Keywords

Evidence for Developmental Programming: Epidemiological and Experimental Evidence

Like many organs, the lung can be affected during its period of development in such a way that its later structure and function are altered. This can occur as a result of altered structural development or altered cellular function as a result of genetic or epigenetic alterations. Many such alterations can impair lung function for much or all of the life span and can also increase the risk of later pulmonary illness. Developmental insults or challenges that have the potential to induce persistent alterations in lung structure or function include physical and metabolic changes in the developmental environment. As the lung undergoes development during both prenatal and early postnatal life, its development can be altered by environmental and/or epigenetic factors that affect the intrauterine environment or the environment during infancy and childhood. Altered development can include gross structural alterations such as lung hypoplasia caused by congenital diaphragmatic hernia (CDH) or oligohydramnios, for example, or more subtle changes that may lead to persistent oxidative stress or inflammation of lung tissue.

In this chapter, we review current knowledge of factors commonly operating during development that can persistently affect the structure and function of the lung , and the underlying processes. It has long been known that impaired lung function during early life is associated with impaired lung function later in life; this is known as “tracking” of lung function [1]. More recently, it has become accepted that impaired lung function during early life can be caused by environmental factors operating during fetal and early postnatal life, and that these can consequently lead to impaired lung function and increased disease risk much later in life: this has become known as “developmental programming” [2, 3]. These two concepts (tracking and developmental programming) are both manifestations of the same phenomenon, as both are caused by persistent changes in the structure and/or function of lung tissue. They are both likely a result of lung development proceeding along a genetically determined program, such that alterations in development at one stage cannot be corrected at a subsequent stage.

Evidence for developmental programming of the lung in humans has been derived from numerous epidemiological and clinical studies and in recent years has become supported and enhanced by an abundance of experimental evidence. Lung diseases of infancy and childhood thought to have fetal or neonatal origins include chronic lung disease of prematurity, lung hypoplasia , persistent pulmonary hypertension (PPH), and asthma. Pulmonary diseases of later life, such as chronic obstructive pulmonary disease (COPD), are thought to have their origins in early life [4]. Similarly, it is likely that accelerated lung aging may be a consequence of altered lung development , such as impaired alveolarization .

Fetal Causes of Altered Lung Development and Later Pulmonary Illness

Fetal Nutrition and Growth

Restricted fetal growth affects around 8–10 % of live births and is recognized as a significant cause of perinatal morbidity [5]. Multiple organs are known to be developmentally affected by fetal growth restriction (FGR), in that their structure and function can be persistently altered. One of the first studies to show a link between FGR and later pulmonary health was an epidemiological study of adult men in the UK [6]; that study showed statistical links between low birthweight (and weight at 1 year) on later lung function and death from respiratory illness. Since that study was published, similar studies from several countries have confirmed that FGR can have a persistent adverse effect on the developing lung [7]. These observations have led to numerous investigations of the underlying processes. The most common causes of FGR are placental insufficiency and a maternal diet that is restricted in calories or protein. Placental insufficiency usually occurs as a result of maternal vascular disease (e.g., hypertension or pre-eclampsia), multiple gestation or maternal tobacco smoking . In the presence of placental insufficiency and FGR, the fetus is likely to be hypoglycemic, hypolipidemic, acidemic, and hypoxemic [8, 9]. Thus, the etiology of altered lung development in the presence of FGR is likely to be multifactorial, and to date it has been difficult to isolate any given factor as being a cause of altered lung development .

Respiratory Effects of FGR Observed During Infancy and Childhood

During infancy, FGR has been associated with respiratory insufficiency and poor lung function [10–13]. It is now clear that this deficit in lung function persists into childhood and beyond, supporting the concept of “tracking” of lung function. School-age children who experienced FGR have poorer lung function, as evidenced by reduced expiratory flow rates and greater airway reactivity, compared to children of the same age born with normal birthweight [14–16]. Many infants with FGR are born preterm, and such infants are found to have even poorer lung function than FGR infants born at term [17]. In twin pairs of children and adolescents, it is apparent that the twin with the lower birthweight, indicative of FGR, has poorer lung function and increased airway reactivity [18, 19]. Such studies of twins are important because they eliminate many maternal and genetic factors.

Respiratory Effects of FGR Observed During Adult Life

Early studies on the relation between fetal growth and adult lung function only had access to data on birthweight. Therefore, such data could have included subjects who were of low birthweight as a result of preterm birth . Later studies have avoided the potentially confounding effects of preterm birth by only including subjects born at term [20]; such studies have in general confirmed a significant relationship between birthweight at term and respiratory illness during adulthood [21–24]. Nine studies of lung function in male and female subjects, performed in several different countries, formed the basis of a meta-analysis of birthweight effects on adult lung function [7]; the analysis revealed a small but significant reduction in adult forced expiratory volume in 1 s (FEV1) with decreasing birthweight. Taken together with lung function data from infants and children, the above studies provide strong evidence that FGR causes changes in lung development that impair lung function throughout postnatal life and may contribute to COPD in adult subjects of low birthweight [25, 26]. Although there are apparently no morphometric analyses of the lungs of adults exposed to FGR, data from animal studies suggest that impaired alveolarization and structural alterations to small conducting airways (see below) play an important mechanistic role.

Experimental Studies

Numerous animal studies using a variety of species have been performed in an attempt to elucidate the mechanisms by which FGR alters fetal lung development [10]. Experimentally, FGR has been induced by maternal food restriction (often protein restriction), mechanical constriction of the uterine or umbilical arteries, maternal hyperthermia or pre-pregnancy ablation of placentation sites.

