Antibiotic use during pregnancy: how bad is it?

Pregnancy is characterized by profound hormonal, immunological, and metabolic changes aimed at supporting the growth of the fetoplacental unit [6]. Interestingly, a healthy pregnancy also induces dramatic changes in the maternal gut microbiota over the course of gestation, with a great expansion of diversity between individuals, an overall increase in Proteobacteria and Actinobacteria, and reduced diversity within each microbiome. Similarly, the vaginal microbiome in pregnancy is different from that of non-pregnant women, with lesser diversity and richness, and with a dominance of Lactobacillus species, Clostridiales, Bacteroidales, and Actinomycetales [7]. When transferred to germ-free mice, third trimester microbiota induced greater adiposity and low-grade inflammation compared to first trimester microbiota [4], and it was recently suggested that alteration of the gut microbiota during pregnancy may also induce complications of pregnancy such as excessive maternal weight gain [8].

Maternal pregnancy complications significantly influence the bacterial composition and diversity of the stool microbiota of premature infants, with these changes persisting during the first year after birth [9]. Antibiotic usage during pregnancy undoubtedly affects the bacterial environment of the mother and of the fetus.

We are rapidly progressing towards understanding the role of bacteria in novel target organs traditionally considered “sterile”, including the central nervous system [10], blood [11], the lower airways [12], the sub-epidermis [13], and the fetoplacental unit [14]. ‘The fetus lies in a sterile environment’ declared the French pediatrician Henry Tissier of the Pasteur Institute in 1900 [15]; since then, the placenta was traditionally believed to form a sterile barrier between the colonized maternal urogenital tract and the fetus. However, recent studies have challenged this assumption. The first demonstration of bacteria in the placenta in the modern scientific literature was published in 1982 by Kovalovszki et al. [16]. Subsequent publications have described the placental microbiome, using culture-dependent and -independent methods, and suggested its relative similarity to the oral microbiome, with bacteria such as Prevotella tannerae and Neisseria being present in the placenta [14] (Fig. 1). Specific components of the placental microbiome have been linked to particular pregnancy complications (Table 1, Fig. 2). Several mechanisms of amniotic fluid colonization have been proposed, including the translocation of vaginal bacteria [17] via the bloodstream or from the oral cavity [18]. Other possible routes include hematogenous spread or bacterial migration from the lower gastrointestinal tract to the lower genitourinary tract (Fig. 1).

The fetal membranes consist of the amniotic membrane, which contains the amniotic fluid, and the chorion, which is in proximity with the decidua parietalis. It has been shown that certain bacteria may be present in fetal membranes without leading to an inflammatory response [16]; Galask et al. [19] demonstrated that bacteria can traverse intact fetal membranes. Other bacteria resident in the fetal membranes are thought to play a role in amniotic inflammation, and thus in the initiation of labor [20].

Some of the studies described above demonstrated very low levels of bacterial DNA in the fetoplacental components. As detection methods become more sensitive, even more rigorous precautions must be taken against contamination when screening for bacterial colonization [21]. Nevertheless, despite the possibility of sample contamination, the physiological and beneficial presence of bacteria in the placenta is gaining widespread acceptance.

Perhaps the most clinically relevant aspect of the pregnancy microbiome is antibiotic treatment during pregnancy. Antibiotics account for 80 % of all prescribed medication in pregnancy [22], yet surprisingly, few published human studies have carefully evaluated the direct effects of antibiotics during pregnancy on either the maternal or fetal microbiome, or evaluated long-term sequelae of such antibiotic use. Thus, there may be a reason for caution in prescribing antibiotics during pregnancy.

In pregnant NOD mice, antibiotic treatment caused alteration of gut microbiota and immunological changes in the intestine of the offspring [23]. In pregnant women, it was demonstrated that antibiotic administration during pregnancy leads to alterations in the vaginal microbiome prior to birth, with long-term effects on the early microbial colonization of the newborn [24] and an association with childhood obesity [25].

