Background
Breast milk has primary importance for feeding of the newborn, because of its nutritional properties and the contribution to the development of host defences [
1]. However, human milk is not sterile and can sometime transmit bacteria. Group B
Streptococcus (GBS) is a leading cause of neonatal infections in developed countries [
2]. Two distinct syndromes are recognized: early-onset disease (EOD, from birth to day 6) and late-onset disease (LOD, from day 7 to 89) [
3]. Several case reports have suggested breast milk as a possible source of GBS LOD [
4]. Nevertheless, the mechanisms of GBS transmission and LOD pathogenesis are not yet clear. GBS gastrointestinal and genitourinary tract colonisation is common, and it ranges in pregnant women from 4 to 36% [
5,
6]. However, only 0.8 to 3.5% of mothers carry GBS in their breast milk [
7,
8].
We report 3 cases of LOD in full-term newborns who were breastfed with GBS contaminated milk. The newborns’ mothers had no signs of mastitis and their rectovaginal swabs were permanently GBS negative (both at prenatal screening and at the time of diagnosis of LOD). These findings suggest that the transmission may occur through a circular mechanism: the newborn (colonised in the throat) could be the initial source of GBS, while the mammary gland could act as a GBS replication site.
Discussion and conclusions
This report deals with three cases of LOD in full-term neonates possibly attributed to the ingestion of breast milk containing GBS. Cases presented with sepsis and/or meningitis at day 9, 17 and 8, respectively. None of the mothers had signs of mastitis and all were GBS-negative at rectovaginal site (both at screening and at the time of diagnosis of LOD). In case 1 and case 2, neonatal and maternal GBS isolates were serotype III. Milk bacterial count was available only in case 3.
Berardi et al. evaluated GBS colonisation in 160 mother-baby pairs. GBS was identified in 53 neonatal throat cultures and 77 neonatal rectal cultures. GBS in breast milk was associated with heavy neonatal colonisation [
10]. However, the mechanism of transmission of LOD through breastfeeding is poorly understood. The retrograde theory hypothesizes that GBS, present in the infant’s throat, colonises the mammary ducts during breast-feeding. GBS load increases in the milk, and in turn the infant is infected during breast-feeding (circular mechanism) [
4]. Alternatively, some authors suggest that GBS might reach the mammary gland through the translocation of bacteria from maternal gut via lymphatics [
11].
A recent review of the literature analysed cases of LOD in which the breast milk was tested positive for GBS [
4]. The review pointed out that the role of breast milk in LOD remains controversial, although the milk would be a more convincing source when LOD occurs in a neonate born to a GBS negative mother, delivered after planned CS and when nosocomial sources are not identifiable. The review also reported that less than half mothers with GBS in breast milk had mastitis [
4] and that most mothers (59%) were GBS negative at antenatal screening. However, Berardi et al. showed that only a few mothers (~ 25%) of neonates with LOD were confirmed GBS negative at rectovaginal site when they were retested at the time of diagnosis of LOD [
12]. Therefore, the proportion of mothers who actually carry GBS at rectovaginal site is certainly underestimated if mothers GBS negative at screening are not retested at the time of diagnosis of LOD.
In the current study, 3 cases of LOD occurred in 3 different hospitals, therefore GBS strains are certainly unrelated. All mothers were confirmed GBS-negative at rectovaginal site. Newborns could have been colonised with GBS at mucosal surface (throat) after birth (from caregivers or environmental sources) and subsequently they could have transmitted GBS to mother’s mammary gland. Indeed, in case 2, the GBS yielded from breast milk and neonate showed an identical genetic profile. In these cases a circular mechanism seems particularly suggestive, whereas a bacterial translocation from maternal gut is unlikely because of persistent negative GBS rectovaginal culture.
Transition from silent breast duct colonisation to active GBS multiplication depends on many factors, such as milk stasis and bacterial load. Some investigators found that mothers with mastitis had higher GBS bacterial load (1.000.000 CFU/ml), than mothers without mastitis (≤100.000 CFU/ml) [
12]. The lower bacterial count could suggest contamination during sampling rather than bacterial active multiplication. In the current study, none of the mothers had evidence of mastitis and milk bacterial count, available only in 1 out of the 3 cultures, was 80.000 CFU/ml. Some studies reported a total bacterial count < 10
6 CFU/ml as the physiological threshold of bacterial load in human milk [
13,
14]. The origin of milk bacteria is still not well understood, but several studies confirmed the existence of a dynamic network between breast-milk and newborn’s oral microbiota [
15]. The causative role of breast milk in LOD results from the interaction of several maternal and neonatal factors. Prematurity, immature immune system, bacterial load and lesions of the intestinal mucosal barrier are recognised risk factors for progression to infection after ingestion of breast milk contaminated with GBS [
4,
16]. This study is subjected to some limitations. Although mothers were apparently negative, we can not firmly rule out that they had a light colonisation, undetected by rectovaginal cultures. However, this eventuality seems unlikely. Furthermore, we have no data regarding the genetic profiles of most GBS isolated from mothers.. Finally, surface cultures were not collected from family members and neonates. This additional information could contribute to understand modes and mechanisms of GBS transmission.
Identifying the underlying mechanisms of postnatal transmission of GBS will be crucial to prevent cases of LOD. To date studies have not recognized the predominant mode (maternal, nosocomial or community) of postnatal GBS transmission. GBS in breast milk can be a risk factor for GBS LOD. The persistent culture-negative maternal GBS status suggests the transmission through a circular mechanism, with the newborn (colonised at the throat) as the initial source of GBS, and the mammary gland as a GBS replication site. High or low bacterial load in breast milk might help distinguish cases in which breast milk is actually infected from cases where breast milk is only “contaminated” during sampling.
Testing maternal GBS rectovaginal status and collecting breast milk culture at the time of LOD diagnosis could help to shed light on the mechanisms of GBS LOD.