Using sheep, a series of studies has shown that FGR, experimentally induced by placental insufficiency, has a profound effect on fetal lung development . In such models, the major factors leading to changes in organ development are likely to be fetal nutrient restriction and hypoxemia, as well as possible elevations in circulating corticosteroids. FGR was induced to coincide with the saccular and early alveolar stages of lung development, which occur before birth in sheep as in humans [27]. The effects of FGR on lung architecture were studied in the near-term fetus , 8-week-old lamb (childhood equivalent), and 2-year-old (adult) sheep. Interestingly, some changes not present before term became apparent with postnatal development. At 8 weeks of age, there were 31 % fewer alveoli in FGR lambs, compared to controls, and this reduction was still evident in adult sheep [28, 29]. In accordance with fewer alveoli, the alveoli of FGR animals were larger at both 8 weeks and 2 years. Thus, FGR induced by late-gestational nutrient restriction led to a permanent reduction in alveolarization , which resulted in a reduction in alveolar surface area relative to both lung volume and body weight [29]. It was also found that the inter-alveolar walls were significantly thicker than in control animals at 8 weeks (73 % thicker) and 2 years (47 % thicker) [29] due to an increased extracellular matrix ; as alveolar walls normally become progressively thinner with lung development, thicker alveolar walls suggest persistently inhibited development. Importantly, the blood–air barrier was also thicker: in adults, it was 43 % thicker after FGR than in control animals [29]. The combination of fewer alveoli, thicker alveolar walls, and a thicker blood–air barrier would be likely to reduce pulmonary compliance and diffusing capacity, as found at 8 weeks [30]. The persistence of such alterations in lung architecture into adulthood would be expected to impair adult lung function and to increase the rate of the age-related decline in lung function.

The findings in sheep are largely consistent with findings in other laboratory species following nutrient restriction coincident with lung development [31–33]. Using rodent models of FGR, it has been shown that the deposition of extracellular matrix proteins, including elastin, is impaired by FGR [32, 34], and that the impairment of alveolarization is likely to be mediated by alterations in retinoic acid receptors [35]. In addition to alterations in alveolar structure, pulmonary surfactant has been reported to be affected by FGR, although findings are not consistent between models [10].

The conducting airways may also be affected by FGR, although few structural studies have been undertaken. In near-term fetal sheep subjected to late-gestation FGR, it was found that airway walls were proportionally thinner than in controls (less cartilage and epithelium ); however, these changes had resolved by 8 weeks after birth [36]. There was also evidence that airway submucosal glands were attenuated in the FGR animals after birth [36].

The mechanisms by which FGR alters lung development , with some changes persisting into adult life, are still under investigation. In most cases, FGR results from placental insufficiency, which induces a complex and variable mix of physiological alterations in the fetus , including reduced nutrient supply, altered levels of growth factors , hypoxemia, and endocrine alterations, including elevated levels of circulating corticosteroids [37]. Each of these factors has the potential to alter aspects of prenatal lung development , including alveolarization . Owing to the postnatal lung’s limited ability to recover from structural deficits that occur during development, many of the alterations induced during fetal (and neonatal) life are likely to persist throughout life. There is now strong evidence that FGR induces epigenetic alterations in genes that are critical to lung development [38].

Amniotic Fluid Volume

The fetal lung develops in an expanded state; indeed, the volume of luminal liquid within the lung during fetal life (fetal lung liquid) is greater than functional residual capacity (FRC) in the postnatal air-filled lung [39]. A remarkable feature of fetal lung development is that the degree of lung expansion by fetal lung liquid has a profound, direct effect on lung growth. This is clearly demonstrated experimentally by the removal of lung liquid leading to cessation of lung growth and lung hypoplasia ; conversely, chronically increasing the degree of fetal lung expansion (e.g., by tracheal occlusion) leads to augmented growth of lung tissue [39].

An inadequate amount of amniotic fluid surrounding the fetus (oligohydramnios) is a common cause of fetal lung hypoplasia . Clinically, oligohydramnios is associated with respiratory insufficiency at birth and increased hospitalization during childhood [40]. Oligohydramnios can be caused by reduced production of amniotic fluid (usually due to fetal renal abnormalities) or by premature rupture of the fetal membranes. The lack of amniotic fluid allows the elastic uterine wall to “force” the fetal trunk into a state of increased flexion; this exaggerated trunk flexion increases intra-abdominal pressure, elevating the fetal diaphragm and compressing the fetal lung, as observed in the sheep and human fetus [41, 42]. Thoracic dimensions are reversibly reduced by experimental oligohydramnios [43]. The degree of lung hypoplasia at the time of birth is related to the duration and severity of the oligohydramnios. Oligohydramnios caused by reduced fetal urine production occurs from earlier in gestation than oligohydramnios resulting from membrane rupture; hence, the effects on the developing lung are likely to be more severe.

Attempts have been made to restore amniotic fluid volume, and hence fetal lung expansion, by infusing an appropriate fluid into the amniotic sac. In an ovine model of oligohydramnios, serial infusions of artificial amniotic fluid led to a reduction in lung hypoplasia and normalized airspace fraction [44]. A recent review of five randomized controlled trials suggests that amnioinfusion leads to a reduction in pulmonary hypoplasia [45]. It is likely that amnioinfusion improves fetal lung development by reducing the degree of fetal trunk flexion, allowing the fetal lungs to expand; it may also improve placental perfusion, resulting in improved oxygenation and nutrient delivery to the fetus [45].

Congenital Diaphragmatic Hernia

A less common cause of fetal lung hypoplasia is CDH, in which a defect in closure of the diaphragm (unilateral or more rarely bilateral) allows the lung to recoil and abdominal contents to enter the chest. The defect normally occurs during embryogenesis, resulting in the lung being unexpanded from an early stage in its development . CDH affects 1–4 per thousand live births, and usually has a poor prognosis; it is fatal in about 30 % of cases.

Fetal lung hypoplasia , if severe, can lead to respiratory distress or death at birth. The hypoplastic lung has a reduced number of airway generations, fewer alveoli , an underdeveloped vascular network, and hence a reduced capacity for gas exchange. The extent to which the hypoplastic lung can recover structurally after birth is unclear, as few studies have examined the potential for catch-up lung growth following intrauterine compromise. Given the limited ability of the postnatal lung to form new alveoli, it is likely that the lungs of survivors of CDH have reduced gas exchange, which may in turn limit exercise capacity. A recent study of almost 100 survivors of CDH aged up to 3 years showed that lung function was impaired; the extent of impairment was related to the degree of lung hypoplasia at birth and the duration of mechanical ventilation [46].

One of the major long-term complications of CDH is PPH, which is thought to be a result of poorly developed vasculature in the lung , and impaired vasodilator mechanisms [47]. PPH is evident in infants and children with CDH, and is likely to persist into adulthood, potentially contributing to adult lung disease; however, current survivors of CDH are still young and therefore there is little information available on the long-term outcomes of CDH survivors.

