<<

Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.82.1.F69 on 1 January 2000. Downloaded from Arch Dis Child Fetal Neonatal Ed 2000;82:F69–F74 F69

CURRENT TOPIC

Lung growth: implications for the newborn infant

Sailesh Kotecha

Introduction Table 1 growth stages Modern neonatal practice has improved the outcome of extremely preterm infants. How- Time (weeks) ever, why some infants require prolonged peri- Embryonic 3–7 ods of respiratory support while others improve Canalicular 7–16 Pseudoglandular 16–26 after a short period of , Saccular 26–36 remains largely speculative. Many risk factors, Alveolar 36 weeks–2 years including or volutrauma due to Postnatal growth 2–18 years mechanical ventilation, , and infection, have been identified for the develop- vessels continues, and by the end of this stage ment of chronic lung disease of prematurity the conducting airways, terminal bronchioles, (CLD). Attempts to minimise these with mod- and primitive acinus, are completed. The ern neonatal practice, including newer ventila- pseudo-stratified columnar epithelium is pro- tory techniques, have had minimal impact on gressively replaced by tall columnar cells in the its incidence. Factors other than barotrauma proximal airways and cuboidal cells in the dis- and oxygen toxicity are likely to be important tal acinar structures. in the development of CLD. During the canalicular stage, which occurs Although our understanding of normal fetal between 16 and 26 weeks in utero, further lung development has increased substantially development of the distal airways into defini- over the past few years, it nevertheless remains tive primary acini occurs. The acinar structures rudimentary, especially in infants who have consist of respiratory bronchioles, alveolar survived neonatal intensive care. Animal mod- ducts, and rudimentary alveoli. Development els have provided many clues to the eVects of of the intracinar capillaries, which are derived interventions in the neonatal unit on the lung from the surrounding mesenchyme, accompa- growth of preterm infants. Normal lung growth nies the evolution of the acinus. Lamellar bod- and some of the abnormalities that may result ies containing proteins and phos- from disordered growth or from medical inter- pholipid in type II pneumocytes can be ventions are reviewed in this article. There are observed lining the acinar tubules at this stage.

a vast number of other factors which influence http://fn.bmj.com/ DiVerentiation into type I pneumocytes occurs lung growth—some, such as fetal and in conjunction with the formation of the lung fluid dynamics, deserve reviews of their alveolar–capillary barrier. own. The saccular phase begins with marked enlargement of the peripheral airways as the Normal lung growth acinar tubules dilate and the walls thin, result- Normal lung development, which occurs as a ing in increased gas exchanging surface area. series of complex tightly regulated events, can Lamellar bodies in type II cells increase and on September 24, 2021 by guest. Protected copyright. be divided into a number of stages (table 1).1–3 further maturation into type I cells occurs. During the earliest embryonic stage, the lung Capillaries are closely associated with type I develops as an outgrowth of the ventral wall of cells, thus reducing the distance between the the primitive foregut endoderm. Epithelial cells future air–blood interface. from the foregut endoderm invade the sur- The secondary alveolar septa are formed rounding mesoderm to form the proximal during the alveolar stage, which occurs from 36 structures of the . Following weeks of gestation until at least 24 months the formation of the trachea and the main postnatally.5 The secondary septa consist of bronchi, the five lobes are formed, and by the projections of connective and a double end of this stage, the 18 major lobules are rec- capillary loop. Alveolar formation and matura- ognisable. evidence suggests that the tion occur, with thinning of the alveolar walls surrounding mesoderm regulates the branch- and remodelling of the double capillary loops ing of the tracheobronchial tree.4 At the end of by apoptosis to form a single capillary loop.6 this stage, the pulmonary arteries develop from During this stage marked proliferation of all Department of Child the sixth aortic arches and accompany the types occurs. Mesenchymal cells proliferate Health, University of Leicester, branching airways. and deposit the necessary extracellular matrix. Leicester LE2 7LX The embryonic phase is followed by the Epithelial cells, especially type I and II S Kotecha pseudo-glandular stage—so-called because the pneumocytes, increase in numbers to line the Correspondence to: epithelial tubules are surrounded by thick mes- alveolar walls, and endothelial cells undergo Dr Sailesh Kotecha enchymal tissue. Branching of the airways and massive growth in the secondary septa with Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.82.1.F69 on 1 January 2000. Downloaded from F70 Kotecha

