- •
Multiple antenatal exposures have effects on the fetal lung—some adverse and other beneficial for postnatal survival.
- •
Antenatal exposures can alter lung development and interact with postnatal exposures.
- •
Early gestational lung maturation is common and promoted by both antenatal steroids and fetal exposure to inflammation.
- •
Trials indicate that antenatal steroid use may expand to late preterm infants and elective cesarean section.
- •
Antenatal steroid is a common exposure that may interact with postnatal care practices in presently unknown ways.
- •
Sepsis with chorioamnionitis is an infrequent event in term and near-term infants.
- •
Fetal exposure to inflammation/chorioamnionitis results in complex immunomodulation that may alter postnatal exposure.
- •
Growth restriction and maternal tobacco and alcohol use alter fetal lung development with effects on post-delivery outcomes.
Overview of Lung Development and Perinatal Events
Lung growth and development are the substrate on which all lung outcomes ultimately depend. This chapter emphasizes four categories of events that can modulate fetal and subsequent postnatal lung development and thus alter lung outcomes for a lifetime ( Fig. 2.1 ). McElrath and colleagues propose that there are two pathologic pathways that result in deliveries at very early gestational age (GA): intrauterine inflammation that is often chronic and aberrations of placentation/vascular development. Other examples of clinically relevant modulators of lung development are small for gestational age/intrauterine growth restriction (SGA/IUGR) and environmental exposures such as maternal tobacco and alcohol use. Lung maturation is a late phase of lung development that can be accelerated by antenatal corticosteroids and by fetal exposure to inflammation. Although infection can induce lung maturation, fetal exposures to acute or chronic chorioamnionitis also can injure the lung. There are two “elephants in the room” for this discussion of events that influence lung development. The first is the concept of what is considered normal. Any discussion of premature lungs is complicated by the lack of a normal comparison group with which to evaluate the impact of the perinatal event of interest. Although the words all or never should be sparingly used in biology and medicine, all very low-birth-weight (VLBW) deliveries must be regarded as adverse pregnancy outcomes. The 24-week GA newborn who does not have respiratory distress syndrome (RDS) is a true wonder of nature.
The second “elephant” is the complexity of the entangled pathways that regulate lung development, injury, and repair for any perinatal event that affects the lung. These three cellular and molecular response programs share signaling pathways superimposed simultaneously or sequentially on the immature lung. This complexity confounds simple interpretations about what mediator is causing which outcome. Finally, outcomes such as bronchopulmonary dysplasia (BPD) and asthma/airway disease in childhood and later life may be initiated by fetal events that then are modulated by postnatal responses of the lungs. An example is early life exposures to viral infections. This biologic complexity generates inconsistencies in clinical data and controversies. In this chapter, we provide our current understanding of how prenatal exposure to various conditions can change postnatal lung function based on both clinical information and animal models.
Lung Development: The Substrate for Adverse Events
Lung development is programmed by the fetus to be sufficiently mature to adapt rapidly to air breathing at birth. The timing of the structural development of the lung in Fig. 2.2 is given as weeks from the last menstrual period and from conception to emphasize the 2-week difference. Infants born by elective cesarean section as early term infants (38 weeks postmenstrual age) have more problems with pulmonary adaptation than infants born at 40 weeks postmenstrual age. Late preterm infants have more lung adaptation problems and more RDS for each week of birth prior to 37 weeks. Finally, lung adaptation abnormalities, RDS, and subsequently BPD become increasingly frequent as GA decreases into the early GA and very early GA categories of preterm infants.
Lung development includes development of the structural elements as well as functional maturation of fluid clearance pathways and the surfactant system. The major structural events are the completion of airway branching by about 18 weeks of gestation, followed by three generations of airway divisions to form respiratory bronchioles, and three more divisions to form alveolar ducts to about 32 to 36 weeks. Subsequently, secondary septation or alveolarization occurs to term and for several years after birth. Septation events are dynamic with about 0.06 × 10 6 distal structures at 18 to 20 weeks that increase to about 100 × 10 6 alveoli at term—a 1700-fold increase. Alveolar numbers increase only about fivefold from the term birth lung to the adult lung. There is essentially no information about the variability of the timing of normal septation in the human lung. It is also not known whether very early lung maturation changes the timing of the later gestational septation events that generate respiratory bronchioles and alveolar ducts. Some forms of pulmonary hypoplasia may result from altered septation and airway development. The injury and repair associated with BPD do inhibit and delay alveolarization of the developing lung.
