Newborn at High Risk of Brain Damage: Euraibi International Workshop, Siena, April 2001 (Biology of the Neonate)

The Newborn at High Risk of Brain Damage EURope Against Infant Brain Injury (EURAIBI) International Workshop Siena, Italy, April 5–7, 2001

Guest Editor

G. Buonocore, Siena

38 figures and 22 tables, 2001

Basel 폷 Freiburg 폷 Paris 폷 London 폷 New York 폷 New Delhi 폷 Bangkok 폷 Singapore 폷 Tokyo 폷 Sydney

This publication was supported by grants from Chiesi Group and Monte dei Paschi di Siena.

It is dedicated to Rodolfo Bracci, my teacher of neonatology and life. I also wish to thank Jean-Pierre Relier, who made this issue possible. Giuseppe Buonocore

S. Karger Medical and Scientific Publishers Basel 폷 Freiburg 폷 Paris 폷 London New York 폷 New Delhi 폷 Bangkok Singapore 폷 Tokyo 폷 Sydney

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Drug Dosage The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.

All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher or, in the case of photocopying, direct payment of a specified fee to the Copyright Clearance Center (see ‘General Information’). © Copyright 2001 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland) Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel ISBN 3–8055–7212–3

Vol. 79, No. 3–4, 2001

Contents

149 Foreword Relier, J.-P. (Paris) 150 Human Placenta as a Source of Neuroendocrine Factors Reis, F.M.; Florio, P.; Cobellis, L.; Luisi, S.; Severi, F.M.; Bocchi, C.; Picciolini, E.; Centini, G.; Petraglia, F. (Siena) 157 Maternal Risk Factors for Fetal and Neonatal Brain Damage Terzidou, V.; Bennett, P. (London) 163 Fetal Endocrine Signals and Preterm Labor Challis, J.R.G. (Cambridge/Toronto); Smith, S.K. (Cambridge) 168 Influence of Maternal Stress on Fetal Behavior and Brain Development Relier, J.-P. (Paris) 172 Caspase-3 Activation after Neonatal Rat Cerebral Hypoxia-Ischemia Wang, X. (Göteborg/Zhengzhou); Karlsson, J.-O. (Göteborg); Zhu, C. (Göteborg/Zhengzhou); Bahr, B.A. (Storrs, Conn.); Hagberg, H.; Blomgren, K. (Göteborg) 180 Free Radicals and Brain Damage in the Newborn Buonocore, G.; Perrone, S.; Bracci, R. (Siena) 187 Effect of Graded Hypoxia on Cerebral Cortical Genomic DNA Fragmentation

in Newborn Piglets Akhter, W.; Ashraf, Q.M.; Zanelli, S.A.; Mishra, O.P.; Delivoria-Papadopoulos, M. (Philadelphia, Pa.) 194 Is Periventricular Leucomalacia a Result of Hypoxic-Ischaemic Injury?

Hypocapnia and the Preterm Brain Greisen, G. (Copenhagen); Vannucci, R.C. (Hershey, Pa.) 201 Chorioamnionitis and Fetal/Neonatal Brain Injury Toti, P.; De Felice, C. (Siena) 205 New Insights into the Pathogenesis of Pulmonary Inflammation in Preterm

Infants Speer, C.P. (Würzburg) 210 Red Blood Cell Involvement in Fetal/Neonatal Hypoxia Bracci, R.; Perrone, S.; Buonocore, G. (Siena)

© 2001 S. Karger AG, Basel Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Access to full text and tables of contents, including tentative ones for forthcoming issues: www.karger.com/journals/bon/bon_bk.htm

213 Early Markers of Brain Damage in Premature Low-Birth-Weight Neonates

Who Suffered from Perinatal Asphyxia and/or Infection Fotopoulos, S.; Pavlou, K.; Skouteli, H.; Papassotiriou, I.; Lipsou, N.; Xanthou, M. (Athens) 219 Prevention of Bilirubin Encephalopathy Bertini, G.; Dani, C.; Pezzati, M.; Rubaltelli, F.F. (Florence) 224 Inflammatory Mediators and Neonatal Brain Damage Saliba, E.; Henrot, A. (Tours) 228 The Biology of Erythropoietin in the Central Nervous System and Its

Neurotrophic and Neuroprotective Potential Dame, C. (Gainesville, Fla.); Juul, S.E. (Seattle, Wash.); Christensen, R.D. (Gainesville, Fla.) 236 Fetal and Neonatal Cerebral Infarcts Marret, S.; Lardennois, C.; Mercier, A.; Radi, S.; Michel, C.; Vanhulle, C.; Charollais, A. (Rouen); Gressens, P. (Paris) 241 Blood Pressure and Tissue Oxygenation in the Newborn Baby at Risk of

Brain Damage Weindling, A.M.; Kissack, C.M. (Liverpool) 246 Monitoring of Antepartum and Intrapartum Fetal Hypoxemia:

Pathophysiological Basis and Available Techniques Clerici, G.; Luzietti, R.; Di Renzo, G.C. (Perugia) 254 Glutamate in Cerebral Tissue of Asphyxiated Neonates during the First

Week of Life Demonstrated in vivo Using Proton Magnetic Resonance Spectroscopy Groenendaal, F.; Roelants-van Rijn, A.M.; van der Grond, J.; Toet, M.C.; de Vries, L.S. (Utrecht) 258 Resuscitation of the Asphyxic Newborn Infant: New Insight Leads to New

Therapeutic Possibilities Saugstad, O.D. (Oslo) 261 Six Years of Experience with the Use of Room Air for the Resuscitation of

Asphyxiated Newly Born Term Infants Vento, M. (Valencia/Alicante); Asensi, M.; Sastre, J. (Valencia); García-Sala, F. (Valencia/Alicante); Viña, J. (Valencia) 268 Psychological Prevention of Early Pre-Term Birth: A Reliable Benefit Mamelle, N.J. (Lyon) for the PPPB Study Group 274 Pharmacotherapeutical Reduction of Post-Hypoxic-Ischemic Brain Injury in

the Newborn Peeters, C.; van Bel, F. (Utrecht)

281 Author Index Vol. 79, No. 3–4, 2001 282 Subject Index Vol. 79, No. 3–4, 2001 283 Author Index Vol. 79, 2001 285 Subject Index Vol. 79, 2001 after 286 Contents Vol. 79, 2001

148

Biology of the Neonate Vol. 79, No. 3–4, 2001

Contents

Foreword It is a great privilege for Biology of the Neonate to have been chosen to publish the proceedings of the conferences held by the leading specialists in neonatology, particularly those concerned with the subject of brain damage in the newborn child. For easily understandable practical reasons, it was unfortunately not possible to gather together the proceedings of all the conferences in a single issue. The proceedings of the other conferences will be published later in future issues. As editor in chief of this review for the last 20 years and neonatologist for the last 37, it is my pleasant duty to introduce the principal protagonists of this meeting, particularly Rodolfo Bracci and Giuseppe Buonocore of the Università degli Studi di Siena, two research workers who, for 30 years, have done so much to promote this difficult field of research on ‘the cerebral sequelae’ of abnormalities in pregnancy and birth. ‘A tout Seigneur, tout honneur’, as the French say. There being no exact English equivalent, the nearest would perhaps be ‘give credit where credit is due!’ Indeed, Rodolfo Bracci was among the first to elucidate details of the oxygenation mechanisms in the fetus and the newborn child. In 1970, Rodolfo Bracci and others published ‘Hydrogen peroxide generation in the erythrocytes of newborn infants’ in Biology of the Neonate. In 1981, again published in Biology of the Neonate, Bracci directed a magnificent work on ‘Fatty acid pattern of the erythrocyte lipids and plasma vitamin E in the first days of life’. Without going into a detailed history of this essential aspect of fetal oxygenation and the disturbances of its physiology during birth, it seems nevertheless right to cite the work of Maria Delivoria, who, after having elucidated the details of the gas exchanges of fetal haemoglobin, during sojourns in Toronto and Denver, undertook a difficult and often thankless project on the cerebral sequelae of hypoxia-ischaemia at the University of Philadelphia. Ever since 1985, her group certainly has been the most prolific in the Western world, producing between 22 and 26 abstracts at the APS-SPR meetings each year, in addition to numerous original publications in the most prestigious reviews. The 1980s saw the arrival of young intelligent and dynamic researchers who, in bringing the work of their precursors to fruition, allow the highest hopes for the future of perinatology. Three of these young researchers were very active in organizing this meeting and form part of the Editorial Board: Giuseppe Buonocore, Ola Didrik Saugstad and a young obstetrician, Felice Petraglia. Giuseppe Buonocore is the one whom I know best and esteem for his great qualities. A faithful pupil of Rodolfo Bracci, he continues to improve our understanding of the consequences of hypoxia as much as those of hyperoxia. Several stays in Maria Delivoria’s unit in Philadelphia have allowed him to make the best use of the extraordinary capacities of certain fundamental researchers of the University of Siena, particularly in the study of the protein changes in hypoxia. Thus, he and others have just published (1999) an article on ‘Hypoxic response of synaptosomal proteins in term guinea pig fetuses’ in the Journal of Neurochemistry. On the occasion of the retirement of Rodolfo Bracci, it was only right that his successor Giuseppe Buonocore be the principal organizer of a meeting of this quality, reviewing over 30 years of research and establishing the bases of practical applications that appear to be possible at present. Moreover, it is this pragmatism which marks out Ola Didrik Saugstad, well known for 20 years for his work on the physiopathological con-

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 2001 S. Karger AG, Basel

Accessible online at: www.karger.com/journals/bon

sequences of ‘oxygenation’ in all its aspects. His remarkable knowledge of all the publications on the physiology as well as the pathology of this aspect of fetal and neonatal life, associated with an extreme clinical prudence, allows him to bring us all the elements concerning changes in oxygen handling at birth. The presence of Felice Petraglia completes this group effectively, in at last introducing obstetrical research capabilities. Initially the young director of a dynamic group in Udine, Felice Petraglia has traversed the world of developmental biology in Europe and on the American continent, taking an interest as much in studying the consequences of maternal stress as the physiology of the fallopian tube or the complex mechanisms of nidation and placental organization. His choice of such capable speakers allows a more complete approach to all the phenomena leading to fetal distress and premature birth, with all their consequences. It cannot be denied that the considerable progress in these areas has resulted in a simultaneous decrease in perinatal mortality and frequency of neurological handicap. Nonetheless, perinatal asphyxia is still responsible across the world for the death of a million newborn babies and a million severely handicapped children each year. In Europe, out of 4 million births a year, 10,000 infants will suffer from cerebral palsy. The etiology of these cerebral complications is multifactorial. In addition to asphyxia at birth, which has become rare, it has been possible to demonstrate the presence of other risk factors, such as psychological stress of the mother, poor management of minor abnormalities, oxygenation birth stress, inflammation, infection and environmental events during pregnancy or even at birth, such as the pain suffered by the newborn child or the sudden separation from the mother. These last two ‘birth accidents’, that is, sudden and lasting separation of the newborn child from the mother and the insufficient comprehension of pain experienced by the newborn child, are doubtless the origin of postnatal neuropsychic complications that are not possible to evaluate completely. With the daily ever-increasing understanding of all these physiopathological mechanisms, it should be possible to achieve better prevention, which nevertheless turns out to be difficult, in face of the particular physiopathological features specific to each pregnancy. The aim of the EURAIBI (Europe Against Infant Brain Injury) association is to gather together the efforts of all centres of neonatology, nursing care, doctors and scientists, in order to determine the elements allowing a precise definition of these risk factors. Hopefully, these efforts will eventually lead to new preventive and therapeutic approaches. Elements that seem to be important are the preparation for pregnancy, respect for the exceptional creative power of the woman, understanding of the extraordinary power of the fetus for recovery and even of the newborn child, all these being conditional on a good parent-child interaction, especially that between mother and child. At a period when humankind has the impression of being able to dominate technology, it is nevertheless common sense prevention which remains the most effective means to avoid the two great causes of cerebral damage, prematurity and retardation of intrauterine growth. It is certainly this elementary prevention applied from the moment of wanting a child and the idea of the pregnancy that will assure a happy pregnancy and birth. Jean-Pierre Relier, Paris

Biol Neonate 2001;79:150–156

Human Placenta as a Source of Neuroendocrine Factors Fernando M. Reis Pasquale Florio Luigi Cobellis Stefano Luisi Filiberto M. Severi Caterina Bocchi Enrico Picciolini Giovanni Centini Felice Petraglia Chair of Obstetrics and Gynecology, University of Siena, Siena, Italy

Key Words Neurohormones W Placenta W Pregnancy W Labor W Gestational diseases

Abstract Progress in the understanding of the physiological and pathological functions of the placenta introduced the concept that the placenta is a neuroendocrine organ, since it shows local production and release of substances analog to neurohormones. These products act as endocrine, paracrine and autocrine factors to control the secretion of other regulatory molecules, including the pituitary hormones of both mother and fetus and their placental counterparts. Furthermore, they may play a role in the regulation of maternal and fetal physiology during pregnancy, ranging from the control of placental anchoring to fetal growth and maturation, fine regulation of uterine blood flow and/or initiation of labor. All this evidence underlines the decisive contribution of the placenta to all phases of gestation, through a range of substances largely exceeding the classically known sex steroids and chorionic gonadotropin, throughout normal pregnancy as well as in the presence of gestational diseases.

rosteroids, monoamines) (table 1). These products act as endocrine, paracrine and autocrine factors to control the secretion of other regulatory molecules, including the pituitary hormones of both mother and fetus and their placental counterparts. Several findings suggest a role of placental neurohormones in the regulation of maternal and fetal physiology during pregnancy, ranging from the control of placental anchoring to fetal growth and maturation, fine regulation of uterine blood flow and/or initiation of labor. However, how placental hormones interact within the maternal uterus, and which precise functions they exert on the maintenance and/or interruption of pregnancy is still largely unknown. Recent studies have provided evidence for a decisive contribution of the placenta to all phases of gestation, through a range of substances largely exceeding the classically known sex steroids and chorionic gonadotropin. The placenta and its accessory membranes, amnion and chorion, although of fetal origin, actually undertake the role of intermediary barriers and active messengers in the maternal-fetal dialog. In the present chapter, we will introduce the main families of placental neurohormones and summarize the cellular localization, the gestationdependent changes and some putative functions attributed to these substances by experimental and clinical studies.

Copyright © 2001 S. Karger AG, Basel

Thyrotropic Axis Introduction

In the past three decades, there has been an accelerated progress in the understanding of the physiological and pathological functions of the placenta, much of which is owed to the discovery of new placental signaling molecules. The developing concept is that the placenta is a neuroendocrine organ, since it shows local production and release of substances analog to neurohormones (neuropeptides, neu-

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0150$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

Thyrotropin-releasing hormone (TRH) is produced by the placenta from early pregnancy until term. The peptide is localized mainly in the syncytiotrophoblast but also in the fetal and maternal blood vessels musculature as well as in the extravillous trophoblast cells [1]. Placental TRH is secreted into both the maternal and fetal circulation, but concentrations are higher in the latter compartment, probably because the rapid degradation catalyzed by proteases is more

Prof. Felice Petraglia, MD Chair of Obstetrics and Gynecology, University of Siena Policlinico ‘Le Scotte’, viale Bracci I–53100 Siena (Italy) Tel. +39 0577 586 601, Fax +39 0577 233 454, E-Mail [email protected]

Table 1. Neuropeptides, neurosteroids and

monoamines produced by the human placenta

Brain peptides

Pituitary-like peptides and proteins

Neurosteroids

Monoamines and adrenal-like peptides

CRF TRH GHRH GnRH Melatonin Colecistokinin Met-enkephalin Dynorphin Neurotensin VIP Galanin Somatostatin CGRP NPY Substance P Endothelin ANP Renin Angiotensin Urocortin

ACTH TSH GH hPL hCG LH FSH ß-endorphin prolactin oxytocin leptin activin follistatin inhibin

progesterone allopregnanolone pregnenolone sulfate 5·-dihydroprogesterone

epinephrine norepinephrine dopamine serotonin adrenomedullin

VIP = Vasoactive intestinal polypeptide; CGRP = calcitonin gene-related peptide; ANP = atrial natriuretic peptide; LH = luteinizing hormone; FSH = follicle-stimulating hormone.

active on the maternal side [1]. The possible paracrine effects of placental TRH remain unclear. There is very little passage of maternal TRH across the placental barrier, but minimal amounts of TRH can elicit an acute thyroid-stimulating hormone (TSH) release by the fetal pituitary [2]. Since placental TRH is predominantly released into the fetal circulation [1], this placental neuropeptide may be involved in the regulation of thyroid function during fetal life. The placenta also produces a TSH-like peptide, named chorionic thyrotropin (hCT) [3]. Although hCT was initially characterized as a bioactive thyrotropin [3], testing of better purified extracts demonstrated extremely low thyrotropic activity. The levels of hCT progressively increase from the first to third trimester, whereas the levels of TSH remain constant, suggesting an independent regulation of pituitary and placental thyrotropins. In addition, the injection of TRH in pregnant women produces an acute release of TSH but fails to induce placental release of hCT [4].

Growth Hormone, Placental Lactogen and Insulin-Like Growth Factor-I

The classical placental hormone with growth hormone (GH)-like activity is chorionic somatotropin, also named human placental lactogen (hPL) [5]. The dual nomenclature reflects its mammosomatotropic characteristics, due to the structural homology with GH and prolactin (PRL). The hPL molecule is a polypeptide of 191 amino acids with 96% homology with GH but no more than 3% of the somatotropic activity of GH. The levels of hPL in the maternal circulation are very low in early pregnancy and increase progressively,

Human Placenta as a Source of Neuroendocrine Factors

showing some correlation with placental weight [6]. Because the halflife of hPL in maternal plasma is very short, the daily placental production of hPL has to be of great magnitude to maintain circulating levels [6]. The secretion of hPL by the placenta is insensitive to several factors known to affect pituitary GH secretion. Like GH, however, the levels of hPL have been reported to rise following hypoglycemia [6]. The role of hPL in pregnancy is related to its metabolic properties, rather than somatotropic or lactogenic effects. In the fasting state, the augmented release of hPL coupled with low insulin levels results in increased lipolysis and decreased glucose uptake by the mother, which protects the fetus from hypoglycemia. Human placental GH, the product of the GH-V gene, is a GHlike peptide synthesized by the syncytiotrophoblast and released into the maternal circulation, where it gradually replaces the pituitary GH from the second trimester of gestation. At least four splicing variants of the GH-V mRNA are expressed in the human placenta [7]. The peptide exhibits both somatogenic and lactogenic activities and is structurally distinguishable from pituitary GH by 13 amino acids [8]. It is released in a constant rather than pulsatile fashion, in contrast to the episodic release of pituitary GH. This continuous secretion appears to have important implications for physiologic adjustment to gestation and especially the control of maternal insulin-like growth factor-I levels [8]. The placenta also expresses a GH-releasing hormone (GHRH), which is identical to the hypothalamic GHRH but is regulated by distinct mechanisms, including differential splicing of an untranslated exon and the activation of tissue-specific gene promoters [9]. The role of placental GHRH in human pregnancy is still unknown; the presence of GHRH receptor in the placenta [10] suggests a possible paracrine action.

Biol Neonate 2001;79:150–156

151

Placental ACTH, also called chorionic corticotropin, is a product of the proopiomelanocortin (POMC) gene and has the same structure as pituitary ACTH, retaining its immunogenic and biologic activity [18, 19]. Placental ACTH is localized to the cytotrophoblast in the first trimester and to the syncytiotrophoblast in the second and third trimesters [20]. There is a significant increase in POMC gene expression in the placenta with the advance of gestation, which is manifested by increasing levels of POMC mRNA as well as immunoreactive ACTH [20]. Among the possible local effects of placental ACTH are the stimulation of placental steroidogenesis [21] and reduction of vascular resistance [22].

Fig. 1. Paracrine control of hCG release by the placental syncytiotrophoblasts. The stimulatory effects of GnRH and activin and the inhibitory effects of inhibin and follistatin resemble the control of pituitary follicle-stimulating hormone release.

Corticotropic Axis

The placenta produces corticotropin-releasing factor (CRF), corticotropin (ACTH) and cortisol during the whole course of gestation. These hormones have important roles in maternal adaptation and fetal development, in addition to their local effects within the placenta, fetal membranes and myometrium [11]. CRF is located predominantly in the syncytial layer of placental villi, but it is also present in the cytotrophoblast, in the epithelial cells and in some cells of the subepithelial layer of the amnion, and in cells of the reticular layer of the chorion. Placental CRF secretion into maternal plasma progressively increases during pregnancy to reach the highest values at term. The local effects of CRF are mediated by two classes of receptors, named type 1 (CRF-R1) and type 2 (CRF-R2). CRF-R1 encompasses at least two isoforms of proteins, · and ß, while CRF-R2 has at least three splice variants characterized. By using reverse transcriptase polymerase chain reaction and in situ hybridization, the · variant of CRF-R1 mRNA and the ß variant of CRF-R2 mRNA have been identified in the placenta, chorion, amnion and decidua [12, 13]. Specifically, CRF-R1· was exclusively localized in syncytiotrophoblast cells which showed an intense hybridization, whereas CRF-R2ß was also present in the cytotrophoblast. There was no quantitative difference between placentas collected from vaginal delivery or cesarean section [13]. The CRF-R2ß mRNA signal was also present within the structure of the villi (blood vessels), chorionic trophoblast and decidual cells, while a less intense or no signal was found in amniotic epithelium. Recently, a new peptide related to CRF has been found in human placenta, named urocortin [14]. Due to its pharmacological characteristics and tissue distribution in the brain, urocortin may be considered a neuropeptide with high affinity to CRF-R2. The human placenta and its related membranes produce urocortin throughout gestation [14, 15]. Placental and decidual cells collected in the first or third trimester express urocortin mRNA. Immunohistochemistry has localized urocortin in syncytial cells of the trophoblast, as well as in the amnion, chorion and decidua [14]. Studies in vitro have demonstrated that urocortin has CRF-like effects on placental cells and tissue explants, such as stimulating ACTH and prostaglandin secretion [16]. In addition, urocortin has a potent vasodilatory effect on the fetal-placental circulation [17].

152

Biol Neonate 2001;79:150–156

Gonadotropic Axis

Human chorionic gonadotropin (hCG), the most classic placental hormone, is a glycoprotein biologically equivalent to pituitary luteinizing hormone. Like all glycoprotein hormones, hCG is composed of two subunits, · and ß. Since the immunological specificity of hCG is conferred by the ß subunit, current clinical assays employ antibodies directed against the ß subunit in order to quantify hCG without cross-reacting with luteinizing hormone. Produced by the syncytiotrophoblast, hCG is detectable in the maternal circulation 8–10 days after ovulation, coinciding with the implantation of the blastocyst [23]. Although much is known about the mechanisms regulating hCG secretion [23], the role of this hormone in the physiology of pregnancy is still a matter of controversy. Apart from its well-known luteotropic role in early pregnancy, it has been suggested that hCG may regulate steroidogenesis both at the placental level and in the fetal adrenals and testis [6]. It has also been noted that hCG stimulates maternal thyroid function and this probably accounts for a transient hyperthyroidism in the first trimester of gestation, particularly in twin or molar pregnancy [24]. Gonadotropin-releasing hormone (GnRH) and its specific receptor are expressed in the human placenta from the first trimester to term. Abundant expression of both hormone and receptor is observed in the cyto- and syncytiotrophoblast, where they are localized in identical cells [25]. Placental GnRH is likely to be one of the paracrine regulators of hCG secretion, as suggested by evidence from in vivo and in vitro experiments [26, 27]. Administration of GnRH to pregnant women elicits a significant increase in hCG in the first trimester, a response seldom observed in the third trimester [27]. This different sensitivity may be explained by the downregulation of GnRH receptors in the term placenta [28]. In this regard, the influence of local GnRH on hCG secretion is better explained by changes in GnRH receptor expression, which parallel the time course of hCG release during pregnancy [28]. The GnRH receptor expressed in the placenta is identical to the pituitary counterpart [29] and binds with high specificity to synthetic GnRH in vitro. Resembling what occurs in the pituitary-ovary axis, the release of placental GnRH and hCG is modulated by inhibins, activins and the activin-binding protein follistatin (fig. 1). Inhibin and activin are dimers composed of · and ß (inhibin) or two ß subunits (activin), which are coexpressed with GnRH in placental villi at term [30]. Activin stimulates and inhibin antagonizes the activin-induced GnRH and hCG secretion by cultured placental cells [31, 32]. In turn, GnRH and hCG stimulate inhibin mRNA expression and immunoreactive inhibin release in cultured placental cells [33]. The human placenta is the source and target of inhibin-related proteins [34, 35]. Inhibin/activin subunits are present in placental

Reis/Florio/Cobellis/Luisi/Severi/Bocchi/ Picciolini/Centini/Petraglia

cells, particularly in the syncytiotrophoblast. Dimeric activins and inhibins are found in placental homogenates and dimeric activin A is localized in both cyto- and syncytiotrophoblast cells from early to term pregnancy. From early pregnancy, activin A levels are higher than those measured during the menstrual cycle. Minimal changes are seen during the first and second trimesters, but an exponential increase comes with the beginning of the third trimester and is followed by an additional increase toward term. Activin A, inhibin A and inhibin B are also present in the amniotic fluid. Activin A is increased in hypertensive complications of pregnancy, preterm labor and gestational diabetes, and is particularly elevated in preeclampsia. Alterations in maternal serum inhibin levels may be indicative of several gestational diseases. For example, women with preeclampsia have high circulating levels of inhibin A and its precursor subunit pro-· C [35].

Prolactin

The maternal decidua and placental trophoblast produce PRL, which is carried through fetal membranes and released into the amniotic fluid [36]. The levels of amniotic fluid PRL increase in parallel with decidual PRL secretion and reach a peak by the sixth month of pregnancy [37]. Although the decidua is the main site of PRL production in the pregnant uterus, PRL and its receptor are also expressed in the trophoblast throughout gestation [38]. The role of decidual/placental PRL in human pregnancy remains unclear, although some hypotheses regarding fluid homeostasis and fetal lung maturation have been suggested [39].

Oxytocin

Oxytocin gene expression has been identified in the amnion, chorion and decidua, and to a lesser extent in the placental trophoblast. The levels of oxytocin mRNA in the decidua increase markedly around the time of labor onset and do not correlate with oxytocin levels in the maternal circulation [40]. A similar increase can be induced in vitro by estradiol stimulation through an estrogen receptor-mediated mechanism [41]. There appears to be no change in oxytocin metabolism around the time of parturition, but oxytocin receptor gene expression is upregulated [42]. These observations indicate that the onset of labor coincides with an increase in the paracrine rather than systemic release of oxytocin. This local oxytocin production seems to be regulated by other paracrine factors, such as CRF, activin A and prostaglandins [43].

Leptin

Leptin is a peptide secreted in the circulation by the adipose tissue that acts on the brain to regulate food intake and energy balance. This peptide is also produced in nonadipose tissues such as the pituitary gland [44] and the placenta. Placental leptin is released into both maternal and fetal compartments, where its levels increase steadily during late gestation [45, 46]. There is controversy about the relative contribution of the placenta to the maternal plasma leptin content. The time course of maternal plasma leptin levels during pregnancy suggests that the placenta has a decisive role, since leptin levels increase by more than two-fold up to 30 weeks and fall imme-

Human Placenta as a Source of Neuroendocrine Factors

Fig. 2. Stress hormones in the maternal circulation at parturition. The sharp increase in the concentrations of CRF, cortisol and NPY around the time of labor reflects acute placental release. The placenta seems to participate in the stress response of human parturition.

diately after delivery [45]. Leptin levels are substantially increased in women with severe preeclampsia, probably as a consequence of increased placental leptin production [47]. The observation that leptin levels in arterial cord blood have a direct correlation with fetal body weight and are lower in newborns with intrauterine growth restriction (IUGR) suggests that leptin levels in the fetal circulation are determined by the fetal adipose mass [46]; however, the placenta is likely to contribute decisively to the fetal leptin pool, since leptin levels are higher in venous than arterial cord plasma and decrease shortly after birth [48]. The existence of leptin receptors in the placenta supports the possibility of local effects, such as the regulation of other placental neuropeptides [49, 50]. Many aspects of the possible role of leptin in gestational diseases remain unclear. For example, placental leptin production is increased in preeclampsia and local hypoxia may be an underlying mechanism [47], but this contrasts with the low leptin levels observed in IUGR [46].

Vasoactive Neuropeptides

A large number of vasoactive neuropeptides is present in human placental tissues. An extensive immunohistochemical study has revealed the presence of vasoactive intestinal polypeptide, calcitonin gene-related peptide, neuropeptide Y (NPY), galanin, somatostatin, metionine-enkephaline and substance P in the decidual layer, and the expression of endothelin-1 in the trophoblast cells of the human placenta at term [51]. Atrial natriuretic peptide is expressed at both the mRNA and protein level in purified term trophoblasts, particularly in extravillous interstitial and cytotrophoblastic shell trophoblasts [52]. NPY is produced by the placenta, mainly by the cytotrophoblast cells, and is found at high concentrations in maternal plasma with no significant change from early to late gestation [53, 54]. NPY is also present in epithelial amnion cells and in the chorionic cytotrophoblast, and is found at high concentrations in the amniotic fluid. The levels of NPY in maternal plasma (but not in amniotic fluid) increase by three-fold during labor [54], suggesting that this peptide may play a role in the stress response of parturition (fig. 2).

Biol Neonate 2001;79:150–156

153

Table 2. Placental neurohormones involved in gestational diseases

CRF Activin A Inhibin A NPY Leptin GnRH

Preterm labor

Preeclampsia

IUGR

++ + – + – +

++ ++ ++ ++ ++ –

++ – + – – –

Adrenomedullin, a vasoactive peptide initially discovered in pheochromocytoma, has recently been identified in plasma and amniotic fluid samples throughout gestation [55]. Immunoreactive adrenomedullin has been localized predominantly in amnion and extravillous trophoblast cells, although a few syncytiotrophoblast cells and endothelial cells may also express the peptide [56]. Like the brain, the human placenta contains a local renin-angiotensin system that is intrinsically regulated. The placenta produces renin, prorenin, angiotensin I, angiotensin II, angiotensin-converting enzyme, angiotensin receptors and angiotensinogen [57]. Although angiotensin II is a potent vasoconstrictor in peripheral vessels, in the placenta it stimulates the release of vasorelaxants such as nitric oxide and parathyroid hormone-related peptide [58]. The low expression of angiotensin receptors in the placenta may be an important component of placental failure in pregnancies complicated by IUGR [58]. The vasoactive peptides of placental origin appear to participate in the control of local blood flow and to provide a compensatory mechanism to the uteroplacental ischemia of preeclampsia [59].

Neurosteroids and Monoamines

Neurosteroid is a generic denomination applied to the steroid hormones which are synthesized within the nervous system, either de novo from cholesterol, or by the metabolism of precursors obtained from an outside source. The placenta is a source of several neurosteroids, comprising progesterone itself, its derivates 5·-pregnane-3·ol-20-one (allopregnanolone) and 5·-dihydroprogesterone, and its precursor pregnenolone sulfate [60, 61]. The levels of allopregnanolone in maternal serum increase progressively during gestation and,

in contrast to progesterone, are augmented in hypertensive complications of pregnancy [62]. Apart from progesterone, the role of placental neurosteroids in the physiology of pregnancy is largely unknown. These hormones may contribute to the neurochemical and behavioral changes of pregnancy and puerperium, since they interfere with GABAergic circuits and have anxiolytic effects [60]. Placental neurosteroids may also contribute to myometrial quiescence, as suggested by their ability to reduce the contraction frequency of human myometrial strips in vitro [63]. The placenta is a source and target of epinephrine, norepinephrine, dopamine and 5-hydroxitryptamin (serotonin). The enzymes involved in monoamine synthesis and metabolism as well as monoamine transporters and receptors have been identified in the placenta [64–68]. Several studies have suggested that local monoamines participate in the regulation of placental function. The placental metabolism and transport of these neurohormones also has an important role in determining the availability and bioactivity of biogenic amines to both mother and fetus. In preeclampsia, there is an increased activity of tyrosine hydroxylase in placental tissue and this is likely to contribute to the higher levels of catecholamines in the maternal circulation [69]. It has been shown that placental norepinephrine transporter mRNA expression is reduced in some gestational diseases, resulting in increased norepinephrine levels in the fetal circulation [64]. The activity of serotonin transporter in placental cells is suppressed by agonistic stimulation of cannabinoid receptors, indicating that placental clearance of serotonin may account for the adverse effects of cannabinoid use during pregnancy [66].

Conclusions

Placental neurohormones may be involved in preterm labor, preeclampsia and IUGR. In particular, the hypersecretion into the maternal circulation of these neuroendocrine factors has been described (table 2). The hypothesis that they represent a causal factor or more probably merely a sign of a placental reactive adaptive response has been made. Supporting both hypotheses is the evidence that some of those factors may be hypersecreted even before the development of the disease. Augmented maternal CRF levels in women who later have preterm labor [11], or ofactivin A in women who develop preeclampsia [70] indicate that a placental derangement exists before the clinical onset of the disease. The role of the placenta in the pathogenesis of preterm labor or preeclampsia is a growing aspect and needs further evaluation.

References 1 Bajoria R, Babawale M: Ontogeny of endogenous secretion of immunoreactive-thyrotropin releasing hormone by the human placenta. J Clin Endocrinol Metab 1998;83:4148–4155. 2 Bajoria R, Peek MJ, Fisk NM: Maternal-to-fetal transfer of thyrotropin-releasing hormone in vivo. Am J Obstet Gynecol 1998;178:264–269. 3 Hershman JM, Starnes WR: Extraction and characterization of a thyrotropic material from the human placenta. J Clin Invest 1969;48:923–929. 4 Kanazawa S, Nakamura A, Saida K, Tojo S: Placento-thyroidal relationship in normal pregnancy. Acta Obstet Gynecol Scand 1976;55:201–205.

154

5 Josimovich JB, MacLaren JA: Presence in human placenta and term serum of highly lactogenic substance immunologically related to pituitary growth hormone. Endocrinology 1962;71:209–215. 6 Jaffe RB: Neuroendocrine-metabolic regulation of pregnancy; in Yen SSC, Jaffe RB, Barbieri RL (eds): Reproductive Endocrinology, ed 4. Philadelphia, W.B. Saunders, 1998, pp 751–784. 7 Boguszewski CL, Svensson PA, Jansson T, Clark R, Carlsson LM, Carlsson B: Cloning of two novel growth hormone transcripts expressed in human placenta. J Clin Endocrinol Metab 1998;83:2878– 2885.

Biol Neonate 2001;79:150–156

8 Alsat E, Guibourdenche J, Couturier A, Evain Brion D: Physiological role of human placental growth hormone. Mol Cell Endocrinol 1998;140: 121–127. 9 Nogues N, Del Rio JA, Perez Riba M, Soriano E, Flavell RA, Boronat A: Placenta-specific expression of the rat growth hormone-releasing hormone gene promoter in transgenic mice. Endocrinology 1997;138:3222–3227. 10 Gaylinn BD: Molecular and cell biology of the growth hormone-releasing hormone receptor. Growth Horm IGF Res 1999;9(suppl A):37–44.

Reis/Florio/Cobellis/Luisi/Severi/Bocchi/ Picciolini/Centini/Petraglia

11 Reis FM, Fadalti M, Florio P, Petraglia F: Putative role of corticotropin-releasing factor in the mechanisms of human parturition. J Soc Gynecol Investig 1999;6:109–119. 12 Karteris E, Grammatopoulos D, Dai Y, Olah KB, Ghobara TB, Easton A, Hillhouse EW: The human placenta and fetal membranes express the corticotropin-releasing hormone receptor 1alpha (CRH1alpha) and the CRH-C variant receptor. J Clin Endocrinol Metab 1998;83:1376–1379. 13 Florio P, Franchini A, Reis FM, Palumbo M, Ottaviani E, Petraglia F: Human placenta, chorion, amnion and decidua express different variants of corticotropin-releasing factor receptor messenger RNA. Placenta 2000;21:32–37. 14 Petraglia F, Florio P, Gallo R, Simoncini T, Saviozzi M, Di Blasio AM, Vaughan J, Vale W: Human placenta and fetal membranes express human urocortin mRNA peptide. J Clin Endocrinol Metab 1996;81:3807–3810. 15 Watanabe F, Oki Y, Ozawa M, Masuzawa M, Iwabuchi M, Yoshimi T, Nishiguchi T, Iino K, Sasano H: Urocortin in human placenta and maternal plasma. Peptides 1999;20:205–209. 16 Petraglia F, Florio P, Benedetto C, Marozio L, Di Blasio AM, Ticconi C, Piccione E, Luisi S, Genazzani AR, Vale W: Urocortin stimulates placental ACTH and prostaglandin release and myometrial contractility in vitro. J Clin Endocrinol Metab 1999;84:1420–1423. 17 Leitch IM, Boura AL, Botti C, Read MA, Walters WA, Smith R: Vasodilator actions of urocortin and related peptides in the human perfused placenta in vitro. J Clin Endocrinol Metab 1998;83:4510– 4513. 18 Waddell BJ, Burton PJ: Release of bioactive ACTH by perifused human placenta at early and late gestation. J Endocrinol 1993;136:345–353. 19 Smith R, Thomson M: Neuroendocrinology of the hypothalamo-pituitary-adrenal axis in pregnancy and the puerperium. Baillières Clin Endocrinol Metab 1991;5:167–186. 20 Cooper ES, Greer IA, Brooks AN: Placental proopiomelanocortin gene expression, adrenocorticotropin tissue concentrations, and immunostaining increase throughout gestation and are unaffected by prostaglandins, antiprogestins, or labor. J Clin Endocrinol Metab 1996;81:4462–4469. 21 Barnea ER, Lavy G, Fakih H, Decherney AH: The role of ACTH in placental steroidogenesis. Placenta 1986;7:307–313. 22 Clifton VL, Read MA, Boura AL, Robinson PJ, Smith R: Adrenocorticotropin causes vasodilatation in the human fetal-placental circulation. J Clin Endocrinol Metab 1996;81:1406–1410. 23 Muyan M, Boime I: Secretion of chorionic gonadotropin from human trophoblasts. Placenta 1997; 18:237–241. 24 Grun JP, Meuris S, De Nayer P, Glinoer D: The thyrotrophic role of human chorionic gonadotrophin (hCG) in the early stages of twin (versus single) pregnancies. Clin Endocrinol (Oxf) 1997;46: 719–725. 25 Wolfahrt S, Kleine B, Rossmanith WG: Detection of gonadotrophin releasing hormone and its receptor mRNA in human placental trophoblasts using in-situ reverse transcription-polymerase chain reaction. Mol Hum Reprod 1998;4:999–1006. 26 Siler-Khodr TM, Khodr GS: Dose response analysis of GnRH stimulation of hCG release from human term placenta. Biol Reprod 1981;25:353– 358. 27 Iwashita M, Kudo Y, Shinozaki Y, Takeda Y: Gonadotropin-releasing hormone increases serum human chorionic gonadotropin in pregnant women. Endocr J 1993;40:539–544.

Human Placenta as a Source of Neuroendocrine Factors

28 Lin LS, Roberts VJ, Yen SS: Expression of human gonadotropin-releasing hormone receptor gene in the placenta and its functional relationship to human chorionic gonadotropin secretion. J Clin Endocrinol Metab 1995;80:580–585. 29 Boyle TA, Belt-Davis DI, Duello TM: Nucleotide sequence analyses predict that human pituitary and human placental gonadotropin-releasing hormone receptors have identical primary structures. Endocrine 1998;9:281–287. 30 Petraglia F, Woodruff TK, Botticelli G, Botticelli A, Genazzani AR, Mayo KE, Vale W: Gonadotropin-releasing hormone, inhibin, and activin in human placenta: Evidence for a common cellular localization. J Clin Endocrinol Metab 1992;74: 1184–1188. 31 Steele GL, Currie WD, Yuen BH, Jia XC, Perlas E, Leung PC: Acute stimulation of human chorionic gonadotropin secretion by recombinant human activin-A in first trimester human trophoblast. Endocrinology 1993;133:297–303. 32 Petraglia F, Vaughan J, Vale W: Inhibin and activin modulate the release of gonadotropin-releasing hormone, human chorionic gonadotropin, and progesterone from cultured human placental cells. Proc Natl Acad Sci USA 1989;86:5114–5117. 33 Qu J, Thomas K: Inhibin and activin production in human placenta. Endocr Rev 1995;16:485–507. 34 Petraglia F, Sawchenko P, Lim AT, Rivier J, Vale W: Localization, secretion, and action of inhibin in human placenta. Science 1987;237:187–189. 35 Petraglia F: Inhibin, activin and follistatin in the human placenta – a new family of regulatory proteins. Placenta 1997;18:3–8. 36 Wu WX, Brooks J, Millar MR, Ledger WL, Saunders PT, Glasier AF, McNeilly AS: Localization of the sites of synthesis and action of prolactin by immunocytochemistry and in-situ hybridization within the human utero-placental unit. J Mol Endocrinol 1991;7:241–247. 37 Wu WX, Brooks J, Glasier AF, McNeilly AS: The relationship between decidualization and prolactin mRNA and production at different stages of human pregnancy. J Mol Endocrinol 1995;14:255– 261. 38 Frasor J, Gaspar CA, Donnelly KM, Gibori G, Fazleabas AT: Expression of prolactin and its receptor in the baboon uterus during the menstrual cycle and pregnancy. J Clin Endocrinol Metab 1999;84:3344–3350. 39 Ben-Jonathan N, Mershon JL, Allen DL, Steinmetz RW: Extrapituitary prolactin: Distribution, regulation, functions, and clinical aspects. Endocr Rev 1996;17:639–669. 40 Chibbar R, Miller FD, Mitchell BF: Synthesis of oxytocin in amnion, chorion, and decidua may influence the timing of human parturition. J Clin Invest 1993;91:185–192. 41 Chibbar R, Wong S, Miller FD, Mitchell BF: Estrogen stimulates oxytocin gene expression in human chorio-decidua. J Clin Endocrinol Metab 1995;80: 567–572. 42 Takemura M, Kimura T, Nomura S, Makino Y, Inoue T, Kikuchi T, Kubota Y, Tokugawa Y, Nobunaga T, Kamiura S: Expression and localization of human oxytocin receptor mRNA and its protein in chorion and decidua during parturition. J Clin Invest 1994;93:2319–2323. 43 Florio P, Lombardo M, Gallo R, Di Carlo C, Sutton S, Genazzani AR, Petraglia F: Activin A, corticotropin-releasing factor and prostaglandin F2 alpha increase immunoreactive oxytocin release from cultured human placental cells. Placenta 1996;17:307–311.

44 Jin L, Burguera BG, Couce ME, Scheithauer BW, Lamsan J, Eberhardt NL, Kulig E, Lloyd RV: Leptin and leptin receptor expression in normal and neoplastic human pituitary: Evidence of a regulatory role for leptin on pituitary cell proliferation. J Clin Endocrinol Metab 1999;84:2903–2911. 45 Sivan E, Whittaker PG, Sinha D, Homko CJ, Lin M, Reece EA, Boden G: Leptin in human pregnancy: The relationship with gestational hormones. Am J Obstet Gynecol 1998;179:1128–1132. 46 Jaquet D, Leger J, Levy Marchal C, Oury JF, Czernichow P: Ontogeny of leptin in human fetuses and newborns: Effect of intrauterine growth retardation on serum leptin concentrations. J Clin Endocrinol Metab 1998;83:1243–1246. 47 Mise H, Sagawa N, Matsumoto T, Yura S, Nanno H, Itoh H, Mori T, Masuzaki H, Hosoda K, Ogawa Y, Nakao K: Augmented placental production of leptin in preeclampsia: Possible involvement of placental hypoxia. J Clin Endocrinol Metab 1998; 83:3225–3229. 48 Yura S, Sagawa N, Mise H, Mori T, Masuzaki H, Ogawa Y, Nakao K: A positive umbilical venousarterial difference of leptin level and its rapid decline after birth. Am J Obstet Gynecol 1998;178: 926–930. 49 Bodner J, Ebenbichler CF, Wolf HJ, Muller-Holzner E, Stanzl U, Gander R, Huter O, Patsch JR: Leptin receptor in human term placenta: In situ hybridization and immunohistochemical localization. Placenta 1999;20:677–682. 50 Henson MC, Swan KF, O’Neil JS: Expression of placental leptin and leptin receptor transcripts in early pregnancy and at term. Obstet Gynecol 1998; 92:1020–1028. 51 Graf AH, Hutter W, Hacker GW, Steiner H, Anderson V, Staudach A, Dietze O: Localization and distribution of vasoactive neuropeptides in the human placenta. Placenta 1996;17:413–421. 52 Graham CH, Watson JD, Blumenfeld AJ, Pang SC: Expression of atrial natriuretic peptide by third-trimester placental cytotrophoblasts in women. Biol Reprod 1996;54:834–840. 53 Petraglia F, Calza L, Giardino L, Sutton S, Marrama P, Rivier J, Genazzani AR, Vale W: Identification of immunoreactive neuropeptide-Y in human placenta: Localization, secretion, and binding. Endocrinology 1989;124:2016–2022. 54 Petraglia F, Coukos G, Battaglia C, Bartolotti A, Volpe A, Nappi C, Segre A, Genazzani AR: Plasma and amniotic fluid immunoreactive neuropeptideY level changes during pregnancy, labor, and at parturition. J Clin Endocrinol Metab 1989;69: 324–328. 55 Di Iorio R, Marinoni E, Letizia C, Villaccio B, Alberini A, Cosmi EV: Adrenomedullin production is increased in normal human pregnancy. Eur J Endocrinol 1999;140:201–206. 56 Marinoni E, Di Iorio R, Letizia C, Villaccio B, Scucchi L, Cosmi EV: Immunoreactive adrenomedullin in human fetoplacental tissues. Am J Obstet Gynecol 1998;179:784–787. 57 Poisner AM: The human placental renin-angiotensin system. Front Neuroendocrinol 1998;19:232– 252. 58 Li X, Shams M, Zhu J, Khalig A, Wilkes M, Whittle M, Barnes N, Ahmed A: Cellular localization of AT1 receptor mRNA and protein in normal placenta and its reduced expression in intrauterine growth restriction. Angiotensin II stimulates the release of vasorelaxants. J Clin Invest 1998;101: 442–454. 59 Holst N, Oian P, Aune B, Jenssen TG, Burhol PG: Increased plasma levels of vasoactive intestinal polypeptide in pre-eclampsia. Br J Obstet Gynaecol 1991;98:803–806.

Biol Neonate 2001;79:150–156

155

60 Dombroski RA, Casey ML, MacDonald PC: 5Alpha-dihydroprogesterone formation in human placenta from 5alpha-pregnan-3beta/alpha-ol-20ones and 5-pregnan-3beta-yl-20-one sulfate. J Steroid Biochem Mol Biol 1997;63:155–163. 61 Le Goascogne C, Eychenne B, Tonon MC, Lachapelle F, Baumann N, Robel P: Neurosteroid progesterone is up-regulated in the brain of jimpy and shiverer mice. Glia 2000;29:14–24. 62 Luisi S, Petraglia F, Benedetto C, Nappi RE, Bernardi F, Fadalti M, Reis FM, Luisi M, Genazzani AR: Serum allopregnanolone levels in pregnant women: Changes during pregnancy, at delivery, and in hypertensive patients. J Clin Endocrinol Metab 2000;85:2429–2433. 63 Lofgren M, Holst J, Backstrom T: Effects in vitro of progesterone and two 5 alpha-reduced progestins, 5 alpha-pregnane-3,20-dione and 5 alphapregnane-3 alpha-ol-20-one, on contracting human myometrium at term. Acta Obstet Gynecol Scand 1992;71:28–33.

156

64 Bzoskie L, Yen J, Tseng YT, Blount L, Kashiwai K, Padbury JF: Human placental norepinephrine transporter mRNA: Expression and correlation with fetal condition at birth. Placenta 1997;18: 205–210. 65 Vaillancourt C, Petit A, Belisle S: Expression of human placental D2-dopamine receptor during normal and abnormal pregnancies. Placenta 1998; 19:73–80. 66 Kenney SP, Kekuda R, Prasad PD, Leibach FH, Devoe LD, Ganapathy V: Cannabinoid receptors and their role in the regulation of the serotonin transporter in human placenta. Am J Obstet Gynecol 1999;181:491–497. 67 Nguyen TT, Tseng YT, McGonnigal B, Stabila JP, Worrell LA, Saha S, Padbury JF: Placental biogenic amine transporters: In vivo function, regulation and pathobiological significance. Placenta 1999; 20:3–11.

Biol Neonate 2001;79:150–156

68 Falkay G, Kovacs L: Expression of two alpha 2adrenergic receptor subtypes in human placenta: Evidence from direct binding studies. Placenta 1994;15:661–668. 69 Manyonda IT, Slater DM, Fenske C, Hole D, Choy MY, Wilson C: A role for noradrenaline in preeclampsia: Towards a unifying hypothesis for the pathophysiology. Br J Obstet Gynaecol 1998;105: 641–648. 70 Lambert-Messerlian GM, Silver HM, Petraglia F, Luisi S, Pezzani I, Maybruck WM, Hogge WA, Hanley-Yanez K, Roberts JM, Neveux LM, Canick JA: Second-trimester levels of maternal serum human chorionic gonadotropin and inhibin a as predictors of preeclampsia in the third trimester of pregnancy. J Soc Gynecol Investig 2000;7:170– 174.

Reis/Florio/Cobellis/Luisi/Severi/Bocchi/ Picciolini/Centini/Petraglia

Biol Neonate 2001;79:157–162

Maternal Risk Factors for Fetal and Neonatal Brain Damage Vasso Terzidou Phillip Bennett Institute of Reproductive and Developmental Biology, Hammersmith Hospital, Imperial College School of Medicine, London, UK

Key Words Preterm labour W Infection W Twins W Placenta W Asphyxia W Cerebral palsy

Abstract Prematurity is probably the major factor associated with brain damage in newborns. Our growing knowledge of the biochemical mechanisms leading to the onset of labour at term allows the biochemical correlates of the epidemiological risk factors for prematurity to be understood. Infection is the major cause of early preterm labour and is now recognised to be a major cause of fetal cerebral damage leading to cerebral palsy. Only some 5% of cerebral palsy is due to intrapartum asphyxia at term. This may occur due to an obstetric catastrophe or through inadequate placental function leading to chronic intrapartum asphyxia. Copyright © 2001 S. Karger AG, Basel

Introduction

Prematurity is probably the major factor associated with brain damage in newborns [1–3]. This was recognised as early as 1843 when William John Little, a London orthopaedic surgeon, published his lectures on Deformities of the Human Frame, in which he noted the association between prematurity and severe asphyxia with convulsions. Among those infants born before 30 weeks who survive, approximately 25% will have a major disability [4]. Another 10% will have some disability and an additional 30% will have cognitive, perceptual and behavioural problems that could interfere with school performance [5–7]. Intraventricular haemorrhage (IVH) occurs in up to 40% of infants born before 35 weeks or with a birth weight ! 1,500 g, as compared with 3–4% of babies born at term [8]. Approximately 5% of infants with IVH will develop cerebral palsy [9]. Preterm new-

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0157$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

borns seem to be at a higher risk for cerebral white matter damage. Neonatal white matter damage, which has been subdivided into periventricular haemorrhagic infarction and periventricular leucomalacia (PVL) on the basis of neuropathological findings, appears to count more in predicting long-term developmental disability.

Mechanisms of Labour

During pregnancy, the uterus expands to accommodate the growing fetus and placenta, without increasing contractility, whilst the cervix remains firm and closed. Throughout pregnancy, ‘pro-pregnancy’ factors operate to inhibit myometrial contractility and to allow myometrial hypertrophy, until, near to term, ‘pro-labour’ factors begin to operate to mediate remodelling of the cervix, to allow it to efface and dilate, and the uterus is stimulated to begin co-ordinated contractions. Steven Lye from Toronto [10] has suggested that labour is the result of the activation of a ‘cassette of contraction-associated proteins’ which act to convert the myometrium from a state of quiescence to a state of contractility. These might include gap junction proteins, oxytocin and prostaglandin receptors, enzymes for the synthesis of prostaglandins or cytokines and also components of cell signalling mechanisms which affect the way in which the uterus responds to receptor activation. It is likely that the factors which control the activation of the ‘cassette of contraction-associated proteins’ also activate factors in the fetal membranes that lead to the production of prostaglandins and cytokines associated with labour, and factors within the cervix which lead to cervical remodelling and ripening.

Phases of Parturition

We currently consider pregnancy to be divided into four parturitional phases. During the first parturitional phase (phase 0), the uter-

Prof. Phillip Bennett Imperial College School of Medicine Institute of Reproductive and Developmental Biology, Hammersmith Hospital Du Cane Road, London W12 0NN (UK) Tel. +44 208 759 42141, Fax +44 208 759 42154, E-Mail [email protected]

us is under strong progesterone repression. During the second phase (phase 1), rising oestrogen and corticotrophin-releasing hormone (CRH) concentrations activate proteins such as cell surface receptors and gap junctions which will be needed for labour itself. CRH also increases the expression of interleukin (IL)-1ß and cyclooxygenase-2. Both McLean et al. [11] and Jones et al. [12] have suggested that the timing of labour is mediated through placental release of CRH, whose concentration in maternal plasma begins to rise about 90 days prior to the onset of labour. Circulating CRH exists principally in an inactive form, bound to a binding protein, CRH-BP, derived from the maternal liver. CRH-BP is itself negatively regulated by CRH, so that as the placental release of CRH increases, maternal liver release of CRH-BP decreases. At about 37 weeks in the average pregnancy, CRH concentrations begin to exceed those of CRH-BP, with, effectively, a large increase in the concentration of biologically active CRH. An important difference between pituitary CRH and placental CRH is that whilst pituitary CRH is under the negative feedback control of cortisol, the placental release of CRH is increased by cortisol [13]. Labour itself arises because a relatively rapid increase in the synthesis of inflammatory mediators and the consequent influx of inflammatory cells causes cervical ripening and stimulates uterine contractions. Intrauterine infection, with associated maternal and fetal inflammatory responses, may account for up to 30% of cases of preterm birth [14]. One of the basic tenets of infection-associated preterm labour is that a number of inflammatory cytokines and other immunomodulatory proteins are released as a response to infection, which in turn affect the local production of autocrine and paracrine uterotonic factors [15, 16]. The results of many studies indicate that bacterial infection of the extraplacental membranes is a potential cause of a significant subgroup of preterm deliveries. They also indicate that although intra-amniotic infections are not always associated with chorioamnionitis, subclinical intrauterine infection appears to be a frequently occurring causative factor. Similarly, there is convincing biochemical and physiologic evidence that the production of cytokines by gestational tissues plays a key role in the mechanism of infection-driven preterm labour. Evidence provided by Romero [15] as well as other investigators has demonstrated a critical role of IL-1 in preterm deliveries complicated by maternal gram-negative infections [17]. Initially, Fidel et al. [18] showed that both IL-1· and IL-1ß were elevated in the amniotic fluid of women presenting with preterm uterine contractions in association with intrauterine bacterial infections. IL-1ß has been shown to exist in abundance in uterine decidua and seems to be a potent stimulus for prostaglandin E2 biosynthesis in human amnion [19, 20]. Expression of IL-1ß at the mRNA level is higher in fetal membranes after labour than before. IL-1, as well as tumour necrosis factor (TNF), is considered to be an early cytokine in the inflammatory response. These early cytokines also stimulate the production of chemokines. IL-6, a primary proinflammatory cytokine (originally identified as interferon-ß), plays a central role in nearly all acute inflammatory responses. Amniotic fluid concentrations of IL-6 appear to correlate with inevitable preterm delivery in association with intrauterine infection better than other parameters [21]. In addition, elevation of serum IL-6 has also been found to be associated with infection-associated preterm labour and inevitable preterm birth [22]. Greig et al. [23] subsequently found that serum IL-6 concentrations were elevated in women with preterm premature rupture of the membranes who delivered soon after membrane rupture.

158

Biol Neonate 2001;79:157–162

Chemokines, as their name implies, are a superfamily of cytokines that are involved in the chemoattraction and infiltration of different types of inflammatory cells (neutrophils, monocytes, macrophages, lymphocytes, mast cells and eosinophils). The first chemokine to be evaluated in pregnancy was IL-8 (neutrophil attractantactivating peptide-1). The levels of IL-8 in the amniotic fluid of women who presented with preterm labour seem to correlate with the concentration of neutrophils invading the amniotic cavity [24].There also seems to be a positive correlation with bacterial colonisation and histological chorioamnionitis, suggesting that IL-8 may attract neutrophils into the uterus with the presence of infection. Laham et al. [25] demonstrated (in vitro) that gestational tissues were capable of synthesising IL-8. Recently, a few studies have suggested the potential role of IL-8 in the ripening of the uterine cervix in preparation for labour. Barclay et al. [26] demonstrated IL-8 immunoreactivity in the human cervix at the time of parturition. These investigations suggest that IL-8 may serve as a potent stimulus for the infiltration of neutrophils and remodelling of the cervical extracellular matrix via elastase and collagenase activities [27].

The Cervix

The cervix is composed mainly of a connective tissue of collagen and elastin embedded in a proteoglycan matrix with very little smooth muscle. The proteoglycan matrix of the cervix consists of heteropolysaccharide side-chains (glycosaminoglycans) covalently linked to a central polypeptide core. At term, near to labour, the collagen of the cervix changes, undergoing collagenolysis. The fibrils become dissociated from their tightly organised bundles and are more widely scattered in an increased amount of ground substance, and there is loosening of the collagen bundles in the cervical stroma. There are also changes in the concentrations of glycosaminoglycans in the cervical matrix at the time of parturition. These biochemical changes in the cervix are reflected as changes in its physical properties [28]. Much of the change in the cervix at term is due to increased collagenase activity. Fibroblasts present in the cervix are capable of synthesising collagenase, but there is also an accumulation of neutrophils which release collagenase into the cervix around the time of parturition. Cervical ripening therefore resembles an inflammatory reaction. It is currently thought that neutrophils are attracted into the cervix at term by the combination of increased prostaglandin synthesis and the neutrophil attractant peptide IL-8, whose expression may be mediated by NFÎB [28–31].

Risk Factors for Preterm Labour

Understanding the mechanisms leading to labour at term helps in our understanding of risk factors for preterm delivery. Many of the epidemiological risk factors will have obvious biochemical correlates. For example, multiple pregnancy has an epidemiological association with prematurity which correlates with its likely effects upon the placental CRH clock and the effects of myometrial stretch. Just as epidemiological risk factors may be additive, so their biochemical risk factors may combine. For example, a twin pregnancy alone will not explain preterm delivery at 28 weeks, but may be a factor in combination with abnormal genital tract bacterial flora, stress or cervical incompetence.

Terzidou/Bennett

Stress

There is growing evidence that either maternal or fetal stress may be associated with preterm labour. Major life events have an association with prematurity, as does the fetal stress of intrauterine growth retardation. A link between maternal stress and preterm labour is suggested by its increased prevalence among unmarried and poor mothers as well as in stressful sociodemographic conditions (such as loss of employment, housing or partner) [32]. Prematurity is also more common among women reporting increased stress or anxiety [33]. The link between fetal stress and preterm labour is suggested by an increased occurrence of placental vascular lesions and intrauterine growth retardation among patients delivering preterm without evidence of intra-amniotic infection [34]. The biochemical pathway through which maternal and fetal stress promotes preterm delivery appears to involve a premature increase in placental, decidual and amniochorionic expression of CRH [35, 36]. Since placental CRH is up-regulated by corticosteroids [12], chronic up-regulation of maternal or fetal corticosteroid concentrations would lead to an early rise in CRH, signalling the early onset of labour. Maternal stress is associated with increased plasma cortisol, and hypoxaemia-induced activation of the fetal hypothalamic axis is associated with an increase in fetal adrenocorticotrophic hormone and cortisol in humans [37], which results in the enhanced CRH production. Moreover, other factors released in response to maternal stress and fetal hypoxaemia, including norepinephrine, angiotensin II, arginine-vasopressin and acetylcholine, also enhance the placental release of CRH [38]. Warren et al. [39] showed higher levels of mean CRH in women who delivered preterm without infection than in women with infectionassociated preterm delivery or those at the same gestational age subsequently delivering at term. It seems possible that sustained stress can have a deleterious effect on the developing brain through different pathways. Not only is it related to preterm delivery, but glucocorticoids (GCs) seem also capable of damaging the hippocampus, a principal neural target site for hormones [40]. Excessive exposure to GCs is capable of endangering rodent hippocampal neurons, impairing their capacity to survive insults such as seizure, hypoxia-ischaemia and hypoglycaemia [41]. There has not been good evidence that GCs have a similar effect in the human brain, but two studies, one in patients with major depression and one in patients with Cushing’s syndrome (both groups have higher circulating levels of GCs), have showed a significant reduction in the volume of hippocampi in both groups when compared with appropriate controls [42]. Clearly, these findings remain preliminary and indicate a complex relationship between perceived stress in pregnancy, the hypothalamic axis, preterm delivery and the possible effects of persistently high levels of GCs. The impact of fetal and maternal stress on fetal/ neonatal brain development remains to be examined more fully in prospective controlled trials.

Multiple Pregnancy and Uterine Distension

The concept that the length of pregnancy is controlled by a placental product, CRH, helps to explain the onset of preterm labour in the case of multiple pregnancy. It would be expected that a larger placental mass would generate higher levels of CRH. Down-regulation of CRH-BP would then occur prematurely, leading to an early rise in free, active CRH in the maternal circulation. If parturitional

Maternal Risk Factors for Fetal and Neonatal Brain Damage

phase 3 is signalled simply by a certain threshold of CRH, or of IL-1ß stimulated by CRH, then the onset of labour would occur earlier. In higher-order multiples, this may be sufficient even if pulmonary maturity associated with platelet activating factor release had not been reached. Although progesterone protects the uterus against the effects of stretch, it is probable that in severe polyhydramnios, the degree of stretch is such that the up-regulation of contraction-associated proteins nevertheless occurs early. It is likely that in addition to gap junction proteins, other contraction-associated proteins will be found to be stretch sensitive [43, 44]. Studies during the last decade [32, 45–47] have shown a 6- to 7-fold increased risk of cerebral palsy in twins. The prevalence of cerebral palsy rose with decreasing birth weight. However, the birth weightspecific prevalence among those of low birth weight (!2,500 g) was not significantly different in singleton compared to multiple births. Among infants weighing 1 2,500 g, there was a significantly higher risk in multiple births. Multiple logistic regression analysis confirmed the importance of twinning as a risk factor for cerebral palsy, regardless of weight and gestational age [47]. The risk is higher for triplets and increases 100-fold for a twin whose co-twin dies in utero [45]. It has been suggested that twinning confers an increased risk of cerebral palsy because of the increased vulnerability of blood flow in utero [48]. It also appears that all monochorionic twins have a higher risk of antenatally acquired cerebral lesions. Bejar et al. [49] found that 30% of monochorionic twins who were delivered before 36 weeks had antenatally acquired necrosis of cerebral white matter. Cerebral necrosis was associated with the presence of VV and AA anastomoses. Some series suggest that the incidence of cerebral lesions is at least similar if not higher in twin-to-twin transfusion syndrome [50].

Cervical Incompetence

Cervical incompetence is considered a possible factor in the aetiology of premature deliveries and recurrent spontaneous abortions. As an entity, it was first described by Rivierus as early as 1658. The diagnosis is suggested by a history of painless second trimester pregnancy loss or when a patient presents in the mid-trimester with cervical dilatation. More recently, ultrasound has been used to evaluate the state of the cervix, with encouraging results [51, 52]. It is probable that many cases of early preterm labour, thought to be associated with cervical incompetence, occur because vaginal pathogens enter the lower pole of the uterus through a modestly open cervix, where they set up an inflammatory reaction.

Infection

Preterm labour and preterm prelabour rupture of membranes appear to be related to maternal infection [15]. For example, chorioamnionitis is associated with a 3-fold increased risk of delivery before term with intact membranes and a 4-fold increased risk with ruptured membranes [53]. Bacterial vaginosis at 23–26 weeks of gestation appears to increase a woman’s risk for preterm delivery by 50% [54]. A randomised trial comparing women with bacterial vaginosis who received antibiotic treatment with untreated peers demonstrated a 27% decrease in preterm delivery [55]. It is thought that in most cases, infection ascends from the vagina, although it may also be transplacental, or introduced during invasive procedures.

Biol Neonate 2001;79:157–162

159

It seems logical that bacterial infection would initiate preterm labour by causing an inflammatory reaction within the uterus, and effectively switching on parturitional phase 3 at an earlier stage. The mechanisms by which bacteria initiate infections are probably multifactorial. Bacteria do not directly synthesise prostaglandins [56], but they do release phospholipases which may liberate arachidonic acid from intracellular lipid pools, therefore increasing the synthesis of prostaglandins. Lamont et al. [57, 58] have demonstrated that bacteria release a substance which increases prostaglandin synthesis in amnion cells. This effect is probably mediated by phospholipase release [59]. An important mechanism by which bacteria might activate inflammatory mediators within the uterus is the action of endotoxin. Endotoxin, or lipopolysaccharide, is a component of the cell wall of gram-negative bacteria which has been shown to stimulate prostaglandins in a variety of tissues. Romero et al. [60] found that early spontaneous labour with preterm premature rupture of the membranes was more likely when endotoxin was detected. Endotoxin stimulates the synthesis of inflammatory cytokines, including IL-1ß, IL-6 and TNF [61], in amnion cells and in intact fetal membranes. Endotoxins also inhibit the release of PAF acylhydrolase, the enzyme which inactivates PAF, from decidual macrophages. Infection seems to be dually involved as a risk factor for fetal brain damage. As we have described, it can be directly involved in the onset of preterm labour. An intrauterine infection can also lead to a fetal inflammatory response, which can be identified chemically as proinflammatory cytokinaemia, or morphologically as vasculitis of the umbilical cord and/or the vessels of the chorionic plate. Adinolfi [62] proposed that cytokines produced by the immune system during the course of maternal infection are harmful to the developing brain of the unborn infant. Leviton [63] extended the hypothesis, suggesting that cytokines such as TNF-· contribute to both preterm delivery and periventricular white matter damage. Women who were found to have increased levels of TNF-· or IL-6 in cervicovaginal swabs were at a 3-fold increased risk of delivering preterm [64]. Most recently, IL-6 levels were found to be significantly elevated in the amniotic fluid of women whose infants later developed IVH or PVL [65]. The simplistic model of the ‘maternofetal unit’ may be an approach to the link between intrauterine infection, preterm birth, IVH and PVL. The uterus, the fetal circulation and the fetal brain can be viewed as three compartments, with the placenta representing the boundary between the mother and the fetus, and the fetal blood-brain barrier (BBB) as the boundary between the systemic circulation and the brain [66]. An infection remote from the brain results in circulating products of infection which easily cross the BBB, either because of the immaturity of the BBB [67, 68] or because these proinflammatory cytokines reduce the efficacy of the BBB [69]. Once those cytokines are in the brain, they can damage developing white matter. The hypothesis that the fetal and neonatal inflammatory response can damage developing white matter is supported by the higher levels of cytokines found in brains of infants who die with morphological evidence of white matter damage [70]. Support for the remote infection hypothesis has come mainly from clinical studies [66]. The local cytokine environment in which white matter damage is induced by products of infection has not yet been adequately explored. One recent preliminary study has indicated that peritoneal administration of endotoxin to the immature rat results in a rapid and marked expression of proinflammatory cytokines in the white and grey matter [71]. Another study showed a dose-dependent cytokine increase in the fetal rat brain after maternal endotoxin administration [72].

160

Biol Neonate 2001;79:157–162

Maternal Hypertension

Approximately 5–7% of all pregnancies are complicated by preeclampsia [73], and maternal hypertension in pregnancy is associated with increased perinatal mortality and morbidity and often results in intrauterine growth retardation. Early epidemiological studies suggested that there is an association between maternal hypertension and cerebral palsy [74], whereas in more recent studies of preterm infants, pre-eclampsia has been associated with a reduced risk of cerebral palsy [75, 76]. The results from the American study [75], however, are difficult to interpret, as most women with pre-eclampsia were treated with magnesium sulphate, which may have a beneficial, protective role [77]. The Australian study [76] was a prospective cohort study that included all live-born infants delivered between 24 and 32 weeks of gestation to mothers who were hypertensive during pregnancy for a 2-year period. Magnesium sulphate was not prescribed as anti-hypertensive medication. 107 infants born to hypertensive mothers and 107 controls matched for gestational age, sex and multiple pregnancy were compared. The incidence of periventricular haemorrhage was significantly lower in the study group. Cerebral palsy was not diagnosed in any infant of a hypertensive mother compared with 5 of the controls. The mean general quotient for the two groups was very similar and no difference was shown in the incidence of minor neuromotor developmental problems.

Neonatal Injury at Term

The commonest intrapartum cause of neonatal brain injury at term is asphyxia. The mechanisms giving rise to intrapartum asphyxia can be conveniently divided into two types. The first is interruption of the fetal oxygen supply by some acute obstetric catastrophe which interferes either with the maternal or fetal blood flow into and out of the placenta. So, for example, acute abruption, in which the placenta separates from its attachment to the uterus, would interfere with the maternal blood supply to the placenta. Uterine rupture may also cause a sudden interruption of fetal oxygenation. An umbilical cord accident might suddenly stop fetal blood flow into the placenta. Such accidents might include umbilical cord prolapse, where the umbilical cord passes through the cervix and into or through the vagina and into the outside air, where the temperature change and compression of the cord by the presenting part leads to slowing or cessation of the supply of fetal blood to the placenta. A cord accident might also occur where the umbilical cord is unusually short or where it is extensively wrapped around the fetal body and becomes stretched and occluded in the final stages of the second stage of labour just before expulsion of the fetus. The umbilical cord may also be compressed during a forceps delivery if it becomes entangled with the instrument. Uterine rupture is most often associated with previous classical caesarean section. In 30% of cases of uterine rupture associated with a classical scar, rupture occurs prior to labour [78]. Previous lower segment caesarean section and the use of oxytocic drugs are generally considered to be risk factors for rupture; however, Rosen et al. [79] found a similar rate of scar dehiscence in patients undergoing trial of labour (1.8%) to those having repeat caesarean section, with the use of oxytocin not increasing the risk. Placental abruption is associated with maternal hypertension, pre-eclampsia, fetal growth restriction and placenta praevia. It commonly occurs, however, in otherwise uncomplicated pregnancies and is almost impossible to predict. Um-

Terzidou/Bennett

bilical cord prolapse occurs commonly in any situation where there is a poor fit between the presenting part and the maternal pelvis, such as with a transverse lie or a breech presentation. It may also occur following the birth of the first of a pair of twins. The second broad category of mechanisms leading to asphyxia may be termed chronic intrapartum asphyxia or chronic intrapartum partial asphyxia. During labour, each contraction of the uterus compresses the maternal blood vessels supplying the placenta to reduce the flow of maternal blood into it. At the peak of a strong contraction, the maternal blood supply to the placenta is reduced and probably stops altogether. The average contraction lasts for 45 s and occurs every 2–3 min. Although the supply of oxygen to the placenta is reduced or cut off during a contraction, in a normal labour, the fetus is unaffected because there is a pool of maternal blood within the placenta during the contraction and because there is sufficient time

between contractions to compensate for the temporary placental hypoxia. If, however, the placenta is abnormal, the intermittent interruption of the maternal blood supply to it during labour may be enough to compromise the fetal acid-base balance. The relevant placental abnormality is failure of the trophoblast cells to adequately invade maternal blood vessels during the first and second trimesters of pregnancy, leading to an abnormally high resistance to maternal blood flow into the placenta. This type of abnormal placentation is commonly associated with both intratuerine growth retardation and the pre-eclampsia syndrome. Compromise of the fetal acid-base balance may also occur where contractions are particularly frequent or unusually forceful, usually when syntocinon is used to augment labour. It may also occur where the maternal placental blood flow is reduced because of maternal hypotension, sometimes seen following epidural anaesthesia.

References 1 Ellenberg JH, Nelson KB: Birth weight and gestational age in children with cerebral palsy or seizure disorders. Am J Dis Child 1979;133:1044–1048. 2 Johnson A, Townshend P, Yudkin P, Bull D, Wilkinson AR: Functional abilities at age 4 years of children born before 29 weeks of gestation. BMJ 1993;306:1715–1718. 3 Wood NS, Marlow N, Costeloe K, Gibson AT, Wilkinson AR: Neurologic and developmental disability after extremely preterm birth. EPICure Study Group. N Engl J Med 2000;343:378–384. 4 Escobar GJ, Littenberg B, Petitti DB: Outcome among surviving very low birthweight infants: A meta-analysis. Arch Dis Child 1991;66:204–211. 5 Hall A, McLeod A, Counsell C, Thomson L, Mutch L: School attainment, cognitive ability and motor function in a total Scottish very-low-birthweight population at eight years: A controlled study. Dev Med Child Neurol 1995;37:1037–1050. 6 Schothorst PF, van Engeland H: Long-term behavioral sequelae of prematurity. J Am Acad Child Adolesc Psychiatry 1996;35:175–183. 7 McCarton CM, Brooks-Gunn J, Wallace IF, Bauer CR, Bennett FC, Bernbaum JC, Broyles RS, Casey PH, McCormick MC, Scott DT, Tyson J, Tonascia J, Meinert CL: Results at age 8 years of early intervention for low-birth-weight premature infants. The Infant Health and Development Program. JAMA 1997;277:126–132. 8 Paneth NP-MJ: The epidemiology of germinal matrix/intraventricular hemorrhage; in Kiely M (ed): Reproductive and Perinatal Epidemiology. Boca Raton, CRC Press, 1990, pp 371–399. 9 Paneth NRR, Kazam E: Brain Damage in Preterm Infant. London, MacKeith Press, 1994, pp 171– 185. 10 Lye S: The initiation and inhibition of labour – Toward a molecular understanding. Semin Reprod Endocrinol 1994;12:284–294. 11 McLean M, Bisits A, Davies J, Woods R, Lowry P, Smith R: A placental clock controlling the length of human pregnancy. Nat Med 1995;1:460–463. 12 Jones SA, Brooks AN, Challis JR: Steroids modulate corticotropin-releasing hormone production in human fetal membranes and placenta. J Clin Endocrinol Metab 1989;68:825–830. 13 Jackson M, Dudley DJ: Endocrine assays to predict preterm delivery. Clin Perinatol 1998;25:837– 857.

Maternal Risk Factors for Fetal and Neonatal Brain Damage

14 Gibbs RS, Romero R, Hillier SL, Eschenbach DA, Sweet RL: A review of premature birth and subclinical infection. Am J Obstet Gynecol 1992;166: 1515–1528. 15 Romero R: Does infection cause premature labor and delivery? Semin Reprod Endocrinol 1994;12: 227–239. 16 Dudley D: Infection, inflammation, and contractions: The role of cytokines in the pathophysiology of preterm labour. Semin Reprod Endocrinol 1994;12:263–272. 17 Dinarello CA: Biologic basis for interleukin-1 in disease. Blood 1996;87:2095–2147. 18 Fidel PL Jr, Romero R, Wolf N, Cutright J, Ramirez M, Araneda H, Cotton DB: Systemic and local cytokine profiles in endotoxin-induced preterm parturition in mice. Am J Obstet Gynecol 1994; 170:1467–1475. 19 Romero R, Mazor M, Brandt F, Sepulveda W, Avila C, Cotton DB, Dinarello CA: Interleukin-1 alpha and interleukin-1 beta in preterm and term human parturition. Am J Reprod Immunol 1992;27:117– 123. 20 Romero R, Baumann P, Gonzalez R, Gomez R, Rittenhouse L, Behnke E, Mitchell MD: Amniotic fluid prostanoid concentrations increase early during the course of spontaneous labor at term. Am J Obstet Gynecol 1994;171:1613–1620. 21 Romero R, Yoon BH, Mazor M, Gomez R, Gonzalez R, Diamond MP, Baumann P, Araneda H, Kenney JS, Cotton DB, et al: A comparative study of the diagnostic performance of amniotic fluid glucose, white blood cell count, interleukin-6, and gram stain in the detection of microbial invasion in patients with preterm premature rupture of membranes. Am J Obstet Gynecol 1993;169:839–851. 22 Laham N, Rice GE, Bishop GJ, Hansen MB, Bendtzen K, Brennecke SP: Elevated plasma interleukin 6: A biochemical marker of human preterm labour. Gynecol Obstet Invest 1993;36:145–147. 23 Greig PC, Murtha AP, Jimmerson CJ, Herbert WN, Roitman-Johnson B, Allen J: Maternal serum interleukin-6 during pregnancy and during term and preterm labor. Obstet Gynecol 1997;90:465– 469. 24 Cherouny PH, Pankuch GA, Romero R, Botti JJ, Kuhn DC, Demers LM, Appelbaum PC: Neutrophil attractant/activating peptide-1/interleukin-8: Association with histologic chorioamnionitis, preterm delivery, and bioactive amniotic fluid leukoattractants. Am J Obstet Gynecol 1993;169: 1299–1303.

25 Laham N, Brennecke SP, Rice GE: Interleukin-8 release from human gestational tissue explants: The effects of lipopolysaccharide and cytokines. Biol Reprod 1997;57:616–620. 26 Barclay CG, Brennand JE, Kelly RW, Calder AA: Interleukin-8 production by the human cervix. Am J Obstet Gynecol 1993;169:625–632. 27 El Maradny E, Kanayama N, Halim A, Maehara K, Sumimoto K, Terao T: Interleukin-8 induces cervical ripening in rabbits. Am J Obstet Gynecol 1994;171:77–83. 28 Calder AA: Prostaglandins and biological control of cervical function. Aust NZ J Obstet Gynaecol 1994;34:347–351. 29 Kelly RW, Leask R, Calder AA: Choriodecidual production of interleukin-8 and mechanism of parturition. Lancet 1992;339:776–777. 30 El Maradny E, Kanayama N, Halim A, Maehara K, Sumimoto K, Terao T: Biochemical changes in the cervical tissue of rabbit induced by interleukin8, interleukin-1beta, dehydroepiandrosterone sulphate and prostaglandin E2: A comparative study. Hum Reprod 1996;11:1099–1104. 31 Osmers RG, Adelmann-Grill BC, Rath W, Stuhlsatz HW, Tschesche H, Kuhn W: Biochemical events in cervical ripening dilatation during pregnancy and parturition. J Obstet Gynaecol 1995;21: 185–194. 32 Williams MA, Mittendorf R, Stubblefield PG, Lieberman E, Schoenbaum SC, Monson RR: Cigarettes, coffee, and preterm premature rupture of the membranes. Am J Epidemiol 1992;135:895– 903. 33 Lobel M, Dunkel-Schetter C, Scrimshaw SC: Prenatal maternal stress and prematurity: A prospective study of socioeconomically disadvantaged women. Health Psychol 1992;11:32–40. 34 Arias F, Rodriquez L, Rayne SC, Kraus FT: Maternal placental vasculopathy and infection: Two distinct subgroups among patients with preterm labor and preterm ruptured membranes. Am J Obstet Gynecol 1993;168:585–591. 35 Sandman CA, Wadhwa PD, Chicz-DeMet A, Dunkel-Schetter C, Porto M: Maternal stress, HPA activity, and fetal/infant outcome. Ann NY Acad Sci 1997;814:266–275. 36 Lockwood CJ: Stress-associated preterm delivery: The role of corticotropin-releasing hormone. Am J Obstet Gynecol 1999;180:S264–S266.

Biol Neonate 2001;79:157–162

161

37 Economides DL, Nicolaides KH, Linton EA, Perry LA, Chard T: Plasma cortisol and adrenocorticotropin in appropriate and small for gestational age fetuses. Fetal Ther 1988;3:158–164. 38 Petraglia F, Coukos G, Volpe A, Genazzani AR, Vale W: Involvement of placental neurohormones in human parturition. Ann NY Acad Sci 1991;622: 331–340. 39 Warren WB, Patrick SL, Goland RS: Elevated maternal plasma corticotropin-releasing hormone levels in pregnancies complicated by preterm labor. Am J Obstet Gynecol 1992;166:1198–1207. 40 Sapolsky RM: Stress, glucocorticoids, and damage to the nervous system: The current state of confusion. Stress 1996;1:1–19. 41 McEwen BS: Re-examination of the glucocorticoid hypothesis of stress and aging. Prog Brain Res 1992;93:365–381. 42 Sapolsky RM: Why stress is bad for your brain. Science 1996;273:749–750. 43 Keelan JA, Coleman M, Mitchell MD: The molecular mechanisms of term and preterm labor: Recent progress and clinical implications. Clin Obstet Gynecol 1997;40:460–478. 44 Miyoshi H: Voltage-clamp studies of gap junctions between uterine muscle cells during term and preterm labour. Biophys J 1996;71:1324. 45 Petterson B, Nelson KB, Watson L, Stanley F: Twins, triplets, and cerebral palsy in births in Western Australia in the 1980s. BMJ 1993;307: 1239–1243. 46 Grether JK, Nelson KB, Cummins SK: Twinning and cerebral palsy: Experience in four northern California counties, births 1983 through 1985. Pediatrics 1993;92:854–858. 47 Pharoah PO, Cooke T: Cerebral palsy and multiple births. Arch Dis Child Fetal Neonatal Ed 1996;75: F174–F177. 48 Scheller JM, Nelson KB: Twinning and neurologic morbidity. Am J Dis Child 1992;146:1110–1113. 49 Bejar R, Vigliocco G, Gramajo H, Solana C, Benirschke K, Berry C, Coen R, Resnik R: Antenatal origin of neurologic damage in newborn infants. II. Multiple gestations. Am J Obstet Gynecol 1990; 162:1230–1236. 50 Cincotta RB, Fisk NM: Current thoughts on twintwin transfusion syndrome. Clin Obstet Gynecol 1997;40:290–302. 51 Iams JD, Johnson FF, Sonek J, Sachs L, Gebauer C, Samuels P: Cervical continence as a continuum: A study of ultrasonographic cervical length and obstetric performance. Am J Obstet Gynecol 1995; 172:1097–1103.

162

52 Heath VC, Souka AP, Erasmus I, Gibb DM, Nicolaides KH: Cervical length at 23 weeks of gestation: The value of Shirodkar suture for the short cervix. Ultrasound Obstet Gynecol 1998;12:318–322. 53 Seo K, McGregor JA, French JI: Preterm birth is associated with increased risk of maternal and neonatal infection. Obstet Gynecol 1992;79:75–80. 54 Hillier SL, Nugent RP, Eschenbach DA, Krohn MA, Gibbs RS, Martin DH, Cotch MF, Edelman R, Pastorek JG 2nd, Rao AV, et al: Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. The Vaginal Infections and Prematurity Study Group. N Engl J Med 1995;333:1737–1742. 55 Hauth JC, Goldenberg RL, Andrews WW, DuBard MB, Copper RL: Reduced incidence of preterm delivery with metronidazole and erythromycin in women with bacterial vaginosis. N Engl J Med 1995;333:1732–1736. 56 Bennett PR, Elder MG, Myatt L: Secretion of phospholipases by bacterial pathogens may initiate preterm labor. Am J Obstet Gynecol 1992;163: 241–242. 57 Lamont RF, Rose M, Elder MG: Effect of bacterial products on prostaglandin E production by amnion cells. Lancet 1985;ii:1331–1333. 58 Lamont RF, Anthony F, Myatt L, Booth L, Furr PM, Taylor-Robinson D: Production of prostaglandin E2 by human amnion in vitro in response to addition of media conditioned by microorganisms associated with chorioamnionitis and preterm labor. Am J Obstet Gynecol 1990;162:819–825. 59 Bennett PR, Elder MG, Myatt L: Secretion of phospholipases by bacterial pathogens may initiate preterm labor (letter). Am J Obstet Gynecol 1990; 163:241–242. 60 Romero R, Kadar N, Hobbins JC, Duff GW: Infection and labor: The detection of endotoxin in amniotic fluid. Am J Obstet Gynecol 1987;157:815– 819. 61 Romero R, Lockwood C, Oyarzun E, Hobbins JC: Toxemia: New concepts in an old disease. Semin Perinatol 1988;12:302–323. 62 Adinolfi M: The maternal-fetal interaction: Some controversies and solutions. Exp Clin Immunogenet 1993;10:103–117. 63 Leviton A: Preterm birth and cerebral palsy: Is tumor necrosis factor the missing link? Dev Med Child Neurol 1993;35:553–558. 64 Inglis SR, Jeremias J, Kuno K, Lescale K, Peeper Q, Chervenak FA, Witkin SS: Detection of tumor necrosis factor-alpha, interleukin-6, and fetal fibronectin in the lower genital tract during pregnancy: Relation to outcome. Am J Obstet Gynecol 1994;171:5–10. 65 Figueroa R: Elevated amniotic fluid interleukin-6 periventricular-leucomalacia and intraventricular haemorrhage. Am J Obstet Gynecol 1996;174: 330.

Biol Neonate 2001;79:157–162

66 Dammann O, Leviton A: Maternal intrauterine infection, cytokines, and brain damage in the preterm newborn. Pediatr Res 1997;42:1–8. 67 Adinolfi M: The development of the human bloodCSF-brain barrier. Dev Med Child Neurol 1985; 27:532–537. 68 Sarnat H: Role of human fetal ependyma. Pediatr Neurol 1992;8:163–178. 69 Megyeri P, Abraham CS, Temesvari P, Kovacs J, Vas T, Speer CP: Recombinant human tumor necrosis factor alpha constricts pial arterioles and increases blood-brain barrier permeability in newborn piglets. Neurosci Lett 1992;148:137–140. 70 Deguchi K, Mizuguchi M, Takashima S: Immunohistochemical expression of tumor necrosis factor alpha in neonatal leukomalacia. Pediatr Neurol 1996;14:13–16. 71 Hagberg HE: Expression of cytokines and chemokines in the immature white and gray matter in response to hypoxia-ischemia or endotoxin (abstract). J Cereb Blood Flow Metab 1999;19:s312. 72 Cai Z, Pan ZL, Pang Y, Evans OB, Rhodes PG: Cytokine induction in fetal rat brains and brain injury in neonatal rats after maternal lipopolysaccharide administration. Pediatr Res 2000;47:64– 72. 73 Sibai BM, Ewell M, Levine RJ, Klebanoff MA, Esterlitz J, Catalano PM, Goldenberg RL, Joffe G: Risk factors associated with preeclampsia in healthy nulliparous women. The Calcium for Preeclampsia Prevention (CPEP) Study Group. Am J Obstet Gynecol 1997;177:1003–1010. 74 Jonas O, Stern LM, Macharper T: A South Australian study of pregnancy and birth risk factors associated with cerebral palsy. Int J Rehabil Res 1989; 12:159–166. 75 Grether JK, Nelson KB, Emery ES 3rd, Cummins SK: Prenatal and perinatal factors and cerebral palsy in very low birth weight infants. J Pediatr 1996; 128:407–414. 76 Gray PH, O’Callaghan MJ, Mohay HA, Burns YR, King JF: Maternal hypertension and neurodevelopmental outcome in very preterm infants. Arch Dis Child Fetal Neonatal Ed 1998;79:F88–F93. 77 Nelson KB, Grether JK: Can magnesium sulfate reduce the risk of cerebral palsy in very low birthweight infants? Pediatrics 1995;95:263–269. 78 Cunningham FG: Injuries to the birth canal; in Cunningham FG (ed): Williams Obstetrics, ed 19. Norwalk, Appleton and Lange, 1993. 79 Rosen MG, Dickinson JC, Westhoff CL: Vaginal birth after cesarean: A meta-analysis of morbidity and mortality. Obstet Gynecol 1991;77:465–470.

Terzidou/Bennett

Biol Neonate 2001;79:163–167

Fetal Endocrine Signals and Preterm Labor John R.G. Challis a, b Stephen K. Smith a a Department of Obstetrics and Gynaecology, University of Cambridge, Cambridge, UK; b Departments of Physiology and Obstetrics and Gynaecology, University of Toronto and CIHR Group in Fetal and Neonatal Health and Development, CIHR Institute of Human Development, Child and Youth Health, Canada

Key Words Preterm labor W Cortisol W Corticotropin-releasing hormone W Prostaglandins W Fetal membranes W Myometrium W Trophoblast

Abstract Increased uterine contractility at term and preterm results from activation and then stimulation of the myometrium. Activation can be provoked by mechanical stretch of the uterus and by an endocrine pathway resulting from increased activity of the fetal hypothalamicpituitary-adrenal axis. Cortisol, derived from the fetal adrenal in cases of intrauterine compromise or from the maternal adrenal in response to stress, or generated locally from cortisone in choriodecidual trophoblasts, provides a crucial link to uterine stimulation. Cortisol contributes to the increased production of prostaglandins (PGs) by fetal membranes and the decidua through the upregulation of PG synthase and the downregulation of PG dehydrogenase enzymes. Cortisol also stimulates placental corticotropin-releasing hormone (CRH) output, although CRH may both relax and stimulate uterine activity depending on the distribution and affinity of its receptor subtypes. Other agents such as cytokines may intercede in this sequence to stimulate PGs and/or CRH, giving rise to a cascade phenomenon that results in preterm birth. Copyright © 2001 S. Karger AG, Basel

Introduction

Preterm birth occurs in approximately 5–10% of all pregnancies. This figure may be higher in certain population groups and has not decreased over the past 20–30 years. Although some preterm births may be elective, approximately 30% occur in association with an underlying infectious process, and about 50% are idiopathic preterm births of unknown cause. Preterm birth is associated with 70% of neonatal deaths and up to 75% of neonatal morbidity. Infants born preterm have an increased incidence of cerebral palsy, neurological

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0163$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

handicap and pulmonary disorders. The costs of caring for preterm babies in the United States have been estimated at around USD 8 billion annually. At present, there are no effective diagnostic indicators of preterm birth, and there are no effective treatments for this condition. Thus, the current aim of research is to understand the underlying biochemistry of the birth process and to utilize that information to develop better diagnostic indicators and improve methods of therapeutic management.

Phases of Parturition

Uterine contractility during pregnancy and parturition can be divided into different phases [1, 2]. Phase 0 of parturition corresponds to pregnancy, a time of relative uterine quiescence. Phase 1 of parturition is associated with activation of uterine function, wherein mechanical stretch or uterotrophic priming leads to upregulation of a cassette of genes required for contractions. These contraction-associated protein (CAP) genes include connexin-43, the major protein of gap junctions, agonist receptors and proteins encoding ion channels. In phase 2 of parturition, the uterus can then be stimulated by uterotonins, including prostaglandins (PGs), oxytocin and corticotropinreleasing hormone (CRH). Phase 3 of parturition includes expulsion of the placenta and the involution process and has been attributed primarily to the effects of oxytocin. The regulation of uterine quiescence during pregnancy (phase 0) has been discussed in several recent reviews [1–4]. Major effectors of myometrial relaxation, acting in a paracrine or endocrine fashion, include progesterone, relaxin, prostacyclin, parathyroid hormonerelated peptide, nitric oxide and CRH, which may both inhibit and stimulate uterine contractility. These agents act in different ways, but in general result in increased intracellular levels of cyclic adenosine monophosphate (cAMP) or cyclic guanosine monophosphate. These nucleotides inhibit intracellular calcium release and reduce the activity of the enzyme myosin light chain kinase, which is required for shortening of the myofilaments. Several current strategies for manag-

John R.G. Challis University of Cambridge, Department of Obstetrics and Gynaecology The Rosie Hospital, Robinson Way Cambridge CB2 2SW (UK) Tel. +44 1223 763100, Fax +44 1223 24881, E-Mail [email protected]

ing preterm labor are directed at increasing intracellular cAMP and/ or reducing the availability of calcium. It is clear now that the activation of myometrial function (phase 1) is driven by the fetal genotype, and effected through two separate but interdependent pathways [2]. The first involves activation of the fetal hypothalamic-pituitary-adrenal (HPA) axis. The second involves mechanical distension of the uterus, leading to stretch-related upregulation of CAP gene expression. Maturation of fetal HPA function during late pregnancy is a consistent developmental change across species, including primates [1– 3]. Extensive studies in fetal sheep have shown increased expression of CRH mRNA in parvocellular neurones of the paraventricular nucleus of the hypothalamus and of proopiomelanocortin mRNA in the pars distalis of the fetal pituitary in late pregnancy. These changes correlate with increased concentrations of adrenocorticotropic hormone (ACTH1–39) in the fetal circulation. ACTH acts on the fetal adrenal to increase the expression of key enzymes required for cortisol production (especially P450C17), and to upregulate ACTH receptors in the fetal adrenal cortex. This allows enhanced binding and coupling to adenylate cyclase, resulting in increased sensitivity of the fetal adrenal gland to further stimulation by ACTH. In primates, similar changes occur. A major difference, however, is that the fetal adrenal gland is divided into an outer definitive cortex that produces aldosterone, a transitional zone that produces cortisol and a large fetal zone that lacks 3ß-hydroxysteroid dehydrogenase (3ß-HSD) and produces the C19¢5 estrogen precursor, dehydroepiandrosterone (sulfate) [DHEA(S)]. Regulation of P450C17 in the transitional zone, however, appears to be effected by ACTH, as in animal species. It has been suggested that during human pregnancy, the level of fetal HPA function is relatively suppressed, because maternal cortisol can cross the placenta into the fetal compartment. At term, the increased output of estrogen stimulates 11ß-HSD type 2 (11ß-HSD-2) in the placental syncytiotrophoblast. This enzyme metabolizes maternal cortisol to cortisone, removes the inhibition and facilitates autonomous fetal HPA maturation. Activation of the fetal HPA axis occurs in the presence of an adverse intrauterine environment, for example, with compromised uteroplacental blood flow and/or conditions of fetal hypoxemia. Fetal sheep made transiently hypoxemic had increased levels of hypothalamic CRH mRNA and pituitary proopiomelanocortin mRNA [1]. In late gestation, hypoxemia also led to increased levels of fetal adrenal ACTH receptor mRNA, consistent with increased overall responsiveness of the fetal HPA axis. When fetuses at two thirds of term gestation were subjected to hypoxemia by repeated umbilical cord occlusion over several days, the adrenal cortisol response relative to the level of ACTH stimulation rose. Other experimental models of sustained, but episodic fetal hypoxemia produce similar fetal hormonal and cardiovascular responses, and may result in a shortened length of gestation. Increases in fetal HPA function in animal species such as sheep lead to changes in the placental output of progesterone before birth. During pregnancy, progesterone is required for uterine growth, but it simultaneously suppresses the expression of CAP genes. At term, in most animals species, the influence of progesterone on the myometrium declines, uterine stretch no longer stimulates uterine growth and the increase in wall tension caused by continued fetal growth becomes translated into increased expression of CAP genes and myometrial activation. Mechanical stretch likely contributes to the greater incidence of preterm birth in pregnancies with multiple fetuses, and may account for the higher incidence of preterm birth in

164

Biol Neonate 2001;79:163–167

pregnancies where the fetal size is large for gestational age. However, in women, there does not appear to be a decline in circulating progesterone concentrations prepartum. We have argued recently that this represents a mechanism to maintain relaxation of the lower uterine segment at the time of birth, while local antagonism of progesterone action in the fundal region of the uterus facilitates the development of uterine contractions predominantly in that region [1].

PGs and Parturition

There is now compelling evidence that increased output of intrauterine PGs contributes to the drive towards myometrial contractility at term and preterm [5, 6]. The stimulus for increased PG synthesis may be a consequence of fetal HPA activation and/or uterine mechanical stretch. It is clear, however, that the role of these substances is primarily in phase 2 rather than phase 1 of parturition. The evidence implicating PGs in the parturition processes is overwhelming. Concentrations of PGs or their metabolites in maternal plasma or amniotic fluid increase at the time of labor before the onset of coordinated uterine contractility. Drugs which block PG synthesis are known to promote myometrial quiescence and may prolong the length of gestation. Furthermore, pregnant mice in which the PG synthase type 1 (PGHS-1) gene has been inactivated by mutation have prolonged gestations [6]. In women, PG production is discretely compartmentalized within the tissues of the pregnant uterus [7]. PGE2 is formed predominantly in the amnion, and its output increases at the time of labor. The chorion and decidua also produce PGE2. The presence in the chorion of the enzyme 15-hydroxyprostaglandin dehydrogenase (PGDH) is presumed to cause metabolism of PGs generated in the amnion and chorion, preventing their passage to the underlying tissues. Thus, in normal pregnancy, those PGs that stimulate myometrial contractions are likely generated either within the decidua, or in the myometrium itself. In some cases of preterm birth, however, PGDH activity in the chorion is diminished and the metabolic barrier is reduced, and PGs generated within the amnion or chorion could then provide the stimulus for myometrial contractions [8]. Studies on PG production from myometrium collected from women at the time of labor have led to divergent findings. Although some investigators have reported increased PG output, most reports have failed to demonstrate increased PG synthesis or PGHS activity in myometrium collected from women in labor at term or preterm. Primary PGs are formed from arachidonic acid through the activity of the PGHS enzyme complex [6]. There are two forms of PGHS: PGHS-1 and PGHS-2. These are separate gene products. PGHS-1 has been described as constitutive, PGHS-2 as inducible by growth factors, cytokines and, paradoxically, in human fetal membranes by glucocorticoids. The expression and activity of PGHS-2 in human amnion and chorion increases at the time of labor; PGHS-2 is also upregulated in amnion collected from patients in preterm labor. Thus, new strategies for inhibiting the myometrium in preterm labor have included the development of specific inhibitors of PGHS-2 activity. It is anticipated that these drugs should be more efficacious than the nonspecific PGHS-1 inhibitors used previously, and will have less cardiovascular and renal side effects in the fetus since they will produce minimal inhibition of PGHS-1. PG action is mediated through specific receptor binding. There are four main subtypes of receptors: EP1, EP2, EP3 and EP4 for PGE2 and FP for PGF2·. EP1 and EP3 receptors mediate contrac-

Challis/Smith

tions by increasing intracellular calcium mobilization and inhibiting intracellular cAMP generation. Conversely, PGE2 acting through EP2 and EP4 receptors increases adenylate cyclase and relaxes smooth muscle. Theoretically, inhibition of uterine activity might be achieved through antagonism of EP1, EP3 or FP action, or through activation of the inhibitory EP2 and EP4 receptors. At present, however, there is little information available concerning the regulation and regional changes in any PG receptor subtype in human myometrial tissue at the time of labor. Increased expression of PGHS-2 occurs in response to a variety of growth factors, including epidermal growth factor, cytokines [interleukin (IL)-1, tumor necrosis factor and IL-6] [5–9]. The action of cytokines appears to be mediated through an NF-ÎB consensus sequence in the promoter region of PGHS-2. This promoter also contains a cAMP response element and a glucocorticoid response element at approximately 760 base pairs from the PGHS-2 transcription start sign. Several years ago, Gibb et al. [10] and Economopoulos et al. [11] showed that human amnion cells maintained in monolayer tissue culture produced increased amounts of PGE2 in a dose-dependent fashion in response to treatment with a synthetic glucocorticoid, dexamethasone. It remains unclear whether this action occurs on amniotic epithelium or fibroblast cells, since both contain glucocorticoid receptors (GRs), and both have been reported to increase the expression of PGHS-2 upon glucocorticoid stimulation. However, in mixed cell cultures, the action of glucocorticoids is clearly dependent upon interaction with GRs and apparently requires activation of protein kinase C. Proinflammatory cytokines also increase PGHS-2 gene expression, mRNA and protein levels and PG output by cultured amnion and chorion cells. Interestingly, anti-inflammatory cytokines such as IL-10 attenuate the stimulatory effect of IL-1ß on PGHS-2 gene expression and activity [12]. Thus, in vivo, it is apparent that the relative amounts of eicosanoids and cytokines produced from an interactive cytokine-eicosanoid cascade will be critical in regulating the final response of the tissue and the level of PG produced [9]. These results also raise the interesting possibility that anti-inflammatory cytokines might be utilized therapeutically to modulate the action of compounds such as IL-1. It is possible that cytokines may contribute to the stimulus for uterine contractility in patients at normal term with a subclinical infection. Cytokines are very clearly involved, however, in precipitating PG synthesis in association with infective processes [5, 7]. Infection-driven preterm labor occurs in the presence of increased concentrations of IL-1, IL-6 and tumor necrosis factor in the amniotic fluid. Experimentally, administration of IL-1 or bacterial endotoxin to pregnant mice provokes premature delivery, and this is associated with increased PG output by intrauterine tissues. Cytokines increase phospholipase A2 and PGHS-2 gene expression in a time- and dosedependent fashion by amnion and choriodecidual tissue maintained in vitro. Thus, it is generally accepted that in the presence of an ascending bacterial infection, organisms pass between the fetal membranes and later reach the amniotic cavity. Bacterial organisms release phospholipases, which may in turn stimulate PG synthesis. They also release endotoxins such as lipopolysaccharide, which provoke PG output directly, or release cytokines from macrophages. These cytokines, in turn, stimulate PGHS-2 expression and PG output from the amnion and/or decidual cells. In addition, IL-1 stimulates the output of other cytokines, including IL-6 and IL-8 from the decidua, thereby establishing a positive feed-forward cascade [9].

Preterm Labor

PG Metabolism

Recently, it has become apparent that the biologically active levels of PG depend not only on the rates of synthesis, but also on the rates of metabolism [12]. Normally, high levels of PGDH expressed in chorion trophoblasts would be expected to effectively metabolize PG generated within the amnion or chorion. However, patients in preterm labor with an underlying infective process have markedly reduced numbers of trophoblasts in the chorion layer and dramatically reduced levels of PGDH activity [8]. In addition, approximately 15% of patients with idiopathic preterm labor have diminished expression of PGDH but a normal presence of trophoblasts. PGDH activity is modestly reduced in the chorion from patients at term, but is markedly diminished in the myometrium and cervix of patients presenting in preterm labor. Thus, reduced PG metabolism appears to be an effective way of increasing PG levels, which may then reach agonist PG receptors in a paracrine fashion. Furthermore, levels of matrix metalloproteinase (MMP)-9 in the chorion are increased in preterm labor. Since this gelatinase enzyme contributes to the controlled degradation of collagen within the fetal membranes, and MMP-9 activity is increased by PGE2, this feed-forward cascade may also help explain the mechanism of preterm premature rupture of the membranes with MMP-9 as the predominant gelatinolytic activity. Recent studies have been directed towards understanding the mechanism by which steroid hormones might regulate PGDH [13]. Surprisingly, these studies have also revealed a mechanism for local progesterone withdrawal within the human fetal membranes. Patel et al. [13] have shown that human chorionic PGDH gene expression and activity is inhibited by glucocorticoids (cortisol, betamethasone and dexamethasone) and is maintained in a tonic fashion by progesterone. Chorionic trophoblasts express the enzyme 3ß-HSD and have the capacity to produce their own progesterone from pregnenolone. Inhibition of the 3ß-HSD enzyme with the drug Trilostane inhibited progesterone output from chorionic trophoblast cells and reduced PGDH mRNA levels. Replacement of progesterone or a synthetic progestagen restored PGDH activity. This effect could be blocked, in part, by a progesterone receptor antagonist. However, the action of progesterone to restore PGDH could also be blocked by a specific GR antagonist. This observation suggested that progesterone produced locally within the chorion acts throughout pregnancy to maintain chorionic PGDH activity. It does so, however, through interacting with GRs. At term, increased availability of endogenous cortisol would displace progesterone from GRs, resulting in loss of the stimulation of PGDH, and also a direct inhibitory effect on PGDH expression. This interaction, whereby the effects of progesterone are mediated through GRs but can be opposed by increased output of glucocorticoid, may provide a mechanism for producing local progesterone withdrawal in the human uterus. We have suggested elsewhere that this activity may be greater in the fundal area, thereby contributing to regionalized changes in uterine contractions [1, 14]. Regulation of PGDH, however, is clearly multifactorial [12]. The enzyme is also downregulated by cytokines, cAMP and CRH. At present, it is unclear whether the effect of CRH on PGDH is mediated by cAMP, since the CRH receptor in human fetal membrane does not appear to be coupled through the G·s protein. Antiinflammatory cytokines such as IL-10 appear to increase PGDH gene expression. Surprisingly, we found that the biologically inactive corticosteroid, cortisone, was almost as effective as cortisol in inhibiting PGDH in chorion cells, but not in placental trophoblast cells [12]. In

Biol Neonate 2001;79:163–167

165

the chorion, the action of cortisone could be blocked by a GR antagonist, and was completely attenuated in the presence of the drug carbonexolone. This drug, an active ingredient of liquorice, inhibits the enzyme 11ß-HSD-1, which is abundantly expressed in chorionic trophoblasts and predominantly converts cortisone to cortisol. Thus, these cells have the potential to form cortisol locally from cortisone, in addition to forming progesterone locally from pregnenolone. Therapeutic regulation of PGDH, theoretically, could be accomplished by steroid hormones, or by drugs which alter the levels of 11ß-HSD-1.

CRH and Preterm Labor

It is now well established that the concentrations of CRH in maternal blood rises progressively during human pregnancy [15, 16]. This rise correlates with increased levels of CRH mRNA and CRH peptide in placental tissue [17]. In the circulation, CRH is largely associated with a high-affinity circulating CRH-binding protein (CRH-BP) produced in the liver, placenta and also at other sites, including the brain. CRH-BP effectively blocks the action of placental CRH on the maternal pituitary and on the myometrium. Near term, and in association with preterm labor, CRH-BP concentrations fall, coincident with the increase in circulating CRH [16]. Thus, it has been suggested that there is a substantial increase in free CRH concentrations in systemic plasma as a component of the triggering of the labor process. Regulation of the placental CRH output is multifactorial and has been reviewed extensively [15]. Briefly, CRH gene expression and CRH output by placental trophoblast cells is paradoxically increased by glucocorticoids. CRH output from the placenta and fetal membranes also increases in response to PGs, cytokines and catecholamines, and is decreased by nitric oxide and progesterone. Karalis and Majzoub [18] have suggested that the inhibitory effect of progesterone is exerted through binding to GR in trophoblast cells. At term, increased levels of cortisol displace progesterone bound to GR and this is reflected as an increase in CRH output. Thus, the mechanism of interaction between progesterone and cortisol in the regulation of CRH is similar to that proposed for the regulation of PGDH. The action of CRH on the intrauterine tissues and myometrium is effected through an extensive network of high-affinity CRH receptors with different specificities. There are two main classes of CRH receptor, CRH-R1 and CRH-R2. In the myometrium, CRH acts by binding to CRH-R1, which is coupled to G·s, leading to the stimulation of cAMP output. Thus, the primary effect of CRH throughout pregnancy is likely to be one of uterine relaxation. The binding affinity of the CRH receptor in human myometrium increases during pregnancy, but then decreases prior to parturition. Studies by Grammatopoulos and Hillhouse [19, 20] have suggested that oxytocin effects this change by upregulating a protein kinase C which phosphorylates the CRH receptor protein, resulting in desensitization and loss of the inhibitory influence of CRH on the myometrium. Therefore, the peptide CRH may act as an inhibitor or stimulant of the myometrium, depending upon the affinity and second messenger of the different receptor species. The differential effects of CRH on the myometrium may also contribute to the regionalization of myometrial activity at term and in the preterm period. Stevens et al. [21] showed that the expression of CRH-R1 in myometrium collected from the lower uterine segment was higher in patients in labor compared to those not in labor. Furthermore, the expression of CRH-R1 was substantially higher in lower-segment compared to fundal myometrium when paired samples of

166

Biol Neonate 2001;79:163–167

tissue from individual patients were examined. Thus, at the time of labor, CRH may promote relaxation of the lower segment, but stimulate activity in the body of the uterus. This stimulatory action could be direct; it could also be indirect, since CRH stimulates the output of PGs by upregulating PGHS-2 and downregulating PGDH in human fetal membranes. There has been much interest recently in the possibility that elevation of the maternal plasma CRH concentration may be used to predict women destined to enter preterm labor. McLean et al. [22] demonstrated elevated maternal plasma concentrations of CRH as early as 14–16 weeks of pregnancy in women who subsequently delivered preterm, and lower concentrations of CRH in the plasma of women who delivered post-term. Korebrits et al. [23] found that maternal plasma CRH concentrations were elevated in patients at 28–32 weeks of gestation with an initial diagnosis of threatened preterm labor who delivered within 48 h. However, CRH concentrations were within the normal range in patients with the same initial diagnosis who proceeded to delivery at term. At the present time, it seems unlikely that a single measurement of maternal plasma CRH will provide an adequate means of predicting the patient who is at risk of preterm labor. However, we and others have suggested that a combination of biochemical tests, including CRH and salivary estriol, combined with measurements of fibronectin may be of sufficient sensitivity and specificity to be of clinical use.

Birth – An Integrated Cascade?

From the preceding discussion, it should be apparent that birth, at term and preterm, results from processes leading to increased PG output. Glucocorticoids have a central role in those processes. Glucocorticoids also stimulate CRH output within the placenta and fetal membranes and CRH similarly upregulates PGHS-2 and downregulates PGDH. The effects of CRH may be modulated by the state of the CRH receptor. Oxytocin appears to play a key role in changing the affinity of CRH receptor interaction. Oxytocin could be derived from the systemic circulation, but also could be derived locally from the chorion and/or decidua. Increased levels of cortisol could be derived from the maternal circulation, for example in association with a maternal stress response, or from the fetus following precocious activation of the fetal HPA axis. In addition, cortisol can be formed locally within chorionic trophoblast cells from the inactive precursor cortisone. The expression of 11ß-HSD-1 enzyme, which effects this conversion, progressively increases in chorionic trophoblasts during human gestation. More recently, it has been shown [Alfaidy N., Challis J.R.G., unpubl. results] that the activity of 11ß-HSD-1 is significantly increased by PGs through a mechanism that involves increased release of intracellular calcium and phosphorylation of the enzyme. In this way, increased production of PGs (PGE2, PGF2·) increases 11ß-HSD-1 activity, leading to more cortisol formation from cortisone (fig. 1). Cortisol in turn increases further PG production. It is evident that with infection, other agents such as cytokines can intercede in this series of loops by stimulating PGHS-2 and downregulating PGDH expression. It is also apparent that the mechanisms which predispose to preterm labor almost certainly vary at different stages of gestation. The incidence of preterm birth in association with chorioamnionitis is higher earlier in pregnancy. Later in gestation, the fetal stress response may predominate. In this situation, fetal HPA activation increases fetal cortisol output, which in turn upregulates

Challis/Smith

placental CRH expression. This is consistent with elevated concentrations of CRH in the umbilical cord plasma of fetuses with intrauterine growth restriction. Recently, it has been shown that placental CRH also drives fetal adrenal steroidogenesis. This leads to increased production of DHEA from the fetal zone of the fetal adrenal. DHEA in turn is aromatized in the placenta to estrogen, thereby contributing to myometrial activation. There are several caveats to utilizing this information to develop better diagnostic tests for the patient at risk of preterm birth. First, one must recognize that the prevention of preterm delivery may not always be in the best interests of the fetus. Second, the causes of preterm birth likely vary at different times of gestation and this needs to be recognized and treated appropriately. It is very difficult to envisage that a single test or a single management strategy will be efficacious for all patients. Hence, many groups of investigators are now turning their attention to microchip gene array techniques for the prediction of premature birth. Ongoing studies in rats have shown that distinct gene clusters are altered at the time of birth, and that there is a greater tendency to decrease rather than increase gene expression at this time. Future studies will need to examine the applicability of this information to human pregnancy. It is possible that gene array techniques can be applied to maternal peripheral blood samples for diagnostic purposes. The ability to predict or diagnose the patient in preterm labor will be invaluable in selecting those women to whom prenatal corticosteroids should be administered to help promote fetal pulmonary maturity. There is increasing concern regarding the use of repeated corticosteroid administration at regular intervals in women who may not actually be at risk of preterm birth. Animal studies and human studies have now demonstrated detrimental effects of glucocorticoids on fetal growth, glucose homeostasis, cardiovascular function and neu-

Fig. 1. Cortisol and PG interactions in human fetal membranes.

ral development. Clinically, the aim should be to restrict the use of corticosteroids and tocolytic treatment to those patients in whom preterm labor has been diagnosed. The purpose of continuing studies in this area will be to achieve those objectives.

Acknowledgments

Work in the authors’ laboratory has been supported by the Canadian Institute of Health Research (CIHR) Group in Fetal and Neonatal Health and Development. We gratefully acknowledge support from the Parke-Davies Fellowship Trust of the University of Cambridge.

References 1 Challis JRG, Matthews SG, Gibb W, Lye SJ: Endocrine and paracrine regulation of birth at term and preterm. Endocrine Rev 2000;21:514–550. 2 Lye SK, Ou C-W, Teoh TG, Erb G, Stevens Y, Casper R, Patel FA, Challis JRG: The molecular basis of labour and tocolysis. Fetal Matern Med Rev 1998;10:121–136. 3 Liggins GC, Thorburn GD: Initiation of parturition; in Lamming GE (ed): Marshall’s Physiology of Reproduction. London, Chapman and Hall, 1994, pp 863–1002. 4 Norwitz ER, Robinson JN, Challis JRG: The control of labor. N Engl J Med 1999;341:660–666. 5 Mitchell MD, Edwin SS, Lundin-Schiller S, Silver RM, Smotkin D, Trautman MS: Mechanism of interleukin-1ß stimulation of human amnion prostaglandin biosynthesis: Mediation via a novel inducible cyclooxygenase. Placenta 1993;14:615– 625. 6 Kniss DA: Cyclooxygenases in reproductive medicine and biology. J Soc Gynecol Invest 1999;6: 285–292. 7 Challis JRG: Characteristics of parturition; in Creasy R, Resnik R (eds): Maternal-Fetal Medicine: Principles and Practice. Philadelphia, W.B. Saunders, 1988, pp 484–497. 8 Sangha RK, Walton JC, Ensor CM, Tai H-H, Challis JRG: Immunohistochemical localization, messenger ribonucleic acid abundance, and activity of 15-hydroxyprostaglandin dehydrogenase in placenta and fetal membranes during term and preterm labor. J Clin Endocrinol Metab 1994;78:982– 989.

Preterm Labor

9 Keelan JA, Sato T, Mitchell MD: Interleukin (IL)-6 and IL-8 production by human amnion: Regulation cytokines, growth factors, glucocorticoids, phorbol esters, and bacterial lipopolysaccharide. Biol Reprod 1997;57:1438–1444. 10 Gibb W, Lavoie JC: Effects of glucocorticoids on prostaglandin formation by human amnion. Can J Physiol Pharmacol 1990;68:671–676. 11 Economopoulos P, Sun M, Purgina B, Gibb W: Glucocorticoids stimulate prostaglandin H synthase type-2 (PGHS-2) in the fibroblast cells in human amnion cultures. Mol Cell Endocrinol 1996;117:141–147. 12 Challis JRG, Patel FA, Pomini F: Prostaglandin dehydrogenase and the initiation of labor. J Perinat Med 1999;27:26–34. 13 Patel FA, Clifton VL, Chwalisz K, Challis JRG: Steroid regulation of prostaglandin dehydrogenase activity and expression in human term placenta and chorio-decidua in relation to labor. J Clin Endocrinol Metab 1999;84:291–299. 14 Sparey C, Robson S, Bailey J, Lyall F, Europe-Finner GN: The differential expression of myometrial connexin-43, cyclooxygenase-1 and -2 and Gs· proteins in the upper and lower segments of the human uterus during pregnancy and labor. J. Clin Endocrinol Metab 1999;84:1705–1710. 15 Petraglia F, Florio P, Nappi C, Genazzani AR: Peptide signaling in human placenta and membranes: Autocrine, paracrine and endocrine mechanisms. Endocr Rev 1996;17:156–186. 16 Linton EA, Perkins AV, Woods RJ, Eben F, Wolfe CD, Behan DP, Potter E, Vale WW, Lowry PJ: Corticotropin releasing hormone-binding protein

17

18

19

20

21

22

23

Biol Neonate 2001;79:163–167

(CRH-BP): Plasma levels decrease during the third trimester of normal human pregnancy. J Clin Endocrinol Metab 1993;76:260–262. Frim DM, Emmanuel RL, Robinson BG, Smas CM, Adler GK, Majzoub JA: Characterization and gestational regulation of corticotropin-releasing hormone messenger RNA in human placenta. J Clin Invest 1988;82:287–292. Karalis K, Majzoub JA: Regulation of placental corticotropin-releasing hormone by steroids. Possible implications in labor initiation. Ann NY Acad Sci 1995;771:551–555. Grammatopoulos D, Hillhouse EW: Activation of protein kinase C by oxytocin inhibits the biological activity of the human myometrial corticotropinreleasing hormone receptor at term. Endocrinology 1999;140:585–594. Grammatopoulos D, Hillhouse EW: Role of corticotropin-releasing hormone in onset of labour. Lancet 1999;354:1546–1549. Stevens Y, Challis JRG, Lye SJ: Corticotropinreleasing hormone receptor subtype 1 (CRH-R1) is significantly upregulated at the time of labor in the human myometrium. J Clin Endocrinol Metab 1998;83:4107–4115. McLean M, Bisits A, Davies J, Woods R, Lowry P, Smith R: A placental clock controlling the length of human pregnancy. Nat Med 1995;1:460–463. Korebrits C, Ramirez MM, Watson L, Brinkman E, Bocking AD, Challis JRG: Maternal corticotropin-releasing hormone is increased with impending preterm birth. J Clin Endocrinol Metab 1998; 83:1585–1591.

167

Biol Neonate 2001;79:168–171

Influence of Maternal Stress on Fetal Behavior and Brain Development Jean-Pierre Relier Hôpital Port-Royal, Paris, France

Key Words Fetus W Maternal stress W Brain W Asphyxia W Psychoaffective interchanges

Abstract The very early establishment of certain sensory faculties turns the fetus into a being capable of perceiving multiple stimuli. This perceptive capability forms part of many interchanges between the mother and her developing child. These interchanges are doubtless not only biological and metabolic in nature, but also sensorial and sensitive. The importance of a good quality of psychoaffective communication between mother and child during pregnancy has been shown to be decisive for fetal growth and also for the perinatal period and further development of the child. Maternal psychological stress leads to adverse pregnancy outcome. Chronic anxiety causes an increased stillbirth rate, fetal growth retardation and altered placental morphology. Experimental studies have demonstrated a relationship between specific episodes of maternal psychological stress and exacerbation of fetal asphyxia in utero. It is concluded that all the psychoaffective interchanges between the mother and child are decisive for harmonious fetal growth and brain development. Copyright © 2001 S. Karger AG, Basel

Introduction

Stress is defined as ‘a breakdown in equilibrium, i.e. homeostasis, that is due to life events’. At the time this definition was coined, only major disturbances of life that led to biophysical adjustments that were easily measurable and that could lead to the ultimate expression, i.e. disease, were accepted as stress. Since the concept of homeostasis depends on the constancy of the milieu interne, it became evident that stress would have to be redefined as all environmental events that modify the milieu interne, which is characteristic of the biology of living beings. This concept undergoes constant modification and is absolutely not static. To undertake a study on the influence of maternal stress on fetal behavior and brain development is undoubtedly to study the ‘bio-

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0168$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

affective’ consequences of acute, subacute and chronic stress on fetal behavior and development. It is also, and perhaps more importantly, over and above the harmful role of stress, an evaluation of whether stress or stresses are perhaps useful or even necessary for harmonious growth. It is also to assess the role of the mother and her means of ‘protecting’ her fetus against the sometimes dramatic consequences of this aggression; in other words, to investigate the usefulness of psychological management in a mother at risk of premature birth.

The Nature of Stress

It is now agreed that fetal growth results from two developmental influences: (1) expression of the genetic inheritance (innate), and (2) epigenetic and environmental factors (acquired). In the setting of neurobiology, the very early acquisition, between 8 and 10 weeks of gestational age, of receptors for the main functions of perception [1], which constitutes the passage from the embryonic to the fetal period, gives a good idea of the importance of these epigenetic or environmental factors in, for example, the quality of cerebral growth. It can thus be considered that stress, an environmental factor, can play a role in the course of these two processes, either by its direct effect on the growth or differentiation of an organ or function, or, more often, by a mixed effect. In the latter case, it is often impossible to determine the main influence permitting or hindering the genomic expression inherited from the genetic legacy of the parents, in addition to the direct action of the environmental factor on cellular growth and differentiation. When talking of pregnancy, we will make a fundamental distinction between retrospective or prospective data on chronic or subacute stress, which is most often psychoaffective or psychosocial, and the spectacular results of certain experiments on acute stress, especially in animals. Considering the four periods of prenatal life and growth, stress could alter the mother-fetus relationship and embryo-fetal growth at the time of (1) creation, (2) conception, (3) the embryonic phase, or (4) the fetal phase.

Prof. Dr. J.-P. Relier Centre de Recherches Biologiques Néonatales, Hôpital Port-Royal F–75674 Paris Cedex 14 (France) Tel. +33 1 584 12151, Fax +33 1 4329 0338 E-Mail [email protected]

Time of Creation

This period, very poorly recognized by clinical doctors, corresponds to what is called ‘grossesse psychique’. For psychoanalysts and psychiatrists, this is the time when a man and woman who are in love have the desire of creation, of having a baby. The infant appears as a real ‘materialization’ of total love between two complementary human beings, man and woman, at three different levels: comportmental, emotional and phantasmic. At this period, the influence of stress could be extremely different. Two different categories of stress are possible: (1) from the immediate environment, often but not always conscious, and (2) from this ‘transgenerational heritage’, which is dominated by the unconscious. The consequences of these stresses are unpredictable. Most obvious is the impossibility of becoming pregnant, leading to the diagnosis of ‘sterility’, which, very often, is only ‘hypofertility’ which could be treated and cured by psychotherapy or other advice.

Time of Conception and Implantation

This is what is called ‘grossesse physique’ due to the union of ovule and spermatozoon in the fallopian tube. The fundamental importance of the fallopian tube in conception and implantation has been pointed out by Gandolfi [2] by summarizing this role on five different postovulatory days. For each of these days there is an action: (1) on the ovule by secretion of specific proteins with long chains and high molecular weight, different for each postovulatory day and active at the time of implantation; (2) on the spermatozoon by preservation of spermobiles, and (3) on the fertilized ovule by an active role in division and growth and activation of genoma. This action is essential for the first steps of growth but almost absent during in vitro fertilization, reducing the efficiency of this technique with possible long-term consequences after birth [3]. Recently, at the meeting of the neonatal society in January 1999, Victor Han [4] reported that it is now possible to isolate IGF1 and IGF2 receptors on the surface of ovula. This distribution varies according to the time of ovulation-conception and implantation in the uterine mucosa. Han also accepts the idea that this distribution is a function of the mother’s psychoaffective quality. This means that psychological stresses in the mother could act on the action of IGF2, which is responsible for the placental vascular organization. These data are relevant and point out the fundamental importance of a real psychoaffective equilibrium of the parents as well as the conscious and unconscious in the quality of early growth of the embryo and placenta, preventing intrauterine growth retardation, prematurity, maternal hypertension, toxemia and early miscarriage.

The Embryonic Period

This period starts at the time of implantation in the uterine mucosa and lasts until the 8th week after ovulation. Embryonic growth is fantastic, since in 8 weeks or 56 days the embryo goes from one cell (fertilized ovule) to more than 8 billion cells. Mainly under genetic influence, this cell explosion is sensitive to the environment through the placenta, which some psychiatrists like Michel Soulé consider as part of the trials during pregnancy ‘mother-fetus-placenta’. As we

Influence of Maternal Stress on Fetal Behavior and Brain Development

pointed out, Han [4] recalls the essential actions of IGF2 in organizing trophoblastic invasion and the placental vascular network. In the literature, many authors have already published studies regarding the importance of a good psychological attitude by the mother in the prevention of intrauterine growth retardation [5–7], mental handicaps [6] and prematurity [8]. In 1962–1963, Grimm [9] and James [10] wrote about the importance of psychological investigations and approaches in the prevention of early habitual and spontaneous abortion.

The Fetal Period

It is during the time of pregnancy that we have the most experimental and clinical data about the consequences of maternal stress on fetal behavior and the neonatal brain. Acute Stress In this section, we will consider only acute emotional or affective psychosensory stress, thus eliminating purely physical stress of an aggressive nature. Here again, we have both human observations and experimental animal data. These two types of data seem complementary, since, in women, acute affective psychosensory stress is very often followed by subacute anguish or anxiety and can lead to a form of ‘rejection’ that is extremely harmful for the fetus, especially if the stress occurs at a critical time in cerebral development. Experimental data are relatively recent. Ronald Myers first took an interest in acute emotional stress in female monkeys in 1972. Among the many reports on acute fetal distress (AFD) due to fetal or maternal strain, one of the most important is the Glasgow report (September 1973) on the triggering of AFD by one single instance of maternal psychological stress [11]. In this series of experiments, Myers confirmed that laboratory signs of fetal asphyxia (fall in PO2 and pH, rise in PCO2) never occurred at the time of surgical placement of the fetal catheters when the mother monkey is anesthetized. On the other hand, the signs of AFD always occurred, in each model, when the monkey awoke, that is to say, when the effects of anesthesia ceased exactly 50 s after awakening. Myers therefore considers that AFD was not due to the surgical procedure or mechanical aggression, but instead to stimulation of the mother’s sympathetic nervous system at the moment she awakes. This leads to the release of catecholamines and medulloadrenal hormones due to the mother’s distress at finding herself attached to the operating chair in a strange environment, giving rise to a stress that is uniquely psychoaffective. Moreover, all laboratory signs of AFD disappear when the monkey is again anesthetized (rise in PO2 and pH, fall in PCO2). Myers refined his experiments by varying the type of psychoaffective stress. Sometimes a stress that he considered to be minor resulted in dramatic changes, requiring renewed anesthetization of the mother to save the fetus. In his presentation, Myers briefly reviewed the literature of the time. He recalled the observations in humans where chronic anxiety leads to intrauterine growth retardation as well as anomalies of morphology and placental vascularization [12]. It is remarkable that already in 1972, Myers identified an important element concerning the role of early abnormalities of the mother-fetus interaction in the establishment of placental vascular lesions that appear only later as AFD or even hypertension or toxemia.