There are several components to this issue. Antibiotic treatment of infectious diseases is one of the greatest advances of modern medicine. Accordingly, antibiotics are widely prescribed during pregnancy as the most important modality for treating and preventing infections. It is estimated that one in five pregnant women in Europe is prescribed at least one antibiotic during pregnancy; in the United States, the rate is double [26]. Nevertheless, prescription of antibiotics should be carefully considered on an individual basis, weighing its benefits versus drawbacks for both the fetus and the mother. It has been shown that administration of certain antibiotics is linked to a significantly higher rate of neonatal necrotizing enterocolitis, although antibiotic treatment is also associated with a reduced rate of lung complications and major cerebral abnormalities, relative to non-antibiotic treated controls [27]. A more recent study published in 2008 demonstrated that the prescription of antibiotics for women in spontaneous preterm labor with intact membranes was associated with an increased risk of cerebral palsy and functional impairment among their children at 7 years of age [28].

As discussed above, the healthy microbiome is important for maintaining a normal pregnancy and, therefore, it has been suggested that we may be using too many antibiotics during pregnancy [29]. A large systematic review concluded that antibiotics during the second and third trimester do not reduce adverse pregnancy outcomes and morbidity [30]. In addition, even a short course of antibiotics perturbs bacterial communities in human hosts [30]. In one study, it was shown that, within 30 days following cessation of antibiotic treatment, fecal microbiota reached an average similarity of 88 % to baseline, with the level rising to 89 % within 60 days [31]; however, the microbiota did not completely return to baseline over the timescale studied. Thus, antibiotics cause an immediate perturbation of the ecosystem, followed by incomplete recovery of the gut microbiome. The response to a given antibiotic is individualized, and may be influenced by prior exposure to the same drug. Accordingly, even a short course of antibiotics may sometimes have a long lasting residual effect on the microbiome, with possible metabolic or immune consequences.

The use of antibiotics during pregnancy has also been associated with increased risk of asthma in early childhood [32‐34], increased risk of childhood epilepsy, and increased risk of childhood obesity [25]. Of course, the argument could be made that the primary maternal infection was the cause for the increased risk of these conditions, rather than the treatment itself. Nevertheless, we suggest that antibiotics in pregnancy may affect the bacterial ecosystem of the mother as well as that of the fetus, and therefore that their use should be carefully considered based on what is known, and what remains unknown, regarding their effects.

Recent studies have demonstrated that priming of the immune system and microbiota-driven immune changes begin in utero and are not – as traditionally believed – induced postnatally by the newborn’s microbiota [35]. These new insights suggest that the maternal microbiota during pregnancy actually drives early postnatal innate immune development [36]. It is becoming clearer that the maternal microbiota, in concert with maternal antibodies, are important in preparing the fetus for host-microbial symbiosis later in life. The mechanisms of this phenomenon are now being explored and involve microbial molecular transfer (without any live bacteria). In addition, maternal antibodies have a dual effect, promoting pathogen neutralization whilst simultaneously enhancing microbial molecular transfer. Gomez de Aguero et al. [36] recently showed that pups of mothers transiently colonized during pregnancy have a greater capacity to avoid inflammation in response to bacterial molecules and penetration of intestinal microbes. Thus, the maternal microbiota plays a role in shaping the postnatal immune system and interferences with maternal microbiota during pregnancy may hinder the natural process of prenatal immune priming.

We believe that the issue of antibiotics during pregnancy is one of the greatest challenges of human microbiome research and certainly deserves increased focus in the form of observational and interventional studies to unravel the role of these drugs in human development.

One may argue for the importance of antibiotics during pregnancy to prevent or treat bacterial infections. Indeed, antibiotics have an important role in improving and promoting health in pregnant women. Nevertheless, as with other therapeutic modalities, overuse may be counter-productive. The realization that antibiotic use in pregnancy may interfere with the delicate balance between the pregnant woman’s microbiota, which is important for normal fetal development, may gradually decrease the enormous, and possibly excessive, utilization of these drugs during pregnancy. In case of proven maternal infection, narrow spectrum antibiotics should be preferred due to their less extensive effects on the microbiome, taking into account the association of prenatal antibiotics with increased risk of childhood asthma, epilepsy, and obesity. Future studies should focus on maternal microbiome manipulations such as personalized probiotics and fecal microbiota transplantation, and their effect on pregnancy outcome, as well as on the use of specific microbial profiling for diagnosis of pregnancy complications and prediction of long-term outcomes of newborn and maternal health.