Recent human anatomic and experimental data demonstrate that the healthy lung probably grows new alveoli and loses old alveoli continuously at a very slow rate. Empirically, very preterm infants with a BPD-associated “arrest” in alveolar septation must be able to grow alveoli or they could not grow and survive. These lungs may “catch up” to have lung volumes or alveolar numbers equivalent to normal lungs. The questions for the future include how alveoli grow after very preterm birth and how lung injury can be prevented.
Lung Maturation
From the clinical perspective, questions about lung maturation have focused on RDS and the surfactant system since the seminal report from Avery and Mead in 1959 that the lungs of infants who died of RDS had less surfactant. Lung maturation also includes epithelial development of ion/water regulation, thinning of the alveolar capillary barrier, and microvascular development. The first challenge is to define the timing of normal lung maturation, which is not an easy task if one assumes that most preterm infants are abnormal. In the 1970s, amniotic fluid was sampled from women with relatively normal gestations to test for lung maturation using surfactant components. The lecithin/sphingomyelin (L/S) ratio was less than 2 until after 34 to 35 weeks of gestation, and phosphatidylglycerol was seldom detected prior to 34 weeks of gestation in normal pregnancies. The true time course and the variability for the timing for normal lung maturation are not known with any precision for the human. However, lung maturity testing and inadvertent experiences with non-indicated cesarean sections prior to 37 weeks demonstrate that lung maturation, defined as absence of RDS, normally occurs after about 34 to 36 weeks in normal pregnancies.
The diagnosis of RDS for very early GA preterm infants has been confounded in clinical series and epidemiologic studies by intubation and ventilation with or without surfactant treatment shortly after birth. These intubated and ventilated infants likely carry the diagnosis of RDS even if they are not receiving supplemental oxygen. Furthermore, if these infants have infection, transient tachypnea of the newborn, a degree of pulmonary hypoplasia, or apnea requiring ventilatory support, they likely will be said to have RDS. The successful use of continuous positive airway pressure (CPAP) to minimize lung injury in very early GA infants demonstrates that many very preterm infants do not have enough RDS to need surfactant or mechanical ventilation. Infants born at 24 to 25 weeks’ GA without RDS are surprisingly common. Because lung maturation (and RDS) is a continuum from severe immaturity to sufficient maturation to avoid RDS, we think that most very preterm infants have some degree of induced lung maturation. The infant born at 24 to 26 weeks of gestation with severe respiratory failure and a poor response to surfactant and who dies soon after birth is an infant with “normal” 25-week lungs or an infant with RDS-plus (RDS plus infection or pulmonary hypoplasia, for example). Very few infants die of RDS in the United States unless they are of extremely low GA.
At the margin of lung maturity in preterm sheep, a surfactant pool size of about 4 mg/kg is sufficient to support normal gas exchange with CPAP, demonstrating that a small amount of surfactant is sufficient to protect the preterm lung from RDS. No good tests are available to quantify the lung maturation status prior to or soon after delivery. The L/S ratio or phosphatidylglycerol measurements in amniotic fluid are no longer commonly available, and other tests such as lamellar body number in amniotic fluid are imprecise. Samples of fetal lung fluid (intubated infants) or gastric aspirates soon after birth could provide information about surfactant and inflammation, but they are not used routinely. An evaluation of the messenger ribonucleic acid (mRNA) in amniotic fluid may provide a maturation profile for multiple fetal organs in the future. A clinical controversy is: Which very preterm infant should receive surfactant/ventilation or CPAP after birth? The controversy is based on the perceived risk for RDS and the ability of these infants to transition if treated by CPAP. The clinical trials demonstrate that the two approaches yield similar outcomes that marginally favor an initial trial of CPAP. However, individualized treatments could be given if the functional potential of the very preterm lung could be assessed prior to delivery. Clearly, very preterm delivery is an event that profoundly changes the developing lung. However, there is no good information about how preterm delivery and breathing, independent of oxygen exposure and injury, change the trajectory of subsequent lung development. Lung stretch from breathing and the striking changes in hormone milieu alter lung development in experimental models. The implication is that independent of injury, lung structure at 40 weeks will be different for infants born at 25 weeks or 35 weeks from those for infants born at 40 weeks.