Biol Neonate 2001;79:168–171

169

Still in the domain of a single unique stress, one cannot help being impressed by the dermatoglyphic abnormalities resulting from maternal stress during the second trimester of pregnancy, which have been demonstrated both in the macaque monkey and in humans. Two publications give representative accounts of this phenomenon. That of Newell-Morris et al. [13] reported that in the macaque monkey, following maternal psychosocial stress, the newborns had dermatoglyphic asymmetry between the right and left hands that was significantly greater than in the control group. This maternal stress did not lead to any significant differences in the duration of pregnancy, birth weight or mean survival. On the other hand, major dermatoglyphic asymmetry between the right and left hands (8 versus 13) was always associated with greater perinatal mortality. The report of Bracha et al. [14] confirmed this finding in humans and threw new light on the epigenetic interactions between the environment and genetics in the occurrence of schizophrenia. The authors pointed out that the second trimester of pregnancy in humans is an important period for the massive migration of neurons to the cortex. In the same period, the dermal cells of the papillae of the fingers migrate to form the dermatoglyphics. In view of the findings of Newell-Morris and other authors, Bracha et al. [14] decided to evaluate the different levels of tolerance to a maternal psychological stress in pairs of monozygotic twins. Twenty-three out of the 30 pairs were discordant for schizophrenia and 7 pairs were normal. The result was extremely interesting: in the twins discordant for schizophrenia, the number (either more or less) of papillary ridges in twins of the same pair differed significantly. That is to say, in the pairs discordant for schizophrenia, no pair had the same number of digital ridges, despite their homozygosity. The authors concluded that twins, even homozygous ones, but with two different amniotic pouches and placentae, can react differently to the same maternal stress during the second trimester of pregnancy, to such an extent that only one of the two twins exhibits the results of the stress, in this case schizophrenia. These findings help to explain the observation of Huttunen and Niskanen [15] on the persistence of behavioral abnormalities long after birth in humans. These authors showed that adults whose fathers died before their birth commit more criminal acts and have more severe psychiatric problems than those whose fathers died in the year following their birth [15]. This finding suggests not only that maternal stress can directly affect the fetus but also that the effect can be more or less permanent. Thus, certain psychiatric pathologies, including some that are very severe and well defined, may be caused by a major disturbance in the mother-infant relationship due to maternal stress during pregnancy. Huttunen and Niskanen [15] stress the gravity of certain traumas that occur between the third and fifth month of pregnancy, during development of the thalamus, which is believed to be an early center of emotion [16]. On the other hand, the protective role of the mother has been well demonstrated by pregnancies that occurred under dramatic conditions in the Second World War as well as in cases of subacute stress where the mother gave birth to a normal term infant [17]. Chronic or Subacute Stress Until recent years, practically all available studies were retrospective. The oldest involved major events such as death in utero or very premature birth with neonatal death in populations subjected to wartime trauma such as emigration, terror and famine [17]. We have already quoted the data about the mother’s stress and intrauterine growth retardation [5, 7], mental handicaps [6] and prematurity [8]. In France, Choquet et al. [18] and Choquet and Ledoux [19] demon-

170

Biol Neonate 2001;79:168–171

strated a relationship between the overall condition of the infant at 3 years of age and that of the mother and even the father during pregnancy. In 1973, Stott [20] had already shown in his series that these pediatric disorders were related to ‘emotional stress experienced by the mother during her pregnancy’, whereas obstetrical pathology had no influence on the occurrence of such disorders. In a prospective study in 1990, Richard [21] made an interesting contribution to this approach. In the Tours University Maternity Hospital, she interviewed 100 randomly selected pregnant women whose psychoaffective status was assessed using two questionnaires, i.e. at 6 months of pregnancy and just before birth. Each infant was examined at least twice, at birth for postnatal problems and when aged between 6 and 8 months to assess mental and physical health. Three groups of mothers were established according to their psychic equilibrium: group A consisted of 47 normal pregnancies with no serious psychological or psychoaffective anomalies, group B of 29 moderately disturbed pregnancies with compensated psychic anomalies, and group C of 24 pregnancies with a major psychoaffective disorder in the form of severe psychological distress such as refusal or abandon of the infant, or death, or any other event that severely perturbed the psychological equilibrium of the mother, and thus the mother-infant relationship. The results show a significant difference between groups A and C for rates of premature birth (4% in group A, 17% in groups B and C), birth complications (12% in group A, 25% in group C), neonatal complications (4% in group A, 13% in group C) and feeding problems in the first week of life (9% in group A, 40% in group C). Differences in the first year were even more marked, as shown in the results for the prevalence of sleep disorders (2% in group A, 20% in group C), feeding problems (13 versus 24%), ENT disorders (13 versus 29%), spasme du sanglot (10 versus 33%), other problems (15 versus 58%) and difficult temperament (30 versus 76%). Finally, the presence of at least two health problems was noted in 8% of infants in group A and in 43% in group C. Richard [21] and many other authors interpret these finding as follows: ‘Any negative influence, such as maternal anxiety, runs the risk of damaging the subtle interaction of the mother-infant pair. The perturbations experienced by the mother during pregnancy can contribute to the maintenance of postnatal maternal anxiety and lead to the development of an affective conflict.’ One thing is certain in our experience of pathological pregnancies, and that is the presence of a common denominator, i.e. ‘maternal stress’, whatever the cause. Very often, this stress is difficult to demonstrate because it is masked by the educational conditioning of the mother. This absence of any manifestation of stress is, unfortunately, not at all equivalent to a stable stress that is compensated and thus less aggressive or even ‘structuring’ for the developing fetus. Some research groups have already succeeded in providing experimental proof of an abnormal secretion of catecholamines in chronic anxiety, and of a relationship between the level of this chronic secretion of catecholamines and the physical development of the infant and his/her behavior towards the mother. Thus, during chronic or subacute psychological stress, there is already a satisfactory physiological and pathogenic explanation of the consequences of maternal or even parental pathology on the development of the fetus. Environmental Stress As noted at the beginning of this paper, study of the consequences of maternal or parental stress on the development of the fetus and infant cannot be limited to quantifiable stress.

Relier

The idea implicit in the title of this paper is the investigation of which particular events during pregnancy can modify the development of the fetus and the infant. Obviously, it is not feasible here to consider, even briefly, the consequences of poisoning or of certain diseases that can be considered a form of stress, either acute or chronic. In the case of the influence of the environment, is it not perhaps more precise to talk of ‘imprint’ rather than stress? There are many imprints that can affect the fate of the individual; they may arise from the socioeducational environment, from emotional experience, from parental messages of permission or prohibition that affect behavioral activity, or, indeed, from the genomic imprint and early environmental effects that will have repercussions on the development and morbidity of the adult [22]. There are critical periods during fetal development that correspond to phases of growth and functional differentiation, during which structures and functions are established. Thus, various kinds of influence, especially nutritional, can affect the development of certain tissues or systems whose repercussions will be felt only later, during adult life. David Barker and his environmental epidemiology unit at Southampton have devoted more than 15 years to ‘the study of the intrauterine environment’. These epidemiological studies, carried out with the necessary rigor, seem to prove that hypertension, lipid disorders, non-insulin-dependent diabetes mellitus and hyperfibrinogenemia, i.e. the classical signs of coronary artery disease, are programmed during fetal life and are the late consequence of a pathological early environment reflected in intrauterine growth retardation and a low weight at birth and at 1 year [23]. Numerous other investigations of correlations have been made, but it is not possible here to give an exhaustive review.

It is interesting to note that Barker’s group [24] reports fingerprint abnormalities in all these affected adults, which appear to be a marker of an anomaly or stress that occurred during intrauterine life, as in the findings of Newell-Morris et al. [13].

Conclusions

Thus, the study of the consequences of maternal stress on fetal development and on the infant tends to take into consideration all the elements that can modify, alter or interrupt for a certain time the quality of the psychoaffective exchange between the mother, the two partners and their developing infant. This rather succinct exposition of the basis of the biology of fetal development completely neglects one aspect of the role of the mother, who can not only totally protect her infant against most of these environmental influences, but can also participate to a great extent in the process of cure, which can lead to a complete recovery thanks to the high degree of plasticity of nearly all the organs characteristic of this period of development. As is well known, particularly in developmental neurobiology, this capacity for recuperation is enormously influenced by the quality of the mother-infant relationship. If some environmental influences can appear to be a real source of stress during fetal development, it has also been demonstrated that ‘in the same way as nutritional or energetic requirements or the quality of the placenta-fetal exchange, the emotion, joy, or sadness of the mother play a fundamental role. In this respect, love, whose impact cannot yet be measured, undoubtedly represents the peripheral or environmental stimulation that is most appropriate to ensure the growth and harmonious equilibrium of the infant’ [25].

References 1 Relier JP: Importance of fetal perceptions in the organisation of mother-fetus interactions. Biol Neonate 1996;69:165–212. 2 Gandolfi F: The role of the fallopian tube. Communication at ‘The Early Human Life’, Vatican City, Sinodo Room, September 6–8, 2000. 3 Relier JP, Monset-Couchard M, Huon C: The neonatologist’s experience of in vitro fertilization risks; in Stephenson P, Wagner MG (eds): Tough Choices. In vitro Fertilization and the Reproductive Technologies. Philadelphia, Temple University Press, 1993, pp 135–143. 4 Han V: Fetal nutrition and the paracrine regulation of fetal growth. Keynote lecture, The Neonatal Society. Joint meeting with the Paediatric Section of the Royal Society of Medicine, London, 20 January, 1999. 5 Sontag LW: The significance of fetal environmental differences. Am J Obstet Gynecol 1941;42:996– 1003. 6 Stott DH: Physical and mental handicaps following a disturbed pregnancy. Lancet 1957;18:1006– 1012. 7 Newton RW, Hunt LP: Psychosocial stress in pregnancy and its relation to low birth weight. Br Med J (Clin Res Ed) 1984;288:1191–1194. 8 Mamelle N, Laumon B, Lazar P: Prematurity and occupational activity during pregnancy. Am J Epidemiol 1984;119:309–322. 9 Grimm ER: Psychological investigations of habitual abortion. Psychosom Med 1962;24:369–378.

Influence of Maternal Stress on Fetal Behavior and Brain Development

10 James WH: The problem of spontaneous abortion. X. The efficacy of psychotherapy. Am J Obstet Gynecol 1963;85:38–40. 11 Myers RE: Production of fetal asphyxia by maternal psychological stress; in Rippmann ET, Stamm H, McEwan HP, Howie P (eds): III International Symposium EPH-Gestosis, Glasgow, 26–29 September, 1973. Glasgow, Organisation Gestosis Press, 1975. 12 Myers RE: The pathology of the rhesus monkey placenta. Acta Endocrinol Suppl (Copenh) 1972; 166:221–257. 13 Newell-Morris LL, Fahrenbruch CE, Sackett GS: Prenatal psychological stress, dermatoglyphic asymmetry and pregnancy outcome in the pigtailed macaque (Macaca nemestrina). Biol Neonate 1989;56:61–75. 14 Bracha HS, Torrey EF, Gottesman II, Bigelow LB, Cunnif C: Second trimester markers of fetal size in schizophrenia: A study of monozygotic twins. Am J Psychiatry 1992;149:1355–1361. 15 Huttunen RO, Niskanen P: Prenatal loss of father and psychiatric disorders. Arch Gen Psychiatry 1978;35:429–431. 16 Trevarthen C: Le développement du cerveau et la psychologie fœtale; in III° Journées européennes ‘naissance et avenir’. Montpellier, 15–16 December, 1994. 17 Antonov AN: Children born during the siege of Leningrad in 1942. J Pediatr 1947;30:250–259.

18 Choquet M, Facy F, Laurent F, Davidson F: Les enfants à risque en âge préscolaire. Mise en évidence par analyse typologique. Arch Fr Pédiatr 1982;39:185–192. 19 Choquet M, Ledoux S: La valeur pronostique des indicateurs de risques précoces. Etude longitudinale des enfants à risque à 3 ans. Arch Fr Pédiatr 1985;42:541–546. 20 Stott DH: Follow-up study from birth of the effects of prenatal stresses. Dev Med Child Neurol 1973; 15:770–787. 21 Richard S: Influence du vécu émotionnel de la femme enceinte sur le tempérament et la santé physique du nourrisson; in Relier JP (ed): Progrès en néonatologie. Paris, Karger, 1990, vol 10, pp 202– 223. 22 Battin J: Empreinte génomique et environnementale précoce dans l’avenir de l’individu; in Relier JP (ed): Progrès en néonatologie. Paris, Karger, 1996, vol 16, pp 241–255. 23 Barker DJP: The intrauterine origins of cardiovascular diseases. Acta Paediatr Suppl 1993;82(suppl 391):93–100. 24 Godfrey KM, Barker DJP, Cloke J, Osmond L: Relation of fingerprints and shape of the palm to fetal growth and adult blood pressure. BMJ 1993; 307:405–409. 25 Relier JP, Laugier J, Salle B: Avant-propos; in: Médecine périnatale (fœtus et nouveau-né). Paris, Flammarion Médecine-Sciences, 1989, XXII.

Biol Neonate 2001;79:168–171

171

Biol Neonate 2001;79:172–179

Caspase-3 Activation after Neonatal Rat Cerebral Hypoxia-Ischemia Xiaoyang Wang a, b Jan-Olof Karlsson c Changlian Zhu a, b Ben A. Bahr d Henrik Hagberg a, e Klas Blomgren a, f a Perinatal

Center, Department of Physiology, Göteborg University, Göteborg, Sweden; b The Third Affiliated Hospital of Henan Medical University, Zhengzhou, Henan, People’s Republic of China; c Institute of Anatomy and Cell Biology, Göteborg University, Göteborg, Sweden; d Department of Pharmaceutical Sciences and the Neurosciences Program, University of Connecticut, Storrs, Conn., USA; e Perinatal Center, Department of Obstetrics and Gynecology, Sahlgrenska University Hospital/Östra, Göteborg, f Perinatal Center, Department of Pediatrics, Göteborg University, The Queen Silvia Children’s Hospital, Göteborg, Sweden

Key Words Caspase W Neonatal rat W Brain W Hypoxia W Ischemia W Apoptosis W Poly(ADP-ribose) polymerase W Inhibitor of caspase-activated DNase W Fodrin

Abstract Caspase-3 is a major effector protease in several apoptotic pathways, but its role in hypoxic-ischemic (HI) brain injury is incompletely understood. Cerebral HI was induced in 7-day-old rats by unilateral carotid artery ligation and exposure to 7.7% oxygen for 55 min. Caspase-3-like activity was significantly increased at 1 h (208%), peaked at 24 h (2,563%) and was still increased 6 days after HI (169%) in the ipsilateral cerebral cortex. Concomitantly, cleavage of the caspase-3 proform (31/33 kD) was detected on immunoblots, producing 29- and 17-kD fragments. Furthermore, significant degradation of the endogenous caspase-3 substrates inhibitor of caspaseactivated DNase (DNA fragmentation factor 45), poly(ADP-ribose) polymerase and fodrin occurred. In conclusion, caspase-3 is activated extensively in the immature brain after HI. The subsequent cleavage of proteins involved in cellular homeostasis and repair may contribute to the process of brain injury. Copyright © 2001 S. Karger AG, Basel

Introduction

Perinatal hypoxic-ischemic (HI) brain injury is an important cause of neurodevelopmental impairment and disability. The relative contributions of necrosis and apoptosis to the injury that devel-

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0172$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

ops after cerebral HI have been a matter of much debate [1]. Apoptosis is a fundamentally important process whereby the brain can eliminate nonfunctional cells during normal development with a minimum of inflammation. In recent years, many of the molecules have been identified that participate in the biochemical pathways mediating the highly ordered disassembly of cells through apoptosis. At the heart of the pathways is a group of cysteine proteases termed caspases, named in order of publication [2]. They are mammalian homologues of the ced-3 gene product [3] and are essential for the execution steps in apoptosis [4]. The caspases can be classified as regulatory or upstream caspases (such as caspase-2, -8 and -9) or as effector or downstream caspases (such as caspase-3, -6 and -7), as defined by their predominant role in the cascade or the size of the prodomain. Regulatory caspases have a long prodomain capable of interacting with other messenger proteins, whereas effector caspases have a short prodomain. A third group of caspases is involved in mediating inflammatory reactions, and the best characterized of these is caspase-1, also called interleukin-1ß-converting enzyme. Caspases share a unique requirement of an aspartate residue in the P1 position of their substrate proteins, resulting in discrete cleavage of a limited number of target proteins. Caspase substrates are typically structural proteins, or proteins involved in cellular homeostasis and repair. All caspases are synthesized as latent proenzymes, which can be cleaved and activated by other caspases, in a cascade-like manner [5]. Caspase-3 is the most abundant of the known caspases in the brain, and appears to play a crucial role both during normal development and in situations of brain injury. Mice devoid of caspase-3 through gene targeting die at 1–3 weeks of age and display a hyperplastic, disorganized brain. Interestingly, the rest of their bodies,

Klas Blomgren Perinatal Center, Institute of Physiology and Pharmacology Göteborg University, PO Box 432 SE–405 30 Göteborg (Sweden) Tel. +46 31 773 3376, Fax +46 31 773 3512, E-Mail [email protected]

including their thymuses, where extensive apoptosis takes place, appeared to be relatively normal [6]. This indicates a particularly important role for caspase-3 in the brain, and that apoptotic pathways may be different in different organs. There are numerous reports where apoptosis-related parameters have been detected after cerebral ischemia [7–11], as well as reports of neuroprotection after administration of caspase inhibitors [12–15]. The immature brain retains its apoptotic machinery to a larger extent than the adult brain, at least as judged by the presence of caspase-3 [16, unpubl. observations]. Immature neurons may therefore be more prone to apoptotic death, while terminally differentiated neurons are more likely to die by necrosis [17]. The aim of the present work was to characterize the time course and extent of caspase-3 activation in a model of neonatal HI.

Materials and Methods

Surgical Procedures Unilateral HI was induced in 7-day-old Wistar F rats of both sexes [18, 19]. The pups were anesthetized with halothane (3.0% for induction and 1.0–1.5% for maintenance) in a mixture of nitrous oxide and oxygen (1:1), and the duration of anesthesia was ! 10 min. The left common carotid artery was cut between double ligatures of prolene sutures (6–0). After the surgical procedure, the wounds were infiltrated with a local anesthetic and the pups were allowed to recover for 1–2 h. The litters were then placed in a chamber perfused with a humidified gas mixture (7.70 B 0.01% oxygen in nitrogen) for 55 min. The temperature of the gas mixture and the chamber was kept at 36 ° C. Following hypoxic exposure, the pups were returned to their biological dams until sacrificed. Sham-operated animals were used as controls. All animal experimentation was approved by the Ethical Committee of Göteborg (No. 225-97). Preparation of Samples The animals were sacrificed by decapitation 1 h, 3 h, 8 h, 24 h, 3 days or 6 days after HI. Sham-operated animals were sacrificed on postnatal day 7, 10 or 13. The number of animals was 6 or 7 for each time point. The brains were rapidly dissected out on a bed of ice, quickly frozen in isopenthane and dry ice and stored at –80 ° C. Cortical tissue rostral to the hippocampus (approximately 50 mg) was dissected out from each hemisphere at –10 ° C. The tissue was homogenized by sonication in 1,000 Ìl of ice-cold 50 mM Tris-HCl (pH 7.3) containing 5 mM EDTA, aliquoted and stored at –80 ° C. The protein concentration was determined after mixing the samples with the sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer according to the method of Karlsson et al. [20].

Perkin-Elmer LS 50B luminescence spectrometer with a microtiter plate attachment, at an excitation wavelength of 380 nm (slit 15 nm) and an emission wavelength of 460 nm (slit 20 nm). The degradation was followed at 2-min intervals for 1–2 h, and Vmax was calculated from the entire linear part of the curve. Standard curves with AMC in the appropriate buffer were used to express the data in pmoles of AMC formed per minute and milligram of protein. Every sample was analyzed 3–4 times and the average value was used as n = 1. Immunoblotting Homogenate samples were mixed with an equal volume of concentrated (3!) SDS-PAGE buffer and heated (96 ° C) for 5 min. Individual or pooled samples corresponding to 20 Ìg of total protein were run on 8–16% Tris-glycine gels (Novex, San Diego, Calif., USA) and transferred to reinforced nitrocellulose (Optitran, 0.2 Ìm, Schleicher & Schuell, Dassel, Germany) or PVDF (Hybond-P, Amersham, UK) membranes. The membranes were blocked in 30 mM Tris-HCl (pH 7.5), 100 mM NaCl and 0.1% Tween 20 (TBS-T) containing 5% fatfree milk powder for 1 h at room temperature or 4 ° C overnight. Primary antibodies were diluted in TBS-T containing 3% BSA and 9 mM NaN3 and incubated with the membranes for 1 h at room temperature. After washing in TBS-T, the membranes were incubated with the appropriate secondary antibodies diluted in blocking buffer for 30 min at room temperature. Immunoreactive species were visualized using Super Signal Western PICO or DURA chemiluminescent substrates (Pierce, Rockford, Ill., USA) and a LAS 1000 cooled CCD camera (Fujifilm, Tokyo, Japan). Immunoreactive bands were quantified in the linear range of signal development using the Image Gauge software (version 3.3, Fujifilm). Alternatively, membranes were exposed using Fuji RX film (Fuji Photo Film Co., Tokyo, Japan). The films were scanned and immunoreactive bands quantified using the same software. Every sample was analyzed 1–4 times, and when multiple determinations were performed, the average value was used as n = 1. Stripping of membranes for reprobing purposes was performed by incubation in 62.5 mM Tris-HCl (pH 6.7), 100 mM ß-mercaptoethanol and 2% SDS at 50 ° C for 30 min. Antibodies The caspase-3 antibody used was H-277 (Santa Cruz, Calif., USA), diluted 1/1,000. The inhibitor of caspase-activated DNase (ICAD) antibody was a kind gift from Dr. Xiaodong Wang, University of Texas Southwestern Medical Center [21], and diluted 1/300. The poly(ADP-ribose) polymerase (PARP) antibody used was clone C-2-101 (Zymed, San Fransisco, Calif., USA), diluted 1/500. The ·fodrin antibody (FG 6090) was obtained from Affiniti Research Products Ltd., Mamhead, UK, and diluted 1/500. Peroxidase-conjugated, anti-rabbit or anti-mouse secondary antibodies were from Vector (Vector Laboratories, Burlingame, Calif., USA).

Fluorometric Assay of Caspase-3-Like Activity Samples of homogenate (50 Ìl) were mixed with 50 Ìl of extraction buffer containing 50 mM Tris-HCl (pH 7.3), 100 mM NaCl, 5 mM EDTA, 1 mM EGTA, 3 mM NaN3, 1 mM PMSF, 1 Ìg/ml pepstatin, 2.5 Ìg/ml leupeptin, 10 Ìg/ml aprotinin and 0.2% CHAPS, on a microtiter plate (Microfluor, Dynatech, Va., USA). After incubation for 15 min at room temperature, 100 Ìl of peptide substrate, 50 ÌM Ac-Asp-Glu-Val-Asp-aminomethyl coumarine (Ac-DEVD-AMC; Enzyme Systems Products, Livermore, Calif., USA) in extraction buffer without inhibitors or CHAPS, but with 4 mM DTT, was added. Cleavage of the substrate was measured at room temperature using a

Statistics The Mann-Whitney U test with Bonferroni correction was used throughout.

Caspase-3 in Neonatal Cerebral Hypoxia-Ischemia

Biol Neonate 2001;79:172–179

Results

Activity Assay Caspase-3 activity (DEVD cleavage) could be detected in the control animals. There was no significant difference between the hemispheres (fig. 1), so the average value of the ipsi- and contralateral

173

Fig. 1. Caspase-3-like activity (DEVD cleavage) in the parietal cortex from control animals and from animals allowed to recover after HI for the time indicated in hours (h) or days (d). P7, 10 and 13 indicate the postnatal day for control animals. The average value B SD from 6–8 animals is shown for the contralateral and ipsilateral hemispheres. The activity is measured as pmoles of AMC released per minute and milligram of protein. The inset shows a magnification of the values for the control animals, to demonstrate the decrease between postnatal day 10 and 13. ° p ! 0.05,°° p ! 0.01 compared with control animals. * p ! 0.05, ** p ! 0.01 comparing the ipsilateral and contralateral hemispheres at the same time point. In the inset, the average value between the contra- and ipsilateral hemispheres from each animal was used as n = 1 when comparing different ages.

Table 1. Results from immunoblots of

individual samples 24 h after HI (n = 7), demonstrating the statistically significant degradation of the caspase-3 proform (32 kD) and the three endogenous substrates PARP, ICAD and fodrin in the ipsilateral hemispheres

Caspase-3

32 kDa (I/C) 29/32 kD (I) 17/32 kD (I)

PARP

116 kD (I/C) 85/116 kD (I)

ICAD

long (45 kD) (I/C) short (3.5 kD) (I/C)

Fodrin

240 kD (I/C) 145+150 kD (I/C) 120 kD (I/C)

P7 24 h P7 24 h P7 24 h

Ratio, %

SD

n

p

97.7 57.4 1.8 10.6 0.13 0.59

14.4 13.2 0.4 4.4 0.04 0.26

7 7 7 7 7 7

0.004

P7 24 h P7 24 h

89.0 43.4 32.6 210.0

19.6 15.2 11.6 110.9

7 7 7 7

P7 24 h P7 24 h

96.5 55.9 107.9 76.1

19.7 13.2 21.6 25.6

7 7 7 7

P7 24 h P7 24 h P7 24 h

91.0 55.0 89.4 508.7 109.9 371.7

57.7 7.5 10.9 115.1 37.6 71.1

7 7 7 7 7 7

0.0017 0.0017 0.004 0.0017 0.0027 0.035 0.013 0.0017 0.006

Comparison is made with postnatal day 7 (P7) control animals. The average values of the ratios between the ipsilateral (I) and the contralateral (C) hemispheres for the proteins specified are listed. In the case of the 29- and 17-kD caspase-3, and the 85-kD PARP degradation products, which cannot be found in the controlateral hemispheres, the ratios listed are between the specific band and the proform.

174

Biol Neonate 2001;79:172–179

Wang/Karlsson/Zhu/Bahr/Hagberg/ Blomgren

hemispheres was used as n = 1 for comparison with the control animals. There was a 49% decrease from postnatal day 10 to 13 (p = 0.0027) (fig. 1). After HI, the activity in the ipsilateral hemisphere was significantly increased already after 1 h of recovery (208%; p = 0.018), reached a peak after 24 h of recovery (2,563%; p = 0.002) and was still increased 3 days (364%; p = 0.006) and 6 days (169%; p = 0.015) after the insult (fig. 1). In comparison, the contralateral hemisphere displayed more moderate changes; the activity was significantly increased 8 h (208%; p = 0.002), 24 h (141%; p = 0.009) and 3 days (162%; p = 0.037) after the insult (fig. 1). Caspase-3 Protein Procaspase-3 displayed a double band, with apparent molecular weights of 33 and 31 kD, respectively (fig. 2a). From 1 h (clearly from 3 h) up to 6 days after HI, the ipsilateral hemispheres also displayed one or two additional bands with apparent molecular weights of approximately 29 and 17 kD, respectively (fig. 2a). These bands were never found in the contralateral hemispheres. The appearance of the 29- and 17-kD bands occurred parallel to a decrease in the 33/31-kD proform (called 32), as indicated by a decreased ipsi-/contralateral ratio (table 1). This processing of the proform to lighter bands reached a maximum 24 h after HI (fig. 2a–c), a time point when the DEVD-cleaving activity also had reached its maximum (fig. 1). Results from densitometric quantification of the proform and fragments in pooled samples (n = 6 or 7) are shown in figures 2b and c. To check the statistical significance, analysis of individual samples after 24 h of recovery was performed. The ratio of the 32-kD proform between the ipsilateral and the contralateral hemispheres changed from 98% in the controls to 57% after 24 h of recovery (p = 0.004) (table 1). The ratio between the 29- and 32-kD bands in the ipsilateral hemispheres increased nearly 6-fold (p = 0.0017), and the ratio between the 17- and 32-kD bands increased more than 4-fold (p = 0.0017) (table 1). Endogenous Substrates Poly(ADP-Ribose) Polymerase. PARP is a DNA repair enzyme, known to be cleaved by caspase-3 [22], and it is a frequently utilized marker of caspase-3 activity. The antibody against PARP stained the 116-kD intact enzyme as well as an 85-kD degradation product. Both these bands could be detected in control animals as well as after HI. There was an upregulation of PARP in both hemispheres after the insult (fig. 3a, b). Degradation of PARP, as judged by the depletion of the intact enzyme and an increase in the 85-kD degradation product, was most obvious after 24 h of recovery (fig. 3a–c). This was reflected as a significantly decreased ratio of the 116-kD intact enzyme between the ipsilateral and the contralateral hemispheres, as well as a more than 6-fold increase in the ratio of the 85- and 116-kD bands in the ipsilateral hemispheres (table 1).

Caspase-3 in Neonatal Cerebral Hypoxia-Ischemia

Fig. 2. a A caspase-3 immunoblot of pooled samples (n = 6 or 7 for

each time point) of parietal cortex from control animals and from animals allowed to recover after HI for the time indicated in hours (h) or days (d). P7, 10 and 13 indicate the postnatal day for control animals. C and I indicate the contra- and ipsilateral hemispheres, respectively. The apparent molecular weights are indicated for the proform (32 kD) and for the two fragments (29 and 17 kD). The 29and 17-kD fragments could only be observed in the ipsilateral hemispheres. The 32-kD band is overexposed in this image to make the 29- and 17-kD fragments more obvious. b Results from densitometric quantification of the bands detected in the ipsilateral hemispheres only, after caspase-3 immunoblotting of pooled samples as in a. The values (OD per Ìg of protein) for the 32-kD proform and for the 29and 17-kD fragments are indicated. There was an approximately 30% (at 1–8 h) to 100% (at 6 days) upregulation of the caspase-3 proform after HI, compared with the controls. Degradation of the caspase-3 proform and increased levels of fragments were detected, most pronounced 24 h after HI. P = Postnatal day; d = day. c The relative ratio between the fragments and the proform in the ipsilateral hemispheres of the pooled samples (as in a), indicating the pronounced production of the 29- and 17-kD fragments around 24 h after HI. P = Postnatal day; d = day.

Biol Neonate 2001;79:172–179

175

(ICAD-L, 45 kD) and a short (ICAD-S, 35 kD) form. The antibody stained two double bands detectable in virtually all samples (fig. 4a). There appeared to be an upregulation of ICAD-L following the insult, concomitant with a pronounced decrease in the ipsilateral hemisphere 24 h after HI. The changes in ICAD-S were similar, but less pronounced (fig. 4b–e). Both ICAD-L and ICAD-S were degraded, but no specific degradation products could be detected in our model using this antibody. A significantly decreased ratio of both ICAD-L and ICAD-S was detected 24 h after HI (table 1). Fodrin. Fodrin (nonerythroid spectrin) is an actin-binding protein and a major component of the plasma membrane skeleton [24]. ·-Fodrin is a well-established substrate of calpains, producing 145and 150-kD cleavage products [25], but has recently also been identified as a caspase-3 substrate, producing a 120-kD cleavage product [26, 27]. The antibody against ·-fodrin stained several bands, including the intact 240-kD protein, and the 145/150- and 120-kD fodrin breakdown products (FBDPs) (fig. 5a). All these bands were detectable in all samples, even the controls, indicating the high turnover rate in the rapidly growing, immature brain [28]. The intact 240-kD ·-fodrin displayed a biphasic decrease (to about one third of controls) in the ipsilateral hemisphere, one 3 h and one 24 h after HI (fig. 5b). There was a corresponding biphasic increase in the ratio between the calpain-induced (145 + 150 kD) and the caspase-3-induced (120 kD) FBDPs and the intact 240-kD fodrin (fig. 5c). However, the calpaininduced cleavage displayed its major peak earlier (3–8 h after HI), whereas the caspase-3-induced cleavage was more prominent later (24 h after HI) (fig. 5b), in accordance with our earlier findings [28, 29, unpubl. observations]. The activation of calpain was further confirmed by using an antibody specific to a 150-kD calpain-induced FBDP [25] (fig. 5a). The individual samples analyzed 24 h after HI revealed a significant decrease in the ratio of the 240-kD fodrin, as well as significant increases in both the 145/150- and the 120-kD FBDPs (table 1).

Discussion

Fig. 3. a A PARP immunoblot of pooled samples of parietal cortex (as in fig. 2) demonstrating substantial cleavage 24 h after HI of the 116-kD intact protein, concomitant with the appearance of the 85kD fragment. C and I indicate the contra- and ipsilateral hemispheres, respectively. P = Postnatal day; d = day. b Results from densitometric quantification of the bands detected in the ipsilateral hemispheres only, after PARP immunoblotting of pooled samples. The values (OD per Ìg of protein) for the 116-kD intact PARP and the 85-kD cleavage product are indicated. P = Postnatal day; d = day. c The relative ratio between the 85-kD fragment and the 116-kD intact PARP in the ipsilateral hemispheres, further demonstrating the sharp peak of PARP cleavage 24 h after HI. P = Postnatal day; d = day.

Inhibitor of Caspase-Activated DNase. ICAD, or DNA fragmentation factor 45 (DFF45), is a caspase-3 substrate that must be cleaved for internucleosomal fragmentation to proceed [21, 23]. ICAD/ DFF45 is both a specific molecular chaperone mediating the correct folding of CAD/DFF40 as well as an inhibitor of CAD/DFF40 when complexed with the DNase. There are two forms of ICAD, a long

176

Biol Neonate 2001;79:172–179

Caspases The caspase-3-like activity in the control animals displayed a significant and rapid decrease between postnatal day 10 and 13 (fig. 1, 2a–c), a time period when the brain growth spurt is leveling out. This is in accordance with earlier findings [16, 17], indicating that the apoptotic machinery is downregulated when neurogenesis and synaptogenesis are finished and the neurons are terminally differentiated. The DEVDase activity displayed a 25-fold increase 24 h after HI, and this extensive activation coincided with considerable degradation of the known caspase-3 substrates ICAD, PARP and fodrin. Furthermore, the peak in DEVD cleavage also coincided with the depletion of the caspase-3 32-kD proform and the appearance of 29and 17-kD cleaved forms in the ipsilateral hemisphere. This massive activation of caspase-3 occurred quite late after the insult, indicating that there may be a substantial time window for the initiation of pharmacological inhibition, provided that the changes induced earlier are not irreversible. Cheng et al. [12] achieved significant neuroprotection using boc-Asp-fmk (a broad-spectrum caspase inhibitor) in the same model of HI in 7-day-old rats, even when boc-Asp-fmk was administered as late as 3 h after HI. Neuroprotection using other broad-spectrum caspase inhibitors (ZVAD or DEVD) has also been demonstrated in models of adult ischemia [14, 15, 30].

Wang/Karlsson/Zhu/Bahr/Hagberg/ Blomgren

Fodrin The biphasic degradation of fodrin indicated that calpains were mainly responsible for the early peak, around 3 h after HI, whereas caspase-3 contributes to a large extent to the peak around 24 h after HI. It has been demonstrated that both calpain and caspase-3 can produce cleavage products of fodrin with an apparent molecular weight of 150 kD [31–34]. The early, major 150-kD FBDP peak in this model can be attributed mainly, if not solely, to calpain, because there was little DEVDase activity and hardly any detectable degradation of the other caspase-3 substrates (PARP and ICAD) at this time point. Furthermore, the antibody specific to the calpain-cleaved 150kD FBDP [25, 29] closely followed the 150-kD band as it appeared using the other antibody staining both intact and cleaved fodrin. Interestingly, when we used the antibody specific for calpain-cleaved FBDP, the biphasic pattern was less obvious [28]. The biphasic pattern became clear only when the ratio between intact and degraded fodrin (fig. 5c) or the degradation of calpastatin [19] was studied. The biphasic (or multiphasic) activation of calpain has been demonstrated earlier [19, 35, 36]. The temporal pattern is different in different species [37]. The early peak(s) are considered to reflect calcium influx and concurrent calpain activation, and the subsequent peak(s) reflect cellular degradation and demise. Poly(ADP-Ribose) Polymerase The DNA repair enzyme PARP was one of the first caspase-3 substrates to be identified [38, 39]. The upregulation of PARP following HI is in accordance with increased poly(ADP-ribose) production after focal [40] and global [41] ischemia. However, we found an upregulation also in the contralateral, nonischemic, undamaged hemisphere. It seems reasonable that a DNA repair enzyme should be targeted during apoptosis, rendering the cells unable to maintain their DNA integrity, but it has also been postulated that PARP activ-

Fig. 4. a An ICAD immunoblot of pooled samples of parietal cortex (as in fig. 3) demonstrating substantial cleavage 24 h after HI of the long (45 + 42 kD) and the short (35 + 32 kD) forms. C and I indicate the contra- and ipsilateral hemispheres, respectively. P = Postnatal day; d = day. b Results from densitometric quantification of the bands detected in the ipsilateral hemispheres only, after ICAD immunoblotting of pooled samples, demonstrating the results for the long form (ICAD-L, 45 + 42 kD). The values (OD per Ìg of protein) for the contralateral and the ipsilateral hemispheres are indicated. The relative upregulation after hypoxia (contralateral) and HI (ipsilateral) is demonstrated (8 h to 6 days after HI). P = Postnatal day; d = day. c The relative ratio of ICAD-L immunoreactivity between the ipsilateral and the contralateral hemispheres, demonstrating the relative depletion of ICAD-L 24 h after HI. The comparison with the contralateral hemisphere was made because no specific cleavage product of ICAD was detected using this antibody. P = Postnatal day; d = day. d Results from densitometric quantification as in b, demonstrating the results for the short form (ICAD-S, 35 + 32 kD). The values (OD per Ìg of protein) for the contralateral and the ipsilateral hemispheres are indicated. P = Postnatal day; d = day. e The relative ratio of ICAD-S immunoreactivity between the ipsilateral and the contralateral hemispheres, demonstrating the relative depletion also of ICAD-S 24 h after HI. P = Postnatal day; d = day.

Caspase-3 in Neonatal Cerebral Hypoxia-Ischemia

Biol Neonate 2001;79:172–179

177

ity contributes to the damage after HI by consuming ATP. This is supported by the findings that PARP inhibitors provide neuroprotection [30, 40, 42], and that PARP-deficient mice show reduced susceptibility to ischemic injury [30]. Inhibitor of Caspase-Activated DNase/DNA Fragmentation Factor 45 Cleavage of ICAD/DFF45 by caspase-3 [21, 23, 43, 44] releases CAD/DFF40, which in turn degrades chromosomal DNA into nucleosomal fragments. ICAD, like PARP, also appeared to be upregulated after the insult, concomitant with extensive degradation in the ipsilateral hemisphere 24 h after HI. Recently, Sakahira et al. [45] reported that ICAD-L, but not ICAD-S, supported the production of functional CAD, suggesting a chaperone-like activity for ICAD-L but not ICAD-S. It seems, however, that the short form still retains inhibitory activity towards CAD/DFF40 [46]. Mouse CAD and ICAD-L are apparently the counterparts of human DFF40 and DFF45, respectively, whereas the human counterpart of mouse ICAD-S has not been identified. This is the first report to our knowledge demonstrating the presence and cleavage of the equivalent to ICAD-S in the rat. Previously, we have demonstrated that active caspase-3 and markers of DNA damage were colocalized as early as 3 h after HI [29].

Summary

In summary, we found evidence for extensive activation of caspase-3 after HI in the immature brain, leading to disruption of the cytoskeletal protein fodrin, inactivation of the DNA repair enzyme PARP and cleavage of ICAD which would lead to the activation of the DNase CAD. These downstream consequences of caspase-3 may contribute to the development of brain injury.

Acknowledgements

Fig. 5. a A fodrin immunoblot of pooled samples of parietal cortex (as in fig. 3) demonstrating substantial cleavage 24 h after HI of the 240-kD intact protein, concomitant with the appearance of the 145/ 150- and the 120-kD fragments. C and I indicate the contra- and ipsilateral hemispheres, respectively. P = Postnatal day; d = day. b Results from densitometric quantification of the bands detected in the ipsilateral hemispheres only, after fodrin immunoblotting of pooled samples. The values (OD per Ìg of protein) for the 240-kD intact fodrin and for the 150- and 120-kD breakdown products are indicated. A biphasic decrease in the intact fodrin is seen, plus an early (3–8 h) peak for the calpain-induced FBDP (145 + 150 kD) and a later peak (24 h) for the caspase-3-induced FBDP (120 kD). P = Postnatal day; d = day. c The relative ratio between the 145 + 150- or the 120-kD FBDP and the intact fodrin in the ipsilateral hemispheres, more clearly demonstrating the biphasic pattern of degradation, with a predominance for calpain cleavage early (3 h) and caspase-3 cleavage later (24 h) after HI.

178

Biol Neonate 2001;79:172–179

We are very grateful to Dr. Xiaodong Wang, University of Texas Southwestern Medical Center, for supplying the antibody against ICAD/DFF45. This work was supported by the Swedish Medical Research Council (13238, 12213, 9455, 5932), the Åhlén Foundation, the Sven Jerring Foundation, the Magnus Bergvall Foundation, the Wilhelm and Martina Lundgren Science Foundation, the Linnéa and Josef Carlsson Foundation, the Tore Nilson Foundation for Medical Research, the United States Army Medical Research and Material Command, and the Frimurare Barnhus Foundation.

Wang/Karlsson/Zhu/Bahr/Hagberg/ Blomgren

References 1 Lee JM, Zipfel GJ, Choi DW: The changing landscape of ischaemic brain injury mechanisms. Nature 1999;399(6738 suppl):A7–A14. 2 Nicholson DW, Thornberry NA: Caspases: Killer proteases. Trends Biochem Sci 1997;22:299–306. 3 Alnemri ES, Livingston DJ, Nicholson DW, Salvesen G, Thornberry NA, Wong WW, Yuan J: Human ICE/CED-3 protease nomenclature. Cell 1996;87:171. 4 Cohen GM: Caspases: The executioners of apoptosis. Biochem J 1997;326:1–16. 5 Stennicke HR, Salvesen GS: Caspases – controlling intracellular signals by protease zymogen activation. Biochim Biophys Acta 2000;1477:299–306. 6 Kuida K, Zheng TS, Na S, Kuan C, Yang D, Karasuyama H, Rakic P, Flavell RA: Decreased apoptosis in the brain and premature lethality in CPP32deficient mice. Nature 1996;384:368–372. 7 Linnik MD, Zobrist RH, Hatfield MD: Evidence supporting a role for programmed cell death in focal cerebral ischemia in rats. Stroke 1993;24: 2002–2008. 8 MacManus JP, Buchan AM, Hill IE, Rasquinha I, Preston E: Global ischemia can cause DNA fragmentation indicative of apoptosis in rat brain. Neurosci Lett 1993;164:89–92. 9 Namura S, Zhu J, Fink K, Endres M, Srinivasan A, Tomaselli KJ, Yuan J, Moskowitz MA: Activation and cleavage of caspase-3 in apoptosis induced by experimental cerebral ischemia. J Neurosci 1998; 18:3659–3668. 10 Chen J, Nagayama T, Jin K, Stetler RA, Zhu RL, Graham SH, Simon RP: Induction of caspase-3like protease may mediate delayed neuronal death in the hippocampus after transient cerebral ischemia. J Neurosci 1998;18:4914–4928. 11 Pulera MR, Adams LM, Liu H, Santos DG, Nishimura RN, Yang F, Cole GM, Wasterlain CG: Apoptosis in a neonatal rat model of cerebral hypoxia-ischemia. Stroke 1998;29:2622–2630. 12 Cheng Y, Deshmukh M, DaCosta A, Demaro JA, Gidday JM, Shah A, Sun Y, Jacquin MF, Johnson EM, Holtzman DM: Caspase inhibitor affords neuroprotection with delayed administration in a rat model of neonatal hypoxic-ischemic brain injury. J Clin Invest 1998;101:1992–1999. 13 Endres M, Namura S, Shimizu-Sasamata M, Waeber C, Zhang L, Gomez-Isla T, Hyman BT, Moskowitz MA: Attenuation of delayed neuronal death after mild focal ischemia in mice by inhibition of the caspase family. J Cereb Blood Flow Metab 1998;18:238–247. 14 Ma J, Endres M, Moskowitz MA: Synergistic effects of caspase inhibitors and MK-801 in brain injury after transient focal cerebral ischaemia in mice. Br J Pharmacol 1998;124:756–762. 15 Hara H, Friedlander RM, Gagliardini V, Ayata C, Fink K, Huang Z, Shimizu-Sasamata M, Yuan J, Moskowitz MA: Inhibition of interleukin 1beta converting enzyme family proteases reduces ischemic and excitotoxic neuronal damage. Proc Natl Acad Sci USA 1997;94:2007–2012. 16 Ni B, Wu X, Su Y, Stephenson D, Smalstig EB, Clemens J, Paul SM: Transient global forebrain ischemia induces a prolonged expression of the caspase-3 mRNA in rat hippocampal CA1 pyramidal neurons. J Cereb Blood Flow Metab 1998;18:248– 256.

Caspase-3 in Neonatal Cerebral Hypoxia-Ischemia

17 Hu BR, Liu CL, Ouyang Y, Blomgren K, Siesjo BK: Involvement of caspase-3 in cell death after hypoxia-ischemia declines during brain maturation. J Cereb Blood Flow Metab 2000;20:1294– 1300. 18 Rice J, Vannucci R, Brierley J: The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann Neurol 1981;9:131–141. 19 Blomgren K, Hallin U, Andersson AL, Puka-Sundvall M, Bahr BA, McRae A, Saido TC, Kawashima S, Hagberg H: Calpastatin is up-regulated in response to hypoxia and is a suicide substrate to calpain after neonatal cerebral hypoxia-ischemia. J Biol Chem 1999;274:14046–14052. 20 Karlsson J, Ostwald K, Kåbjorn C, Andersson M: A method for protein assay in Laemmli buffer. Anal Biochem 1994;219:144–146. 21 Liu X, Zou H, Slaughter C, Wang X: DFF, a heterodimeric protein that functions downstream of caspase-3 to trigger DNA fragmentation during apoptosis. Cell 1997;89:175–184. 22 Lazebnik YA, Kaufmann SH, Desnoyers S, Poirier GG, Earnshaw WC: Cleavage of poly(ADP-ribose) polymerase by a proteinase with properties like ICE. Nature 1994;371:346–347. 23 Enari M, Sakahira H, Yokoyama H, Okawa K, Iwamatsu A, Nagata S: A caspase-activated DNase that degrades DNA during apoptosis, and its inhibitor ICAD. Nature 1998;391:43–50. 24 Levine J, Willard M: Fodrin: Axonally transported polypeptides associated with the internal periphery of many cells. Cell Biol 1981;90:631–642. 25 Bahr B, Tiriveedhi S, Park G, Lynch G: Induction of calpain-mediated spectrin fragments by pathogenic treatments in long-term hippocampal slices. J Pharmacol Exp Ther 1995;273:902–908. 26 Pörn-Ares MI, Samali A, Orrenius S: Cleavage of the calpain inhibitor, calpastatin, during apoptosis. Cell Death Differ 1998;5:1028–1033. 27 Wang KK, Posmantur R, Nadimpalli R, Nath R, Mohan P, Nixon RA, Talanian RV, Keegan M, Herzog L, Allen H: Caspase-mediated fragmentation of calpain inhibitor protein calpastatin during apoptosis. Arch Biochem Biophys 1998;356:187– 196. 28 Blomgren K, Kawashima S, Saido T, Karlsson J, Elmered A, Hagberg H: Fodrin degradation and subcellular distribution of calpains after neonatal rat cerebral hypoxic-ischemia. Brain Res 1995; 684:143–149. 29 Zhu C, Wang X, Hagberg H, Blomgren K: Correlation between caspase-3 activation and three different markers of DNA damage in neonatal cerebral hypoxia-ischemia. J Neurochem 2000;75:819– 829. 30 Endres M, Wang ZQ, Namura S, Waeber C, Moskowitz MA: Ischemic brain injury is mediated by the activation of poly(ADP-ribose)polymerase. J Cereb Blood Flow Metab 1997;17:1143–1151. 31 Jänicke R, Ng P, Sprengart M, Porter A: Caspase-3 is required for alpha-fodrin cleavage but dispensable for cleavage of other death substrates in apoptosis. J Biol Chem 1998;273:15540–15545.

32 Nath R, Raser K, Stafford D, Hajimohammadreza I, Posner A, Allen H, Talanian R, Yuen P, Gilbertsen R, Wang K: Non-erythroid alpha-spectrin breakdown by calpain and interleukin 1 beta-converting-enzyme-like protease(s) in apoptotic cells: Contributory roles of both protease families in neuronal apoptosis. Biochem J 1996;319:683–690. 33 Vanags D, Pörn-Ares M, Coppola S, Burgess D, Orrenius S: Protease involvement in fodrin cleavage and phosphatidylserine exposure in apoptosis. J Biol Chem 1996;271:31075–31085. 34 Wang K, Posmantur R, Nath R, McGinnis K, Whitton M, Talanian R, Galntz S, Morrow J: Simultaneous degradation of alphaII- and betaIIspectrin by caspase 3 (CPP32) in apoptotic cells. J Biol Chem 1998;273:22490–22497. 35 Seubert P, Lee K, Lynch G: Ischemia triggers NMDA receptor-linked cytoskeletal proteolysis in hippocampus. Brain Res 1989;492:366–370. 36 Saido T, Yokota M, Nagao S, Yamaura I, Tani E, Tsuchiya T, Suzuki K, Kawashima S: Spatial resolution of fodrin proteolysis in postischemic brain. J Biol Chem 1993;268:25239–25243. 37 Kitagawa K, Matsumoto M, Saido TC, Ohtsuki T, Kuwabara K, Yagita Y, Mabuchi T, Yanagihara T, Hori M: Species differences in fodrin proteolysis in the ischemic brain. J Neurosci Res 1999;55:643– 649. 38 Tewari M, Quan LT, O’Rourke K, Desnoyers S, Zeng Z, Beidler DR, Poirier GG, Salvesen GS, Dixit VM: Yama/CPP32 beta, a mammalian homolog of CED-3, is a CrmA-inhibitable protease that cleaves the death substrate poly(ADP-ribose) polymerase. Cell 1995;81:801–809. 39 Nicholson DW, Ali A, Thornberry NA, Vaillancourt JP, Ding CK, Gallant M, Gareau Y, Griffin PR, Labelle M, Lazebnik YA: Identification and inhibition of the ICE/CED-3 protease necessary for mammalian apoptosis. Nature 1995;376:37–43. 40 Tokime T, Nozaki K, Sugino T, Kikuchi H, Hashimoto N, Ueda K: Enhanced poly(ADP-ribosyl)ation after focal ischemia in rat brain. J Cereb Blood Flow Metab 1998;18:991–997. 41 Love S, Barber R, Wilcock GK: Neuronal accumulation of poly(ADP-ribose) after brain ischaemia. Neuropathol Appl Neurobiol 1999;25:98–103. 42 Takahashi K, Greenberg JH, Jackson P, Maclin K, Zhang J: Neuroprotective effects of inhibiting poly(ADP-ribose) synthetase on focal cerebral ischemia in rats. J Cereb Blood Flow Metab 1997;17: 1137–1142. 43 Sakahira H, Enari M, Nagata S: Cleavage of CAD inhibitor in CAD activation and DNA degradation during apoptosis. Nature 1998;391:96–99. 44 Liu X, Zou H, Widlak P, Garrard W, Wang X: Activation of the apoptotic endonuclease DFF40 (caspase-activated DNase or nuclease). Oligomerization and direct interaction with histone H1. J Biol Chem 1999;274:13836–13840. 45 Sakahira H, Enari M, Nagata S: Functional differences of two forms of the inhibitor of caspase-activated DNase, ICAD-L, and ICAD-S. J Biol Chem 1999;274:15740–15744. 46 Gu J, Dong RP, Zhang C, McLaughlin DF, Wu MX, Schlossman SF: Functional interaction of DFF35 and DFF45 with caspase-activated DNA fragmentation nuclease DFF40. J Biol Chem 1999; 274:20759–20762.

Biol Neonate 2001;79:172–179

179

Biol Neonate 2001;79:180–186

Free Radicals and Brain Damage in the Newborn Giuseppe Buonocore

Serafina Perrone Rodolfo Bracci

Institute of Preventive Pediatrics and Neonatology, University of Siena, Siena, Italy

Key Words Newborn W Brain damage W Free radical injury

Abstract Newborns and particularly preterm infants are at high risk of oxidative stress and they are very susceptible to free radical oxidative damage. Indeed, there is evidence of an imbalance between antioxidant- and oxidant-generating systems which causes oxidative damage. The brain may be especially at risk of free radical-mediated injury because neuronal membranes are rich in polyunsaturated fatty acids and because the human newborn has a relative deficiency of brain superoxide dismutase and glutathione peroxidase. The brain of the term fetus is at higher risk of oxidative stress than that of the preterm fetus, as a consequence of its higher concentration of polyunsaturated fatty acids and the maturity of the N-methyl-D-aspartate receptor system at term. There seems to be a maturation-dependent window of vulnerability to free radical attack during oligodendrocyte development. Early in its differentiation, the oligodendrocyte may be vulnerable because of active acquisition of iron for differentiation at a time of relative delay in the development of certain key antioxidant defenses in the brain. Excess free iron and deficient iron-binding and -metabolizing capacity are additional features favoring oxidant stress in premature infants. Free radicals may be generated by different mechanisms, such as ischemia-reperfusion, neutrophil and macrophage activation, Fenton chemistry, endothelial cell xanthine oxidase, free fatty acid and prostaglandin metabolism and hypoxia. Reactive oxidant production by these different mechanisms contributes in a piecewise manner to the pathogenesis of perinatal brain injury. Copyright © 2001 S. Karger AG, Basel

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0180$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

Introduction

Free radicals are highly reactive chemical molecules containing one or more unpaired electrons. They donate or take electrons from other molecules in an attempt to pair their electrons and generate a more stable species. Oxygen-derived free radicals, collectively termed reactive oxygen species (ROS), are normally produced in living organisms. When overproduced, they are important mediators of cell and tissue injury [1, 2]. Free radicals are relatively unstable and certain enzymes and small-molecular-weight molecules with antioxidant capabilities have a protective effect [3]. There is therefore a critical balance between free radical generation and antioxidant defenses. Oxidative stress in vivo is a degenerative process caused by the overproduction and propagation of free radical reactions. Free radical reactions lead to the oxidation of lipids, proteins and polysaccharides and to DNA damage (fragmentation, apoptosis, base modifications and strand breaks), and therefore have a wide range of biologically toxic effects [4, 5]. Intracellular and extracellular antioxidant systems protect against free radical-induced damage. Transferrin (Tf), ceruloplasmin, vitamin C, vitamin E, uric acid, bilirubin, sulfhydryl groups and other unidentified antioxidants contribute to the total antioxidant capacity of extracellular fluids [6]. Oxidative stress exists and tissue damage is possible when there are low levels of antioxidants or increased free radical activity [1]. Newborns and particularly preterm infants are at high risk of oxidative stress and they are very susceptible to free radical oxidative damage [4]. Indeed, there is evidence of an imbalance between antioxidant- and oxidant-generating systems which causes oxidative damage [7]. At birth, the newborn encounters an environment much richer in oxygen (PO2 100 Torr) than the intrauterine environment (20– 25 Torr). This 4- to 5-fold increase is exacerbated by the low efficiency of natural antioxidant systems in the newborn, especially the preterm newborn [8]. Neonatal plasma has an antioxidant profile with low levels of glutathione peroxidase, superoxide dismutase, ß-carotene, riboflavin, ·-proteinase, vitamin E, selenium, copper, zinc, ceruloplasmin, Tf and other plasma factors [9–11]. The brain may be

Prof. Giuseppe Buonocore Institute of Preventive Pediatrics and Neonatology University of Siena, viale Bracci, 36 I–53100 Siena (Italy) Tel. +39 0577 586 523, Fax +39 0577 586 182, E-Mail [email protected]

especially at risk of free radical-mediated injury, because neuronal membranes are rich in polyunsaturated fatty acids and because the human newborn, especially if preterm, has a relative deficiency of brain superoxide dismutase and glutathione peroxidase [12]. The brain of the term fetus is at higher risk of oxidative stress than that of the preterm fetus as a consequence of its higher concentration of polyunsaturated fatty acids and the maturity of the N-methyl-Daspartate receptor system at term [13, 14]. There seems to be a maturation-dependent window of vulnerability to free radical attack during oligodendrocyte development [15]. Early in its differentiation, the oligodendrocyte may be vulnerable because of active acquisition of iron for differentiation at a time of relative delay in the development of certain key antioxidant defenses in the brain [16, 17]. Excess free iron and deficient iron-binding and -metabolizing capacity are additional features favoring oxidant stress in premature infants [18, 19]. Free radicals may be generated by different mechanisms, such as ischemia-reperfusion, neutrophil and macrophage activation, Fenton chemistry, endothelial cell xanthine oxidase, free fatty acid and prostaglandin metabolism and hypoxia [20–24].

Hypoxia-Asphyxia

Hypoxia-asphyxia plays a key role in the perinatal period. Although the consequences of hypoxia-asphyxia can be easily observed, the specific pathologic processes preceding the onset of irreversible cerebral damage are not well understood and appear to be a combination of several complex mechanisms [25]. Early events in the hypoxia-induced response trigger tyrosine phosphorylation cascades involving many enzymes and substrates. Studies performed in guinea pig cerebral cortical synaptosomes suggest that hypoxia remodels the signaling pathway by inducing quantitative and qualitative changes in protein phosphorylation [26]. Many experimental studies have demonstrated free radical production and oxidative damage due to hypoxia in fetal life [27–29]. We recently demonstrated a direct relation between the degree of hypoxia and the severity of oxidative damage in plasma of newborn infants at birth [30]. In the developing brain, hypoxia results in an increase in anaerobic metabolism, leading to a rapid rise in levels of lactic acid and oxygen free radicals [31, 32]. Mitochondrial oxidative metabolism, nitric oxide (NO), phospholipid metabolism, iron, proteolytic and inflammatory pathways are other potential sources of intracellular free radicals and ROS [33–35]. Free radicals may cause lipid peroxidation of immature myelin sheaths and lipid peroxides are themselves free radicals [36].

such as SP-22 and mitochondrial respiration itself [39–41]. Superoxide radicals produced by the respiratory chain are readily dismutated by mitochondrial superoxide dismutase, producing H2O2 [42]. Under conditions in which mitochondrial superoxide generation increases, or when antioxidant systems are depleted, H2O2 may accumulate, leading to a condition of mitochondrial oxidative stress. In this situation, H2O2 may react with mitochondrial Fe2+, resulting in the formation of the highly reactive hydroxyl radical (HO) via the Fenton reaction [43]. The basal rate of mitochondrial superoxide generation may be altered by different physiological or pathological conditions [37]. Specifically, when mitochondria are loaded with Ca2+, uncoupling of mitochondrial respiration from ADP phosphorylation increases mitochondrial production of ROS [44]. Thus, accumulation of ROS and oxidative stress is initiated early during hypoxia-ischemia due to the dramatic increase in cytosolic Ca2+ concentrations, which upsets mitochondrial handling of Ca2+. Normal Ca cycling across the inner mitochondrial membrane serves to regulate mitochondrial enzymes such as pyruvate dehydrogenase and ·-oxyglutarate dehydrogenase [35]. However, when intracellular Ca increases over the set point for net calcium accumulation or when the calcium release pathway is stimulated by prooxidants, cycling may become excessive and lead to increased ROS production, loss of mitochondrial membrane potential, structural alterations of the inner mitochondrial membrane and inhibition of adenosine 5)-triphosphate (ATP) synthesis [35]. ROS generation increases due to disorganization of the mitochondrial respiratory chain, since most components of this system are integral inner mitochondrial membrane proteins. Mitochondria are particularly sensitive to hypoxic injury and play a central role in both apoptosis and necrosis [45]. Severe mitochondrial damage releases a flood of Ca2+ and ROS into the cytosol, leading to the disruption of plasma membrane integrity and cell damage [35]. They may also directly activate caspase-9 and trigger apoptotic cell death [46].

Role of Energy Metabolism and Calcium

ROS are normally generated in a continuous manner by the mitochondrial respiratory chain [33, 37]. In normal mitochondria, oxygen is reduced to water by cytochrome C oxidase in four consecutive one– electron steps. Thus, the production of superoxide radicals (O 2W ) occurs during the operation of complex I and complex II in the mitochondrial electron transport chain, and at least at the level of coen– zyme Q as a result of the semiquinone state (UQ W ) of ubiquinone donating electrons to molecular oxygen [38]. Mitochondria have an efficient antioxidant system composed of superoxide dismutase, glutathione peroxidase, glutathione reductase, glutathione, NAD(P) transhydrogenase, NADPH, vitamins E and C, thiol peroxidases

The normal functioning of the brain is essentially dependent on an adequate oxygen supply to maintain energy metabolism. At the cell level, cerebral hypoxia-ischemia sets in motion a cascade of biochemical events commencing with a shift from oxidative to anaerobic metabolism, which leads to an accumulation of NADH, FADH and lactic acid and H+ ions [47] (fig. 1). Anaerobic glycolysis does not provide sufficient energy, resulting in the depletion of high-energy phosphate reserves, including ATP [48]. During moderate hypoxemia, the fetus maintains adequate levels of ATP by speeding up the rate of anaerobic glycolysis, but an acute reduction in the fetal oxygen supply leads to the breakdown of energy metabolism in a few minutes [49–51]. The Na+/K+ pump stops working through lack of energy. The transcellular ion pump fails, leading to loss of membrane potential and an influx of Na+, Ca+ and Cl – . Intracellular accumulation of Na+ and Cl – ions leads to swelling of the cells as water enters by osmosis (cytotoxic cell edema) [52]. Calcium buildup in the cytoplasm occurs by other mechanisms besides massive influx due to the extreme extraintracellular concentration gradient. Calcium enters the cytosol by the activation of voltage-dependent channels [25]. Calcium is released by mitochondria stimulated by the increase in intracellular Na+ and free fatty acids. It is also released by the endoplasmic reticulum through the depletion of ATP [53]. It enters through agonist-dependent channels such as amino-hydroxyl-methyl-isoxa-

Neonatal Oxidative Brain Injury

Biol Neonate 2001;79:180–186

Mitochondrial Production of Free Radicals

181

Fig. 1. Cellular mechanisms of free radical production during hypoxia. PL = Phospholipase; IP3 = inositol triphosphate; XDH = xanthine dehydrogenase; XO = xanthine oxidase; FFA = free fatty acids; FR = free radicals.

zole propionate, kainate and N-methyl-D-aspartate receptors [54]. The intracellular buildup of calcium has many consequences. One damaging effect is the activation of phospholipases A2 and C [55]. These reactions lead to membrane phospholipid hydrolysis, producing free radicals, disrupting cell and organelle membranes, increasing permeability and altering ionic distribution. Phospholipase C also catalyzes reactions leading to the production of inositol triphosphate, a second messenger that releases calcium from the endoplasmic reticulum, and diacylglycerol, which decreases calcium-sodium exchange [53, 56]. Both reactions further augment calcium concentrations in the cell and amplify its deleterious effects, creating a vicious circle that ultimately destroys the cell.

Free Radicals during the Reperfusion Phase

During cerebral ischemia, the cutback in oxidative phosphorylation rapidly diminishes reserves of high-energy phosphates [48]. High levels of adenosine and hypoxanthine accumulate in a few minutes. During reperfusion, these metabolic products are further metabolized by xanthine oxidase to xanthine and uric acid [57], resulting in their build up in blood and tissues such as the brain [58]. The activity of xanthine oxidase in the resting brain is very low [59], but during cerebral ischemia, a massive conversion of xanthine dehydrogenase to xanthine oxidase takes place, regulated by the calciumdependent protease calpain [60]. The breakdown of hypoxanthine by

182

Biol Neonate 2001;79:180–186

xanthine oxidase in the presence of oxygen produces a flood of superoxide radicals [61, 62]. These are then converted to hydrogen peroxide by superoxide dismutase. Hydrogen peroxide and tissue iron can then combine to form hydroxyl radicals by the Haber-Weiss reaction [63]. Xanthine oxidase concentrates in endothelial cells lining the cerebral microvasculature, targeting the blood-brain barrier for oxidative attack [64]. Accelerated arachidonic acid metabolism in brain tissue and leukocyte activation after ischemia also produce large quantities of oxygen radicals [21]. Although there is more evidence that blood vessels are the main source of free radicals in cerebral ischemia-reperfusion, neurons also generate superoxide in response to anoxia by activated neutrophils and microglia [65]. Free radicals impair transmembrane enzyme Na+/K+-ATPase activity, especially in cortical synaptosomal membranes, resulting in persistent membrane depolarization and excessive release of the excitatory amino acid glutamate [29, 66]. Cerebellar granule cells produce superoxide when exposed to the excitatory amino acid N-methyl-D-aspartate [67]. Besides being neurotoxic, glutamate is also toxic to oligodendroglia, via free radical effects [68]. Glutamate enters the cell in exchange for cystine. Intracellular cystine depletion is followed by a drop in glutathione levels and the cells die of oxidative stress [68]. Studies have shown that free radicals are also implicated in initiating delayed cell death by apoptosis after cerebral ischemia [69, 70]. An enormous amount of data suggests that oxidative stress plays a role in the initiation of apoptosis [70–73], but it is not yet clear

Buonocore/Perrone/Bracci

exactly how ROS trigger this response. Like Ca2+, oxidative stress may either promote or inhibit apoptosis, depending on the degree of the insult [74]. Kim et al. [75] demonstrated that hydrogen peroxide and NO directly inhibit caspase activity in vivo. The close relationship between oxidative stress and mitochondrial function also suggests that overproduction of ROS leads to ATP depletion and apoptosis [76]. Oxidative stress may also disrupt intracellular Ca2+ homeostasis, inhibiting apoptosis [77].

Iron is a versatile and highly reactive element. By virtue of its two common valences, iron (II) (ferrous) and iron (III) (ferric), it has access to a wide range of redox potentials spanning the standard redox potential range from B300 to –500 mV [91]. This property underlies its essential biological role in oxygen transport and many electron transfer reactions [92]. Normally, iron is safely sequestered in transport proteins such as Tf and lactoferrin and stored in proteins such as ferritin (Ft) and haemosiderin [93]. The plasma copper-containing protein ceruloplasmin acts in concert with the above proteins, catalyzing the oxidation of reactive ferrous ions to less reactive ferric ions which bind Tf [94, 95]. In healthy adults, plasma Tf is approximately one third loaded with iron, and the protein retains a considerable ability to

bind iron salts. Tf can bind 2 mol of iron per mole of protein, and when iron is correctly loaded on its high-affinity binding sites, it is not available as a growth factor for tissues, or as a prooxidant factor [96, 97]. Since iron ions cannot exist in plasma, the term ‘free iron’ has been introduced to indicate a low-molecular-mass iron form, free of high-affinity binding to Tf [96]. Free iron seems to occur in plasma, complexed to citrate, lactate or phosphate or loosely bound to albumin or other proteins [98]. During situations of iron overload and low plasma pH, as occurs during ischemia, Tf releases its iron and chelatable forms of Fe (iron ions or redox-active complexes of iron) escape sequestration in biological systems, producing free radicals [99, 100]. These free radicals may release even more iron by mobilizing it from ferritin. This may lead to a cascade of iron release and free radical production, causing extensive cell damage [101]. We recently observed higher intraerythrocyte free iron levels in infants with asphyxia [102]. Iron may be released from hemoglobin in erythrocytes as result of oxidative stress [103]. Since the erythrocyte is a target of extracellular free radicals, free iron release may follow extracellular oxidative stress caused by superoxide anion release due to phagocyte activation [104]. Intraerythrocyte free iron concentrations appear to be a reliable marker of red cell oxidative stress and an indicator of the risk of oxidative injury in other tissues. Indeed, free radicals are linked to neonatal oxidative stress and are involved in severe diseases such as retinopathy, bronchopulmonary dysplasia, intraventricular hemorrhage and hypoxic-ischemic encephalopathy [4, 105]. In these oxidative stress-related pathologies, iron is released from iron stores and may cause cell damage by lipid and protein peroxidation. The highest values of lipid and protein peroxidation have been found in hypoxic newborns. The more severe the hypoxia, the higher the intraerythrocyte free iron release, free radical production and oxidative damage [30, 106]. The newborn infant is very susceptible to free iron-induced oxidative damage [107]. Plasma from a high percentage of normal term and preterm neonates has recently been shown to contain free iron, as if Tf were fully loaded with iron [108–110]. The iron-binding capacity of cerebrospinal fluid was also found to be low (low Tf concentrations), and high concentrations of vitamin C and low concentrations of ceruloplasmin in cerebrospinal fluid suggested that most of the iron was in its active ferrous form [111, 112]. Since iron-positive reactive glia collect near damaged tissue, iron accumulation is a sensitive indicator of injury [113]. In damaged areas, there is an increase in iron-positive reactive glia starting about 8 h after recovery and an earlier (4–8 h after hypoxia-ischemia) increase in microglia, especially around cortical blood vessels [114]. The perivascular distribution of iron reaction products is a consistent finding 1 week after recovery. The consequences of selective vascular injury include secondary ischemia and blood-brain barrier disruption. After asphyxia in newborn infants, there is an increase in intraerythrocyte and plasma free iron, significantly correlated with neurodevelopmental outcome [109]. Leakage of plasma free iron into the brain through a damaged barrier may occur and is particularly damaging, as it is taken up directly by cells in a manner that is independent of Tf. Additional sources of free iron could be iron released by heme catabolism and iron released from storage protein by oxidative stress. During ischemia-reperfusion, ROS are generated by mitochondrial dysfunction, exocytotoxic insult, metabolism of arachidonic acid, inflammation and stimulation of NOS and xanthine oxidase [115]. Oxidative stress may also result from iron delocalization induced by the superoxide anion, acidosis and anoxia [102, 106].

Neonatal Oxidative Brain Injury

Biol Neonate 2001;79:180–186

Nitric Oxide

NO is a free radical synthesized by NO synthase (NOS) in endothelial cells and neurons in response to rises in intracellular calcium concentrations [78]. NOS produces NO, citrulline and water from arginine, NADPH and oxygen [79]. NO and superoxide radicals combine to produce peroxynitrite, which spontaneously decomposes to form hydroxyl radicals, nitrogen dioxide and NO+2 [80]. Three types of NOS are known: neuronal NOS (NOS 1), inducible NOS (NOS 2) and endothelial NOS (NOS 3) [81–83]. However, NOS occurs in a wide variety of other cell types. Although NOS 1 is the principal neuronal form of NOS and the predominant form in the nervous system, all three forms of NOS are reported to be expressed in some populations of neurons. In addition, NOS 1, NOS 2 and possibly NOS 3 have been detected in astrocytes, and NOS 1 in oligodendrocytes and microglia. The activity of all three forms of NOS increases during ischemia; NOS 1 and 3 within minutes and NOS 2 after several hours [84]. Since there is no oxygen available during ischemia, NO cannot be synthesized until the reperfusion phase. Likewise, many superoxide radicals are produced in mitochondria by xanthine oxidase and other pathways during and especially after ischemia. During reperfusion, NO and superoxide radicals combine to produce peroxynitrite, leading to the formation of more potent radicals [85, 86]. Other potentially damaging metabolites of NO include the nitrogen dioxide radical NO2 and nitryl chloride, formed by the reaction of nitrite, an end product of NO metabolism, with hypochlorous acid, itself produced by the action of myeloperoxidase in neutrophils [87, 88]. Experimental studies have demonstrated that the initial NO-mediated vasodilation and enhanced perfusion that result from the activation of NOS 3 are neuroprotective, at least during the first 2 h of ischemic insult [89]. However, the overall effects of enhanced NOS 1 and NOS 2 activity after ischemia are detrimental [90].

Iron Toxicity

183

Acidosis during cerebral ischemia potentiates oxidative neuronal death resulting from impaired antioxidant enzyme functions and increased intracellular free iron levels [116]. Enhanced proteolytic activity occurring in injured tissue also releases iron from storage proteins [117]. When non-protein-bound iron gains access to the extracellular space, its uptake by cells is enhanced by intracellular calcium and paradoxically also by increased levels of intracellular iron [118]. The toxicity of iron is inversely proportional to the availability of ferritin to sequester and detoxify ferrous ion, and directly proportional to the quantity of hydrogen peroxide available to produce hydroxyl radicals by the Fenton reaction [99, 119]. After hypoxia, the expression of Tf receptors (TfRs) on brain macrophages increases [120]. Hypoxia-reoxygenation is known to increase iron content and iron release in the extracellular space, causing injury to the periventricular white matter where microglial cells are known to preponderate [121]. It has been suggested that TfRs are involved in acquiring excess iron from the extracellular spaces, probably for storage; they therefore presumably help protect the brain from the toxic effects of excess iron. Normal axonal transport of brain iron has also been reported to be disrupted in anoxia-ischemia, leading to increased accumulation of iron in the white matter [121]. The increased expression of TfRs along with accumulation of iron in microglial cells is a protective mechanism to facilitate the active uptake of excess iron that may be released by iron-rich oligodendrocytes, or may accumulate due to disruption of its normal transport after hypoxic insult [119].

Role of Infections

Ischemia and subsequent reperfusion can set off an inflammatory reaction in the brain [122, 123]. IL-1, IL-6, transforming growth factor and fibroblast growth factor appear to be formed in activated microglia [65, 124]. They are thought to mediate the migration of inflammatory cells in reperfused tissue. Increased expression of the adhesion molecules P- and E-selectin and ICAM-1 on endothelial cells and integrins on leukocytes cause granulocytes to attach to the endothelium, migrate through the vessel wall and accumulate in the

interstitium [68, 125, 126]. There, after further activation by cytokines, they synthesize oxygen radicals, especially superoxide radicals, that proceed to damage neuronal tissue. So, reperfusion injury has been attributed to free radicals produced by neutrophils at the site of damage, but this is only part of the story. Neutrophils, which are activated by C5a and IL-8 released by ischemic tissue [127], damage reperfused tissue by other mechanisms in addition to free radical production, and the damaged tissue has other sources of free radicals besides neutrophils [128]. Recent clinical studies suggest that perinatal brain damage is closely associated with intrauterine infection before or at birth [129– 131]. However, it remains unclear whether fetal brain damage due to endotoxemia is the result of cerebral hypoperfusion caused by circulatory decentralization or is caused directly by endotoxins on cerebral tissue. Lipopolysaccharide (LPS)-induced effects on fetal circulation seem to play a central role in the development of fetal brain damage due to intrauterine infection [132]. A direct toxic effect of LPS on immature brain tissue seems unlikely; however, delayed activation of LPS-sensitive pathways involved in apoptosis-like cell death and damage limited to a small subgroup of cells, such as oligodendrocyte progenitors, cannot yet be excluded [132].

Conclusions

The relationship between free radical production and perinatal brain damage is complex. It is clear that free radical damage results from many pathogenic influences; hypoxia, ischemia-reperfusion, neutrophil and macrophage activation, Fenton chemistry, endothelial cell xanthine oxidase, phospholipid metabolism, NO, mitochondrial oxidative metabolism, iron and proteolytic pathways are all implicated. Reactive oxidant production by these different mechanisms contributes in a piecewise manner to the pathogenesis of perinatal brain injury, but each mechanism is only one of the many factors responsible. Each step in the oxidative cascade has become a potential target for therapy. The multiplicity of pathways and processes involved suggests that there is considerable potential for additive or synergistic benefit from combined therapies.

References 1 Halliwell B: Free radicals, antioxidants and human disease: Curiosity, cause, or consequence? Lancet 1994;344:721–724. 2 Fridovich I: Oxygen toxicity: A radical explanation. J Exp Biol 1998;201:1203–1209. 3 Halliwell B: Oxygen radicals as key mediators in neurological disease: Fact or fiction? Ann Neurol 1992;32:S10–S15. 4 Saugstad OD: Mechanisms of tissue injury by oxygen radicals: Implications for neonatal disease. Acta Paediatr 1996;85:1–4. 5 Sarker AH, Watanabe S, Seki S, Akiyama T, Okada S: Oxygen radical-induced single-strand DNA breaks and repair of the damage in a cell-free system. Mutat Res 1995;337:85–95. 6 Jacob RA: The integrated antioxidant system. Nutr Res 1995;15:755–766.

184

7 Phylactos AC, Leaf AA, Costeloe K, Crawford MA: Erythrocyte cupric/zinc superoxide dismutase exhibits reduced activity in preterm and low-birthweight infants at birth. Acta Paediatr 1995;84: 1421–1425. 8 Frank L, Sosenko RS: Development of lung antioxidant enzyme system in late gestation: Possible implications for the prematurely born infant. J Pediatr 1987;110:9–14. 9 Haga P, Hunde G: Selenium and vitamin E in cord blood from preterm and full-term infants. Acta Paediatr Scand 1978;67:735–739. 10 Bracci R, Buonocore G, Talluri B, Berni S: Neonatal hyperbilirubinemia. Evidence for a role of the erythrocyte enzyme activities involved in the detoxification of oxygen radicals. Acta Paediatr Scand 1988;77:349–356. 11 Gophinathan V, Miller NJ, Milner AD, RiceEvans CA: Bilirubin and ascorbate: Antioxidant activity in neonatal plasma. FEBS Lett 1994;349: 197–200.

Biol Neonate 2001;79:180–186

12 Inder TE, Graham P, Sanderson K, Taylor B: Lipid peroxidation as a measure of oxygen free radical damage in the very low birthweight infant. Arch Dis Child Fetal Neonatal Ed 1994;70:F107–F111. 13 Mishra OP, Delivoria-Papadopoulos M: Modification of modulatory sites of NMDA receptor in the fetal guinea pig brain during development. Neurochem Res 1992;17:1223–1228. 14 Crawford MA, Sinclair AJ: Nutritional influences in the evolution of mammalian brain; in: Lipids, Malnutrition and the Developing Brain. Ciba Foundation Symposium. Amsterdam, Associated Scientific Publishers, 1972, pp 267–292. 15 Inder TE, Volpe JJ: Mechanisms of perinatal brain injury. Semin Neonatol 2000;5:3–16. 16 Ozawa H, Nishida A, Mito T, Takashima S: Development of ferritin-positive cells in cerebrum of human brain. Pediatr Neurol 1994;10:44–48.

Buonocore/Perrone/Bracci

17 Takashima S, Kuruta H, Mito T, Houdou S, Konomi H, Yao R, Onodera K: Immunohistochemistry of superoxide dismutase-1 in developing human brain. Brain Dev 1990;12:211–213. 18 Sullivan JL: Iron Metabolism and Oxygen Radical Injury in Premature Infants. Boca Raton, CRC Press, 1992. 19 Evans PJ, Evans R, Kovar IZ, Holton AF, Halliwell B: Bleomycin-detectable iron in the plasma of premature and full-term neonates. FEBS Lett 1992;303:210–212. 20 McCord JM: Oxygen-derived free radicals in postischemic tissue injury. N Engl J Med 1985;312: 159–163. 21 Babior BM: Oxygen-dependent microbial killing by phagocytes. N Engl J Med 1978;298:659–668. 22 Chan PH, Fishman RA: Transient formation of superoxide radicals in polyunsatured fatty-acidinduced brain swelling. J Neurochem 1980;35: 1004–1007. 23 Kukreja RC, Kontos HA, Hess ML, Ellis EF: PGH synthase and lipoxygenase generate superoxide in the presence of NADH or NADPH. Circ Res 1986; 59:612–619. 24 Mishra OP, Delivoria-Papadopoulos M: Cellular mechanisms of hypoxic injury in the developing brain. Brain Res Bull 1999;48:233–238. 25 Delivoria-Papadopoulos M, Mishra OP: Mechanisms of cerebral injury in perinatal asphyxia and strategies for prevention. J Pediatr 1998;132:S30– S34. 26 Buonocore G, Liberatori S, Bini L, Mishra OP, Delivoria-Papadopoulos M, Pallini V, Bracci R: Hypoxic response of synaptosomal proteins in term guinea pig fetuses. J Neurochem 1999;73: 2139–2148. 27 Maulik D, Numagami Y, Ohnishi ST, Mishra OP, Delivoria-Papadopoulos M: Direct detection of oxygen free radical generation during in utero hypoxia in the fetal guinea pig brain. Brain Res 1998;798: 166–172. 28 Mishra OP, Delivoria-Papadopoulos M: Lipid peroxidation in developing fetal guinea pig brain during normoxia and hypoxia. Brain Res Dev Brain Res 1989;45:129–135. 29 Radzan B, Marro PJ, Tammela O, Goel R, Mishra OP, Delivoria-Papadopoulos M: Selective sensitivity of synaptosomal membrane function to cerebral cortical hypoxia in newborn piglets. Brain Res 1994;600:308–314. 30 Buonocore G, Perrone S, Longini M, Terzuoli L, Bracci R: Total hydroperoxide and advanced oxidation protein products in preterm hypoxic babies. Pediatr Res 2000;47:221–224. 31 White BC, Wiegenstein JG, Winegar CD: Brain ischemic anoxia. Mechanism of injury. JAMA 1984;251:1586–1590. 32 Ikeda T, Choi BH, Yee S, Murata Y, Quilligan EJ: Oxidative stress, brain white matter damage and intrauterine asphyxia in fetal lambs. Int J Dev Neurosci 1999;17:1–14. 33 Boveris A, Change B: The mitochondrial generation of hydrogen peroxide. Biochem J 1973;134: 707–716. 34 Kakinuma K, Minakami S: Effects of fatty acids on superoxide radical generation in leukocytes. Biochim Biophys Acta 1978;538:50–59. 35 Taylor DL, Edwards D, Mehmet H: Oxidative metabolism, apoptosis and perinatal brain injury. Brain Pathol 1999;9:93–117. 36 Hasegawa K, Yoshioka H, Sawada T: Lipid peroxidation in neonatal mouse brain subjected to two different types of hypoxia. Brain Dev 1991;13: 101–103. 37 Turrens JF: Superoxide production by the mitochondrial respiratory chain. Biosci Rep 1997;17: 3–8.

Neonatal Oxidative Brain Injury

38 Kowaltowski AJ, Vercesi AE: Mitochondrial damage induced by conditions of oxidative stress. Free Radic Biol Med 1999;26:463–471. 39 Watabe S, Hiroi T, Yamamoto Y, Fujioka Y, Hasegawa H, Yago N, Takahashi SY: SP-22 is a thioredoxin-dependent peroxide reductase in mitochondria. Eur J Biochem 1997;249:52–60. 40 Guidot DM, Repine JE, Kitlowski AD, Flores SC, Nelsol SK, Wright RM, McCord JM: Mitochondrial respiration scavenges extramitochondrial superoxide anion via a nonenzymatic mechanism. J Clin Invest 1995;96:1131–1136. 41 Radi R, Turrens JF, Chang LY, Bush KM, Crapo JD, Freeman BA: Detection of catalase in rat heart mitochondria. J Biol Chem 1991;266:22028– 22034. – 42 Fridovich I: Superoxide anion radical (OW2), superoxide dismutases and related matters. J Biol Chem 1997;272:18515–18517. 43 Sutton HC, Winterbourn CC: On the participation of higher oxidative states of iron and copper in Fenton reactions. Free Radic Biol Med 1989;6:53– 60. 44 Kowaltowski AJ, Castilho RF, Vercesi AE: Opening of the mitochondrial permeability transition pore by uncoupling or inorganic phosphate in the presence of Ca2+ is dependent on mitochondrialgenerated reactive oxygen species. FEBS Lett 1996; 378:150–152. 45 Kroemer G, Dallaporta B, Resche-Rigon M: The mitochondrial death/life regulator in apoptosis and necrosis. Annu Rev Physiol 1998;60:619–642. 46 Petit PX, Susin SA, Zamzami N, Mignotte B, Kroemer G: Mitochondria and programmed cell death: Back to the future. FEBS Lett 1996;369: 7–13. 47 Palmer C, Brucklalcher RM, Christensen MA, Vannucci RC: Carbohydrate and energy metabolism during the evolution of hypoxic-ischemic brain damage in the immature rat. J Cereb Blood Flow Metab 1990;10:227–235. 48 Yager JY, Brucklalcher RM, Vannucci RC: Cerebral energy metabolism during hypoxia-ischemia and early recovery in immature rats. Am J Physiol 1992;262:H672–H677. 49 Berger R, Jensen A, Krieglstein J, Steigelman JP: Cerebral energy metabolism in immature and mature guinea pig fetuses during acute asphyxia. J Dev Physiol 1992;18:125–128. 50 Berger R, Jensen A, Krieglstein J, Steigelman JP: Cerebral energy metabolism in fetal guinea pigs during moderate maternal hypoxemia at 0.75 of gestation. J Dev Physiol 1993;19:193–196. 51 Berger R, Gjedde A, Heck J, Muller E, Krieglstein J, Jensen A: Extension of the 2-deoxyglucose method to the fetus in utero: Theory and normal values for the cerebral glucose consumption in fetal guinea pigs. J Neurochem 1994;63:271–279. 52 Vannucci RC, Christensen MA, Yager JY: Nature, time-course and extent of cerebral edema in perinatal hypoxic-ischemic brain damage. Pediatr Neurol 1993;9:29–34. 53 Renchrona S, Westerberg E, Akesson B, Siesjo BK: Brain cortical fatty acids and phospholipids during and following complete and severe incomplete ischemia. J Neurochem 1982;38:84–93. 54 Monaghan DT, Bridges RJ, Cotman CW: The excitatory amino acid receptors: Their classes, pharmacology and distinct properties in the function of the central nervous system. Annu Rev Pharmacol Toxicol 1989;29:365–402. 55 de Courten-Myers GM, Fogelson HM, Kleinholz M, Myers RE: Hypoxic brain and heart injury thresholds in piglets. Biomed Biochim Acta 1989; 48:S143–S148. 56 Huang H-M, Gibson GE: Phosphatidylinositol metabolism during in vitro hypoxia. J Neurochem 1989;52:830–835.

57 McCord JM: Oxygen-derived free radicals in postischemic tissue injury. N Engl J Med 1985;312: 159–163. 58 De Haan HH, Ijzermans ACM, de Haan J, Van Belle H, Hasaart THM: Effects of surgery and asphyxia on levels of nucleosides, purine bases, and lactate in cerebrospinal fluid of fetal lambs. Pediatr Res 1994;36:595–600. 59 Al-Khalidi UAS, Chaglassian TH: The specific distribution of xanthine oxidase. J Biochem 1965;97: 318–320. 60 Kinuta Y, Kimura M, Itokawa M, Ishikawa M, Kikuchi H: Changes in xanthine oxidase in ischemic rat brain. J Neurosurg 1989;71:417–420. 61 Fridovich I: The biology of oxygen radicals. The superoxide radical is an agent of oxygen toxicity; superoxide dismutases provide an important defense. Science 1978;201:875–880. 62 Fridovich I: Superoxide radical: An endogenous toxicant. Annu Rev Pharmacol Toxicol 1983;23: 239–257. 63 Halliwell B, Gutteridge JC: Role of free radicals and catalytic metal ions in human disease: An overview; in Packer AN (ed): Methods in Enzymology. San Diego, Academic Press, 1990, pp 1–85. 64 Betz AL: Identification of hypoxanthine transport and xanthine oxidase activity in brain capillaries. J Neurochem 1985;44:574–579. 65 McRae A, Gilland E, Bona E, Hagberg H: Microglia activation after neonatal hypoxic-ischemia. Brain Res Dev Brain Res 1995;84:245–252. 66 Schiff SJ, Somjen GG: Hyperexcitability following moderate hypoxia in hippocampal tissue slices. Brain Res 1985;337:337–340. 67 Lafon-Cazal M, Pietri S, Culcasi M, Bockaert J: NMDA-dependent superoxide production and neurotoxicity. Nature 1993;364:535–537. 68 Palmer C: Hypoxic-ischemic encephalopathy. Therapeutic approaches against microvascular injury, and role of neutrophils, PAF and free radicals. Clin Perinatol 1995;22:481–517. 69 Buttke TM, Sandstrom PA: Oxidative stress as a mediator of apoptosis. Immunol Today 1994;15: 7–10. 70 Saikumar P, Dong Z, Weinberg JM, Venkatachalam MA: Mechanisms of cell death in hypoxia-reoxygenation injury. Oncogene 1998;17:3341–3349. 71 MacManus JP, Buchan AM, Hill IE, Rasquinha I, Preston E: Global ischemia can cause DNA fragmentation indicative of apoptosis in rat brain. Neurosci Lett 1993;164:89–92. 72 Fernandez A, Kiefer J, Fosdick L, Mc Conkey DJ: Oxygen radical production and thiol depletion are required for Ca2+ mediated endogenous endonuclease activation in apoptotic thymocytes. J Immunol 1995;155:5133–5139. 73 Polyak K, Xia Y, Zweier JL, Kinzler KW, Vogelstein BA: A model for p53-induced apoptosis. Nature 1997;389:300–305. 74 McConkey DJ: Biochemical determinants of apoptosis and necrosis. Toxicol Lett 1998;99:157–168. 75 Kim YM, Talanian RV, Billiar TR: Nitric oxide inhibits apoptosis by preventing increases in caspase-3 like activity via two distinct mechanisms. J Biol Chem 1997;272:31138–31148. 76 Richter C: Pro-oxidants and mithocondrial Ca2+: Their relationship to apoptosis and oncogenesis. FEBS Lett 1993;325:104–107. 77 Orrenius S, McConkey DJ, Bellomo G, Nicotera P: Role of Ca2+ in toxic cell killing. Trends Pharmacol Sci 1989;10:281–285. 78 Garthwaite J: Glutamate, nitric oxide and cell-cell signaling in the nervous system. Trends Neurosci 1991;14:60–67. 79 East SJ, Garthwaite J: NMDA receptor activation in rat hippocampus induces cGMP formation through the L-arginine-nitric oxide pathway. Neurosci Lett 1991;123:17–19.

Biol Neonate 2001;79:180–186

185

80 Beckman JS, Koppenol WH: Nitric oxide, superoxide and peroxynitrite: The good, the bad, and ugly. Am J Physiol 1996;271:C1424–C1437. 81 Doyle CA, Slater P: Localization of neuronal and endothelial nitric oxide synthase isoforms in human hippocampus. Neuroscience 1997;76:387– 395. 82 Kugler P, Drenckhahn D: Astrocytes and Bergmann glia as an important site of nitric oxide synthase 1. Glia 1996;16:165–173. 83 Merrill JE, Murphy SP, Mitrovic B, MackenzieGraham A, Dopp JC, Ding M, Griscavage J, Ignarro LJ, Lowestein CJ: Inducible nitric oxide synthase and nitric oxide production by oligodendrocytes. J Neurosci Res 1997;48:372–384. 84 Dalkara T, Mosckowitz MA: Neurotoxic and neuroprotective roles of nitric oxide in cerebral ischemia. Int Rev Neurobiol 1997;40:319–336. 85 Crow JP, Beckman JS: The importance of superoxide in nitric-oxide-dependent toxicity: Evidence for peroxynitrite-mediated injury. Adv Exp Med Biol 1996;387:147–161. 86 Szabò C, Ohshima H: DNA damage induced by peroxynitrite: Subsequent biological effects. Nitric Oxide 1997;1:373–385. 87 Eiserich JP, Cross CE, Jones AD, Halliwell B, van der Vliet A: Formation of nitrating and chlorinating species by reaction of nitrite with hypoclorous acid. A novel mechanism for nitric oxide-mediated protein modification. J Biol Chem 1996;271: 19199–19208. 88 Eiserich JP, Hristova M, Cross CE, Jones AD, Halliwell B, van der Vliet A: Formation of nitric oxide-derived inflammatory oxidants by myeloperoxidase in neutrophils. Nature 1998;391: 393–397. 89 Iadecola C: Bright and dark sides of nitric oxide in ischemic brain injury. Trends Neurosci 1997;20: 132–139. 90 Dalkara T, Mosckowitz MA: The complex role of nitric oxide in the pathophysiology of focal cerebral ischemia. Brain Pathol 1994;4:49–57. 91 Crichton RR, Ward RJ: Iron metabolism – new perspectives in view. Biochemistry 1992;31: 11255–11264. 92 Gutteridge JMC, Quinlan GJ: Antioxidant protection against organic and inorganic oxygen radicals by normal human plasma: The important primary role for iron-binding and iron-oxidising proteins. Biochim Biophys Acta 1993;1156:144–150. 93 O’Connell M, Halliwell B, Moorhouse CP, Aruoma OI, Baum H, Peters TJ: Formation of hydroxyl radicals in the presence of ferritin and haemosiderin. Is haemosiderin formation a biological protective mechanism? Biochem J 1986;234:727–731. 94 Gutteridge JM, Stocks J: Caeruloplasmin: Physiological and pathological perspectives. Crit Rev Clin Lab Sci 1981;14:257–329. 95 Gutteridge JMC: Inhibition of the Fenton reaction by the protein caeruloplasmin and other copper complexes. Assessment of ferroxidase and radical scavenging activities. Chem Biol Interact 1985;56: 113–120. 96 Weinberg ED: The iron-with-holding defense system. Am Soc Microbiol News 1993;59:559–562. 97 Gutteridge JMC, Paterson SK, Segal AW, Halliwell B: Inhibition of lipid peroxidation by the ironbinding protein lactoferrin. Biochem J 1981;199: 259–261.

186

98 Grootveld M, Bell JD, Halliwell B, Aruoma OI, Bomford A, Sadler PJ: Non-transferrin-bound iron in plasma or serum from patients with idiopathic hemochromatosis. Characterization by high performance liquid chromatography and nuclear magnetic resonance spectroscopy. J Biol Chem 1989;264:4417–4422. 99 Lefnesky EJ: Tissue iron overload and mechanisms of iron-catalyzed oxidative injury. Adv Exp Med Biol 1994;366:129–146. 100 Ying W, Han S-K, Miller JW, Swanson RA: Acidosis potentiates oxidative neuronal death by multiple mechanisms. J Neurochem 1999;73: 1549–1556. 101 McCord JM: Iron, free radicals and oxidative injury. Semin Hematol 1998;35:5–12. 102 Buonocore G, Zani S, Sargentini I, Gioia D, Signorini C, Bracci R: Hypoxia-induced free iron released in the red cells of newborn infants. Acta Paediatr 1998;87:77–81. 103 Ferrali M, Signorini C, Ciccoli L, Comporti M: Iron release and membrane damage in erythrocytes exposed to oxidizing agents, phenylhydrazine, divicine and isouramil. Biochem J 1992; 285:295–301. 104 Buonocore G, Gioia D, De Filippo M, Picciolini E, Bracci R: Superoxide anion release by polymorphonuclear leukocytes in whole blood of newborns and mothers during the peripartal period. Pediatr Res 1994;36:619–622. 105 Saugstad OD: Oxygen toxicity in the neonatal period. Acta Paediatr Scand 1990;79:881–892. 106 Buonocore G, Zani S, Perrone S, Caciotti B, Bracci B: Intraerythrocyte nonprotein-bound iron and plasma malondialdehyde in the hypoxic newborn. Free Radic Biol Med 1998;25:766– 770. 107 Bracci R, Buonocore G: The antioxidant status of erythrocytes in preterm and term infants. Semin Neonatol 1998;3:191–197. 108 Moison RMW, Palinckx JJS, Roest M, Houdikamp E, Berger HM: Induction of lipid peroxidation of pulmonary surfactant by plasma of preterm babies. Lancet 1993;341:79–82. 109 Berger HM, Mumby S, Gutteridge JMC: Ferrous ions detected in iron-overloaded cord blood plasma from preterm and term babies: Implications for oxidative stress. Free Radic Res 1995;22: 555–559. 110 Gutteridge JM, Mumby S, Koizumi M, Taniguchi N: ‘Free’ iron in neonatal plasma activates aconitase: Evidence for biologically reactive iron. Biochem Biophys Res Commun 1996;229:806– 809. 111 Gutteridge JMC: Iron and oxygen radicals in brain. Ann Neurol 1992;32(suppl):S16–S21. 112 Gutteridge JMC: Ferrous ions detected in cerebrospinal fluid by using bleomycin and DNA damage. Clin Sci (Colch) 1992;82:315–320. 113 Palmer C, Menzies SL, Roberts RL, Pavlick G, Connor JR: Changes in iron histochemistry after hypoxic-ischemic brain injury in the neonatal rat. J Neurosci Res 1999;56:60–71. 114 Ivacko JA, Sun R, Silverstein FS: Hypoxic-ischemic brain injury induces an acute microglial reaction in perinatal rats. Pediatr Res 1996;39: 39–47.

Biol Neonate 2001;79:180–186

115 Chan PH: Role of oxidants in ischemic brain damage. Stroke 1996;27:1124–1129. 116 Oubidar M, Boquillon M, Marie C, Schreiber L, Bralet J: Ischemia-induced brain iron delocalization: Effect of iron chelators. Free Radic Biol Med 1994;16:861–867. 117 Rothman RJ, Serroni A, Farber JL: Cellular pool of transient ferric iron, chelatable by deferoxamine and distinct from ferritin, that is involved in oxidative cell injury. Mol Pharmacol 1992;42: 703–710. 118 Cozzi A, Santambrogio P, Levi S, Arosio P: Iron detoxifying activity of ferritin. FEBS Lett 1990; 277:119–122. 119 Kaur C, Ling EA: Increased expression of transferrin receptors and iron in amoeboid microglial cells in postnatal rats following an exposure to hypoxia. Neurosci Lett 1999;262:183–186. 120 Adcock LM, Yamashita Y, Goddard-Finegold J, Smith CV: Cerebral hypoxia-ischemia increases microsomal iron in newborn piglets. Metab Brain Dis 1996;1:359–367. 121 Dietrich RB, Bradley WG: Iron accumulation in the basal ganglia following severe ischemicanoxic insults in children. Radiology 1998;68: 203–206. 122 Fellman V, Raivio KO: Reperfusion injury as the mechanism of brain damage after perinatal asphyxia. Pediatr Res 1996;41:599–606. 123 Rothwell NJ, Hopkins SJ: Cytokines and the nervous system. II. Actions and mechanisms of action. Trends Neurosci 1995;18:130–136. 124 Gehrmann J, Bonnekoh P, Miyazawa T, Oschlies U, Dux E, Hossmann K-A, Kreutzberg GW: The microglial reaction in the rat hippocampus following global ischemia: Immuno-electron microscopy. Acta Neuropathol (Berl) 1992;84:588– 595. 125 Etzioni A: Adhesion molecules – their role in health and disease. Pediatr Res 1996;39:191– 198. 126 Matsuo Y, Onodera H, Shiga Y, Nakamura M, Ninomya M, Kihara T, Kogure K: Correlation between myeloperoxidase-quantified neutrophil accumulation and ischemic brain injury in the rat. Effect of neutrophil depletion. Stroke 1994; 25:1469–1475. 127 Ambrosio G, Tritto I: Reperfusion injury: Experimental evidence and clinical implications. Am Heart J 1999;138:69–75. 128 Babior BM: Phagocytes and oxidative stress. Am J Med 2000;99:33–44. 129 De Felice C, Toti P, Stumpo M, Laurini RN, Picciolini E, Todros T, Buonocore G, Bracci R: Early neonatal brain injury in histological chorioamnionitis. J Pediatr, in press. 130 Toti P, De Felice C, Palmeri ML, Villanova M, Martin JJ, Buonocore G: Inflammatory pathogenesis of cortical polymicrogyria: An autopsy study. Pediatr Res 1998;44:291–296. 131 Yoon BH, Kim CJ, Romero R, Jun JK, Park KH, Choi ST, Chi JG: Experimentally induced intrauterine infection causes fetal brain white matter lesions in rabbits. Am J Obstet Gynecol 1997; 177:797–802. 132 Berger R, Garnier Y: Perinatal brain injury. J Perinat Med 2000;28:261–285.

Buonocore/Perrone/Bracci

Biol Neonate 2001;79:187–193

Effect of Graded Hypoxia on Cerebral Cortical Genomic DNA Fragmentation in Newborn Piglets Waseem Akhter Qazi M. Ashraf Santina A. Zanelli Om P. Mishra Maria Delivoria-Papadopoulos Department of Pediatrics, MCP Hahnemann University and St. Christopher’s Hospital for Children, Philadelphia, Pa., USA

Key Words Apoptosis W Hypoxia W DNA fragmentation W Brain W Newborn

Abstract Previous studies have shown that hypoxia is associated with modification of the cerebral cortical nuclear membrane, leading to increased intranuclear calcium. The increased intranuclear calcium activates calcium-dependent endonucleases, resulting in DNA fragmentation. The present study tests the hypothesis that the fragmentation of neuronal genomic DNA increases with an increase in the degree of cerebral tissue hypoxia. Sixteen newborn piglets were anesthetized, ventilated and divided into normoxic and hypoxic groups with varying degrees of hypoxia. Cerebral hypoxia was documented biochemically by measuring tissue levels of ATP and phosphocreatine. Isolation of cerebral cortical neuronal nuclei and DNA and their purity was confirmed by standard techniques. DNA samples were separated by electrophoresis on 1% agarose gel and stained with ethidium bromide. In the hypoxic samples, multiple low-molecular-weight DNA fragments were present as a smear pattern from 200 to 2,000 base pairs. Levels of high-energy phosphates were compared to the area of each smear for each animal to correlate the degree of hypoxia with the degree of DNA fragmentation. DNA fragmentation increased when high-energy phosphate levels decreased. We conclude that there is a critical threshold value of oxidative metabolism beyond which there are progressive changes in the cortical neuronal cells, leading to DNA fragmentation. Copyright © 2001 S. Karger AG, Basel

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 2001 S. Karger AG, Basel

Accessible online at: www.karger.com/journals/bon

Introduction

Neuronal death may either be a natural phenomenon or a pathological process. The loss of neurons is a normal process during central nervous system development [1, 2]. The failure to establish appropriate trophic connections leads to developmental switches which may trigger cell death. Pathological neuronal death may be the result of a variety of conditions, the most important being inflammatory (infection), traumatic, metabolic (hypoglycemia) and respiratory (hypoxia, ischemia). In recent years, a large number of studies have been carried out in order to understand cellular and molecular mechanisms of hypoxic-ischemic neuronal death. Hypoxic-ischemic neuronal death has long been considered to represent necrosis, characterized by passive cell swelling, rapid energy loss, generalized disruption of internal homeostasis leading to eventual lysis of the nucleus, organelles and plasma cell membranes and the release of intracellular components that induce a local inflammatory response, resulting in edema and injury to the neighboring cells. However, recent studies suggest that necrosis is only one of the mechanisms of cell death following hypoxia or ischemia to the brain. Programmed cell death, or apoptosis, also appears to contribute to cell death following hypoxia-ischemia, especially cell death that occurs days to weeks following the insult [3, 4]. Apoptosis is an active process that requires the activation of a genetic program and specific endonucleolytic digestion of nuclear DNA. In contrast to necrosis, programmed cell death is characterized by cell shrinkage, coarse chromatin aggregation with extensive nuclear DNA fragmentation, nuclear pyknosis and extrusion of membrane-bound cytoplasmic fragments or apoptotic bodies, but is not associated with the lysis

Waseem Akhter, MD Department of Pediatrics, Division of Neonatology MCP-Hahnemann University Hospital, Room #701, 7th Floor, 3300 Henry Ave Philadelphia, PA 19129 (USA) Tel. +1 215 842 4960, Fax +1 215 843 3505, E-Mail [email protected]

of the plasma membrane [5, 6]. Studies in cell culture models have demonstrated that hypoxia can trigger programmed cell death [7]. Programmed cell death as assessed by the cleavage of genomic DNA has also been shown to occur in the brain following focal [3, 8, 9] and global ischemia [4, 10–13]. The mechanism by which hypoxia causes DNA fragmentation has been extensively studied but is not well understood. It has been demonstrated that, during hypoxia, modification of the N-methylD-aspartate (NMDA) receptor leads to an increase in the intracellular Ca2+ concentration [14, 15] and is associated with the generation of oxygen free radicals [16–18]. Oxygen radicals may in turn alter the membrane structure and lead to changes in the function of the membrane components, such as receptors, by altering the surrounding lipid milieu. Free radicals may also interact with the nuclear membrane to modify membrane proteins, including the nuclear high-affinity Ca2+-ATPase and the inositol 1,3,4,5-tetrakisphosphate (IP4) and inositol 1,4,5-triphosphate (IP3) receptors, thus altering the mechanisms of intranuclear Ca2+ homeostasis and potentially leading to increased intranuclear Ca2+ concentrations. Increased intranuclear calcium can activate Ca2+-,Mg2+-dependent endonucleases [19–22] which cleave the genomic DNA at specific endonucleosomal sites, resulting in a specific DNA fragmentation pattern characteristic of programmed cell death. The present study was designed to test the hypothesis that the neuronal genomic DNA fragmentation correlates with the degree of cerebral tissue hypoxia in newborn piglets.

Materials and Methods

Animal Preparation The experimental protocol was approved by the Institutional Animal Care and Use Committee of MCP Hahnemann University. Studies were conducted in anesthetized, ventilated and instrumented newborn piglets, 2–5 days of age. Anesthesia was induced with 4% halothane, which was lowered to 1% during surgery while allowing the animals to breathe spontaneously. Lidocaine (2%) was injected locally before instrumentation for endotracheal tube insertion and femoral arterial and venous catheter insertion. After instrumentation, the use of halothane was discontinued, and anesthesia was maintained with nitrous oxide (79%), oxygen (21%) and fentanyl (50 Ìg/kg/h) throughout the experiment. Tubocurarine (0.3 mg/kg) was administered after placing the animal on a volume ventilator. Arterial blood gases, heart rate and blood pressure were recorded in all animals throughout the study. The core body temperature was maintained at 38.5–39 ° C with a warming blanket. Baseline measurements were obtained in both groups for 1 h after surgery to ensure normal arterial pressures and blood gas values. After stabilization following surgery, the 11 piglets assigned to the hypoxic group were exposed to different degrees of hypoxia by varying the inspired oxygen concentration (FiO2 ranging from 0.15 to 0.05) for 1 h, while the 5 piglets assigned to the normoxic group were ventilated to maintain a PaO2 of between 80 and 90 mm Hg. At the end of the experiment, the cortical brain tissue was removed within 5 s, a portion of it was placed in liquid nitrogen and stored at –80 ° C for biochemical studies and the rest of the brain was placed in a buffer containing 0.32 M sucrose, 10 mM Tris-HCl and 1 mM MgCl2 at 4 ° C for the isolation of nuclei.

188

Biol Neonate 2001;79:187–193

Isolation of Neuronal Nuclei Cerebral cortical nuclei were isolated according to the method described by Giuffrida et al. [23], with some modification. Cortical tissue was homogenized in 15 volumes of a medium containing 0.32 M sucrose, 10 mM Tris-HCl (pH 6.8) and 1 mM MgCl2 in a Dounce-type glass homogenizer (200 Ìm clearance) in 22 strokes. The homogenate was filtered through Nylon mesh bolting cloth (mesh 110) and centrifuged at 850 g for 10 min. The resulting pellet was resuspended in homogenizing buffer. The suspension was mixed with a medium containing 2.4 M sucrose, 10 mM Tris-HCl (pH 6.8) and 1 mM MgCl2 to achieve a final concentration of 2.1 M sucrose, which increases the yield of large neuronal nuclei. The nuclei were then purified by centrifugation at 53,000 g for 60 min. All procedures were carried out at 0–4 ° C. The above method of neuronal nuclei isolation yields a 90–95% pure preparation [24]. The purity of each nuclear preparation was assessed by phase contrast microscopy (Olympus). DNA Extraction DNA was isolated from normoxic and hypoxic neuronal nuclei according to the method described by Higuchi and Linn [25]. Firstly, 0.5 ml of cerebral cortical nuclei were centrifuged and resuspended in 0.7 ml of 50 mM Tris-HCl (pH 8.0), 100 mM EDTA and 0.5% sodium dodecyl sulfate and incubated with 35 Ìl of a fresh 10 mg/ml solution of proteinase K (Boehringer Mannheim, Germany) at 55 ° C overnight on a gently rocking platform. Following incubation, the digest was extracted with 0.7 ml of Tris-saturated phenol (pH 8.0) (Boehringer Mannheim) by shaking gently to completely mix the phases. The phases were separated by centrifugation and the aqueous phase was transferred to another tube. Extraction was repeated with 0.7 ml of phenol-chloroform (1:1). The samples were gently mixed and centrifuged, and the aqueous phase was transferred to another tube, avoiding interphase. The DNA was then precipitated by adding 70 Ìl of 3 M sodium acetate (pH 6.0) (i.e. 1/10 of the volume) and 0.7 ml of 100% ethanol at room temperature and shaking gently to mix thoroughly. DNA seen as stringy precipitate was pelleted by centrifugation and washed with 70% ethanol to remove traces of sodium dodecyl sulfate and phenol. After removing as much ethanol as possible, the DNA was air dried overnight and suspended with 0.1 ml of 10 mM Tris (pH 8.0) and 1mM EDTA. The DNA content was measured spectrophotometrically by absorbance at 260 nm and the purity was confirmed by a ratio of 1 1.7 at 260/280 nm. Agarose Gel Electrophoresis Nuclear DNA in equal amounts (0.5 Ìg) was dissolved in a mixture of 9 Ìl of Tris-EDTA and 1 Ìl of gel loading buffer [0.25% bromophenol blue, 0.25% xylene cyanol FF, 30% (v/v) glycerol] and then loaded into a well of a 1% agarose gel in Tris-boric acid-EDTA (TBE) buffer (89 mM Tris-boric acid, 2 mM EDTA, pH 8.0). Electrophoresis was performed in TBE for 50 min at 80 V. After electrophoresis, the gel was stained with ethidium bromide (0.5 Ìg/ml) in TBE (Sigma, St. Louis, Mo., USA) and analyzed using the Gel Doc 1000 system (BioRad, Hercules, Calif., USA). A 1-kb ready-load DNA ladder (Life Tech, Frederick, Md., USA) was used as a standard size marker. Determination of High-Energy Phosphates To assess the level of cerebral hypoxia in newborn piglets, tissue concentrations of ATP and phosphocreatine were determined with an enzyme-coupled assay as described by Lamprecht et al. [26]. A

Akhter/Ashraf/Zanelli/Mishra/ Delivoria-Papadopoulos

weighed portion of frozen brain tissue was powdered under liquid nitrogen and deproteinized by the addition of 6% (w/v) perchloric acid. The extract was allowed to thaw on ice and was centrifuged at 2,000 g for 5 min. Aliquots of supernatant were neutralized with triethanolamine-potassium bicarbonate and centrifuged at 2,000 g for 5 min. ATP and phosphocreatine concentrations were determined in a 1-ml assay medium containing 50 mM triethanolamine, 5 mM MgCl2, 1 mM EDTA, 2 mM glucose, 400 Ìl of neutralized extract and 20 Ìl of NADP. Hexokinase (10 Ìl) was added and readings were taken at 0, 5, 10, 15 and 20 min until a steady state was reached. The ATP concentration was calculated from the increase in absorbance at 340 nm for the 20-min period following the addition of 20 Ìl of hexokinase. Twenty microliters of ADP and 20 Ìl of creatine kinase were added and readings were taken at 5-min intervals until a steady state was reached. The phosphocreatine concentration was calculated from the increase in absorbance at 340 nm between 0 and 20 min after the addition of creatine kinase. Statistical Analysis Statistical analysis was performed using an unpaired t test to compare the normoxic and hypoxic DNA fragments. A p value of ! 0.05 was considered statistically significant. All values are presented as the mean B standard deviation. The exponential form of nonlinear regression was used to analyze the relation between ATP, phosphocreatine and DNA fragmentation.

Results

Five normoxic and 11 hypoxic newborn piglets were studied at graded levels of hypoxia. PaO2 decreased significantly within the first 5 min following the reduction in FiO2. PaCO2 did not change significantly throughout the experiment. Arterial pH decreased significantly for all animals following 1 h of hypoxia, indicating a metabolic acidosis in the presence of normal PaCO2. There was also a trend of an increase in heart rate as the hypoxia progressed, but in a few cases it dropped to a lower rate, probably because of uncompensated metabolic acidosis and hypoxemia.Gel electrophoresis of DNA from both the normoxic and hypoxic group showed a significant difference in terms of fragmentation compared to a standard 1-kb DNA ladder. DNA from normoxic piglets contained one high-molecular-weight DNA band (fig. 1). In contrast, DNA from the graded hypoxia group showed both high- and low-molecular-weight fragments ranging from 200 to 2,000 base pairs, demonstrating fragmentation of nuclear DNA, following 1 h of hypoxia in newborn piglets. Analysis of DNA fragments by Gel Doc 1000 (BioRad) showed that the peak area corresponding to the amount of fragmentation was significantly higher in the hypoxic group (1,676.2 B 595.8 OD/mm) as compared to the normoxic group (165.9 B 62.7 OD/mm) (p ! 0.001). Representative Gel Doc scans of DNA fragments from 3 normoxic and 6 hypoxic newborn piglets are shown in figure 2.The relationship between DNA fragmentation secondary to graded hypoxia and high-energy phosphate levels (table 1) was examined by plotting the peak areas of DNA fragments from the normoxic and hypoxic groups against ATP and phosphocreatine levels. As shown in figures 3 and 4, the degree of neuronal cortical genomic DNA fragmentation secondary to hypoxia correlated exponentially with both the ATP and phosphocreatine concentration (r = 0.88). However, there was no significant increase in fragmentation until the ATP and phosphocreatine levels decreased by 1 50% compared to baseline levels.

Effect of Hypoxia on Cerebral Cortical DNA Fragmentation in Newborn Piglets

Fig. 1. Agarose gel electrophoresis of neuronal nuclear DNA from normoxic and hypoxic newborn piglets. DNA was extracted by the phenol-chloroform-isoamyl alcohol method, electrophoresed on 1% agarose gel and stained with ethidium bromide. A ready-load 1-kb DNA ladder was used as standard (lane 1). Lanes 2–7 represent DNA samples from hypoxic brain and lanes 8 and 9 from normoxic brain of newborn piglets.

Discussion

Programmed cell death as assessed by the cleavage of genomic DNA has been shown to occur in the brain following focal as well as global ischemia [3, 4, 7–9]. Studies conducted in newborn rats following either mild (15 min) or severe (60 min) unilateral hypoxia-ischemia showed that selective neuronal loss produced by 15 min of injury was associated with apoptosis and DNA laddering 3 days after hypoxia, while 60 min of hypoxia-ischemia accelerated this process, resulting in signs of DNA degradation 10 h after hypoxia [27]. In another study using the 4-vessel occlusion model following global ischemia for 30 min, no evidence of DNA fragmentation was observed by TUNEL staining 24 h after ischemia. However, 72 h after ischemia, DNA fragmentation was observed in hippocampal CA1 neurons [28]. Recent studies have also demonstrated that oxygen deprivation alone could induce DNA fragmentation in cortical neurons in vitro [29]. The present study shows that, in the newborn, fragmentation of neuronal genomic DNA occurs immediately following hypoxia, and that the degree of fragmentation correlates with the severity of cerebral hypoxia. The data also suggest that there is a threshold for the energy depletion (ATP and phosphocreatine) of approximately 50%, beyond which DNA fragmentation is observed. This is in contrast to earlier studies regarding the effect of hypoxia on plasma membrane Na+-,K+-ATPase activity and lipid peroxidation in newborn piglets, which indicated a threshold in energy depletion of 30%, beyond which changes occur through activation of the NMDA receptor, phospholipases and the production of free radicals [30]. Although the biochemical and pathological consequences of hypoxic-ischemic brain injury have been extensively studied, the mechanisms of cellular and intranuclear processes such as structural alterations in chromosomal DNA are not well understood. Two distinct patterns of DNA degradation during hypoxic-ischemic brain cell death have been observed [31]. The first is random cleavage of the DNA, including nick formation in a single strand. The second pattern is characterized by double-stranded DNA degradation into oligonucleosomal fragments. These two DNA fragmentation patterns are visualized on agarose gel electrophoresis as a ‘smear’ or ‘ladder’, respectively. The ladder pattern of DNA fragmentation has been the

Biol Neonate 2001;79:187–193

189

Fig. 2. Analysis of DNA fragments in neuronal nuclei of newborn piglets. The peak amplitude area of DNA fragments

of normoxic and hypoxic neuronal nuclei of newborn piglets is shown. The peak amplitude area (i.e. area under the curve) is expressed as OD/mm.

Table 1. Cerebral tissue high-energy

phosphates and peak amplitude area of DNA fragments from normoxic and hypoxic newborn piglets

ATP PCr Peak 1 Ìmol/g brain Ìmol/g brain OD/mm 4.236 4.996 4.483 5.432 4.075 2.860 1.811 1.572 1.250 1.119 1.054 1.092 0.966 0.923 0.824 0.649

4.220 4.419 3.951 4.993 3.366 2.590 1.832 0.706 1.050 0.803 0.612 0.937 0.802 0.957 0.700 0.782

61.73 153.19 83.68 158.25 106.74 118.73 166.39 220.93 415.73 1,500.20 1,008.95 86.20 14.76 21.54 8.60 4.20

Peak 2 OD/mm 35.44 43.40 86.69 20.06 6.85 230.85 843.89 1,363.02 941.35 218.03 808.25 1,125.61 146.65 1,326.21 36.92

Peak 3 OD/mm

Peak 4 OD/mm

64.18

9.65 1,717.01

2,318.86 95.85 1,175.28 479.30 741.52

674.53

637.89

PCr = Phosphocreatine.

190

Biol Neonate 2001;79:187–193

Akhter/Ashraf/Zanelli/Mishra/ Delivoria-Papadopoulos

Mean OD/mm 60.82 40.78 54.12 14.33 8.25 688.86 505.14 791.97 678.54 1,500.20 613.49 1,062.94 512.30 468.54 654.59 380.77

Fig. 3. Relation between ATP level and DNA fragmentation in neuronal nuclei of newborn piglets. The total peak amplitude area, expressed as OD/mm, is correlated with the levels of ATP. Each point represents one individual animal. r is the correlation coefficient.

Fig. 4. Relation between phosphocreatine level and DNA fragmenta-

focus of most studies, due to its association with apoptosis, a nonnecrotic mode of cell death [5, 32]. However, it is becoming clear that morphological characteristics of apoptosis are not always associated with ladder-type DNA fragmentation [33]. This has also been supported by the observation of ladder-type DNA fragments in some necrotic cells [34]. It has been proposed that the cleavage of DNA at its internucleosomal linker regions is produced by a specific endonuclease that is Ca2+ dependent [21, 32, 35]. However, the molecular mechanisms of endonuclease activation, as well as the relationship between the production of the two patterns of DNA fragmentation, are not clear. A possible explanation could be the competition between endonuclease and proteinase. Increased intracellular Ca2+ may result in proteinase activation leading to the digestion of histone proteins and consequently the degradation of nucleosomal structure. If proteinase activation precedes the activation of endonuclease, the disintegration of histone proteins will allow random access of endonuclease to host DNA. Under these conditions, nonspecific DNA fragmentation will follow, producing a ‘smear’ pattern [22]. In previous studies, smear patterns were observed during acute hypoxia, suggesting that this mechanism might be operative. On the other hand, if the activation of endonucleases precedes that of proteinase, DNA fragmentation will occur specifically at internucleosomal linker regions, producing a ‘ladder’ of nucleosome-sized fragments. In hypoxic brain injury, both types of nonspecific and specific fragments may be produced, due to the activation of both proteinases and endonucleases as well as the production of oxygen free radicals. The mechanisms that trigger the cascade of events leading to programmed cell death following hypoxia are not well understood. Ca2+ appears to be a critical mediator of hypoxic cell death, leading to free radical generation and subsequent peroxidation of the nuclear membrane [18, 36], as well as alteration of Ca2+ entry mechanisms and a potential increase in the intranuclear Ca2+ concentration. Intranu-

clear Ca2+ signals are required for the regulation of nuclear processes such as gene transcription, cell cycle regulation, synthesis, repair and fragmentation of nuclear DNA [36–38]. These nuclear processes require that appropriate Ca2+ signals be transduced across the nuclear envelope. Studies suggest that the nuclear envelope serves as a nucleocytoplasmic barrier to calcium and insulates the nucleus from large changes in the calcium concentration in the cytosol [39]. The nuclear envelope consists of an inner and outer nuclear membrane interrupted by nuclear pores. The outer nuclear membrane contains high-affinity Ca2+-ATPase activity and the IP4 receptors, while IP3 is located on the inner membrane [38–40]. Thus, ATP and IP4 induce Ca2+ uptake into the nuclear envelope, with the subsequent release of Ca2+ into the nucleoplasm through the IP3 receptor. Several studies have demonstrated a gradient between nuclear and cytosolic compartments, suggesting that calcium movement in and out of the nucleus is a regulated process rather than one of passive diffusion [38–40]. A free cytosolic Ca2+ concentration below 0.1 ÌM may serve as a regulator of the Ca2+-ATPase pathway by which ATP facilitates Ca2+ entry into the nucleus. However, this pathway is not operative when the extranuclear Ca2+ concentration is between 1 and 10 ÌM. At this higher concentration of calcium, IP4-mediated calcium uptake predominates. During hypoxia, the increase in intracellular calcium may lead to an increase in intranuclear calcium, resulting in activation of Ca2+-dependent endonucleases and subsequent fragmentation of nuclear DNA. Alteration of the intranuclear Ca2+ concentration in response to tissue hypoxia may alter the expression of the proapoptotic gene bax and the antiapoptotic gene bcl-2. Bcl-2 and Bax proteins are expressed in neurons of the central and peripheral nervous system [41, 42]. The active form of the Bcl-2 protein, which promotes cell survival, forms a heterodimer with the Bax protein, which promotes cell death by allowing the activation of caspases [43]. Thus, the ratio of Bcl-2 and Bax appears to determine the susceptibility of cells to apo-

Effect of Hypoxia on Cerebral Cortical DNA Fragmentation in Newborn Piglets

Biol Neonate 2001;79:187–193

tion in neuronal nuclei of newborn piglets. The total peak amplitude area, expressed as OD/mm, is correlated with the levels of phosphocreatine. Each point represents one individual animal. r is the correlation coefficient.

191

ptotic stimuli [44, 45]. Bax is thought to contribute to the vulnerability of neurons to apoptotic cell death induced by exposure to gamma radiation, glutamate and kainate [9, 13, 44]. Expression of bcl-2 is downregulated in cerebellar granule neurons in culture exposed to excitotoxic concentrations of glutamate, and in hippocampal neurons in vivo following ischemic damage. Systemic administration of kainate induced a 45% decrease in bcl-2 immunoreactivity and a three-fold increase in bax mRNA. Bax and Bcl-2 are also thought to play a role in cell death following hypoxia and ischemia. Several studies have shown that cerebral ischemia alters the expression of Bax and Bcl-2 [9, 11, 46, 47]. Although the precise mechanism of delayed neuronal death after ischemia is not clear, increased Bcl-2 expression has been observed in surviving neurons after cerebral ischemia [46, 48], suggesting that the expression of Bcl-2, at least in part, determines the fate of neurons following an ischemic injury. In summary, neuronal nuclear DNA fragmentation correlates with the degree of cerebral hypoxia in newborn piglets. However,

DNA fragmentation was observed when high-energy phosphates decreased by more than 50% from the baseline. This observation is in contrast to our previous studies, which demonstrated that neuronal plasma membrane Na+-,K+-ATPase activity decreased when cerebral energy levels decreased by 1 30% compared to baseline. We speculate that while energy-dependent biochemical processes at the plasma membrane require high levels of ATP (mM ), the energy requirement for intranuclear functions is lower (ÌM ); thus, a higher degree of hypoxia is necessary to induce changes in intranuclear processes and fragmentation of genomic DNA.

Acknowledgments

This study was supported by NIH grants HD-20337 and HD38079. The authors wish to express their gratitude to Mrs. Joanna A. Kubin for her expert technical assistance.

References 1 Deckworth TL, Johnson EM: Temporal analysis of events associated with programmed cell death (apoptosis) of sympathetic neurons deprived of nerve growth factor. J Cell Biol 1993;123:1207– 1222. 2 Diana A, Setzu M, Sirigu S, Diaz G: Nuclear patterns of apoptotic and developing neurons of superior cervical ganglion of newborn rat. Int J Dev Neurosci 1993;11:773–780. 3 Linnik MD, Zobrist RH, Hatfield MD: Evidence supporting a role for programmed cell death in focal cerebral ischemia in rats. Stroke 1993;24: 2002–2008. 4 Ferrer I, Tortosa A, Macaya A, Sierra A, Moreno D, Munell F, Blanco R, Squier W: Evidence of nuclear DNA fragmentation following hypoxiaischemia in the infant rat brain, and transient forebrain ischemia in the adult gerbil. Brain Pathol 1994;4:115–122. 5 Wyllie AH, Kerr JFR, Currie AR: Cell death: The significance of apoptosis. Int Rev Cytol 1980;68: 251–306. 6 Columbano A: Cell death: Current difficulties in discriminating apoptosis and necrosis in the context of pathological processes in vivo. J Cell Biochem 1995;58:181–190. 7 Rosenbaum DM, Michaelson M, Batter DK, Doshi P, Kessler JA: Evidence for hypoxia induced programmed cell death of cultured neurons. Ann Neurol 1994;36:864–870. 8 Dragunow M, Beilharz E, Sirimanne E, Lawlor P, Williams C, Bravo R, Gluckman P: Immediateearly gene protein expression in neurons undergoing delayed death, but not necrosis, following hypoxic-ischaemic injury to the young rat brain. Brain Res Mol Brain Res 1994;25:19–33. 9 Gillardon F, Lenz C, Waschke KF, Krajewski S, Reed JC, Zimmermann M, Kuschinsky W: Altered expression of Bcl-2, Bcl-X, Bax and c-Fos colocalizes with DNA fragmentation and ischemic cell damage following middle cerebral artery occlusion in rats. Brain Res Mol Brain Res 1996;40:254– 260. 10 Nitatori T, Sato N, Waguri S, Karasawa Y, Araki H, Shibanai K, Kominami E, Uchiyama Y: Delayed neuronal death in the CA1 pyramidal layer of the gerbil hippocampus following transient ischemia in apoptosis. J Neurosci 1995;15:1001– 1011.

192

11 Honkaniemi J, Massa SM, Breckinridge M, Sharp FR: Global ischemia induces apoptosis-associated genes in hippocampus. Brain Res Mol Brain Res 1996;43:79–88. 12 Mehmet H, Yue X, Squier MV, Lorek A, Cady E, Penrice J, Sarraf C, Wylezinska M, Kirkbride V, Cooper C, Brown GC, Wyatt JS, Reynolds EOR, Edwards AD: Increased apoptosis in the cingulate sulcus of newborn piglets following transient hypoxia-ischaemia is related to the degree of high energy phosphate depletion during the insult. Neurosci Lett 1994;181:121–125. 13 Kitada S, Krajewski S, Miyashita T, Krajewski M, Reed JC: Gamma-radiation induces upregulation of Bax protein and apoptosis in radiosensitive cells in vivo. Oncogene 1996;12:187–192. 14 Frandsen A, Schousboe A: Excitatory amino acidmediated cytotoxicity and calcium homeostasis in cultured neurons. J Neurochem 1993;60:1202– 1211. 15 Zanelli SA, Numagami Y, McGowan JE, Mishra OP, Delivoria-Papadopoulos M: NMDA receptormediated calcium influx in cerebral cortical synaptosomes of the hypoxic guinea pig fetus. Neurochem Res 1999;24:437–446. 16 Lafon-Cazal M, Pietri S, Culcasi M, Bockaert J: NMDA-dependent superoxide production and neurotoxicity. Nature 1993;364:535–537. 17 Numagami Y, Zubrow AB, Mishra OP, DelivoriaPapadopoulos M: Lipid free radical generation and brain cell membrane alteration following nitric oxide synthase inhibition during cerebral hypoxia in the newborn piglet. J Neurochem 1997;69:1542– 1547. 18 Maulik D, Numagami Y, Ohnishi ST, Mishra OP, Delivoria-Papadopoulos M: Direct measurement of oxygen free radicals during in utero hypoxia in the fetal guinea pig brain. Brain Res 1998;798: 166–172. 19 Hameed A, Olsen KJ, Lee MK, Lichtenheld MG, Podack ER: Cytolysis by Ca-permeable transmembrane channels: Pore formation causes extensive DNA degradation and cell lysis. J Exp Med 1989; 169:765–777. 20 Jones DP, McConkey DJ, Nicoreta P, Orrenius S: Calcium-activated DNA fragmentation in rat liver nuclei. J Biol Chem 1989;264:6398–6403.

Biol Neonate 2001;79:187–193

21 Cohen JJ, Duke RC: Glucocorticoid activation of a calcium-dependent endonuclease in thymocyte nuclei leads to cell death. J Immunol 1984;132:38– 42. 22 Tominaga T, Kure S, Narisawa K, Yoshimoto T: Endonuclease activation following focal ischemic injury in the rat brain. Brain Res 1993;608:21–26. 23 Giuffrida AM, Cox D, Mathias AP: RNA polymerase activity in various classes of nuclei from different regions of rat brain during post-natal development. J Neurochem 1975;24:749–755. 24 Austoker J, Cox D, Mathias AP: Fractionation of nuclei from brain by zonal centrifugation and a study of the ribonucleic acid polymerase activity in the various classes of nuclei. Biochem J 1972;129: 1139–1155. 25 Higuchi Y, Linn S: Purification of all forms of HeLa cell mitochondrial DNA and assessment of damage to it caused by hydrogen peroxide treatment of mitochondria or cells. J Biol Chem 1995; 270:7950–7956. 26 Lamprecht W, Stein P, Heinz F, Weisser H: Creatine phosphate; in Bergmeyer HU (ed): Methods of Enzymatic Analysis. New York, Academic Press, vol 4, pp 1777–1778. 27 Beilharz EJ, Williams CE, Dragunoe M, Sirimane ES, Gluckman PD: Mechanisms of delayed cell death following hypoxic-ischemic injury in the immature rat: Evidence for apoptosis during selective neuronal loss. Brain Res Mol Brain Res 1995;29: 1–14. 28 Clemens JA, Stephenson DT, Dixon EP, Smalstig EB, Mincy RE, Rash KS, Little SP: Global cerebral ischemia activates nuclear factor-ÎB prior to evidence of DNA fragmentation. Brain Res Mol Brain Res 1997;48:187–196. 29 Copin JC, Reola LF, Chan TYY, Li Y, Epstein CJ, Chan PH: Oxygen deprivation but not a combination of oxygen, glucose, and serum deprivation induces DNA degradation in mouse cortical neurons in vitro: Attenuation by transgenic overexpression of CuZn-superoxide dismutase. J Neurotrauma 1996;13:233–244. 30 DiGiacomo JE, Pane CR, Gwiazdowski S, Mishra OP, Delivoria-Papadopoulos M: Effect of graded hypoxia on brain cell membrane injury of newborn piglets. Biol Neonate 1992;61:25–32.

Akhter/Ashraf/Zanelli/Mishra/ Delivoria-Papadopoulos

31 Tominaga T, Kure S, Yoshimoto T: DNA fragmentation in cortical freeze injury of rats. Neurosci Lett 1992;139:265–268. 32 Arends MJ, Morris RG, Wyllie AH: Apoptosis. The role of the endonuclease. Am J Pathol 1990; 136:539–608. 33 Cohen GM, Sun XM, Snowden RT, Dinsdale D, Skilletter DN: Key morphological features of apoptosis may occur in the absence of internucleosomal DNA fragmentation. Biochem J 1992;286:331– 334. 34 Fukuda K, Kojiro M, Chiu JF: Demonstration of extensive chromatin cleavage in transplanted Morris hepatoma 7777 tissue: Apoptosis or necrosis? Am J Pathol 1993;142:935–946. 35 Ishida R, Akiyoshi H, Takahashi T: Isolation and purification of calcium and magnesium dependent endonuclease from rat liver nuclei. Biochem Biophys Res Commun 1974;56:703–710. 36 Maulik D, Kuo MF, Mishra OP, Delivoria-Papadopoulos M: The effect of hypoxia on the nuclear membranes in cerebral cortex of the guinea pig fetus. Pediatr Res 1998;42:1068.

