EPIDEMIOLOGICAL AND GENETIC STUDY OF RESPIRATORY DISTRESS SYNDROME IN PRETERM INFANTSSpecific aspects of twin and multiple births

Abstract in Finnish

RIITTAMAR TTILA

Department of Paediatrics,University of Oulu

OULU 2003

RIITTA MARTTILA

EPIDEMIOLOGICAL AND GENETIC STUDY OF RESPIRATORY DISTRESS SYNDROME IN PRETERM INFANTSSpecific aspects of twin and multiple births

Academic Dissertation to be presented with the assent ofthe Faculty of Medicine, University of Oulu, for publicdiscussion in the Auditorium L12 of the Department ofPaediatrics, on December 12th, 2003, at 12 noon.

OULUN YLIOPISTO, OULU 2003

Copyright © 2003University of Oulu, 2003

Supervised byProfessor Mikko Hallman

Reviewed byProfessor Vineta FellmanProfessor Marjo-Riitta Järvelin

ISBN 951-42-7214-5 (URL: http://herkules.oulu.fi/isbn9514272145/)

ALSO AVAILABLE IN PRINTED FORMATActa Univ. Oul. D 766, 2003ISBN 951-42-7213-7ISSN 0355-3221 (URL: http://herkules.oulu.fi/issn03553221/)

OULU UNIVERSITY PRESSOULU 2003

Marttila, Riitta, Epidemiological and genetic study of respiratory distress syndromein preterm infants. Specific aspects of twin and multiple birthsDepartment of Paediatrics, University of Oulu, P.O.Box 5000, FIN-90014 University of Oulu,Finland Oulu, Finland2003

Abstract

Respiratory distress syndrome, RDS, is a multifactorial lung disease of premature infants. The maincause of RDS is a deficiency of pulmonary surfactant, a lipoprotein mixture required to reduce surfacetension at the air-liquid interface and to prevent generalized atelectasis of the alveolar ducts andalveoli. Prematurity is the most important factor predisposing to RDS. During the past decade thenumber of multiple pregnancies has increased significantly as a result of diversified infertilitytreatments and advanced maternal age. Due to the considerably higher rate of preterm births ofmultiples compared to singletons, RDS is one of the major causes of morbidity among them.

The objectives of the present research were to evaluate the incidence and risk factors of RDS intwins compared to singletons, and to assess the role of SP-A and SP-B gene variations and gene-environment interactions in the susceptibility to the disease in a population of preterm twins andhigher order multiples.

This research showed that during the past fifteen years the gestational age-specific incidence ofRDS has declined. Twin infants do not have increased risk of RDS except when born very immaturelyat a very early gestational age. The presenting twin is less susceptible to RDS than the non-presentingtwin or singleton infant after 28 weeks of gestation until term. Additionally the SP-B Ile131Thrpolymorphism was shown to affect the susceptibility specifically in the presenting twin. The role ofSP-A polymorphisms in the risk of RDS in twins turned out to be different from that in singletons.The major allele and genotype of SP-A1 were associated with a decreased risk of RDS in near-termtwin infants. The threonine allele in SP-B Ile131Thr appeared interactively with SP-A1 to associatewith the risk of RDS both in twins and in singletons: associating with a lower risk of RDS insingletons at very early gestation, but surprisingly associating with a protection in twins and multiplesfrom RDS near term. The risk of RDS, defined by the interaction of SP-A and SP-B alleles, wasadditionally associated with the fetal mass. Thus, the difference in the susceptibility to RDS inpremature singletons and multiples may depend on the size of the conceptus.

In an evaluation of the genetic risk factors for RDS, the classical twin study method comparingthe concordance of a disease between MZ and DZ twins underestimates the extent of heredity. Severalpredominant intrauterine and perinatal environmental factors contribute to disease susceptibilityregardless of zygosity and are suspected to override the hereditary components of RDS. Twingestation was shown to be an effect modifier in the genetic susceptibility to RDS.

Keywords: gene polymorphism, multiple pregnancy, prematurity, preterm birth, pulmonary surfactant-associated proteins, respiratory distress syndrome

Marttila, Riitta, Ennenaikaisena syntyneen lapsen hengitysvaikeusoireyhtymänepidemiologinen ja geneettinen tutkielma: erityisen tarkastelun kohteena kaksos- jamonisikiöiset synnytykset Lastentautien klinikka, Oulun yliopisto, PL 5000, 90014 Oulun yliopisto2003Oulu, Finland

Tiivistelmä

Vastasyntyneen hengitysvaikeusoireyhtymä, RDS-tauti, on hyvin monitekijäinen ennenaikaisenasyntyneiden lasten sairaus. Taudin tärkein syy on keuhkojen pintajännitystä alentavan aineen,surfaktantin, puute ennenaikaisessa synnytyksessä. Surfaktantti on rasva-proteiiniseos, joka alentaakeuhkorakkuloiden pintajännitystä ja estää rakkuloiden kasaan painumisen hengityksen aikana.Epäkypsyys on tärkein RDS-taudin kehittymiselle altistava tekijä. Viimeisen vuosikymmenen aikanamonisikiöiset raskaudet ovat lisääntyneet merkittävästi keinohedelmöitysten ja synnyttäjienkohonneen iän seurauksena. RDS-tauti on monisikiösynnytyksissä tärkeä sairastavuutta aiheuttavatekijä, koska monisikiöraskaudet päättyvät huomattavasti useammin ennenaikaisesti kuinyksösraskaudet.

Tämän tutkimuksen tarkoituksena oli selvittää RDS-taudin ilmaantuvuutta ja riskitekijöitäkaksosilla verrattuna yksösiin, tutkia surfaktanttiproteiinien A ja B geenivaihtelua sekä geenien jaympäristötekijöiden vuorovaikutusta kaksosten ja monisikiöisistä raskauksista syntyneiden lastenRDS-taudissa.

Tutkimus osoitti, että viimeisten viidentoista vuoden aikana RDS-taudin raskausikäkohtainenilmaantuvuus on vähentynyt. Vain hyvin epäkypsillä, ennenaikaisilla kaksosilla on lisääntynyt riskisairastua RDS-tautiin verrattuna yksösiin. 28 raskausviikon jälkeen ensimmäisenä syntyvälläkaksosella on pienempi riski saada RDS-tauti kuin toisena syntyvällä tai yksösellä. Lisäksisurfaktanttiproteiini B:n (SP-B) geenin eksonin 4 monimuotoisuus osoittautui vaikuttavan RDS-taudin alttiuteen erityisesti ensin syntyvällä kaksosella. Surfaktanttiproteiini A:n (SP-A) geeninyhteys RDS-taudin riskiin osoittautui olevan erilainen kaksosilla verrattuna yksösiin. Lähestäysiaikaisilla kaksosilla SP-A1 geenin valta-alleeli ja genotyyppi liittyivät RDS-taudinvähentyneeseen riskiin. SP-B geenin eksonin 4 treoniini alleeli osoittautui liittyvän yhdessä SP-A1valtagenotyypin kanssa RDS-taudin riskiin sekä yksösillä että kaksosilla: hyvin ennenaikaisillayksösillä geenien yhteisvaikutus lisäsi riskiä, kun taas lähes täysiaikaisilla kaksosilla sama geenienyhteisvaikutus yllättäen suojeli RDS-taudilta. SP-A ja SP-B geenien vuorovaikutuksen kauttavälittyvä RDS-taudin riski oli lisäksi yhteydessä sikiön kokoon. Siten ennenaikaisena syntyneidenyksösten ja monisikiöisten RDS-taudin alttiuden ero voi olla riippuvainen sikiön koosta.

Klassinen kaksostutkimusmenetelmä, jossa verrataan identtisten ja ei-identtisten kaksostensairastuvuutta, aliarvioi perimän määräämien tekijöiden osuutta RDS-taudin synnyssä. Useathallitsevat kohdunsisäiset ja synnytykseen liittyvät ympäristötekijät vaikuttavat RDS-taudin syntyynja kumoavat perimän osuutta. Kaksosuus osoittautui olevan periytymisen vaikutusta muokkaavatekijä RDS-taudissa.

Asiasanat: ennenaikainen synnytys, epäkypsyys, geenien monimuotoisuus, hengitysvaikeusoireyhtymä, monisikiöraskaus, surfaktanttiproteiinit

Acknowledgements

The present work was carried out at the Department of Paediatrics, University of Oulu, during the years 1997–2003.

I express my deepest gratitude to my supervisor, Professor Mikko Hallman, M.D., Head of the Department of Paediatrics for giving me the opportunity to do my thesis work under his guidance. I thank him for his positive attitude, kind patience and trust in this work during these years. I also express my appreciation to Professor Matti Uhari, M.D., for teaching me consistent and critical attitude towards scientific thinking and methodology. I kindly thank Professor Jaakko Kaprio, M.D., Head of the Department of Public Health, University of Helsinki, for guiding me in the field of epidemiology.

I wish to express my cordial thanks to my second supervisor Marja-Leena Pokela, M.D., for her support and encouragement, friendship and guidance in the treatment of the tiniest newborns. Warm thanks are due to my hard-working co-author Ritva Haataja, Ph.D., for her excellent knowledge of genetics and positively critical attitude towards this research. We have shared many moments of hope — and despair — during these years. I am specifically obliged to Mika Rämet, M.D., Johan Löfgren, Med.Cand., and Meri Rova, M.Sc., in the Research Laboratory of Paediatrics for their kind help and co-authorship. Susan Guttentag, M.D., is acknowledged for her contribution. I thank sincerely Ms. Maarit Hännikäinen, Ms. Elsi Jokelainen and Ms. Mirkka Ovaska for their patient and excellent lab work.

I am deeply grateful to Professor Marjo-Riitta Järvelin, M.D., Head of the Department of Epidemiology and Public Health, Imperial College, University of London, and Professor Vineta Fellman, M.D., Head of the Department of Paediatrics and Neonatology, University of Lund, for their prompt review of the manuscript of this thesis and for their expertise. I thank warmly Juha Turtinen, M.Sc., who has helped me numerous times with statistics and computer software. I owe my gratitude to Ms. Eija Rautio and Ms. Marjatta Paloheimo for their kind assistance and supportive attitude towards me and my work. Sirkka-Liisa Leinonen, Lic.Phil., is acknowledged for the skilful revision of English language of the thesis and the original publications.

I wish to express my warm thanks to Docent Maila Koivisto, M.D., my “official” teacher in neonatology. She has taught me a lot with her persistence in struggling for the most preterm infants as well as with her continuous enthusiasm in scientific work. I also thank her for all the help and advice as a co-author. I express my sincere gratitude to Docent Outi Tammela, M.D., University of Tampere, for flexible co-operation and for her contribution to sample collection. Docent Mika Gissler, M.D., and Dr. Martti Virtanen, M.D., in the National Research and Development Center for Welfare and Health Care (STAKES), and Docents Kirsti Heinonen, Sami Ikonen and Pentti Kero are kindly acknowledged for their contribution to this thesis.

As I have been mostly living and working elsewhere, my closest colleagues in the neonatal ward of Oulu University Hospital have helped me a lot with the collection of the study material. I thank especially Eija Anttila, M.D., Tuula Kaukola, M.D., Timo Saarela, M.D., and Marita Valkama, M.D., for valuable help, support and friendship both in clinical practise and in scientific work during these years. During my hasty visits now and then we have had time for inspiring discussions, too. I have enjoyed working in the neonatological ward 55, and I thank the entire nursing staff for smooth and encouraging co-operation. The staff in the Department of Obstetrics and Gynecology also made a great effort to collect the blood samples. I am grateful to my colleagues and friends in the paediatric departments of Seinäjoki Central Hospital and University of Oulu for their flexibility and supportive attitude. I express my sincere thanks to all the tiny newborns and their parents for participating in these studies. Hopefully this work will take us a small step closer to understanding and countering the challenges of preterm birth and the consequent disorders.

I thank warmly my long-standing friends Anu-Maaria Hämäläinen, Merja Joro, Sirpa Koski-Tervo and Tarja Nevala for their concern and care. I owe my warm gratitude to my parents, Terttu and Heimo Juntunen, and my sister Marja-Leena for their loving support and practical help throughout my endless years of studying. Marja-Leena has shared the ups and downs of scientific work with me while finishing her own doctoral thesis.

Finally, I owe my loving gratitude to my family. I thank my husband Timo for his patience and for providing opportunities for me to work where-ever and whenever I needed to. I thank our children Juho, Emmi, Tuukka and Hanna for the joy and optimism they spread around. They are the most precious things in my life and I dedicate this dissertation to them.

This work was supported financially by the Foundation of Paediatric Research in Finland, the Alma and K.A. Snellman Foundation, Oulu, Finland, and the Research Funds of the University of Oulu, the University of Tampere and Seinäjoki Central Hospital, which all are kindly appreciated.

Seinäjoki, October 2003

Riitta Marttila

Abbreviations

AGA appropriate for gestational age (between – 2SD and + 2 SD) ARDS adult respiratory distress syndrome bp base pair BPD bronchopulmonary dysplasia CAP congenital alveolar proteinosis CI confidence interval CLD chronic lung disease COPD chronic obstructive pulmonary disease CPAP continuous positive airway pressure CRD carbohydrate recognition domain DPPC dipalmitoyl phosphatidylcholine DNA deoxyribonucleic acid DZ dizygotic ELBW extreme low birth weight (< 1000 g) HLA human leukocyte antigen ICD International Classification of Diseases IL interleukin IUGR intrauterine growth restriction kb kilobase kDA kilodalton LBW low birth weight (< 2500 g) LD linkage disequilibrium LPS lipopolysaccharide L/S lecithin to sphingomyelin MZ monozygotic mRNA messenger ribonucleic acid OR odds ratio p short arm of a chromosome p probability PC phosphatidylcholine PCR polymerase chain reaction

PG phosphatidylglycerol PIF pulmonary interstitial fibrosis PNM perinatal mortality PPROM preterm premature rupture of fetal membranes q long arm of a chromosome RDS respiratory distress syndrome RFLP restriction fragment length polymorphism RSV respiratory syncytial virus SGA small for gestational age SNP single nucleotide polymorphism SP surfactant protein TM tubular myelin TLR toll-like receptor TNF tumor necrosis factor UTR untranslated region VLBW very low birth weight (< 1500 g)

List of original papers

This thesis is based on the following articles, which are referred to in the text by their Roman numerals: I Marttila R, Kaprio J & Hallman M (2003) Respiratory distress syndrome in twin

infants compared to singletons. Submitted for publication. II Koivisto M, Marttila R, Saarela T, Pokela M-L, Jouppila J & Hallman M (2003)

Changing incidence and outcome of infants with respiratory distress syndrome in the 1990s: a population based survey. Submitted for publication.

III Marttila R, Haataja R, Rämet M, Löfgren J & Hallman M (2003) Surfactant protein B polymorphism and respiratory distress syndrome in premature twins. Hum Gen 112:18–23.

IV Marttila R, Haataja R, Pokela M-L, Tammela O & Hallman M (2003) Surfactant protein A gene locus and respiratory distress syndrome in Finnish premature twin pairs. Ann Med 35:344–352.

V Marttila R*, Haataja R*, Guttentag S & Hallman M (2003) Surfactant protein A and B genetic variants in respiratory distress syndrome in singletons and twins. Am J Resp Crit Care Med Aug 168:1216-1222.

* Contributed equally to this work

Contents

Abstract Acknowledgements Abbreviations List of original papers Contents 1 Introduction .................................................................................................................15 2 Review of the literature................................................................................................16

2.1 Human lung development.......................................................................................16 2.2 Composition, function and metabolism of pulmonary surfactant...........................18 2.3 Pulmonary surfactant protein A ..............................................................................19

2.3.1 Structure of SP-A.............................................................................................19 2.3.2 The human SP-A gene .....................................................................................20 2.3.3 Processing and function of SP-A.....................................................................21

2.4 Pulmonary surfactant protein B ..............................................................................22 2.4.1 Structure of SP-B.............................................................................................22 2.4.2 The human SP-B gene .....................................................................................23 2.4.3 Secretion and function of SP-B .......................................................................24

2.5 Regulation of SP-A and SP-B expression...............................................................24 2.6 Predicting the risk of RDS with surfactant component analysis.............................25 2.7 Clinical picture of RDS in preterm infants .............................................................26 2.8 Pathophysiology of RDS ........................................................................................26 2.9 Epidemiology of RDS ............................................................................................27 2.10 Clinical risk factors for RDS ................................................................................28

2.10.1 Preterm premature rupture of membranes and intrauterine infection ............29 2.10.2 Intrauterine growth ........................................................................................30 2.10.3 Maternal diseases...........................................................................................30 2.10.4 Mode of delivery ...........................................................................................31 2.10.5 Twins and higher order gestations .................................................................31

2.10.5.1 Incidence of twins...................................................................................31 2.10.5.2 Zygosity of twins ....................................................................................32 2.10.5.3 RDS-related morbidity of twins..............................................................32

2.11 Genetic predisposition to RDS..............................................................................33 2.11.1 Epidemiological evidence..............................................................................33 2.11.2 Hereditary SP-B and SP-C deficiency ...........................................................34 2.11.3 Allelic association studies..............................................................................35 2.11.4 SP-A gene polymorphism ..............................................................................37 2.11.5 SP-B gene polymorphism ..............................................................................37 2.11.6 Surfactant protein genes associated with other pulmonary diseases than RDS .........................................................................................37

2.12 Prevention, treatment and prognosis of RDS........................................................38 3 Objectives of the present research ...............................................................................40 4 Subjects and methods...................................................................................................41

4.1 The study populations and the study design ...........................................................41 4.2 Epidemiological approach ......................................................................................43 4.3 Genetic studies........................................................................................................44 4.4 DNA samples and genotyping ................................................................................46 4.5 Ethical considerations.............................................................................................46 4.6 Statistical analysis...................................................................................................47

5 Results .........................................................................................................................49 5.1 Population-based epidemiology and risk factors for RDS (I, II) ............................49

5.1.1 Incidence of RDS (I, II) ...................................................................................49 5.1.2 Risk factors for RDS in twins (I).....................................................................50 5.1.3 RDS-related mortality and morbidity (II) ........................................................52 5.1.4 Outcome at one year of age (II).......................................................................52

5.2 Concordance of RDS in MZ and DZ twin pairs (III, IV) .......................................52 5.3 SP-A and SP-B polymorphisms and RDS in twins (III, IV)...................................53 5.4 Interaction by the SP-B gene and plurality on the association between the SP-A gene and RDS (V) ...................................................................................55

6 Discussion....................................................................................................................60 6.1 Epidemiological aspects .........................................................................................60 6.2 Methods in genetic studies .....................................................................................61 6.3 Concordance of RDS between MZ and DZ twin pairs ...........................................63 6.4 Role of SP-A and SP-B polymorphisms in RDS of twins ......................................64 6.5 Interaction by the SP-B gene and plurality on the association between the SP-A gene and RDS..........................................................................................65

7 Summary and future perspectives................................................................................67 References

1 Introduction

Respiratory distress syndrome, RDS, is a multifactorial lung disease of premature infants. It is characterized by respiratory failure and deficient gas exchange within the first few hours after birth unless effective treatment is instituted. The main cause of RDS is a deficiency of pulmonary surfactant, a lipoprotein mixture that is required to reduce surface tension at the air-liquid interface and to prevent generalized atelectasis of the alveolar ducts and alveoli.

Functional prematurity is the most important factor predisposing to RDS. The incidence of RDS decreases as a function of the length of gestation and RDS only occasionally manifests at term birth. During the 1990s the rate of preterm deliveries, defined as delivery between 22 weeks and 37 completed weeks of gestation, increased from 5.4% in 1990 to 6.1% in 1999 in Finland (Gissler et al 2002). At the same time the number of multiple pregnancies increased significantly as a result of diversified infertility treatments and advanced maternal age. In the late 1990s 1.6 % of all births were multiple births. Multiple pregnancies result in much higher rate of preterm births than singleton pregnancies, as almost half of the deliveries occur preterm compared to 4–5 % in singleton pregnancies. The high rate of preterm births in twin and higher order gestations leads to disproportionately high overall neonatal morbidity.

As twin and multiple infants represent approximately 20% of all infants born preterm, RDS is one of the major causes of morbidity among them. Maternal diseases, gender and ethnicity are known factors affecting the risk of RDS. There is also evidence of a more specific genetic background, as surfactant protein SP-A and SP-B genes have been suggested to associate with RDS.

The objectives of the present research were to evaluate the incidence and risk factors of RDS in twins compared to singletons, and to assess the role of SP-A and SP-B gene variations and gene-environment interactions in the susceptibility to RDS disease in a population of preterm twins and higher order multiples.

2 Review of the literature

2.1 Human lung development

The human lung is a derivative of the primitive foregut and appears by 22 days of gestation as an enlargement of the caudal end of the laryngotracheal sulcus. In humans, the airways are well formed by 20 weeks gestation. The development of lungs can be divided into five overlapping stages of organogenesis: early embryonic (3–7 weeks of gestation), pseudoglandular (5–17 weeks of gestation), canalicular (16–26 weeks of gestation), saccular (24–38 weeks of gestation) and alveolar (36 weeks of gestation to 2 years of postnatal age) (Burri 1997, Hackett & Gitlin 1997, Kotecha 2000). The stages of organogenesis together with the expression of the surfactant proteins A (Ballard et al 1986) and B (Liley et al 1989) during human gestation are shown in Figure 1.

17

Fig. 1. The stages of lung organogenesis are shown above (modified from Burri 1997). Lung development continues after preterm delivery. The expression of surfactant protein genes begins early in the fetus and increases towards term gestation.

During the initial embryonic stage, the lung develops as an overgrowth of the foregut endoderm. Epithelial cells from the foregut endoderm invade the surrounding mesoderm to form the proximal structures of the respiratory tract. The trachea, the main bronchi and the major lung lobules are formed, as are also the pulmonary arteries. Epithelial tubules are surrounded by thick mesenchymal tissue during the pseudoglandular phase. By the end of this stage, the conducting airways, terminal bronchioles and primitive acinus are completed. During the canalicular phase the gas-exchanging tissue becomes visible in light microscopy, and lamellar bodies containing surfactant proteins and phospholipids in type II pneumocytes can be observed lining the acinar tubules. Vascularization with the concomitant formation of the alveolar-capillary barrier for gas exchange takes place. The secretion of surface-active material enables a prematurely born infant to survive. The saccular phase is characterized by marked enlargement of the peripherial airways and development of primitive saclike alveoli called saccules, which increase the gas-exchanging surface area. Lamellar bodies in type II cells increase, and further maturation into type I alveolar epithelial cells takes place. Capillaries lie close to type I cells, thus reducing the distance at the air-blood interface. The first alveoli begin to appear by 32 weeks of gestation, and by term, an average of 30% of the adult number of alveoli have developed. The formation of thin secondary alveolar septa, remodelling of the capillary bed and marked proliferation of all cell types occur primarily postnatally. During this alveolar phase of lung development, the gas-exchanging surface area increases tenfold, from 20–50 million alveoli at birth to an average of 300 million by adulthood.

