Abortus Berulang Trapi Dan Penatalaksanaan

0415421306-FM 4/3/07 3:31 PM Page i
Recurrent Pregnancy Loss

0415421306-FM 4/3/07 3:31 PM Page ii
SERIES IN MATERNAL-FETAL MEDICINE
Forthcoming
Vincenzo Berghella, Obstetric Evidence Based Guidelines
ISBN 9780415701884
Vincenzo Berghella, Maternal-Fetal Evidence Based Guidelines
ISBN 9780415432818
Of related interest
Joseph J Apuzzio, Anthony M Vintzelos, Leslie Iffy, Operative Obstetrics
ISBN 9781842142844
Isaac Blickstein, Louis G Keith, Prenatal Assessment of Multiple Pregnancy
ISBN 9780415384247
Tom Bourne, George Condous, Handbook of Early Pregnancy Care
ISBN 9781842143230
Gian Carlo Di Renzo, Umberto Simeoni, The Prenate and Neonate: The transition to extrauterine life
ISBN 9781842140444
Asim Kurjak, Guillermo Azumendi, The Fetus in Three Dimensions: Imaging, Embryology and Fetoscopy
ISBN 9780415375238
Asim Kurjak, Frank A Chervenak, Textbook of Perinatal Medicine, second edition
ISBN 9781842143339
Catherine Nelson-Piercy, Handbook of Obstetric Medicine, third edition
ISBN 9781841845807
Dario Paladini, Paolo Volpe, Ultrasound of Congenital Fetal Anomalies
ISBN 9780415414449
Donald M Peebles, Leslie Myatt, Inflammation and Pregnancy
ISBN 9781842142721
Felice Petraglia, Jerome F Strauss, Gerson Weiss, Steven G Gabbe, Preterm Birth: Mechanisms, Mediators,
Prediction, Prevention and Interventions
ISBN 9780415392273
Ruben A Quintero, Twin-Twin Transfusion Syndrome
ISBN 9781842142981
Baskaran Thilaganathan, Shanthi Sairam, Aris T Papageorghiou, Amor Bhide, Problem Based Obstetric
Ultrasound
ISBN 9780415407281
0415421306-FM 4/3/07 3:31 PM Page iii
Recurrent Pregnancy Loss

Causes, Controversies and
Treatment
Edited by
Howard JA Carp MB BS FRCOG

Professor, Department of Obstetrics and Gynecology
Sheba Medical Center
Tel Hashomer
and
Sackler School of Medicine
Tel Aviv University
Tel Aviv
Israel
0415421306-FM 4/3/07 3:31 PM Page iv
This edition published in the Taylor & Francis e-Library, 2008.

“To purchase your own copy of this or any of Taylor & Francis or Routledge’s
collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.”
© 2007 Informa UK Ltd
First published in the United Kingdom in 2007 by Informa Healthcare, Telephone House, 69–77 Paul Street, London, EC2A 4LQ.
Informa Healthcare is a trading division of Informa UK Ltd. Registered office, 37/41 Mortimer Street, London W1T 3JH. Registered
in England and Wales number 1072954.
Tel: +44 (0)20 7017 5000

Fax: +44 (0)20 7017 6336
Website: www.informahealthcare.com
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of the publisher or in accordance
with the provisions of the Copyright, Designs and Patents Act 1988 or under the terms of any licence permitting limited copying issued
by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1P 0LP.
Although every effort has been made to ensure that all owners of copyright material have been acknowledged in this publication, we
would be glad to acknowledge in subsequent reprints or editions any omissions brought to our attention.
Although every effort has been made to ensure that drug doses and other information are presented accurately in this publication, the
ultimate responsibility rests with the prescribing physician. Neither the publishers nor the authors can be held responsible for errors
or for any consequences arising from the use of information contained herein. For detailed prescribing information or instructions on
the use of any product or procedure discussed herein, please consult the prescribing information or instructional material issued by
the manufacturer.
A CIP record for this book is available from the British Library.
Library of Congress Cataloging-in-Publication Data
Data available on application
ISBN 0-203-93167-X Master e-book ISBN
ISBN-10: 0 415 42130 6 (Print Edition)