Antenatal Corticosteroids
The antenatal administration of corticosteroids for women at risk for preterm delivery at 24 to 34 weeks of gestation is the standard of care. This therapy was developed to reduce the risk of RDS, but it also decreased the incidences of intraventricular hemorrhage and death. The therapy is supported by two National Institutes of Health Consensus Conferences, and an extensive meta-analysis. Nevertheless, there are controversies about repeated treatments, the drug and dose to be used, and the responses of selected populations of patients. We briefly review the effects of antenatal corticosteroids on fetal lungs and identify some of the remaining questions.
Maternal corticosteroid treatments have pleiotropic effects on the lung and other fetal organ systems. Corticosteroids upregulate families of genes and downregulate other genes. The net responses in an organ such as the lung are changes in anatomy, physiology, and clinical outcomes ( Table 2.1 ). The fetal primate lung responds to a 3-day maternal corticosteroid treatment, beginning at 63 days of gestation with a thinning of the mesenchyme and enlargement of air spaces ( Fig. 2.3 ). The functional effect is to increase lung gas volume. This increase in lung gas volume reflects two events: mesenchymal thinning and an interruption in alveolar septation such that saccular/alveolar numbers decrease and the sizes increase. These anatomic changes in the lung to increase lung gas volumes occur within 15 hours and prior to an increase in alveolar surfactant in fetal sheep. Antenatal steroids increase the risk for pneumothorax for about 24 hours after treatment in preterm sheep and rabbits by reducing the rupture pressures of the lung, presumably by thinning the mesenchyme without increasing support structures.
| Anatomy/biochemistry | Thin mesenchyme of alveolar-capillary structures Increased saccular/alveolar gas volumes Decreased alveolar septation Increased antioxidant enzymes Increased surfactant |
| Physiology | Increased compliance Improved gas exchange Decreased epithelial permeability Protection of preterm lung from injury during resuscitation |
| Interactions with exogenous surfactant | Improvement in surfactant treatment responses Improvement in surfactant dose-response curve Decreased inactivation of surfactant |
| Clinical | Decreased incidence of respiratory distress syndrome No effect on incidence of bronchopulmonary dysplasia Decreased mortality |
The physiologic changes that accompany the changes in lung structure and increased surfactant are increased lung compliance, improved gas exchange, and decreased permeability of the air space epithelium. Antenatal corticosteroid treatments improve surfactant treatment responses in animal models and in infants. The fetal corticosteroid exposure decreases the amount of surfactant to achieve better responses because surfactant composition is altered. Endogenous surfactant is less susceptible to inactivation by proteinaceous pulmonary edema. The decreased epithelial permeability also protects surfactant from inactivation. Lung injury during resuscitation with high tidal volumes was decreased by fetal exposure to corticosteroids in preterm surfactant-deficient sheep. The surfactant lipids and proteins do not increase in fetal sheep until 4 to 7 days after corticosteroid treatment. Therefore the early physiologic effects of antenatal corticosteroids result primarily from anatomic and other associated effects. The clinical correlates are a decrease in RDS and improved surfactant treatment responses. There was no consistent decrease in BPD in the placebo-controlled trials of antenatal corticosteroids conducted before 1990 or in clinical series since that era ( Fig. 2.4 ). Antenatal corticosteroids decrease mortality and salvage a population of infants who are at high risk for BPD, and the great majority of infants at the highest risk for BPD are exposed to maternal corticosteroids. Any difference in a BPD outcome will depend on the comparison group, and currently 80% to 90% of the at-risk population of extremely low-birth-weight (ELBW) infants have been exposed to antenatal steroids. Those not exposed will have different antenatal histories, perinatal management, and outcomes ( Table 2.2 ).
| Characteristic | Percentage of Population Affected from 1993 to 2012 | % Change from 1993 to 2012 |
|---|---|---|
| Antenatal corticosteroids | 80 | 62% Increase |
| Antenatal antibiotics | 66 | 33% Increase |
| Cesarean delivery | 56 | 19% Increase |
| Small for gestational age | 7 | No change |
| Multiple births | 24 | 8% Increase |
a A total of 34,636 infants received care at National Institutes of Child Health and Human Development (NICHD) centers from 2003 to 2012.