A number of hormones and growth factors, induced by fetal breathing movements and other events, interactively regulate normal lung growth and maturation. The exact

5 10 15 20 25 30 35 400,0

0,2

0,4

0,6

0,8

SP-B

SP-AEmbryonal stage

Pseudoglandular stage

GESTATIONAL AGE (WEEKS)

EXTRAUTERINE LUNG DEVELOPMENTAND GROWTH

INTRAUTERINE LUNG DEVELOPMENT

Alveolar stage

Saccular stage

Canalicular stage

18

mechanisms are still largely unknown (Jobe 1999, Kotecha 2000). Lung development and growth can be altered as a result of various antenatal and perinatal events, such as an inflammation caused by intrauterine infection (Jobe & Ikegami 2001). A deficiency or impaired function of the regulatory elements, such as the vascular endothelial growth factor (VEGF), affects pulmonary vascular growth and alveolarization (Grover et al 2003) as well as the differentiation of type II cells (Compernolle et al 2002).

2.2 Composition, function and metabolism of pulmonary surfactant

The maintenance of alveolar structure during the cyclic changes in lung volume is critical for normal respiration. The tendency of the alveolus to collapse at the end of expiration is mainly due to the high surface tension generated by the aqueous layer lining the alveolar epithelium. Alveolar stability is achieved by the pulmonary surfactant present at the air-liquid interface which reduces surface tension and prevents atelectasis (Hawgood & Clements 1990). A schematic representation of the structure of an alveolus is shown in Figure 2. Surfactant components also participate in the clearance of foreign material and in the host defence mechanisms and innate immunity of the lung (Crouch & Wright 2001).

Fig. 2. Structure of an alveolus (modified from Hawgood & Clements 1990).

Human surfactant consists of approximately 80% phospholipids, 8% neutral lipids (cholesterol and free fatty acids) and 12% proteins and its composition is fairly constant across mammalian species (King 1984). The most abundant phospholipid is

19

phosphatidylcholine (PC), especially its disatured form dipalmitoylphosphatidylcholine (DPPC). DPPC is the component of surfactant that reduces surface tension. The other lipid components of the surfactant (phosphatidylglycerol, phosphatidylinositol and cholesterol) and the hydrophobic surfactant proteins facilitate the adsorption of DPPC. The amount of DPPC increases concurrently with the appearance of lamellar bodies after 22 weeks of gestation in the fetal human lung (Clements 1977).

Approximately half of the proteins consists of contaminating protein from plasma or lung tissue (Jobe 1993). In addition there are four surfactant proteins (SPs) expressed by respiratory epithelial cells, designated as SP-A, SP-B, SP-C and SP-D. SP-A and SP-D are large glycosylated water-soluble proteins and members of the calcium-dependent carbohydrate-binding collectin family and they have a role in the host defence of the lung (Crouch & Wright 2001). SP-A is also important in the organization and function of the surfactant complex regulating surfactant recycling and secretion. SP-B and SP-C are highly hydrophobic small peptides that confer surface tension-lowering properties and are important for the adsorption and spreading of the surfactant (Whitsett & Weaver 2002). They are the only surfactant proteins in the animal-based surfactant replacement preparations used in RDS, as the purification process removes the hydrophilic proteins SP-A and SP-D from the completed product.

Surfactant components are synthesized, secreted and recycled by type II epithelial cells in the alveolus (Rooney et al 1994). With the exception of SP-A, surfactant proteins are synthesized in polyribosomes, modified in the endoplasmic reticulum, Golgi apparatus and multivesicular bodies and stored in lamellar bodies before secretion. Surfactant phospholipids are synthesized in the endoplasmic reticulum, transported through the Golgi apparatus into multivesicular bodies and packaged into lamellar bodies. After exocytosis of lamellar bodies, surfactant phospholipids, in the presence of SP-A, SP-B and Ca2+, are organized into a lattice structure called tubular myelin (TM), which forms a lipid-rich layer at the air-liquid interface of the alveolus. Most of the extracellular surfactant is taken up by type II cells, catabolized and transported into lamellar bodies for recycling. Alveolar macrophages also take part in the catabolization process of the surfactant components (Gurel et al 2001).

Because only SP-A (the most abundant SP) and SP-B (essential for postnatal lung function) gene polymorphisms were studied for their associations with RDS in the present project, the review of the literature on surfactant proteins concentrates mainly on SP-A and B.

2.3 Pulmonary surfactant protein A

2.3.1 Structure of SP-A

Sequence comparison of the cDNAs of SP-A isolated from human and several animal species shows extensive homology (White et al 1985, Boggaram et al 1988, King et al 1989, Korfhagen et al 1992). The primary structure of SP-A in humans consists of four

20

structural domains: a short N-terminal segment, a collagen-like region, a hydrophobic neck domain and a carbohydrate recognition domain (CRD) homologous to all calcium-dependent (C-type) mammalian lectins. The combination of a collagen-like region and a lectin domain is one of the defining structural features of the collectin family of proteins shown to function as antibody-independent opsonins (Hoppe & Reid 1994). Human SP-A is synthesized as 26 kDa monomer, and after processing it forms a large multimeric 650 kDa-protein with a molecular weight of 32–36 kDA consisting of 248 amino acids. It exists as an octadecameric (6 trimers) bouquet-like molecule within the alveolus (McCormack et al 1998). Each trimer is suggested to consist of two SP-A1 and one SP-A2 gene products linked through their collagen-like domains (Voss et al 1991). Two cysteine residues in the N-terminal domain are required for multimer formation.

2.3.2 The human SP-A gene

The human SP-A locus on the long arm of chromosome 10 consists of two functional genes, SP-A1 and SP-A2, with a pseudogene in the middle (Hoover & Floros 1998). The functional genes encode isoforms containing minor amino acid differences in the mature protein (Katyal et al 1992). The human SP-A genes differ in their exon-intron organization and are separated by only 40 kb from each other (Hoover & Floros 1998). There is strong linkage disequilibrium between the functional SP-A genes (Floros et al 1996). The schematic construction of the SP-A protein and gene is shown in Figure 3.

Fig. 3. Diagram of the gene and protein structure of SP-A. The Roman numerals designate the coding exons. Both SP-A genes have four coding exons. The thin line represents the introns. The structural domains of the mature SP-A are also shown.

Gene

mRNA

III IV

4.7 kb

Mature SP-A32-36 kDa

III3´5´

ATG TGA

N-terminal

Collagen

Neck

CRDDomains

21

Both SP-A1 and SP-A2 are highly polymorphic. Single-nucleotide polymophisms (SNPs) occur throughout the coding sequence, resulting in amino acid substitutions or silent changes within the genes (McGormick et al 1994). The alleles of the SP-A1 gene are denoted as 6An and the SP-A2 alleles as 1An. The alleles within each gene differ from each other in one or several SNPs in the coding exons, as shown in Table 1 and Table 2. There are differences between SP-A1 and SP-A2 alleles in their ability to stimulate TNF-α production in cell lines in vitro (Wang et al 2000), but the impact of allelic differences on the surfactant function and immune responses of the mature SP-A are unknown. There is also splicing variability in the 5´untranslated region (UTR) and sequence variability in the 3´-UTR of SP-A1. According to an in vitro study the human 3´-UTR variants play a differential role in SP-A mRNA levels and in the response to dexamethasone (Wang et al 2003a).

Table 1. SNP combinations of the most frequent alleles of SP-A1 in a Finnish population of newborn infants (Rämet et al 2000).

SP-A1 alleles frequency Val19Ala T/C

Val50Leu G/C

Pro 62 A/G

Thr133 A/G

Arg219Trp C/T

6A 0.04 C (Ala) C (Leu) G G C (Arg) 6A2 0.60 T (Val) G (Val) A A C (Arg) 6A3 0.28 T (Val) C (Leu) A A C (Arg) 6A4 0.09 T (Val) C (Leu) G A T (Trp)

Table 2. SNP combinations of the most frequent alleles of SP-A2 in a Finnish population of newborn infants (Rämet et al 2000).

SP-A2 alleles frequency Asn9Thr A/C

Ala91Pro G/C

Ser140 C/T

Gln223Lys C/A

1A 0.04 C (Thr) C (Pro) C C (Gln) 1A0 0.57 A (Asn) G (Ala) C C (Gln) 1A1 0.16 C (Thr) G (Ala) T A (Lys) 1A2 0.13 C (Thr) G (Ala) C C (Gln) 1A3 0.03 A (Asn) G (Ala) T A (Lys)

2.3.3 Processing and function of SP-A

SP-A is the most abundant surfactant protein, accounting for approximately 50 % of total protein. Apart from alveolar type II cells, SP-A mRNA and protein are expressed in epithelial Clara cells and serous cells of tracheo-bronchial glands (Khoor et al 1993, Wong et al 1996). SP-A expression has also been localized within the Eustachian tube epithelium, and the mature protein has been found in the granules of epithelial cells (Paananen et al 2001a). The extrapulmonary expression of SP-A may be species-specific (Rubio et al 1995). In humans, SP-A expression is considered to be limited to the respiratory system (Madsen et al 2003). SP-A is synthesized as a preprotein and is

22

extensively modified in the endoplasmic reticulum (ER) and the Golgi apparatus. Only approximately 1% of the surfactant in lamellar bodies consists of SP-A, and there is evidence that it is secreted into the alveolar space via several routes (Ikegami et al 1992).

Most previous studies have focused on the role of SP-A in surfactant structure and function, but further recent research has shown that the main function of SP-A is in pulmonary host defence (Hoppe & Reid 1994). In vitro and in vivo studies have shown SP-A to be important for the formation of tubular myelin (Poulain et al 1992). It improves the surface activity of surfactant by binding to surfactant lipids and enhancing their adsorption at the air-liquid interface (Hawgood et al 1987). It also has a regulatory role in the secretion of surfactant, and specifically, of phospholipids by type II cells (Rice et al 1987).

Surprisingly, studies on SP-A gene-targeted mice (SP-A -/-) revealed no apparent abnormalities in lung function suggesting that other mechanisms take over in the absence of SP-A (Korfhagen et al 1996, Ikegami et al 1998). However, the lungs of SP-A (-/-) mice had markedly decreased tubular myelin formation and cleared bacterial antigens less efficiently than SP-A wild-type mice. According to a recent study on gene-targeted mice SP-A is also necessary for the pulmonary uptake and metabolism of DPPC (Jain et al 2003).

Both lung collectins, SP-A and SP-D, bind to lipids and carbohydrates and interact with specific cell surface receptors, micro-organisms and leukocytes as well as modulate the functions of phagocytic cells in vitro and in vivo (Haagsman 1998, Korfhagen et al 1998, Crouch & Wright 2001). Studies on mice have shown that SP-A inhibits cytokine production in inflamed lung (Chabot et al 2003). It has been suggested that the surfactant-related and immunomodulatory functions of SP-A overlap, and disturbances in either of them lead to altered homeostasis of the lung (Hawgood & Poulain 2001).

2.4 Pulmonary surfactant protein B

2.4.1 Structure of SP-B

The structural organization of SP-B mRNA is highly conserved among mammalian species (Whitsett et al 1995). Translation of human SP-B mRNA results in the synthesis of a precursor protein in which the mature SP-B protein is flanked by N- and C-terminal peptides. Both of these peptides contain saposin-like domains that are identified by six conserved cysteine residues (Patthy 1991). These domains may facilitate the processing of the SP-B preproprotein into a mature peptide during its transit through the secretory pathway (Lin et al 1996). Complete processing of a 381 amino acid precursor into a mature 79 amino acid peptide is considered to take place exclusively in type II epithelial cells in the alveoli (Weaver & Whitsett 1989). Recently SP-B mRNA and protein were shown to be expressed in the Eustachian tube epithelium (Paananen et al 2001b), but the phospholipid composition of the surfactant is different from that of the lungs (Paananen et al 2002). Mature SP-B is a small homodimer of approximately 18 kDa with a reduced

23

molecular mass of 8 kDA, which comprises 1 to 2 % of surfactant (Weaver 1998). The precise conformation of SP-B is not known, and the structural domains are presumed to consist of five amphipathic helices (Weaver & Conkright 2001). Figure 4 shows the gene and protein structure of SP-B.

Fig. 4. Gene and protein structure of SP-B. The Roman numerals designate the coding exons. The thin line represents the introns. The branched structures are potential glycosylation sites.

2.4.2 The human SP-B gene

The human SP-B gene on chromosome 2p12-p11.2 consists of 11 exons with the coding sequences for the mature SP-B in the exons 6 and 7 (Pilot-Matias et al 1989). Within the lung the gene is expressed in type II cells in the alveoli and in bronchiolar Clara cells. A number of polymorphisms have been identified in the SP-B gene (Glasser et al 1987, Floros et al 1995, Lin et al 1998, Lin et al 2000). A coding polymorphism at the end of exon 4, T/C +1580, leading to substitution of threonine for isoleucine at codon 131, is presumably the most remarkable, because it forms a consensus signal for N-terminal linked glycosylation at this site of the SP-B propeptide. This, in turn, may affect the processing, folding and secretion of the surfactant protein (Haataja et al 2000, Wang et al 2003b). A length polymorphism with deletion or insertion of specific sequence motifs in intron 4 of SP-B has been characterized as well as SNPs in the 5´and 3´ untranslated regions (Pilot-Matias et al 1989), but the impacts of these variations on SP-B mRNA and on mature SP-B are unknown.

Gene

mRNA2 kb

I II III IV V VI VII VIII IX X

11 kb

PreproSP-B

Signal

N-terminal

N-flanking

C-terminal

45 kDa

ProSP-BY

Mature SP-B8 kDa

5´ 3´

ATG TGA

(Y)

Mature peptide

24

2.4.3 Secretion and function of SP-B

Glycosylated SP-B precursor protein is synthesized and transported from ER into the Golgi apparatus and multivesicular bodies and packaged into lamellar bodies with surfactant phospholipids (Guttentag et al 1998). The contents of lamellar bodies are then secreted into the alveoli, where they interact with SP-A to form tubular myelin.

Hydrophobic SP-B is the only surfactant-associated protein indispensable for postnatal lung function and survival (Whitsett & Weaver 2002). SP-B is critical to the organization of surfactant phospholipids in lamellar bodies and the generation of the surfactant mono- and bilayers that reduce surface tension. It is essential for maintaining the integrity of alveolar epithelial cells, for tubular myelin formation (Whitsett et al 1995, Pryhuber 1998) and for the processing of SP-C (Vorbroker et al 1995). It promotes surface adsorption of DPPC and enhances the stability of phospholipid molecules (Hawgood et al 1998).

SP-B may also possess anti-inflammatory properties (Miles et al 1999), protect against oxygen-induced lung injury (Tokieda et al 1999) and regulate the immune responses in the airways (van Iwaarden et al 2001). A recent study on transgenic mice showed that SP-B-B protects against endotoxin-induced lung inflammation by enhancing surfactant function, decreasing the influx of inflammatory cells and lowering cytokine levels (Epaud et al 2003).

2.5 Regulation of SP-A and SP-B expression

Expression of the genes encoding SPs is developmentally and hormonally regulated, and complex interactive mechanisms are involved. Several transcription factors that are critical determinants of lung morphogenesis per se are also critical regulators of surfactant gene expression (Boggaram et al 1988, Bohinski et al 1994, DeFelice et al 2003). SP-A gene transcription is initiated in fetal lung in the second trimester and reaches its maximum just prior to birth (Ballard et al 1986). SP-A gene expression in fetal lung tissue is regulated by factors that increase the formation of cyclic AMP by enhancing the rate of type II cell differentiation (Mendelson et al 1998, Gonzales et al 2002).

Corticosteroids appear to have a complex biphasic dose-dependent effect on SP-A gene expression. According to in vitro studies, when present at low levels they cause induction of SP-A gene transcription, but in elevated concentrations, they temporarily decrease SP-A mRNA stability (Boggram et al 1989). In addition the SP-A1 and SP-A2 genes are regulated differently by corticosteroids (Kumar & Snyder 1998). It has been proposed that with increased lung maturation there is a decrease in the responsiveness of the fetal lung to the inhibitory effects of corticosteroids on SP-A gene expression (Mendelsson et al 1998). The physiological role of endogenous corticosteroids in the regulation of SP-A gene expression in fetal lung remains unclear, however.

During the past few years inflammatory mediators have been studied widely for their role in lung development and in the regulation of surfactant synthesis. In fetal lung

25

explants, interferon-γ increases total SP-A mRNA, but there are marked differences in the response between explants (Karinch et al 1998). In the human pulmonary adenocarcinoma cell line recombinant interleukin-6 (IL-6) initiates the synthesis of SP-A mRNA and protein (Shimoya et al 2000). Recently, inhaled nitric oxide (NO) was shown to increase both SP-A and SP-B mRNA content in newborn lambs (Stuart et al 2003). The effect of cytokines on surfactant synthesis seems to be gestational age-dependent: in very immature lung, IL-1 strongly upregulates the expression of SP-A, whereas towards term gestation TNF-α, lipopolysaccharide (LPS), and the combination of cytokines and LPS suppress the expression of SP-A (Odom & Ballard 1997, Bry et al 1997, Väyrynen et al 2002).

Expression of the SP-B and SP-C genes begins before the differentiation of type II epithelial cells. Both SP-B and SP-C mRNA have been detected as early as the end of the first trimester. By 24 weeks of gestation, the SP-B mRNA level has reached 50% of that of adults (Liley et al 1989), while the level of mature SP-B is less than 5% of the adult level at that time (Beers et al 1995). Similarly to SP-A expression, different hormonal, inflammatory and growth factors regulate SP-B gene expression according to in vitro studies. Corticosteroids increase SP-B mRNA and enhance the accumulation of SP-B mRNA in human fetal explants (Liley et al 1989). IL-1 induces the synthesis of SP-B in immature lung (Väyrynen et al 2002), whereas TNF-α decreases SP-B mRNA and SP-B protein (Pryhuber et al 1996).

Expression of surfactant protein genes fades when type II cells undergo terminal differentiation into type I cells, which constitute most of the gas-exchanging surface in the alveolus.

2.6 Predicting the risk of RDS with surfactant component analysis

Surfactant components secreted by the fetal lung accumulate in amniotic fluid, and the proportions of phospholipids change during fetal development. The lecithin-to-sphingomyelin (L/S) ratio and the presence of phosphatidylglyserol (PG) and phosphatidylcholine (PC) in amniotic fluid have been used to predict lung maturation (Gluck & Kulovich 1973, Hallman et al 1976, Torday et al 1979). In pregnancies complicated by pre-eclampsia, for example, their value in predicting RDS is unreliable. Additionally amniocentesis required for the analysis may lead to acute complications threatening the fetus (Hallman 2002). Gastric aspirate for surfactant analysis immediately after birth have been used for the prediction of RDS (Fiori et al 2001), but this has not been adopted into clinical use.

Experimentally, the levels of SP-A in amniotic fluid have been used to predict the risk of RDS. SP-A secretion increases sharply with advancing gestational age (Hallman et al 1988). Cho et al measured SP-A levels in cord blood sera from 48 infants born at gestational ages < 32 weeks. They found that a SP-A level over 10 ng/ml in cord blood was significantly related to a non-RDS outcome (Cho et al 2000). In another study the SP-A content in sera from newborn infants was measured with an enzyme-linked immunosorbent assay system. The level of serum SP-A was shown to increase with

26

advancing gestation, but no definite cut-off level could be pointed out because of inadequate sample size (Kaneko et al 2001).

Mature SP-B is first detectable in amniotic fluid at 31 weeks of gestation and its specificity for the prediction of RDS is considered low (Pryhuber et al 1991).

2.7 Clinical picture of RDS in preterm infants

Respiratory distress syndrome in a newborn develops within the first few hours after preterm birth. It manifests as deficient gas exchange and progressive respiratory failure, unless effective treatment is instituted. Infants with RDS present with clinical signs of respiratory distress that include tachypnea, grunting, retractions and cyanosis accompanied by increasing oxygen requirements. Physical findings include rales, poor gas exchange, use of diaphragmatic and other accessory muscles of breathing, nasal flaring, and abnormal patterns of respiration that may be complicated by apnoea. Chest radiograph is characterized by diffuse reticulogranular infiltrates, atelectasis and air bronchograms, often progressing to severe bilateral opacity. However, radiographic patterns in RDS are variable and may not reflect the degree of respiratory compromise.

The infant attempts to avoid alveolar collapse by prolonging and increasing expiratory pressures by breathing against partially closed glottis, which causes the grunting noise characteristic of RDS, but also seen in other respiratory disorders. The increasing oxygen requirements and the need for continuous positive airway pressure (CPAP) or ventilatory support with continuous distending pressure persist for two to three days or even longer. The clinical course depends on the severity of RDS and the maturity of the infant at birth. In uncomplicated RDS, typically seen in more mature infants, recovery is rapid during the first two weeks of life. In most cases, the surfactant content increases and reaches the normal level by one week of postnatal age (Hallman et al 1994). The administration of early exogenous surfactant has alleviated the course of “classical” RDS. The most premature infants are still at greatest risk for severe RDS and frequently develop complications, including air leak, infection, patent ductus arteriosus and central nervous system hemorrhage, all of which contribute to prolonged requirements for oxygen and ventilatory support.

2.8 Pathophysiology of RDS

Pathologic findings include atelectasis, high-permeability pulmonary edema, pulmonary vascular congestion, pulmonary hemorrhage and direct injury to the respiratory epithelium by inflammatory cells. The previous term “hyaline membrane disease” refers to the histologic findings of pathognomic hyaline membranes in the gas-exchanging alveoli seen post-mortem. Hyaline membranes are fibrin-rich clots representing plasma components and pulmonary epithelial debris. In uncomplicated, mild RDS these membranes disappear quickly. In ventilated newborns, the healing process is altered and

27

delayed, and hyaline membranes may remain prominent. In most severe cases of RDS, progressive scarring, fibrosis and emphysema of the alveoli and airways lead to chronic lung disease (CLD).

Discovery of high surface tension in the terminal airways of infants dying of hyaline membrane disease was the first demonstration of surfactant deficiency (Avery & Mead 1959). Subsequently, qualitative and quantitative abnormalities in the surfactant complex were demonstrated in infants who developed RDS after preterm birth (Hallman 1982). The importance of the surfactant complex and some of its critical components was further confirmed when surfactant given at birth or in established RDS ameliorated respiratory difficulty considerably (Jobe 1983). The ultimate cause of RDS is a failure to generate an adequate surfactant layer at the alveolar lining after preterm birth. The high capillary permeability allows plasma-derived components and edema fluid to deteriorate surfactant function. Alveolar ventilation is further impaired and RDS complicated by the relatively weak respiratory muscles and the compliant chest wall of the premature infant. Pulmonary vascular resistance is increased due to diminished oxygenation and respiratory and metabolic acidosis. Right-to-left shunting through the ductus arteriosus and foramen ovale and intrapulmonary ventilation-perfusion mismatch further aggravate hypoxemia. The causes of the striking variations in the severity of RDS at a given state of gestation remain largely unknown, however.