subsequent remodelling to form a single capil- epithelial cell proliferation and the resulting lary loop from a double one. The net result is a branching of the airways. Disruption of its great increase in gas exchanging surface area receptor FGF-R2 in epithelial cells of the and maturation of cells which will respond to airways results in blockage of dichotomous the postnatal environment. branching of the conducting airways. By Due to the diYculties of estimating alveolar contrast, transforming growth factor â numbers at birth, numbers ranging from 20 (TGF-â) inhibits branching morphogenesis, million to 50 million have been quoted. A final epithelial cell growth, and diVerentiation of number of around 300 million is reached by fetal lung explants.11 TGF-â decreases with adulthood. increasing gestation, which removes the inhibi- tory eVects of this growth factor and allows Regulation of lung growth branching to proceed. Other growth factors Most of our knowledge about lung growth is which may be important are listed in table 2. A derived from the study of animals who often more comprehensive review of growth factors have very diVerent timing of morphological and their importance in normal lung growth is 4–7 lung growth compared with humans. For discussed elsewhere. instance, in sheep most of the alveolar develop- Many of these growth factors are produced ment occurs before birth. Rats and mice may by the mesenchyme surrounding the lung epi- be more useful models of human lung growth thelial cells. Indeed, the mesenchyme directs as most alveolar development occurs postna- the ultimate destiny of the epithelial cells. For tally. instance, salivary epithelium grown on mam- Despite widespread interest in this area, our mary mesenchyme results in mammary gland understanding of the mechanisms involved in morphology, and transposition of the bronchial normal lung growth remains limited.7 Table 2 mesenchymal to the peripheral airways results shows the increasing list of transcriptional and in a bronchial-like morphology. Further growth factors which are implicated in normal to the importance of the mesenchyme in lung growth. Hepatic nuclear factor- â directing the epithelial development is given by 3 the presence of the mRNA of growth factors in (HNF3â) seems to be required for the forma- tion of the foregut from which the primitive the mesenchyme and the corresponding pro- lung bud is derived.8 Genetic disruption of tein in epithelial cells—for example, KGF and IGF. The mesenchymal–epithelium interac- HNF3â disrupts formation of the foregut endoderm and its derivatives, including the tions may result from direct cell to cell contact lung.9 In human neonatal lung, it is present in by soluble molecules, including growth factors type II pneumocytes as well as ciliated and (paracrine) or by cell–extracellular matrix 10 interactions. non-ciliated epithelial cells. HNF3â also influences expression of other nuclear factors including transcription factor 1 (TTF- 1). TTF-1 mRNA is detected in rat primordial Factors which may aVect lung growth lung and the protein has been detected as early DEVELOPMENTAL ABNORMALITIES as 11 weeks of gestation in human . Lung growth may be aVected by several factors. TTF-1 seems to increase expression of the sur- During development of the pulmonary tree, factant proteins, at least in vitro, and its laryngeal, tracheal, or oesophageal atresia; tra- ablation by genetic targeting impairs lung mor- cheal stenosis; tracheo-oesophageal atresia or http://fn.bmj.com/ phogenesis, resulting in hypoplastic lung with fistula; pulmonary agenesis; arterio-venous mal- poorly diVerentiated epithelium and poor gas formations or congenital lung cysts (including exchanging areas.7 Interactions between the bronchogenic cysts) may develop during the transcription factors are likely to be more com- embryonic stage. Pulmonary sequestration, pul- plex than described above and very tightly monary hypoplasia or lymphangectasia, con- regulated. genital cystic adenomatous malformations, and As with transcription factors, our under- lung cysts may develop during the pseudo- on September 24, 2021 by guest. Protected copyright. standing of the role of growth factors remains glandular stage. Failure of the pleuro-peritoneal in its infancy. The number of growth factors membranes to close at this stage may lead to the identified continues to increase (table 2), but formation of a congenital diaphragmatic hernia their exact role in both normal lung develop- (CDH). During the canalicular stage, pulmo- ment and in abnormal repair processes after nary hypoplasia may be seen, often secondary to acute or chronic lung disease, remains largely oligiohydramnios or prolonged rupture of mem- rudimentary. Proliferation of cells forming the branes. occurring during the respiratory airways seems to depend on several canalicular stage is very likely to lead to severe of these growth factors, including keratinocyte respiratory distress because of poorly developed growth factor (KGF). KGF seems to promote peripheral airways and poor maturity of cells important to lung maturation—for example, Table 2 Growth factors that may have a role in normal lung growth poor surfactant production by type II cells and inadequate responses to increased Fibroblast growth factor Branching Keratinocyte growth factor (FGF-7) Branching ambient oxygen. Poor development of the Insulin-like growth factor Early branching alveolar–capillary interface during the saccular Platelet derived growth factor (B) Lung growth phase may result in alveolar–capillary dysplasia, Platelet derived growth factor (A) Early lung branching Epidermal growth factor/Transforming growth factor á Branching and growth and other abnormalities during this period Transforming growth factor â Branching morphogenesis include pulmonary hypoplasia, acinar dysplasia, Vascular endothelial growth factor Angiogenesis and vasulogenesis and respiratory distress syndrome if the fetus is Granculocyte macrophage-colony stimulating factor ?Surfactant recirculation delivered after preterm labour. Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.82.1.F69 on 1 January 2000. Downloaded from Lung growth: implications for the newborn infant F71