A number of questions remain about the appropriate use and benefits of antenatal corticosteroids. Current practice is to treat with either 12 mg of betamethasone acetate plus betamethasone phosphate given as a 2-dose treatment separated by 24 hours or a 4-dose 12-hour interval treatment with dexamethasone phosphate. These drugs are not equivalent, having different and complex pharmacokinetics in the mother and the fetus. In animal models, fetal lung maturation is induced with low and prolonged fetal exposures to betamethasone, but not with single high-dose exposures to betamethasone. The presently recommended drug doses probably cause peak fetal exposures to betamethasone phosphate or dexamethasone phosphate that are higher than necessary, and treatment intervals may not be optimal. Research is needed to optimize the benefits of antenatal corticosteroid treatments while minimizing risks.
Trials with antenatal corticosteroids were completed by about 1990, and only about 100 randomly assigned infants were delivered prior to 28 weeks’ GA. However, in clinical practice, the use of corticosteroids is routine for infants who are to be offered intensive care of GAs as low as 22 to 23 weeks. Because most of these extremely preterm infants have a diagnosis of RDS (appropriate or not), a decrease in RDS may not be a good indicator of benefit. For example, Garite and coworkers found no decrease in RDS but did document decreased severity of RDS in early GA infants. As demonstrated in Fig. 2.3 , the primate lung can respond to antenatal corticosteroids at very early (previable human equivalent) GAs, as do explants of human fetal lung at less than 20 weeks of gestation. Antenatal corticosteroids should be used if a very preterm delivery is likely to occur within the next 7 days if the goal is for the infant to survive.
Although the neonatal community is primarily concerned about the lung outcomes of very preterm infants, large populations of term infants delivered by elective cesarean section and late preterm infants delivered for multiple indications require respiratory care after birth. Sinclair pointed out in 1995 that the early clinical trials demonstrated lung benefits for infants exposed to antenatal corticosteroids who were delivered after 32 to 34 weeks of gestation. The respiratory morbidities RDS and transient tachypnea of the newborn increase greatly for deliveries at less than 38 weeks’ GA for even modest prematurity. The recently completed randomized controlled trial of corticosteroid use for deliveries between 34 0 and 34 6 weeks enrolled 2831 patients and demonstrated that steroids improved the combined primary outcome of need for respiratory support, stillbirth, or death ( Table 2.3 ). The benefit was for respiratory morbidity, but an unanticipated complication was hypoglycemia. There presently is no follow-up for this trial. The benefits are modest and the risks are poorly defined.
| Placebo ( N = 1400) | Antenatal Steroid ( N = 1427) | P Value | |
|---|---|---|---|
| Gestational age at delivery (weeks) | 36.1 | 36.1 | — |
| Primary outcome (%) | 14.4 | 11.6 | .023 |
| Severe respiratory morbidity (%) | 12.1 | 8.1 | <.001 |
| Surfactant treatment (%) | 3.1 | 1.8 | .031 |
| Hypoglycemia (%) | 15 | 24 | <.001 |
A provocative randomized trial of antenatal corticosteroids using betamethasone (Celestone) for 998 women scheduled for elective cesarean delivery demonstrated that respiratory distress after use of corticosteroids decreased from 11.4% to 5.4% for deliveries at 37 weeks and from 1.5% to 0.6% for deliveries at 39 weeks. A second trial used an untested treatment schedule of 3 doses of 8-mg dexamethasone given 12 hours apart prior to elective cesarean section at 38 weeks. The steroid treatment decreased admission to the neonatal intensive care unit from 3.9% to 1.6% in 1290 randomly assigned women with a number needed to treat of 43. These trials identify much larger populations of women for treatment with antenatal corticosteroids beyond prematurity. There are minimal long-term outcome data to assure safety for such widespread use.
Repeated courses of antenatal corticosteroids are conceptually attractive because many early GA fetuses are not delivered within 7 days of maternal treatment and the benefits of therapy seem to decrease with time after treatment. In one report, about 60% of women who received an initial treatment with corticosteroids delivered more than 7 days later. In animal models, second courses of corticosteroids progressively increase the indicators of fetal lung maturation. The risks are adverse effects of repeated fetal exposures on fetal somatic and brain growth, effects that occur in animal models ( Fig. 2.5 ). A meta-analysis of 4733 randomized pregnancies in trials reported a modest benefit for RDS and severe lung disease with modest growth effects for repeated courses of treatment. A subsequent trial of 1858 women randomly assigned to a 14-day retreatment interval versus no retreatment found no benefit for RDS, mortality, or other outcomes, but a decrease in birth weights and head circumferences. Reports of 2-year neurodevelopment outcomes are reassuring for infants exposed to fewer than four courses of antenatal corticosteroids. Another option is a rescue treatment when a woman has received an initial treatment and again has a high risk of delivery prior to 34 weeks. There are no recommendations from learned societies about the use of repeated corticosteroid treatments.