Effect of Hypoxia on Cerebral Cortical DNA Fragmentation in Newborn Piglets

37 Steinhardt RA, Alderton J: Intracellular free calcium rise triggers nuclear envelope breakdown in the sea urchin embryo. Nature 1988;332:364–366. 38 Santella L, Carafoli E: Calcium signaling in the cell nucleus. FASEB J 1997;11:1091–1109. 39 Al-Mohanna FA, Caddy KWT, Boisover SR: The nucleus is insulated from large cytosolic calcium ion changes. Nature 1994;367:745–750. 40 Humbert JP, Matter N, Artault JC, Koppler P, Malviya AN: Inositol 1,4,5-triphosphate receptor is located to the inner nuclear membrane vindicating regulation of nuclear calcium signaling by inositol 1,4,5-triphosphate. J Biol Chem 1996;271: 478–485. 41 Krajewski S, Krajewski M, Shabaik A, Miyashita T, Wang HG, Reed JC: Immunohistochemical determination of in vivo distribution of Bax, a dominant inhibitor of Bcl-2. Am J Pathol 1994;145: 1323–1336. 42 Merry DE, Veis EDJ, Hickey WF, Korsmeyer SJ: bcl-2 protein expression is widespread in the developing nervous system and retained in the adult PNS. Development 1994;120:301–311. 43 Chinnaiyan AM, O’Rourke K, Lane BR, Dixit VM: Interaction of CED-4 with CED-3 and CED9: A molecular framework for cell death. Science 1997;275;1122–1126.

44 Gillardon F, Wickert H, Zimmermann M: Up-regulation of bax and down-regulation of bcl-2 is associated with kainate-induced apoptosis in mouse brain. Neurosci Lett 1995;192:85–88. 45 Yin XM, Oltvai ZN, Korsmeyer SJ: BH1 and BH2 domains of Bcl-2 are required for inhibition of apoptosis and heterodimerization with Bax. Nature 1994;369:321–323. 46 Chen J, Zhu RL, Nakayama M, Kawaguchi K, Jin K, Stetler RA, Simon RP, Graham SH: Expression of the apoptosis-effector gene, Bax, is up-regulated in vulnerable hippocampal CA1 neurons following global ischemia. J Neurochem 1996;67:64–71. 47 Chen J, Graham SH, Nakayama M, Zhu RL, Jin K, Stetler RA, Simon RP: Apoptosis repressor genes Bcl-2 and Bcl-x-long are expressed in the rat brain following global ischemia. J Cereb Blood Flow Metab 1997;17:2–10. 48 Chen J, Graham SH, Chan PH, Lan JQ, Zhou GJ, Simon RP: bcl-2 is expressed in neurons that survive focal ischemia in the rat. Neuroreport 1995:6: 394–398.

Biol Neonate 2001;79:187–193

193

Biol Neonate 2001;79:194–200

Is Periventricular Leucomalacia a Result of Hypoxic-Ischaemic Injury? Hypocapnia and the Preterm Brain

Gorm Greisen a Robert C. Vannucci b a Department

of Neonatology, Rigshospitalet, Copenhagen, Denmark; b Department of Paediatrics (Paediatric Neurology), The Pennsylvania State University College of Medicine, The Milton S. Hershey Medical Centre, Hershey, Pa., USA

Key Words Cerebral blood flow W Hypoxia-ischaemia W Brain damage W Mechanical ventilation W Prematurity

Abstract Decrease in the arterial partial pressure of carbon dioxide (PaCO2) causes a reduction in cerebral blood flow in humans and in most animal species; in adults as well as in newborns and even in fetal life. Severely decreased PaCO2 increases cerebral lactate production, modifies spontaneous electric brain activity, and may decrease the metabolic rate of oxygen. A relation between very low PaCO2 and brain injury, however, has not been shown in adult humans or full-term newborn infants, nor in perinatal animals. In contrast, an association between low PaCO2 and cerebral palsy and white matter injury in preterm infants has been reported repeatedly. A cause-andeffect relation is suggested by data from the immature rat: brain damage induced by ligation of a carotid artery can be reduced by adding CO2 to the inspired gas and hence avoiding the consequences of spontaneous hyperventilation. This may be relevant for the clinical care of preterm infants, since PaCO2 to a large extent is a function of respiratory management. The questions to be addressed are whether hypocapnia sensitises the brain to hypoxaemia, and also whether the escape mechanisms are less effective in the preterm human brain. Copyright © 2001 S. Karger AG, Basel

Introduction

Most of what we know about the influence of the partial pressure of oxygen (PO2) and carbon dioxide (PCO2) in the blood on cerebral blood flow (CBF) and oxygen metabolism was learned 40 years ago

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0194$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

when the newly developed quantitative methods were used to meet the challenges of aviation medicine [1]. The question of the regulation of CBF was re-examined in the immature brain 20 years later to find a more rational basis for the evolving active perinatal care. The assumption that the vasculature of the immature brain would show significantly different responses from those in the mature brain was in general not confirmed [2]. To a certain degree, the research effort has slowed down in recent years because of lack of results of direct clinical importance. The purposes of this paper are (1) to review the physiology of the cerebral effects of hypocapnia, (2) to summarise the evidence of a relation between hypocapnia due to hyperventilation of preterm neonates and brain damage, and (3) to identify areas of new research.

Acute Cerebral Effects of Hyperventilation

Effect on Cerebral Energy Metabolism The changes in cerebral metabolism which result from hypocapnia are complex. Moderate hyperventilation does not reduce the cerebral metabolism of oxygen in healthy, young men [1]. Extreme hypocapnia affects cerebral energy metabolism through a degree of tissue hypoxia, as indicated by: (1) a decrease in brain and cerebral venous blood PO2 [3]; (2) an increase in brain and cerebrospinal fluid lactate and lactate/puruvate ratios [4]; (3) an increase in brain tissue NADH/NAD+ ratios [5], and (4) an increase in the cerebral metabolic rate for glucose. The latter suggests a stimulation of anaerobic glycolysis [6]. Thus, in the adult rat, the accumulation of metabolic acids reverts the alkaline intracellular pH, which is observed at moderate levels of hypocapnia [arterial partial pressure of CO2 (PaCO2) = 3.5 kPa], to normal or even mildly subnormal (acidosis) when PaCO2 falls below 2.7 kPa [4, 7] (fig. 1). Despite these alterations, high-energy phos-

Gorm Greisen Department of Neonatology 5021 Rigshospitalet, Blegdamsvej 9 DK–2100 Copenhagen Ø (Denmark) Tel. +45 3545 4320, Fax +45 3545 5025, E-Mail [email protected]

Fig. 1. Cerebral glycolytic and tricarboxylic intermediates during hypocapnia in adult rats. The columns represent the

mean tissue concentration of 5–6 animals. * p ! 0.05, ** p ! 0.001 compared to normocapnia. ·-KG = ·-ketoglutaric acid. (Derived from data of MacMillan and Siesjö [5].)

phate reserves are relatively well preserved, even during extreme hypocapnia, owing in part to a suppression of cerebral energy utilisation [7]. Findings similar to those of adult animals have been observed in newborn dogs and lambs exposed to hyperventilation. Normally, the newborn brain consumes a small amount of lactate, but hyperventilation induces a net production [8, 9]. Hypocapnia to PaCO2 of 2.0– 2.3 kPa leads to a 20% reduction in the cerebral metabolic rate for oxygen in newborn dogs [10] and pigs [11], but not in newborn sheep [9]. Taken together, the findings suggest a partial shift from aerobic to anaerobic glycolysis to maintain normal tissue energy stores. Effect on Neurotransmission Severe hypocapnia (PaCO2 = 1.5 kPa) modifies the NMDA receptor channel complex in newborn piglets [12], possibly facilitating excitation, whereas acidosis on the other hand may inhibit excitation [13]. Effect on Cerebral Function Hyperventilation causes dizziness. Slow activity in the EEG increases and the alpha-band decreases during hyperventilation in adult humans. These changes resembles those seen during hypoxaemic hypoxia. The changes have been shown to be more marked at PaCO2 levels of 2.0 kPa than at an arterial oxygen-haemoglobin saturation of 60% [14]. A similar comparison has recently been reported in newborn lambs, also showing a more marked effect of hypocapnia [15]. It is not clear if the slowing of the EEG indicates that hypocapnia to a PaCO2 below 2 kPa induces a more severe tissue hypoxia compared to the levels of arterial hypoxaemia tested, or if the differ-

Hypocapnia and the Preterm Brain

ence is due to other effects of CO2 on the EEG. Interestingly, hyperventilation did not affect the EEG in newborn dogs [10]. A study of active hyperventilation in adult humans showed that 100% O2 at 3 atmospheres could reverse the EEG slowing [16]. Hence, severe hypocapnia interferes with cerebro-electrical function, apparently through tissue hypoxia. For comparison, however, it should be noted here that ischaemia induced by arterial occlusion which is sufficiently severe to completely abolish electrical activity does not necessarily produce neuronal loss [17]. Effect on Cerebral Oxygenation Tissue PO2 falls to about one third of normocapnic values with severe hyperventilation as measured by miniaturised PO2 electrodes placed in the cortex [3] or white matter in the adult dog [18], or on the surface of the brain in newborn pigs as measured by phosphorescence imaging [19]. This may in part be due to the fact that hypocapnic alkalosis shifts the oxygen-haemoglobin dissociation curve to the left (the classical Bohr effect), and hence oxygen unloading occurs at a lower PO2. Alkalosis stimulates the synthesis of 2,3-DPG, which will tend to normalise the oxygen-haemoglobin dissociation curve, but this mechanism is less effective for fetal haemoglobin than adult haemoglobin and occurs too late [20] to protect against the effects of acute hypocapnia. Effect on CBF Hypocapnia reduces brain oxygenation mainly by constricting cerebral resistance vessels and reducing CBF. A fall in CBF induced by active or passive hyperventilation was first described by Kety and Schmidt [1] in human adults. The reactivity was reported to be less

Biol Neonate 2001;79:194–200

195

The CBF-CO2 relation is S-shaped. There are few reports of studies going below 1.5 kPa, and it has not been tested if the slope at the lowest levels of CO2 is statistically significantly different from zero, i.e. if a true threshold exists.

Interactions between Hyperventilation and Other Acute Disturbances

Fig. 2. Schematic illustration of the effects of arterial hypotension

and hypoxaemia on CBF-CO2 reactivity. The normal S-shaped relation of CBF to PaCO2 is indicated. At blood pressures below the autoregulatory plateau, i.e. hypotension, the CBF is low and the CBF-CO2 reactivity is very blunted. Noticeably, at a very low PaCO2, the CBF is not further reduced by hypotension. During severe hypoxaemia, CBF is high and hypercapnia does not increase CBF further. The effects of hypoxaemia during severe hypocapnia have not been well investigated.

strong in newborn dogs [21]; the reactivity, however, was 19%/kPa change in CO2 when expressed relative to the level of normocapnic CBF, which is not so different from the normal, adult value of 30%/ kPa. Rather low CBF-CO2 reactivity has also been reported in rat [22] and dog pups [23], and in piglets [24] shortly after birth. We found a reduced CBF-CO2 reactivity on the first day of life in mechanically ventilated infants [25]. In spontaneously breathing preterm infants, however, the CBF reactivity to short-term inhalation of 2–3% CO2 was 59%/kPa [26], and in mechanically ventilated preterm infants of 2 or more days of age, CBF reactivity to changes in ventilator settings was 67%/kPa [95% confidence interval (CI) 13– 146], whereas spontaneous changes in PCO2 were associated with a CBF-CO2 reactivity of 52%/kPa (95% CI 24–86) [27]. When PaCO2 is chronically changed, CBF returns to baseline over a period of many hours, in parallel with normalisation of the pH of cerebrospinal fluid [28]. There is evidence that such adaptation begins within a few hours in newborn piglets [29]. The fact that there is an association between day-to-day variation in CBF and PaCO2 in preterm infants [30], although smaller than the acute CBF-CO2 reactivity, i.e. 22% compared to 30%, suggests that full adaptation may take more than a day in preterm infants, but adaptation has not been studied directly in the preterm human brain.

196

Biol Neonate 2001;79:194–200

Interaction between CBF-CO2 and CBF-Mean Arterial Blood Pressure Normal CBF-mean arterial blood pressure reactivity exhibits a range in which changes in systemic blood pressure do not change CBF (the autoregulatory plateau). CBF increases with pressure above this plateau. This plateau may be narrowed or may completely disappear as a result of cerebral disease or insult (the ‘pressure passive state’). In a report of six adult patients with brain tumours or stroke, short-term hyperventilation was able to restore the normal reactivity to moderate blood pressure increases [31]. Accidental hyperventilation of very-low-birth-weight infants in the delivery room was associated with a lower frequency of intraventricular haemorrhage, which was thought to be a result of a reduction in cerebral hyperperfusion during high blood pressure peaks [32]. At PaCO2 below 4 kPa, decreases in blood pressure from the baseline of 100 mm Hg to 40 mm Hg did not reduce CBF further in adult dogs [33], and at a blood pressure at the lower end of the autoregulatory plateau (50 mm Hg), hyperventilation to 2.7 kPa did not decrease CBF [34, 35]. Similarly, at blood pressures below 38 mm Hg, hyperventilation to PaCO2 of 2 kPa did not change CBF in newborn piglets [36]. This means that hyperventilation does not appear to sensitise the brain to arterial hypotension. Interaction between CBF-CO2 and CBF-O2 Hypoxaemia abolishes the normo- to hypercapnic CBF-CO2 reactivity in the fetal lamb [37], just as it abolishes the pressure autoregulation [33], whereas the evidence of an interaction between hypoxaemic and hypocapnic responses of CBF is conflicting. Hyperoxaemia was found not to increase cortical PO2 in adult dogs during moderate hyperventilation reducing PaCO2 to approximately 3 kPa [3], whereas hyperoxaemia did increase white matter PO2 in dogs during severe hyperventilation to a PaCO2 of 1.8 kPa [18]. Abrupt elevation to high altitude results in spontaneous hyperventilation due to hypoxic chemoreceptor stimulation. The resulting decrease in PaCO2 blunts the hypoxic vasodilatation. Adding CO2 to inspired air in adult humans to obtain normocapnia doubled the CBF increase associated with elevation from sea level to 4,000 m altitude [38]. Similarly, in ducks, the effects of hyperventilation to a PaCO2 of 2 kPa on cerebral blood volume and NADH/NAD+ ratio were not significantly different at a low inspired oxygen fraction (0.05–0.1, simulating high altitude flying) compared to the effects at a normal inspired oxygen fraction (0.21) [39]. A general model of CBF regulation, including the influences of PaO2 and PaCO2, perfusion pressure and the cerebral metabolic rate of oxygen validated in adult dogs, predicts an approximate halving of the CBF-CO2 reactivity during severe hypoxaemia [40]. Hyperventilation to a PaCO2 of 2 kPa during severe hypoxaemia (arterial oxygen-haemoglobin saturation = 50%) in newborn piglets increased the arterio-venous difference for oxygen as much as it did during normoxia [41]. The interactions discussed above are illustrated in figure 2.

Greisen/Vannucci

The Molecular Basis of CBF-CO2 and CBF-O2 Reactivity

The principal part of CBF-CO2 reactivity seems to be mediated through pH, i.e. the H+ concentration. Perivascular pH has a direct effect on the membrane potential of arterial smooth muscle cells, since the extracellular H+ concentration is the main determinant of the potassium conductance of the plasma membrane of cerebral arterial smooth muscle cells and hence the outward K+ current [42]. Therefore, when hypercapnia decreases pH in the perivascular space in the brain before the buffering by HCO+3 occurs, the K+ outflow from the smooth muscle cell increases, and the cell hyperpolarises and relaxes and causes vasodilatation. Furthermore, an increased extracellular – and to a lesser degree intracellular – H+ concentration reduces the conductance of voltage-dependent Ca2+ channels, which also induces relaxation [43]. It is likely that there are other – modulating – mechanisms of CBF-CO2 reactivity, as a block of neuronal nitric oxide (NO) synthetase halves hypercapnic vasodilatation in the adult rat brain [44]. The hypercapnic response could be normalised by adding an NO donor [45], suggesting that a basal level of NO is necessary for the pH to act, at least in some species. The role of prostanoids [46, 47] is less clear. The fact that indomethacin abolishes the CBF-CO2 response in preterm infants [48] may not indicate a role of prostanoids, since ibuprofen does not [49]. The molecular basis of the CBF-O2 reactivity is to a large extent different from that of the CBF-CO2 reactivity. The CBF-O2 reactivity depends on an intact endothelium and NO production. Oxygen also has a selective and direct action on ATP-sensitive K+ channels of the arterial smooth muscle cell membrane [42], whereas pH influences K+ flow through all types of channels. Hypoxia induces tissue lactacidosis as reviewed above, and this indirect pH effect constitutes a point of interaction between CBF-CO2 reactivity and CBF-O2 reactivity. Unfortunately, these insights into the molecular basis of the reactivities do not yet allow quantitative predictions regarding the interaction between hypocapnia and hypoxaemia.

Fig. 3. The uncorrected Bayley mental development index (MDI) at

18 months in a small group of very-low-birth-weight infants who had been inadvertently hyperventilated to a PaCO2 below 2.0 kPa (15 mm Hg) during the first 24 h of life, compared to a group of mechanically ventilated infants with PCO2 remaining above 3.3 kPa (normoventilated), and a group of infants breathing spontaneously. The three children showing signs of cerebral palsy (CP) at 18 months were all in the hyperventilated group (p ! 0.05). (Derived from Greisen et al. [51].)

Hyperventilation and Brain Injury

Clinical Data Fifteen years ago, in a cohort study, we found cerebral palsy in 3 of 7 very-low-birth-weight infants who had been inadvertently hyperventilated to a minimum PaCO2 of 2 kPa (15 mm Hg) or less during the first 24 h of life, compared to none of 7 infants with a minimal PaCO2 above 3.3 kPa (25 mm Hg) (fig. 3). The mean gestational age in the two groups was 29.8 and 31 weeks, respectively [50, 51]. In the same year, a retrospective study of 15 cases of cystic periventricular leucomalacia and 15 matched controls showed that antenatal haemorrhage and PaCO2 values below 3.3 kPa during the first 72 h of life were more frequent in the cases compared to the controls. The mean gestational age was 28.4 weeks [52]. In 1991, in an autopsy study, 12 of 18 newborn infants with pontosubicular necrosis had a PaCO2 of less than 2.7 kPa (20 mm Hg), compared to 1 of 16 newborns without pontosubicular necrosis [53]. The mean gestational age in both groups was 30 weeks. In 1992, in a case-control study of 134 surviving very-low-birth-weight infants with major brain ultrasound abnormalities and 117 controls, one or more PaCO2 values below 2.2 kPa (17 mm Hg) during the first 72 h was associated with large periventricular cysts with an odds ratio of 4.2 (95% CI 1.5–11.9) and with cerebral palsy at 19 months to 8 years of age with an odds ratio of 4.1

(95% CI 1.4–11.9) [54]. Three additional studies support the association between hypocapnia and cystic leucomalacia [55–57], whereas one retrospective study of a consecutive group of 200 very-low-birthweight infants with a median gestational age of 29 weeks demonstrated no association between prolonged flare or precystic or cystic periventricular leucomalacia and PaCO2 below 3 kPa, but rather with PaCO2 above 7 kPa in the first days of life [58]. It is unclear whether the different definition of periventricular leucomalacia explains this discrepancy. It must be stressed that several of the studies reviewed above, as well as studies not reviewed here, have identified other pre- and neonatal factors which are important predictors of periventricular leucomalacia and/or cerebral palsy. In contrast to the apparent association between hypocapnia and brain injury in preterm infants, follow-up of full-term infants treated with hyperventilation to control pulmonary hypertension has suggested only an inconsistent risk of brain injury, considering the severity of the conditions treated with hyperventilation. In one study, all eight infants who had PaCO2 below 2 kPa and who were followed to 12 months of age or more were neurologically normal and had normal developmental scores [59]. In another study, 3 of 11 infants had

Hypocapnia and the Preterm Brain

Biol Neonate 2001;79:194–200

197

Fig. 4. Composite diagram showing the severity of hypoxic-isch-

aemic brain damage at varying concentrations of carbon dioxide in immature rats. Seven-day-old postnatal rats were subjected to cerebral hypoxia-ischaemia breathing either 0, 3, 6 or 9% carbon dioxide. The bars represent the numbers of brains as a percentage of the total in each experimental group.

subtle signs of cerebral palsy and 2 had subnormal developmental scores [60]. In a later study of 11 infants with PaCO2 below 2.7 kPa examined at 12 months of age or more, 1 of 11 had mild hemiparesis, and all 9 infants who were tested had normal developmental scores [61]. In a prospective follow-up of 21 infants for 12 months, however, severe neurologic impairment was documented in 4 and mild to moderate impairment in another 7. The risk of impairment was associated with the duration of hyperventilation [62]. Controlled hyperventilation is used in the neuro-intensive care of older children and adults, typically achieving a PaCO2 of 3.3 kPa, because this reduces intracranial pressure and potentially reduces the risk of cerebral herniation, even though it may worsen cerebral ischaemia either globally [63] or locally in areas already marginally ischaemic due to local injury [64]. The only randomised controlled study, involving 40 patients in each group, failed to show any statistically significant difference at follow-up after 1 year [65]. The chance of a good outcome in the hyperventilated low-risk patients was 44%, compared to 57% in the normoventilated low-risk patients. Interestingly, in a third group treated with hyperventilation and intravenous base (tromethamin), the chance of a good outcome was 67%. Experimental Data The clinical setting of premature infants requiring mechanical ventilation for respiratory distress syndrome, which may lead to hypocapnia in association with systemic hypoxaemia, has important similarities to the alterations in systemic acid-base balance which occur in immature rats undergoing hypoxic-ischaemic brain damage [66]. When 7-day-old rats are exposed to 8% O2, the hypoxaemia leads to lactacidaemia and hyperventilation. The resulting hypocapnia completely compensates for the metabolic acidaemia, and blood pH remains normal. To ascertain the contribution of hypocapnia to

198

Biol Neonate 2001;79:194–200

the hypoxic-ischaemic brain damage ipsilateral to the ligated common carotid artery in this animal model (the Levine preparation [67]), CO2 was added to the inspired air [68]. Without added CO2, PaCO2 averaged 3.5 kPa. In rat pups exposed to 3% CO2, PaCO2 averaged 5.3 kPa, while in those animals exposed to 6 or 9% CO2, PaCO2 averaged 7.6 and 8.9 kPa, respectively. Arterial PO2 was essentially identical in all the hypoxic groups. The neuropathologic findings demonstrated a protective influence of CO2 (fig. 4). Specifically, 3% CO2 was better than none, and 6% better than 3%, whereas 9% did not give any further protection. These are the first animal data supporting the clinical association between hypocapnia and brain damage in preterm infants reviewed in the first section of this paper. Since hypocapnia aggravated and hypercapnia protected the immature rat from hypoxic-ischaemic brain damage, further investigations were conducted to ascertain CBF and metabolic correlates in the same experimental protocol [69]. CBF at 90 min of hypoxia-ischaemia was 63% of that of controls in the hypocapnic rat pups. CBF was better preserved (80–85% of controls) in the normo- and hypercapnic animals. Cerebral glucose utilisation during hypoxia-ischaemia without supplemental CO2 was minimally increased over that of controls, whereas the rate nearly doubled in the normo- and hypercapnic animals. Brain glucose concentrations during hypoxia-ischaemia were lowest and lactate highest in the hypocapnic rat pups compared to the normo- and hypercapnic animals, indicating that glucose was consumed oxidatively in the latter groups rather than by anaerobic glycolysis, as apparently occurred in the hypocapnic animals. Brain ATP and phosphocreatine during hypoxia-ischaemia were nearly completely exhausted in the hypocapnic rats but were similar to the control values in the other two groups. CSF glutamate, assumed to be a reflection of the brain extracellular fluid concentration, was increased by 60% in the hypo- and normocapnic rats at 2 h of hypoxia-ischaemia but was increased by only 14% in the hypercapnic animals. The data indicate that during hypoxia-ischaemia in the immature rat, CBF is better preserved during normo- and hypercapnia; the greater oxygen delivery supports cerebral glucose utilisation and oxidative metabolism for better maintenance of tissue highenergy phosphate reserves compared with hypocapnic animals. An inhibition of glutamate secretion into the synaptic cleft and its attenuation of NMDA receptor activation would further protect the hypercapnic animal from hypoxic-ischaemic brain damage [13].

Possible Reasons for the Apparent Specific Vulnerability of Preterm Infants

In the follow-up studies reviewed above, the values of PaCO2 used to define hyperventilation were somewhat lower for the preterm infants compared to the other patients. The differences were minor, however, and due to clinical variability there must be a great overlap of values among the studies. In clinical practice, episodes of hypoxaemia, e.g. during endotracheal suctioning, may occur more often in inadvertently hyperventilated preterm infants who are mechanically ventilated because of cardio-respiratory illness than in adult patients who are hyperventilated in the course of neuro-intensive care. Hypoxaemia, on the other hand, is likely to be more frequent and more severe in full-term infants who are hyperventilated for persistent pulmonary hypertension. Therefore, a maturational difference in vulnerability to hypocapnia is indicated.

Greisen/Vannucci

Although the main regulation of CBF takes place at the level of the intraparenchymal arteries and the arterioles, extracranial and pial arteries contribute significantly to cerebrovascular resistance. The is the concept of dual regulation [70]. Hypocapnia-induced resistance in the proximal arteries would not be within the reach of local acidosis and hence such an escape mechanism would not be effective. In adult man, intracranial arteries of 0.5–1.0 mm exhibit moderate CO2 reactivity [71], and in adult cats and dogs, pial arteries do [72]. In fetal lambs, even the carotid artery dilates markedly during asphyxia [73], indicating reactivity to hypoxia or hypocapnia or both. It may be hypothesised that the preterm human brain is placed at risk in two ways. Firstly, inhibition of cerebral arterial vasoconstriction may not be well developed before term [74]; this may explain why damage is seen only in the immature brain. Secondly, the shorter distributor arteries needed by the animal brain because of its relatively smaller size compared to the human brain may explain why even immature animals exposed to simple hyperventilation do not develop evidence of brain damage. This needs further study.

Conclusion

There is no convincing evidence in adult humans or perinatal animals that hyperventilation per se causes a reduction in cerebral oxygen and substrate delivery so severe as to interfere significantly with normal energy metabolism. The evidence indicates that during severe hypocapnia, when CBF falls below a critical level, stimulation of anaerobic glycolysis produces cellular metabolic acids, which would tend to relax cerebral vessels, i.e. an escape mechanism. So, there are three possible explanations for the statistical association between hypocapnia and brain damage and developmental deficit in preterm infants: (1) hypocapnia sensitises the brain to hypoxaemia, which may be less easily detected in preterm infants; (2) the escape mechanism in the human preterm infant is less effective, or (3) the relation between hypocapnia and brain damage in preterm infants is not a causal one.

References 1 Kety SS, Schmidt CF: The effects of active and passive hyperventilation on cerebral blood flow, cerebral oxygen consumption, cardiac output, and blood pressure of normal young men. J Clin Invest 1946;25:107–119. 2 Pryds O: Control of cerebral circulation in the high-risk neonate. Ann Neurol 1991;30:321–329. 3 Sugioka K, Davis DA: Hyperventilation with oxygen – a possible cause of cerebral hypoxia. Anesthesiology 1960;21:135–143. 4 Granholm L, Siesjö BK: The effects of hypercapnia and hypocapnia upon the cerebrospinal fluid lactate and puruvate concentrations and upon the lactate, puruvate ATP, ADP, phosphocreatine and creatine concentration of cat brain tissue. Acta Physiol Scand 1969;75:257–266. 5 MacMillan V, Siesjö BK: The influence of hypocapnia upon intracellular pH and upon some carbohydrate substrates, amino acids and organic phosphates in the brain. J Neurochem 1973;21: 1283–1299. 6 Plum F, Posner JB: Blood and cerebrospinal fluid lactate during hyperventilation. Am J Physiol 1967;212:864–870. 7 Kogure K, Busto R, Matsumoto A, Scheinberg P, Reinmuth OM: Effect of hyperventilation on dynamics of cerebral energy metabolism. Am J Physiol 1975;228:1862–1867. 8 Young RSK, Yagel SK: Cerebral physiological and metabolic effects of hyperventilation in the neonatal dog. Ann Neurol 1984;16:337–342. 9 Rosenberg AA: Response of the cerebral circulation to profound hypocarbia in neonatal lambs. Stroke 1988;19:1365–1370. 10 Reuter JH, Disney TA: Regional cerebral blood flow and cerebral metabolic rate of oxygen during hyperventilation in the newborn dog. Pediatr Res 1986;20:1102–1106. 11 Hansen NB, Nowicki PT, Miller RR, Malone T, Bickers RG, Menke JA: Alteration in cerebral blood flow and oxygen consumption during prolonged hypocarbia. Pediatr Res 1986;20:147–150. 12 Graham EM, Apostolou M, Mishra OP, DelivoriaPapadopoulos M: Modification of the N-methyl-Daspartate (NMDA) receptor in the brain of newborn piglets following hyperventilation induced ischemia. Neurosci Lett 1996;218:29–32.

Hypocapnia and the Preterm Brain

13 Tombaugh CG, Sapolsky RM: Evolving concepts about the role of acidosis in ischaemic neuropathology. J Neurochem 1993;61:793–803. 14 Van der Worp HB, Kraaier V, Wieneke GH, Van Huffelen AC: Quantitative EEG during progressive hypocarbia and hypoxia. Hyperventilationinduced EEG changes reconsidered. Electroencephalogr Clin Neurophysiol 1991;79:335–341. 15 van de Bor M, Meinesz J, Benders M, Steendiijk P, Cardozo RL, van Bel F: Electrocortical brain activity during repeated hypoxia and hypocarbia in newborn lambs (abstract). Pediatr Res 1996;40: 520. 16 Cohen PJ, Reivich M, Greenbaum L: The electroencephalogram of awake man during hyperventilation: Effects of oxygen at three atmospheres (absolute) pressure. Anesthesiology 1966;27:211–212. 17 Astrup J: Energy-requiring cell functions in the ischemic brain. J Neurosurg 1982;56:482–497. 18 Kennealy JA, McLennan JE, Loudon RG, McLaurin RL: Hyperventilation-induced cerebral hypoxia. Am Rev Respir Dis 1980;122:407–411. 19 Wilson DF, Pastuszko A, DiGiacomo JE, Pawlowski M, Schneiderman R, Delivoria-Papadopoulos M: Effect of hyperventilation on oxygenation of the brain cortex in newborn piglets. J Appl Physiol 1991;70:2691–2696. 20 Wimberley PD: Fetal hemoglobin, 2,3-diphosphoglycerate and oxygen transport in the newborn preterm infant. Scand J Clin Lab Invest Suppl 1982; 160:92–104. 21 Hernandez MJ, Brennan RW, Vannucci RC, Bowman GS: Cerebral blood flow and oxygen consumption in the newborn dog. Am J Physiol 1978; 234:R209–R215. 22 Purin VR, Syutkina EV: Local cerebral blood flow velocity in the newborn rat in normo- and hypercapnia. Bull Exp Biol Med 1977;84:139–141. 23 Shapiro HM, Greenberg JH, Naughton KVH, et al: Heterogeneity and cerebral metabolic rate of oxygen during hyperventilation in the newborn dog. J Appl Physiol 1980;49:113–118. 24 Haaland K, Karlsson B, Skovland E, Lagerkrantz H, Thoresen M: Postnatal development of the cerebral blood flow velocity response to changes in CO2 and mean arterial blood pressure in the piglet. Acta Paediatr 1995;84:1414–1420.

25 Pryds O, Greisen G, Lou H, Friis-Hansen B: Heterogeneity of cerebral vasoreactivity in preterm infants supported by mechanical ventilation. J Pediatr 1989;115:638–645. 26 Leahy FAN, Cates D, MacCallum M, Rigatto H: Effect of CO2 and 100% oxygen on cerebral blood flow in preterm infants. J Appl Physiol 1980;48: 468–472. 27 Greisen G, Trojaborg W: Cerebral blood flow, PaCO2 changes, and visual evoked potentials in mechanically ventilated, preterm infants. Acta Paediatr Scand 1987;76:394–400. 28 Lassen NA: Control of cerebral circulation in health and disease. Circ Res 1974;34:749–760. 29 Hansen NB, Nowicki PT, Miller RR, Malone T, Bickers RG, Menke JA: Alterations during prolonged hypocarbia. Pediatr Res 1986;20:147–150. 30 Pryds O, Greisen G: Effect of PaCO2 and haemoglobin concentration on day to day variation of CBF in preterm neonates. Acta Paediatr Scand Suppl 1989;360:33–36. 31 Paulson OB, Olesen J, Christensen MS: Restoration of autoregulation of cerebral blood flow by hypocapnia. Neurology 1972;22:286–293. 32 Lou HC, Phibbs RH, Wilson SL, Gregory GA: Hyperventilation at birth may prevent early periventricular haemorrhage (letter). Lancet 1982; i:1407. 33 Häggendal E, Johansson B: Effects of arterial carbon dioxide tension and oxygen saturation on cerebral blood flow autoregulation in dogs. Acta Physiol Scand Suppl 1965;258:27–53. 34 Harper AM, Glass HI: Effects of alterations in the arterial carbon dioxide tensions on the blood flow through the cerebral cortex at normal and low arterial blood pressure. J Neurol Neurosurg Psychiatry 1965;28:449–452. 35 Artru AA, Colley PS: Cerebral blood flow responses to hypocapnia during hypotension. Stroke 1984;15:878–883. 36 Whitelaw A, Karlson BR, Haaland K, Dahlin I, Steen PA, Thoresen M: Hypocapnia and cerebral ischaemia in hypotensive newborn piglets. Arch Dis Child 1991;66:1110–1114. 37 Kjellmer I, Karlsson K, Olsson T, Rosén KG: Cerebral reactions during intrauterine asphyxia in the sheep. I. Circulation and oxygen consumption in the fetal brain. Pediatr Res 1974;8:50–57.

Biol Neonate 2001;79:194–200

199

38 Lassen NA: Increase of cerebral blood flow at high altitude. Its possible relation to AMS. Int J Sports Med 1992;13:S47–S48. 39 Bickler PE, Koh SO, Severinghaus JW: Effects of hypoxia and hypocapnia on brain redox balance in ducks. Am J Physiol 1989;257:R132–R135. 40 Wilson DA: On the contribution of respiratory gases to cerebral autoregulation; in Harrison DK, Delpy DT (eds): Oxygen Transport to Tissue XIX. New York, Plenum Press, 1998, pp 695–702. 41 Brun NC, Moen A, Saugstad OD, Greisen G: Cerebro-venous oxygen content is further reduced by hyperventilation during hypoxemia in newborn piglets (abstract). Pediatr Res 1995;38:427. 42 Pearce WJ, Harder DR: Cerebrovascular smooth muscle and endothelium; in Mraovitch S, Sercombe R (eds): Neurophysiological Basis of Cerebral Blood Flow Control: An Introduction. London, John Libbey, 1996, pp 153–158. 43 Aalkjaer C, Poston L: Effects of pH on vascular tension: Which are the important mechanisms? J Vasc Res 1996;33:347–359. 44 Wang Q, Pelligrino DA, Baughman VL, Koenig HM, Albrecht RF: The role of neuronal nitric oxide synthetase in regulation of cerebral blood flow in normocapnia and hypercapnia in rats. J Cereb Blood Flow Metab 1995;15:774–778. 45 Iadecola C, Zhang F: Permissive and obligatory roles of NO in cerebrovascular responses to hypercapnia and acetylcholine. Am J Physiol 1996;271: R990–R1001. 46 Wagerle LC, Mishra OP: Mechanism of CO2 response in cerebral arteries of the newborn pig: Role of phospholipase, cyclooxygenase, and lipooxygenase pathways. Circ Res 1988;62:1019–1026. 47 Rama GP, Parfenova H, Leffler CW: Protein kinase Cs and tyrosine kinases in permissive action of prostacyclin on cerebrovascular regulation in newborn pigs. Pediatr Res 1996;41:83–89. 48 Edwards AD, Wyatt JS, Ricardsson C, Potter A, Cope M, Delpy DT, Reynolds EOR: Effects of indomethacin on cerebral haemodynamics in very preterm infants. Lancet 1992;i:1491–1495. 49 Patel J, Roberts I, Azzopardi D, Hamilton P, Edwards AD: Cerebral haemodynamic effects of indomethacin and ibuprofen during treatment for patent ductus arteriosus (abstract). Pediatr Res 1997;42:389. 50 Greisen G, Munck H, Lou H: May hypocarbia cause ischaemic brain damage in the preterm infant? (letter). Lancet 1986;ii:460.

200

51 Greisen G, Munck H, Lou H: Severe hypocarbia in preterm infants and neurodevelopmental deficit. Acta Paediatr Scand 1987;76:401–404. 52 Calvert SA, Hoskins EM, Fong KW, Forsyth SC: Etiological factors associated with the development of periventricular leucomalacia. Acta Paediatr Scand 1987;76:254–259. 53 Hashimoto K, Takeuchi Y, Takashima S: Hypocarbia as a pathogenic factor in pontosubicular necrosis. Brain Dev 1991;13:155–157. 54 Graziani LJ, Spitzer AR, Michell DG, Merton DA, Stanley C, Robinson N, McKee L: Mechanical ventilation in preterm infants: Neurosonographic and developmental studies. Pediatrics 1992;90:515– 522. 55 Ikonen RS, Janas MO, Koivikko MJ, Laippala P, Kuusinen EJ: Hyperbilirubinemia, hypocarbia and periventricular leucomalacia in preterm infants: Relationship to cerebral palsy. Acta Paediatr Scand 1992;81:802–807. 56 Iida K, Takashima S, Takeuchi Y: Etiologies and distribution of neonatal leukomalacia. Pediatr Neurol 1992;8:205–209. 57 Wiswell TE, Graziani LJ, Kornhauser MS, Stanley C, Merton DA, McKee L, Spitzer AR: Effects of hypocarbia on the development of cystic periventricular leukomalacia in premature infants treated with high-frequency jet ventilation. Pediatrics 1996;98:918–924. 58 Trounce JQ, Shaw DE, Levene MI, Rutter N: Clinical risk factors and periventricular leucomalacia. Arch Dis Child 1988;63:17–22. 59 Brett C, Deckle M, Leonard CH, Clark C, Sniderman S, Roth R, Ballard R, Clyman R: Developmental follow-up of hyperventilated neonates. Preliminary observations. Pediatrics 1981;68:588– 591. 60 Bernbaum JC, Russell P, Sheridan PH, Gewitz MH, Fox WW, Peckham GJ: Long term follow-up of newborns with persistent pulmonary hypertension. Crit Care Med 1984;12:579–583. 61 Ferrara B, Johnson DE, Chang PN, Thompson TR: Efficacy and neurological outcome of profound hypocapnic alkalosis for the treatment of persistent pulmonary hypertension in infancy. J Pediatr 1984;105:457–461. 62 Bifano EM, Pfannenstiel A: Duration of hyperventilation and outcome in infants with persistent pulmonary hypertension. Pediatrics 1988;81:657– 661.

Biol Neonate 2001;79:194–200

63 Obrist WD, Langfitt TW, Jaggi JL, Cruz J, Gennarelli TA: Cerebral blood flow and metabolism in comatose patients with acute head injury. Relationship to intracranial hypertension. J Neurosurg 1984;61:241–253. 64 Stringer WA, Hasso AN, Thompson JR, Hinshaw DB, Jordan KG: Hyperventilation-induced cerebral ischemia in patients with acute brain lesions: Demonstration by xenon-enhanced CT. AJNR Am J Neuroradiol 1993;14:475–484. 65 Muizelaar JP, Marmarou A, Ward JD, Kontos HA, Choi SC, Becker DP, Gruemer H, Young HF: Adverse effects of prolonged hyperventilation in patients with severe head injury: A randomised clinical trial. J Neurosurg 1991;75:731–739. 66 Welch FA, Vannucci RC, Brierley JB: Columnar alterations of NADH fluorescence during hypoxiaischemia in immature rat brain. J Cereb Blood Flow Metab 1982;2:221–228. 67 Rice JE, Vannucci RC, Brierley JB: The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann Neurol 1981;9:131–141. 68 Vannucci RC, Towfighi J, Heitjan DF, Brucklacher RM: Carbon dioxide protects the perinatal brain from hypoxic-ischemic damage: An experimental study in the immature rat. Pediatrics 1995; 95:868–874. 69 Vanucci RC, Brucklacher RM, Vannucci SJ: Effect of carbon dioxide on cerebral metabolism during hypoxia-ischemia in the immature rat. Pediatr Res 1997;42:24–29. 70 Faraci FM, Heistad DD: Regulation of large cerebral arteries and cerebral microvascular pressure. Circ Res 1989;66:8–17. 71 Huber P, Handa J: Effect of contrast material, hypercapnia, hyperventilation, hypertonic glucose and papaverine on the diameter of the cerebral arteries. Angiographic determination in man. Invest Radiol 1967;2:17–32. 72 Busija DW, Heistad DD, Marcus ML: Continuous measurement of cerebral blood flow in anesthetized cats and dogs. Am J Physiol 1981;241:H228– H234. 73 Malcus P, Kjellmer I, Lingman G, Marsal K, Thiringer K, Rosén K: Diameters of the common carotid artery and aorta change in different directions during acute asphyxia in the fetal lamb. J Perinat Med 1991;19:259–267. 74 Wagerle LC, Moliken W, Russo P: Nitric oxide and beta-adrenergic mechanisms modify contractile responses to norepinephrine in ovine fetal and newborn cerebral arteries. Pediatr Res 1995;38:237– 242.

Greisen/Vannucci

Biol Neonate 2001;79:201–204

Chorioamnionitis and Fetal/Neonatal Brain Injury Paolo Toti Claudio De Felice Institute of Pathology and Department of Preventive Pediatrics and Neonatology, University of Siena, Siena, Italy

Key Words Chorioamnionitis W Brain injuries W Infant/newborn diseases W Pregnancy complications (infectious) W Placental histology

Abstract Chorioamnionitis (CA) is the leading cause of preterm birth and neonatal complications. Even in the absence of a proven infection, fetuses and neonates present a systemic inflammatory response which can be identified by radiological and morphological examination of the thymus. The frequent occurrence of brain injury in neonates with CA is probably linked to systemic, unspecific mechanisms which have not yet been completely clarified. Only by relating placental pathology to clinical evaluation of the newborn will it be possible to achieve a better understanding of these infections and to reduce long-term morbidity and mortality. Copyright © 2001 S. Karger AG, Basel

Introduction

Chorioamnionitis (CA) is the most frequent cause of fetal death during the second half of pregnancy [1, 2]. It also represents the leading cause of preterm birth, perinatal complications and fetal and neonatal death [3–6]. In fact, histologic signs of CA can be detected in more than 50% of women who deliver prematurely, and most patients have no clinical signs of infection [3, 7, 8]. Here, we briefly summarize our data and those from the literature, indicating that CA is also associated with perinatal brain injury and represents the most frequent cause of acquired brain damage occurring in the perinatal period.

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0201$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

Histological CA

CA is a pathological entity that is easily identifiable by histological examination after delivery [1, 2]. It is characterized by the presence of acute inflammatory infiltrate in the placenta, membranes and umbilical cord. However, the presence of leukocytes in the membranes has been misinterpreted for a long time, as a chemotactic effect of meconium was assumed [9]. It is now clear that CA causes fetal suffering that can frequently induce meconium discharge [1]. CA is an acute inflammatory reaction in which principally polymorphonucleate leukocytes participate. Eosinophils are found at times, but only after protracted infection; macrophages may participate to a variable extent and plasma cells are generally absent in the membrane. Leukocytes come from two sources: the intervillous space (which are thus maternal) and fetal surface blood vessels. In early gestation, especially prior to the 20th week of gestation, the leukocytes are mainly of maternal origin; by midtrimester, the fetus begins to be capable of producing leukocytes that participate in the inflammatory response [1]. CA may cause preterm birth. Induction of labor may be caused by several mechanisms, indirectly due to fetal suffering, by the liberation of prostaglandins and chemokines in the decidua associated with decidual colonization, granulocyte infiltration and decidua necrosis, or by premature rupture of membranes (PROM) [3, 10–12]. In fact, the presence of polymorphonucleates reduces the resistance of membranes and may lead to premature rupture. Among preterm labor pregnancy placentas delivered within 1 h of membrane rupture, 61% had membrane inflammation [13, 14]. We can therefore conclude that inflammation frequently precedes PROM and preterm labor. The presence of polymorphonucleates in the membranes does not necessarily mean fetal infection, yet 36% of cases of preterm labor pregnancies which delivered within 1 h of membrane rupture had fetal vasculitis [13]. However, after membrane rupture, severe fetal infection occurs within 24–48 h. Fetal infection may also occur without membrane rupture, since germs may colonize the amniotic fluid

Paolo Toti, MD Institute of Pathological Anatomy and Histology University of Siena, Via delle Scotte, 6 I–53100 Siena (Italy) Tel. +39 0577 233 240, Fax +39 0577 48268, E-Mail [email protected]

and reach the fetus through the skin and mucosae of the respiratory and digestive tracts. When chronic aspiration pneumonia occurs, the lungs contain granulocytes intermixed with amnion squames, along with lymphocytes and plasma cells. Moreover, a strong correlation exists between CA and otitis media, since the widely patent eustachian tubes of premature infants are believed to be a possible point of entry of infected aspirated amniotic fluid [1]. Bacteria may also reach the fetus via the bloodstream, entering the fetal circulation through the villi capillaries (in cases of acute villitis and intervillositis), via the large vessels of the fetal face (chorionitis) and the umbilical cord vessels (funisitis). CA may induce the presence of maternal symptoms. The most important clinical symptoms of CA are high temperature, uterine tenderness, fetal tachycardia 1 160 beats/min, maternal tachycardia 1 100 beats/min, foul-smelling vaginal discharge, white blood cell count 1 12,500/mm3 with 1 85% granulocytes and C-reactive protein 1 0.8 mg/dl [3, 15, 16]. The criteria for a clinical diagnosis of CA include the presence of at least two of the symptoms mentioned above. However, the presence of such unspecific symptoms does not necessarily indicate CA. In our experience, only slightly more than half (54%) of pregnancies with clinical CA had histological CA, and clinical evidence identified histological CA with a sensitivity of 24% and specificity of 87% (positive and negative likelihood ratios of 1.93 and 0.86, respectively) [17]. Indeed, inflammation of the fetal adnexa is a significantly underdiagnosed condition, on the basis of clinical criteria alone, since only 24% of mothers with CA have clinical signs [17, 18]. Histological evaluation of the placenta and membranes is an easy and simple method that nowadays is mandatory to diagnose CA, in order to study the effects of inflammatory reaction in the preterm and term baby. However, the etiology and pathogenesis are not completely understood. Bacteria may invade the uterus by migration from the abdominal cavity through the fallopian tubes, inadvertent needle contamination during amniocentesis or chorionic villus sampling, hematogenous spread through the placenta, or passage through the cervix from the vagina [19]. The pathological findings that the degree of inflammation is most significant in the membrane portion in the proximity of the point of rupture, and that it is twin A that invariably has CA or whose membranes are more severely inflamed, indicate the ascending nature of the infection [1]. Several clinical studies also support the ascending nature of this infection. For instance, infections of the lower genital and urinary tract such as vaginosis and vaginitis are associated with increased risk for preterm delivery, which is reduced by early identification and treatment [11, 20, 21]. Antimicrobial therapy when used in the expectant management of preterm PROM is associated with prolongation of pregnancy and a reduction in the diagnosis of maternal and infant morbidity [22]. Histopathological evaluation is not, however, able to clarify the etiology of such inflammatory infiltrate. In our files, we have positive placental cultures in 64% of cases with clinical CA, confirmed by histological examination, whereas when there were no maternal symptoms, cultures showed positivity in only 39% of cases. The most commonly identified bacteria are vaginal organisms of relatively low virulence, such as Ureaplasma uralyticum, group B streptococci and bacteroides species, as in other reports [19]. The presence of placental inflammation without demonstrable germs is a topic not yet completely clarified. It is also not clear when the bacteria ascend from the vagina. Intrauterine infection may occur quite early in pregnancy and

202

Biol Neonate 2001;79:201–204

remain undetected for months [19]. For example, the detection of U. urealyticum or high levels of interleukin-6 in the amniotic fluid obtained for chromosomal analysis at 15–18 weeks of gestation is associated with spontaneous preterm delivery [19, 23–25]. These chronic upper genital infections are not associated with fever, a tender uterus or peripheral blood leukocytosis [19], yet may be responsible for symptoms such as vaginal bleeding [26].

CA and Fetal and Neonatal Modification

As discussed above, the presence of CA does not necessarily mean fetal and neonatal infection. However, an increasing body of evidence suggests that CA induces several identifiable modifications both in fetuses and neonates, due to an unspecific inflammatory response. Modifications in inflammatory cytokine levels, such as interleukin-6, tumor necrosis factor-·, interleukin-6, interleukin-8, interleukin-1ß, interleukin-1· and granulocyte colony-stimulating factor, have been described in mothers, fetuses, preterm and term neonates and amniotic fluid [10, 19, 27–30], associated with CA. In the meantime, autopsy studies have shown activation of the fetal immune system, such as the augmentation of fetal granulopoiesis [31]. We have shown that neonatal morphological modifications associated with CA are also identifiable in vivo. In fact, a conventional anteroposterior supine chest X-ray performed at birth (median 2 h, range 0.2–6 h) is able to detect the cardiothymic silhouette and therefore is able to measure the width of the thymus at the level of the carina. The ratio between the transverse diameter of the cardiothymic image and that of the thorax (CT/T) showed a significant decrease in thymic size in very-low-birth-weight neonates with CA [32]. Such an easy and simple method is sensitive enough to be effective in distinguishing between infants born to mothers with CA and to controls. An autopsy study performed on fetuses and neonates with CA confirmed the radiological data. The main morphological findings included: (1) decreased organ volume; (2) a reduced corticomedullary ratio; (3) significant changes in the relationship between thymic parenchyma and thymic interstitial tissue with resulting increased organ complexity; (4) severe reduction of thymocytes, and (5) other degenerative processes such as monocyte/macrophage infiltration of Hassall’s bodies. Similar morphological data, corresponding to acute thymic involution, were found in preterm and term neonates who had died because of proven sepsis [33]. The abovereported data (i.e. reduced CT/T ratio in neonates born to mothers with CA, confirmed by autopsy morphological study) indicate that the morphology of the thymus could be valuable as a highly informative easy-to-get index of fetal stress in clinical evaluation of the newborn. The molecular mechanisms and biological mediators leading to thymus shrinkage are not yet completely clarified. However, thymus shrinkage is of a reactive nature, and it is known to be reversible if the baby survives. Long-term effects of thymus shrinkage and rebound on the developing immune system are not known at present. The morphological alterations of the thymus described above only represent one aspect of severe multiorgan disease occurring in histologically determined CA. The systemic inflammatory response is an independent risk factor for the occurrence of severe neonatal morbidity, defined by the presence of respiratory distress syndrome, suspected or proven neonatal sepsis, pneumonia, bronchopulmonary dysplasia, intraventricular hemorrhage, periventricular leukomalacia or necrotizing enterocolitis [30].

Toti/De Felice

Brain Damage

The prenatal and perinatal development of the human brain is a complex process during which countless neurons, axonic fibers and blood vessels are transported, structurally differentiated and functionally interconnected [34]. These evolving processes are interdependent and damage to one will have developmental repercussions on the subsequent differentiation of the others. Therefore, the vulnerability of the developing brain and the pathogenesis of ensuing neurological sequelae remain poorly understood. Only by studying different perinatally acquired neocortical lesions in prematurely born infants who survived them for different periods of time (hours, days, weeks or years), may it be possible to indirectly reconstruct some aspects of their pathogenesis, subsequent developmental impact and outcome. Several lines of evidence indicate that CA is associated with severe acquired brain damage, both in the prenatal and early postnatal periods. For instance, evidence suggests that 70–80% of cases of cerebral palsy are due to prenatal factors and that birth asphyxia plays a relatively minor role [35, 36]. Prematurity itself does not appear to be an independent risk factor, since preeclampsia has been shown to be strongly protective against cerebral palsy in preterm infants [37, 38]. Preterm infants whose amnion is acutely inflamed are at a 3- to 4-fold greater risk than their peers of developing intraventricular hemorrhage [39]. A meta-analysis review has recently concluded that CA is a risk factor for both cerebral palsy and cystic periventricular leukomalacia, although it is highlighted that using a clinical definition of CA produces heterogeneous results [36]. A recent study by our group has examined this issue; we have classified a cohort of preterm and term neonates according to the presence/absence of clinical and/or histological signs of CA. After checking for potential confounders, histological but not clinical CA was found to be significantly associated with the established neurological endpoints (p ! 0.001). Neonates with histologically proven CA showed an increased risk of periventricular echodensity [¯2 = 40.1, p ! 0.0005; unadjusted relative risk (RR) 4.6, 95% confidence interval (CI) 2.6–8.2], white matter echolucency (¯2 = 30.8, p ! 0.0005; unadjusted RR 6.9, 95% CI 2.4–19.6), ventriculomegaly (¯2 = 31.7, p ! 0.0005; unadjusted RR 4.8, 95% CI 2.1–10.7), intraventricular hemorrhage 63 (¯2 = 58.1, p ! 0.0001; unadjusted RR 6.9, 95% CI 3.3–14.2) and seizures (¯2 = 15.4, p = 0.0015; unadjusted RR 3.6, 95% CI 1.4–8.9). CA and brain damage remained significantly associated, independently of gestational age (p ! 0.001). Conversely, a significant inverse relationship was found between negative histology for CA and the study endpoints (odd ratios ^0.51, 95% CI upper limits ^0.82, p ^ 0.005). The outcome of babies whose mothers had clinical or subclinical CA did not show any differences [17]. The brain damage may happen at birth, in the early neonatal period, and also in utero during the pregnancy. In the latter case, brain injury may be morphologically indistinguishable from a malformation. Polymicrogyria is a cortical abnormality usually classified among neuron migration disorders that may be found in a number of congenital autosomal recessive syndromes associated with multiorgan abnormalities (e.g. Zellweger syndrome, congenital muscular dystrophies), yet many cases appear to be sporadic [40–42]. In these cases, it has been hypothesized that intrauterine brain laminar destruction may result in polymicrogyria [33, 43, 44]. The question of whether this represents a primary malformation or rather a destructive lesion with secondary malformation continues to be debated. Mari`n-Padil-

CA and Brain Injury

la [34], in studying the developing brain by rapid Golgi preparation, has shown that periventricular hemorrhage, which is more frequent in CA, is followed by the degeneration of radial glial fibers. The destruction of radial glial fibers will result in impairment of the migration of cells from the germinal matrix to the cortex, causing nonspecific and variable cytoarchitectural alterations of the overlying gray matter, such as displacement and disorientation of neurons, abnormal neurons, laminar disorganization anomalies in the intrinsic circuitry, vascular anomalies, focal heterotopias, foci of ulegyria, microgyria, agyria/pachigyria and leptomeningeal heterotopias [34]. We have also studied the possibility of a nonlethal infection in the fetal adnexa eventually giving rise to distant brain defects, such as micropolygyria. We have found the presence of different histological alterations in the primitive cortical architecture, both isolated and combined (undulation of the cortical ribbon, untimely cortical folding/molecular layer fusion and neuronal loss), associated with severe inflammatory lesions of the fetal adnexa. Such damage was usually restricted to the cortex, yet some fetuses also showed hemorrhage in the periventricular area. Brain anomalies were found in 78% of cases, while they were never present in healthy controls. In all the cases, severe CA was the cause of fetal death and abortion; we hypothesize that the brain lesions we describe could have eventually given rise to cortex polymicrogyria had the fetuses survived [45]. The cortical lesion we have described, represented by granular invasion of the molecular layer, shows striking morphological resemblance to ‘status verrucosus simplex’, an aspect first described by Retzius in 1895 (quoted by Larroche [46]). He proposed that status verrucosus simplex represents a transient, physiological stage of the developing cortical cell layout. Since status verrucosus simplex does not present a constant feature, as it would if it were physiological, an alternative explanation could be that even the fetuses studied a century ago by Retzius were aborted because of CA.

Conclusions

Both subclinical and clinical CA are associated with preterm delivery and related neonatal complications. Moreover, they can lead to a fetal inflammatory response which contributes to neonatal brain injury [6, 36, 39, 47–49], but may also be responsible for intrauterine damage. Information that could be derived from histological study of such a tissue as readily available as the placenta would be valuable in improving our understanding and clinical management of the highrisk preterm and term neonate. Having more information about the history of uterine infections before and during pregnancy and about mechanisms of the maternal and fetal response to bacterial invasion is crucial to a better understanding of these infections. This, in turn, is expected to stimulate the research into better treatments to reduce spontaneous preterm delivery and its associated long-term morbidity and mortality.

Acknowledgments

This study was supported by grants from the Italian Ministry of Education MURST (funding for research of national interest) and from the University of Siena.

Biol Neonate 2001;79:201–204

203

References 1 Benirschke K, Kaufmann P: Pathology of the Human Placenta, ed 3. New York, Springer-Verlag, 1995, pp 537–623. 2 Altshuler G: Role of the placenta in perinatal pathology (revisited). Pediatr Pathol Lab Med 1996;16:207–233. 3 Arias F, Victoria A, Cho K, Kraus F: Placental histology and clinical characteristics of patients with preterm premature rupture of membranes. Obstet Gynecol 1997;89:265–271. 4 Alexander JM, McIntire DM, Leveno KJ: Chorioamnionitis and the prognosis for term infants. Obstet Gynecol 1999;94:274–278. 5 Dexter SC, Malee MP, Pinar H, Hogan JW, Carpenter MW, Vohr BR: Influence of chorioamnionitis on developmental outcome in very low birth weight infants. Obstet Gynecol 1999;94:267–273. 6 Leviton A, Paneth N, Lynne Reuss M, Susser M, Allred EN, Damman O: Maternal infection, fetal inflammatory response, and brain damage in very low birth weight infants. Pediatr Res 1999;46:566– 575. 7 Hillier SL, Martius J, Krohn M, Kiviat N, Holmes KK, Eschenbach DA: A case-control study of chorioamnionic infection and histologic chorioamnionitis in prematurity. N Engl J Med 1988;319: 972–978. 8 Russell P: Inflammatory lesions of the human placenta. I. Clinical significance of acute chorioamnionitis. Am J Diagn Gynecol Obstet 1979;1:127– 137. 9 Widholm O, Meyer B, Numers CV: Inflammation of the umbilical cord in cases of foetal asphyxia of unknown clinical etiology. Gynaecologia 1963; 155:385–399. 10 Hillier SL, Witkin SS, Krohn MA, Watts DH, Kiviat NB, Eschenbach DA: The relationship of amniotic fluid cytokines and preterm delivery, amniotic fluid infection, histologic chorioamnionitis and chorioamnion infection. Obstet Gynecol 1993; 81:941–948. 11 Hack M, Merkatz IR: Preterm delivery and low birth weight – a dire legacy (editorial). N Engl J Med 1995;333:1772–1774. 12 Mercer BM, Lewis R: Preterm labor and preterm premature rupture of the membranes. Infect Dis Clin North Am 1997;11:177–201. 13 Hansen AR, Collins MH, Genest D, Heller D, Schwarz S, Banagon P, Allred EN, Leviton A: Very low birthweight infant’s placenta and its relation to pregnancy and fetal characteristics. Pediatr Dev Pathol 2000;3:419–430. 14 Hansen AR, Collins MH, Genest D, Heller D, Shen-Schwarz S, Banagon P, Allred EN, Leviton A: Very low birthweight placenta: Clustering of morphologic characteristics. Pediatr Dev Pathol 2000; 3:431–438. 15 Ohlsson A, Wang E: An analysis of antenatal tests to detect infection in preterm premature rupture of the membranes. Am J Obstet Gynecol 1990;162: 809–818. 16 Newton ER, Piper J, Peairs W: Bacterial vaginosis and intraamniotic infection. Am J Obstet Gynecol 1997;176:672–677. 17 De Felice C, Toti P, Laurini RN, Stumpo M, Picciolini E, Todros T, Tanganelli P, Buonocore G, Bracci R: Early neonatal brain injury in histologic chorioamnionitis. J Pediatr, in press.

204

18 Hobel C: Clinical aspects of infection as a cause of prematurity; in Saling E (ed): Perinatology, Nestlé Series. New York, Raven Press, 1992, vol 26, pp 111–127. 19 Goldenberg RL, Hauth JC, Andrews WW: Intrauterine infection and preterm delivery. N Engl J Med 2000;342:1500–1507. 20 Hillier SL, Nugent RP, Eschenbach DA, Krohn MA, Gibbs RS, Martin DH, Cotch MF, Edelman R, Pastorek JG 2nd, Rao AV, et al: Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. The Vaginal Infections and Prematurity Study Group. N Engl J Med 1995;333:1737–1742. 21 Hauth JC, Goldenberg RL, Andrews WW, DuBard MB, Copper RL: Reduced incidence of preterm delivery with metronidazole and erythromycin in women with bacterial vaginosis. N Engl J Med 1995;333:1732–1736. 22 Mercer BM, Arheart KL: Antimicrobial therapy in expectant management of preterm premature rupture of the membranes. Lancet 1995;346:1271– 1279. 23 Horowitz S, Mazor M, Romero R, Horowitz J, Glezerman M: Infection of the amniotic cavity with Ureaplasma urealyticum in the midtrimester of pregnancy. J Reprod Med 1995;40:375–379. 24 Ghidini A, Jenkins CB, Spong CY, Pezzullo JC, Salafia CM, Eglinton GS: Elevated amniotic fluid interleukin-6 levels during the early second trimester are associated with greater risk of subsequent preterm delivery. Am J Reprod Immunol 1997;37: 227–231. 25 Wenstrom KD, Andrews WW, Hauth JC, Goldenberg RL, DuBard M, Cliver S: Elevated secondtrimester amniotic fluid interleukin-6 levels predict preterm delivery. Am J Obstet Gynecol 1998; 178:546–550. 26 De Felice C, Toti P, Picciolini E, Massafra C, Pecciarini L, Palmeri MLD, Bracci R: High incidence of histologic chorioamnionitis in women with gestational vaginal bleeding. Acta Obstet Gynecol Scand 1997;76:85–86. 27 Stallmach T, Hebisch G, Joller-Jemelka HI, Orban P, Schwaller J, Engelmann M: Cytokine production and visualized effects in the feto-maternal unit. Quantitative and topographic data on cytokines during intrauterine disease. Lab Invest 1995; 73:384–392. 28 Yoon BH, Jun JK, Romero R, Park KH, Gomez R, Choi J-H, Kim I-O: Amniotic fluid inflammatory cytokines (interleukin-6, interleukin-1ß, and tumor necrosis factor-·), neonatal brain white matter lesions, and cerebral palsy. Am J Obstet Gynecol 1997;177:19–26. 29 Dammann O, Leviton A: Maternal intrauterine infection, cytokines, and brain damage in the preterm newborn. Pediatr Res 1997;42:1–8. 30 Gomez R, Romero R, Ghezzi F, Yoon H, Mazor M, Berry SM: The fetal inflammatory response syndrome. Am J Obstet Gynecol 1998;179:194– 202. 31 Stallmach T, Karolyi L: Augmentation of fetal granulopoiesis with chorioamnionitis during the second trimester of gestation. Hum Pathol 1994; 25:244–247.

Biol Neonate 2001;79:201–204

32 De Felice C, Toti P, Santopietro R, Stumpo M, Pecciarini L, Bagnoli F: Small thymus in very low birth weight infants born to mothers with subclinical chorioamnionitis. J Pediatr 1999;135:384– 386. 33 Toti P, De Felice C, Stumpo M, Schürfeld K, Di Leo L, Vatti R, Bianciardi G, Buonocore G, Seemayer TA, Luzi L: Acute thymic involution in fetuses and neonates with chorioamnionitis. Hum Pathol 2000;31:1121–1128. 34 Mari`n-Padilla M: Developmental neuropathology and impact of perinatal brain damage. I: Hemorrhagic lesions of neocortex. J Neuropathol Exp Neurol 1996;55:758–773. 35 Perlman JM: Intrapartum hypoxic-ischemic cerebral injury and subsequent cerebral palsy: Medicolegal issues. Pediatrics 1997;99:851–859. 36 Wu YW, Colford JM: Chorioamniositis as a risk factor for cerebral palsy. A meta-analysis. JAMA 2000;284:1417–1424. 37 Murphy DJ, Sellers S, MacKenzie IZ, Yudkin PL, Johnson AM: Case-control study of antenatal and intrapartum risk factors for cerebral palsy in very preterm singleton babies. Lancet 1995;346:1449– 1454. 38 Nelson KB, Grether JK: Can magnesium sulfate reduce the risk of cerebral palsy in very low birthweight infants? Pediatrics 1995;95:263–269. 39 Dammann O, Leviton A: Infection remote from the brain, neonatal white matter damage, and cerebral palsy in the preterm infant. Semin Pediatr Neurol 1998;5:190–201. 40 Harding B, Copp AJ: Malformations; in Graham DI, Lantos PL (eds): Greenfield’s Neuropathology, ed 6. London, Edward Arnold, 1997, pp 397–533. 41 Ferrie CD, Jackson GD, Giannakodimos S, Panayiotopous CP: Posterior agyria-pachygyria with polymicrogyia: Evidence for an inherited neuronal migration disorder. Neurology 1995;45:150–153. 42 Dobyns WB, Truwit CL: Lissencephaly and other malformations of cortical development: 1995 update. Neuropediatrics 1995;26:132–147. 43 Marret S, Mukendi R, Gadisseux JF, Gressens P, Evrard P: Effect of ibotenate on brain development: An excitoxic mouse model of microgyria and posthypoxic-like lesions. J Neuropathol Exp Neurol 1995;54:358–370. 44 Volpe JJ: Neurology of the Newborn, ed 3. Philadelphia, W.B. Saunders, 1995, pp 53–69. 45 Toti P, De Felice C, Palmeri MLD, Villanova M, Martin JJ, Buonocore G: Inflammatory pathogenesis of cortical polymicrogyria: An autopsy study. Pediatr Res 1998;44:291–296. 46 Larroche JC: Cytoarchitectonic abnormalities (abnormalities of cell migration); in Vinken PJ, Bruyn GW (eds): Handbook of Clinical Neurology: Congenital Malformations of the Brain and Skull, Part I. Amsterdam, Elsevier, 1977, p 488. 47 Grether JK, Nelson KB: Maternal infection and cerebral palsy in infants of normal birth weight. JAMA 1997;278:207–211. 48 O’Shea TM, Klinepeter KL, Meis PJ, Dillard RG: Intrauterine infection and the risk of cerebral palsy in very low-birthweight infants. Paediatr Perinat Epidemiol 1998;12:72–83. 49 Adinolfi M: Infectious diseases in pregnancy, cytokines and neurological impairment: An hypothesis. Dev Med Child Neurol 1999;35:549–553.

Toti/De Felice

Biol Neonate 2001;79:205–209

New Insights into the Pathogenesis of Pulmonary Inflammation in Preterm Infants Christian P. Speer University Children’s Hospital, Würzburg, Germany

Key Words Preterm infants W Chronic lung disease W Bronchopulmonary dysplasia W Inflammation W Oxidative damage W Cytokines

Abstract Chronic lung disease (CLD) and bronchopulmonary dysplasia are associated with a significant inflammatory response of the airways and the interstitium of the lungs. Besides inflammatory cells, various cytokines, lipid mediators, proteolytic enzymes and toxic oxygen radicals may play an essential role in the pathogenesis of this disease. Intrauterine exposure to chorioamnionitis or proinflammatory cytokines has been shown to induce a pulmonary and systemic inflammatory response in the fetus. In this subgroup, antenatal infection may prime the lung such that minimally injurious postnatal events provoke an excessive inflammatory response in the airways and the pulmonary tissue. Inflammation and lung injury most certainly affect normal alveolization and pulmonary vascular development in preterm infants with CLD. Copyright © 2001 S. Karger AG, Basel

Introduction

During the past decade, our understanding of the pathogenesis of chronic lung disease (CLD) and bronchopulmonary dysplasia (BPD) has expanded considerably. The principal risk factors which have clearly been identified are lung immaturity, baro- and volutrauma, initiation of mechanical ventilation, oxygen toxicity, prenatal and nosocomial infections as well as increased pulmonary blood flow secondary to a patent ductus arteriosus [1]. In addition, there is a new category of preterm infants with CLD who initially have minimal or absent signs of respiratory distress syndrome (RDS) but who subse-

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0205$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

quently develop oxygen dependency and ventilatory needs within 14 days [2]. Most likely, these infants have been exposed to chorioamnionitis or early postnatal pulmonary or systemic infection. These aspects and the current pathogenetic concepts of CLD/BPD will be summarized in this article. However, we have to be cautious about these interpretations, since the exact pathogenetic mechanisms of lung injury and repair and the interference with normal lung growth are incompletely understood. Nevertheless, there is growing evidence that a complex inflammatory reaction takes place in the airways and the interstitium of the immature lungs. This pulmonary inflammation is characterized by an accumulation of various cells and inflammatory mediators which eventually contribute to an increased microvascular permeability and to lung damage [3–8].

Recruitment and Activation of Neutrophils and Macrophages

It has been convincingly demonstrated that preterm infants who developed BPD had much higher numbers of inflammatory cells in airway samples compared with infants who recovered from RDS [9]. The predominant cell identified in the bronchoalveolar lavage fluid at the early stage of inflammation was the neutrophil [10–14]. With resolution of RDS, the neutrophil count in tracheobronchial aspirates decreased, but it persisted in those infants who developed CLD, a finding which has been associated with the development of CLD [4]. By approximately 4 days of postnatal age, alveolar macrophages have been shown to reach the maximum concentration in the airway samples of infants with RDS, and to persist in infants with BPD [13, 15, 16]. Murch et al. [17] were able to demonstrate that CD68- and MAC-387-positive macrophages and neutrophils were the predominant cell population in the lung tissues of infants who had died during the early stages of RDS. Besides various detrimental effects on

Christian P. Speer, MD, FRCP(E) Professor of Pediatrics, Director and Chairman University Children’s Hospital, Josef-Schneider-Strasse 2 D–97080 Würzburg (Germany) Tel. +49 931 201 5830, Fax +49 931 201 5833, E-Mail [email protected]

pulmonary epithelial and endothelial integrity, neutrophils and macrophages play an essential role in host defense mechanisms of the preterm infant and the newborn. Slightly impaired neutrophil and macrophage functions [18, 19], combined with some deficits in the humoral host defense mechanisms, may directly contribute to the increased susceptibility of preterm infants to pulmonary and systemic nosocomial infections [20]. It has been convincingly demonstrated that airway secretions of infants with RDS have a high chemotactic activity, and, moreover, that infants who develop CLD exhibit a higher chemotactic activity at a postnatal age of 5 days when compared with patients with RDS [11]. The anaphylatoxin C5a [21], leukotriene B4 and IL-8, important chemoattractants for human neutrophils, have been detected in the bronchoalveolar fluid of these infants. The concentrations of C5a, leukotriene B4 and IL-8 were found to be significantly higher in those infants who developed CLD compared to babies with RDS [21–23]. IL-8 is probably the most important chemotactic factor in the lung [11, 24–29]. Increased IL-8 levels in tracheal aspirates of infants with developing CLD clearly preceded the marked neutrophil influx observed in these infants [28]. Additional factors with chemotactic properties, such as platelet-activating factor, ICAM-1, 5-hydroxyeicosatetraenoic acid, fibronectin and elastin degradation products, have been identified in the airways of infants with CLD or BPD [23, 27]. Three potent ß-chemokines which induce the chemotaxis of monocytes and macrophages have been detected in the bronchoalveolar lavage fluid of infants with RDS and BPD: macrophage inflammatory protein-1·, monocyte chemotactic protein and growth-related protein-· [17, 30]. Significantly higher concentrations of macrophage inflammatory protein-1· in particular were associated with the later development of pulmonary fibrosis. Before migrating from the circulation into the lung, the neutrophil must adhere to the vascular endothelium. Endothelial cells express various adhesion molecules or selectins (i.e. E-selectin) which bind to complementary sites on the neutrophils (L-selectin). IL-8 plays an important role in the increased expression of neutrophil cell surface receptors (ß2-integrins; e.g. CDIIb/CD18) and it is also thought to exert control over the regulation of endothelial cell adhesion molecules, among them the ligand ICAM-1 [31]. In infants with CLD, increased plasma levels of soluble ICAM-1 and E-selectin have been documented; in addition, concentrations of L-selectin and ICAM-1 in tracheal aspirates were increased when compared with controls [26, 31]. These studies provide indirect evidence for the recruitment of circulating neutrophils into pulmonary tissue and airways. In addition, it has been clearly shown that activated compounds of plasma protein systems are able to affect the alveolar-capillary membrane directly and indirectly by sequestration of activated neutrophils and platelets in the pulmonary vascular bed. Brus et al. [32, 33] and Saugstad et al. [34] found a simultaneous activation of clotting, fibrinolysis, kinin-kallikrein and the complement system in neonatal RDS. This activation was accompanied by an activation of neutrophils and platelets as indicated by an increased release of neutrophil elastase and thromboxane (see below). Once this activation process has started, it might become a complicated self-perpetuating process which – at least to some extent – may cause endothelial injury in the pulmonary vascular bed and, additionally, in the systemic circulation. It is intriguing to speculate that this process could also involve cerebral vessels; this aspect is addressed in more detail in other articles of this special issue of Biology of the Neonate.

206

Biol Neonate 2001;79:205–209

Cytokines, Mechanical Ventilation, Chorioamnionitis and Postnatal Infection

Besides IL-8, other proinflammatory cytokines such as TNF-·, IL-1 and IL-6 seem to be important mediators in the early inflammatory response by recruiting and activating inflammatory cells. Increased levels of TNF-·, IL-1, IL-1ß2, IL-6 and IL-8 have been demonstrated by a number of investigators in bronchoalveolar fluid from infants who developed CLD [16, 25, 35, 36]. These cytokines are synthesized by alveolar macrophages, fibroblasts, type II pneumocytes and endothelial cells upon stimulation by hypoxia, hyperoxia, endotoxin and other factors [37, 38]. Mechanical ventilation with large tidal volumes has been a wellrecognized cause of pulmonary inflammation [39, 40]. Recently, Tremblay et al. [41] studied the effect of mechanical ventilation on the generation of various inflammatory and anti-inflammatory mediators in a rat lung model in the presence or absence of lipopolysaccharide-induced sepsis. Different levels of lung lavage cytokines, as well as differential tissue expression of an immediate early response gene (g-fos), were observed with the various ventilation strategies. The highest levels of inflammatory mediators were seen with those ventilatory strategies with no positive end expiratory pressure. The authors concluded that mechanical ventilation can have a significant influence on the inflammatory events of the normal or septic lung, and thus may play a role in either initiating or propagating a local as well as a systemic inflammatory response [41]. In addition, in vitro cyclic cell stretch was shown to upregulate the production and release of IL-8 by human alveolar epithelium in the absence of structural cell damage [42]. In a recent study with baboons, no consistent differences in tracheal TNF-·, IL-1, IL-6 and IL-10 were found in animals treated with high-frequency oscillatory ventilation compared with animals supported with conventional mechanical ventilation [43]. However, higher IL-8 concentrations and macrophage numbers were identified in conventionally ventilated baboons [43]. In preterm infants treated with high-frequency oscillatory ventilation or conventional ventilation, no differences in tracheobronchial IL-8 levels have been observed [44]. Recent epidemiological data support a strong link between chorioamnionitis and lung injury in utero and the subsequent development of BPD; the current knowledge has recently been summarized [45–47]. Watterberg et al. [49] reported that infants born very prematurely to women with clinical chorioamnionitis had a lower incidence of RDS but a greater risk of BPD than a matched group of comparable infants [48]. Moreover, IL-1 levels were increased in tracheal lavage fluid of infants exposed to chorioamnionitis. The proinflammatory cytokine IL-1 has been shown to induce the synthesis of surfactant proteins and to improve postnatal lung function in various animal models by intraamniotic injection [49]. In contrast, in another study, increased TNF-· levels in human amniotic fluid during amnionitis were associated with a higher incidence of neonatal RDS, suggesting a possible role of TNF-· in the surfactant systems and/or alveolar integrity of the immature lung [50]. Fetal exposure to proinflammatory cytokines such as TNF-·, IL-1ß, IL-6 and IL-8, however, was clearly identified as a risk factor for BPD [51–53]. In addition, elevated levels of IL-6 in fetal cord blood at birth, indicating a fetal inflammatory response, were shown to be an independent risk factor for the development of CLD/BPD. Our most recent findings demonstrated that chorioamnionitis was associated with an intrauterine inflammatory response of the fetal lung characterized by a marked infiltration of neutrophils and

Speer

macrophages as well as an increased expression of IL-8 mRNA in the interstitial tissue and the bronchoalveolar epithelium [54]. These data indicate that the injury responsible for CLD – at least in a subset of infants – may begin prior to birth. In a recent publication, it has been hypothesized that antenatal infection/inflammation ‘primes’ the lung such that minimally injurious events provoke an excessive inflammatory response which could promote damage in the preterm lung [55]. In addition to chorioamnionitis, a clear association between systemic infections, colonization of the airways with various microbial pathogens and the development of CLD, especially in very-low-birthweight infants with mild RDS or absent acute pulmonary disease, has been established [35, 48, 56]. In this population, the development of late episodes of persistent ductus arteriosus, usually associated with a nosocomial infection, has been shown to play a primary role in the pathogenesis of CLD [57, 58]. Increased serum concentrations of 6ketoprostaglandin F1· have been identified in infants with nosocomial infections. These vasoactive mediators, which are most likely produced by inflammatory cells, effectively prevent ductal closure [59]. In addition, IL-1 has been found in high concentrations in bronchoalveolar fluid of infants with perinatal colonization of the airways even on the first day of life [56, 60]. It is not surprising, however, that different types of mediators can be detected in the bronchoalveolar secretions of infants with colonized airways and/or pulmonary infection [61–63]. Unfortunately, it is not yet clear if there are any differences in the profile of inflammatory mediators or in the sequence of inflammatory events between infants with colonized airways and those with proven pulmonary infection. Moreover, there is no definite proof if any difference exists between the inflammatory reaction evoked by microbes, lipopolysaccharide or unspecific stimuli like hyperoxia and/or barotrauma in the immature lung of preterm infants. Utilizing reverse transcriptase polymerase chain reaction, a number of investigators have detected mRNA for IL-1·, IL-1ß, IL-6, IL-8 and TNF-· expressed by bronchoalveolar cells of infants with RDS and CLD (macrophages, neutrophils, endothelial cells and other cells) [24, 36, 64]. The production of the proinflammatory cytokines TNF-·, IL-1ß and IL-8 is regulated in part by the anti-inflammatory cytokine IL-10. In a recent study, IL-10 mRNA was undetectable in most of the airway samples from preterm infants with RDS, but it was present in bronchoalveolar cells of term infants with meconium aspiration syndrome. The susceptibility of the preterm infant to CLD may in part reflect an inability to regulate inflammation through the expression of the anti-inflammatory cytokine IL-10 [24]. It is noteworthy, however, that lung inflammatory cells of preterm infants are capable of responding to IL-10 exposure with a reduction of IL-1ß and IL-8 expression in a cell culture system [65]. The interstitial inflammatory response in preterm infants with RDS has been carefully studied by immunohistochemical analysis of whole lung lobes, obtained at postmortem from infants who had died from acute RDS in the first week of life [17]. The results demonstrated a rapid increase from birth in the mucosal density of macrophages and TNF-·-immunoreactive cells, maximal in those who died at or after 72 h of age. The infiltration by TNF-·-positive macrophages was found to be associated with a striking loss of endothelial basement membrane, and interstitial glycosaminoglycans, which was almost complete by 48–72 h. The role of transforming growth factorß and other mediators of fibrogenesis (i.e. fibronectin) has recently been addressed [66, 67].

Pathogenesis of Pulmonary Inflammation in Preterm Infants

Proteolytic Damage and Oxygen Toxicity

A number of investigators have evaluated the possible role of elastase, a powerful neutral proteinase stored in the azurophilic granules of neutrophils, in the pathogenesis of acute lung disease and CLD in preterm infants. Pulmonary tissue elastin is the primary substrate of neutrophil elastase. Under normal circumstances, elastase is rapidly bound and inactivated by ·1-proteinase, which protects the alveolar-capillary unit from proteolytic damage by forming elastase·1-proteinase [68]. In tracheobronchial secretions of infants with RDS and BPD, increased concentrations of free elastase and low activities of ·1-proteinase have been detected [9, 66, 69, 70]. It has been suggested that an imbalance between elastase and ·1-proteinase within the airways may be a hallmark of lung injury in preterm infants [71]. ·1-Proteinase is presumably inactivated by proteolytic cleavage of oxidized ·1-proteinase in the pulmonary tissue and by complex formation of elastase with ·1-proteinase [71, 72]. In the presence of free iron, the generation of hydroxyl radicals has recently been demonstrated in tracheobronchial secretions of preterm infants; the majority of ventilated babies had a high concentration of free iron in their airway samples [73]. In addition, free elastase was shown to prime macrophages for an increased production of toxic oxygen metabolites [74]. Furthermore, oxidative stress has been found to upregulate matrix metalloproteinase-9 activity, a type IV collagenase causing disruption of the extracellular matrix [75]. Clinical data on the presence of free elastase in tracheobronchial fluid of infants with RDS and BPD are somewhat contradictory and have been discussed elsewhere [8]. There is no doubt that free elastase unabated by ·1-proteinase inhibition contributes to the development of acute and chronic lung injury. Increased urinary excretion of desmosine, an elastolytic degradation product of mature cross-linked elastic fibers, was identified in infants with free elastase levels in tracheobronchial secretions [76]. Degradation of lung elastic fibers may lead to impaired alveolar septation. This is of particular concern in the light of recent studies showing that alveolar septation is markedly reduced in the lungs of infants with severe BPD [77]. Moreover, disruption of sulfated glycosaminoglycans, changes in hyaluronan deposition and increased laminin concentrations in lung secretions [78] have been attributed to elastolytic destruction [79, 80]. One of the most important pathophysiological features of RDS and CLD is the increased alveolar-capillary permeability [81]. At a postnatal age of 10–14 days, infants who later developed CLD had a drastic increase in albumin concentrations in bronchoalveolar secretions when compared with infants who recovered from RDS [11]. This abnormal lung permeability is pathognomonic for the early stage of CLD, and it is clearly associated with a deterioration of lung function and a worsening of the clinical situation. During the inflammatory process which has been described – at least in part –, several factors may have detrimental effects on microvascular permeability: direct effects of inflammatory cells and mediators on the alveolar and capillary membrane, including cytokines, toxic oxygen radicals generated by phagocytes or other sources, inactivation of the surfactant system by various serum proteins, modulation of vascular perfusion in the inflamed area or increased shunting via the ductus arteriosus, microbial colonization and infection of the airways [51]. A variety of lipid mediators with well-defined direct effects on microvascular permeability, including leukotrienes, prostacyclin, platelet-activating factor and endothelin-1, have been demonstrated in the airways of infants with BPD [82–86]. Some cytokines may indirectly affect the alveolar-capillary integrity by activat-

Biol Neonate 2001;79:205–209

207

ing inflammatory cells for increased release and secretion of proteolytic enzymes (elastase, collagenase, cathepsin, metalloproteinase and other granule constituents such as defensins) or by priming phagocytes for increased release of toxic oxygen radicals [87]. Resting and stimulated alveolar macrophages of infants with BPD were shown to generate increased amounts of hydrogen peroxide when compared with cells of controls [15]. The possible role of oxidative

stress and the detrimental effects of lipid peroxidation of pulmonary tissue and its association with the development of CLD has recently been reviewed by Saugstad [88]. In general, the pulmonary inflammatory autoinjury is mediated by an arsenal of both cellular and humoral mediators [7]. There is growing evidence that the inflammatory process may affect normal alveolization and pulmonary vascular development in preterm infants with CLD [89].

References 1 Van Marter LJ, Allred EN, Pagano M: Do clinical markers of barotrauma and oxygen toxicity explain interhospital variation in rates of chronic lung disease? Pediatrics 2000;105:1194–1201. 2 Bancalari E, Gonzalez A: Clinical course and lung function abnormalities during development of neonatal chronic lung disease; in Bland RD, Coalson JJ (eds): Chronic Lung Disease in Early Infancy. New York, Marcel Dekker, 2000, pp 41–64. 3 Groneck P, Speer CP: Inflammatory mediators and bronchopulmonary dysplasia. Arch Dis Child Fetal Neonatal Ed 1995;73:F1–F3. 4 Kotecha S: Cytokines in chronic lung disease of prematurity. Eur J Pediatr 1996;155(suppl 2): S14–S17. 5 Özdemir A, Brown MA, Morgan WJ: Markers and mediators of inflammation in neonatal lung disease. Pediatr Pulmonol 1997;23:292–306. 6 Pierce MR, Bancalari E: The role of inflammation in the pathogenesis of bronchopulmonary dysplasia. Pediatr Pulmonol 1995;19:371–378. 7 Zimmermann JJ: Bronchoalveolar inflammatory pathophysiology of bronchopulmonary dysplasia. Clin Perinatol 1995;22:429–456. 8 Speer CP, Groneck P: Oxygen radicals, cytokines, adhesion molecules and lung injury in neonates. Semin Neonatol 1998;3:219–228. 9 Merritt TA, Cochrane CG, Holcomb K, Bohl B, Hallman M, Strayer D, Edwards DK, Gluck L: Elastase and ·1-proteinase inhibitor activity in tracheal aspirates during respiratory distress syndrome. J Clin Invest 1983;72:656–666. 10 Arnon S, Grigg J, Silverman M: Pulmonary inflammatory cells in ventilated preterm infants: Effect of surfactant treatment. Arch Dis Child 1993;69:44– 48. 11 Groneck P, Götze-Speer B, Oppermann M, Eiffert H, Speer CP: Association of pulmonary inflammation and increased microvascular permeability during the development of bronchopulmonary dysplasia: A sequential analysis of inflammatory mediators in respiratory fluids of high-risk preterm infants. Pediatrics 1994;93:712–718. 12 Kotecha S, Chan B, Azam N, Silverman M, Shaw RJ: Increase in interleukin-8 and soluble intercellular adhesion molecule-1 in bronchoalveolar lavage of premature infants with chronic lung disease. Arch Dis Child Fetal Neonatal Ed 1995;72:F90– F96. 13 Ogden BE, Murphy SA, Saunders GC, Pathak D, Johnson JD: Neonatal lung neutrophils and elastase/proteinase inhibitor imbalance. Am Rev Respir Dis 1984;130:817–821. 14 Jackson JC, MacKenzie AP, Chi EY, Standaert TA, Truog WE, Hodson WA: Mechanisms for reduced total lung capacity at birth and during hyaline membrane disease in premature newborn monkeys. Am Rev Respir Dis 1990;142:413–419. 15 Clement A, Chadelat K, Sardet A, Grimfeld A, Tournier G: Alveolar macrophage status in bronchopulmonary dysplasia. Pediatr Res 1988;23: 470–473.

208

16 Rindfleisch MS, Hasday JD, Taciak V, Broderick K, Viscardi RM: Potential role of interleukin-1 in the development of bronchopulmonary dysplasia. J Interferon Cytokine Res 1996;16:365–373. 17 Murch SH, Costeloe K, Klein NJ, Rees H, McIntosh N, Keeling JW, MacDonald TT: Mucosal tumor necrosis factor-· production and extensive disruption of sulfated glycosaminoglycans begin within hours of birth in neonatal respiratory distress syndrome. Pediatr Res 1996;40:484–489. 18 Speer CP, Johnston RB: Neutrophil function in newborn infants; in Polin RA, Fox WW (eds): Fetal and Neonatal Physiology. Philadelphia, Saunders, 1998, vol 2, pp 1954–1960. 19 Grigg J, Riedler J, Robertson CF, Boyle W, Uren S: Alveolar macrophage immaturity in infants and young children. Eur Respir J 1999;14:1198–1205. 20 Stoll BJ, Gordon T, Korones SB, Shankaran S, Tyson JE, Bauer CR, Fanaroff AA, Lemons JA, Donovan EF, Oh W, Stevenson DK, Ehrenkranz RA, Papile LA, Verter J, Wright L: Early-onset sepsis in very low birth weight neonates: A report from the National Institute of Child Health and Human Development Neonatal Research Network. J Pediatr 1996;129:72–80. 21 Groneck P, Oppermann M, Speer CP: Levels of complement anaphylatoxin C5a in pulmonary effluent fluid of infants at risk for chronic lung disease and effects of dexamethasone treatment. Pediatr Res 1993;34:586–590. 22 Kunkel SL, Standiford T, Kasahara K, Strieter RM: Interleukin-8 (IL-8): The major neutrophil chemotactic factor in the lung. Exp Lung Res 1991; 17:17–23. 23 Stenmark KR, Eyzaguirre M, Westcott JY, Henson PM, Murphy RC: Potential role of eicosanoids and PAF in the pathophysiology of bronchopulmonary dysplasia. Am Rev Respir Dis 1987;136:770–772. 24 Jones CA, Cayabyab RG, Kwong KY, Stotts C, Wong B, Hamdan H, Minoo P, DeLemos RA: Undetectable interleukin (IL)-10 and persistent IL8 expression early in hyaline membrane disease: A possible developmental basis for the predisposition to chronic lung inflammation in preterm newborns. Pediatr Res 1996;39:966–975. 25 Jo´nsson B, Tullus K, Brauner A, Lu Y, Noack G: Early increase of TNF· and IL-6 in tracheobronchial aspirate fluid indicator of subsequent chronic lung disease in preterm infants. Arch Dis Child Fetal Neonatal Ed 1997;77:F198–F201. 26 Speer CP: Inflammatory mechanisms in neonatal chronic lung disease. Eur J Pediatr 1999;158(suppl 1): S18–S22. 27 Little S, Dean T, Bevin S, Hall M, Ashton M, Church M, Warner J, Shute J: Role of elevated plasma soluble ICAM-1 and bronchial lavage fluid IL-8 levels as markers of chronic lung disease in premature infants. Thorax 1995;50:1073–1079.

Biol Neonate 2001;79:205–209

28 Munshi UK, Niu JO, Siddiq MM, Parton LA: Elevation of interleukin-8 and interleukin-6 precedes the influx of neutrophils in tracheal aspirates from preterm infants who develop bronchopulmonary dysplasia. Pediatr Pulmonol 1997;24:331–336. 29 Takasaki J, Ogawa Y: Interleukin 8 and granulocyte elastase in the tracheobronchial aspirate of infants without respiratory distress syndrome or intrauterine infection and development of chronic lung disease. Acta Paediatr Jpn 1997;39:437–441. 30 Inwald DP, Costeloe K, Murch SH: High concentrations of GRO-· and MCP-1 in bronchoalveolar fluid of infants with respiratory distress syndrome after surfactant. Arch Dis Child Fetal Neonatal Ed 1998;78:F234–F235. 31 Kotecha S, Silverman M, Shaw RJ, Klein M: Soluble L-selectin concentration in bronchoalveolar fluid obtained from infants who develop chronic lung disease of prematurity. Arch Dis Child Fetal Neonatal Ed 1998;78:F143–F147. 32 Brus F, van Oeveren W, Okken A, Bambang Oetomo S: Activation of the plasma clotting, fibrinolytic, and kinin-kallikrein system in preterm infants with severe idiopathic respiratory distress syndrome. Pediatr Res 1994;36:647–653. 33 Brus F, van Oeveren W, Okken A, Bambang Oetomo S: Activation of circulating polymorphonuclear leukocytes in preterm infants with severe idiopathic respiratory distress syndrome. Pediatr Res 1996; 39:456–463. 34 Saugstad OD, Buo L, Johansen HT, Roise O, Aasen AO: Activation of the plasma kallikrein-kinin system in respiratory distress syndrome. Pediatr Res 1992;32:431–435. 35 Groneck P, Schmale J, Soditt V, Stützer J, GötzeSpeer B, Speer CP: Bronchoalveolar inflammation following airway infection in preterm infants with chronic lung disease. Pediatr Pulmonol, in press. 36 Kotecha S, Wilson L, Wangoo A, Silverman M, Shaw RJ: Increase in interleukin (IL)-1· and IL-6 in bronchoalveolar lavage fluid obtained from infants with chronic lung disease of prematurity. Pediatr Res 1996;40:250–256. 37 Kotecha S: Pathophysiology of chronic lung disease or prematurity. Biol Neonate 2000;78:236– 237. 38 Jobe AH, Ikegami M: Mechanisms initiating lung injury in the preterm. Early Hum Dev 1998;53:81– 94. 39 Muscedere JG, Mullen JBM, Gan K, Slutsky AS: Tidal ventilation at low airway pressures can augment lung injury. Am J Respir Crit Care Med 1994;149:1327–1334. 40 Dreyfuss D, Saumon G: Ventilator-induced lung injury: Lessons from experimental studies. Am J Respir Crit Care Med 1998;157:294–323. 41 Tremblay L, Valenza F, Ribeiro SP, Li J, Slutsky AS: Injurious ventilatory strategies increase cytokines and c-fos m-RNA expression in an isolated rat lung model. J Clin Invest 1997;99:944–952.

Speer

42 Vlahakis NE, Schroeder MA, Limper AH, Hubmayr RD: Stretch induces cytokine release by alveolar epithelial cells in vitro. Am J Physiol 1999; 277:L167–L173. 43 Yoder BA, Siler-Khodr T, Winter VT, Coalson JJ: High frequency oscillatory ventilation – effects on lung function, mechanics, and airway cytokines in the immature baboon model for neonatal chronic lung disease. Am J Respir Crit Care Med 2000;162: 1867–1876. 44 Thome U, Götze-Speer B, Speer CP, Pohlandt F: Comparison of pulmonary inflammatory mediators in preterm infants treated with intermittent positive pressure ventilation or high frequency oscillatory ventilation. Pediatr Res 1998;44:330– 337. 45 Lyon A: Chronic lung disease of prematurity. The role of intra-uterine infection. Eur J Pediatr 2000; 159:798–802. 46 Hallman M: Cytokines, pulmonary surfactant and consequences of intrauterine infection. Biol Neonate 1999;76(suppl 1): 2–9. 47 Vigneswaran R: Infection and preterm birth: Evidence of a common causal relationship with bronchopulmonary dysplasia and cerebral palsy. J Paediatr Child Health 2000;36:293–296. 48 Watterberg KL, Demers LM, Scott SM, Murphy S: Chorioamnionitis and early lung inflammation in infants in whom bronchopulmonary dysplasia develops. Pediatrics 1996;97:210–215. 49 Bry K, Lappalainen U, Hallman M: Intraamniotic interleukin-1 accelerates surfactant protein synthesis in fetal rabbits and improves lung stability after premature birth. J Clin Invest 1997;99:2992– 2999. 50 Hitti J, Krohn MA, Patton DL, Tarcy-Hornoch P, Hillier SL, Cassen EM, Eschenbach DA: Amniotic fluid tumor necrosis factor-· and the risk of respiratory distress syndrome among preterm infants. Am J Obstet Gynecol 1997;177:50–56. 51 Yoon BH, Romero R, Jun JK, Park KH, Park JD, Ghezzi F, Kim BI: Amniotic fluid cytokines (interleukin-6, tumor necrosis factor-alpha, interleukin1 beta, and interleukin-8) and the risk for the development of bronchopulmonary dysplasia. Am J Obstet Gynecol 1997;177:825–830. 52 Ghezzi F, Gomez R, Romero R, Yoon BH, Edwin SS, David C, Janisse J, Mazor M: Elevated interleukin-8 concentrations in amniotic fluid of mothers whose neonates subsequently develop bronchopulmonary dysplasia. Eur J Obstet Gynecol Reprod Biol 1998;78:5–10. 53 Yoon BH, Romero R, Kim KS, Park JS, Ki SH, Kim BI, Jun JK: A systemic fetal inflammatory response and the development of bronchopulmonary dysplasia. Am J Obstet Gynecol 1999;181: 773–779. 54 Schmidt B, Cao L, Mackensen-Haen S, Kendziorra H, Klingel K, Speer CP: Chorioamnionitis and inflammation of the fetal lung. Am J Obstet Gynecol, in press. 55 Jobe AH: Intrauterine cytokine activation and the role of infection. Biol Neonate 2000;78:244–246. 56 Groneck P, Götze-Speer B, Speer CP: Inflammatory bronchopulmonary response of preterm infants with microbial colonisation of the airways at birth. Arch Dis Child Fetal Neonatal Ed 1996;74:F51– F55. 57 Cordero L, Ayers LW, Davis K: Neonatal airway colonization with gram-negative bacilli: Association with severity of bronchopulmonary dysplasia. Pediatr Infect Dis J 1997;16:18–23.

Pathogenesis of Pulmonary Inflammation in Preterm Infants

58 Rojas MA, Gonzalez A, Bancalari E, Claure N, Poole C, Silva-Neto G: Changing trends in the epidemiology and pathogenesis of neonatal chronic lung disease. J Pediatr 1995;126:605–610. 59 Gonzales A, Sosenko IRS, Chandar J, Hummler H, Claure N, Bancalari E: Influence of infection on patent ductus arteriosus and chronic lung disease in premature infants weighing 1000 grams or less. J Pediatr 1996;128:470–478. 60 Wang EEL, Matlow AG, Ohlsson A, Nelson SC: Ureaplasma urealyticum infections in the perinatal period. Clin Perinatol 1997;24:91–105. 61 Buck C, Gallati H, Pohlandt F, Bartmann P: Increased levels of tumor necrosis factor alpha (TNFalpha) and interleukin 1 beta (IL-1 beta) in tracheal aspirates of newborns with pneumonia. Infection 1994;22:238–241. 62 Grigg J, Barber A, Silverman M: Increased levels of bronchoalveolar fluid interleukin-6 in preterm ventilated infants after prolonged rupture of membranes. Am Rev Respir Dis 1992;145:782–786. 63 Willmot RW, Kassab JT, Kilian PL, Benjamin WR, Douglas SD, Wood RE: Increased levels of interleukin-1 in bronchoalveolar washings from children with bacterial pulmonary infections. Am Rev Respir Dis 1990;142:365–368. 64 LoMonaco MB, Barber CM, Sinkin RA: Differential cytokine mRNA expression by neonatal pulmonary cells. Pediatr Res 1996;39:248–251. 65 Kwong KYC, Jones CA, Cayabyab R, Lecart C, Khuu N, Rhandhawa I, Hanley JM, Ramanathan R, de Lemos RA, Minoo P: The effects of IL-10 on proinflammatory cytokine expression (IL-1ß and IL-8) in hyaline membrane disease (HMD). Clin Immunol Immunopathol 1998;88:105–113. 66 Gerdes JS, Harris MC, Polin RA: Effects of dexamethasone and indomethacin on elastase, ·1-proteinase inhibitor, and fibronectin in bronchoalveolar lavage fluid from neonates. J Pediatr 1988;113: 732–737. 67 Kotecha S, Wangoo A, Silverman M, Shaw RJ: Increase in the concentration of transforming growth factor beta-1 in bronchoalveolar lavage fluid before development of chronic lung disease of prematurity. J Pediatr 1996;128:464–469. 68 Speer CP, Ninjo O, Gahr M: Elastase-alpha-1-proteinase inhibitor in early diagnosis of neonatal septicemia. J Pediatr 1986;108:987–990. 69 Sveger KT, Svenningsen N: Protease inhibitors in bronchoalveolar lavage fluid from neonates with special references to secretory leukocyte protease inhibitor. Acta Paediatr 1992;81:757–759. 70 Yoder MC, Chua R, Tepper R: Effect of dexamethasone on pulmonary inflammation and pulmonary function of ventilator-dependent infants with bronchopulmonary dysplasia. Am Rev Respir Dis 1991;143:1044–1048. 71 Merritt TA, Stuard ID, Puccia J, Edwards DK, Finkelstein J, Shapiro DL: Newborn tracheal aspirate cytology: Classification during respiratory distress syndrome and bronchopulmonary dysplasia. J Pediatr 1986;98:949–956. 72 Speer CP, Ruess D, Harms K, Herting E, Gefeller O: Neutrophil elastase and acute pulmonary damage in neonates with severe respiratory distress syndrome. Pediatrics 1993;91:794–799. 73 Gerber CE, Bruchelt G, Stegmann, Schweinsberg F, Speer CP: Presence of bleomycin-detectable free iron in the alveolar system of preterm infants. Biochem Biophys Res Commun 1999;257:218–222.

74 Speer CP, Pabst M, Hedegaard HB, Rest RF, Johnston RB: Enhanced release of oxygen metabolites by monocyte-derived macrophages exposed to proteolytic enzymes: Activity of neutrophil elastase and cathepsin G. J Immunol 1984;133:2151– 2156. 75 Schock BC, Sweet DG, Ennis M, Warner JA, Young IS, Halliday HL: Oxidative stress and increased type-IV collagenase levels in bronchoalveolar lavage fluid from newborn babies. Pediatr Res, in press. 76 Bruce MC, Wedig KE, Jentoft N, Martin RJ, Cheng PW, Boat TF, Fanaroff AA: Altered urinary excretion of elastin cross-links in premature infants who develop bronchopulmonary dysplasia. Am Rev Respir Dis 1985;131:568–572. 77 Margraf LR, Tomashefiski JF Jr, Bruse MC, Dahms B: Morphometric analysis of the lung in bronchopulmonary dysplasia. Am Rev Respir Dis 1991;143:391–400. 78 Alnahhas MH, Karathanasis P, Kriss VM, Pauly TH, Bruce MC: Elevated laminin concentrations in lung secretions of preterm infants supported by mechanical ventilation are correlated with radiographic abnormalities. J Pediatr 1997;131:555– 560. 79 Juul SE, Kinsella MG, Jackson JC, Truog WE, Standaert TA, Hodson WA: Changes in hyaluronan deposition during early respiratory distress syndrome in premature monkeys. Pediatr Res 1994; 35:238–243. 80 Murch SH, MacDonald TT, Walker-Smith JA, Levin M, Lionetti P, Klein NJ: Disruption of sulphated glycosaminoglycans in intestinal inflammation. Lancet 1993;341:711–714. 81 Jefferies AL, Coates G, O’Brodovich H: Pulmonary epithelial permeability in hyaline-membrane disease. N Engl J Med 1984;311:1075–1080. 82 Davidson D, Drafta D, Wilkens BA: Elevated urinary leukotriene E4 in chronic lung disease of extreme prematurity. Am J Respir Crit Care Med 1995;151:841–845. 83 Gaylord MS, Smith ZL, Lorch V, Blank ML, Snyder H: Altered platelet-activating factor levels and acetylhydrolase activities are associated with increasing severity of bronchopulmonary dysplasia. Am J Med Sci 1996;312:149–154. 84 Kojima T, Hattori K, Hirata Y, Aoki T, SasaiTakedatsu M, Kino M, Kobyashi Y: Endothelin-1 has a priming effect on production of superoxide anion by alveolar macrophages: Its possible correlation with bronchopulmonary dysplasia. Pediatr Res 1996;39:112–116. 85 Mirro R, Armstead W, Leffler C: Increased airway leukotriene levels in infants with severe bronchopulmonary dysplasia. Am J Dis Child 1990;144: 160–161. 86 Niu JO, Munshi UK, Siddiq MM, Parton LA: Early increase in endothelin-1 in tracheal aspirates of preterm infants: Correlation with bronchopulmonary dysplasia. J Pediatr 1998;132:965–970. 87 Speer CP, Ambruso DR, Grimsley JA, Johnston RB: Oxidative metabolism in cord blood monocytes and monocyte-derived macrophages. Infect Immun 1985;50:919–921. 88 Saugstad OD: Bronchopulmonary dysplasia and oxidative stress: Are we closer to an understanding of the pathogenesis of BPD? Acta Paediatr 1997; 86:1277–1282. 89 Coalson J, Winter V, Siler-Khodr T, Yoder B: Neonatal chronic lung disease in extremely immature baboons. Am J Respir Crit Care Med 1999;160: 1333–1346.

Biol Neonate 2001;79:205–209

209

Biol Neonate 2001;79:210–212

Red Blood Cell Involvement in Fetal/Neonatal Hypoxia Rodolfo Bracci Serafina Perrone Giuseppe Buonocore Institute of Preventive Pediatrics and Neonatology, University of Siena, Siena, Italy

Key Words Fetus W Newborn W Hypoxia W Erythrocyte W Oxidative stress

Abstract Free radical release plays an important role in the development of brain injury following hypoxic-ischemic encephalopathy. It causes endothelial cell damage and anomalies in NMDA receptors, synaptosome structure and astrocyte function. Mitochondrial dysfunctions caused by asphyxia, reperfusion after ischemia, arachidonic acid cascade, catecholamine metabolism and phagocyte activation are known sources of reactive oxygen species, particularly the superoxide anion (O2–). O2– mainly induces peroxidation by the Fenton/Haber Weiss reaction or via iron-oxygen complexes. Since both reactions require reactive heavy metals, non-protein-bound iron (NPBI) is essential for the induction of lipid peroxidation. Experimental studies have demonstrated the neurotoxicity of iron in ischemiareperfusion. Normal axonal transport of brain iron is also reported to be disrupted in hypoxia-ischemia, leading to a buildup of iron in the white matter. The free iron content of erythrocytes (ICRBC) is considered a marker of oxidative stress. Free iron release is accompanied by the oxidation of membrane proteins and the appearance of senescent antigen, as measured by autologous IgG binding. Our preliminary results suggest a significant positive correlation between plasma free iron and the number of nucleated red cells in cord blood, currently considered a reliable index of lasting intrauterine asphyxia but also possessing a high predictive value for poor neurodevelopmental outcome. The rate of erythropoiesis and the entity of ICRBC are related to the degree of asphyxia and the probability of neurological impairment. Since even an increase in NPBI during asphyxia is related to a poor outcome, iron released by red cells could possibly also contribute to NPBI levels. Copyright © 2001 S. Karger AG, Basel

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0210$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

Introduction

Erythrocytes were the first cells of newborns to reveal the susceptibility of the neonate to oxidative stress. Studies on vitamin E deficiency and increased Heinz body formation in the newborn red cell in the presence of oxidizing agents were followed by reports of deficiencies in enzymes detoxifying reactive oxygen species (ROS) [1, 2]. This knowledge and new details on the nature of the red cell membrane, especially its lipid composition, have clarified the major causes of oxidative hemolysis in the neonate [2–6]. Recently, the knowledge of the complex system by which the erythrocyte adapts to extrauterine life has led to the identification of counterbalancing factors, some of which are involved in protection against oxidative stress. Examples include the protective effects of bilirubin, highly responsive glutathione recycling and probably membrane protein at birth [6–10]. Evaluation of neonatal red cell resistance to free radicals has been considered crucial for identifying babies at high risk of free radical-mediated diseases, taking into account the fact that the oxidative injury to red cells depends upon the intrinsic erythrocyte characteristics, the total antioxidant power of plasma and the amount of prooxidant factors [11]. After the demonstration of the key role of free radicals and ROS in the development of brain damage following hypoxia-ischemia, some considerations regarding a cell which is both a target and a source of oxidative stress are justified.

Iron as a Factor Inducing Oxidative Damage in Red Cells and Other Tissues

Free radical release plays an important role in the development of brain injury following hypoxic-ischemic encephalopathy. It causes endothelial cell damage and anomalies in NMDA receptors, synaptosome structure and astrocyte function [12–17]. Mitochondrial dysfunctions caused by asphyxia, reperfusion after ischemia, arachidonic acid cascade, catecholamine metabolism and phagocyte activation are known sources of ROS, particularly the superoxide anion (O –2) [16]. O –2 mainly induces peroxidation by the Fenton/Haber Weiss reaction or via iron-oxygen complexes [17].

Prof. Rodolfo Bracci Institute of Preventive Pediatrics and Neonatology University of Siena, viale Bracci, 36 I–53100 Siena (Italy) Tel. +39 0577 586542, Fax +39 0577 586182, E-Mail [email protected]

Since both reactions require reactive heavy metals, non-proteinbound iron (NPBI) is essential for the induction of peroxidation of membrane lipids and plasma lipoproteins [16, 17]. Experimental studies have demonstrated the key role of NPBI in inducing neurodegeneration [18]. The neurotoxicity of iron in ischemia-reperfusion has been demonstrated in tissue cultures and in vivo. Hypoxia-reoxygenation is known to increase the iron content in the extracellular spaces, damaging the periventricular white matter [19]. Normal axonal transport of brain iron is also reported to be disrupted in hypoxia-ischemia, leading to a buildup of iron in the white matter [20]. The role of iron in the brain damage involved in hypoxic-ischemic encephalopathy of the newborn is demonstrated by the benefits of chelating agents used in animal models [21, 22]. The chelating agent deferoxamine has recently been found to have beneficial effects on the ATP-dependent Na+,K+ pump of the cortical cell membrane in the early post-hypoxia-ischemia period [23]. However, the chelating agents used in these experiments are not suitable for acute care situations because of their toxicity [24]. Iron deficiency could also increase the vulnerability of some areas of brain to hypoxic-ischemic injury [25]. In human neonates, high NPBI levels in the plasma of asphyxiated newborns and an association between high NPBI levels and poor outcome at the age of 1 year suggests that iron is involved in the development of brain injury following asphyxia [26–29]. The normally low levels of transferrin in the newborn may be a predisposing factor, although severe asphyxia and acidosis seem to be sufficient to cause high plasma NPBI levels [28]. It is interesting that Dorrepaal et al. [28] found levels of NPBI to be the most significant variable in relation to neurodevelopmental outcome. Other authors, who did not measure NPBI, also found high plasma markers of oxidative stress, such as hydroperoxides, in severely asphyxiated newborns [30–32]. Recent studies suggest that the index of lipoprotein lipid peroxidation is a very reliable marker of intrauterine hypoxia [33]. Asphyxiated newborns also have high plasma advanced oxidative protein products, which means that not only plasma lipoproteins but also plasma proteins are targets of oxidation [34].

Red Cell Oxidative Stress and Fetal/Neonatal Asphyxia

Our previous study on the free iron content of erythrocytes (ICRBC) demonstrated significantly higher values in cases of acidosis and asphyxia [35]. A significant positive correlation between hypoxanthine and ICRBC and a significant negative correlation between pH and ICRBC were found [35]. ICRBC correlated with plasma levels of free iron [36]. In a subsequent study, a significant positive correlation was found between ICRBC and malondialdehyde, demonstrating a relationship between ICRBC and plasma lipoprotein lipid peroxidation in asphyxiated infants [37]. ICRBC is considered a marker of oxidative stress, since iron is released in the free form when erythrocytes are incubated with a number of oxidizing agents [38, 39]. ICRBC appears to play a key

role, since chelating agents protect the red cell under the same incubation conditions [38]. Release is accompanied by the oxidation of membrane proteins and the appearance of senescent antigen, as measured by autologous IgG binding [40]. Increased ICRBC has been reported in various red cell abnormalities, including sickle cell anemia and thalassemia [41, 42]. Free radical release in these red cells is complex and does not necessarily involve xenobiotics or other exogenous oxidative stress [41]. The nature of redox-active iron in thalassemic red cells is unclear. Iron hemichrome band 3 cluster association which promotes free radical formation and further hemoglobin oxidation has been suggested [43]. Studies on hemoglobin-bound oxygen have demonstrated that the autoxidation of hemoglobin is greater under hypoxic conditions, when increased production of superoxide anions has also been observed [44]. The apparently contradictory occurrence of free iron release and oxidative stress under hypoxic conditions, well known in mammalian cells, may therefore occur even in red cells with impaired metabolic potential due to loss of organelles. Our preliminary results suggest a significant positive correlation between plasma free iron and the number of nucleated red cells in cord blood [45]. This finding is important, since the number of nucleated red cells is currently considered a reliable index of lasting intrauterine asphyxia [46–51]. The period between the induction of high erythropoietin secretion caused by hypoxia and the appearance of nucleated red cells in the circulation is at least 36–48 h [52]. Our study on the nucleated red blood cell count (NRBC) at birth and follow-up after 3 years not only showed a high NRBC at birth in hypoxic newborns but also a high predictive value for poor outcome [47]. Recent observations confirmed a high NRBC in various conditions of fetal distress [53]. It is unclear whether NPBI in the asphyxiated newborn infant is due to high levels of iron released from storage because of the very low plasma and tissue pH peculiar to the asphyxiated fetus and newborn or to a less stable iron-binding capacity of the plasma of neonates compared to adults. The recent observation of rapid disappearance of NPBI and a rise in plasma iron-binding antioxidant activity while transferrin levels and ferroxidase activity remain stable suggested to Lindeman et al. [54] the possibility of changes in the ability of transferrin to bind iron after birth. Since the release of ferritin from tissues is unlikely unless extremely severe tissue damage occurs, the source of high plasma NPBI levels may be the red cells [55]. Indeed, it has been demonstrated that red cell free iron in the active form is released after lysis [56]. Accentuated hemopoiesis may predispose red cell free iron release, causing red cell damage and the release of iron in the active form. In conclusion, fetal and neonatal asphyxia is followed by severe changes in erythropoiesis and red cell characteristics. The rate of erythropoiesis and the entity of ICRBC are related to the degree of asphyxia and the probability of neurological impairment. Since even an increase in NPBI during asphyxia is related to poor outcome, iron released by red cells could possibly also contribute to NPBI levels.

References 1 Gordon HH, Nitowsky HM, Cornblath M: Studies of tocopherol deficiency in infants and children. I. Hemolysis of erythrocytes in hydrogen peroxide. Am J Dis Child 1955;90:669. 2 Bracci R, Buonocore G: The antioxidant status of erythrocytes in preterm and term infants. Semin Neonatol 1998;3:191–197.

Red Cells and Hypoxia

3 Neefjes VME, Evelo CTA, Baars LGM, Blanco CE: Erythrocyte glutathione S transferase as a marker of oxidative stress at birth. Arch Dis Child Fetal Neonatol Ed 1999;81:F130–F133.

4 Neerhout RC: Erythrocyte lipids in the neonate. Pediatr Res 1968;2:172–178. 5 Matovcik LM, Chiu D, Lubin B, Mentzer WC, Lane PA, Mohandas N, Schrier SL: The aging process of human neonatal erythrocytes. Pediatr Res 1986;20:1091–1096.

Biol Neonate 2001;79:210–212

211

6 Jain SK: Presence of phosphatidylserine in the outer membrane bilayer of newborn human erythrocytes. Biochem Biophys Res Commun 1986;136: 914–920. 7 Mireles LC, Lum MA, Dennery PA: Antioxidant and cytotoxic effects of bilirubin on neonatal erythrocytes. Pediatr Res 1999;45:355–362. 8 Clahsen PC, Moison RMW, Holtzer AJ, Berger HM: Recycling of glutathione during oxidative stress in erythrocytes of the newborn. Pediatr Res 1992;32:399–402. 9 Bracci R, Martini G, Buonocore G, Talluri B, Berni S, Ottaviani MF, Picchi MP, Casini A: Changes in erythrocyte properties during the first hours of life: Electron spin resonance of reacting sulfydryl groups. Pediatr Res 1988;24:391–395. 10 Sandhu IS, Ware K, Grisham MB: Peroxyl radicalmediated hemolysis: Role of lipid, protein and sulfhydryl oxidation. Free Radic Res Commun 1992; 16:111–122. 11 Boda V, Finck B, Durken M, Comments J, Hellwege HH, Kohlschutter A: Monitoring erythrocyte free radical resistance in neonatal blood microsamples using a peroxyl radical-mediated haemolysis test. Scand J Clin Lab Invest 1998;58:317–322. 12 Kunstmann S, Mertsch K, Blasig IE, Grune T: High metabolic rates of 4-hydroxynonenal in brain capillary endothelial cells during hypoxia/reoxygenation. Brain Res 1996;740:353–355. 13 Razdan B, Marro PJ, Tammela O, Goel R, Mishram OP, Delivoria-Papadopopulos M: Selective sensitivity of synaptosomal membrane function to cerebral cortical hypoxia in newborn piglets. Brain Res 1993;600:308–314. 14 Aizenman E, Hartnett KA, Reynolds JJ: Oxygen free radicals regulate NMDA receptor function via a redox modulatory site. Neuron 1990;5:841–846. 15 Sorg O, Horn TFW, Yu NC, Gruol DL, Bloom FE: Inhibition of astrocyte glutamate uptake by reactive oxygen species: Role of antioxidant enzymes. Mol Med 1997;3:431–440. 16 Bracci R: Oxygen toxicity; in Kurjak A (ed): Textbook of Perinatal Medicine. London, Parthenon, 1998, pp 78–89. 17 Schafer FQ, Qian SY, Buettner GR: Iron and free radical oxidations in cell membranes. Cell Mol Biol (Noisy-le-Grand) 2000;46:657–662. 18 Shoham S, Youdim MBH: Iron involvement in neural damage and microgliosis in models of neurodegenerative diseases. Cell Mol Biol (Noisy-leGrand) 2000;46:743–760. 19 Adcock LM, Yamashita Y, Goddard-Finegold J, Smith CV: Cerebral hypoxia-ischemia increases microsomal iron in newborn piglets. Metab Brain Dis 1996;11:359–367. 20 Dietrich RB, Bradley WG: Iron accumulation in the basal ganglia following severe ischemic-anoxic insults in children. Radiology 1998;68:203–206. 21 Rosenthal RE, Chanderbhan R, Marshall G, Fiskum G: Prevention of post-ischemic brain lipid conjugated diene production and neurological injury by hydroxyethyl starch-conjugated deferoxamine. Free Radic Biol Med 1992;12:29–33. 22 Shadid M, Buonocore G, Groenendaal F, Moison R, Ferrali M, Berger HM, Van Bel F: Effect of deferoxamine and allopurinol on non-protein-bound iron concentrations in plasma and cortical brain tissue of newborn lambs following hypoxia-ischemia. Neurosci Lett 1998;248:5–8. 23 Groenendaal F, Shadid M, McGowan JE, Mishra OP, Van Bel F: Effects of deferoxamine, a chelator of free iron, on Na+,K+-ATPase activity of cortical brain cell membrane during early reperfusion after hypoxia-ischemia in newborn lambs. Pediatr Res 2000;48:560–564.

212

24 Bauer M, Feucht K, Ziegenfuss T, Marzi I: Attenuation of heat shock-induced hepatic microcirculatory disturbances by the use of a starch-deferoxamine conjugate for resuscitation. Crit Care Med 1995;23:316–322. 25 Rao R, de Ungria M, Sullivan D, Wu P, Wobken JD, Nelson CA, Georgieff MK: Perinatal brain iron deficiency increases the vulnerability of rat hippocampus to hypoxic ischemic insult. J Nutr 1999;129:199–206. 26 Lindeman JHN, Houdkamp E, Lentjes EGW, Poorthuis BJH, Berger HM: Limited protection against iron-induced lipid peroxidation by cord blood plasma. Free Radic Res Commun 1992;16: 285–294. 27 Evans PJ, Evans R, Kovar IZ, Holton AF, Halliwell B: Bleomycin-detectable iron in plasma of premature and full-term neonates. FEBS Lett 1992; 303:210–212. 28 Dorrepaal CA, Berger HM, Benders MJN, van Zoeren-Grobben D, Van De Bor M, Van Bel F: Non-protein-bound iron in postasphyxial reperfusion injury of the newborn. Pediatrics 1996;98: 883–889. 29 Kime R, Gibson A, Yong W, Hider R, Powers H: Chromatographic method for the determination of non-transferrin-bound iron suitable for use on the plasma and bronchoalveolar lavage fluid of preterm babies. Clin Sci (Colch) 1996;91:633–638. 30 Schmidt H, Grune T, Muller R, Siems WG, Wauer RR: Increased levels of lipid peroxidation products malondialdeyde and 4-hydroxynonenal after perinatal hypoxia. Pediatr Res 1996;40:15–20. 31 Pitkanen OM, Hallman M, Anderson S: Correlation of free oxygen radical-induced lipid peroxidation with outcome in very low birth weight infants. J Pediatr 1990;116:760–764. 32 Varsilia E, Hallman M, Anderson S: Free radicalinduced lipid peroxidation during the early neonatal period. Acta Pediatr 1994;83:692–695. 33 Kaya H, Oral B, Dittrich R, Ozkaya O: Lipid peroxidation in umbilical arterial blood at birth: The effects of breech delivery. Br J Obstet Gynecol 2000;107:982–986. 34 Buonocore G, Perrone S, Longini M, Terzuoli L, Bracci R: Total hydroperoxide and advanced oxidation protein products in preterm hypoxic babies. Pediatr Res 2000;47:221–224. 35 Buonocore G, Zani S, Sargentini I, Gioia D, Signorini C, Bracci R: Hypoxia-induced free iron released in the red cells of newborn infants. Acta Paediatr 1998;87:77–81. 36 Buonocore G, Perrone S, Paffetti P, Longini M, Ciccoli L, Signorini C, Rossi V, Comporti M, Bracci R: Intraerythrocyte non protein bound iron release and plasma oxidative stress in the newborn. Pediatr Res, in press. 37 Buonocore G, Zani S, Perrone S, Caciotti B, Bracci B: Intraerythrocyte nonprotein-bound iron and plasma malondialdehyde in the hypoxic newborn. Free Radic Biol Med 1998;25:766–770. 38 Ferrali M, Signorini C, Ciccoli L, Comporti M: Iron release and membrane damage in erythrocytes exposed to oxidizing agents, phenylhydrazine, divicine and isouramil. Biochem J 1992;285:295– 301. 39 Ciccoli L, Signorini C, Alessandrini C, Ferrali M, Comporti M: Iron release, lipid peroxidation and morphological alterations of erythrocytes exposed to acrolein and phenylhydrazine. Exp Mol Pathol 1994;60:108–118.

Biol Neonate 2001;79:210–212

40 Signorini C, Ferrali M, Ciccoli L, Sugherini L, Magnani A, Comporti M: Iron release, membrane protein oxidation and erythrocyte ageing. FEBS Lett 1995;362:165–170. 41 Greinberg L, Rachmilewitz EA, Kitrossky N, Chevion M: Hydroxyl radical generation in ß-thalassemic red blood cells. Free Radic Biol Med 1995; 18:611–615. 42 Ciccoli L, Signorini C, Scarano C, Rossi V, Bambagioni S, Ferrali M, Comporti M: Iron release in erythrocytes from patients with ß-thalassemia. Free Radic Res 1999;30:407–413. 43 Repka T, Shalev O, Reddy R, Yuan J, Abramov A, Racmilewitz EA, Low PS, Hebbel RP: Nonrandom association of free iron with membranes of sickle and ß-thalassemic erythrocytes. Blood 1993;82: 3204–3210. 44 Balagopalakrishna C, Manoharan PT, Abugo OO, Rifkind JM: Production of superoxide from hemoglobin-bound oxygen under hypoxic conditions. Biochemistry 1996;35:6393–6398. 45 Buonocore G, Perrone S, De Marco L, Paffetti P, Longini M, Ciccoli L, Bracci R: Plasma non protein bound iron in newborn infants at birth: A reliable new index of fetal redox stress. Pediatr Res, in press. 46 Thilaganathan B, Athanasiou S, Ozmen S, Creighton S, Watson NR, Nicolaides KH: Umbilical cord blood erythroblast count as an index of intrauterine hypoxia. Arch Dis Child Fetal Neonatal Ed 1994;70:F192–F194. 47 Buonocore G, Perrone S, Gioia D, Gatti MG, Massafra C, Agosta R, Bracci R: Nucleated red blood cell count at birth as an index of perinatal brain damage. Am J Obstet Gynecol 1999;181:1500– 1505. 48 Green DW, Lyon J, Ackerman NB, Mimouni F: Nucleated erythrocyte count in newborn infants with left-sided congenital diaphragmatic hernia: Relationship with the need of extracorporeal membrane oxygenation and survival. J Pediatr 1995; 127:131–133. 49 Bernstein PS, Minior VK, Divon MY: Neonatal nucleated red blood cell counts in small-for-gestational age fetuses with abnormal umbilical artery Doppler studies. Am J Obstet Gynecol 1997;177: 1079–1084. 50 Baschat AA, Gembruch U, Reiss I, Gortner L, Harman CR, Weiner CP: Neonatal nucleated red blood cell counts in growth-restricted fetuses: Relationship to arterial and venous Doppler studies. Am J Obstet Gynecol 1999;181:190–195. 51 Yeruchimovich M, Dollberg S, Green DW, Mimouni F: Nucleated red blood cells in infants of smoking mothers. Obstet Gynecol 1999;93:403– 406. 52 Bondurant MC, Lind RN, Koury MJ, Ferguson ME: Control of globin gene transcription by erythropoietin in erythroblasts from Friend virus-infected mice. Mol Cell Biol 1985;5:675–683. 53 Hanlon-Lundberg KM, Kirby RS: Nucleated red blood cells as a marker of acidemia in term neonates. Am J Obstet Gynecol 1999;181:196–201. 54 Lindeman JH, Lentjes EG, van Zoeren-Grobben D, Berger HM: Postnatal changes in plasma ceruloplasmin and transferrin antioxidant activities in preterm babies. Biol Neonate 2000;78:73–76. 55 Hwang J, Krebs C, Huynh BH, Edmondson DE, Theil EC, Penner-Hahn JE: A short Fe-Fe distance in peroxodiferric ferritin: Control of Fe substrate versus cofactor decay? Science 2000;287:122–125. 56 Ferrali M, Signorini C, Ciccoli L, Comporti M: Iron released from an erythrocyte lysate by oxidative stress is diffusible and in redox active form. FEBS Lett 1993;319:40–44.

Bracci/Perrone/Buonocore

Biol Neonate 2001;79:213–218

Early Markers of Brain Damage in Premature Low-Birth-Weight Neonates Who Suffered from Perinatal Asphyxia and/or Infection S. Fotopoulos K. Pavlou H. Skouteli I. Papassotiriou N. Lipsou M. Xanthou B’ Neonatal Intensive Care Unit, ‘Aghia Sophia’ Children’s Hospital, Athens, Greece

Key Words Low birth weight W Nucleated red blood cells W Interleukin-6 W Interleukin-1ß W Tumour necrosis factor-· W Periventricular leucomalacia

Abstract We studied 57 low-birth-weight premature neonates, of whom 29 suffered from perinatal asphyxia and/or infection, while the remaining 28 did not and served as controls. We measured peripheral nucleated red blood cell (NRBC) absolute numbers as well as interleukin (IL)-1ß, IL-6 and tumour necrosis factor (TNF)-· cytokine serum levels at 24 h postnatally and on days 3 and 7 following birth. Fourteen of the asphyxiated/infected neonates and 12 controls had neurologic assessments at the corrected postnatal age of 18 months. We found NRBC absolute numbers and serum IL-1ß and IL-6 cytokine levels at 24 h postnatally to be significantly higher in neonates with perinatal asphyxia/infection than in the controls (p = 0.022, p = 0.036 and p = 0.037, respectively). TNF-· levels did not differ. Neurologic examination at the corrected postnatal age of 18 months showed 8 out of the 14 children who had been asphyxiated/ infected as neonates to have abnormal findings, while 12 children who were used as controls during their neonatal period were normal. Abnormal neurologic findings correlated with high NRBC counts and IL-1ß and IL-6 levels at 24 h postnatally. In conclusion, increased NRBC counts and proinflammatory cytokine levels in asphyxiated/infected neonates represent early markers for subsequent neurologic impairment. Copyright © 2001 S. Karger AG, Basel

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0213$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

Introduction

Despite marked improvements in perinatal care, brain injury and damage remain high, particularly in very-low-birth-weight (LBW) infants. Thus, of those born more than 10 weeks before the due date who survive, approximately 25% will have a major disability [1]. Of babies born under 1,500 g, 85% survive, and of the survivors, approximately 5–15% exhibit major spastic motor deficits. Cerebral palsy occurs 20–30 times more often in LBW than in term infants [2]. These LBW infants, apart from being at high risk for motor dysfunction, may present with other neurological impairments, such as mental retardation, cortical visual deficits and epileptic seizures [3]. In this age group, the major neuropathologies for post-asphyxia spastic motor deficits are periventricular leucomalacia and periventricular haemorrhagic infarction, which are lesions primarily to the white matter [4]. Indicators commonly used to recognize birth asphyxia do not allow assessment of the severity of asphyxia which is sufficient to pose a risk of irreversible brain injury. Thus, moderately and briefly low Apgar scores are not related to poor neurologic outcome [5], and most acidotic newborns do not become neurologically symptomatic [6]. Therefore, new early markers of perinatal asphyxial brain damage are being investigated. Several lines of evidence indicate that inflammation is involved in the pathogenesis of ischaemic brain injury [7], in which cytokines have been found to play a significant role [8, 9]. Furthermore, during recent years, a strong association has been found between perinatal infection and white matter damage in the preterm newborn [9–11]. Cytokines triggered by bacteria of the uterine cavity seem to play here again a very important role in causing brain damage [12, 13]. As both perinatal asphyxia and infection appear to damage the brain by a final common pathway that involves proinflammatory cytokines, we decided to investigate peripheral blood cytokine levels in both asphyxiated and/or infected neonates shortly after birth and to define

Dr. M. Xanthou Director, B’ Neonatal Intensive Care Unit ‘Aghia Sophia’ Children’s Hospital GR–11527 Athens (Greece) Tel./Fax +30 1 7488 685, E-Mail [email protected]

their association with the subsequent neurologic development of these infants. Another marker of brain damage which has been described is the rise in nucleated red blood cells (NRBC) in peripheral neonatal blood following birth asphyxia [14–18]. Since severe infections cause ischaemia and hypoxia, the second aim of our study was to investigate this early marker of brain damage both in asphyxiated and septicaemic neonates.

