Bronchopulmonary Dysplasia: The Role of Oxidative Stress

The developments in molecular medicine are transforming respiratory medicine. Leading clinicians and scientists in the world have brought their knowledge and experience in their contributions to this book. Clinicians and researchers will learn about the most recent advances in a variety of lung diseases that will better enable them to understand respiratory disorders. This treatise presents state of the art essays on airways disease, neoplastic diseases, and pediatric respiratory conditions. Additionally, aspects of immune regulation, respiratory infections, acute lung injury/ARDS, pulmonary edema, functional evaluation in respiratory disorders, and a variety of other conditions are also discussed. The book will be invaluable to clinicians who keep up with the current concepts, improve their diagnostic skills, and understand potential new therapeutic applications in lung diseases, while scientists can contemplate a plethora of new research avenues for exploration.

practices regarding oxygen-saturation targets. Indeed, published literature cites acceptable saturation ranges from 84% to 98% (Garg et al., 1988;Moyer-Mileur et al., 1996;Sekar & Duke, 1991;Walsh, 2003;Zanardo et al., 1995). Accordingly, the physiological definition determined BPD at 36 weeks of correct age as follows: 1) In all infants treated with mechanical ventilation, continuous positive airway pressure, or supplemental oxygen at FiO 2 > 0.30, without additional testing; 2) If the FiO 2 < 0.30, infants are to be gradually weaned to room air, in a timed stepwise fashion; those who cannot maintain an SaO 2 ≥ 88% are diagnosed with BPD, unless they pass a timed, continuously monitored oxygen reduction test. An oxygen saturation 80% to 87% for 5 minutes, or < 80% for 1 minute, indicates BPD. If all SaO 2 measurements over 15 minutes ≥ 96%, or if instead, all SaO 2 measurements in a 60-minute period > 88%, the infant is deemed not to have BPD (Walsh et al., 2003). To evaluate the impact of the physiological definition on BPD rates, 1598 consecutively born preterm infants (birth weight 501-1249 g) in hospital at 36 weeks' PMA were prospectively assessed and assigned an outcome using both the clinical and physiological definitions of BPD. The NICHD neonatal network centers demonstrated that many babies who, according to the nursing staff, required oxygen were able to maintain an SaO 2 > 90% on room air. Though 560 (35%) had clinical BPD (oxygen use at 36 weeks), only 398 (25%) had physiological BPD (as defined above) (Walsh et al., 2004).

Structural lung changes
As described by Northway et al. (1967), the histological features of classical BPD included prominent interstitial fibrosis, alveolar overdistention alternating with regions of atelectasis, and airway abnormalities such as squamous metaplasia and excessive muscularization. On the other hand, the "new" BPD shows histological features consistent with developmental arrest and impaired alveolar development (Husain et al., 1998): alveoli are fewer in number and larger in diameter than normal; the fibrosis, squamous metaplasia and excessive airway muscularization seen in classical BPD are conspicuously absent; airway and microvascular growth are affected. A short comparative study by Bhatt et al. (2001) found decreased levels of vascular endothelial growth factor (VEGF) and angiogenic receptors Flt-1 and Tie-2 in infants who died from BPD vs. from other causes. The authors concluded that the lungs from infants with BPD showed abnormal development of alveolar microvessels (abnormal placement in the alveolar septa) and that the capillaries were frequently dilated, changes attributable to low VEGF and associated receptors (Bhatt et al., 2000(Bhatt et al., , 2001. Controls were five children born at term who died at a mean of 3.4  1.3 days, whereas the five BPD subjects were born at 27  2 weeks' gestation, received FiO 2 > 0.5 during 37  33 days, and died at 65  34 days.