ISBN-13: 978 0 415 42130 0
Distributed in North and South America by

Taylor & Francis
6000 Broken Sound Parkway, NW (Suite 300)
Boca Raton, FL 33487, USA
Within Continental USA

Tel: 1 (800) 272 7737; Fax: 1 (800) 374 3401
Outside Continental USA
Tel: (561) 994 0555; Fax: (561) 361 6018
Email: [email protected]
Distributed in the rest of the world by

Thomson Publishing Services
Cheriton House
North Way
Andover, Hampshire SP10 5BE, UK
Tel: +44 (0)1264 332424
Email: [email protected]
Composition by CEPHA IMAGING PVT LTD, Bangalore, India.

0415421306-FM 4/3/07 3:31 PM Page v
Contents
List of Contributors vii
Foreword xi
Bruno Lunenfeld
Preface xiii
Howard JA Carp
1. Epidemiology of recurrent pregnancy loss 1

Ole B Christiansen
2. Signaling between embryo and mother in early pregnancy: Basis for development of tolerance 15
Eytan R Barnea
3. Genetics of spontaneous abortions 23

Joe Leigh Simpson
Debate: Should fetal karyotyping be performed in RPL?

3a. For Howard JA Carp 35
3b. Against Zvi Borochowitz 39
Debate: Should PGD be performed in RPL?

3c. For Howard JA Carp, and JG Grudzinskas 45
3d. Against Patricio Donoso, Andre Van Steirteghem, and Paul Devroey 49
3e. Opinion: Should CVS or amniocentesis be performed in RPL without screening? 55
Howard Cuckle
4. Does the maternal immune system regulate the embryo’s response to teratogens? 59
Arkady Torchinsky and Vladimir Toder
5. Fetal structural malformations – embryoscopy 67
Thomas Philipp
6. Endocrinology of pregnancy loss 79
Stefano Luisi, Lucia Lazzeri, and Andrea Riccardo Genazzani
Debate: Should progesterone supplements be used?

6a. For Jerome H Check 89
6b. Against Shazia Malik and Lesley Regan 93
Debate: Should hCG supplementation be used?

6c. For James Walker 97
6d. Against Siobhan M Quenby and Roy G Farquharson 101
7. Antiphospholipid syndrome – pathophysiology 107

Gilad Twig, Yaniv Sherer, Miri Blank, and Yehuda Shoenfeld
8. Diagnosis of aPL-associated abortions 115

Marighoula Varla-Leftherioti
v
0415421306-FM 4/3/07 3:31 PM Page vi
CONTENTS
9. Management of antiphospholipid syndrome in pregnancy 119

Wendell A Wilson and Nigel Harris
10. Defects in coagulation factors leading to recurrent pregnancy loss 127

Aida Inbal and Howard JA Carp
Debate: Should thromboprophylaxis be used in hereditary thrombophilias with RPL?

10a. For Benjamin Brenner 139
10b. Against Pelle G Lindqvist 143
11. Uterine anomalies and recurrent pregnancy loss 147

Daniel S Seidman and Mordechai Goldenberg
12. Immunobiology of recurrent miscarriage 165

Debate: Should immunotherapy be used in RPL?