2.9 Epidemiology of RDS

Despite the progress made in perinatal care, RDS continues to be the major reason for increased mortality and morbidity among preterm infants. The surviving premature infants are, furthermore, at risk of long-term morbidity, i.e. chronic lung disease of prematurity and neurosensory disorders (Dammann & Leviton 1997, Hallman 1999).

RDS-related mortality represents approximately 10–15% of all neonatal deaths. In the United States, infant mortality (mortality from birth up to one year of age) due to RDS declined from 0.8 in 1987 to 0.4 per thousand live births in 1995 (Malloy et al 2000). In Finland, the overall infant mortality rate declined from 5.6 to 3.4 per thousand live births, perinatal mortality (PNM) (stillbirths and deaths before one week of postnatal age) from 7.8 to 5.6 per thousand births and early neonatal mortality (deaths during the first week of life) from 3.1 to 2.1 per thousand live births between 1990 and 1999 (Gissler et al 2002). There are no published studies on RDS-related mortality in Finland.

The published incidence figures of RDS vary widely. In a Swedish study, the incidence of RDS in newborn infants was only 3.3 per thousand (Hjalmarson 1981), while in an American study it was 18.3 per thousand live-born infants (Robertson et al 1992). In Finland the average incidence of RDS was 6 per thousand live-born infants and 120 per thousand infants born preterm (Koskinen et al 1999). Studies on the influence of advanced perinatal care and the increasing number of multiple gestations to the overall incidence of RDS are scanty, and the analyses are limited to very or extremely low birth weight infants (ELBW, < 1000 g) or singletons (Horbar et al 2002). In Finland, the overall incidence of RDS among those ELBW surviving for more than 12 hours was 760

28

per thousand (Tommiska et al 2001). Due to the delayed maturation of their lungs male infants have a higher incidence of RDS and higher RDS-related mortality compared to females (Farrell & Wood 1976, Khoury et al 1985).

2.10 Clinical risk factors for RDS

RDS is associated with preterm birth at a time when the biochemical and structural functions of the lung are still underdeveloped. Figure 5 shows an overview of complex components involved in the susceptibility to RDS.

Fig. 5. Schematic representation of complex relationships between the environmental, inherent and genetic factors that affect the risk of RDS in preterm infants.

The degree of prematurity is the main factor determining the risk of RDS. The frequency of RDS decreases as a function of the length of gestation, and term infants are very rarely affected by RDS (Usher et al 1971, Wax et al 2002). There is evidence that preterm birth per se results in altered alveolarization that affects lung function (Pelkonen et al 1997, Hjalmarson & Sandberg 2002).

The preterm infant is more prone to asphyxia, hypoxia, hypothermia and hypotension, all of which are likely to impair surfactant synthesis and to increase alveolar capillary permeability. Severe asphyxia is considered to be present when umbilical arterial pH is ≤ 7.10 and/or Apgar score at 5 minutes is ≤ 4. During fetal asphyxia, lung perfusion falls, resulting in ischemic damage to pulmonary capillaries. After recovery pulmonary

RDS

Disease phenotypes

Inherent factors

Other regulatory genes

Surfactant protein genes

Ethnicity GenderInbornindividualcharacteristics

Degree of prematurity Placental diseases Maternal diseases Antenatal and perinatal careBirth order

Inflammatory mediators

Transcriptional genes?

Growth factors

Transcriptionfactors

Genetic variation

Inherentfactors

Composition and functionof surfactant

Environment

29

hyperperfusion occurs and protein-rich fluid leaks out of the damaged capillaries. This inhibits surfactant activity, leading to RDS.

Agenesia of the urinary tract and subsequent deficient urination by the fetus, lack of fetal breathing and oligohydramnion are conditions leading to diminished lung fluid production and subsequently, to pulmonary hypoplasia and associated RDS.

Spontaneous preterm birth, induced preterm birth due to maternal or fetal condition, and preterm premature rupture of membranes (PPROM) are the main conditions leading to preterm birth. Subclinical infection with intact membranes is presumably a major reason for spontaneous, or idiopathic, preterm birth (Moutquin 2003).

2.10.1 Preterm premature rupture of membranes and intrauterine infection

Preterm premature rupture of membranes is defined as premature rupture of fetal membranes before 37 comleted weeks of gestation. Furman and coworkers compared 376 infants with PPROM to 1950 infants without PPROM and found no differences in morbidity from RDS between the groups. They concluded that RDS was affected by prematurity itself rather than by the occurrence of PPROM (Furman et al 2001).

Subclinical infection has been implicated in the mechanism of membrane rupture. Up to 50%, or even more, of fetuses delivered before 30 weeks of gestation have chronic chorioamnionitis, as proved by histologic and microbiologic examinations of amniotic fluid and fetal membranes (Goldenberg et al 2000). Intrauterine infection activates cytokines and other inflammatory mediators that trigger the cascade leading to preterm labour (Gomez et al 1995, Greci et al 1998). The role of infection in the susceptibility to RDS is controversial. Severe and chronic infection is generally considered deleterious. Very early onset of respiratory distress is usually a manifestation of perinatally acquired severe infection leading to ARDS (Becroft et al 1976). However, PPROM and associated infection may expose the fetus to stress and increase the concentrations of different hormones, cytokines and growth factors, which, in turn, induce surfactant synthesis, accelerate lung maturation and protect from RDS (Lassus et al 1999, Shimoya et al 2000, Kallapur et al 2001, Sims et al 2002). Infants born to mothers with histologic chorioamnionitis had a decreased incidence of RDS, but an increased incidence of chronic lung disease, compared to infants of mothers without chorioamnionitis (Watterberg et al 1996). Adjustment for gestational age was not conducted due to the small sample size of 52 preterm infants. Signs of infection and increased cytokine production associated with RDS have also been shown in preterm births with intact membranes (Hitti et al 2001).

30

2.10.2 Intrauterine growth

Studies on the effect of intrauterine growth on RDS are controversial. Restricted growth of the fetus due to placental insufficiency induces hypoxemia and increases the production of corticosteroids, which, in turn, may facilitate pulmonary maturation (Laatikainen et al 1988, Braems et al 1998). According to Ley and coworkers the effect of intrauterine growth restriction (IUGR) on the risk of RDS depends on the gestational age of the infant. In very premature infants (under 28 weeks of gestation) it increased the risk, whereas between 29 and 32 weeks it decreased the risk (Ley et al 1997). Piper and coworkers reported no differences in the rate of RDS between small for gestational age (SGA) and appropriate for gestational age (AGA) infants after stratification for gestational age (Piper et al 1996). In a large study based on infants delivered at a regional hospital in Dallas, USA between 1988 and 1996, preterm SGA singleton infants had increased risk of RDS compared to AGA infants (McIntire et al 1999).

2.10.3 Maternal diseases

Some studies report a higher incidence of RDS in preterm infants of hypertensive than non-hypertensive mothers, which could, however, be explained by the absence of labour before caesarean section (Tubman et al 1991, Wolf et al 1993). In another study preterm infants of mothers with pre-eclampsia had less RDS compared to infants of non-hypertensive mothers (Shah et al 1995). In a matched cohort study of 225 infants of pre-eclamptic women maternal pre-eclampsia per se did not accelerate lung maturation or have any other salutary effect on the postnatal course of infants born at 24 to 35 weeks of gestation (Friedman et al 1995). Expectant management of pre-eclampsia remote from term in selected patients is recommended to reduce the risk of RDS and other neonatal complications due to preterm birth associated with pre-eclampsia (Friedman et al 1999).

In animal models, insulin delayed the maturation of alveolar type II cells and decreased the proportions of saturated phophatidylglyserol (PG) in surfactant (Gross et al 1980, Ojamo et al 1990). Surfactant protein A mRNA levels were decreased in diabetic pregnancy (Moglia & Phelbs 1996). Abnormal surfactant synthesis, particularly a delay in the appearance of PG, has also been reported in human infants of diabetic mothers (Ojomo et al 1990). The most recent studies report normal surfactant patterns and fetal lung maturation in normoglycemic, metabolically controlled diabetic mothers (Piazze et al 1999). The improved control of the disease during pregnancy makes it possible to postpone the delivery until term and thus RDS is no longer of such great concern.

31

2.10.4 Mode of delivery

Caesarean section carried out before the onset of labour is considered to increase the risk of RDS (Reed 1978, Goldberg et al 1981, Ziadeh & Badriat 2000). Infants born by caesarean section have a larger residual volume of lung fluid and a smaller residual capacity and, consequently, secrete less surfactant into the alveolar space. During vaginal delivery, the chest of the infant is squeezed and part of the fetal lung fluid is removed. The adrenergic stimulation associated with vaginal labour also releases surfactant into the airways (Walters & Olver 1978). Cephalic presentation in vaginal deliveries is regarded optimal. Especially in twin deliveries breech presentation of the presenting infant is considered a risk factor for asphyxia and adverse outcome, and caesarean delivery is usually preferred. Trial of vaginal delivery is safe, however, in twin deliveries with the presenting twin in cephalic presentation, and the non-presenting twin in cephalic or breech presentation (Winn et al 2001, Caukwell & Murphy 2002).

2.10.5 Twins and higher order gestations

2.10.5.1 Incidence of twins

The advanced use of ovulation induction and other infertility treatments during the past decade has increased the incidence of twins and higher order multiples. In the United States, between 1980 and 1999, the overall twin and multiple birth ratio increased by 59%. At the same time twin and multiple infant mortality declined more than that of singletons in all birth weight categories. Consequently, the mortality rates of very low birth weight (VLBW) and low birth weight (LBW) infants were lower among multiples than among singletons (Russell et al 2003). Similar decline in the PNM of preterm twins under 32 weeks of gestation was seen in a cohort population study including births between 1985 and 1996 in Canada (Joseph et al 2001). In Finland, the incidence of multiple births was 11 per thousand in the 1980s, but increased up to 17 per thousand in 1998 (Gissler et al 2002).

Twin and higher order gestations have a much higher rate of preterm births than singleton pregnancies. Approximately 40–50% of twins are born prematurely, accounting for 14–15% of all preterm births according to a previous study (Powers & Kiely 1994). Due to the high rate of preterm births, infants from multifetal gestations have much higher postnatal morbidity compared to singletons.

The greater risk of preterm delivery for twins is considered to be due to uterine over-distension (Neilson et al 1988, Rouse et al 1993). Decreased uterine blood flow, hormonal factors and subclinical chorioamnionitis, secondary to ascending infection associated with premature cervical ripening and PPROM, have also been suggested as factors leading to preterm delivery among twins (Romero et al 1990).

32

2.10.5.2 Zygosity of twins

Identical twins originate at some stage during the first two weeks following fertilization, as the early embryo splits into two parts that subsequently develop separately and give rise to two individuals. These two siblings originate from the same zygote and are thus called monozygotic (MZ) twins. They are genetically identical and, theoretically, any difference existing between MZ co-twins should originate from environmental influences. When at least two embryos originate from two different zygotes, they are called dizygotic (DZ) in the case of twins, and polyzygotic in the case of higher order multiples. Genetically, these individuals are no more similar than any other pair of full siblings. The similarity between two DZ twins can vary widely, but 50% genetic similarity is usually considered an average (Bomsel-Helmreich & Mufti 1995).

2.10.5.3 RDS-related morbidity of twins

The growth and development of a twin fetus is affected not only by the intrauterine factors present in a singleton pregnancy but also by the interaction with the co-twin fetus and by the placental, circulatory or other factors associated with the multiple pregnancy per se. The average weight of a newborn twin is less than that of a singleton of the same gestational age. It has been shown that the divergence in growth pattern between singleton and twin infants becomes visible at 30 weeks of gestation (Alexander et al 1998). After that, the fetal growth of twins slows down markedly, their body weight lagging disproportionately behind head circumference. The fetal growth patterns also differ by gender; female twins are lighter than male twins at each gestational age in both same-sex and opposite-sex pairs.

The considerable birth weight discordance between twins is associated with increased mortality and morbidity (Demissie et al 2002). Severely divergent fetal growth increases risk of fetal demise and, consequently, leads to induction of labour and preterm birth (Hollier et al 1999). Discordance in birth weight is most often attributable to a fetal growth restriction, frequently in the second-born twin (Cooperstock et al 2000). Studies comparing the outcome of the smaller versus the larger twin of the pair are contradictory, but the large twin is usually considered to possess greater risks.

MZ twins are assumed to have higher perinatal mortality and morbidity compared to DZ twins or singletons (Benirschke & Kim 1973). According to recent studies the risk for an adverse outcome depends on chorionicity rather than zygosity (Sherer et al 2001), with monochorionic monozygotic twins being at the greatest risk. Monochorionic twins represent approximately two-thirds of all MZ twins. Specifically alterations in placental circulation and the associated discordance of growth in monochorionic twins have serious implications for development and survival (Victoria et al 2001).

Morbidity from RDS has been shown to be similar (Friedman et al 1997, Gardner et al 1995, Wolf et al 1992) or higher (Donovan et al 1998) in preterm twins compared to preterm singletons and corrected for gestational age. When the L/S ratio was used as an index of fetal lung maturation, amniotic fluid studies revealed that pulmonary maturity

33

was biochemically comparable between preterm twin and singleton pregnancies (Winn et al 1992). At term gestation comparison of the mean L/S ratios of twins to those of uncomplicated singleton pregnancies showed that fetal lung maturation was concordant in twins and occurred several weeks earlier in them (Leveno 1984).

Most studies have shown that the second-born, or non-presenting twin, has a higher risk of RDS compared to the presenting twin (Lankenau 1976, Ziadeh & Badria 2000), but only in vaginal deliveries (Arnold et al 1987, Hacking et al 2001). Arnold et al concluded that a predisposition to distress at birth does not explain the excess risk, but the non-presenting twin does not benefit from the salutary effects of labour to the same extent as the presenting twin.

2.11 Genetic predisposition to RDS

A genetic risk for respiratory distress in infancy has been suggested in reports of family clusters of affected infants, studies of different ethnic groups and gender, characterization of infants with inherited deficiency of surfactant protein B and association studies using the candidate gene approach on surfactant protein polymorphisms.

2.11.1 Epidemiological evidence

According to a study published in 1965, mothers who gave birth to LBW infants could be divided into low- and high-risk groups based on the risk of RDS in the offspring (Graven & Mesenheimer 1965). An unknown genetic factor was suggested to be involved in the susceptibility to the disease, as the two groups did not differ in any perinatal characteristics.

Lankenau studied retrospectively 111 families with an infant affected by RDS. The risk for maternal half-sibs was the same as that for full-sibs, while the risk for paternal half-sibs was minimal (Lankenau 1976). Infants of maternal aunts possessed an excess risk of RDS as well. The author concluded that there is a genetically determined maternal factor predisposing infants to RDS. There is also recurrence of low birth weight in siblings unrelated to other predisposing perinatal factors (Khoury et al 1989). According to a study of 400 mothers by Nagourney and coworkers, preterm infants born to women with a previous preterm infant affected by RDS are at an increased risk of RDS, suggesting an important genetic predisposing factor (Nagourney et al 1996).

A twin study showed a higher concordance rate (i.e. frequency of twin pairs where both were affected) of RDS in MZ (85%) compared to DZ (44%) twin pairs. The study was done in a rather small study population of 31 twin pairs of known zygosity, including twin pairs of opposite sexes (Myriathopoulos et al 1971).

Ethnicity affects the risk of RDS. Even though black infants have a higher rate of preterm births, the incidence of RDS is lower compared to white infants of the same gestational age (Farrell& Wood 1976, Hulsey et al 1993). Black infants are considered to

34

be more responsive to antenatal corticosteroid administration and having accelerated lung maturation, or else there are qualitative differences in surfactants between races. However, the difference in the susceptibility to RDS between races was independent of the L/S ratio (Richardson & Torday 1994). RDS-related neonatal mortality has improved more for white than for black infants with very low birth weight after the introduction of surfactant therapy for RDS (Hamvas et al 1996).

2.11.2 Hereditary SP-B and SP-C deficiency

It was noticed a decade ago that term infants unable to produce SP-B due to a genetic defect develop fatal neonatal respiratory disease unresponsive to exogenous surfactant (Nogee et al 1993). Subsequent work demonstrated that these infants were homozygous for a frame-shift mutation that involved a 1-bp -deletion and 3-bp -insertion at codon 121 in exon 4 of the SP-B gene (121ins2). This caused a hereditary SP-B deficiency, one form of congenital alveolar proteinosis (CAP) (Nogee et al 1993). The frequency of the 121ins2 mutation was counted to be one allele per 1000–3300 individuals in two cohorts of American population (Cole et al 2000). A variety of other, even more rare mutations in the SP-B gene have been found which may associate with reduced or no mature SP-B (Nogee et al 2000).

Studies on homozygous gene-targeted mice SP-B (-/-) dying of surfactant deficiency and respiratory distress immediately after birth have shown that complete deficiency of SP-B disrupts tubular myelin and lamellar body formation and the processing of SP-C (Clark et al 1995, Vorbroker et al 1995). Heterozygous SP-B +/- mice have decreased lung compliance and air trapping, but the processing of SP-B and lamellar body formation remain sufficient (Clark et al 1997). Genetic and biochemical studies of heterozygous human infants and adults as well as gene-targeted animal studies suggest that approximately 50% of normal SP-B synthesis may be sufficient for normal pulmonary function at birth (Clark et al 1997, Klein et al 1998).

SP-B influences critically the intracellular processing of SP-C, since, in the absence of SP-B, no mature SP-C is recovered in the airways (Melton et al 2003). Abnormal processing of SP-C may indeed prove to be involved in the pathogenesis of severe respiratory failure, since a mutation in the SP-C gene was recently associated with familial interstitial lung disease (Nogee et al 2001). Gene-targeted SP-C (-/-) mice do not develop neonatal respiratory distress, but the stability of the surfactant isolated from them is significantly decreased (Glasser et al 2001).

Surfactant containing only SP-C has been shown to function well in premature lambs and rabbits suffering from RDS (Davis et al 1998), but SP-C is not able to compensate for a lack of SP-B in humans.

35

2.11.3 Allelic association studies

The first allelic association study of RDS suggested the presence of HLA-linked susceptibility genes, with over-representation of the HLA antigen alleles A3 and B14 in RDS compared to controls (Hafez et al 1989). This finding has not been confirmed in later studies, however. The genes coding for SP-A and SP-B have been regarded as the most plausible candidate genes associating with RDS, because these proteins have been considered essential for surfactant function and metabolism.

Table 3 shows case-control allelic association studies and transmission disequilibrium test (TDT) analyses that have been conducted so far in the human SP-A and SP-B genes, and which have been suspected to associate with RDS and other pulmonary diseases.

Table 3. Studied polymorphisms in the human SP-A and SP-B genes that have been suspected to associate with RDS and other pulmonary diseases

Location of the polymorphism

Study design Study populations Result Comments Reference

SP-B intron 4 Case-control 82 RDS, 137 non-RDS infants

Insertion/deletion variants ↑ in RDS

Study groups not adjusted for gestational age

Floros et al 1995

SP-B intron 4 Case-control 40 RDS, 274 non-RDS infants

12 groups of alleles carrying 3 to 18 motifs

Allele frequency differences evident only in racially mixed groups

Veletza et al 1996

SP-A, SP-B intron 4

Case-control 144 RDS, 98 non-RDS infants

SP-A2 1A0 freq ↑ in RDS in white infants, synergistic association with SP-B intron 4?

Ethnically heterogenous study group, association evident only in subgroups

Kala et al 1998

SP-A, SP-B intron 4

Matched case-control

88 RDS, 88 non-RDS, 225 term non-RDS infants

SP-A 6A2-1A0 freq ↑, 6A3-1A1 freq ↓ in RDS

Associations evident specifically in infants < 32 weeks

Rämet et al 2000

SP-A, SP-B intron 4, SP-B exon 4

Case-control 188 RDS, 500 preterm 475 term non-RDS infants

SP-A 6A2-1A0 freq ↑, 6A3-1A1 freq ↓ in RDS, association restricted according to gestational age and SP-B Ile131Thr genotype

p-values corrected for multiple testing

Haataja et al 2000

SP-A, SP-B Case-control 52 ARDS, 25 at risk, but no ARDS, 46 healthy controls

SP-B exon 4 Thr allele freq ↑ in idiopathic ARDS

Associations evident only in disease subgroups, healthy control group not in H-W equilibrium

Lin et al 2000

36

Table 3. Continued.

Location of the polymorphism

Study design Study populations Result Comments Reference

SP-A Case-control 107 tuberculotic, 71 contacts, 101 healthy control adults

SP-A 6A4 and 1A3 freq ↑ in tuberculosis

Floros et al 2000

SP-B Case-control 243 RDS, 369 non-RDS infants

SP-B intron 4 deletion variant ↑ in white males, insertion variant↑ in black females

Associations evident only in racial subgroups

Floros et al 2001

SP-A Case-control TDT-analysis

249 RDS, 335 non-RDS infants TDT: 32 trios

SP-A 6A2-1A0 freq ↑, 6A3-1A1 freq ↓ in RDS both in black and white infants

Ethnically heterogenous study group in TDT analysis

Floros et al 2001

SP-A TDT-analysis 76 RDS trios SP-A 6A2-1A0 freq ↑, 6A3-1A1 freq.↓ in RDS

p-values corrected for multiple testing

Haataja et al 2001

SP-A, SP-B Case-control 97 COPD, 82 smoker control, 99 non-smoker control adults

SP-B exon 4 Thr allele freq ↑ in COPD

Smoking was a significant confounder

Guo et al 2001

SP-A Matched case-control

86 RSV, 95 control infants

SP-A2 1A3 freq ↑ and 1A freq ↓ in RSV infants

Löfgren et al 2002

SP-B intron 4 Observational 29 preterm infants with deletion or insertion, 111 with wild type

RDS ↑ and BPD ↑ in infants with genetic variations

Makri et al 2002

SP-B intron 4 Matched case-control

118 chronic bronchitis or COPD, 118 matched controls, 118 healthy adults

Size variants ↑ in COPD, specifically in acute respiratory failure

Associations evident only in disease subgroup

Seifart et al 2002

SP-A, SP-B Case-control 84 IPF, 194 healthy control adults

SP-B exon 4 Thr allele freq ↑ in IPF

Selman et al 2003

37

2.11.4 SP-A gene polymorphism

There is evidence that allelic heterogeneity of SP-A is associated with susceptibility to RDS in premature infants. The SP-A1 allele 6A2, the SP-A2 allele 1A0 and corresponding haplotype 6A2-1A0 are most frequently associated with an increased risk of RDS (Kala et al 1998, Rämet et al 2000, Haataja et al 2000, Floros et al 2001, Haataja et al 2001). The problems arising from ethnic heterogeneity and racial differences in allele frequencies can be minimized by studying homogenous populations.