Epidemiology
BPD remains the most prevalent and one of the most serious long-term sequelae of preterm birth (Fanaroff et al., 2007). There is considerable variation in reported rates, however, depending upon the centre. Among 4213 infants born in 2003 at 24-31 weeks' gestation in 10 different European regions, the rate of BPD (oxygen requirement at 36 weeks' PMA) was anywhere from 10.5% to 21.5% (Zeitlin et al., 2008). A 2010 NICHD Neonatal Research Network report on neonatal outcomes of extremely preterm infants assessed 9575 infants born at extremely low gestational ages (22-28 weeks) and very low birth weights (401-1500 g) at network centers between January 1, December 31, 2007. Including babies with mild BPD (oxygen therapy for 28 days but use of room air at 36 weeks), the incidence of BPD as determined by the severity-based definition was 68%; traditional definition, 42%; physiologic definition, 40% (Stoll et al., 2010).

The preterm lung: Set-up for injury
Human lung development proceeds in five regulated stages: embryonic (3-7 weeks' gestation), pseudoglandular (7-17 weeks'), canalicular (17-27 weeks'), saccular (28-36 weeks') and alveolar and microvascular maturation (36 weeks' gestation to at least 2 years after birth). The lungs of preterm infants born at 24-28 weeks' gestation are in the late canalicular or early saccular stages and therefore cannot support efficient gas exchange. Branching and expansion of air spaces to form saccules and thinning of mesenchyme occur later in gestation, as do the formation of alveoli and the synthesis of surfactant by type II alveolar cells which only commence in late gestation. Any injury to the lung at the early stages of development can potentially alter the developmental process, leading to long-term pulmonary sequelae (Chakraborty et al., 2010). Whereas fetal development is predicated on a hypoxic environment, at birth the oxidative load is sharply increased. At the same time, oxygen demands increase abruptly. The baby born at term easily adapts to this transition in most cases but for the preterm infant, the intra-to extra-uterine transition is not without risks. Among the reasons why the preterm infant is more likely to experience oxidative injury than more mature newborns and older children are the following: 1) intracellular defences against oxidative stress are still poorly d e v e l o p e d ; 2 ) t h e p r e t e r m i n f a n t i s o f t e n , f o r v a r i o u s r e a s o n s , e x p o s e d t o h i g h concentrations of supplemental oxygen; and 3) the fetus and premature infant are also susceptible to inflammation and infection that may lead to increased oxidative stress (Saugstad, 2010). It may therefore be instructive to look at some of the molecules implicated in oxidative stress, while drawing parallels with the corresponding processes in BPD. This added insight will contribute to delimiting specific sources of oxidant molecules that may contribute to the development of BPD, a topic we will explore later in the chapter in relation to BPD risk factors.

Oxidative stress
In utero, the arterial pressure of oxygen (PaO 2 ) is close to 30 mm Hg. After birth, with the baby breathing in ambient air, the PaO 2 rises to 75 mm Hg. This greater oxygen load increases the concentration of dissolved oxygen available for oxidative phosphorylation in the mitochondria, organelles that release 1-3% of oxygen in the form of reactive oxygen species (ROS). Inspired oxygen (O 2 ) is a diatomic molecule with two free electrons (O-O). This molecule has the highest half-cell reduction potential (E hc ) in vivo (E hc for the ½O 2 /H 2 O couple = 0.816 V). Consequently, dissolved O 2 readily accepts an electron () from donors such as polyunsaturated fatty acids or ascorbic acid, generating the free radical superoxide anion (O-O or O 2  − ). This transformation of O 2 into O 2  − is spontaneous, generating the oxidized form of vitamin C (dehydroascorbate, DHA) and/or the by-products of fatty acid oxidation (lipid peroxides, aldehydes such as malondialdehyde or 4-hydroxy-2-nonenal (HNE), or isoprostanes). The reaction may also be catalyzed by nicotinamide adenine dinucleotide phosphate (NADPH) oxidase. Thus, the inspiration of diatomic oxygen leads to an increase in the cellular concentration of free radicals (O 2  − ) as well as free O 2 , which will contribute to metabolic regulation by hydroxylation of several biologically active molecules. For instance, O 2 is essential for the degradation of hypoxia-inducible factor-1 (HIF-1; HIF-1 activates transcription of the gene encoding VEGF, an important growth factor for angiogenesis. This process is impaired in BPD (Husain et al., 1998;Bhatt et al., 2000Bhatt et al., , 2001. Figure 1 shows a number of oxidative-reduction (redox) reactions of interest.