12a. Paternal leukocyte immunization – For David Clark 179
12b. Intravenous immunoglobulin – For Carolyn B Coulam 185
12c. Against Raj Rai 189
13. Infections and recurrent pregnancy loss 193

David Alan Viniker
14. Midtrimester loss – the role of cerclage 207

Israel Hendler and Howard JA Carp
15. Midtrimester loss and viability 219

Victor YH Yu
16. Obstetric outcomes after recurrent pregnancy loss 231

Howard JA Carp
17. Coping with recurrent pregnancy loss: Psychological mechanisms 243

Keren Shakhar
18. Methodological issues in evidence-based evaluation of treatment for recurrent miscarriage 255
Salim Daya
19. Investigation protocol for recurrent pregnancy loss 269

Howard JA Carp
20. A patient’s perspective 281

Mindy Gross
Index 285
vi
0415421306-FM 4/3/07 3:31 PM Page vii
Contributors
Eytan R Barnea MD FACOG Howard JA Carp MB BS FRCOG
Chairman, The Society for the Investigation of Professor, Department of Obstetrics and
Early Pregnancy Gynecology
Clinical Associate Professor of Obstetrics, Sheba Medical Center
Gynecology and Reproductive Sciences Tel Hashomer
UMDNJ/Robert Wood Johnson Medical School and
Cherry Hill, NJ Sackler School of Medicine
USA Tel Aviv University
Tel Aviv
Miri Blank PhD Israel
Associate Professor of Immunology
Sackler Faculty of Medicine Jerome H Check MD PhD
Tel-Aviv University Robert Wood Johnson Medical School at Camden
and Melrose Park, PA
Center for Autoimmune Diseases and
Sheba Medical Center University of Medicine and Dentistry of
Tel-Hashomer New Jersey
Israel Newark, NJ
USA
Zvi Borochowitz MD
Professor of Pediatrics and Medical Genetics, Ole B Christiansen MD DMSC
Director, Simon Winter Institute for Human Fertility Clinic
Genetics Rigshospitalet
and Copenhagen
Institute for Human Genetics Denmark
Technion-Rappaport Faculty of Medicine
Haifa David Clark MD PhD DiplABIM FRCPC FRCPED
Israel Emeritus Professor of Medicine
McMaster University
Benjamin Brenner MD Hamilton
Professor of Hematology, Technion Israel Institute Ontario
of Technology Canada
Director, Thrombosis and Hemostasis Unit
Rambam Medical Center Carolyn Coulam MD
Haifa Director, Millennova Immunology Laboratories;
Israel Director, Pregnancy Loss Center
and
Director of Research, Rinehart Center for
Reproductive Medicine
Chicago, IL
USA
vii
0415421306-FM 4/3/07 3:31 PM Page viii
LIST OF CONTRIBUTORS
Howard Cuckle MA MSc DPhil Mordechai Goldenberg MD

Adjunct Professor of Obstetrics and Gynecology Professor, Chaim Sheba Medical Center
Columbia University Tel-Hashomer
New York, NY and
USA The Sackler School of Medicine
and Tel Aviv University
Emeritus Professor Tel Aviv
University of Leeds Israel
Leeds
UK Mindy Gross
Ra′anana
Salim Daya MBC hB MSc FRCSC Israel
Hamilton
Ontario JG Grudzinskas MD
Canada The London Bridge Fertility
Gynaecology and Genetics Centre
Paul Devroey MD PhD London
Professor, Centre for Reproductive Medicine UK
University Hospital
Dutch-speaking Brussels Free University Nigel Harris MD DM
Belgium Vice Chancellor, University of West Indies
Kingston
Patricio Donoso MD Jamaica
Centre for Reproductive Medicine
University Hospital Israel Hendler MD
Dutch-speaking Brussels Free University Assistant Professor, Wayne State University
Belgium Detroit, MI
USA
Roy G Farquharson MD FRCOG and
Clinical Director, Liverpool Women’s Hospital Sheba Medical Center
Liverpool Tel Hashomer
UK Israel
Andrea Riccardo Genazzani MD PhD Aida Inbal MD