Antenatal Infection/Inflammation
Overview of Fetal Inflammation
The human fetus is normally considered to be in an environment protected from infection. However, the human fetus can be exposed to a variety of pathogens, which may initiate an inflammatory process in the placenta, chorioamnion, or fetus. For example, human fetuses are exposed to viral pathogens as a consequence of maternal viremia. The patterns of fetal injury after exposure to agents such as varicella and cytomegalovirus depend on the period of gestation during which the infection occurs. Similarly, the fetus can acquire a spirochete infection with syphilis or a parasitic infection with toxoplasmosis secondary to maternal infection, and each causes characteristic syndromes depending on the gestational timing of exposure. These infections are not generally viewed as predominantly inflammatory, although the fetal injury and immune responses have inflammatory characteristics. Asphyxia with injury to fetal tissue also causes inflammation as part of the injury and the repair process. Similarly, normal labor is associated with an increase in proinflammatory mediators. Both innate and acquired inflammatory responses of the fetus are generally considered less robust than those in the child or adult because the response systems in the fetus are immature and pregnancy is an immune-suppressive environment. For example, fetal inflammatory responses to pathogens such as group B streptococcus and Listeria monocytogenes are blunted, resulting in severe infection and often death of the fetus or newborn. The most common fetal infectious exposure is to chorioamnionitis, which is associated with preterm labor and delivery. In this section, we identify the questions and controversies about the associations of chorioamnionitis with a range of effects on the fetal and newborn lung.
Diagnosis of Chorioamnionitis
Chorioamnionitis can be either a clinical syndrome or a silent, indolent process. The clinical diagnosis of chorioamnionitis is made when a pregnant woman has a constellation of findings that include fever, a tender uterus, an elevated blood granulocyte count, and bacteria and/or inflammatory cells in amniotic fluid and often preterm or prolonged rupture of membranes. Clinical chorioamnionitis is an imprecise diagnosis that has little prognostic or treatment value. The diagnosis of clinical chorioamnionitis is frequently made for near-term or term labors, and occasionally infection can be caused by highly virulent organisms. Before 30 weeks of gestation, clinical chorioamnionitis is most often diagnosed after attempts to delay preterm delivery or with preterm prolonged rupture of membranes. Another method to diagnose chorioamnionitis is by histopathology of the chorioamnion with inflammation, indicating histologic chorioamnionitis . The amount of infiltration of the chorioamnion by inflammatory cells and the intensity of secondary changes are used to grade the severity of the fetal exposure to inflammation. Inflammation of the cord, called funisitis , is generally considered to indicate a more advanced inflammatory process that involves the fetus. Another diagnostic approach is to culture amniotic fluid or fetal membranes for organisms or to assay amniotic fluid for proinflammatory inflammatory mediators such as tumor necrosis factor α (TNF-α) and interleukin-1 (IL-1) and IL-6. With the recognition that only a minority of organisms in the human biome can be cultured, polymerase chain reaction (PCR) and DNA sequencing techniques are being used to demonstrate that chorioamnionitis is often polymicrobial with organisms that cannot be cultured. Technologies to identify multiple proteins in biologic fluids also are being adapted to develop proteomic biomarkers for chorioamnionitis in amniotic fluid. These technologies have the potential to rapidly diagnose inflammation and to identify specific organisms. Such approaches will change the understanding of fetal exposures to inflammation and pathology related to specific organisms.