The newborn term or near term infant will CORTICOSTEROIDS have respiratory distress if the underlying The eVects of antenatal and postnatal adminis- developmental abnormality is clinically signifi- tration of corticosteroids have recently been cant and may be compromised further by post- reviewed.18 Antenatal administration of natal disorders such as persistent pulmonary corticosteroids accelerates lung growth by sev- hypertension of the newborn, meconium aspi- eral mechanisms, including maturation of type ration, respiratory distress syndrome, etc. II pneumocytes, thinning of the double capil- Development of the lung will also be aVected lary loops during the saccular and alveolar by developmental abnormalities of other or- stages of lung development, and partial sup- gans. Both anencephaly and renal hypoplasia pression of the formation of secondary septa. cause pulmonary hypoplasia as do abnormali- Although the normal thinning of the double ties of the thoracic cage—asphyxiating thoracic capillary loops to form thin gas exchanging dystrophy. Inherited genetic disorders may also walls occurs rapidly,18 thus increasing alveolisa- aVect normal lung growth—for example, tion, the final numbers of alveoli are 19 increasing numbers of children with surfactant decreased. Antenatal corticosteroids may also protein B (SPB) deficiency are being reported12 decrease somatic growth, so also aVecting lung and congenital alveolar proteinosis may also growth. Similarly, postnatal administration of have its origins in abnormal and corticosteroids to newborn animals with doses recirculation of surfactant. At least some similar to those used in human infants seems to children with alveolar proteinosis may have an accelerate alveolisation with decreased final 20 abnormal inheritance of the gene for GM–CSF numbers of alveoli. It is somewhat reassuring or its receptor.13 that antenatal treatment with corticosteroids in The mechanisms responsible for the devel- humans seems not to aVect lung function at 7 21 opmental abnormalities mentioned above are years of age, although it should be noted that poorly understood. Genetic factors seem to be lung function is a very insensitive marker of important in some disorders such as alveolar lung growth. proteinosis, and genetic defects of the sur- factant system are increasingly being reported. NUTRITION The factors responsible for developmental By increasing or decreasing the litter size, abnormalities, such as congenital diaphrag- nutrition can be increased or decreased artifi- matic hernia, are likely to be multifactorial, cially. Enhanced nutrition increases surface with interactions occurring between the feto- area, but not septation, and decreased nutrition during fetal life seems to impair septation and maternal genetic make up and the feto- 22 placental environment. surface area, but not the final alveolar size. However, prolonged deprivation of nutrition may increase both surface area and alveolar FETAL BREATHING AND LUNG FLUID size.22 Restriction of protein intake during fetal Both fetal breathing and the volume of development decreases somatic growth as well amniotic fluid are critical to the development as reducing .23 24 Specific lung and growth of the lung. In general, any restric- volumes—that is, volume per body weight— tion to either results in lung hypoplasia, and were, however, increased which suggests that overdistension of the lung by lung fluid the lung is less aVected by malnutrition than 114 improves lung growth. the rest of the body, or that the lack of elasticity http://fn.bmj.com/ Factors which may restrict normal fetal results in overdistended lungs.23 24 Interest- breathing include abnormalities of the chest ingly, re-feeding results in catchup growth, with wall; intrathoracic space occupying lesions normalisation of specific lung volume, suggest- such as congenital diaphragmatic hernia and ing that the potential for recovery is present decreased amniotic fluid result in lung hypo- despite severe fetal malnutrition.23 Fetal mal- plasia. In a series of experiments Wigglesworth nutrition therefore seems to decrease lung vol- and Desai clearly showed that fetal breathing umes but not the maturation of the pulmonary on September 24, 2021 by guest. Protected copyright. was critical to normal lung growth as ablation airways. of the phrenic nerve substantially decreased Postnatal food restriction may exacerbate lung growth.15 Similarly, decreased amniotic lung injury, as has been shown by Langley and fluid as a result of decreased production—as in Kelly: 72 hour food restriction resulted in renal hypoplasia—or increased leak due to pre- increased mortality in preterm guinea pigs mature rupture of membranes, decreases lung exposed to whereas mortality was growth. Whether this is due to restricting fetal unaVected in animals exposed to air.25 Al- breathing or to properties of the fluid itself is though acute lung injury was shown, this was unclear. In contrast, increased fluid in the fetal not thought to be due to an alteration of the lung seems to promote or accelerate lung antioxidant enzymes because these were unaf- growth: in laryngeal atresia the lungs are fected by starvation. increased in volume, surface area, and alveolar 16 17 numbers. Tracheal plugging has been used OXYGEN TOXICITY in both animal models and even in humans to Both and hyperoxia disrupt septation improve lung growth. The exact mechanisms and the ultimate gas exchanging surface area. underlying the influence of fetal breathing and Although newborn animals are more resistant amniotic fluid on lung growth are not entirely to hyperoxia than adults,26 increased oxygen, clear.114Certainly, stretch is likely to be impor- via the formation of tant as this releases many growth factors which including and hydroxyl ions, are important to lung growth. severely disrupts alveolisation in animal Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.82.1.F69 on 1 January 2000. Downloaded from F72 Kotecha