In the Finnish population, the SP-A 6A2 and 1A0 alleles were associated with an increased risk, whereas the 6A3 allele was associated with a decreased risk of RDS, in very premature infants under 32 weeks of gestation (Rämet et al 2000). The associations were further determined by the SP-B Ile131Thr genotype. The homozygous SP-B genotype Thr/Thr interactively with certain SP-A alleles and genotypes was associated with an increased risk of RDS (Haataja et al 2000). A family study based on a transmission disequilibrium test (TDT) confirmed the association between the SP-A gene polymorphism and RDS (Haataja et al 2001). In adults, the SP-A 6A2-1A0 genotype may associate with low levels of SP-A mRNA (Karinch et al 1997). Furthermore, a low SP-A level has been shown to associate with the severity of RDS, and infants who died from RDS had very low levels of SP-A (deMello et al 1993, Moya et al 1994).

2.11.5 SP-B gene polymorphism

A highly polymorphic region containing variable nucleotide tandem repeat sequences within intron 4 of the SP-B gene was studied by Floros and coworkers. They observed differences in the distribution of alleles in different ethnic groups (Veletza et al 1996). Either insertion or deletion of extra motifs was found in preterm infants suffering from RDS compared to controls (Floros et al 1995). A subsequent study showed that this intron 4 polymorphism was associated with RDS interactively with an SP-A gene polymorphism (Kala et al 1998).

Neither of the two SP-B polymorphisms studied, i.e the intron 4 variant or the exon 4 Ile131Thr polymorphism, was associated directly with RDS or prematurity in a Finnish population consisting mainly of singleton infants (Rämet et al 2000, Haataja et al 2001).

2.11.6 Surfactant protein genes associated with other pulmonary diseases than RDS

In adults surfactant protein B polymorphism Ile131Thr have been associated with ARDS (Lin et al 2000), chronic obstructive pulmonary disease (COPD)(Guo et al 2001) and idiopathic pulmonary fibrosis (IPF)(Selman et al 2003). SP-A and SP-D gene

38

polymorphisms have been associated with RSV bronchiolitis (Lofgren et al 2002, Lahti et al 2002) and tuberculosis (Floros et al 2000).

2.12 Prevention, treatment and prognosis of RDS

Prematurity is the most important factor predisposing to RDS and the most efficient way to reduce the incidence of RDS would thus be to prolong gestation. Unfortunately, however, this cannot be always achieved. Furthermore the progress made in antenatal and neonatal intensive care during the past two decades has enabled the survival of even more premature infants.

Corticosteroids were first used antenatally to accelerate the synthesis of surfactant and to prevent RDS in premature infants in 1972 (Liggins & Howie 1972), but for fear of adverse effects they were only adopted into wider use 20 years later. Prophylactic administration of corticosteroids has been shown to reduce the incidence of RDS by 50% in certain gestational age groups (NIH consensus, Crowley 2003). It is debated, however, whether prophylactic corticosteroids have any beneficial effect on preterm twins and higher order multiples (Murphy et al 2002, Crowley 2003). Additionally, corticosteroids have been shown, time- and dose-dependently, to interfere with alveolarization (Jobe et al 1998). Studies on animal models or preterm infants have revealed no beneficial effect of repetitive doses of corticosteroids (Polk et al 1997, McEvoy et al 2002). There are ongoing multicenter studies to find out the optimal dosage of antenatal steroids.

Better control of gestational diabetes, pre-eclampsia and other maternal diseases affecting the pregnancy outcome would also decrease the rate of preterm deliveries (Hauth et al 1995) and the related morbidities of infants born preterm. Expectant management of elective deliveries without labour (Wax et al 2002) and without documented lung maturity reduces the risk of RDS. In a meta-analysis the use of maternal systemic antibiotics in PPROM was associated with a delay in delivery and a reduction in neonatal infections, use of surfactant and oxygen therapy (Kenyon et al 2003). However, another meta-analysis failed to demonstrate any benefit or harm for neonatal outcome in spontaneous preterm labour with intact membranes and without evidence of clinical infection (King & Flenady 2002). The analysis raised concerns about the increased neonatal mortality among those who received antibiotics. In a retrospective study, antenatal corticosteroid significantly decreased the incidence of RDS in preterm infants in the presence of histologic chorionamnionitis (Elimian et al 2000). Associated use of antibiotics and antenatal corticosteroid in PPROM seems to reduce the risk of RDS following preterm delivery (Lovett et al 1997, Vermillion et al 2000), but the optimal timing and dosage of combined therapies remain to be evaluated.

The introduction of surfactant has significantly decreased overall and RDS-related mortality and morbidity in VLBW infants (Hamvas et al 1996). Administration of exogenous surfactant directly into the airways has been a common practice to treat RDS of preterm infants during the last decade and such prophylactic use is increasing among the most premature infants. It has been shown that treatment with exogenous surfactant also stimulates endogenous surfactant synthesis in premature infants suffering from RDS

39

(Bunt et al 2000). Advanced ventilatory support, including administration of continuous distending pressure via the nasal route (De Paoli et al 2003), patient triggered ventilation as well as high frequency ventilation and nitric oxide treatment in elective severely ill infants are all practices used to treat RDS and its complications.

During the 1990s, corticosteroid treatment, mainly dexamethasone in relatively high doses, became into wide use postnatally in order to improve lung function, to facilitate endotracheal extubation and to prevent CLD. Towards the end of the decade, however, follow-up studies provided evidence of abnormal neurodevelopment, especially in infants treated within the first two or three days (Halliday et al 2003), and the use of postnatal corticosteroids became more deliberate. There are only a few studies that have aimed to evaluate the effects of postnatal inhaled steroids on lung function and their role hence remains to be elucidated (Shah et al 2003). Despite the progress in care, CLD, or bronchopulmonary dysplasia (BPD), continues to be a common sequel of lung damage in infants born preterm. Late neurosensory disorders also associate with prematurity and RDS-morbidity (Hallman 1999).

3 Objectives of the present research

The complex relationships between genetic, inherent and environmental factors contribute to the development of RDS in premature infants. Multiple pregnancies involve a considerably higher rate of preterm births than singleton pregnancies, and consequently, RDS is the major cause of morbidity among premature twins and multiples.

The objectives of the present research were:

1. To evaluate the incidence and risk factors of RDS in twins compared to singletons

2. To estimate the changes in the incidence of RDS and RDS-related morbidity due to progress made in antenatal and perinatal care during the past decade

3. To estimate the genetic component of RDS in a twin concordance study comparing monozygous and dizygous twin pairs

4. To evaluate the association of SP-A and SP-B polymorphisms and RDS in twin infants

5. To assess the role of SP-A and SP-B gene variations, and the gene-gene and gene-environment interactions on the risk of RDS in a population of premature twins and higher order multiples compared to singletons

4 Subjects and methods

4.1 The study populations and the study design

The present study is based on an epidemiological and a genetic approach to evaluating the incidence and risk factors of RDS in preterm infants. The data on the study populations are shown summarized in Table 4. Additionally the geographical distribution of the study populations and their proportion to each other are demonstrated in Figure 6. Detailed descriptions of the subjects and the methods are also presented in the original articles (I–V).

42

Table 4. Summary of infants in the individual studies

Study population Total number of live-born

infants

Number of infants with

RDS

Gestational age of RDS infants weeks (sd)

Male/Female ratio of RDS infants

Population-based registry study I All

Singleton infants 850 406 3 945 31.4 (4.0) 0.60 Twin pairs 23 278 1 025 30.0 (3.3) 0.56

Preterm (≥ 24weeks and < 37 weeks) Singleton infants 36 043 3 535 30.8 (3.4) 0.59 Twin pairs 9 498 1 003 29.8 (3.1) 0.56

Regional epidemiologic study II Singleton infants 57 357 373 31.2 (3.7) 0.54 Twin pairs 1 633 a 96 30.8 (2.7) 0.65 Higher-order multiples 19 31.0 (2.5) 0.47

Genetic studies III and IV Altogether 209 preterm twin pairs 418 209 30.5 (2.8) 0.56 Concordance study of twins III, SP-BIle131Thr polymorphism 100 preterm twin pairs 200 149 30.6 (2.8) 0.52 SP-A polymorphism IV 198 preterm twin pairs 396 189 30.4 (2.7) 0.58

Interaction between SP-A and SP-B V Singleton infants 441 119 30.5 (3.3) 0.66 Twin pairs 366 172 30.4 (2.8) 0.59 Higher-order multiples 114 73 30.8 (1.9) 0.64

a Twin and higher-order multiples altogether

43

Fig. 6. Geographical distribution of the study populations. Retrospective twin pairs for the genetic studies were recruited through each of the five university hospitals (Helsinki, Kuopio, Oulu, Tampere and Turku). The diagram demonstrates the study populations in the individual genetic studies III–V in proportion to each other.

4.2 Epidemiological approach

The population-based registry study (I) was based on data from the Finnish Medical Birth Registry. All twin and singleton pregnancies and births registered since the founding of the register in 1987 until the end of the year 2000 were evaluated. The register covers over 98% of all births in Finland, and it includes data on pregnancy, delivery and the condition of all live-born neonates and stillborn fetuses weighing 500 gram or more or having gestational age of 22 weeks or more. The whole study population consisted of 853 831 singleton and 23 691 twin infant births. Specifically 36 043 singleton and 9 498 twin infants born preterm were evaluated. The diagnosis of RDS was coded according to the International Classification of Diseases ninth revision (ICD9) (769) in 1987–1990, dichotomously as present/not present in 1991–1995 and as ICD10 code P 22.0 in 1996–2000. Gestational age, birth year, mode of delivery, gender, birth order, birth weight and intra-pair difference in the birth weight of twins were recorded. The concordance of RDS, i.e. both twins being affected, was compared between same-sex and opposite-sex pairs.

The regional epidemiologic study (II) was a population-based survey of the incidence, clinical course and outcome of respiratory distress syndrome in the 1990s. The study population consisted of 58 990 infants, 57 357 singletons and 1 633 multiples, born in the catchment area of Oulu University Hospital, in Northern Finland, between January 1st,

Population-based registry study I

Regional epidemiological study II

SP-B III

SP-A IV

Interaction of SP-A and SP-B V

44

1990 and December 31st, 1999. The clinical characteristics of the neonates admitted to intensive care were prospectively recorded on admission, and the diagnosis was coded in each case on discharge. RDS was defined according to the published clinical (grunting respiration, retractions, nasal flaring, need for supplemental oxygen for > 48 h, or need for exogenous surfactant therapy), radiographic (diffuse reticulogranular pattern and air bronchograms) and pathologic criteria (diffuse atelectasis and hyaline membranes). The hospital records of all infants born alive before 33 completed weeks of gestation and all others suspected or documented to have RDS were individually reviewed and evaluated by the authors of the original manuscript. The RDS-associated morbidity of infants and outcome up to one year of age were recorded.

4.3 Genetic studies

The twin study (III) on the concordance of RDS and the role of SP-B gene variation in MZ and DZ twins consisted of twin pairs born in Finland between 1987 and 1996. The power calculations for the size of a study population needed were hampered by the gestational age dependency of RDS. Our calculations were based on a previous twin study by Myrianthopoulos et al 1971. They found significantly higher concordance of RDS in MZ (0.85) compared to DZ twin pairs (0.44), suggesting a genetic contribution to RDS. They had, however, included both same-sex and opposite-sex DZ twin pairs. Based on our calculations, in order to avoid the confounding effect of gender, we would have needed 90–100 same-sex twin pairs in order to achieve a concordance difference of the same degree as in the previous study with α = 0.05 and β = 0.20. This number of preterm twin pairs was estimated to be attainable only by recruiting twin pairs both prospectively and retrospectively.

Among twin pairs of the same sex born in Finland before 37 weeks of gestation in 1987–1996, those with at least one of them suffering from RDS were identified from the Finnish Medical Birth Register. Altogether 351 twin pairs were found, of whom188 twin pairs born in one of the five university hospitals in Finland were studied. The clinical data of the twin pairs was reviewed and the diagnosis of RDS was confirmed. None of the newborns were treated with surfactant prophylactically. Of the 188 pairs, 26 were excluded because of stillbirth or the early death of at least one of the infants before the diagnosis of RDS was established (n = 12), the parents were not of Finnish origin (n = 4) or the diagnosis of RDS was not confirmed (n = 10). A total of 162 twin pairs were eligible for analysis. Clinical data for several neonatal variables (gestational age, gender, birth order, birth weight, postnatal surfactant use and duration of oxygen therapy) were obtained. Maternal records were also reviewed for age, parity, underlying diseases, medication, cause of preterm birth, antenatal glucocorticoid treatment, and mode of delivery. Histological findings of the placenta and fetal membranes were recorded when available. The difference in the concordance rates of RDS between monozygotic (MZ) and dizygotic (DZ) twin pairs, as presumed evidence for genetic influence, was evaluated. The parents of the twin pairs were asked to obtain double buccal smear samples for DNA analysis from their children according to written instructions and to

45

send them to the research laboratory for genetic analysis. Altogether 100 preterm twin pairs of same-sex were successfully genotyped for the SP-B polymorphisms. The 62 pairs with no DNA available for analysis did not differ in population characterstics from the 100 twin pairs analyzed.

The genetic association study on surfactant protein A gene (IV) included premature twin infants born in Finland during 1987–2000. The study was designed to evaluate the associations between SP-A allelic variants and RDS in a population consisting of 198 premature twin pairs. From 102 retrospectively recruited preterm twin pairs born during 1987–1996, buccal smear samples were obtained and analyzed. In addition, 96 premature twin pairs born in the University Central Hospitals of Oulu and Tampere and Seinäjoki Central Hospital during 1997–2000 were prospectively recruited, and umbilical cord blood samples were obtained and analyzed for SP-A1 and SP-A2 SNPs. The medical records of all preterm twins were reviewed to confirm or exclude the diagnosis of RDS based on clinical, radiographic and pathologic criteria.

The interaction by the SP-B gene and plurality on the association between the SP-A gene and RDS (V). Based on our earlier observations in different sets of Finnish neonates, the allelic associations between the SP-A or SP-B genes and RDS were not similar throughout the population (Ramet et al 2000, Haataja et al 2000). The conflicting observations made in singleton and twin subpopulations of the similar ethnicity are likely to be affected by multiple birth. Furthermore, the published studies on differences in the incidence of RDS between singletons and twins matched for gestational age are contradictory (Wolf et al 1992, Donovan et al 1998). Thus, we wanted to assess in further detail the interaction between the SP-A and SP-B genetic variants as a genetic determinant of RDS separately in subpopulations of prematurely born singletons versus multiples. DNA samples from 119 singleton infants with RDS, 322 without RDS and 245 multiple infants with RDS and 235 without RDS were genotyped for the the SP-A1 and SP-A2 SNPs, SP-B exon 4 Ile131Thr and SP-B intron 4 size variants and studied for association with RDS, prematurity and fetal size. Blood samples collected prospectively (n = 594) during 1997–2000 and 327 retrospectively recruited premature singletons and multiples born during 1987–1996, from whom buccal smear samples or a blood spot dried on a filter paper were available, were included in the study.

Additionally experimental data on the different glycosylation of preSP-B in human fetal lung tissue is presented to explain the potential differences in the prosessing of SP-B conferred to the different SP-B exon 4 Ile131Thr genotypes.

Infants were excluded from the genetic studies III–V if one or both of the parents were of non-Finnish origin or if the infants had received intrauterine blood transfusion. All DNA samples were genotyped for SP-A1, SP-A2 and SP-B exon 4 polymorphisms and intron 4 size variants, but occasional failure in genotyping caused slight variation in the number of analyzed cases.

46

4.4 DNA samples and genotyping

There were altogether 594 umbilical cord blood specimens obtained prospectively from preterm infants during 1997–2000. Samples of 0.5–3.0 ml of whole blood were used to isolate genomic DNA with the Purogene DNA Isolation Kit (Gentra Systems). From 276 twin and multiple infants born during 1987–1996 duplicate buccal smear samples from the cheek mucous membrane were collected retrospectively into sterile swab collection tubes. DNA was extracted from the swabs using Chelex 100 medium (Bio-Rad). Appropriate contamination controls for all reagents were included in each series of DNA preparation. In addition, 51 retrospectively collected samples of blood spots dried on a filter paper were included in the series. The sample disk punched from the paper was purified with DNA Purification Solution (Gentra Systems) and used directly as a template. A blank paper disk treated similarly was included in each series to control for potential DNA cross-contamination.

Polymerase chain reaction (PCR) amplification was used to carry out the genetic analyses. The PCR reactions were controlled for purity by inclusion of a water sample in each series of amplification. The SP-A genotypes were determined using a restriction fragment length polymorphism (RFLP) method based on converted gene-specific PCR primers (Di Angelo et al 1999). The samples were digested with appropriate restriction enzymes. SPs for five codons for the SP-A1 gene, and four codons for the SP-A2 gene were analyzed for the SP-A genotypes. After PCR amplification, the SP-B Ile131Thr genotypes were digested with the restriction enzyme TaaI before analyses. SP-B intron 4 size variants were analyzed directly after PCR amplification (Rämet et al 2000).

Twin zygosity was determined by using four highly informative tetranucleotic repeat markers in addition to genotyping of the SP-A and SP-B polymorphisms. Dizygosity was inferred if twins differed in any of the markers.

Additionally, human fetal lung explants were cultured and genotyped for the SP-B Ile131Thr polymorphism. Homozygous Thr/Thr and Ile/Ile explants were labelled and processed for immunoprecipitation as described (Guttentag et al 1998). The proteins were immunoprecipitated with a rabbit polyclonal antibody to human SP-B and the products separated using a 12% NuPAGE Bis-Tris gel with a MES buffer system. SP-B fragments were visualized by Storm Phosphorimager.

4.5 Ethical considerations

The Finnish Ministry for Social Affairs and Health gave a permission to conduct the population study I and genetic studies on a nationwide basis. In order to obtain an adequate population of twins the National Birth Registry was evaluated. The prospective and retrospective samples for genetic studies were collected during 1996–2000 from the districts served by the University Hospitals of Oulu (III–V), Tampere (III–V), Turku (III–V), Kuopio (III–V), and Helsinki (III–V), as well as Seinäjoki central hospital (V). Families with preterm twins or multiples were contacted via the respective university hospitals and asked for buccal smear samples for genetic studies. The ethical committees

47

of the participating centres approved the study protocols. The clinical data and the neonatal and maternal histories were obtained from the medical records of each hospital. Written informed consent was obtained from the parents of the infants for the blood and buccal smear samples to be used in genetic studies.

4.6 Statistical analysis

In the epidemiological studies I and II the overall differences in the binary and category variables between singletons and twins were tested by 2 × k with trend χ2 tests. Continuous variables were tested by Student´s t-test for unpaired samples if the variables were normally distributed and by the non-parametric Mann-Whitney U-test if they were not. Odds ratios (ORs) and 95% confidence intervals (CIs) for the differences between singletons and twins, male and female newborns and first-born and second-born twins were calculated with the Mantel-Haenszel method after stratification by gestational age, mode of delivery or year of birth. The independent samples t-test and the binomial test were used to calculate the proportional differences with 95% CIs of the concordance rates of RDS between the different zygosity groups. Multiple regression analyses, with twin pairs clustered, i.e. taking into account the dependence within twin pairs, were used to estimate the available birth characteristics as risk factors for RDS. Discordant twin pairs (only one of the pair being affected) were additionally analyzed with conditional logistic regression analysis (considered as matched case-control pairs).

The t-test for independent samples and the binomial test were used to calculate the proportion differences with 95% CIs in the concordance rates of RDS between MZ and DZ twin pairs.

In the genetic association studies III–V the differences in the allele and genotype frequencies between the RDS and no-RDS infants were evaluated using 2 × 2 contingency tables and χ2 analysis. Fisher`s exact test was used when the expected cell frequency was < 5. Bonferroni´s corrections were made for multiple comparisons by multiplying the statistically significant p values by the number of pairwise comparisons being made (V). ORs and 95% CIs were calculated with the Woolf logit method to assess the risk of RDS conferred by a particular allele and genotype. The observed genotype frequencies were compared with the expected Hardy-Weinberg distributions by χ2 tests. The ORs between genetic factors and RDS, and stratified by zygosity (MZ/DZ), degree of prematurity (< or ≥ 32 weeks of gestation), antenatal glucocorticoid treatment (yes/no), birth order (first/second born), and gender (male/female) in order to explore effect modification, were compared with the homogeneity of the odds ratio (HOR) test (IV, V). Multiple logistic regression analyses were used to determine whether the different clinical variables or the homozygosity or heterozygosity of particular alleles/genotypes of SP-A or B influenced the risk of RDS. The twin pairs were clustered in the analyses. The analyses were adjusted for zygosity, degree of prematurity, antenatal glucocorticoid treatment, birth order and gender.

Statistical analyses were performed by the SPSS versions 9.0 to 11.0 for Windows (basic statistical calculations, logistic regression analysis, Mantel-Haenszel

48

stratifications), Arcus Quick Stat (χ2-tests for allele frequencies), StatXact-3 for Windows (HOR test), Intercooled STATA version 7.0 (logistic regression analysis with clustered data of twin pairs) and Egret for Windows (conditional logistic regression analysis).

5 Results

5.1 Population-based epidemiology and risk factors for RDS (I, II)

5.1.1 Incidence of RDS (I, II)

Although the overall number of births per year in Finland decreased from 60 223 in 1987 to 57 371 in 2000, the rate of preterm births increased from 5.6% to 6.3%, respectively, ranging from 4.5% to 4.9% in singletons and from 39.0% to 49.0% in twins. 4970 RDS cases were found among the 873 784 singleton and twin infants born alive during the years 1987–2000, accounting for an overall RDS incidence of 5.7 per 1000 live-born infants: 5.0 for singletons and 39.5 for the first-born and 50.9 for the second-born twin infants, respectively. The incidence of RDS decreased from 114 in 1987–1990 to 102 per 1000 live-born preterm infants in 1996–2000 in the whole population of preterm infants and, specifically, from 114 to 83 among the first-born twin infants.

When the catchment area of Oulu University Hospital was studied, altogether 488 infants with RDS were identified among 58 990 live-born infants, accounting for an overall incidence of RDS 8.3 per 1000 live births. Figure 7 shows birth weight-specific incidence rates of RDS in the two consequent year periods 1990–1995 and 1996–1999. The incidence of RDS was 6.7 per 1000 singletons and 70.4 per 1000 multiple live-born infants, respectively. The overall incidence of RDS was 8.7/1 000 live births in 1990–95 and 7.6 in 1996–99 (p = 0.15). The incidence of RDS changed towards lower gestational age due to the decreased incidence rate among preterm infants born after 30 weeks. The gestational age-specific incidence of RDS decreased significantly from 1990–1995 to 1996–1999 (OR 0.7, 95% CI 0.5–0.9, p = 0.005). The most marked difference in the RDS incidence rate between the two periods was seen in the infants born at 31 and 32 weeks (58% vs. 36% and 46% vs. 28%) and in the birth weight classes of 1 750–1 999 g (28% vs. 11%) and 2 000–2 249 g (8% vs. 4%).