The chorioamnion is fetal tissue, and the amniotic fluid surrounding the fetus is in direct contact with the fetal gut, skin, and lung. Therefore the fetus will be exposed to inflammation if there is histologic chorioamnionitis or if the amniotic fluid contains mediators of inflammation. The Venn diagram in Fig. 2.6 illustrates the diagnostic conundrum. Clinical chorioamnionitis does not correlate well with the subsequent diagnosis of histologic chorioamnionitis, and an amniotic fluid diagnosis of infection may or may not predict chorioamnionitis associated with preterm delivery. PCR-based analyses of amniotic fluid call into question the assumption that fetal colonization with organisms is abnormal and will cause preterm delivery. Gerber and associates demonstrated that 11% of 254 presumably normal amniotic fluid samples collected at 15 to 19 weeks of gestation for genetic analysis were PCR-positive for Ureaplasma urealyticum . Although 17 of the 29 Ureaplasma -positive pregnancies had preterm labor, only 2 fetuses were delivered before 34 weeks of gestation. Perni and colleagues analyzed 179 amniotic fluid samples and found that 13% were positive for Ureaplasma and 6% were positive for Mycoplasma hominis , and in 28 of the 33 pregnancies with positive amniotic fluid samples infants were not delivered preterm. Deep sequencing of the microbial genome reveals “colonization” of the amniotic fluid in some women without apparent adverse effects, suggesting the presence of an amniotic biome. Attempts to extensively culture the placenta/chorioamnion have recovered multiple organisms of low virulence that include vaginal flora. The severity of the chorioamnionitis does not correlate well with the organisms but tends to be more severe with Ureaplasma and Mycoplasma species. Furthermore, in many preterm deliveries, polymicrobial organisms are recovered by culture or PCR from the amniotic fluid. The unknowns are the association of organisms with pregnancies that are not delivered preterm and the variety of organisms that can be identified by PCR. For example, Steel and coworkers used a fluorescent probe for a common 16S ribosomal RNA bacterial sequence and identified organisms deep within the membranes of all preterm deliveries and many term deliveries. These results suggest that the human pregnancy can tolerate colonization/infection with low-pathogenicity organisms. The provocative question is: Does the fetus need exposure to a biome for normal development?
There is no clear answer to the question “What is chorioamnionitis?” The multiple ways to make the diagnosis are not necessarily congruent. Furthermore, if one accepts that chorioamnionitis results from colonization/infection, then the diagnosis is imprecise in the extreme in relation to how infectious diseases are generally diagnosed. The diagnosis of an infection includes the identity of the organism, an estimate of the duration of infection, its intensity, and specific sites of involvement. The diagnosis of chorioamnionitis contains none of these elements. Research is now linking genetically determined inflammatory response characteristics of the mother and fetus with prematurity. The chronic indolent chorioamnionitis associated with prematurity may result from the interaction of the environment and the genetically determined immunomodulatory characteristics of the mother and fetus. Challenges for the future are how to better diagnose and to understand what makes patients susceptible to chorioamnionitis, and how to quantify the severity potential for fetal injury from the chorioamnionitis.
Clinical Pulmonary Outcomes of Fetal Exposure to Inflammation/Infection
An important perspective is how infrequent severe sepsis and pneumonia are in large delivery populations where the diagnosis of clinical chorioamnionitis will apply to 5% to 10% of the population. A clinical experience from Dallas is representative: of 23,321 deliveries; 7% (1660) of cases were diagnosed as clinical chorioamnionitis and 1571 infants were asymptomatic. Only 61 (0.26%) infant were ultimately thought to be sick and in need of the neonatal intensive care unit, primarily for respiratory and sepsis symptoms. The occurrence of culture-proven sepsis is very infrequent and of 396,586 infants, 389 (0.1%) had early-onset sepsis; 60% of these had clinical chorioamnionitis. Most of the septic infants were born preterm and had symptoms ( Fig. 2.7 ). Another concern is a sepsis/pneumonia risk for infants born before term after prolonged rupture of membranes (PROM). Should women with PROM have immediate delivery on presentation or expectant management with maternal treatment with antibiotics? A recent trial demonstrated that expectant management was associated with fewer cesarean sections, less RDS, and no difference in neonatal sepsis. While a risk of infection is associated with chorioamnionitis, the risk is very low for asymptomatic infants and overtreatment with antibiotics is prevalent.
The fetal effects of chorioamnionitis are much more frequent in preterm populations. A decreased incidence of RDS was associated with preterm PROM, a surrogate marker for chorioamnionitis, as early as 1974. Watterberg and associates reported in 1996 that ventilated preterm infants exposed to histologic chorioamnionitis had a lower incidence of RDS but a higher incidence of BPD than infants not exposed to chorioamnionitis. Furthermore, the initial tracheal aspirates from infants exposed to chorioamnionitis contained proinflammatory mediators such as IL-1, IL-6, and IL-8, indicating that the lung inflammation was of antenatal origin. Other reports support that association. Clinical chorioamnionitis was associated with decreased death in all infants born at or before 26 weeks of gestation in the United Kingdom and Ireland in 1995. Hannaford and colleagues identified U. urealyticum as an organism of fetal origin that was associated with a decreased risk of RDS. Lahra and coworkers noted, in a population of 724 preterm infants, that RDS was decreased for infants exposed to histologic chorioamnionitis (odds ratio [OR] 0.49, 95% confidence interval [95% CI] 0.31–0.78) or chorioamnionitis plus funisitis (OR 0.23, 95% CI 0.15–0.35) relative to no chorioamnionitis. This group also reported its 13-year experience that histologic chorioamnionitis (with or without funisitis) was associated with a decreased risk of BPD (OR 0.58, 95% CI 0.51–0.67). A recent systematic meta-analysis showed that histologic chorioamnionitis was associated with BPD, but there was low confidence in the findings because of evidence of publication bias.