Table 1. Clinical characteristics of the neonates studied

Asphyxiated/infected Controls (n = 29) (n = 28) Birth weight, g Gestational age, weeks Asphyxiated Infected Asphyxiated/infected

1,381B120 31.80B0.84 20 4 5

1,592B57 32.07B0.45 – – –

Patients and Methods

Patients Fifty-seven appropriate for gestational age LBW neonates were studied during the first 48 h of their lives at a mean postnatal age of 24 h. Twenty-nine of them were asphyxiated and or infected and the remaining 28, who had no perinatal asphyxia or infection, were used as controls. The clinical characteristics of these neonates are shown in table 1. In order to define a newborn as asphyxiated, several factors were taken into consideration [18–21]: (1) from the maternal side: prolapsed umbilical cord, placenta abruptio and maternal shock [19, 20], and (2) from the fetal/neonatal side: progressive fetal heart rate abnormalities, fetal acidosis (scalp pH ! 7.20), neonatal acidosis (umbilical cord arterial pH ! 7.10), very low Apgar scores and the need for resuscitation with ventilation for 1 3 min [18, 21]. In order to define a newborn as having perinatal infection, maternal risk factors were taken into consideration, such as prolonged rupture of membranes or chorioamnionitis [22, 23]. In addition, infected newborns were diagnosed as having either proven septicaemia (with positive blood cultures) or suspected septicaemia (with at least 3 clinical and 3 laboratory relevant findings). Neonates with evidence of major congenital malformations or inborn errors of metabolism were ineligible for enrollment, as were those who had blood group incompatibility or had mothers with diabetes. The study was approved by the Hospital Research Committee and informed written parental consent was obtained before neonates were entered in the study. Methods The absolute numbers of peripheral NRBC were determined from 1 ml of blood obtained through venipunctures or indwelling umbilical catheters. Blood samples from the neonates were collected 24 h postnatally, as well as on the 3rd and 7th postnatal days. Complete blood cell counts were performed using the Bayer H1 autoanalyser. The NRBC counts were estimated optically by measuring the numbers of NRBC per 5,000 red blood cells [24]. The results were expressed as absolute numbers of NRBC ! 109/litre. For the determination of tumour necrosis factor (TNF)-·, interleukin (IL)-1ß and IL-6 serum levels, 1.5 ml of blood was collected as above at the same time periods. The ELISA method was used with commercially available kits (for TNF-·, Biosure-Medgenix, sensitivity 3 pg/ml, and for IL-6 and IL-1ß, Diaclone Research, MRS, sensitivity 2 and 5 pg/ml, respectively). Cerebral ultrasound examinations were performed using convex transducers (frequency 5 and 7.5 MHz) with standard sagittal and coronal views. Ultrasounds were obtained from all the babies studied on postnatal days 1, 3, 7 and weekly thereafter until discharge. All neonatal ultrasounds were performed by the same radiologist. The neurologic status of the infants was assessed using the Amiel-Tison method [25] by a paediatric neurologist unaware of the

214

Biol Neonate 2001;79:213–218

Table 2. NRBC absolute numbers and cytokine serum levels at 24 h

postnatally in asphyxiated/infected neonates and controls

NRBC, !109/litre IL-1ß, pg/ml IL-6, pg/ml TNF-·, pg/ml

Asphyxiated/infected Controls (n = 29) (n = 28)

p

7,805B2,888 3.0B1.1 51B13 37B5

0.022 0.036 0.037 n.s.

737B223 0.41B0.36 20B5 43B2

n.s. = Not significant.

perinatal problems, at the corrected postnatal ages of 6, 12 and 18 months. The 18-month neurologic examination was taken into consideration for outcome evaluation. Twenty-six of the 57 initially studied neonates were assessed, as 9 had died and the remaining 22 were lost to follow-up. Statistical Analysis The data were analysed for statistically significant differences by the two-tailed Student t test. The results are expressed as mean values plus or minus the standard error of the mean (SEM). Analysis was performed using the SPSS software package and p values ! 0.05 were considered significant.

Results

The birth weights and gestational ages did not differ significantly between the asphyxiated/infected neonates and the controls (table 1). Nine of the asphyxiated/infected neonates died, while none of the controls did. At a mean postnatal age of 24 h, the absolute numbers of NRBC in the peripheral blood of asphyxiated/infected neonates were significantly higher compared with those of the controls (p = 0.022) (table 2). During the following days, the absolute NRBC numbers in neonatal blood showed a progressive fall in both groups (fig. 1). The asphyxiated/infected neonates had significantly higher IL-6 and IL1ß serum levels than the controls (p = 0.037 and p = 0.036, respectively) at a mean postnatal age of 24 h (table 2, fig. 2). TNF-· serum levels did not differ between the two groups of neonates studied.

Fotopoulos/Pavlou/Skouteli/Papassotiriou/ Lipsou/Xanthou

Fig. 1. Absolute numbers of NRBC ! 109/litre (mean B SEM) in asphyxiated/infected neonates and controls during the first week of life. * p = 0.023 when comparing the NRBC absolute numbers between asphyxiated/infected neonates and controls at 24 h postnatally.

Fig. 2. IL-6 and IL-1ß serum levels (mean B SEM) in asphyxiated/

Table 3. Abnormal neurologic and ultra-

Table 4. NRBC absolute numbers and serum cytokine levels at 24 h

sound findings in children who attended follow-up

postnatally in neonates with normal and abnormal neurologic development

Follow-up

infected neonates and controls 24 h postnatally. * p = 0.037 when comparing IL-6 levels between asphyxiated/infected neonates and controls. ** p = 0.036 when comparing IL-1 levels between asphyxiated/infected neonates and controls.

Asphyxiated/ Controls infected

Total children 14 Abnormal NF 8 Hypertonia 4 Hypotonia 2 Cerebral palsy 2 Abnormal UF 9 ↑ Echogenicity 5 IVH : II 1 IVH: III 1 IVH: III+PVL 2

12 0 – – – 2 1 1 – –

NF = Neurologic findings; UF = ultrasound findings; IVH = intraventricular haemorrhage; PVL = periventricular leucomalacia.

Neurologic examination at the corrected age of 18 months revealed abnormal findings in 8 of the 14 children who were asphyxiated/infected as neonates. Abnormalities included spastic cerebral palsy in 2 children and less severe deficits, such as generalized hypotonia and lower limb hypertonia, in 6 children. No neurologic abnormalities were found in 12 children who served as controls as neonates (table 3).

NRBC and Cytokines as Markers of Brain Damage in LBW Neonates

NRBC, ! 109/litre IL-1ß, pg/ml IL-6, pg/ml

Neurologically abnormal

Neurologically normal

p

2,196B592 1.00B0.38 58B28

525B317 0.30B0.01 20B12

0.034 0.049 0.07

The peripheral NRBC numbers at a mean postnatal age of 24 h were significantly higher in the neonates who developed neurologic abnormalities than in those who had a normal neurologic development (p = 0.034) (table 4, fig. 3). The serum levels of IL-1ß were higher during the same time period in the neonates who developed neurologic abnormalities (p = 0.049) (fig. 4). The serum levels of IL-6 were also higher in the neonates who developed neurologic abnormalities than in those who did not; however, the increase did not reach statistical significance (p = 0.07) (fig. 4). No differences in the serum levels of TNF-· were found between neonates who later developed normally and those who showed neurologic abnormalities. From the children who attended follow-up, 9 out of the 14 asphyxiated/infected and 2 out of the 12 controls had abnormal ultrasounds (table 3). However, the ultrasound findings in the controls revealed only mild abnormalities, such as transient periventricular echogenicities that cleared in the first postnatal week and unilateral grade II intraventricular haemorrhage (table 3). No correlation was found between abnormal ultrasound findings and cytokine IL-1, IL-6 and TNF-· levels or NRBC absolute numbers.

Biol Neonate 2001;79:213–218

215

Fig. 3. Absolute numbers of NRBC ! 109/litre (mean B SEM) at

Fig. 4. IL-6 and IL-1ß serum levels at 24 h postnatally in neonates

24 h postnatally in neonates who developed neurologic abnormalities and in those who did not. * p = 0.034 when comparing the values of neonates who developed neurologic abnormalities with those who did not.

who developed neurologic abnormalities and in those who did not. * p = 0.049 when comparing IL-1ß values of neonates who developed neurologic abnormalities with those who did not.

Discussion

In our study, we found the absolute NRBC numbers to be significantly higher in the asphyxiated/infected newborns than in the controls. Our findings are in agreement with those reported in the literature, according to which, raised NRBC levels indicate fetal hypoxia in term [14, 17] and preterm newborns [15, 16, 18]. We have also found that asphyxiated/infected neonates had significantly higher cytokine IL-6 and IL-1ß serum levels than their controls. Proinflammatory cytokines have been found to be increased in the serum [26, 27] and cerebrospinal fluid [7, 21] of newborns with perinatal hypoxia-ischaemia. Recent studies have also shown a strong association between perinatal infection and proinflammatory cytokines [10, 11, 28–31]. In two of these studies [30, 31], increased IL-6 levels were found in the umbilical cord blood of premature newborns. Experimental studies indicate that the immunoinflammatory system is involved in the biochemical cascade leading to injury in the immature brain after hypoxia-ischaemia [32–34]. The inflammatory reaction triggered by ischaemia during early reperfusion consists of a large influx of polymorphonuclear neutrophils, followed by monocytes [35] and microglia activation [36]. In addition, experimental models have recently shown that the expression of mRNA and bioactive protein for the proinflammatory cytokines IL-1, IL-6 and TNF-· increases following hypoxia-ischaemia [37, 38]. Apart from the local effect at the level of the brain, birth asphyxia affects the entire body and is often associated with multiple organ failure which may result in a substantial systemic increase in cytokines. During perinatal infections, cytokines are present in the uterus, fetal circulation and the brain. The passage from the circulation to the brain is easy, either because of the immaturity of the blood-brain barrier [39] or because proinflammatory cytokines reduce the efficacy of the blood-brain barrier [40]. When cytokines of either asphyxiated or infected fetuses reach the brain they may induce further cytokine production by microglia, astrocytes and leucocytes that have invaded the area. These cytokines

216

Biol Neonate 2001;79:213–218

might damage germinal matrix epithelium, causing intraventricular haemorrhage [41], or the developing white matter [42], causing cerebral palsy. Proinflammatory cytokines might damage developing white matter by adversely influencing oligodendrocytes, astrocytes and myelin, by inducing intravascular coagulation and/or thrombosis and vasoconstriction or by inducing the production of other cytokines like platelet-activating factor [43]. Furthermore, cytokines in the developing brain play a role in the mediation of neurotoxicity by nitric oxide [44] and excitatory amino acids [45]. As we have seen, infection and hypoxia-ischaemia are both associated with inflammation-related abnormalities of the white matter [46]. They share some characteristics, including elevated levels of proinflammatory cytokines and adhesion molecules [9, 47]. Leviton et al. [8] presented a two-component model of brain white matter damage in preterm neonates. The first component is the insult, which comprises infection and/or hypoxia-ischaemia, which are both associated with inflammation-related abnormalities in the white matter. The second is a developmental component, which comprises immaturity of the ependymal/endothelial, oligodendroglial and endogenous protection systems. They postulated that both insults and developmentally regulated protective mechanisms play a role in the aetiology of white matter damage. Of the 14 neonates with perinatal asphyxia and/or infection who we followed up, 9 developed neurologic abnormalities, while all our controls had normal neurologic assessments. Furthermore, when we compared serum cytokine levels at 24 h postnatally between the neonates who subsequently proved to be neurologically normal or abnormal, we found that the neonates who developed neurologic abnormalities had significantly higher IL-1 levels, while the levels of IL-6 showed a rise which did not reach statistical significance. Perinatal asphyxia has been associated with brain damage [5, 26, 48, 49]. Recently, a strong association has also been found between intrauterine infection and white matter damage in the preterm newborn, leading to cerebral palsy [9–11]. Thus, elevated cytokine levels in the amniotic fluid [12] and umbilical cord [13] have been found to be associated with periventricular leucomalacia and cerebral palsy.

Fotopoulos/Pavlou/Skouteli/Papassotiriou/ Lipsou/Xanthou

Studies in newborn kittens have shown that intraperitoneal injection of bacterial endotoxin causes brain white matter damage [50]. Finally, immunohistochemical studies of infants suffering periventricular leucomalacia have demonstrated increased TNF-· and IL-6 in brain tissue [51]. We found erythroblast numbers to be significantly higher, 24 h postnatally, in the peripheral blood of neonates who developed neurologic abnormalities than in the controls. The association between high peripheral blood erythroblast counts and neurologic impairment has been reported in the literature [17, 18]. Nine of the 14 asphyxiated/infected children who attended follow-up had neonatal ultrasound abnormalities, while mild abnormalities were encountered in only 2 of the controls. We found no correlation between abnormal ultrasound findings and cytokine levels or

NRBC absolute numbers at 24 h postnatally. Although ultrasonography is a very good test for brain haemorrhage, ventricular dilatation and cystic periventricular leucomalacia, it is not specific for hypoxicischaemic lesions [52]. The use of magnetic resonance imaging, even during the first postnatal days, may offer more promising correlations [53]. In conclusion, increased absolute NRBC numbers and serum proinflammatory cytokine levels in asphyxiated/infected neonates represent early markers for subsequent neurologic impairment. These markers should be considered when interventions to prevent brain damage are being applied. Furthermore, strategies to prevent neonatal brain damage should include such modulations of fetal and neonatal inflammatory responses as anti-inflammatory cytokines, cytokine-binding proteins and cytokine receptor blockers [54].

References 1 Escobar GJ, Littenberg B, Petitt DB: Outcome among surviving very low birthweight infants: A meta-analysis. Arch Dis Child 1991;66:204–211. 2 Murphy DJ, Seller S, MacKenzie IZ, Yudkin PL, Johnson AM: Case-control study of antenatal and intrapartum risk factors for cerebral palsy in very preterm single babies. Lancet 1995;346:1449– 1454. 3 Hack M, Friedman H, Fanaroff A: Outcome of extremely low birth weight infants. Pediatrics 1996;98:931–937. 4 Perlman J: White matter injury in the preterm infant: An important determination of abnormal neurodevelopment outcome. Early Hum Dev 1998;53:99–120. 5 Vannucci RC: Current and potentially new management strategies for perinatal hypoxic-ischemic encephalopathy. Pediatrics 1990;85:961–968. 6 Goldberg RN, Morosco P, Bauer CR, Bloom FL, Curless RG, Burke B, Bancalari E: Use of barbiturate therapy in severe perinatal asphyxia: A randomized controlled trial. J Pediatr 1986;109:851– 856. 7 Martin-Ancel A, Garcia-Alix A, Pascual-Salcedo D, Cabanas F, Valcarce M, Quero J: Interleukin-6 in the cerebrospinal fluid after perinatal asphyxia is related to early and late neurological manifestations. Pediatrics 1997;100:789–794. 8 Leviton A, Paneth N, Reuss ML, Susser M, Allred EN, Dammann O, Kuban K, Van Marter LJ, Pagano M: Maternal intrauterine infection, fetal inflammatory response and brain damage in very low birth weight infants. Pediatr Res 1999;46:566– 575. 9 Dammann O, Leviton A: Maternal intrauterine infection, cytokines, and brain damage in the preterm newborn. Pediatr Res 1997;42:1–8. 10 Grether JK, Nelson KB, Emery ES III, Cummins SK: Prenatal and perinatal factors and cerebral palsy in very low birth weight infants. J Pediatr 1996; 128:407–414. 11 Murphy DJ, Hope PL, Johnson A: Neonatal risk factors for cerebral palsy in very preterm babies: Case-control study. BMJ 1997;314:404–408. 12 Yoon BH, Jun JK, Romero R, Park KH, Gomez R, Choi JH, Kim IO: Amniotic fluid inflammatory cytokines (interleukin-6, interleukin-1beta, and tumor necrosis factor-alpha), neonatal brain white matter lesions, and cerebral palsy. Am J Obstet Gynecol 1997;177:19–26.

NRBC and Cytokines as Markers of Brain Damage in LBW Neonates

13 Yoon BH, Romero R, Yang SH, Jun JK, Kim IO, Choi JH, Syn HC: Interleukin-6 concentrations in umbilical cord plasma are elevated in neonates with white matter lesions associated with periventricular leukomalacia. Am J Obstet Gynecol 1996; 174:1433–1440. 14 Naeye RL, Localio AR: Determining the time before birth when ischemia and hypoxemia initiated cerebral palsy. Obstet Gynecol 1995;86:713–719. 15 Green DW, Hendon B, Mimouni FB: Nucleated erythrocytes and intraventricular hemorrhage in preterm neonates. Pediatrics 1995;96:475–478. 16 Leikin E, Verma U, Klein S, Tejani N: Relationship between neonatal nucleated red blood cell counts and hypoxic-ischemic injury. Obstet Gynecol 1996;87:439–443. 17 Korst LM, Phelan JP, Ahn MO, Martin GI: Nucleated red blood cells: An update on the marker for fetal asphyxia. Am J Obstet Gynecol 1996;175: 843–846. 18 Buonocore G, Perrone S, Gioia D, Gatti MG, Massafra C, Agosta R, Bracci R: Nucleated red blood cell counts at birth as an index of perinatal brain damage. Am J Obstet Gynecol 1999;181:1500– 1505. 19 Perlman JM: Markers of asphyxia and neonatal brain damage. N Engl J Med 1999;341:364–365. 20 Badawi N, Kurinczuk JJ, Keogh JM, Alessandri LM, O’Sullivan F, Burton PR, Pemberton PJ, Stanley FJ: Intrapartum risk factors for newborn encephalopathy: The Western Australian case-control study. BMJ 1998;317:1554–1558. 21 Savman K, Blennow M, Gustafson K, Tarkowski E, Hagberg H: Cytokine response in cerebrospinal fluid after birth asphyxia. Pediatr Res 1998;43: 746–751. 22 Singh B, Merchant P, Walker CR, Kryworuchko M, Mitoma FD: Interleukin-6 expression in cord blood of patients with clinical chorioamnionitis. Pediatr Res 1996;39:976–979. 23 Berner R, Niemeyer CM, Leititis JU, Funke A, Schwab C, Rau U, Richter K, Tawfeek MSK, Clad A, Brandis M: Plasma levels and gene expression of granulocyte colony-stimulating factor, tumor necrosis factor-·, interleukin (IL)-1ß, IL-6, IL-8 and soluble intercellular adhesion molecule-1 in neonatal early onset sepsis. Pediatr Res 1998;44:469– 477. 24 Paterakis G, Lykopoulou L, Papassotiriou I, Stamouylakatou A, Loukopoulos D: Direct estimation of nucleated red cell counts in thalassemia and sickle cell disease by the R-1000 (SYSMEX) apparatus (abstract). Blood 1994;84:557.

25 Amiell-Tison C: Neuromotor status; in Taeusch HW, Yogman MW (eds): Follow-Up Management of the High-Risk Infant. Boston, Little, Brown, 1987, pp 115–126. 26 Nelson KB, Dambrosia JM, Grether JK, Phillips TM: Neonatal cytokines and coagulation factors in children with cerebral palsy. Ann Neurol 1998;44: 665–675. 27 Miller LC, Isa S, Lopreste G, Schaller JG, Dinarello CA: Neonatal interleukin-1ß, interleukin-6 and tumor necrosis factor: Cord blood levels and cellular production. J Pediatr 1990;117:961–965. 28 O’Shea TM, Klinepeter KL, Meis PJ, Dillard RG: Intrauterine infection and the risk of cerebral palsy in very low-birthweight infants. Paediatr Perinat Epidemiol 1998;12:72–83. 29 Spinillo A, Capuzzo E, Orcesi S, Stronati M, Di Mario M, Fazzi E: Antenatal and delivery risk factors simultaneously associated with neonatal death and cerebral palsy in preterm infants. Early Hum Dev 1997;48:81–91. 30 Yoon BH, Romero R, Kim CJ, Jun JK, Gomez R, Choi JH, Syn HC: Amniotic fluid interleukin-6: A sensitive test for antenatal diagnosis of acute inflammatory lesions of preterm placenta and prediction of perinatal morbidity. Am J Obstet Gynecol 1995;172:960–970. 31 Weeks JW, Reynolds L, Taylor D, Lewis J, Wan T, Gall SA: Umbilical cord blood interleukin-6 levels and neonatal morbidity. Obstet Gynecol 1997;90: 815–818. 32 Palmer C: Hypoxic-ischemic encephalopathy. Therapeutic approaches against microvascular injury and role of neutrophils, PAF and free radicals. Clin Perinatol 1995;22:481–517. 33 Fellman V, Raivio KO: Reperfusion injury as the mechanism of brain damage after perinatal asphyxia. Pediatr Res 1997;41:599–606. 34 Ghezzi P, Dinarello CA, Bianchi M, Rosandrich ME, Repine JE, White CW: Hypoxia increases production of interleukin-1 and tumor necrosis factor by human mononuclear cells. Cytokine 1991;3: 189–194. 35 Garcia JH, Liu KF, Yoshida Y, Lian J, Chen S, Zoppo GJ: Influx of leukocytes and platelets in an evolving brain infarct. Am J Pathol 1994;144:188– 199. 36 McRae A, Gilland E, Bona E, Hadberg H: Microglia activation after neonatal hypoxic-ischemia. Dev Brain Res 1994;84:245–252.

Biol Neonate 2001;79:213–218

217

37 Szaflarski J, Burtrum D, Silverstein FS: Cerebral hypoxia-ischemia stimulates cytokine gene expression in perinatal rats. Stroke 1995;26:1–8. 38 Hagberg H, Gilland E, Bona E, Hansson LA, Hahn-Zoric M, Holst M, McRae A, Soder O: Enhanced expression of interleukin (IL)-1 and IL-6 messenger RNA and bioactive protein after hypoxia-ischemia in neonatal rats. Pediatr Res 1996;40: 603–609. 39 Adinolfi M: The development of the human bloodCSF-brain barrier. Dev Med Child Neurol 1985; 27:532–537. 40 Megyeri P, Abraham CS, Temesvari P, Kovacs J, Vas T, Speer CP: Recombinant human tumor necrosis factor alpha constricts pial arterioles and increases blood-brain barrier permeability in newborn piglets. Neurosci Lett 1992;148:137–140. 41 Seo K, McGregor JA, French JI: Preterm birth is associated with increased risk of maternal and neonatal infection. Obstet Gynecol 1992;79:75–80. 42 Perlman JM: Markers of asphyxia and neonatal brain injury. N Engl J Med 1999;341:364–365.

218

43 Leviton A: Preterm birth and cerebral palsy: Is tumor necrosis factor the missing link. Dev Med Child Neurol 1993;35:553–558. 44 Lee SC, Dickson DW, Lin W, Brosnan CF: Induction of nitric oxide synthase activity in human astrocytes by interleukin-1ß and interferon-Á. J Neuroimmunol 1993;46:19–24. 45 Hattori H, Wasterlain CG: Excitatory amino acids in the developing brain, ontogeny, plasticity and excitotoxicity. Pediatr Neurol 1990;6:219–228. 46 Dammann O, Leviton A: Brain damage in preterm newborns: Might enhancement of developmentally regulated endogenous protection open a door for prevention. Pediatrics 1999;104:541–550. 47 Silverstein FS, Barks JDE, Hagan P, Liu XH, Ivacko J, Szaflarski J: Cytokines and perinatal brain injury. Neurochem Int 1997;30:375–383. 48 Volpe JJ: Neurology of the Newborn, ed 3. Philadelphia, Saunders, 1994, pp 876–884. 49 Perlman JM, Risser R: Can asphyxiated infants at risk for neonatal seizures be rapidly identified by current high-risk markers? Pediatrics 1996;97: 456–462.

Biol Neonate 2001;79:213–218

50 Gilles FH, Leviton A, Kerr CS: Endotoxin leucoencephalopathy in the telencephalon of the newborn kitten. J Neurol Sci 1976;27:183–191. 51 Yoon BH, Romero R, Kim CJ, Koo JN, Choe G, Syn HC, Chi JG: High expression of tumor necrosis factor-alpha and interleukin-6 in periventricular leukomalacia. Am J Obstet Gynecol 1997;177: 406–411. 52 Hope PL, Gould SJ, Howard S, Hamilton PA, Costello AM, Reynolds EOR: Precision of ultrasound diagnosis of pathologically verified lesions in the brains of very preterm infants. Dev Med Child Neurol 1988;30:457–471. 53 Battin M, Maalouf EF, Counsell S, Herithy AH, Edwards AS: Magnetic resonance imaging of the brain of premature infants. Lancet 1997;349: 1741. 54 Dammann O, Leviton A: Brain damage in preterm newborns: Biological response modification as a strategy to reduce disabilities. J Pediatr 2000;136: 433–438.

Fotopoulos/Pavlou/Skouteli/Papassotiriou/ Lipsou/Xanthou

Biol Neonate 2001;79:219–223

Prevention of Bilirubin Encephalopathy Giovanna Bertini Carlo Dani Marco Pezzati Firmino F. Rubaltelli Department of Critical Care Medicine and Surgery, Neonatal Intensive Care Unit, Careggi University Hospital, Florence, Italy

Key Words Bilirubin W Kernicterus W Bilirubin encephalopathy W Prematurity W Transcutaneous bilirubin evaluation

Abstract Prevention of bilirubin encephalopathy is based on the detection of infants at risk of developing a significant hyperbilirubinemia. This task can be accomplished by performing a simple umbilical cord blood test, such as blood group, Rh, Coombs’ test and glucose-6phosphate dehydrogenase, in order to detect hemolytic diseases. In preterm infants, the prevention of hyperbilirubinemia with phototherapy is a relatively simple task, since these infants are cared for in hospital. Early hospital discharge of full-term neonates represents a major concern. The management of neonatal jaundice requires that therapy begins when total serum bilirubin levels are significantly below the levels at which kernicterus is considered an immediate threat. Unfortunately, determination of serum bilirubin is a painful procedure, and is not very accurate since there is a high variability in laboratory measurements. The accuracy and precision of a new transcutaneous bilirubin measurement, comparable to the standard of care laboratory test, makes the daily evaluation of transcutaneous bilirubin measurement a useful tool in distinguishing physiological from nonphysiological hyperbilirubinemia, and determining the bilirubin increment in the first days of life. Full-term neonates who lose a significant amount of weight are especially at risk of significant hyperbilirubinemia and must be treated with ad libitum feeding and intensive phototherapy. Copyright © 2001 S. Karger AG, Basel

Introduction

In 1992, Newman and Maisels [1] suggested a less aggressive approach to the evaluation and treatment of jaundice in term neonates, affirming that many types of the then-recommended treatments were unnecessary, that fewer laboratory tests should be per-

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0219$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

formed and that higher treatment thresholds should be utilized. These suggestions were welcomed with great enthusiasm by many authors [2–9] and were incorporated by the American Academy of Pediatrics in their recommendations for the treatment of hyperbilirubinemia in the term neonate which were published 2 years later [10]. Unfortunately, in recent years, a few new cases of kernicterus have appeared and have been discussed in the literature, thus reopening the debate on the guidelines for the prevention of nuclear jaundice, which is the aim of hyperbilirubinemia treatment in neonates [11–15]. However, it should be noted that the majority of cases described in recent years have been neonates who had total serum bilirubin 1 30 mg/dl at the time of diagnosis and were therefore well above the recommended treatment threshold, demonstrating that the guidelines are not insufficiently aggressive, but rather that there is some carelessness on the part of clinicians with regard to their application [16]. Some writers have suggested that there has probably been a loss of historical memory with regard to the illness, resulting in the acquisition of a greater sensitivity towards the damaging effects of a possible treatment rather than towards the neurotoxicity of bilirubin [17]. It is also likely that the increasingly earlier release of neonates from maternity wards has a great influence on the difficulty in following the guidelines, so much so that the onset of severe jaundice is one of the major causes of the rehospitalization of neonates [18]. Certainly, it is now known that the maximum peak of bilirubin is evident at about the third or fourth day of life, yet more and more maternity wards release neonates 48 h after birth, with the risk of being unable to identify which neonates have significantly high bilirubin levels. Furthermore, the actual level of serum bilirubin above which cerebral damage is caused must still be determined, also because we still do not possess such precise diagnostic tools with which to easily measure bilirubin, that is, the amounts of bilirubin which are not bound to albumin, acting as the true cause of bilirubin encephalopathy. Therefore, the identification of factors favoring the onset of severe jaundice as well as the possibility of establishing specific and sensitive diagnostic methodologies which allow us to accurately monitor high-risk situations are extremely important.

Firmino F. Rubaltelli, MD Neonatal Intensive Care Unit Careggi University Hospital, Viale Morgagni, 85 I–50134 Florence (Italy) Fax +39 055 41 29 00, E-Mail [email protected]

Fig. 1. Normogram of transcutaneous bilirubin percentiles of healthy full-term male newborns. –––– = Mean; – – – = upper or lower limit for 95% confidence interval; W W W W W W = upper or lower limit for 99% confidence interval.

Clinical-Instrumental Diagnosis

For over 40 years, the measurement of serum has acted as the ‘clinical gold standard’ in the diagnosis and treatment of neonatal hyperbilirubinemia. The laboratory methodologies available (methods based on diazo reaction, direct spectrophotometry and reflectance spectrophotometry) still do not provide diagnostic precision due to considerable inter- and intralaboratory variability [19–21]; only HPLC is the method of reference that, more than anything, provides for the dosage of the various fractions, total, mono- and disconjugated, unconjugated and ‰-bilirubin, but unfortunately this methodology is not available in the clinical setting. In recent years, the need for instruments that measure bilirubin and provide precision monitoring, thus saving time during diagnosis and pointless neonatal suffering, has promoted the construction of new instruments for the transcutaneous measurement of bilirubin. These types of devices were already in use in the 1980s but had considerable clinical limitations with regard to race, gestational age (g.a.) and postnatal age of the subject under examination and the fairly high degree of intra- and interoperator variability recorded [22]. It was due to the aforementioned reasons that verification through laboratory analyses was required and that these instruments could only be used as a screening device. The new generation of instruments has sought to obviate these obstacles and are equipped with an algorithm that eliminates interference related to hemoglobin, cutaneous thickness and the type of skin pigmentation. Recently, Tayaba et al. [23] reported good

220

Biol Neonate 2001;79:219–223

results with a new device (Chromatics ColorMate III) that compares the change in the yellow coloration of the newborn’s skin to a baseline skin color measurement. The drawback to this method is that every baby born, regardless of their likelihood of developing subsequent hyperbilirubinemia, would require a baseline measurement. The additional complexity and cost of this practice is prohibitive for widespread use. Excellent results (also in terms of manageability) have been obtained with a new portable instrument that measures transcutaneous bilirubin, BiliCheckTM (SpectRx Inc., Norcross, Ga., USA) [24]. This device utilizes the entire spectrum of visible light (from 380 to 760 nm) reflected by the skin. White light is transmitted into the skin of the newborn and the reflected light is collected for analysis. By mathematically isolating the light absorption of certain interfering factors (hemoglobin, melanin and dermal thickness), the absorption of light due to the presence of bilirubin in the capillary beds and subcutaneous tissue can be isolated by spectral subtraction. In theory, this will allow for an unbiased measurement that is independent of the race, age and weight of the newborn. In a multicenter study performed with this device, the close correlation between BiliCheck and HPLC was shown to be equivalent to that between HPLC and the laboratory, so much so that the possibility of using the instrument not only as a screening device but also in place of total serum bilirubin testing should be considered. More than anything, transcutaneous bilirubin levels are certainly closer to free bilirubin levels, in that the mechanism by which bilirubin is deposited in the skin is

Bertini/Dani/Pezzati/Rubaltelli

Fig. 2. Normogram of transcutaneous bilirubin percentiles of healthy full-term female newborns. –––– = Mean; – – – = upper or lower limit for 95% confidence interval; W W W W W W = upper or lower limit for 99% confidence interval.

similar to the mechanism by which bilirubin passes through the hematoencephalic barrier. One may then hypothesize that the measurement of skin coloration could be a better predictive index of cerebral damage with relation to serum bilirubin concentration [25]. Through the use of this instrument in our neonatal unit and neonatal intensive care unit, we have set the goal of outlining percentiles for bilirubin levels within the first 5 days of life in vaginally born healthy term neonates who display no risk factors for the onset of significant hyperbilirubinemia. The aim was to have normal values as a reference point so that the early detection of at-risk neonates with the onset of severe jaundice could be performed. The neonates were divided according to sex and g.a. into the following 12 subpopulations: (1) females with g.a. 637 and ! 38 weeks; (2) females with g.a. 638 and ! 39 weeks; (3) females with g.a. 639 and ! 40 weeks; (4) females with g.a. 640 and ! 41 weeks; (5) females with g.a. 641 and ! 42 weeks; (6) females with g.a. 642 weeks; (7) males with g.a. 637 and ! 38 weeks; (8) males with g.a. 638 and ! 39 weeks; (9) males with g.a. 639 and ! 40 weeks; (10) males with g.a. 640 and ! 41 weeks; (11) males with g.a. 641 and ! 42 weeks, and (12) males with g.a. 642 weeks. In all neonates, transcutaneous bilirubin measurements were taken during the first 5 days of life. The first measurement was taken 24 h after birth and then measurements were taken successively at intervals of 24 h. Neonates released from the maternity ward before day 5 returned to our clinic each day at the scheduled time for transcutaneous bilirubin measurement. Neonates with serum bilirubin

The onset of severe neonatal jaundice is closely associated with certain universally known risk factors that are demonstrated especially in Rh factor incompatibility (which is fortunately extremely rare today), ABO incompatibility and glucose-6-phosphate dehydrogenase (G-6-PD) deficiency. It is for this very reason that it would be beneficial to determine blood type and Rh factor, perform the Coombs’ test and determine G-6-PD in cord blood at birth. However, there is a high percentage of neonates (55–56%) who have severe hyperbilirubinemia but do not display the known risk factors [26, 27]. The controversy in the literature regarding hyperbilirubinemia of unknown origin is whether breast-feeding increases the incidence of jaundice in the first days of life. Numerous authors [26–31]

Prevention of Bilirubin Encephalopathy

Biol Neonate 2001;79:219–223

levels requiring phototherapy were excluded from the study. Our results are still preliminary, but they are allowing us to outline the percentiles of transcutaneous bilirubin of the subpopulations examined (fig. 1, 2), with the exception of the females and males with g.a. 637 and ! 38 weeks and with g.a. 642 weeks, due to the scarcity of the number of patients in these g.a. groups. Once the data are complete, it will be possible to compare them with a table of normal bilirubin levels in the first week of life for each g.a., so that early detection can be achieved in those neonates requiring closer monitoring and special prevention of the probable risk factors.

Risk Factors

221

have reported a striking association between breast-feeding and significant hyperbilirubinemia, but other reports [32, 33] have not substantiated this observation. One of our soon-to-be-published studies [34], conducted on 2,174 neonates, demonstrates a statistically significant positive correlation between patients with a total serum bilirubin concentration 1 12.9 mg/dl (221 Ìmol/l) and supplementary feeding; in contrast, breast-fed neonates did not present a higher frequency of significant hyperbilirubinemia in the first days of life. Newborns with significant hyperbilirubinemia demonstrated a greater weight loss after birth when compared to the overall studied population, and infants given mixed feeding lost more weight than breastfed and formula-fed newborns, indicating the important role of fasting in the pathogenesis of neonatal hyperbilirubinemia. However, best gaussian fitting of our data suggests that a small subpopulation of breast-fed infants has a higher serum bililrubin peak than bottlefed infants. These infants, when starved and/or dehydrated, could probably be at high risk of bilirubin encephalopathy. It is likely that in the pathogenesis of significant hyperbilirubinemia in this subpopulation of breast-fed neonates, genetic factors may have an important role, as in Asian populations, for Gilbert’s syndrome and G-6-PD deficiency [35–37]. It is well known that environmental factors influence neonatal jaundice. The method of delivery also influences serum bilirubin concentration: cesarean sections preserve newborn infants from the development of neonatal hyperbilirubinemia, while vacuum extractor births are strongly associated with the onset of significant jaundice during the first days of life, probably due to the consequent development of hemorrhaging.

The Damaging Neurological Mechanism Caused by Hyperbilirubinemia

The structure and properties of unconjugated bilirubin and the way it is transported in the serum are aspects that are intrinsically tied to one another and to the neurotoxic mechanism of bilirubin. Unconjugated bilirubin is transported in blood that is mainly bound to albumin in the form of anion bilirubin. This molecule binds with phospholipids and is soluble in polar solvents. The bilirubin that is not bound to albumin (free bilirubin), which represents a small portion of physiological pH, binds with hydrogen ions, forming bilirubinic acid, which is insoluble in nonpolar solvents and tends to bind together and fall. The formation of this acid seems to be the principal damaging mechanism of the neuronal plasmatic membrane [38]. The bilirubin then passes through the hematoencephalic barrier and binds with phospholipids of the mitochondrion membrane, the endoplasmic reticulum and the nucleus, forming bilirubinic acid and resulting in neuronal failure; this is correlated with the probable damage of oxidative phosphorylation [39] and the homeostasis of the excitatory amino acids [40]. In short, the factors that can have an effect on neurological damage caused by bilirubin are tied to the concentration of bilirubin, the quantity of serum albumin, its capabilities and its capacity to bind with bilirubin, the concentration of hydrogen ions, the state of the hematoencephalic barrier and neuronal susceptibility. In turn, some clinical and/or metabolic conditions (asphyxia, metabolic acidosis, serum hyperosmolarity, hypercarbia, convulsions and meningitis) can have an effect on the formation of bilirubinic acid, damage the hematoencephalic barrier and favor the deposit of bilirubin on a neuronal level. What must be more precisely clarified is the problem of neuronal susceptibility; damage caused by bilirubin has a certain topographical distribution predominantly

222

Biol Neonate 2001;79:219–223

involving the corpus subthalamicum, hippocampus and adjacent olfactory areas, striate bodies, thalamus, globus pallidus, putamen, inferior clivus, cerebellar nuclei and cranial nerve nuclei. Certain hypotheses have been formulated (areas with increased cerebral flow or concomitant neurological damage or with increased endogenous production of billirubin are at greater risk while, in contrast, areas where bilirubin is degraded by bilirubin oxidase are protected), but no definitive conclusion has been reached.

Types of Treatment

Phototherapy and exchange transfusions remain the predominant treatments for neonatal hyperbilirubinemia. The advent of phototherapy was a moment of great importance in the history of this pathology, in that it dramatically decreased the need for exchange transfusions, a methodology frequently encumbered with complications. The line of conduct, however, must vary according to whether a term neonate or a preterm neonate is concerned. In preterm infants, the typical clinical feature of kernicterus is seen very rarely, and kernicterus is now a very infrequent postmortem observation. It is very difficult to give guidelines for the treatment of jaundice in very-low-birth-weight infants, other than to keep the serum bilirubin levels at a lower level than in full-term infants (e.g. 10 mg/dl lower than in full-term babies). In the term neonate, it is appropriate to recommend beginning phototherapy for bilirubin levels 1 10 mg/dl on the first day, 1 15 mg/dl on the second day and 1 17 mg/dl thereafter. Phototherapy should not be stopped prior to 24 h, or, at any rate, if the bilirubin level was 1 17 mg/dl, when it has dropped below 15 mg/dl. Exchange transfusions should be taken into consideration for bilirubin levels between 20 and 25 mg/dl on the third day and for levels that are lower on the first and second day of life. For phototherapy-treated infants, a new determination of total serum bilirubin is recommended 24 h after the discontinuation of phototherapy in order to evaluate a possible rebound. Exchange transfusions should be taken into consideration for bilirubin levels between 20 and 25 mg/dl on the third day and for levels that are lower on the first and second day of life. Intensive whole-body phototherapy and ad libitum feeding are employed as first-line treatment in every infant readmitted to hospital for severe jaundice, even when an exchange transfusion is planned. Intravenous immune globulin reduces jaundice in many cases of neonatal isoimmunization [41, 42]. The mechanism of intravenous immune globulin is probably similar to that found in neonatal isoimmune thrombocytopenia, such as the blockade of immunoglobulin constant fragment receptors and the resultant inhibition of hemolysis of antibody-coated erythrocytes. The recommended dose is 500 mg/ kg body weight, infused over a 2-hour period. The infant must be closely monitored for possible adverse effects, with particular attention to heart rate and blood pressure. Informed consent must be obtained from parents, since this treatment is still experimental. There are many pharmacological approaches to the prevention and/or treatment of neonatal hyperbilirubinemia. The most comforting results have been obtained through the use of mesoporphyrin, which blocks the heme oxygenase, a key enzyme in the synthesis of bilirubin [43]. In particular, a single small dose of Sn-mesoporphyrin, administered shortly after birth, can significantly moderate the severity of hyperbilirubinemia and reduce the need for phototherapy [44].

Bertini/Dani/Pezzati/Rubaltelli

References 1 Newman TB, Maisels MJ: Evaluation and treatment of jaundice in the term newborn: A kinder, gentler approach. Pediatrics 1992;89:809–818. 2 Wennberg RP: Bilirubin recommendations present problems: New guidelines simplistic and untested. Pediatrics 1992;89:821–822. 3 Merestein GB: New bilirubin recommendations questioned. Pediatrics 1992;89:822–823. 4 Poland RL: In search of a ‘gold standard’ for bilirubin toxicity. Pediatrics 1992;89:823–824. 5 Cashore WJ: Hyperbilirubinemia: Should we adopt a new standard of care? Pediatrics 1992;89: 824–826. 6 Gartner LM: Management of jaundice in the well baby. Pediatrics 1992;89:826–827. 7 Brown AK, Seidman DS, Stevenson DK: Jaundice in healthy, term neonates: Do we need new action levels or new approaches? Pediatrics 1992;89:827– 829. 8 Johnson L: Yet another expert opinion on bilirubin toxicity. Pediatrics 1992;89:829–831. 9 Valaes T: Bilirubin toxicity: The problem was solved a generation ago. Pediatrics 1992;89:819– 821. 10 Practice parameter: Management of hyperbilirubinemia in the healthy term newborn. American Academy of Pediatrics. Provisional Committee for Quality Improvement and Subcommittee on Hyperbilirubinemia. Pediatrics 1994;94:558–565. 11 Penn AA, Enzmann DR, Hahn JS, Stevenson DK: Kernicterus in a full term infant. Pediatrics 1994; 93:10003–10006. 12 Washington EC, Ector W, Abboud M, Ohning B, Holsen K: Hemolytic jaundice due to G6PD deficiency causing kernicterus in a female newborn. South Med J 1995;88:776–779. 13 MacDonald M: Hidden risks: Early discharge and bilirubin toxicity due to glucose-6-phosphate dehydrogenase deficiency. Pediatrics 1995;96:734– 738. 14 Sola A, Kitterman J: Changes in clinical practices and bilirubin encephalopathy in ‘healthy term newborns’. Pediatr Res 1995;37:145A. 15 Ebbesen F: Recurrence of kernicterus in term and near-term infants in Denmark. Acta Paediatr 2000;89:1213–1217. 16 Newman TB, Maisels MJ: Less aggressive treatment of neonatal jaundice and reports of kernicterus: Lessons about practice guidelines. Pediatrics 2000;105:242–245.

Prevention of Bilirubin Encephalopathy

17 Hansen TWR: Kernicterus in term and near-term infants – the specter walks again. Acta Paediatr 2000;89:1155–1157. 18 Maisels MJ, Kring E: Length of stay, jaundice, and hospital readmission. Pediatrics 1998;101:995– 998. 19 Vreman HJ, Verter J, Oh W, Fanaroff AA, Wright LL, Lemons JA, Shankaran S, Tyson JE, Korones SB, Bauer CR, Stoll BJ, Papile LA, Donovan EF, Ehrenkranz RA, Stevenson DK: Interlaboratory variability of bilirubin measurements. Clin Chem 1996;42:869–873. 20 Schreiner RL, Glick MR: Interlaboratory bilirubin variability. Pediatrics 1982;69:277–281. 21 Johnson KR: An external quality assessment scheme for total bilirubin. Ann Clin Biochem 1988;25:78–84. 22 Dai J, Parry DM, Krahn J: Transcutaneous bilirubinometry: Its role in the assessment of neonatal jaundice? Clin Biochem 1997;30:1–9. 23 Tayaba R, Gribetz I, Holzman IR: Noninvasive estimation of serum bilirubin. Pediatrics 1998; 102:E28. 24 Rubaltelli FF, Gourley GR, Loskamp N, Modi N, Roth-Kleiner M, Sender A, Vert P: Transcutaneous bilirubin measurement: A multi-centre evaluation of a new device. Pediatrics, in press. 25 Knudsen A, Brodersen R: Skin color and bilirubin in neonates. Arch Dis Child 1984;64:605. 26 Maisels MJ, Gifford K: Neonatal jaundice in fullterm infants. Am J Dis Child 1983;137:561–562. 27 Maisels MJ, Gifford K: Normal serum bilirubin levels in the newborn and the effect of breast-feeding. Pediatrics 1986;78:837–843. 28 Osborn LM, Reiff MI, Bolus R: Jaundice in the full-term neonate. Pediatrics 1984;73:520–525. 29 Maisels MJ, Gifford K, Antle CE, Leib GR: Jaundice in the healthy newborn infant: A new approach to an old problem. Pediatrics 1988;81:505– 511. 30 Linn S, Schoenbaum SC, Monson RR, Rosner B, Stubblefield PG, Ryan KJ: Epidemiology of neonatal hyperbilirubinemia. Pediatrics 1985;75:770– 774. 31 Schneider AP: Breast milk jaundice in the newborn. JAMA 1986;255:3270–3274. 32 Rubaltelli FF: Unconjugated and conjugated bilirubin pigments during perinatal development. IV. The influence of breast-feeding on neonatal hyperbilirubinemia. Biol Neonate 1993;64:104–109. 33 Nielsen HE, Haase P, Blaabjerg J, Stryhn H, Hilden J: Risk factors and sib correlation in physiological neonatal jaundice. Acta Paediatr Scand 1987; 76:504–511.

34 Bertini G, Dani C, Tronchin M, Rubaltelli FF: Is breast feeding favoring early neonatal jaundice? Pediatrics, in press. 35 Akaba K, Kimura T, Sasaki A, Tanabe S, Ikegami T, Hashimoto M, Umeda H, Yoshida H, Umetsu K, Chiba H, Yuasa I, Hayasaka K: Neonatal hyperbilirubinemia and mutation of the bilirubin uridine diphosphate-glucuronosyltransferase gene: A common missense mutation among Japanese, Koreans and Chinese. Biochem Mol Biol Int 1998;46: 21–26. 36 Yamamoto K, Sato H, Fujiyama Y, Doida Y, Bamba T: Contribution of two missense mutations (G71R and Y486D) of the bilirubin UDP glycosyltransferase (UGTLA1) gene to phenotypes of Gilbert’s syndrome and Crigler-Najjar syndrome type II. Biochim Biophys Acta 1998;1406:267–273. 37 Kaplan M, Muraca M, Hammerman C, Vilei MT, Leiter C, Rudensky B, Rubaltelli FF: Bilirubin conjugation, reflected by conjugated bilirubin fractions, in glucose-6-phosphate dehydrogenase-deficient neonates: A determining factor in the pathogenesis of hyperbilirubinemia. Pediatrics 1998; 102:E37. 38 Brodersen R, Stern L: Deposition of bilirubin acid in the central nervous system: A hypothesis for the development of kernicterus. Acta Paediatr Scand 1990;79:12–19. 39 Hansen TWR, Mathiesen SBW, Walaas SI: Bilirubin has widespread inhibitory effects on protein phosphorylation. Pediatr Res 1996;39:1072–1077. 40 Roseth S, Hansen TWR, Fonnum F, Walaas SI: Bilirubin inhibits transport of neurotransmitters in synaptic vesicles. Pediatr Res 1998;44:312–316. 41 Rubo J, Albrect K, Lasch P, et al: High-dose intravenous immune globulin therapy for hyperbilirubinemia caused by Rh hemolytic disease. J Pediatr 1992;121:93–97. 42 Hammerman C, Kaplan M, Vreman HJ, Stevenson DK: Intravenous immune globulin in neonatal ABO isoimmunization: Factors associated with clinical efficacy. Biol Neonate 1996;70:69–74. 43 Rubaltelli FF, Dario C, Zancan L: Congenital nonobstructive, nonhemolytic jaundice: Effect of tinmesoporphyrin. Pediatrics 1995;95:942–944. 44 Kappas A, Drummond GS, Henschke C, Valaes T: Direct comparison of Sn-mesoporphyrin, an inhibitor of bilirubin production, and phototherapy in controlling hyperbilirubinemia in term and nearterm newborns. Pediatrics 1995;95:468–474.

Biol Neonate 2001;79:219–223

223

Biol Neonate 2001;79:224–227

Inflammatory Mediators and Neonatal Brain Damage Elie Saliba Anne Henrot Department of Neonatology, INSERM U 316, University of Tours, Tours, France

Key Words Cytokines W Newborn W Premature W Hypoxia-ischemia W White matter damage

Abstract Inflammatory mediators are multifunctional cytokines that play important roles both in normal central nervous system (CNS) development and in the response of the brain to diverse forms of injury. Interleukin (IL)-1ß, tumor necrosis factor-· and IL-6 are among the best-characterized early-response cytokines. Recent data suggest that they may be synthesized and secreted by several CNS cell types, including microglia, astrocytes and neurons. Biological effects of these cytokines that could influence the progression of injury in the brain include stimulating the synthesis of other cytokines and neuronal injury mediators such as nitric oxide synthase, inducing leukocyte infiltration and the expression of adhesion molecules, influencing glial gene expression and damaging oligodendrocytes. In the immature brain, proinflammatory cytokines might lead to white matter damage during prenatal intrauterine infection and contribute to progressive neuronal damage in acute brain injury evoked by cerebral hypoxia-ischemia. Interrupting the proinflammatory cascade might limit the extent of irreversible injury. Copyright © 2001 S. Karger AG, Basel

Introduction

It is now well established that infection and inflammation play an important role in the pathogenesis of neonatal brain damage [1]. Brain cells can produce cytokines and chemokines, and can express adhesion molecules that enable an in situ inflammatory reaction. The accumulation of neutrophils early after brain injury is also believed to contribute to the degree of tissue loss. Intrauterine infection is now recognized as one important initiator of premature rupture of the membranes and as a risk factor for premature birth, especially birth

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0224$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

before 30 weeks of gestational age [2]. Such infections appear to lead to a fetal inflammatory response, which may damage developing white matter. This hypothesis is supported by epidemiological and experimental studies. Low gestational age is associated with a high risk of cerebral palsy and also with a high frequency of indicators of intrauterine infection [3]. Peritoneal administration of endotoxin to the immature rat results in the marked expression of cytokines in the white matter [4]. Brain ischemia is also associated with the expression of inflammatory mediators in the central nervous system (CNS). Hypoxia-ischemia induces an inflammatory response in the immature and adult rodent brain, characterized by the expression of interleukin (IL)-1 and tumor necrosis factor (TNF)-· [5, 6]. In this review, recent observations and basic scientific evidence are assembled to support the hypothesis that inflammatory mediators are involved in the pathogenesis of neonatal brain damage.

Cytokines in the Normal Brain

Several cytokines and their receptors have been identified in the brain and in cerebrospinal fluid. Historically, the stimulatory action of IL-1 on the hypothalamic-pituitary-adrenal axis and the unexpected localization of IL-1ß immunoreactivity neurons in the human hypothalamus have called attention to the role of cytokines in the brain. Since then, cytokines have been shown to play a key role in mediating communication between the endocrine and immune system. In attempting to understand the potential action of cytokines in the pathogenesis of perinatal brain injury, it is important to acknowledge that cytokines may also play an important role in normal brain development [7]. They are involved in the regulation of lineage commitment and cellular differentiation in the CNS. TNF-· is a pleiotropic cytokine released by many cell types. TNF-· and its receptors are present in the CNS. TNF-· can be synthesized and released by astrocytes, microglia and some neurons, and TNF can induce the proliferation of astrocytes. TNF has been shown to influence neuro-

Elie Saliba, MD Pediatric Intensive Care Hôpital Clocheville, Boulevard Tonnellé F–37000 Tours (France) Tel. +33 247 47 8093, Fax +33 247 47 3856, E-Mail [email protected]

nal progenitor cell proliferation and differentiation. Inducers of TNF production include IL-1, IL-2, interferons, endotoxin and mitogens. Suppressors of TNF production include IL-4, IL-6, transforming growth factor (TGF)-ß and dexamethasone. In the CNS, TNF-· induces fever and sickness behavior and directly induces the secretion of corticotropin-releasing hormone. IL-1, a pleiotropic cytokine with multiple biological actions in various organs, including the CNS, has been the focus of many studies addressing the action of cytokines in the brain. The IL-1 system consists of two receptor types, the IL-1 type 1 and type 2 receptors, and three peptides, IL-1·, IL-1ß and IL-1 receptor antagonist (IL1ra), which is a unique naturally occurring antagonist. IL-1 and IL-1ß activate the receptor, resulting in biological action, while IL-1ra completely blocks the receptor, preventing biological action. IL-1ß is converted from the pro-IL-1ß form to biologically active IL-1ß by ILconverting enzyme. CNS actions of IL-1 include the regulation of temperature, food intake and neuroendocrine function. IL1-ß can be produced in the CNS from various cellular elements, including microglia, astrocytes, neurons and endothelium. IL-1 also has neurotrophic properties; these neurotrophic effects may be mediated by the stimulation of nerve growth factor production. IL-1· and IL-1ß regulate the survival of fetal neurons in vitro. Direct intracerebral injection of IL-1 has been shown to stimulate astrogliosis and angiogenesis in the developing rodent brain. Similarly, direct intracerebral injection of ciliary neurotrophic factor and TNF-· synergistically stimulates both astrogliosis and the microglial response in neonatal rat brain. The responsiveness of immature glia and endothelial cells to such cytokines suggests that they may also influence normal maturational processes at this developmental age. IL-6 is also a pleiotropic cytokine that synergizes with IL-1 and TNF to costimulate immune responses. The major activities of IL-6 include inducing the acute-phase response in the liver and enhancing B cell replication, differentiation and immunoglobulin production. Many cell types, such as T and B lymphocytes, monocytes, endothelial cells and fibroblasts, can produce IL-6. The gene for IL-6 is located on human chromosome 7 and its expression may be induced by a variety of stimuli, including TNF and IL-1. IL-6 has prominent effects on the CNS; these central effects include activation of the hypothalamic-pituitary-adrenal axis, reduction of food intake, induction of fever and neuronal growth. Astrocytes and microglia can be stimulated in vitro to produce IL-6. TGF-ß was discovered as a growth factor for fibroblasts that promotes wound healing. TGF-ß has antiproliferative effects; it also suppresses the production of most cytokines that have chemoattractant activity for leukocytes and fibroblasts, and of cytokines that are produced by monocytes or macrophages. TGF-ß and several members of the TGF-ß superfamily, including bone morphogenetic proteins, growth and differentiation factors and glial cell line-derived neurotrophic factors, may play key roles in the development, repair and survival of neurons. Increased levels of TGF-ß1 mRNA and changes in its distribution patterns have been observed in various regions of the brain after ischemic insult and in response to kainic acid-induced injuries, where it may influence the postischemic neuron-glial interaction. Possible synthesis of TGF-ß in the CNS was inferred by high levels of TGF-ß in the cerebrospinal fluid of patients with multiple sclerosis in remission relative to patients in the active phase. TGF-ß1 mRNA levels have either been detected in the brain at very low levels, or found to be undetectable under basal conditions. It remains unclear whether the overproduction of TGF-ß1 in neurological conditions has primarily beneficial or detrimental effects in the CNS.

Cytokines and Perinatal Brain Injury

Cytokines during States of Infection and Inflammation

Systemic cytokines cause clear biological effects mediated through the CNS, such as fever, anorexia and activation of the hypothalamic-pituitary-adrenocortical axis. Therapeutic uses of cytokines and prolonged administration of interferons, ILs and TNF are described to be accompanied by a range of toxic effects that vary from mild flu-like symptoms to CNS toxicity. Cytokines can act in the brain through one or more of the following mechanisms: disruption of the brain-blood barrier, penetration into the brain through circumventricular organs, or in situ synthesis in the CNS. Several cell types within the brain are able to secrete cytokines: microglia, astrocytes, endothelial cells and neurons. In addition, there is also evidence to support the involvement of peripherally derived cytokines in brain inflammation. The blood-brain barrier is permeable to cells of the immune lineage [8]. Peripherally derived mononuclear phagocytes, T lymphocytes, natural killer cells and polymorphonuclear leukocytes, which produce and secrete cytokines, can all contribute to CNS inflammation and gliosis. The role of the cytokine cascade in the induction of preterm labor and premature rupture of the membranes has been discussed in detail by many authors [9–11]. Chorioamnionitis is associated with a 3-fold increased risk of delivery before term with intact membranes and a 4-fold increased risk of preterm birth due to prelabor rupture of the membranes. Bacterial vaginosis at 23–26 weeks of gestation increases the risk of preterm delivery by 50% [12]. TNF-· and IL-6 have been shown to be elevated in the amniotic fluid of pregnant women with chorioamnionitis [13]. Some of these intraamniotic cytokines that link intrauterine infection with preterm delivery seem to be of fetal origin. Fetal umbilical vein cytokine levels, but not maternal serum values, seem to correlate with the presence and severity of chorioamnionitis and umbilical vasculitis. In pregnancies complicated by premature rupture of the membranes, fetal plasma IL-6 levels were significantly higher in the infants who had a positive amniotic fluid culture than in those whose amniotic fluid was sterile. These observations support the hypothesis that the cytokine cascade leading to labor has fetal origins [14, 15]. IL-6 levels are higher in the cord blood of newborns whose mothers had clinical and histological chorioamnionitis than in unexposed controls [16]. The blood serum concentration of IL-6 is higher in infants with confirmed prenatal infection than in those with infections acquired after birth [17].

Cytokines as Mediators of Brain Injury

In the remainder of this paper, we will summarize the role of some cytokines (TNF-·, IL-1ß and IL-6) in brain injury. Ischemic Brain Injury Elevated TNF-· levels have been repeatedly shown in various experimental models of brain injury such as systemic kainic acid administration or intracerebroventricular injection of the excitotoxin ibotenic acid [18, 19]. Elevated expression of TNF-· mRNA and protein occurs shortly (1–3 h) after middle cerebral artery occlusion in rats [20]. Intracerebral injections of TNF-· 24 h before middle cerebral artery occlusion exacerbate ischemia-induced tissue injury [21]. TNF-· may prime the brain for subsequent damage by activating the capillary endothelium to a proadhesive state, possibly through the upregulation of surface endothelial adhesion molecules.

Biol Neonate 2001;79:224–227

225

The early expression of TNF-· mRNA preceding leukocyte infiltration suggests that TNF-· may be involved in this response. Data obtained in adult animals suggest that IL-1ß contributes directly to the pathogenesis of ischemic brain injury. An increase in IL-1ß mRNA expression has been shown to occur after several types of injury to the brain, including acute ischemic injury, kainate excitotoxicity and lipopolysaccharide endotoxin [22]. IL-1ß mRNA expression has been shown to increase after transient brain ischemia in the rat [23]. The exacerbation of ischemic brain injury caused by exogenous IL-1ß administration has also been observed [24]. The expression of IL-1ra and IL-1 receptor mRNA after focal brain ischemia has also been reported [25]. The upregulation of IL-1ra mRNA after ischemic injury suggests that IL-1ra may serve as a defense system to attenuate IL-1-mediated brain injury. Studies on the effects of exogenous IL-1ra have indicated that the blockage of endogenous IL-1 actions dramatically reduces brain damage caused by acute ischemic injury and excitatory amino acid infusion. Martin et al. [26] reported the first evidence that IL-1ra could protect the neonatal brain from injury. Focal Ischemia Induces Gene Expression in the Brain In response to ischemia, numerous proinflammatory genes are upregulated, including cytokines, chemokines and adhesion molecules that are likely to play a role in neutrophil and mononuclear cell infiltration that accumulate within vessels and tissues. The inflammatory response to tissue injury occurs after the excitotoxic cascade and contributes to the ongoing evolution of tissue injury. Immediate early response genes are expressed within 2 h, and then inflammatory cytokines are expressed and drive the inflammatory response [27]. Potential cellular targets for cytokines in the early postinjury period include astrocytes, microglia and neurons. IL-1ß stimulates gliosis and induced nitric oxide synthase [28]. IL-1ß also may directly influence endothelial gene expression; in vitro, IL-1ß stimulates endothelial cell expression of intercellular adhesion molecule-1 (ICAM-1) [29]. In addition, IL-1ß may exert direct effects on neurotransmission. IL-1ß and TNF-· may stimulate microglia to synthesize other cytokines that amplify the injury response. Tissue remodeling factors (e.g. TGF-ß) are expressed later and are associated with the healing of infarcted tissue. Hypoxia-Ischemia in the Immature CNS Hypoxia-ischemia in the immature CNS selectively stimulates IL-1ß and TNF-· gene expression in brain regions susceptible to irreversible injury in perinatal rats [5]. Production of IL-1ß and TNF-· is also induced in the developing brain by excitotoxin administration [18]. In term neonates, the magnitude of IL-6 levels in the cerebrospinal fluid after perinatal asphyxia is related to the severity of early neonatal hypoxic-ischemic encephalopathy and late neurological outcome [30].

later developed ultrasonographic echolucency and 21% in 33 controls. Inflammatory cytokines such as TNF-· damage oligodendrocytes [36]. Moreover, IL-6 directs the development of O-2A precursor cells in vitro away from becoming myelinating oligodendrocytes toward the path leading to astrocyte development [37]. These observations might explain both the hypomyelination and abundance of astrocytes in the brains of infants with WMD [38]. Maternal Antenatal Infection and Cerebral Palsy Nelson et al. [39] identified 31 full-term children born between 1983 and 1985 who developed cerebral palsy. They showed that many proinflammatory cytokines, chemokines and autoimmune and coagulation factors were higher in patients than in controls. In summary, cytokines in the amniotic fluid and fetal and neonatal blood, as a ‘humoral’ characteristic of the fetal inflammatory response, seem to increase the risk of neonatal brain injury and cerebral palsy. Recent observations from Damman and Leviton [40] support the hypothesis that fetal vasculitis, a ‘morphologic’ characteristic of fetal infection, is also associated with increased risk of brain injury.

Neuroprotection by Cytokine Inhibition

Anti-Inflammatory Cytokines Some recent reports have shown the neuroprotective effects of the anti-inflammatory cytokine IL-10 in both stroke and head trauma in experimental models [41, 42]. These findings may suggest that antiinflammatory modulatory cytokines may be of therapeutic potential in perinatal hypoxic-ischemic brain injury or to reduce inflammation-induced damage to preterm white matter. Receptor Antagonists As discussed earlier, many studies have shown the protective effects of IL-1ra in brain injury [25, 26]. In neonatal rats, exogenous IL-1ra reduces hypoxic-ischemic brain injury, and in adult rats, intracerebroventricular administration of recombinant IL-1ra produced a marked reduction in brain damage induced by focal stroke. Of interest are data showing that peripheral administration of IL-1ra reduces brain injury, suggesting a potential use in clinical studies [43]. Likewise, several studies have shown that blocking TNF-· results in reduced infarct volume and attenuated ICAM-1 expression. Pentoxifylline, a methylxanthine that reduces TNF-· production or TNF receptor 1, which acts by competing with TNF-· at the receptor, improves neurologic outcome, reduces the disruption of the blood-brain barrier and protects neurons from delayed cell death in animal models of head trauma [44].

Conclusion

White Matter Damage in the Preterm Newborn Yoon et al. [31, 32] showed that elevated cytokine levels in the amniotic fluid and umbilical cord are associated with white matter damage (WMD). TNF-· and IL-6 are significantly more often present in the brains of infants who die with WMD than among those who die without WMD [33]. Furthermore, infants with elevated IL-6 levels (111 pg/ml) in cordocentesis serum were twice as likely as infants with lower levels to have intraventricular hemorrhage and periventricular leukomalacia [34]. Martinez et al. [35] reported that the rate of positive amniotic fluid culture was 64% in 14 infants who

226

Biol Neonate 2001;79:224–227

A rapidly expanding body of evidence implicates diverse proinflammatory cytokines as potential mediators of perinatal brain injury. Hypoxia-ischemia and fetal infection appear to share some characteristics, including elevated levels of cytokines and adhesion molecules. The relevance of the actions of cytokines to a variety of neurological disorders has opened a potentially fruitful area of research and therapeutic development.

Saliba/Henrot

References 1 Dammann O, Leviton A: Maternal intrauterine infection, cytokines, and brain damage in the preterm newborn. Pediatr Res 1997;42:1–8. 2 Gomez R, Romero R, Edwin SS, David C: Pathogenesis of preterm labor and preterm premature rupture of membranes associated with intra-amniotic infection. Infect Dis Clin North Am 1997; 11:135–179. 3 O’Shea TM, Klinepeter KL, Meis PS, Dillard RG: Intrauterine infection and the risk of cerebral palsy in very low-birthweight infants. Paediatr Perinat Epidemiol 1998;12:72–83. 4 Cai Z, Pan ZL, Pang Y, Evans OB, Rhodes PG: Cytokine induction in fetal rat brains and brain injury in neonatal rats after maternal lipopolysaccharide administration. Pediatr Res 2000;47:64– 72. 5 Szaflarski J, Burtrum D, Silverstein FS: Cerebral hypoxia-ischemia stimulates cytokine gene expression in perinatal rats. Stroke 1995;26:1093–1100. 6 Garcia JH, Liu KK, Ye ZR: Cytokines and reperfusion in ischemic stroke. Brain Pathol 1997;7:1151– 1161. 7 Merrill JE: Tumor necrosis factor alpha, interleukin 1 and related cytokines in brain development: Normal and pathological. Dev Neurosci 1992;14: 1–10. 8 Giulian D, Chen J, Ingeman JE, George JK, Noponen M: The role of mononuclear phagocytes in wound healing after traumatic injury to the adult mammalian brain. J Neurosci 1989;9:4416–4429. 9 Gomez R, Hezzi F, Romero R, Munoz H, Tolosa JE, Rojas I: Premature labor and intra-amniotic infection. Clinical aspects and role of the cytokines in diagnosis and pathophysiology. Clin Perinatol 1995;22:281–342. 10 Seo K, McGregor JA, French JI: Preterm birth is associated with increased risk of maternal and neonatal infection. Obstet Gynecol 1992;79:75–80. 11 Mazor M, Cohen J, Romero R, Ghezzi F, Tolosa JE, Gomez R: Cytokines and preterm labor. Fetal Matern Med Rev 1995;7:207–233. 12 Hillier SL, Nugent RP, Eschenbach DA, Krohn MA, Gibbs RS, Martin DH, Cotch MF, Edelman R, Pastorek JG 2nd, Rao AV, McNellis D, Regan JA, Carey JC, Klebanoff MA: Association between bacterial vaginosis and preterm delivery of a lowbirth-weight infant. The Vaginal Infections and Prematurity Study Group. N Engl J Med 1995;333: 1737–1742. 13 Greig PC, Ernest JM, Teot L, Erikson M, Talley R: Amniotic fluid interleukin-6 levels correlate with histologic chorioamnionitis and amniotic fluid cultures in patients in premature labor with intact membranes. Am J Obstet Gynecol 1993;169: 1035–1044. 14 Salafia CM, Sherer DM, Spong CY, Lencki S, Eglinton GS, Parkash V, Marley E, Lage JM: Fetal but not maternal serum cytokine levels correlate with histologic acute placental inflammation. Am J Perinatol 1997;14:419–422. 15 Romero R, Gomez R, Ghezzi F, Yoon BH, Mazor M, Edwin SS, Berry SM: A fetal systemic inflammatory response is followed by the spontaneous onset of preterm parturition. Am J Obstet Gynecol 1998;179:186–193.

Cytokines and Perinatal Brain Injury

16 Singh B, Merchant P, Walker CR, Kryworuchko M, Diaz-Mitoma F: Interleukin-6 expression in cord blood of patients with clinical chorioamnionitis. Pediatr Res 1996;39:976–979. 17 Drews K, Szczapa J, Zak J, Andrzejewska R, Zak L, Mackiewicz A: Blood serum concentration of Creactive protein and interleukin-6 in diagnosis of neonatal infections. Ann NY Acad Sci 1995;762: 398–399. 18 Minami M, Kuraishi Y, Satoh M: Effects of kainic acid on messenger RNA levels of IL-1ß, IL-6, TNF· and LIF in the rat brain. Biochem Biophys Res Commun 1991;176:593–598. 19 Alvarez XA, Franco A, Fernandez-Novoa L, Carabelos R: Effects of neurotoxic lesion in histaminergic neurons on brain tumor necrosis factor levels. Agents Actions 1994;41:C70–C72. 20 Liu T, Clark RK, McDonnell PC, Young PR, White RF, Barone FC, Feuerstein GF: Tumor necrosis factor-· expression in ischemic neurons. Stroke 1994,25:1481–1488. 21 Barone FC, Arvin B, White RF, Miller A, Webb CL, Willette RN, Lysko PC, Feuerstein GZ: Tumor necrosis factor-·. A mediator of focal ischemic brain injury. Stroke 1997;28:1233–1244. 22 Buttini M, Boddeke H: Peripheral lipopolysaccharide stimulation induces interleukin-1ß messenger RNA in rat brain microglial cells. Neuroscience 1995;65:523–550. 23 Yabuuchi K, Minami M, Katsumata S, Yamzaki A, Satoh M: An in situ hybridization study on interleukin-1ß mRNA induced by transient forebrain ischemia in the rat brain. Brain Res Mol Brain Res 1994;26:135–142. 24 Yamasaki K, Matsuura N, Shozuhara H, Onodera H, Itoyama Y, Kogure K: Interleukin-1 as a pathogenetic mediator of ischemic brain damage in the rats. Stroke 1995;26:676–681. 25 Wang X-K, Barone FC, Aiyar NV, Feuerstein GZ: Increased interleukin-1 receptor and interleukin-1 receptor antagonist gene expression after focal stroke. Stroke 1997;28:155–162. 26 Martin D, Chinookoswong FN, Miller G: The interleukin-1 receptor antagonist (rhIL-1ra) protects against cerebral infarction in a rat model of hypoxia-ischemia. Exp Neurol 1994;130:362–367. 27 Barone FC, Feuerstein GZ: Inflammatory mediators and stroke: New opportunities for novel therapeutics. J Cereb Blood Flow Metab 1999;19:819– 834. 28 Lee SC, Dickson DW, Liu W, Brosnan CF: Induction of nitric oxide synthase activity in human astrocytes by interleukin-1ß and interferon-Á. J Neuroimmunol 1993;46:19–24. 29 Wong D, Dorovini-Zis K: Upregulation of intercellular adhesion molecule-1 (ICAM-1) expression in primary cultures of human brain microvessel endothelial cells by cytokines and lipopolysaccharide. J Neuroimmunol 1992;39:11–22.

30 Martin-Ancel A, Garcia-Alix A, Pascula-Salcedo D, Cabanas F, Valcarce M, Quero J: Interleukin-6 in the cerebrospinal fluid after perinatal asphyxia is related to early and late neurological manifestations. Pediatrics 1997;5:785–794. 31 Yoon BH, Jun JK, Romero R, Park KH, Gomez R, Choi JH, Kim IO: Amniotic fluid inflammatory cytokines (interleukin-6, interleukin-1 beta, and tumor necrosis factor-alpha), neonatal brain white matter lesions and cerebral palsy. Am J Obstet Gynecol 1997;177:19–26. 32 Yoon BH, Romero R, Ysang SH, Jun JK, Kim IO, Choi JH: Interleukin-6 concentrations in umbilical cord plasma are elevated in neonates with white matter lesions associated with periventricular leukomalacia. Am J Obstet Gynecol 1996;174:1433– 1440. 33 Yoon BH, Romero R, Kim CJ, Koo JN, Choe G, Syn HC, Chi JG: High expression of tumor necrosis factor-· and interleukin-6 in periventricular leukomalacia. Am J Obstet Gynecol 1997;177: 406–411. 34 Gomez R, Romero R, Ghezzi F, Yoon BH, Mazor M, Berry SM: The fetal inflammatory response syndrome. Am J Obstet Gynecol 1998;179:194– 202. 35 Martinez E, Figuora R, Garry D, Visintainer P, Patel K, Verma U: Elevated amniotic fluid interleukin-6 as a predictor of neonatal periventricular leukomalacia and intraventricular hemorrhage. J Matern Fet Invest 1998;8:101–107. 36 Selmaj KW, Raine CS: Tumor necrosis factor mediates myelin and oligodendrocyte damage in vitro. Ann Neurol 1988;23:339–346. 37 Kahn MA, de Vellis J: Regulation of an oligodendrocyte progenitor cell line by the interleukin-6 family of cytokines. Glia 1994;12:87–98. 38 Leviton A, Gilles F: Ventriculomegaly, delayed myelination, white matter hypoplasia, and ‘periventricular’ leukomalacia: How are they related? Pediatr Neurol 1996;15:127–136. 39 Nelson KB, Dambrosia JM, Grether JK, Philippe TM: Neonatal cytokines and coagulation factors in children with cerebral palsy. Ann Neurol 1998;44: 665–675. 40 Damman O, Leviton A: The role of the fetus in perinatal infection and neonatal brain damage. Curr Opin Pediatr 2000;12:99–104. 41 Spera PA, Ellison JA, Feuerstein GZ, Barone FC: IL-10 reduces rat brain injury following focal stroke. Neurosci Lett 1998;251:189–192. 42 Knoblach SM, Faden AL: Interleukin-10 improves outcome and alters proinflammatory cytokine expression after experimental brain injury. Exp Neurol 1998;153:143–151. 43 Relton JK, Martin D, Thompson RC, Russell DA: Peripheral administration of interleukin-1 receptor antagonist inhibits brain damage after focal cerebral ischemia in the rat. Exp Neurol 1996;138: 206–213. 44 Shohami E, Bass E, Wallach D, Yamin A, Gallily R: Inhibition of tumor necrosis factor alpha (TNF·) activity in rat brain is associated with cerebroprotection after closed head injury. J Cereb Blood Flow Metab 1996;16:378–384.

Biol Neonate 2001;79:224–227

227

Biol Neonate 2001;79:228–235

The Biology of Erythropoietin in the Central Nervous System and Its Neurotrophic and Neuroprotective Potential Christof Dame a Sandra E. Juul b Robert D. Christensen a a Division b Division

of Neonatology, Department of Pediatrics, University of Florida, Gainesville, Fla., of Neonatology, Department of Pediatrics, University of Washington, Seattle, Wash., USA

Key Words Erythropoietin W Erythropoietin receptor W Erythropoietin mimetic peptide W Neurodevelopment W Neuroprotection W Asphyxia W Intraventricular hemorrhage W Ischemia W Stroke

Abstract This review summarizes published as well as preliminary data on the biology of erythropoietin (Epo) in the developing and mature human central nervous system (CNS). Both Epo receptor (Epo-R) and Epo gene expression underlie developmental changes and a brain-specific regulation. These features suggest a different role of Epo in normal brain development than in neuroprotection and neuronal tissue repair after brain injury. Epo concentrations in the cerebrospinal fluid may have primary paracrine effects. While the transport of Epo across the intact blood brain barrier (BBB) is generally limited in humans, systemically produced or administrated Epo may cross during BBB dysfunction. Summarized data of the in vivo and in vitro effects of Epo in the CNS show significant neuroprotective and neurotrophic effects of this molecule. These effects are mediated by several mechanisms, including the activation of a variety of genes and their consecutive protein production. Therapeutic strategies involving activation of the CNS Epo-R are discussed, including the potential use of Epo mimetic peptides. Copyright © 2001 S. Karger AG, Basel

Introduction

A new field of scientific and clinical interest has been opened up based on the finding that erythropoietin (Epo) has substantial nonhematopoietic effects [1–6]. As we learn more about the plasticity of

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0228$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

pluripotent stem cells and their potential use for organ repair, certain nonhematopoietic effects of Epo which contribute to the specific microenvironment for stem cells become very compelling [7, 8]. It seems likely that future clinical practice involving neonatal neuroprotection and brain repair might utilize the nonhematopoietic effects of recombinant Epo (rEpo) or other substances such as Epo mimetic peptides (EMPs) that activate the Epo receptor (Epo-R) in the central nervous system (CNS). This review summarizes published studies and additional preliminary data of the authors on the biology of Epo in the CNS. It focuses primarily on human data, but it includes animal data and in vitro data on issues that cannot be investigated in humans.

Expression of the Epo-R in the CNS

The biological effects of Epo in the CNS are based on the expression of its specific receptor. Epo-R expression is abundant in the embryonic, fetal and adult brain, as has been shown in rats, mice, monkeys and humans [9–13]. Epo-R expression has also been documented in several neuronal cell lines, including NT2 cells (a human committed neuronal precursor cell line which can be induced to differentiate into postmitotic neurons), PC12 cells (a rat pheochromocytoma cell line which can differentiate into a neuronal phenotype), SN6 cells (a cholinergic hybridoma cell line with neuronal properties) and SK-N-MC cells (a human neuroblastoma cell line) [14–17]. These have allowed the establishment of various models to investigate the biological effects of Epo in the CNS [3, 4, 13, 18, 19]. Using immunofluorescence, it has been shown in unstimulated individual SK-N-MC cells that the Epo-R is localized not only on the plasma membrane but also in the cytoplasma. It is unclear whether this observation results from an internalization of the Epo-R, but it is of

Robert D. Christensen, MD Division of Neonatology, University of Florida College of Medicine PO Box 100296, JHMHC Gainesville, FL 32610-0296 (USA) Tel. +1 352 392 4195, Fax +1 352 392 4533, E-Mail [email protected]

interest for the exact understanding of mechanisms involved in the neuroprotective effects of Epo [19]. With immunochemical analysis in PC12 cells, the solubilized Epo-R proved to be smaller (62 kD) than that on rat erythroid cells (68 kD). Moreover, rEpo affinity to the Epo-R on neuronal cells was found to be lower than on erythroid cells [3]. The temporal and cellular distribution of the Epo-R in the embryonic, fetal and adult human CNS has been identified by RTPCR and immunohistochemistry. In the embryonic cerebral hemisphere 5–6 weeks after conception, Epo-R expression has been localized to undifferentiated neuroepithelial cells in the periventricular germinal zone [13, 20]. At 10 weeks of gestation, Epo-R was localized primarily to the subventricular zone, the neuropil and the cortical plate, while only low expression was found in the ventricular and matrix zone. Staining of the hippocampus and nucleus caudatus was weak or nonreactive at this stage of development. As development proceeds, the diffuse staining of broad zones of the developing neocortex seen in early gestation is replaced by more specific cellular staining of increasingly differentiated cells, e.g. the Cajal Retzius neurons. Epo-R expression can be detected in subpopulations of astrocytes within and around brain capillaries [13, 21]. Epo-R immunoreactivity has been found within the astrocytic foot processes surrounding the capillaries and also within endothelial cells [21]. Intense immunoreactivity for Epo-R has also been found in neurons (in a pattern restricted to the somata and proximal dendrites) and the choroid plexus [20, 21]. The pattern of cellular staining seems to be shifted from astrocytes, which are predominantly stained for the Epo-R early in development, to neurons in the mature brain [13, 20]. In adult humans, Epo-R gene expression has been described in the hippocampus, amygdala and temporal cortex [10, 20]. However, the level of Epo-R expression (amol/Ìg of total RNA) in human adult bone marrow is twofold higher than the level detected in adult brain as determined by a quantitative PCR [22]. Developmental regulation of Epo-R expression in the CNS has been confirmed in animal models. While Epo-R expression can be detected on embryonic day 10.5 in the mouse brain at similar levels as in hematopoietic tissue, Epo mRNA levels in the CNS decrease significantly with development [23]. Since Epo-R transcripts have a relatively short half-life (90 min in erythroid cells), the gene must be continuously transcribed to maintain high levels of Epo-R transcripts [24]. It has been shown that a region flanking the human Epo-R proximal promoter (exon 1 to exon 2) is necessary to actively drive Epo-R expression [12]. The proximal promoter of the Epo-R gene is functional in the brain, but detailed analysis of human and mouse Epo-R brain transcripts indicates an alternate processing or reduction in splicing efficiency compared with Epo-R on erythroid progenitor cells. In the human brain, Epo-R transcripts can also be initiated by a region far upstream of the Epo-R proximal promoter (exon A to exon B). The expression of these transcripts, in comparison to transcripts from the Epo-R-encoding regions (exon 1 to exon 2), is developmentally regulated. In the adult brain, transcripts of the proximal promoter region are expressed at levels 10-fold lower than in the developing brain, while in bone marrow, the relative amount of transcripts from the proximal region (exon 1 to 2) is clearly higher than in the developing and mature brain [22]. However, in transgenic mice, human mature Epo-R transcripts are induced twofold in the brain, and expressed 5-fold in the spleen as a hematopoietic organ, but not in the liver. These data indicate that the biological effectiveness of Epo in the CNS may be regulated at least in part by Epo-R expression.

In vivo and in vitro data show that in response to hypoxia or anemic stress, neuronal Epo-R expression is modified by a relatively greater increase in transcripts containing the appropriately spliced 5) coding region (exons 1 to 2) than in upstream transcripts (exon A to B). This indicates a shift towards increased sensitivity to Epo under injury [22]. These data confirm a tissue-specific and developmentally regulated Epo-R expression with remarkable uniqueness to the brain. Mice with a transgenic homozygous deletion of the Epo-R (EpoR–/–) die at day 13.5 of gestation due to a defect in ‘definitive’ erythropoiesis. However, the effect of the Epo-R null mutation in the CNS of these animals has not been investigated in detail. Limited data show that in the Epo-R–/– mouse, the proliferation of brain-derived cells, measured on embryonic day 11.5–13.5, is normal, whereas in the same animals, the number of proliferating cardiac myocytes in the ventricular myocardium is significantly decreased [25]. In contrast, a crucial role of endogenous brain Epo was shown in rats which were treated with soluble Epo-R (sEpo-R). This experiment was based on the concept that binding of endogenous Epo by sEpo-R would render Epo less available to the developing CNS cells. Indeed, treated animals showed neuronal degeneration and impaired learning ability [26].

Neurotrophic and Neuroprotective Effects of Erythropoietin

Biol Neonate 2001;79:228–235

Epo mRNA Expression in the CNS

Epo gene expression in the developing human CNS has been investigated in detail. Five to six weeks after conception, Epo reactivity can be localized by immunohistochemical staining in the periventricular germinal zone [20]. Ten weeks after conception, Epo stained most prominently in the ventricular and periventricular zones of the cerebral wall [20]. Later in development (around 20 weeks after conception), neuronal cytoplasmatic staining was noticed particularly in the thalamus, the hippocampus, the lateral geniculate body, the cortex and the spinal cord. The analysis suggests a pronounced staining with the progression of development [20]. We have also shown that during mid- and late gestation, Epo mRNA is expressed in various areas (myelencephalon, metencephalon, diencephalon and telencephalon) of the developing human CNS. Areas expressing the Epo gene include the spinal cord, cerebellum, pituitary gland, basal ganglia, thalamus, corpus geniculatum, corpus amygdalum, thalamus, hippocampus and cortex cerebri [27]. It is unknown whether the adenohypophysis and/or neurohypophysis express the Epo gene. Recently, we described a significantly higher level of expression of Epo mRNA per gram of tissue in the liver than in the brain. However, the amount of Epo mRNA per microgram of total RNA in the liver or kidney was not always higher than in various areas of the CNS of the same fetus [27]. The finding that Epo is expressed to a degree comparable to that in the liver and kidneys, which are the most relevant sites for circulating Epo at this stage of development, may indicate the impact of the biology of Epo in the developing CNS. This conclusion is strongly supported by evidence that in normoxic mice, high constitutive Epo mRNA expression occurs in the brain in comparison to the kidney and liver [10]. Variations in the levels and localization of Epo mRNA expression and the immunohistochemial staining for Epo in the developing human brain also suggest that Epo is involved in the neurogenesis of the brain. The Epo gene is expressed in astrocytes, but also in neuronal precursors and mature neurons and the choroid plexus [4, 13, 28].

229

In the human adult brain, similar quantitative Epo mRNA expression has been detected in all investigated areas, including the temporal cortex, the amygdala and the hippocampus [10]. Epo gene expression has also been found in adults with cerebral tumors, such as frontal meningioma and cerebellar hemangioblastoma [29, 30]. In vitro, Epo mRNA expression can be investigated in different cell lines, as in undifferentiated and differentiated NT2 cells and PC12 cells [13, 18].

redox state is involved in the regulation of Epo mRNA expression in neurons and astrocytes [28]. Like the hypoxia-induced stimulation of the Epo gene in the CNS, Epo production in astrocyte cell cultures can be stimulated in a dose-dependent manner by insulin and insulin-like growth factors (IGFs), particularly IGF-I [36]. This stimulation is not oxygen dependent and appears to be caused by the activation of the tyrosine kinase signal transduction pathway. Comparison with Hep3B cell cultures or uterine tissue cultures shows that the effect of insulin and IGFs on Epo production is specific for the brain [36].

Regulation of the Epo Gene in the CNS

Epo gene expression in the monkey and mouse brain, and in primary cell cultures, can be upregulated by hypoxia [9, 10]. Detailed analysis showed a tissue-specific regulation of the Epo gene. Two hours after exposure to 8% oxygen, Epo mRNA expression in monkey brain was increased 3-fold, whereas competitive RT-PCR analysis revealed a more than 200-fold increase in Epo mRNA in the kidneys. In a hypoxic mouse model (4 h of exposure to 0.1% carbon monoxide), Epo gene expression in the brain was induced 20-fold, compared to a 200-fold induction in the kidney. In primary astrocyte cultures, Epo mRNA levels can be upregulated more than 100-fold by hypoxia (1% oxygen) [10]. The Epo gene possesses a 50-base pair hypoxia response element (HRE) 120 base pairs downstream from the polyA signal. This HRE consists of three segments; one is the highly conserved binding site (5)CGTGC3)) of hypoxia-inducible factor-1 (HIF-1), which is located near the 5) end of the HRE. HIF-1 is a heterodimer consisting of HIF-1· and HIF-1ß. HIF-1· increases HIF-1 expression by hypoxia-induced stabilization, resulting in an increase in Epo mRNA expression. HIF-1ß is an aryl hydrocarbon receptor nuclear translocator-1 (ARNT-1) which is most likely constitutively expressed [31]. In the CNS of rats, HIF-1· is not only expressed with ARNT-1, but also with the recently identified ARNT-2, which is a cerebral translocator homolog of ARNT-1 with a selective neuronal expression [32]. Further, two splice variants of HIF-1· are expressed in the brain, one of which dimerizies with ARNT-2 even more avidly than with ARNT-1. However, the resulting heterodimer complex (HIF-1·/ ARNT-2) does not recognize the HIF-binding site of the Epo gene in the CNS [33]. The middle segment and the 3) segment of the HRE are important for the function of Epo gene regulation in humans. Although the middle segment is not well conserved in humans, a mutation of this region abolishes the function of the HRE. The 3) segment is a 2-bp DR-2 binding site, and Epo gene regulation can be modified by a variety of proteins binding to that region [31]. The role of this element of the HRE in the CNS has not yet been elucidated. It is strongly suggested that additional factors can substantially influence (as positive or negative regulating factors) Epo gene expression in the CNS, as described for hepatocyte nuclear factor-4, which increases the hypoxic inducibility of hepatic Epo gene expression [34]. Animal data show significant differences in the regulation of Epo gene expression between the kidney and the CNS. In murine kidneys, Epo mRNA expression increases with a maximum induction 2 h after the onset of hypoxia, followed by a decrease to 30% of the maximum level 8 h after the onset of continuous hypoxia. In contrast, the hypoxic stimulation of Epo mRNA expression in the cerebellum reached a peak at 4 h and remained similarly high for 24 h under continuous hypoxia [35]. Recently published data showed that, as in hepatoma cell lines, an oxygen-sensing mechanism that involves the

230

Biol Neonate 2001;79:228–235

Epo Protein in the Cerebrospinal Fluid

Epo protein is detected in cerebrospinal fluid (CSF) of human neonates and adults; the data from previous studies are summarized in table 1. In neonates, Epo is present in the CSF of both term and preterm infants (! 0.6–21 U/ml). Neonates treated with rEpo (1,200 U/kg/week s.c. or 1,400 U/kg/week i.v.) in order to prevent transfusions do not have higher levels of Epo in the CSF than do controls [37]. Epo CSF concentrations in infants are obviously lower than in preterm and term neonates [37]. Low Epo concentrations are still detectable in adults [37–40]. Information about Epo protein production in the CNS is provided from analysis of Epo concentrations simultaneously in CSF and plasma. Such samples were obtained from neonates and children with asphyxia, meningitis or intraventricular hemorrhage and controls. Both CSF and plasma Epo concentrations were significantly increased in asphyxia, whereas neonates suffering from intraventricular hemorrhage had elevated Epo concentrations in the CSF, but not in the plasma. In contrast, patients with meningitis had normal Epo concentrations in both CSF and plasma [41]. This raises the question of whether Epo expression in the CNS may be suppressed (or not upregulated) by inflammation, a question of substantial interest in regard to the role of rEpo in neuronal protection or repair. In adults with traumatic brain injury, Epo concentrations in the CSF correlate with the degree of blood brain barrier (BBB) dysfunction, as estimated by the CSF/serum albumin ratio [38]. The finding of somewhat higher Epo concentrations in the CSF of patients with depression, compared with controls, and the observation that this difference resolved after antidepressant therapy, is not understood [39]. In the CSF of adult patients investigated 5 months after brain injury due to infarction or bleeding, Epo concentrations were still moderately increased in comparison to controls [39]. While these data suggest that in humans, neither endogenous Epo nor rEpo cross the BBB, recently published animal data lead to the opposite conclusion. The differences could be specific to the species, or might depend on the degree of BBB damage or dysfunction. Brines et al. [21] showed that rEpo crosses the intact BBB in healthy adult rats by a specific and saturable transport mechanism. Biotinylated rEpo was detected surrounding the lumen of the capillaries, but not the large vessels, in the rat brain 5 h after intraperitoneal administration. In a subsequent experiment, the result was confirmed by documenting an increase in Epo concentrations (100 mU/ml) in rat CSF within 30 min after administration of 5,000,000 mU/kg [21]. We investigated the crossing of the intact BBB by rEpo under the condition of long-term application of a lower dose. Healthy neonatal rats received rEpo (200,000 mU/kg/day s.c.) for 7 days (experiment A) or 12 days (experiment B). Paired serum and CSF samples were obtained 24 h after the last dose (experiment A) or within 1–12 h

Dame/Juul/Christensen

Table 1. Epo concentrations in human CSF Study

Method

Patients

rEpo treatment

Epo concentration in serum Epo concentration in CSF or plasma, mU/ml mU/ml [mean B SD; [mean B SD; median (range)] median (range)]

Marti et al. 1997 [38]

RIA

Adults (n = 5; 25–59 years) with traumatic brain injury; paired plasma/CSF specimens (n = 15)

no

39.5B3.5

1.12B0.11 (n = 9) (intact BBB) 1.33 (n = 2) (moderate BBB dysfunction) 1.6B0.27 (n = 4) (severe BBB dysfunction)1

Juul et al. 1997 [37]

ELISA (R&D)

Preterm and term neonates (n = 25; GA 24 weeks to term); 5 of 25 with paired samples Preterm and term neonates (n = 23; GA 24 weeks to term) Controls (n = 30; 1 month to 60 years)

no

(1–56)

7.6B1.4 5.0 (! 0.6–21) (0.6–12) 9.9B2.4 (0.6–21.0) 2.8B1.2 0.6 (! 0.6–32)2

Adults with depressions (n = 13)

no

Nakamura RIA et al., 1998 [39] (Chugai)

yes no

3.21B0.46 after treatment: 1.56B0.343 1.8B0.32

Adults with old cerebrovascular diseases (n = 10)4 Controls (n = 10) Juul et al. 1999 [41]

Buemi et al. 2000 [40]

1 2 3 4 5

ELISA (R&D)

ELISA (Genzyme)

0.98B0.26

Neonates with asphyxia (n = 16)

no

Neonates with IVH (n = 11; BW ! 1,500 g) Neonates (n = 13), infants and children with meningitis (n = 13; 6 weeks to 12 years) Preterm and term controls (n = 11) Preterm and term controls (n = 20; GA 24 weeks to term) and healthy children (n = 20; 1 month to 12 years)

no

adult (44 years)

yes basal: 12.9 (6,000 IU) after 1 min: 6,960 after 60 min: 6,660

no

yes no

1,806B1,254 (6.7–18,501) 29.8B14.3 (4.7–164.2) 29.1B81.6 (3.0–416.0)

225B155 (6.7–2,350)5 20.6–9.9 (4.3–63.1) 14B45.7 (1.1–255.0)

(11–306.5) 5.6B0.8

(0.6–12) 5.8B0.7

basal: 5.8 after 1 min: 4.6 after 60 min: 4.1

RIA = Radioimmunoassay; GA = gestational age; IVH = intraventricular hemorrhage; BW = body weight. BBB dysfunction is defined by an elevated CSF/serum albumin ratio. In all samples of individuals older than 5 months, the CSF Epo concentration was lower than 3.0 mU/ml. After 5 months of treatment with imipramine and/or nortriptyline. Samples were obtained 5 months after bleeding or infarction. A positive correlation between Epo concentrations in plasma and CSF was found (r = 0.996, p ! 0.0001).

(experiment B). The results in the treated animals and controls are summarized in table 2. Despite significantly higher Epo serum concentrations and hematocrits compared to controls, Epo was detectable in the CSF of only 2 rEpo-treated animals (10 and 21 mU/ml, both in experiment A). In both cases, the CSF was visibly contaminated with blood. These data indicate a limited capacity to transport rEpo over the intact BBB.

In a rabbit model of subarachnoidal hemorrhage-induced acute cerebral ischemia, animals systemically treated with rEpo (1,000 U/ kg/day i.p.) had significantly higher Epo concentrations in CSF (23 days after initiation of ischemia) than placebo-treated or control animals. This suggests that rEpo may cross the BBB in higher amounts if brain injury has occurred [42]. Masuda et al. [4] partially purified Epo produced by primary rat brain cultures. Purified brain Epo had a higher biological activity on

Neurotrophic and Neuroprotective Effects of Erythropoietin

Biol Neonate 2001;79:228–235

231

Table 2. Epo concentrations (median and range) in the serum and CSF of neonatal rats

Group 1 (rEpo)

Experiment Dose

Hct, %

Epo in serum mU/ml

Epo in CSF mU/ml

A (n = 5) B (n = 6)

rEpo 200 U/kg for 7 days rEpo 200 U/kg for 12 days

37.5 (25.6–41.0) 36.4 (31.3–39.7)

24.2 (16.2–48.0) 502 (412–570)

9.2 (! 1.5–21.0) ! 1.5 (all ! 1.5)

saline

28.7 (23.3–30.9)

! 1.5 (! 1.5–8.8)

! 1.5

Group 2 (control) (n = 4) Hct = Hematocrit.

the Epo-dependent growth of erythroid colonies than did Epo isolated from rat serum. Western blot analysis showed that a smaller molecular size and a different sialysation could account for a higher receptor affinity and higher biological activity [4].

In vivo Effects of rEpo in Models of Experimental Brain Injury

In multiple mouse, rat, gerbil or rabbit models, the in vivo effects of rEpo in the brain have been investigated. These studies show that Epo can improve the functional properties and diminish the anatomic destruction after brain injury. The role of Epo in normal brain has been demonstrated by evidence of neuronal degeneration and learning disability as a consequence of the sEpo-R-mediated binding of endogenous Epo protein from CSF. Further evidence that Epo is responsible for CNS-protective effects is derived from experiments in which the administration of sEpo-R into the CNS resulted in an exacerbation of ischemic stress with increased neuronal injury. Infusion of rEpo into the lateral ventricle of gerbils with occluded common carotid arteries prevented ischemia-induced learning disability and rescued hippocampal CA1 neurons from lethal damage [26]. The maintenance of neuronal functions has also been shown in a model of stroke-prone spontaneously hypertensive rats with permanent occlusion of the middle cerebral artery. Intraventricular rEpo administration preserved place navigation, limited the degree of cortical infarction and supported neuronal survival in the characteristically secondarily impaired ventroposteral thalamic nucleus. The crucial role of the Epo-R in the CNS was confirmed by in situ hybridization, showing upregulation and continuous expression of the Epo-R (24 h after brain injury) in the periphery (ischemic penumbra) of the cerebrocortical infarct area [43]. Additional experiments in a mouse stroke model induced by occlusion of the middle cerebral artery showed significant reduction of the infarct volume in animals receiving rEpo systemically (i.p.) 24 h before, or up to 3 h, but not 6 or 9 h after infarction. Optimal effects were found with doses ranging from 450 to 5,000 U/kg body weight [21]. Evidence is also given for markedly reduced inflammatory infiltration after systemic rEpo administration in a mouse stroke model with a blunt trauma to the calvaria [21]. In rats with experimentally induced autoimmune encephalomyelitis (using guinea pig myelin basic protein and complete Freund’s adjuvant), systemic administration of rEpo delays the onset and the severity of clinical symptoms. This suggests that Epo affects both the inflammatory and immune responses in the brain [21]. rEpo also modulates the neuronal excitability in a mouse model with kainate-induced seizures. Mice treated

232

Biol Neonate 2001;79:228–235

24 h before kainate infusion showed a significant delay of onset of status epilepticus with a markedly reduced motor involvement and a significant reduction in the mortality rate compared with controls. Since this effect could not be shown if rEpo was administrated only 30 min before the kainate application, it is suggested that an array of genes affected by rEpo are responsible for the biological effect [21]. Other studies have showed that rEpo also promoted the survival of septal cholinergic neurons in adult rats which had undergone fimbria-fornix transsections [18]. There is also evidence that rEpo has neuroprotective effects in rabbits with subarachnoidal hemorrhage [44]. These in vivo studies allow the conclusion that rEpo acts in the CNS by a variety of mechanisms, including antiapoptotic and antiinflammatory mechanisms, and immune or electrophysiological modulation.

In vitro Effects of rEpo on Primary Brain-Derived Cell Cultures and Neuronal Cell Lines

In vitro experiments using primary cell/organ cultures or undifferentiated and differentiated cell lines confirmed the results of in vivo studies and allow conclusions regarding the mechanisms by which Epo mediates its cell-protective effects in the CNS. rEpo decreases cell death when cultured neurons and neuronal precursors (taken from the murine cortex, hippocampus or whole brain, or NT2, hNT, SN6 or PC12 cell cultures) are exposed to hypoxia, glutamate or UV irradiation in a dose-dependent manner. To achieve an optimal neuroprotective effect, which can be completely reversed by the coapplication of sEpo-R or anti-Epo antibody, exposure of the cells to rEpo is necessary 4–8 h prior to cell injury [11, 26, 45, 46]. However, the exposure to rEpo for a short period (5 min or less) is also sufficient to elicit protective effects on neurons [11]. In contrast to neurons, significant protection of cultured rat cortical astroglia exposed to hypoxia with glucose deprivation could not be shown [47]. The neuroprotective effects of rEpo have been shown using different methods to detect cell death or injury, including the release of lactate dehydrogenase or nuclear matrix protein into the supernatant, the detection of DNA fragmentation by Klenow reaction, the quantification of trypan blue exclusion, the detection of cell viability by the fluorescence of Alamar blue dye and the analysis of intact synapses. Different mechanisms have been identified as mediating the effects of Epo in the CNS. The binding of Epo on PC12 and SN6 cells increases the intracellular concentrations of monoamines [3]. In primary cultured hippocampal and cerebral cortical neurons, it has been confirmed that Epo does not suppress the N-methyl-D-aspartate

Dame/Juul/Christensen

receptor-mediated increase in intracellular Ca2+ concentrations, which is believed to cause glutamate toxicity as a major cause of neuron death by ischemia [11, 26]. Epo may reduce the formation of free radicals or antagonize their toxicity [26]. PC12 and SK-N-MC cell experiments confirmed that rEpo stimulates cell viability by the activation of Ca2+ channels and mitogen-activated protein kinase [19, 45]. Because rEpo exerts the greatest neuroprotective effects when administered prior to cell injury, it has been concluded that de novo synthesis of proteins is required for exertion of the neuroprotective effects of Epo. It is suggested that Epo thereby increases the activity of antioxidant enzymes such as superoxide dismutase, glutathione peroxidase and catalase in neurons [26]. As in the in vivo situation, the induction of Epo-R expression also partly accounts in NT2 cells for an increased sensitivity and responsiveness to Epo. This effect is mediated by the activation of transcripts from the upstream region and the proximal promoter and is associated with a shift toward appropriate and effective splicing/processing of Epo-R transcripts (exon 1 to exon 2) [22]. In rEpo-treated NT2 cells, differential display PCR allowed us to identify 26 different regulated DNA sequences, including ATP synthase A chain gene homologons, a sequence encoding pre-mRNA splicing factors (SR family) [48]. Preliminary data from the analysis of changes in gene expression in PC12 cells treated with rEpo compared with untreated cells indicated a significant induction of the expression of various genes. This included members of the growth and differentiation systems as well as factors involved in apoptosis, including NF-Îb, IKKa, akt and several Bcl family members [49]. The concept that Bcl-x family members or JAK/STAT transcription factors (e.g. STAT5) are critically involved is supported by findings in erythroid progenitor cells [50–52]. Since Epo-R are also expressed on brain endothelial cells, macroglia/macrophage-like cells and reactive astrocytes, Epo might also support repair mechanisms by the modulation of angiogenesis and tissue oxygenation in the border zone of the ischemic area [53]. The variety of effects mediated by rEpo in the CNS is wide. rEpo augments choline acetyltransferase in cultured mouse embryonic primary septal neurons [18]. Epo affects in vitro neuronal activity by an increase in the dopamine release and tyrosine hydroxylase activity in PC12 cells, both directly or indirectly induced by Ca2+-dependent protein kinases or enzymes. Epo may also stimulate dopamine release, partly by nitric oxide production [45].

Neurotrophic Effects of Epo

An amino sequence comparison between Epo, thrombopoietin, nerve growth factor-ß and neutrophins revealed areas of high homology [54]. A 17-mer peptide with potential neurotrophic function has been identified in the Epo protein, which extends from amino acid 29 to the AB loop of the molecule [55]. Neurotrophic effects of rEpo have been shown on septal cholinergic neurons in vivo (adult rat) and in vitro (primary culture from mouse brain-derived cells and SN6 cells) [18]. This effect has been confirmed (serum and nerve growth factor free) in PC12 cell cultures and NT2 cell cultures [17] [Dame et al., unpubl. data]. As shown by the significant increase in BrdU incorporation, indicating a higher mitotic activity, in NT2 cells exposed to UV irradiation, it is suggested that the neurotrophic effects may contribute to the presentation of a lower degree of anatomic damage in in vivo models of induced brain injury.

Neurotrophic and Neuroprotective Effects of Erythropoietin

Therapeutic Strategies

Clinical data on the neuroprotective or neurotrophic effects of rEpo in neonates are not available. Such data cannot be taken from the controlled trials of rEpo administration to preterm neonates, since those studies were conducted to prevent or treat anemia of prematurity, and hypoxemic and ill patients were generally excluded from enrollment. Similarly, in adults, no data on the clinical application of rEpo as a neuroprotective agent are available. However, intense discussion on the neuroprotective properties of Epo at the 5th International Lübeck Conference on the Pathophysiology and Pharmacology of Erythropoietin (June 2000) suggested that the first clinical data will come from adult stroke patients treated with rEpo to improve tissue oxygenation. The first treatment strategies should respect the critical window, which may be open for up to 3 h after acute brain injury, to achieve a successful therapeutic intervention. It remains unknown to what degree the BBB will limit the clinical benefit of rEpo administrated systemically. The systemic delivery of rEpo as a neurotherapeutic agent over the BBB would have the advantage that it is available to the capillary endothelium, and thereby present everywhere in the brain. In contrast, intrathecally applied rEpo would be highly localized. The therapeutic applicability of rEpo as a neuroprotective agent may be limited, since crossing of the (intact) BBB may be restricted, due to a saturable transport mechanism or the function of the BBB. An alternative could be the use of EMPs. These peptides are a family of Epo-R agonists that mimic the erythropoietic effects of rEpo in vitro and in vivo. These peptides contain between 9 and 22 amino acids and do not share sequence homology with Epo [56, 57]. The opportunity to modulate biological characteristics of EMPs (such as increasing Epo-R affinity and biological potency by dimerization of the EMP) and the potential for synergy make these agents very attractive for therapeutic purposes in specific conditions. Therefore, we recently compared the neurotrophic and neuroprotective effects of EMP1 (20 amino acids; molecular weight 2,092 D) with those of rEpo (37,000 D) in undifferentiated NT2 cells. We found that EMP1 has similar neuroprotective effects as rEpo. These effects were present irrespective of the mechanism of injury (hypoxia or UV irradiation). As for rEpo, a pretreatment interval of at least 8 h is necessary to achieve significant neuroprotective effects. This may indicate that the expression of antiapoptotic genes has a critical role in the effects mediated by brain Epo-R activation [46]. Our data demonstrate an important difference between rEpo and EMP1 on neuronal cells that are in the process of differentiation. Pretreatment with rEpo in vitro results in greater cell proliferation (in sets with and without artificial cell injury) in comparison both to controls and EMP1-treated cells. This effect may be mediated by the 17-amino acid neurotrophic sequence in the Epo molecule [55]. We speculate that the neurotrophic sequence of Epo might not be exclusively mediated by the activation of the Epo-R. These data suggest that a selective use of rEpo and EMP1 could be of interest in specific strategies utilizing the effects of Epo-R activation in the CNS. Before any clinical inference can be made about the feasibility or practicability of EMP1 as a neurotherapeutic agent, in vivo data must be obtained. Little is known about the pharmacology of EMP1, other than that it has a half-life of 8 h in isolated mouse serum [57]. However, our preliminary data indicate substantial differences in the kinetics and biological availability of rEpo and EMP1. In addition to the experiments described above (table 2), one group of healthy

Biol Neonate 2001;79:228–235

233

Sprague-Dawley rat pups received a single dose of EMP1 [50 Ìg (equivalent dose to 200 U/kg rEpo) or 100 Ìg i.p., n = 12 and 6 animals, respectively]. Using liquid chromatography mass spectral analysis, EMP1 was undetectable in serum and CSF obtained 1 and 2 h following intraperitoneal injection of EMP1 [Dame et al., unpubl. data]. This may be the result of poor resorption, rapid binding and denaturation, or ‘first pass’ clearance. Further studies are ongoing to elucidate the pharmacokinetics of EMP1.

opmental differences in Epo-R gene expression in the CNS, and the specific regulation of Epo-R transcripts in response to hypoxia, also indicate a different biology of Epo in nonhematopoietic organs. Cell proliferation or cell-protective mechanisms can be selectively induced and controlled during development or in organ injury. This makes rEpo or EMPs very attractive agents for nonhematopoietic applications in organ protection and repair.

Acknowledgment Summary

The specific regulation of Epo gene expression, and the data on Epo protein concentrations in the CSF, suggest that Epo generally acts in the CNS as a paracrine factor. The tissue-specific and devel-

Christof Dame was supported by the Deutsche Forschungsgemeinschaft (DA 484/1-1).

References 1 Anagnostou A, Lee ES, Kessimian N, Levinson R, Steiner M: Erythropoietin has a mitogenic and positive chemotactic effect on endothelial cells. Proc Natl Acad Sci USA 1990;87:5978–5982. 2 Tan CC, Eckardt KU, Firth JD, Ratcliffe PJ: Feedback modulation of renal and hepatic erythropoietin mRNA in response to graded anemia and hypoxia. Am J Physiol 1992;263:F474–F481. 3 Masuda S, Nagao M, Takahata K, Konishi Y, Gallyas F Jr, Tabira T, Sasaki R: Functional erythropoietin receptor of the cells with neural characteristics. Comparison with receptor properties of erythroid cells. J Biol Chem 1993;268:11208–11216. 4 Masuda S, Okano M, Yamagishi K, Nagao M, Ueda M, Sasaki R: A novel site of erythropoietin production. Oxygen-dependent production in cultured rat astrocytes. J Biol Chem 1994;269:19488– 19493. 5 Wald MR, Borda ES, Sterin-Borda L: Mitogenic effect of erythropoietin on neonatal rat cardiomyocytes: Signal transduction pathways. J Cell Physiol 1996;167:461–468. 6 Yasuda Y, Nagao M, Okano M, Masuda S, Sasaki R, Konishi H, Tanimura T: Localization of erythropoietin and erythropoietin-receptor in postimplantation mouse embryos. Dev Growth Differ 1993;35:711–722. 7 Schmitt RM, Bruyns E, Snodgrass HR: Hematopoietic development of embryonic stem cells in vitro: Cytokine and receptor gene expression. Genes Dev 1991;5:728–740. 8 Heberlein C, Fischer KD, Stoffel M, Nowock J, Ford A, Tessmer U, Stocking C: The gene for erythropoietin receptor is expressed in multipotential hematopoietic and embryonal stem cells: Evidence for differentiation stage-specific regulation. Mol Cell Biol 1992;12:1815–1826. 9 Digicaylioglu M, Bichet S, Marti HH, Wenger RH, Rivas LA, Bauer C, Gassmann M: Localization of specific erythropoietin binding sites in defined areas of the mouse brain. Proc Natl Acad Sci USA 1995;92:3717–3720. 10 Marti HH, Wenger RH, Rivas LA, Straumann U, Digicaylioglu M, Henn V, Yonekawa Y, Bauer C, Gassmann M: Erythropoietin gene expression in human, monkey and murine brain. Eur J Neurosci 1996;8:666–676. 11 Morishita E, Masuda S, Nagao M, Yasuda Y, Sasaki R: Erythropoietin receptor is expressed in rat hippocampal and cerebral cortical neurons, and erythropoietin prevents in vitro glutamate-induced neuronal death. Neuroscience 1997;76:105–116.

234

12 Liu C, Shen K, Liu Z, Noguchi CT: Regulated human erythropoietin receptor expression in mouse brain. J Biol Chem 1997;272:32395– 32400. 13 Juul SE, Anderson DK, Li Y, Christensen RD: Erythropoietin and erythropoietin receptor in the developing human central nervous system. Pediatr Res 1998;43:40–49. 14 Pleasure SJ, Lee VM: NTera 2 cells: A human cell line which displays characteristics expected of a human committed neuronal progenitor cell. J Neurosci Res 1993;35:585–602. 15 Greene LA, Tischler AS: Establishment of a noradrenergic clonal line of rat adrenal pheochromocytoma cells which respond to nerve growth factor. Proc Natl Acad Sci USA 1976;73:2424–2428. 16 Hammond DN, Wainer BH, Tonsgard JH, Heller A: Neuronal properties of clonal hybrid cell lines derived from central cholinergic neurons. Science 1986;234:1237–1240. 17 Tabira T, Konishi Y, Gallyas F Jr: Neurotrophic effect of hematopoietic cytokines on cholinergic and other neurons in vitro. Int J Dev Neurosci 1995;13:241–252. 18 Konishi Y, Chui DH, Hirose H, Kunishita T, Tabira T: Trophic effect of erythropoietin and other hematopoietic factors on central cholinergic neurons in vitro and in vivo. Brain Res 1993;609:29– 35. 19 Assandri R, Egger M, Gassmann M, Niggli E, Bauer C, Forster I, Gorlach A: Erythropoietin modulates intracellular calcium in a human neuroblastoma cell line. J Physiol 1999;516:343–352. 20 Juul SE, Yachnis AT, Rojiani AM, Christensen RD: Immunohistochemical localization of erythropoietin and its receptor in the developing human brain. Pediatr Dev Pathol 1999;2:148–158. 21 Brines ML, Ghezzi P, Keenan S, Agnello D, de Lanerolle NC, Cerami C, Itri LM, Cerami A: Erythropoietin crosses the blood-brain barrier to protect against experimental brain injury. Proc Natl Acad Sci USA 2000;97:10526–10531. 22 Chin K, Yu X, Beleslin-Cokic B, Liu C, Shen K, Mohrenweiser HW, Noguchi CT: Production and processing of erythropoietin receptor transcripts in brain. Brain Res Mol Brain Res 2000;81:29–42. 23 Liu ZY, Chin K, Noguchi CT: Tissue specific expression of human erythropoietin receptor in transgenic mice. Dev Biol 1994;166:159–169. 24 Wickrema A, Krantz SB, Winkelmann JC, Bondurant MC: Differentiation and erythropoietin receptor gene expression in human erythroid progenitor cells. Blood 1992;80:1940–1949.

Biol Neonate 2001;79:228–235

25 Wu H, Lee SH, Gao J, Liu X, Iruela-Arispe ML: Inactivation of erythropoietin leads to defects in cardiac morphogenesis. Development 1999;126: 3597–3605. 26 Sakanaka M, Wen TC, Matsuda S, Masuda S, Morishita E, Nagao M, Sasaki R: In vivo evidence that erythropoietin protects neurons from ischemic damage. Proc Natl Acad Sci USA 1998;95:4635– 4640. 27 Dame C, Bartmann P, Wolber EM, Fahnenstich H, Hofmann D, Fandrey J: Erythropoietin gene expression in different areas of the developing human central nervous system. Brain Res Dev Brain Res 2000;125:69–74. 28 Bernaudin M, Bellail A, Marti HH, Yvon A, Vivien D, Duchatelle I, Mackenzie ET, Petit E: Neurons and astrocytes express EPO mRNA: Oxygensensing mechanisms that involve the redox-state of the brain. Glia 2000;30:271–278. 29 Trimble M, Caro J, Talalla A, Brain M: Secondary erythrocytosis due to a cerebellar hemangioblastoma: Demonstration of erythropoietin mRNA in the tumor. Blood 1991;78:599–601. 30 Bruneval P, Sassy C, Mayeux P, Belair MF, Casadevall N, Roux FX, Varet B, Lacombe C: Erythropoietin synthesis by tumor cells in a case of meningioma associated with erythrocytosis. Blood 1993; 81:1593–1597. 31 Semenza GL: HIF-1: Mediator of physiological and pathophysiological responses to hypoxia. J Appl Physiol 2000;88:1474–1480. 32 Drutel G, Heron A, Kathmann M, Gros C, Mace S, Plotkine M, Schwartz JC, Arrang JM: ARNT2, a transcription factor for brain neuron survival? Eur J Neurosci 1999;11:1545–1553. 33 Drutel G, Kathmann M, Heron A, Gros C, Mace S, Schwartz JC, Arrang JM: Two splice variants of the hypoxia-inducible factor HIF-1alpha as potential dimerization partners of ARNT2 in neurons. Eur J Neurosci 2000;12:3701–3708. 34 Zhang W, Tsuchiya T, Yasukochi Y: Transitional change in interaction between HIF-1 and HNF-4 in response to hypoxia. J Hum Genet 1999;44: 293–299. 35 Chikuma M, Masuda S, Kobayashi T, Nagao M, Sasaki R: Tissue-specific regulation of erythropoietin production in the murine kidney, brain, and uterus. Am J Physiol Endocrinol Metab 2000;279: E1242–E1248.

Dame/Juul/Christensen

36 Masuda S, Chikuma M, Sasaki R: Insulin-like growth factors and insulin stimulate erythropoietin production in primary cultured astrocytes. Brain Res 1997;746:63–70. 37 Juul SE, Harcum J, Li Y, Christensen RD: Erythropoietin is present in the cerebrospinal fluid of neonates. J Pediatr 1997;130:428–430. 38 Marti HH, Gassmann M, Wenger RH, Kvietikova I, Morganti-Kossmann MC, Kossmann T, Trentz O, Bauer C: Detection of erythropoietin in human liquor: Intrinsic erythropoietin production in the brain. Kidney Int 1997;51:416–418. 39 Nakamura T, Ebihara I, Shimada N, Koide H: Elevated levels of erythropoietin in cerebrospinal fluid of depressed patients. Am J Med Sci 1998;315: 199–201. 40 Buemi M, Allegra A, Corica F, Floccari F, D’Avella D, Aloisi C, Calapai G, Iacopino G, Frisina N: Intravenous recombinant erythropoietin does not lead to an increase in cerebrospinal fluid erythropoietin concentration. Nephrol Dial Transplant 2000;15:422–423. 41 Juul SE, Stallings SA, Christensen RD: Erythropoietin in the cerebrospinal fluid of neonates who sustained CNS injury. Pediatr Res 1999;46:543–547. 42 Alafaci C, Salpietro F, Grasso G, Sfacteria A, Passalacqua M, Morabito A, Tripodo E, Calapai G, Buemi M, Tomasello F: Effect of recombinant human erythropoietin on cerebral ischemia following experimental subarachnoid hemorrhage. Eur J Pharmacol 2000;406:219–225.

Neurotrophic and Neuroprotective Effects of Erythropoietin

43 Sadamoto Y, Igase K, Sakanaka M, Sato K, Otsuka H, Sakaki S, Masuda S, Sasaki R: Erythropoietin prevents place navigation disability and cortical infarction in rats with permanent occlusion of the middle cerebral artery. Biochem Biophys Res Commun 1998;253:26–32. 44 Buemi M, Grasso G, Corica F, Calapai G, Salpietro FM, Casuscelli T, Sfacteria A, Aloisi C, Alafaci C, Sturiale A, Frisina N, Tomasello F: In vivo evidence that erythropoietin has a neuroprotective effect during subarachnoid hemorrhage. Eur J Pharmacol 2000;392:31–34. 45 Koshimura K, Murakami Y, Sohmiya M, Tanaka J, Kato Y: Effects of erythropoietin on neuronal activity. J Neurochem 1999;72:2565–2572. 46 Dame C, Zhao Y, Christensen RD, Juul SE: A comparison of the neuroprotective effects of Epo mimetic peptide 1 (EMP1) with erythropoietin (Epo) in NT2 cells. Ann Hematol 2000;79:B10. 47 Sinor AD, Greenberg DA: Erythropoietin protects cultured cortical neurons, but not astroglia, from hypoxia and AMPA toxicity. Neurosci Lett 2000; 290:213–215. 48 Juul S, Ferguson R, Christensen R: Erythropoietin induces differential gene expression in NT2 cells. Pediatr Res 1999;45:263A. 49 Renzi MJ, Farrell FX, Bittner A, Galindo JE, Jolliffe LK: Analysis of changes in gene expression in PC-12 cells treated with erythropoietin. Blood 2000;96(suppl 2):13b. 50 Gregory T, Yu C, Ma A, Orkin SH, Blobel GA, Weiss MJ: GATA-1 and erythropoietin cooperate to promote erythroid cell survival by regulating bcl-xL expression. Blood 1999;94:87–96.

51 Silva M, Benito A, Sanz C, Prosper F, Ekhterae D, Nunez G, Fernandez-Luna JL: Erythropoietin can induce the expression of bcl-x(L) through Stat5 in erythropoietin-dependent progenitor cell lines. J Biol Chem 1999;274:22165–22169. 52 Bittorf T, Seiler J, Ludtke B, Buchse T, Jaster R, Brock J: Activation of STAT5 during EPO-directed suppression of apoptosis. Cell Signal 2000; 12:23–30. 53 Bernaudin M, Marti HH, Roussel S, Divoux D, Nouvelot A, MacKenzie ET, Petit E: A potential role for erythropoietin in focal permanent cerebral ischemia in mice. J Cereb Blood Flow Metab 1999; 19:643–651. 54 Li B, Dai W: Thrombopoietin and neurotrophins share a common domain. Blood 1995;86:1643– 1644. 55 Campana WM, Misasi R, O’Brien JS: Identification of a neurotrophic sequence in erythropoietin. Int J Mol Med 1998;1:235–241. 56 Johnson DL, Farrell FX, Barbone FP, McMahon FJ, Tullai J, Hoey K, Livnah O, Wrighton NC, Middleton SA, Loughney DA, Stura EA, Dower WJ, Mulcahy LS, Wilson IA, Jolliffe LK: Identification of a 13 amino acid peptide mimetic of erythropoietin and description of amino acids critical for the mimetic activity of EMP1. Biochemistry 1998;37:3699–3710. 57 Wrighton NC, Farrell FX, Chang R, Kashyap AK, Barbone FP, Mulcahy LS, Johnson DL, Barrett RW, Jolliffe LK, Dower WJ: Small peptides as potent mimetics of the protein hormone erythropoietin. Science 1996;273:458–464.

Biol Neonate 2001;79:228–235

235

Biol Neonate 2001;79:236–240

Fetal and Neonatal Cerebral Infarcts Stéphane Marret a Caroline Lardennois a Anne Mercier a Sophie Radi a Christine Michel a Catherine Vanhulle a Aude Charollais a P. Gressens b a Department b Department

of Neonatal Medicine, Clinique de Pédiatrie et de Puériculture, Hôpital Charles Nicolle, Rouen, of Neuropediatrics, Hôpital Robert-Debré, Paris, France

Key Words Infant W Stroke W Porencephaly W Seizure W Thrombophilia W Cerebral palsy

Abstract Focal arterial infarction in the full-term newborn is an important cause of acquired cerebral lesions in the perinatal period. Clinical motor seizures, most often unifocal, are the nearly constant disclosing symptom confirmed by focal EEG abnormalities. A multifactorial physiopathology is usual, including genetic and perinatal environmental factors. In the past decade, various acquired or genetic thrombophilias have been discussed as risk factors. For several of the involved mechanisms, the excitotoxic cascade could represent a common final pathway leading to neuronal cell death. Early magnetic resonance imaging studies and EEG help to identify the newborns with strokes who are likely to develop hemiplegia and disabilities at school. Protection of the human fetal brain remains difficult, since the triggering factor initiating the excitotoxic cascade is rarely observed. Treatment of seizures is nevertheless necessary, because it seems that they accelerate anoxia-induced neuronal death in animal models of focal hypoxic ischemia. Copyright © 2001 S. Karger AG, Basel

Introduction

Recent advances in the follow-up of pregnancies and perinatal care have improved the survival rate of newborns. However, the prevalence of cerebral palsy and cognitive sequelae in infancy

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0236$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

remains high despite this progress. However, the panorama of different perinatal brain injuries associated with neurological sequelae has changed in the last few years. Single perinatal or neonatal arterial distribution infarctions are more often recognized and probably participate in the altered distribution of brain damage in infancy. This paper will focus on cerebral arterial and border-zone infarcts, most often diagnosed in full-term newborns. It will not cover the description of deep paraventricular hemorrhagic venous infarcts observed in preterm babies, nor neonatal venous thrombosis involving the lateral sinus.

Anatomical Presentation of Cerebral Infarcts in Neonates

Porencephalic cysts or strokes may be the expression of focal cerebrovascular necrosis in near-term fetuses or neonates [1]. Their pathology is unknown and only a few autopsy studies are available because the newborns affected rarely die. With the use of brain imaging techniques such as computerized tomography (CT) scans, several reports have documented their occurrence in the perinatal period. The etiology and pathogenesis of porencephaly and hydranencephaly are still controversial. Yakovlev and Wadsworth [2] applied the term porencephaly to a bad developmental process of cerebral cavitation due to localized agenesis of the cortex resulting in a cleft or a cavity communicating with lateral ventricles. However, it is more widely accepted that porencephaly and hydranencephaly could result from the destruction of brain tissue in specific arterial territories of the carotid circulation. Observation of a porencephalic cyst on a cranial CT scan done in the first few hours or days after birth suggests an antenatal event (fig. 1). However, the first cranial CT scan done in

Prof. Stéphane Marret Department of Neonatal Medicine, Clinique de Pédiatrie et de Puériculture Hôpital Charles Nicolle, 1, rue de Germont F–76031 Rouen Cedex (France) Tel. +33 2 3288 8097, Fax +33 2 3288 8633, E-Mail [email protected]

Fig. 1. Postnatal CT scan in a girl showing a porencephalic cyst. An antenatal diagnosis of ventricular dilatation was made during pregnancy.

the first few days after birth more often reveals a large wedge-shaped hypodense area, suggesting a recent perinatal ischemic event. Larger lesions may progress to cavity formation, but smaller lesions form a scar with localized atrophy. Infarcts are characterized as being in the territory of the main arteries or in a border-zone distribution in the watershed area between end fields of main artery territories. The left middle cerebral artery territory is most often involved and therefore suggests a thromboembolic origin. The main branch, cortical branches or lenticulostriate branches could be involved. The lesions are classified according to the possible combination of damage in hemispheric and subcortical structures such as the brain stem, thalamus, basal ganglia or internal capsule.

Physiopathological Factors Associated with Perinatal Cerebral Strokes

Impaired or absent blood flow within the distribution of single or multiple major cerebral vessels is the main factor responsible for cerebral infarcts. Alteration of cerebral blood flow could be due to pre- or postnatal focal vascular occlusion secondary to embolic, thrombotic and ischemic events or generalized systemic circulatory insufficiency. The multifactorial physiopathology of perinatal/neonatal strokes is widely emphasized [3–5]. Several environmental and genetic factors have been described in association with perinatal strokes. Nongenetic risk factors include viral or bacterial maternally transmitted septicemias, placental dysfunctions, intrauterine growth retardation, preeclampsia, twin pregnancy, feto-fetal transfusion in twin pregnancy, hydrops fetalis, exposure to alcohol or vasoactive

Neonatal Strokes

drugs such as cocaine, maternal diabetes, maternal traumatism, maternal hypotension, alloimmune thrombocytopenia, peripartal asphyxia and postnatal factors (patent foramen ovale, persistent pulmonary hypertension, polycythemia, renal venous thrombosis). Genetic diseases in which strokes have been described are extremely rare, and include fibromuscular dysplasia and congenital metabolic diseases [6] such as mitochondrial disorders, homocystinuria and organic acidurias, ornithyl-carbamyl-phosphate deficiency and carbamyl-phosphate synthetase deficiency and carbohydrate-deficient glycoprotein syndrome, among others. Recent studies have focused on the genetic and acquired risks of thromboembolism [7–9]. In a retrospective study on 24 neonates and children with porencephaly, Debus et al. [8] observed two genetic defects in 3 of these 24 patients and one in 13 others: heterozygous factor V Leiden mutation (n = 3); protein C deficiency type I (n = 6); increased lipoprotein Lp(a) (n = 3), and protein S type I deficiency (n = 1). In a large multicenter case-control prospective study [9], a significant association between acquired and genetic prothrombotic risk factors and symptomatic ischemic stroke in full-term neonates was found: 62 of 91 infants with cerebral infarcts (68%) had at least 1 prothrombotic risk factor compared with 44 control patients (24%) [odds ratio (OR), confidence interval (CI): 6.7 (3.84–11.67)]. Lipoprotein Lp(a) was the most important risk factor in the neonatal period, as in children up to 6 months of age [OR (CI): 4.84 (2.16–10.88)]. Protein C deficiency was observed in 6 out of 91 infants. Heterozygous factor V Leiden G1691A mutations were associated with a significant OR in neonatal strokes [OR (CI): 3.95 (1.72–9)]. The factor V Leiden mutation was found in combination with increased Lp(a) or increased acquired anticardiolipin antibodies. Additional triggering factors were reported in 54% of neonates. However, prothrombin G20210A variants and the methylenetetrahydrofolate reductase T677T genotype were not significantly increased. Neither antithrombin deficiency nor protein S deficiency was found. Therefore, this study underlines the multifactorial etiology of symptomatic strokes in full-term newborns.