The fetal lungs can be expanded by occlusion of the trachea , due to the continued production of fetal lung liquid by the distal lung epithelium [39]. In animal studies, tracheal occlusion has been shown to increase pulmonary luminal volume, thereby rapidly accelerating lung tissue growth and alveolarization [48]. Currently, there is an ongoing clinical trial of tracheal occlusion performed endoscopically under local anesthesia in fetuses diagnosed with CDH [49].

Fetal Exposures

Fetal exposure to an increasing number of chemicals is now known to alter lung development , with long-term consequences. These include the products of maternal tobacco smoking , alcohol ingested by the mother, antenatal corticosteroids, and pro-inflammatory cytokines. Each of these exposures has been reported to alter lung development in the fetus , with potential long-term consequences for lung function and respiratory health.

Maternal Tobacco Smoking and Nicotine

Tobacco smoking during pregnancy has long been known to impact upon placental function, fetal lung development , and postnatal respiratory health in exposed offspring. Fetal exposure to the products of maternal smoking leads to an increased postnatal risk of pulmonary infections, wheezing, reduced lung function (low FEV1), and asthma [50–52]. The products of maternal smoking that are the most important causal factors in altering fetal lung development are likely to be nicotine and carbon monoxide, both of which readily cross the placenta, although many other chemicals may also play a role. Numerous experimental studies have been performed to identify the mechanisms by which maternal smoking alters lung development: these have largely focused on the role of whole smoke or nicotine administered to the mother [53]. Such studies have shown that fetal nicotine exposure permanently impairs alveolarization and reduces the ability of exposed offspring to protect themselves against environmental stressors. Nicotine has been shown to be genotoxic, to increase the production of reactive oxygen species (ROS), and to reduce the antioxidant capacity of the lung [54]. Thus, the fetal lungs develop in an environment of oxidant–antioxidant imbalance, with concomitant adverse effects of oxidants and nicotine on cell integrity [54]. As a result of fetal disturbances to lung development, exposed offspring are more prone to develop respiratory disease s such as asthma and emphysema later in life [53, 54]. Recent studies have shown that retinoic signalling in the fetal lung may be perturbed by maternal smoking [50], and that collagen deposition may be increased via upregulated transforming growth factor (TGF)-β signalling [55]. It is now considered likely that maternal smoking is capable of inducing alterations in the fetal epigenome that could result in life-long alterations in lung function [56].

Maternal Alcohol Consumption

Recent studies suggest that a substantial number of women consume alcohol during pregnancy, with 30 % of women in the USA reporting consumption of alcohol at some stage during their pregnancy and 8.3 % reporting consumption of four or more alcoholic drinks on one occasion at least once during pregnancy [57]. This is cause for concern, as experimental studies indicate that ethanol readily crosses the placenta [58]. Indeed, chronic maternal alcohol consumption can lead to a spectrum of developmental defects in the offspring, collectively termed fetal alcohol spectrum disorders (FASD). The incidence of FASD is estimated to be 1 in 100 live births, although the prevalence could be as high as 2–5 % [59, 60]. Fetal alcohol syndrome is the most severe form of FASD and is characterized by growth restriction, craniofacial dysmorphology, and central nervous system dysfunction. Although most research into the effects of fetal ethanol exposure has been focused on the developing brain, it has also been associated with congenital abnormalities of the respiratory tract [61] and an increased susceptibility to respiratory infection s in childhood [62]. In animal models, exposure to high levels of ethanol throughout gestation or at mid-gestation causes reduced fetal lung growth [63, 64]; however, moderate exposure in late gestation has no effect [65]. Despite a lack of effect on lung growth or alveolarization , moderate ethanol exposure in late gestation increases collagen deposition in the ovine fetal lung, decreases surfactant protein (SP) levels, and alters surfactant phospholipid composition [65–67]; such changes could compromise lung function after birth. Evidence also suggests that prenatal ethanol exposure may reduce the ability of the lungs to clear pathogens, as indicated by decreases in ciliary beat frequency [66] and pro-inflammatory cytokine gene expression [65], an increase in apoptosis of alveolar macrophages and a decrease in alveolar macrophage function both under normal conditions and in response to infection [68–70]. A recent study of prepubescent sheep exposed to ethanol in late gestation, however, reported no significant differences in lung growth, lung architecture, pro-inflammatory cytokine gene expression, and surfactant phospholipid composition, although SP-D mRNA levels were increased [67]. As ethanol exposure did not continue to birth or after birth, this study suggests that the developing lung has some capacity to undergo repair in the absence of further exposure. As SP-D has immunoregulatory roles and roles in surfactant phospholipid homeostasis [39], the observed increase in SP-D mRNA levels following ethanol exposure suggests that innate pulmonary immunity and lung function may be persistently altered. Further studies investigating the long-term effects of fetal ethanol exposure on the lung are warranted.

Antenatal Exposure to Corticosteroids

During the last several decades, corticosteroids have been routinely administered to pregnant women who show signs of preterm delivery (8–12 % of pregnancies). These agents, primarily betamethasone and dexamethasone, are administered during preterm labor to enhance lung maturation in the preterm fetus , thereby increasing its chances of survival. The principal beneficial effect of corticosteroids for preterm survival is the stimulation of surfactant production and release by the immature lung; this occurs as a result of an increase in the number and maturity of alveolar epithelial type ll cells [71]. It has long been recognized that, as well as stimulating surfactant production, corticosteroids inhibit secondary septation and alveolarization in the immature lung when administered before or during the process of alveolarization [72]. The effects of corticosteroids during development can be long lasting in many tissues [73], depending on the timing and duration of corticosteroid exposure, and the lung is also likely to be permanently affected, potentially as a result of epigenetic modifications [74]. This persistent effect has been clearly demonstrated in rats exposed to corticosteroids during early postnatal life [75] and is thought to result from a persistent imbalance between cell proliferation and cell death in secondary septa [76]. Because such a high proportion of infants who are born preterm (i.e., birth prior to 37 completed weeks of gestation) are exposed to corticosteroids, the adverse effects of corticosteroids on structural development of the lung, and hence on later lung function , are likely to be a significant factor in the etiology of respiratory illness and reduced lung function during all stages of life.