Superoxide anion
The dismutation of the superoxide anion (O 2  − ) into O 2 and H 2 O 2 (2 O-O + 2H +  O-O + HO-OH) may be either spontaneous or catalyzed by a superoxide dismutase (SOD). In preterm infants, the pulmonary activity of SOD is suspected to be immature. As reported by Lee Frank in several animal species (mice, hamster, rat, guinea pig), the pulmonary activity of SOD, catalase, and glutathione peroxidase are only 10-15% of that in term babies, in preterm newborns < 32 weeks of human-equivalent gestation (Frank & Sosenko, 1987a, 1987bFrank, 1991). As a result, the levels of O 2  − may be higher in preterm than term neonates. Furthermore, the oxidant property of O 2  − is not related to the attraction of an electron from a common antioxidant such as ascorbate, but to the donation of an electron to a free transition metal such as ferric iron (Fe 3+ ) in a Haber-Weiss reaction (O 2  − + Fe 3+  O 2 + Fe 2+ ). The resulting ferrous ion (Fe 2+ ) from this reaction reacts rapidly with hydrogen peroxide (H 2 O 2 ) in a Fenton reaction to generate Fe 3+ , OH − and OH. This hydroxyl radical (OH) is among the most reactive of molecules, leading to the oxidation of proteins, lipids and DNAs. Therefore, high oxygen supplementation coupled with low SOD activity add to oxidative stress, and this may be evidenced by an increase in the by-products of lipid peroxidation (lipid peroxides, malondialdehyde, HNE, alkanes such as ethane and pentane, and isoprostanes) and/or of protein oxidation (carbonyl compounds, o-dityrosine). Newborn infants receiving O 2 supplementation have demonstrably elevated levels of markers of oxidative stress such as exhaled ethane and pentane (Nycyk et al., 1998;Pitkanen et al., 1990), serum HNE (Ogihara et al., 1999), F 2a -isoprostanes in tracheal aspirate (Cotton et al., 1996) or in plasma (Ahola et al., 2004), protein-carbonyl in bronchoalveolar fluid (Gladstone & Levine et al., 1994) or o-dityrosine in urine (Kelly & Lubec, 1995;Lubec et al., 1997). It has been suggested that some of these markers may be higher in the first few days of life in preterm infants who will develop BPD as compared to those who will not (Gladstone & Levine et al., 1994;Hodgman et al., 1969). Hence, reducing the O 2  − levels in preterm neonates has been a seductive approach to BPD prevention. Indeed, a randomized study of human recombinant SOD administered intratracheally in the first 24 hours to preterm infants at high risk (birth weight 600-1200g) has been associated with a lower incidence of respiratory illnesses such as wheezing, asthma and pulmonary infections (Davis et al., 1993;Davis et al., 2003).