In contrast, other reports associate chorioamnionitis with poor pulmonary and other outcomes. Hitti and associates, for example, reported that high levels of TNF-α in amniotic fluid predicted prolonged postnatal ventilation, suggesting early and persistent lung injury from chorioamnionitis. Ramsey and colleagues also demonstrated that chorioamnionitis increased neonatal morbidities. Laughon and coworkers, after extensively evaluating and culturing the placentas of 1340 infants born before 28 weeks of gestation, found no association between histologic chorioamnionitis, funisitis, or specific organisms and the initial oxygen requirements of the infants or subsequent development of BPD. The Canadian Neonatal Network also reported that clinical chorioamnionitis was not predictive of RDS or BPD.
These discrepant reports need to be understood within the complexities of the diagnosis of chorioamnionitis as well as the factors contributing to the diagnosis of RDS or BPD. Van Marter and associates evaluated the outcomes of ventilated and VLBW infants and found that chorioamnionitis was associated with a decreased incidence of BPD (OR 0.2). However, BPD was increased if the infant had been exposed to chorioamnionitis and either was mechanical ventilated for more than 7 days (OR 3.2) or had postnatal sepsis (OR 2.9). Lahra and coworkers noted the same associations in an unselected population of 761 infants with gestations less than 30 weeks. BPD was lower in infants exposed to histologic chorioamnionitis than in infants without chorioamnionitis, as noted previously. However, the combination of histologic chorioamnionitis and postnatal sepsis increased the risk for BPD (OR 1.98, 95% CI 1.15–3.39). These reports demonstrate that antenatal and postnatal exposures interact to change outcomes such as BPD. Been and colleagues reported that newborns exposed to chorioamnionitis with fetal involvement had more severe RDS and impaired surfactant treatment responses. In contrast, infants exposed to chorioamnionitis without fetal involvement had minimal lung disease. The severity of the chorioamnionitis and postnatal interventions confound simple correlations between chorioamnionitis and outcomes such as RDS and BPD.
Other studies have explored the associations of antenatal inflammation with postnatal lung outcomes with measurements of proinflammatory cytokines in cord plasma and tracheal aspirates collected shortly after birth. In general, cord plasma from early gestational deliveries had higher proinflammatory cytokine levels than cord plasma from term deliveries, but the median values were not greatly different, suggesting little useful resolution between the preterm and term populations. Although Ambalavanan and coworkers could detect differences in blood cytokines collected within 4 hours of birth for infants in whom BPD developed from those without BPD, the resolution between the populations was not clinically useful for the prediction of risk of BPD. Similarly, Paananen and associates found higher selected cord plasma cytokine levels in infants exposed to severe chorioamnionitis. The cord cytokine levels decreased with age for infants at lower risk for BPD, but cord cytokine levels were not reliable predictors of BPD. De Dooy and colleagues could predict chorioamnionitis from IL-8 levels in tracheal aspirates collected soon after birth, but the clinical utility of that information also is unclear. Been and colleagues did find that vascular endothelial growth factor levels in initial tracheal aspirates were predictive of BPD. However, there is no compelling evidence that measurements of proinflammatory mediators in cord plasma or tracheal aspirates will identify high-resolution biomarkers for either chorioamnionitis or the pulmonary outcomes RDS and BPD.