50

Fig. 7. Birth weight-specific incidence of RDS in the two periods, 1990–95 and 1996–99. When adjusted for birth weight, the incidence of RDS was significantly lower in the latter compared to the previous period (OR 0.7, 95% CI 0.5–0.9, p = 0.001). This was especially obvious in the birth weight groups of 1 750–1 999 g and 2 000–2 249 g. The numbers of infants with RDS and all infants born in each birth weight category are shown above the histograms.

5.1.2 Risk factors for RDS in twins (I)

Due to the significantly higher rate of preterm births, twin and multiple infants were over-represented in the overall incidence rates. When the data were stratified by plurality and gestational age, the incidence rate of RDS in twin infants differed significantly compared to that of singletons. The incidence rates were counted per 1000 live-born preterm singleton infants, first-born and second-born twins, respectively. In very preterm infants born before 28 weeks, the incidence of RDS was higher in twins compared to singletons of similar gestational age. On the contrary, in near-term infants born between the weeks 32 and 36, the risk of RDS was lower in twins than in singletons of similar gestational age. Compared across the whole set of data, first-born twins had a lower gestational age-adjusted incidence of RDS compared to second-born twins (OR 0.6, 95% CI 0.5–0.8). As a whole, there were no significant differences between second-born twins and singleton

< 750 750-999 100-1249 1250-1499 1500-1749 1750-1999 2000-2249 2250-2499 2500-2749 2750-2999

0

10

20

30

40

50

60

70

80

90

2750-2999

2500-2749

2250-2499

2000-2249

1750-1999

1500-1749

1250-1499

1000-1249

750-999

<750

6/15

3516

/231

7

9/75

7

14/1

104

12/3

64

22/5

37

8/20

625

/306

13/1

17

49/1

7825/7

9

42/1

2122/5

4

35/8

116/3

131/5

5

29/3

641

/52

23/2

824/2

7

Birth weight (g)

Freq

uenc

y of

RD

S (%

) 1990-1995 1996-1999

51

infants (OR 1.1, 95% CI 0.9–1.2), as shown in Figure 8. This result was even more pronounced when only the latter five-year period of 1996–2000 was considered.

Fig. 8. Incidence of RDS in premature singleton and first- and second-born twin infants. The gestational age-adjusted incidence of RDS between first-born and second-born twins and between first-born twins and singletons differed significantly (OR 0.6, 95% CI 0.5–0.7 and OR 0.7, 95% CI 0.6–0.8, respectively), but no difference was seen between second-born twins and singletons (OR 1.1, 95% CI 1.0–1.2).

Same-sex pairs, both males and females, were born more prematurely than opposite-sex pairs (mean difference for gestational age -0.4 weeks, 95% CI -0.5 to -0.2). There was no difference in the gestational age-adjusted risk of RDS (OR 1.1, 95% CI 0.9–1.3) or in the gestational age-adjusted concordance of RDS (OR 1.3, 95% CI 0.9–1.7) between same-sex and opposite-sex twin pairs. When stratified by gender, female infants from opposite-sex pairs had a slightly lower risk of RDS compared to infants from same-sex pairs (OR 0.8, 95% CI 0.6–1.0) regardless of birth order. Multiple regression analysis showed that vaginal delivery (OR 0.7, 95% CI 0.6–0.9), female sex (OR 0.7, 95% CI 0.5–0.9), being born first (OR 0.2, 95% CI 0.1–0.3) and being the lighter of the twins (OR 0.3, 95% CI 0.2–0.5) were independent factors protecting from RDS.

24-25 26-27 28-29 30-31 32-33 34-35 36-0

100

200

300

400

500

600

700

800

900

Inci

denc

e of

RD

S(p

er 1

000

pret

erm

live

sin

glet

on/tw

in b

irths

)

Gestational age (weeks)

Singleton First-born twin Second-born twin

52

5.1.3 RDS-related mortality and morbidity (II)

According to the regional study II during the 1990s, RDS-related neonatal mortality decreased parallel to overall neonatal mortality, accounting for 15% of all neonatal deaths, (i.e deaths during the first 28 days). All infants who died from RDS or its complications were born before 28 completed weeks, with the exception of four cases in the beginning of 1990s. The mean gestational age at birth of the infants who died during the neonatal period were 27.2 weeks (range 23–33) in 1990–1995 and 25.6 weeks (range 23–27) in 1996–1999.

The duration of oxygen therapy shortened (8.0 vs. 5.5 days) and the incidence of pneumothorax decreased (9.7% vs. 4.1%), whereas the rate of chronic lung disease at 36 weeks of postconceptional age (16.4% vs. 16.7%) and at one year of corrected age (9.2% vs. 8.2%) remained unchanged.

5.1.4 Outcome at one year of age (II)

Only 8 (1.9%) out of 488 RDS infants did not participate the follow-up by one year of corrected age. Late pulmonary and neurosensory sequelae of the infants affected with RDS remained at the same level between the two periods. Somatic growth in terms of height and head circumference expressed as standard deviations from the mean remained below the mean. Weight, which was expressed relative to height as a percentage of normal, showed that the infants settled about two percent below the mean.

5.2 Concordance of RDS in MZ and DZ twin pairs (III, IV)

The recruitment of retrospective twin pairs covered all university hospital districts in Finland, and preterm twins with no DNA available for analysis did not differ in population characteristics from the twin pairs analyzed. Of the 200 twin infants analysed, RDS was diagnosed in 149 (75%). The concordance rates of RDS were 54% and 44% in the MZ and DZ same-sex pairs, respectively. The concordance difference of 10% was not statistically significant (95% CI–0.1 to 0.3, p = 0.32). Stratification by gender or gestational age did not affect the concordance rates or differences between MZ and DZ twins. When 96 prospectively recruited twin pairs were added to the concordance analysis, the concordance difference between MZ and DZ same-sex pairs was 5% (95% CI -0.1 to 0.2, p = 0.62).

53

5.3 SP-A and SP-B polymorphisms and RDS in twins (III, IV)

In the twin concordance study, the SP-A and SP-B genes were evaluated for potential associations with the susceptibility to RDS in 100 same-sex pairs. The SP-B Ile131Thr polymorphism was shown to associate with RDS. The threonine allele was associated with an increased risk of RDS (OR 2.2, 95% CI 1.4–3.5, p = 0.0014). This was particularly evident in first-born male twins (OR 6.2, 95% CI 2.4–16.3, p < 0.001). Figure 9 shows the allele frequencies in RDS and no-RDS infants by birth order. Neither the SP-A alleles or the SP-B intron 4 polymorphism differed between RDS and no-RDS infants or between MZ and DZ twins in the whole population or in any of the subgroups determined by gender or gestational age. The degree of prematurity (OR 2.0, 95% CI 1.1–3.7, p = 0.021) and birth order (second-born OR 3.1, 95% CI 1.3–7.4, p = 0.009) were the clinical variables that influenced the risk of RDS, when the data were analyzed with clustered multiple logistic regression.

Fig. 9. SP-B exon 4 Thr allele frequencies in no-RDS and RDS infants among first- and second-born twins. The horizontal line represents the allele frequency among Finnish premature singletons. The p values, odds ratios (OR) and 95% confidence intervals (CI), which illustrate the association of the Thr allele with RDS, are shown below the histograms.

First born Second born0

10

20

30

40

50

0.91.10.6-2.2

0.0052.31.3-3.9

POR95% CI

SP-B

Thr

alle

le fr

eque

ncy

%

NoRDS RDS

54

The prospectively and retrospectively recruited twin pairs did not differ significantly in any clinical characteristics. When they were pooled together, there were altogether 198 premature twin pairs. DNA samples from all of the 396 twin infants were genotyped for SP-A1 and SP-A2 polymorphisms. Among the twin pairs, the SP-A1 allele frequency distribution as a whole differed significantly between the no-RDS and RDS infants (p = 0.030). The frequencies for 6A2 were 0.65 in the no-RDS group and 0.57 in the RDS group (p = 0.027), and those for 6A3 0.20 and 0.28 (p = 0.0074), respectively. The SP-A1 genotype 6A2/6A2 frequencies were 0.45 in the no-RDS and 0.31 in the RDS twin infants (p = 0.0042), and the SP-A haplotype 6A2-1A0 frequencies were 0.62 and 0.53 (p = 0.016), respectively.

The association between RDS and the SP-A gene locus was dependent on the degree of prematurity. Very preterm twin infants (< 32 weeks of gestation) showed no detectable differences in the distribution of SP-A1 or SP-A2 alleles, SP-A genotypes or haplotypes according to RDS status. However, among the near-term twin infants (from 32 to 36 weeks of gestation), the SP-A1 allele, genotype and SP-A haplotype frequency differences were more pronounced. The frequencies in no-RDS and RDS twin infants were 0.69 and 0.63 (p = 0.029) for allele 6A2, 0.17 and 0.25 (p = 0.080) for the allele 6A3, 0.49 and 0.28 (p = 0.0070) for the homozygous genotype 6A2/6A2, and 0.65 and 0.57 (p = 0.16) for the haplotype 6A2-1A0, respectively.

It is known that SP-A gene expression is regulated by glucocortiocoids (Boggaram et al 1989). The results of the HOR tests showed effect modification by antenatal glucocorticoid treatment on the SP-A allele frequencies (p = 0.015, 0.022, 0.076 and 0.22 for frequencies of the SP-A1 alleles 6A2, 6A3, genotype 6A2/6A2 and SP-A haplotype 6A2-1A0, respectively). When the data were stratified according to the glucocorticoid treatment, the most striking differences were seen in the infants with no steroid therapy. The SP-A1 genotype 6A2/6A2 frequency was 0.51 in the no-RDS infants compared to 0.27 in the infants with RDS (OR 0.35, CI 0.2–0.6, p < 0.001, corrected for multiple comparisons). The SPA1 genotype 6A2/6A2 frequencies in infants with and without RDS in two gestational age groups and stratified by glucocorticoid treatment are illustrated in Figure 10. Stratification according to zygosity, birth order, growth restriction or gender did not have any detectable effect on the association between RDS and SP-A polymorphisms.

55

Fig. 10. Homozygous genotype 6A2/6A2 frequencies in no-RDS and RDS twin infants according to glucocorticoid treatment in all twins and in the two gestational age groups. The p values, odds ratios (OR) and 95% confidence intervals (CI), which illustrate the association of the SP-A1 6A2/6A2 genotype with RDS, are shown below the histograms. The corrected p value in multiple comparison is shown in parentheses.

In order to control confounding factors, the data were analyzed with logistic regression, with the twin pairs clustered. After adjustment for the clinical variables, the homozygous 6A2/6A2 genotype of SP-A1 compared to any other SP-A1 genotype was associated with a decreased risk of RDS (OR 0.57, CI 0.3–0.9, p = 0.027). This type of association was even stronger in the twin pairs born near term (OR 0.34, CI 0.2–0.7, p = 0.004).

5.4 Interaction by the SP-B gene and plurality on the association between the SP-A gene and RDS (V)

No significant differences in the allele distributions of the SP-A or SP-B genes were shown between the RDS and no-RDS infants when singletons and multiple infants were considered together. However, the HOR test showed the presence of a profound effect modification by multiple birth in the SP-A gene polymorphisms (p = 0.0002, 0.001, 0.0006, 0.0001 and 0.17 for the frequencies of the alleles 6A2, 6A3, 1A0, the genotype

No steroid Steroid -- No steroid Steroid -- No steroid Steroid0

10

20

30

40

50_

0.260.490.14-1.73

0.0049 (0.020)0.310.14-0.72

0.601.370.52-3.07

0.741.250.34-4.61

0.780.910.48-1.73

<0.0010.350.20-0.63

4/16

21/52

10/37

53/98

23/61

11/34

17/65

4/13

27/7732/86

27/102

57/111

POR95% CI

>32 weeks< 32 weeks

All twins

SP-A

1 ge

noty

pe 6

A2 /6A2 fr

eque

ncy

% NoRDS RDS

56

6A2/6A2 and haplotype 6A2-1A0, respectively). The SP-B Ile131Thr polymorphism differed significantly between RDS and no-RDS only in the first-born infants from twin and multiple gestations, the threonine allele being directly associated with increased risk of RDS (OR 1.72, 95% CI 1.18–2.52, p = 0.005) and the isoleucine allele with decreased risk of RDS (OR 0.58, 95%CI 0.40–0.85, p = 0.005), as shown in the twin concordance study. No differences between the RDS and no-RDS infants were found in the distributions of the intron 4 polymorphism in infants born from singleton or multiple gestations even after separate evaluation.

As expected, the distributions of both the SP-A1 and the SP-A2 genes between RDS and no-RDS infants differed significantly when infants from singleton and multiple gestations were analyzed separately. This relationship was influenced by the SP-B Ile131Thr genotype. The SP-A1 allele 6A2 was over-represented in RDS of singletons when the SP-B genotype was Thr/Thr (frequencies 0.71 in RDS and 0.50 in no-RDS, p = 0.009). In contrast the SP-A1 allele 6A2 tended to be under-represented in RDS of multiples when the SP-B genotype was Ile/Thr (frequencies 0.53 in RDS and 0.67 in no-RDS, p = 0.012) or Thr/Thr (frequencies 0.54 in RDS and 0.64 in no-RDS, p = 0.12). The SP-B Ile131Thr genotype-dependent association between SP-A1 homozygous 6A2/6A2

genotype and RDS was even more evident, as shown in Figure 11.

Fig. 11. Interaction of the SP-A and SP-B genes in all premature infants born from singleton and multiple gestations: Frequencies of the genotype 6A2/6A2 among 322 singleton and 235 multiple infants without RDS and 119 singleton and 245 multiple infants with RDS. The infants were divided into subgroups based on their SP-B Ile131Thr genotype. The p values corrected for multiple comparisons, odds ratios (ORs) and 95% confidence intervals (CIs) which illustrate the association of the 6A2/6A2 genotype with RDS in the different SP-B genotype groups; in singletons with the SP-B Thr/Thr genotype, p = 0.006, OR 4.5, CI 1.8–11.3; in multiples with the Ile/Thr genotype p = 0.05, OR 0.5, CI 0.3–0.9; and in multiples with the Thr/Thr genotype p = 0.04, OR 0.4, CI 0.2–0.8.

Ile/Ile Ile/Thr Thr/Thr -- Ile/Ile Ile/Thr Thr/Thr0

10

20

30

40

50MultiplesSingletons

SP-A

1 6A

2 /6A

2 gen

otyp

e fr

eque

ncy

%

SP-B Ile131Thr genotype

NoRDS RDS

57

In a previous study consisting mainly of singletons (Haataja et al 2001), the interaction between the SP-A and SP-B exon 4 polymorphisms was strongly dependent on the degree of prematurity. Accordingly, the data were analyzed separately for premature infants born before (very premature) and after 32 weeks of gestation (near term). Among the very premature singleton infants, the SP-B Ile131Thr genotype defined the cases with a significant association between the SP-A allele 6A2 and RDS. Regardless of the SP-B Ile131Thr genotype, no significant association was evident between the SP-A1 alleles and RDS among very premature multiples. On the contrary, among near-term multiples, the presence of the SP-B genotype Ile/Thr or Thr/Thr defined the significant association between the SP-A1 allele 6A2 and the genotype 6A2/6A2 and RDS. In near-term singletons, the SP-A and/or SP-B alleles/genotypes/haplotypes had no detectable association with RDS.

The results of multiple logistic regression analyses supported the findings on allelic and genotype associations. Gestational age (< and ≥ 32 weeks of gestation), birth order and gender were included as confounding factors in the analyses. In singletons, a homozygous 6A2/6A2 genotype was associated with an increased risk of RDS when the SP-B genotype was Thr/Thr (OR 0.27, 95% CI 0.06–0.80, p = 0.018). In multiples, on the other hand, a homozygous 6A2/6A2 genotype decreased the risk of RDS when the SP-B genotype was Ile/Thr or Thr/Thr (OR 0.42, 95% CI 0.22–0.80, p = 0.008). This was especially obvious in near-term multiples (OR 0.29, 95% CI 0.09–0.90, p = 0.032).

Table 5 demonstrates the main findings of genetic studies III–V. SP-A1 6A2/6A2 and SP-B exon 4 Thr/Thr genotype frequencies in twin infants with and without RDS, and stratified by gestational age, are shown.

Table 5. Frequencies of SP-A1 6A2/6A2 and SP-B Thr/Thr genotypes in genetic studies III–V

Genotype Non-RDS RDS p a OR (95% CI) SP-B Thr/Thr 0.14 0.28 0.04 2.5 (1.0–6.0)

< 32 weeks of gestation 0.12 0.27 0.07 (0.1) 2.9 (0.8–10.4) ≥ 32 weeks of gestation 0.15 0.27 0.2 (0.4) 2.0 (0.6–7.1)

SP-A1 6A2/6A2 0.45 0.31 0.004 0.5 (0.3–0.7) < 32 weeks of gestation 0.32 0.32 0.6 (1.0) 1.0 (0.5–2.1) ≥ 32 weeks of gestation 0.49 0.28 0.003 (0.007) 0.4 (0.2–0.7)

SP-B Thr/Thr and SP-A1 6A2/6A2 Singleton infants 0.21 0.55 0.002 (0.006) 4.5 (1.8–11.3) < 32 weeks of gestation 0.27 0.65 0.06 (0.4) 4.9 (0.9–25.7) ≥ 32 weeks of gestation 0.20 0.43 0.08 (0.5) 3.0 (0.9–10.4)

Twin infants 0.49 0.25 0.01 (0.04) 0.4 (0.2–0.8) < 32 weeks of gestation 0.25 0.26 0.6 (1.0) 1.0 (0.2–4.2) ≥ 32 weeks of gestation 0.56 0.24 0.01 (0.06) 0.3 (0.1–0.8)

a The corrected p values for multiple comparisons are shown in parentheses

In a comparison of singleton and multiple pregnancies, the major SP-A1 allele 6A2 and the SP-B Ile131Thr genotype appeared to have contrasting effects on the risk of RDS defined on the basis of the length of pregnancy. Therefore, the pregnancies were analyzed on the basis of intrauterine size. The birth weight of singletons and the combined birth

58

weight of multiples were considered, since accurate information on placental weight and intra-amniotic volume was not available. As shown in Figure 12, there was a continuous association between fetal weight and the risk of RDS, which was confined to the carriers of the major SP-A1 allele 6A2 and the SP-B genotype Ile/Thr or Thr/Thr.

Fig. 12. Influence of the birth weight of singletons or the combined weight of multiples on the interaction between the SP-A and SP-B genes in predicting susceptibility to RDS. Altogether 313 singleton and 344 twins or multiples with either SP-B Ile/Thr or Thr/Thr genotype were analysed. The birth weights and the combined birth weights were categorised into two groups based on the medians. Both premature singletons and multiples were categorised into two subgroups: below or above the median birth weight / combined birth weight. In the singletons, the median birth weight and the interquartiles were 2220 g, 1565 g, and 2707 g, respectively. In the multiples, the median birth weight and the interquartiles were 3625 g, 2920 g, and 4605 g, respectively. Upward and downward arrows illustrate the increased and decreased risk of RDS associated with the SP-A allele 6A2. The odds ratios (ORs) and 95% confidence intervals (CIs), which illustrate the association of the 6A2 allele with RDS in the combined SP-B genotype Ile/Thr and Thr/Thr groups are shown. They were in singletons under the median birth weight they were OR = 1.8, CI 1.1–2.9, p = 0.02; in singletons over the median birth weight OR 1.1, CI 0.6–2.1, p = 0.72; in multiples under the median birth weight OR = 0.7, CI 0.5–1.2, p = 0.18; in multiples over the median birth weight OR 0.6, CI 0.4–1.0, p = 0.04. An asterisk (*) above a plot identifies a significant result. There were no significant associations with SP-A and RDS in the SP-B Ile/Ile group (data not shown).

According to sequence data, the peptide encoded by the Thr allele but not by the Ile allele contains a consensus sequence (Asn-X-Thr) for N-linked glycosylation at Asn129, suggesting that the Ile131Thr variation results in genetic variation of the molecular phenotype of proSP-B. In an experimental design utilizing human lung tissue, the molecular size of human proSP-B was shown to be genotype-specific, with the presumably glycosylated Thr/Thr product migrating more slowly than the unglycosylated

Singl bw under median singl bw over medianmultiples bwsum under medianmultiples bwsum over median

0,4

0,6

0,8

1,0

1,2

1,4

1,6

1,8

2,0

2,2

2,4

2,6

2,8

In the presence of the SP-B Ile/Thr or Thr/Thr genotype

Odd

s ra

tio fo

r RDS

Multiple pregnanciesSingleton pregnancies

*

*

SP-A 6A2 alleledecreasing the risk

SP-A 6A2 alleleincreasing the risk

Above medianBelow medianAbove medianBelow median

59

Ile/Ile product. Furthermore, the NH2–terminal fragment was shown to be glycosylated in proSP-B from Thr/Thr lung, but not from Ile/Ile lung, whereas both forms were detected in heterozygous Ile/Thr lung.

6 Discussion

6.1 Epidemiological aspects

The rate of preterm births increased during the 1990s, due to the increasing incidence of pregnancies with assisted conception and the high rate of multiples in these pregnancies (Blickstein & Keith 2002). It has also been shown that in vitro fertilization twins have more likely a higher incidence of preterm birth and prematurity-related respiratory complications (Nassar et al 2003). According to a Canadian study the increase of preterm deliveries is partly related to the population level phenomenon of delayed childbearing (Tough et al 2002).

During the 1990s, more premature infants survived and the incidence of RDS shifted towards more immature infants. The gestational age-adjusted incidence of RDS declined significantly. The overall incidence of RDS, however, did not decline as much as could have been expected as a consequence of the introduction of novel treatments. During the period studied, however, prophylactic surfactant was not in use. Combined use of antenatal corticosteroid, postnatal early surfactant and distending pressure via the nasal route has been shown to be effective in preterm infants (Verder et al 1999, DePaoli et al 2003). Even though new therapies do not remarkably decrease the incidence of RDS, at least in very premature infants, they alleviate the disease, reduce the risk of pulmonary leakage and decrease the incidence of intraventricular hemorrhage (Horbar et al 2002).