models.27 28 Even after recovery from hyperoxic tors such as compromised lung growth of the exposure, persisting abnormalities remain in fetus itself may contribute toward the develop- lung morphology. Recently, Warner et al ment of CLD, especially if this requires a need reported that mice exposed to 85% oxygen for for respiratory support at birth. There are few 28 days had decreased pulmonary septation, data on CLD in humans, but we have reported increased terminal space diameter, and de- an increase at birth of the potent pro-fibrotic

creased surface area. These eVects may be growth factor TGF-â1 in infants who develop mediated through an increase in pulmonary CLD. In addition to promoting synthesis and inflammation as pro-inflammatory cytokines deposition of the extracellular matrix, this also were greatly increased.29 In the baboon model, inhibits branching morphogenesis which may using preterm animals, oxygen toxicity pro- then explain the decreased alveolisation seen in duces a similar picture of decreased alveolisa- CLD.37 These data suggest that perinatal or tion, decreased surface area, increased alveolar antenatal factors may be important in the size and increased pulmonary inflammation.30 development of CLD. Further work in this area Pulmonary inflammation is likely to be impor- is impeded by the lack of adequate methods to tant in disrupting normal alveolisation but the assess lung growth in the human fetus or new- exact sites and mechanisms whereby septation born infant. is disrupted by hyperoxia are not well under- A model such as that shown in fig 1 may be stood. useful in helping us to understand some of the underlying mechanisms that lead to the devel- PULMONARY INFLAMMATION opment of CLD. In the infant who develops Margraf et al have shown that lung growth, as respiratory distress syndrome, respiratory sup- estimated by alveolisation, surface area, and port with mechanical ventilation and oxygen mean linear diameter (which reflects airway treatment are likely to be required. As shown size), is substantially aVected in infants who die by the baboon model,30 lung growth is aVected from chronic lung disease of prematurity by these postnatal factors. The infant who is 31 (CLD). Both pre- and postnatal factors are well at birth may sometimes progress to oxygen likely to decrease alveolisation. The baboon dependency (often called Wilson-Mikity syn- model of CLD, where the animals are delivered drome). The newborn infant does not develop at a pre-determined time, suggests that postna- respiratory distress syndrome due to adequate tal factors can produce disease similar to CLD, surfactant and antioxidant systems. However, 30 including decreased alveolisation. Mechani- other risk factors for the development of CLD, 32 cal ventilation may be one important factor, such as infection, gastro-oesophageal reflux, or but others such as oxygen toxicity and infection fluid overload may exacerbate abnormal prena- may all aVect lung growth (fig 1). The common tal fetal lung development. Although the infant pathway whereby lung growth is aVected is can adapt at birth, continuing minimal insults likely to be mediated by pulmonary inflamma- may compromise the lung such that oxygen tion. Pulmonary inflammation in infants has treatment is needed. Proving such an hypoth- 33–35 been reported by many groups. It is tempt- esis is diYcult because current methods are ing to speculate that the risk factors for CLD very poor at assessing lung growth— result in pulmonary inflammation as a com- alveolisation, for example—at birth, and lung mon pathway through which lung growth is function tests do not have the sensitivity to compromised (fig 1). Ante- and perinatal detect subtle lung growth abnormalities. http://fn.bmj.com/ factors, which have been poorly investigated in humans, may also be equally important in Strategies to mature the lung compromising lung growth in preterm infants. Antenatal corticosteroids can result in matura- The uterine environment, especially if there is tion of the surfactant and antioxidant systems, infection due to chorioamnionitis, is associated but may also accelerate lung maturation.18 36 with the development of CLD, but other fac- Thyrotropin releasing hormone (TRH) in ani- mal models seems to mature the surfactant on September 24, 2021 by guest. Protected copyright. Prenatal system but not the antioxidant enzymes.38 In Developmental Feto-maternal clinical practice, TRH does not seem to aVect the incidence of neonatal respiratory distress Nutrition 39 40 Genetic syndrome, mortality, or CLD. Recently there have been several publications on tra- cheal ligation for congenital diaphragmatic Endocrine Lung growth hernia (CDH) to promote maturation of the lungs.41 The data are encouraging in animal models, but in humans the results have been Pulmonary inflammation uniformly poor. Tracheal ligation of the human fetus with CDH has been attempted in North Other America, but since the infant might be born Infection Oxygen toxicity Barotrauma PDA ¥ GOR ¥ Fluid with an occluded airway this also remains overload highly experimental (Albanese C, presentation RDS at the European Respiratory Society, Geneva, 1998).42 Prematurity Postnatal The techniques available to promote lung growth postnatally remain poor. Extracorpor- Figure 1 Model of lung growth. Top half shows prenatal factors that may be important in normal lung growth; each may cause inflammation, so aVecting lung growth. Bottom half eal membrane oxygenation remains a rescue shows postnatal factors that may aVect lung growth. treatment for infants with severe respiratory Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.82.1.F69 on 1 January 2000. Downloaded from Lung growth: implications for the newborn infant F73