Metabolic Basis of Focal Cerebral Infarcts

From in vitro studies and in vivo animal models of focal brain injury, many neurobiological mechanisms have been shown to be involved in neural cell damage and neuritic process breakings observed in brain lesions of neonates. Neural cell death could be due to an excess of cytokines, free radical overproduction, extracellular matrix damage, excitatory amino acid accumulation or growth factor deficiencies. For several of these mechanisms, the excitotoxic cascade could represent a common final pathway [10, 11]. After an initial insult such as hypoxic ischemia, hypoglycemia or infectious disease, an excessive amount of glutamate is released and results in a massive influx of ions, in particular calcium and zinc, through the N-methyl-D-aspartate (NMDA) glutamate receptor of the postsynaptic neuronal membrane. Abnormal amounts of intracellular calcium trigger metabolic pathways (nitric oxide synthesis, lipid peroxidation, free radical synthesis, platelet-activated factor and protease activation, DNA fragmentation) which lead to neural cell death. They also participate in oxidative stress.

Biol Neonate 2001;79:236–240

237

Brain Developmental Aspects and Sensitivity of the Gray Matter

The type of acute brain lesion and the final effect on the brain depend on the stage of development at which it was inflicted. Immature gray matter, i.e. cortex and basal ganglia, of the full-term newborn is more sensitive to brain injury than gray matter of the preterm newborn [1, 12, 13]. Observational and anatomical studies in humans have shown that focal cortical injury or diffuse asphyxia are more frequent in fullterm newborns than in preterm newborns. This relative sparing of the cerebral cortex and subcortical white matter could be explained by the presence of interarterial meningeal anastomoses, which do not regress until term. Several animal models in different species (rat, cat, dog, monkey) help us to understand developmental aspects of brain maturation and mechanisms of brain injuries [14]. The most popular model consists of unilateral carotid ligation at the age of 7 days in rat pups, followed by a period of recovery, and then a variable period of hypoxia in a sealed chamber with delivery of specific oxygen concentrations [15]. Severe neuronal damage was found in the cerebral cortex ipsilateral to the side of the carotid artery occlusion. Unilateral carotid ligation in 2-, 4-, 10-, 14-, 20- and 30day-old rats resulted in a maximal anatomical pattern of neuronal necrosis at 10 days of age, which approximately corresponds to the cortex of a human full-term newborn. We recently established an animal model of excitotoxic lesions in the developing mouse brain [12]. Brain damage was induced by intracortical injections of ibotenate, a glutamatergic agonist, at different steps of brain development and neuronal maturation. This study was the first to report developmental changes in the response of white and gray matter of the neonatal mouse brain to excitotoxicity. When administered at birth soon after the completion of neuronal migration to the supragranular layers, ibotenate induced neuronal death in cortical layers V–VI. The lesion mimics microgyria, a cortical lesion observed between 14 and 28 weeks of human gestation. When administered on postnatal day 5 (P5) or P10, after all neurons have completed migration in the neocortex, ibotenate produced a severe neuronal loss in all neocortical layers II, III, IV, V and VI, mimicking focal cortical necrosis observed in preterm newborns above 32 weeks of gestation and in term newborns. Furthermore, ibotenate induced the formation of white matter cysts between P2 and P10 with a 100% peak of intensity and frequency at P5; these lesions mimic periventricular leukomalacias. Although ibotenate is able to activate both NMDA and metabotropic glutamate receptors, all ibotenate-induced brain lesions were inhibited by treatment with a specific NMDA antagonist (D,L-2-amino-7-phosphoheptanoic acid), but not by an antagonist of phosphoinositide-linked metabotropic receptors [L(+)-2-amino-3-phosphonopropionic acid]. The ibotenate model, applied at consecutive developmental steps, therefore represents a well-defined tool. Therefore, several physiopathological mechanisms potentially involved in perinatal brain damage could be investigated and programs for neuroprotection could be envisaged. We can conclude from observational studies in human newborns and from animal models of focal injury that the cortical gray matter of the full-term newborn is characterized by a regression of meningeal anastomosis, a functional maturation of neurons and of the receptors of neurotransmitters, a predominance of excitatory synapses over the inhibitory synapses and a high oxygen consumption.

238

Biol Neonate 2001;79:236–240

These data strengthen the hypothesis of the high vulnerability of gray matter in the full-term newborn brain compared to the preterm newborn. Moreover, this step of brain development is characterized by neuritogenesis, synaptogenesis, neural differentiation and processes of stabilization such as neural apoptosis, which play a significant role in the establishment of the neural network; these processes could also be altered due to gray matter damage.

Clinical Presentation of Perinatal Strokes

During the prenatal period, porencephaly or a hyperechogenic focus can be observed in routine ultrasonographic studies done in the third trimester of pregnancy. However, in the case of antenatal physiopathological factors described above, ultrasonographic studies may be oriented towards the brain [16]. After birth, perinatal stroke is relatively common in the full-term neonate but difficult to recognize by clinical examination alone [4, 5, 17, 18]. Uneventful pregnancies are often observed. The first hours after birth can be unremarkable. In a few cases, infants require respiratory support for early signs of perinatal asphyxia with hypotonia, hyperexcitability or reduced alertness. The majority of full-term newborns presenting with a stroke suffer repetitive unifocal motor seizures in the first 24–48 h. After seizures, neurological examination may be normal, or show an asymmetric tone. In asphyxiated infants, generalized hypotonia can be observed. Cerebral infarct is revealed on the CT scan or the magnetic resonance imaging done in the screening of seizures. A clear EEG shows evidence of focal or lateralized functional abnormalities. The lateralized EEG abnormalities are not specific for cerebral infarct, but their presence emphasizes the value of neonatal EEG in signaling the presence of focal parenchymal injury [19]. Background activity is usually normal. However, in the case of severe asphyxia or some rare metabolic congenital diseases, abnormal background activity can be observed. In a series of 154 newborns with neonatal seizures admitted to the Department of Neonatology of the Centre Hospitalier Universitaire of Rouen between 1989 and 2000, 22 cases (14.5%) were attributed to a perinatal stroke. In contrast to the full-term newborn, arterial infarcts in the preterm baby are less commonly reported [20, 21]. Most of them are discovered on routine cranial ultrasonographic scans performed during their admission to the intensive care unit. Clinical seizures are only observed in anecdotal reports. Most of the seizures are EEG seizures and clinically silent. Neurological examination is always normal. Therefore, their frequency can be underestimated.

Early Prognostic Indicators and Neurologic Outcome

None of the adverse antenatal or perinatal factors was significantly associated with abnormal outcome [22–24]. Initial neonatal clinical examination was not predictive of the outcome. Neurological examination at discharge, EEG and magnetic resonance imaging were better predictors of the outcome. Abnormal backgrounds on EEG and involvement of the internal capsule and/or basal ganglia combined with hemisphere injury on magnetic resonance imaging were significantly associated with hemiplegia (fig. 2). The incidence of hemiplegia reported in the literature varies from 8 to 100% [17]. Global neurodevelopmental delay is uncommon, but speech delay seems more frequent in children with a left hemisphere

Marret/Lardennois/Mercier/Radi/Michel/ Vanhulle/Charollais/Gressens

Fig. 2. T1-weighted magnetic resonance

imaging in a girl with no genetic factors for familial thrombopenia, showing a stroke.

infarct. Behavioral difficulties and epilepsy at school age can be seen in children who display motor difficulties. In a recent study of 46 children who had had a neonatal stroke [24], the neurodevelopmental outcome was normal in 33% of the 46 cases. Cerebral palsy was observed in 48% of the 46 infants with a cerebral infarct, cognitive impairment in 41%, visual impairment in 15% and seizure disorder in 46%.

Prevention of Strokes and Neuroprotection

An ideal strategy for stroke prevention should take place prenatally. However, this is difficult, since the triggering ischemic event is rarely observed. When a possible etiologic factor has been identified, serial ultrasonographic studies must be done to detect the occurrence of a stroke. There is evidence from newborn rats and preterm mon-

keys that the brain may reorganize to achieve complete functional compensation if cortical damage occurs early [25]. Additional postnatal therapies could optimize functional compensation at a time of rapid brain growth and development. Several drugs such as inhibitors of free radicals, inhibitors of the NMDA receptor, anticytokine molecules or inhibitors of apoptosis have been tested in animal models of stroke with great benefits. In an excitotoxic model of cortical necrosis obtained at P5 and P10 in mouse pups, we observed significant neuroprotection with kynurenic acid (an antagonist of the NMDA receptor), an inhibitor of nitric oxide synthase and vasointestinal peptide, a growth factor [26, 27]. Magnesium sulfate had a significant protective effect at P5 but not at P10, a stage of brain development which corresponds to a full-term human brain [28, unpubl. data]. Adverse effects, biodisponibilities and penetration of the blood-brain barrier by new drugs are of paramount importance in the design of new trials in humans. Therefore, only a few strategies can be addressed in human preterm newborns and their mothers before delivery. Fluorinated glucocorticoids or aspirin could be good candidate neuroprotective molecules to limit the extent of inflammatory and excitotoxic processes. Seizure treatment seems useful. It is unclear if adverse neurodevelopmental outcomes can occur as a consequence of seizures and can be prevented by treatment. Seizures and, above all, repetitive seizures and status epilepticus are associated with a fall in the concentration of ATP and an excessive release of excitatory amino acids; they could therefore interfere with brain development, particularly in the limbic system [1]. Immediate prognosis is particularly related to the background EEG and to the underlying neurological disease. However, prolonged or poorly controlled neonatal seizures have been associated with worse outcome than infrequent or readily controlled seizures [29]. Cognitive abnormalities observed in some children with a perinatal stroke in a private series imply that brain lesions spread beyond the cortical infarct zone [5]. Moreover, the long-term cognitive development (spelling, arithmetic, memory impairments) of children who had benign neonatal convulsions could be affected even if they had a normal motor development and an intellectual coefficient up to 80 [30]. In these cases, we could hypothesize that an underlying brain lesion has contributed to seizure occurrence or that the seizure itself could have altered brain development. However, environmental enrichment could also influence cognitive development. Nevertheless, experimental and clinical data justify anticonvulsant drug initiation with either phenobarbitone or phenytoin to control seizures. Potential adverse effects of anticonvulsant drugs require their termination as soon as possible.

References 1 Volpe JJ: Neurology of the Newborn, ed 3. Philadelphia, Saunders, 1995. 2 Yakovlev PI, Wadsworth RD: Schizencephalies. Study of the congenital clefts in the cerebral mantle. I: Clefts with fused lips. II: Cleft with hydrocephalus and lips separated. J Neuropathol Exp Neurol 1946;5:116–130. 3 Koeflen W, Freund M, Varnholt V: Neonatal stroke involving the middle cerebral artery in term infants: Clinical presentation, EEG and imaging studies, and outcome. Dev Med Child Neurol 1995;37:294–312.

Neonatal Strokes

4 Clancy R, Malin S, Laraque D, Baumgart S, Younkin D: Focal motor seizures heralding stroke in full-term neonates. Am J Dis Child 1985;139:601– 606. 5 Vanhulle C, Marret S, Parain D, Samson-Dollfus D, Fessard C: Convulsions néonatales focalisées et infarctus artériel cérébral. Arch Pediatr 1998;5: 404–408. 6 Sperl W, Felber S, Skladal D, Wermuth B: Metabolic stroke in carbamyl phosphate synthetase deficiency. Neuropediatrics 1997;28:229–234.

7 Thorarensen O, Ryan S, Hunter J, Younkin D: Factor V Leiden mutation: An unrecognized cause of hemiplegic cerebral palsy, neonatal stroke, and placental thrombosis. Ann Neurol 1997;3:372– 375. 8 Debus O, Koch HG, Kurlemann G, Sträter R, Vielhaber H, Weber P, Nowak-Göttl U: Factor V Leiden and genetic defects of thrombophilia in childhood porencephaly. Arch Dis Child Fetal Neonatal Ed 1998;78:F121–F124.

Biol Neonate 2001;79:236–240

239

9 Günther G, Junker R, Sträter R, Schobess R, Kurnik K, Kosch A, Nowak-Göttl U: Symptomatic ischemic stroke in full-term neonates: Rôle of acquired and genetic prothrombotic risk factors. Stroke 2000;31:2437–2441. 10 McDonald JW, Johnston MV: Physiological and pathophysiological roles of excitatory amino acids during central nervous system development. Brain Res Brain Res Rev 1990;15:41–70. 11 Lipton SA, Rosenberg PA: Excitatory amino acids as a final common pathway for neurologic disorders. N Engl J Med 1994;330:613–622. 12 Marret S, Mukendi R, Gadisseux JF, Gressens P, Evrard Ph: Effect of ibotenate on brain development: An excitotoxic mouse model of microgyria and posthypoxic-like lesions. J Neuropathol Exp Neurol 1995;54:358–370. 13 Reddy K, Mallard C, Guan J, Marks K, Bennet L, Gunning M, Gunn A, Gluckman P, Williams C: Maturational change in the cortical response to hypoperfusion injury in the fetal sheep. Pediatr Res 1998;43:674–682. 14 Tuor UI, Del Bigio MR, Chumas PD: Brain damage due to cerebral hypoxia/ischemia in the neonate: Pathology and pharmacological modification. Cerebrovasc Brain Metab Rev 1996;8:159– 193. 15 Rice JEI, Vannucci RC, Brierley JB: The influence of immaturity on hypoxic-ischemic brain damage in the rat. Ann Neurol 1981;9:131–141.

240

16 Amato M, Hüppi P, Hersckowitz N, Huber P: Prenatal stroke suggested by intrauterine ultrasound and confirmed by magnetic resonance imaging. Neuropediatrics 1991;22:100–102. 17 Govaert P, Matthys E, Zecic A, Roelens F, Oostra A, Vanzieleghem B: Perinatal cortical infarction within middle cerebral artery trunks. Arch Dis Child Fetal Neonatal Ed 2000;82:F59–F63. 18 Mercuri E, Cowan F: Cerebral infarction in the newborn infant: Review of the literature and personal experience. Europ J Paediatr Neurol 1999;3: 255–263. 19 Samson-Dollfus D (ed): Electroencéphalographie de l’enfant. Paris, Masson, 1998. 20 Pierrat V, Cneude F, Duquennoy C, Lequien P: Infarctus artériel cérébral: Diagnostic échographique et particularités sémiologiques chez le nouveau-né prématuré. Arch Pediatr 1996;3:137–140. 21 De Vries LS, Groenendaal F, Eken P, Van Hasstert IV, Rademaker KJ, Meiners LC: Infarcts in the vascular distribution of the middle cerebral artery in preterm and fullterm infants. Neuropediatrics 1997;28:88–96. 22 Trauner DA, Mannino FL: Neurodevelopmental outcome after neonatal cerebrovascular accident. J Pediatr 1986;108:459–461. 23 Mercuri E, Rutherfort M, Cowan F, Pennock J, Counsell S, Papadimitriou M, Azzopardi D, Bydder G, Dubowitz L: Early prognostic indicators of outcome in infants with neonatal cerebral infarction: A clinical, electroencephalogram, and magnetic resonance imaging study. Pediatrics 1999; 103:39–45.

Biol Neonate 2001;79:236–240

24 Sreenan C, Bhargava R, Robertson C: Cerebral infarction in the term newborn: Clinical presentation and long-term outcome. J Pediatr 2000;137: 351–355. 25 Kolb B, Whishaw IQ: Fundamentals of Human Neuropsychology, ed 3. New York, Freeman, 1990, pp 671–711. 26 Marret S, Bonnier C, Raymackers JM, Delpech A, Evrard Ph, Gressens P: Glycine antagonist and NO synthase inhibitor protect the developing mouse brain against neonatal excitotoxic lesions. Pediatr Res 1999;45:337–342. 27 Gressens P, Marret S, Hill MJ, Brenneman E, Gozes I, Fridkin M, Evrard P: Vasoactive intestinal peptide prevents excitotoxic cell death in the murine developing brain. J Clin Invest 1997;100: 390–397. 28 Marret S, Gressens P, Gadisseux JF, Evrard P: Prevention by magnesium of excitotoxic neuronal death in the developing brain: An animal model for clinical intervention studies. Dev Med Child Neurol 1995;37:473–484. 29 Evans D, Levene M: Neonatal seizures. Arch Dis Child Fetal Neonatal Ed 1998;78:F70–F75. 30 Temple CM, Dennis J, Carney R, Sharich J: Neonatal seizures: Long-term outcome and cognitive development among ‘normal’ survivors. Dev Med Child Neurol 1995;37:109–118.

Marret/Lardennois/Mercier/Radi/Michel/ Vanhulle/Charollais/Gressens

Biol Neonate 2001;79:241–245

Blood Pressure and Tissue Oxygenation in the Newborn Baby at Risk of Brain Damage A. Michael Weindling Christopher M. Kissack University of Liverpool, Liverpool Women’s Hospital, Liverpool, UK

Key Words Blood pressure W Tissue oxygenation W Cerebral oxygenation W Newborn baby W Neonate W Neonatal intensive care W Brain damage W Neurodisability W Disability

Abstract A cardinal aim of neonatal intensive care is the maintenance of an adequate oxygen supply to the tissues, particularly the brain. This process depends on several factors. These include an adequate blood oxygen content, blood flow to the tissues and the ability of cells to extract and utilise oxygen. Oxygen carriage depends on ventilation and haemoglobin concentration and type. Blood flow depends on cardiac output (in turn dependent on cardiac contractility, heart rate, blood pressure and vascular resistance). Different tissues also have different oxygen demands depending on their oxygen consumption, which are likely to vary within the tissue itself and with the activity of the infant. This paper discusses evidence that suggests that even in preterm neonates, cerebral blood flow may be independent of blood pressure, and that even very low cerebral blood flow seems to be consistent with healthy survival. Evidence is considered that cardiac output rather than blood pressure may be more important in determining brain tissue oxygenation. We have found a negative correlation between cardiac output and cerebral oxygen extraction in preterm infants, but no relationship between mean arterial blood pressure and cerebral oxygen extraction. Copyright © 2001 S. Karger AG, Basel

Introduction

The link between blood pressure and brain damage in young babies has been considered in two ways: an association with presumed loss of the ability to autoregulate, and an association with low blood pressure. The first concept led to the exploration of ways to

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0241$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

predict autoregulation and to use the presence or absence of autoregulation as a way of assessing prognosis; there have been no studies aimed at restoring lost autoregulation. The second led to a series of studies to discover the most effective way to maintain blood pressure. Recently, doubts have been cast on the importance of maintaining blood pressure, since normal cerebral blood flow has been observed in the face of low blood pressure measurements [1]. A possible reason for this is that when there is systemic hypotension, autoregulation acts to maintain cerebral blood flow, but, when this mechanism is at its limit, increased oxygen extraction at a cellular level may compensate. This paper will set out the evidence for autoregulation in the newborn infant, review the work that has been done concerning the maintenance of blood pressure, explore the importance of cardiac output and finally consider a mechanism for avoiding brain injury when there is systemic hypotension and loss of autoregulation.

Autoregulation

In 1959, Lassen [2] proposed the concept of cerebral autoregulation – the mechanism whereby cerebral blood flow is maintained in spite of variations in cerebral perfusion pressure and therefore systemic blood pressure. It was based on observations in the 1930s by Fog [3, 4], who studied pial vessel diameter in response to various stimuli. The range of mean blood pressure over which autoregulation operates in the adult is thought to be between about 60 and 150 mm Hg, with a response time of between 3 and 15 s [5–7]. Prompted by the observations of Reivich et al. [8], who used an autoradiographic technique to show that cerebral blood flow was reduced in asphyxiated monkeys, Lou et al. [9], in 1977, studied the cerebral haemodynamics of asphyxiated sheep in utero and found that cerebral blood flow was increased but that autoregulation had been lost.

Dr. A. Michael Weindling, Professor of Perinatal Medicine University of Liverpool, Neonatal Unit, Liverpool Women’s Hospital Crown Street, Liverpool L8 7SS (UK) Tel. +44 151 702 4119/4055, Fax +44 151 702 4024/4082 E-Mail [email protected]

neonates, increases in blood pressure had no effect on cerebral blood flow. Most recently, Anthony [15] defined two mechanisms for autoregulation, each operating in a different time frame: there is beatto-beat variability, occurring at about 1–2 Hz (described by Perlman et al. [16, 17]), and there are slow wave cyclical changes at 3–6 cycles/min [18, 19]. The latter are considered to be constant finetuning processes, and further evidence for this mechanism was provided by Boylan et al. [20].

Hypotension and Brain Damage

Fig. 1. The relationship between cerebral blood flow and blood pressure in sick infants (data of Lou et al. [10]). The data have been replotted to consider only ‘sick’ infants, i.e. excluding those with Apgar scores at 5 min above 7 and pH above 7. B1 and B2 represent successive measurements in the same infant.

Two years later, Lou et al. [10] investigated the effect of blood pressure changes on cerebral blood flow in 19 sick term babies (mean birth weight 3.20 kg) in whom autoregulation might have been expected to be impaired. Because of the unique nature of these data, they are worth considering in some detail, and the data have been re-analysed. Cerebral blood flow was measured using 133Xe injected into the carotid artery from an umbilical arterial catheter passed into the arch of the aorta. Five babies in the study had Apgar scores of 7 or less at 5 min and the majority of the others were also acidaemic, only 7 having pH values above 7.25 at the time that cerebral blood flow was measured. The median cerebral blood flow was 40.0 ml/100 g/ min (range 30–55) in 5 babies who were normotensive (systolic blood pressure 1 59 mm Hg) and who had normal Apgar scores. (Two other groups independently observed similar cerebral blood flow rates in term infants using a completely different methodology, i.e. strain gauge jugular venous occlusion plethysmography [11, 12].) Cerebral blood flow in neonates is therefore about two thirds that reported for normal adults using the 133Xe technique, i.e. 64.9 B 9 ml/100 g/min [13]. In the study of Lou et al. [10], 9 hypotensive babies (systolic blood pressure ! 46 mm Hg) had even lower cerebral blood flow rates (median 22.0 ml/100 g/min, range 12–32). Of these 9 infants, 7 were also acidaemic (pH ! 7.25) and 8 had a 5-min Apgar score below 7. The difference between the measurement of cerebral blood flow in the normal infants and those who were considered to be hypotensive is significant. Replotting the data of Lou et al. [10] and excluding babies with a 5-min Apgar score above 7 and pH 1 7.25 (fig. 1), the relationship between blood pressure and cerebral blood flow remained, i.e. cerebral blood flow varied with blood pressure and autoregulation seemed to have been abolished. Data from Seri et al. [14] also suggest that preterm infants do not autoregulate, but that this is not the case with healthy infants born at term. They found that in five 1-day-old preterm infants, changes in cerebral blood flow transiently paralleled dopamine-induced alterations in systemic blood pressure, indicating impaired autoregulation of cerebral blood flow in these infants. By contrast, in eight term

242

Biol Neonate 2001;79:241–245

The cerebral white matter of the preterm human infant is particularly vulnerable to hypoperfusion due to hypotension. The evidence for this comes from animal studies of the newborn puppy [21] and from studies of the changing anatomy of the human cerebrovasculature [22]. Generally, the suggestion is that there is an association between hypotension and peri- and intraventricular haemorrhage, although Gronlund et al. [23] found that a raised blood pressure was significantly associated with peri-intraventricular haemorrhage in preterm newborn infants. Low et al. [24] found that preterm newborn babies with either hypotension or hypoxaemia were about 2.5 times more likely to develop intraventricular haemorrhage, ventriculomegaly or hyperechoic brain parenchymal lesions than those who were not hypotensive or hypoxaemic. The probability of a major abnormal neurodevelopmental outcome increased from 8% in newborns with no hypotension or hypoxaemia, to 53% in children with both hypotension and hypoxaemia [25]. There has been considerable work done in preterm infants on maintaining blood pressure. Gill and Weindling [26] found that around 20% of very-low-birth weight infants admitted to a neonatal intensive care unit became hypotensive within 24 h of their admission [26]. A randomised controlled trial showed that dopamine was more effective at raising blood pressure than plasma protein fraction. The median dose of dopamine needed to increase the blood pressure to at least the 10th centile was 7.5 Ìg/kg/min and there was no benefit above 10 Ìg/kg/min. Two other separate randomised controlled trials have shown that dopamine is more effective than dobutamine at raising blood pressure [27, 28]. It also seems that a low-dose dopamine infusion improves urine output in very immature infants [29]. This is probably achieved through improved renal perfusion. Cardiac output was not measured in this study, but is likely to be important. There is evidence that cardiac output may be a more relevant determinant of cerebral perfusion and oxygenation than blood pressure, and that hypotension may be compensated by increased cardiac output. Lopez et al. [30] found that a combination of dopamine and dobutamine was particularly effective in raising cardiac output in preterm infants with respiratory distress syndrome. Both Lopez et al. [30] and Pladys et al. [31] found that hypotensive infants had increased vascular resistance, and that low blood pressure was associated with a high cardiac output.

Hypotension, Cardiac Output and Oxygen Extraction

Meek et al. [32] found that a low cerebral blood flow on the first day of life, but not hypotension, was associated with the subsequent development of severe intraventricular haemorrhage. Cerebral blood

Weindling/Kissack

flow in the infants with germinal matrix haemorrhage-intraventricular haemorrhage was 7.0 ml/100 g/min, compared with 12.2 ml/ 100 g/min in those without haemorrhage. Fractional oxygen extraction (FOE) represents the balance between oxygen delivery to a tissue and its consumption, i.e. the proportion of oxygen delivered that is extracted by the tissue (usually about 30%). It is calculated by the difference between arterial oxygen content (estimated from arterial oxygen saturation) and venous oxygen content (from venous oxygen saturation). Cerebral FOE (CFOE) can be calculated from the arterial oxygen saturation, measured by pulse oximetry, and by the cerebral venous oxygen saturation, measured using near infrared spectroscopy with partial jugular venous occlusion. A Liverpool study [33] aimed to determine CFOE in preterm infants during hypotension, during moderate anaemia and with changes in the PaCO2. Mean B SD CFOE was similar in control (0.292 B 0.06), anaemic (0.310 B 0.08) and hypotensive (0.278 B 0.06) neonates. After anaemic neonates were transfused, CFOE decreased to 0.274 B 0.05 (p = 0.02). In the hypotensive neonates, there was no relationship between CFOE and blood pressure. There were, however, changes in CFOE when physiological changes occurred over a relatively short period: CFOE decreased after blood transfusion and increased with decreasing PaCO2. As no change in CFOE was seen during hypotension, it was speculated that cerebral oxygen delivery might have been maintained by cerebral blood flow autoregulation. Cardiac output was not measured in this study, but is being measured in a study currently being undertaken to examine the relationships between cardiac output, blood pressure, cerebral oxygenation and cerebral injury in sick, preterm low-birth weight infants.

Near Infrared Spectroscopy

Near infrared spectroscopy is a validated and now widely applied research tool. Photons in the near infrared region of the electromagnetic spectrum have good penetration of dense mammalian tissues, including skin and bone. The principle of transmitting near infrared photons across intact tissues to assess oxygenation status was first described in 1977 [34]. Its application in the neonatal intensive care unit began some time later, initially using the oxidation status of cytochrome aa3 as an indicator of tissue oxygenation [35]. The measurement of the relative quantities of the chromophores haemoglobin (Hb) and oxyhaemoglobin (HbO) in the circulation supplying the tissues in question came later [36]. It is the measurement of these chromophores that may be used in the calculation of cerebral venous saturation, oxygen extraction, blood flow, oxygen delivery and oxygen consumption. Hb and HbO have different levels of absorption of near infrared photons according to the wavelength used. Knowing the amount of photons transmitted into and emerging from the tissue permits calculation of the absorbance of photons by the relevant chromophore. The concentration of the chromophore in the tissue may then be calculated from a modification of the Beer-Lambert equation:

loss per centimetre of travel through a one molar solution of the chromophore in question at the wavelength employed, DPF is the differential path length factor, to correct for the differing paths the photons take to reach the receiving optode as they are scattered by the tissue under study, and K is a constant related to the geometry of the tissue under study. Because K varies with each study and cannot be determined, the absolute concentration of a chromophore cannot be calculated. Fortunately, as it remains constant within each study, changes in chromophore concentrations with respect to an arbitrary baseline can be determined. Hence: ¢C = ¢absorbance/(D ! E ! DPF). Changes in the concentrations of cerebral levels of chromophores may be used to calculate values for cerebral venous saturation, but first that change in concentration must be induced. Early work using near infrared spectroscopy achieved this by tilting the infant under study head down [36, 37], with only limited success in one study [37]. Later studies have demonstrated that partial jugular venous occlusion is more reliable [38]. The transmitting and receiving optodes are secured in a frontotemporal arrangement on the infant’s head, and light is excluded using an occlusive blanket. The near infrared spectroscope locks in a set level of laser activity to produce a sufficient quantity of photons arriving at the receiving optode. This results in an arbitrary baseline, against which relative changes can be measured. Gentle unilateral pressure on the external jugular vein causes venous pooling within the cerebral circulation, and consequently a measurable rise in the quantity of cerebral venous HbO and Hb. This change in HbO and Hb may be used to calculate a cerebral venous saturation. Typical results are displayed in figure 2. After 5 s to obtain a suitably steady baseline, a partial jugular venous occlusion was performed. For each of the 10 subsequent half-second intervals, the change in HbO (¢HbO) and Hb (¢Hb) is calculated. This technique has been shown to be preferable to using ¢HbO and ¢Hb at the peak of the curve, or to calculating the area under the curve [38]. The subsequent averages of the 10 values for ¢HbO and ¢Hb are used to calculate cerebral venous saturation, using the following equation: SVO2 = ¢HbO/(¢HbO + ¢Hb)

where C is the unknown concentration of the chromophore, A is the absorbance of photons at a particular wavelength, D is the distance between transmitting and receiving optodes, E is the molar extinction coefficient for the chromophore in question, defined as the light

where SVO2 is the cerebral venous saturation, ¢HbO is the average change in HbO over the 5 s following partial jugular venous occlusion and ¢Hb is the average change in Hb over the 5 s following partial jugular venous occlusion. This method of measuring cerebral venous saturation using partial jugular venous occlusion has been validated by comparison with invasive co-oximetry of jugular bulb blood obtained during cardiac catheterisation in 15 children [38]. A Bland-Altman analysis showed the mean difference between the two techniques was 1.5%, with limits of agreement between –12.8 and +15.9%. Cerebral blood flow may be measured using the Fick principle, the delivery of an intravascular tracer. By briefly increasing the level of inspired oxygen, a bolus of HbO is delivered to the infant, and this is detected in the cerebral circulation using near infrared spectroscopy. This technique for measuring cerebral blood flow has been validated by comparison with the 133Xe clearance technique [39], demonstrating good agreement between the two methods. Our aim has been to describe the relationships that exist between parameters of cerebral oxygenation, cardiac function and blood pres-

Blood Pressure and Tissue Oxygenation

Biol Neonate 2001;79:241–245

C = A/(D ! E ! DPF ! K)

243

Fig. 2. Typical data from near infrared measurement. A partial jugular venous occlusion results in a change in the amount of HbO and Hb which may be used to calculate a venous HbO saturation.

sure on the first day after birth in sick, preterm and very-low-birth weight infants. We studied ventilated newborn infants with gestation less than 32 weeks and weight less than 1,500 g, and who had an umbilical arterial catheter in situ for the measurement of mean arterial blood pressure. Each infant had a measurement of cerebral venous saturation, followed by a measurement of cardiac output by two-dimensional and Doppler ultrasound, during the first 12 h after delivery. Arterial saturation was measured using pulse oximetry. CFOE was calculated using the formula: CFOE = SAO2 – SVO2/SAO2 where SAO2 is arterial oxygen Hb saturation and SVO2 is venous oxygen Hb saturation. We found a negative correlation between cardiac output and cerebral oxygen extraction in all 26 infants recruited, but no relationship between mean arterial blood pressure and cerebral oxygen extraction. This relationship between cardiac output and cerebral oxygenation would suggest that cerebral blood flow is directly linked to cardiac output, but not necessarily blood pressure. This fits with the observations of Tyszczuk et al. [1], who found that cerebral blood flow was maintained in spite of low blood pressure.

Seven infants with very high cerebral oxygen extraction above the 95th centile (0.4) [33] had significantly lower cardiac output than those with normal oxygen extraction. There was a particularly strong negative correlation between oxygen extraction and cardiac output in this subgroup, whereas those infants with normal CFOE showed no relationship between CFOE and cardiac output. Despite the association between cardiac output and cerebral oxygenation in infants with high CFOE, there were no demonstrable relationships between blood pressure and cardiac output, or blood pressure and oxygen extraction. Consequently, the measurement of blood pressure in these infants gave no indication that they were in a low cardiac output state and that high oxygen extraction was necessary to maintain tissue oxygenation and cellular metabolism.

Conclusion

The main purpose of neonatal intensive care that is aimed at preserving cerebral function should not be simply to preserve cerebral perfusion, but to preserve cerebral oxygenation, i.e. the ability to utilise oxygen at the cellular level. Evidence is now emerging to suggest that cardiac output and tissue oxygen extraction may be more important than blood pressure in maintaining cerebral metabolism.

References 1 Tyszczuk L, Meek J, Elwell C, Wyatt JS: Cerebral blood flow is independent of mean arterial blood pressure in preterm infants undergoing intensive care. Pediatrics 1998;102:337–341. 2 Lassen NA: Cerebral blood flow and oxygen consumption in man. Physiol Rev Am J Physiol 1959; 39:183–233. 3 Fog M: Cerebral circulation. The reaction of the pial arteries to a fall in blood pressure. Arch Neurol Psychiatry 1937;37:351–364.

244

4 Fog M: Cerebral circulation II. Reaction of pial arteries to increase in blood pressure. Arch Neurol Psychiatry 1939;41:260–268. 5 Paulson OB, Strandgaard S, Edvinsson L: Cerebral autoregulation. Cerebrovasc Brain Metab Rev 1990;2:161–192. 6 Florence G, Seylaz J: Rapid autoregulation of cerebral blood flow: A laser-Doppler flowmetry study. J Cereb Blood Flow Metab 1992;12:674–680. 7 Aaslid R, Lindegaard KF, Sorteberg W, Nornes H: Cerebral autoregulation dynamics in humans. Stroke 1989;20:45–52.

Biol Neonate 2001;79:241–245

8 Reivich M, Kovach AG, Spitzer JJ, Sandor P: Cerebral blood flow and metabolism in hemorrhagic shock in baboons. Adv Exp Med Biol 1972; 33:19–26. 9 Lou HC, Lassen NA, Friis-Hansen B: Low cerebral blood flow in hypotensive perinatal distress. Acta Neurol Scand 1977;56:343–352. 10 Lou HC, Friis-Hansen B: Arterial blood pressure elevations during motor activity and epileptic seizures in the newborn. Acta Paediatr Scand 1979; 68:803–806.

Weindling/Kissack

11 Cross KW, Dear PR, Hathorn MK, Hyams A, Kerslake DM, Milligan DW, Rahilly PM, Stothers JK: An estimation of intracranial blood flow in the new-born infant. J Physiol 1979;289:329–345. 12 Cooke RW, Rolfe P, Howat P: A technique for the non-invasive estimation of cerebral blood flow in the newborn infant. J Med Eng Technol 1977;1: 263–266. 13 Lassen NA, Menon O: The cerebral blood flow in man determined by the use of radioactive krypton. Acta Physiol Scand 1955;33:30–49. 14 Seri I, Rudas G, Bors Z, Kanyicska B, Tulassay T: Effects of low-dose dopamine infusion on cardiovascular and renal functions, cerebral blood flow, and plasma catecholamine levels in sick preterm neonates. Pediatr Res 1993;34:742–749. 15 Anthony MY: Cerebral autoregulation in sick infants. Pediatr Res 2000;48:3–5. 16 Perlman JM, McMenamin JB, Volpe JJ: Fluctuating cerebral blood-flow velocity in respiratory-distress syndrome. Relation to the development of intraventricular hemorrhage. N Engl J Med 1983; 309:204–209. 17 Perlman JM, Goodman S, Kreusser KL, Volpe JJ: Reduction in intraventricular hemorrhage by elimination of fluctuating cerebral blood-flow velocity in preterm infants with respiratory distress syndrome. N Engl J Med 1985;312:1353–1357. 18 Anthony MY, Evans DH, Levene MI: Cyclical variations in cerebral blood flow velocity. Arch Dis Child 1991;66:12–16. 19 Coughtrey H, Rennie JM, Evans DH: Postnatal evolution of slow variability in cerebral blood flow velocity. Arch Dis Child 1992;67:412–415. 20 Boylan GB, Young K, Panerai RB, Rennie JM, Evans DH: Dynamic cerebral autoregulation in sick newborn infants. Pediatr Res 2000;48:12–17. 21 Rivkin MJ, Volpe JJ: Strokes in children. Pediatr Rev 1996;17:265–278.

Blood Pressure and Tissue Oxygenation

22 Pape KE, Wigglesworth JS: Haemorrhage, Ischaemia and the Perinatal Brain. Clinics in Developmental Medicine 69/70. London, Spastics International Medical Publications, William Heinemann Medical Books, 1979. 23 Gronlund JU, Korvenranta H, Kero P, Jalonen J, Valimaki IA: Elevated arterial blood pressure is associated with peri-intraventricular haemorrhage. Eur J Pediatr 1994;153:836–841. 24 Low JA, Froese AB, Smith JT, Galbraith RS, Sauerbrei EE, Karchmar EJ: Hypotension and hypoxemia in the preterm newborn during the four days following delivery identify infants at risk of echosonographically demonstrable cerebral lesions. Clin Invest Med 1992;15:60–65. 25 Low JA, Froese AB, Galbraith RS, Smith JT, Sauerbrei EE, Derrick EJ: The association between preterm newborn hypotension and hypoxemia and outcome during the first year. Acta Paediatr 1993; 82:433–437. 26 Gill AB, Weindling AM: Randomised controlled trial of plasma protein fraction versus dopamine in hypotensive very low birthweight infants. Arch Dis Child 1993;69:284–287. 27 Greenough A, Emery EF: Randomized trial comparing dopamine and dobutamine in preterm infants. Eur J Pediatr 1993;152:925–927. 28 Klarr JM, Faix RG, Pryce CJ, Bhatt-Mehta V: Randomized, blind trial of dopamine versus dobutamine for treatment of hypotension in preterm infants with respiratory distress syndrome. J Pediatr 1994;125:117–122. 29 Emery EF, Greenough A: Efficacy of low-dose dopamine infusion. Acta Paediatr 1993;82:430– 432. 30 Lopez SL, Leighton JO, Walther FJ: Supranormal cardiac output in the dopamine- and dobutaminedependent preterm infant. Pediatr Cardiol 1997; 18:292–296.

31 Pladys P, Wodey E, Beuchee A, Branger B, Betremieux P: Left ventricle output and mean arterial blood pressure in preterm infants during the 1st day of life. Eur J Pediatr 1999;158:817–824. 32 Meek JH, Tyszczuk L, Elwell CE, Wyatt JS: Low cerebral blood flow is a risk factor for severe intraventricular haemorrhage. Arch Dis Child Fetal Neonatal Ed 1999;81:F15–F18. 33 Wardle SP, Yoxall CW, Weindling AM: Determinants of cerebral fractional oxygen extraction using near infrared spectroscopy in preterm neonates. J Cereb Blood Flow Metab 2000;20:272–279. 34 Jobsis FF: Noninvasive, infrared monitoring of cerebral and myocardial oxygen sufficiency and circulatory parameters. Science 1977;198:1264– 1267. 35 Brazy JE, Lewis DV, Mitnick MH, Jobsis vander Vliet FF: Noninvasive monitoring of cerebral oxygenation in preterm infants: Preliminary observations. Pediatrics 1985;75:217–225. 36 Wyatt JS, Cope M, Delpy DT, Wray S, Reynolds EO: Quantification of cerebral oxygenation and haemodynamics in sick newborn infants by near infrared spectrophotometry. Lancet 1986;ii:1063– 1066. 37 Skov L, Pryds O, Greisen G, Lou H: Estimation of cerebral venous saturation in newborn infants by near infrared spectroscopy. Pediatr Res 1993;33: 52–55. 38 Yoxall CW, Weindling AM, Dawani NH, Peart I: Measurement of cerebral venous oxyhemoglobin saturation in children by near-infrared spectroscopy and partial jugular venous occlusion. Pediatr Res 1995;38:319–323. 39 Bucher HU, Edwards AD, Lipp AE, Duc G: Comparison between near infrared spectroscopy and 133Xenon clearance for estimation of cerebral blood flow in critically ill preterm infants. Pediatr Res 1993;33:56–60.

Biol Neonate 2001;79:241–245

245

Biol Neonate 2001;79:246–253

Monitoring of Antepartum and Intrapartum Fetal Hypoxemia: Pathophysiological Basis and Available Techniques G. Clerici R. Luzietti G.C. Di Renzo Centre of Reproductive and Perinatal Medicine, University of Perugia, Perugia, Italy

Key Words Fetal hypoxemia W Antepartum fetal monitoring W Intrapartum fetal monitoring W Perinatal asphyxia W Cardiotocography W Fetal pulse oximetry W Fetal intrapartum electrocardiogram W Fetal hemodynamics W Doppler technology

Abstract The challenge of obstetric surveillance is to identify those fetuses whose physiological defence mechanisms are compromised, in order to be able to act before decompensation has occurred. During the antenatal period, the evaluation of fetal hemodynamic adaptation to hypoxemia and the assessment of its chronological evolution by Doppler technology are crucial. During the intrapartum period, the relative inaccessibility of the fetus and the complexity of the pathophysiology of fetal oxygenation make it difficult to obtain and interpret information on the fetal response to labor stress. Due to the limitations of cardiotocography, additional information is required for appropriate decision making during labor. Current evidence suggests that modern technology applied to fetal surveillance can provide useful additional information that can improve our capacity to interpret fetal reactions to labor events. Copyright © 2001 S. Karger AG, Basel

Introduction

Fetal asphyxia can be defined as the combination of a lack of oxygen, metabolic acidosis and impaired organ function. It is important to remember that it occurs as a result of a cascade of events over time [1]. The first step is represented by hypoxemia, which is a reduction in the oxygen carried in the blood as a result of a decreased pO2 and decreased oxygen content. When defense mechanisms fail to compensate for the decrease in blood oxygen content, hypoxia develops in the tissues. The fetus can

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0246$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

still supplement energy production with anaerobic metabolism and maintain organ function. Anaerobic metabolism, however, only produces one fifth of the ATP produced aerobically and also leads to the production of lactate, which tends to accumulate, giving rise to metabolic acidosis. When energy production can no longer be maintained in this way, asphyxia develops and the risk of organ failure and tissue damage increases [1]. The diagnosis of asphyxia at birth requires the assessment of blood gas acid-base values and the hypoxia-related morbidity of the newborn in the neonatal period. Cord metabolic acidemia with a pH ! 7.00 and a base deficit 612 mmol/l is regarded as a marker of significant fetal metabolic adjustments to intrapartum hypoxia and a level above which moderate or severe complications may occur [2]. In order to avoid the impact of respiratory acidosis on base deficit calculations, buffer changes should be calculated in the extracellular fluid [3].

Pathophysiology of the Antepartum Fetal Response to Chronic Hypoxemia

The placenta is the ‘lung of the fetus’. In the placenta, oxygen from maternal red cells is exchanged for carbon dioxide coming from the fetal circulation. Deoxygenated fetal blood reaches the chorionic villi via the two umbilical arteries and oxygenated fetal blood returns to the fetus via the single umbilical vein. The partial pressure of oxygen in the fetal blood is much lower than in the adult (fig. 1). However, due to the high fetal cardiac output, organ blood flow, hemoglobin concentration and affinity of fetal hemoglobin for oxygen, the fetus normally has a surplus of oxygen available for its energy requirements. Fetal hypoxemia may be the result of different feto-maternal pathophysiological processes which can produce completely different fetal hemodynamic modifications, not only in relation to the quality but particularly in relation to the chronology of the hemodynamic

Dr. G.C. Di Renzo Department of Obstetrics, Gynecology and Pediatrics University of Perugia, Policlinico Monteluce I–06122 Perugia (Italy) Tel. +39 075 572 0563, Fax +39 075 572 9271, E-Mail [email protected]

Fig. 1. Oxygen transfer from mother to fetus. The numbers refer to the partial pressure of oxygen (mm Hg) in inspired air, maternal alveolar air, maternal arterial and venous blood, umbilical venous and arterial blood and fetal venous blood. Modified from: Richardson BS, Canadian Medical Protective Association.

events. However, fetal antepartum oxygen deficiency is mostly due to placental vascular insufficiency, and it is important to point out that fetal hypoxemia-acidemia is part of the terminal pathway starting from placental functional and structural alterations through to fetal growth restriction, potentially leading to fetal damage or intrauterine fetal death. Sonography and, particularly, Doppler ultrasound technologies can help the obstetrician in the evaluation of the antepartum well-being of the fetus, and of fetal hemodynamic adaptations to different maternal and fetal pathophysiological conditions leading to fetal hypoxemia. Several mechanisms are involved in the beginning of the processes which lead to fetal hemodynamic changes, from adaptation to decompensation, during hypoxemia: feto-maternal immunologic tolerance alterations, failure of the endothelial vasodilator tone control (possibly alterations in the NO system), reduction of maternal plasmatic expansion, increased maternal blood viscosity at a low shear rate, inappropriate trophoblastic invasions with histological, morphological and functional placental alterations and others. All these processes are involved in the hemodynamic alterations in both uterine and umbilical arteries which characterize fetuses with fetal growth restriction [4–10]. When the structural and functional placental alterations appear and/or increase, the fetus adapts itself to this situation with decreased growth, alterations in behavior (i.e. decrease in the episodes of body movements) and hemodynamic changes in order to maintain the supply of oxygen and substrates for tissues with active metabolism such as the brain, heart and adrenals [11]. Only when the obstruction of placental vessels is greater than 60% is there a detectable and clear alteration in the umbilical artery velocity waveform profile [10]. Thus, when a particular level of pO2 is reached, there is a redistribution of the fetal blood flow. These hemodynamic modifications, which are known as the ‘brain sparing effect’ and which produce a ‘fetal hemodynamic centralization’, are thought

to be protective against hypoxic insult and consist of vasodilatation with an increase in blood flow in the fetal structures which are most sensitive to hypoxemia (such as the brain, adrenals and coronary arteries) and a decrease in the blood supply in the peripheral vascular districts such as pulmonary, intestinal, cutaneous, renal and skeletal vessels [11–19]. These changes in arterial perfusion are mediated by neuronal stimulation, either directly through stimulation of the vagal center or through chemoreceptors in the aorta and in the carotid arteries. If the uteroplacental vascular bed alterations persist, this produces a further increase in the impedance to flow in the umbilical artery and in the fetal aorta and, mainly as a result of the hypoxemia, in the renal artery. Moreover, these factors cause a further increase in the hypoxemic fetal status, balanced by a more pronounced fetal blood flow redistribution with lowest impedance to flow values in the cerebral vessels. This ‘centralization of blood flow’ influences cardiac hemodynamics with a decreased left ventricle afterload due to the cerebral vasodilatation and an increased right ventricle afterload due to the systemic vasoconstriction. This phase is characterized by the extreme response of the fetus to hypoxemia, which may lead to the decompensatory phase (fig. 2). The last phase is characterized by the impairment of fetal cardiac function, which is unable to balance all the factors mentioned above. Due to the persistent severe hypoxemia and the consequent polycythemia and increased blood viscosity, there is an impairment of fetal cardiac contractility, which is the most important factor leading to the terminal decompensatory phase. The impairment of cardiac function causes a decrease in the cardiac afterload and an increase in the cardiac preload, leading to an increase in the atrioventricular gradient and abnormal ventricular filling with an increase in venous pressure beyond the inferior vena cava, hepatic and ductus venosus circulation throughout the umbilical vein blood flow. Moreover, during this stage, the reduced cardiac output and the high blood viscosity

Monitoring of Antepartum and Intrapartum Fetal Hypoxemia

Biol Neonate 2001;79:246–253

247

Fig. 2. Pathophysiology of the hemodynamic redistribution induced by fetal hypoxemia due to uteroplacental vascular insufficiency. PG = Prostaglandins; EDRF = endothelium derived relaxing factor.

also cause a reduction in the cerebral perfusion, leading to the disappearance of the so-called ‘brain sparing’. The disappearance of the latter may also be induced by a mechanical mechanism brought about by the edema caused by the brain damage from the hypoxic insult [14].

increase in the impedance to flow in the umbilical artery and in the fetal peripheral vessels and a minor decrease in the cerebral vessels), and the main hemodynamic change which it is possible to detect with Doppler technology is the decrease in the impedance to flow values characterizing the subcortical segment (M2) of the MCA, but the clinical usefulness of this hemodynamic event is still unclear [13].

Techniques of Antepartum Fetal Surveillance: Doppler Evaluation of Fetal Hemodynamic Adaptation and Its Chronology

Early Stage of Fetal Blood Flow Redistribution Fetal Doppler velocimetry shows an increase in the impedance to flow values as expressed by an increase in the PI of the umbilical artery and also of the aorta and the renal artery. The increase in the vascular resistance of the aorta is probably related to different factors, such as the increase in vascular impedance in the umbilicoplacental vessels and arterial vasoconstriction of peripheral vessels due to progressive hypoxemia. During this phase, it is possible to observe some hemodynamic modifications which involve the whole fetal organism. These modifications are related to the substantial redistribution of the cardiac output that goes in the direction of the tissues which are important for fetal life. The inversion of the cerebroplacental ratio, the so-called ‘brain sparing effect’, is the most evident hemodynamic effect. At this stage, a statistically significant increase in the blood flow and a decrease in the resistance in all the cerebral vessels examined can be seen, while, due to the hemodynamic redistribution, a decrease in the peripheral flow in the umbilical artery, abdominal aorta, renal artery, femoral artery and other vessels can be observed, with high impedance to flow values [11–18]. The ratio between the PI of the MCA and the PI of the umbilical artery, the so-called ‘cerebroplacental ratio’, can be considered as the Doppler flow expression of the ‘brain sparing effect’; a decrease in this ratio below two standard deviations is a sign of incipient severe hypoxemia, and in its presence, it is possible to observe anomalies of the fetal biophysical profile, reduction of fetal heart rate variability and a reduction in amniotic fluid volume [16]. During this stage, the PI of the umbilical artery and the fetal aorta is elevated, but Doppler velocimetry frequency values continue to be positive throughout the whole cardiac cycle, even in the end-diastolic

Doppler Silent Stage Considering the processes which lead to conclamate uteroplacental vascular insufficiency, the fetal hemodynamic profile might remain ‘normal’ even for a long period of time; the umbilical artery velocity waveform shows a positive blood flow pattern throughout the whole cardiac cycle and the impedance to flow values expressed as the pulsatility index (PI) is normal, with a nonsignificant increase. The Doppler velocimetry of the remaining main fetal vessels and districts (particularly the aorta, renal artery, femoral artery, cerebral vessels, etc.) is also in the normal range, with nonsignificant alterations. Under these ‘normal’ conditions, the mean PI of the middle cerebral artery (MCA) is found to be higher than that of either the internal carotid artery or the anterior cerebral artery, while that of the posterior cerebral artery is usually found to be lower than that of the MCA and anterior cerebral artery, and is higher than that of the umbilical artery. The MCA, because of its size and the simplicity of its sampling, has been one of the most investigated cerebral vessels and it appears to be one of the most sensitive to initial hypoxemia [12]; in particular, it seems that its subcortical segment (M2) responds earlier than the proximal part of the vessel (M1). Besides, the ratio between the flow indices of the two parts of the vessel (M2/ M1) became lower than two standard deviations in the presence of an initial fetal hypoxemic status [13]. In conclusion, the alteration in the uteroplacental vascular bed and the alterations in the placental metabolites and gas exchange, in this initial stage, produce only minor and not significant fetal hemodynamic modifications (minor

248

Biol Neonate 2001;79:246–253

Clerici/Luzietti/Di Renzo

Table 1. Different stages of fetal blood flow redistribution during hypoxemia due to uteroplacental vascular insuffi-

ciency Fetal vessels

Doppler silent stage

Early stage of hemodynamic redistribution

Advanced stage Decompensation of hemodynamic stage redistribution

Middle cerebral artery-M2 Middle cerebral artery-M1 Anterior cerebral artery Posterior cerebral artery Internal carotid artery Common carotid artery Aorta Umbilical artery Renal artery External iliac artery Femoral artery Pulmonary valve Aortic valve Mitral valve Tricuspid valve Inferior vena cava Ductus venosus Umbilical vein

↓ =↓ =↓ =↓ =↓ =↓ =↑ =↑ =↑ =↑ =↑ = = = = = = =

↓ ↓ ↓ ↓ ↓ ↓ ↑ ↑ ↑ ↑ ↑ ↑ ↓ = = = = =

↓ ↓ ↓ ↓ ↓ ↓ ↑ (ADF) ↑ (ADF) ↑ ↑ ↑ ↑ ↓ = = ↑= ↑= =

↑ ↑ ↑ ↑ ↑ ↑ ↑ (RDF) ↑ (RDF) ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑ (RDF) ↑ (pulsation)

Arrows indicate the changes in impedance to flow. ADF = Absent diastolic flow; RDF = reverse diastolic flow.

phase. On the other hand, and at the same time, it is possible to find high-velocity frequencies during diastole in all cerebral vessels, suggesting an increase in the fetal cerebral blood flow. Advanced Stage of Fetal Hemodynamic Redistribution This phase is essentially characterized by a further increase in the impedance to flow in the umbilical artery, in the fetal aorta and in the renal artery. Looking at the umbilical artery flow velocity profile, a decrease in the diastolic frequencies is observed progressing towards the absence of diastolic flow; the end-diastolic frequency disappears first, but subsequently, the lack of blood flow is evident in the whole diastolic phase. This usually occurs when 80% of villi arterioles are occluded [10]. Aortic velocity waveforms also exhibit a similar pattern with an absence of diastolic frequencies, usually preceding those observed in the umbilical artery. At the same time, the impedance to flow values in the cerebral vessels shows a further decrease, leading to the lowest PI values in this district as a result of concurrent maximal vasodilatation of cerebral vessels [14]. Moreover, during this phase, it is possible to find a relative decrease in the right cardiac output and an increased left cardiac output characterized by increased time to peak velocity in the aorta and a decrease in the same parameter in the pulmonary arteries, suggesting a preferential shift of cardiac output in favor of the left ventricle, leading to improved perfusion to the brain. Decompensatory Phase During this phase, the cardiac output and the peak velocity of the main arterial trunks gradually decline and, as a consequence, cardiac

Monitoring of Antepartum and Intrapartum Fetal Hypoxemia

filling is impaired, suggesting a progressive deterioration of cardiac function. Therefore, these factors cause changes that induce hemodynamic alterations in all cardiovascular districts (intracardiac, arterial and venous districts). The incipient heart failure produces a decreased cardiac output which causes a decrease in the peak velocity of the outflow tracts leading to reverse flow in the aorta, in the umbilical artery and lastly, as a terminal sign, in many other arterial vessels such as the cerebral vessels [19]. During this phase, the increased viscosity of the fetal blood, the decrease in the cardiac output and, probably, the cerebral edema, produce a decrease in brain perfusion, shown by the decrease in blood velocity especially during diastole and, thus, the disappearance of the ‘brain sparing effect’ [14–19]. At the same time, the impairment of the cardiac contractility causes an increased atrioventricular gradient, abnormal ventricular filling underlined by a decrease in the E/A ratio (E: pick due to ventricular diastole; A: pick due to atrial systole) of the atrioventricular blood flow velocity waveforms. The increased pressure gradient in the right atrium leads, during atrial contractions, to the evidence of reverse flow in the ductus venosus and to a high percentage of abnormal reverse flow in the inferior vena cava. The next step is the extension of the abnormal reversal of blood flow from the inferior vena cava beyond the ductus venosus and the hepatic circulation into the umbilical vein, causing typical end-diastolic pulsations in this vessel. It has been observed that this hemodynamic pattern is associated with the onset of severe fetal heart rate anomalies and with severe acidemia at birth [16]. Table 1 summarizes the different hemodynamic patterns during different stages of fetal hypoxemia.

Biol Neonate 2001;79:246–253

249

Table 2. CTG classification [24]

CTG classification

Baseline heart frequency

Variability reactivity

Decelerations

Normal CTG

110–150 bpm

5–25 bpm accelerations

early decelerations uncomplicated variable decel. with a duration of ! 60 s and loss of ! 60 beats

Intermediary CTG

100–110 bpm 1 25 bpm without accelerations uncomplicated variable decel. with a 150–170 bpm ! 5 bpm for 1 40 min duration of ! 60 s and loss of 1 60 beats short bradycardia episode A combination of several intermediary observations will result in an abnormal CTG

Abnormal CTG

150–170 bpm and reduced variability 1 170 bpm persistent bradycardia

Preterminal CTG

Total lack of variability and reactivity with or without decelerations or bradycardia

! 5 bpm for 1 60 min sinusoidal pattern

Pathophysiology of the Fetal Response to Acute Hypoxemia during Labor

The events of labor frequently expose the fetus to episodes of reduction of placental blood flow and consequent reduction of oxygen delivery. Cord compression, for example, can be responsible for impairment of oxygen delivery to the fetus by altering the fetal myocardial preload and afterload. Another common factor responsible for intermittent reduction or interruption of feto-maternal gas exchange is represented by uterine contractions. The rise in intramyometrial pressure during uterine contractions can affect fetal perfusion by compression of the spiral arteries supplying the intervillous space of the placenta. Several maternal factors can impair appropriate intrapartum fetal oxygenation, such as maternal hypotension, maternal respiratory depression, anesthetic agents and drugs. Other more rare acute events associated with labor are placental abruption and cord prolapse [20]. The fetal ability to adapt to hypoxemia involves multiple defense mechanisms [for a more comprehensive review, see ref. 1]. Through the activation of cardiovascular reactions, the fetus can compensate for hypoxemia by increasing the blood flow to the most important organs, i.e. the brain, heart and adrenals, thereby counteracting the decreasing oxygen content. Hypoxemia causes a decrease in the fetal heart rate and an increase in blood pressure, secondary to an intense vasoconstriction at the level of the skin, muscles and gut. This allows a greater proportion of the cardiac output to be distributed to high-priority organs, so that oxygen delivery to central organs can be maintained despite hypoxemia [21]. Aerobic metabolism can be maintained in this situation until the oxygen content of arterial blood is decreased by 70% [22]. A second line of defense is represented by the metabolic compensatory mechanisms. When cardiovascular mechanisms can no longer compensate for hypoxemia, aerobic metabolism can be integrated by the anaerobic metabolism of the glucose stores accumulated as glycogen. The importance of anaerobic metabolism in maintaining organ functions during hypoxia has long been established and depends on the prehypoxial content of glycogen in the heart and liver [23]. Anaerobic glycogenolysis, however, leads to the production of

250

Biol Neonate 2001;79:246–253

complicated variable decel. with a duration of 1 60 s repeated late decelerations

lactate, which tends to accumulate in the tissues, giving rise to increasing metabolic acidemia. It is only when these compensatory mechanisms are insufficient or exhausted that asphyxia will develop and along with it the possibility of central nervous system damage and handicap.

Techniques of Intrapartum Fetal Surveillance

Cardiotocography Continuous fetal heart rate and uterine contraction recording (cardiotocography; CTG) is widely used to assess fetal well-being during labor [24] (table 2). However, this method has limitations. A normal CTG trace reflects optimal fetal oxygenation and is reassuring regarding fetal conditions. The significance of fetal heart rate changes is often unclear and therefore difficult to interpret. In the clinical scenario, this can result in unnecessary interventions for suspected fetal hypoxia or inappropriate delay in action with potentially disastrous consequences for the fetus [25]. Some of these difficulties can be overcome by better training of medical and midwifery staff. Evidence also suggests that the use of expert systems for decision support would provide a valuable contribution to improving the detection and clinical management of cases with abnormal CTG patterns [26]. However, it is also evident that there are situations where CTG changes are not specific enough for the presence of fetal hypoxia and additional information is necessary for appropriate decision making. Pulse Oximetry Fetal pulse oximetry monitoring provides a quantitative, direct and real-time measurement of fetal arterial oxygen saturation (SpO2) [27]. The term oxyhemoglobin describes hemoglobin when all of the available binding sites of hemoglobin are fully bound with oxygen. Hemoglobin molecules not carrying oxygen are referred to as deoxyhemoglobin. Oxygen saturation monitoring measures the ratio of hemoglobin molecules bound with oxygen (oxyhemoglobin) to the total amount of hemoglobin molecules available to bind with oxygen (oxyhemoglobin plus deoxyhemoglobin) [27]. Oxyhemoglobin and deoxyhemoglobin differ in their absorption of red and infrared light.

Clerici/Luzietti/Di Renzo

The pulsatile changes in absorption of red and infrared light are used to determine the SpO2 of fetal blood [28]. The oximeter sensors used in the fetus (reflectance pulse oximetry) have light-emitting diodes and a photodetector that are adjacent to one another on a flexible probe, and the absorption of light is determined from the light that scatters back to the tissue surface. The probe is placed, when the membranes have been ruptured and cervical dilatation is more than 2 cm, between the fetal cheek and the uterine wall. The percentage oxygen saturation values are printed continuously on the CTG paper. Experimental studies have demonstrated that it is when SpO2 levels fall below 30% that we assist in an increase in lactate production and decrease in pH [29]. Similar findings have been reported in human fetuses, comparing SpO2 levels and acidemia from the fetal scalp [30]. A level of SpO2 of 30% is therefore used clinically as the cut-off above which fetuses are considered not to be hypoxemic. Clinical observational studies have evaluated the usefulness of this information in discriminating nonreassuring CTG fetal heart rate patterns during labor [31, 32]. These studies showed that the predictive value of intrapartum fetal pulse oximetry can be favorably compared with that of fetal scalp blood analysis and pH of cord vein blood at birth. A recent US multicenter randomized trial of 1,010 laboring women with a nonreassuring fetal heart rate tracing showed a reduction in emergency cesarean sections from 10 to 5% with the use of CTG plus pulse oximetry. However, unexpectedly, the study also showed an increase in the section rate for failure to progress in the test group (19 versus 9%), and the overall section rates were not different between the test and control groups [33]. The current literature reveals somewhat diverging views on the information available from fetal pulse oximetry during labor. Pulse oximetry appears to be a screening technique that can help in discriminating nonreassuring fetal heart rate tracing. However, the issue still to be resolved is the ability of CTG plus pulse oximetry to allow a diagnosis of fetal metabolic acidosis. Thus, we may have a situation where the two parameters in combination may not be specific enough to enable the obstetrician to grade the impact of hypoxemia on fetal organ function [34, 35]. Therefore, it may be difficult to rely only on the actual level of oxygenation. Instead, it may be more rewarding to try to interpret the reactions taking place in a high-priority organ like the heart or the brain. ST Waveform Analysis of the Fetal Electrocardiogram ST analysis has emerged not as an alternative to CTG but as a support tool to allow more accurate interpretation of intrapartum events. The fetal electrocardiogram (ECG) is readily obtainable during labor from the same scalp electrode used to obtain the fetal heart rate [36], utilizing a dedicated CTG plus fetal ECG monitor (STAN® S 21, Neoventa Medical AB, Gothenburg, Sweden). Numerous experimental animal studies have clarified the pathophysiology of ST waveform changes on the fetal ECG during hypoxia [37, 38]. The ST segment and T wave relate to the repolarization of myocardial cells in preparation for the next contraction. This repolarization process is energy consuming. An increase in T wave height, quantified by the ratio between T and QRS amplitudes, occurs when the energy balance within the myocardial cells threatens to become negative. A negative energy balance indicates a situation in which the

Monitoring of Antepartum and Intrapartum Fetal Hypoxemia

Fig. 3. The ECG with a schematic presentation of hypoxia-related

changes. The T/QRS measurement is also indicated. From: Sundström AK, Rosén D, Rosén KG: Fetal Surveillance, Neoventa Medical AB, with permission.

amount of oxygen supplied to the cells no longer covers the energy required for metabolic activity. During hypoxia, this balance becomes negative and the cells produce energy by the beta-adrenoceptor-mediated anaerobic breakdown of glycogen reserves. The ability of these cells to produce energy in this manner and thereby maintain myocardial function is a vital compensatory defense mechanism. This process not only produces lactic acid but also potassium ions, which affect the myocardial cell membrane potential and cause a rise in the ST waveform. ST depression with negative T waves has been observed during hypoxia experiments in experimentally growth-retarded guinea pigs [39]. Clinically, these changes have emerged as a specific sign of myocardial hypoxic stress. They reflect a myocardium that is not able or had not had the time to mobilize its defense to hypoxemia. The result is a decrease in myocardial activity and a risk of cardiovascular failure (fig. 3). In conclusion, the evidence from experimental work indicates that ST waveform elevation reflects compensated myocardial stress and a switch to anaerobic myocardial metabolism. A progressive rise in the T/QRS ratio represents continuing anaerobic metabolism with a risk of eventual decompensation due to the depletion of myocardial glycogen stores and a progressive metabolic acidosis. Persistently biphasic and negative waveform changes indicate myocardial decompensation as a result of direct myocardial ischemic hypoxia. Clinical analysis of ST waveform changes is assisted by a specifically developed computerized ST log function that provides direct statements on specific significant ST events, to provide additional user support [40]. This pathophysiological model of interpretation has led to the development of specific clinical action guidelines that have been tested in several observational and randomized control studies [41– 44]. These studies demonstrated the high sensitivity of CTG plus ST for predicting fetal acidosis, associated with a significant increase in positive predictive values as compared with CTG only. The first randomized trial comparing CTG only with CTG plus ST analysis in 2,400 cases [44] showed a 46% reduction in operative interventions for fetal distress in the CTG plus ST arm of the trial.

Biol Neonate 2001;79:246–253

251

The results from the recent Swedish randomized trial on CTG alone versus CTG plus ST analysis (4,495 cases) [45] showed a 60% reduction in the number of cases with metabolic acidosis (defined as cord artery pH ! 7.05 and base deficit in the extracellular fluid 1 12 mmol/l) in the CTG plus ST arm of the trial, accompanied by a 25% reduction in operative interventions for fetal distress as compared with the CTG only arm, with no increase in operative deliveries for other reasons. The trial protocol allowed for an interim analysis after 1,600 cases. This analysis showed frequent breaches of protocol, as clinical management in the CTG plus ST arm was conducted according to the ‘old’ CTG information. The result of this lack of compliance was not only more operative interventions but also babies being exposed to unnecessary intrauterine hypoxia, with 2 babies requiring neonatal intensive care. After retraining and enhanced experience with ST analysis that allowed a more rigorous application of the CTG plus ST clinical action protocol, it was possible to obtain an even more pronounced reduction in metabolic acidosis (75%) in the second half of the trial, with no babies admitted to the neonatal intensive care unit and a decrease of 44% in the operative delivery rate for fetal distress. These results confirm the capacity of ST waveform analysis to provide diagnostic information on developing hypoxia during labor, which can lead to a significant improvement in fetal outcome.

Conclusions

The challenge of obstetric surveillance is to identify those fetuses whose physiological defence mechanisms are compromised, in order to be able to act before decompensation has occurred.

During the antenatal period, the evaluation of fetal hemodynamic adaptation to hypoxemia and the assessment of its chronological evolution by Doppler technology are crucial. This assists in planning appropriate obstetrical management and in reducing the risks of fetal damage. In addition to conventional Doppler evaluation of fetal arterial districts, which is important for the diagnosis of fetal hemodynamic adaptation to hypoxemia, it is important to consider also the intracardiac and venous hemodynamics. The evaluation of the output tracts, the atrioventricular flow and vessels like the ductus venosus, inferior vena cava and umbilical vein, provides more detailed information on the incipient failure of the compensatory mechanism of the fetus, because this heralds the development of right heart failure due to myocardial hypoxia. During the intrapartum period, the relative inaccessibility of the fetus and the complexity of the pathophysiology of fetal oxygenation make it difficult to obtain and interpret information on the fetal response to labor stress. Due to the limitations of CTG, additional information is required for appropriate decision making during labor. The results of clinical randomized studies show the capacity of modern technology applied to fetal surveillance, and in particular the analysis of fetal ECG, to provide useful additional information that can improve our ability to interpret fetal reactions to labor events. The achievement of a significant improvement in intrapartum fetal surveillance is, however, related not only to the availability of more specific information but also to the capacity to make better use of the information available. This requires clinical skills, a knowledge of fetal physiology and an understanding of the technical basis and limitations of the methodologies of intrapartum monitoring.

References 1 Greene KR, Rosén KG: Intrapartum asphyxia; in Levene MI, Lilford RJ (eds): Fetal and Neonatal Neurology and Neurosurgery. New York, Churchill Livingstone, 1995. 2 MacLennan A: A template for defining a causal relation between acute intrapartum events and cerebral palsy: International consensus statement. BMJ 1999;319:1054–1059. 3 Rosén KG, Murphy K: How to assess fetal metabolic acidosis from cord samples. J Perinat Med 1991;19:221–226. 4 Jauniaux E, Jurkovic D, Campbell S, Hustin J: Doppler ultrasonographic features of the developing placental circulation: Correlation with anatomic findings. Am J Obstet Gynecol 1992;166: 585–587. 5 Warwick BG, Trudinger BJ, Baird PJ: Fetal umbilical artery flow velocity waveforms and placental resistance: Pathological correlation. Br J Obstet Gynaecol 1985;92:31–38. 6 Nordenvall M, Ullberg U, Laurin J, Lingman G, Sandstedt B, Ulmsten U: Placental morphology in relation to umbilical artery blood velocity waveforms. Eur J Obstet Gynecol Reprod Biol 1991;40: 179–190. 7 Trudinger BJ, Giles WB, Cook CM: Uteroplacental blood flow velocity-time waveforms in normal and complicated pregnancy. Br J Obstet Gynaecol 1985;92:39–45.

252

8 Trudinger BJ, Giles WB, Cook CM: Flow velocity waveforms in the maternal uteroplacental and fetal umbilical placental circulations. Am J Obstet Gynecol 1985;152:155–163. 9 Trudinger BJ, Giles WB, Cook CM, Bombardieri J, Collins L: Fetal umbilical artery flow velocity waveforms and placental resistance: Clinical significance. Br J Obstet Gynaecol 1985;92:23–30. 10 Trudinger BJ, Stevens D, Connelly A, Hales JRS, Alexander G, Bradley L, Fawcett A, Thompson RS: Umbilical artery flow velocity waveforms and placental resistance: The effects of the embolization of the umbilical circulation. Am J Obstet Gynecol 1987;157:1443–1448. 11 Mari G, Deter RL: Middle cerebral artery flow velocity waveforms in normal and small-for-gestational-age fetuses. Am J Obstet Gynecol 1992;166: 1262–1270. 12 Veille JC, Penry M: Effect of maternal administration of 3% carbon dioxide on umbilical artery and fetal renal and middle cerebral artery Doppler waveforms. Am J Obstet Gynecol 1992;167:1668– 1671. 13 Luzi G, Coata G, Caserta G, Cosmi EV, Di Renzo GC: Doppler velocimetry of different sections of the fetal middle cerebral artery in relation to perinatal outcome. J Perinat Med 1996;24:327–334.

Biol Neonate 2001;79:246–253

14 Clerici G, Luzi G, Di Renzo GC: Cerebral circulation from healthy to IUGR and distressed fetus: What happens and how we can explain it; in Kurjak A, Di Renzo GC (eds): Modern Methods of the Assessment of Fetal and Neonatal Brain. Rome, CIC, 1996, pp 36–50. 15 Bilardo CM, Snijders RM, Campbell S, Nicolaides KH: Doppler study of fetal circulation during longterm maternal hyperoxygenation for severe early onset intrauterine growth retardation. Ultrasound Obstet Gynecol 1991;1:250–257. 16 Scherjon SA, Smolders-DeHaas H, Kok JH, Zonderwan HA: The ‘brain sparing’ effect: Antenatal cerebral Doppler findings in relation to neurologic outcome in very preterm infants. Am J Obstet Gynecol 1993;169:169–175. 17 Weiner Z, Farmakides G, Schulman H, Penny B: Central and peripheral hemodynamic changes in fetuses with absent end-diastolic velocity in umbilical artery: Correlation with computerized fetal heart rate pattern. Am J Obstet Gynecol 1994;170: 509–515. 18 Van Den Wijngaard JA, Groenenberg IAL, Wladimiroff JW, Hop WCJ: Cerebral Doppler ultrasound of the human fetus. Br J Obstet Gynaecol 1989;96:845–849. 19 Sepulveda W, Shennan AH, Peek MJ: Reverse end-diastolic flow in the middle cerebral artery: An agonal pattern in the human fetus. Am J Obstet Gynecol 1996;174:1645–1647.

Clerici/Luzietti/Di Renzo

20 Greene KR: Intrapartum fetal monitoring: CTG, ECG and fetal blood sampling; in Rodeck CH, Whittle M (eds): Fetal Medicine: Basic Science and Clinical Practice. London, Churchill Livingstone, 2000. 21 Cohn HE, Sachs EJ, Heymann MA, Rudolph AM: Cardiovascular responses to hypoxemia and acidemia in fetal lambs. Am J Obstet Gynaecol 1974; 120:817–824. 22 Rurak DW, Richardson BS, Patrick JE, Carmichael L, Homan J: Oxygen consumption in the fetal lamb during sustained hypoxemia with progressive acidemia. Am J Physiol 1990;258:R1108– R1115. 23 Dawes GS, Mott JC, Shelley HJ: The importance of cardiac glycogen for the maintenance of life in foetal lambs and newborn animals during anoxia. J Physiol 1959;146:516–538. 24 FIGO guidelines for the use of fetal monitoring. Int J Gynaecol Obstet 1987;25:159–167. 25 Larsen JF: Why has conventional intrapartum cardiotocography not given the expected results? J Perinat Med 1996;24:15–23. 26 Greene KR: Intelligent fetal heart rate computer systems in intrapartum surveillance. Curr Opin Obstet Gynecol 1996;8:123–127. 27 Dildy GA, Clark SL, Loucks CA: Intrapartum fetal pulse oximetry: Past, present and future. Am J Obstet Gynecol 1996;175:1–9. 28 Mannheimer PD, Fein ME, Casciani JR: Physiooptical considerations in the design of fetal pulse oximetry sensors. Eur J Obstet Gynecol Reprod Biol 1997;72(suppl): S9–S19. 29 Richardson B, Carmichael L, Homan J, Patrick J: Cerebral oxidative metabolism in fetal sheep with prolonged, graded hypoxaemia. Soc Gynecol Invest 1990;35:899–900.

Monitoring of Antepartum and Intrapartum Fetal Hypoxemia

30 Luttkus A, Fengler TW, Friedman W, Dudenhausen JW: Continuous monitoring of fetal oxygen saturation by pulse oximetry. Obstet Gynecol 1995; 85:183–186. 31 Carbonne B, Langer B, Goffinet F, Audibert F, Tardif D, Le Goueff F, Laville M, Maillard F: Multicenter study on the clinical value of fetal pulse oximetry. II. Compared predictive values of pulse oximetry and fetal blood analysis. The French Study Group on Fetal Pulse Oximetry. Am J Obstet Gynecol 1997;177:593–598. 32 Di Renzo GC, Branconi F, Anceschi MM, Ferrari G, Cenci F, Mandruzzato GP, Mantegazza G: Italian group study on fetal pulse oximetry. Fetal Diagn Ther 1998;13:46–47. 33 The Mallinckrodt Fetal Oximetry System Clinical Study: www.healthymotherandbaby.com/professionals/trial.html 34 Richardson B, Nodwell A, Webster K, Alshimmiri M, Gagnon R, Natale R: Fetal oxygen saturation and fractional extraction at birth and the relationship to measures of acidosis. Am J Obstet Gynecol 1998;178:572–579. 35 Arikan GM, Scholz HS, Haeusler MC, Giuliani A, Haas J, Weiss PA: Low fetal oxygen saturation at birth and acidosis. Obstet Gynecol 2000;95:565– 571. 36 Lindecrantz K, Lilja H, Widmark C, Rosén KG: The fetal ECG during labour. A suggested standard. J Biomed Eng 1988;10:351–353. 37 Rosén KG, Isaksson O: Alterations in fetal heart rate and ECG correlated to glycogen, creatine phosphate and ATP levels during graded hypoxia. Biol Neonate 1976;30:17–24.

38 Rosén KG, Dagbjartsson A, Henriksson BA, Lagercrantz H, Kjellmer I: The relationship between circulating catecholamines and ST waveform in the fetal lamb electrocardiogram during hypoxia. Am J Obstet Gynecol 1984;149:190–195. 39 Widmark C, Jansson T, Lindecrantz K, Rosén KG: ECG waveform, short term heart rate variability and plasma catecholamine concentrations in response to hypoxia in intrauterine growth retarded guinea pig fetuses. J Dev Physiol 1991;15:161– 168. 40 Rosén KG, Luzietti R: Intrapartum fetal monitoring: Its basis and current developments. Prenat Neonat Med 2000;5:1–14. 41 Arulkumaran S, Lilja H, Lindecrantz K, Ratnam SS, Thavarasah AS, Rosén KG: Fetal ECG waveform analysis should improve fetal surveillance in labour. J Perinat Med 1990;18:13–22. 42 Luzietti R, Erkkola R, Hasbargen U, Mattsson L, Thoulon J-M, Rosén KG: European Community multi-Center Trial ‘Fetal ECG Analysis During Labor’: ST plus CTG analysis. J Perinat Med 1999; 27:431–440. 43 Luzietti R, Rosén KG: ST waveform analysis of the fetal ECG and intrapartum hypoxia. XVII European Congress of Perinatal Medicine. Prenat Neonat Med 2000;5:30. 44 Westgate J, Harris M, Curnow JSH, Greene KR: Plymouth randomized trial of cardiotocogram only versus ST waveform plus cardiotocogram for intrapartum monitoring in 2,400 cases. Am J Obstet Gynecol 1993;169:1151–1160. 45 Amer-Wåhlin I, Norén H, Hellsten C, et al: Randomised controlled trial of CTG versus CTG+ST analysis of the fetal ECG. Int J Gynaecol Obstet 2000;70:35.

Biol Neonate 2001;79:246–253

253

Biol Neonate 2001;79:254–257

Glutamate in Cerebral Tissue of Asphyxiated Neonates during the First Week of Life Demonstrated in vivo Using Proton Magnetic Resonance Spectroscopy Floris Groenendaal a Ariadne M. Roelants-van Rijn a Jeroen van der Grond b Mona C. Toet a Linda S. de Vries a a Department b Department

of Neonatology, Wilhelmina Children’s Hospital, University Medical Center, of Radiology, University Medical Center, Utrecht, The Netherlands

Key Words Neonate W Hypoxia-ischemia W Proton magnetic resonance spectroscopy W Glutamate

Abstract We tested the hypothesis that glutamate (Glx) levels as demonstrated by proton magnetic resonance spectroscopy (1H-MRS) are elevated in brain tissue of neonates with severe hypoxic-ischemic encephalopathy (HIE). Studies were performed in 26 neonates (median gestational age 40.5 weeks, range 36.7–42.4 weeks; median birth weight 3,360 g, range 2,180–4,200 g). The median postnatal age at the time of testing was 2.5 days (range 1–7 days). HIE was scored according to Sarnat as grade I (n = 4), grade II (n = 15) or grade III (n = 7). Results for neonates with mild to moderate HIE (group 1) were compared to those with severe HIE (group 2). After magnetic resonance imaging, 1H-MRS was performed in a single volume of interest including the basal ganglia. An echo time of 31 ms was used. After curve-fitting procedures, peak area ratios of different brain metabolites were calculated. The median total Glx/ N-acetylaspartate ratio was 1.21 (range 0.64–3.25) in group 1 versus 1.55 (range 1.10–2.75) in group 2 (p = 0.035). The median total Glx/ choline ratio was 1.33 (range 0.71–2.52) in group 1 versus 2.14 (range 1.21–3.55) in group 2 (p = 0.019). We concluded that during the first days of life, Glx was elevated in the basal ganglia of neonates with severe HIE. Copyright © 2001 S. Karger AG, Basel

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0254$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

Introduction

Animal experiments have emphasized the role of the excitotoxic amino acid glutamate (Glx) in causing progression of neurological damage following brain hypoxia-ischemia [1, 2]. Levels of Glx have been found to be raised in the cerebrospinal fluid of asphyxiated human neonates, and are correlated with the severity of hypoxicischemic encephalopathy (HIE) [3]. With short echo times (TEs), e.g. 31 ms, Glx can be detected in vivo in cerebral tissue by proton magnetic resonance spectroscopy (1H-MRS) [4, 5]. A recent paper by Pu et al. [6] suggested that 1H-MRS-detectable levels of Glx were elevated in neonates with severe HIE. However, they used a TE of 135 ms for the detection of ·-Glx, which is not the optimal TE for analysis of Glx. Previous studies have shown changes in the cerebral metabolism of human neonates and animals following hypoxia-ischemia [7–10]. Using 1H-MRS, decreases in N-acetylaspartate (NAA)/choline (Cho) (NAA/Cho) ratios have been shown to predict poor neurodevelopment [7] An increase in brain lactate, which under normal circumstances is present in only very small amounts in the neonatal brain and is hardly detectable by 1H-MRS at term age, has also been found to predict a poor outcome [8, 11]. These studies used TEs of 272 or 136 ms [7–9, 11]. With shorter TEs, e.g. 30 ms, lipids and macromolecules can be identified. Especially in ischemic brain areas, these metabolites with resonances of 0.9–1.3 ppm will be massively elevated, thereby complicating the detection of lactate at 1.33 ppm [12]. Therefore, short TEs have not been used frequently in neonatal studies.

Dr. Floris Groenendaal Department of Neonatology, Room KE 04.123.1 Wilhelmina Children’s Hospital/University Medical Center Utrecht PO Box 85090, NL–3508 AB Utrecht (The Netherlands) Tel. +31 302 504 545, Fax +31 302 505 320, E-Mail [email protected]

Table 1. Clinical data

Gestational age at birth, weeks Postnatal age at scan, days Birth weight, g Apgar score 1 min 5 min aEEG normal abnormal not recorded Sarnat grade I II III Died

Group 1 (n = 19)

Group 2 (n = 7)

40.3 (36.7–42.4) 3 (1–7) 3,300 (2,180–4,200) 2 (0–9) 4.5 (2–9) 14 3 2 4 15 0 0

40.6 (40–42.1) 2 (1–4) 3,460 (2,500–3,870) 1 (1–3) 1 (1–6) 0 7 0 0 0 7 5

Results are shown as median and range in parentheses. Sarnat grade according to Sarnat and Sarnat [29].

The aim of the present study was the detection of Glx in brain tissue of neonates with hypoxia-ischemia. We tested the hypothesis that Glx was elevated in neonates with severe HIE (Sarnat grade III) compared to neonates with mild to moderate HIE (Sarnat grade I or II).

neous oxygen saturation were monitored using standard equipment (Nonim, Minneapolis, Minn., USA). The respiratory rate was alo measured (Philips ACS-NT, Best, The Netherlands). Twenty of the 21 surviving infants were seen in the follow-up clinic. As 8 infants were younger than 6 months at the most recent examination, follow-up data were not used in the present study.

Patients and Methods

Patients Studies were performed in 26 neonates (median gestational age 40.5 weeks, range 36.7–42.4 weeks) with perinatal asphyxia. The patients were diagnosed as having perinatal asphyxia when at least three of the following criteria were met: abnormal fetal heart rate patterns, need for resuscitation at birth with an Apgar score ! 5 at 5 min, meconium-stained amniotic fluid, and pH of the umbilical artery ! 7.10. All infants were admitted to the Neonatal Intensive Care Unit of the Wilhelmina Children’s Hospital. Amplitude-integrated EEG (aEEG) was recorded. The background pattern was scored at 24 h according to the criteria of Toet et al. [13] as normal (continuous or discontinuous normal voltage) or abnormal (suppression burst, low voltage, flat trace). Patient characteristics are shown in table 1. Group 1 included 19 patients with Sarnat grade I or II HIE, while group 2 included 7 patients with grade III HIE. Magnetic resonance imaging (MRI) and 1H-MRS studies were performed at a median postnatal age of 2.5 days (range 1–7 days), i.e. after the hypoxic-ischemic insult. MRI examinations were performed for clinical reasons, and 1H-MRS was added to the scan protocol following informed parental consent. The study was approved by the Medical Ethical Committee of the Wilhelmina Children’s Hospital/University Medical Center Utrecht. The patients were transferred from the Neonatal Intensive Care Unit to the MRI unit in a transport incubator. If considered necessary, the infants were sedated for the examination with a combination of intramuscular pethidine (2 mg/kg body weight), chlorpromazine (0.5 mg/kg body weight) and promethazine (0.5 mg/kg body weight). We have previously demonstrated this regimen to be safe and effective [7]. During the MRI studies, heart rate and transcuta-

Brain Glutamate in Asphyxiated Neonates

MRI and 1H-MRS Studies were performed with a 1.5-tesla Philips ACS-NT system. MRI included spin echo sagittal T1 [repetition time (TR)/TE: 512/ 15 ms], axial T2 (TR/TE: 5,931/150 ms) and IR weighted (TR/time to inversion/TE: 2,500/800/30 ms) scans. Thereafter, one volume of interest was selected using a point-resolved spectroscopy sequence, including the left basal ganglia. The size of the volume was dependent on the size of the basal ganglia, but was mostly 20 ! 20 ! 20 mm3 (anterior-posterior, left-right and feet-head directions). Contact with the periventricular white matter and the lateral ventricle was avoided. For 1H-MRS, 64 signals were averaged (512 data points), and a bandwidth of 1,000 Hz was used. TR/TE was 2,000/31 and 144 ms. The data were processed by applying Lorenz-Gaussian windowing in the time domain (Gaussian broadening 6 Hz, exponential narrowing 4 Hz) for noise reduction and spectral resolution enhancement, followed by zero filling to 4,096 data points. In the present study, only the data for a TE of 31 ms are reported. Curve fitting was performed using MRUI software, including VARPRO/AMARES (EC Human Capital and Mobility/Networks program, Universitat Autònoma, Barcelona, Spain). Peaks were identified by the operator as ·-Glx (3.75 ppm), myoinositol (3.50 ppm), Cho (3.25 ppm), creatine (3.0 ppm), ß- and Á-Glx (2.0–2.5 ppm) or NAA (2.02 ppm) (fig. 1). The linewidth of ß- and Á-Glx was set at 6–9 Hz, depending on the linewidth of NAA. Time domain algorithm ratios of peak areas were calculated and included ·-Glx/ NAA, ·-Glx/Cho, total (t)-Glx/NAA, t-Glx/Cho and NAA/Cho. t-Glx consisted of the area of the ·-, ß- and Á-Glx areas. Statistical Analysis A Mann-Whitney U test was used to compare metabolite ratios of patients of group 1 (mild to moderate HIE, i.e. Sarnat grade I and II)

Biol Neonate 2001;79:254–257

255

with those of group 2 (severe HIE, i.e. grade III), and of patients with a normal versus an abnormal aEEG pattern. Analysis was performed using SPSS version 8.0 for Windows software (SPSS Inc., Chicago, Ill., USA). A p value of less than 0.05 was considered statistically significant.

Results

MRI Abnormalities were detected in all 7 neonates with severe HIE (group 2) and consisted of changes in the basal ganglia and thalami (n = 5) and/or cortical abnormalities (n = 3). Of the 19 neonates in group 1, 11 had a normal scan, whereas 5 had changes in the basal ganglia, 2 had changes in the periventricular white matter and 1 had a hemorrhage in the posterior fossa.

Fig. 1. 1H-MRS on the second day of life in a neonate with grade III

HIE. Peaks identified for curve fitting are ·-Glx, myoinositol, Cho, creatine, ß- and Á-Glx and NAA. Peaks that were not fitted included lactate at 1.33 ppm, propan-1,2-diol at 1.1 ppm and mobile lipids and macromolecules (0.8–1.5 ppm). mI = Myoinositol; Cr = creatine.

1H-MRS Table 2 shows ·-Glx/NAA, ·-Glx/Cho, t-Glx/NAA, t-Glx/Cho and NAA/Cho ratios of the two HIE groups. Differences in t-Glx/ NAA and t-Glx/Cho between the two HIE groups were significant. Table 3 shows these same ratios for the two aEEG groups. Again, differences in t-Glx/NAA and t-Glx/Cho between the two aEEG groups were significant.

Discussion

Table 2. Results of 1H-MRS according to the severity of HIE

·-Glx/NAA ·-Glx/Cho tGlx/NAA t-Glx/Cho NAA/Cho

Group 1

Group 2

p value

0.57 (0.03–1.41) 0.68 (0.04–1.24) 1.21 (0.64–3.25) 1.33 (0.71–2.52) 1.20 (0.49–1.72)

0.73 (0.24–1.04) 1.13 (0.31–1.23) 1.55 (1.10–2.75) 2.14 (1.21–3.55) 1.29 (0.89–1.57)

NS NS 0.035 0.019 NS

Results are shown as median and range in parentheses. Group 1 included patients with Sarnat grade I or II HIE. Group 2 included patients with Sarnat grade III HIE. NS = Not significant.

Table 3. Results of 1H-MRS according to aEEG results

·-Glx/NAA ·-Glx/Cho t-Glx/NAA t-Glx/Cho NAA/Cho

aEEG normal (n = 14)

aEEG abnormal (n = 10)

p value

0.57 (0.03–0.86) 0.67 (0.04–1.13) 1.12 (0.64–1.84) 1.27 (0.71–2.52) 1.23 (0.79–1.62)

0.73 (0.24–1.41) 0.84 (0.31–1.23) 1.51 (1.08–3.25) 1.82 (1.21–3.55) 1.25 (0.49–1.72)

NS NS 0.012 0.010 NS

Results are shown as median and range in parentheses. NS = Not significant.

256

Biol Neonate 2001;79:254–257

The results of the present study confirmed the hypothesis that Glx is elevated in neonates with severe HIE compared to neonates with mild to moderate HIE. Glx is an important neurotransmitter in the cascade leading to neuronal cell damage following hypoxia-ischemia [14]. It activates postsynaptic receptors, resulting in opening of the Ca2+ channels [15, 16]. Calcium plays a crucial role in the formation of oxygen free radicals and neuronal damage [17, 18]. The outcome of neonates with moderate HIE (Sarnat grade II) may be variable, and aEEG has a better predictive value in the first 24 h after birth [13, 19, 20]. We therefore analyzed aEEG and 1H-MRS results. In neonates with severe aEEG abnormalities 24 h after birth, t-Glx/ NAA and t-Glx/Cho ratios were elevated. In a recent study, we were unable to demonstrate elevated Glx levels at the end of the first week of life [21]. In the present study, however, patients were studied earlier. Glx levels were elevated only during the first 2 days of life in the cerebrospinal fluid of newborn infants with HIE [3]. We suggest that both extracellular levels of Glx as detected in the cerebrospinal fluid and intracellular levels of Glx as detected with 1H-MRS are elevated, indicating enhanced Glx metabolism. Pu et al. [6] also reported elevated levels of ·-Glx in neonates with HIE and a poor outcome. The conclusions of their study were based on data obtained at a TE of 135 ms. In our experience, curve fitting of ·-Glx is extremely difficult at this TE, and we considered this TE to be unreliable for ·-Glx measurements. In the present study, differences in ratios of ·-Glx/NAA and ·-Glx/Cho did not reach significance, probably because of the minor contribution of ·-Glx to the t-Glx signal. Although the Glx signal also consists of contributions by glutamine and gamma-aminobutyric acid, Glx is considered the main constituent of the Glx resonance [22]. Previously, many studies have demonstrated decreased NAA/Cho ratios and elevated lactate/NAA ratios in neonates with HIE and a poor neurodevelopmental outcome [7, 11, 23–26]. In the present

Groenendaal/Roelants-van Rijn/ van der Grond/Toet/de Vries

study, NAA/Cho was not decreased in neonates with severe HIE. Firstly, the measurements in the present study may have been too early to show a drop in NAA/Cho, since NAA levels decrease a few days after cerebral hypoxia-ischemia [27]. Secondly, we recently reported that the prediction of outcome by means of decreased NAA/Cho ratios was much better with longer TEs [21]. Transverse relaxation times of brain metabolites may be different the day after hypoxia-ischemia, and this may influence metabolite ratios [28]. However, these effects are not expected to be of major significance at short TEs. In the present study, NAA/Cho ratios at a TE of 144 ms were not significantly different between group 1 and 2, whereas lactate/NAA ratios were significantly higher in group 2 (data not shown). This suggests that the present examinations were performed before NAA decreased. It was impossible to calculate absolute concentrations instead of metabolite ratios, since this would include T1 and T2 measurements in all individual patients. The measuring time was limited in these

very ill neonates to approximately 1 h. Using the water signal as an internal standard as has been used by others was considered inappropriate, as neonates with severe HIE may have different degrees of cerebral edema [11]. We concluded that Glx/NAA and Glx/Cho ratios were elevated during the first days of life in the basal ganglia of neonates with severe HIE.

Acknowledgements

Ariadne Roelants-van Rijn was supported by a donation (Sterproject) of the University Medical Center Utrecht, The Netherlands. The authors thank Marco Nijenhuis and all other technicians of the MR institute for their enthusiastic cooperation.

References 1 Ueda Y, Obrenovitch T, Lok S, Sarna G, Symon L: Efflux of glutamate produced by short ischemia of varied severity in rat striatum. Stroke 1992;23: 253–259. 2 Velasco I, Tapia R, Massieu L: Inhibition of glutamate uptake induces progressive accumulation of extracellular glutamate and neuronal damage in rat cortical cultures. J Neurosci Res 1996;44:551– 561. 3 Hagberg H, Thornberg E, Blennow M, Kjellmer I, Lagercrantz H, Thiringer K, Hamberger A, Sandberg M: Excitatory amino acids in the cerebrospinal fluid of asphyxiated infants: Relationship to hypoxic-ischemic encephalopathy. Acta Paediatr 1993;82:925–929. 4 Mason GF, Pan JW, Ponder SL, Twieg DB, Pohost GM, Hetherington HP: Detection of brain glutamate and glutamine in spectroscopic images at 4.1 T. Magn Reson Med 1994;32:142–145. 5 Prost RW, Mark L, Mewissen M, Li S-J: Detection of glutamate/glutamine resonances by 1H magnetic resonance spectroscopy at 0.5 tesla. Magn Reson Med 1997;37:615–618. 6 Pu Y, Li Q, Zeng C, Gao J, Qi J, Luo D, Mahankali S, Fox P, Gao J: Increased detectability of alpha brain glutamate/glutamine in neonatal hypoxicischemic encephalopathy. AJNR Am J Neuroradiol 2000;21:203–212. 7 Groenendaal F, Veenhoven RH, van der Grond J, Jansen GH, Witkamp TD, de Vries LS: Cerebral lactate and N-acetyl-aspartate/choline ratios in asphyxiated full-term neonates demonstrated in vivo using proton magnetic resonance spectroscopy. Pediatr Res 1994;35:148–151. 8 Hanrahan JD, Cox IJ, Edwards AD, Cowan FM, Sargentoni J, Bell JD, Bryant DJ, Rutherford MA, Azzopardi D: Persistent increases in cerebral lactate concentration after birth asphyxia. Pediatr Res 1998;44:304–311. 9 Peden CJ, Cowan FM, Bryant DJ, Sargentoni J, Cox JJ, Menon DK, Gadian DG, Bell JD, Dubowitz LM: Proton MR spectroscopy of the brain in infants. J Comput Assist Tomogr 1990;14:886– 894.

Brain Glutamate in Asphyxiated Neonates

10 Penrice J, Lorek A, Cady EB, Amess PN, Wylezinska M, Cooper CE, D’Souza PD, Brown GC, Kirkbride V, Edwards AD, Wyatt JS, Reynolds EOR: Proton magnetic resonance spectroscopy of the brain during acute hypoxia-ischemia and delayed cerebral energy failure in the newborn piglet. Pediatr Res 1997;41:795–802. 11 Leth H, Toft PB, Peitersen B, Lou HC, Henriksen O: Use of brain lactate levels to predict outcome after perinatal asphyxia. Acta Paediatr 1996;85: 859–864. 12 Saunders D, Howe F, Van den Boogaart A, Griffiths J, Brown MM: Discrimination of metabolite from lipid and macromolecule resonances in cerebral infarction in humans using short echo proton spectroscopy. J Magn Reson Imaging 1997;7: 1116–1121. 13 Toet MC, Hellström-Westas L, Groenendaal F, Eken P, de Vries LS: Amplitude integrated EEG at 3 and 6 hours after birth in full term neonates with hypoxic-ischaemic encephalopathy. Arch Dis Child Fetal Neonatal Ed 1999;81:F19–F23. 14 Hagberg H, Lehmann A, Sandberg M, Nystrom B, Jacobson I, Hamberger A: Ischemia-induced shift of inhibitory and excitatory amino acids from intra- to extracellular compartments. J Cereb Blood Flow Metab 1985;5:413–419. 15 Choi DW: Cerebral hypoxia: Some new approaches and unanswered questions. J Neurosci 1990;10:2493–2501. 16 Siesjo BK: Cell damage in the brain: A speculative synthesis. J Cereb Blood Flow Metab 1981;1:155– 185. 17 Brooks KJ, Kauppinen RA: Calcium-mediated damage following hypoxia in cerebral cortex ex vivo studied by NMR spectroscopy. Evidence for direct involvement of voltage-gated Ca2+-channels. Neurochem Int 1993;23:441–450. 18 Budd S: Mechanisms of neuronal damage in brain hypoxia/ischemia: Focus on the role of mitochondrial calcium accumulation. Stroke 1998;29:1048– 1057. 19 Levene MI, Sands C, Grindulis H, Moore JR: Comparison of two methods of predicting outcome in perinatal asphyxia. Lancet 1986;i:67–69.

20 Mellits ED, Holden KR, Freeman JM: Neonatal seizures. II. A multivariate analysis of factors associated with outcome. Pediatrics 1982;70:177–185. 21 Roelants-van Rijn AM, van der Grond J, de Vries LS, Groenendaal F: Value of 1H-MRS using different echo times in neonates with cerebral hypoxiaischemia. Pediatr Res, in press. 22 Ross BD: Biochemical considerations in 1H spectroscopy. Glutamate and glutamine; myo-inositol and related metabolites. NMR Biomed 1991;4:59– 63. 23 Peden CJ, Rutherford MA, Sargentoni J, Cox JJ, Bryant DJ, Dubowitz LM: Proton spectroscopy of the neonatal brain following hypoxic-ischemic injury. Dev Med Child Neurol 1993;35:502–510. 24 Hanrahan JD, Cox IJ, Edwards AD, Cowan FM, Sargentoni J, Bell JD, Bryant DJ, Rutherford MA, Azzopardi D: Persistent increases in cerebral lactate concentration after birth asphyxia. Pediatr Res 1998;44:304–311. 25 Robertson NJ, Cox IJ, Cowan FM, Counsell SJ, Azzopardi D, Edwards AD: Cerebral intracellular lactic alkalosis persisting months after neonatal encephalopathy measured by magnetic resonance spectroscopy. Pediatr Res 1999;46:287–296. 26 Barkovich AJ, Baranski K, Vigneron D, Partridge JC, Hallam DK, Hajnal BL, Ferriero DM: Proton MR spectroscopy for the evaluation of brain injury in asphyxiated, term neonates. AJNR Am J Neuroradiol 1999;20:1399–1405. 27 Gideon P, Henriksen O, Sperling B, Christiansen P, Olsen TS, Jørgensen HS, Arlien-Søborg P: Early time course of N-acetylaspartate, creatine and phosphocreatine, and compounds containing choline in the brain after acute stroke. Stroke 1992;23: 1566–1572. 28 Cady EB, Lorek A, Penrice J, Wylezinska M, Cooper CE, Brown GC, Owen-Reece H, Kirkbride V, Wyatt JS, Osmund E, Reynolds EO: Brainmetabolite transverse relaxation times in magnetic resonance spectroscopy increase as adenosine triphosphate depletes during secondary energy failure following acute hypoxia-ischaemia in the newborn piglet. Neurosci Lett 1994;182:201–204. 29 Sarnat HB, Sarnat MS: Neonatal encephalopathy following fetal distress. A clinical and electroencephalographic study. Arch Neurol 1976;33:696– 705.

Biol Neonate 2001;79:254–257

257

Biol Neonate 2001;79:258–260

Resuscitation of the Asphyxic Newborn Infant: New Insight Leads to New Therapeutic Possibilities Ola Didrik Saugstad Department of Pediatric Research, The National Hospital, University of Oslo, Oslo, Norway

Key Words Birth asphyxia W Brain injury W Newborn resuscitation W Oxidative stress W Reoxygenation injury

Abstract The basic mechanisms leading to cell death in birth asphyxia are becoming better known. Some of these are excitotoxicity, inflammation and oxidative stress. In the so-called therapeutic window – between the primary and secondary energy failure – modulation of these processes may be beneficial, reducing apoptosis and perhaps necrosis. In order to reduce oxidative stress, reoxygenation with low oxygen concentrations, even as low as room air, might be beneficial. Increased oxidative stress might have long-term effects on brain growth and development and there is evidence indicating that exposure to 100% oxygen after birth for only a few minutes might have long-term effects. New guidelines for newborn resuscitation have recently been published but more research is needed in this field, especially regarding resuscitation of preterm infants, where few data exist. Copyright © 2001 S. Karger AG, Basel

Introduction

Out of the approximately 130 million annual births worldwide, it has been calculated that 4 million suffer from birth asphyxia, and of these, 1 million die and a similar number develop some sequelae [1]. Whether or not these figures are correct, they nevertheless indicate the huge extent of the problem of birth asphyxia. The incidence of birth asphyxia is higher in developing than in so-called developed countries. Still, in the latter, 2–6 births per thousand develop hypoxic ischemic encephalopathy [2], representing between 8,000 and 25,000 infants in the EU area. Of these, many develop severe injuries.

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0258$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

The figures given above must, however, be interpreted with caution. First of all, because the definition of asphyxia is still under debate, and secondly, since the cause of neonatal encephalopathy seems to be uncertain. One definition of birth asphyxia in use is based on the finding of three of the four following criteria: (1) pH in umbilical arterial cord blood ! 7.00; (2) Apgar score ! 4 after more than 5 min; (3) multiple organ damage, and (4) hypoxic ischemic encephalopathy [3]. It therefore has become clear that the diagnosis of birth asphyxia can only be made retrospectively [4]; it is the sequence of symptoms and signs and how the brain and other organs react over time that indicate the diagnosis. Therefore, neither the diagnosis nor prognosis can be decided until some time after birth. To make the matter even more complicated, it was recently suggested that only 20% of babies developing neonatal encephalopathy suffered from birth asphyxia [5]. This illustrates some of the complications in this field. It is difficult to obtain robust figures regarding the incidence of asphyxia because we really do not fully know what we are discussing when the topic is birth asphyxia.

Primary and Secondary Energy Depletion

It is now well established that energy depletion occurs in two phases [6, 7]. The primary energy failure takes place in conjunction with the insult. The secondary energy phase typically occurs between 8 and 48 h after the insult. In between, the energy status often normalizes. During the primary phase, cell loss is related directly to hypoxia with exhaustion of energy metabolism. Three mechanisms have been identified as being involved in cell death in this phase: (1) depolarization due to hypoxia, which causes an influx of sodium and passive chloride entry with entry of water leading to cell swelling and, if severe enough, lysis and cell death; (2) intracellular accumulation of calcium because of excessive calcium entry due to activation of Na+/ K+ ion channels and NMDA receptor channel activation by gluta-

Ola Didrik Saugstad Department of Pediatric Research, The National Hospital University of Oslo, N–0027 Oslo (Norway) Tel. +47 23 07 27 90/94, Fax +47 23 07 27 80 E-Mail [email protected], Homepage: http://www.uio.no/olasF/

mate, and (3) damage of cell membranes by oxygen free radicals in the immediate reoxygenation/reperfusion phase. Secondary or delayed neuronal death is triggered by events occurring in the primary phase and is associated with hyperexcitability and cytotoxic edema from about 6 to 100 h after the insult. The mechanisms involved in injury in this phase are (1) excitotoxicity, (2) apoptosis and (3) the cytotoxic action of activated microglia [7, 8]. In this period, a secondary hyperemia is found which is considered neuroprotective [7].

Therapeutic Strategies

The degree of secondary energy failure strongly predicts neurodevelopmental outcome. This illustrates why asphyxia is a retrospective diagnosis and it also indicates that asphyxic injury may be prevented by intervention in the time between the primary and secondary energy failure. Treatment strategies therefore concentrate on preventing or modulating (1) reoxygenation injury, (2) excitotoxicity and (3) apoptosis. Figure 1 shows a simplified outline of some important factors regulating apoptosis. It is clear that hypoxia/reoxygenation and increased oxidative stress may induce apoptosis both directly and indirectly. Bcl-2, an important protective agent of apoptosis, is also a powerful antioxidant. During hypoxia in neonatal rat cardiomyocytes, the expression of bcl-2 and p53 is upregulated. The latter is a powerful apoptotic inducer. During reoxygenation, bcl-2 is downregulated and p53 is further upregulated [9]. Depolarization of the mitochondrial membrane due to hypoxia/reoxygenation releases cytochrome C, and the stage might thus be set for caspase activation which eventually leads to execution of the death program. These processes may be blocked by antioxidants at several sites. In experimental studies, caspase inhibition has been shown to induce a smaller posthypoxic brain infarct [10].

Reducing Oxygen Free Radical Production

Instead of blocking the actions of oxygen free radicals, another approach might be to reduce their production. We have tested this out by reducing the oxygen concentration from 100 to 21% during reoxygenation following hypoxia. In fact, in numerous animal studies performed by our group as well as others, it has been demonstrated that room air resuscitation seems to be as efficient as the use of 100% oxygen [for a review, see ref. 11]. For instance, the free radical production in the lungs of newborn piglets subjected to apnea and reoxygenation is significantly higher when 100% oxygen is administrated compared with room air [12]. In a pilot study including 84 children, we were able to show that room air is as efficient as 100% oxygen for resuscitation regarding normalization of heart rate, Apgar scores and acid/base variables [13]. In order to explore this issue in more depth, we organized the Resair 2 study, in which 288 and 321 infants, respectively, were resuscitated with room air or 100% oxygen. Neonatal mortality was 5% lower in the room air group, with the odds ratio reduced 2/3 in favor of room air; however, this was only borderline significant. In the oxygen group, both time to the first cry and breath were significantly delayed by approximately 0.5 min, indicating that oxygen depresses ventilation [14]. In a recent study by Vento et al. [15], it was confirmed that room air normalizes the respi-

Resuscitation of the Asphyxic Newborn Infant

Fig. 1. Reactive oxygen species (ROS) may induce apoptosis directly

or indirectly by depolarization of the mitochondrial membrane. Antioxidants therefore might modulate apoptosis at various steps in this process.

ratory pattern earlier than 100% oxygen. Further, these authors found an increased oxidative stress 1 month after birth in infants reoxygenated with 100% oxygen, while it was normalized in room air-exposed babies. This important finding indicates that the resuscitation procedure might influence long-term growth and development. It is, for instance, clear that a number of ligands, such as tumor necrosis factor-·, interleukin-1ß, platelet-derived factor and others, may use reactive oxygen species. Further, oxygen radicals act in signal transduction by stimulating calcium signaling and activating transcription factors, induce apoptosis, induce bone resorption, inhibit viral replication, and induce proinflammatory cytokines as well as adhesion molecules. An increased oxidative stress in the growing and developing brain therefore might have serious unwanted effects. It is indeed dramatic that even a brief exposure to oxygen – in fact only a few minutes on average – might induce such long-lasting effects. These findings therefore must prompt us to initiate new studies investigating the effects of room air versus oxygen resuscitation.

Biol Neonate 2001;79:258–260

259

New Guidelines for Resuscitation of Newborns

In the last 2–3 years, new guidelines have been published regarding resuscitation of the newly born [16–18]. The World Health Organisation now states that oxygen supplementation is not necessary for so-called basic resuscitation, but ‘when the newborn’s colour does not improve despite effective ventilation, oxygen should be given if available’ [16]. The new guidelines issued by the American Heart Association (AHA) and American Academy of Pediatrics (AAP) [18] states that ‘if assisted ventilation is required, 100% oxygen should be delivered by positive pressure ventilation’. This is more or less in line with the old guidelines from 1992. The following statement is, however, new: ‘If supplemental oxygen is not available, resuscitation of the newly born infant should be initiated with positive pressure ventilation and room air.’ The careful but perhaps conservative approach of the AHA/AAP to this issue is justified by the fact that not enough data regarding room air resuscitation of the newly born infant exist.

Resuscitation of Preterm Infants

There are no specific guidelines regarding resuscitation of preterm infants. The reason for this is that no or very few data exist. This is paradoxical, since it is known that the premature infant has a greater need than the term infant for resuscitation after birth [19]. The World Health Organisation states that the same principles should be applied to the preterm as the term infant [16]. The International Liaison Committee on Resuscitation has not drawn up any specific recommendations for the preterm infant [17], while the AHA/AAP [18] states that rapid boluses of volume expanders or hyperosmolar solu-

tions should be avoided. In the near future, it will therefore not only be highly important but even crucial to collect scientific data regarding resuscitation of preterm infants, especially emphasizing their special needs regarding heat loss and immaturity per se. From a theoretical point of view, the preterm infant might be even more vulnerable to high oxygen concentrations than the term infant. Thus, perhaps one should be even more careful in administrating excessive concentrations of oxygen to the preterm infant.

Conclusion

New insight into the pathophysiology of birth asphyxia has given us the opportunity to prevent permanent injury by modulating basic molecular processes, some of which have been mentioned in this article. A problem is, however, that a clear definition and understanding of the concept of birth asphyxia still do not exist. Nonetheless, I believe that combinations of therapeutic regimes, for instance blocking excitotoxicity and oxygen radicals, modulating inflammation as well as apoptosis, will represent future therapeutic approaches. A simple and cheap way to modulate many of the processes leading to permanent injury might be by resuscitating the asphyxiated newborn more efficiently. This might be done by reoxygenating them with lower concentrations than 100% oxygen, perhaps as low as room air. In the future, there will probably not be written general guidelines for newborn resuscitation since each child has to be treated individually, tailoring the resuscitation to each subject’s special needs. This might, for instance, be achieved by continuously measuring oxygen saturation and blood flow perfusion during resuscitation. However, the optimal levels still have to be defined for the term and especially for the preterm infant.

References 1 World Health Organisation: Child Health and Development: Health of the Newborn. Geneva, World Health Organisation, 1991. 2 Levene ML, Kornberg J, Williams THC: The incidence and severity of post-asphyxial encephalopathy in full-term infants. Early Hum Dev 1985;11: 21–26. 3 Use and abuse of the Apgar score. Committee on Fetus and Newborn, American Academy of Pediatrics, and Committee on Obstetric Practice, American College of Obstetricians and Gynecologists. Pediatrics 1996;98:141–142. 4 Nelson KB, Emery ES: Birth asphyxia and the neonatal brain: What do we know and when do we know it? Clin Perinatol 1993;20:327–344. 5 Adamson SJ, Alessandri LM, Badawi N, Burton PR, Pemberton PJ, Stanley F: Predictors of neonatal encephalopathy in full term infants. BMJ 1995; 311:598–602. 6 Lorek A, Takei Y, Cady EB, Wyatt JS, Penrice J, Edwards AD, Peebles D, Wylezinska M, OwenReece H, Kirkbride V, et al: Delayed (‘secondary’) cerebral energy failure after acute hypoxia-ischemia in the newborn piglet: Continuous 48-hour studies by phosphorous magnetic resonance spectroscopy. Pediatr Res 1994;36:699–706.

260

7 Inder TE, Volpe JJ: Mechanisms of perinatal brain injury. Semin Neonatol 2000;5:3–16. 8 Fellman V, Ravio KO: Reperfusion injury as a mechanism of brain damage after perinatal asphyxia. Pediatr Res 1997;41:599–606. 9 Shen J-G, Quo X-S, Jiang B, Li M, Xin W-J, Zhao B-L: Chinonin, a novel drug against cardiomyocyte apoptosis induced by hypoxia and reoxygenation. Biochim Biophys Acta 2000;1500:217–226. 10 Johnston MV, Trescher WH, Ishida A, Nakajima W: Novel treatments after experimental brain injury. Semin Neonatol 2000;5:75–86. 11 Saugstad OD: Resuscitation with room-air or oxygen supplementation. Clin Perinatol 1998;25:741– 756. 12 Kondo M, Itoh S, Isobe K, Kondo M, Kunikata T, Imai T, Onishi S: Chemiluminescence because of the production of reactive oxygen species in the lungs of newborn piglets during resuscitation periods after asphyxiation load. Pediatr Res 2000;47: 524–527. 13 Ramji S, Ahuja S, Thirupuram S, Rootwelt T, Rooth G, Saugstad OD: Resuscitation of asphyxic newborn infants with room air or 100% oxygen. Pediatr Res 1993;34:809–812. 14 Saugstad OD, Rootwelt T, Aalen O: Resuscitation of asphyxiated newborn infants with room air or oxygen: An international controlled trial: The Resair 2 study. Pediatrics 1998;102:e1.

Biol Neonate 2001;79:258–260

15 Vento M, Garcia-Sala F, Vina J, Asensi MA, Sastre J: The use of room air ventilation enhances the establishment of a sustained spontaneous respiratory pattern after perinatal asphyxia. Pediatr Res 2000;47:438A. 16 World Health Organisation: Basic Newborn Resuscitation: A Practical Guide. Geneva, World Health Organisation, 1998. 17 Kattwinkel J, Niermeyer S, Nadkarni V, Tibballs J, Phillips B, Zideman D, Van Reempts P, Osmond M: ILCOR advisory statement: Resuscitation of the newly born infant. An advisory statement from the pediatric working group of the International Liaison Committee on Resuscitation. Circulation 1999;99:1927–1938. 18 Niermeyer S, Kattwinkel J, Van Reempts P, Nadkarni V, Phillips B, Zideman D, et al: International guidelines for neonatal resuscitation: An excerpt from guidelines 2000 for cardiopulmonary resuscitation and emergency cardiovascular care: International consensus on science. Pediatrics 2000;106: e29. 19 Finer NN, Horbar JD, Carpenter JH: Cardiopulmonary resuscitation in the very low birth weight infant: The Vermont Oxford Network experience. Pediatrics 1999;104:428–434.

Saugstad

Biol Neonate 2001;79:261–267

Six Years of Experience with the Use of Room Air for the Resuscitation of Asphyxiated Newly Born Term Infants Max Vento a Miguel Asensi b Juan Sastre b Fernando Garcı´a-Sala a José Viña b a Departamento

de Pediatrı´a y Neonatologı´a, Hospital Virgen del Consuelo, Valencia, y Facultad de Medicina, Universidad de Alicante, Alicante, b Departamento de Fisiologı´a, Facultad de Medicina, Universidad de Valencia, Valencia, España

Key Words Asphyxia W Resuscitation W Room air W Newly born infant W Oxidative stress W Glutathione

Abstract In the last 6 years, 830 asphyxiated newly born term infants have been resuscitated with room air (RAR; n = 304) or 100% oxygen (OxR; n = 526) in our hospital. We have studied the time to onset of a regular respiratory pattern, morbidity and mortality, blood gases, reduced glutathione (GSH) and oxidised glutathione (GSSG) and antioxidant enzymes in these infants. No significant differences in the effectiveness of either gas sources or in the final outcome have been found. The RAR group required a shorter time of positive pressure ventilation to attain a spontaneous pattern of respiration. The OxR group showed hyperoxaemia during resuscitation, which was positively correlated with increased GSSG concentrations. Significant oxidative stress was found in the OxR group at 28 days of postnatal life when compared with normal control infants and the RAR group. Oxygen concentrations used during the resuscitation of newly born infants should be strictly monitored. Copyright © 2001 S. Karger AG, Basel

Introduction

Perinatal asphyxia accounts for 19% of all perinatal deaths, as well as for severe long-term sequelae such as cerebral palsy, seizures or delay in development [1, 2]. The foetal to neonatal transition is characterised by a complex series of physiologic events which are not always adequately accomplished. Thus, approximately 5–10% of newly born infants require

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0261$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

some type of active resuscitation at birth [3], and approximately 1– 10% are reported to require assisted ventilation [4]. The most recent updated guidelines published as a result of an international consensus of experts propose the use of 100% oxygen in the resuscitation of asphyctic newly born infants in order to achieve normoxia. However, if supplemental oxygen is not available, room air ventilation is considered a valid alternative [5]. However, there has been an increasing body of evidence in the last years that casts doubts on the use of pure oxygen as the preferable gas source to achieve normoxia after perinatal asphyxia. Criticism of the use of pure oxygen in the resuscitation of asphyxiated newly born infants is essentially based upon: (1) the experimental evidence proportioned by the more in depth knowledge of the profound mechanisms involved in the ischaemia-reperfusion phenomenon occurring during perinatal asphyxia [6]; (2) the deleterious consequences of hyperoxia [7], and (3) the theory of energy failure due to mitochondrial damage caused by free radicals [8]. Moreover, the results of several prospective clinical trials that have compared the effectiveness of the use of 100% oxygen versus room air in the resuscitation of the asphyctic newly born infant [9–11] have added substantial evidence in support of the suitability of the latter in order to attain a rapid and sustained recovery. In 1994, we joined the Resair 2 Trial [10], and since then, we have increasingly used room air for the resuscitation of moderately to severely asphyxiated newly born infants. We have reported our clinical and experimental results on different occasions [11–14]. Thus, the present report is a comprehensive summary of our experience during 6 consecutive years (1994–1999) in the use of room air as an alternative to pure oxygen in the resuscitation of asphyxiated newly born term infants. On most occasions, the gas source used for resuscitation was not revealed to those undertaking resuscitation, in order to eliminate this bias from the study. However, this was not always the

Max Vento, MD, PhD Jefe de Servicio de Pediatrı´a y Neonatologı´a Hospital Virgen del Consuelo, Callosa de Ensarria´, 12 E–46007 Valencia (Spain) Tel. +34 96 317 4000, Fax +34 96 317 7870, E-Mail [email protected]

Table 1. Total deliveries, type of asphyxia (moderate or severe), type of resuscitation (RAR or OxR) and mortality distributed per year, from 1994 to 1999

Year

1994 1995 1996 1997 1998 1999 Total

Total deliveries 2,540 2,388 2,699 2,330 2,616 2,459 15,002

Severely RAR asphyxiated 4 6 5 8 3 4 30

2 2 3 4 2 3 16

Deceased OxR

Deceased Moderately RAR asphyxiated

Deceased OxR

Deceased

0 0 0 1 0 0 1

1 2 1 1 0 1 6

0 0 0 0 0 0 0

0 1 0 1 0 1 3

2 4 2 4 1 1 14

154 133 141 135 110 127 800

0 0 55 74 83 76 288

154 133 86 61 27 51 512

Severe asphyxia: Apgar score at one minute ! 3; arterial pH at delivery ! 7.0; bradycardia at 1 min ! 60 bpm; hypotonia non-responsive to external stimuli; apnoea; Apgar score at 5 min ! 5. Moderate asphyxia: Apgar score at 1 min of 3–5; arterial pH at delivery ! 7.15; bradycardia at 1 min ! 80 bpm; hypotonia non-responsive to external stimuli; apnoea.

case, especially in the most recent years, when resuscitation with room air has increasingly become a routine procedure in our neonatal unit.

Subjects and Methods

Study Population During a period of time of 6 years (1994–1999), a total of 15,002 term deliveries have been attended in our hospital. Table 1 illustrates the total and the yearly distribution of the number of deliveries, the number and type (severe or moderate) of cases of asphyxia, the type of resuscitation employed and total mortality. Gestations were controlled in our obstetric outpatient clinic and deliveries carried out in the obstetric ward of our hospital. The attending obstetricians were informed of the ongoing clinical trial and thus of the possibility of the babies being resuscitated either with room air (RAR) or 100% oxygen (OxR). During gestation and at admission, parents were informed of the possibility of having their babies randomly resuscitated by RAR or OxR and of our previous published experience. Informed consent was required in all cases for participation in the different studies. From a total of 15,002 term newborn infants, 830 were enrolled in the studies. Eligible neonates showed clinical and biochemical signs of moderate to severe asphyxia as described in table 1. Moderate asphyxia, affecting 800 newly born infants, represented 5.3% of all deliveries. A total of 30 infants presented with severe asphyxia and represented 0.2% of all deliveries. Gas sources were located at the wall and connected to an oxygen blender that was not visible for the resuscitation team. The nurse in charge could switch from 21 to 100% oxygen following an aleatoric number in a sequential manner, which corresponded to room air or 100% oxygen in each asphyxiated baby. Nurses provided the neonatologists with the bag and mask for resuscitation, which were connected to the corresponding gas mixture (room air or 100% oxygen). Thus, the members of the resuscitating team were unaware of the type of gas they were using on each different infant when we did the blinded studies. In addition, the maximal pressure ventilation, gas

262

Biol Neonate 2001;79:261–267

flow and the ventilatory rate were limited (as described in Clinical Proceedings below). When asphyxiated infants were not blindly resuscitated, the resuscitation team requested a precise type of gas mixture. The Ethical Committee of the Hospital Virgen del Consuelo (Valencia, Spain) approved the study protocol. The parents’ written consent was obtained in each case when the mother was admitted to the obstetric ward before delivery. Clinical Proceedings Immediately after birth, all infants were put under a radiant heating unit (Baby Resuscitator Draeger Inc.) and resuscitated following the usual procedures of our nursery [15]. A specific gas mixture (room air or 100% oxygen) was assigned to each infant at random, or by requirement of the resuscitation team when the baby was not entering a blinded trial. In every case, the gas mixture employed was recorded. In order to avoid differences between infants, the maximal gas flow was limited to 6 litres/min, the inspiratory pressure was limited to 40 mbar and the ventilation frequency was kept below 30 rpm. Once the initial resuscitating procedures (at approximately 1 min of postnatal life) had stabilised the patient, nurses placed the probes to monitor the infants’ clinical parameters (ECG, temperature, respiration and pulse oximetry). Neonatal nurses determined the Apgar score and clinical parameters at 1-min intervals beginning immediately after cord clamping. The clinical parameters recorded usually are: (1) occurrence of the first cry, defined as the first audible cry spontaneously emitted by the infant, and (2) onset of a regular respiratory pattern, defined as the establishment of a spontaneous and sustained respiratory pattern of efficacious respiratory movements allowing the maintenance of adequate clinical parameters (heart and respiratory rate) and haemoglobin saturation above 90%, thus needing no additional intervention from the resuscitation team. Blood samples were drawn from the umbilical artery prior to detachment from the placenta, and afterward from the radial artery, peripheral vein or arterialised capillary immediately after finishing resuscitation, and/or at 15 min, 24 h, 72 h and even at 4 weeks of postnatal age, depending on the individual study.

Vento/Asensi/Sastre/Garcı´a-Sala/Viña

Neurological follow-up, including clinical evaluation, ultrasound and EEG, was performed at 1 month of postnatal life in all cases. In severely asphyxiated infants, sequential follow-up studies are still being performed in the outpatient clinic and have included CAT scan and MRI when needed. Analytical Assays Routine biochemical determinations were performed in the clinical laboratory of the hospital following the standard procedures. Reduced glutathione (GSH) and oxidised glutathione (GSSG) were determined in whole blood as we have previously described [16, 17]. Glutathione peroxidase (GPx), superoxide dismutase (SOD) and catalase activities were determined in erythrocytes as described by Flohé and Güzzler [18], Flohé and Otting [19] and Aebi [20], respectively. Blood was collected into heparinised tubes and then immediately centrifuged for 10 min at 500 g and 4 ° C. Plasma was removed and erythrocytes were washed twice with 0.9% NaCl. The supernatant was aspirated and the cell pellet was hemolyzed with distilled water. GPx, SOD and catalase activities were assayed in the hemolysate. Enzymic activities were expressed per gram of haemoglobin content in the hemolysate. Fig. 1. The number of asphyxiated newly born infants resuscitated

Statistics Statistical analysis was performed using nonparametric statistics, since the data obtained did not have a normal distribution; thus, we used the Mann-Whitney test for non-paired samples and the KruskalWallis test for more than two non-paired (independent) samples. Data were analysed using the GB-STAT computer program (Dynamics Microsystems Inc.).

with the two types of resuscitation (RAR or OxR) during the period between 1994 and 1999. Simple regression lines, including the r values for tendencies, are depicted in the graph.

Results

Epidemiological Data Table 1 shows the number of deliveries distributed according to the year, type of asphyxia and type of resuscitation (RAR or OxR) during the 6 years reviewed. The number of deliveries of severely asphyxiated infants and moderately asphyxiated infants remained quite stable throughout this period of time. However, as depicted in figure 1, there has been an increased tendency towards the use of room air in the last 3 years, outweighing the number of infants resuscitated with 100% oxygen. Table 1 shows that no significant differences in the mortality of moderately and severely asphyxiated infants were found in relation to the type of resuscitation. Additionally, as seen in figure 2, no significant differences were found for 1-min and 5-min Apgar scores between the RAR and the OxR groups. Onset of Spontaneous Respiration and Blood Gases In a subset of infants (OxR, n = 28; RAR, n = 28), blood gases at delivery in the umbilical artery, at the end of the resuscitation procedure and at approximately 15–16 min of postnatal life were determined [14]. Table 2 shows the mean and standard deviation for blood gases in the control, RAR and OxR groups at delivery, at the end of the resuscitation procedure (5–6 min of postnatal life) and at 15–16 min of postnatal life. Values for arterial partial pressure of oxygen (PaO2) in the umbilical artery were significantly lower in both asphyxiated groups than in the control group. However, after 5–6 min of positive pressure ventilation, the RAR group showed a signifi-

Experience with the Use of Room Air in Perinatal Resuscitation

Fig. 2. One- and 5-min Apgar scores (median B 5–95 centiles) of

asphyxiated newly born infants resuscitated by RAR or OxR. * p ! 0.05 vs. controls, ** p ! 0.001 vs. controls. 1 = RAR severely asphyxiated; 2 = OxR severely asphyxiated; 3 = RAR moderately asphyxiated; 4 = OxR moderately asphyxiated.

cantly lower PaO2 than the OxR group (RAR vs. OxR: 88 B 9 vs. 157 B 27 mm Hg; p ! 0.001). This difference was still significant at 15– 16 min of postnatal life, once the resuscitation procedure had already ceased for approximately 10 min (RAR vs. OxR: 92 B 12 vs. 115 B 19 mm Hg; p ! 0.001). No significant differences were found for pH and PaCO2.

Biol Neonate 2001;79:261–267

263

Fig. 3. Time needed for the onset of a sustained respiratory pattern

Fig. 4. GSH/GSSG ratio in asphyxiated newly born infants resusci-

in moderately asphyxiated neonates resuscitated either by RAR or OxR compared to the non-asphyxiated control group. ** p ! 0.01 vs. controls, # p ! 0.05 vs. the RAR group.

tated by RAR or OxR. ** p ! 0.01 vs. controls, RAR group.

##

p ! 0.01 vs. the

Table 2. Blood gases determined intrapartum in the umbilical artery and at different postnatal times in asphyxiated newly born infants resuscitated by RAR or OxR

Postnatal life

Umbilical artery 5–6 min 15–16 min

1

PaO2, mm Hg

pH1

PaCO2, mm Hg

controls (n = 65)

OxR (n = 69)

RAR (n = 65)

87B13 n.d. n.d.

56B15** 57B18** 157B27 88B9## 114B19 92B12##

controls (n = 65)

OxR (n = 69)

RAR (n = 65)

controls (n = 65)

OxR (n = 69)

RAR (n = 65)

45B3.2 n.d. n.d.

55B6.7 48B6.7 43B5.5

58B2.5 52B3.7 47B7.1

7.34B0.04 n.d. n.d.

7.11B0.04* 7.27B0.12 7.37B0.07

7.09B0.04* 7.31B0.09 7.42B0.03

Values are expressed as mean B standard deviation. n.d. = Not determined. * p ! 0.05 vs. control group, ** p ! 0.001 vs. control group, ## p ! 0.001 vs. OxR group. pH is expressed as [H+] to be averaged.

Table 3. Effect of resuscitation of asphyctic newly born infants by RAR or OxR on blood glutathione status and antioxidant enzymes in

erythrocytes Umbilical artery

GSH, ÌM GSSG, ÌM SOD, U/g Hb Catalase, K/mg Hb GPx, U/g Hb

72 h postnatally

28 days postnally

controls (n = 26)

RAR (n = 19)

OxR (n = 21)

controls

RAR

OxR

controls

RAR

OxR

1,005 (157) 20.6 (5.3) 1.5 (0.3) 182 (28) 47 (11)

970 (160) 50.2** (6.9) 2.5* (0.7) 226* (44) 46 (11)

1,290 (160) 59.2** (5.4) 2.7* (0.9) 245* (61) 46 (9)

970 (160) 29.4 (10.6) 1.8 (0.3) 211 (40) 49 (14)

788 (120) 63.9* (13.6) 2.3* (0.7) 277* (55) 64* (13)

945 (146) 94.2*# (9.4) 3.7*# (0.6) 384**# (91) 67* (13)

927 (158) 19.2 (10.7) 1.6 (0.4) 165 (30) 50 (10)

999 (137) 18.8 (9.4) 1.9 (0.8) 149 (31) 45 (5)

836 (173) 56.7**## (13.0) 2.7**# (0.9) 294**# (83) 54 (10)

Values are expressed as mean and SD in parentheses. * p ! 0.05 vs. control group, ** p ! 0.01 vs. control group, # p ! 0.05 vs. RAR group, ## p ! 0.01 vs. RAR group.