Hydrogen peroxide
As noted in Figure 1, H 2 O 2 is generated following high oxygen supplementation. Chemically, H 2 O 2 is a relatively stable molecule that can diffuse passively through cell membranes. Its oxidation reactions occur in two ways, one by accepting an electron from ferrous iron (Fe 2+ ) to generate the free radical hydroxyl (OH), the other by oxidizing sulfhydryl or thiol groups (R-SH) on protein. By its high affinity for thiol, H 2 O 2 is considered an important player in the regulation of several metabolic pathways (Winterbourn & Hampton, 2008). Of interest to BPD, H 2 O 2 can activate nuclear factor kappa B (NF-kB) (Flohé et al., 1997;Haddad, 2002;Takada et al., 2003), upregulating the transcription of genes encoding proinflammatory cytokines (Randell et al., 1990). H 2 O 2 also contributes to the stability of HIF-1 (Bonello et al., 2007;Chen Y Shi, 2008;Haddad, 2002;López-Lázaro, 2006;Simon, 2006), a transcription factor involved in angiogenesis. It is therefore important that the intracellular level of H 2 O 2 be tightly regulated. The intracellular concentration of H 2 O 2 depends on the balance between production from the dismutation of superoxide anions catalyzed by manganese superoxide dismutase (MnSOD) (Buettner et al., 2006), and detoxification by catalase and/or glutathione peroxidase. Catalase has a high catalytic activity but relative low affinity for H 2 O 2 (K m of 1.1 M) (Jones & Suggett, 1968), whereas glutathione peroxidase has a K m close to 1 M (Flohéa & Branda, 1969). With the exception of erythrocytes (Gaetani et al., 1996), catalase is present in peroxisomes and mitochondria. Glutathione peroxidase, however, is present in the cytosol, where it is an efficient regulator of the intracellular level of H 2 O 2 . Reduction of H 2 O 2 by glutathione peroxidase implies a conversion of glutathione (GSH) to its disulfide form (GSSG). The cell exerts tight control over the intracellular concentrations of GSH and GSSG in order to maintain the appropriate redox environment for the various cellular processes to occur efficiently. Indeed, the redox potential is a component of the Gibbs free energy equation that predicts the feasibility of a chemical reaction. Several biochemical pathways are dependent on the intracellular redox potential, including NF-B activation and HIF-1 levels as discussed earlier (Bonello et al., 2007;Chen & Shi, 2008;Land & Wilson, 2005;López-Lázaro, 2006;Roy et al., 2008). In the presence of a large peroxide load or sustained generation of peroxides, the formation of GSSG can exceed the capacity of glutathione reductase to recycle it into GSH, and the redox potential will change to a more oxidized state. Oxygen (O 2 ) supplementation as well as hydrogen peroxide (H 2 O 2 ) from parenteral nutrition can lead to modulation of: 1) transcription factors such as hypoxia-inducible factor-1 (HIF-1) and nuclear factor kappa B (NF-B), important players in the pathogenesis of BPD; 2) levels of oxidative stress markers (isoprostanes, peroxides, lipid aldehydes, alkanes); 3) activity of thiol-sensitive proteins (R-SH); and 4) redox potential of glutathione (GSH), as influenced by glutathione peroxidase (GPx) and the intracellular concentrations of reduced (GSH) and oxidized (GSSG) glutathione.

Redox potential of glutathione
The redox potential is dependent of the concentration of GSH and GSSG according to the Nernst equation: ∆E =∆E°· (RT/nF)· log ([GSH] 2 /[GSSG]) (Schafer & Buettner, 2001). In cells extracted from the endotracheal aspirate of intubated newborns, the level of glutathione increases with gestational age and female sex, being lower in preterm and male infants . The sex is a significant risk factor for BPD, as BPD affects more boys than girls (Ambalavanan et al., 2008;Ambalavanan & Novak, 2003;Darlow & Horwood, 1992;Kraybill et al., 1989;Stroustrup & Trasande, 2010). The low glutathione concentration measured in preterm newborns  is associated with an oxidized redox potential. Low blood level of glutathione were also reported in preterm neonates with chronic lung disease (White et al., 1994). Recently, Chessex et al. (2010) demonstrated a correlation between BPD severity in preterm infants (26  1 weeks' gestation) and blood redox potential measured one week after birth: the more oxidized the redox potential, more severe the disease. As previously reported (Schafer & Buettner, 2001), the redox potential acts as a switch for a number of metabolic pathways, inducing cellular proliferation, differentiation or death (apoptosis) (Figure 2). During organ development, cells must pass through the various cell cycle stages in order to allow for continued remodelling. This process is essential to proper lung development (Bruce et al., 1999;Luyet et al., 2000). Consequently, the redox potential must also cycle continuously (Figure 3). The proliferation phase is accompanied by a higher metabolic rate leading to increased generation of ROS. These ROS in turn favour a shift of the redox potential toward a more oxidized status, inducing the differentiation phase. Alternatively, the oxidized status may 1) induce apoptosis, which favours tissue remodelling, and 2) activate redox-sensitive factors inducing the transcription of genes that encode enzymes involved in glutathione synthesis and GSSG recycling (glutathione reductase). This last event will shift the redox potential toward a more reduced state, beginning a new cell cycle. Fig. 2. The redox environment influences metabolic processes. By modulation from a more reduced to a more oxidized state, the redox environment allows activation and inactivation of various metabolic processes controlling the cell cycle. A reduced state favours cellular proliferation whereas an oxidized state favours differentiation. A more oxidized state will also induce cell death by apoptosis. (Adapted from Schafer FQ et al., 2001) The link between redox potential and BPD can be explained by an exacerbated apoptosis rate caused by an abnormally elevated redox potential (Luyet et al., 2000). Lung samples from premature baboons with BPD (Das et al., 2004) showed a large number of apoptotic events. In newborn guinea pigs given parenteral nutrition for 4 days, the alveolar count was 20% lower when the nutritive solution was infused without light protection, peroxide concentration being higher in light-exposed solutions (Section 5.2 below) (Lavoie et al., , 2005(Lavoie et al., , 2008. On histology, 30% of alveolar cells were in an apoptotic state . During normal alveolar development, however, about 10% of cells die by apoptosis (Luyet et al., 2000), in order to thin the septa between alveoli for more efficient gas exchange (Bruce et al., 1999;Luyet et al., 2000). Various factors may contribute to a shift in redox potential to a more oxidized state ( Figure  3, dashed line). An induced or sustained oxidized status favours the apoptosis process, leading to a loss of tissue such as observed in BPD (Das et al., 2004;Lavoie et al., , 2005Lavoie et al., , 2008. In preterm infants, these factors are oxygen supplementation, parenteral nutrition (containing peroxides), and inflammation. A normal oscillation of the redox environment occurs over time, between a more reduced and a more oxidized state. As the environmental redox potential changes, pulmonary development is supported by cell proliferation, differentiation, and controlled apoptotic events. An excessively oxidized redox environment, such as that caused by oxidative stress, will favour an apoptotic phase, leading to loss of cells and impaired development.