distress, but can also buy time to stabilise the 5 Thulbeck WM. Postnatal human lung growth. Thorax 1982;37:564–71. child with CDH before performing surgery. 6 Schittny JC, Djonov D, Fine A, Burri PH. Programmed cell Corticosteroids can accelerate lung growth in death contributes to postnatal lung development. Am J Cell Mol Biol 1998;18:786–93. newborn rats with decreased final numbers of 7 MacDonald JA. Lung growth and development. New York: alveoli.18 20 The implications of this observation Marcel Decker, 1997. 8 Sasaki H, Hogan BLM. HNF 3 beta as a regulator of fetal on human newborn infants are diYcult to plate development. Cell 1994;70:103–15. assess as few objective means of assessing lung 9 Weinstein DC, Ruiz I, Altaba A, et al. The winged-helix transcription factor HNF-3beta is required for notochord growth are available. Whether ventilatory development in the mouse embryo. Cell 1994;78: methods such as continuous positive airway 575–88. 10 Ikeda K, Shaw-White JR, Wert S, et al. Hepatocyte nuclear (CPAP), given for prolonged periods factor 3 activates transcription of TTF-1 in respiratory to distend the airways, can improve lung epithelial cells. Mol Cell Biol 1996;16:3626–36. 11 Serra R, Pelton RW. TGF beta inhibits branching morpho- growth postnatally is also diYcult to assess in genesis and N-myc expression in lung bud organ cultures. human babies. Zhang et al recently reported Development 1994;120;2153–61. 12 Nogee LM, deMello DE, Dehner LP, Colten HR. Brief that CPAP at 6.0 cm H2O in ferrets given tra- report: Deficiency of protein B in cheotomies results in increased total lung congenital alveolar proteinosis. N Engl J Med 1993;328:406–10. capacity together with increased lung weight, 13 Dirksen U, Nishinakamura R, Groneck P, Hattenhorst U, total lung protein and DNA, suggesting that Nogee L, Murray R, Burdach S. Human pulmonary alveolar proteinosis associated with a defect in GM-CSF/ this mode of treatment may increase lung vol- IL-3/IL-5 receptor common beta chain expression. J Clin umes at a critical time in newborn animals.43 Invest 1997;100:2211–17. 14 Hooper SB, Harding R. Fetal lung fluid: a major The data must, however, be interpreted with determinant of the growth and functional development caution as an increase in total lung capacity of the fetal lung. Clin Exp Pharmacol Physiol 1995;22: 235–47. may not necessarily be due to improved 15 Wigglesworth JS, Desai R. Is fetal respiratory function a pulmonary function or lung growth. Similarly, major determinant of perinatal survival? Lancet 1982;i:264–7. perflurocarbons have been used in animal 16 Wigglesworth JS, Desai R, Hislop AA. Fetal lung growth in models to distend the lung in the hope of laryngeal atresia. Pediatr Pathol 1987;7:515–25. 44 17 Silver MM, Thurston WA, Patrick JE. Perinatal pulmonary improved lung growth. Improved distension hyperplasia due to laryngeal atresia. Hum Pathol may only increase lung volume, but both lung 1988;19:110–13. 18 Vyas JR, Kotecha S. The EVects of antenatal and postnatal growth and its function must be improved if corticosteroids on the preterm lung. Arch Dis Child Fetal this technique is to be recommended. Neonatal Ed 1997;77:F147–50. 19 Beck JC, Mitzner W, Johnson JWC, et al. Betamethasone Whether any of the techniques or methods and the rhesus fetus: eVect on lung morphometry and described above have a future in promoting connective tissue. Pediatr Res 1981;15:235–40. lung maturation in human infants remains to 20 Tschanz SA, Damke BM, Burri PH. Influence of postnatally administered glucocorticoids on rat lung growth. Biol be seen. It must be emphasised that the Neonate 1995;68:229–45. risk:benefits ratio for any new intervention 21 Wiebicke W, Poynter A, Chernick V. Normal lung growth following antenatal dexamethasone treatment for respira- must be fully assessed because the interven- tory distress syndrome. Pediatr Pulmonol 1988;5:27–30. tions themselves may be associated with 22 Massaro M, Massaro GD. Regulation of the architectural development of the lung. In: Crystal RG, West JB, Barnes significant adverse eVects, such as infection, PJ, eds. The lung: scientific foundations.New York: Raven bleeding, or need for intubation. Press, 1997: 1030. 23 Kalenga M, Tschanz SA, Burri PH. Protein deficiency and In summary, the factors which are important the growing rat lung: I. Nutritional findings and related in lung growth are clearly complex, with inter- lung volumes. Pediatr Res 1995;37:783–8. 24 Kalenga M, Tschanz SA, Burri PH. Protein deficiency and actions of various tissues occurring in the face the growing rat lung: II. Morphometric analysis and mor- of multiple regulatory elements. Fluid in the phology. Pediatr Res 1995;37:789–95.

