that IgG was functionally attenuated compared to IgA and IgM [35]. In another 2 studies, low titers of IgG 0-3 months after delivery were reported [23,81]. Contrarily to the previous studies, Fox. A. et al. reported that they detected anti-S IgG antibodies in 75% of participants, with 13% of them being present in high titers [43]. The factors that influence their titers in HM are not all clear, but there is a positive correlation between IgG titers in HM and serum [29,35,47].

Post Vaccination Immune Response
The vaccination of pregnant or lactating women against SARS-CoV-2 also induces the secretion of specific anti-spike antibodies in HM. However, there are some differences in this immune response compared to the one after natural infection. First of all, vaccination mainly induces an IgG response and less of an IgA response [22,25,32,38,42,. After mRNA vaccination, a 10-fold and 100-fold increase in IgA and IgG titers was observed, respectively [47]. Vaccination induces no significant increase in secretory antibodies' titers. IgA seems to be almost exclusively of systemic and not mucosal origin [55]. Pietrasanta et al. measured two subtypes of specific anti-S IgA antibodies, IgA1, which has systemic origin, and IgA2, which is mainly detected in mucosal secretions, and they observed that the antibodies were mainly IgA1 [53]. These differences in immune response are probably due to the intramuscular route of vaccine administration [50,67].

Anti-SARS-CoV-2 IgG Immunoglobulins
IgG titers were detected in 87-100% of women post vaccination [50,62,65,66,72,77]. Only in one study was a moderate IgG immune response (43% of women) observed [73]. IgG titers in HM increase after each dose [38,47,50,52,55,63,67,71,74,75]. The peak of anti-S1-IgG titers occurs about 1-2 weeks after the 2nd dose and after that they wane [48,[51][52][53]56,62,65,66,71]. A recent longitudinal study reported that IgG antibodies' half lives in HM are about 2 months [45]. In other relevant studies, no significant difference in IgG titers was observed between 30 and 60 days post vaccination [57,80]. Even IgG levels wane over time, but remain in detectable levels 2 [57], 3 [38,53,72] and 6 months post vaccination [49,73]. Contrarily with total IgG levels, secretory IgG antibodies continuously increase even 6 months after the first dose of mRNA vaccines [54]. The main IgG subclasses in HM after 2 doses of mRNA-based vaccines are IgG1 and IgG3, which are the main subclasses of IgG that emerge after viral infections [45,97]. Interestingly, Agostinis C. et al., in a recent study, demonstrated that the presence of anti-S IgG in the HM of vaccinated lactating women is capable of activating in vitro the complement [80]. All of these data indicate that vaccinated mothers are capable of providing protective antibodies to their offspring for a long time after their primary vaccination.
As for the factors influencing antibody levels in HM, numerous studies have confirmed a significant positive correlation between IgG titers in HM and maternal serum [25,29,45,47,48,51,56,[62][63][64]69,75,77,79,80]. A positive correlation between the lactation period and the total antibody titers was also reported in a study by Trofin F. et al. (lactating period between 3 and 36 months) [57]. However, this was not confirmed in other studies during the lactation period between 1.5 and 23 months [50,73]. A negative correlation was reported with parity [57], while no correlation was confirmed between IgG levels and maternal age or BMI [57,69,74].

Anti-SARS-CoV-2 IgA Immunoglobulins
IgA immunoglobulins in detectable levels are present in 75-95% of vaccinated women at 2 weeks after the 2nd dose [42,63,65,66,74]. In a prospective study, though, detectable levels were present in just 36% of women post vaccination [73]. Yet, it is not clear if IgA rises mainly after the 1st dose, without any additional increase after the 2nd dose [38,47,63,67,73,74], or whether their titers present a biphasic model after vaccination [42,52,55,65,66,70,71]. Many studies indicate waning IgA titers about one month after the 1 st vaccine dose [44,49,51,56]. Contrarily, Juncker HG et al. reported that IgA titers peaked 2 weeks after the 1 st dose, and then they waned until the 2 nd dose, when they finally reached a second peak 5 days after vaccination [70]. Ricciardi et al. studied secretory IgA and they observed peak titers at 3 weeks after the 2nd dose, while their concentration significantly decreased at 6 months post vaccination [54]. Perez SE et al. detected IgA antibodies in about 50% and 25% of samples at 1 and 3 months post vaccination, respectively [62]. Finally, Narayanaswamy et al. observed no difference in anti-RBD IgA median titers before and after vaccination [50].
A positive correlation was found between IgA titers in HM and IgG titers in maternal serum [51,62,69,74]. A positive correlation may also exist between the IgA titers in HM and the IgA titers in maternal serum [25,47,64]; however, this finding is not supported by all studies [74]. The lactation period is also another factor which has an impact on IgA titers, and is not clearly defined by the included studies. A positive impact (lactation period between 3 and 36 months) [57] was reported in one study, while in others, either a negative (lactation period <18 months) [63] or no impact (lactation period 1.5-23 months for infants) [50] was reported. A negative correlation was found between antibody concentration and parity [57], and there was no correlation with maternal age [57,74].