264

Biol Neonate 2001;79:261–267

Vento/Asensi/Sastre/Garcı´a-Sala/Viña

Fig. 5. Polynomial regression correlating PaO2 values with the GSSG concentration in erythrocytes in asphyxiated newly born infants resuscitated by RAR or OxR. The GSSG concentration correlates positively with PaO2 when the latter values are above physiological levels (6100 mm Hg), independent of the type of resuscitation (RAR or OxR) employed.

Figure 3 [11] depicts the time required for the onset of a regular respiratory pattern in a group of 28 asphyxiated newborn infants resuscitated by RAR and 28 newborn infants resuscitated by OxR. As shown, the non-asphyxiated control group established a regular respiratory pattern significantly faster than the RAR and OxR neonates. Furthermore, differences between both experimental groups, as shown in figure 3, are also statistically significant. Thus, the RAR group needed a significantly shorter time of positive pressure ventilation in order to achieve a sustained respiratory pattern than did the OxR group. In a subsequent study done with a larger group of infants [14], we confirmed this finding. The onset of a regular pattern of respiration was significantly shorter in the RAR group (n = 65) as compared with the OxR group (n = 69) (RAR vs. OxR: 4.6 B 1.5 vs. 6.6 B 2.1 min; p ! 0.005). Glutathione Metabolism and Antioxidant Enzymes We measured the GSH/GSSG ratio, as indicative of oxidative stress, from the immediate postnatal period to the 28th day of postnatal life [11] in a group of asphyxiated newly born infants resuscitated with room air or 100% oxygen. Table 3 shows that the GSSG concentrations were significantly higher in the umbilical artery in all the asphyxiated infants before resuscitation procedures were initiated when compared to normal control infants. After 72 h of postnatal life, GSSG was still significantly higher in the RAR and OxR groups as compared with the non-asphyxiated control group. Furthermore, GSSG was also significantly greater (p ! 0.01) in the OxR group (94.2 ÌM/g Hb) than in the RAR group (63.9 ÌM/g Hb).

Experience with the Use of Room Air in Perinatal Resuscitation

Table 3 shows that, after 28 days of postnatal life, GSSG in the RAR group and in the non-asphyxiated control group did not show significant differences. However, table 3 also shows that the GSSG concentration was much higher (p ! 0.01) in the OxR group (56.7 ÌM/g Hb) than in non-asphyxiated controls (19.2 ÌM/g Hb) and in the RAR group (18.8 ÌM/g Hb). Figure 4 depicts the evolution of the GSH/GSSG ratio throughout the study [11]. It shows the significant differences between the experimental groups and the non-asphyxiated control group immediately before cord clamping (day 0) and at 3 days of postnatal life (day 3). The GSH/GSSG ratio of the RAR group at 28 days of postnatal life shows no differences compared with the non-asphyxiated control group. However, figure 4 also shows that the GSH/GSSG ratio of the OxR group is significantly different, not only from the non-asphyxiated control group, but also from the RAR group, revealing a persistent oxidant stress in this group (fig. 4). In different studies, it has caught our attention that there was a protracted oxidative stress in babies ventilated with 100% oxygen and that, coincidentally, their GSSG concentrations were significantly higher than in the control groups. Thus, we undertook a prospective determination of serial PaO2 values in asphyxiated newly born infants during the resuscitation procedure. OxR patients showed higher values of PaO2, often above physiological values, after 5–6 min of positive pressure ventilation. These hyperoxaemic PaO2 values lasted even after the patient had been stabilised for 10 min. However, RAR infants showed PaO2 values within the physiologic range. Figure 5 depicts the correlation between PaO2 values in arterial blood samples during resuscitation and GSSG concentrations. The PaO2 values below 100 mm Hg did not correlate with their corresponding GSSG concentrations. However, PaO2 values in the hyperoxic range, above 100 mm Hg, showed a very significant correlation with their corresponding GSSG concentrations. Since the RAR infants did not attain PaO2 values greater than 100 mm Hg, a significant correlation between the PaO2 values of the RAR group and their respective GSSG concentration was not found. Concomitantly, we performed SOD, catalase and GPx determinations in the aforementioned study [10]. The results are shown in table 3. In the first determination in the umbilical artery, a significant increase in SOD and catalase was evidenced in the asphyxiated infants, before initiation of resuscitation, when compared with the non-asphyxiated group. Table 3 shows that there were significant differences at 72 h of postnatal life for SOD, catalase and GPx between the asphyxiated OxR and RAR groups and non-asphyxiated controls. However, table 3 also shows that the activities of the antioxidant enzymes SOD and catalase in the OxR group are remarkably higher than in the RAR group at 72 h of postnatal life. Table 3 also shows that at the last determination, at 28 days of postnatal life, there were still higher SOD and catalase activities in the OxR group than in the non-asphyxiated control and RAR groups, while there were no differences between the latter two groups.

Discussion

In a very recent publication [5], the International Guidelines 2000 Conference on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care recommended the use of 100% oxygen in the resuscitation of asphyxiated newly born infants. However, the International Liaison Committee on Resuscitation indicates that there is accumulated evidence suggesting that lower inspired oxygen concen-

Biol Neonate 2001;79:261–267

265

trations may be useful under certain circumstances. Finally, the use of room air ventilation if supplemental oxygen is unavailable and positive pressure ventilation is needed is recommended. These statements, although very cautiously made indeed, are based upon a large body of experimental and clinical evidence gathered during the last decades. Research in animals and in humans has not only demonstrated the suitability of low oxygen concentrations in the resuscitation of asphyxiated newly born infants, but has also underscored the risks inherent in the use of pure oxygen [6, 7]. Pathophysiological alterations and tissue damage caused by the use of high oxygen concentrations during positive pressure ventilation are attributed to a great extent to hyperoxia [6] and to an increased production of oxygen free radicals [7]. Hyperoxia has been demonstrated to be the cause of a series of detrimental effects upon the respiratory reflex and cerebral circulation [21–23]. Moreover, very recent experimental work [24] has confirmed an increased production of free radicals in the pulmonary tissue in newly born piglets resuscitated with pure oxygen as compared to those resuscitated with room air. This model can be considered a useful replica of the real circumstances experienced by the asphyxiated newly born infant undergoing perinatal resuscitation [24]. Additionally, it is widely recognised that the delayed phase of injury before secondary energy failure during the ischaemia-reperfusion process in perinatal asphyxia is not associated with a deficit in tissue oxygenation, but nevertheless there is impaired mitochondrial function with the production of damaging free radicals [25]. These reactive oxygen species and their product, lipid peroxides, are thought to be among the important causes of cell membrane destruction and cell damage [25–27]. In addition, mitochondrial dysfunction may have far-reaching effects, leading not only to energy depletion and free radical generation, but also possibly to the direct activation of caspase 9 and triggering of the apoptotic execution of the cell [8]. Indeed, there is evidence to suggest that mitochondrial dysfunction may persist for a long time after hypoxia-ischaemia [28]. Our data [11] suggest that oxidative stress secondary to hyperoxia may last for at least the first 4 weeks of postnatal life (fig. 4). We hypothesise that protracted inflammation through the activation of leukocytes triggered by a pro-oxidant status may be the cause. In recent years, hundreds of newly born infants with severe and moderate asphyxia have been satisfactorily resuscitated using room air [9, 10, 11, 14]. In previous studies, no differences were found respecting epidemiological (mortality and morbidity) and clinical parameters in RAR and OxR infants [9–11]. Our personal data for the last 6 years, shown in table 1 [unpubl. data], are fully coincident with the results of the Resair 2 study [10]. Moreover, the analysis of the data has suggested some advantages of RAR, with respect to an earlier occurrence of the first cry or a tendency towards a higher Apgar score at 5 min [10]. In two previous reports [11, 14], we confirmed that RAR infants needed a shorter time of positive pressure ventilation in order to attain a sustained and spontaneous pattern of respiration (fig. 3). After analysing a total of 830 resuscitated newly born infants, we found that the mean time needed for the onset of a sustained pattern of respiration was on average more than 2 min longer in the OxR group than in the RAR group. Additionally, we found PaO2 values above physiological levels (1 100 mm Hg) in a significant number of OxR infants (fig. 5). Several minutes (B 10 min) after cessation of the positive pressure ventilation, hyperoxic values in arterial blood were still present. Thus, these infants received a significantly greater amount of oxygen per kilogram of body weight than the RAR infants did. During hyperoxia, alterations in the glutathione metabolism

266

Biol Neonate 2001;79:261–267

were detected. High GSSG concentrations correlated positively with the high levels of PaO2 [13]. This occurred in spite of a significant increase in the activity of glutathione reductase and the activities of antioxidant enzymes such as SOD, catalase or GPx [13]. These findings suggest that the antioxidant defence system of the newly born infants may be overwhelmed by a burst of free radicals of oxygen generated during the reperfusion phase in a hyperoxic atmosphere. These findings are consistent with previous findings in relation to glutathione metabolism [10]. Glutathione is the most important intracellular antioxidant, and is responsible for the redox status of the cells. The quotient GSH/GSSG reflects with great accuracy the pro-oxidant to antioxidant balance [29]. We have found that asphyxia provokes an imbalance in the GSH/GSSG ratio, in the sense of an increase in GSSG, thus reflecting a pro-oxidant situation [10]. However, RAR infants were able to re-establish an adequate GSH/GSSG ratio more rapidly (fig. 4). In contrast, OxR infants maintained an altered GSH/GSSG ratio for as long as 4 weeks of postnatal life, suggesting a protracted oxidative stress (fig. 4). The long-lasting alteration of the GSH/GSSG ratio indicates a protracted situation of oxidative stress. Oxygen free radicals modulating the cellular redox status act as cell signalling mediators for inflammation and apoptosis. An increased production of oxygen free radicals may cause differentiated responses of the cell proliferation activity, ranging from increased proliferation to apoptosis depending upon the degree of alteration of the redox status. Thus, a vast array of signalling molecules can be activated or inactivated by the cellular redox status changes induced by oxygen free radicals, thus inducing the transcription of antioxidant and/or proinflammatory genes as well as the caspase cascade [30, 31]. Consequently, the effects of a protracted oxidative stress may go far beyond the immediate tissue damage, and have long-term consequences.

Conclusions

For a period of 6 years, a total of 304 out of 830 asphyxiated newly born infants were resuscitated satisfactorily with room air in our maternity unit. No differences regarding morbidity or mortality have been detected as compared with the use of 100% oxygen for resuscitation. Thus, in the first place, we conclude that both methods of resuscitation seem to be equally effective. RAR infants gave their first cry earlier and attained a sustained pattern of spontaneous respiration more rapidly, thus needing bag and mask ventilation for a shorter period of time. Thus, RAR seems to achieve positive results more rapidly. OxR infants showed significantly higher arterial PaO2 than the RAR infants during the first 15 min of postnatal life. Arterial hyperoxaemic values correlated significantly with increased concentrations of GSSG. Thus, RAR avoids hyperoxaemia in the first minutes of life. Finally, oxidative stress, as measured by the GSH/GSSG quotient, was re-established to normal levels more rapidly in the RAR group as compared with the OxR group, thus indicating a protracted oxidative stress in the latter group, with unknown consequences. Our findings suggest that there is need for more strict control of the amount of oxygen given to newly born asphyxiated infants during resuscitation. Further studies should be undertaken.

Vento/Asensi/Sastre/Garcı´a-Sala/Viña

References 1 World Health Report. Geneva, World Health Organisation, 1995. 2 Child Health Development: Health of the Newborn. Geneva, World Health Organisation, 1991. 3 Saugstad OD: Practical aspects of resuscitating asphyxiated newborn infants. Eur J Pediatr 1998; 157(suppl 1):S11–S15. 4 Palme-Kilander C: Methods of resuscitation in low Apgar-score newborn infants: A national survey. Acta Paediatr 1992;81:739–744. 5 International Guidelines for Neonatal Resuscitation: An Excerpt From the Guidelines 2000 for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care: International Consensus on Science. Pediatrics 2000;106:e29. 6 Fellman V, Raivio KO: Reperfusion injury as the mechanism of brain damage after perinatal asphyxia. Pediatr Res 1997;41:599–606. 7 Saugstad OD: Resuscitation with room-air or oxygen supplementation. Clin Perinatol 1998;25:741– 756. 8 Edwards AD, Azzopardi DV: Perinatal hypoxiaischemia and brain injury. Pediatr Res 2000;47: 431–432. 9 Ramji S, Ahuja S, Thirupuram S, Rootwelt T, Aalen O: Resuscitation of asphyxiated newborn infants with room air or 100% oxygen. Pediatr Res 1993;34:809–812. 10 Saugstad OD, Rootwelt T, Aalen O: Resuscitation of asphyxiated newborn infants with room air or oxygen: An international controlled trial: The Resair 2 Study. Pediatrics1998;102:e1.

Experience with the Use of Room Air in Perinatal Resuscitation

11 Vento M, Asensi M, Sastre J, Garcı´a-Sala F, Pallardo FV, Viña J: Resuscitation with room air instead of 100% oxygen prevents oxidative stress in moderately asphyxiated term neonates. Pediatrics, in press. 12 Sastre J, Asensi M, Rodrigo F, Pallardo´ FV, Vento M, Viña J: Antioxidant administration to the mother prevents oxidative stress associated with birth. Life Sci 1994;54:2055–2059. 13 Vento M, Asensi M, Garcı´a-Sala F, Catala´ J, Viña J: Oxidative stress caused by fetal-neonatal transition in humans. Pediatr Res 1995;38:460A. 14 Vento M, Asensi M, Garcı´a-Sala F, Viña J, Asensi MA, Sastre J: The use of room air ventilation enhances the establishment of a sustained spontaneous respiratory pattern after perinatal asphyxia. Pediatr Res 2000;47:438A. 15 American Academy of Pediatrics/American Heart Association: Textbook of Neonatal Resuscitation. Dallas, American Heart Association, 1994. 16 Asensi M, Sastre J, Pallardo´ FV, Lloret S, Lehner M, Garcı´a de la Asuncio´n J, Viña J: Ratio of reduced to oxidised glutathione as indicator of oxidative stress status and DNA damage. Methods Enzymol 1999;299:267–276. 17 Asensi MA, Sastre J, Pallardo´ FV, Garcı´a de la Asuncio´n J, Estrela JM, Viña J: A high-performance liquid chromatography method for measurement of oxidized glutathione in biological samples. Anal Biochem 1994;217:323–328. 18 Flohé L, Güzzler WA: Assays on glutathione peroxidase. Methods Enzymol 1984;105:114–121. 19 Flohé L, Otting F: Superoxide dismutase assays. Methods Enzymol 1984;105:93–104.

20 Aebi H: Catalase in vitro. Methods Enzymol 1984; 105:121–126. 21 Hutchinson AA: Recovery from hypnoea in preterm lambs: Effects of breathing air or oxygen. Pediatr Pulmonol 1987;3:317–323. 22 Mortola JP, Frapell PB, Dotta A, Matsuoka T, Fox G, Weeks S, Mayer D: Ventilatory and metabolic responses to acute hyperoxia in newborns. Am Rev Respir Dis 1992;146:11–15. 23 Lundstrøm K, Pryds O, Greisen G: Oxygen at birth and prolonged cerebral vasoconstriction in preterm infants. Arch Dis Child 1995;73:F81–F86. 24 Nakai A, Asakura H, Taniuchi Y, Koshino T, Araki T, Siesjö BK: Effect of ·-phenyl-N-tert-butyl nitrone (PBN) on fetal cerebral energy metabolism during intrauterine ischemia and reperfusion in rats. Pediatr Res 2000;47:451–456. 25 Siesjö BK, Siesjö P: Mechanisms of secondary brain injury. Eur J Anaesthesiol 1996;13:247–268. 26 Kontos HA: Oxygen radicals in cerebral vascular injury. Circ Res 1985;57:508–516. 27 Siesjö BK, Agardh C-D, Bengtsson F: Free radicals and brain damage. Cerebrovasc Brain Metab Rev 1989;1:165–211. 28 Taylor DL, Edwards AD, Mehmet H: Oxidative metabolism, apoptosis and perinatal brain injury. Brain Pathol 1999;9:93–117. 29 Beckman KB, Ames BN: The free radical theory of aging matures. Physiol Rev 1998;78:547–581. 30 Finkel T: Oxygen radicals and signaling. Curr Opin Cell Biol 1998;10:248–253. 31 Kowaltowski AJ, Vercesi A: Mitochondrial damage induced by conditions of oxidative stress. Free Radic Biol Med 1999;26:463–471.

Biol Neonate 2001;79:261–267

267

Biol Neonate 2001;79:268–273

Psychological Prevention of Early Pre-Term Birth: A Reliable Benefit Nicole J. Mamelle

for the PPPB Study Group

INSERM Unit 369, Research Group on Epidemiology of Growth and Development, Lyon, France The members of the PPPB Study Group are as follows: Obstetricians: M. Berlanda, J.M. Thoulonb, G. Mellierb, M. Uzanc, G. Hathemd, D. Raudrante, B. Jacquetinf, B. Mariag, E. Papiernikh, C. Racineti. Psychologists: M. Seguillaa, b, B. Igertc, H. Donzed, M.A. Lafonde, P. Grunbergf, B. Le Cozannetg, C. Brigaudiota, B. Bereni-Mazoukh, A. Poizati. Statisticians: F. Munozj, O. Rivièrej. Centres:

aCentre

Hospitalier Lyon-Sud, Pierre Benite; bHôpital Ed Herriot, Lyon; cCentre Hospitalier J. Verdier, Bondy; Hospital of Bluets, Paris; eHôpital de l’Hôtel Dieu, Lyon; fHôpital de l’Hôtel Dieu, Clermont-Ferrand; gCentre Hospitalier Intercommunal, Villeneuve St Georges; hHôpital de Port-Royal, Paris; iHôpital Nord, Grenoble; jINSERM Unit 369, Research Group on Epidemiology of Growth and Development, Lyon. dMaternity

Key Words Pre-term labour W Pre-term birth W Prevention W Psychotherapeutic support W Epidemiological assessment

Abstract Objectives: After a previous study had shown the existence of psychological risk factors of pre-term delivery, we designed a study aimed at assessing the effect of psychotherapeutic support of pregnant women hospitalised with pre-term labour, followed by a second multicentric study aimed at demonstrating the reliability of such an intervention. Methods: Both studies were conducted in two successive cohorts of patients hospitalised with pre-term labour at 18– 35 weeks of gestation. The initial study comprised 157 patients in each group, whereas the reliability study comprised 191 patients in the experimental group versus 202 in the control group. In each experimental group, the patients were offered psychotherapeutic support in addition to the usual clinical management. The psychological support included interviews with a psychologist and a collaborative work plan implemented with the nursing staff. Results: The analysis, conducted in the ‘intention to treat’ manner, shows a significant decrease in the early pre-term birth rate (! 35 weeks) from 25.7 to 5.9% (p ! 0.0001). After controlling for confounding factors, the adjusted relative risk was 0.16 [95% confidence interval (CI) = 0.07–0.37]. These results were confirmed, at a lesser level, in the reliability study, where the early pre-term birth rate changed from 15.7 to 7.2% (p ! 0.02) and the adjusted relative risk was 0.35

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0268$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

(95% CI = 0.16–0.78). Conclusion: This study offers new and major results related to the prevention of delivery before 35 weeks of gestation, both in the initial study as well as in the reliability study. Thus, providing this type of psychological support to women hospitalised for pre-term labour, in the context of antenatal care, can help to avoid early pre-term births and their complications in terms of brain damage and neuropsychological development. Copyright © 2001 S. Karger AG, Basel

Introduction

During the last decade, numerous programs have been promoted to specifically address the reduction of pre-term births by social support. They were designed at different levels of prevention. Certain programs were devoted to assessing the high risks of pre-term birth in ordinary populations based on individual characteristics which are both immutable and amenable (first level), whereas other programs were aimed at preventing the onset of pre-term labour in women at high risk (second level), or others still at delaying early delivery in women in pre-term labour (third level). At the first level, prevention programs were devoted to identifying women at a high risk of pre-term birth, by means of a risk assessment system. Because most characteristics, such as age, race, family status, socioeconomic level and previous pregnancy history, cannot be changed, the objective is to modify the amenable risk factors, such as strenuous working conditions, toxic substance use or

N. Mamelle, Directeur de recherche INSERM U 369 151 cours Albert-Thomas F–69424 Lyon Cedex 03 (France) Tel. +33 4 7268 1950, Fax +33 4 7268 1951, E-Mail [email protected]

impaired partner support. The results of interventions based on social support in ordinary risk populations were disappointing [1]. At the second level, programs were aimed at preventing the onset of pre-term labour in women at a high risk. Some involved education to identify signs and symptoms of pre-term labour by self-palpation or home uterine monitoring and counselling in case of symptoms [2]. Others were based on home visiting systems by midwives or nurses or by family or social workers [3–11]. In a review of prevention programs in the United States, Alexander et al. [12] observed that investigations of high-risk patients produced little effect. From a large and recent meta-analysis of randomised assays, no beneficial effect was observed either [13]. One of the problems which are faced in these studies is related to the definition and measure of the ‘social support’, which may take many forms. At the third level, only a few programs were devoted to preventing pre-term births among women in pre-term labour, a situation in which medical treatment is often required. Here, the major problem is the lack of a generally accepted definition of pre-term labour and the variation concerning its management. In one study, conducted in Paris, no reduction in the rate of pre-term births was obtained in women with moderate threatened pre-term labour and in those with decreasing risk of pre-term labour after discharge from hospital [14]. However, Peacock et al. [15] noted a beneficial effect of emotional support by the partner, relatives or neighbours. Following an old observational study based on exercise and relaxation techniques which mentioned a reduction in the frequency and intensity of contractions and a uterine quiescence in women in pre-term labour, a recent quasi-experimental study reported a positive response to relaxation [16, 17]. In a similar vein, the groups of Omer et al. [18] and Mehl [19] used a hypnotic relaxation technique in a small number of patients in order to extend the period between the onset of pre-term birth and the date of delivery. In a recent review of the literature on social support and pregnancy outcome, Hoffman and Hatch [20] suggest that negative results are particularly observed in trials which were not conceived with intimate support in mind. The programs designed around the concept of emotional support by the partner, family or neighbours might have more positive results than the medico-social support provided by midwives. In view of these differing results regarding the potential benefits of social support for women in pre-term labour, we hypothesised that pre-term labour is the physical expression of a difficult pregnancy experience. In a previous study, we used a double psychological and epidemiological approach to inventory psychological distress and emotional response to events or environmental conditions in pregnant women who delivered prematurely. The psychological approach allowed the outlining of six psychological dimensions, as follows: pregnancy’s effects on the body, feelings of fulfilment during pregnancy, attitudes toward daily life and behaviour while expecting, the role of the baby’s father, family ties and maternal identification, and beliefs and superstitions [21]. The epidemiological approach, based on a large prospective study and a self-administered questionnaire built up from these previously outlined dimensions, showed a significant relationship between our ‘psychological score’ and the risk of pre-term birth [22]. This study contributed to a better understanding of psychological factors that may affect pregnant women and be associated with pre-term birth and allowed us to propose psychotherapeutic support especially devoted to preventing pre-term delivery. Therefore, we planned to conduct a study aimed at assessing the effectiveness of such psychotherapeutic support of women at risk of

pre-term birth. Indeed, it had appeared as a priority, from the French Sentinel network, to prevent pre-term births, especially among women who experienced a threatened pre-term labour; 15% of pregnant women have a threatened pre-term labour, and 25% of these deliver prematurely, so that 40% of pre-term deliveries can be attributable to pre-term labour [23]. An initial study was conducted in two successive cohorts of pregnant women with signs of pre-term labour, with or without hospitalisation, and showed a significant decrease in the rate of pre-term delivery, from 25.7% in the control group to12.3% in the experimental group which benefited from psychotherapeutic support. This result was confirmed after accounting for confounding factors [24]. Moreover, a specific analysis performed in the subgroup of hospitalised women in pre-term labour showed a significant decrease in the rate of pre-term birth from 39.6% in the control group to 18.7% in the experimental group [25]. The aim of the present paper is firstly to prove the benefit of such psychotherapeutic support in preventing early pre-term birth in women hospitalised for pre-term labour, from a re-analysis of this initial study, and secondly, to assess the reliability of this effectiveness in other maternity wards. As a consequence, such psychotherapeutic support in women at risk of an early pre-term birth would appear effective as an antenatal prevention measure of brain damage.

Prevention of Pre-Term Birth

Biol Neonate 2001;79:268–273

Subjects and Methods

Study Population Initial Study. The study population consisted of pregnant women with premonitory symptoms of pre-term labour, regardless of whether these clinical conditions necessitated hospitalisation. The clinical symptoms of pre-term labour were defined as follows: signs of uterine maturation (cervix shorter than 2 cm, patent internal os, ballooning of lower segment, signs of engagement), with or without painful contractions. All these changes have previously been shown to be good predictive signs of pre-term delivery [26]. Reliability Study. The study population consisted of pregnant women hospitalised between the 18th and the 35th week of pregnancy with symptoms of pre-term labour as previously defined in the initial study. Study Design Initial Study. Two successive cohorts were selected in the LyonSud Maternity Hospital in Lyon between 1992 and 1995. The first cohort comprised all the pregnant women with signs of pre-term labour during the first 18 months of the study. They were followed up with the usual therapeutic procedures (control group). The second cohort comprised all the hospitalised women during the second period of 18 months. They were offered psychological support in addition to the usual therapeutic procedures (experimental group). Reliability Study. A multicentric design was chosen in order to ensure diversity of patient recruitment. Eight maternity hospitals were included in the protocol, from Paris, Lyon, Bondy, ClermontFerrand, Villeneuve Saint Georges and Grenoble, and two successive cohorts of patients were included between 1996 and 1997 (see above for the composition of the study group). The control group was constituted during a first period of 6 months and the experimental group during a second period of 6 months, 1 year later. Patients in the experimental group were offered additional psychotherapeutic support, as in the initial study.

269

Description of Intervention The patients in the experimental group were offered psychological support comprising an initial meeting with the psychologist in charge of the study, followed by one or more interviews, as requested. In parallel with this direct intervention for the women, a co-operative work plan was implemented with the nursing staff. The intervention was designed on the basis of psychoanalytic concepts describing pregnancy as an identity crisis, comparable to adolescence, ‘through which one passes awakening latent anxiety and conflicts’, and on the content analysis of interviews with women who experienced pre-term birth in our previous etiological study [22, 27–29]. After the initial interview with its psychoanalytic orientation, an individual support program was proposed for each woman, based on whether the patient requested it and/or the nature of the problem observed by the psychologist. The intervention also comprised co-operative work with nursing staff in charge of the patient. Assessment Criterion In each study, the effectiveness criterion for the intervention was deemed the rate of pre-term delivery, i.e. before the 37th week of gestation, in the experimental group, regardless of whether the women met the psychologist and agreed to have psychological support, compared to the control group. In this paper, a secondary criterion was the rate of early pre-term birth, defined as delivery before the 35th week of gestation. Number of Subjects Required Initial Study. The number of patients required to demonstrate the effectiveness of the intervention was calculated on the basis of a 25% pre-term birth rate in the entire ‘control group’ and an expected reduction in the pre-term birth rate in the ‘experimental group’ by half, leading to 250 in each total group (· = ß = 0.05), i.e. about 150 patients in each subgroup of hospitalised women. Reliability Study. In this study, which only comprised hospitalised women, the number of patients required was calculated on the basis of a 30% pre-term birth rate in the ‘control group’ and an expected reduction by half in the ‘experimental group’, i.e. 200 in each group. Inclusion of Patients and Follow-Up Initial Study. All the patients who experienced symptoms or signs of pre-term labour were included in the study based on obstetrical considerations, with or without hospitalisation. After a 1-day delay necessary for initial medical care, the hospitalised women were referred to the psychologist in charge of the study. The baseline characteristics of the patients, age, gravidic complications, gestational age, contractions and signs of uterine maturation were registered at the entry to the study. At each following visit or hospitalisation, the gravidic complications and treatments were noted. Reliability Study. All the hospitalised patients with symptoms or signs of pre-term labour were included and followed up with the same care protocol as in the initial study. Statistical Analysis In each study, the analysis was performed in an ‘intent to treat’ manner to test the effectiveness of the offer of psychotherapeutic support made to all women, regardless of whether they actually met the psychologist and accepted the psychological support. The results were expressed as the crude and adjusted relative risks of pre-term birth in the experimental group compared to the control group, by considering, successively, gestational age ! 37 weeks and ! 35 weeks.

270

Biol Neonate 2001;79:268–273

Results

Initial Study Inclusion of Patients. The study dealt respectively with two successive samples of 323 women included in the first period (January 1992 to July 1993) who were treated with the usual management and 309 women included in the second period (November 1993 to May 1995) who had the same therapeutic management plus an offer of psychological support. Within both these groups, 157 women per group were hospitalised. Out of the 157 hospitalised women of the second sample, 125 (79%) were referred to the psychologist and, of these, 117 (94%) accepted the psychotherapeutic support. The number of interviews ranged from 1 to 8. Comparison of Baseline Characteristics of Hospitalised Patients. Characteristics of age, family status and the presence or absence of pathological events related to pre-term delivery were comparable for the two groups, except for maternal age, which was slightly lower in the experimental compared to the control group. The gestational age at hospitalisation, indicative of pre-term labour, was also equally distributed between the two groups, as were the clinical signs, i.e. short cervix, patent internal os, signs of engagement. However, a ballooned lower segment and contractions experienced as painful were slightly more frequent in the experimental group than among controls. The course of pregnancy and the rate of prescription of beta-mimetic agents were not different between the two groups. Effectiveness of Intervention. As specified in the study design, the effectiveness of intervention was tested in the ‘intention to treat’ manner by comparing the pre-term birth rates observed in the experimental and control groups, including all the women in the former group, whether or not they had actually received psychological support. The pre-term delivery rate (!37 weeks) observed among the hospitalised women of the experimental group was 18.5%, versus 39.6% in the control group (p ! 0.001). When considering early preterm delivery (!35 weeks), the sample was reduced to patients hospitalised before 35 weeks of gestation. The early pre-term birth rate was 5.9% in the experimental group, versus 25.7% in the control group (p ! 0.0001) (table 1). The estimate of the crude odds ratio was 0.35 (95% CI = 0.21–0.58) when considering pre-term birth (!37 weeks) and 0.18 (95% CI = 0.09–0.39) when considering early pre-term birth (!35 weeks) (table 2). These results remained stable when taking into account the following prognostic factors: maternal age, existence of painful contractions and ballooned lower segment at hospitalisation (which were variably distributed between the two groups), and gestational age at hospitalisation (a major prognostic factor). The logistic regression model led to adjusted odds ratios of 0.32 (95% CI = 0.19–0.56) when considering pre-term birth (!37 weeks) and 0.16 (95% CI = 0.07– 0.37) when considering early pre-term birth (!35 weeks) (table 2). Reliability Study Inclusion of Patients. The study was initially implemented in eight maternity units with the same protocol and recommendations to clinicians and psychologists. In two units, problems occurred with the inclusion of the patients, collection of medical data and with respect to the study design; there was no admission register, there were difficulties in collecting data from the medical records, and modification of hospitalisation criteria in relation to a change of medical staff occurred (Hôpital de Port-Royal of Paris and Hôpital Nord of Grenoble). Furthermore, in one of these two maternity units, the protocol was only partly respected, as psychologists did not

Mamelle

Table 1. Initial study: rates of pre-term birth (! 37 weeks) and early pre-term birth (! 35

weeks) in the experimental group as compared to the control group in hospitalised women in pre-term labour (Lyon, France, 1992–1995)

Gestational age ! 37 weeks Gestational age ! 35 weeks1 1

Statistical significance

18.5 5.9

p ! 0.0001 p ! 0.0001

39.5 25.7

In the subgroup of women hospitalised before 35 weeks of gestational age.

implement the collaborative work with medical and nursing staff, an important part of the intervention. For these reasons, these two centres had to be discarded from the data analysis. The inclusion of patients in the control group took place from January to June 1996, and those in the experimental group from January to June 1997. Globally, 393 patients were included into the study: 113 from the Ed. Herriot Hospital of Lyon, 91 from the J. Verdier Hospital of Bondy, 72 from the Maternity Hospital of Bluets in Paris, 67 from the Hôtel Dieu Hospital of Lyon, 35 from the Hôtel Dieu Hospital of Clermont-Ferrand and 15 from the Hospital of Villeneuve Saint Georges. The 202 patients included in the control group were treated with the usual management and the 191 patients included in the experimental group received the same management plus an offer of psychotherapeutic support. Out of the 191 hospitalised women from the second sample, 166 (87%) were referred to the psychologist and, of these, 161 (97%) accepted the psychotherapeutic support. Among these, the number of interviews ranged from 1 to 12. The mean number of interviews was 2.5 per patient and ranged from 1.5 to 3, depending on the maternity unit. Comparison of Baseline Characteristics of Hospitalised Patients. Comparisons of the initial characteristics of patients (age, family status, clinical signs or symptoms, pathological events, gestational age at hospitalisation) were performed with data collected from each maternity unit, in order to compare the patient recruitment from one maternity unit to another, and from the first period to the second one. The initial characteristics of the patients were rather different from one maternity unit to another; this variation was a reflection of the care level of the maternity units (3 maternity units of care level III, 1 of care level II and 2 of care level I). As a consequence of this variation in patient recruitment, the pre-term birth rates observed in the control groups of the 6 maternity units ranged from 25 to 37.5%. Comparisons were then made between the control and experimental groups, as a whole. The initial characteristics and gestational age at hospitalisation were equally distributed between both groups, as were the clinical signs, i.e. short cervix, ballooned lower segment and signs of engagement. However, patent internal os and painful contractions were significantly more frequent in the experimental group than among controls. Effectiveness of Intervention. As in the initial study, the effectiveness of intervention was tested in the ‘intention to treat’ manner, by comparing the pre-term birth rates observed in the experimental and control groups. The pre-term delivery rate (!37 weeks) observed among the hospitalised women of the experimental group was 19.9%, versus 30.2% in the control group (p ! 0.02). When considering early

Prevention of Pre-Term Birth

Experimental group: Control group: pre-term birth pre-term birth (n = 157), % (n = 157), %

Table 2. Initial study: estimate of relative risks of pre-term birth

(! 37 weeks) and early pre-term birth (! 35 weeks) in the experimental group as compared to the control group, before and after adjustment for prognostic factors, in hospitalised women in pre-term labour (Lyon, France, 1992–1995) Odds ratio

95% CI

Statistical significance

Gestational age ! 37 weeks: experimental/control group Before adjustment 0.35 0.21–0.58 p ! 0.0001 After adjustment1 0.33 0.19–0.56 p ! 0.0001 Gestational age ! 35 weeks2: experimental/control group Before adjustment 0.18 0.09–0.39 p ! 0.0001 After adjustment1 0.16 0.07–0.37 p ! 0.0001 1 After adjustment for maternal age, painful contractions, ballooned lower segment and gestational age at hospitalisation. 2 In the subgroup of women hospitalised before 35 weeks of gestational age.

pre-term deliveries (!35 weeks), the sample was reduced to patients hospitalised before 35 weeks of gestation. The early pre-term birth rate was 7.2% in the experimental group, versus 15.7% in the control group (p ! 0.02) (table 3). The estimates of crude odds ratios were 0.57 (95% CI = 0.36–0.91) when considering pre-term birth (!37 weeks) and 0.42 (95% CI = 0.21–0.83) when considering early preterm birth (!35 weeks) (table 4). A variation in this last odds ratio was noted from 0.23 to 0.60 depending on the maternity unit, which was partly related to the variable proportion of women who met the psychologist and accepted psychological support. As noted previously, these results remained stable when taking into account the following prognostic factors: patent internal os and painful contractions (which were variably distributed between the two groups), and gestational age at hospitalisation (a major prognostic factor). The logistic regression model led to adjusted odds ratios of 0.42 (95% CI = 0.21–1.0) when considering pre-term birth (!37 weeks) and 0.35 (95% CI = 0.16–0.78) when considering early preterm birth (!35 weeks) (table 4).

Biol Neonate 2001;79:268–273

271

Table 3. Reliability study: rates of pre-term birth (! 37 weeks) and very pre-term birth (! 35

weeks) in the experimental group as compared to the control group, in hospitalised women in pre-term labour (Paris, Lyon, Bondy, Clermont-Ferrand, Villeneuve St Georges, France, 1996–1997)

Gestational age ! 37 weeks Gestational age ! 35 weeks1 1

(! 37 weeks) and early pre-term birth (! 35 weeks) in the experimental group as compared to the control group, before and after adjustment for prognostic factors, in hospitalised women in pre-term labour (Paris, Lyon, Bondy, Clermont-Ferrand, Villeneuve St Georges, France, 1996–1997) 95% CI

Statistical significance

Gestational age ! 37 weeks: experimental/control group Before adjustment 0.57 0.36–0.91 p ! 0.02 After adjustment1 0.60 0.35–1.00 p = 0.05 Gestational age ! 35 weeks2: experimental/control group Before adjustment 0.42 0.21–0.83 p ! 0.02 After adjustment1 0.35 0.16–0.78 p ! 0.01 1 After adjustment for painful contractions, patent internal os and gestational age at hospitalisation. 2 In the subgroup of women hospitalised before 35 weeks of gestational age.

Discussion

Observational and experimental studies concerning the potential effect of psychosocial support conducted among women in pre-term labour are contradictory. Because of the putative negative results concerning the benefits of social support for women with high obstetrical risks, we used a support program based on psychotherapeutic interviews. Bydlowski [29] developed the notion of ‘psychic transparency’, a mode of psychological functioning in pregnant women which suggests that psychotherapeutic intervention can be effective in this population, even if performed over a very short time. The psychotherapeutic approach enabled women to verbalise painful or fearful concerns in order to make them less harmful. This oral elaboration turned out to be particularly useful for women with a previous poor obstetrical history or those who had experienced stressful or traumatising events during their present pregnancy. One or more interviews gave the patients the opportunity to revisit emotions related to pre-

272

Biol Neonate 2001;79:268–273

Control group: pre-term birth (n = 202), %

Statistical significance

19.9 7.2

30.2 15.7

p ! 0.02 p ! 0.02

In the subgroup of women hospitalised before 35 weeks of gestational age.

Table 4. Reliability study: estimate of relative risks of pre-term birth

Odds ratio

Experimental group: pre-term birth (n = 191 ), %

vious events, and allowed them to complete pregnancy more peacefully. The analysis of the interviews demonstrated that pre-term labour often occurred at a time when the woman failed to find the support required for her psychical safety, in her environment, after which she lost confidence in her capacity to fulfill the pregnancy. Of particular interest, the psychological support was very well accepted by the hospitalised patients in pre-term labour (94% in the initial study and 97% in the reliability study), showing the true desire of these hospitalised patients to get psychological support. From a biological point of view, in order to explain such a beneficial psychotherapeutic support, Omer and Everly [30] suggest a mechanism based on an increase in catecholamine levels under the effect of stress and anxiety, leading to increased uterine activity, and conclude that anxiety-reducing interventions may reduce catecholamine levels and thus help to inhibit contractions. Other hypotheses, developed by Majzoub et al. [31] and Hobel et al. [32], deal with the role of corticotropin-releasing factor as a biological intermediate between psychological stress and the occurrence of pre-term labour or delivery. In the initial study, we had chosen a study design based on two cohorts of women observed during two successive periods in the same maternity ward. Although the quality of this type of study is less robust than that of a controlled randomised trial, we chose this recruitment modality to avoid a contamination bias that would be inevitable in an intervention involving all the maternity staff in addition to the psychologist in charge of the interviews. No changes occurred in the patient recruitment pattern or in the procedures implemented in the two periods, so that good comparability of the initial characteristics of women in both groups was achieved. This fact allowed us to have confidence in the results. However, it was of paramount interest to show the reliability of such an intervention in various maternity units with various patient recruitment and medical procedures. The reliability study, conducted in eight maternity hospitals, shows the difficulties involved in implementing the same protocol based on psychotherapeutic support involving many persons caring for patients, i.e. medical and nursing staff and psychologists. These difficulties may also be a reflection of the place taken by psychologists in hospital services. An important point of our protocol is to reunify somatic and psychological dimensions in the minds of patients. Regular interviews with a psychologist and collaborative work with nurses and medical staff fostered the creation of therapeutic alliances that were used to help the women regain a more self-assured vision of their pregnancy and their image

Mamelle

as a successful mother-to-be. Concerning the effectiveness of the intervention, we showed that, globally, psychological support remains beneficial for women hospitalised with pre-term labour, even though the results appeared to be less spectacular. However, the decrease remains important when deliveries before 35 weeks of gestation are considered, the adjusted estimate of relative risk being 0.35 in the reliability study versus 0.16 in the initial study. We could make the hypothesis that this is partly due to differences in psychotherapeutic approaches from one psychologist to another and/or to variable acceptability of the psychologist’s work in medical care techniques, from one maternity ward to another. We observed an obvious range of the odds ratios, but we did not plan to analyse results centre by centre, because of the lack of statistical power in such a partial analysis. However, we can make the hypothesis that antenatal psychological support in women at high risk of pre-term delivery may have a beneficial effect on the establishment of the future mother-infant relationship, even though the delivery occurs prematurely. We believe that an effort must be made to give special

training to psychologists who want to work in the maternity field and to medical and nursing staff to better understand the benefits of working in collaboration with psychologists. In conclusion, this paper provides new and major results related to the prevention of deliveries before 35 weeks of gestation, both in the initial study as well as in the reliability study. Thus, providing this type of psychological support to women hospitalised for pre-term labour, in the context of antenatal care, can help to avoid early preterm births and their complications in terms of brain damage or neuropsychological development.

Acknowledgements

This assessment was supported by grants from the Mutuelle Générale de l’Education Nationale (MGEN) and from the Société d’Etudes et de Soins pour les Enfants Paralysés et Polymalformés (SESEP).

References 1 Andersen HF, Freda MC, Damus K, Brustman L, Merkatz IR: Effectiveness of patient education to reduce preterm delivery among ordinary risk patients. Am J Perinatol 1989;6:214–217. 2 Dyson DC, Crites YM, Ray DA, Armstrong MA: Prevention of preterm birth in high-risk patients: The role of education and provider contact versus home uterine monitoring. Am J Obstet Gynecol 1991;164:756–762. 3 Heins HC Jr, Nance NW, McCarthy BJ, Efird CM: A randomized trial of nurse-midwifery prenatal care to reduce low birth weight. Obstet Gynecol 1990;75:341–345. 4 Olds DL, Henderson CR, Tatelbaum R, Chamberlin R: Improving the delivery of prenatal care and outcomes of pregnancy: A randomized trial of nurse home visitation. Pediatrics 1986;77:16–28. 5 Bryce R, Stanley F, Garner B: Randomized controlled trial of antenatal social support to prevent preterm birth. Br J Obstet Gynaecol 1991;98: 1001–1008. 6 Oakley A, Rajam L, Grant A: Social support and pregnancy outcome. Br J Obstet Gynaecol 1990; 97:155–162. 7 Spira N, Audras F, Chapel A, Debuissone E, Jacquelin J, Kirchoffer C, Lebrun C, Prudent C: Domiciliary care of pathological pregnancies by midwives. Comparative controlled study on 996 women (in French). J Gynecol Obstet Biol Reprod (Paris) 1981;10:543–548. 8 Villar J, Farnot U, Barros F, Victoria C, Langer A, Belizan JM: A randomized trial of psychosocial support during high-risk pregnancies. The Latin American Network for Perinatal and Reproductive Research. N Engl J Med 1992;327:1266–1271. 9 Spencer B, Thomas H, Morris J: A randomized controlled trial of the provision of a social support service during pregnancy: The South Manchester family worker project. Br J Obstet Gynaecol 1989; 96:281–288. 10 Dawson AJ, Middlemiss C, Coles EC, Gough NA, Jones ME: A randomized study of a domiciliary antenatal care scheme: The effect on hospital admissions. Br J Obstet Gynaecol 1989;89:1319– 1322.

Prevention of Pre-Term Birth

11 Langer A, Victora C, Victora M, Barros F, Farnot U, Belizan J, Villar J: The Latin American trial of psychosocial support during pregnancy: A social intervention evaluated through an experimental design. Soc Sci Med 1993;36:495–507. 12 Alexander GR, Weiss J, Hulsey TC, Papiernik E: Preterm birth prevention: An evaluation of programs in the United States. Birth 1991;18:160– 169. 13 Hodnett ED: Support from caregivers during atrisk pregnancy; in Kierse MJCN et al (eds): Pregnancy and Child Module in the Cochrane Database of Systematic Reviews. Oxford, The Cochrane Collaboration, issue 2, update software, 1995. 14 Blondel B, Breart G, Llado J, Chartier M: Evaluation of the home-visiting system for women with threatened preterm labor: Results of a randomized controlled trial. Eur J Obstet Gynecol Reprod Biol 1990;34:47–58. 15 Peacock JL, Bland JM, Anderson HR: Preterm delivery: Effects of socioeconomic factors, psychological stress, smoking, alcohol and caffeine. BMJ 1995;311:531–536. 16 Derom H: Prevention of premature labour through physiotherapy. J Psychosom Obstet Gynaecol 1983;2–3:171. 17 Janke J: The effect of relaxation therapy on preterm labor outcomes. J Obstet Gynecol Neonatal Nurs 1999;28:255–263. 18 Omer H, Palti Z, Friedlander D: Evaluating treatments for preterm labor: Possible solutions for some methodological problems. Eur J Obstet Gynecol Reprod Biol 1986;22:229–236. 19 Mehl LE: Psychobiosocial intervention in threatened premature labor. Pre Perinat Psychol 1988;3: 41–52. 20 Hoffman S, Hatch MC: Stress, social support and pregnancy outcome: A reassessment based on recent research. Paediatr Perinat Epidemiol 1996;10: 380–405. 21 Audras de la Bastie M: Recherche des facteurs psychologiques dans l’étiologie de la prématurité. Psychanal Univ 1984;9:259–280.

22 Mamelle N, Measson A, Munoz F, Audras de la Bastie M, Gerin P, Hanauer MT, Collet P, Guyotat J: Development and use of a self-administered questionnaire for assessment of psychologic attitudes toward pregnancy and their relation to a subsequent premature birth. Am J Epidemiol 1989; 130:989–998. 23 Mamelle N, Lehingue Y, Munoz F, Miginiac M, Beranger C, Tounissoux D: Le Réseau Sentinelle AUDIPOG – Indicateurs de santé périnatale en 1994. J Gynecol Obstet Biol Reprod (Paris) 1996; 25:568–576. 24 Mamelle N, Segueilla M, Munoz F, Berland M: Prevention of preterm birth in patients with symptoms of preterm labor – the benefits of psychologic support. Am J Obstet Gynecol 1997;177:947–952. 25 Mamelle N, Segueilla M, Munoz F, Berland M: Prevention of pre-term birth – the role of psychological support. Prenat Neonat Med 1998;3:165– 169. 26 Bouyer J, Papiernik E: Facteurs de risque relevés lors des consultations prénatales; in Bouyer J (ed): La prématurité – enquête périnatale de Haguenau. Paris, INSERM, 1987, pp 25–58. 27 Racamier PC, Sens C, Carretier L: La mère et l’enfant dans la psychose du post-partum. Evol Psychiatr (Paris) 1961;8:525–570. 28 Bibring GL: Some considerations on the psychological processus in pregnancy. Psychoanal Study Child 1959;14:113–123. 29 Bydlowski M: La transparence psychique de la grossesse. Etud Freud 1991;32:135–142. 30 Omer H, Everly GS Jr: Psychological factors in preterm labor: Critical review and theoretical synthesis. Am J Psychiatry 1988;145:1507–1513. 31 Majzoub JA, McGregor JA, Lockwood CJ, Smith R, Taggart MS, Schulkin J: A central theory of preterm and term labor: Putative role for corticotropin-releasing hormone. Am J Obstet Gynecol 1999;180:S232–S241. 32 Hobel CJ, Dunkel-Schetter C, Roesch SC, Castro LC, Arora CP: Maternal plasma corticotropinreleasing hormone associated with stress at 20 weeks’ gestation in pregnancies ending in preterm delivery. Am J Obstet Gynecol 1999;180:S257– S263.

Biol Neonate 2001;79:268–273

273

Biol Neonate 2001;79:274–280

Pharmacotherapeutical Reduction of Post-Hypoxic-Ischemic Brain Injury in the Newborn Cacha Peeters Frank van Bel Department of Neonatology, Wilhelmina Children’s Hospital/University Medical Center, Utrecht, The Netherlands

Key Words Brain W Newborn W Hypoxia-ischemia W Pharmacological neuroprotection

Abstract Perinatal hypoxia-ischemia (PHI) is a major cause of morbidity and mortality. A substantial part of PHI-related brain damage occurs upon reperfusion and reoxygenation by the excess production of excitatory amino acids, free (pro)radicals and the release of cytokines, triggering programmed cell death. In this respect, several neuroprotective agents have been investigated in neonatal animal models, providing evidence for their usefulness in PHI. Several agents have been shown to be neuroprotective in neonatal animal hypoxia-ischemia models, but only a few agents have been used in clinical studies on term newborns. Although some general information will be provided with respect to focal hypoxia-ischemia and neuroprotective agents, this paper focuses on the investigated neuroprotective agents for global PHI and reperfusion brain injury in the newborn, categorized by their mode of action. Future experimental and clinical trials with promising neuroprotective agents need to be performed, including long-term follow-up to monitor long-term consequences. Moreover, well-designed combinations of neuroprotective agents with or without other neuroprotective strategies such as brain hypothermia should be given consideration for producing the most promising results in reducing post-hypoxic-ischemic reperfusion injury of the newborn brain.

motor and mental retardation, learning disabilities and/or epilepsy [4]. It is an important health problem with considerable social consequences, since many of these patients will be institutionalized because of their demanding care. PHI is characterized by a variable period of global hypoxia-ischemia followed by reperfusion and reoxygenation. This distinguishes PHI from adult stroke, which causes focal ischemia with or without reperfusion. During the actual hypoxic-ischemic insult of PHI, the most sensitive neurons will die. Although restoration of blood and oxygen supply to the brain is necessary to limit ischemic neuronal damage, the renewed availability of oxygen during the reperfusion phase gives rise to the activation of multiple biochemical pathways (which are partly interrelated), ultimately leading to secondary energy failure and substantial cell death [5, 6] (fig. 1). The target for acute pharmacotherapeutical intervention following PHI are the neurons that are still viable after the primary insult, but which will die because of secondary energy failure due to excess production of excitatory amino acids, free oxygen radicals and proradicals and the release of cytokines, triggering programmed cell death [6, 7]. Several agents have been shown to be neuroprotective in neonatal animal hypoxia-ischemia models, but only a few agents have been used in pilot studies for term newborns. This review discusses the most important investigated neuroprotective drugs, categorized by mode of action.

Voltage-Sensitive Calcium Channel and N-MethylD-Aspartate Channel Blockers and Glutamate Inhibition

Copyright © 2001 S. Karger AG, Basel

Introduction

Perinatal hypoxia-ischemia (PHI) affects about 3–6 per 1,000 term neonates in the Western world [1–3]. The long-term consequences of this injury are post-hypoxic-ischemic encephalopathy,

ABC

© 2001 S. Karger AG, Basel 0006–3126/01/0794–0274$17.50/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/journals/bon

Loss of calcium homeostasis in the neuronal cell is the key event in relation to PHI-induced brain damage. Under normal conditions, neurons maintain their (resting) intracellular free calcium level at extremely low concentrations [8]. This process is oxygen and ATP dependent [9]. When oxygenation fails, ATP formation drops and excessive calcium influx occurs through voltage-sensitive calcium channels, leading to the production of excitatory amino acids such as

Frank van Bel, MD Department of Neonatology, Room KE.04.123.1 Wilhelmina Children’s Hospital, PO Box 85090 NL–3508 AB Utrecht (The Netherlands) Tel. +31 30 250 4545, Fax +31 30 250 5320, E-Mail [email protected]

Fig. 1. Postulated biochemical pathways after global PHI with reperfusion/reoxygenation and modes of intervention. A Calcium blockers, NMDA channel blockers and glutamate inhibition. B NPBI chelation, free radical scavengers and xanthine oxidase inhibition. C NOS inhibition. D Anti-inflammatory pathways. E Anti-apoptotic pathways. F Growth factor therapy. VSCC = Voltage-sensitive calcium channels; eNOS = endothelial NOS; nNOS = neuronal

NOS; iNOS = inducible NOS.

glutamate [10]. This causes activation of the N-methyl-D-aspartate (NMDA) channel receptors, leading to further influx of calcium into the cell, triggering numerous and very complex cellular changes, such as an excessive production of free radicals, especially during reperfusion and reoxygenation (see below; fig. 1A). The regulation of calcium entry into the neuronal cells by inhibition of voltage-sensitive calcium channels has been extensively investigated, especially in relation to adult stroke. The calcium channel blockers nimodipine and flunarizine are important examples, and gave promising results for reducing infarct size when administered after permanent and transient focal ischemia in experimental studies [7, 11]. However, clinical studies have showed much less impressive results [12, 13]. A clinical study in asphyxiated newborns using nicardipine, an agent similar to nimodipine, as a rescue treatment was discontinued because of severe adverse hemodynamic effects of this agent [14]. A blocker of the NMDA-related glutamate receptor, MK-801, significantly improved neuronal outcome after hypoxia-ischemia in fetal sheep [15], but because of unacceptable adverse effects such as psychomimetic syndromes, sedation, catatonia and concerns about potential neurotoxicity, this agent cannot be used for clinical trials [16]. Magnesium sulfate, used as an NMDA channel blocker, showed no marked protective effects in newborn animals after severe hypoxiaischemia [17, 18]. A multicenter trial using magnesium sulfate in PHI neonates as a rescue treatment was prematurely discontinued because of unacceptable hypotension [19].

During the PHI insult, mitochondrial function is compromised and nonacceptance of electrons leads to free radical leakage from the mitochondria into the cytoplasm, causing lipid peroxidation of the polyunsaturated fatty acid-rich neuronal membranes [20]. The low intracellular pH during the actual period of hypoxia-ischemia, in concert with the already formed free radicals, gives rise to the release of metal ions from their binding proteins. Non-protein-bound iron (NPBI), in particular, is a very potent proradical which, via the superoxide-driven Haberman-Weiss reaction, strongly contributes to the formation of the very toxic hydroxyl free radical [21, 22]. Upon reperfusion, the renewed availability of oxygenated blood to the previously ischemic tissue causes a further and overwhelming formation of free oxygen radicals. This process is already ‘primed’ by the excessive calcium influx during the actual hypoxic-ischemic insult and accelerates intensively upon reperfusion and reoxygenation [23]. The calcium-induced production of free radicals and phospholipase A2 and C are involved in the breakdown of neuronal cell membrane phospholipids, releasing free fatty acids, especially arachidonic acid [24]. Arachidonic acid stimulates the enzyme cyclooxygenase and catalyzes the formation of the prostaglandin intermediate PGG2, which liberates the free radical superoxide. Superoxide can contribute directly and indirectly, for instance via the above-mentioned

Perinatal Hypoxic-Ischemic Brain Injury

Biol Neonate 2001;79:274–280

Non-Protein-Bound Iron Chelation, Xanthine Oxidase Inhibition and Free Radical Scavengers

275

Haberman-Weiss reaction and reaction with nitric oxide (see below), to free radical damage of the neuronal cells. Finally, the conversion of hypoxanthine, formed in large amounts during the actual asphyctic insult, to uric acid by xanthine oxidase upon reoxygenation gives rise to further formation of large amounts of superoxide [25]. See figure 1B for a schematic summary. Elimination of transition metals and especially NPBI seems fundamental to interruption of the formation of hydroxyl radicals. This can be achieved by iron chelators, e.g. deferoxamine, which has been successfully used in neonatal animals to diminish oxidative damage [21, 26–28]. In studies of our own group in newborn lambs and piglets, deferoxamine as a rescue treatment decreased free radical formation [29, 30], preserved electrical brain activity and brain metabolism [31] and prevented the decrease in Na+-,K+-ATPase activity of cortical cells after hypoxia-ischemia [32]. NPBI formation is indeed related to brain damage after severe asphyxia in human newborns, as shown by Dorrepaal et al. [33]. However, caution is warranted with deferoxamine in newly born humans, since it showed negative circulatory effects in preterm baboons [34]. Another agent with NPBIchelating effects is allopurinol [35, 36]. The prevention of the conversion of arachidonic acid to prostanoids by cyclooxygenase inhibitors such as indomethacin results in a decreased production of free radicals and protects the brain energy metabolism and electrical cortical activity in the newborn animal [31, 37, 38]. Moreover, indomethacin and other related agents, such as ibuprofen, reduce the number of circulating neutrophils, the amount of cytokine production and the plugging of the cerebral microcirculation [6, 39, 40] (see also Anti-Inflammatory Pathways below). Furthermore, in experimental studies with both focal as well as global ischemia, the use of superoxide dismutase and catalase showed a reduction of cerebral brain damage [41, 42]. It is generally believed that these enzymes exert their action from within the cerebral microcirculation, since they do not penetrate easily through cell membranes or the blood-brain barrier. No clinical studies involving this free radical pathway in neonates have been performed yet to the best of our knowledge. Brain edema and the production of superoxide and hydrogen peroxide are related to xanthine oxidase activity, and its inhibition reduces the production of these free radical species [25, 43]. Patt et al. [44] showed that the inhibition of xanthine oxidase in a gerbil model reduced free radical formation and brain edema. Palmer et al. [45] reported that rescue treatment with allopurinol, a xanthine oxidase inhibitor, was neuroprotective in immature rats in a dose-dependent manner [46], even when given up to 4 h after the actual hypoxic insult. In neonatal lambs and to a lesser extent in newborn piglets, allopurinol has been shown to have a preserving effect on brain metabolism and electrical cortical brain activity [30, 38]. It remains uncertain whether or not xanthine oxidase inhibition or direct free radical scavenging can explain this neuroprotective effect [36]. It has been shown that allopurinol, at least at a high concentration, can act as a scavenger of hydroxyl and as a metal ion chelator [35, 47]. Tirilazad and ebselen are free radical-reducing agents which have been tested in larger experimental and clinical studies in adults with stroke. Tirilazad is a lipid peroxidation inhibitor which acts as a free radical scavenger. In a number of clinical studies in adults with stroke, tirilazad showed only a very limited success [48, 49]. Ebselen is currently under investigation, but its efficacy still has to be proved [50].

276

Biol Neonate 2001;79:274–280

Inhibition of Nitric Oxide

Nitric oxide (NO W ) is a free radical produced by nitric oxide synthase (NOS) in cerebral endothelial cells, astrocytes and neuronal cells in a constitutional way. It unfolds diverse indispensable physiological intracellular calcium-dependent activities such as vasodilatation (endothelial NOS), modulation of neurotransmission, promotion of synaptogenesis and remodeling, and is involved in long-term potentiation (neuronal NOS). Another isoform of NOS, inducible NOS, which is calcium independent and inducible by cytokines, is present in white blood cells, macrophages and microglia [51]. During hypoxia-ischemia, but in particular upon and after reperfusion and reoxygenation, all three isoforms of NOS increase considerably, initiating increased formation of NO W . Endothelial NOSinduced NO W production has a neuroprotective function on the one hand, because it induces vasodilation, thus enhancing perfusion of the brain during PHI. On the other hand, excessive NO W production by neuronal NOS directly upon and after reperfusion after global PHI has direct and indirect deleterious effects on neuronal cells. This is because of the NO W -induced damage to neuronal cell DNA and the subsequent excess formation of superoxide (as indicated in the above section) to form peroxynitrite (ONOO – ) and eventually the very toxic hydroxyl free radical [51, 52]. Moreover, 4–12 h after reperfusion, excessive production of NO W occurs in brain tissue by the upregulation of inducible NOS in infiltrating neutrophils, macrophages and microglia, which can last for days [51, 53]. See figure 1C for a schematic summary. Given the direct and indirect deleterious effects of excessive production of NO W , NOS inhibition may ameliorate (perinatally) hypoxia-ischemia-related brain damage. Inhibition of NOS production using nonselective NOS inhibitors during the early post-hypoxicischemic period has been reported to reduce free radical-mediated reperfusion injury to the (newborn) brain [54, 55]. A study in severely asphyxiated newborn lambs did indeed show less free (pro)radical production and preservation of cerebral blood flow, brain metabolism and electrical brain activity in the first hours after hypoxiaischemia in those lambs treated with low-dose nonselective NOS inhibitors [56, 57]. However, several recent studies have reported that nonselective NOS inhibitors, especially those with prominent endothelial NOS inhibition characteristics, prevent adequate posthypoxic-ischemic brain perfusion, eventually leading to greater production of free radicals [58–60]. Selective inhibition of neuronal NOS with 7-nitroindazole in the gerbil [61] and 2-iminobiotin in the newborn piglet [62], and of inducible NOS with aminoguanidine in neonatal rats, proved to be neuroprotective after global hypoxia-ischemia [63, 64]. Further experimental and clinical research seems warranted to prove the protective potential of selective NOS inhibition in relation to neonatal reperfusion injury of the brain after severe PHI.

Anti-Inflammatory Pathways

Upon reperfusion, the damaged brain tissue releases inflammatory mediators, such as thrombin, histamine and reactive oxygen species, which cause the activation of blood vessel endothelium. The activated endothelium expresses adhesion molecules which migrate to the cell surface. P-selectin facilitates low-affinity rolling of the leukocytes on the endothelial surface. Intercellular adhesion molecule 1, upregulated after interleukin (IL)-1 and tumor necrosis factor (TNF)-

Peeters/van Bel

· activation, and integrin-ß2 molecules, conformationally changed by chemoattractants such as platelet-activating factor (PAF) and IL8, cause high-affinity adhesion between leukocytes and the endothelial cell. Adherent leukocytes secrete cytotoxic factors, leading to microvascular damage and infiltration of the damaged brain area [40, 65] (fig. 1D). IL-1ß and TNF-· are important mediators of the immune and inflammatory response, as has been shown by Szaflarski et al. [66] and Minami et al. [67]. This was supported by the finding that hypoxia-ischemia-related brain injury resulted in a transient marked increase in the expression of IL-1ß and TNF-· mRNA in brain regions susceptible to irreversible brain injury [68]. The neuroprotective effects that are shown using antagonists to the IL-1 receptor and TNF-· receptor after focal ischemia also provide evidence that IL-1ß and TNF-· play an important role in the response of the developing brain to post-hypoxic-ischemic brain injury [69]. PAF is a product of phospholipase A2 that is generated during reperfusion and reoxygenation when arachidonic acid metabolites are broken down. PAF exacerbates reperfusion injury by enhancing the inflammatory response [70] and stimulates neutrophils to release oxygen free radicals, thereby further enhancing neutrophil adhesion to the vessel wall [71]. These results suggest that PAF antagonists may harbor promising therapeutical potential. A similar role in limiting post-hypoxic-ischemic damage has been postulated for transforming growth factor-ß (TGF-ß), another member of the cytokine family. TGF-ß mRNA peaks in the brain a few days after hypoxia-ischemia, paralleling the time profile of monocyte/macrophage infiltration [72]. A neuroprotective role of TGF-ß has been proposed, since it increases neuronal survival after exposure to excitatory amino acids [73]. However, exogenous TGF-ß administration was able to reduce the infarct size only minimally in a rabbit model of cerebral ischemia [74]. Glial cell line-derived neurotrophic factor, another member of the TGF-ß family, was shown to reduce delayed neuronal death after adenovirus-mediated gene transfer in gerbils after transient global ischemia [75]. Nuclear factor-ÎB (NF-ÎB) is a transcription factor which resides in its inactivated form in the cytoplasm. Following activation after cerebral ischemia and reperfusion injury, NF-ÎB acts on genes for cytokines, adhesion molecules, NOS, cyclooxygenase-2, metalloproteinase-9 and perhaps apoptotic genes [76]. Depending on whether activation of NF-ÎB in neuronal cells is transient or permanent after hypoxia-ischemia, the effects could be protective or apoptotic [77]. However, inhibition of NF-ÎB by CVT-634, a proteasome inhibitor, has been shown to be neuroprotective in focal transient ischemia models in rats [78]. Further research directed at the newborn (animal) has to be done to determine the exact role of NF-ÎB in PHI.

Antiapoptotic Pathways

Apoptotic characteristics are often encountered in the process of delayed neuronal death after global PHI. The interplay between apoptosis and inflammation is likely to be a key event in cerebral hypoxic-ischemic cell injury. Caspases, a family of cysteinyl-aspartate proteases, are essential players in apoptotic death. Some serve as initiators (caspase-8, -9, -6) and some as executioners (caspase-3, -7), whereas others seem to be involved in promoting inflammation (caspase-4, -5) [79]. Caspase-1 is involved both in apoptosis as well as inflammation through the intermediate of the proinflammatory cytokine IL-1ß (fig. 1E). Inhibition of caspase-1-like activity was shown to be neuroprotective in a mouse model of transient cerebral isch-

Perinatal Hypoxic-Ischemic Brain Injury

emia [80]. Rabuffetti et al. [81] showed that a long-lasting reduction of apoptosis and decrease of the amount of proinflammatory cytokines could be achieved with caspase-1 inhibition by Ac-Tyr-ValAla-Asp-chloromethyl ketone. Pan-caspase inhibition with boc-aspartyl (OMe)-fluoromethylketone given either intracerebroventricularly or systemically after the hypoxic-ischemic insult proved to be neuroprotective in a perinatal rat model of hypoxia-ischemia [82]. These findings suggest that caspase inhibitors may be able to provide benefit over a prolonged therapeutic window after the hypoxic-ischemic event in the still-developing brain, probably by reestablishing a normal trophic state for neuronal cells at risk.

Growth Factor Therapy

The ability of the developing central nervous system to recover from hypoxic-ischemic injury is dependent on the production of neurotrophic and neurite-promoting factors (fig. 1F). Several growthpromoting factors have been isolated, such as insulin-like growth factor (IGF), basic fibroblast growth factor, hypoxia-inducible factor (HIF), epidermal growth factor (EGF) and brain-derived neurotrophic factor. IGF-1 is an anabolic pleiotrophic factor essential for postnatal brain development. It has been shown that IGF-1 mRNA is induced in infant rats after transient hypoxia-ischemia. Exogenous IGF-1, given centrally to adult rats following a similar hypoxic-ischemic insult, reduced the amount of neuronal loss [83]. Furthermore, it has been shown that the IGF-binding proteins are involved in altering the bioavailability and effect of IGF-1 and/or IGF-2 in the late phase of neuronal recovery and repair [84], suggesting a modulating role of IGFs in hypoxic-ischemic brain injury. Recent data report that daily subcutaneous growth hormone administration in a perinatal rat model of hypoxia-ischemia provides moderate neuroprotection [85]. All these results suggest a role for the IGF-1/growth hormone axis in the neurochemical process leading to hypoxic-ischemic brain injury. Similar results with other neurotrophic factors have been reported. Basic fibroblast growth factor has been shown to be effective in preventing neuronal injury after NMDA-induced neurotoxicity and hypoxia-ischemia in vivo [86]. HIF-1 is a transcription factor which is responsible for the activation of hypoxia-ischemia-inducible genes like erythropoietin, some glucose transporters and vascular endothelial growth factor (VEGF) [87]. Following transient global ischemia in the rat, a clear induction of HIF-1 and VEGF was demonstrated, suggesting a potential role in adaptive processes after hypoxia-ischemia, such as angiogenesis and neuroprotection [88]. Furthermore, in a neonatal hypoxia-ischemia model, upregulation of heparin-binding EGF was demonstrated immediately after the insult, suggesting that heparin-binding EGF modulates brain cell damage after hypoxia-ischemia [89]. Intracerebroventricular injection with brain-derived neurotrophic factor showed a rapid phosphorylation of the tyrosine kinase receptors in 7-day-old rats and demonstrated evident reduction of neuronal loss. In the 21-day-old rat, however, there was only moderate phosphorylation and protection [90]. Although the specific mode of action of neurotrophic factors following perinatal asphyxia is not fully understood, it is clear that modulation of the brain damage can be achieved, especially in the still-developing brain, if the necessary neurotrophic and neurite-promoting factors are available after the hypoxic-ischemic insult. See figure 1 for a schematic summary.

Biol Neonate 2001;79:274–280

277

What’s Next? Combination Treatment?

It is increasingly accepted that brain damage related to PHI occurs for a substantial part after reperfusion and reoxygenation, providing a window in which the impact of the cerebral reperfusion injury to the brain can be reduced by intervening in the above-mentioned destructive pathways. Experimental studies in the immature brain have indeed showed some promising results, but until now, clinical studies with pharmacotherapeutical interventions in asphyxiated neonates have been scarce. Furthermore, from studies in adults with stroke, we can learn that results with neuroprotective agents which were promising in the experimental setting are often disappointing in the clinical setting [7]. Insufficient information with respect to the therapeutic time window, the optimal dosage, duration of treatment and safety are mentioned as being responsible for this discrepancy. Regarding PHI in particular, another important consideration must be added to this list, namely the fact that information is often lacking about the nature of the actual hypoxic-ischemic insult, e.g. At what point in time did the insult start? What was the duration

of the insult? and Were there other disorders predisposing to PHI? These are all factors which can influence the efficacy of early postnatal pharmacological treatment of PHI, and the lack of answers to these questions urges for more properly conducted pharmacokinetic trials and stricter criteria to define those hypoxic-ischemic insults which are eligible for pharmacological protection. Furthermore, it is conceivable that intervening in one particular pathway will not completely prevent excessive calcium influx into the neuronal cell. Rather than a single therapy directed at one of the potentially destructive pathways, combinations of drugs intervening at different levels in the cascade might achieve greater reduction of the brain injury. Since it has been shown that hypothermia postpones secondary energy failure, the use of direct hypothermia could prolong the window for pharmacotherapeutic intervention [91]. Animal studies dealing with focal ischemia have shown that combination therapies of neuroprotective agents have synergistic effects [92–94]. To the best of our knowledge, no reliable combination intervention studies are being performed now in clinical PHI trials. However, this may be a clinically rewarding means of intervention with possibly exciting results.

References 1 Mulligan JC, Painter MJ, O’Donoghue PA, MacDonald HM, Allen AC, Taylor PM: Neonatal asphyxia. II. Neonatal mortality and long-term sequelae. J Pediatr 1980;96:903–907. 2 Levene ML, Kornberg J, Williams THC: The incidence and severity of post-asphyxial encephalopathy in full-term infants. Early Hum Dev 1985;11: 21–26. 3 MacDonald HM, Mulligan JC, Allen AC, Taylor PM: Relationship of obstetric and neonatal complications to neonatal mortality in 38,405 consecutive deliveries. J Pediatr 1980;96:898–902. 4 Volpe JJ: Hypoxic-ischemic encephalopathy: Clinical aspects; in Volpe JJ (ed): Neurology of the Newborn, ed 3. Philadelphia, W.B. Saunders, 1994, pp 314–369. 5 McCord JM: Oxygen-derived free radicals in postischemic tissue injury. N Engl J Med 1985;312: 159–163. 6 Palmer C: Hypoxic-ischemic encephalopathy: Therapeutic approaches against microvascular injury, and role of neutrophils, PAF, and free radicals. Clin Perinatol 1995;22:481–517. 7 De Keyser J, Sulter G, Luiten PG: Clinical trials with neuroprotective drugs in acute ischaemic stroke: Are we doing the right thing? Trends Neurosci 1999;22:535–540. 8 Siesjö BK: Cerebral circulation and metabolism. J Neurosurg 1984;60:883–908. 9 Vannucci RC: Experimental biology of cerebral hypoxia-ischemia: Relation to perinatal brain damage. Pediatr Res 1990;27:317–326. 10 McDonald JW, Johnston MV: Physiological and pathophysiological roles of excitatory amino acids during central nervous system development. Brain Res 1990;15:41–70. 11 Luiten PGM: Calcium homeostasis, nimodipine and stroke; in ter Horst GJ, Korf J (eds): Clinical Pharmacology of Cerebral Ischemia. Totowa, Humana Press, 1997, pp 69–99. 12 Franke CL, Palm R, Dalby M, Schoonderwaldt HC, Hantson L, Eriksson B, Lang-Jenssen L, Smakman J: Flunarizine in stroke treatment (FIST): A double-blind, placebo-controlled trial in Scandinavia and The Netherlands. Acta Neurol Scand 1996;93:56–60.

278

13 Wahlgren NG, Ranasinha KW, Rosolacci T, Franke CL, Van Erven PM, Ashwood T, Claesson L: Clomethiazole acute stroke study (CLASS): Results of a randomized, controlled trial of clomethiazole versus placebo in 1360 acute stroke patients. Stroke 1999;30:20–28. 14 Levene MI, Gibson NA, Fenton AC, Papathoma E, Barnett D: The use of a calcium-channel blocker, nicardipine, for severely asphyxiated newborn infants. Dev Med Child Neurol 1990;32:567–574. 15 Tan WK, Williams CE, Gunn AJ, Mallard CE, Gluckman PD: Suppression of post ischemic epileptiform activity with MK-801 improves neuronal outcome in fetal sheep. Ann Neurol 1992;32:677– 682. 16 Levene MI: Role of excitatory amino acid antagonists in the management of birth asphyxia. Biol Neonate 1992;62:248–251. 17 Galvin KA, Oorschot DE: Postinjury magnesium sulfate treatment is not markedly neuroprotective for striatal medium spiny neurons after perinatal hypoxia/ischemia in the rat. Pediatr Res 1998;44: 740–745. 18 Greenwood K, Cox P, Mehmet H, Penrice J, Amess PN, Cady EB, Wyatt JS, Edwards AD: Magnesium sulfate treatment after transient hypoxiaischemia in the newborn piglet does not protect against cerebral damage. Pediatr Res 2000;48: 346–350. 19 Levene MI, Blennow M, Whitelaw A, Hanko E, Fellman V, Hartly R: Acute effects of two different doses of magnesium sulphate in infants with birth asphyxia. Arch Dis Child Fetal Neonatal Ed 1995; 73:F174–F177. 20 Phillis JW: A radical view of cerebral ischemic injury. Prog Neurobiol 1994;42:441–448. 21 Gutteridge JM, Richmond R, Halliwell B: Inhibition of the iron-catalysed formation of hydroxyl radicals from superoxide and of lipid peroxidation by desferrioxamine. Biochem J 1979;184:469– 472. 22 Halliwell B: Reactive oxygen species and the cerebral nervous system. J Neurochem 1992;59:1609– 1623.

Biol Neonate 2001;79:274–280

23 Halliwell B, Gutteridge JM, Cross CE: Free radicals, antioxidants, and human disease: Where are we now? J Lab Clin Med 1992;119:598–620. 24 Wieloch T, Siesjö BK: Ischaemic brain injury: The importance of calcium lipolytic activities, and free fatty acids. Pathol Biol 1984;30:269–277. 25 Palmer C, Vannucci RC, Towfighi J: Reduction of perinatal hypoxic ischemic brain damage with allopurinol. Pediatr Res 1990;27:332–336. 26 Hedlund BE, Hallaway PE: High dose systemic iron chelation attenuates reperfusion injury. Biochem Soc Trans 1993;21:340–343. 27 Palmer C, Roberts RL, Bero C: Deferoxamine posttreatment reduces ischemic brain injury in neonatal rats. Stroke 1994;25:1039–1045. 28 Hurn PD, Koehler RC, Blizzard KK, Traystman J: Deferoxamine reduces early metabolic failure associated with severe cerebral ischemic acidosis in dogs. Stroke 1995;26:688–695. 29 Shadid M, Buonocore G, Groenendaal F, Moison RMW, Ferrali M, Berger HM, Van Bel F: Effect of deferoxamine and allopurinol on non-protein bound iron concentrations in plasma and cortical brain tissue of newborn lambs following hypoxiaischemia. Neurosci Lett 1998;248:5–8. 30 Peeters C, Ioroi T, Borst I, Groenendaal F, Van Bel F: 2-Iminobiotin, allopurinol and deferoxamine reduce post hypoxic-ischemic changes in cerebral hemodynamics and electric brain activity in newborn piglets. Pediatr Res 2000;47:61A. 31 Shadid M, Moison RMW, Steendijk P, Hiltermann L, Berger HM, Van Bel F: The effect of antioxidative combination therapy on post hypoxic-ischemic perfusion, metabolism, and electrical activity of the newborn brain. Pediatr Res 1998;44: 119–124. 32 Groenendaal F, Shadid M, McGowan JE, Mishra OP, Van Bel F: Effects of deferoxamine, a chelator of free iron, on Na+, K+-ATPase activity of cortical brain cell membrane during early reperfusion after hypoxia-ischemia in newborn lambs. Pediatr Res 2000;48:560–564. 33 Dorrepaal CA, Berger HM, Benders MJNL, Van Zoeren-Grobben D, Van de Bor M, Van Bel F: Nonprotein-bound iron in postasphyxial reperfusion injury of the newborn. Pediatrics 1996;98: 883–889.

Peeters/van Bel

34 DeLemos RA, Roberts RJ, Coalson JJ, DeLemos JA, Null DM, Gerstmann DR: Toxic effects associated with the administration of deferoxamine in the premature baboon with hyaline membrane disease. Am J Dis Child 1990;144:915–919. 35 Ko KM, Godin DV: Inhibition of transition metal ion-catalysed ascorbate oxidation and lipid peroxidation by allopurinol and oxypurinol. Biochem Pharmacol 1990;40:803–809. 36 Van Bel F, Shadid M, Moison RMW, Dorrepaal CA, Fontijn J, Monteiro L, Van de Bor M, Berger HM: Effect of allopurinol on postasphyxial free radical formation, cerebral hemodynamics and electrical brain activity. Pediatrics 1998;101:185– 193. 37 Pourcyrous M, Leffler CW, Bada HS: Brain superoxide generation in asphyxiated piglets and the effect of indomethacin at therapeutic doses. Pediatr Res 1993;34:366–369. 38 Van Bel F, Klautz RJM, Steendijk P, Schipper IB, Teitel DF, Baan J: The influence of indomethacin on the autoregulatory ability of the cerebral vascular bed in the newborn lamb. Pediatr Res 1993;34: 178–181. 39 Grice SC, Chappell ET, Prough DS, Whitley JM, Su M, Watkins WD: Ibuprofen improves cerebral blood flow after global cerebral ischemia in dogs. Stroke 1987;18:787–791. 40 Hudome S, Palmer C, Roberts RL, Mauger D, Housman C, Towfighi J: The role of neutrophils in the production of hypoxic-ischemic brain injury in the neonatal rat. Pediatr Res 1997;41:607–616. 41 Liu TH, Beckman JS, Freeman BA: Polyethylene glycol-conjugated superoxide dismutase and catalase reduce ischemic brain injury. Am J Physiol 1989;256:H589–H593. 42 Rosenberg AA, Murdaugh E, White CW: The role of oxygen free radicals in postasphyxia cerebral hypoperfusion in newborn lambs. Pediatr Res 1989;26:215–220. 43 Godin DV, Ko KM: Allopurinol and ischemia/ reperfusion injury: New use for an old drug? Can J Cardiol 1991;7:163–169. 44 Patt A, Harken AH, Burton LK: Xanthine oxidasederived hydrogen peroxide contributes to ischemia reperfusion-induced edema in gerbil brains. J Clin Invest 1988;81:1556–1559. 45 Palmer C, Towfighi J, Roberts R, Heitjan DF: Allopurinol administered after inducing hypoxiaischemia reduces brain injury in 7-day-old rats. Pediatr Res 1993;33:405–411. 46 Palmer C, Smith MB, Williams GD: Allopurinol preserves cerebral energy metabolism during perinatal hypoxic-ischemic injury and reduces brain damage in a dose dependent manner. J Cereb Blood Flow Metab 1991;11:S144–S149. 47 Moorhouse PC, Grootveld M, Halliwell B, Quinlan JG, Gutteridge JM: Allopurinol and oxypurinol are hydroxyl radical scavengers. FEBS Lett 1987; 213:23–28. 48 A randomized trial of tirilazad mesylate in patients with acute stroke (RANTTAS). The RANTTAS Investigators. Stroke 1996;27:453–458. 49 Fleishaker JC, Peters GR: Pharmacokinetics of tirilazad and U-89678 in ischemic stroke patients receiving a loading regimen and maintenance regimen of 10 mg/kg/day of tirilazad. J Clin Pharmacol 1996;36:809–813. 50 Yamaguchi T, Sano K, Takakura K, Saito I, Shinohara Y, Asano T, Yasuhama H: Ebselen in acute ischemic stroke: A placebo-controlled, doubleblind clinical trial. Ebselen Study Group. Stroke 1998;29:12–17. 51 Love S: Oxidative stress in brain ischemia. Brain Pathol 1999;9:119–131. 52 Beckman JS, Koppenol WH: Nitric oxide, superoxide and peroxynitrite: The good, the bad and the ugly. Am J Physiol 1996;271:C1424–C1437.

Perinatal Hypoxic-Ischemic Brain Injury

53 Iadecola C, Zhang F, Xu X: Inhibition of inducible nitric oxide synthase ameliorates cerebral ischemic damage. Am J Physiol 1995;268:R286–R292. 54 Dawson VL, Dawson TH, Bartley DA, Uhl CR, Snijder SH: Mechanisms of nitric oxide mediated neurotoxicity in primary brain cultures. J Neurosci 1993;13:2651–2661. 55 Hamada Y, Hayakawa T, Hattor H, Mikawa J: Inhibitor of nitric oxide synthesis reduces hypoxicischemic brain damage in the neonatal rat. Pediatr Res 1994;35:10–14. 56 Dorrepaal CA, Van Bel F, Moison RMW, Shadid M, Van de Bor M, Steendijk P, Berger HM: Oxidative stress during post hypoxic-ischemic reperfusion in the newborn lamb: The effect of nitric oxide synthesis inhibition. Pediatr Res 1997;41:321– 326. 57 Dorrepaal CA, Shadid M, Steendijk P, Van der Velde ET, Van de Bor M, Baan J, Van Bel F: Effect of post-hypoxic-ischemic inhibition of nitric oxide synthesis on cerebral blood flow, metabolism and electrocortical brain activity in newborn lambs. Biol Neonate 1997;72:216–226. 58 Ashwal S, Cole DJ, Osborne TN, Pearce WJ: Dual effects of L-NAME during transient focal cerebral ischemia in spontaneously hypertensive rats. Am J Physiol 1994; 267:H276–H284. 59 Groenendaal F, de Graaf RA, van Vliet G, Nicolay K: Effects of hypoxia-ischemia and inhibition of nitric oxide synthase on cerebral energy metabolism in newborn piglets. Pediatr Res 1999;45:827– 833. 60 Marks KA, Mallard CE, Roberts I, Williams CE, Gluckman PD, Edwards AD: Nitric oxide synthase inhibition and delayed cerebral injury after severe cerebral ischemia in fetal sheep. Pediatr Res 1999; 46:8–12. 61 O’Neill MJ, Hicks C, Ward M: Neuroprotective effects of 7-nitroindazole in the gerbil model of global cerebral ischaemia. Eur J Pharmacol 1996; 310:47–49. 62 Peeters C, Veldhuis W, Borst I, Nicolay K, Groenendaal F: 2-Iminobiotin, a neuronal NOS inhibitor preserves cerebral energy status and prevents vasogenic edema following hypoxia-ischemia in newborn piglets. Pediatr Res 2000;47:464A. 63 Higuchi Y, Hattori H, Kume T, Tsuji M, Akaike A, Furusho K: Increase in nitric oxide in the hypoxicischemic neonatal rat brain and suppression by 7nitroindazole and aminoguanidine. Eur J Pharmacol 1998;342:47–49. 64 Tsuji M, Higuchi Y, Shiraishi K, Kume T, Akaike A, Hattori H: Protective effect of aminoguanidine on hypoxic-ischemic brain damage and temporal profile of brain nitric oxide in neonatal rats. Pediatr Res 2000;47:79–83. 65 Jean WC, Spellman SR, Nussbaum ES, Low W: Reperfusion injury after focal cerebral ischemia: The role of inflammation and the therapeutic horizon. Neurosurgery 1998;43:1382–1397. 66 Szaflarski J, Burtrum D, Silverstein FS: Cerebral hypoxia-ischemia stimulates cytokine gene expression in perinatal rats. Stroke 1995;26:1093–1100. 67 Minami M, Kuraishi Y, Yabuuchi K, Yamazaki A, Satoh M: Induction of interleukin-1 beta mRNA in rat brain after transient forebrain ischemia. J Neurochem 1992;58:390–392. 68 Silverstein FS, Barks JD, Hagan P, Liu XH, Ivacko J, Szaflarski J: Cytokines and perinatal brain injury. Neurochem Int 1997;30:375–383. 69 Kadoya C, Domino EF, Yang GY, Stern JD, Betz AL: Preischemic but not postischemic zinc protoporphyrin treatment reduces infarct size and edema accumulation after temporary focal cerebral ischemia in rats. Stroke 1995;26:1035–1038.

70 Zimmerman GA, McIntyre TM, Mehra M, Prescott SM: Endothelial cell-associated platelet-activating factor: A novel mechanism for signaling intercellular adhesion. J Cell Biol 1990;110:529– 540. 71 Matsuo Y, Kihara T, Ikeda M, Ninomiya M, Onodera H, Kogure K: Role of platelet-activating factor and thromboxane A2 in radical production during ischemia and reperfusion of the rat brain. Brain Res 1996;709:296–302. 72 Wiessner C, Gehrmann J, Lindholm D, Topper R, Kreutzberg GW, Hossmann KA: Expression of transforming growth factor-ß1 and interleukin-1ß mRNA in rat brain following transient forebrain ischemia. Acta Neuropathol (Berl) 1993;86:439– 446. 73 Prehn JHM, Backhauss C, Krieglstein J: Transforming growth factor-ß1 prevents glutamate neurotoxicity in rat neocortical cultures and protects mouse neocortex from ischemic injury in vivo. J Cereb Blood Flow Metab 1993;13:521–525. 74 Gross CE, Howard DB, Dooley RH, Raymond SJ, Fuller S, Bednar MM: TGF-ß1 post-treatment in a rabbit model of cerebral ischemia. Neurol Res 1994;16:465–470. 75 Yagi T, Jikihara I, Fukumura M, Watabe K, Ohashi T, Eto Y, Hara M, Maeda M: Rescue of ischemic brain injury by adenoviral gene transfer of glial cell line-derived neurotrophic factor after transient global ischemia in gerbils. Brain Res 2000;885:273–282. 76 Carroll JE, Hess DC, Howard EF, Hill WD: Is nuclear factor-kappaB a good treatment target in brain ischemia/reperfusion injury? Neuroreport 2000;11:R1–R4. 77 Clemens JA, Stephenson DT, Yin T, Smalstig EB, Panetta JA, Little SP: Drug-induced neuroprotection from global ischemia is associated with prevention of persistent but not transient activation of nuclear factor-kappaB in rats. Stroke 1998;29:677– 682. 78 Buchan AM, Li H, Blackburn B: Neuroprotection achieved with a novel proteasome inhibitor which blocks NF-kappaB activation. Neuroreport 2000; 11:427–430. 79 Denner L: Caspases in apoptotic death. Exp Opin Invest Drugs 1999;8:37–50. 80 Hara H, Friedlander RM, Gagliardini V, Ayata C, Fink K, Huang Z, Shimizu-Sasamata M, Yuan J, Moskowitz MA: Inhibition of interleukin 1beta converting enzyme family proteases reduces ischemic and excitotoxic neuronal damage. Proc Natl Acad Sci USA 1997;94:2007–2012. 81 Rabuffetti, M, Sciorati C, Tarozzo G, Clementi E, Manfredi AA, Beltramo M: Inhibition of caspase1-like activity by Ac-Tyr-Val-Ala-Asp-chloromethyl ketone induces long-lasting neuroprotection in cerebral ischemia through apoptosis reduction and decrease of proinflammatory cytokines. J Neurosci 2000;20:4398–4404. 82 Cheng Y, Deshmukh M, D’Costa A, Demaro JA, Gidday JM, Shah A, Sun Y, Jacquin MF, Johnson EM, Holtzman DM: Caspase inhibitor affords neuroprotection with delayed administration in a rat model of neonatal hypoxic-ischemic brain injury. J Clin Invest 1998;101:1992–1999. 83 Gluckman PD, Klempt ND, Guan J, Mallard EC, Sirimanne E, Dragunow M, Klempt M, Singh M, Williams CE, Nikolocs K: A role for IGF-1 in the rescue of CNS neurons following hypoxic-ischemic injury. Biochem Biophys Res Commun 1992;182: 593–599. 84 Klempt ND, Klempt M, Gunn AJ, Singh K, Gluckman PD: Expression of insulin-like growth factorbinding protein 2 (IGF-BP2) following transient hypoxia-ischemia in the infant rat brain. Brain Res Mol Brain Res 1992;15:55–61.

Biol Neonate 2001;79:274–280

279

85 Gustafson K, Hagberg H, Bengtsson BA, Brantsing C, Isgaard J: Possible protective role of growth hormone in hypoxia-ischemia in neonatal rats. Pediatr Res 1999;45:318–323. 86 Nozaki K, Finklestein SP, Beal MF: Basic fibroblast growth factor protects against hypoxia-ischemia and NMDA neurotoxicity in neonatal rats. J Cereb Blood Flow Metab 1993;13:221–228. 87 Bergeron M, Gidday JM, Yu AY, Semenza GL, Ferriero DM, Sharp FR: Role of hypoxia-inducible factor-1 in hypoxia-induced ischemic tolerance in neonatal rat brain. Ann Neurol 2000;48:285–296. 88 Jin KL, Mao XO, Nagayama T, Goldsmith PC, Greenberg DA: Induction of vascular endothelial growth factor and hypoxia-inducible factor-1alpha by global ischemia in rat brain. Neuroscience 2000; 99:577–585.

280

89 Tanaka N, Sasahara M, Ohno M, Higashiyama S, Hayase Y, Shimada M: Heparin-binding epidermal growth factor-like mRNA expression in neonatal rat brain with hypoxic/ischemic injury. Brain Res 1999;827:130–138. 90 Cheng Y, Gidday JM, Yan Q, Shah AR, Holtzmann DM: Marked age-dependent neuroprotection by brain derived neurotrophic factor against neonatal hypoxic-ischemic brain injury. Ann Neurol 1997;41:521–529. 91 Guan J, Gunn AJ, Sirimanne ES, Tuffin J, Gunning MI, Clark R, Gluckman PD: The window of opportunity for neuronal rescue with insulin-like growth factor-1 after hypoxia-ischemia in rats is critically modulated by cerebral temperature during recovery. J Cereb Blood Flow Metab 2000;20: 513–519.

Biol Neonate 2001;79:274–280

92 Stuiver BT, Douma BR, Bakker R, Nyakas C, Luiten PG: In vivo protection against NMDA-induced neurodegeneration by MK-801 and nimodipine: Combined therapy and temporal course of protection. Neurodegeneration 1996;5:153–159. 93 Lyden PD, Lonzo L: Combination therapy protects ischemic brain in rats: A glutamate antagonist plus a gamma-aminobutyric acid antagonist. Stroke 1994;25:189–196. 94 Auer RN: Combination therapy with U74006F (tirilazad mesylate), MK-801, insulin and diazepam in transient forebrain ischaemia. Neurol Res 1995;17:132–136.

Peeters/van Bel

Author Index Vol. 79, No. 3–4, 2001

Akhter, W. 187 Asensi, M. 261 Ashraf, Q.M. 187 Bahr, B.A. 172 Bel, F. van 274 Bennett, P. 157 Bertini, G. 219 Blomgren, K. 172 Bocchi, C. 150 Bracci, R. 180, 210 Buonocore, G. 180, 210 Centini, G. 150 Challis, J.R.G. 163 Charollais, A. 236 Christensen, R.D. 228 Clerici, G. 246 Cobellis, L. 150 Dame, C. 228 Dani, C. 219 De Felice, C. 201 Delivoria-Papadopoulos, M. 187 Di Renzo, G.C. 246 Florio, P. 150 Fotopoulos, S. 213 Garcı´a-Sala, F. 261 Greisen, G. 194 Gressens, P. 236 Groenendaal, F. 254 Grond, J. van der 254 Hagberg, H. 172 Henrot, A. 224

Mamelle, N.J. 268 Marret, S. 236 Mercier, A. 236 Michel, C. 236 Mishra, O.P. 187 Papassotiriou, I. 213 Pavlou, K. 213 Peeters, C. 274 Perrone, S. 180, 210 Petraglia, F. 150 Pezzati, M. 219 Picciolini, E. 150 PPPB Study Group 268 Radi, S. 236 Reis, F.M. 150 Relier, J.-P. 149, 168 Roelants-van Rijn, A.M. 254 Rubaltelli, F.F. 219 Saliba, E. 224 Sastre, J. 261 Saugstad, O.D. 258 Severi, F.M. 150 Skouteli, H. 213 Smith, S.K. 163 Speer, C.P. 205 Terzidou, V. 157 Toet, M.C. 254 Toti, P. 201 Vanhulle, C. 236 Vannucci, R.C. 194 Vento, M. 261 Viña, J. 261 Vries, L.S. de 254

Juul, S.E. 228 Karlsson, J.-O. 172 Kissack, C.M. 241

Wang, X. 172 Weindling, A.M. 241 Xanthou, M. 213

Lardennois, C. 236 Lipsou, N. 213 Luisi, S. 150 Luzietti, R. 246

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 2001 S. Karger AG, Basel

Accessible online at: www.karger.com/journals/bon

Zanelli, S.A. 187 Zhu, C. 172

281

Subject Index Vol. 79, No. 3–4, 2001

Antepartum fetal monitoring 246 Apoptosis 172, 187 Asphyxia 157, 168, 228, 261

Gestational diseases 150 Glutamate 254 Glutathione 261

Bilirubin 219 – encephalopathy 219 Birth asphyxia 258 Blood pressure 241 Brain 168, 172, 187, 274 – damage 180, 194, 241 – injuries 201, 258 Bronchopulmonary dysplasia 205

Hypoxia 172, 187, 210 Hypoxia-isch(a)emia 194, 224, 254, 274

Cardiotocography 246 Caspase 172 Cerebral blood flow 194 – oxygenation 241 – palsy 157, 236 Chorioamnionitis 201 Chronic lung disease 205 Corticotropin-releasing hormone 163 Cortisol 163 Cytokines 205, 224 Disability 241 DNA fragmentation 187 Doppler technology 246 Epidemiological assessment 268 Erythrocyte 210 Erythropoietin 228 – mimetic peptide 228 – receptor 228 Fetal hemodynamics 246 – hypoxemia 246 – intrapartum electrocardiogram 246 – membranes 163 – pulse oximetry 246 Fetus 168, 210 Fodrin 172 Free radical injury 180

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 2001 S. Karger AG, Basel

Accessible online at: www.karger.com/journals/bon

Infant 236 Infant/newborn diseases 201 Infection 157 Inflammation 205 Inhibitor of caspase-activated DNase 172 Interleukin-1ß 213 Interleukin-6 213 Intrapartum fetal monitoring 246 Intraventricular hemorrhage 228 Ischemia 172, 228 Kernicterus 219 Labor 150 Low birth weight 213 Maternal stress 168 Mechanical ventilation 194 Myometrium 163 Neonatal intensive care 241 – rat 172 Neonate 241, 254 Neurodevelopment 228 Neurodisability 241 Neurohormones 150 Neuroprotection 228 Newborn 180, 187, 210, 224, 274 – baby 241 – resuscitation 258 Newly born infant 261 Nucleated red blood cells 213 Oxidative damage 205 – stress 210, 258, 261

Perinatal asphyxia 246 Periventricular leucomalacia 213 Pharmacological neuroprotection 274 Placenta 150, 157 Placental histology 201 Poly(ADP-ribose)polymerase 172 Porencephaly 236 Pregnancy 150 – complications (infectious) 201 Premature membrane rupture 224 Prematurity 194, 219 Pre-term birth 268 – infants 205 – lab(o)ur 157, 163, 268 Prevention 268 Prostaglandins 163 Proton magnetic resonance spectroscopy 254 Psychoaffective interchanges 168 Psychotherapeutic support 268 Reoxygenation injury 258 Resuscitation 261 Room air 261 Seizure 236 Stroke 228, 236 Thrombophilia 236 Tissue oxygenation 241 Transcutaneous bilirubin evaluation 219 Trophoblast 163 Tumour necrosis factor-· 213 Twins 157 White matter damage 224

Author Index Vol. 79, 2001

Abrams, R.M. 113 Ahonen, A. 27 Akhter, W. 187 Ameno, K. 91 Ameno, S. 91 Artlich, A. 21 Asensi, M. 261 Ashraf, Q.M. 187

Faggian, D. 87 Favaro, F. 87 Fidler, N. 15 Florio, P. 150 Fotopoulos, S. 213 Friedrich, W. 73 Fujii, T. 39

Baba, K. 39 Bahr, B.A. 172 Battaglia, F.C. 54 Bauer, K. 113 Bauer, R. 113 Bel, F. van 274 Bennett, P. 157 Berényi, E. 67 Bertini, G. 219 Bessler, H. 103 Bhiladvala, M. 21 Blomgren, K. 172 Bocchi, C. 150 Bogner, P. 67 Bracci, R. 180, 210 Buonocore, G. 180, 210 Carbone, T. 61 Carreiras, M.M. 5 Centini, G. 150 Challis, J.R.G. 163 Charollais, A. 236 Christensen, R.D. 228 Chung, M. 54 Clerici, G. 246 Cobellis, L. 150 Cuculich, P.S. 9

Jo´nsson, B. 21 Juul, S.E. 228

England, S. 61 Ertl, T. 109

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Hadzsiev, K. 109 Hagberg, H. 172 Hamai, Y. 39 Hegyi, T. 61 Hiatt, M. 61 Holownia, P. 79 Honour, J.W. 79 Huang, X. 113 Iacovidou, N. 1 Ijiri, I. 91 Itou, T. 140 Iwamoto, A. 46

Dame, C. 228 Dani, C. 219 Daskalaki, A. 1 David, J.C. 131 De Amici, D. 97 De Felice, C. 201 Delivoria-Papadopoulos, M. 187 Dellagrammaticas, H.D. 1 Delmonte, P. 97 DeLozier, K.A. 9 Di Renzo, G.C. 246 Do´czi, T. 67

ABC

Garcia-Sala, F. 261 Gasparoni, A. 97 Gerhardt, K.J. 113 Giacomin, C. 87 Gozal, D. 122 Greisen, G. 194 Gressens, P. 236 Groenendaal, F. 254 Grond, J. van der 254 Grongnet, J.F. 131 Gustafsson, L.E. 21

© 2001 S. Karger AG, Basel

Accessible online at: www.karger.com/journals/bon

Karlsson, J.-O. 172 Kato, M. 34 Kimura, H. 34, 91 Kinoshita, H. 91 Kiso, Y. 46 Kissack, C.M. 241 Kleinfeld, A.M. 61 Kobayashi, K. 39 Kodama, H. 46 Koivisto, M. 27 Kozuma, S. 39 Kubota, T. 91 Kusakabe, K. 46

Lanning, P. 27 Lardennois, C. 236 Lipsou, N. 213 Lönnqvist, P.A. 21 Luisi, S. 150 Luzietti, R. 246 Machida, Y. 39 Mamelle, N.J. 268 Marı´n, R. 5 Marret, S. 236 Martinotti, L. 97 Marumo, G. 39 Mellen, B.G. 9 Mercier, A. 236 Michel, C. 236 Midgley, P.C. 79 Mishra, O.P. 187 Moore, M. 79 Morikawa, A. 34 Morikawa, Y. 46 Morioka, H. 46 Mukamoto, M. 46 Nako, Y. 34 Nicolussi, S. 87 Nyu´l, Z. 67 Oates, N. 79 Ohki, Y. 34 Ohyu, J. 39 Okada, T. 46 Okai, T. 39 Papadimitriou, M. 1 Papadoyannis, M. 1 Papassotiriou, I. 213 Pavlou, K. 213 Peeters, C. 274 Perrone, S. 180, 210 Petraglia, F. 150 Piccolini, E. 150 Plebani, M. 87 PPPB Study Group 268 Proverbio, F. 5 Proverbio, T. 5 Punsky, I. 103 Pytel, J. 109 Radi, S. 236 Ramajoli, F. 97 Ramajoli, I. 97

283

Reis, F.M. 150 Relier, J.-P. 149, 168 Repa, I. 67 Roelants-van Rijn, A.M. 254 Rubaltelli, F.F. 219 Rüdiger, M. 73 Russell, K. 79 Rüstow, B. 73 Ryo, E. 39

Takashima, S. 39 Taketani, Y. 39 Teng, C. 54 Terzidou, V. 157 Timmerman, M. 54 Toet, M.C. 254 Tokuyama, K. 34 Torniainen, P. 27 Torres, J.E. 122 Toti, P. 201

Sakai, T. 140 Saliba, E. 224 Salobir, K. 15 Sasaki, F. 46 Sastre, J. 261 Saugstad, O.D. 258 Schmalisch, G. 73 Sedin, G. 67 Severi, F.M. 150 Shacham, D. 103 Shaw, J.C.L. 79 Shenai, J.P. 9 Sirota, L. 103

284

Skouteli, H. 213 Smith, J. 79 Smith, S.K. 163 Speer, C.P. 205 Stibilj, V. 15 Sugisawa, H. 140 Sulyok, E. 67, 109 Szabo, I. 109

Unno, N. 39

Biol Neonate Vol. 79, 2001

Vainionpää, L. 27 Vajda, Z. 67 Valkama, A.M. 27 Vanhulle, C. 236 Vannucci, R.C. 194 Vento, M. 261 Vina, J. 261 Vincze, O. 109 Vries, L.S. de 254 Wang, X. 172 Wauer, R. 73 Weinberger, B. 61 Weindling, A.M. 241 Wilkening, R.B. 54 Xanthou, M. 213 Zanardo, V. 87 Zanelli, S.A. 187 Zhang, X. 91 Zhu, C. 172 Zizzi, S. 97

Author Index

Subject Index Vol. 79, 2001

Amino acid exchange 54 Animal 131 Anoxic tolerance 122 Antepartum fetal monitoring 246 Apoptosis 172, 187 Asphyxia 157, 168, 228, 261 Autoinhalation 21 Autoresuscitation 122 Basement membrane 34 Behavioral state 113 Bilirubin 219 – encephalopathy 219 Birth asphyxia 258 Blood pressure 241 Brain 131, 168, 172, 187, 274 – damage 180, 194, 241 – injuries 201, 258 – single photon emission computed tomography 27 – water 67 Bronchopulmonary dysplasia 9, 34, 205 Bupivacaine 97 Caesarean section 97 Cardiotocography 246 Caspase 172 Cerebral blood flow 194 – oxygenation 241 – palsy 27, 157, 236 Cholesterol/phospholipid ratio 5 Chorioamnionitis 201 Chronic lung disease 9, 205 Colostrum 15, 87, 140 Cord blood 91 – occlusion 39 Corticotropin-releasing hormone 163 Cortisol 79, 163 Cortisone 79 Cranial ultrasound 27 Cytokines 61, 103, 205, 224 Delivery route 87 Developing kidney 46 Dexamethasone 9 Dilatation 1 Disability 241 DNA fragmentation 187 Doppler technology 246 Electrocortical activity 113 ß-Endorphin 87

ABC Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

© 2001 S. Karger AG, Basel

Accessible online at: www.karger.com/journals/bon

Epidemiological assessment 268 Epidermal growth factor 46 – – – receptor 46 Erythrocyte 210 Erythropoietin 228 – mimetic peptide 228 – receptor 228 Essential fatty acids 15 Exhaled gas 21 Fetal heart rate 113 – hemodynamics 246 – hypoxemia 246 – intrapartum electrocardiogram 246 – liver 54 – membranes 163 – pulse oximetry 246 – sheep 113 Fetus 168, 210 Fodrin 172 Free radical injury 180 Gestational diseases 150 Glucocorticosteroid 9 Glutamate 54, 122, 254 Glutamine 54 Glutathione 261 H1-NMR relaxometry 67 Heart 131 HSP-90 131 Human milk 15 Hyaluronan 67 Hyponatremia 109 Hypoxia 172, 187, 210 Hypoxia-isch(a)emia 194, 224, 254, 274 Ibuprofen 103 Impulse noise stimulation 113 Infant 236 – blood 91 Infant/newborn diseases 201 Infection 157 Inflammation 205 Inhibitor of caspase-activated DNase 172 Interleukin-1ß 61, 213 Interleukin-6 61, 213 Intrapartum fetal monitoring 246 Intraventricular hemorrhage 228 Ischemia 172, 228 Isoflurane 97

285

Kernicterus 219 Kidney 131 L/S ratio 73 Labor 150 – pain 87 Lactation 15 Lewis antigen level 91 – blood type 91 – gene 91 Lipid peroxide 39 Liver 131 Low birth weight 213 Lung 61, 131 Maternal stress 168 Mechanical ventilation 194 Mononuclear cells 103 Myometrium 163 Na,K-ATPase 5 Nasal cycle 21 Neonatal intensive care 241 – jaundice 97 – rat 172 Neonate 1, 61, 79, 241, 254 Neurodevelopment 228 Neurodisability 241 Neurohormones 150 Neuroprotection 228 Newborn 73, 180, 187, 210, 224, 274 – baby 241 – resuscitation 258 Newly born infant 261 Nitric oxide 21 – – synthase 122 N-methyl-D-aspartate 122 Nucleated red blood cells 213

Polymorphonuclear leukocytes 140 Porencephaly 236 Postnatal development 91 Preeclampsia 5 Pregnancy 150 – complications (infectious) 201 Premature membrane rupture 79, 224 – infant 27 Prematurity 9, 194, 219 Pre-term birth 268 Preterm infant 109, 205 – lab(o)ur 157, 163, 268 – newborns 103 Prevention 268 Proliferating cell nuclear antigen 46 Prostaglandins 163 Proton magnetic resonance spectroscopy 254 Psychoaffective interchanges 168 Psychotherapeutic support 268 Red blood cell ghosts 5 Renal pelvis 1 Reoxygenation injury 258 Respiration 122 Resuscitation 261 Retinopathy of prematurity 9 Room air 261 Secretor gene 91 Seizure 236 Sensorineural hearing loss 109 Sevoflurane 97 Stroke 228, 236 Surface tension 73 Surfactant deficiency 73 – inhibition 73 99Tcm-ECD

Oxidative damage 205 – stress 210, 258, 261 Parturition 54 Perinatal asphyxia 246 – rats 46 Periventricular leucomalacia 39, 213 Phagocytosis 140 Pharmacological neuroprotection 274 Placenta 54, 150, 157 Placental histology 201 Plasma 79 Pneumonia 73 Poly(ADP-ribose)polymerase 172

286

Biol Neonate Vol. 79, 2001

27 Thrombophilia 236 Tissue oxygenation 241 Transcutaneous bilirubin evaluation 219 Trophoblast 163 Tumor necrosis factor-· 61, 213 Twins 157 Type IV collagen 34 Unbound free fatty acids 61 Urinary tract infection 1 Urine metabolites 79 White matter damage 224

Subject Index

Contents Vol. 79, 2001

No. 2

No. 1

Original Papers

Original Papers 1

Mild Dilatation of Renal Pelvis in Term Neonates with Urinary Tract Infection Dellagrammaticas, H.D.; Iacovidou, N.; Papadimitriou, M.; Daskalaki, A.; Papadoyannis, M. (Athens)

5

Preeclampsia and Na,K-ATPase Activity of Red Blood Cell Ghosts from Neonatal and Maternal Blood Carreiras, M.M.; Proverbio, T.; Proverbio, F.; Marín, R. (Caracas)

9

15

Postnatal Dexamethasone Treatment and Retinopathy of Prematurity in Very-Low-Birth-Weight Neonates

Single Breath Analysis of Endogenous Nitric Oxide in the Newborn

Brain Single Photon Emission Computed Tomography at Term Age for Predicting Cerebral Palsy after Preterm Birth

Elevated Type IV Collagen in Bronchoalveolar Lavage Fluid from Infants with Bronchopulmonary Dysplasia

Generation of Periventricular Leukomalacia by Repeated Umbilical Cord Occlusion in Near-Term Fetal Sheep and Its Possible Pathogenetical Mechanisms Marumo, G.; Kozuma, S.; Ohyu, J.; Hamai, Y.; Machida, Y.; Kobayashi, K.; Ryo, E.; Unno, N.; Fujii, T.; Baba, K.; Okai, T.; Takashima, S.; Taketani, Y. (Tokyo)

46

Perinatal Development of the Rat Kidney: Proliferative Activity and Epidermal Growth Factor Okada, T.; Iwamoto, A.; Kusakabe, K.; Mukamoto, M. (Sakai); Kiso, Y. (Yamaguchi); Morioka, H.; Kodama, H.; Sasaki, F.; Morikawa, Y. (Sakai)

54

Net Amino Acid Flux across the Fetal Liver and Placenta during Spontaneous Ovine Parturition Timmerman, M. (Rotterdam); Teng, C.; Wilkening, R.B.; Chung, M.; Battaglia, F.C. (Denver, Colo.)

61

Effects of Perinatal Hypoxia on Serum Unbound Free Fatty Acids and Lung Inflammatory Mediators Weinberger, B.; Carbone, T.; England, S. (New Brunswick, N.J.); Kleinfeld, A.M. (San Diego, Calif.); Hiatt, M.; Hegyi, T. (New Brunswick, N.J.)

67

Brain Water and Proton Magnetic Resonance Relaxation in Preterm and Term Rabbit Pups: Their Relation to Tissue Hyaluronan Sulyok, E.; Nyúl, Z. (Pécs); Bogner, P.; Berényi, E.; Repa, I. (Kaposvár); Vajda, Z.; Dóczi, T. (Pécs); Sedin, G. (Uppsala)

20

Midgley, P.C. (London/Edinburgh); Holownia, P. (London); Smith, J.; Moore, M. (Edinburgh); Russel, K.; Oates, N.; Shaw, J.C.L.; Honour, J.W. (London)

87 Labor Pain Effects on Colostral Milk Beta-Endorphin Zanardo, V.; Nicolussi, S.; Giacomin, C.; Faggian, D.; Favaro, F.; Plebani, M. (Padua)

Ohki, Y.; Kato, M.; Kimura, H.; Nako, Y.; Tokuyama, K.; Morikawa, A. (Maebashi)

39

Metabolites in Preterm Infants

Concentrations of Lactating Mothers

Valkama, A.M.; Ahonen, A.; Vainionpää, L.; Torniainen, P.; Lanning, P.; Koivisto, M. (Oulu)

34

Rüdiger, M.; Friedrich, W.; Rüstow, B.; Schmalisch, G.; Wauer, R. (Berlin)

79 Plasma Cortisol, Cortisone and Urinary Glucocorticoid

Fatty Acid Composition of Human Colostrum in Slovenian Women Living in Urban and Rural Areas

Artlich, A.; Jónsson, B.; Bhiladvala, M.; Lönnqvist, P.A.; Gustafsson, L.E. (Stockholm)

27

Pneumonia

Cuculich, P.S.; DeLozier, K.A.; Mellen, B.G.; Shenai, J.P. (Nashville, Tenn.)

Fidler, N.; Salobir, K.; Stibilj, V. (DomzZ ale)

21

73 Disturbed Surface Properties in Preterm Infants with

Erratum

© 2001 S. Karger AG, Basel Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Access to full text and tables of contents, including tentative ones for forthcoming issues: www.karger.com/journals/bon/bon_bk.htm

91 Lewis and Secretor Gene Effects on Lewis Antigen and

Postnatal Development of Lewis Blood Type Ameno, S. (Kagawa); Kimura, H. (Tokyo); Ameno, K.; Zhang, X.; Kinoshita, H.; Kubota, T.; Ijiri, I. (Kagawa)

97 Can Anesthesiologic Strategies for Caesarean Section Influence

Newborn Jaundice? A Retrospective and Prospective Study De Amici, D.; Delmonte, P.; Martinotti, L.; Gasparoni, A.; Zizzi, S.; Ramajoli, I.; Ramajoli, F. (Pavia)

103 Ibuprofen Affects Pro- and Anti-Inflammatory Cytokine

Production by Mononuclear Cells of Preterm Newborns Sirota, L.; Shacham, D.; Punsky, I.; Bessler, H. (Petah Tiqva)

109 Hyponatremia and Sensorineural Hearing Loss in Preterm

Infants Ertl, T.; Hadzsiev, K.; Vincze, O.; Pytel, J.; Szabo, I.; Sulyok, E. (Pécs)

113 Effects of Impulse Noise Stimulation on Electrocorticogram

and Heart Rate Bauer, R. (Jena); Gerhardt, K.J.; Abrams, R.M.; Huang, X. (Gainesville, Fla.); Bauer, K. (Jena)

122 Brainstem Nitric Oxide Tissue Levels Correlate with

Anoxia-Induced Gasping Activity in the Developing Rat Gozal, D.; Torres, J.E. (Louisville, Ky.)

131 Perinatal Expression of Heat-Shock Protein 90 in Different

Regions of the Brain and in Non-Neural Tissues of the Piglet David, J.C.; Grongnet, J.F. (Rennes)

140 Promoting Effect of Colostrum on the Phagocytic Activity of

Bovine Polymorphonuclear Leukocytes in vitro Sugisawa, H.; Itou, T.; Sakai, T. ( Kanagawa)

219 Prevention of Bilirubin Encephalopathy Bertini, G.; Dani, C.; Pezzati, M.; Rubaltelli, F.F. (Florence)

No. 3–4 The Newborn at High Risk of Brain Damage. EURope Against Infant Brain Injury (EURAIBI) International Workshop. Siena, Italy, April 5–7, 2001 Guest Editor: G. Buonocore, Siena

224 Inflammatory Mediators and Neonatal Brain Damage Saliba, E.; Henrot, A. (Tours) 228 The Biology of Erythropoietin in the Central Nervous System

and Its Neurotrophic and Neuroprotective Potential Dame, C. (Gainesville, Fla.); Juul, S.E. (Seattle, Wash.); Christensen, R.D. (Gainesville, Fla.)

149 Foreword Relier, J.-P. (Paris)

236 Fetal and Neonatal Cerebral Infarcts Marret, S.; Lardennois, C.; Mercier, A.; Radi, S.; Michel, C.; Vanhulle, C.; Charollais, A. (Rouen); Gressens, P. (Paris)

150 Human Placenta as a Source of Neuroendocrine Factors Reis, F.M.; Florio, P.; Cobellis, L.; Luisi, S.; Severi, F.M.; Bocchi, C.; Picciolini, E.; Centini, G.; Petraglia, F. (Siena)

241 Blood Pressure and Tissue Oxygenation in the Newborn Baby

157 Maternal Risk Factors for Fetal and Neonatal Brain Damage Terzidou, V.; Bennett, P. (London)

246 Monitoring of Antepartum and Intrapartum Fetal Hypoxemia:

163 Fetal Endocrine Signals and Preterm Labor Challis, J.R.G. (Cambridge/Toronto); Smith, S.K. (Cambridge) 168 Influence of Maternal Stress on Fetal Behavior and Brain

Development Relier, J.-P. (Paris)

172 Caspase-3 Activation after Neonatal Rat Cerebral Hypoxia-

at Risk of Brain Damage Weindling, A.M.; Kissack, C.M. (Liverpool)

Pathophysiological Basis and Available Techniques Clerici, G.; Luzietti, R.; Di Renzo, G.C. (Perugia)

254 Glutamate in Cerebral Tissue of Asphyxiated Neonates during

the First Week of Life Demonstrated in vivo Using Proton Magnetic Resonance Spectroscopy Groenendaal, F.; Roelants-van Rijn, A.M.; van der Grond, J.; Toet, M.C.; de Vries, L.S. (Utrecht)

Ischemia

258 Resuscitation of the Asphyxic Newborn Infant: New Insight

Wang, X. (Göteborg/Zhengzhou); Karlsson, J.-O. (Göteborg); Zhu, C. (Göteborg/Zhengzhou); Bahr, B.A. (Storrs, Conn.); Hagberg, H.; Blomgren, K. (Göteborg)

261 Six Years of Experience with the Use of Room Air for the

180 Free Radicals and Brain Damage in the Newborn Buonocore, G.; Perrone, S.; Bracci, R. (Siena) 187 Effect of Graded Hypoxia on Cerebral Cortical Genomic DNA

Fragmentation in Newborn Piglets Akhter, W.; Ashraf, Q.M.; Zanelli, S.A.; Mishra, O.P.; Delivoria-Papadopoulos, M. (Philadelphia, Pa.)

194 Is Periventricular Leucomalacia a Result of Hypoxic-Ischaemic

Leads to New Therapeutic Possibilities Saugstad, O.D. (Oslo)

Resuscitation of Asphyxiated Newly Born Term Infants Vento, M. (Valencia/Alicante); Asensi, M.; Sastre, J. (Valencia); García-Sala, F. (Valencia/Alicante); Viña, J. (Valencia)

268 Psychological Prevention of Early Pre-Term Birth:

A Reliable Benefit Mamelle, N.J. (Lyon) for the PPPB Study Group

274 Pharmacotherapeutical Reduction of Post-Hypoxic-Ischemic

Injury? Hypocapnia and the Preterm Brain

Brain Injury in the Newborn

Greisen, G. (Copenhagen); Vannucci, R.C. (Hershey, Pa.)

Peeters, C.; van Bel, F. (Utrecht)

201 Chorioamnionitis and Fetal/Neonatal Brain Injury Toti, P.; De Felice, C. (Siena) 205 New Insights into the Pathogenesis of Pulmonary Inflammation

281 Author Index Vol. 79, No. 3–4, 2001 282 Subject Index Vol. 79, No. 3–4, 2001

in Preterm Infants Speer, C.P. (Würzburg)

210 Red Blood Cell Involvement in Fetal/Neonatal Hypoxia Bracci, R.; Perrone, S.; Buonocore, G. (Siena)

283 Author Index Vol. 79, 2001 285 Subject Index Vol. 79, 2001

213 Early Markers of Brain Damage in Premature Low-Birth-Weight

Neonates Who Suffered from Perinatal Asphyxia and/or Infection Fotopoulos, S.; Pavlou, K.; Skouteli, H.; Papassotiriou, I.; Lipsou, N.; Xanthou, M. (Athens)

IV

Biol Neonate Vol. 79, 2001

Contents

Digestion 1998;59(suppl 4):1–12

University Department of Medicine The Royal Infirmary, Manchester, UK

A Framework for the Aetiogenesis of Chronic Pancreatitis

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Key Words Xenobiotics Cytochrome P450 Glutathione Antioxidants Signal transduction Chronic pancreatitis

Abstract The traditional ductal model for the development of chronic pancreatitis leaves many questions unanswered and it has not facilitated management. An alternate philosophy centres on the acinar cell as the site of mounting oxidant stress, usually as a result of steady exposure to xenobiotics that induce cytochrome P450 mono-oxygenases while depleting glutathione: ductal changes are regarded as secondary, disease-compounding manifestations of the oxidant environment. Within this framework each burst of oxidant stress jeopardises exocytosis, to trigger an attack of pancreatitis by interfering with the methionine-to-glutathione transsulphuration pathway, which interacts closely with ascorbate and selenium. The resulting diversion of free radical oxidation products into the pancreatic interstitium causes mast cells to degranulate, thereby provoking inflammation, the activation of nociceptive axon reflexes, and profibrotic interactions.

Joan M. Braganza

OOOOOOOOOOOOOOOOOOOOOO

Background Chronic pancreatitis is a progressive and generally painful inflammatory process. It has been on the clinical map for more than 200 years; yet until recently there has been little prospect of specific medical treatment for the cardinal symptom, pain. This stalemate suggests fundamental misconceptions as to the aetiogenesis of the disease and the cause of pain. In regard to the former, the concept of ‘primary ductal lithiasis’ has held sway for some time and was reviewed at the World Congress of Gastroenterology in 1986 [1]; with reference to the latter, the on-going practice of total pancreatectomy to control intractable pain suggests that viable acini are a prerequisite for its production. This paper focuses on an alternate concept wherein the pancreatic acinar cell is regarded as

ABC

© 1998 S. Karger AG, Basel 0012–2823/98/5910–0001$15.00/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

the site of mounting oxidant stress, usually as a result of exposure to xenobiotics, that is to say, exogenous lipophilic compounds. The stageposts in the development of the new philosophy can be tracked from successive reports over the past 15 years [2–8]. However, the seeds for the concept were actually sown in the earliest case reports. Thus, in 1788 Cawley [9] described a patient with calcific chronic pancreatitis who also had subtle disturbances in the liver and kidneys; Wagner’s famous patient Beethoven who died in 1827 had macronodular cirrhosis and calcareous renal papillary necrosis together with the noncalcific variant of chronic pancreatitis [10]; and this triad of diseased organs was mentioned again in 1878 by Friedrich [11] who made the link with alcohol. Whereas the liver has long been known as the main detoxification site for xeno-

Dr. J.M. Braganza, DSc MSc FRCP FRCPath University Department of Medicine The Royal Infirmary, Manchester M13 9WL (UK) Tel./Fax +44 0161 276 4168 E-Mail [email protected]

otics. By reference to structural and secretory aspects of pancreatic organisation, it becomes possible to appreciate how the reactivation of a latent xenobiotic-metabolising capability in pancreatic acinar cells can cascade the series of changes leading to chronic active inflammatory fibrosis of the gland.

Frame of Reference

Fig. 1. Microcirculatory arrangements in the liver with three functional zones (1, 2, 3) determined by the arrangement of the central vein (CV) and portal tract (PT), contrasted with that in the pancreas. Composite figure reproduced from Timbrell [24] and Foulis [23] with kind permission from the authors and publishers.

biotics that enter via the gastrointestinal tract, and the kidney shares the responsibility for their safe disposal, it is now clear that the extrahepatic metabolism of xenobiotics may be very significant in pathological terms [12] and that the pancreas is a versatile, but also vulnerable, drugmetabolising organ [7]. The new concept is analogous to carbon tetrachloride [13] and also paracetamol hepatotoxicity that result from the loss of protection by glutathione (GSH) of critical protein thiols, but with the difference that in chronic pancreatitis this problem arises over many years, as a result of repetitive and simultaneous exposures to several xenobi-

2

Digestion 1998;59(suppl 4):1–12

Structural Considerations Several facets are potentially relevant to the development of chronic pancreatitis [14]. A common ampullary channel in some 65% of individuals could allow toxic substances excreted in bile to reflux into the pancreas, either directly or by way of the duodenum; in this event, as also when high ductal pressure is generated during endoscopic retrograde cholangiopancreatography (ERCP), the singlelayered epithelium of the smallest intralobular ducts would yield. However, the construction of many acini as anastomosing networks mitigates against intralobular obstruction. The tight junctions between acinar cells lie on the lateral surface adjacent to the lumen, so that there is a wide surface area for communication with the interstitial space. Junctional complexes are just one among many factors that interact to maintain the polarity of secretory epithelia; others include the actin cytoskeleton, distinctive domains at apical and basolateral aspects of the cell surface membrane, and adhesion molecules that facilitate attachment between cells and the substratum [15]. The pancreatic interstitium contains many adipocytes [16, 17], including retinol-rich cells with the capacity to metamorphose into myofibroblast-like stellate cells [18], and numerous mast cells that lie between basement membrane of capillaries and plasma membrane of acinar cells [16], juxtaposed to membrane receptors for a tissue-type plasminogen activator [19] and also urokinase-plasminogen activator [20]. Immune systems are not prominent and there is neither a substantial HLA display [21] nor secretion of immunoglobulins [22]. The pancreas has a generous arterial supply, but its pattern of branching leaves the periphery of the lobule vulnerable in ischaemic states [23], in the same way as leaves zone 3 hepatocytes vulnerable [24] (fig. 1). A variable portion of the blood supply first passes to the islets of Langerhans, constituting a limited ‘portal’ circulation. An extensive lymphatic drainage originates in the perilobular connective tissue and gains access to the surface of the gland by channels that run alongside the blood vessels. Parasympathetic fibres via the vagus nerve provide part

Braganza

of the secretory stimulus to the pancreas; sympathetic fibres innervate the vasculature and carry pain-mediating fibres. Secretory Aspects The acinar cell is a prototype of a polarised secretory unit: occupancy of secretagogue receptors in the basolateral membrane, under a facilitating neural influence, results in the bulk discharge by exocytosis of stored zymogen granules at the apical pole into the centroacinar space. This is accomplished by a protein transport apparatus that is used by all eukaryotes and involves trafficking among membrane-bound compartments, facilitated by vesicular carriers [25]. Fusion of transport vesicles with their respective target membranes depends upon soluble adaptor molecules, the N-ethylmaleimide (NEM)-sensitive fusion factor (NSF) and the soluble NSF attachment proteins (SNAPs). Compartmental specificity is provided by SNAP membrane receptors (SNAREs) and also by the family of Rab proteins. All secretory cells possess ‘constitutive’ pathways and the exocrine pancreas is no exception: two pathways have been identified at the apical pole and one at the basolateral pole [26]. Different requirements for the adaptor molecules have been identified in apical and basolateral transport and it has been shown that the basolateral route is used, perhaps exclusively, by newly formed enzymes [27]. From the latter route, enzymes partition into lymphatics or directly into the bloodstream [28]. Whereas the NSF/ SNAP-dependent pathway remains in the ‘on’ position in constitutive events, exocytosis is regulated by superimposed ‘fusion clamps’ that are lifted in response to second messenger stimulation. In addition to the well-known messenger systems, recently recognised participants include reactive oxygen species, nitric oxide, free radical oxidation products (FROPs) such as those derived from arachidonic acid, and the related substance, platelet-activating factor (PAF) [29, 30]. Assembly of the SNARE apparatus at the cytoplasmic leaflet of the apical membrane and formation of the exocytosis pore mark completion of a successful round of signal transduction, but there are other factors that play crucial roles in enzyme discharge and/or in the retrieval of zymogen granule membranes from the centro-acinar space by endocytosis. (1) The methylation of membrane components, in particular of a prenylated cysteine residue [31], appears to be a prerequisite for exocytosis and depends upon the methyl donors S-adenosylmethionine (SAMe) and choline, which interact with folate and vitamin B12, betaine, ATP and homocysteine in the methio-

nine transsulphuration pathway (fig. 2). (2) As in other cells [32], GSH derived from this pathway is another absolute requirement [33], not least by sparing critical protein thiols [34], one or other of which may qualify as the elusive NEM-sensitive fusion factor in the exocytosis machinery [35]. (3) It is increasingly evident that GP-2 (a glycosyl phosphatidyl inositol-linked protein) anchored to the apical plasma membrane, plays a critical role in coupling exocytosis of zymogen granules to endocytosis of shed granule membranes [36]. (4) There is increasing evidence too that the cystic fibrosis transmembrane conductance regulator protein (CFTR) is involved in vesicle transport and membrane recycling, possibly by a direct action in facilitating exocytosis in the pancreatic acinar cell [37] where it is now known to be present [38], and also indirectly, by mediating the transfer of bicarbonate across the intralobular ductules, overcoming the natural acidity of the intra-acinar space [39]. Alkalinization ensures the

Aetiogenesis of Chronic Pancreatitis

Digestion 1998;59(suppl 4):1–12

Fig. 2. The intracellular methionine transsulphuration pathway: SAMe = sulphadenosylmethione; MTA = methylthioadenosine; SAH = sulfadenosylhomocysteine; GSH = glutathione in its reduced bioactive form; GSH.Px = glutathione peroxidase; GSSG = glutathione in its oxidised form in combination with biological oxidants; GSSR = glutathione in combination with xenobiotic metabolites; GSH.Rx = glutathione reductase; Glu-6-PO4 = glucose-6-phosphate dehydrogenase catalysed actions in the pentose phosphate shunt. Modified from Braganza [8].