Intrauterine Inflammation (Chorioamnionitis)

In many pregnancies, the fetus is exposed to inflammation of the surrounding membranes (chorion and amnion) as a result of bacterial infection; this is often referred to as chorioamnionitis. Intrauterine inflammation is present in up to 60 % of pregnancies that result in preterm birth , and it is now considered that inflammation plays a causal role in the etiology of preterm birth [77]. In such pregnancies, the fetus is exposed to inflammatory mediators over prolonged periods, and it is now clear that these mediators can enter the fetal lungs [78]. Since the first demonstration that intrauterine inflammation improves respiratory function in the preterm neonate [79], numerous studies have investigated the impact of intrauterine inflammation on fetal lung development . Such studies have shown that lipopolysaccharide (LPS), a component of the Gram-negative bacterial cell wall, and pro-inflammatory cytokines accelerate the maturation of alveolar epithelial type ll cells in the preterm fetus, such that they produce more surfactant ; in addition, pulmonary vascular development is enhanced [80]. It has also been reported that intrauterine inflammation inhibits pulmonary elastin deposition and alveolarization [81]; these effects may be transient, as other studies have not shown a persistent effect of intrauterine inflammation on lung morphometry [82, 83]. The long-term effects of intrauterine inflammation on the lungs have received very little attention. However, a recent study in sheep has shown that it can lead to persistent changes in gene expression of pro-inflammatory cytokines in the postnatal lung, as well as reductions in protein expression of SP-A and -C [82]; together, these changes may impair innate immunity of the postnatal lung. It is clear that more research is needed regarding the pulmonary programming effects of intrauterine inflammation.

Neonatal Causes of Lung Disease or Altered Lung Development

Early postnatal environments that interfere with normal lung development and maturation may result in persistent alterations in lung structure and function, increasing the risk for respiratory disease in adult life. The first two years of life are an important time for lung development , and many studies have shown that a large proportion of adult respiratory diseases stem from events that occur during infancy and early childhood. In the sections below, we provide an overview of some of the factors in the neonatal period that can alter lung development and may be key contributors to adult lung disease.

Preterm Birth

Preterm birth, defined as birth prior to 37 completed weeks of gestation, affects approximately 11 % of pregnancies worldwide [84]. Preterm birth can be categorized into very preterm birth (<32 weeks’ gestation) and extremely preterm birth (<28 weeks’ gestation). As lung development occurs throughout gestation, and critical stages of lung development occur late in gestation, babies born preterm are born with structurally and functionally immature lungs. Very preterm infants are born during the saccular stage of lung development, which is prior to the formation of definitive alveoli . Extremely preterm babies can be born during the canalicular (pre-saccular) stage of lung development and therefore are born with a thick lung interstitium and prior to the production of surfactant . Owing to lung immaturity, infants born more than 6–8 weeks before term usually experience respiratory insufficiency at birth, referred to as the respiratory distress syndrome (RDS). Therefore, infants who are born very preterm or extremely preterm usually require respiratory support to ensure their survival; the degree of support and the length of stay in hospital are proportional to the degree of prematurity [85]. Respiratory support of preterm infants can involve supplemental oxygen therapy and/or assisted ventilation, including continuous positive airway pressure (CPAP), nasal intermittent positive pressure ventilation (nIPPV), or mechanical ventilation via an endotracheal tube. Improvements in the respiratory care of preterm infants, especially the introduction of exogenous surfactant therapy, more gentle ventilation strategies and more cautious use of supplemental oxygen, have resulted in an increase in their survival, in particular those who are born at earlier gestational ages. Despite the necessity of this respiratory intervention, the incidence of chronic lung disease or bronchopulmonary dysplasia (BPD) in this population remains high [86]. BPD is characterized by disrupted lung development, in particular alveolar simplification (fewer and larger alveoli), variable interstitial fibrosis , and increased airway smooth muscle (see Fig. 5.1) [87, 88].

Fig. 5.1
figure 1

Histological representation of the alveolar region of a lung from (a; ×25 magnification) a term infant, and (b; ×50 magnification) a surfactant -treated preterm infant with BPD. Photomicrographs of a preterm infant with “healed” BPD, demonstrating (c; ×30 magnification) diffuse alveolar septal fibrosis (bottom) with enlarged alveoli (top), and (d; ×100 magnification) hyperplastic smooth muscle (arrow) surrounding a bronchiole (modified from [196])

Full size image

Although BPD is a multifactorial disease, major factors in its etiology are thought to be inflammation , oxidative stress , and inappropriate repair mechanisms due to prolonged exposure to high oxygen concentrations (hyperoxia) during supplemental oxygen therapy or mechanical ventilation [89–91]. Mechanical ventilation can contribute to lung injury by inducing inflammation as a result of repetitive tissue stretch (shear stress) and over-inflation (barotrauma) of incompliant lung tissue; particularly damaging is repetitive opening of collapsed alveoli in atelectatic regions of the surfactant deficient lung [92]. Numerous other fetal and neonatal factors associated with preterm birth can contribute to the pathogenesis of BPD or alterations in lung development; such factors include the severity of prematurity, exposure to antenatal or postnatal corticosteroids, prenatal infection or inflammation, postnatal infection, intrauterine growth restriction, and inadequate postnatal nutrition (see Fig. 5.2) [91, 93]. Regardless of etiology, it is now apparent that persistent changes within the lungs of preterm infants, especially those who developed BPD, lead to an increased risk of poor lung function in childhood and late adolescence, as indicated by reductions in FEV1, forced mid-expiratory flow (FEF25–75 %), peak expiratory flow (PEF), transfer across the lung of carbon monoxide (TLCO), and increased airway resistance [94–99]. Individuals who were born preterm also have an increased risk of reduced exercise capacity [97, 98], asthma [94, 99–101], COPD [102], and respiratory infection s [103–105] later in life. The contribution of hyperoxia, mechanical ventilation, and corticosteroids to altered lung development, and later lung disease following preterm birth will be discussed in more detail in the following sections.

Fig. 5.2
figure 2

Factors that are likely to contribute to respiratory distress syndrome (RDS) in infants who are born preterm, and factors that may lead to the subsequent development of chronic lung disease (CLD) [93]

Full size image

Hyperoxia

Preterm infants are often administered hyperoxic gas to increase blood oxygen levels. Oxygen is added to inhaled gas to reach a target arterial oxygen saturation (SaO2) of 85–99 %, with the aim of preventing tissue hypoxia. In the past, the use of high oxygen concentrations (up to 100 % O2) was common, although concentrations were reduced once it became apparent that high concentrations could lead to tissue injury [106–108]. Currently, preterm babies are normally resuscitated using room air (21 % O2) and then the fraction of inspired oxygen (FiO2) is increased until adequate blood oxygenation levels are achieved [109]. However, as fetuses normally have an arterial partial pressure of oxygen (PaO2) of approximately 25–30 mmHg and a SaO2 of 70–80 %, exposure of an immature lung with an immature antioxidant system to even low supplemental oxygen concentrations may be detrimental. It is difficult to determine the contribution of neonatal hyperoxia alone to the long-term pulmonary outcomes of preterm birth and BPD. However, oxygen supplementation during the neonatal period in infants born very preterm has been identified as an independent risk factor for asthma in childhood [101]; further, prolonged oxygen therapy is associated with low FEV1 and the use of inhaled bronchodilators in very-low-birth-weight (preterm) children [96], suggesting an impact on the conducting airways .