The inconsistent clinical correlates most likely result from the imprecise nature of the diagnosis of chorioamnionitis and its association with different populations of infants. An example of the inconsistency is the diagnosis of fetal exposures by histologic chorioamnionitis or by blood culture for Ureaplasma collected from the cord at delivery and the outcomes of RDS and BPD for the same cohort of consecutive patients ( Table 2.4 ). The associations with histologic chorioamnionitis and culture positivity for Ureaplasma for BPD are the opposite of those for RDS in the same cohort of patients. The diagram in Fig. 2.8 may help frame the question about the variable outcome. A progressive chorioamnionitis caused by virulent organisms may cause severe postnatal lung and systemic inflammation with the outcomes of more severe RDS, BPD, or sepsis/death. Such outcomes are relatively infrequent in VLBW infants who are not stillborn. Fewer than 2% of VLBW infants have positive blood culture results at birth. Chronic, indolent chorioamnionitis caused by organisms such as Ureaplasma can induce lung maturation (less RDS), but that early maturation may be associated with more BPD. These associations may depend on how the diagnosis of chorioamnionitis is made (clinical, histopathologic, other), and the population of infants studied (ventilated only, all VLBW infants, other selected populations). In an attempt to better establish a cause-and-effect relationship, Viscardi and associates correlated the intensity of the inflammatory response to chorioamnionitis in the fetal membranes with the clinical outcome of BPD ( Fig. 2.9 ). More severe chorioamnionitis at delivery predicted a higher incidence and greater severity of BPD.
| Percentage of Population | |||
|---|---|---|---|
| Yes | No | P Value | |
| Histologically Diagnosed Chorioamnionitis | |||
| Respiratory distress syndrome | 61 | 73 | .008 |
| Bronchopulmonary dysplasia | 17 | 17 | NS |
| Fetal inflammatory response syndrome | 44 | 18 | .001 |
| Positive Result of Cord Blood Culture | |||
| Respiratory distress syndrome | 66 | 65 | NS |
| Bronchopulmonary dysplasia | 27 | 10 | .001 |
| Fetal inflammatory response syndrome | 41 | 26 | .007 |
a A comparison of fetal exposures diagnosed by histologic chorioamnionitis or positive results of cord blood cultures for Ureaplasma or Mycoplasma .
Experimental Results: The Link Between Fetal Exposure to Inflammation and Lung Maturation and Lung Remodeling
Animal models have consistently demonstrated that fetal exposure to inflammation causes lung injury and induces lung maturation. In a strict sense, lung maturation should probably be called “dysmaturation” because the beneficial effects of improved lung mechanics are also accompanied by altered lung development. Bry and colleagues first demonstrated in 1997 that inflammation induced lung maturation by intraamniotic injection of IL-1α. Our group found that intraamniotic injection of the proinflammatory mediator endotoxin from Escherichia coli in sheep caused chorioamnionitis (inflammatory cells and increased IL-1β and IL-6 mRNA expression in the chorioamnion), inflammatory cells in amniotic fluid, and increased IL-8 protein levels in amniotic fluid. The chorioamnionitis was accompanied by inflammation of the fetal lung, as demonstrated by recruitment of granulocytes to the fetal lung tissue and air spaces within 24 hours and expression of multiple proinflammatory mediators ( Fig. 2.10 ). Apoptosis of lung cells increased at 24 hours, and proliferation increased at 3 days. This lung inflammation/injury sequence included multiple indicators of lung microvascular injury—epithelial nitric oxide synthase and vascular endothelial growth factor decreased, and medial smooth muscle hypertrophied. Thus intraamniotic endotoxin caused lung inflammation and an injury sequence that resulted in lung remodeling.
There are very few studies of respiratory muscle function in preterms. Recently Song et al. reported that intraamniotic endotoxin exposure in preterm fetal lambs resulted in transient inflammation in fetal diaphragm followed by atrophic gene expression resulting in impaired diaphragm contractility.
Inflammation was associated with the induction of the mRNAs for the surfactant proteins within 12 to 24 hours, persistent elevation of those mRNAs for weeks, and an increase in alveolar surfactant proteins and lipids with improved lung function within 5 to 7 days. The improvement in lung function was accompanied by a decrease in mesenchymal tissue and an increase in potential gas volume in the fetal lung. The residual effects of the injury at 7 days were greater thickness of the pulmonary microvessels and a reduction in secondary septation of the alveoli. However, the net effect was a lung that was easier to ventilate because of improved compliance and that had better gas exchange ( Fig. 2.11 ). Of note, the lung injury followed by maturation sequence did not result from “fetal stress,” because fetal blood cortisol levels did not increase. Intraamniotic endotoxin also increased surfactant protein mRNA in the primate. Lung inflammation results in accelerated lung maturation.