25 Langley SC, Kelly FJ. EVect of food restriction on http://fn.bmj.com/ airways maintained at a minimal pressure hyperoxia-induced lung injury in preterm guinea pig. Am J seems to be important and any interference Physiol 1992;263:L357–62. 26 Johnston CJ, Wright TW, Reed CK, Finkelstein JN. with the mechanism of amniotic fluid produc- Comparison of adult and newborn pulmonary cytokine tion will aVect normal lung growth. Factors mRNA expression after hyperoxia. Exp Lung Res 1997;23:537–52. aVecting intrapulmonary fluid include renal 27 Randell SH, Mercer RR, Young SL. Postnatal growth of abnormalities, patency of the airways, space pulmonary acini and alveoli in normal and oxygen- exposed rats studied by serial section reconstructions. Am occupying lesions, abnormal airway contrac- J Anat 1989;186:55–68. tions (including abnormal breathing patterns), 28 Thibeault DW, Mabry S, Rezaiekhaligh M. Neonatal on September 24, 2021 by guest. Protected copyright. pulmonary oxygen toxicity in the rat and lung changes. inadequate thoracic volume and inadequate Pediatr Pulmonol 1990;9:96–108. fluid secretion. Contractions seem to be 29 Warner BB, Stuart LA, Papes RA, Wispe JR. Functional and pathological eVects of prolonged hyperoxia in neonatal important in the expression of both surfactant mice. Am J Physiol 1998;275:L110–17. and growth factors essential for normal fetal 30 Coalson JJ, Winter V, deLemos RA. Decreased alveolariza- tion in baboon survivors with broncho-pulmonary dyspla- lung growth. Abnormalities of any of these fac- sia. Am J Respir Crit Care Med 1995;152:640–6. tors are likely to adversely aVect lung growth in 31 Margraf LR, Tomashefski JF, Bruce MC, Dahms BB. Mor- phometric analysis of the lung in bronchopulmonary dys- the fetus, predisposing the newborn infant to plasia. Am Rev Respir Dis 1991;143:391–400. respiratory distress at birth. 32 Hislop AA, Wigglesworth JS, Desai R, Aber V.The eVects of preterm delivery and mechanical ventilation on human lung growth. Early Hum Dev 1987;15:147–64. I thank professors David Field and Michael Silverman for their 33 Kotecha S, Chan B, Azam N, Silverman M, Shaw RJ. comments. Increase of interleukin-8 and soluble ICAM-1 in broncho- alveolar lavage fluid from preterm infants who develop chronic lung disease of prematurity. Arch Dis Child Fetal 1 Thurlbeck WM. Pre- and postnatal organ development. In: Neonatal Ed 1995;72:F90–6. Chernick J, Mellins RB, eds. Basic