Prematurity remains the leading cause of neonatal morbidity and mortality, accounting for $26 billion a year in immediate costs, despite the implementation in obstetrics of a host of risk stratification algorithms and strategies for risk reduction, including the use of some medications.
Dr. Michal A. Elovitz
It now is questionable whether injectable 17-alpha hydroxyprogesterone caproate (Makena) truly is efficacious in women who’ve had a prior spontaneous preterm birth (sPTB) – a Food and Drug Administration advisory committee last year recommended withdrawing it from the market based on results of an FDA confirmatory study. Even if the drug were efficacious, only a small percentage of the women who have an sPTB have had a prior one. The majority of sPTB occurs among women without such a history.
Vaginal progesterone appears to confer some protection in women found to have a short cervix during the second trimester, but this approach also has limited reach: Only 9% of women with sPTB had an antecedent short cervix in a 2017 study.1 Like a history of sPTB, screening for short cervical length is a potentially helpful strategy for risk reduction, but it is not a strategy that will significantly impact the overall rate of prematurity.
We’ve fallen short in our goals to significantly reduce the public health impact of prematurity partly because we still do not understand the exact pathways and mechanisms by which sPTB occurs. The main working paradigm for myself and many other researchers over the past 2 decades has centered on infection in the uterus triggering inflammation, followed by cervical remodeling and ripening. Research in animal models, as well as human clinical trials targeting various infections and inflammation, have led to some insights and discoveries, but no successful interventions.
In the past decade, however, our research framework for understanding sPTB incorporates new questions about immunologic, microbiological, and molecular/cellular events that happen in the cervicovaginal space. We’ve learned more about the cervicovaginal microbiota, and most recently, our research at the University of Pennsylvania has elucidated the role that nonoptimal bacteria play in disrupting the cervical endothelial barrier and initiating the process of cervical remodeling that likely precedes sPTB.
We also know that this association is stronger in black women and may help explain some of the observed racial disparities in sPTB. Although more research is needed to determine specific therapeutic strategies, new doors are open.
Host immune-microbial interactions
This new research paradigm has involved stepping back and asking basic questions, such as, what do we really know about the cervicovaginal space? In actuality, we know very little. We know little about the immune function of the vaginal and cervical epithelial cells in pregnancy, for instance, and there is a large gap in knowledge regarding the biomechanics of the cervix – a remarkable organ that can change shape and function in a matter of minutes. Studies on the biomechanics of the cervix during pregnancy and in labor are still in their infancy.
However, lessons can be drawn from research on inflammatory bowel disease and other disorders involving the gut. In the gastrointestinal tract, epithelial cells have been found to act as sentinels, forming a mucosal barrier against bacterial pathogens and secreting various immune factors. Research in this field also has shown that microbes living in the gut produce metabolites; that these microbial metabolites may be the key messengers from the microbial communities to the epithelial barrier; and that the microbes, microbial metabolites, and immune responses are responsible for triggering inflammatory processes in the tissues underneath.
In 2011, Jacques Ravel, PhD, who was part of the National Institutes of Health’s Human Microbiome Project, characterized the vaginal microbiome of reproductive-age women for the first time.2 His paper classified the vaginal microbial communities of approximately 400 asymptomatic women of various ethnicities into five “community state types” (CSTs) based on the predominant bacteria found in the cervicovaginal space.3
On the heels of his research, Dr. Ravel and I launched an NIH-funded study involving a prospective cohort of 2,000 women with singleton pregnancies – the Motherhood & Microbiome cohort – to look at the cervicovaginal microbiota, the local immune response, and the risk of sPTB.4 Cervicovaginal samples were collected at 16-20 weeks’ gestation and during two subsequent clinical visits. From this cohort, which was composed mostly of African American women (74.5%), we conducted a nested case-controlled study of 103 cases of sPTB and 432 women who delivered at term, matched for race.
We carefully adjudicated the deliveries in our 2,000-person cohort so that we homed in on sPTB as opposed to preterm births that are medically indicated for reasons such as fetal distress or preeclampsia. (Several prior studies looking at the associations between the cervicovaginal microbiome had a heterogeneous phenotyping of PTB that made it hard to draw definitive conclusions.)
Our focus in assessing the microbiome and immunologic profiles was on the samples collected at the earliest time points in pregnancy because we hoped to detect a “signature” that could predict an outcome months later. Indeed, we found that the nonoptimal microbiota, known in microbiological terms as CST IV, was associated with about a 150% increased risk of sPTB. This community comprises a dominant array of anaerobic bacteria and a paucity of Lactobacillus species.
We also found that a larger proportion of African American women, compared with non–African American women, had this nonoptimal microbiota early in pregnancy (40% vs. 15%), which is consistent with previous studies in pregnancy and nonpregnancy showing lower levels of Lactobacillus species in the cervicovaginal microbiome of African American women.
Even more interesting was the finding that, although the rate of sPTB was higher in African American women and the effect of CST IV on sPTB was stronger in these women, the risk of sPTB couldn’t be explained solely by the presence of CST IV. Some women with this nonoptimal microbiome delivered at term, whereas others with more optimal microbiome types had sPTBs. This suggests that other factors contribute to African American women having a nonoptimal microbiota and being especially predisposed to sPTB.
Through the study’s immunologic profiling, we found a significant difference in the cervicovaginal levels of an immune factor, beta-defensin 2, between African American women who delivered at term and those who had a sPTB. Women who had a sPTB, even those who had higher levels of Lactobacillus species, had lower levels of beta-defensin 2. This association was not found in non–African American women.
Beta-defensin 2 is a host-derived antimicrobial peptide that, like other antimicrobial peptides, works at epithelial-mucosal barriers to combat bacteria; we have knowledge of its action from research on the gut, as well as some studies of the vaginal space in nonpregnant women that have focused on sexually transmitted infections.
Most exciting for us was the finding that higher levels of beta-defensin 2 appeared to lower the risk of sPTB in women who had a nonoptimal cervicovaginal microbiota. There’s an interplay between the host and the microbiota, in other words, and it’s one that could be essential to manipulate as we seek to reduce sPTB.