www.intechopen.com
Lung Diseases -Selected State of the Art Reviews 494

Sources of oxidant molecules in BPD
Oxygen supplementation and parenteral nutrition are exogenous sources of oxidant molecules affecting the preterm neonate. Inflammation, however, is an endogenous source of oxidants and its role is complex. Indeed, inflammation can either be a source or consequence of oxidative stress. In this section, we will analyze each of these sources for their potential role in BPD.

Oxygen supplementation
Oxygen supplementation increases ROS generation in the lungs. For this reason, and because of the potential effect of oxidative stress on the development of BPD, it has been hypothesized that high O 2 concentration in inspired air is linked to the development of BPD in preterm neonates (Northway et al., 1967). This hypothesis is supported by animal studies. In rats, exposure to 95% O 2 during the first week of life resulted in a 13% reduction in pulmonary alveolar surface area at 40 days (Randell et al., 1990). However, the impact of oxygen has recently been questioned. Although major clinical advances such as the use of surfactant and continuous positive airway pressure (CPAP) have led to a reduction in oxygen supplementation, their impact on lessening the incidence of BPD has been only about 3% per year between 1993 and 2006, for a global reduction of 30% in 13 years (Stroustrup & Trasande, 2010). This relatively weak contribution was confirmed by studies in newborn preterm baboons, where a reduction in the fraction of inspired oxygen (FiO 2 ) from 80-100% to 21-50% had no significant impact on the levels of fibrosis and alveolar hypoplasia (Coalson et al., 1995(Coalson et al., , 1999. Similarly, a 2010 study of 1316 human infants born at less than 28 weeks' gestation reported a non-significant effect of ventilation strategy leading to a lower oxygen saturation (85-89% versus 91-95%) on the incidence of BPD (SUPPORT Study Group, 2010). Furthermore, The use of high-dose antioxidants scavenging free radicals (vitamins C and E) did not have any protective effect against alveolar hypoplasia (Berger et al., 1998). Free radicals were therefore not the major player in BPD. If higher O 2 in inspired air could lead to a greater cellular concentration of H 2 O 2 that is not quenched by vitamins C or E, the apparently weak effect of oxygen supplementation on BPD development must be explainable by another source of oxidant molecules masking the impact of O 2 . It is noteworthy that the major risk factor for BPD is gestational age; the lower the age, the greater the incidence of BPD. Coincidentally, the more premature the infant, the greater is his dependence on parenteral nutrition. A 2011 study showed that preterm infants developing BPD received more parenteral than enteral nutrition (Wemhöner et al., 2011). In the various studies on BPD, including those in baboons, the gestation ages of the subjects were such that the infants likely required parenteral nutritive support, a major source of ROS and particularly of H 2 O 2 (Laborie et al., 1998;. In itself, parenteral nutrition may be sufficient to induce the development of BPD. In fact, however, it is highly probable that both parenteral nutrition and oxygen supplementation induce oxidative stress, by both increasing the intracellular concentration of H 2 O 2 and modifying the redox potential of glutathione .