Numerous experimental studies have shown that neonatal inhalation of hyperoxic gas can persistently alter the development of alveoli , conducting airways and pulmonary vasculature. Specifically, neonatal hyperoxia leads to fewer and larger alveoli (reviewed in [110]), altered alveolar epithelial cell differentiation [111], altered cellular composition of the bronchiolar epithelium [112–114], altered surfactant composition [111, 115], increased airway smooth muscle [116–119], decreased bronchiolar collagen [112], decreased pulmonary capillary density [120], and remodelling of the lung extracellular matrix [111, 121] and small pulmonary arteries [120]. Consequently, neonatal hyperoxia is associated with later reductions in lung function , including altered airway resistance and reactivity, dynamic lung compliance, tissue elastance and tissue damping, and pulmonary hypertension, as evidenced by right ventricular hypertrophy [111, 112, 120, 122]. Importantly, some of the effects of neonatal hyperoxia on the lung are long-lasting, with alterations in lung structure and function observed in early and advanced adulthood [111, 112, 118, 122–125]. Interestingly, recent studies and reviews have indicated that the oxygen concentration used and the duration of oxygen exposure are important determinants of lung injury [110, 111, 119].

Lung injury and alterations to lung structure that are induced by neonatal exposure to hyperoxia are thought to be due to oxidative stress and a subsequent pro-inflammatory response (see Fig. 5.3) (reviewed in [110]). Experimentally, neonatal treatments aimed at minimizing ROS and exploiting antioxidant defense, as well as treatments that target inflammatory cell influx into the lung and reduce the degree of lung inflammation , have shown promise in improving lung structure following neonatal hyperoxia [110]. Due to the importance of pulmonary vascular development in regulating alveolarization , treatments promoting angiogenesis following neonatal hyperoxia have also received much attention [110]. Recently, the use of stem/progenitor cells to repair or prevent lung injury has also been trialled in experimental models, and clinical trials for their use in preterm infants are currently underway, as discussed below.

Fig. 5.3
figure 3

Potential pathways whereby exposure of the developing lung to hyperoxia leads to alterations in the structure and function of the airways, pulmonary vasculature and alveoli , thus contributing to impaired lung function later in life. ASM airway smooth muscle , VEGF vascular endothelial growth factor [110]

Full size image

Mechanical Ventilation

Owing to lung immaturity, preterm infants usually require assisted ventilation. Different ventilation strategies are employed depending on the requirement of the infant; these range from noninvasive strategies, including those involving constant airway pressure (i.e., CPAP using a constant or variable flow, or nasal cannula with low or high flow) or variable airway pressure (i.e., nIPPV, nasal positive airway pressure or nasal high-frequency ventilation), to invasive positive pressure ventilation requiring endotracheal intubation, including patient-triggered ventilation, volume target ventilation, and high-frequency oscillation [126]. It is now recognized that invasive ventilation strategies can increase the incidence of BPD and lung injury, which can increase the risk for persistent lung disease in infants born preterm. Furthermore, intubation is associated with trauma at the site of intubation (i.e., larynx and upper trachea ), as well as bacterial colonization, and sepsis [126]. Noninvasive ventilation strategies have been shown to reduce these adverse effects associated with mechanical ventilation, improve gas exchange, and reduce the incidence of BPD [127]. Thus, not only are noninvasive ventilation strategies now used for weaning infants off mechanical ventilation to reduce extubation failure, they are increasingly being used as primary methods of respiratory support for preterm infants with respiratory insufficiency. However, even noninvasive techniques have been associated with adverse effects, including pneumothorax, pulmonary over-inflation, increased work of breathing, decreased cardiac output, as well as nasal complications and gastrointestinal distension and perforation [126]. Large randomized controlled trials are required to determine the beneficial effects of different modes of noninvasive ventilation on long-term respiratory and non-respiratory outcomes.

Ventilator-induced lung injury is believed to be a result of either volutrauma or barotrauma, which is injury related to over-expansion of the lungs (due to high volumes or pressures), atelectrauma, which is injury caused by repeated opening of collapsed alveoli , or biotrauma, which is injury caused by up-regulation of pro-inflammatory responses to mechanical ventilation, or a combination of these factors [126]. In the clinical setting, it is difficult to determine the precise cause of lung injury in preterm infants due to the multifactorial and individualized nature of their care, both before and after birth. Experimentally, it has been shown in rodents and sheep that even brief mechanical ventilation (2–24 h) of the developing lung alone (i.e., without other confounding factors) leads to impaired formation of alveoli and lung microvessels, as indicated by decreases in secondary crest density, alveolar number, and CD31 staining, and increases in airspace and alveolar size; these changes in lung structure are likely due to disruption of elastin deposition in the lung [92, 128–130] and altered expression of other genes/proteins that regulate alveolar formation and angiogenesis, including platelet derived growth factor-A (PDGF-A), tenascin-C, and vascular endothelial growth factor receptor 2 (VEGF-R2) (see Fig. 5.4) [92, 131]. Studies have also identified numerous transcription factors and genes that are dysregulated by mechanical ventilation, contributing to lung injury and altered lung development ; likely factors include those involved in TGF-β signalling, of which elastin is a target, mechanotransduction and the immune response [132–134]. Furthermore, recent studies have shown that invasive mechanical ventilation in preterm lambs causes an increase in histone deacetylatase (HDAC) 1 and subsequent genome-wide histone hypoacetylation in the lung, indicating alterations in the epigenetic regulation of gene expression [135]. Insulin-like growth factor-1 (IGF-1) is one gene whose expression is increased in mechanically ventilated preterm lambs and in infants who have died from RDS or BPD, and which appears to be epigenetically modified by mechanical ventilation [135]. Further studies are required to identify other genes that are modified by mechanical ventilation and the specific pathways leading to their alteration [93]. A greater understanding of the mechanisms by which mechanical ventilation alters gene expression and leads to altered lung development will enable targeted interventions to be developed. In this regard, histone acetylation is increased and alveolar formation is improved when preterm lambs that are mechanically ventilated are treated with HDAC inhibitors [135].