Director of Obstetrics and Gynecology Associate Professor of Hematology
Hospital S Chiara Sackler Faculty of Medicine
University of Pisa Tel Aviv University
Pisa and
Italy Director of Thrombosis and Hemostasis Unit
Rabin Medical Center
Petah Tikva
Israel
viii
0415421306-FM 4/3/07 3:31 PM Page ix
Lucia Lazzeri MD Lesley Regan MD FRCOG

Section of Obstetrics and Gynecology Professor of Obstetrics and Gynaecology
Policlinics Le Scotte Imperial College London
University of Siena St Mary’s Hospital
Siena London
Italy UK
Pelle G Lindqvist MD PhD Daniel S Seidman MD MMSc

Associate Professor and Senior Consultant Professor, Chaim Sheba Medical Center
Malmö University Hospital Tel-Hashomer
Malmö and
Sweden Sackler School of Medicine
Tel Aviv University
Stefano Luisi MD PhD Tel Aviv
Section of Obstetrics and Gynecology Israel
Policlinics Le Scotte
University of Siena Keren Shakhar PhD
Siena Department of Oncological Sciences
Italy Mount Sinai School of Medicine
New York, NY
Shazia Malik MBChB (hon), MRCOG USA
St Mary’s Hospital NHS Trust
London Yaniv Sherer MD
UK Instructor in Internal Medicine
Sackler Faculty of Medicine
Thomas Philipp MD Tel Aviv University
Assistant Professor Tel Aviv
Ludwig Boltzman Institute of Gynecology and and
Obstetrics Center for Autoimmune Diseases
Danube Hospital Sheba Medical Center
Vienna Tel-Hashomer
Austria Israel
Siobhan M Quenby MD MRCOG Yehuda Shoenfeld MD

Senior Lecturer, University Department of Professor of Internal Medicine
Obstetrics and Gynaecology Sackler Faculty of Medicine,
Liverpool Women’s Hospital Incumbent of the Laura Schwartz Kipp Chair for
Liverpool Research of Autoimmune Diseases
UK Tel-Aviv University
Tel Aviv
Raj Rai BSc MD MRCOG and
Senior Lecturer and Consultant Gynaecologist Director, Department of Medicine ‘B’ and Center
Imperial College London for Autoimmune Diseases
St Mary’s Hospital Sheba Medical Center
London Tel-Hashomer
UK Israel
ix
0415421306-FM 4/3/07 3:31 PM Page x
Joe Leigh Simpson MD Marighoula Varla-Leftherioti MD PhD