Parenteral nutrition
Parenteral nutrition consists of the intravenous administration of a solution containing amino acids, dextrose, electrolytes, vitamins and lipids. Parenteral nutrition is essential for the nutritional support of the preterm infant, bypassing a gastrointestinal system whose immaturity severely limits the natural feeding process. Although parenteral nutrition is sufficient to support growth in the child, the instability of the nutrients in solution favours the generation of undesirable molecules. The admixture of redox-sensitive elements such as amino acids (tryptophan, tyrosine and others), polyunsaturated fatty acids, and vitamin C, in the presence of a strong oxidizing molecule such as dissolved oxygen, will induce oxidation of the nutrients and the formation of their consequent derivatives. For instance, peroxidation of omega-6 polyunsaturated fatty acids will yield lipid hydroperoxides and HNE (Massarenti et al., 2004;Silvers et al., 2001), while vitamin C and dissolved oxygen will produce H 2 O 2 (Laborie et al., 1998;. As vitamin C is the most powerful antioxidant found in parenteral nutrition, the main source of peroxides in parenteral nutrition would appear to be the multivitamin preparation (Laborie et al., 1998;. Furthermore, this solution contains riboflavin, a photosensitive molecule. In the presence of light, photoexcited riboflavin catalyzes a peroxide-producing reaction (Laborie et al., 1998). The simple act of adequately shielding parenteral nutrition solutions from ambient light halved the concentration of peroxides in the infused solution (Chessex et al., 2001;Laborie et al., 1998Laborie et al., , 1999Laborie et al., , 2000Lavoie et al., , 2007 as well as in the urine of preterm infants (Bassiouny et al., 2009;Chessex et al., 2001). Adequate photoprotection of parenteral nutrition has also been reported to reduce the incidence of chronic lung disease (Bassiouny et al., 2009) or BPD  in premature infants. As administered in neonatal units, without adequate photoprotection, parenteral nutritive solutions are contaminated with several molecules having the potential to perturb the redox status of the lung, i.e. lipid hydroperoxides (Silvers et al., 2001), HNE (personal communication of Lavoie JC, 2011), ascorbylperoxide Maghdessian et al., 2010), and H 2 O 2 (Laborie et al., 1998;Silvers et al., 2001). All these molecules are detoxified by the glutathione system. Since glutathione levels are low in preterm infants , these molecules can conceivably overwhelm the glutathione system, allowing the redox potential to shift toward an oxidized state. Infusion of parenteral nutrition without light protection for 4 days in newborn guinea pigs was associated with: 1) a loss of glutathione , 2) a more oxidized glutathione redox potential (Lavoie et al., 2008), and 3) a lower alveolar count (Lavoie et al., , 2005(Lavoie et al., , 2008, as compared to animals infused with a fully photoprotected solution. A recent study demonstrated that the blood glutathione redox potential measured in 7-day-old preterm infants (26  1 weeks' gestation) was correlated with the severity of BPD; a more oxidized status was measured in the most severe cases . Therefore, current knowledge suggests that each oxidant molecule affecting the glutathione system, whether from oxygen supplementation or from parenteral nutrition, may contribute to the development of BPD.