Fig. 5.4
figure 4

Potential pathways whereby prolonged cyclic stretch of the lung caused by positive-pressure mechanical ventilation alters the formation of the pulmonary vasculature and alveoli . TGF-β transforming growth factor-β, VEGF-R2 vascular endothelial growth factor receptor 2 [92]

Full size image

In addition to altering alveolar structure, a short period (2–12 h) of positive pressure ventilation of the immature (prenatal) sheep lung induces injury and remodelling of the small conducting airways (i.e., bronchioles), including epithelial cell detachment, alterations in the areas of epithelium and collagen in intact airways, and an increase in airway smooth muscle [129, 130, 136]. There is evidence that airway remodelling persists in the absence of further ventilation, with changes in epithelial area, epithelial cell proliferation and apoptosis, airway smooth muscle and collagen, and a decrease in the number of bronchiolar-alveolar attachments [129, 130, 136]. There is also evidence, however, that the fetal lung is capable of spontaneous repair following ventilation-induced injury, with normalization of alveolar structure in the absence of continued ventilation [129, 130], although this recovery could potentially be restricted to the fetal lung. Changes in alveolarization and airway wall structure following mechanical ventilation could contribute to the increased risk of poor lung function and obstructive lung disease reported in individuals who were born preterm. Indeed, airway reactivity is increased 48 h after brief mechanical ventilation of rat pups [137]. Further studies investigating the long-term effects of mechanical ventilation of the immature lung on respiratory health are needed.

Corticosteroids

Corticosteroids have been used in preterm infants with evolving or established BPD to reduce lung inflammation and improve gas exchange; corticosteroids are used to reduce the need for oxygen supplementation [91] and to decrease the risk of retinopathy of prematurity and death [138]. Postnatal corticosteroids, predominantly dexamethasone, are generally administered systemically and can be given early (≤7 days after birth) or late (>7 days). Although systemic corticosteroid therapy can have short-term benefits, it has been linked to adverse long-term outcomes, including detrimental effects on lung structure, neurodevelopment, cardiovascular function, and the gastrointestinal system [91, 138]. Although many of these effects have been linked to the prolonged use of high-dose corticosteroids (especially dexamethasone), studies have indicated that even low-dose corticosteroids may have adverse effects [91]. Inhaled corticosteroids have been trialled in preterm infants with the aim of reducing the side effects associated with systemic corticosteroid use, although these have had limited success [91, 138]. Studies investigating the optimal timing, dose, and type of corticosteroids given to preterm infants to improve short-term outcomes and reduce adverse long-term outcomes are warranted.

Nutrition

Adequate nutrition plays a key role in normal lung development . Experimentally, reduced nutrition during lung development before and after birth can affect alveolarization and bronchiolar structure [28, 31, 139]. The effects of undernutrition during fetal life on lung development have been described above. Lung development continues after birth and it has been shown in sheep that, at maturity, animals that experienced postnatal growth restriction have a smaller lung volume and a smaller pulmonary surface area, relative to lung and body size, than animals that grow normally [140]. Undernutrition during the early postnatal period has been shown to disrupt lung development, especially when combined with exposure to hyperoxia [112, 141, 142]. This is of particular importance in the context of very preterm infants who form most of their alveoli after birth. In addition, very preterm infants have a high risk of developing BPD, and it is known that malnutrition represents one of the many factors contributing to BPD [143]. Maintaining a careful balance of nutritional support for very and extremely preterm infants represents a considerable challenge. Interventions that are necessary to support the survival of these preterm infants, combined with developmental complications they may encounter (e.g., prolonged orotracheal intubation, medications that alter gastrointestinal motility, altered palatal groove formation, irritability due to neurological status, gastrointestinal immaturity) make feeding by mouth a problematic task; therefore, nutritional support is often provided via parenteral or enteral routes [144]. However, most extremely preterm infants receive minimal enteral feeding during the first two weeks of life, due to the abovementioned problems, and their nutritional requirements need to be supplemented with parenteral feeding. Recently, it was reported that a critical minimal amount of enteral feeding is required to prevent the development of BPD [145]. Preterm infants who developed BPD received less enteral feeding during the first two weeks of life, even though it was compensated by parenteral nutrition [145]. Furthermore, it has been reported that, compared to preterm infants without BPD, those with BPD have a greater energy expenditure (up to 25 % above total caloric needs), likely due to their increased work of breathing, stress, inflammation , medications, and the perceived need for catch-up growth [143, 144].

Another aspect of nutritional management of preterm infants that has been associated with an increased risk for BPD is fluid intake. It has been reported that high fluid intake during the first days after birth may increase the risk of developing BPD, through the persistence of a patent ductus arteriosus and a higher fluid content in the pulmonary interstitial tissue; these complications could contribute to decreased lung compliance and increase the need for respiratory support [143, 146]. Overall, it is suggested that adequate nutritional management of preterm infants with evolving or established BPD can be maintained by adopting enteral nutrition as soon as possible, avoiding excess fluid administration during the first days after birth, and progressively increasing energy intake [147].

The use of specific micronutrients, such as vitamins A, C, and E, inositol, iron, selenium, glutamine, cysteine, and methionine, may be beneficial in the nutritional management of preterm infants to prevent or lessen the severity of BPD (reviewed in [144]). Indeed, optimal feeding and growth in preterm infants remains a challenge, and overnutrition of the preterm infant to achieve catch-up growth may actually be detrimental to long-term health [148]. Although accelerated growth of the preterm infant has beneficial effects on neurodevelopment, it has also been associated with many metabolic and cardiovascular implications that present in childhood and adulthood [148]. Further research is needed in order to identify the optimal method of managing preterm nutrition with the goal of enhancing lung growth, preventing BPD, and avoiding long-term, potentially harmful programming effects of catch-up growth.

Although infants born preterm are faced with a vulnerable period during which nutrition can greatly affect their lung development and subsequent lung health, nutritional status can also impact upon the mature lung. Emphysema is commonly observed in patients with anorexia nervosa and was also noted at autopsies of humans who died from starvation [149]. These findings support the concept that lung structure/function is dependent upon nutritional status.