The regional incidence figures of RDS in Oulu University Hospital were somewhat higher than national figures. The incidence rate of RDS recorded in the study based on the National Birth Registry may, to some extent, underestimate the true incidence rate of RDS in the whole population. There may be some missing diagnoses in the registry especially among the most premature infants who died early in the postnatal period. Although RDS is a well known disorder, its definition is not straightforward, and some variation in the diagnostic criteria between neonatal units may occur. The validity of registry-based RDS diagnoses proved to be good in twin pairs recruited retrospectively to the genetic studies, however. The practise of antenatal corticosteroid administration to the mother was quite uniform in Finland during the last decade covered here. However, even though prophylactic surfactant was not given in the delivery room during the study

61

period, subsequent use of surfactant was one of the diagnostic criteria. Pulmonary infection and the administration of surfactant immediately in response to respiratory distress may result in minor misclassifications.

The risk of developing RDS is strongly dependent on gestational age at birth. The rate of preterm births in Finland is low compared to other industrialized countries (Robertson et al 1992), which explains the rather low incidence figures of RDS. On the basis of regional epidemiological study carried out in the Oulu University Hospital, RDS-related neonatal mortality decreased parallel to the decrease in overall neonatal mortality. However, RDS mortality was still considerable, representing 15% of all neonatal deaths. The proportion of infants with supplemental oxygen at 36 postconceptional weeks remained constant throughout the study period from 1990 to 1999, which is consistent with the reports published elsewhere (Jobe 1999). Additional long-term morbidity remained unchanged and considerably high.

Most previous studies on neonatal outcomes have compared singletons and twins across the whole range of gestational ages. However, because the degree of prematurity is the main risk factor for RDS, it is essential to stratify the data by gestational age (Kilpatrick et al 1996). We found out that, after taking into account the gestational age, twins are not at a substantially higher risk of RDS compared to singletons except at a very early gestational age. On the contrary, after 28 weeks of gestation, first-born twins had a lower incidence of RDS compared to both second-born twins and singletons. Twins have been shown to have higher fetal lung maturity values compared to singletons after 31 weeks of gestation (McElrath et al 2000). In addition, pulmonary maturity is usually synchronous in twins (Leveno et al 1984). It is possible that the maturation of lungs in twins precedes that of singletons due to hormonal, growth or inflammatory factors influencing the intrauterine environment.The second-born, or non-presenting infant may not benefit from the salutary effects of vaginal delivery to the same extent as the presenting twin and may be inherently at a higher risk for respiratory distress (Arnold et al 1987).

The finding on the lack of concordance difference between same-sex twins compared to opposite-sex twins refutes the proposal that genetic factors uniformly influence the risk of RDS in twin pregnancies. The same-sex twin pairs that were born more preterm than opposite-sex pairs did not have a higher gestational age-specific incidence or concordance rate of RDS. Chorionicity rather than zygosity may affect the risk of RDS. It is likely that among the most premature twins, monochorionic monozygous twin pairs, who are known to possess the greatest risk of pregnancy complications, are overrepresented. It can be concluded that environmental factors predominate over genetic ones in the predisposition of twins to RDS.

6.2 Methods in genetic studies

The sample size determination was performed before the twin concordance study. Twin pairs were recruited also retrospectively in order to obtain adequate number of twin pairs. The retrospectively recruited pairs did not differ from prospectively recruited in any of

62

the clinical characteristics. Even after combining retrospectively and prospectively recruited twin pairs, the concordance difference between MZ and DZ pairs was not significant. Lack of adequate hypothesis concerning the magnitude of genetic influence or the distribution of individual alleles in the study population posed restrictions on power calculations of the genetic studies. Furthermore the limitation to achieve extensive sample size in a population of preterm twins hampered the stratification into subpopulations.

Comparison concerning the concordance of diseases between MZ and DZ twin pairs have been used to define the role of heredity in disease etiology (Kaprio et al 2000). If MZ twins, with their complete genetic identity, have a significantly higher concordance than DZ twins, with their mean 50% genetic similarity, genetic factors are considered to influence the susceptibility to a disease. The interpretation of classic twin studies is not always so straightforward, however. It is usually assumed that MZ and DZ twins have an equal environment, and there does not exist major genetic-environmental interactions. Under these assumptions, the estimation of heredity, i.e. the proportion of etiology attributable to genetic factors, can be measured as the fraction of population variance (i.e. the underlying liability to disease) due to genetic variance (MacGregor 2000). Thus we could roughly calculate the heredity component as the concordance difference between MZ and DZ twin pairs multiplied by two. Statistical modelling with genetic, shared and individual environmental effects also assumes the absence of interaction effects and is thus not directly applicable to RDS, either.

Even though the concordance study failed to show a major hereditary impact on RDS, it does not mean that genetic analysis is worthless in assessing the etiologic factors of RDS. Twins do not share a completely identical intrauterine environment (Phillips 1993). Intrapair differences due to antenatal events may exist irrespective of zygosity (Martin et al 1997). The presenting twin is more prone to PPROM and intrauterine infection, and may therefore be less susceptible to RDS (Arnold et al 1987). When estimating the risk of RDS in twin and higher-order gestations, twinning itself is an effect modifier, which hamperes the assessment of the genetic component in the disease risk.

Association studies with candidate genes have been widely used to evaluate the genetic component of complex diseases. Several conflicting aspects must be regarded when interpreting the results. The most common type of genetic variation originates from a single base mutation, when one nucleotide is substituted for another, referred to as a single-nucleotide polymorphism (SNP). An association study aims to find the disease-predisposing or protective alleles by statistically comparing allele frequencies between a case and a control population. Statistical evidence for an association may result from different situations (Cardon & Palmer 2003). Firstly SNP may be functional and directly affect the expression of a phenotype. Both SP-A and SP-B are essential for proper lung function. In populations of mainly singletons the common polymorphisms of the SP-A and SP-B genes have been found out to associate with RDS (Rämet et al 2000, Haataja et al 2000, Floros et al 2001). There is also evidence that the SP-B Ile131Thr polymorphism may affect the glycosylation prosess of proSP-B and this in turn may affect the processing of mature SP-B (Haataja et al 2000, Wang et al 2003b). Thus we considered meaningful to study these polymorphisms in preterm twins and multiples. Secondly the allele studied may correlate with, or be in linkage disequilibrium with, a causative allele located nearby. Thirdly the association may be due to confounding or selection bias. The

63

recruitment of retrospective twin pairs covered all university hospital districts in Finland, and twins with no DNA available for analysis did not differ in population characteristics from the twin pairs analyzed. Additionally almost all parents of preterm twin pairs, who were recruited prospectively, gave their consent to conduct the genetic studies. Thus we considered the present studies be fairly free from selection bias. The candidate gene approach has also been criticized because of the non-reproducibility of the results and its limited ability to include all possible causative factors in complex diseases. Identical phenotypes resulting from different constellation of genes, a large number of different genes, interactions between genes, and the effect of one gene being only modest are factors that further complicate the interpretation of the results (Tabor et al 2002). Additionally a single gene may contain polymorphisms that influence the risk of several diseases (genetic parsimony). In situations where an association between a candidate gene and a disease is physiologically meaningful, the approach is better than linkage studies detecting genes of minor effect (Tabor et al 2002). Furthermore lingage analysis requires families with affected individuals in multiple generations. A matched case-control setting is usually preferred, and in case of twin pairs, discordant twins would be optimal, if birth order, gender and twinning itself were associated with RDS. Adequately large samples of RDS-discordant preterm twins for genetic studies are hard to obtain. An unmatched case-control design allows the use of a larger study population, and confounding factors can be controlled by stratification and logistic regression analysis (Sham 1998), as proceeded in the present studies. Irrespective of disease status, allele frequencies are known to vary widely between populations. The Finnish population is historically isolated and genetically comparatively homogenous (de la Chapelle et al 1993). In the present studies only infants of Finnish origin were included and population stratification was thus avoided.

6.3 Concordance of RDS between MZ and DZ twin pairs

According to the results of the twin studies III–IV, there was no significant concordance difference between the MZ and DZ twin pairs suggesting a lack of significant overall genetic impact on RDS. The result was similar after correction for gestational age or gender. In contrast, both the SP-B Ile131Thr and the SP-A polymorphisms were significantly associated with RDS in the same twin populations. This apparent discrepancy between the results of the concordance study and the candidate gene analyses can be explained on the basis of the lack of uniformity in the shared intrauterine environment of twins and the overriding role of environmental factors in twins compared to singletons. In addition, the observed low concordance rate in MZ twins may be due to the generally lower risk of RDS in the first-born than in the second-born twin infants. The inflammatory mediators and growth factors released in active preterm labour accelerate fetal lung maturity (Bry et al 1997, Jobe et al 2000) and, regardless of zygosity or genetic contribution, may protect the presenting twin against RDS, even without manifest infection. Therefore, birth order in twin pregnancy is a confounding factor when evaluating the genetic component of RDS in twin infants.

64

The previous study showing a higher concordance of RDS in 13 MZ compared to 18 DZ twin pairs included DZ twin pairs of opposite sexes (Myrianthopoulos et al 1971). The high concordance rate in MZ compared to DZ twin pairs was likely due to bias, because male sex is associated with an increased risk of RDS (Khoury et al 1985). In the present concordance study, only same-sex twin pairs were included.

6.4 Role of SP-A and SP-B polymorphisms in RDS of twins

Being significantly over-represented among the first-born no-RDS twins the Ile allele and Ile/Ile genotype was associated with a significantly lower rate of RDS than the Thr allele and Thr/Thr genotype in both MZ and DZ twin pairs. The present study is the first to indicate a direct allelic association between the SP-B Ile131Thr polymorphism and RDS. The SP-B Ile131Thr polymorphism was associated with RDS in the presenting twin. The SP-B Ile131Thr polymorphism may be an important determinant of lung maturity, especially in the first-born infant. This may partly explain the first-born infant´s lower risk of RDS. Recently, SP-B has been shown to take part in lung defence against infection (van Iwaarden et al 2001, Epaud et al 2003), but the mechanism by which the SP-B gene polymorphism might influence the function of mature SP-B is speculative. The substitution of threonine for isoleucine in the SP-B peptide at amino acid 131 affects the N-linked glycosylation site (Asn-X-Ser/Thr) of proSP-B. Glycosylation may influence the proSP-B folding and sorting and thereby the secretion of surfactant (Haataja et al 2000, Wang et al 2003b). This could be tested using pulse-chase experiments on cultures from human fetal lung. SP-B polymorphism and birth order serve as an example of genetic and environmental interaction as a determinant of the phenotype. It can be speculated whether SP-B in the association of the Ile131 Thr polymorphism and RDS in the presenting twin functions as a protein involved in inflammation. The SP-B Thr allele has been associated with pulmonary diseases in adults (Lin et al 2000, Guo et al 2000, Selman et al 2003).

In the study of SP-A polymorphisms on twins we found that, in healthy preterm twin infants, particularly ones born at 32 weeks of gestation or later, the major allele 6A2, the homozygous genotype 6A2/6A2 of SP-A1 and the 6A2-1A0 haplotype of SP-A were over-represented even after adjustment for zygosity, birth order, gender and corticosteroid therapy. Antenatal corticosteroid therapy served as an effect modifier on the association between SP-A polymorphisms and RDS among preterm twin pairs. Since corticosteroid therapy is mainly indicated during imminent very premature birth, gestation and corticosteroid treatment were interdependent factors. Therefore, some of the associations displayed may be due to a selection bias. However, the genotype of the twin individual affected the risk of RDS even after separate adjustments for gestational age and antenatal corticosteroid treatment with the clinical variables. According to in vitro studies, the responsiveness of SP-A expression to corticosteroid is dependent on the length of gestation (Boggaram et al 1989). It has also been shown that the responsiveness of different alleles of SP-A to synthetic corticosteroid is allele-specific (Kumar & Snyder 1998), the SP-A1 alleles being more responsive to corticosteroid than SP-A2 (Hoover et

65

al 1999). Corticosteroid suppresses the inflammatory functions frequently associated with premature birth, including the expression of TNF-α (Shima et al 1999). The SP-A protein may restrict or promote the inflammatory processes potentially influencing the risk of RDS (Song & Phelps 2000). The observed associations between RDS, premature birth, corticosteroid and SP-A genotype may be due to inflammatory events. However, the mechanism by which the major SP-A genotype and haplotype alter the interaction of the inflammatory events and antenatal lung maturation remains open.

In twin pregnancies, the association between the SP-A polymorphism and the risk of RDS is apparently different from that seen in singleton pregnancies. The reason for this remains speculative. In twin pregnancies, birth order and other confounding environmental factors not present in singleton pregnancies potentially contribute to the susceptibility to RDS. The length of gestation is closely associated with fetal size. When the cumulative fetal weight or total intrauterine content is considered, twin and singleton pregnancies are quite different. If expressed on the basis of cumulative fetal weight the genotype-associated susceptibility to RDS appears to make up a continuum. With the low weight in the case of very immature singletons the main SP-A allele is associated with RDS, while with the moderate weight of many near-term singleton infants and very premature twins the genotype has little influence. In the presence of excessive intrauterine distention in near-term twin pregnancies, the SP-A 6A2 allele is associated with a lower risk of RDS. Hormones and growth factors associating with uterine distension influence both the onset of labour and the differentiation of the surfactant system (Hallman et al 2001). Because SP-A is important both for the proper function of surfactant and for the innate immunity of the lungs, it can be speculated that the most common SP-A allele and haplotype may function differently depending on the length of gestation.

6.5 Interaction by the SP-B gene and plurality on the association between the SP-A gene and RDS

The interactive associations of SP-A and SP-B polymorphisms with RDS were latent and intertwined with acquired factors, particularly the degree of prematurity, birth order and twinning. Despite the lack of overall direct genetic influence on the risk of RDS, the segments of the study population had strong gene-gene and gene-environment interactions. The degree of prematurity was always an overriding risk factor. SP-A and SP-B have an interactive role in maintaining the surface activity in vitro, and they are both essential components of the highly surface-active tubular myelin fraction (Possmayer et al 2001). According to an in vitro study low levels of SP-A mRNA, confined to the major SP-A2 allele 1A0, have been observed to associate with respiratory distress (Floros & Hoover 1998). According to the present study, the risk of RDS, defined by the SP-A genotype among the carriers of the SP-B Thr allele, is restricted to the subgroup carrying either the homozygous Thr/Thr or also the heterozygous Thr/Ile genotype. This study on interaction between the SP-A and SP-B common genetic variants confirmed that the genetic predisposition to RDS in multiples is different from that seen

66

in singleton infants. Almost 50% of multiple pregnancies end in preterm birth, whereas preterm birth in singleton pregnancies is uncommon and often associated with placental disease. The risk of RDS, defined by the interaction of SP-A and SP-B polymorphisms, was associated with the fetal mass. Thus, the difference in the susceptibility to RDS in premature singletons and multiples may depend on the size of the conceptus. A single hormone or cytokine may influence the expression of SP-A in very immature lung, whereas the same agonist may have an opposite effect on the same gene towards term (Odom et al 1997, Liley et al 1988, Glumoff et al 2000). Furthermore, specific combinations of hormones or cytokines may have concentration-dependent additive effects (Väyrynen et al 2002), making the underlying molecular mechanisms even more complex and difficult to elucidate at the phenotypic level.

Figure 13 shows a schematic framework of current knowledge of SP-A and B gene polymorphisms, which have been found to associate with the risk of RDS and other pulmonary diseases.

Fig. 13. Schematic framework illustrating the current knowledge of SP-A and SP-B gene polymorphisms associating with the risk of RDS and other pulmonary diseases. SP-B intron 4 size variants are represented as ∆i4.

7 Summary and future perspectives

Despite the progress made in antenatal and neonatal care, RDS continues to be a major factor causing increased morbidity and mortality among preterm infants. Advances in molecular genetics have enabled identification of candidate genes and interactions between genes affecting the susceptibility to RDS and have clarified some of the genetic background of RDS. Allelic variation and interaction of the SP-A and SP-B genes have been shown to associate with the disease susceptibility.

The present study improves our understanding of factors involved in the susceptibility to RDS in preterm infants. The study confirms that the genetic susceptibility to RDS is heterogenous and involves complex genetic and environmental relationships. The degree of prematurity is the most important determinant of RDS, but there are individual differences between infants of the same gestational age. This research showed that the gestational age-specific incidence of RDS has decreased after the introduction of new treatments. Twin infants do not have an increased risk of RDS, except at very early gestational age. The presenting twin is less susceptible to RDS than the non-presenting infant or singleton infant towards term gestation. A direct allelic association between the SP-B Ile131Thr polymorphism and RDS was evident only in the presenting twin. Both presentation and the polymorphism of the SP-B gene are controlled by a complex network of inflammatory and growth factors modulating lung development and growth. The role of SP-A polymorphisms in the risk of RDS in twins turned out to be different from that in singletons. The major allele and genotype of SP-A1 were associated with a decreased risk of RDS in near term twin infants. The threonine allele in SP-B Ile131Thr appeared to associate, interactively with SP-A1, with the risk of RDS both in twins and in singletons, with an increased risk of RDS in singletons at very early gestation, but with a lower risk of RDS in twins and multiples near term. The risk of RDS, defined by the interaction of SP-A and SP-B alleles, was associated with the fetal mass. Thus, the difference in the genetic susceptibility to RDS in premature singletons and multiples may depend on the size of the conceptus. Twin gestation modified the association between SP-A and SP-B gene polymorphisms and RDS. Environmental factors turned out to predominate over genetic ones in the disease susceptibility among twins.

When evaluating the genetic risk factors for RDS in twins, the classical twin study method of comparing MZ and DZ twins for the concordance of a disease is not directly

68

applicable. Several predominant intrauterine and perinatal environmental factors contribute to disease susceptibility regardless of zygosity and are suspected to partly override the heredity components of RDS.

The present results explain part of the susceptibility to RDS in different subgroups of preterm infants. There are several open issues concerning the processing of surfactant proteins and the role of surfactant protein gene polymorphisms modulating the process. The precise molecular mechanisms and interactions by which lung surfactant protein genes and their common variants contribute to lung function in health and disease remain to be elucidated. The future perspective into epidemiology, too, would require methods using genetic associations to test hypotheses about the causal pathways in disease etiology. Genome-wide scans, followed by positional cloning, may enable the identification of novel genes influencing the development of RDS and other pulmonary diseases. Genes for corticosteroid receptors, for cytokines and their receptors, for growth factors and for transcription factors that also regulate the expression of surfactant proteins might be potential candidates.

References

Alexander GR, Kogan M, Martin J & Papiernik E (1998) What are the fetal growth patterns of singletons, twins, and triplets in the United States? Clin Obstet Gynecol 41:114–125.

Arnold C, McLean FH, Kramer MS & Usher RH (1987) Respiratory distress syndrome in second-born versus first-born twins. A matched case-control analysis. N Engl J Med 317:1121–5.

Avery ME & Mead J (1959) Surface properties in relation to atelectasis and hyaline membrane disease. Am J Dis Child 97:517–23

Ballard PL, Hawgood S, Liley H, Wellenstein G, Gonzales LW, Benson B, Cordell B & White RT (1986) Regulation of pulmonary surfactant apoprotein SP 28-36 gene in fetal human lung. Proc Natl Acad Sci U S A 83:9527–9531.

Becroft DM, Farmer K, Mason GH, Morris MC & Stewart JH (1976) Perinatal infections by goup B beta-hemolytic streptococci. Br J Obstet Gynecol 83:960–966.

Beers MF, Shuman H, Liley HG, Floros J, Gonzales LW, Yue N & Ballard PL (1995) Surfactant protein B in human fetal lung: developmental and glucocorticoid regulation. Pediatr Res 38:668–75.

Benirschke K & Kim CK (1973) Multiple pregnancy. N Engl J Med 288:1329–1336. Blickstein I & Keith LG (2002) Iatrogenic multiple pregnancy. Semin Neonatol 7:169–176. Boggaram V & Mendelson CR (1988) Transcriptional regulation of the gene encoding the major

surfactant protein (SP-A) in rabbit fetal lung. J Biol Chem 263:19060–19065. Boggaram V, Smith ME & Mendelson CR (1989) Regulation of expression of the gene encoding

the major surfactant protein (SP-A) in human fetal lung in vitro. Disparate effects of glucocorticoids on transcription and on mRNA stability. J Biol Chem 264:11421–7.

Bohinski RJ, Di Lauro R & Whitsett JA (1994) The lung-specific surfactant protein B gene promoter is a target for thyroid transcription factor 1 and hepatocyte nuclear factor 3, indicating common factors for organ-specific gene expression along the foregut axis. Mol Cell Biol 14:5671–5681.

Bomsel-Helmreich O & Mufti W (1995) Mechanism of monozygosity and double ovulation. In: Keilth LG, Papiernik E, Keith DM & Luke B (eds) Multiple birth: Epidemiology, gestation and perinatal outcome. The Parthenon Publishing Group, New York, pp. 25–40.

Braems GA, Han VK & Challis JR (1998) Gestational age-dependent changes in the levels of mRNAs encoding cortisol biosynthetic enzymes and IGF-II in the adrenal gland of fetal sheep during prolonged hypoxemia. J Endocrinol 159:257–264.

Bry K, Lappalainen U & Hallman M (1997) Intraamniotic interleukin-1 accelerates surfactant protein synthesis in fetal rabbits and improves lung stability after premature birth. J Clin Invest 99:2992–9.

Bunt JE, Carnielli VP, Janssen DJ, Wattimena JL, Hop WC, Sauer PJ & Zimmermann LJ (2000) Treatment with exogenous surfactant stimulates endogenous surfactant synthesis in premature infants with respiratory distress syndrome. Crit Care Med 28:3383–3388.

70Burri P (1997) Structural aspects of prenatal and postnatal development and growth of the lung. In:

McDonald (ed) Lung Growth and Development. Marcel Dekker Inc., New York, pp. 1–35. Cardon LR & Palmer LJ (2003) Population stratification and spurious allelic association. Lancet

361:598–604. Caukwell S & Murphy DJ (2002) The effect of mode of delivery and gestational age on neonatal

outcome of the non-cephalic- presenting second twin. Am J Obstet Gynecol 187:1356–1361. Chabot S, Salez L, McCormack FX, Touqui L & Chignard M (2003) Surfactant protein A inhibits

lipopolysaccharide-induced in vivo production of interleukin-10 by mononuclear phagocytes during lung inflammation. Am J Respir Cell Mol Biol 28:347–353.

Cho K, Matsuda T, Okajima S, Matsumoto Y, Sagawa T, Fujimoto S & Kobayashi K (2000) Prediction of respiratory distress syndrome by the level of pulmonary surfactant protein A in cord blood sera. Biol Neonate 77:83–87.

Clark JC, Wert SE, Bachurski CJ, Stahlman MT, Stripp BR, Weaver TE & Whitsett JA (1995) Targeted disruption of the surfactant protein B gene disrupts surfactant homeostasis, causing respiratory failure in newborn mice. Proc Natl Acad Sci U S A 92:7794–7798.