Differences in Immune Response after Infection or Vaccination
Higher IgG titers were observed in the HM of vaccinated women compared to those who were previously infected with SARS-CoV-2 [22,25,32]. In the case of IgA, the data are ambiguous, with some studies reporting higher antibody titers after vaccination [27], some reporting lower [52] and some reporting no significant difference after vaccination or infection [42]. As has already been mentioned, infection mainly induces IgA and, to a lesser extent, the IgM immune response, while vaccination primarily stimulates the production of IgG antibodies. IgA and IgG are found mainly in secretory form and they are released by mammary tissue to HM. After a natural infection with SARS-CoV-2, B-cells stimulated in the lymphoid tissues of the respiratory tract migrate to the mammary gland and release secretory antibodies into HM [93]. Contrarily, post vaccination antiSARS-CoV-2 antibodies in HM are derived from maternal serum [11]. Secretory IgA antibodies are attached to the gastrointestinal tract of breastfeeding infants, where they bind with local microorganisms and block their penetration [4]. On the other hand, IgG antibodies produced post vaccination have been found to be capable of activating the complement. Therefore, infants can benefit from both types of antibodies in different ways [80].
As for the type of immunoglobulins, vaccination induces anti-spike protein antibodies. Infection, on the other hand, mainly induces the anti-N protein. A relative study by Bobik T.V. et al. reported significantly lower levels of the anti-RBD antibody compared to anti-N antibodies in the HM of lactating women infected during the third trimester [19]. L.Bode et al., in a recent study, reported that IgA antibodies in the HM of infected women are mainly directed against the N protein (about 43%) and less against the S protein (about 24%), and there was heterogeneity in the type and quantity of antibodies.

Differences in Immune Response According to Vaccine Type
No difference between detected antibody titers was described in women vaccinated with Moderna or Pfizer/BioNtech vaccines [32,52,57,63,68,88]. Only in one study of Gray KJ et al. were higher IgA titers detected in the Moderna group after the 2nd dose [67]. However, compared to the adenovirus-vectored vaccines, mRNA vaccines induce higher titers of antibodies [52,58,61]. Yang X. et al. reported that IgG and IgA levels were detectable in 86-100% and 52-71%, respectively, in mRNA-vaccinated women, and in 33-38% and 17-23%, respectively, in the adenovirus-vector-vaccinated women [58]. The Moderna vaccine also induces significantly higher titers of secretory antibodies compared with the rest of the vaccines [58]. Yet, six months post vaccination, no significant difference in antibody titers was observed among three types of vaccines [88].
In a study of 38 infected lactating women, no correlation between specific anti-SARS-CoV-2 antibody levels and the neutralization capacity of the HM was observed [31]. However, this is not in line with what other studies support. The neutralizing capacity against SARS-CoV-2 seems to be greater in the HM of infected women compared to the pre-pandemic controls, and it also seems to be positively correlated with antibody titers [18,27,30,32,36,43,44,82]. In a relevant study, 62% of the samples had neutralizing antibodies in vitro. Contrarily, the samples collected from the pre-pandemic controls had no neutralizing capacity [36]. Pace RM et al. reported that the HM neutralizing capacity post infection is significantly correlated with anti-RBD antibody levels [30].
Antibody titers also seem to influence the neutralizing capacity of the HM of vaccinated women [44,62]. As for variants, vaccination seems to be more beneficial against the Wuhan-Hu-1 strain and less against the Beta, Gamma and Delta variants. The binding capacity of antibodies is reduced by 30% to these strains [49]. In another study, less of a neutralizing capacity was reported against the Beta variant compared to D614G, Alpha and Gamma [50].
Our systematic review has some limitations. SARS-CoV-2 is a new virus. In the early stages of the pandemic, many studies were conducted with small sample sizes and lacked inclusion criteria. Furthermore, most of the tests used were homemade ELISA assays, which lacked standardization in terms of measurement units. Due to the great heterogeneity of the studies included, not only regarding the type of laboratory test, but also the differences in the timing of sampling between studies and study subjects (pregnant or lactating women), a meta-analysis was not deemed appropriate and instead we focused on presenting a systematic review of the available literature.

Conclusions
Infection with SARS-CoV-2 and vaccination against the virus elicit a maternal immune response in breastfeeding mothers with a short period of 1-2 weeks. Consequently, these mothers can transmit specific immunoglobulins with neutralizing capacity to their infants via HM. Therefore, it is recommended that breastfeeding is encouraged in mothers infected or vaccinated, as their HM can provide infants with specific antibodies even months after infection or vaccination.
Infection primarily induces an IgA-mediated immune response, while vaccination mainly elevates IgG immunoglobulins. Our data suggest that both IgA and IgG immunoglobulins contribute to the neutralizing capacity of HM, indicating clinical benefits for infants who receive HM from vaccinated or infected women. However, further studies are required to determine whether factors such as infection severity, lactation period, parity, maternal age and BMI have an impact on the levels of immunoglobulins in HM.