Respiratory Infections

The link between early-life respiratory infection s, particularly viral infections, and subsequent increased risk of recurrent wheezing and asthma is being increasingly investigated. Viral respiratory infections during infancy have been shown to contribute to the development of wheezing in childhood. In particular, the occurrence of symptomatic rhinovirus infection during infancy (<12 months of age) has been identified as the most significant risk factor for the development of wheezing 1 year after infection [150] and into early childhood [151]. Infection with respiratory syncytial virus (RSV) in infancy has also been linked with wheezing in pre-school children, albeit with a weaker relationship compared to rhinovirus infections [151]. A later study confirmed these findings, showing that bronchiolitis caused by viruses other than RSV was associated with a greater risk of recurrent wheezing (within the first 2 and 3 years after hospitalization for bronchiolitis) compared to RSV-associated wheezing [152]. Not only has rhinovirus infection been associated with recurrent wheezing, but rhinovirus infection during the first 3 years of life is also significantly associated with reduced lung function (lower FEV1, percent predicted FEV1, FEV0.5, FEF25–75 %, FEV1/forced vital capacity (FVC), and FEV0.5/FVC) at 8 years of age compared to children with non-wheezing rhinovirus illnesses and no rhinovirus illnesses [153]. Furthermore, this study showed that children with RSV-induced wheezing did not have significant alterations in any of the lung function measurements, indicating that infections in early childhood caused by rhinovirus, rather than RSV, are significant predictors of decreased lung function in childhood [153]. Not only are viral respiratory infections during early postnatal life associated with subsequent wheezing, but they often precede the development of childhood asthma. In a cohort of children that was followed up to 6 years of age after specific viral respiratory wheezing illnesses in early childhood (within the first 3 years of life), rhinovirus-associated wheezing illnesses were the most significant predictors of the subsequent development of asthma [154]. This, however, does not exclude RSV as a potential risk factor for asthma. Recent studies showed that severe RSV-induced bronchiolitis during infancy (<12 months of age) is associated with an increased risk of asthma at 6 and 7 years of age [155, 156]. Furthermore, RSV infection in infancy was shown to be associated with a 3.2-fold increase in the risk of wheezing and impaired lung function (lower FEV1, FVC, and FEV1/FVC, and increased respiratory system resistance) at age 6 [155]. Another recent study reported on a study population at 15–18 years of age, and compared patients hospitalized for bronchiolitis in infancy (<24 months of age) caused by RSV or rhinovirus to a control population [157]. This study showed that patients hospitalized in infancy for bronchiolitis had an increased risk of asthma in adolescence; of the bronchiolitis patients, those who had rhinovirus infection had a higher risk of self-reported asthma, but not doctor-diagnosed asthma, than those who had RSV infection [157]. Studies that have extended further into adulthood have found that hospitalization for bronchiolitis in infancy (<24 months of age) is associated with an increased risk of asthma, and an increased use of asthma medication at 28–31 years [158]. In that study, doctor-diagnosed asthma was present in adulthood in one-third of former bronchiolitis patients, whereas asthma was present in only 10 % of controls. An earlier study of the same cohort demonstrated that the prevalence of asthma at 28–31 years was similar to that recorded 10 years earlier at 18–21 years of age [159], indicating the early-onset and persistence of long-term sequelae following early-life respiratory infection. However, these studies did not differentiate between RSV and rhinovirus bronchiolitis. The link between early-life respiratory infections and subsequent long-term respiratory morbidity has also been highlighted in a follow-up study that showed an association between RSV bronchiolitis in the first year of life and increased prevalence of asthma or recurrent wheeze at 18 years of age; 39 % of RSV subjects vs. 9 % of control subjects [160]. A reduction in spirometric airway function was reported in the RSV cohort compared to controls, regardless of whether the RSV subjects presented with or without asthma or recurrent wheeze [160].

Bacterial colonization of the hypopharynx with S. pneumonia, H. influenza, or M. catarrhalis in healthy neonates at 4 weeks of age has been associated with an increased risk of pneumonia and bronchiolitis in early life (up to 3 years) independent of asthma [161]; this suggests that an altered microbiome of the neonatal airway may affect the maturation of the immune system, leading to an increased risk of pneumonia and bronchiolitis. Furthermore, radiologically confirmed pneumonia in the first 2 years of life has been associated with asthma or asthma-like symptoms in pre-school children (4–5 years of age) [162]. This is confirmed in a meta-analysis that showed an overall risk of major long-term respiratory sequelae following childhood pneumonia (before 5 years of age) in non-hospitalized children of 5.5 %; this risk was almost three times greater in children who had been hospitalized, with adenovirus pneumonia having the highest risk (approximately 55 %) [163]. Major respiratory sequelae were defined as restrictive lung disease, obstructive lung disease, and/or bronchiectasis, and the most common type of major outcome reported was restrictive lung disease [163]. These findings indicate that children diagnosed with pneumonia should be considered at risk of long-term respiratory sequelae, even after acute symptoms have resolved.

Environmental Exposures

Many studies have highlighted the deleterious effects of a suboptimal environment on lung development (reviewed in [164]). Harmful environmental exposures include both outdoor pollutants (e.g., industrial and motor vehicle emissions) and indoor pollutants (e.g., environmental tobacco smoke (ETS) and fungal spores). In particular, ETS exposure in early life has been identified as a significant risk factor for poor pulmonary outcome in later life. A recent meta-analysis indicated that passive exposure to ETS (in both pre- and postnatal life) increases the incidence of wheeze and asthma in children and adolescents by at least 20 % [165]. Maternal smoking was associated with a 52 % increase in the risk of wheeze and a 20 % increase in the risk of asthma at 5–18 years of age [165]. Not only has ETS exposure during infancy been associated with increased risk of wheeze and asthma, but a meta-analysis has also shown it to increase the risk of lower respiratory infection s, particularly bronchiolitis [166]. All types of passive ETS exposure (pre- and postnatal maternal, paternal, both parents, any household member smoking), but particularly maternal smoking, led to a significant increase in the risk of infants developing lower respiratory infections in the first 2 years of life [166]. This meta-analysis highlights the importance of limiting ETS exposure during both gestation and childhood to prevent the onset of wheezing, asthma, and respiratory infections. In contrast to the large number of reports on effects of ETS exposure, there are few reports on the effects of exposure to traffic-related air pollution during infancy and childhood and its effects on later respiratory morbidity. A recent systematic review and meta-analysis showed that early childhood exposure to traffic-related air pollution was associated with the development of asthma throughout childhood up to 12 years of age, with the magnitude of risk increasing with age [167]. However, further studies are needed to investigate the relation between effects of exposure at specific ages and adult outcomes.