Clark JC, Weaver TE, Iwamoto HS, Ikegami M, Jobe AH, Hull WM & Whitsett JA (1997) Decreased lung compliance and air trapping in heterozygous SP-B-deficient mice. Am J Respir Cell Mol Biol 16:46–52.

Clayton D & McKeigue PM (2001) Epidemiological methods for studying genes and environmental factors in complex diseases. Lancet 358:1356–1360.

Clements JA (1977) Kinetics of surface active material in fetal lung. In: Hudson WA (ed) Development of the lung. Marcel Dekker, New York, pp. 346–366.

Cole FS, Hamvas A & Nogee LM (2001) Genetic disorders of neonatal respiratory function. Pediatr Res 50:157–162.

Compernolle V, Brusselmans K, Acker T, Hoet P, Tjwa M, Beck H, Plaisance S, Dor Y, Keshet E, Lupu F, NEmery B, Dewerchin M, Van Veldoven P, Plate K, Moons L, Collen D & Carmeliet P (2002) Loss of HIF -2alpha and inhibition of VEGF impair felt lung maturation, whereas treatment with VEGF prevents fatal respiratory distress in premature mice. Nat Med 8(7):702–710.

Cooperstock MS, Tummaru R, Bakeqwell J & Schramm W (2000) Twin birth weight discordance and risk of preterm birth. Am J Obstet Gynecol 183:63–67.

Crouch E & Wright JR (2001) Surfactant proteins a and d and pulmonary host defense. Annu Rev Physiol 63:521–554.

Crowley P (2000) Prophylactic corticosteroids for preterm birth. Cochrane Database Syst Rev. The Cochrane Library.

Crowley P (2003) Antenatal corticosteroids-current thinking. BJOG 110 Suppl 20:77–78. Dammann O & Leviton A (1997) Maternal intrauterine infection, cytokines, and brain damage in

the preterm newborn. Pediatr Res 42:1–8. Davis AJ, Jobe AH, Hafner D & Ikegami M (1998) Lung function in premature lambs and rabbits

treated with a recombinant SP-C surfactant. Am J Respir Crit Care Med 157:553–559. de la Chapelle A (1993) Disease gene mapping in isolated human populations: the example of

Finland. J Med Genet 30:857–865. De Paoli AG, Morley C & Davis PG (2003) Nasal CPAP for neonates: what do we know in 2003?

Arch Dis Child Fetal Neonatal Ed 88:F168–F172. DeFelice M, Silberschmidt D, DiLauro R, Xu Y, Wert SE, Weaver TE, Bachurski CJ, Clark JC &

Whitsett JA (2003) TTF-1 phosphorylation is required for peripheral lung morphogenesis, perinatal survival, and tissue specific gene expression. J Biol Chem Jun 26 [Epub ahead of print].

deMello DE, Heyman S, Phelps DS & Floros J (1993) Immunogold localization of SP-A in lungs of infants dying from respiratory distress syndrome. Am J Pathol 142:1631–1640.

Demissie K, Ananth CV, Martin J, Hanley ML, MacDorman MF & Rhoads GG (2002) Fetal and neonatal mortality among twin gestations in the United States: the role of intrapair birth weight discordance. Obstet Gynecol 100:474–480.

71DiAngelo S, Lin Z, Wang G, Phillips S, Ramet M, Luo J & Floros J (1999) Novel, non-radioactive,

simple and multiplex PCR-cRFLP methods for genotyping human SP-A and SP-D marker alleles. Dis Markers 15:269–281.

Donovan EF, Ehrenkranz RA, Shankaran S, Stevenson DK, Wright LL, Younes N, Fanaroff AA, Korones SB, Stoll BJ, Tyson JE, Bauer CR, Lemons JA, Oh W & Papile LA (1998) Outcomes of very low birth weight twins cared for in the National Institute of Child Health and Human Development Neonatal Research Network's intensive care units. Am J Obstet Gynecol 179:742–749.

Elimian A, Verma U, Beneck D, Cipriano R, Visintainer P & Tejani N (2000) Histologic chorioamnionitis, antenatal steroids, and perinatal outcomes. Obstet Gynecol 96:333–336.

Epaud R, Ikegami M, Whitsett JA, Jobe AH, Weaver TE & Akinbi HT (2003) Surfactant protein B inhibits endotoxin-induced lung inflammation. Am J Respir Cell Mol Biol 28:373–378.

Farrell PM & Wood RE (1976) Epidemiology of hyaline membrane disease in the United States: analysis of national mortality statistics. Pediatrics 58:167–176.

Fiori HH, Linderholm B, Fiori RM & Robertson B (2001) Computerized image analysis of bubbles in gastric aspirate for prediction of respiratory distress syndrome. Acta Paediatr 90:1402–1404.

Floros J, Veletza SV, Kotikalapudi P, Krizkova L, Karinch AM, Friedman C, Buchter S & Marks K (1995) Dinucleotide repeats in the human surfactant protein-B gene and respiratory-distress syndrome. Biochem J 305 ( Pt 2):583–590.

Floros J, DiAngelo S, Koptides M, Karinch AM, Rogan PK, Nielsen H, Spragg RG, Watterberg K & Deiter G (1996) Human SP-A locus: allele frequencies and linkage disequilibrium between the two surfactant protein A genes. Am J Respir Cell Mol Biol 15:489–498.

Floros J & Hoover RR (1998) Genetics of the hydrophilic surfactant proteins A and D. Biochim Biophys Acta 1408:312–322.

Floros J, Lin HM, Garcia A, Salazar MA, Guo X, DiAngelo S, Montano M, Luo J, Pardo A & Selman M (2000) Surfactant protein genetic marker alleles identify a subgroup of tuberculosis in a Mexican population. J Infect Dis 182:1473–1478.

Floros J & Fan R (2001) Surfactant protein A and B genetic variants and respiratory distress syndrome: allele interactions. Biol Neonate 80 Suppl 1:22–25.

Friedman SA, Schiff E, Kao L & Sibai BM (1995) Neonatal outcome after preterm delivery for preeclampsia. Am J Obstet Gynecol 172:1785–1788.

Friedman SA, Schiff E, Kao L, Kuint J & Sibai BM (1997) Do twins mature earlier than singletons? Results from a matched cohort study. Am J Obstet Gynecol 176:1193–1196.

Friedman SA, Schiff E, Lubarsky SL & Sibai BM (1999) Expectant management of severe preeclampsia remote from term. Clin Obstet Gynecol 42:470–478.

Furman B, Shoham-Vardi I, Bashiri A, Erez O & Mazor M (2001) Preterm premature rupture of membranes is not an independent risk factor for neonatal morbidity. J Matern Fetal Med 10:107–111.

Gardner MO, Goldenberg RL, Cliver SP, Tucker JM, Nelson KG & Copper RL (1995) The origin and outcome of preterm twin pregnancies. Obstet Gynecol 85:553–557.

Gissler M, Vuori ER & Ritvanen A (2002) Finnish Reproduction Statistics 2000. National Research and Development Centre for Welfare and Health (STAKES) 1–136.

Glasser S.W, Korfhagen TR, Weaver T, Pilot-Matias T, Fox JL & Whitsett JA (1987) cDNA and deduced amino acid sequence of human pulmonary surfactant- associated proteolipid SPL(Phe). Proc Natl Acad Sci U S A 84:4007–4011.

Glasser SW, Burhans MS, Korfhagen TR, Na CL, Sly PD, Ross GF, Ikegami M & Whitsett JA (2001) Altered stability of pulmonary surfactant in SP-C-deficient mice. Proc Natl Acad Sci U S A 98:6366–6371

Gluck L & Kulovich MV (1973) Lecithin-sphingomyelin ratios in amniotic fluid in normal and abnormal pregnancy. Am J Obstet Gynecol 115:539–546.

Glumoff V, Väyrynen O, Kangas T & Hallman M (2000) Degree of lung maturity determines the direction of the interleukin-1- induced effect on the expression of surfactant proteins. Am J Respir Cell Mol Biol 22:280–288.

Goldberg JD, Cohen WR & Friedman EA (1981) Cesarean section indication and the risk of respiratory distress syndrome. Obstet Gynecol 57:30–32.

72Goldenberg RL, Hauth JC & Andrews WW (2000) Intrauterine infection and preterm delivery. N

Engl J Med 342:1500–1507. Gomez R, Ghezzi F, Romero R, Munoz H, Tolosa JE & Rojas I (1995) Premature labor and intra-

amniotic infection. Clinical aspects and role of the cytokines in diagnosis and pathophysiology. Clin Perinatol 22:281–342.

Gonzales LW, Guttentag SH, Wade KC, Postle AD & Ballard PL (2002) Differentiation of human pulmonary type II cells in vitro by glucocorticoid plus cAMP. Am J Physiol Lung Cell Mol Physiol 283:L940–951.

Graven SN & Mesenheimer HR (1965) Respiratory distress syndrome and the high risk mother. Am J Dis Child 109:489–494.

Greci LS, Gilson GJ, Nevils B, Izquierdo LA, Qualls CR & Curet LB (1998) Is amniotic fluid analysis the key to preterm labor? A model using interleukin-6 for predicting rapid delivery. Am J Obstet Gynecol 179:172–178.

Gross I, Smith GJ, Wilson CM, Maniscalco WM, Ingleson LD, Brehier A & Rooney SA (1980) The influence of hormones on the biochemical development of fetal rat lung in organ culture. II. Insulin. Pediatr Res 14:834–838.

Grover TR, Parker TA, Zenge JP, Markham NE, Kinsella JP & Abman SH (2003) Intrauterine hypertension decreases lung VEGF expression and VEGF inhibition causes pulmonary hypertension in the ovine fetus. Am J Physiol Lung Cell Mol Physiol 284(3):L508–517.

Guo X, Lin HM, Lin Z, Montano M, Sansores R, Wang G, DiAngelo S, Pardo A, Selman M & Floros J (2001) Surfactant protein gene A, B, and D marker alleles in chronic obstructive pulmonary disease of a Mexican population. Eur Respir J 18:482–490.

Gurel O, Ikegami M, Chroneos ZC & Jobe AH (2001) Macrophage and type II cell catabolism of SP-A and saturated phosphatidylcholine in mouse lungs. Am J Physiol Lung Cell Mol Physiol 280:L1266–1272.

Guttentag SH, Beers MF, Bieler BM & Ballard PL (1998) Surfactant protein B processing in human fetal lung. Am J Physiol 275:559–566.

Haagsman HP, White RT, Schilling J, Lau K, Benson BJ, Golden J, Hawgood S & Clements JA (1989) Studies of the structure of lung surfactant protein SP-A. Am J Physiol 257:421–429.

Haagsman HP (1998) Interactions of surfactant protein A with pathogens. Biochim Biophys Acta 1408:264–277.

Haataja R, Rämet M, Marttila R & Hallman M (2000) Surfactant proteins A and B as interactive genetic determinants of neonatal respiratory distress syndrome. Hum Mol Genet 9:2751–2760.

Haataja R, Marttila R, Uimari P, Lofgren J, Rämet M & Hallman M (2001) Respiratory distress syndrome: evaluation of genetic susceptibility and protection by transmission disequilibrium test. Hum Genet 109:351–355.

Hackett BP & Gitlin JD (1997) Role of transcription factors in the development of the pulmonary epithelium. In: McDonald JA (ed) Lung Growth and Development. Marcel Dekker Inc., New York, pp. 55–80.

Hacking D, Watkins A, Fraser S, Wolfe R & Nolan T (2001) Respiratory distress syndrome and birth order in premature twins. Arch Dis Child Fetal Neonatal Ed 84:F117–121.

Hafez M, el-Sallab S, Khashaba M, Risk MS, el-Morsy Z, Bassiony MR, el-Kenawy F & Zaghloul W (1989) Evidence of HLA-linked susceptibility gene(s) in respiratory distress syndrome. Dis Markers 7:201–208.

Halliday HL, Ehrenkranz RA & Doyle LW (2003) Early postnatal (<96 hours) corticosteroids for preventing chronic lung disease in preterm infants. Cochrane Database Syst Rev. The Cochrane Library.

Hallman M, Kulovich M, Kirrkpatrick E, Sugarman RG & Gluck L (1976) Phosphatidylinositol and phosphatidylglyserol in amniotic fluid: indices of lung maturity. Am J Obstet Gynecol 125:613–617.

Hallman M (1982) Lung surfactant phospholipids in the foetus, newborn, and in the adult; evidence of abnormality in respiratory failure. Ann Chir Gynaecol Suppl 196:19–23.

Hallman M, Arjomaa P, Mizumoto M & Akino T (1988) Surfactant proteins in the diagnosis of fetal lung maturity. I. Predictive accuracy of the 35 kD protein, the lecithin/sphingomyelin ratio, and phosphatidylglycerol. Am J Obstet Gynecol 158:531–535.

73Hallman M, Merritt TA & Bry K (1994) The fate of exogenous surfactant in neonates with

respiratory distress syndrome. Clin Pharmacokinet 26:215–232. Hallman M (1999) Cytokines, pulmonary surfactant and consequences of intrauterine infection.

Biol Neonate 76 Suppl 1:2–9. Hallman M, Glumoff V & Rämet M (2001) Surfactant in respiratory distress syndrome and lung

injury. Comp Biochem Physiol A Mol Integr Physiol 129:287–294. Hallman M (2002) Is rapid surfactant analysis at birth feasible in the diagnostics of respiratory

distress syndrome? Acta Paediatr 91:370–372. Hamvas A, Wise PH, Yang RK, Wampler NS, Noguchi A, Maurer MM, Walentik CA, Schramm

WF & Cole FS (1996) The influence of the wider use of surfactant therapy on neonatal mortality among blacks and whites. N Engl J Med 334:1635–1640.

Hauth JC, Goldenberg RL, Andrews WW, DuBard MB & Copper RL (1995) Reduced incidence of preterm delivery with metronidazole and erythromycin in women with bacterial vaginosis. N Engl J Med 333:1732–1736.

Hawgood S, Benson BJ, Schilling J, Damm D, Clements JA & White RT (1987) Nucleotide and amino acid sequences of pulmonary surfactant protein SP 18 and evidence for cooperation between SP 18 and SP 28-36 in surfactant lipid adsorption. Proc Natl Acad Sci U S A 84:66–70.

Hawgood S & Clements JA (1990) Pulmonary surfactant and its apoproteins. J Clin Invest 86:1–6. Hawgood S, Derrick M & Poulain F (1998) Structure and properties of surfactant protein B.

Biochim Biophys Acta 1408:150–160. Hawgood S & Poulain FR (2001) The pulmonary collectins and surfactant metabolism. Annu Rev

Physiol 63:495–519. Hitti J, Tarczy-Hornoch P, Murphy J, Hillier SL, Aura J & Eschenbach DA (2001) Amniotic fluid

infection, cytokines, and adverse outcome among infants at 34 weeks' gestation or less. Obstet Gynecol 98:1080–1088.

Hjalmarson O (1981) Epidemiology and classification of acute, neonatal respiratory disorders. A prospective study. Acta Paediatr Scand 70:773–783.

Hjalmarson O & Sandberg K (2002) Abnormal lung function in healthy preterm infants. Am J Respir Crit Care Med 165:83–87.

Hollier LM, McIntire DD & Leveno KJ (1999) Outcome of twin pregnancies according to intrapair birth weight differences. Obstet Gynecol 94:1006–1010.

Hoover RR & Floros J (1998) Organization of the human SP-A and SP-D loci at 10q22-q23. Physical and radiation hybrid mapping reveal gene order and orientation. Am J Respir Cell Mol Biol 18:353–362.

Hoover RR, Thomas KH & Floros J (1999) Glucocorticoid inhibition of human SP-A1 promoter activity in NCI-H441 cells. Biochem J 340 ( Pt 1):69–76.

Hoppe HJ & Reid KB (1994) Collectins--soluble proteins containing collagenous regions and lectin domains--and their roles in innate immunity. Protein Sci 3:1143–1158.

Horbar JD, Badger GJ, Carpenter JH, Fanaroff AA, Kilpatrick S, LaCorte M, Phibbs R & Soll RF (2002) Trends in mortality and morbidity for very low birth weight infants, 1991–1999. Pediatrics 110:143–151.

Hulsey TC, Alexander GR, Robillard PY, Annibale DJ & Keenan A (1993) Hyaline membrane disease: the role of ethnicity and maternal risk characteristics. Am J Obstet Gynecol 168:572–576.

Ikegami M, Lewis JF, Tabor B, Rider ED & Jobe AH (1992) Surfactant protein A metabolism in preterm ventilated lambs. Am J Physiol 262:765–772.

Ikegami M, Korfhagen TR, Whitsett JA, Bruno MD, Wert SE, Wada K & Jobe AH (1998) Characteristics of surfactant from SP-A-deficient mice. Am J Physiol 275:247–254.

Jain D, Dodia C, Bates SR, Hawgood S, Poulain FR & Fisher AB (2003) SP-A is necessary for increased clearance of alveolar DPPC with hyperventilation or secretagogues. Am J Physiol Lung Cell Mol Physiol 284:L759–765.

Jobe A (1983) Respiratory distress syndrome--new therapeutic approaches to a complex pathophysiology. Adv Pediatr 30:93–130.

Jobe AH (1993) Pulmonary surfactant therapy. N Engl J Med 328:861–868.

74Jobe AH, Newnham J, Willet K, Sly P & Ikegami M (1998) Fetal versus maternal and gestational

age effects of repetitive antenatal glucocorticoids. Pediatrics 102:1116–1125. Jobe AH & Ikegami M (2001) Antenatal infection/inflammation and postnatal lung maturation and

injury. Respir Res 2:27–32. Jobe AJ (1999) The new BPD: an arrest of lung development. Pediatr Res 46:641–643. Joseph KS, Marcoux S, Ohlsson A, Liu S, Allen AC, Kramer MS & Wen SW (2001) Changes in

stillbirth and infant mortality associated with increases in preterm birth among twins. Pediatrics 108:1055–1061.

Kala P, Koptides M, DiAngelo S, Hoover RR, Lin Z, Veletza V, Kouretas D & Floros J (1997) Characterization of markers flanking the human SP-B locus. Dis Markers 13:153–167.

Kala P, Ten Have T, Nielsen H, Dunn M & Floros J (1998) Association of pulmonary surfactant protein A (SP-A) gene and respiratory distress syndrome: interaction with SP-B. Pediatr Res 43:169–177.

Kallapur SG, Willet KE, Jobe AH, Ikegami M & Bachurski CJ (2001) Intra-amniotic endotoxin: chorioamnionitis precedes lung maturation in preterm lambs. Am J Physiol Lung Cell Mol Physiol 280:L527–536.

Kaneko K, Shimizu H, Arakawa H & Ogawa Y (2001) Pulmonary surfactant protein A in sera for assessing neonatal lung maturation. Early Hum Dev 62:11–21.

Kaprio J, Verkasalo PK & Koskenvuo M (2000) Sample selection and outcome definition in twin studies: Experiences from the Finnish twin cohort study. In: Spector TD, Snieder H, MacGregor AJ (eds) Advances in twin and sib-pair analysis. Oxford University Press, London, pp. 23–33.

Karinch AM, deMello DE & Floros J (1997) Effect of genotype on the levels of surfactant protein A mRNA and on the SP-A2 splice variants in adult humans. Biochem J 321 ( Pt 1):39–47.

Karinch AM, Deiter G, Ballard PL & Floros J (1998) Regulation of expression of human SP-A1 and SP-A2 genes in fetal lung explant culture. Biochim Biophys Acta 1398:192–202.

Katyal SL, Singh G & Locker J (1992) Characterization of a second human pulmonary surfactant-associated protein SP-A gene. Am J Respir Cell Mol Biol 6:446–452.

Kenyon S, Boulvain M & Neilson J (2003) Antibiotics for preterm rupture of membranes (Cochrane Review). Cochrane Database Syst Rev. The Cochrane Library.

Khoor A, Gray ME, Hull WM, Whitsett JA & Stahlman MT (1993) Developmental expression of SP-A and SP-A mRNA in the proximal and distal respiratory epithelium in the human fetus and newborn. J Histochem Cytochem 41:1311–1319.

Khoury MJ, Marks JS, McCarthy BJ & Zaro SM (1985) Factors affecting the sex differential in neonatal mortality: the role of respiratory distress syndrome. Am J Obstet Gynecol 151:777–782.

Khoury MJ, Calle EE & Joesoef RM (1989) Recurrence of low birth weight in siblings. J Clin Epidemiol 42:1171–1178.

Kilpatrick SJ, Jackson R & Croughan-Minihane MS (1996) Perinatal mortality in twins and singletons matched for gestational age at delivery at > or = 30 weeks. Am J Obstet Gynecol 174:66–71.

King J & Flenady V (2002) Prophylactic antibiotics for inhibiting preterm labour with intact membranes. Cochrane Database Syst Rev. The Cochrane Library.

King R (1984) Isolation and chemical composition of pulmonary surfactant. In: Robertson B, Van Golde L, Batenburg J (eds) Pulmonary surfactant. Elsevier, Amsterdam, pp. 1–15.

King RJ, Simon D & Horowitz PM (1989) Aspects of secondary and quaternary structure of surfactant protein A from canine lung. Biochim Biophys Acta 1001:294–301.

Klein JM, Thompson MW, Snyder JM, George TN, Whitsett JA, Bell EF, McCray PB, Jr. & Nogee LM (1998) Transient surfactant protein B deficiency in a term infant with severe respiratory failure. J Pediatr 132:244–248.

Korfhagen TR, Bruno MD, Glasser SW, Ciraolo PJ, Whitsett JA, Lattier DL, Wikenheiser KA & Clark JC (1992) Murine pulmonary surfactant SP-A gene: cloning, sequence, and transcriptional activity. Am J Physiol 263:546–554.

Korfhagen TR, Bruno MD, Ross GF, Huelsman KM, Ikegami M, Jobe AH, Wert SE, Stripp BR, Morris RE, Glasser SW, Bachurski CJ, Iwamoto HS & Whitsett JA (1996) Altered surfactant function and structure in SP-A gene targeted mice. Proc Natl Acad Sci U S A 93:9594–9599.

75Korfhagen TR, LeVine AM & Whitsett JA (1998) Surfactant protein A (SP-A) gene targeted mice.

Biochim Biophys Acta 1408:296–302. Koskinen R, Merilainen J, Gissler M & Virtanen M (1999) Finnish perinatal statistics 1997-1998.

National Research and Development Centre for Welfare and Health (STAKES) 1–128. Kotecha S (2000) Lung growth: implications for the newborn infant. Arch Dis Child Fetal Neonatal

Ed 82:F69–74. Kumar AR & Snyder JM (1998) Differential regulation of SP-A1 and SP-A2 genes by cAMP,

glucocorticoids, and insulin. Am J Physiol 274:177–185. Laatikainen TJ, Räisänen IJ & Salminen KR (1988) Corticotropin-releasing hormone in amniotic

fluid during gestation and labor and in relation to fetal lung maturation. Am J Obstet Gynecol 159:891–895.