Inflammation
The third major risk factor for BPD is inflammation, a significant source of ROS (Federico et al., 2007;Pereda et al., 2006). Several pro-inflammatory cytokines have been detected in aspirated fluids from infants with BPD, the concentration increasing as a function of assisted ventilation duration and level of oxygen supplementation (Bose et al., 2008). As previously demonstrated, exposure to high amounts of O 2 favours the production of H 2 O 2 , a known activator of transcription factor NF-B (Flohé et al., 1997;Haddad, 2002;Takada et al., 2003), which in turn upregulates the expression of several pro-inflammatory cytokine genes (Federico et al., 2007;Pereda et al., 2006). The oxygen-cytokine connection was further supported by research showing that oxygen supplementation induced an inflammatory response in preterm infants . The association between BPD and inflammation may therefore be explained by an initial oxidative stress followed by a local increase in H 2 O 2 concentration. However, other researchers have argued that an inflammatory process independent of the variation in inspired oxygen concentration could also induce BPD, for example chorioamnionitis (Gien & Kinsella, 2011). Paananen R et al (2009) reported that elevated plasma concentrations of IL-6, a pro-inflammatory cytokine, and IL-10, an anti-inflammatory cytokine, on the first day of life were indicative of greater BPD risk, independently of previous exposure to chorioamnionitis (39% of the 128 preterm neonates in the cohort had had chorioamnionitis; incidence of BPD in cohort, 25%). The lack of correlation between an initial inflammatory process and BPD development was confirmed in 2010 in a study investigating the association between chorioamnionitis and BPD (Prendergast, et al., 2010). From the 71 preterm infants developing BPD, 41 had been exposed to chorioamnionitis and/or funisitis. Their results, however, showed a significant correlation between the severity of BPD and gestational age or birth weight. Thus, endogenous infection does not seem to be connected to the development of BPD while cytokines are, underlining a possible implication of oxidative stress early in life.

Strategies for prevention/treatment
Under the hypothesis that glutathione, by its very involvement in the cellular redox environment, could be a key player in BPD development, one strategy to prevent BPD development or reduce its severity would be to preserve or increase the intracellular concentration of glutathione. It is noteworthy that the low levels of glutathione observed in preterm infants  are not due to a defective enzymatic process. Indeed, GSH synthesis is very active, even in newborns of 26 weeks' gestation . The defect comes rather from the immaturity of the cellular transport system of cysteine (Lavoie et al., 2002), an amino acid whose low intracellular availability limits the synthesis of glutathione (Deneke & Fanburg, 1989). This fact may explain the failure of intravenous administration of N-acetylcysteine to prevent the development of BPD in extremely low birth weight newborns (Ahola et al., 2003). If it is difficult to increase the intracellular concentration of glutathione, one must at least prevent its consumption by reducing oxidative stress. This can be partly achieved by monitoring blood oxygen saturation levels to prevent excessive oxygen supplementation. Prevention of inflammation will help as well. However, limiting peroxide contamination in parenteral nutrition is essential. Though photoprotection of the solution may be difficult to institute in the clinical setting, the process must be initiated in the pharmacy department at the time of compounding and continued until bedside. New nutritive strategies leading to improvements in the nutritive quality of parenteral products, reducing the oxidation of nutrients and preventing the generation of oxidant molecules, will have a positive impact on the incidence of BPD.

Conclusions/perspectives
Presently, no therapy exists for BPD (Gien & Kinsella, 2011) and its prevention is difficult. The etiology is multifactorial. This chapter focused on the part played by oxidative stress, in particular the glutathione redox potential. While a number of oxidant sources can contribute to the shift in redox potential toward a more oxidized state, several BPD-related factors were found to have an impact, among them oxygen supplementation, parenteral nutrition, and inflammation. Modification of even one of these factors may decrease the incidence of BPD, but the best practice remains to administer a combination of new measures, as suggested by Geary C et al. (2008), including early use of surfactant and nasal continuous positive airway pressure for ventilatory support, as well as lowered oxygen saturation targets and better nutritive support. It is remarkable that all associations between biochemical markers and BPD have been observed with parameters measured in the first days of life (Ahola et al., 2004;Geary et al., 2008;Gladstone & Levine, 1994;Lavoie et al., 2008;Ogihara et al., 1999;Pitkanen et al., 1990;Welty, 2001). The first week of life, in both infants and animal models, seems be a critical window during which all efforts to reduce oxidative stress must be pursued.