Professor of Obstetrics and Gynecology Head of the Immunobiology Department
Professor of Molecular and Human Genetics ‘Helena Venizelou’ Maternity Hospital
Baylor College of Medicine Athens
Houston, TX Greece
USA
David Alan Viniker MD FRCOG
Vladimir Toder MD PhD Consultant Obstetrician and Gynaecologist
Professor in Embryology Whipps Cross University Hospital
Sackler School of Medicine Loughton
Tel-Aviv University UK
Ramat-Aviv
Israel James Walker MD FRCOG FRCP Edin FRCPS Glas
Professor of Obstetrics and Gynaecology
Arkady Torchinsky MD PhD DSc St James’s University Hospital
Principal Research Associate Leeds
Sackler School of Medicine UK
Tel-Aviv University
Ramat-Aviv Wendell A Wilson MD FRCP
Israel Professor of Medicine, Arthritis and
Rheumatology
Gilad Twig MD PhD Center of Excellence
Center for Autoimmune Diseases Louisiana State University Medical Center
Sheba Medical Center New Orleans, LA
Tel-Hashomer USA
and
Sackler Faculty of Medicine Victor YH Yu MD MSc FRACP FRCP FRCP Ed FRCP Glas
Tel-Aviv University Professor of Neonatology
Tel Aviv Clinical Director of Ritchie Centre for Baby
Israel Health Research
Monash University
Andre Van Steirteghem MD PhD Melbourne
Professor Emeritus Australia
Centre for Reproductive Medicine
University Hospital
Dutch-Speaking Brussels Free University
Belgium
x
0415421306-FM 4/3/07 3:31 PM Page xi
Foreword
‘Children are the anchors that hold a mother to life’ This book will be welcomed by many investiga-
tors and clinicians working in the field of recurrent
Sophocles, Phaedra
pregnancy loss. There are chapters governing basic
In almost all traditions, the importance of procre- scientific topics such as cytokines, mechanisms of
ation is inherent in man’s very creation; both Old action of antiphospholipid antibodies, and signal-
and New Testaments of the Bible refer to the tragic ing between mother and fetus. The major advances
plight of barren women, eloquently describing the in genetics, including pregestational diagnosis,
pain and agony of childlessness. However, records immunology, endocrinology, and thrombotic
dated far earlier than the Bible confirm that fertility mechanisms, have been described in depth. The
has been a constant fundamental priority and pre- methodology of clinical research and the applica-
occupation, in all societies, throughout the ages of tion of evidence-based medicine to clinical practice
man. Fertility symbols are clearly identified in the have been explained comprehensively. The prob-
relics of prehistoric times, of ancient civilizations in lems of midtrimester loss and late obstetric compli-
all parts of the world, a recognition of the concept cations are aired, including the problems associated
that man’s existence depends upon the renewal of fer- with extreme prematurity and possible resulting
tility. The above quotation was written by Sophocles handicaps, and recent views on the role of cerclage
2500 years ago. The ancient Canaanites and Greeks in prevention. However, as is inevitable in clinical
had gods of fertility – Ashtarte and Hermes. Today, practice, there are many controversies, leaving the
infertility is recognized as a disease by the World clinician in a quandary, as how to help the patient.
Health Organization and by numerous healthcare Hence, there are six debates, and one opinion article,
providers throughout the world. Recurrent preg- that try to bring the relevant points before the
nancy loss represents one aspect of disordered fer- clinician, to aid in deciding the most appropriate
tility. Recurrent pregnancy loss has been described management. However, we must never forget that
as the ‘orphan’ of infertility, as this condition is often at the end of the line is a patient. Therefore, the
overlooked in the larger process of research and chapter on psychological mechanisms and the con-
management of fertility. Recurrent pregnancy loss is nection between psychological mechanisms and the
a heterogeneous condition, with numerous causes immune and other systems is welcome. The story told
and numerous treatment options. It is multidiscipli- by the patient in Chapter 22 is most touching, and
nary, involving gynecology, genetics, endocrinology, reminds us of the real problem at hand.
immunology, pediatrics, and internal medicine. It is hoped that this book will be read by specialists
Whatever the cause and possible treatment, the psy- working in recurrent pregnancy loss clinics and
chological implications are enormous. Both partners associated disciplines who wish to keep up to date,
may feel that they have failed in their parenting role. and generalists who wish to gain a comprehensive
Couples have divorced with mutual recriminations, view of developments in the field. The ‘half-life’ of
each blaming the other. Even when pregnancy does scientific knowledge has been said to be ten years.
succeed, it may be fraught with the fear of another However, many advances are occurring so quickly
loss. This anxiety is multipled when the diagnosis that a ‘half-life’ of ten years seems to be out of date.
remains unexplained. It is to be hoped that the advances in scientific and
xi
0415421306-FM 4/3/07 3:31 PM Page xii
FOREWORD
clinical knowledge will continue at this pace, in and to allow those still unable to have children to
order to improve the management of the patients fulfil this most basic of human desires.
Professor Bruno Lunenfeld MD PhD FRCOG
FACOG (Hon) P.O.G.S. (Hon)
Professor Emeritus at Faculty of Life Sciences

Bar-Ilan University
President of the International Society for the Study
of the Aging Male (ISSAM)
General Secretary of The Asian–Pacific Initiative
on Reproductive Endocrinology (ASPIRE)
Member of the Israel Government National
Council for Obstetrics, Genetics and Neonatology
xii
0415421306-FM 4/3/07 3:31 PM Page xiii
Preface
Prof Howard JA Carp MB BS FRCOG