mechanisms of respiratory 34 Bagchi A, Viscardi RM, Taciak V, Ensor JE, McCrea KA, disease: Cellular and integrative. Philadelphia: BC Decker, Hasaday JD. Increased activity of interleukin-6 but not 1991:23–35. tumour necrosis factor-á in lung lavage of preterm infants 2 Adamson IY. Development of lung structure. In: Crystal is associated with the development of bronchopulmonary RG, West JB, Barnes PJ, eds. The lung: scientific foundations. dysplasia. Pediatr Res 1994;36:244–52. New York: Raven Press, 1997: 993–1001. 35 Kotecha S. Cytokines in chronic lung disease of prematu- 3 Hilfer SR. Morphogenesis of the lung: control of embryonic rity. Eur J Paediatr 1996;155:514–7. and fetal branching. Annu Rev Physiol 1996;58:93–113. 36 Watterberg KL, Demers LM, Scott SM, Murphy S. Chori- 4 Post M. Tissue interactions. In: Crystal RG, West JB, oamnionitis and early lung inflammation in infants in Barnes PJ, eds. The lung: scientific foundations. New York: whom bronchopulmonary dysplasia develops. Pediatrics Raven Press, 1997:1003–101. 1996:97:210–5. Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.82.1.F69 on 1 January 2000. Downloaded from F74 Kotecha

37 Kotecha S, Wangoo A, Silverman M, Shaw RJ. Increase in 41 Benachi A, Chailley-Heu B, Delezoide AL, et al. transforming growth factor-â1 in broncho- Lung growth maturation after tracheal occlusion in alveolar lavage fluid obtained from infants with chronic diaphragmatic hernia. Am J Respir Crit Care Med lung disease of prematurity. J Pediatr 1996;128:464–9. 1998;157:921–7. 38 Frank L. Development of the antioxidant defences in fetal 42 Davis CF, Sabharwal AJ. Managment of congenital life. Semin Neonatol 1998;3:173–82. diaphragmatic hernia. Arch Dis Child Fetal Neonatal Ed 39 Ballard RA, Ballard PL, Cnaan A, et al. Antenatal 1998;79:F1–3. thyrotropin-releasing hormone to prevent lung disease in 43 Zhang S, Garbutt V, McBride JT. Strain-induced growth of preterm infants. N Engl J Med 1998;338:493–8. the immature lung. J Appl Physiol 1996;81:1471–6. 40 ACTOBACT study group. Australian collaborative trial of 44 Nobuhara KK, Fauza DO, DiFiore JW, et al. Continuous antenatal thyrotropin releasing hormone (ACTOBAT) for intrapulmonary distension with perfluorocarbon acceler- prevention of neonatal . Lancet ates neonatal (but not adult) lung growth. J Ped Surg 1995;345:877–82. 1998;33:292–8. http://fn.bmj.com/ on September 24, 2021 by guest. Protected copyright.