3

Fig. 3. Inhibition of trypsin by dithiothreitol followed by biphasic reactivation and subsequent inhibition by incremental additions of sodium periodate (NaIO4). The results are expressed as a percentage of the initial trypsin activity indicated by the arrow labelled A; the arrow at B indicates the initial inhibition of trypsin caused by addition of dithiothreitol; the line BC represents the reactivation of trypsin and the curve CDE the subsequent inhibition of trypsin by incremental additions of the oxidant. Redrawn from Steven and Al-Habib [43] and reproduced with kind permission of the author and publishers.

solubility of secreted enzymes and enables GP-2 to be detached from the lumenal aspect of the apical membrane, so that shed zymogen membranes can re-enter the cell by endocytosis [39, 40]. Lysosomal enzymes are incompletely separated from food digestive enzymes within the Golgi apparatus, and therefore small amounts are found in normal pancreatic juice [41]. The premature activation of trypsinogen by lysosomal cathepsin B within the secretory pathway is prevented by the co-production of pancreatic secretory trypsin inhibitor in an amount that can inhibit 20% of potential tryptic activity; should this capacity be exceeded, trypsin quickly activates two proteases, mesotrypsin and enzyme Y, that immediately degrade both trypsin and zymogens. This activation-to-degradation sequence is also effected by lysosomal enzymes within crinophagic vacuoles that appear in adverse circumstances [41]. Should leaky lysosomal membranes allow active trypsin to enter the cytoplasm, it would be immediately inhibited by GSH in a non-stoichometric reversible reaction involving thiol-disulphide exchange [42–44]. We have confirmed that 1 mM GSH, which is the lower limit of the

4

Digestion 1998;59(suppl 4):1–12

concentration that is expected in the cytosol of acinar cells [45], immediately inhibits 20 ÌM trypsin which is in excess of that expected in zymogen granules [Chaloner and Braganza, unpubl.]. This crucial mechanism of serine protease inhibition, first described some 30 years ago [42], and further elucidated by Steven and co-workers at Manchester, UK [19, 43, 44], has not hitherto been explicitly stated in the context of pancreatitis. These investigators went on to show that the thiol-inhibited trypsin could be readily re-activated by a variety of oxidants, including sodium periodate (fig. 3), mercury compounds and cystine, the oxidised form of cysteine. However, the use of high concentrations resulted in irreversible cleavage of the significant disulphide bond with loss of trypsin activity. Should active trypsin overcome this further layer of defence and enter the interstitium, ·1-proteinase inhibitor is available; should trypsin enter the bloodstream, ·2macroglobulin is a further deterrent. The degradation of trypsinogen within crinophagic vacuoles can be tracked by measuring its activation peptide in tissue homogenates or acinar cells in culture [46], but the appearance of the peptide in serum or urine during experimental pancreatitis seems to signify activation within the interstitium [46, 47]. Drug Metabolism It is now established that the normal pancreas contains a dormant machinery for xenobiotic metabolism, involving phase I cytochrome P450 (CYP) biotransforming enzymes [48] and phase II conjugation reactions that often involve GSH [12, 24]. Exposure to a xenobiotic results in the phenomenon of ‘enzyme induction’ which is brought about by a rapid increase in cellular protein and phosphatidylcholine with increased activity of mitochondrial heme synthetase, so that additional amounts of heme and also of heme oxygenase are produced; the former is an integral part of CYP, the latter not only catalyses the formation of bilirubin but also belongs to the family of heat shock proteins that is now known to buttress antioxidant defence [49]. Ancillary proteins such as the glutathione-S-transferases B which bind anions are also induced, but not necessarily in parallel. Expansion of the endoplasmic reticulum membranes may accompany these phenomena. Enzyme induction is facilitated by diets that are rich in linoleic acid; sources include corn, linseed and peanut oil. The enzyme glycine N-methyltransferase of the methionine transsulphuration pathway (fig. 2) apparently serves as a mediator of CYP1A1 gene induction [50].

Braganza

The metabolism of xenobiotics by CYP usually results in their detoxification, but many drugs and chemicals inadvertently undergo ‘bioactivation’, threatening the viability of cells when phase II conjugating systems are swamped. Clearly, the yield of oxidants would be higher if the relevant CYP isoenzymes were to be already induced or if GSH stores were already depleted. Paracetamol poisoning is a case in point. Ultrastructural clues to toxic metabolite strain include plasma membrane blebbing, dilatation of endoplasmic reticulum membranes, accumulation of products such as fatty globules, cytoskeletal disorganisation, and an excess of lipofuscin indicating membrane lipid peroxidation [12, 51]. Persistent oxidant strain may be followed by phenotypic transdifferentiation [52], or redifferentiation into tubular complexes [52, 53]. The pancreas has a high turnover of GSH [54] and it actively synthesises this tripeptide [55], for its aforementioned secretory and protease inhibitory roles, and also as an antioxidant, removing hydrogen and lipid peroxides and xenobiotic electrophiles. In this antioxidant capacity, it is just one aspect of a sophisticated defence strategy [56].

The Disease Aetiology and Aspects of Pathobiology As with any chronic inflammatory process, the defining triad on histology is parenchymal loss, fibrosis, and an infiltrate of lymphocytes with some plasma cells and eosinophils, but in this instance with the proviso that the inflammatory component disappears when all acini are effaced [14]. Ductal lesions are inconstant [14]. Chronic pancreatitis may occur in numerous settings, but in practice these reduce to cystic fibrosis, alcoholism, miscellaneous associations among which vasculitis is notable, and idiopathic disease. The autosomal recessive disease cystic fibrosis, due to mutations in the CFTR gene at 7q31, invariably causes chronic pancreatitis-like lesions [14] that begin in utero [57]. The importance of a full quota of CFTR protein for pancreatic integrity is shown by the finding of a 2.5-fold increase over the expected gene carrier frequency among patients with idiopathic chronic pancreatitis [58]. By contrast, the autosomal-dominant trait associated with hereditary pancreatitis displays the phenomenon of ‘variable penetrance’: a mutation in the gene for cationic trypsinogen at 7q35 has recently been identified and it has been deduced that the R117H mutation results in a form of trypsin that resists degradation [59]. Preliminary studies do not show an

Aetiogenesis of Chronic Pancreatitis

increased mutation rate of this gene among cohorts with the much commoner acquired form of chronic pancreatitis [60]. In developed countries chronic pancreatitis is often equated with alcoholism, but alcohol is actually a weak risk factor on its own in that an average of 15 years usually elapses before the first symptom in men who drink 6150 g daily [3]. In any series 20–30% of cases are idiopathic; the proportion is higher among children and elderly people and may reach 95% in tropical zones of the Far East. However, this categorisation does not take into account the suspected importance of dietary cyanogenic glycosides in tropical pancreatitis [61], or the proposed autoimmune connection in the uncommon duct destructive sclerosing variant [62], or of the identification of smoke constituents, volatile petrochemical products and insufficiency of micronutrient antioxidants as independent risk factors, irrespective of geography [63–69]. The natural history of chronic pancreatitis is well known, dominated by ‘attacks’ that are indistinguishable from acute pancreatitis, and mounting background pain. The latter problem is best explained by an increase in parenchymal fluid pressure [70] and the expression of tachykinins, substance P and calcitonin gene-related peptide (CGRP), within nociceptive nerve fibres that display breached perineurium and cytotoxic T lymphocytes in the vicinity [71]. Any framework for the aetiogenesis of the disease should also rationalise trends shown by hospital admission statistics: a higher rate among African Americans than in other ethnic groups in the USA [72]; the steady increase in the 25-year period from around 1955 in developed countries, which is generally ascribed to alcoholism but can also be seen to follow by some 10 years a marked increase in consumption of linoleic acid [73]; and the steady fall-off in admission rate at Kerala, South India, apparently without a change in the consumption of cassava (manioc, tapioca), a dietary staple that is rich in cyanogenic glycosides [74]. Theories on Pathogenesis These have been reviewed [1, 3, 7, 8]. Four theories centre on the duct system, namely, primary ductal lithiasis, stagnation leading to intraductal calculi, primary autoimmune attack on ducts, and reflux via a common ampullary channel. These concepts are inherently flawed because ductal lesions are inconstant [14]. There is a school of thought that regards recurrent fat necrosis [17] and recurrent autodigestion episodes [59] as the basis for chronic pancreatitis, but this perception would not rationalise the biphasic pattern of acinar cell and pancreatic

Digestion 1998;59(suppl 4):1–12

5

juice disturbances in chronic pancreatitis, an early phase of hypertrophy associated with increased enzyme, GP-2 and calcium secretion, followed by atrophy and exocrine secretory impairment [8]. The same criticisms can be levied against the other three theories [8], of which one focuses on chronic ischaemia, another on toxic metabolites from alcohol in the acinar cell, and the last on cyanide toxicity.

Chronic Pancreatitis: Casualty of Detoxification Reactions In this concept, oxidant stress in pancreatic acinar cells is regarded as the ‘necessary but insufficient’ intermediate phenotype, in Koch’s terminology [75], with the qualifier that the problem usually results from three interacting factors, namely, CYP induction [63, 76], concurrent exposure to a chemical that undergoes bioactivation [65– 67], and, above all, insufficiency of micronutrients that are required to sustain GSH stores [8]. The qualifying clauses help to explain why patients on anticonvulsant drugs rarely develop the disease [6, 7], why isolated selenium deficiency causes only fibrosis [6, 7], and why profound oxidant strain, but with reduced CYP activity in kwashiorkor results in non-inflammatory acinar effacement [77]. It follows that if the diagnosis of chronic pancreatitis is indisputable in any of these settings, some other explanation is likely to be forthcoming [67, 78]. The concept does allow for a slow build-up of reactive oxygen species (ROS) alone, for example, in elderly individuals whose antioxidant intake falls, or in patients with vasculitis, in whom ischaemia-reperfusion injury may increase oxidant load. In this latter setting, the ability of ROS to derange the structure, and hence the immunogenicity of Á-globulin [79] could be relevant to immune-mediated injury [62, 80]. Cytoplasmic vacuolation and heavy lipofuscin deposits, indicating oxidant stress [5, 7, 51, 81], are among the earliest alterations in acinar cells and, not infrequently, hepatocytes show similar changes [5] (fig. 4). The particular susceptibility of the pancreas, shown by the experimental production of injury within a short time of exposure to ethanol or fatty acids [55], whereas liver [5] and kidney [82] involvement generally remain subclinical, is rationalised by the substantially lower concentration and turnover of GSH in the pancreas [54], and also by the inhalation route of xenobiotic entry [65–67]. Interestingly, inhalation exposure to occupational hydrocarbons has been proposed in the development of glomerulone-

6

Digestion 1998;59(suppl 4):1–12

phritis [83]. Further inhalation toxicity could result in CYP induction in islet cells [76] which, in turn, may have a bearing both on the development of diabetes [12] and on the patchy distribution of chronic pancreatitis lesions, in that a ‘portal’ circulation involves some acini but not others. By analogy to microcirculatory territories in the liver (fig. 1), it is also possible that acini in the periphery of the lobule are particularly vulnerable if, as in zone 3 hepatocytes, they are exposed to the lowest oxygen tension, favouring a reductive mode of CYP action [24]. Experimental studies show that toxic metabolite strain causes acinar cells to redifferentiate into tubular complexes, as occurs early in the course of chronic pancreatitis [53]. And, finally, the predilection of the head of the pancreas to the worst lesions of chronic pancreatitis [78, 84] could result from the additional oxidant strain caused by reflux of toxic metabolite-laden bile [5] into the pancreatic duct, as was suggested by pancreatic histology in one report [84]. Within this disease model, bursts of oxidant stress disrupt the methionine transsulphuration pathway (fig. 2) [85, 86], robbing the cell of activated methyl groups [31] and also GSH [34, 85, 86] and thereby jeopardising exocytosis [31–33]. In this respect, it is interesting that GSH seems to mitigate carbon tetrachloride hepatotoxicity by protecting a critical cysteine moiety in SAMe [87]. The blockade to exocytosis triggers a sequence of disturbances to produce the clinical syndrome of a pancreatitis attack [4, 8, 35]. Experimental studies suggest that the acinar cell strives in three ways to compensate for the impedance to secretion at its apical pole: by enhanced crinophagy and autophagy to rid the cell of preformed enzymes in an activation-to-degradation sequence; by diverting newly synthesised enzymes, along with FROPs and PAF, into the interstitium; and by shutting off enzyme synthesis. The entry of FROPs and PAF into the interstitium causes the rapid degranulation of mast cells [16], with the release of products that would quickly elicit the five classical signs of inflammation, not least intense pain due to activation of local axon reflexes with the production of substance P and CGRP. Meanwhile, the injured acinar cell soon becomes a microcosm of the inflammatory response, oxidant strain activating the critical transcription factor NFÎB and, thereby, increasing the synthesis of heat shock proteins, cytokines, and also pancreatitis-associated protein (PAP) which has bacteriostatic potential and shares some homology with the calciumsolubilising protein lithostatin [88]. Should the exocytosis blockade remain in place, because of prior nutritional deficiencies and/or because infiltrating neutrophils add to

Braganza

a

c

b

d Fig. 4. Changes in the pancreas (a, b) and liver (c, d) indicating oxidant stress in patients with chronic pancreatitis. Upper frames show microvesiculation in acinar cells and hepatocytes; lower frames show masses of lipofuscin as black deposits (L) in each cell type [5].

the oxidant load when they are hyperstimulated by cell casts or refluxed bile salts [4, 35, 84], a further wave of inflammation would be expected when the newly synthesised cytokines find their way into the interstitium. The products released from mast cells have the capacity to activate the ‘contact system’, and this would rationalise two very recent findings. First, analysis of admission samples from African patients with a first attack of pancreatitis, among whom a substantial number have underlying chronic pancreatitis, showed profound activation of fibrinolysis that was not dependent upon the presence of active trypsin or active thrombin [89]. Second, an immunocytochemical study of surgical specimens from patients with necrotizing pancreatitis showed enhanced expression of urokinase plasminogen activator receptor and the plasmin system [90], as had previously been

shown in a study of established chronic pancreatitis [20]. There is no evidence of pathological trypsin activation at the inception of a pancreatitis attack, whether clinical or experimental. However, lysosomal membranes would be expected to become leaky in the oxidant environment of the acinar cell. When this happens, any active trypsin that has not yet been fully degraded in lysosomes could enter the cytosol but would only remain active if GSH is substantially depleted. However, trypsin should not remain active for long because, in vitro, persistent oxidant strain deactivates the enzyme irreversibly (fig. 3). Extreme oxidant strain disrupts mitochondrial membrane potential and depletes ATP to cause cell death by necrosis, rather than apoptosis [91], augmenting an already aberrant inflammatory response that is being driven by free radicals [35].

Aetiogenesis of Chronic Pancreatitis

Digestion 1998;59(suppl 4):1–12

7

Fig. 5. Schematic representation of the evolution of calcifying chronic pancreatitis. The pyramidal structures represent acinar cells during the development of the disease. Square brackets indicate concentrations of substances. Circles with a plus symbol indicate increases and circles with a minus symbol indicate decreases. HCO3 = bicarbonate; FRA = free radical activity; HSP = heat shock proteins; PAR-2 = G-protein coupled receptor at plasma membrane, that

is cleaved by pancreatic trypsin and mast cell tryptase; UPA-R = urokinase plasminogen activator-receptor; U-PA = urokinase plasminogen activator; TT-PA = tissue-type plasminogen activator characterised by cleavage of the synthetic substrate benzoatase; SP = substance P; CGRP = calcitonin gene-related peptide; PAP = pancreatitis-associated protein; PSP = pancreatic stone protein also called lithostatin; E = enzymes; Ca++ = calcium.

CYP induction is envisaged as an important factor in the development of ‘large-duct’ calcifying disease [8], because the expanded endoplasmic reticulum in response to CYP2 induction would result in protein and GP-2 hypersecretion [40] in the early stage of the disease, when the exocytosis machinery recovers between bursts of free radical activity. The precipitation of protein in the duct system is encouraged by the dissociation between exocytosis and GP-2 facilitated endocytosis of shed granule membranes [36] and by hypersecretion of lactoferrin. This iron-trapping antioxidant is synthesised by acinar cells, perhaps to compensate for micronutrient antioxidant insufficiency. However, it is a potent chemotaxin, explaining neutrophil influx in large-duct disease [17], including the duct-destructive autoimmune variant [62]. In the natural progression of the inflammatory response, activation of the innate arc is soon followed by activation of the immune arc and the latter becomes increasingly dominant when the inflammatory stimulus persists. This sequence rationalises the HLA display [21] with secretion of immunoglobulin A [22] and other hallmarks of an immune response [80]. The tendency for protein aggregation would be compounded by hypersecretion of another antioxidant, mucin, and also by reduced production of bicarbonate: these features suggest that ductal epithelium

is caught up in the oxidative environment of the gland, and also that the CFTR protein may be a free radical target, not only in bicarbonate-secreting cells [39] but in the acinar cell itself [37, 38, 58]. Provided that patients’ lifestyles do not change, it would be expected that methyl groups and GSH in acinar cells would be steadily depleted (fig. 2), so that any further blast of free radical activity, whether caused by an alcoholic binge or exposure to volatile hydrocarbons [65–67], would more readily precipitate an attack and, furthermore, that attacks would become more frequent as time goes by (fig. 5). With each episode, FROPs and PAF would enter the interstitium, causing mast cells to degranulate, causing chemotaxis, with activation of collagenases and also of plasminogen activators at the cell surface, by thiol-disulphide exchange [19, 42–44], to facilitate tissue remodelling. It has recently been shown that proteinase activated receptor-2 (PAR-2), a G-protein coupled receptor that is cleaved to expose a tethered ligand, is readily activated by mast cell tryptase or by trypsin [92], highlighting the shared substrate specificities of these two enzymes and suggesting to me that tryptase may be the true activator of trypsinogen within the pancreatic interstitium during experimental or clinical pancreatitis (fig. 5). PAR-2 activation may represent a further attempt

8

Digestion 1998;59(suppl 4):1–12

Braganza

It has been recommended that any hypothesis for a causal connection in a polygenic disease should fulfil a set of postulates derived from Koch’s classical work on tuberculosis [75]. The proposal that oxidant stress in acinar cells is the ‘obligate intermediate phenotype’ in chronic pancreatitis largely fulfils the criteria [8]. (1) The relationship is mechanistically plausible: The scheme in figure 5 represents the evolution of large-duct calcifying disease. CYP-mediated oxidant stress without prior enzyme induction or prior antioxidant insufficiency leads to small-duct disease, as shown by animal models (discussed below). (2) Oxidant stress precedes the disease: This evidence has come from studies of outwardly healthy people in

Soweto, South Africa, where alcoholic disease predominates, and areas of South India where dietary nitriles and volatile petrochemical products are suspected risk factors. In the last context, it is interesting to note that the fall in incidence of chronic pancreatitis among the highly literate population of Kerala, South India, has occurred in parallel with electrification of the province, reducing inhalation exposure to kerosene fumes from lamps. Many years ago it was noted that duodenal aspirates from apparently healthy children in this province contained high concentrations of lactoferrin, as high as in alcoholics with chronic pancreatitis at Marseilles in France [96]. Recent studies in the adjacent province of Andhra Pradesh [65] suggest that lactoferrin hypersecretion may represent an attempt to compensate for the low bioavailability of vitamin C, resulting from hostile culinary practices. Similarly, heightened lipid peroxidation, found by serum analysis, in ‘healthy’ adults at Soweto was accompanied by negligible levels of vitamin C, due to inaffordability of fruit [69]. Further, studies of asymptomatic chronic alcoholics in the USA, Europe, the UK and South Africa show a milder degree of the same biochemical disturbances as occur in patients with alcoholic chronic pancreatitis and which can be rationalised in terms of oxidant stress; for example, increased lactoferrin in duodenal aspirates [96] and an increase in free radical activity with reduction in blood antioxidants [69]. (3) Conventional treatment does not ameliorate oxidant strain: This evidence has been published, from studies of serum and also duodenal aspirates [97]. (4) Antioxidant supplements should ameliorate symptoms: Placebo-controlled trials [86] and an audit of longterm treatment at Manchester [98] fulfils this clause, findings endorsed in preliminary observations at Soweto (Segal, pers. commun.) and Kerala (Pai, pers. commun.) and by a 12-month therapeutic study from the Czech Republic [99]. In this context, the failure of allopurinol therapy in a report from the USA is hardly surprising, as it has no impact on xenobiotic metabolite damage via CYP, and does not justify the national society denial of an antioxidant trial [100]. The importance of oxidants and interstitial mast cells in the progression of chronic pancreatitis is further supported by the ameliorating effect of high-dose vitamin E (·-tocopherol) on liver fibrosis [95] and in patients with recurrent pancreatitis by mast cell stabilisers [101, 102], respectively. (5) Disease genes should co-segregate with oxidant stress-facilitating genes: Many candidate genes are suggested by the new model [8], for example, CFTR [58]. The familial clustering of cases in tropical zones, where cystic

Aetiogenesis of Chronic Pancreatitis

Digestion 1998;59(suppl 4):1–12

by the acinar cell to overcome the exocytosis blockade, in that the tethered ligand has been shown to elicit a prompt increase in ionised calcium in dispersed acini [92]. A key role for mast cells in the progression of chronic pancreatitis is further underlined by a very recent study describing the apparent paradox that the expression of preprotachykinin-A gene was no different in normal and chronic pancreatitis tissue [93], whereas substance P expression is clearly elevated in nociceptive nerve endings [71]: that is exactly what would be expected from mast cell-elicited axon reflexes. The juxtaposition of mast cells [16], adipocytes [16, 17] including precursor stellate-myofibroblasts [18], two classes of plasminogen activator [19, 20], and activated inflammatory cells including resident macrophages with infiltrating neutrophils and activated T lymphocytes is poised to favour fibrosis under conditions of oxidant strain [94, 95]. There is currently a flurry of interest in molecular mechanisms of pancreatic fibrosis, but space constraints preclude coverage of this burgeoning field in the present review. Suffice to say that the findings to date could be predicted from what is known of fibrosis in liver disease [94, 95]. Finally, the oxidant stress template for the evolution of chronic pancreatitis rationalises the enhanced risk of pancreatic cancer and also the peculiar vulnerability of African Americans to both diseases [7, 72]: in regard to the former connection, certain xenobiotic electrophiles are known to be pro-carcinogenic, whereas the latter connection may reflect the higher intrinsic activity of certain CYP isoenzymes in other Black races, exemplifying the principle of genetic polymorphism in CYP function.

The New Koch’s Postulates

9

fibrosis is rare, does not exclude the possibility that one or other amino acid residue, such as phenylalanine, in the CFTR protein is a free radical target in the same way as oxidant attack on a critical methionine residue inactivates ·1 proteinase inhibitor. However, the recognised polymorphism in CYP function indicates the need for a molecular epidemiology study of CYP isoenzymes in these communities of the Far East and South Africa. Other candidate genes are those for heme oxygenase, GP-2, GSH and linked enzymes [8], lithostatin and PAP. It may well be that a mutation in the cationic trypsinogen gene requires GSH depletion in order to express its diseaseconferring potential, in that low antioxidant status identified vulnerable members among hereditary pancreatitis kindreds [103]. (6) It should be possible to produce an animal model of chronic pancreatitis by inducing oxidant stress: This integral aspect of Koch’s classical work has been dropped from the new recommendations, but it is of interest that dietary deprivation of both choline and methionine (fig. 2), or exposure to carbon tetrachloride when combined with low dietary methionine [13], or combined treatment with a high corn oil diet and a CYP inducer [104, 105] or an injection of the organotin dibutyltin di-

chloride [106] produce small duct disease. These protocols share the property of inciting oxidant stress. It is likely that a period of prior enzyme induction or antioxidant insufficiency will generate a model of large-duct disease.

Concluding Comments Chronic pancreatitis emerges as a paradigm of a disease that is triggered by disruption of a critical biochemical pathway, in this case the methionine transsulphuration route that is crucial for signal transduction in the acinar cell. New aspects that should be urgently investigated include the key role envisaged for mast cells (fig. 5) and the suspected importance of inhalation toxicology. Firstline treatment for chronic pancreatitis by way of antioxidant supplementation has now become a reality [98]; however, it is the prospect for disease prophylaxis that is especially gratifying. Acknowledgements Immense thanks are due to Ms. Tina Knott for expert secretarial help and to Mrs. Sandra Roe for preparing the figures.

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

References 1 Sarles H: Aetiopathogenesis and definition of chronic pancreatitis. Dig Dis Sci 1986;31:91S– 107S. 2 Braganza JM: Pancreatic disease: A casualty of hepatic ‘detoxification’? Lancet 1983; ii:1000– 1003. 3 Braganza J: The pancreas. Recent Adv Gastroenterol 1986;6:251–280. 4 Braganza JM: 1998 Free radicals and pancreatitis; in Rice Evans, Dormandy T (eds): Free Radicals: Chemistry, Pathology and Medicine. London, Richelieu Press, 1988, pp 357–381. 5 Braganza JM: The role of the liver in exocrine pancreatic disease. Int J Pancreatol 1988b; 3(suppl):519–542. 6 Braganza JM (ed): The Pathogenesis of Pancreatitis. Manchester, Manchester University Press, 1991. 7 Braganza JM: Toxicology of the pancreas; in Ballantyne B, Turner P, Marrs TC (eds): Textbook of General and Applied Toxicology. New York, MacMillan, 1993, pp 663–714. 8 Braganza JM: The pathogenesis of chronic pancreatitis. Q J Med 1996;89:243–250. 9 Cawley T: A singular case of diabetes, consisting entirely in the quantity of urine: With an inquiry into the different theories of the disease. Lond Med J 1788;9:286–308.

10

10 London SJ: Beethoven: Case report of a titan’s last crisis. Arch Intern Med 1964;113:442– 448. 11 Friedrich N: Disease of the pancreas; in Ziemssen H (ed): Cyclopaedia of the Practice of Medicine. New York, William Wood, 1878, pp 549–630. 12 Cohen GM (ed): Target Organ Toxicity. Florida, CRC Press, 1986 13 Veghelyi PV, Kemeny TT: Protein metabolism and pancreatic function; in de Reuck AVS, Cameron MP (eds): Ciba Foundation Symposium on the Exocrine Pancreas. London, J & A Churchill, 1962, pp 329–349. 14 Longnecker DS: Pathology of pancreatitis; in Braganza JM (ed): The Pathogenesis of Pancreatitis. Manchester, Manchester University Press, 1991, pp 3–18. 15 Fish EM, Molitoris BA: Alterations in epithelial polarity and the pathogenesis of disease states. N Engl J Med 1994;330:1580–1588. 16 Hollender LF, Lehnert P, Wanke M: Acute Pancreatitis. Baltimore, Urban & Schwarzenberg, 1983. 17 Kloppel G, Maillet B: Chronic pancreatitis: Evolution of the disease. Hepatogastroenterology 1991;38:408–412.

Digestion 1998;59(suppl 4):1–12

18 Apte MV, Haber PS, Applegate TL, Norton ID, McCaughan, Korsten MA, Pirola RC, Wilson JS: Periacinar stellate shaped cells in rat pancreas: Identification, isolation, and culture. Gut 1998;43:128–133. 19 Steven FS, Benbow EW: Differential inhibition of a cell surface protease on normal and tumour cells. Anticancer Res 1992;12:393–398. 20 Freiss H, Cantero D, Graber H, Tang WH, Guo X, Kashiwagi M, Zimmerman A, Gold L, Korc M, Büchler MW: Enhanced urokinase plasminogen activation in chronic pancreatitis suggests a role in its pathogenesis. Gastroenterology 1997;113:904–913. 21 Bedossa P, Bacci J, Lemaigre G, Martin E: Lymphocyte subsets and HLA-DR expression in normal pancreas and chronic pancreatitis. Pancreas 1990;5:415–420. 22 Emmrich, Seyfarth M, Conradi P, Plath F, Sparmann, Lohr, Liebe S: Immunoglobulin A in pancreatic juice and pancreatic tissue of patients with chronic pancreatitis. Gut 1998; 42:436–441. 23 Foulis A K: Histological evidence of initiating factors in acute necrotising pancreatitis in man. J Clin Pathol 1980;33:1125–1131. 24 Timbrell JA: Principles of Biochemical Toxicology, 2nd edition. London, Taylor & Francis, 1991.

Braganza

25 Rothman JE: Mechanisms of intracellular protein transport. Nature 1994;372:55–63. 26 Arvan P, Castle JD: Phasic release of newly synthesized secretory proteins in the unstimulated rat exocrine pancreas. J Cell Biol 1987; 104:243–252. 27 Cook LJ, Musa OA, Case RM: Intracellular transport of pancreatic enzymes. Scand J Gastroenterol 1996;31(suppl 219:)1–5. 28 Anderson RJL, Braganza JM, Case RM: Routes of protein secretion in the isolated perfused cat pancreas. Pancreas 1990;5:394–400. 29 Van Der Vliet A, Bast A: Effect of oxidative stress on receptors and signal transmission. Chem-Biol Interact 1992;85:95–116. 30 Nigam S, Kunkel G, Prescott SM (eds): Platelet-Activating Factor and Related Lipid Mediators 2. New York, Plenum Press, 1996. 31 Capdevila A, Decha-Umphai W, Song K-H, Borchardt RT, Wagner C: Pancreatic exocrine secretion is blocked by inhibitors of methylation. Arch Biochem Biophys 1997;345:47– 55. 32 Brown LAS: Glutathione protects signal transduction in type II cells under oxidant stress. Am J Physiol 1994;266:L172–L177. 33 Stenson WF, Lobos E, Wedner HJ: Glutathione depletion inhibits amylase release in guinea pig pancreatic acini. Am J Physiol 1983; 244:G273–G277. 34 Lüthen RE, Grendell JH: Thiol metabolism in acute pancreatitis: Trying to make the pieces fit. Gastroenterology 1994;107:888–892. 35 Braganza JM, Chaloner C: Acute pancreatitis. Curr Opin Anaesthesiol 1995;8:126–131. 36 Tenner S, Freedman SD: Chronic ethanol administration selectively impairs endocytosis in the rat exocrine pancreas. Pancreas 1998;17: 127–133. 37 Quesnel LB, Jaran AS, Braganza JM: Antibiotic accumulation and membrane trafficking in cystic fibrosis cells. J Antimicrob Chemother 1998;41:215–221. 38 Kopelman H, Ferretti E, Gauthier C, Goodyear PR: Rabbit pancreatic acini express CFTR as a cAMP activated chloride efflux pathway. Am J Physiol 1995;269:C626–C631. 39 Scheele GA, Fukuoka SI, Kern HF, Freedman SD: Pancreatic dysfunction in cystic fibrosis occurs as a result of impairments in luminal pH, apical trafficking of zymogen granule membranes and solubilization of secretory enzymes. Pancreas 1996;12:1–9. 40 Freedman SD, Sakamoto K, Venu RP: GP2, the homologue to the renal cast protein uromodulin, is a major component of intraductal plugs in chronic pancreatitis. J Clin Invest 1993;92:83–90. 41 Rinderknecht H: Pancreatic secretory enzymes; in Go VWL, Di Magno EP, Gardner JD, Lebenthal E, Reber HA, Scheele GA (eds): The Pancreas, Biology, Pathobiology, and Disease. New York, Raven Press, 1993, pp 219– 252. 42 Neurath H, Walsh KA, Winter WP: Evolution of structure and function of proteases. Science 1967;158:1638–1644.

Aetiogenesis of Chronic Pancreatitis

43 Steven FS, Al-Habib A: Inhibition of trypsin and chymotrypsin by thiols: Biphasic kinetics of reactivation and inhibition induced by sodium periodate addition. Biochim Biophys Acta 1979;568:408–415. 44 Steven FS, Griffin MM: Studies on the molecular mechanism of mersalyl and 4-aminophenylmercuric acetate re-activation of trypsin-thiol complexes. Eur J Biochem 1980;109:567–573. 45 Hwang C, Sinskey AJ, Lodish HF: Oxidised redox state of glutathione in the endoplasmic reticulum. Science 1992;257:1496–1502. 46 Mithöfer K, Fernandez del Castillo C, Rattner D, Warshaw AL: Subcellular kinetics of early trypsinogen activation in acute rodent pancreatitis. Am J Physiol 1988;274:G71–G79. 47 Fernandez-del Castillo C, Schmidt J, Warshaw AL, Rattner DW: Interstitial protease activation is the central event in progression to necrotizing pancreatitis. Surgery 1994;116:497–504. 48 Guengerich FP: Characterisation of human microsomal cytochrome P-450 enzymes. Annu Rev Pharmacol Toxicol 1989;29:241–264. 49 Stocker R: Induction of heme-oxygenase as a defence against oxidative stress. Free Rad Res Commun 1990;9:101–112. 50 Raha A, Joyce J, Gusky S, Bresnick E: Glycine N-methyltransferase is a mediator of cytochrome P4501A1 gene expression. Arch Biochem Biophys 1995;322:395–404. 51 Orrenius, S, Ormstad K, Thor H, Jewell SA: Turnover and functions of glutathione studied with isolated hepatic and renal cells. Fed Proc 1983;42:3177–3188. 52 Scarpelli DG: Toxicology of the pancreas. Toxicol Appl Pharmacol 1989;101:543–554. 53 Bockman DE: Toward understanding pancreatic disease: From architecture to cell signalling. Pancreas 1995;11:324–329. 54 Githens S: Glutathione metabolism in the pancreas compared with that in the liver, kidney and small intestine. Int J Pancreatol 1991;8: 97–109. 55 Neuschwander-Tetri BA, Presti ME, Wells LD: Glutathione synthesis in the exocrine pancreas. Pancreas 1997;14:342–349. 56 Gutteridge JMC: Lipid peroxidation and antioxidants as biomarkers of tissue damage. Clin Chem 1995;41:1819–1828. 57 Stein AA, Porta E, Powers S, Leather R, Linton P, Patterson P: Studies on surgical biopsies of pancreas and liver in four cases of cystic fibrosis. Ann Surg 1963;157:516–524. 58 Sharer N, Schwarz M, Malone G, Howarth A, Painter J, Super M, Braganza JM: Mutations of the cystic fibrosis gene in patients with chronic pancreatitis. N Engl J Med 1998;339:645–652. 59 Gorry, MC, Gabbaizedeh D, Furey W, Gates LK Jr, Preston RA, Aston CE, Zhang Y, Ulrich C, Ehrlich GD, Whitcomb DC: Mutations in the cationic trypsinogen gene are associated with recurrent acute and chronic pancreatitis. Gastroenterology 1997;113:1063–1068. 60 Keim V, Teich N, Reich A, Mössner J: Trypsinogen mutations of exons 2 and 3 in hereditary and chronic-alcoholic pancreatitis (abstract). Gastroenterology 1998;114(suppl): A473.

61 McMillan DE, Geervarghese PH: Dietary cyanide and tropical malnutrition diabetes. Diabetes Care 1979;2:202–208. 62 Ectors N, Maillet B, Aerts R, Geboes K, Donner A, Borchard F, Lankisch P, Stolte M, Lüttges J, Kremer B, Kloppel G: Non-alcoholic duct destructive chronic pancreatitis. Gut 1997;41:263–268. 63 Acheson DWK, Hunt LP, Rose P, Houston JB, Braganza JM: Factors affecting the accelerated clearance of theophylline and antipyrine in patients with exocrine pancreatic disease. Clin Sci 1989;76:377–387. 64 Talamini G, Bassi C, Falconi M, Frulloni L, Di Francesco V, Vaona B, Bovo P, Rigo L, Castagnini A, Angelini G, Vantini I, Pederzoli P, Cavallini G: Cigarette smoking: An independent risk factor in alcoholic pancreatitis. Pancreas 1996;12:131–137. 65 Braganza JM, John S, Padamalayam I, Mohan V, Viswanathan M, Chari S, Madanagopalan M: Xenobiotics and tropical pancreatitis. Int J Pancreatol 1990;7:231–245. 66 McNamee R, Braganza JM, Hogg J, Leck I, Rose P, Cherry N: Occupational exposure to hydrocarbons and chronic pancreatitis: A casereferent study. Occup Environ Med 1994;51: 631–637. 67 Uden S, Acheson DWK, Reeves J, Worthington HV, Hunt LP, Brown S, Braganza JM: Antioxidants, enzyme induction and chronic pancreatitis: A reappraisal following studies in patients on anticonvulsants. Eur J Clin Nutr 1988;42:561–569. 68 Braganza JM, Schofield, D, Snehalatha C, Mohan V: Micronutrient antioxidant status in tropical compared to temperate-zone pancreatitis. Scand J Gastroenterol 1993;28:1098– 1104. 69 Segal, Gut A, Schofield D, Shiel N, Braganza JM: Micronutrient antioxidant status in black South Africans with chronic pancreatitis: Opportunity for prophylaxis. Clin Chim Acta 1995;239:71–79. 70 Ebbehøj N, Borly L, Bulow J, Rasmussen SG, Madsen R: Evaluation of pancreatic tissue fluid pressure and pain in chronic pancreatitis: A longitudinal study. Scand J Gastroenterol 1990;25:462–466. 71 Büchler M, Weihe E, Freiss H, Malfertheiner P, Bockman E, Muller S, Nohr D, Beger HG: Changes in peptidergic innervation in chronic pancreatitis. Pancreas 1992;7:183–192. 72 Lowenfels AB: Epidemiology of diseases of the pancreas: Clues to understanding and preventing pancreatic diseases; in Grendell JH, Forsmark CE (eds): Controversies and Clinical Challenges in Pancreatic Diseases. Bethesda, American Gastroenterological Association, Bethesda, 1998, pp 9–13. 73 Hollander D, Tarnawski A: Dietary essential fatty acids and the decline in peptic ulcer disease: A hypothesis. Gut 1986;27:239–242. 74 Balakrishnan V: Tropical pancreatitis: Epidemiology, pathogenesis and etiology; in Balakrishnan V (ed): Chronic Pancreatitis in India. Trivandrum, St Joseph’s Press, 1987, pp 79– 86.

Digestion 1998;59(suppl 4):1–12

11

75 Korner PI, Swales JD: The role of resistance arteries in the pathogenesis of hypertension; in Mulvany MJ (ed): Resistance Arteries, Structure and Function. Amsterdam, Excerpta Medica, 1991, pp 39–43. 76 Foster JR, Idle JR, Hardwick JP, Barrs R, Scott P, Braganza JM: Induction of drug metabolising enzymes in human pancreatic cancer and chronic pancreatitis. J Pathol 1993;169:457– 463. 77 Golden MHN, Ramdath DD: Free radicals in the pathogenesis of kwashiorkor. Proc Nutr Soc 1987;46:53–68. 78 Sharer NM, Taylor PM, Linaker BD, Gutteridge JMC, Braganza JM: Safe and successful use of vitamin C to treat painful calcific chronic pancreatitis despite iron overload from primary haemochromatosis. Clin Drug Invest 1995;10:310–315. 79 Wickens DG, Norden AG, Lunec J, Dormandy TL: Fluorescence changes in human gamma globulin induced by free radical activity. Biochim Biophys Acta 1983;742:607–616. 80 Cavallini G: Is chronic pancreatitis a primary disease of the pancreatic ducts? A new pathogenic hypothesis. Ital J Gastroenterol 1993;25: 391–396. 81 Day CP, James OFW: Steatohepatitis: A tale of two ‘hits’? Gastroenterology 1998;114:842– 845. 82 Basso D, Panozzo MP, Plebani M, Meggiato T, Zaninotto M, Piccoli A, Fogar P, Ferrara C, Del Favero G, Burlina A: Lipid peroxidation and renal tubular damage in chronic pancreatic diseases: Is there any relationship? J Med 1994; 25:91–104. 83 Pai P, Hindell P, Stevenson A, Mason H, Bell GM: Genetic variants of microsomal metabolism and susceptibility to hydrocarbon-associated glomerulonephritis. Q J Med 1997;90: 693–698. 84 Sandilands D, Jeffrey IJM, Haboubi NY, MacLennan I, Braganza JM: Abnormal drug metabolism in chronic pancreatitis: Treatment with antioxidants. Gastroenterology 1990;98: 766–772. 85 Mårtensson J, Bolin T: Sulfur amino acid metabolism in chronic relapsing pancreatitis. Am J Gastroenterol 1986;81:1179–1184.

12

86 Bilton D, Schofield D, Mei G, Kay PM, Bottiglieri T, Braganza JM: Placebo-controlled trials of antioxidant therapy including S-adenosylmethionine in patients with recurrent non-gallstone pancreatitis. Drug Invest 1994;8:10–20. 87 Corrales F, Giménez A, Alvarez L, Caballerı´a J, Pajares MA, Andreu H, Parés A, Mato JM, Rodés J: S-adenosylmethionine treatment prevents carbon tetrachloride-induced S-adenosylmethionine synthetase inactivation and attenuates liver injury. Hepatology 1992;4:1022– 1027. 88 Ortiz EM, Dusetti NJ, Vasseur S, Malka D, Bodeker H, Dagorn JC, Iovanna JL: The pancreatitis-associated protein is induced by free radicals in AR4-2J cells and confers cell resistance to apoptosis. Gastroenterology 1998; 114:808–816. 89 Chaloner C: Investigation toward a further understanding of pancreatitis and trypsinogen activation. PhD Thesis, The University of Manchester, UK, 1998. 90 Freiss H, Duarte R, Kleeff J, Fukuda A, Tang W-H, Graber H, Schilling M, Zimmerman A, Korc M, Büchler MW: The plasminogen activator/plasmin system is up-regulated after acute necrotizing pancreatitis in human beings. Surgery 1998;124:79–86. 91 Kaplowitz N: Hepatotoxicity of herbal remedies: Insights into the intricacies of plant-animal warfare and cell death. Gastroenterology 1997;113:1408–1412. 92 Glasgow RE, Kong W, Corvera CU, Mulvihill SJ, Bunnett NW: Expression and potential functions of proteinase activated receptor-2 (PAR-2) in pancreatic acinar cells (abstract). Gastroenterology 1998;114:A461. 93 Di Febbo C, Di Mola SF, Baccante G, Porreca E, Innocenti P, Freiss H, Büchler MW: Changes in cytokine but not tachykinin gene expression in patients with chronic pancreatitis (abstract). Gastroenterology 1998;114:A453. 94 Gressner AM: Perisinusoidal lipocytes and fibrogenesis. Gut 1994;35:1331–1333. 95 Houglum K, Venkataramani A, Lyche K, Chojkier M: A pilot study of the effects of d-·-tocopherol on hepatic stellate cell activation in chronic hepatitis C. Gastroenterology 1997; 113:1069–1073.

Digestion 1998;59(suppl 4):1–12

96 Balakrishnan V, Sauniere JF, Hariharan M, Sarles H: Diet, pancreatic function, and chronic pancreatitis in South India and France. Pancreas 1988;3:30–35. 97 Guyan PM, Uden S, Braganza JM: Heightened free radical activity in pancreatitis. Free Radical Biol Med 1990;8:347–354. 98 Whiteley G, Kienle A, Lee S, Taylor P, Schofield D, Braganza JM, McCloy R: Micronutrient antioxidant therapy in the non-surgical management of painful chronic pancreatitis: Long-term observations (abstract). Pancreas 1994;9:A807. 99 Dı´teˇ P, Prˇecechteˇlova´ M, Soˇska V, Lata G: Oxygen radicals and a long-term antioxidant therapy in chronic pancreatitis patients (abstract). Digestion 1998;59(suppl 3):504. 100 Warshaw AL, Banks PA, Ferna´ndez-del Castillo C: AGA technical review: Treatment of pain in chronic pancreatitis. Gastroenterology 1998;115:765–776. 101 Matteo A, Sarles H: Is food allergy a cause of acute pancreatitis: Case report. Pancreas 1990;5:234–257. 102 Hardo PG, Axon ATR: Danazol improves chronic pancreatic pain. J R Soc Med 1993; 86:359. 103 Prasad M, Wyllie R, van Lente F, Steffen RM, Kay MH: Antioxidants in hereditary pancreatitis. Am J Gastroenterol 1996;91: 1558–1562. 104 Tsukamoto H, Towner SJ, Yu GSM, French SW: Potentiation of ethanol-induced pancreatic injury by dietary fat. Am J Pathol 1988;131:246–257. 105 Rutishauser SCB, Ali AE, Jeffrey IJM, Hunt LP, Braganza JM: Towards an animal model of chronic pancreatitis: Pancreatobiliary secretion in hamsters on long-term treatment with chemical inducers of cytochromes P450. Int J Pancreatol 1995;18:117–126. 106 Sparmann G, Merkord J, Jäschke A, Nizze H, Jonas L, Löhr M, Liebe S, Emmrich J: Pancreatic fibrosis in experimental pancreatitis induced by dibutyltin dichloride. Gastroenterology 1997;112:1664–1672.

Braganza

Digestion 1998;59(suppl 4):13–24

Matthew A. Wallig University Department of Veterinary Pathobiology, College of Veterinary Medicine, Urbana, Ill., USA

Xenobiotic Metabolism, Oxidant Stress and Chronic Pancreatitis Focus on Glutathione

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Key Words Xenobiotics Chronic pancreatitis Glutathione Oxidant stress

Abstract Chronic pancreatitis, although relatively rare in the Western World, is common in certain tropical zones where staple crops such as cassava are rich in cyanogenic glycosides. This paper reviews the evidence for a cyanide connection, with reference to experimental studies using another plant nitrile, crambene; and then examines the hypothesis that chronic pancreatitis represents a manifestation of uncoordinated detoxification reactions between pancreatic cytochrome P450 mono-oxygenases and phase II conjugating enzymes, resulting in the irreversible consumption of glutathione in the acinar cell. The conclusion is that the central role of disrupted pancreatic glutathione status, as a result of ‘xenobiotic stress’, in the evolution of chronic pancreatitis cannot be overestimated. This position contrasts with that in acute pancreatitis, in which glutathione depletion has a pivotal role too, but occurs as a result of ‘stress’ from reactive oxygen species. OOOOOOOOOOOOOOOOOOOOOO

Background Chronic pancreatitis is a relentlessly progressive disease that causes untold misery in its victims. It usually presents initially as acute pancreatitis does, with an episode of excruciating abdominal pain and hyperamylaseaemia; however, after full clinical recovery, the pathological sequelae of inflammatory fibrosis and patchy acinar atrophy persist. Further attacks can be anticipated, along with worsening background pain, and there is a propensity to form intraductal calcium carbonate stones and for diabetes to occur, with an increased risk of pancreatic cancer. Treatment is palliative, often necessitating resective pancreatic surgery. While relatively rare in most Western countries, where alcoholism is the dominant aetiological factor, the idiopathic form of the disease is a substantial problem in parts of the developing world, in particular in certain areas of

ABC

© 1998 S. Karger AG, Basel 0012–2823/98/5910–0013$15.00/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

the tropics. In these areas, calcifying chronic pancreatitis often begins in childhood and may affect several family members, characteristics that suggest a genetic predisposition and/or toxicity from commonplace exogenous factors that are nowadays called xenobiotics. This paper examines the hypothesis that ‘oxidant stress’, from altered xenobiotic metabolism rather than from reactive oxygen species alone, depletes glutathione and associated antioxidant defence systems in the pancreatic acinar cell, to set in motion a series of changes leading to chronic pancreatitis [1]. The review begins with a consideration of plant nitrile toxicity because cyanogenic glycosides in dietary staples have been postulated as an aetiological factor in tropical pancreatitis [2, 3], and will then dwell upon the complexities of glutathione (GSH) homeostasis and cytochrome P450 (CYP) function, before considering the implications of oxidant stress for acute pancreatitis [4, 5], and animal models for each disease.

Dr. M. Wallig, DVM, PhD, Diplomate ACVP Department of Veterinary Pathobiology University of Illinois at Urbana-Champaign 2001 South Lincoln Avenue, Urbana, IL 61802 (USA) Tel. +1 (217) 344 8750, Fax +1 (217) 244 7421, E-Mail [email protected]

Toxicity of Plant Nitriles and Relation to Pancreatitis The good geographic match between the occurrence of tropical chronic pancreatitis and the consumption of cassava (tapioca, manioc), a plant rich in the cyanogenic glycosides linamarin and methyl linamarin, formed the basis for the cyanide toxicity theory [2], which was subsequently extended to include alcohol-related disease [3]. The hypothesis is that hydrogen cyanide, released from the plant after ingestion, damages the exocrine pancreas. The theory has been discredited by the lack of a cassava connection in some tropical areas [6, 7], the apparently innocuous nature of cassava in other zones [8], and the declining incidence of chronic pancreatitis in the province of Kerala, South India, in the past two decades [9], although dietary patterns have not changed in this region where the cassava connection was first postulated [2]. Other xenobiotics have now been identified and may be more relevant [7], for example, petrochemical products in fumes from kerosene lamps and cookers, or substances in vehicle emissions; smoke from cigarettes and firewood, and cooking oils high in C18:2 fatty acids. Nevertheless, experimental studies of plant nitrile toxicity provide insights into potential mechanisms of xenobiotic-related damage to the exocrine pancreas. In theory, the toxicity of cyanogenic glycosides may be due to liberated cyanide, toxicity of the intact parent molecule, a non-cyanide product generated when the glycoside is metabolised via CYP, or permutations and combinations of these factors. Cassava varies widely in its cyanide content, depending on whether the leaves, tuber or both are consumed, and on how they are prepared during cooking. Thus, cyanide is released when the plant is macerated, chopped or cut up for food preparation, dissociating from an unstable ‘cyanohydrin’ that forms when endogenous plant ß-glucosidases cleave the glucose during processing [10]. Although much of the cyanide consumed in cassava is already released as a result of food preparation, further cyanide release can occur from linamarin after ingestion. While the long-term effects of chronic cyanide exposure on the central nervous system, thyroid gland and retina are well known and well characterized, to date no direct link between chronic cyanide exposure per se and pancreatitis has been made either in human or in animal studies. However, long-term feeding of cassava itself can produce chronic pancreatitis-like morphological changes in rat and rabbit exocrine pancreas [11, 12]. This paradox suggests either that the non-cyanide moiety is the true pancreatic toxin, or that an additional factor, most

14

Digestion 1998;59(suppl 4):13–24

likely low sulphur amino acid content in the diet or tissue, contributes to damage. The former suggestion is supported by work investigating the toxicities of a series of closely related nitriles, each differing only by the R-groups on the end of the molecule opposite the cyano-group. It was found that the episulfide R-group is associated with nephrotoxicity, an allylic bond is associated with pancreatic toxicity, while no R-group is associated with substantial release of cyanide [13, 14]. The latter suggestion is supported by the need to supplement diets with sulphur amino acids when rats are reared on cassava for long periods [15]. This is in keeping with the detoxifying role of sulphur in converting cyanide to thiocyanate. Animal experiments using the naturally occurring cruciferous compound, 1-cyano-2-hydroxy-3-butene (crambene), were among the earliest investigations to disclose the vital role that GSH plays in determining organ toxicity from plant glycosides in general, cyanogenic or noncyanogenic [13]. Acute administration of crambene to rats causes an early phase of hypersecretion of both pancreatic juice and bile [16], akin to the changes that have been documented in chronic pancreatitis [1, 17]. Morphologically, there is profound oedematous pancreatitis with widespread acinar cell apoptosis in both mice and rats [13, 18, 19] preceded, several hours earlier, by a 80% depletion of GSH [18]. Repeated doses lead to pancreatic atrophy, with a lymphohistiocytic inflammation resembling the early stages of human chronic pancreatitis [13]. The suggestion of a cause-and-effect relationship between GSH depletion and pancreatic damage is upheld by the preservation of pancreatic morphology at doses of crambene producing a less profound depletion of GSH [20] and by protection from damage by caerulein when GSH is induced in the pancreas prior to treatment. Further, in vitro studies have revealed that crambene not only oxidises GSH to GSSG but also cleaves GSH into cysteinylglycine and glutamate via a glutathione S-transferase (GST)-mediated process [21], while studies in rats show a differential effect of the compound on GSH synthesis between pancreas and liver [22].

Glutathione Homeostasis Glutathione, the most important and abundant nonprotein thiol in cells, has multifarious vital functions [23– 28]. It is one of the major cellular and extracellular defences against oxidant stress, being protective against hydrogen peroxide that is produced as a natural consequence of mitochondrial respiration and other physiologi-

Wallig

Fig. 1. GSH synthesis (Á-glutamyl cycle). 5-OP = 5-Oxoprolinase; DP = dipeptidase. If the transported amino acid in the GGT reaction is cysteine (or cystine, which is reduced to cysteine intracellularly), the Á-glutamylcysteine can participate in GSH synthesis by bypassing the GCS step. If the transported amino acid is glycine, it can also participate in GSH synthesis at the GS step. Adapted from Meister [24].

cal processes, against reactive electrophiles that are inadvertently generated when certain xenobiotics are processed via CYP (as discussed below), and against lipid peroxides that may be generated as a result of membrane attack by either of these classes of reactive intermediates. GSH has other functions that impact upon the oxidant status in the cell. It is critically important for the maintenance of intracellular redox status by preserving the appropriate GSH/GSSG ratio, which is far higher in the cytosol (30:1 to 100:1) than within the endoplasmic reticulum and regulated secretory pathway (1:1 to 3:1), where disulphide bond formation is a pre-requisite for proper protein folding [27]. This ratio also ensures the integrity of the cytoskeleton, facilitates signal transduction [29, 30], and maintains normal mitochondrial function [28]. In the exocrine pancreas, with its huge turnover of digestive proteins containing numerous disulphide linkages, the role of GSH in serving as a reservoir and a transport form of cysteine/cystine is also of paramount importance [24, 31]. GSH is a tripeptide (L-Á-glutamyl-L-cysteinylglyceine) synthesised and metabolised by the reactions of the Á-glutamyl cycle (fig. 1). GSH synthesis results from consecutive actions of Á-glutamyl cysteinyl synthetase (GCS) and GSH synthetase, both requiring ATP, with GSH exerting feedback inhibition on the first step. The metabolism of GSH, and also of GSSG and mixed disulphides, is catalysed by Á-glutamyl transpeptidase (GGT), which enables transfer of the Á-glutamyl moiety to a range of acceptors including cysteine, cystine or GSH itself [32]. It is now

clear that cellular turnover of glutathione is associated with its transport, in the form of GSH, out of cells, which explains the preferential location of GSH intracellularly whereas a major fraction of GGT is on the external surface of cell membranes. As GSH moves across cell membranes, Á-glutamyl amino acids are formed and transported into cells, serving as substrates of Á-glutamyl cyclotransferase, which leads to the release of the corresponding amino acids along with 5-oxo-L-proline. This metabolite is then converted to L-glutamate in an ATP-dependent reaction, and the cysteinylglyceine formed is split by dipeptidase, completing the cycle. Two of the enzymes in the cycle also function in the metabolism of S-substituted GSH derivatives that arise either spontaneously, or in reactions catalysed by GST, when GSH forms conjugates with electrophiles that may be generated during the processing of xenobiotics by CYP. It should be noted that, in the absence of significant GGT activity, substantial amounts of GSH appear extracellularly, underlining the participation of GGT in the transmembrane delivery of GSH [24]. As a result of all these interactions, the intracellular level of GSH is maintained in the millimolar range (0.5–10 mM), whereas plasma contains micromolar concentrations. Within the cytosol, the bulk of detoxification of hydrogen peroxide and lipid peroxides is handled by the selenium-dependent enzyme GSH peroxidase, which uses two GSH molecules to reduce hydrogen peroxide to water while converting GSH to GSSG. This oxidised product is rapidly reduced back to GSH by a riboflavin-dependent

Xenobiotic Metabolism and Chronic Pancreatitis

Digestion 1998;59(suppl 4):13–24

15

enzyme, GSH reductase, utilising electrons from the interlinking NADPH-NADP and pentose phosphate shuttles. Thus, both GSH and GSSG concentrations are maintained in the desirable range, until GSSG accumulates to unsatisfactory levels under conditions of oxidative stress, when GSSG is actively excreted [33]. A further protection against lipid peroxides is afforded by certain GST isoenzymes, producing GSSG which is then recycled via GSH reductase. In contrast to the conservation of GSH in reactions involving hydrogen or lipid peroxides, GSH conjugates with xenobiotic electrophiles are rapidly excreted from cells, resulting in the irretrievable loss of both GSH and the cysteine that it contains [24, 25]. Thus, the removal of xenobiotic metabolites by GSH, although vitally protective, is a ‘two-edged sword’ that can result in permanent loss of GSH from the cell, a loss that may not be replaceable in a heavily stressed cell in time to prevent disruption of thiol status. The pancreas has the fourth highest concentration of GSH among body organs, some 2 Ìmol/g tissue [34, 35], representing approximately half the concentration in the kidney and a quarter of that in the liver; and it has the third highest turnover of GSH [34]. It is now known that acinar cells actively synthesize GSH [35]. However, a unique feature, in comparison to liver and kidney, is the absence or very low activity of GCS and low levels of GCS mRNA [22, 34, 35], although GCS is the first and ratelimiting enzyme in the GSH synthetic pathway [24]. In compensation, the pancreatic acinar cell does have the second highest GGT activity in the body, not only on its basolateral and apical membranes, but also internally, on zymogen granule membranes [34]. The high level of GGT activity on the basolateral membrane suggests the enzyme-coupled translocation of GSH and GSSG as well as cysteine and cystine from plasma (fig. 1), obviating the need for complete dependency on GCS [36] and facilitating the efficient transport and recycling of GSH [37]. Another peculiarity of the pancreas is that, despite its high requirement for cysteine, it has, compared to liver, relatively limited activities of enzymes in the transsulphuration pathway down from methionine via cystathionine [38], so that the acinar cell must also depend upon GSH for a continuous supply of this amino acid. This dependence is underlined by the speed with which an injection of GSH monoethylester increases the cysteine concentration within the pancreas [39]. In addition, pancreatic acinar cells have high levels of sulphydryl oxidase, which catalyses the formation of disulphide bonds within proteins, producing hydrogen peroxide as a byproduct [40]. Finally, it is noteworthy that the cholecystokinin agonist,

16

Digestion 1998;59(suppl 4):13–24

caerulein, when given at a physiologic dose of 0.1 Ìg/kg, produces significant depletion of GSH for 8–12 h [41], emphasising its role in signal transduction, but also showing that the gland is in a state of oxidant stress under physiological conditions.

Glutathione Depletion, Reactive Oxygen Species, and Acute Pancreatitis In the past two decades, a body of evidence has accumulated that clearly shows ‘oxidant stress’ to be a common, if not universal, feature of acute pancreatitis. The phrase implies ‘stress’ on the antioxidant defence system in cells, whether due to deficiency of micronutrient sources of antioxidants including GSH, saturation of antioxidant defence systems, or altered production rates of reactive species. At least three lines of evidence support the notion that oxidant stress is causally connected with experimental acute pancreatitis: (1) heightened free radical production; (2) depletion of antioxidant systems; (3) amelioration of pancreatic injury by free radical scavengers [42]. This evidence is germane to the present review because of the clinical identity of a pancreatitis ‘attack’, whether acute pancreatitis or an exacerbation of chronic pancreatitis [1]. The first line of evidence has been adduced by measuring the products of lipid peroxidation in a variety of models [4, 5], including the new dibutyltin model, which demonstrates the transition from classical acute pancreatitis to chronic pancreatitis within just 60 days [43]. However, the direct demonstration of oxygen free radical activity by dye-coupled fluorescence in the caerulein hyperstimulation model which typically causes oedematous pancreatitis [44], by chemiluminescence in this model and also in the retrograde taurocholate model of haemorrhagic pancreatitis [45], and by electron spin resonance spectroscopy in the CDE model [46] provide unequivocal evidence that oxidant stress is involved at a very early stage in the evolution of experimental acute pancreatitis. The depletion of antioxidant systems in pancreatitis has been investigated in the caerulein model [47, 48]. Depletion of non-protein sulphydryls, roughly equivalent to GSH depletion, was followed by an increase in lipid peroxidation products and depletion of protein thiols [47]. The inhibition of GGT by the inhibitor AT-125 has been observed to exacerbate the GSH depletion [41, 48]. Of interest, the degree of GSH depletion is roughly equivalent to that found in crambene-induced oedematous pancreatitis [18]. Recent in vitro studies show how quick-

Wallig

ly GSH is depleted under conditions that more closely simulate human acute pancreatitis, as caused by an acute alcoholic binge or hypertriglyceridaemia [35]. Notwithstanding all these observations, the idea that GSH depletion is the initial step in the development of acute pancreatitis has not gone undisputed; in one study the subcellular redistribution of enzymes into lysosomes with activation of trypsinogen, and also the early histological lesions, preceded the expected drop in GSH concentrations after caerulein administration [49]. The third line of evidence, namely the amelioration of pancreatic injury by free radical scavengers, was first demonstrated more than a decade ago [42]: in an ex vivo model utilising the isolated perfused canine pancreas, wherein injury was initiated by free fatty acid infusion, or partial ductal obstruction in the presence of secretin stimulation, or ischaemia followed by reperfusion, the addition of allupurinol to the perfusate, in order to reduce the production of the superoxide anion by inhibiting xanthine oxidase, resulted in a decrease in interstitial oedema and hyperamylaseaemia. The same group of investigators went on to show that the infusion of acetaldehyde, the first product along the oxidative pathway of ethanol metabolism, was innocuous unless the conversion from xanthine dehydrogenase to xanthine oxidase was first encouraged by a period of ischaemia, whereupon it produced the changes of oedematous pancreatitis, once again indicating the possible involvement of the superoxide anion [50]. The results of later experiments have been reviewed many times over [4, 5, 51]. In general, variable results have been reported following treatment with allopurinol, or the antioxidant enzymes, superoxide dismutase (SOD) and/or catalase, in the caerulein model, with many authors reporting partial improvement in the disturbances of acinar cell vacuolation and hyperamylaseaemia. By contrast, the antioxidants desferrioxamine (which traps iron), dimethyl sulfoxide (which traps the hydroxyl radical), or the selenium compound, ebselen, have only minimal effect when acute haemorrhagic pancreatitis is induced by a choline-deficient ethionine-supplemented (CDE) diet in mice. The same is true for retrograde infusion of taurocholate into the pancreatic duct in rats. Any improvement that has been seen has been in counteracting the inflammatory phase of pancreatitis, not in minimising the disturbances in the pancreatic acinar cell itself [5]. Underlining the critical importance of GSH depletion in the initiation phase of acute pancreatitis, it has now been shown that treatment with glutathione monoethylester, a GSH analogue that is directly taken up by cells, partially abrogates GSH depletion, acinar cell lesions

and hyperamylaseaemia in the hyperstimulation model, whereas a cysteine pro-drug, L-2-oxothiazolidine-4-carboxylate, produces variable results [52]. Zinc supplementation has a similar result, buttressing pancreatic GSH while increasing metallothionine levels [53]. Further, a synthetic analogue of ascorbate has been shown to ameliorate damage in both caerulein and CDE models [54], underlining the close interaction between ascorbate and GSH via redox and non-redox shuttles. Considering all these points, it is perhaps not surprising that the abrupt depletion of GSH seems to be sufficient in itself to initiate the changes of acute pancreatitis, provided that protein thiols are not protected by such protocols as a prior dose of pentamedion [55]. For obvious reasons, it is impossible to prove that GSH depletion, coupled with oxidant stress, underlies the development of acute pancreatitis in man. The finding of high concentrations of lipid peroxidation products and chemiluminescence in admission blood samples [56, 57], some 15 h after disease onset, could be dismissed as being post hoc, especially because the parenteral administration of two GSH precursors, S-adenosylmethionine and Nacetylcysteine, failed to alter outcome in a controlled clinical trial [58]. However, the long-term protection against recurrences of pancreatitis in patients with type I hyperlipidaemia, using a combination of methionine with vitamin C and selenium, offers strong indirect evidence in support of the idea that GSH depletion is a key step in disease development [59]. Table 1 lists some of the ways in which GSH may be depleted in the pancreas, but it should be noted that it is not known exactly how GSH and protein thiol depletion trigger acute pancreatitis, nor is it possible to say with any certainty which of the numerous potential sources of reactive oxygen species contribute to thiol depletion in the human disease.

Xenobiotic Metabolism and Chronic Pancreatitis

Digestion 1998;59(suppl 4):13–24

Glutathione Depletion, Xenobiotic Electrophiles, and Chronic Pancreatitis Although it is possible to produce a degree of pancreatic fibrosis and chronic inflammation by modifying hyperstimulation or CDE protocols for acute pancreatitis, the aetiological factors that are linked with human chronic pancreatitis suggest that in this disease the long-term depletion of GSH and, consequently, the inability to cope with the normal physiological stresses of pancreatic function, result from altered xenobiotic metabolism via CYPs [1]. This hypothesis was first proposed in 1983 [60], and was based initially on the observation that bile from

17

Table 1. Partial list of etiologies of GSH depletion in the pancreas1 Source of depletion

Etiology

Mechanism

Decreased synthesis Sulfur amino acid deficiency Protein deficiency GGT inhibition GCS inhibition

poor diet poor diet AT-125 buthionine sulfoximine, ↓ ATP

lack of substrate (cysteine) lack of substrate (cysteine) ↓ transport of precursors direct inhibition of GCS, lack of co-factor

Enhanced oxidation Oxidant stress (↑ O –2W , ↑ H2O2 )

hypoxia selenium deficiency poor diet poor diet ↑ dietary unsaturated fat high-fat diets ethanol hypoxia dietary and environmental xenobiotics, drugs biliary reflux inflammation vitamin E deficiency

mitochondrial dysfunction, xanthine oxidase activation, producing ↑ O –2W , ↑ H2O2 generation of free radicals via ↑ H2O2 ↑ cytosolic free radicals ↑ membrane free radicals ↑ lipid peroxides, CYP induction ↑ lipid peroxides, CYP induction CYP-mediated ↑ in free radicals ↑ O –2W , ↑ H2O2 generating membrane free radicals direct injury to membranes by toxins or metabolites surfactants, peroxidation products in bile generation of O 2–W , H2O2 by leukocytes ↑ membrane free radicals

hypoxia, etc. ↑ xenobiotic exposure CYP induction leading to ↑ intermediates cholecystokinin stimulation

↑ efflux of excess GSSG from cell covalent binding of GSH to metabolite and efflux of conjugate from cell oxidation during secretion, efflux into juice

↓ GSH peroxidase activity Vitamin C deficiency Vitamin E deficiency ↑ lipid peroxidation

Increased consumption Oxidant stress GSH conjugation ‘Physiologic’ loss 1

Based on studies cited in the text.

patients with pancreatitis contained high concentrations of lipid membrane-derived free radical oxidation products [61]. This was coupled with the realization that many substances which induce hepatic phase I metabolism, including cigarette smoke constituents, certain prescribed drugs, alcohol and high-fat diets, are also important aetiologic factors in pancreatitis [1]. The phase I system is made up predominantly of CYP mono-oxygenases, formerly known as mixed function oxidases. The enzyme exists in numerous forms, mainly in the liver, but also in the kidney, adrenal glands, lung, nasal mucosa and other tissues. CYPs have evolved to metabolise endogenous and exogenous lipophilic substances usually by attaching an activated oxygen or hydroxyl group to the substrate, thereby solubilizing it for excretion or for further metabolism by the phase II detoxification system of predominantly conjugating enzymes. Certain CYPs (e.g. CYP1A1, CYP1A2, CYP2E1 and CYP3A) have the unfortunate tendency to bioactivate xenobiotics, that is to say, to make them more toxic and

18

Digestion 1998;59(suppl 4):13–24

reactive. This problem is exacerbated if the relevant CYP subfamily were to be ‘induced’ to start with. In these circumstances, reactive intermediates, if produced with sufficient rapidity and in sufficient quantity, could overwhelm the capacity of the phase II system to further metabolize the intermediates, via conjugation or reduction reactions. The shortfall allows interactions of the electrophiles with lipids, proteins and nucleic acids to produce adducts or, in many cases, free radical intermediates, thereby causing oxidant stress. CYP 2E1 is especially pertinent because not only is it inducible by ethanol, but also it is known to generate reactive oxygen species from ethanol itself [62, 63]. It stands to reason that the prior antioxidant status of an individual would be a strong factor in determining whether, and to what extent, non-biological intermediates would inflict cellular damage. It is thus of interest that poor antioxidant profiles in blood samples are a feature in outwardly healthy individuals from areas where chronic pancreatitis is endemic, namely, Soweto in South

Wallig

Africa [64] and Madras in South India [65]. These observations suggest that the high concentrations of free radical oxidation products in quiescent-phase samples of serum and duodenal juice [66], and the profound GSH depletion in surgically resected specimens from patients with established disease [5] may indeed be pre hoc phenomena. The initial hypothesis centred on altered xenobiotic metabolism in the liver [1, 17] and pharmacokinetic investigations were designed to test this notion. In a study of patients in England, increased clearances and decreased half-lives of both theophylline and antipyrine were found in the majority, indicating CYPs induction [67]; later, increased theophylline clearance was found in a group of patients with tropical pancreatitis [68] and in a group of African patients with alcoholic disease [69]. However, theophylline clearance values in the controls from the tropical area were substantially lower than in the other two areas, which suggests a role for genetic polymorphism in CYPs expression and hence activities of theophylline-metabolising isoenzymes. These studies were extended to examine the excretion of D-glucaric acid as a marker of phase II metabolism of xenobiotics: levels in patients were higher than in controls [68, 70], which, by inference, indicates enhanced phase I metabolism [70]. Direct evidence was then obtained from an immunocytochemical semi-quantification study of bioactivating CYPs in hepatic biopsy specimens; these studies showed an increase above the normal content of CYP1A2, CYP3A and CYP2E1 not only in hepatocytes, but also in bile duct epithelium [71]. A causative role for increased phase I hepatic metabolism in the evolution of chronic pancreatitis presupposes that reactive intermediates would escape the liver by diffusion or transport into the bloodstream or in refluxed bile in sufficient quantities to be directly toxic to the pancreas. This is very unlikely if the phase II conjugating systems in the liver are working at par, since there are adequate systems such as the GSTs, sulfotransferases and glycuronyl transferases, along with systems to transport the resulting metabolites. The one exception is the byproducts of lipid peroxidation, which can be excreted into bile in forms that retain oxidant activity [17]. The inference that the exocrine pancreas is damaged during the de novo processing of xenobiotics was underlined when the surgical diversion of bile, laden with lipid oxidation products, failed to prevent recurrences of pancreatitis in 3 young patients with idiopathic disease [72]. It has now come to light that the human pancreas contains many of the same CYPs that are responsible for generating reactive intermediates in the liver [71, 73], including CYP2E1, the enzyme that is induced by ethanol [74,

75] and CYP3A, which catalyses the generation of the reactive, carcinogenic metabolite of aflatoxin B1 [76]. Although studies in the rat pancreas show that the activities and content of CYPs of the 1A family (3-methylcholanthrene inducible) and 2E family (ethanol inducible) are !1% and !0.1% that of the liver, respectively [74], both forms are highly inducible in the pancreas, F130fold for 1A1 and 3.6- to 5.1-fold for 2E1 [74, 75]. A similar pattern is present in the hamster pancreas [76]. Given the rich arterial blood supply of the pancreas, it is conceivable that under certain conditions sufficient exposure and uptake of xenobiotics could occur that would result in delivery of a sufficiently large load of reactive metabolites to severely compromise the metabolic and antioxidant defences of the pancreas. This risk would be higher if xenobiotics arrived by the inhalation route, avoiding first pass through the liver. In keeping with this possibility, a case-control study in England identified regular close exposure to volatile hydrocarbons in the occupational environment of patients with chronic pancreatitis, among whom some 50% drank little or no alcohol [77]. A preliminary study in South India has also implicated volatile petrochemical products in the domestic environment [7]. In further support of direct pancreatic toxicity by xenobiotics, studies using pancreatic microsomes or isolated acinar cells show ready activation of both benzo(a)pyrene and heretocyclic amines [78–80]. The information in table 2 facilitates an appreciation of some key points from immunocytochemical localisation studies of pancreatic CYPs [71, 73, 81]. First, the distribution of CYPs between species is very different for some isoenzymes, but similar for others, illustrating the potential pitfalls in extrapolating between data from human and non-human species. Second, all CYP isoenzymes seem to increase upon exposure of experimental animals to chemical inducers and carcinogenic xenobiotics [81, 82]. In humans, two studies have been reported of pancreatic specimens from patients with chronic pancreatitis; in both there was induction of CYP1A and CYP3A [71, 73], with an increase in CYP2E1 in the study in which alcohol was the exclusive aetiological factor [73] but not in the study in which idiopathic disease was equally represented [71]. Third, the inducibility of most CYPs in the pancreas serves to exacerbate any tendency for localised bioactivation of xenobiotics and thus to increase the chance for endogenous antioxidant and xenobiotic defence systems to be overwhelmed. Taken together, these data have ominous implications for the pancreas as a whole, but for the acinar cell in particular, when under ‘xenobiotic stress’.

Xenobiotic Metabolism and Chronic Pancreatitis

Digestion 1998;59(suppl 4):13–24

19

Table 2. Immunolocalization of CYPs in the pancreas1 CYP type

Acinar cells2

Ducts

Islets

1A1

absent/weak in rodents strong in humans

absent/weak in rodents moderate in humans

weak in rodents strong in humans

1A2

moderate in humans

moderate in humans

strong in humans

2E1

strong in rodents strong in humans

weak in rodents strong in humans

strong in rodents weak/moderate in humans

2C

moderate in humans

weak in humans

moderate in humans

3A

weak/moderate in rodents moderate/strong in humans

negative in rats weak/moderate in hamsters moderate in humans

negative in rodents moderate/strong in humans

1

The table is a compilation of results from several studies cited in the text. Absent, weak, etc. refer to the intensity of staining observed in histologic sections of pancreas stained with antibodies against the various CYP forms listed.

2

Table 3. Immunolocalization of GSTs in the pancreas1

GST type

Acinar cells2

Centroacinar/ductules/ ducts

Islets

·GSTs Subunit 1 Subunit 2 ÌGSTs GSTs  GSTs

absent to weak negative to weak absent to weak absent to weak strong

moderate to strong moderate to strong3 moderate to strong4 strong4 strong

absent to weak absent to weak weak moderate5 weak

1

The table is a compilation of results from several studies cited in the text. Absent, weak, etc. refer to the intensity of staining observed in histologic sections of pancreas stained with antibodies against the various GST classes listed. 3 Staining very strong in centroacinar cells. 4 Includes strong staining of duct contents. 5 Positive staining confined to cells at periphery of islet. 2

This potential vulnerability is even more striking when the expression of GST enzymes, the major phase II enzymes dealing with reactive electrophilic intermediates, is examined (table 3) [71, 82–86]. Even though pancreatic GST activity is relatively high in rodents, being 25–50% of that in the liver [84, 85], the distribution of the major classes of GSTs ·, Ì and  is primarily in the duct system or in the lumenal content, which cannot compensate for the relatively weak activity in the acinar cells and islets [81, 82, 85]. The mismatch in distribution between CYPs and GSTs in the acinar cell theoretically puts it at increased risk for damage by self-generated reactive xenobiotic intermediates. In this regard, it is worth noting the relatively low acinar cell activity and content of · GSTs, the class that is usually associated with the high-

20

Digestion 1998;59(suppl 4):13–24

est peroxidase activity [86, 87], again underlining its vulnerability, but now to lipid peroxidative attack. As a final point, and returning to the plant nitrile crambene, it has recently been shown that induction of pancreatic GSTs by subtoxic doses of the nitrile results in a 1.5- to 2.5-fold increase in all GST subunits, but without any change in distribution, so that acinar cells remain isolated and unprotected [88]. It might be thought that this rather bleak picture provided by immunohistochemistry might be mitigated with the identification of the   class of GST, 5-5, an enzyme that protects against epoxide intermediates in all major cell types in the pancreas, including acinar cells [71, 89]. However, in a study of human chronic pancreatitis, which included specimens from patients with pancreatic cancer,

Wallig

Table 4. Summary of characteristics of three potential models of CP Features

Dibutyltin toxicosis

High-fat diet + CYP inducers or ethanol

Zinc toxicosis

Model species

rat

rat, hamster

cow, sheep, pig, chicken, duck

Histopathology Acinar cell degeneration Acinar cell apoptosis/necrosis Acinar atrophy Interstitial inflammation Fibrosis Ductal obstruction

+ + + + + +

+ + + + + +

+ + (sheep, pig, chicken, duck) + (cow, sheep, pig) + (sheep) + (all species) + (sheep)

Biochemical status Enzyme content

?

↓ juice, ↓ protein, ↓ amylase, ↓ lipase, ↓ trypsinogen, ↑ HCO3 ↑ amylase (chicken)

↑ amylase, ↑ lipase

↓ flow, ↓ juice, ↓ protein, ↓ HCO3 ↑ cationic trypsinogen

Oxidant stress Lipid peroxidation Antioxidant enzyme status GSH status

+ ↓ SOD, ↓ GSH peroxidase ?

+ ? ?

+ conflicting results (chicken) ?

Xenobiotic metabolism CYP induction GST induction

? ?

potentiated ?

? ?

Interactions with other factors

ethanol exacerbates

CYP inducers exacerbate damage by xenobiotics (e.g. ethanol)

no effect of supplemental selenium vitamin E

Proposed pathogenesis

bile duct damage and obstruction, direct damage to acinar cells

oxidant and xenobiotic stress potentiated

direct damage to acinar cells

Other organs affected

liver (biliary tree)

none

liver, kidney, blood, adrenal, skin, small intestine

Serum parameters

GST 5-5 was not induced whereas several of the phase I bioactivating enzymes were strongly induced [71]. It is hoped that another recently identified GST, a microsomal membrane-bound isoenzyme (mGST) in the apical cytoplasm of the acinar cell [90] and thought to be specifically protective against lipid peroxidation [91], might be found to mitigate xenobiotic stress. Only time will tell.

New Indicator Models of Chronic Pancreatitis Three models of chronic pancreatitis are now available which faithfully replicate the triad for the pathological diagnosis, namely, patchy acinar cell loss, an infiltrate of chronic inflammatory cells, and fibrosis (table 4). The dibutyltin model is particularly attractive because it mimics virtually every facet of the human disease [1, 72]: an

Xenobiotic Metabolism and Chronic Pancreatitis

initial phase of oedematous acute pancreatitis; independent involvement of the biliary system with changes that are conducive to cholestasis; lesions more severe in the head of the pancreas than in the tail; tubular complexes representing dedifferentiated acini; a transition within 14–21 days from an acute inflammatory infiltrate to the lymphohistiocytic infiltrate characteristic of chronic inflammation; and intense fibrosis by 60 days, indicating imbalance between the deposition of collagen and its removal [43, 92, 93]. It is amazing that all these problems can follow just one dose of an organotin compound that is known to be processed by CYPs, with the production of free radical intermediates [43], such that the activity of GSH peroxidase is decreased as early as one hour after exposure, with a transient depression in SOD activity but increased lipid peroxidation [93]. The further finding that ethanol co-administration exacerbates the morphological

Digestion 1998;59(suppl 4):13–24

21

and biochemical lesions, at doses of ethanol that are innocuous by themselves [94], suggests the involvement of CYP2E1. The widespread use of dibutyltin dichloride, and other members of this organotin family, as plastic stabilisers in the synthesis of polyvinylchloride and polyurethane, and also as pesticides, endorses the suggestion that occupational chemicals may be critically important in the aetiogenesis of human chronic pancreatitis [72, 77]. A second model exploits the facilitatory effect of C18:2 fatty acids on CYPs function [17, 95, 96]. The combination of a high corn oil diet with chronic ethanol consumption caused chronic pancreatitis-like lesions in the rat, whereas excess of any one of these substances alone was ineffective [95]. Studies in hamsters investigated the effect of chemical inducers (phenobarbitone, ß-naphthoflavone) against a background of a corn oil-enriched diet and produced functional and morphological changes in the liver and pancreas that were similar, but not identical, to those associated with human chronic pancreatitis [96]. A third potential model, zinc toxicosis, is well known in veterinary circles, and exemplifies the principle that divalent metals, such as dibutyltin and zinc, depend upon GSH for their detoxification [23]. In all species, inflammatory fibrosis and acinar atrophy have been documented with, in the more intensively studied species, hyperamylaseaemia, depressed concentrations of pancreatic enzymes in the gland and decreased protein and bicarbonate in pancreatic juice [97, 98]: in addition, necrosis and sloughing off of ductular epithelia has been noted in sheep [99].

Concluding Comments

GSH is not only a key antioxidant, but also essential for signal transduction and the delivery of digestive enzymes, which means that glutathione dysregulation would have drastic effects on pancreatic integrity. The corollary that ancillary antioxidant systems, including micronutrients such as vitamin C and antioxidant enzymes such as GSH(selenium)-peroxidase, exert a GSH-sparing facility has already been exploited in developing first line treatment for chronic pancreatitis by way of antioxidant supplementation [100]. The further corollary that a deficiency in any of these accessory pathways would shunt GSH away from its more physiological functions in the pancreas is exemplified in the high incidence of chronic pancreatitis in South Africa and South India [64, 65]. Along the same lines, it is now possible to see why the increasingly poor diets of established alcoholics eventually fail to deliver micronutrients that are required to deal with the increasing oxidant load from induced CYP2E1. In regard to the second component, ‘xenobiotic stress’, the aetiological factors for chronic pancreatitis are now readily rationalised; these include cigarette smoke, volatile hydrocarbons in the occupational or domestic environment and diets rich in C18:2 fatty acids [1]. Unfortunately, the exocrine pancreas is a ‘hostage to fortune’, being a xenobiotic-metabolizing organ in its own right but with the disadvantage of being without the full resources that the liver possesses to deal with any inadvertently bioactivated compounds. The death knell for the pancreatic acinar cell is sounded when these two ‘final common pathways’ come together, in a self-perpetuating cycle of destructive oxidant stress. Acknowledgements

From the evidence reviewed in this paper, the pivotal and interlinking roles of disrupted GSH status and ‘xenobiotic stress’ in the evolution of chronic pancreatitis cannot be overestimated. In regard to the first component,

The author wishes to acknowledge the invaluable assistance of Yvonne Wingfield finding, collecting and collating the information needed for the preparation of the manuscript. The author also wishes to thank Dr. Joan M. Braganza for her valuable suggestions and ‘leads’ in the conceptual development of the manuscript.

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

References 1 Braganza JM: The pathogenesis of chronic pancreatitis. Q J Med 1996;89:243–250. 2 McMillan DE, Geervarghese PH: Dietary cyanide and tropical malnutrition diabetes. Diabetes Care 1979;2:202–208. 3 Pitchumoni CS, Jain NK, Lowenfles AB, DiMagno EP: Chronic cyanide poisoning: Unifying concept for alcoholic and tropical pancreatitis. Pancreas 1988;3:220–222. 4 Braganza JM: Experimental acute pancreatitis. Curr Opin Gastroenterol 1990;6:763–768.

22

5 Schoenberg MH, Birk D, Beger HG: Oxidative stress in acute and chronic pancreatitis. Am J Clin Nutr 1995;62:1306S–1314S. 6 Narendranathan M, Cheriyan A: Lack of association between cassava consumption and tropical pancreatitis syndrome. J Gastroenterol Hepatol 1994;9:282–285. 7 Braganza JM, John S, Padmalayam I, Mohan V, Viswanathan M, Chari S, Madanagopalan M: Xenobiotics and tropical chronic pancreatitis. Int J Pancreatol 1990;5:231–245.

Digestion 1998;59(suppl 4):13–24

8 Abubakare A, Taylor R, Gill GV, Alberti KGMM: Tropical or malnutrition-related diabetes: A real syndrome? Lancet 1986;i:1135– 1138. 9 Geevarghese PJ: Calcific Pancreatitis. Bombay, Varghese Publishing House, 1986. 10 Montgomery RD: Cyanogens; in Liener IE (ed): Toxic Constituents of Plant Foodstuffs. New York, Academic Press, 1980, pp 143– 160.

Wallig

11 Geldof AA, Becking JLF, De Vries CD, Van Der Ven E: Histopathological changes in rat pancreas after fasting and cassava feeding. In Vivo 1992;6:545–552. 12 Shenoy KT, Leena KB, Nair RB: Pancreatic changes (enzymes and histological) in an experimental model fed with cassava-based diet; in: Tropical Tuber Crops: Problems, Prospects and Future Strategies. Lebanon, Science Publishers, 1996, chap 65, pp 458–466. 13 Wallig MA, Gould DH, Fettman MJ: Selective pancreatotoxicity in the rat induced by the naturally occurring plant nitrile, 1-cyano-2-hydroxy-3-butene. Food Chem Toxicol 1988;26: 137–147. 14 Wallig MA, Gould DH, Fettman MJ: Comparative toxicities of the naturally occuring nitrile 1-cyano-3,4-epithiobutene and the synthetic nitrile n-valeronitrile in rats: differences in target organs, metabolism and toxic mechanisms. Food Chem Toxicol 1988;26:149–157. 15 Maner JH, Go´mez G: Implications of cyanide toxicity in animal feeding studies using high cassava rations; in: Chronic Cassava Toxicity: Proceedings of an Interdisciplinary Workshop. London, Int Dev Res Centre Monogr, IDRC010c, 1973, pp 113–120. 16 Maher M, Chernenko G, Barrowman JA: The acute pancreatotoxic effects of the plant nitrile, 1-cyano-2-hydroxy-3-butene. Pancreas 1991;6: 168–174. 17 Braganza JM: The role of the liver in exocrine pancreatic disease. Int J Pancreatol 1988; 3:S19–S42. 18 Wallig MA, Jeffery EH: Enhancement of pancreatic and hepatic glutathione levels in rats during cyanohydroxybutene intoxication. Fundam Appl Toxicol 1990;14:144–159. 19 Bhatia M, Wallig MA, Hofbauer B, Lee H-S, Frossard J-L, Steer ML, Saluja AK: Induction of apoptosis in pancreatic acinar cells reduces the severity of acute pancreatitis. Biochem Biophys Res Commun 1998;246:476–483. 20 Wallig MA, Kore AN, Crawshaw J, Jeffery EH: Separation of the toxic and glutathione-enhancing effects of the naturally occurring nitrile, cyanohydroxybutene. Fundam Appl Toxicol 1992;19:598–606. 21 Davis MA, Wallig MA, Jeffery EH: In vitro metabolism of cyanohydroxybutene: Formation of glutathione-S-transferase catalyzed product. Res Commun Chem Pathol Pharmacol 1993;79:343–353. 22 Davis MA, Wallig MA, Eaton D, Borroz KI, Jeffery EH: Differential effect of cyanohydroxybutene on glutathione synthesis in liver and pancreas of male rats. Toxicol Appl Pharmacol 1993;123:257–264. 23 Kossower NS, Kossower EM: The glutathioneglutathione disulfide system; in Pryor WA (ed): Free Radicals in Biology. New York, Academic Press, 1976, vol II, pp 55–84. 24 Meister A: Metabolism and function of glutathione; in Dolphin D, Avarmovic O, Poulson R (eds): Glutathione: Chemical, Biochemical and Medical Aspects. New York, Wiley, 1983, part A, p 367.

Xenobiotic Metabolism and Chronic Pancreatitis

25 Orrenius S, Moldeus P: The multiple roles of glutathione in drug metabolism. Trends Pharmacol Sci 1984;5:432–435. 26 Deneke SM, Fanburg BL: Regulation of cellular glutathione. Am J Physiol 1989;257:L163– L173. 27 Hwang C, Sinskey AJ, Lodish HF: Oxidized redox state of glutathione in the endoplasmic reticulum. Science 1992;257:1496–1501. 28 Griffith OW, Meister A: Origin and turnover of mitochondrial glutathione. Proc Natl Acad Sci USA 1985;82:4668–4672. 29 Scheele G, Jacoby R: Conformational changes associated with proteolytic processing of presecretory proteins allow gluathione-catalyzed formation of native disulfide bonds. J Biol Chem 1982;257:12277–12282. 30 Stenson WF, Lobos E, Wedner HJ: Glutathione depletion inhibits amylase release in guinea pig pancreatic acini. Am J Physiol 1983; 244:G273–G277. 31 Meister A, Anderson ME, Hwang O: Intracellular cysteine and glutathione delivery systems. J Am Coll Nutr 1986;5:137–151. 32 Thompson GA, Meister A: Utilization of Lcystine by the Á-glutamyl transpeptidase-Á-glutamyl cyclotransferase pathway. Proc Natl Acad Sci USA 1975;72:1985–1988. 33 Akerboom TPM, Bilzer M, Sies H: The relationship of biliary glutathione disulfide efflux and intracellular glutathione content in perfused rat liver. J Biol Chem 1982;257:4248– 4252. 34 Githens S: Glutathione metabolism in the pancreas, compared to liver, kidney and small intestine. Int J Pancreatol 1991;8:97–109. 35 Neuschwander-Tetri BA, Presti ME, Wells LD: Glutathione synthesis in the exocrine pancreas. Pancreas 1997;14:342–349. 36 Sweiry JH, Sastre J, Vina J, Elasasser-HP, Mann GE: A role for gamma-glutamyl transpeptidase and the amino acid transport system xc – in cysteine transport by a human pancreatic duct cell line. J Physiol (Lond) 1995;485: 167–177. 37 Meredith MJ, Williams GM: Intracellular glutathione cycling by gamma-glutamyl transpeptidase in tumorigenic and nontumorigenic cultured rat liver cells. J Biol Chem 1986;261: 4986–4992. 38 Mudd SH, Finkelstein JD, Irreverre F, Laster L: Transsulfuration in mammals: Microassays and tissue distribution of three enzymes of the pathway. J Biol Chem 1965;240:4382–4392. 39 Anderson ME, Powrie F, Puri RN, Meister A: Glutathione monoethyl ester: Preparation, uptake by tissues and conversion to glutathione. Arch Biochem Biophys 1985;239:538–548. 40 Clare DA, Pinnix IB, Lecce JG, Horton HR: Purification and properties of sulfhydryl oxidase from bovine pancreas. Arch Biochem Biophys 1988;265:351–361. 41 Neuschwander-Tetri BA, Ferrell LD, Sulshabote RJ, Grendell JH: Glutathione monoethyl ester ameliorates caerulein-induced pancreatitis in the mouse. J Clin Invest 1992;89:109– 116.

42 Sanfey H, Bulkley GB, Cameron JL: The pathogenesis of acute pancreatitis: The source and role of oxygen-derived free radicals in three different experimental models. Ann Surg 1985;20:633–639. 43 Weber H, Merkord J, Jonas L, Wagner A, Schröder H, Käding U, Werner A, Dummler W: Oxygen radical generation and acute pancreatitis: effects of dibutyltin dichloride/ethanol and ethanol on rat pancreas. Pancreas 1995;11:382–388. 44 Suzuki H, Suematsu M, Miura S, Asako H, Kurose I, Ishii H, Houzawa S, Tsuchiya M: Xanthine oxidase-mediated intracellular oxidative stress in response to caerulein in rat pancreatic acinar cells. Pancreas 1993;8:465–470. 45 Gough DB, Boyle B, Joyce WP, Delaney CP, McGeeney KF, Gorey TF, Fitzpatrick JM: Free radical inhibition and serial chemiluminescence in evolving experimental pancreatitis. Br J Surg 1990;77:1256–1259. 46 Nonaka A, Manabe T, Asano N, Imanishi T, Tamura K, Tobe T, Sugiura Y, Makino K: Direct ESR measurement of free radicals in mouse pancreatic lesions. Int J Pancreatol 1989;5:203–211. 47 Dabrowski A, Gabrylewicz A, Chwieko M: Products of lipid peroxidation and changes in sulfhydryl compounds in pancreatic tissue of rats with caerulein-induced acute pancreatitis. Biochem Med Metab Bio 1991;46:10–16. 48 Lüthen R, Grendell JH: Thiol metabolism and acute pancreatitis: trying to make the pieces fit. Gastroenterology 1994;107:888–892. 49 Grady T, Saluja A, Kaiser A, Steer M: Edema and intrapancreatic trypsinogen activation precede glutathione depletion during caerulein pancreatitis. Am J Physiol 1996;34:G20–G26. 50 Nordback IH, MacGowan S, Potter JJ, Cameron JL: The role of acetaldehyde in the pathogenesis of acute alcoholic pancreatitis. Ann Surg 1991;214:671–678. 51 Sweiry J, Mann GE: Role of oxidative stress in the pathogenesis of acute pancreatitis. Scand J Gastroenterol 1996;31(suppl 219):10–15. 52 Lüthen R, Grendell JH, Häussinger D, Niederau C: Beneficial effects of L-2-oxothiazolidine-4-carboxylate on cerulein pancreatitis in mice. Gastroenterology 1997;112:1681–1691. 53 Wang Z, Iguchi H, Ohshio G, Imamura T, Okada N, Tanaka T, Imamura M: Increased pancreatic metallothionein and glutathione levels: Protecting against cerulein- and taurocholate-induced acute pancreatitis. Pancreas 1996;13:73–83. 54 Nonaka A, Manabe T, Tobe T: Effect of a new synthetic ascorbic acid derivative, as a free radical scavenger, on the development of acute pancreatitis in mice. Gut 1991;32:528–532. 55 Lüthen RE, Haussinger D, Quante M, Grendell JH, Niederau C: Pentamedion: Beneficial effects on acute pancreatitis afforded by protection of pancreatic protein thiols (abstract). Gastroenterology 1997;112:A460. 56 Braganza JM, Scott P, Bilton D, Schofield D, Chaloner C, Scheele N, Hunt LP, Bottiglieri T: Evidence for early oxidative stress in acute pancreatitis: Clues for correction. Int J Pancreatol 1995;17:69–81.

Digestion 1998;59(suppl 4):13–24

23

57 Tsai K, Wang S-S, Chen T-S, Kong C-W, Chang F-Y, Lee F-D, Lu F-J: Oxidative stress: an important phenomenon with pathogenetic significance in the progression of acute pancreatitis. Gut 1998;42:850–855. 58 Sharer NM, Scott PD, Deardon DJ, Lee SH, Taylor PM, Braganza JM: Clinical trial of 24 hours’ treatment with glutathione precursors in acute pancreatitis. Clin Drug Invest 1995;10: 147–157. 59 Sharer NM, Rameh B, Braganza JM: Control by antioxidant therapy of recurrent pancreatitis from type I hyperlipidemia (abstract). Gut 1997;41(suppl 3):A12. 60 Braganza JM: Hypothesis: Pancreatic disease: A casualty of hepatic detoxification? Lancet 1983;ii:1000–1003. 61 Braganza JM, Wickens DG, Cawood P, Dormandy TL: Lipid-peroxidation (free-radicaloxidation) products in bile from patients with pancreatic disease. Lancet 1983;ii:375–378. 62 Ekstrom G, Ingelman-Sundberg M: Rat liver microsomal NADPH-supported oxidase activity and lipid peroxidation dependent on ethanol-inducible cytochrome P-450 (P450IIE1). Biochem Pharmacol 1989;38:1313–1319. 63 Lieber CS: Cytochrome P4502E1: Its physiological and pathological role. Physiol Rev 1997; 77:517–544. 64 Segal I, Gut A, Schofield D, Shiel N, Braganza JM: Micronutrient antioxidant status in black South Africans with chronic pancreatitis: Opportunity for prophylaxis. Clin Chim Acta 1995;239:71–79. 65 Braganza JM, Schofield D, Snehalatha C, Mohan V: Micronutrient antioxidant stasis in tropical compared to temperate-zone pancreatitis. Scand J Gastroenterol 1993;28:1098–1104. 66 Guyan PM, Uden S, Braganza JM: Heightened free radical activity in pancreatitis. Free Radical Biol Med 1990;8:347–354. 67 Acheson DWK, Rose P, Houston JB, Braganza JM: Induction of cytochromes P-450 in pancreatic disease: consequence, coincidence or cause? Clin Chim Acta 1985;153:73–84. 68 Chaloner C, Sandle LN, Mohan V, Snehalatha C, Viswanathan M, Braganza JM: Evidence for induction of cytochrome P450I in patients with tropical pancreatitis. Int J Clin Pharmacol Ther Toxicol 1990;28:235–240. 69 Gut A, Chaloner C, Schofield D, Sandle LR, Purmasir M, Segal I, Braganza JM: Evidence of toxic metabolic stress in black South Africans with chronic pancreatitis. Clin Chim Acta 1995;236:145–153. 70 Sandle LN, Braganza JM: An evaluation of the low-pH enzymatic assay of urinary D-glucaric acid and its use as a marker of enzyme induction in exocrine pancreatic disease. Clin Chim Acta 1987;162:245–256. 71 Foster JR, Idle JR, Hardwick JP, Bars R, Scott P, Braganza JM: Induction of drug-metabolizing enzymes in human pancreatic cancer and chronic pancreatitis. J Pathol 1993;169:457–463. 72 Sandilands D, Jeffery IJM, Haboubi NY, MacLennan IAM, Braganza JM: Abnormal drug metabolism in chronic pancreatitis: Treatment with antioxidants. Gastroenterology 1990:88:766–772.

24

73 Wacke R, Kirchner A, Prail F, Nizze H, Schmidt W, Fischer U, Nitschke F-P, Adam U, Fritz P, Belloc C, Drewelow B: Up-regulation of cytochrome P450 1A2, 2C9 and 2E1 in chronic pancreatitis. Pancreas 1998;16:521–528. 74 Kessova IG, DeCarli LM, Lieber CS: Inducibility of cytochromes P-4502E1 and P-4501A1 in the rat pancreas. Alcohol Clin Exp Res 1998; 22:501–504. 75 Norton ID, Apte MV, Haber PS, McCaughan GW, Pirola RC, Wilson JS: Cytochrome P4502E1 is present in rat pancreas and is induced by chronic ethanol administration. Gut 1997;42:426–430. 76 Weibkin P, Schaeffer BK, Longnecker DS, Curphy TJ: Oxidative and conjugative metabolism of xenobiotics by isolated rat and hamster acinar cells. Drug Metab Dispos 1984;12:427– 431. 77 McNamee R, Braganza JM, Hogg J, Leck I, Rose P, Cherry NM: Occupational exposure to hydrocarbons and chronic pancreatitis: A case referent study. Occup Environ Med 1994;51: 631–637. 78 Iqbal Z, Varnes ME, Yoshida A, Epstein SS: Metabolism of benzo(a)pyrene by guinea pig pancreatic microsomes. Cancer Res 1977;37: 1011–1015. 79 Black O, Murrill E, Pallas F: Pancreatic metabolism of benzo(a)pyrene in vitro and in vivo in the Long-Evans rat. Res Commun Chem Pathol Pharmacol 1980;29:291–307. 80 Anderson KE, Hammons GJ, Kadlubar FF, Potter JD, Kaderlik KR, Ilett KF, Minchin RF, Teitel CH, Chou H-C, Martin MV, Guengerich FP, Barone GW, Lang NP, Peterson LA: Metabolic activation of aromatic amines by human pancreas. Carcinogenesis 1997;18:1085–1092. 81 Baron J, Voigt JM, Whitter TB, Kawabata TT, Knapp SA, Guengerich FP, Jakoby WB: Identification of intratissue sites for xenobiotic activation and detoxication. Adv Exp Med Biol 1986;197:119–144. 82 Moore MA, Makino T, Tsuchida S, Sato K, Ichihara A, Amelizad Z, Oesch F, Konishi Y: Altered drug metabolizing potential of acinar cell lesions induced in rat pancreas by hydroxyaminoquinoline 1-oxide. Carcinogenesis 1987;8:1089–1094. 83 Ketterer B, Taylor J, Meyer D, Pemble S, Coles B, ChuLin X, Spencer S: Some functions of glutathione transferases; in Tew KD, Pickett CB, Mantle TJ, Mannervik B, Hayes JD (eds): Structure and Function of Glutathione STransferases. Boca Raton, CRC Press, 1993, pp 15–27. 84 Black O, Howerton BK: Glutathione S-transferase activity in rat pancreas. J Natl Cancer Inst 1984;72:121–123. 85 March TH, Jeffery EH, Wallig MA: Characterization of rat pancreatic glutathione S-transferases by chromatofocusing, reverse-phase highperformance liquid chromatography, and immunohistochemistry. Pancreas 1998;17:217– 228. 86 Oesch F, Gath I, Ingrashi T, Glatt H, OeschBartlomowicz B, Thomas H: Role of the wellknown basic and recently discovered acidic glutathione S-transferases in the control of ge-

Digestion 1998;59(suppl 4):13–24

87

88

89

90

91

92

93

94

95

96

97

98

99

100

notoxic metabolites; in Winner CM (ed): Biological Reactive Intermediates IV. New York, Plenum Press, 1990, pp 25–39. Nijhoff WA, Peters WHM: Quantification of induction of rat oesophageal, gastric and pancreatic glutathione and glutathione S-transferases by dietary anticarcinogens. Carcinogenesis 1994;15:1769–1772. March TH, Jeffery EH, Wallig MA: The cruciferous nitrile, crambene, induces rat hepatic and pancreatic glutathione S-transferases. Toxicol Sci 1998;42:82–90. Mannervik B, Widersten M, Board PG: Glutathione-linked enzymes in detoxication reactions; in Taniguchi N (ed): Glutathione Centennial: Molecular Perspectives and Clinical Implications. New York, Academic Press, 1989, pp 23–34. Otieno MA, Baggs RB, Hayes JD, Anders MW: Immunolocalization of microsomal glutathione S-transferase in rat tissue. Drug Metab Dispos 1997;25:12–20. Mosialou E, Ekstrom G, Adang AEP, Morgenstern R: Evidence that rat liver microsomal glutathione transferase is responsible for glutathione-dependent protection against lipid peroxidation. Biochem Pharmacol 1993; 45:1645–1651. Sparmann G, Merkord J, Jäschke A, Nizze H, Jonas L, Löhr M, Liebe S, Emmrich J: Pancreatic fibrosis in experimental pancreatitis induced by dibutyltin dichloride. Gastroenterology 1997;122:1664–1672. Weber H, Merkord J, Jonas L, Wagner A, Schröder H, Käding U, Werner A, Dummier W: Oxygen radical generation and acute pancreatitis: Effects of dibutyltin/ethanol and ethanol on rat pancreas. Pancreas 1998;11: 382–388. Merkord J, Weber H, Jonas L, Nizze H, Hennighausen G: The influence of ethanol on long-term effects of dibutyltin dichloride (DBTC) in pancreas and liver of rats. Human Environ Toxicol 1998;17:144–150. Tsukamoto H, Towner SJ, Yu GSM, French SW: Potentiation of ethanol-induced pancreatic injury by dietary fat: Induction of chronic pancreatitis by alcohol in rats. Am J Pathol 1998;131:246–257. Rutishauser SCB, Ali AM, Jeffery IJM, Hunt LP, Braganza JM: Toward an animal model of chronic pancreatitis. Int J Pancreatol 1995; 18:117–126. Allen JG, Masters HG, Peet RL, Mullins KR, Lewis RD, Skirrow SZ, Fry J: Zinc toxicity in ruminants. J Comp Pathol 1983;93:363–377. Lü J, Combs GF: Effect of excess dietary zinc on pancreatic exocrine function in the chick. J Nutr 1988;118:681–689. Smith BL, Embling PP: Sequential changes in the development of the pancreatic lesion of zinc toxicosis in sheep. Vet Pathol 1993;30: 242–247. Uden S, Bilton D, Nathan L, Hunt LP, Main C, Braganza JM: Antioxidant therapy for recurrent pancreatitis: Placebo-controlled trial. Aliment Pharmacol Ther 1990;4:357–351.

Wallig

Digestion 1998;59(suppl 4):25–35

Isidor Segal Gastrointestinal Unit, Chris Hani Baragwanath Hospital, Johannesburg, South Africa

Pancreatitis in Soweto, South Africa Focus on Alcohol-Related Disease

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Key Words Acute pancreatitis Chronic pancreatitis Soweto African Blacks

Abstract Both acute pancreatitis and chronic pancreatitis now appear to be endemic at Soweto, South Africa, and they carry a substantial toll in terms of morbidity and mortality. Case-control studies identified the same three environmental factors in each disease, namely, heavy alcohol consumption, marked exposure to occupational chemicals and low intake of fruit (a major source of vitamin C). This congruity, and parallel trends on blood biochemical analysis indicating heightened free radical activity coupled with poor antioxidant status, suggest that the two diseases may be part of a pathobiological spectrum that is linked by pancreatic oxidant stress. Further, asymptomatic chronic alcoholics had plasma glutathione concentrations that were midway between the values in non-alcoholic controls and patients with chronic pancreatitis, being significantly different from each. And, finally, apparently healthy Sowetans were actually in a state of oxidant stress that was tied in with their very poor vitamin C status, and lower serum selenium concentrations than in the UK. These data, and evidence that both antioxidants mitigate against alcoholic toxicity in experimental studies, may offer scope for disease prophylaxis in this unprivileged community. OOOOOOOOOOOOOOOOOOOOOO

Background Today acute pancreatitis and chronic pancreatitis are common diseases at Soweto, South Africa, and they cause considerable mortality and morbidity. Alcohol excess is the obvious aetiological factor but several observations question traditional interpretations on its modus operandi. Also in contrast to standard teaching, acute pancreatitis and chronic pancreatitis at Soweto appear to be part of a pathobiological spectrum, rather than being utterly different entities. This paper traces the emergence of pancreatitis in this area and briefly describes local peculiarities

ABC

© 1998 S. Karger AG, Basel 0012–2823/98/5910–0025$15.00/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

in presentation, before focussing upon theories for the development of alcoholic pancreatitis. Thereafter, collaborative studies in support of the new ‘oxidant stress’ concept are described in some detail.

Soweto, and the Emergence of Pancreatitis The township of Soweto, on the outskirts of Johannesburg, is served by the Chris Hani Baragwanath Hospital and various polyclinics: the hospital with 3,200 beds is the largest in the world. The current population of Soweto is

Prof. Isidor Segal, MD, FRCP Gastrointestinal Unit, Chris Hani Baragwanath Hospital PO Bertsham 2013 Johannesburg (South Africa) Tel. +27 11 933 1355, Fax +27 11 933 8921, E-Mail [email protected]

approximately 3.5 million, a complex polyglot of largely immigrant and mainly lower working class people, but with an emerging highly urbanised and better trained upper working class and middle class [1]. The population is relatively balanced in terms of males and females and has a young age structure indicative of the current high population growth rate. Urbanisation and industrialisation is proceeding at an alarmingly high rate in South Africa. Soweto has a continuing influx of rural Black people who merge into and contribute towards an ever-growing permanently urbanised population. In a way it is analogous to that which occurred during the industrial revolution in England. Historically, African Blacks consumed traditional home-brewed beer of low alcohol content, approximately 3%, which was associated with tissue siderosis [2]. In 1950, Walker and Arvidsson [3] conclusively showed that the source of excess ingested iron was leaching of the element from utensils during the alcohol fermentation process. In 1962 the legislation was repealed which formerly forbade the sale of Western-type alcohol to Blacks [2, 4], and African men markedly increased their consumption of spirits, notably brandy and fortified wine, with a corresponding decrease in traditional beverages. Now the pattern of liver disease changed. Whereas heavy deposits of haemosiderin were the hallmark of micronodular cirrhosis associated with iron overload, the new pattern conformed to alcoholic liver disease in Western societies, namely, alcoholic hepatitis or micronodular cirrhosis associated with steatonecrosis and alcoholic hyaline [2, 4, 5]. In keeping with these morphological differences, measurements of liver iron in autopsy studies showed that material obtained in 1976 had 40% less iron than that in material examined in 1959–1960 [5]. The emergence of pancreatitis at Soweto can be tracked from hospital admission reports over the past 70 years. Beyers [6], in a review of surgical diseases at Johannesburg during the 5-year period 1921–1926, stated that no cases of pancreatitis were observed in African Blacks. Thirty years later, Keeley [7] published a review of gastrointestinal diseases at Baragwanath Hospital and concluded that acute pancreatitis was frequently sought, but rarely encountered. A trickle of patients with calcific chronic pancreatitis began in the 1970s, such that 28 patients were identified by 1977 [8]. In the 3-year period 1981–1983 there were 55 new cases [9], the increased admission rate occurring some 20 years after Africans had access to Western-type alcohol [4]. Approximately twothirds still drank mainly home-brews, in addition to Western-type spirits, as was evidenced by high serum ferritin

26

Digestion 1998;59(suppl 4):25–35

levels in 65% and high liver iron in 64% of patients tested; in the other third, who mainly drank Western-type spirits, both these measurements were within normal limits [9]. A further survey in 1988 of admissions during the previous 4.5 years identified 90 new cases [4], with a male to female ratio of 6:1, a mean age of 40 years, and mean alcohol intake of 180 g/day for an average of 15 years. Nowadays, however, it is not uncommon to diagnose some 5 new chronic pancreatitis patients per month in the hospital which would amount to 150 cases each year, leaving aside those who present with an attack of apparently acute pancreatitis. This later increase in disease frequency parallels urbanisation and industrialisation at Soweto.

Clinical Idiosyncrasies The traditional classification of pancreatitis is into ‘acute’ and ‘chronic’ forms, wherein the descriptors are used in a pathological rather than temporal sense. Although both diseases generally manifest with an episode of agonising abdominal pain with hyperamylaseaemia, a retrospective diagnosis of ‘acute pancreatitis’ is properly retained only when morphological and functional tests of the exocrine pancreas are within normal limits after full clinical recovery from the attack; if either modality is abnormal, the diagnosis is changed to ‘chronic pancreatitis’, implying a pathological picture of irregular inflammatory sclerosis with destruction and loss of exocrine parenchyma, whether focal, segmental or diffuse. Ductal distortion and intraductal calculi, although producing rather dramatic ERCP and computed tomography images, are not part of the diagnostic definition of chronic pancreatitis because the duct system is within normal limits in some 30% of patients [10, 11]. It can be extremely difficult to diagnose chronic pancreatitis pre-operatively in this subgroup, although suspicion may be raised by impaired exocrine secretory capacity in some of them [10]; random needle biopsies to look for the patchy lesions is like searching for a needle in a haystack [12]. Paradoxically, pancreatic pain is no less, and may be considerably greater among patients with ‘small-duct disease’ than in those with a hugely dilated duct system and, while surgery remains the mainstay of treatment for intractable pain, total pancreatectomy may be the only option for these patients [13]. It has been considered axiomatic that acute pancreatitis does not lead to chronic pancreatitis except when the main pancreatic duct is disrupted during a vicious acute attack [14].

Segal

A 12-month audit of patients with a first attack of pancreatitis was undertaken at Baragwanath Hospital in 1994 [15]. The study group of 136 patients included 108 men and 28 women, of mean age 40 years. Alcohol was the predominant aetiological factor in 83% and biliary disease in 7%, while no cause could be identified in the others. Substantial morbidity was experienced by 32% of patients, from metabolic derangements, alcohol-related symptoms, respiratory impairment or, as in 10% of cases, local complications of pseudocysts, pancreatic necrosis or abscess. The overall mortality rate was 8%. A follow-up after an average of 9.3 months revealed serious morbidity in two-thirds of patients: 52 suffered severe abdominal pain, 36% had substantial weight loss, 18% had clear abnormalities on pancreatic ultrasound scans, and 31% had exocrine pancreatic impairment as gauged by faecal chymotrypsin levels. It could be estimated that 30–40% of these patients actually had underlying chronic pancreatitis, judging by the presence of at least two among the three criteria of weight loss, abdominal pain and low faecal chymotrypsin. It is very likely that the 100 or so patients with chronic pancreatitis, including those who present with an attack and others who arrive via the routine clinic, together with the further 100 cases of ‘acute pancreatitis’ each year represent the tip of an iceberg, in that Sowetans are stoical by nature and may neither seek medical therapy nor be admitted to hospital, and several still prefer traditional healers [16]. A clinical study in 1977 charted the presenting features of patients with chronic pancreatitis arriving via the clinic [8]. Pain was the dominant feature in 60% of patients, and obstructive jaundice in 33%, due to constriction of the intra-pancreatic portion of the common bile duct. Diabetes was identified in 23% of cases and was difficult to control because compliance with treatment was poor. Steatorrhoea was usually mild despite marked pancreatic insufficiency because traditional diets are high in carbohydrate but low in both fat and protein; gross fat maldigestion was only revealed when patients were deliberately tested with a high-fat load. Pulmonary tuberculosis was discovered in 25% of patients and was usually in an advanced stage [9]. Both internal and external pancreatic fistulas were not uncommon. The former presented with ascites, pleural effusions or pseudocysts, and sometimes with complex connections among these fluid-filled spaces. A significant correlation was noted between abnormalities on serum biochemistry at admission and the severity of ductal disease and rupture as gauged by ERCP. Whereas a tiny leak

from the pancreatic duct and normal serum biochemistry were usually accompanied by ready closure of the fistula in response to conventional medical treatment after a mean of about 30 days, sometimes assisted by octreotide, surgery was generally required when there were two or more sites of extravasation on the initial ERCP, and/or serum albumin was depressed and/or the ratio of fluid to total serum protein was high. In one study there were two deaths among 23 patients [17]. External pancreatic fistulas occurred after abdominal trauma, surgery to the pancreas, following catheter drainage of pseudocysts or as a complication of acute pancreatitis. Spontaneous closure was unusual, but octreotide therapy was highly efficacious, often achieving closure by 3 days; occasionally total parenteral nutrition was used as an adjuvant measure [18]. This experience concurs with that of others [19, 20]. This short description of some clinical features in patients with pancreatitis at Soweto serves to illustrate the ferocity of an attack, the lingering morbidity in a substantial number in whom underlying chronic pancreatitis was likely, and the high frequency of pulmonary tuberculosis and also diabetes diagnosed at initial presentation. A disquieting mortality rate of 15% over a 3-year period in one survey of patients with chronic pancreatitis [9] may still be an underestimate, because many patients are lost to follow-up. Patients with chronic pancreatitis are known to be at increased risk of pancreatic cancer [21], but the magnitude of this risk could not be assessed at Soweto because of the high early mortality rate and poor attendance at follow-up clinics.

Pancreatitis in Soweto

Digestion 1998;59(suppl 4):25–35

Alcohol and Pancreatic Injury Numerous experimental studies have been done and reviewed [22], as have clinical studies attempting to address key facets of the alcoholic pancreatitis problem in man [23, 24]. First, it is necessary to explain why only few people seem to be vulnerable among the thousands who consume excessive amounts of ethanol. Second, and in contradistinction to the first point, studies of patients with alcoholic chronic pancreatitis indicate that there is no threshold for alcoholic toxicity to the gland [14]. Third, there should be a plausible reason why alcoholism is a major aetiological factor for ‘acute pancreatitis’, and also for ‘chronic pancreatitis’ in the Western World, Japan, Brazil and South Africa [14]. Fourth, although the majority of patients with alcoholic chronic pancreatitis have the classical calcifying form of the disease, there are patients on record with the small-duct variant [10, 11, 13], or the

27

newly characterised duct-destructive cancer-like variant which has autoimmune overtones [25]. Fifth, there is the enigma that the clinicopathological spectrum of chronic pancreatitis, including the composition of calculi [26], is identical in alcoholic disease and in non-alcoholic disease, irrespective of geography. Any proposed mechanism must have the plasticity to accommodate all these findings, while also taking into account the absence of any convincing evidence to implicate a number of potentially predisposing factors. These include dietary macronutrients, drinking pattern and type of beverage, blood group antigens and HLA serotype, ·1 antitrypsin level and phenotype, apolipoprotein E phenotypes [23, 24], distribution of lithostatin (pancreatic stone protein) [27], or cystic fibrosis genotype [28]. Any hypothesis should also accommodate the finding that ‘trypsinogen activation’ in the pancreas is a very early phenomenon in experimental acute pancreatitis [29], as can now be inferred from the release of the activation peptide. This finding has been taken as an endorsement for pancreatic autodigestion as the all-important trigger in every attack of pancreatitis, whether clinical or experimental. The discovery that a mutation in the cationic trypsinogen gene underlies hereditary pancreatitis, and the further deduction that the abnormal form of the activated enzyme resists degradation by trypsin-like proteases, has been seen as rubberstamping this orthodox philosophy [30]. This latest development has also been put forward in support of an emerging school of thought that recurrent episodes of acute pancreatitis may, after all, give way to chronic pancreatitis via a necrosis-to-fibrosis sequence [30–32]. Over the years, the same sets of mechanisms have been invoked in alcoholic acute pancreatitis and chronic pancreatitis [22]. There are three main propositions: (1) the flow-reflow concept, wherein sphincter of Oddi dysfunction allows reflux of enterokinase-rich bile or intestinal juice into the pancreatic duct, with trypsinogen activation as a consequence; (2) the protein plug philosophy, wherein precipitates of cell casts and protein are encouraged by the presence of unduly low bicarbonate and citrate, trypsin inhibitor and lithostatin concentrations in pancreatic juice and also by increased levels of lysosomal enzymes, lactoferrin and GP-2 (glycosyl phosphatidyl inositol protein from zymogen granule membranes) [22, 33–37]; (3) the toxic metabolite concept [38] in which alcohol is thought to be directly injurious to the acinar cell, not least by altering lipid metabolism [22] so that small fat droplets appear in the cytoplasm and also between membranes of the rough endoplasmic reticulum [33]. It is probably fair

28

Digestion 1998;59(suppl 4):25–35

to say that the first two sets of mechanisms are now out of vogue, turning the focus on alcohol metabolism by the pancreas. It is known that ethanol enters cells by diffusion, and that the rate of uptake in pancreatic acinar cells is similar to that in hepatocytes [39]. However, in contrast, the pancreas has a very low capacity for the oxidative metabolism of ethanol by the sequential action of alcohol dehydrogenase and acetaldehyde dehydrogenase (fig. 1) [39, 40]. Oxidative metabolites that are generated in the liver do appear in the bloodstream albeit at low concentrations: it is thus of interest that acetaldehyde has been shown to damage the pancreas in an isolated perfused pancreas model [41]. It is also possible that, as in injury to hepatocytes, pancreatic injury may be initiated, or compounded, by antibody reactions against acetaldehyde adducts at the plasma membranes [42]. Recently, there has been excitement over the possibility that the non-oxidative metabolism of alcohol by esterification with fatty acids to yield fatty acid ethyl esters (FAEE), may be responsible for the pancreatic toxicity of alcohol. In support of this suggestion, the pancreas has been found to have the highest levels of FAEE and also of the enzyme involved in FAEE production [43]. Further, the infusion into the carotid artery of low-density lipoprotein particles reconstituted with FAEE were found to target the exocrine pancreas selectively, producing the lesions that are associated with mild acute pancreatitis, including acinar cell vacuolation, ‘trypsinogen activation’ and oedema [44]. Another proposal [45–47] is that alcoholic damage to the pancreas results from its interaction with cytochrome P450 monooxygenases (CYP), specifically CYP2E1 which has been extensively examined in the context of alcoholic liver injury [48]. Whereas the metabolism of ethanol by CYP is a very minor pathway when doses of ethanol are small, it becomes increasingly important at large doses; further, even the smallest dose is a potent CYP2E1 inducer, thereby accelerating the processing of other xenobiotics by this CYP isoenzyme. Although CYP evolved to facilitate the passage of endogenous and exogenous (xenobiotic) lipophilic substrates out of cells, by increasing their hydrophilicity, it is increasingly realised that phase I metabolism via CYP may yield a toxic metabolite. Animal experiments show that in this event prior CYP2E1 induction by ethanol magnifies risk [49, 50]; in clinical medicine, paracetamol poisoning is a good example. Unless phase II conjugating pathways are robust, cell viability is under threat, not only by reactive oxygen species that are linked to CYP function but, more importantly, by reactive xenobiotic metabolites that cause irrevers-

Segal

Fig. 1. Potential ways in which ethanol may injure the pancreas. Figure adapted by Braganza from Gut et al. [40]. ROS = Reactive oxygen species; FAEE = fatty acid ethyl esters; GSH = glutathione; iSO4 = inorganic sulphate; CYP2E1 = the ethanol-inducible form of cytochrome P450.

When originally proposed 15 years ago [45], the hypothesis was that CYP induction in the liver and the entry of toxic xenobiotic intermediates into the pancreas, by way of refluxed bile or the bloodstream, initiated chronic pancreatitis, irrespective of putative aetiological factor, and also non-gallstone acute pancreatitis. In support of this concept, analysis of serum and duodenal aspirates, following full clinical recovery after a pancreatitis attack, showed high concentrations of lipid-based free radical oxidation products in both sets of patients [53]. When diversion of abnormal bile, laden with these products, failed to abort attacks in 3 young patients with idiopathic chronic pancreatitis [54], the hypothesis was modified to include concurrent CYP induction and toxic metabolite stress in the pancreas itself [46, 47]. Over the years, a body of evidence has accumulated showing that the CYP machinery is present in the pancreas and that several isoenzymes are highly inducible [55]. This generalisation holds true for the human pan-

creas and immunocytochemical studies have confirmed that bioactivating CYP isoenzymes are indeed induced in pancreatic specimens from patients with chronic pancreatitis [56, 57]. Those isoforms include CYP1A which metabolises smoke constituents and also potential carcinogens such as benzo(a)pyrene, CYP3A which bioactivates aflatoxin, and CYP2E1 which processes numerous drugs and solvents. The last of these enzymes was clearly induced in a study in which all patients had alcoholic disease [57], but not in the earlier study in which several patients had idiopathic disease [56]. Animal experiments have now confirmed pancreatic CYP2E1 induction by ethanol [58], and also show that ethanol administration causes pancreatic oxidant stress, with alteration in redox state and mitochondrial damage [59]. Finally, it has been shown that the toxicity of ethanol, especially when combined with high dietary unsaturated fat, is related to production of free radical intermediates [60] and that this combination now provides a reproducible animal model of chronic pancreatitis [61]. The Manchester hypothesis, as currently stated [62], envisages oxidant stress in pancreatic acinar cells as the initiator of pancreatic injury in non-gallstone acute pancreatitis and, almost without exception, in chronic pancreatitis. Three component causes of oxidant stress have been identified: (1) CYP induction, whether by ethanol, constituents of cigarette smoke, high dietary C18:2 fatty acids, or anticonvulsant drugs; (2) concurrent exposure to

Pancreatitis in Soweto

Digestion 1998;59(suppl 4):25–35

ible loss of glutathione (GSH). Studies of hepatocytes show that lipid membrane peroxidation, loss of calcium compartmentation, derangement of mitochondrial function and interference in vesicle movement interact to cause cell death [51, 52].

The Manchester Hypothesis

29

a volatile chemical, often in the occupational environment [63], which undergoes bioactivation via CYP; (3) above all, a shortfall in GSH because habitual diets fail to deliver sufficient amounts of methionine and vitamin C to meet the increased demand [64]. The first component is considered to be important in the development of large-duct calcifying disease, because expansion of the endoplasmic reticulum accompanies induction of the broad CYP2 family, and results in a non-specific increase in the synthesis of a cell’s normal secretory products, whether albumin and very low density lipoprotein by the hepatocyte, or digestive protein by the pancreatic acinar cell [46, 47]. The precipitation of protein in the duct system is known to be facilitated by the hypersecretion of (apo)lactoferrin and mucin; these excesses can be rationalised with the recognition of their antioxidant capability, but it has perhaps been insufficiently stressed that impaired bicarbonate secretion, possibly as a result of oxidant stress in the centro-acinar space, could be a major contributor in the formation of intraductal protein plugs. The oxidant stress concept envisages any attack of pancreatitis as representing a breakdown of the signal transduction pathway that normally results in exocytosis of enzymes from pancreatic acinar cells, such that the zymogens and also pro-inflammatory free radical oxidation products are re-routed into the interstitium [65]. Support for this philosophy has come with the recognition of the key role of pancreatic thiols in driving the regulated secretory pathway [66]. It has further been suggested that reactive xenobiotic metabolites predispose to chronic pancreatitis by causing the irreversible loss of cellular GSH, whereas GSH is quickly refurbished when it detoxifies biological oxygen metabolites that have been repeatedly implicated in experimental acute pancreatitis [55]. Within the oxidant stress template, increasing background pain is regarded as a consequence of persistently misdirected secretion of pro-inflammatory substances into the interstitial space, because of continued impedance to exocytosis, with the resultant increase in interstitial pressure, collagen synthesis and excitation of nociceptive nerve endings [62]. A range of pharmacokinetic, immunocytochemical, occupational, biochemical and secretory studies at Manchester helped to construct this disease model. The corollary that antioxidant supplementation should ameliorate symptoms has been realised in a placebo-controlled trial [67] and also by follow-up of some 100 patients for an average of 5 years [68].

30

Digestion 1998;59(suppl 4):25–35

Xenobiotic Stress and Pancreatitis: Studies at Soweto The rapid increase in alcoholic pancreatitis at Soweto in the past decade, in line with urbanisation and industrialisation, is philosophically in tune with the Manchester hypothesis. Further, the suggestion from hospital admission statistics that both acute pancreatitis and chronic pancreatitis may now be endemic at Soweto, and the aggressive nature of each disease could, in theory, be rationalised by inaffordability of antioxidant-rich foodstuffs. Accordingly, collaborative pilot studies were undertaken in four groups: non-alcoholic controls, consecutive patients with newly diagnosed chronic pancreatitis, asymptomatic chronic alcoholics, and consecutive patients with a first attack of apparent acute pancreatitis. The studies in non-alcoholic controls revealed two peculiarities: (1) subclinical oxidative stress, wherein heightened free radical activity was coupled with poor micronutrient antioxidant status [69, 70]; (2) hypercoagulability, because of a low concentration of plasminogen activator inhibitor [Douglas, Segal, Braganza, unpubl. obs.]. In regard to the first outcome, plasma vitamin C levels were very poor, due to inaffordability of fresh fruit and vegetables, while serum selenium concentrations were also lower than in European centres. However, the concentrations of ß-carotene, ·-tocopherol and also linoleic acid in serum conformed to reference ranges from Manchester (table 1), indicating a good quality of dietary fat. Further, the concentration of urinary inorganic sulphate, which is a gauge of long-term sulphur amino acid intake conformed to levels in Manchester controls. Methionine is one source of cysteine, the rate-limiting component in the synthesis of GSH; thus, the plasma concentration of GSH, which is a rough-and-ready index of GSH homeostasis, was compatible with values in Manchester controls (table 1). Ascorbate, the bioactive form of vitamin C, closely interacts with GSH in cells, quickly buttressing GSH concentration in times of need [71, 72]: the studies in Sowetan controls revealed that this back-up device would not be available should the demand for GSH increase, for example, as a result of CYP induction, in that a high percentage of the available ascorbate was already oxidised (table 1). Since ascorbate is a key protector against oxidation of plasma lipids [73], it was not surprising to find heightened lipid peroxidation in the control studies. It is now known that the pancreas has a very high rate of glutathione turnover and that it actively synthesises GSH; however, the concentration of GSH in the pancreas is only a quarter of that in the liver [74], such that the GSH

Segal

Table 1. Summary of pilot studies testing the oxidant stress concept at Soweto in 15 controls and 14 patients with chronic pancreatitis [40, 69, 70] Soweto

Lipid isomerization (P) Linoleic acid, Ìmol/l 9,11 isomer, Ìmol/l Molar ratio, % Lipid peroxidation (S) Lipid peroxides, Ìmol/l Glutathione availability (P) GSH + GSSG, Ìmol/l Ascorbate consumption (P) Vitamin C, Ìmol/l Ascorbate, Ìmol/l Molar ratio of inactive form, % Selenium (S), nmol/l ß-Carotene (S), nmol/l ·-Tocopherol (S), mmol/mol cholesterol Inorganic sulphate (U), mmol/l D-Glucaric acid (U), mmol/mol creatinine Theophylline clearance (S), ml/kg/h

Soweto vs. Manchester

controls

CP

difference

922 17.0 1.80

802 22.1 2.89

NS NS !0.002

2.52

3.27

!0.05

6.24

2.19

0.0001

17.6 13.6 21.9 1,329 130 5.01 1.40 2.61 61

10.5 4.54 57.0 848 45 2.82 0.80 3.85 63

!0.02 !0.001 !0.002 !0.001 !0.001 !0.001 !0.05 !0.02 NS

controls

NS NS NS !0.0001 (Swo↑) NS !0.0001 (Swo↓) !0.0002 (Swo↓) NS !0.01 (Swo↓) NS NS NS NS NS

CP

NS !0.05 (Mcr↑) !0.05 (Mcr↑) !0.001 (Mcr↑) NS !0.05 (Swo↓) !0.001(Swo↓) !0.0005 (Swo↑) NS NS !0.05 (Swo↓) !0.05 (Swo↓) NS !0.001 (Mcr↑)

Data as medians; comparisons against Manchester data from published reports [50, 67, 87, 88], using two-tailed non-parametric tests: P = Plasma; S = serum; U = urine; Swo = Soweto; Mcr = Manchester. The upward or downward arrow indicates the direction of the significant difference.

requirements for normal functions of the acinar cell could be readily compromised if it took on the additional role of xenobiotic detoxification. In the group with alcoholic chronic pancreatitis, studied in the phase of background pain rather than during an acute exacerbation, there was evidence of further ascorbate oxidation, and this was accompanied not only by a further increase in lipid peroxidation, but also by depletion of plasma GSH, suggesting that GSH in the acinar cell was being excessively utilised. This interpretation is supported by the reduction in urinary inorganic sulphates, and the mobilisation, probably compensatory, of the glucuronic acid phase II pathway of drug metabolism. Studies of iron and iron-binding proteins confirmed that the heightened free radical activity was not due to iron overload [75], while pharmacokinetic studies of theophylline, a drug that is largely processed by CYP1A, suggested that environmental chemicals other than alcohol contributed to toxic metabolite stress in at least 2 among the group of 14 patients [40]. This deduction was reinforced when a structured questionnaire was administered to 35 consecutive patients, with 3 age- and gender-matched

controls for each case. Univariate group comparison identified heavier alcohol intake (p = 0.0001), increased cigarette smoking (p ! 0.017), marked exposure to occupational chemicals (p ! 0.001) and lower intake of fruit (p = 0.01) in the patients. A logistic regression analysis underlined the high risk from combined exposure to alcohol and chemicals, higher still when cigarette smoking was considered in the equation, and extremely high when fruit intake was low [76]. In the group of asymptomatic chronic alcoholics, numbering 21 individuals who drank at least 150 g ethanol daily, levels of vitamin C and selenium were as low as in the group with chronic pancreatitis, suggesting equally poor diets. However, the concentration of plasma GSH lay midway between the values in controls and patients with chronic pancreatitis, being significantly different from each (3.86 Ìmol/l, p ! 0.01 vs. non-alcoholic controls, p ! 0.05 vs. chronic pancreatitis), while the concentrations of both inorganic sulphate and D-glucaric acid in urine remained within normal limits [77]. These data lend further support to the notion that the development of chronic pancreatitis in the heavily polluted city of Soweto

Pancreatitis in Soweto

Digestion 1998;59(suppl 4):25–35

31

may be linked with xenobiotics additional to alcohol [40, 76]. In groups with an attack of apparent acute pancreatitis, several interesting findings emerged. (1) Studies of admission blood samples in 24 patients gave evidence of profound oxidant strain [78], with excessive activation of innate relative to immune inflammatory response [79]. (2) Heightened fibrinolysis, indicating the presence of active plasmin, was a striking feature even in patients with mild pancreatitis, and seemed to be independent of both trypsinogen activation, as gauged by measurement of the activation peptide of carboxypeptidase B in urine, and thrombin activation as indicated by the concentration of soluble fibrin [80]. (3) Following recovery from the attack in a further group of 30 patients, the same questionnaire was administered as was used in the previous study of chronic pancreatitis, and the information was compared with that from 30 age- and gender-matched controls. The patients had markedly higher alcohol consumption, but also higher exposure to occupational chemicals (p = 0.055) and lower fruit intake (p ! 0.05). Excessive alcohol consumption was for ^15 years in 62% of the patients. Smoking, binge drinking, alcohol consumption for 115 years and vegetable intake did not discriminate between the two groups [Segal, Ally, Becker, unpubl. observations]. Collectively, these preliminary studies support the new oxidant stress concept for the evolution of pancreatitis, and they provide a rational explanation for the observation that when apparent acute pancreatitis develops in chronic alcoholics, the chance of underlying chronic pancreatitis is high, but the disease may only be revealed after further attacks over several years [32]. The poor antioxidant status and intrinsic hypercoagulability of outwardly healthy Sowetans helps to rationalise the aggressive nature of a pancreatitis attack [15], in that both oxidative stress and the deposition of fibrin degradation products contribute to multisystem organ failure, not least the adult respiratory distress syndrome [65]. The high frequency of pancreatic calculi at the time of diagnosis in patients with chronic pancreatitis is also potentially rationalised by heightened oxidation of ascorbate, leading to the compensatory mobilisation of lactoferrin and mucin. Prior deficiency of selenium in these circumstances may sow the seeds for painful disease, judging by observations in patients at Manchester [81], while multiple deficiencies may prejudice DNA repair mechanisms and thus sow the seeds for pancreatic cancer. The studies also confirm previous work from South Africa showing low plasma vitamin E concentration that could not always be explained by exocrine pancreatic failure [82].