Potential Targets for Cell-Based Therapies: Neonatal Treatments for Preterm Infants

The use of cell-based therapies to prevent or repair neonatal lung injury in very preterm infants has now evolved from experimental studies using animal models to phase I and II clinical trials in very preterm infants at risk of BPD. The rationale for the current clinical trials is based on knowledge gained from studies examining the perturbation of pulmonary stem cells in neonatal lung injury, as well as experimental studies in animal models that found stem cell-based therapies to be beneficial. Studies in animal models and in humans have shown that altered function or depletion of resident stem cells within the developing lung likely contributes to the pathogenesis of BPD. In rodent models of BPD, in which newborn pups are exposed to high concentrations of oxygen, lung epithelial cells, mesenchymal stromal cells (MSCs), and endothelial progenitor cells (EPCs) were shown to be significantly reduced in number and also displayed altered differentiation potential [123, 168–170]. Similar perturbations to resident stem cells have been reported in tracheal aspirates and umbilical cord blood samples taken from preterm infants [171–175]. The presence of MSCs in tracheal aspirates has been reported as an indicator of BPD development , and these MSCs were shown to demonstrate a lung-specific gene expression and to secrete pro-inflammatory cytokines; this suggests that the MSCs in the tracheal aspirates originated from the lungs and have the capacity to contribute to lung injury [171, 172]. In samples of umbilical cord blood from preterm infants, endothelial colony forming cells (ECFCs; a subset of EPCs) were shown to have an increased susceptibility to in vitro exposure to hyperoxic gas, and were found in reduced numbers among BPD infants compared to non-BPD preterm infants [173–175]. This suggests that a reduction in the levels of angiogenic progenitor cells may contribute to disrupted vasculogenic mechanisms during lung development and lead to impaired vascular and alveolar growth. In response to the findings of stem cell perturbation in BPD, experimental studies have focused on their therapeutic replacement with exogenous-derived stem cells (reviewed in [176]). One of the most commonly researched cell types is the MSC, which can be easily sourced from bone marrow, adipose tissue, muscle, peripheral blood, umbilical cord blood, umbilical cord Wharton’s jelly, and placenta. Numerous components of neonatal lung injury have been attenuated by the administration of MSCs, including inflammation , vascular damage, impaired alveolarization , and fibrosis [170, 177–188]. Furthermore, MSCs have been reported to have low engraftment and differentiation into the lung, suggesting a paracrine-mediated therapeutic effect. This has been supported by studies showing similar protective properties of cell-free conditioned media from MSCs in experimental models of neonatal lung injury. Other cell types (ECFCs, angiogenic cells, and amnion epithelial cells) have also been examined in experimental studies and have shown promise in their therapeutic benefit for neonatal lung injury [189–194].

The overall aim of experimental studies on neonatal lung regeneration has been to generate evidence to create a rationale for translating experimental findings into the clinic. Encouragingly, there has recently been translation of findings from animal studies into clinical trials in preterm infants, with the results of the first phase I clinical trial (NCT01297205) of stem cell therapy in preterm infants being published [195]. The authors assessed the safety and feasibility of human umbilical cord blood-derived MSCs, administered to nine very preterm infants who were at risk for developing BPD. The infants were born at 24–26 weeks of gestation (630–1030 g) and needed continuous ventilatory support that could not be decreased due to significant respiratory distress prior to enrolment into the study. Three patients were given a low dose of MSCs (1 × 107 cells/kg) via the trachea , and six were given a high dose (2 × 107 cells/kg). Treatment-related serious adverse effects were assessed for 84 days following transplantation of the MSCs. The authors reported no serious adverse effects or dose-limiting toxicity associated with transplantation of the MSCs. There was no significant difference in serious adverse effects between the low-dose and high-dose MSC group or between the MSC group and a matched-comparison group, except in BPD severity; the MSC group had significantly lower BPD severity than the matched-comparison group. Furthermore, the inflammatory markers matrix metalloproteinase (MMP)-9, interleukin (IL)-6, IL-8, tumor necrosis factor (TNF)-α, and TGF-β were reduced in tracheal aspirate fluid after MSC transplantation. The findings of this study are important in establishing the feasibility, safety, and efficacy of MSC treatment for BPD in preterm infants. Follow-up studies are currently underway to assess the long-term safety of MSC transplantation in the same preterm infants reported in the phase I clinical trial, at approximately 2 years of age (NCT01632475) and 5 years of age (NCT02023788). Furthermore, a randomized, double-blind, multicenter, phase II clinical trial (NCT0182957) has commenced that will investigate the efficacy and safety of a low dose of MSCs versus a control group for the treatment of BPD in preterm infants; this trial will be followed up with another study (NCT01897987) investigating the long-term safety and efficacy of the phase II trial. It is anticipated that findings from these current clinical trials will pave the way for the future treatment of neonatal lung diseases using stem cell-based therapies.

Conclusions

The lung develops during prenatal and early postnatal life, and is therefore vulnerable to adverse environmental conditions during the embryonic, fetal, and postnatal stages of lung development . The peripheral lung, including alveoli , the pulmonary microcirculation, and small conducting airways , is especially vulnerable to injury or altered development . Owing to the lung’s limited capacity for repair , lung injury, or altered lung development caused by a suboptimal fetal or early postnatal environment can persist into later life, adversely affecting adult lung function and increasing the risk of respiratory illness. Changes in lung structure can be due to a direct developmental effect or to changes in gene and protein expression caused by alterations in the nutritional, physical/mechanical, oxidative, or immunological environments. In addition, a range of environmental factors may induce epigenetic alterations in key genes that regulate lung development to cause persistent alterations in lung structure and/or function. Although there are currently no treatments capable of repairing the injured lung, there is now encouraging experimental evidence on the effectiveness of cell-based therapies to prevent or repair injury to the immature lungs, with clinical trials now underway in infants who are born very preterm and are at risk of developing chronic lung disease.