Lahti M, Löfgren,J, Marttila R, Renko M, Klaavuniemi T, Haataja R, Rämet M & Hallman M (2002) Surfactant protein D gene polymorphism associated with severe respiratory syncytial virus infection. Ped Res 51:696–699.

Lankenau HM (1976) A genetic and statistical study of the respiratory distress syndrome. Eur J Pediatr 123:167–177.

Lassus P, Ristimäki A, Ylikorkala O, Viinikka L & Andersson S (1999) Vascular endothelial growth factor in human preterm lung. Am J Respir Crit Care Med 159:1429–1433.

Leveno KJ, Quirk JG, Whalley PJ, Herbert WN & Trubey R (1984) Fetal lung maturation in twin gestation. Am J Obstet Gynecol 148:405–411.

Ley D, Wide-Swensson D, Lindroth M, Svenningsen N & Marsal K (1997) Respiratory distress syndrome in infants with impaired intrauterine growth. Acta Paediatr 86:1090–1096.

Liggins GC & Howie RN (1972) A controlled trial of antepartum glucocorticoid treatment for prevention of the respiratory distress syndrome in premature infants. Pediatrics 50:515–525.

Liley HG, White RT, Benson BJ & Ballard PL (1988) Glucocorticoids both stimulate and inhibit production of pulmonary surfactant protein A in fetal human lung. Proc Natl Acad Sci U S A 85:9096–9100.

Liley HG, White RT, Warr RG, Benson BJ, Hawgood S & Ballard PL (1989) Regulation of messenger RNAs for the hydrophobic surfactant proteins in human lung. J Clin Invest 83:1191–1197.

Lin S, Akinbi HT, Breslin JS & Weaver TE (1996) Structural requirements for targeting of surfactant protein B (SP-B) to secretory granules in vitro and in vivo. J Biol Chem 271:19689–19695.

Lin Z, deMello DE, Wallot M & Floos J (1998) An SP-B gene mutation responsible for SP-B deficiency in fatal congenital alveolar proteinosis: evidence for a mutation hotspot in exon 4. Mol Genet Metab 64:25–35.

Lin Z, Pearson C, Chinchilli V, Pietschmann SM, Luo J, Pison U & Floros J (2000) Polymorphisms of human SP-A, SP-B, and SP-D genes: association of SP-B Thr131Ile with ARDS. Clin Genet 58:181–191.

Lovett SM, Weiss JD, Diogo MJ, Williams PT & Garite TJ (1997) A prospective, double-blind, randomized, controlled clinical trial of ampicillin-sulbactam for preterm premature rupture of membranes in women receiving antenatal corticosteroid therapy. Am J Obstet Gynecol 176:1030–1038.

Löfgren J, Rämet M, Renko M, Marttila R & Hallman M (2002) Association between surfactant protein A gene locus and severe respiratory syncytial virus infection in infants. J Infect Dis 185:283–289.

MacGregor AJ (2000) Practical approaches to account for bias and confounding in twin data. In: Spector TD, Snieder H & MacGregor A.J. (eds) Advances in twin and sib-pair analysis. Greenwich Medical Media Ltd., London, pp. 36–52.

Madsen J, Tornoe I, Nielsen O, Koch C, Steinhilber W & Holmskov U (2003) Expression and localization of lung surfactant protein A in human tissues. Am J Respir Cell Mol Biol. May 3 [Epub ahead of print].

Makri V, Hospes B, Stoll-Becker S, Borkhardt A & Gortner L (2002) Polymorphism of surfactant protein B encoding genes: modifiers of the course of neonatal respiratory distress syndrome? Eur J Pediatr 161: 604–608.

76Malloy MH & Freeman DH (2000) Respiratory distress syndrome mortality in the United States,

1987 to 1995. J Perinatol 20:414–420. Martin N, Boomsma D & Machin G (1997) A twin-pronged attack on complex traits. Nat Genet

17:387–392. McCormack FX (1998) Structure, processing and properties of surfactant protein A. Biochim

Biophys Acta 1408:109–131. McCormick SM, Boggaram V & Mendelson CR (1994) Characterization of mRNA transcripts and

organization of human SP-A1 and SP-A2 genes. Am J Physiol 266:L354–366. McElrath TF, Norwitz ER, Robinson JN, Tanasijevic MJ & Lieberman ES (2000) Differences in

TDx fetal lung maturity assay values between twin and singleton gestations. Am J Obstet Gynecol 182:1110–1112.

McEvoy C, Bowling S, Williamson K, Lozano D, Tolaymat L, Izquierdo L, Maher J & Helfgott A (2002) The effect of a single remote course versus weekly courses of antenatal corticosteroids on functional residual capacity in preterm infants: a randomized trial. Pediatrics 110:280–284.

McIntire DD, Bloom SL, Casey BM & Leveno KJ (1999) Birth weight in relation to morbidity and mortality among newborn infants. N Engl J Med 340:1234–1238.

Melton KR, Nesslein LL, Ikegami M, Tichelaar JW, Clark JC, Whitsett JA & Weaver TE (2003) SP-B deficiency causes respiratory failure in adult mice. Am J Physiol Lung Cell Mol Physiol 285: L543–549.

Mendelson CR, Gao E, Li J, Young PP, Michael LF & Alcorn JL (1998) Regulation of expression of surfactant protein-A. Biochim Biophys Acta 1408:132–149.

Miles PR, Bowman L, Rao KM, Baatz JE & Huffman L (1999) Pulmonary surfactant inhibits LPS-induced nitric oxide production by alveolar macrophages. Am J Physiol 276:186–196.

Moglia BB & Phelps DS (1996) Changes in surfactant protein A mRNA levels in a rat model of insulin-treated diabetic pregnancy. Pediatr Res 39:241–247.

Moutquin JM (2003) Classification and heterogeneity of preterm birth. BJOG 110 Suppl 20:30–33. Moya FR, Montes HF, Thomas VL, Mouzinho AM, Smith JF & Rosenfeld CR (1994) Surfactant

protein A and saturated phosphatidylcholine in respiratory distress syndrome. Am J Respir Crit Care Med 150:1672–1677.

Murphy DJ, Caukwell S, Joels LA & Wardle P (2002) Cohort study of the neonatal outcome of twin pregnancies that were treated with prophylactic or rescue antenatal corticosteroids. Am J Obstet Gynecol 187:483–488.

Myrianthopoulos NC, Churchill JA & Baszynski AJ (1971) Respiratory distress syndrome in twins. Acta Genet Med Gemellol (Roma) 20:199–204.

Nagourney BA, Kramer MS, Klebanoff MA & Usher RH (1996) Recurrent respiratory distress syndrome in successive preterm pregnancies. J Pediatr 129:591–596.

Nassar AH, Usta IM, Rechdan JB, Harb TS, Adra AM & Abu-Musa AA (2003) Pregnancy outcome in spontaneous twins versus twins who were conceived through in vitro fertilization. Am J Obstet Gynecol 189:512–518.

Neilson JP, Verkuyl DA, Crowther CA & Bannerman C (1988) Preterm labour in twin pregnancies: prediction by cervical assessment. Obstet Gynecol 72:719–723.

NIH Consensus Statement (1995) Effect of corticosteroids for fetal maturation on perinatal outcomes. NIH Consensus Development Panel on the Effect of Corticosteroids for Fetal Maturation on Perinatal Outcomes. JAMA 273:413–418.

Nogee LM, de Mello DE, Dehner LP & Colten HR (1993) Brief report: deficiency of pulmonary surfactant protein B in congenital alveolar proteinosis. N Engl J Med 328:406–410.

Nogee LM, Wert SE, Proffit SA, Hull WM & Whitsett JA (2000) Allelic heterogeneity in hereditary surfactant protein B (SP-B) deficiency. Am J Respir Crit Care Med 161:973–981.

Nogee LM, Dunbar AE, III, Wert SE, Askin F, Hamvas A & Whitsett JA (2001) A mutation in the surfactant protein C gene associated with familial interstitial lung disease. N Engl J Med 344:573–579.

Odom MJ & Ballard PL (1997) Developmental and hormonal regulation of the surfactant system. Lung growth and development. pp. 495–556.

Ojomo EO & Coustan DR (1990) Absence of evidence of pulmonary maturity at amniocentesis in term infants of diabetic mothers. Am J Obstet Gynecol 163:954–957.

77Paananen R, Sormunen R, Glumoff V, van Eijk M & Hallman M (2001a) Surfactant proteins A and

D in Eustachian tube epithelium. Am J Physiol Lung Cell Mol Physiol 281:L660–667. Paananen R, Glumoff V, Sormunen R, Voorhout W & Hallman M (2001b) Expression and

localization of lung surfactant protein B in Eustachian tube epithelium. Am J Physiol Lung Cell Mol Physiol 280:L214–220.

Paananen R, Postle AD, Clark G, Glumoff V & Hallman M (2002) Eustachian tube surfactant is different from alveolar surfactant: determination of phospholipid composition of porcine eustachian tube lavage fluid. J Lipid Res 43:99–106.

Patthy L (1991) Homology of the precursor of pulmonary surfactant-associated protein SP- B with prosaposin and sulfated glycoprotein. J Biol Chem 266:6035–6037.

Pelkonen AS, Hakulinen AL & Turpeinen M (1997) Bronchial lability and responsiveness in school children born very preterm. Am J Respir Crit Care Med 156(4 Pt1):1178–1184.

Phillips DI (1993) Twin studies in medical research: can they tell us whether diseases are genetically determined? Lancet 341:1008–1009.

Piazze JJ, Anceschi MM, Maranghi L, Brancato V, Marchiani E & Cosmi EV (1999) Fetal lung maturity in pregnancies complicated by insulin-dependent and gestational diabetes: a matched cohort study. Eur J Gynekol Reprod Biol 83(2):145–150.

Pilot-Matias TJ, Kister SE, Fox JL, Kropp K, Glasser SW & Whitsett JA (1989) Structure and organization of the gene encoding human pulmonary surfactant proteolipid SP-B. DNA 8:75–86.

Piper JM, Xenakis EM, McFarland M, Elliott BD, Berkus MD & Langer O (1996) Do growth-retarded premature infants have different rates of perinatal morbidity and mortality than appropriately grown premature infants? Obstet Gynecol 87:169–174.

Polk DH, Ikegami M, Jobe AH, Sly P, Kohan R & Newnham J (1997) Preterm lung function after retreatment with antenatal betamethasone in preterm lambs. Am J Obstet Gynecol 176:308–315.

Possmayer F, Nag K, Rodriguez K, Qanbar R & Schurch S (2001) Surface activity in vitro: role of surfactant proteins. Comp Biochem Physiol A Mol Integr Physiol 129:209–220.

Poulain FR, Allen L, Williams MC, Hamilton RL & Hawgood S (1992) Effects of surfactant apolipoproteins on liposome structure: implications for tubular myelin formation. Am J Physiol 262:L730–739.

Powers WF & Kiely JL (1994) The risks confronting twins: a national perspective. Am J Obstet Gynecol 170:456–461.

Pryhuber GS, Hull WM, Fink I, McMahan MJ & Whitsett JA (1991) Ontogeny of surfactant proteins A and B in human amniotic fluid as indices of fetal lung maturity. Pediatr Res 30:597–605.

Pryhuber GS, Bachurski C, Hirsch R, Bacon A & Whitsett JA (1996) Tumor necrosis factor-alpha decreases surfactant protein B mRNA in murine lung. Am J Physiol 270:714–721.

Pryhuber GS (1998) Regulation and function of pulmonary surfactant protein B. Mol Genet Metab 64:217–228.

Reed DM, Bakketeig LS & Nugent RP (1978) The epidemiology of respiratory distress syndrome in Norway. Am J Epidemiol 107:299–310.

Rice WR, Ross GF, Singleton FM, Dingle S & Whitsett JA (1987) Surfactant-associated protein inhibits phospholipid secretion from type II cells. J Appl Physiol 63:692–698.

Richardson DK & Torday JS (1994) Racial differences in predictive value of the lecithin/sphingomyelin ratio. Am J Obstet Gynecol 170:1273–1278.

Robertson PA, Sniderman SH, Laros RK, Jr., Cowan R, Heilbron D, Goldenberg RL, Iams JD & Creasy RK (1992) Neonatal morbidity according to gestational age and birth weight from five tertiary care centers in the United States, 1983 through 1986. Am J Obstet Gynecol 166:1629–1641.

Romero R, Shamma F, Avila C, Jimenez C, Callahan R, Nores J, Mazor M, Brekus CA & Hobbins JC (1990) Infection and labor. VI. Prevalence, microbiology, and clinical significance of intraamniotic infection in twin gestations with preterm labor. Am J Obstet Gynecol 163:757–761.

Rooney SA, Young SL & Mendelson CR (1994) Molecular and cellular processing of lung surfactant. FASEB J 8:957–967.

78Rouse DJ, Skopec GS & Zlatnik FJ (1993) Fundal height as a predictor of preterm twin delivery.

Obstet Gynecol 81:211–214. Rubio S, Lacaze-Masmonteil T, Chailley-Heu B, Kahn A, Bourbon JR & Ducroc R (1995)

Pulmonary surfactant protein A (SP-A) is expressed by epithelial cells of small and large intestine. J Biol Chem 270:12162–12169.

Russell RB, Petrini JR, Damus K, Mattison DR & Schwarz RH (2003) The changing epidemiology of multiple births in the United States. Obstet Gynecol 101:129–135.

Rämet M, Haataja R, Marttila R, Floros J & Hallman M (2000) Association between the surfactant protein A (SP-A) gene locus and respiratory-distress syndrome in the Finnish population. Am J Hum Genet 66:1569–1579.

Seifart C, Plagens A, Brodje D, Muller B, von Wichert P & Floros J (2002) Surfactant protein B intron 4 variation in German patiens with COPD and acute respiratory failure. Dis Markers 18: 129–136.

Selman M, Lin HM, Montano M, Jenkins AL, Estrada A, Lin Z, Wang G, DiAngelo SL, Guo X, Umstead TM, Lang CM, Pardo A, Phelps DS & Floros J (2003) Surfactant protein A and B genetic variants predispose to idiopathic pulmonary fibrosis. Hum Genet 113:542–550.

Shah DM, Shenai JP & Vaughn WK (1995) Neonatal outcome of premature infants of mothers with preeclampsia. J Perinatol 15:264–267

Shah SS, Ohlsson A, Halliday H & Shah VS (2003) Inhaled versus systemic corticosteroids for the treatment of chronic lung disease in ventilated very low birth weight preterm infants (Cochrane Review). Cochrane Database Syst Rev. The Cochrane Library.

Sham P (1998) Statistics in human genetics. Arnold Applications of Statistics, London. Sherer DM (2001) Adverse perinatal outcome of twin pregnancies according to chorionicity:

review of the literature. Am J Perinatol 18:23–37. Shima H, Ohshiro K, Taira Y, Miyazaki E, Oue T & Puri P (1999) Antenatal dexamethasone

suppresses tumor necrosis factor-alpha expression in hypoplastic lung in nitrofen-induced diaphragmatic hernia in rats. Pediatr Res 46:633–637.

Shimoya K, Taniguchi T, Matsuzaki N, Moriyama A, Murata Y, Kitajima H, Fujimura M & Nakayama M (2000) Chorioamnionitis decreased incidence of respiratory distress syndrome by elevating fetal interleukin-6 serum concentration. Hum Reprod 15:2234–2240.

Sims EJ, Vermillion ST & Soper DE (2002) Preterm premature rupture of the membranes is associated with a reduction in neonatal respiratory distress syndrome. Am J Obstet Gynecol 187:268–272.

Song M & Phelps DS (2000) Interaction of surfactant protein A with lipopolysaccharide and regulation of inflammatory cytokines in the THP-1 monocytic cell line. Infect Immun 68:6611–6617.

Stuart RB, Ovadia B, Suzara VV, Ross PA, Thelitz S, Fineman JR & Gutierrez JA (2003) Inhaled Nitric Oxide Increases Surfactant Protein Gene Expression in the Intact Lamb. Am J Physiol Lung Cell Mol Physiol 285: L628–633.

Tabor HK, Risch NJ & Myers RM (2002) Opinion: Candidate-gene approaches for studying complex genetic traits: practical considerations. Nat Rev Genet 3:391–397.

Tokieda K, Ikegami M, Wert SE, Baatz JE, Zou Y & Whitsett JA (1999) Surfactant protein B corrects oxygen-induced pulmonary dysfunction in heterozygous surfactant protein B-deficient mice. Pediatr Res 46:708–714.

Tommiska V, Heinonen K, Ikonen S, Kero P, Pokela ML, Renlund M, Virtanen M & Fellman V (2001) A national short-term follow-Up study of extremely low birth weight infants born in Finland in 1996-1997. Pediatrics 107:E2.

Torday J, Carson L & Lawson EE (1979) Saturated phosphatidylcholine in amniotic fluid and prediction of the respiratory-distress syndrome. N Engl J Med 301:1013–1018.

Tough SC, Newburn-Cook C, Johnston DW, Svenson LW, Rose S & Belik J (2002) Delayed childbearing and its impact on population rate changes in lower birth weight, multiple birth, and preterm delivery. Pediatrics 109:399–403.

Tubman TR, Rollins MD, Patterson C & Halliday HL (1991) Increased incidence of respiratory distress syndrome in babies of hypertensive mothers. Arch Dis Child 66:52–54.

79Usher RH, Allen AC & McLean FH (1971) Risk of respiratory distress syndrome related to

gestational age, route of delivery, and maternal diabetes. Am J Obstet Gynecol 111:826–832. van Iwaarden JF, Claassen E, Jeurissen SH, Haagsman HP & Kraal G (2001) Alveolar

macrophages, surfactant lipids, and surfactant protein B regulate the induction of immune responses via the airways. Am J Respir Cell Mol Biol 24:452–458.

Veletza SV, Rogan PK, TenHave T, Olowe SA & Floros J (1996) Racial differences in allelic distribution at the human pulmonary surfactant protein B gene locus (SP-B). Exp Lung Res 22:489–494.

Verder H, Albertsen P, Ebbesen F, Greisen G, Robertson B, Bertelsen B, Agertoft L, Djernes B, Nathan E & Reinholdt J (1999) Nasal continuous positive airway pressure and early surfactant therapy for respiratory distress syndrome in newborns of less than 30 weeks´gestation. Pediatrics 103:E24.

Vermillion ST, Soper DE, Bland ML & Newman RB (2000) Effectiveness of antenatal corticosteroid administration after preterm premature rupture of the membranes. Am J Obstet Gynecol 183:925–929.

Victoria A, Mora G & Arias F (2001) Perinatal outcome, placental pathology, and severity of discordance in monochorionic and dichorionic twins. Obstet Gynecol 97:310–315.

Vorbroker DK, Profitt SA, Nogee LM & Whitsett JA (1995) Aberrant processing of surfactant protein C in hereditary SP-B deficiency. Am J Physiol 268:647–656.

Voss T, Melchers K, Scheirle G & Schafer KP (1991) Structural comparison of recombinant pulmonary surfactant protein SP-A derived from two human coding sequences: implications for the chain composition of natural human SP-A. Am J Respir Cell Mol Biol 4:88–94.

Väyrynen O, Glumoff V & Hallman M (2002) Regulation of surfactant proteins by LPS and proinflammatory cytokines in fetal and newborn lung. Am J Physiol Lung Cell Mol Physiol 282:L803–810.

Walters DV & Olver RE (1978) The role of catecholamines in lung liquid absorption at birth. Pediatr Res 12:239–242.

Wang G, Phelps DS, Umstead TM & Floros J (2000) Human SP-A protein variants derived from one or both genes stimulate TNF-alpha production in the THP-1 cell line. Am J Physiol Lung Cell Mol Physiol 278:946–954.

Wang G, Guo X & Floros J (2003a) Human SP-A 3´-UTR variants mediate differential gene expression in basal levels and in response to dexamethasone. Am J Physiol Lung Cell Mol Physiol 284:L738–748.

Wang G, Christensen ND, Wigdahl B, Guttentag SH & Floros J (2003b) Difference in N-linked glycosylation between the C and T alleles at the SNP1580 of the human surfactant protein B: Disease implication. Biochem J Pt:179–184.

Watterberg KL, Demers LM, Scott SM & Murphy S (1996) Chorioamnionitis and early lung inflammation in infants in whom bronchopulmonary dysplasia develops. Pediatrics 97:210–215.

Wax JR, Herson V, Carignan E, Mather J & Ingardia CJ (2002) Contribution of elective delivery to severe respiratory distress at term. Am J Perinatol 19:81–86.

Weaver TE & Whitsett JA (1989) Processing of hydrophobic pulmonary surfactant protein B in rat type II cells. Am J Physiol 257:100–108.

Weaver TE (1998) Synthesis, processing and secretion of surfactant proteins B and C. Biochim Biophys Acta 1408:173–179.

Weaver TE & Conkright JJ (2001) Function of surfactant proteins B and C. Annu Rev Physiol 63:555–578.

White RT, Damm D, Miller J, Spratt K, Schilling J, Hawgood S, Benson B & Cordell B (1985) Isolation and characterization of the human pulmonary surfactant apoprotein gene. Nature 317:361–363.

Whitsett JA, Nogee LM, Weaver TE & Horowitz AD (1995) Human surfactant protein B: structure, function, regulation, and genetic disease. Physiol Rev 75:749–757.

Whitsett JA & Weaver TE (2002) Hydrophobic surfactant proteins in lung function and disease. N Engl J Med 347:2141–2148.

Winn HN, Romero R, Roberts A, Liu H & Hobbins JC (1992) Comparison of fetal lung maturation in preterm singleton and twin pregnancies. Am J Perinatol 9:326–328.

80Winn HN, Cimino J, Powers J, Roberts M, Holcomb W, Artal R & Amon E (2001) Intrapartum

management of nonvertex second-born twins: a critical analysis. Am J Obstet Gynecol 185:1204–1208.

Wolf EJ, Vintzileos AM, Rosenkrantz TS, Rodis JF, Lettieri L & Mallozzi A (1992) A comparison of pre-discharge survival and morbidity in singleton and twin very low birth weight infants. Obstet Gynecol 80:436–439.

Wolf EJ, Vintzileos AM, Rosenkrantz TS, Rodis JF, Salafia CM & Pezzullo JG (1993) Do survival and morbidity of very-low-birth-weight infants vary according to the primary pregnancy complication that results in preterm delivery? Am J Obstet Gynecol 169:1233–1239.

Wong CJ, Akiyama J, Allen L & Hawgood S (1996) Localization and developmental expression of surfactant proteins D and A in the respiratory tract of the mouse. Pediatr Res 39:930–937.

Ziadeh SM & Badria LF (2000) Effect of mode of delivery on neonatal outcome of twins with birthweight under 1500 g. Arch Gynecol Obstet 264:128–130.