32

Digestion 1998;59(suppl 4):25–35

The results in the group of asymptomatic chronic alcoholics is particularly relevant to the development of chronic pancreatitis at Soweto in that the fall in plasma GSH could represent a ‘half-way house’ in the progression towards chronic pancreatitis of CYP-mediated oxidant strain (table 1). This interpretation rationalises the changes reported in pancreatic acinar cells and also in pancreatic juice composition in studies of asymptomatic chronic alcoholics in Europe and the USA. Those findings include small fat droplets in acinar cells [38], hypersecretion of lactoferrin [37] and an increase in concentration of lysosomal enzymes [36], which could reflect oxidant attack on lysosomal membranes. The reported increase in concentration and output of protein in chronic alcoholics [36, 37] is now readily understood on the basis of CYP induction, but it is not yet clear why trypsin inhibitor concentration is reduced in pancreatic juice, nor why the normal 2:1 ratio of cationic to anionic trypsinogen is reversed [36].

Xenobiotic Stress and Tropical Chronic Pancreatitis As at Soweto, chronic pancreatitis is endemic in certain tropical zones of the Far East where alcohol is not implicated. In these areas, it is not uncommon for symptoms to begin in childhood and for many family members to be affected. Also as at Soweto, calcifying disease is predominant and chronic pancreatitis runs an aggressive course towards diabetes and premature death. The condition has been well described in studies from Kerala, South India, which were the first to suggest that hydrogen cyanide, derived from cyanogenic glycosides in the dietary staple, cassava, might be responsible for pancreatic damage [83]. More recent studies at Madras, in the neighbouring state of Andhra Pradesh, where dietary staples are not generally laden with cyanogenic glycosides, have produced results that are in keeping with the new oxidant stress philosophy. As at Soweto, the community at large had very poor vitamin C bioavilability, but this could be traced to hostile culinary practices rather than to inaffordability of fruit and vegetables [84]. A predisposition to oxidant stress, as suggested by this finding, explains the previous observation that the concentration and output of lactoferrin in duodenal aspirates from children at Kerala were as high as in patients with alcoholic chronic pancreatitis from Marseilles in France [85]. In the studies at Madras, serum selenium concentrations were far higher than

Segal

noted in Sowetan controls, being as high as in controls at Manchester [86], but urinary inorganic sulphate levels were very similar in the three cohorts. The last finding undermines the proposed hydrogen cyanide connection, because sulphur amino acids hold the key to cyanide detoxification by producing thiocyanates. The paradox that experimental feeding of cassava leads to exocrine pancreatic injury [87], whereas hydrogen cyanide does not, is now rationalised by studies of another plant nitrile, which strongly suggest that the non-cyanide moiety is the true pancreatic toxin [88]. As with other xenobiotics, so too with plant nitriles, there is some evidence for bioactivation via CYP. In this context, drug metabolism studies showed significantly lower theophylline clearance levels in Madras controls [89], than in controls at Soweto [40] or Manchester [90], exemplifying the principle of polymorphism in CYP function. In patients with chronic pancreatitis at Madras, theophylline clearance was significantly accelerated, and there was an increase in urinary concentration of D-glucaric acid, the combination suggesting toxic metabolite strain [89]: the former finding was in line with studies at Manchester where the ratio of idiopathic to alcoholic disease is around 50:50 [90], but not at Soweto where alcoholic disease predominates [40]; the latter phenomenon was noted in all three centres [40, 89, 91] in line with the commonality of phase II pathways of drug metabolism. The implication from the Madras studies was that xenobiotic stress was both important and relevant to the aetiogenesis of chronic pancreatitis. Detailed social histories in 79 consecutive patients then identified regular close exposure to volatile petrochemical products, principally kerosene fumes in lamps and cookers, and also smoke constituents from burning firewood [92]. These exposures were accompanied by, and would account for, reduced concentrations of micronutrient antioxidants [93]. However, in contrast to data from Soweto [70] and Manchester [80], selenium levels were relatively preserved in the patients at Madras [93]. The observation that selenium concentration tended to be particularly low in patients with painful disease at Manchester [81], may help to explain why patients at Madras do not experience as much pain as they might be expected to have at a time when viable acinar tissue was still present [93].

Pancreatitis in Soweto

Concluding Comments The oxidant stress philosophy accommodates experimental and clinical observations on alcoholic injury to the exocrine pancreas. Some would say that the fatal flaw in the concept is that it does not attempt to accommodate pancreatic autodigestion, a concept that seems to have a wide appeal. However, as cogently argued by one prominent pancreatologist [94], ‘trypsinogen activation’ within crinophagic vacuoles in the acinar cell should not be equated with ‘autodigestion’, because there is powerful evidence that this activation, en route to the complete degradation of trypsin by lysosomal enzymes, is a physiological phenomenon to rid the cell of excess material, as occurs when exocytosis is blocked at the inception of a pancreatitis attack [65]. The new observations linking hereditary chronic pancreatitis with a form of cationic trypsinogen that would resist complete degradation, should it become prematurely activated [30], is potentially rationalised in terms of free radical pathology, in that vulnerable members can seemingly be identified by their poor antioxidant profiles [95]. The similarity in free radical marker and antioxidant profiles in the pilot studies of acute pancreatitis and chronic pancreatitis at Soweto suggest that these two conditions may indeed be part of a pathobiological spectrum, linked by gradations in acinar cell GSH status, with a greater degree of GSH depletion in chronic pancreatitis as a result of conjugation reactions with xenobiotic metabolites [62]. The most exciting outcome of the studies at Soweto, and also at Madras, is that prophylaxis against chronic pancreatitis may be possible by the simple measure of a daily tablet of vitamin C, perhaps fortified with selenium at Soweto, and with ß-carotene at Madras [70]. The protection conferred by these substances in experimental studies of alcoholic toxicity [55] provides scientific support for this proposal.

Acknowledgement The author acknowledges the assistance of Dr. J. Braganza with the manuscript.

Digestion 1998;59(suppl 4):25–35

33

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

References 1 Segal I, Dubb AA, Ou Tim L, Solomon A, Sottomayor MCCG, Zwane EM: Duodenal ulcer and working class mobility in an African population in South Africa. Br Med J 1978;i:469– 472. 2 Isaacson C: Iron overload; in Isaacson C (ed): Pathology of a Black African Population. Berlin, Springer, 1982, pp 37–63. 3 Walker ARP, Arvidsson UB: Iron overload in the South African Bantu. Trans R Soc Trop Med Hyg 1953;47:536–548. 4 Segal I, Lerios M, MacPhail AP, Di Bisceglie, Grieve TP: The genesis of chronic pancreatitis in the South African black population. S Afr Med J 1988;74:385–386. 5 MacPhail AP, Simon MO, Torrance JD, Charlton MD, Bothwell TH, Isaacson C: Changing patterns of dietary iron overload in black South Africans. Am J Clin Nutr 1979;32:1272–1278. 6 Beyers CF: Incidence of surgical diseases among the Bantu races of South Africa. J Med Assoc S Afr 1927;1:606–612. 7 Keeley KJ: Alimentary disease in the Bantu. Med Soc Proceed 1958;4:281–286. 8 Ou Tim L, Segal I, Lawson HH: Chronic calcifying pancreatitis at Baragwanath Hospital (abstract). S Afr J Sci 1977;73(suppl 1):XI. 9 Segal I, Leios M, Grieve T: The emergence of chronic calcifying pancreatitis in a developing country; in Gyr K, Singer MV, Sarles H (eds): Pancreatitis: Concepts and Classification. Amsterdam, Excepta Medica, 1984, pp 417–420. 10 Braganza JM, Hunt LP, Warrick F: Relationship between pancreatic exocrine function and ductal morphology in chronic pancreatitis. Gastroenterology 1982;82:1341–1347. 11 Heij HA, Obertop H, Schmitz PIM, Van Blankenstein M, Westbrook DL: Evaluation of the secretin-cholecystokinin test for chronic pancreatitis by discriminant analysis. Scand J Gastroenterol 1986;21:35–40. 12 Braganza JM: Does your patient have pancreatic disease? J R Coll Physicians Lond 1981; 16:13–22. 13 Walsh TN, Rode J, Theiss BA, Russell RCG: Minimal change chronic pancreatitis. Gut 1992;33:1566–1571. 14 Sarles H, Sahel J, Staub JL, Bourry J, Laugier R: Chronic pancreatitis; in Howat HT, Sarles H (eds): The Exocrine Pancreas. Philadelphia, Saunders, 1979, pp 402–439. 15 John KD, Segal I, Hassan H, Levy RD, Amin M: Acute pancreatitis in Sowetan Africans. Int J Pancreatol 1997;9:207–210. 16 Segal I, Ou Tim L: The witchdoctor and the bowel. S Afr Med J 1979;56:308–310. 17 Parekh D, Segal I: Pancreatic ascites and effusion. Arch Surg 1992;127:707–712. 18 Segal I, Parekh D, Lipschitz J, Gecelter G, Myburgh JA: Treatment of pancreatic ascites and external pancreatic fistulas with a long acting somatostatin analogue (Sandostatin). Digestion 1993;54(suppl 1):53–58. 19 Cameron JL: Chronic pancreatic ascites and pancreatic pleural effusion. Gastroenterology 1978;74:134–140.

34

20 Perdezoli P, Bassi C, Falconi M, Albrign R, Vantini I, Micciolo R: Conservative treatment of external pancreatic fistulas with parenteral nutrition alone or in combination with continuous intravenous infusion of somatostain, glucagon or calcitonin. Surg Gynecol Obstet 1986; 163:428–432. 21 Lowenfels, AB, Maisonneuve P, Cavallini G, Ammann RW, Lankisch PG, Andersen JR, Dimagno EP, Andrén-Sandberg Å, Domellöf L: Pancreatitis and the risk of pancreatic cancer. N Engl J Med 1993;328:1433–1477. 22 Singh M, Halis S: Ethanol and the pancreas. Gastroenterology 1990;98:1051–1062. 23 Worning H: Etiologic aspects of chronic pancreatitis. Review of current theories and experimental evidence. Int J Pancreatol 1989;5:1–9. 24 Wilson JS, Pirola RC: The drinker’s pancreas: molecular mechanisms emerge. Gastroenterology 1997;113:355–358. 25 Ectors N, Maillet B, Aerts R, Greboes K, Donner A, Brochard F, Lankisch P, Stolte M, Lüttges, Kremer B, Klöppel G: Non-alcoholic duct destructive chronic pancreatitis. Gut 1997;41:263–268. 26 Pitchumoni CS, Viswanathan KV, Geevarghese PJ, Banks PA: Ultrastructural and elemental composition of human pancreatic calculi. Pancreas 1987;2:152–158. 27 Schmiegel W, Burchert M, Kalthoff H, Roeder C, Butzow G, Grimm H, Kremer B, Soehendra N, Schreiber HW, Theile HG, Greten H: Immunological characterization and quantitative distribution of pancreatic stone protein in sera and pancreatic secretions in pancreatic disorders. Gastronterology 1990;99:1421–1430. 28 Sharer N, Schwartz M, Malone G, Howarth A, Painter J, Super M, Braganza JM: Mutations of the cystic fibrosis gene in patients with chronic pancreatitis. N Engl J Med 1998;339:645–652. 29 Steer ML, Meldolesi J: The cell biology of experimental pancreatitis. N Engl J Med 1987; 316:144–150. 30 Whitcomb DC, Gorry MC, Preston RA, William F, Sossenheimer MJ, Ulrich CD, Martin SP, Gates LK Jr, Amann ST, Toskes PP: Hereditary pancreatitis is caused by a mutation in the cationic trypsinogen gene. Nat Genet 1996; 14:141–145. 31 Klöppel G, Maillet B: Pathology of acute and chronic pancreatitis. Pancreas 1993;8:659– 670. 32 Ammann RW, Muellhaupt B: Progression of alcoholic acute to chronic pancreatitis. Gut 1994;35:552–556. 33 Bockman DE, Singh M, Laugier R, Sarles H: Alcohol and integrity of the pancreas. Scand J Gastroenterol 1985;20(suppl 112):106–113. 34 Freedman DS, Sakamoto K, Venu RP: GP2, the homologue to the renal cast protein uromodulin, is a major component of intraductal plugs in chronic pancreatitis. J Clin Invest 1993;92:83–90.

Digestion 1998;59(suppl 4):25–35

35 Sarles H, Bernard JP: Lithostathine and pancreatic lithogenesis. Gastroenterol Int 1991;4: 130–134. 36 Rinderknecht H, Stace NH, Renner IG: Effects of chronic alcohol abuse on exocrine pancreatic secretion in man. Dig Dis Sci 1985;30:65–71. 37 Brugge WR, Burke CA: Inter-digestive lactoferrin and trypsin secretion in early and late stages of alcoholic pancreatic disease; in Gyr KE, Singer MV, Sarles H (eds): Pancreatitis: Concepts and Classification. Amsterdam, Excerpta Medica, 1984, pp 177–181. 38 Noronha M, Ferreira De Almeida MJ, Dreiling DA, Bordalo O: Alcohol and the pancreas. Am J Gastroenterol 1981;76:114–119. 39 Clementé F, Estival A, Durand S, Ribet A: Biochemical events in rat pancreatic cells in acute and chronic alcohol intoxications; in Gyr KE, Singer MV, Sarles H (eds): Pancreatitis: Concepts and Classification. Amsterdam, Excerpta Medica, 1984, pp 111–115. 40 Gut A, Chaloner C, Schofield D, Sandle LR, Purmasir M, Segal I, Braganza JM: Evidence of toxic metabolite stress in black South Africans with chronic pancreatitis. Clin Chim Acta 1995;236:145–153. 41 Nordback IH, MacGowan S, Potter JJ, Cameron JL: The role of acetaldehyde in the pathogenesis of acute alcoholic patients. Ann Surg 214;6:671–678. 42 Yokoyama H, Ishii H, Nagata S, Kato S, Kamegaya K, Tsuchiya M: Experimental hepatitis induced by ethanol after immunization with acetaldehyde adducts. Hepatology 1993;17: 14–19. 43 Laposata EA, Lange LG: Presence of nonoxidative ethanol metabolism in human organs commonly damaged by ethanol abuse. Science 1986;231:497–499. 44 Werner J, Laposata M, Fernandez-Del Castillo C, Saghir M, Iozzo RV, Lewandrowski KB, Warshaw AL: Pancreatic injury in rats induced by fatty acid acid ethyl ester, a nonoxidative metabolite of alcohol. Gastroenterology 1997; 113:286–294. 45 Braganza JM: Hypothesis. Pancreatic disease: A casualty of hepatic ‘detoxification’? Lancet 1983;ii:1000–1003. 46 Braganza JM: The pancreas. Recent advances in Gastroenterology 1996;6:251–280. 47 Braganza JM: The role of the liver in exocrine pancreatic disease. Int J Pancreatol 1988; 3(suppl):519–542. 48 Lieber CS: Biochemical and molecular basis of alcohol-induced injury to liver and other tissues. N Engl J Med 1988;319:1619–1650. 49 Sato A, Nakajima F, Koyama F: Effects of chronic ethanol consumption on hepatic metabolism of aromatic and chlorinated hydrocarbons in rats. Br J Ind Med 1980;37:382–386. 50 Strubelt O: Interaction between alcohol and other hepatotoxic agents. Biochem Pharmacol 1989;29:1445–1449. 51 Orrenius S: Biochemical mechanisms of toxicity. Trends Pharmacol Sci 1985;Fest suppl:15– 20.

Segal

52 Török N, Marks D, Hsiao K, Oswald BJ, McNiven MA: Vesicle movement in rat hepatocytes is reduced by ethanol exposure: Alterations in microtubule-based motor enzymes. Gastroenterology 1997;113:1938–1948. 53 Guyan PM, Uden S, Braganza JM: Heightened free radical activity in pancreatitis. Free Radical Biol Med 1990;8:347–354. 54 Sandilands D, Jeffrey IJM, Haboubi NY, MacLennan I, Braganza JM: Abnormal drug metabolism in chronic pancreatitis: Treatment with antioxidants. Gastroenterology 1990;98: 766–772. 55 Braganza JM: Toxicology of the pancreas; in Ballantyne B, Turner P, Marrs TC (eds): Textbook of General and Applied Toxicology. New York, MacMillan, 1993, pp 663–714. 56 Foster JR, Idle JR, Hardwick JP, Bars R, Scott P, Braganza JM: Induction of drug metabolising enzymes in human pancreatic cancer and chronic pancreatitis. J Pathol 1993;169:457– 463. 57 Wacke R, Kirchner A, Prail F, Nizze H, Schmidt W, Fischer U, Nitschke F-P, Adam U, Fritz P, Belloc C, Drewelow B: Up-regulation of cytochrome P450 1A2, 2C9 and 2E1 in chronic pancreatitis. Pancreas 1998;16:521– 528. 58 Norton ID, Apte MV, Haber PS, McCaughan GW, Pirola RC, Wilson JS: Cytochrome P4502E1 is present in rat pancreas and is induced by chronic ethanol administration. Gut 1997;42:426–430. 59 Altomare E, Grattagliano I, Vendemiale G, Palmieri V, Palasciana G: Acute ethanol administration induces oxidative changes in rat pancreatic tissue. Gut 1996;38:742–746. 60 Clot P, Parola M, Bellomo G, Dianzani U, Carini R, Tabone M, Arico S, Ingelman-Sundberg M, Albano E: Plama membrane hydroxyethyl radical adducts cause antibody-dependant cytotoxicity in rat hepatocytes exposed to alcohol. Gastroenterology 1997;113:265–276. 61 Tsukamoto H, Towner SJ, Yu GSM, French SW: Potentiation of ethanol-induced pancreatic injury by dietary fat: Induction of chronic pancreatitis by alcohol in rats. Am J Pathol 1998;131:246–257. 62 Braganza JM: The pathogenesis of chronic pancreatitis. Q J Med 1996;89:243–250. 63 McNamee R, Braganza JM, Hogg J, Leck I, Rose P, Cherry N: Occupational exposure to hydrocarbons and chronic pancreatitis: A case referent study. Occup Environ Med 1994;51: 631–537. 64 Uden S, Acheson DWK, Reeves J, Worthington HV, Hunt LP, Brown S, Braganza JM: Antioxidants, enzyme induction, and chronic pancreatitis: A reappraisal following studies in patients on anticonvulsants. Eur J Clin Nutr 1988;42:561–569. 65 Braganza JM, Chaloner C: Acute pancreatitis. Curr Opin Anaesthesiol 1995;8:126–131. 66 Luthen R, Grendell JH: Thiol metabolism and acute pancreatitis: Trying to make the pieces fit. Gastroenterology 1994;107:888–892.

Pancreatitis in Soweto

67 Uden S, Schofield D, Miller PF, Day JP, Bottiglieri T, Braganza JM: Antioxidant therapy for recurrent pancreatitis: Biochemical profiles in a placebo-controlled trial. Aliment Pharmacol Ther 1992;6:229–240. 68 Whiteley G, Kienle A, Lee S, Taylor P, Schofield D, Braganza JM, McCloy RF: Micronutrient antioxidant therapy in the non-surgical management of painful chronic pancreatitis: Long-term observations (abstract). Pancreas 1994;9:807. 69 Gut A, Shiel N, Kay PM, Segal I, Braganza JM: Heightened free radical activity in blacks with chronic pancreatitis at Johannesburg, South Africa. Clin Chim Acta 1994;230:189–199. 70 Segal I, Gut A, Schofield D, Sheil N, Braganza JM: Micronutrient antioxidant status in black South Africans with chronic pancreatitis: Opportunity for prophylaxis. Clin Chim Acta 1995;239:71–79. 71 Winkler BS: Unequivocal evidence in support of the non-enzymatic redox coupling between glutathione/glutathione disulphide and ascorbic acid/dehydroxyascorbic acid. Biochim Biophys Acta 1992;1117:287–290. 72 Martensson J, Griffiths OW, Meister A: Glutathione ester delays the onset of scurvy in ascorbate deficient guinea pigs. Proc Natl Acad Sci USA 1993;90:317–321. 73 Frei B, Stocker R, Ames BN: Antioxidant defenses and lipid peroxidation in human blood plasma. Proc Natl Acad Sci USA 1988;85: 9748–9752. 74 Githens S: Glutathione metabolism in the pancreas, compared to liver, kidney and small intestine. Int J Pancreatol 1991;8:97–109. 75 Segal I, Sharer NM, Kay PM. Gutteridge JMC, Braganza JM: Iron, ascorbate and copper status of Sowetan blacks with calcific chronic pancreatitis. Q J Med 1996;89:45–53. 76 Segal I, Braganza JM, Gut A, Ally R, Maberti P, Mokoena L, Van’t-Hoff A, De Beer M: Alcohol, nutritional factors, and occupational xenobiotics in chronic pancreatitis: A case control study (abstract). S Afr Med J 1993;83:A780. 77 Schofield D, Ally R, Segal I, Turner S, Braganza JM: Investigation of micronutrient antioxidant status as a potential determinant of susceptability to chronic pancreatitis among alcoholics at Soweto (abstract). Gut 1995;37:A29. 78 John KD, Chaloner C, Khan AA, Schofield D, Braganza JM, Hassan H, Segal I: Evidence for oxidative stress in African blacks with acute pancreatitis (abstract). Hepatogastroenterol 1996;43(suppl):A45. 79 John KD, Chaloner C, Braganza JM, Segal I: Discordant inflammatory responses in Sowetan Africans with acute pancreatitis (abstract). Gut 1995;37(suppl 12):A37. 80 Chaloner C: Investigation toward a further understanding of pancreatitis and trypsinogen activation. PhD Thesis, The University of Manchester, UK, 1998.

81 Braganza JM, Hewitt CD, Day JP: Serum selenium in patients with chronic pancreatitis: Lowest values during painful exacerbations. Trace Elements Med 1988;5:79–84. 82 Kalvaria I, Labadrios D, Shephard GS, Visser L, Marks IN: Biochemical vitamin E deficiency in chronic pancreatitis. Int J Pancreatol 1986; 1:119–128. 83 McMillan DE, Geevarghese PH: Dietary cyanide and tropical malnutrition diabetes. Diabetes Care 1979;2:202–208. 84 Braganza JM, John SS, Padmalyalam V, Mohan M, Viswanathan M, Chari S, Madanogopalan M: Xenobiotics and tropical pancreatitis. Int J Pancreatol 1990;7:231–245. 85 Balakrishnan V, Sauniere JF, Hariharan M, Sarles H: Diet, pancreatic function and chronic pancreatitis in South India and France. Pancreas 1988;3:30–35. 86 Yadav S, Day JP, Mohan V, Snehalatha C, Braganza JM: Selenium and diabetes in the tropics. Pancreas 1991;6:528–533. 87 Geevarghese PJ: Epidemiology of chronic pancreatitis: in Balakrishnan V (ed): Chronic Pancreatitis in India. Trivandrum, St Joseph’s Press, 1987, pp 67–71. 88 Wallig MA, Gould DH, Fettman MJ: Selective pancreatotoxicity in the rat induced by the naturally occurring plant nitrile 1-cyano 2-hydroxy 3-butene. Food Chem Toxicol 1988;26: 137–147. 89 Chaloner C, Sandle LN, Mohan V, Snehalatha C, Viswanathan M, Braganza JM: Evidence for induction of cytochrome P450 in patients with tropical chronic pancreatitis. Int J Clin Pharmacol Ther Toxicol 1990;28:235–240. 90 Acheson DWK, Hunt LP, Rose P, Houston JB, Braganza JM: Factors contributing to the accelerated clearance of theophylline and antipyrine in adults with exocrine pancreatic disease. Clin Sci 1989;76:377–385. 91 Sandle LN, Braganza JM: An evaluation of the low-pH enzymatic assay of urinary D-glucaric acid and its use as marker of enzyme induction in exocrine pancreatic disease. Clin Chim Acta 1987;162:245–256. 92 Braganza JM, Schofield D, Snehalatha C, Mohan V: Micronutrient antioxidant status in tropical compared with temperate-zone chronic pancreatitis. Scand J Gastroenterol 1993;2:1098–1104. 93 Yadav S, Day JP, Mohan V, Snehalatha C, Braganza JM: Selenium and diabetes in the tropics. Pancreas 1991;6:528–533. 94 Rinderknecht H: Acute nectrotizing pancreatitis and its complications: an excessive reaction of natural defence mechanisms? In Braganza JM (ed): The Pathogenesis of Pancreatitis. Manchester, Manchester University Press, 1991, pp 86–100. 95 Prasad M, Wyllie R, van Lente F, Steffen RM, Kay MH: Antioxidants in hereditary pancreatitis. Am J Gastroenterol 1996;91:1558–1562.

Digestion 1998;59(suppl 4):25–35

35

Digestion 1998;59(suppl 4):36–48

Rory McCloy Manchester Royal Infirmary, Manchester, UK

Chronic Pancreatitis at Manchester, UK Focus on Antioxidant Therapy

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Key Words Chronic pancreatitis Free radicals Oxidative stress Antioxidant therapy Selenium Vitamin C Methionine Pancreatic surgery

Abstract The Manchester ‘oxidant stress’ hypothesis for the development of pancreatitis accommodates published information on both chronic pancreatitis and acute pancreatitis. Oxidant stress, mainly from reactive xenobiotic metabolites, is perceived as the pivotal pre-morbid problem in chronic pancreatitis and, by depleting glutathione, targets the exocytosis mechanism of the pancreatic acinar cell. Inhalation exposure to petrochemical products is identified as an independent risk factor in patients at Manchester Royal Infirmary, where some 50% of patients referred have non-alcoholic disease. This paper describes the development of antioxidant therapy, using supplements of methionine, vitamin C and selenium, and its validation in a placebo-controlled trial, followed by a retrospective cross-sectional study in 94 consecutive patients for an average of 30 months. Antioxidant therapy emerges as a safe and effective medical alternative to surgery for painful chronic pancreatitis. OOOOOOOOOOOOOOOOOOOOOO

Background Disabling pain is the main therapeutic challenge in chronic pancreatitis [1–4]. The disease generally presents as an ‘attack’ that is indistinguishable from acute pancreatitis and with the same risks. Further attacks tend to follow, such that the disease maims from increasing pain and attempts to palliate it, whether by addictive analgesis or by surgery. The unquantifiable effects from the intractable pain of chronic pancreatitis can leave patients unable to cope socially, whether this involves employment or looking after a family, or in the case of young people, making progress in full time education or training schemes. All too frequently lifestyles are eroded to the point of chronic

ABC

© 1998 S. Karger AG, Basel 0012–2823/98/5910–0036$15.00/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

invalidity [1, 2]. Furthermore, there is an increased risk of pancreatic cancer, the fifth most common cause of deaths from cancer in the UK. The consensus view is that several factors contribute to background pain [5]. Principally, these are an increase in interstitial fluid pressure irrespective of ductal morphology; active inflammation, whether clinical or subclinical; irritation of nociceptive nerve endings in the gland; creeping periductal fibrosis leading to strictures with ductal hypertension, over and above that which may be due to intraductal calculi; inflammatory masses or cysts. In the absence of a reproducible animal model, treatment modalities have been empirical. Medical measures have not been consistently successful [6]. Hence pancreat-

Mr. R.F. McCloy, BSc, MD, FRCS University Department of Surgery, Manchester Royal Infirmary Manchester M13 9WL (UK) Tel. +44 161 276 4534, Fax +44 161 273 3428 E-Mail [email protected]

✽

Fig. 1. Two case histories of patients with chronic pancreatitis illustrating occupational exposure to chemicals and the effects of cigarette and alcohol intake; adapted from Braganza [17] with permission. Attacks of pain stop after withdrawal of these factors [16].

ic surgery has for long been the gold standard against which any new treatment has had to be assessed and, mechanistically, has been directed at relieving ductal hypertension, blocking nociceptive nerves, or removing part or all of the gland [3]. Coeliac nerve blockade and, more recently, thoracoscopic sympathectomy [7] have been used, with variable results. It is widely accepted that pancreatic surgery is the only recourse for intractable pain and also that some patients proceed to near-total pancreatectomy [1–4, 8–10]. This extreme measure is the only realistic option for ‘small duct’ disease or when calculi are diffuse and may be required should pain continue despite duct drainage operations in those with ‘large duct’ disease [8, 10]. A Whipple-type procedure gives good results when there is an inflammatory mass in the head of the gland [11]. However, overall, the associated early and late morbidity and indeed mortality from pancreatic surgery, even in experienced centres, is an unsatisfactory trade-off for a benign condition [1–3]. This paper describes the development and validation of a new medical treatment, using antioxidant supplements [12–14].

Chronic Pancreatitis: Antioxidant Therapy

Observations at Manchester Royal Infirmary, 1975–1985, and Hypothesis Development In 1975, scrutiny of departmental records indicated a 3-fold increase in the annual admission rate since 1955, now with a markedly lower age at presentation. This pattern suggested the disease-promoting influence of some commonplace environmental factor. Alcoholism could not be incriminated in some 50% of individuals. Instead, the 3-fold increase in consumption of linoleic acid through increasing use of corn oil was striking [Fraine, Rose, Braganza, unpubl]. So too were occupational and social histories in patients with non-alcoholic disease (fig. 1), including those in whom attacks continued although they had become teetotal [16, 17]. Collectively, these observations suggested the involvement of cytochrome P450 (CYP) mono-oxygenases, microsomal enzymes that utilise reactive oxygen species to metabolise many endogenous lipophilic substrates but also a huge range of exogenous lipids, nowadays called xenobiotics (fig. 2). The metabolism of xenobiotics usually results in their detoxification but occasionally a sub-

Digestion 1998;59(suppl 4):36–48

37

Fig. 2. Schematic representation of free radical producing and quenching mechanisms in cells. Reproduced from Braganza [26], with permission.

strate may undergo bioactivation, and in this event, prior ‘enzyme induction’, for example, by ethanol or an anticonvulsant drug, greatly increases the yield of the reactive metabolite. Now, whether or not tissue damage ensues will depend upon prior antioxidant status and this, in turn, depends upon habitual pre-morbid diets. Should antioxidant supply fall short of need, ‘oxidant stress’ ensues, and readily deranges the polarity of secretory epithelia in a variety of ways. The transsulphuration pathway of methionine metabolism seems to be critical for the integrity of the pancreatic acinar cell and a highly vulnerable target for free radical attack, which quickly depletes methyl groups, ATP and glutathione (GSH) [14, 17–20]. The finding of high concentrations of lipid-based free radical oxidation products (FROPs) in serum and also in duodenal bile, in the relatively asymptomatic interval between pancreatitis attacks, confirmed persisting oxidant stress in patients with chronic pancreatitis [21]. A series of studies then identified three component causes for the problem (fig. 3): (1) CYP induction, especially of

38

Digestion 1998;59(suppl 4):36–48

polycyclic aromatic hydrocarbon-processing enzymes, as by cigarette smoke [22]; (2) regular close exposure to volatile petrochemical products in the occupational environment [23], and above all (3) habitually lower intakes, than in an equally enzyme-induced group of epilepsy patients on anticonvulsant drugs, of methionine and vitamin C which are required to sustain GSH stores [24]. It was natural to assume that the liver was the primary site of oxidant stress, because it undertakes the bulk of CYP-mediated reactions, and that pancreatic injury occurred by way of toxic metabolites in refluxed bile. However, despite clear evidence of damage to hepatocytes, and also to portal tracts [13, 15, 17], surgical diversion of FROP-laden bile failed to prevent further attacks in three young patients with idiopathic disease [25]. The inference that xenobiotic-mediated injury in the pancreas itself is the real problem in chronic pancreatitis [13, 14] was supported by two sets of observations. First, an immunolocalization study of drug metabolising enzymes in surgical biopsies of the pancreas confirmed induction of the phase I enzymes CYP1A2, CYP3A and NADPH-CYP-oxidore-

McCloy

Fig. 3. Chronic pancreatitis: Due to unquenched reactive intermediates? Reproduced from Braganza [26], with permission.

ductase, but not of the phase II enzyme, GSH-S-transferase 5-5 which facilitates the removal of toxic metabolites by conjugation with GSH [15]. Second, pancreatic acinar cells showed signs of oxidant strain, namely, cytoplasmic microvesiculation [25], as accompanies a breakdown in the exocytosis apparatus early in the course of experimental acute pancreatitis [19, 20]. This second finding seemed to be tantamount to saying that the basic mechanisms leading to an attack of pancreatitis are the same, whether the end result is recurrent acute pancreatitis (that is, restitio ad integram after full clinical recovery in the absence of complications) or chronic pancreatitis (fig. 4) [14]. There is now a wealth of experimental evidence to indicate that a burst of free radical activity is indeed tied in with the disruption of exocytosis and also that highly proinflammatory products are shunted into the interstitium (fig. 4), explaining the rapidity and also aberrant nature of the inflammatory response [14, 20, 26]. These experimental studies mimic transient obstruction to drainage of pancreatic secretion as in gallstone pancreatitis, hyperlipidaemia-induced acute pancreatitis in man, and hyperstimulating conditions, for example, when acute pancreatitis accompanies dietary surfeit after a period of fasting [26]. The Manchester clinical studies suggested that after recovery from an attack, should antioxidant intake fail to keep abreast with the production rate of reactive metabolites, the stage is set for chronic pancreatitis. As time goes

by and the secretory polarity of the acinar cell increasingly shifts in favour of the basolateral route, it would be expected that interstitial fluid pressure increases, nociceptive nerve endings are excited and inflammatory fibrosis creeps on to cause ductal hypertension from periductal fibrosis. These factors are known to interact in the causation of pancreatic pain. The ability of radicals to activate early genes could underlie the increased risk of pancreatic cancer. If the scheme in figure 4 were to be valid, a combination of antioxidants to sustain levels of both activated methyl groups and GSH in the acinar cell should prevent recurrences not only of acute pancreatitis but also of acute-on-chronic pancreatitis.

Chronic Pancreatitis: Antioxidant Therapy

Digestion 1998;59(suppl 4):36–48

Antioxidant Therapy: Pilot Studies and Clinical Trials The dietary studies in patients with chronic pancreatitis had identified the need for methionine and vitamin C. The former is an essential amino acid, although there is a limited capacity for its resynthesis from homocysteine via folate-vitamin B12 and choline-betaine systems; the latter interacts closely with GSH and vice versa, through redox and non-redox shuttles, and also with vitamin E. Accordingly, these two substances, together with selenium [24– 26], were tested between 1983 and 1985 in exploratory

39

4

Fig. 4. The evolution of recurrent acute pancreatitis and chronic pancreatitis characterized, respectively, by full recovery after an attack(s) or progressive loss of secretory mass (hatched cell outlines). E = Pancreatic enzymes; FRA = burst of free-radical activity jeopardizing exocytosis; FROPs = freeradical oxidation products; PAF = plateletactivating factor; PAP = pancreatitis-associated protein. Early in the attack an excess of trypsinogen, not trypsin, accompanies hyperamylasaemia. Reproduced from Braganza [14], with permission. Fig. 5. The age distribution (numbers of patients and cumulative percent) at referral (a) and at first attack of pancreatitis (b) in 94 patients with chronic pancreatitis.

40

Digestion 1998;59(suppl 4):36–48

5

McCloy

Table 1. Exploratory studies of global antioxidant therapy [26] Name

Sex

Present Age at age onset

ERCPa

Chronic pancreatitis SA M 13 MPi M 22 BH M 23 DB M 24 CA M 30 CT F 30 KC F 33 GM M 34

9 14 17 19 16 23 27 19

ADPD(C) MOP MOP(C) MOP MOP(C) MIP MIP MOP(C)

HG LM PM AL MH JL EHa BM EF WH

32 33 32 40 35 45 49 68 61 78

MIP MOP ADPO(C) ADPD normal divisum ADPD(C) ADPO ADPD ADPD

Recurrent acute pancreatitis DC F 29 AH F 46 MPh F 47 EHu F 57

27 36 42 42

normal normal normal normal

AN

60

normal

F F M M F F F F F F

M

37 39 40 47 49 50 58 68 69 80

65

Xenobiotic substancesb

Start of treatment

Last attack

Tablets

nil specific nil specific cigarettes, diesel exhaust, solvents cigarettes, [diesel/petrol exhaust] diesel exhaust, solvents nil specific oral contraceptive pill cigarettes, antionconvulsants, diesel exhaust oral contraceptive pill cigarettes, azathioprine, steroids cigarettes, [diesel exhaust] spray paints nil specific general anaesthetic !6 photocopier inks/solvents NSAID NSAID anticonvulsants, spray weedkillers

Dec. 85 July 84 Oct. 84 Dec. 83 Apr. 84 Dec. 85 July 85 Jan. 84

Jan. 86 May 87 Jun. 85 Nov. 83 Aug. 85 June 86 Nov. 86 June 84

2 6 6 2 6 3 4 6

0 12 8 0 8 0 8 8

Oct. 84 June 83 Sep. 85 Dec. 84 July 85 Apr. 86 Jan. 86 May 86 Dec. 84 Sep. 86

May 84 June 83 Mar. 86 Nov. 84 Feb. 87 Dec. 86 Dec. 86 June 87 Jan. 85 July 86

1 3 6 3 3 3 6 3 3 6

0 2 12 0 0 8 8 8 1 8

NSAID, frusemide cigarettes, NSAID, frusemide NSAID, cyclosporin cigarettes, diesel exhaust, spray polishes [diesel exhaust]

June 86 Jan. 85 June 85 Oct. 85

Dec. 86 Dec. 85 Apr. 85 Jan. 87

3 6 3 6

8 8 0 8

Jan. 85

June 85

2

0

SeACE methionine

a

ERCP changes: MIP = minimal changes; MOP = moderate changes; ADP = abnormal changes with dilated (D) or obstructed (O) main duct; C = pancreatic calculi. Histories usually by occupational physicians. b NSAID = Non-steroidal anti-inflammatory drugs. [ ] Exposure stopped after last attack. Reproduced from Braganza [26] with permission.

dose-ranging studies using commercially available preparations [26]. There was no tablet that contained all three substances, nor any compound antioxidant tablet that did not also contain vitamins A and E. The investigation (table 1) showed that most patients required tablets of methionine together with a compound antioxidant tablet for symptom relief and that, although the effective doses varied widely between patients, success was often associated with 6 tablets per day of selenium ACE (Wassen, Leatherhead, UK) and 8 of methionine (2 g, Evans, Horsham, UK). In these early studies patients with exocrine pancreatic failure or renal insufficiency were excluded. Background pain was controlled by this treatment in patients with chronic pancreatitis and further recurrences of pancreatitis were aborted, irrespective of pancreatogram appearances, through to 1990 when table 1 was compiled [26].

Other anecdotal reports signalled successful treatment with antioxidant supplementation in a child with chronic pancreatitis [27] and after failed pancreatic resection in an elderly woman [28], or after biliary bypass surgery [25]. It was surprising that patients with a grossly distorted duct system, strictures or calculi gained relief without the need for surgery. It was curious too that patients with recurrent acute pancreatitis (non-gallstone) sometimes required higher doses than in those with chronic pancreatitis (table 1), although the degree of oxidant strain between attacks seemed to be lower [21]. This finding hinted at the possibility that the difference between acute pancreatitis and chronic pancreatitis may be rather more complex, perhaps involving differential vulnerability to oxidant stress among enzymes concerned with transsulphuration, transmethylation or GSH turnover [14].

Chronic Pancreatitis: Antioxidant Therapy

Digestion 1998;59(suppl 4):36–48

41

Table 2. Synthesis of data from Uden et al. [30, 31] on a 20-week, double-blind, placebo-controlled, switch-over trial with active treatment, selenium, ß-carotene, vitamin C, vitamin E and methionine Outcome

After placebo

After active

Pvs.A

Attacks Background pain, VAS Pain diaries, 2nd 5 weeks

6 34.1 (0–100) 1.01 (0.12–6.26)

0 18.3 (0–65.8) 0.90 (0.14–3.8)

0.032 8.6 (0.05, 18.8) p = 0.049 0.30 (0.04, 0.74) p = 0.03

110B16 188 (22–903) 20B7.7 44B15 2.80B0.87

p ! 0.001 p ! 0.001 p ! 0.001 p ! 0.05 p ! 0.05

Selenium, Ìg/l ß-Carotene, Ìg/l Vitamin E, mg/l SAMe, mmol/l 9,11 LA)/linoleic acid, %

83B15 42 (0–186) 11B5.7 60B20 3.17B1.14

(1) The previous dietary studies indicated the need for the aqueous-phase antioxidants but there was no commercial preparation that did not also contain the lipid-phase antioxidants. (2) 29 patients recruited, 6 early drop-outs, 23 completed trial but incomplete adherence to protocol in 3 patients. (3) VAS = visual analogue scoresheets incorporating the 11 best descriptors of pancreatic pain in the local vocabulary; data as mean B SD or median (ranges); 2-tailed tests for significance; pain change expressed as median differences (95% confidence intervals); analysis by intention to treat, and reanalysis restricted to 20 patients, 15 with chronic pancreatitis, 5 with recurrent acute pancreatitis. (4) Somatic anxiety, depressive symptoms, emotionality, worry, and pain belief assessments were also done but were not discriminatory. (5) Baseline blood levels of selenium, ß-carotene and vitamin E were lower than in healthy controls while SAMe and 9,11 LA) (a free radical-incitable isomer of linoleic acid, 9,12 LA) were higher, the combination indicating oxidative stress. Blood antioxidant levels were unchanged by placebo. The methodology for ascorbate was not available at this time.

By 1986, methods to measure serum selenium had been set up and it was shown that supplementation generally led to peak serum levels within 4–8 weeks, holding steady thereafter. It was noted also that patients with very painful chronic pancreatitis had the lowest selenium levels [29]. In addition, antioxidant supplementation had no impact on other causes of pain, e.g. spastic colon, costochondritis, peptic ulcer [26]. A 20-week, double-blind, placebo-controlled, switchover trial was set up and ran from July 1985 for 18 months, the duration and number of patients being limited by the desirability of keeping the clinical investigator constant for the duration of the trial [30]. Consecutive patients were considered if they had at least two documented attacks in the previous year and/or constant pain that was suggestive of a pancreatic origin, provided that they were not already addicted to narcotic analgesics or required immediate pancreatic surgery to decompress the bile duct, to drain a pseudocyst or to exclude pancreatic cancer. Patients were asked not to alter their way of life, so far as possible. Allocation to placebo or active treatment in the first 10 weeks was made from random number tables by a senior pharmacist, and the trial was not decoded until the study had been completed. Active treatment was given as 14 tablets per day in divided

42

Digestion 1998;59(suppl 4):36–48

doses (6 tablets selenium ß-carotene CE, Wassen, and 8 tablets of 250 mg methionine, Evans) delivering total daily amounts of 600 Ìg organic selenium, 9,000 IU (5.4 g) ß-carotene, 0.54 g vitamin C, 270 IU vitamin E and 2 g methionine. The clinical outcome [30] in 20 patients who completed the protocol was examined well before the results of blood analyses were known [31]: the latter validated the crossover design of the study (table 2). Active treatment was associated with clinical improvement, over and above a strong placebo effect, and this was matched by correction of the subnormal group values for the measured antioxidants. Sulphadenosylmethionine (SAMe) is the first metabolite of methionine down the transsulphuration pathway [14]; the high baseline level of SAMe in serum from the patients [30] is best explained as indicating oxidant stress-induced interference in the intracellular methionine-to-GSH pathway, causing SAMe to backtrack into the bloodstream. This explanation is supported by a study of patients with an exacerbation of alcohol-related pancreatitis [32], and strengthened by the normalisation of serum SAMe values concurrently with protection of linoleic acid against free radical-induced conversion to its 9, 11, isomer during active treatment (table 2). The wide confidence intervals reflect the small number of patients and the decision to restrict the dose of

McCloy

methionine to 2 g daily although several patients in the exploratory study required twice as much (table 1). There were no side effects that could be attributed to treatment. A second trial involving 20 patients with chronic or recurrent acute non-gallstone pancreatitis ran from 1987 for 18 months and used only SAMe (Bioresearch, Milan, Italy), in divided doses amounting to 2.4 g/day, randomised in a placebo crossover design [33]. A third trial over the next 18 months used the same design and tested SAMe combined with selenium, ß-carotene tablets (Wassen) [33]. Neither regimen improved outcome. By the time of the second trial, the methods to measure vitamin C and its bioactive fraction, ascorbate, had been set up: levels of oxidatively altered vitamin C and linoleic acid were unchanged by active treatment. The third trial was abandoned after enrolment of 14 patients when three patients had a clear-cut attack of pancreatitis while on subsequent ‘open’ treatment with SAMe, selenium and ß-carotene. By a process of elimination, vitamin C and methionine were identified as the critical ingredients in the successful outcome of the initial trial [31, 33]. Furthermore, it was confirmed that improvement did not depend on the ERCP appearance [30]. The implication was that the ductal changes are a secondary, disease-compounding, problem in chronic pancreatitis, with little impact on pain [14]. This conclusion, if valid, suggests that endoscopic therapy, whether directed at removing calculi from the duct or placing stents across strictures [34, 35], is both unnecessary and misdirected.

Long-Term Surgical Audit of Antioxidant Therapy, 1983–1992 A cross-sectional audit was performed in June 1992 of patients attending the Pancreato-Biliary Unit at Manchester Royal Infirmary: a preliminary analysis has been reported in abstract [36]. The case notes of patients with painful chronic pancreatitis, diagnosed according to standard criteria, who had received antioxidant therapy (AOT) for variable periods prior to this date were reviewed and patients were omitted if they fell into certain exclusion categories as defined in the previous reports on AOT (painless disease throughout; exocrine pancreatic failure; psychological dependence on a prescribed narcotic analgesic; current pain more likely due to non-pancreatic origin [26, 30, 33]; chronic renal failure; pregnancy; pancreatic cyst or pseudocyst of diameter 15 cm [26]; pancreatic cancer so strongly suspected that surgical resection for diagnosis was considered; follow-up of less

Chronic Pancreatitis: Antioxidant Therapy

Fig. 6. The duration of follow-up (numbers of patients and cumulative percent) in 94 patients with chronic pancreatitis on long-term AOT.

than 2 months because unpublished laboratory data showed that normalisation of selenium and vitamin C levels could not be expected before 4–6 weeks). Ninety-four patients were included in the final analysis, 60 males and 34 females, with a mean age of 45 years (median 45, range 8–83). Figure 5 shows the age distribution at referral and at the time of first attack. Follow-up of the 94 patients was for a mean period of 30 months (median 19.5, range 2–131); 22% of patients were followed for 5 or more years (fig. 6). It was accepted that the comprehensive assessments of pain and quality of life made in the first placebo-controlled trial [30] were impractical in an out-patient setting. Instead, crude and subjective impressions of pain were graded as ‘no pain’, ‘clear improvement’ or ‘no change’, together with information on analgesic usage where available. The patients’ weight preAOT and at the last attendance, their ability to work and engage in their usual social activities, the number of days spent in hospital in the year pre-AOT, and the total number of days in hospital on AOT during the period of follow-up were determined. All patients received the AOT regimen that most often afforded symptomatic relief in the exploratory studies [26]. The resources for blood assays to assess assimilation of each prescribed antioxidant did not become available until some months after the audit date. The only independent gauge of concordance with AOT was a strong signal for vitamin C on urine dipstick testing at clinic visits. Non-compliance was a declared problem in only four patients, of whom two consumed excessive amounts of alcohol.

Digestion 1998;59(suppl 4):36–48

43

Exposure to environmental factors of suspected importance in the aetiology of chronic pancreatitis were recorded in all-or-none terms to cover the 2-year period before the first symptom. Occupational exposure to volatile chemicals for 4 or more hours per day [23] was present in 68% of cases, 610 cigarettes per day [37] in 67%, and consumption of 680 g alcohol daily by men or 660 g by women in 45%. Hypertriglyceridaemia was present in 9% and a past history of hyperparathyroidism was noted in 2% of cases. More than one among these factors was present in 35% of cases and no factor in 10%. Biliary tract disease is usually regarded as being coincidental or consequent upon the development of chronic pancreatitis: 12% of patients had undergone cholecystectomy prior to treatment, but only 2% for gallstone disease. The presence of gallstones was confirmed in a further 21 (22%) patients at the time of initial referral. Overall, pre-referral surgery involving the biliary tract, pancreas, stomach or duodenum had been undertaken in 32 patients (34%). Imaging studies revealed that 85% of patients had ‘large-duct’ disease (moderate or advanced change pancreatitis, and/or pancreatic calculi) [13], and 15% had ‘small-duct’ disease [9, 13]. An inflammatory mass was present in 10% and a pseudocyst in 14% of cases. Before treatment, 16% of patients had diabetes mellitus and 20% gave a clinical description of steatorrhoea although none had exocrine pancreatic failure shown by pancreatic function testing. Statistical comparisons were made using Confidence Limits (95%), Student’s t test (two-tailed p), ¯2 test (twotailed p), Wilcoxon matched pairs test and Spearman rank correlation coefficient. Differences were regarded as significant when p ^ 0.05. No patient required duct decompression or resective surgery to control pancreatic pain during 248 patientyears of follow-up on AOT. The total number of days spent in hospital while on AOT was significantly lower than the total in 1-year preceding treatment (median 4 days, range 0–82 days with 95% Confidence Interval 2.2% vs. 18, 0–150 days, CI 5.5%, p ! 0.05). On treatment, 78% of patients become pain-free (p ! 0.001) and a further 7% had a substantial reduction in pain. Of the 19 patients who were still with pain, including 2 patients who failed to take their tablets as prescribed, 7 (7% of the total group) had intermittent pain compared to continuous pain pre-AOT. Two patients had continuous pain on AOT compared to 29 before AOT (p ! 0.001), and AOT had no impact on pain in 10 patients. Among these 12 patients, there were 6 with pancreatic cysts/pseudocysts, including 2 patients who failed to comply with treatment,

44

Digestion 1998;59(suppl 4):36–48

and 4 who did comply but in whom the fluid collection did not resolve. Of 13 with fluid collections (one per patient), 2 resolved and 1 reduced in size on AOT. An inflammatory pancreatic mass was shown pre-AOT in 10 patients. There was resolution of the mass in 5 cases, a clear reduction in one, no change in one and no follow-up information on the other 3. Among the 69 patients in whom body weights were available at referral and post-AOT, a significant increase in median weight was found (medians and ranges: 64, 24– 108 kg vs. 67, 31–103 kg, p ! 0.001), representing a 3.7% increase. There was a significant negative correlation between body weight at referral and percentage change in weight, with patients who weighed !66 kg at referral gaining weight and heavier patients losing weight on AOT. Social outcome was gauged from the ability of patients to return to their pre-morbid activities and was achieved in 88% of the patients. Of the 76 patients previously employed, 67 (88%) were back at work and 53 of them (80%) were doing the same job. Of the 42 patients who drank excessive amounts of alcohol, a third continued to drink as previously, half had abstained altogether and the others had reduced their intake to ‘safe’ limits. Despite these impressive outcomes, there was no significant change in analgesic usage. There was no way of knowing retrospectively whether analgesic usage postAOT was for residual pancreatic pain, or the fear of it, or for a concomitant illness such as arthritis. There were no consistent side effects that could reasonably be attributed to treatment. One patient developed symptoms of schizophrenia, as has been reported previously [24, 26]: he drank large amounts of alcohol and worked in an environment which exposed him to diesel exhaust fumes on a daily basis so that he was on 12 tablets of methionine, that is, 3 g/day (patient P.M. in table 1). Further investigation revealed a strong family history of paranoid schizophrenia and a literature search showed that large doses of methionine, usually around 8 g daily, can precipitate organic psychoses in patients with a family history [38]. The obvious merit of AOT demonstrated by this audit is the lack of need for duct decompression or resective pancreatic surgery in these 94 patients during 248 patientyears of follow-up, and hence a reduction in perioperative morbidity, mortality and the cost of in-patient stay. These results compare very favourably with the traditional surgical approach. Surgery carries a perioperative mortality which varies from 0.5% for duodenum-preserving local pancreatic resections (Beger and Frey procedures) to 16% for total pancreatectomy, while pain relief can be expected in about 90% cases over 5 years from operation [3,

McCloy

a

b

c

d Fig. 7. A woman with haemochromatosis and calcific chronic pancreatitis with an inflammatory mass in the head of pancreas (a) which resolved on AOT [41] (b). A man with alcohol-related calcific chronic pancreatitis with compression of the duodenal loop, in 1995 (c) which resolved on AOT by 1998 while he was still drinking some 60 g alcohol daily (d).

39]. Morbidity after surgery is high with significant local complications in 10–20% cases even after local resection [3, 39]. Insulin dependence can increase in up to 50% cases, depending on the proportion of gland resected and exocrine failure developes in about one third of cases [3]. Further, it is accepted that even a total pancreatectomy may not eliminate pain because of the complex interactions among personality disturbances, intercurrent disease, analgesic dependence and alcohol abuse that collectively determine pain perception [1, 2, 5, 26, 30, 40]. Clearly, any first-line treatment for painful chronic pancreatitis should not in itself inflict injury, thus ruling out pancreatic surgery in initial management. While this principle is stated in a recent technical review [40], the authors’ concluding recommendations on pain control are

incomprehensible in the light of their preceding literature analysis which is woefully inadequate and frankly misleading in regard to antioxidant therapy, and also because they ignore the expanding experimental evidence for a rather closer relationship between the evolution of acute pancreatitis and chronic pancreatitis than hitherto perceived [14, 17, 20, 26, 30–32].

Chronic Pancreatitis: Antioxidant Therapy

Digestion 1998;59(suppl 4):36–48

Current Management Policy at the Manchester Royal Infirmary As a consequence of AOT, and confirmed by the Manchester long-term audit, pancreatic surgery for painful chronic pancreatitis is obsolete in this centre. Dramatic

45

improvement can be achieved by the parenteral route to control pain and to shrink an inflammatory mass in the head of the pancreas (fig. 7a, b) and, if treatment is started early, the need for surgery to decompress the stomach in the presence of duodenal compression may also be obviated (fig. 7c, d). Disquiet about administering vitamin C in the presence of iron overload, which was unsuspected until the urgent CT scan 2 days after admission in the young woman who provided the images for figure 7, has been overcome with the realisation that by co-administering N-acetylcysteine, or methionine presumably, no free iron was released [41]. We now reserve intervention for drainage of large pancreatic cysts/pseudocysts, for gastric bypass if still considered necessary after 6 months on AOT, for biliary bypass if jaundice returns after palliative stenting for this time, or when there is suspicion of pancreatic cancer despite negative results on microscopy of material obtained by ultrasound-guided percutaneous biopsy. Until 1993, it was necessary to administer 14 tablets daily in order to deliver the required amounts of organic selenium and vitamin C and the minimum amount (2 g) of methionine (table 1). Concordance with treatment was thus very difficult even for rational patients with idiopathic disease, especially when they began to realise the likely need for treatment indefinitely, well after their symptoms have been controlled. Patients suffering alcohol dependency are in general well-meaning but unreliable, and treatment is doomed to failure in patients who remain addicted to narcotic analgesics even when previous resective surgery leaves little or no pancreatic tissue. In order to improve the chance of concordance, avenues were explored in the quest for a tablet formulation that could incorporate methionine and reduce delivery to 4 tablets per day. This was achieved by discussions with Pharma Nord, Vejle, Denmark, so that a single tablet (Bio-Antox, now renamed Antox) was produced, containing selenium (organic) 75 Ìg, ß-carotene 3 mg, vitamin C 150 mg, vitamin E 47 mg and methionine 400 mg. The majority of patients are stabilised on one tablet 4 times daily. Apart from the cosmetic effect of hypercarotenaemia noticed by a few patients, and dyspepsia in others, the new formulation has proved to be safe and efficacious in a 5-year experience. Antox has also been used successfully to control attacks of pancreatitis in patients with type I hyperlipidaemia [42], underscoring the likelihood of a common trigger in ‘acute’ and ‘chronic’ pancreatitis, namely, oxidant stress in pancreatic acinar cells [14, 17, 26]. It is current Unit policy that blood antioxidant profiles (serum selenium, ß-carotene, and ·-tocopherol, plasma

46

Digestion 1998;59(suppl 4):36–48

vitamin C and GSH, whole blood GSH) are monitored at baseline, 3 months, 6 months and then annually. In children, urine is screened for defective metabolism of branched chain amino acids. Clearly, the results of the first assessment are not available until the first follow-up visit between 10 and 12 weeks. More often than not, in patients with chronic renal failure blood selenium levels have reached the upper limit of the reference range and the same is true for ß-carotene and sometimes vitamin C. In this event, only two Antox tablets are given each day, in divided doses. If GSH levels fail to rise within 6 months additional methionine 2 g daily is prescribed and it is ascertained that serum zinc concentration is sustained [43]. A few patients with normal renal function can reduce their Antox tablets to two or three per day based on their biochemical profiles. Monitoring of blood antioxidant profiles assures safety, confirms concordance and allows for individually tailored therapy, especially in the 10% of patients regarded as treatment failures. This has happened in patients who continue to drink huge amounts of alcohol or whose work exposes them to a large number of volatile chemicals. It has also proved difficult to treat certain patients with idiopathic recurrent acute pancreatitis and it may be that these patients require GSH itself, in the form of the monoethylester, as has been reported in animal studies [44]. We now also routinely check levels of the enzyme glucose-6-phosphate dehydrogenase, especially in patients from the Middle East, Far East or Africa. The smooth working of this enzyme as part of the pentose phosphate shunt, which feeds into the NADPH-NADP shuttle and from there into the GSH-selenium-reductase to GSH-selenium-peroxidase cycle, is of course vital for the success of treatment (fig. 2). Investigations are underway to ensure that AOT does not lead to a build-up of blood homocysteine levels. Several social problems have become apparent to us in the past decade. Patients become afraid to persevere in their work environment, when they realise that attacks coincide with periods at work [16], but compensation for such exposures will require confirmatory studies in animals and so far the field of inhalation toxicology has yet to touch the pancreas. Many patients become disheartened when their treatment is met with ridicule by other clinicians.

Concluding Comments Although greeted with scepticism by some and utter disbelief by others, the oxidant stress hypothesis has, in fact, quietly been gaining ground. Thus, the main tenets of

McCloy

the concept, namely, that damage is linked to CYPmediated processing of xenobiotics within the exocrine pancreas, is now endorsed. There is already a large literature showing that the exocrine pancreas in every species possesses the full CYP machinery and two studies, in addition to the pioneering work at Manchester [15], now confirm that this generalisation holds true for the human pancreas [45, 46]. One of these studies is of particular interest, because it involved patients with alcoholic chronic pancreatitis exclusively and showed clear induction of CYP2E1 [45]. Experimental work has confirmed induction of this isoenzyme in alcohol-fed rats [47] and alcoholic toxicity has now been unequivocally linked with oxidant strain, disorganising mitochondrial function [48]. Furthermore, analysis of human blood samples and of resected pancreatic tissue has confirmed deranged GSH status and heightened lipid peroxidation [49]. Increased lipid peroxidation has been demonstrated in Italian patients with chronic pancreatitis, with heightened levels during relapses [50]. Finally, studies from two areas where chronic pancreatitis is endemic, namely, Soweto in South Africa [51] and Madras, in South India [52], have nicely demonstrated the applicability of the oxidant stress con-

cept. In both areas biochemical findings indicated toxic metabolite stress and also very poor vitamin C status but, in Soweto this was from inaffordability of fresh fruit and vegetables whereas in Madras it was due to hostile culinary practices. Despite these developments, and preliminary reports appearing from other centres showing poor antioxidant profiles in patients with chronic pancreatitis [53–55], prejudice by pancreatologists [40] may deny patients the opportunity for treatment, or even prophylaxis (for example, in patients with epilepsy who may engage in pastimes that expose them to volatile chemicals [24]). That would be a pity!

Acknowledgements The help of Graham Whiteley, Adriene Kienle, Paul Taylor and Stephen Lee in preparing the data for the long-term AOT audit is gratefully acknowledged. The development of the Manchester concept of the aetiopathogenesis of chronic pancreatitis, and subsequently antioxidant therapy, is a tribute to 20 years of research and clinic endeavours by Dr. Joan Braganza – it has been a privilege to work with her.

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

References 1 Trapnell JE: Chronic relapsing pancreatitis: A review of 64 cases. Br J Surg 1979;66:471– 475. 2 Moossa A: Surgical treatment of chronic pancreatitis: An overview. Br J Surg 1987;74:661– 667. 3 Frey CF: Current management of chronic pancreatitis. Adv Surg 1995;28:337–270. 4 Andren-Sandberg A: Pain relief in pancreatic disease. Br J Surg 1997;84:1041–1042. 5 Ihse I: Pancreatic pain. Br J Surg 1990;77:121– 122. 6 Worning H: Chronic pancreatitis: Pathogenesis, natural history and conservative treatment. Clin Gastroenterol 1984;3:871–894. 7 Andren-Sandberg A, Zoucas E, Lillo-Gil R, Gyllstedt E, Ihse I: Thoracoscopic splanchnicectomy for chronic, severe pancreatic pain. Semin Laparosc Surg 1996;3:29–33. 8 Beger HG, Buchler M, Ditschuneit H, Malfertheiner P (eds): Chronic Pancreatitis. Heidelberg, Springer, 1990. 9 Walsh TN, Rode J, Theis BA, Russell RCG: Minimal change chronic pancreatitis. Gut 1992;33:1566–1571. 10 Fleming WR, Williamson RC: Role of total pancreatectomy in the treatment of patients with end-stage chronic pancreatitis. Br J Surg 1995;82:1409–1412.

Chronic Pancreatitis: Antioxidant Therapy

11 Izbicki JR, Bloechle C, Knoefel WT, Kuechler T, Binmoeller KF, Broelsch CE: Duodenumpreserving resection of the head of the pancreas in chronic pancreatitis. A prospective, randomized trial. Ann Surg 1995;221:350–358. 12 Braganza JM: Pancreatic disease: A casualty of hepatic ‘detoxification’? Lancet 1983;ii:1000– 1003. 13 Braganza JM: The pancreas. Recent Advances in Gastroenterology. London, Churchill-Livingstone, 1986;6:251–280. 14 Braganza JM: The pathogenesis of chronic pancreatitis. Q J Med 1996;89:243–250. 15 Foster JR, Idle JR, Hardwick JP, Bars R, Scott P, Braganza JM: Induction of drug metabolising enzymes in human pancreatic cancer and chronic pancreatitis. J Pathol 1993;169:457– 463. 16 Braganza JM, Jolley JE, Lee WR: Occupational volatile chemicals and pancreatitis: A link? Int J Pancreatol 1986;1:9–19. 17 Braganza JM: Toxicology of the pancreas; in Ballantyne B, Turner P, Marrs TC (eds): Textbook of General and Applied Toxicology. New York, MacMillan, 1993, pp 663–714. 18 Stenson WF, Lobos E, Wedner HJ: Glutathione depletion inhibits amylase release in guinea pig pancreatic acini. Am J Physiol 1983; 244:G273–G277.

19 Veghely PV, Kemeny TT: Protein metabolism and pancreatic function; in de Reuck AVS, Cameron MP (eds): Ciba Foundation Symposium on the Exocrine Pancreas. London, J & A Churchill, 1962, pp 329–349. 20 Braganza JM, Chaloner C: Acute pancreatitis. Curr Opin Anaesthesiol 1995;8:126–131. 21 Guyan PM, Uden S, Braganza JM: Heightened free radical activity in pancreatitis. Free Radical Biol Med 1990;8:347–354. 22 Acheson DWK, Hunt LP, Rose P, Houston JB, Braganza JM: Factors contributing to the accelerated clearance of theophylline and antipyrine in adults with exocrine pancreatic disease. Clin Sci 1989;76:377–385. 23 McNamee R, Braganza JM, Hogg J, Leck I, Rose P, Cherry N: Occupational exposure to hydrocarbons and chronic pancreatitis: A casereferent study. Occup Environ Med 1994;51: 631–637. 24 Uden S, Acheson DWK, Reeves J, Worthington HV, Hunt LP, Brown S, Braganza JM: Antioxidants, enzyme induction and chronic pancreatitis: A re-appraisal following studies in patients on anti-convulsants. Eur J Clin Nutr 1988;42:561–569.

Digestion 1998;59(suppl 4):36–48

47

25 Sandilands D, Jeffrey IJM, Haboubi NY, MacLennan I, Braganza JM: Abnormal drug metabolism in chronic pancreatitis: Treatment with antioxidants. Gastroenterology 1990;98: 766–772. 26 Braganza JM (ed): The Pathogenesis of Pancreatitis. Manchester, Manchester University Press, 1991. 27 Braganza JM, Thomas A, Robinson A: Antioxidants to treat chronic pancreatitis in childhood? Case report and possible implications for pathogenesis. Int J Pancreatol 1988;3:309– 316. 28 Braganza JM, Jeffrey IGM, Foster J, McCloy RF: Recalcitrant pancreatitis: Eventual control by antioxidants. Pancreas 1987;2:489–494. 29 Braganza JM, Hewitt CD, Day JP: Serum selenium in patients with chronic pancreatitis; lowest values during painful exacerbations. Trace Elements Med 1988;5:79–84. 30 Uden S, Bilton D, Nathan L, Hunt LP, Main C, Braganza JM: Antioxidant therapy for recurrent pancreatitis: Placebo-controlled trial. Aliment Pharmacol Ther 1990;4:357–371. 31 Uden S, Schofield G, Miller PF, Day JP, Bottiglieri T, Braganza JM: Antioxidant therapy for recurrent pancreatitis: Biochemical profiles in a placebo-controlled trial. Aliment Pharmacol Ther 1992;6:229–240. 32 Martensson J, Bolin T: Sulfur amino acid metabolism in chronic relapsing pancreatitis. Am J Gastroenterol 1986;81:1179–1184. 33 Bilton D, Schofield D, Mei G, Kay PM, Bottiglieri T, Braganza JM: Placebo-controlled trials of antioxidant therapy including S-adenosulmethionine in patients with recurrent non-gallstone pancreatitis. Drug Invest 1994;8:10–20. 34 Burdick JS, Hogan WJ: Chronic pancreatitis: Selection of patients for endoscopic therapy. Endoscopy 1991;23:155–159. 35 Martin RF, Hanson BL, Bosco JJ, Erkkinen JF, Broaddus SB, Goldfarb WB, Howell DA: Combined modality treatment of symptomatic pancreatic ductal lithiasis. Arch Surg 1995;130: 375–380.

48

36 Whiteley G, Kienle A, Lee S, Taylor P, Schofield D, Braganza J, McCloy R: Micronutrient antioxidant therapy in the non-surgical management of painful chronic pancreatitis: Longterm observations (abstract). Pancreas 1994;9: A807. 37 Talamani G, Bassi C, Falconi M, Frulloni L, Di Francesco V, Vaona B, Bovo P, Rigo L, Castagnini A, Angelini G, Vantini I, Pederzoli P, Cavallini G: Cigarette smoking: An independent risk factor in alcoholic pancreatitis. Pancreas 1996;12:131–137. 38 Cohen S, Nichols A, Wyatt R: The administration of methionine to chronic schizophrenic patients: A review of 10 studies. Biol Psychiatry 1974; 8:209–225. 39 Izbicki JR, Bloechle C, Knoefel WT, Kuechler T, Binmoeller KF, Broelsch CE: Duodenumpreserving resection of the head of the pancreas in chronic pancreatitis. Ann Surg 1995;221: 350–358. 40 Warshaw AL, Banks PA, Ferna´ndez-del Castillo C: AGA technical review: Treatment of pain in chronic pancreatitis. Gastroenterology 1998; 115:765–776. 41 Sharer NM, Taylor PM, Lineker BD, Gutteridge JMC, Braganza JM: Safe and successful use of vitamin C to treat painful calcific chronic pancreatitis despite ion overload from primary haemochromatosis. Clin Drug Invest 1995;10:310–315. 42 Sharer NM, Rameh B, Braganza JM: Control by antioxidant therapy (AOT) of recurrent pancreatitis from type I hyperlipidaemia. Gut 1997;41(suppl. 3):A12. 43 Wang ZH, Iguchi H, Oshio G, Inamura T, Okada N, Tanaka T, Imamura M: Increased pancreatic metallothionine and glutathione levels: Protecting against cerulein- and taurocholate-induced acute pancreatitis in rats. Pancreas 1996;13:173–183. 44 Lüthen R, Grendell JH: Thiol metabolism and acute pancreatitis: Trying to make the pieces fit. Gastroenterology 1994;107:888–892. 45 Wacke R, Kirchner A, Prall F, Nizze H, Schmidt W, Fischer U, Nitsschke FP, Adam U, Fritz P, Belloc C, Drewelow B: Up-regulation of cytochrome P450 1A2, 2C9, and 2E1 in chronic pancreatitis. Pancreas 1998;16:521– 528.

Digestion 1998;59(suppl 4):36–48

46 Baron J, Voight JM, Whitter TB, Kawabata TT, Knapp SA, Guengerich FP, Jakoby WB: Identification of intratissue sites for xenobiotic activation and detoxification. Adv Exp Med Biol 1986;197:119–144. 47 Norton ID, Apte MV, Haber PS, McCaughan GW, Pirola RC, Wilson JS: Cytochrome P4502E1 is present in rat pancreas and is induced by chronic ethanol administration. Gut 1997;42:426–430. 48 Altomare E, Grattagliano I, Vendemiale G, Palmieri V, Palasciano G: Acute ethanol administration induces oxidative changes in rat pancreatic tissue. Gut 1996;38:742–7. 49 Schoenberg MH, Büchler MW, Pietrzyk C, Uhl W, Birt D, Eisele S, Marzinzig M, Beger HG: Lipid peroxidation and glutathione metabolism in chronic pancreatitis. Pancreas 1995;10: 36–43. 50 Basso D, Panozzo MP, Fabris C, del Favero G, Meggiato T, Fogar P, Meani A, Faggian D, Plebani M, Burlina A, Naccarato R: Oxygen derived free radicals in patients with chronic pancreatic and other digestive diseases. J Clin Pathol 1990;43:403–405. 51 Segal I, Gut A, Schofield D, Sheil N, Braganza JM: Micronutrient antioxidant status in black South Africans with chronic pancreatitis: Opportunity for prophylaxis. Clin Chim Acta 1995;239:71–79. 52 Braganza JM, Schofield D, Snehalatha C, Mohan V: Micronutrient antioxidant status in tropical compared to temperate-zone pancreatitis. Scand J Gastroenterol 1993;28:1098– 1104. 53 Van Gossum A, Closset P, Noel E, Cremer M, Neve J: Deficiency in antioxidant factors in patients with alcohol-related chronic pancreatitis. Dig Dis Sci 1996;41:1225–1231. 54 Twersky Y, Bank S, Greenberg R: Endogenous antioxidants in chronic pancreatitis. Pancreas 1989;4:646. 55 Dı´teˇ P, Preˇcechtelova´ M, Sosˇka V, Lata J: Oxygen radicals and long-term antioxidant therapy in chronic pancreatitis patients (abstract). Digestion 1998;59(suppl 3):504.

McCloy

Digestion1998;59(suppl 4):49–59

Department of Child Health, Singleton Hospital, Swansea, UK

Paediatric and Hereditary Aspects of Chronic Pancreatitis

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

Key Words Cystic fibrosis Hereditary pancreatitis Hyperlipidaemia Chronic pancreatitis

Abstract Cystic fibrosis is by far the commonest cause of chronic pancreatitis in children, but pancreatitis itself is only rarely its presenting feature. In this paper an hypothesis for the development of the pancreatic lesions is presented. Impaired activation of pancreatic proteases in the small intestine is perceived as the pivotal problem that leads to continual feedback release of cholecystokinin, thus, in effect, causing a chronic hyperstimulation pancreatitis with intra-acinar activation of zymogens and, when bicarbonate secretion falls, precipitation of ‘Reg’ and other proteins in the duct system. This position contrasts with that in hereditary pancreatitis in which a mutation in the cationic trypsinogen gene leads to a form of trypsin that resists degradation by mesotrypsin and enzyme Y. A survey of the literature suggests that oxidant stress is a plausible contributor to pancreatic injury in both these diseases and in several other conditions linked with childhood pancreatitis.

John A. Dodge

OOOOOOOOOOOOOOOOOOOOOO

Background With the exception of cystic fibrosis, chronic pancreatitis is rarely diagnosed in childhood. Recurrent pancreatitis may occur in surgical disorders such as congenital malformations of the pancreato-biliary drainage system and enteric duplication cysts [1–3]. Obstruction to the pancreatic drainage system is a common feature of these conditions, which typically present with chronic or recurrent abdominal pain and therefore belong to the group of obstructive diseases defined according to the MarseilleRome classification of pancreatitis. The other two types comprise acute pancreatitis and chronic pancreatitis, the former being properly defined by complete restitution of pancreatic structure and function after full recovery from an attack [4]. Case reports have rarely been accompanied

ABC

© 1998 S. Karger AG, Basel 0012–2823/98/5910–0049$15.00/0

Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: http://BioMedNet.com/karger

by histological information and some describe acute attacks, others recurrent episodes or persistent pain with varying abdominal findings. Similar symptoms may result from biliary tract obstruction by calculi, particularly in sickle cell disease. Further discussion of these obstructive lesions is clearly outside the scope of this paper. Hereditary chronic pancreatitis and cystic fibrosis are dealt with in more detail below. There remains a small group of children with chronic pancreatitis which is described as idiopathic, and comprises the patients remaining after all other known causes have been excluded (table 1). Some inherited disorders in which chronic pancreatitis is a feature may present during childhood, but more commonly diagnosis is delayed until adult life. These are also included in table 1. The clinical presentation of pancreatitis in childhood is usually by one or more

Prof. J.A. Dodge, MD, FRCP, FRCPCH, DCH Department of Child Health Singleton Hospital Swansea SA2 8QA (UK) Tel. +44 1792 285464/286130, Fax +44 1792 285244, E-Mail [email protected]

Table 2. Acute/recurrent pancreatitis in childhood. ? indicates uncertain status Drugs

Periampullary obstruction

Infections

Trauma

Many, including Salicylates Paracetamol Cytoxic drugs ? corticosteroids Immunosuppressives Thiazides Sodium valproate Tetracyclines Erythromycin

Gallstones (including haematological) Choledochal cyst Pancreatic duct obstruction (e.g. parasitic) Congenital anomalies of pancreas (e.g. annular pancreas) Enteric duplication cysts

EB virus Mumps Measles Cytomegalovirus Influenza A Mycoplasma Leptospirosis Malaria

Blunt injury including child abuse Endoscopic retrograde cholangiopancreatography

Metabolic

Miscellaneous

Inflammatory/ systemic disease

Hyperlipidaemias (types I, IV, V) Hypercalcaemia

Refeeding pancreatitis

Haemolytic-uraemic syndrome Reye’s syndrome Kawasaki disease Inflammatory bowel disease? Henoch-Schönlein purpura

Table 1. Chronic pancreatitis. ? indicates uncertain status Chronic pancreatitis in childhood Cystic fibrosis Hereditary chronic pancreatitis Tropical calcific pancreatitis Inborn errors of metabolism, e.g. type I hyperlipidaemia ? Pancreas divisum ? Selenium/zinc deficiency Idiopathic Chronic heritable pancreatitis diagnosed mainly in adult life Hyperlipidaemias (type I) Partial lipodystrophy Wilson’s disease Haemochromatosis ·1 antitrypsin deficiency

episodes of abdominal pain, vomiting and perhaps prostration. Cystic fibrosis is by far the commonest cause in children, but pancreatitis itself is only rarely a presenting feature although its consequence, namely malabsorption, is. There are many causes of acute pancreatitis in children. These are listed in table 2, and will not be considered further, except insofar as they may relate to chronic pancreatitis.

50

Digestion1998;59(suppl 4):49–59

Cystic Fibrosis The Disease Cystic fibrosis is characterised clinically by sticky mucus secretions which tend to obstruct small epitheliallined passages in organs derived from the primitive gut, hence the widely used but inadequate synonym of ‘mucoviscidosis’. In the lungs, this leads to stagnation of secretions, bacterial colonisation and bronchiectasis, and the commonest mode of death in cystic fibrosis is respiratory failure. In the foetal intestine, obstruction by inspissated meconium leads to the condition of meconium ileus, which affects between 10 and 20% of cystic fibrosis infants. While formerly a major cause of death in the first weeks of life, meconium ileus is now usually treatable either by surgery or ‘Gastrografin’ enemas, following which the prognosis is no worse than that for other children with cystic fibrosis. The disease derives its name from the pancreatic lesions, namely, cysts and fibrosis, although it is by no means certain that the pathological changes are initiated by intraluminal obstruction within the pancreatic ducts. Inheritance of cystic fibrosis is autosomal recessive, and about 4% of the population of the United Kingdom are carriers of the mutated gene. This gives rise to an incidence of cystic fibrosis of approximately 1:2,500 live

Dodge

Pancreatitis in the Cystic Fibrosis Infant The pancreatic changes in infants with cystic fibrosis depend upon whether the observations were made on surgically obtained material during life [13, 14], or on post-

mortem material [15]. A surgical study involved a neonate aged 3 weeks and four infants aged 3, 5, 6 and 11 months, each with clinically proved cystic fibrosis who were surgically explored in an effort to seek anatomical obstruction in pancreatic ducts and extrahepatic biliary tree. In the 2 youngest cases, the pancreas appeared oedematous with petechae and extensive superficial haemorrhages; in the child aged 5 months, the gland showed haemorrhages but no oedema and in the oldest 2 children there was neither oedema nor haemorrhages. On light microscopy, each pancreas showed varying degrees of interstitial inflammation, parenchymal atrophy and dilatation of acini and fibrosis. These alterations seemed to parallel to some extent the age of the patients. Thus, the pancreas of the youngest patient revealed a heavy subacute inflammatory infiltrate in the interstitium, areas of well-preserved acini, moderate loss of parenchymal tissue, and predominantly perilobular fibrosis that had a loose appearance. Some ductules and acini were dilated and contained abundant thin secretion and few neutrophils. Beyond the age of 7 months, the disease had progressed to extensive perilobular fibrosis with mild chronic inflammatory cell infiltration, flattening of the acinar epithelium, duct ectasia and inspissated secretions plugging ductules and acini with extensive acinar atrophy. By electron microscopy in these surgical specimens, acini of widely different dimensions and shapes were found in every biopsy, their lumina varying considerably from one acinus to another. Almost normal acinar configuration was found in the 3-week-old patient. The endoplasmic reticulum was generally well preserved in each biopsy although the spaces bounded by the smooth sides of the lamellae were often greatly dilated and occupied by a homogeneous material of very low electron density. A close relation could be observed between the preservation of acinar cells and the number of round or oblong zymogen granules; as a rule, in the greatly atrophic cells these zymogen granules were very scanty. Some of the granules seemed to undergo dissolution in the apical portion of the cytoplasm. In the lumina of some well-preserved acini there was evidence of secretion, inasmuch as a dense homogeneous material was found. Mitochondria usually appeared normal. The post-mortem pancreatic features in the neonatal period vary widely, and in exceptional cases they may be virtually normal. The earliest change described is focal dilatation of acini and ductules, progressing to widespread involvement of the exocrine pancreas with flattening of the epithelium and plugging of ductules and acini by inspissated secretions. Acinar atrophy and periductal

Paediatric Chronic Pancreatitis

Digestion1998;59(suppl 4):49–59

births. It is predominantly found in people of Caucasian extraction. The current mean survival of affected individuals is in excess of 30 years in Western Europe and North America, which compares with a mean life expectancy measured in months when the disease was first described in the late 1930s. It may be wondered why such a potentially lethal mutation is so common, and various theories have been proposed. The most convincing heterozygote advantage is the apparent protection enjoyed by carriers against the effects of cholera toxin [5], Escherichia coli enterotoxin [6], and typhoid [7]. Cholera and typhoid have, of course, been the cause of major fatal epidemics in the past, and E. coli infection is also a significant cause of severe diarrhoeal disease in young children. The gene is found on the long arm of chromosome 7, the precise locus being 7q31. It is made up of 250 kb with 27 exons [8]. It contains the code for a protein which is inserted into the apical surface of epithelial cell membranes, and which has chloride channel properties. This protein is known as the cystic fibrosis transmembrane conductance regulator (CFTR) [9]. More than 700 mutations of the CFTR gene have been identified, the most common being a 3-bp deletion known as delta F508. Worldwide, this mutation accounts for 66% of the total, but the frequency of delta F508 varies from 27% in Turkish cystic fibrosis patients to 89% in Denmark. The various mutations may be divided into five classes: (i) defective transcription of mRNA; (ii) defective protein processing (including delta F508); (iii) defective regulation of the protein; (iv) defective conduction; (v) reduced amount of protein [10]. The first three classes of mutations are considered severe, while the fourth and fifth classes usually produce mild late-onset disease. However, it is difficult to directly correlate disease severity with genotype, although a consistent finding has been the association of delta F508 and pancreatic disease. A minority of patients heterozygous for this mutation retain some pancreatic function at birth, but ongoing chronic pancreatitis sooner or later leads to pancreatic insufficiency [11]. Approximately 85% of cystic fibrosis patients have clinical signs of exocrine pancreatic insufficiency at the time of diagnosis. The remainder are classified as pancreatic sufficient. Patients with pancreatic insufficiency appear to have more severe lung disease and higher sweat chloride levels than those who are pancreatic sufficient [12].

51

and interacinar fibrosis are evidence of severe involvement [15]. Histologically, the appearances of a burnt-out pancreas in cystic fibrosis are indistinguishable from those of end-stage chronic pancreatitis from other causes, but nevertheless there are quite interesting and unexplained histological differences between individual cases. In some, the predominant feature is fibrosis, while in others the glandular tissue has largely been replaced by fat. Pancreatic calcification is seen in a few, and this may follow clinical episodes of acute pancreatitis in patients who retained partial pancreatic function, but it may also occur without such a history. The islets of Langerhans are preserved until late in the process, and immunofluorescent studies on autopsy material suggest that there is a continual process of islet destruction and new cell formation. Interestingly, new endocrine tissue appears to arise from paraductal tissue, rather than by replacement and repair within existing islets [16]. The great majority of newborns with cystic fibrosis apparently retain the ability to produce pancreatic enzymes, because the serum levels of trypsinogen and lipase are characteristically elevated. This is presumed to reflect damage to pancreatic acini and absorption of retained enzyme precursors into the circulation. There is no necessary inverse correlation between the serum level of pancreatic enzyme and the amount of enzyme entering the duodenum via the ductal system, which may be partially or completely occluded, but of course, when the pancreas is so damaged that synthesis of enzymes ceases, then such enzymes or proenzymes will no longer be detectable either in the blood or the duodenum. The serum immunoreactive trypsinogen measured by radioimmunoassay or by ELISA is approximately 5–10 times higher in the blood of neonates with cystic fibrosis compared with healthy controls. This is the basis of the best current neonatal screening test for cystic fibrosis, and it is positive in more than 90% of affected infants. As stated above, immunoreactive trypsinogen measurement should not be regarded as a test of pancreatic exocrine function except insofar as it indicates whether the acini are able to secrete enzymes or not. Progression of Pancreatitis The rate at which the destructive process in the pancreas progresses varies considerably between individuals. In the majority who will eventually become pancreatic insufficient, serum trypsinogen levels fall rapidly during the first few months, and may reach subnormal or undetectable levels within the first 3 years [17]. This process is generally clinically silent, apart from the evidence of increasing steatorrhoea, but some individuals, including a

52

Digestion1998;59(suppl 4):49–59

small number who are already pancreatic insufficient clinically, experience bouts of acute pancreatitis during which the serum trypsinogen is once again elevated, along with lipase and amylase [18, 19]. Pathophysiology There is evidence for pancreatic dysfunction during intrauterine life. PAS-positive material, with an irregular distribution, was noted in pancreatic acini of a human foetus of 19 weeks’ gestation. The ducts were empty, indicating that the earliest lesions in cystic fibrosis are not ductular but involve acinar cells and their secretion [20]. There is decrease in the pancreatic enzyme concentrations in the amniotic fluid of pregnancies with a cystic fibrosis foetus. While this could possibly be due to early duct obstruction [21], surgical biopsies from a 3-week-old foetus did not show this to be a prominent feature [14] although autopsy specimens in the neonatal period showed inspissated PAS-positive material not only within acini but also within the intra- and extra-lobular ducts. There was a decrease in the ratio of acinar to connective tissue [22]. The same study found no difference in the exocrine pancreas between patients with or without meconium ileus, indicating that pancreatic involvement is not the cause of the intestinal obstruction. This has more recently been confirmed by observations in transgenic cystic fibrosis mice, which died of intestinal obstruction without evidence of pancreatic disease [23]. It is noteworthy that long-lived surviving cystic fibrosis mice do later develop abnormal pancreatic function [24]. It has long been recognised that the meconium of infants with cystic fibrosis has an unusually high content of albumin. This was initially thought to be due to defective pancreatic secretion consequent on duct obstruction, but in fact the concentration of trypsin-like immunoreactivity in the meconium of the newborn with cystic fibrosis is normal or elevated compared with that of healthy newborns. However, although present in abundance, the cystic fibrosis meconium trypsin in cystic fibrosis is devoid of catalytic activity whether or not it is complexed with ·1 proteinase inhibitor. It was concluded that the enzyme was therefore either in zymogen form or was intrinsically inactive [25]. Protein hypersecretion by the cystic fibrosis pancreas has been observed by numerous authors. This is generally attributed to an impaired secretion of fluid [26– 28], and there is no doubt that the exchange of chloride and bicarbonate in pancreatic ducts is affected by the abnormal function of CFTR as a chloride channel. However, it is possible that protein hypersecretion by the acinar cells occurs in response to some form of biofeedback

Dodge

from the intestine, where the trypsinogen present is demonstrably inactive [25]. Since the discovery of the CFTR gene and its associated protein in 1989, there has been little interest in the socalled ‘CF factors’ which were the subject of much research during the previous two decades. Serum from cystic fibrosis patients was shown to increase mucin secretion in vitro and to disrupt ciliary movement in respiratory epithelium and fresh water mussels, among several biological systems. It has since been suggested that some of the effects were mediated by high concentrations of immunoglobulins or other inflammatory markers in serum, but they may equally well have been reflecting the sort of heightened drive which appears to be operating in the pancreas. There are a number of interesting non-enzymatic proteins present in pancreatic secretion which may shed further light on these questions when their functions have been fully elucidated. One is the pancreatitis-associated protein, so called because it appears to be expressed only when the pancreas is damaged. Blood levels of this protein are therefore elevated during acute and chronic pancreatitis, constituting a marker of ongoing pancreatic disease [29], and its measurement has been proposed as an alternative to immunoreactive trypsinogen measurement for the diagnosis of cystic fibrosis by neonatal screening. Early results have been encouraging. Both tests produce some false positives, but it is interesting that with pancreatitisassociated protein a large proportion of the false positives prove to be heterozygous for a CFTR mutation [Dagorn, pers. commun.]. Another protein secreted by the pancreas is variously known as pancreatic stone protein, pancreatic thread protein, pancreatic protein X, and now, perhaps most accurately, as Reg protein. The latter name is derived from its role in the regeneration of pancreatic beta cells. Recent studies have shown that Reg protein is present in regenerating islets of Langerhans [30]. Studies by Figarella and her colleagues demonstrated that this protein, which is present in the pancreatic juice of patients suffering from chronic pancreatitis, is a product of the proteolytic cleavage of a soluble glycoprotein of 19 kDa with liberation of a small glycopeptide [31]. Studies of the major proteins in precipitates from pancreatobiliary secretions showed that they are related to these proteins and that their solubility was uniquely affected by concentration at neutral pH, leading to the conclusion that these proteins are therefore very susceptible to precipitation in the concentrated ductal secretion in cystic fibrosis, where bicarbonate secretion is impaired [32].

From these observations in humans, a plausible hypothesis can be built up, somewhat along the following lines. Early in the course of cystic fibrosis, and usually in utero, trypsinogen activation is impaired, due in part to failure of alkalinisation of the duodenum, and perhaps also the thick mucus on the intestinal mucosal surface hindering accessibility of pancreatic trypsinogen to enterokinase. This stimulates an increased release of cholecystokinin, with subsequent increased secretion of pancreatic enzymes. It is possible that pancreatic enzymes are prematurely activated within pancreatic acinar cells in these circumstances, judging by experimental studies of hyperstimulation pancreatitis in which intra-acinar activation of zymogens is brought about by a serine protease [33]. Precipitation of Reg protein in the pancreatic ductules, in the absence of chloride and bicarbonate [34], leads to blockage of ducts. Decreased passage of enzymes into the duodenum would aggravate cholecystokinin release and produce a vicious cycle, only ended when the acini are so damaged that they can no longer respond. Activation of human cationic trypsinogen without addition of any activating enzyme or calcium occurs at pH 5.0, which is very close to lysosomal pH, suggesting that lysosomes could also play a role in the premature activation of trypsinogen in pancreatitis [35]. Further, a prominent feature of experimental pancreatitis is a massive inflammatory response by neutrophils and phagocytes, with local release of their own digestive enzymes, cytokines and reactive oxygen species. It may be these agents, rather than pancreatic enzymes, which are responsible for the damage to acinar cells [36]. Using in vitro models of cystic fibrosis in the rat exocrine pancreas, investigations have similarly led to a unifying hypothesis which proposes that pancreatic dysfunction in cystic fibrosis occurs as a result of progressive acidification of the acinar and duct lumen, which leads to secondary defects in apical trafficking of zymogen granule membranes and in solubilisation of secretory zymogens. Such a process could occur spontaneously in cystic fibrosis because CFTR functions both as a chloride channel and as a regulator of membrane recycling, which involves exocytosis followed by endocytosis of shed granule membranes [37]. Under normal conditions, the stimulated pancreas initially releases prepackaged enzyme by means of exocytosis [38]. A membrane trafficking defect at the apical plasma membrane in cystic fibrosis could lead to a shift in secretory membranes from zymogen storage pools in the cytoplasm to the apical surface, where the zymogens are released. However, in the absence of adequate bicarbonate, and chloride, secretion in the ducts, retrieval

Paediatric Chronic Pancreatitis

Digestion1998;59(suppl 4):49–59

53

of exocytic membranes into the cell interior for re-use in subsequent secretion is progressively impaired as the pH of the acinar lumen decreases from alkaline to acidic levels. The acinar lumen becomes massively dilated, and the apical pole of the acinar cell is lost. These appearances have been confirmed in the acini of CFTR null mice by electron microscopy, and are thought to represent the earliest changes seen in the acini of humans with cystic fibrosis. In this mouse model, the ductules are not obstructed, and the acinar cell lesions are therefore thought to result from the low pH of the acinar lumen consequent upon the impaired secretion of bicarbonate and chloride by the duct epithelium [39]. Studies of rabbit pancreatic acini, using the polymerase chain-reaction technique, have demonstrated the presence of CFTR and confirmed that the protein was responsible for cAMP-activated chloride efflux [40]. This technique does not as yet appear to have been used in the human pancreas, in which studies using the Northern blot technique showed the predominant site of CFTR to be in the apical domain of the intralobular duct cell, with no CFTR detectable in pancreatic acinar cells, islets or nerves [41]. The CFTR chloride channel functions at the duct cell apical membrane to allow chloride to be exchanged for bicarbonate, and the absence of functional CFTR leads to decreased bicarbonate and fluid secretion into the proximal pancreatic duct lumen [34]. As described above, the resultant acidification at the secretory apex of the acinar cell both impairs enzyme solubilisation and membrane trafficking without the need to invoke a primary acinar cell defect. A second consequence of defective bicarbonate secretion into the ducts is the precipitation of Reg protein which appears to be a major cause of inspissation. However, the new studies of rabbit pancreatic acini [40] now raise the possibility that CFTR may indeed be playing a more direct role in the regulation of vesicular transport and membrane recycling in the pancreatic acinar cell: although the rabbit studies did not indicate a need for functional CFTR for enzyme and zymogen granule secretions [40], it is of interest that studies of antibiotic uptake and disposal in nasal epithelial cells from cystic fibrosis patients indicated normal endocytosis but defective exocytosis [42]. Modifying Factors Although pancreatic disease is the one major feature of cystic fibrosis which has been shown to be genotype-related, there are nevertheless exceptions and the occasional patient homozygous for delta F508 remains pancreatic sufficient for many years. The rate of progress towards

54

Digestion1998;59(suppl 4):49–59

pancreatic insufficiency may depend on random, nongenetic factors within the organ itself, and probably on other genetic and environmental factors which are not yet determined. For example, the lack of obstruction in the pancreatic ductules of the CFTR (–/–) mouse is believed to be due to a calcium-regulated chloride conductance in the pancreatic ducts which is not found in humans. It is, however, likely that variable function of other chloride channels, inherited quite independently of CFTR and under alternative control, may be present in some individuals with cystic fibrosis. Again, using the CFTR null mouse as an example, it was noted that survival was prolonged in animals fed a liquid rather than a solid diet, although the reasons are not clearly understood [24]. It is therefore conceivable that dietary variations may have some effect in modifying the course of the disease in humans. In addition, there are suggestions that oxidant stress may play a significant role in hastening the progress of pancreatitis [43, 44] and, if this is correct, it is likely that a high intake of anti-oxidant vitamins in early life might have an ameliorating effect. Cystic Fibrosis Heterozygotes and Pancreatitis Reference has already been made to the fact that among newborn infants screened for cystic fibrosis by the immunoreactive trypsinogen test there is an excess of babies who carry a single copy of the delta F508 mutation. They are asymptomatic but presumably the elevated zymogen level in serum indicates that they have a transient, mild form of pancreatitis or pancreatic dysfunction at the time of birth [45]. A similar elevation of pancreatitis-associated-protein has also been seen in newborn cystic fibrosis heterozygotes [Dagorn, pers. commun.]. It is not known whether Reg protein levels are also increased, although they are elevated in cystic fibrosis patients, whether pancreatic sufficient or with pancreatic insufficiency. However, Reg protein has been shown by immunohistochemical staining to be present in duodenal enterocytes [46] and its serum level is not proportionate to that of trypsinogen or lipase, suggesting that it is not uniquely of pancreatic origin and hence may have an endocrine or paracrine function [Figarella, pers. commun.). If newborn cystic fibrosis heterozygotes are susceptible to subclinical pancreatic derangement, what happens to them when they grow up? A recent study of 134 consecutive patients with chronic pancreatitis disclosed 18, including 12 non-alcoholics, with a CFTR mutation on one chromosome, which respresented a 2.5-fold increase over the control frequency (!0.001). In addition, 14 patients

Dodge

(10%) had the 5T allele on one chromosome, which was twice the expected population incidence. This minor change in a coding sequence is not a true mutation, but it results in decreased synthesis of (normal) CFTR which may therefore prove inadequate under conditions of stress. None of the 134 patients had sufficient clinical or biochemical features for a diagnosis of CF, but the conclusion was that a mutation in the CFTR gene, even a single copy, is a risk factor for chronic pancreatitis [47]. A smaller study of 27 patients with idiopathic pancreatitis, seemingly including some with recurrent acute pancreatitis and normal endoscopic pancreatograms, supports this conclusion [48]. Therapeutic Opportunities Progression of chronic pancreatitis to the stage of pancreatic insufficiency is seemingly inevitable in most cases of cystic fibrosis. However, what little evidence we have concerning pancreatic function in young infants with the disease suggests that more than one third have substantial preservation of pancreatic function at the time of diagnosis by neonatal screening [17]. Preservation of this residual function might be extended by interrupting the secretory drive on the pancreas by giving supplementary pancreatic enzymes [25], or by antioxidant treatment [43]. Increased lipid peroxidation is a feature of established cystic fibrosis lung disease [44, 49] and may be mitigated by antioxidants [50, 51], but such treatment would need to be started early in infancy if it were to rescue the pancreas. Recent evidence of a protective effect of nitric oxide in caerulein-induced acute pancreatitis in mice adds further support for the involvement of inflammatory cells in the process, because manoeuvres to increase nitric oxide were effective only in the intact animal and not on isolated acini, suggesting an indirect microcirculatory action including inhibition of leucocyte activation [52].

Hereditary Pancreatitis The Disease This is a rare condition characterised by recurrent bouts of pancreatitis. There are excellent descriptions of a relatively small number of families, in whom this form of pancreatitis is inherited as an autosomal dominant condition, with a penetrance of about 80% [53]. A strong family history can usually be found, and the majority of affected individuals experience their first symptoms before the age of 20. In addition to severe abdominal pain, the attacks may be accompanied by fever, and there is marked eleva-

Paediatric Chronic Pancreatitis

tion of serum amylase [54]. Presentation may also be with steatorrhoea, as in the case of a 17-month-old boy whose mother and brother also had steatorrhoea and pancreatic calcification [55]. In one of the largest studies of this condition, a search in England and Wales identified seven families with 72 patients. Penetrance was about 80%. The mean age of onset was 13.6 years, with two peaks at 5 and 17 years. It was thought that the second peak represented affected individuals whose symptoms were precipitated by alcohol. In 5 of the 7 families, there were patients with onset in childhood as well as others with adult onset. It was reassuring that only 4 of the 72 patients had life-threatening disease [56]. Established chronic pancreatitis was associated with pancreatic insufficiency (5.5%), diabetes mellitus (12.5%) and pancreatic pseudocysts (5.5%). Two had proven portal vein thrombosis and it was suspected in 3 others. Attacks were precipitated by alcohol, a high fat intake or by emotional disturbances. Unless exocrine pancreatic failure or diabetes mellitus developed, patients seem to improve in later life [56]. Pancreatic calculi are a frequent finding in hereditary pancreatitis and may require surgical removal. In one family with 5 cases, the calculi were composed of more than 95% calcium salts, but in 10 other families the stones were mainly composed of degraded amorphous residues of lithostathine, namely, Reg protein [57]. The question of an enhanced risk of pancreatic carcinoma in patients with hereditary pancreatitis was raised by observations in a single English family, in which pancreatitis occurred in members of four successive generations [58]. The cancer risk was explored in a large international hereditary pancreatitis study group co-ordinated by Lowenfels. Records on 246 patients (125 males and 121 females) were reviewed and in 218 the diagnosis appeared to be highly probable, whereas in 28 it was less certain. The mean age of onset of symptoms was 13.9 B 12.2 years. Eight pancreatic adenocarcinomas occurred, compared with an expected incidence of 0.15, i.e. a more than 50-fold increase. The cumulative risk of pancreatic cancer occurring by the age of 70 years was 40%, which increased to 75% in patients with a paternal inheritance pattern [59]. Genetics The apparent involvement of Reg protein and duct obstruction, together with the identification of the CFTR gene on chromosome 7, led investigators to search intensively in this part of the genome for the hereditary pancreatitis gene. Studies of highly informative microsatellite markers in a French family with 47 of 147 individuals

Digestion1998;59(suppl 4):49–59

55

suffering from hereditary pancreatitis found linkage to markers fairly close to the CFTR gene. The authors noted that the gene encoding carboxypeptidase A1, which is a pancreatic exopeptidase, had been mapped to the same region [60]. Another study of a large American pedigree located the gene at 7q35 [61]. At least eight trypsinogen genes are found in this region. The common mutation in hereditary pancreatitis was found to be an arginine-histidine substitution designated R117H [62]. Identical mutations were found in five separate kindreds, raising the possibility of a founder effect. Although four American families were unaware of any relationship, subsequent haplotyping confirmed the probability that they had a common ancestor. The fifth family was Italian, and their haplotype was different, suggesting that the same mutation had occurred at least twice. The R117H mutation was not found among 140 unrelated controls, which means that it could be used as a simple screening test for hereditary pancreatitis. However, a recent study of 8 unrelated families with hereditary pancreatitis found that only 6 were affected with the R117H mutation. In the other 2 families, not only was the mutation absent but studies of linkage to 7q35 suggested locus heterogeneity [63]. Pathogenesis It is more than a century since Chiari [64] first suggested that pancreatitis results from inappropriate activation of pancreatic proenzymes, but evidence for activation of trypsinogen within the acinar cell was obtained in experimental studies of acute pancreatitis only a decade ago [65]. It should be stated that trypsinogen activation does not necessarily equate to autodigestion, because the activation process is normally carried through to complete degradation of the enzyme within the safe confines of lysosomes [36]. Nevertheless, the finding of the R117H mutation has been accepted as proof that pancreatic autodigestion does occur during pancreatitis. The site of the R117H mutation in the cationic trypsinogen gene is critical for trypsin inactivation by mesotrypsin and enzyme Y. Thus, it was suggested that this specific mutation confers resistance to trypsin hydrolysis and presumably results in uncontrolled activation of trypsinogen and other zymogens, leading to autodigestion of the pancreas [62]. There would thus appear to be an important difference between the pathogenesis of pancreatitis in this condition, in which activation of cationic trypsinogen is presumed to occur within the acinar cell, and that in cystic fibrosis, where mechanisms to get rid of any prematurely activated cationic trypsinogen appear to be intact [25].

56

Digestion1998;59(suppl 4):49–59

However, the intermittent nature of episodes of hereditary pancreatitis suggests that trigger factors (diet, drugs, alcohol, stress, duct obstruction) are necessary to initiate attacks. Furthermore, within the same hereditary pancreatitis kindred, members who developed the disease had lower vitamin E and selenium levels and higher activities of the antioxidant enzyme superoxide dismutase than did their asymptomatic relatives or unrelated controls [66]. It seems likely that the high superoxide dismutase levels are a response to a persistent subclinical release of oxygen free radicals that were not immediately quenched because of the low selenium and vitamin E levels. This finding raises the possibility that supplementation with selenium and/or vitamin E may be beneficial [66].

Hyperlipidaemias Risk of Pancreatitis This group of conditions is characterised by elevated blood levels of chylomicrons or other categories of lipid. Acute pancreatitis is a recognised feature of types I, IV and V [67], and chronic pancreatitis has also been described, particularly in type I hyperlipoproteinaemia, which is also known as lipoprotein lipase deficiency [68]. This is an autosomal-recessive condition, and the full clinical disease is manifested by attacks of abdominal pain with nausea and vomiting, hepatosplenomegaly, eruptive xanthomas, and occasional jaundice. Blood samples show a milky appearance of the plasma, in which the chylomicron count is raised, along with hypercholesterolaemia and low levels of ·- and ß-lipoproteins. There is decreased plasma postheparin lipolytic activity, which may also be found in heterozygotes who may show slight elevation of lipids. Precocious atherosclerosis is not usually a feature, although there is conflicting evidence on this point. Although onset is usually in childhood, in one case the diagnosis was first made at the age of 75 years. The patient had suffered from recurrent abdominal pain, nausea and vomiting since childhood and had been hospitalised during acute attacks several times a year for nearly 30 years before the underlying problem was identified. The acute episodes were diagnosed as pancreatitis. It was thought that his long survival was related to absolute abstinence from alcohol and a low fat diet [69]. The Biochemical Disorder The usual basic defect is deficiency of lipoprotein lipase, which is also called triacylglycerol acylhydrolase [70]. A similar phenotype can also result from deficiency

Dodge

of the activator of lipoprotein lipase, apolipoprotein C-II [71]. As with cystic fibrosis, the nature of the protein abnormality varies somewhat between patients. Thus, there may be absence of lipoprotein lipase (class I) or defective protein which either binds (class II) or does not bind (class III) to heparin [72]. The gene for lipoprotein lipase is located on 8p22 and codes for a protein of 475 amino acids. It is very similar to hepatic lipase and pancreatic lipase [73]. Numerous mutations have been described, and mouse models have been created. The Heterozygous State Heterozygotes do not usually display the gross clinical features of lipoprotein lipase deficiency, although enzyme activity is only 50% of normal. This may not suffice to keep the plasma triglyceride concentration within normal limits when stress is placed on the plasma lipid transport system, for example, after fatty meals or during pregnancy [74]. Secondary factors include obesity, hyperinsulinaemia, lipid-raising drugs, and independently inherited noninsulin-dependent diabetes mellitus [75]. Contrary to earlier reports, recent work suggests that, at least with some mutations, heterozygotes are not protected against atherosclerosis and may even have a predisposition to ischaemic heart disease [76, 77]. Why Pancreatitis? It has been suggested that high serum triglyceride levels lead to pancreatic injury via the breakdown of excessive triglyceride within the pancreatic microcirculation, and the release of free fatty acids. If intracellular antioxidant defence mechanisms are inadequate, lipid peroxidation

will occur on exposure to oxygen free radicals. Oxidation of unsaturated fatty acids in cell membranes then occurs in a domino fashion, and cell membranes disintegrate. Free radicals are chemotactic for polymorphonuclear leucocytes in the tissues involved, and activated polymorphs themselves secrete a variety of active substances – including further oxygen radicals – that exacerbate tissue damage. Furthermore, release of arachidonic acid from cell membranes is followed by synthesis of prostaglandins and leukotrienes, which further enhance inflammation [67]. The pathways involved in the pathogenesis of acute episodes of pancreatitis are clearly complex, with the probability of one or more vicious circles, and it is difficult to identify the exact factors which may initiate, as distinct from potentiate, the process. However, experimental studies suggest that oxidant stress may be a triggering factor [78], by rapidly depleting pancreatic glutathione [79].

Concluding Comments Recurrent or chronic pancreatitis is a recognised feature of many other conditions. These include Wilson’s disease [80], partial lipodystrophy [81], ·1-antitrypsin deficiency [82], haemochromatosis [83], hyperparathyroidism [84], Pearson’s syndrome [85] which is a rare disorder with persistent lactic acidosis and mitochondrial aberrations [85], and organic acidaemias [86]. A plausible case can be made for oxidant stress as a major contributor to pancreatic injury in all these disorders, as also in cystic fibrosis, hereditary pancreatitis and lipoprotein lipase deficiency [87].

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

References 1 Zeigler DW, Long JA, Phillippart AI, Kelin MD: Pancreatitis in childhood. Ann Surg 1988; 207:257–261. 2 Roberts IM: Disorders of the pancreas in children. Gastroenterol Clin N Am 1990;19:963– 973. 3 Materne R, Clapuyt P, Saint-Martin C, Jespers S, Barrea C, de Ville de Goyet J, Gosseye S, Sokal E: Gastric cystic duplication communicating with a bifid pancreas: A rare case of recurrent pancreatitis. J Pediatr Gastroenterol Nutr 1998;27:102–105. 4 Sarles H, Adler G, Dani R, Frey C, Gullo L, Harada H, Martin E, Norohna M, Scuro LA: Classifications of pancreatitis and definitions of pancreatic diseases. Digestion 1989;43:234– 236.

Paediatric Chronic Pancreatitis

5 Gabriel SE, Brigman KN, Koller BH, Boucher RC, Stutts MJ: Cystic fibrosis heterozygote resistance to cholera toxin in the cystic fibrosis mouse model. Science 1994;266:107–109. 6 Goldstein JL, Sahi J, Bhuva M, Layden TJ, Rao MC: Escherichia coli heat stable enterotoxin-mediated colonic chloride secretion is absent in cystic fibrosis. Gastroenterology 1994; 107:950–956. 7 Pier GB, Grout, M, Zaidi T, Meluleni, G, Mueschenborn SS, Banting G, Ratcliff R, Evans MJ, Colledge WH: Salmonella typhi uses CFTR to enter synovial epithelial cells. Nature 1998;393:79–82.

8 Kerem B-S, Rommens JM, Buchanan JA, Markiewicz D, Cox TK, Chakravarti A, Buchwald M, Tsui L-C: Identification of the cystic fibrosis gene: Genetic analysis. Science 1989;245: 1073–1080. 9 Riordan JR: The cystic fibrosis transmembrane conductance regulator. Annu Rev Physiol 1993;55:609–630. 10 Welsh M, Smith A: Molecular mechanisms of CFTR chloride channel dysfunction in cystic fibrosis. Cell 1993;73:1251–1254. 11 Borgo G, Mastella G, Gasparini P, Zorganello A, Doro P, Pignatti PF: Pancreatic function and gene deletion F508 in cystic fibrosis. J Med Genet 1990;27:665–669.

Digestion1998;59(suppl 4):49–59

57

12 Gaskin K, Gurtwitz D, Durie PR, Corey M, Levison H, Forstner G: Improved respiratory prognosis in CF patients with normal fat absorption. J Pediatr 1982;100:857–862. 13 Porta EA, Stein AA, Patterson P: Ultrastructural changes in the pancreas and liver in cystic fibrosis. Am J Clin Pathol 1964;42:451–465. 14 Stein AA, Porta E, Powers S Jr, Leather R, Linton P, Patterson P: Studies on surgical biopsies of pancreas and liver in four cases of cystic fibrosis: Morphologic and histochemical. Ann Surg 1963;157:516–524. 15 Oppenheimer EH, Esterly JR: Pathology of cystic fibrosis: Review of the literature and comparison with 146 autopsied cases. Perspect Pediatr Pathol 1975;2:241–278. 16 Craig BG, Johnston CF, Kerr JI, Redmond AOB, Dodge JA, Buchanan KD: Evidence for pancreatic endocrine cell neoformation in cystic fibrosis. Can J Physiol Pharmacol 1986; (suppl):399. 17 Waters DL, Dorney SFA, Gaskin KJ, Gruca MA, O’Halloran M, Wilcken B: Pancreatic function in infants identified as having cystic fibrosis in a neonatal screening program. N Engl J Med 1990;322:303–308. 18 Atlas AB, Orenstein SR, Orenstein DM: Pancreatitis in young children with cystic fibrosis. J Pediatr 1992;120:756–759. 19 Fernando del Rosario J, Putnam PE, Orenstein DM: Chronic pancreatitis in a patient with cystic fibrosis and clinical pancreatic insufficiency. J Pediatr 1995;126:951–952. 20 Boué A, Muller F, Nezelof C, Aubry MC, Dumez Y, Oury JF, Duchatel F, Boué J: Prenatal diagnosis in 200 pregnancies with a 1-in-4 risk of cystic fibrosis. Hum Genet 1986;74:288– 297. 21 Carrère J, Muller F, Boué A: Levels and molecular forms of immunoreactive trypsin and chymotrypsin in aminiotic fluids from normal and cystic fibrosis fetuses: evidence for a lack of activation of proteolytic zymogens in the cystic fibrosis fetus. J Pediatr Gastroenterol Nutr 1992;14:198–203. 22 Imrie JR, Fagan DG, Sturgess JM: Quantitative evaulation of the development of the exocrine pancreas in cystic fibrosis and control infants. Am J Pathol 1979;95:697–708. 23 Dorin JR, Dickinson P, Alton EWAJ, Smith SN, Geddes DM, Stevenson BJ, Kimber WL, Fleming S, Clarke AR, Hooper ML, Anderson L, Beddington RSP, Porteous DJ: Cystic fibrosis in the mouse by targeted insertional mutagenesis. Nature 1992;359:211–214. 24 Ip WF, Bronsveld I, Kent G, Corey M, Durie PR: Exocrine pancreatic alterations in longlived surving cystic fibrosis mice. Pediatr Res 1996;40:242–249. 25 Figarella C, Carrère J: The evolution of pancreatic disease in cystic fibrosis; in Dodge JA, Brock DJH, Widdicombe JH (eds): Cystic Fibrosis: Current Topics. Chichester, Wiley, 1994, vol 2, pp 254–275. 26 Kopelman H, Durie, PR, Gaskin K, Weizman Z, Forstner G: Pancreatic fluid secretion and protein hyperconcentration in cystic fibrosis. N Engl J Med 1985;312:329–334.

58

27 Knauff RE, Adams JA: Proteins and mucoproteins in the duodenal fluids of cystic fibrosis and control subjects. Clin Chim Acta 1968;19: 245–248. 28 Lebenthal E, Lee PC: Effect of pancreozymin and secretin on intraluminal enterokinase, trypsin and chymotrypsin activities of cystic fibrosis and control children. Digestion 1982; 23:39–47. 29 Keim V, Iovanna J, Orelle B, Verdier JM, Busing M, Hopt U, Dagorn JC: A novel exocrine protein associated with pancreas transplantation in humans. Similarities to rat pancreatitisassociated protein (PAP). Gastroenterology 1992;103:248–254. 30 Watanabe T, Yonekura H, Terazano K, Yamomoto H, Oramoto H: Complete nucleotide sequence of human Reg gene and its expression in normal and tumoral tissues: The Reg protein, pancreatic stone protein and pancreatic thread protein are one and the same product of the gene. J Biol Chem 1990;265:7432–7439. 31 Guy-Grotte O, Amouric M, Figarella C: Characterisation and N-terminal sequence of a degradation product of 14000 molecular weight isolated from human pancreatic juice. Biochem Biophys Res Commun 1984;125:516–523. 32 Forstner GG, Vesely SM, Durie PR: Selective precipitation of 14 kDa stone/thread proteins by concentration of pancreatico biliary secretions: relevance to pancreatic ductal obstruction, pancreatic failure and CF. J Pediatr Gastroenterol Nutr 1989;8:313–320. 33 Leach SD, Modlin IM, Scheele GA, Gorelick FS: Intracellular activation of digestive zymogens in rat pancreatic acini: stimulation by high doses of cholecystokinin. J Clin Invest 1991;87: 362–366. 34 Kopelman H, Corey M, Gaskin K, Durie P, Weizman Z, Forstner G: Impaired chloride secretion, as well as bicarbonate secretion, underlies the fluid secretory defect in the cystic fibrosis pancreas. Gastroenterology 1988;95: 349–355. 35 Scheele GA, Fukuoka SI, Kern HF, Freedman SD: Pancreatic dysfunction in cystic fibrosis occurs as a result of impairments in luminal pH, apical trafficking of zymogen granule membranes and solubilization of secretory enzymes. Pancreas 1996;12:1–9. 36 Rinderknecht H: Acute necrotizing pancreatitis and its complications: an excessive reaction of natural defence mechanisms?; in Braganza J (ed): The Pathogenesis of Pancreatitis. Manchester, Manchester University Press, 1991, pp 86–100. 37 Bradbury NA, Jilling T, Berta G, Sorscher EJ, Bridges RJ, Kirk KL: Regulation of plasma membrane recycling by CFTR. Science 1992; 256:530–531. 38 Adelson JW, Miller PE: Pancreatic secretion by non parallel exocytosis: Potential resolution of a long controversy. Science 1985;228:993– 996. 39 De Lisle RC: Increased expression of sulfated gp300 and acinar tissue pathology in pancreas of CFTR (–/–) mice. Am J Physiol 1995;268: G717–723.

Digestion1998;59(suppl 4):49–59

40 Kopelman H, Ferratti E, Gauthier C, Goodyer PR: Rabbit pancreatic acini express CFTR as a cAMP-activated chloride efflux pathway. Am J Physiol 1995;269:C626–C631. 41 Marino CR, Matovcik LM, Gorelick FS, Cohn JA: Localisation of the cystic fibrosis transmembrane conductance regulator in pancreas. J Clin Invest 1991;88:712–716. 42 Quesnel LB, Jaran AS, Braganza JM: Antibiotic accumulation and membrane trafficking in cystic fibrosis cells. J Antimicrobial Chemother 1998;41:215–221. 43 Salh B, Webb K, Guyan PM, Day JP, Wickens D, Griffin J, Braganza JM, Dormany TL: Aberrant free radical activity in cystic fibrosis. Clin Chim Acta 1989;181:65–74. 44 Winklhofer-Roob BM: Oxygen free radicals and antioxidation in cystic fibrosis: The concept of an oxidant-antioxidant imbalance. Acta Paediatr 1994;(suppl 395):49–57. 45 Laroche D, Travert G: Abnormal frequency of · F508 mutation in neonatal transitory hypertrypsinaemia. Lancet 1991;337:55. 46 Senegas-Balas F, Figarella C, Amouric M, GuyCrotte O, Bertrand C, Balas D: Immunohistochemical demonstration of a pancreatic secretory protein of unknown function in the human duodenum. J Histochem Cytochem 1991;39: 915–919. 47 Sharer N, Schwarz M, Malone G, Howarth A, Painter J, Super M, Braganza J: Mutations of the cystic fibrosis gene in patients with chronic pancreatitis. N Engl J Med 1998;339;645–652. 48 Cohn JA, Friedman KJ, Noone PG, Knowles MR, Silverman LM, Jowell PS: Relation between mutations of the cystic fibrosis gene and idiopathic pancreatitis. N Engl J Med 1998;339:653–658. 49 Portal BC, Richard M-J, Faure HS, Hadjian AJ, Favier AE: Altered antioxidant status and increased lipid peroxidation in children with cystic fibrosis. Am J Clin Nutr 1995;61:843– 847. 50 Winklhofer-Roob BM, Puhl H, Khoschsorur G, Van t Hof MA, Esterbauer H, Shmerling D: Enhanced resistance to oxidation of low density lipoproteins and decreased lipid peroxide formation during ß-carotene supplementation in cystic fibrosis. Free Radical Biol Med 1995; 18:849–859. 51 Lepage G, Champagne J, Ronco N, Lamarre A, Osberg I, Sokol RJ, Roy CC: Supplementation with carotenoids corrects increased lipid peroxidation in children with cystic fibrosis. Am J Clin Nutr 1996;64:87–93. 52 Werner J, Ferna´ndez-del-Castillo C, Rivera JA, Kollias N, Lewandrowski KB, Rattner DW, Warshaw AL: On the protective mechanisms of nitric oxide in acute pancreatitis. Gut 1998;43: 401–407. 53 Sossenheimer MJ, Aston CE, Preston RA, Gates LK Jr, Ulrich CD, Martin SP, Zhang Y, Gorry MC, Ehrlich GD, Whitcomb DC: Clinical characteristics of hereditary pancreatitis in a large family, based on high risk haplotype. Am J Gastroenterol 1997;92:1113–1116. 54 Singer M, Cohen FB: Hereditary chronic relapsing pancreatitis. J Newark Beth Israel Hospital 1966;21:121–126.

Dodge

55 Mann TP, Rubin J: Familial pancreatic exocrine dysfunction with pancreatic calcification. Proc R Soc Med 1969;62:326. 56 Sibert JR: Hereditary pancreatitis in England and Wales. J Med Genet 1978;15:189–201. 57 Sarles H, Camarena J, Bernard JP, Sahel H, Laugier R: Two forms of hereditary chronic pancreatitis. Pancreas 1996;12:138–141. 58 Lewis MPN, Gazet J-C: Hereditary calcific pancreatitis in an English family. Br J Surg 1993;80:487–488. 59 Lowenfels AB, Maisonneuve P, DiMagno EP, Elitsur Y, Gates LK Jr, Perrault J, Whitcomb DC: The International Hereditary Pancreatitis Study Group: Hereditary pancreatitis and the risk of pancreatic cancer. J Natl Cancer Inst 1997;89:442–446. 60 Le Bodic L, Bignon J-D, Raguenes O, Mercier B, Georgelin T, Schnee M, Soulard F, Gagne K, Bonneville F, Muller J-Y, Bachner L, Ferec C: The hereditary pancreatitis gene maps to long arm of chromosome 7. Hum Molec Genet 1996;5:549–554. 61 Whitcomb DC, Preston RA, Aston CE, Sossenheimer MJ, Barua PS, Zhang Y, Wong-Chong A, White GJ, Wood PG, Gates LK, Jr, Ulrich C, Martin SP, Post J C, Ehrlich GC: A gene for hereditary pancreatitis maps to chromosome 7q35. Gastroenterology 1996;110:1975–1980. 62 Whitcomb DC, Gorry MC, Preston RA, Furey W, Sossenheimer MJ, Ulrich CD, Martin SP, Gates LK Jr, Amann ST, Toskes PP, Liddle R, McGrath K, Uomo G, Post JC, Ehrlich GD: Hereditary pancreatitis is caused by a mutation in the cationic trypsinogen gene. Nature Genet 1996;14:141–145. 63 Dasouki MJ, Cogan J, Summar ML, Neblitt W III, Foroud R, Koller D, Phillips JA III: Heterogeneity in hereditary pancreatitis. Am J Med Genet 1998;77:47–53. 64 Chiari H: Über Selbstverdauung des menschlichen Pankreas. Zt Heilkunde 1896;17:69–96. 65 Steer ML, Meldolesi J: The cell biology of experimental pancreatitis. N Engl J Med 1987; 316:144–150.

Paediatric Chronic Pancreatitis

66 Prasad M, Wyllie R, Van Lente F, Steffen RM, Kay MH: Antioxidants in hereditary pancreatitis. Am J Gastroenterol 1996;91:1558–1562. 67 Cameron JL, Capuzzi DM, Zuidema GD, Margolis S: Acute pancreatitis with hyperlipaemia: Evidence of a persistent defect in lipid metabolism. Am J Med 1974;56:482–489. 68 Prasad M, Wyllie R, Caulfield M, Steffen R, Kay M: Chronic pancreatitis in late childhood and adolescence. Clin Pediatr (Philadelphia) 1994;33:88–94. 69 Hoeg JM, Osborne JC Jr, Gregg RE, Brewer HB Jr: Initial diagnosis of lipoprotein lipase deficiency in a 75-year-old man. Am J Med 1983;75:889–892. 70 Havel RJ, Gordon RS: Idiopathic hyperlipaemia: metabolic studies in an affected family. J Clin Invest 1960;39:1777–1790. 71 Breckenridge WC, Little AC, Steiner G, Chow A, Poapst M: Hypertriglyceridemia associated with deficiency of C-11 apoprotein in plasma lipoproteins. N Engl J Med 1978;298:1265– 1273. 72 Auwerx JH, Babirak SP, Fujimoto WY, Iverius PH, Brunzell JD: Defective enzyme protein in lipoprotein lipase deficiency. Eur J Clin Invest 1989;19:433–437. 73 Deeb SS, Peng R: Structure of the human lipoprotein lipase gene. Biochemistry 1989;28: 4131–4135. 74 Ma Y, Liu M-S, Ginzinger D, Frohlich J, Brunzell JD, Hayden MR: Gene-environment interaction in the conversion of a mild-to-severe phenotype in a patient homozygous for a ser172-to-cys mutation in the lipoprotein lipase gene. J Clin Invest 1993;91:1953–1958. 75 Wilson DE, Hata A, Kwong LK, Lingam A, Shuhua J, Ridinger DN, Yeager C, Kaltenborn KC, Iverius P-H, Lalouel J-M: Mutations in exon 3 of the lipoprotein lipase gene segregating in a family with hypertriglyceridemia, pancreatitis, and non-insulin-dependent diabetes. J Clin Invest 1993;92:203–211. 76 Benlian P, De Gennes JL, Foubert L, Zhang H, Gagne SE, Hayden M: Premature atherosclerosis in patients with familial chylomicronemia caused by mutations in the lipoprotein lipase gene. N Engl J Med 1996;335:848–854.

77 Wittrup HH, Tybjaerg-Hansen A, Abildgaard S, Steffensen R, Schnohr P, Nordesgaard BG: A common sustitution (asn291ser) in lipoprotein lipase is associated with increased risk of ischemic heart disease. J Clin Invest 1997;99: 1606–1613. 78 Sanfey H: Oxygen free radicals in experimental pancreatitis; in, Braganza JM (ed): The Pathogenesis of Pancreatitis. Manchester, Manchester University Press, 1991, pp 66–85. 79 Neuschwander-Tetri BA, Presti ME, Wells LD: Glutathione synthesis in the exocrine pancreas. Pancreas 1997;14:342–349. 80 Weizman Z, Picard E, Barki Y, Moses S: Wilson’s disease associated with pancreatitis. J Pediatr Gastroenterol Nutr 1988;7:931–933. 81 Smith PM, Morgans ME, Clark CG, LennardJones JE, Gunnlaugsson O, Jonasson TA: Lipodystrophy, pancreatitis, and eosinophilia. Gut 1975;16:230–234. 82 Mihas AA, Hirschowitz BE: Alpha1-antitrypsin and chronic pancreatitis. Lancet 1976;ii: 1032–1033. 83 Shimizu M, Hirokawa M, Manabe T: Histological assessment of chronic pancreatitis at necropsy. J Clin Pathol 1996;49:913–915. 84 Bess MA, Edis AJ, van Heerden JA: Hyperparthyroidism and pancreatitis. J Am Med Assoc 1980;243:246–257. 85 Pearson HA, Lobel JS, Kochoshis SA, Naiman JL, Windmiller J, Lammi AT, Hoffman R, Marsh JC: A new syndrome of refractory sideroblastic anemia with vacuolization of marrow precursors and exocrine pancreatic dysfunction. J Pediatr 1979;95:976–984. 86 Kahler SG, Sherwood WG, Woolf D, Lawless ST, Zaritsky A, Bonham J, Taylor CJ, Clarke JTR, Durie P, Leonard JV: Pancreatitis in patients with organic acidemias. J Pediatr 1994;124:239–243. 87 Braganza JM: The pathogenesis of chronic pancreatitis. Q J Med 1996;89:243–250.

Digestion1998;59(suppl 4):49–59

59

Subject Index

Acute pancreatitis 16, 25 African Blacks 25 Alcohol 21, 27 Antioxidant therapy 9, 22, 36 Antioxidants 1 Chronic pancreatitis 1, 13, 25, 36, 49 Cyanide 13 Cystic fibrosis 5, 49 Cytochrome P450 1, 13, 28, 37 Free radicals 36 Glutathione 1, 13 Hereditary pancreatitis 5, 10, 49 Hyperlipidaemia 46, 49 Methionine 1, 36 Oxidant stress 1, 13, 25, 36, 49 Pancreatic surgery 36 Plant nitriles 14 Selenium 31, 36 Signal transduction 1 Soweto 25 Thiol-disulphide exchange 4 Vitamin C 19, 29, 36 Xenobiotics 1, 13, 29

ABC Fax + 41 61 306 12 34 y-Mail [email protected] www.karger.com

© 1998 S. Karger AG, Basel

Accessible online at: http://BioMedNet.com/karger