Professor, Department of Obstetrics and Gynecology,
Sheba Medical Center, Tel Hashomer, and
Sackler School of Medicine,Tel Aviv University,
Tel Aviv, Israel
Recurrent pregnancy loss is a vexing clinical prob- also clinically useful, and which may improve the
lem as the cause often remains unexplained despite care of these couples.
the major advances in genetics and immunology. The book is planned for general gynecologists,
Treatment is often controversial and ranges from and specialists working in the field. Each con-
“masterly inactivity”, to an approach which could tributing author is an authority on a specific area of
be considered as “over aggressive”. The problem is recurrent pregnancy loss. I would like to thank each
distressing to couples, who understandably expect author for the time and effort taken in preparing
answers and solutions, and frustrating for the the manuscripts to make publication of this book
physician who often does not have these answers, possible. I would also like to thank those responsi-
particularly in the face of ever-changing and con- ble in a more indirect way for the publication of
flicting recommendations. This book tries to sum- this book; my teachers over the years, particularly
marize those controversies, and discuss the scientific Prof. Shlomo Mashiach for his constant help and
basis for various causes of pregnancy loss in depth, encouragement in my work in recurrent pregnancy
and to debate the various treatment modalities loss. Thanks go to my collaborators, Prof ’s Yehuda
which have been used in recent years. Hence, it is Shoenfeld, Aida Inbal, Ephraim Gazit, and Vladimir
hoped that this book describes the accumulating Toder, and special recognition goes to the greatest
data in a way which is both scientifically sound and teachers of all, the patients.
xiii
0415421306-FM 4/3/07 3:31 PM Page xiv
0415421306-Ch01 3/29/07 5:25 PM Page 1
1. Epidemiology of recurrent
pregnancy loss
Ole B Christiansen
INTRODUCTION at least three consecutive miscarriages, whereas

recurrent pregnancy loss (RPL) could also include
Epidemiology can be defined as ‘the scientific study pregnancy losses up to gestational week 28. However,
of disease frequency, determinants of disease, and the there is no consensus regarding the definition of
distribution of disease in a population’. The deter- recurrent miscarriage or RPL.2 Pregnancy losses
minants of disease considered in epidemiological after week 20 are rare, so defining recurrent miscar-
studies are normally demographic parameters (age, riage and RPL as above will result in almost identical
sex, occupation, and economic status), in addition populations.
to some clinical parameters relevant for the specific Unfortunately, many studies include women with
disease (e.g., tobacco and alcohol consumption, and only two previous miscarriages in studies of recur-
reproductive and family history) – all information rent miscarriage/RPL, which, from an epidemiological
that can be obtained through registers and ques- point of view, is very problematic. This issue will be
tionnaires, whereas parameters requiring special discussed later.
interventions such as blood samples are normally
not included in purely epidemiological studies.
EPIDEMIOLOGICAL PARAMETERS
RELEVANT FOR RPL
DEFINITION OF MISCARRIAGE AND RECURRENT
PREGNANCY LOSS OCCURRENCE
The term ‘miscarriage’ (or ‘abortion’) is used to The incidence of RPL is the number of new women
describe a pregnancy that fails to progress, resulting each year (or in some other defined period) suffer-
in death and expulsion of the embryo or fetus. ing their third consecutive pregnancy loss, while the
The generally accepted definition stipulates that the prevalence of RPL is the number of women in a
fetus or embryo should weigh 500 g or less – a stage population who, at a specific timepoint, have had
that corresponds to a gestational age of up to three or more consecutive pregnancy losses. The
20 weeks, according to the World Health Organization incidence or prevalence is often expressed as a rate
(WHO).1 Unfortunately, this definition is not used of those individuals being at risk for the disorder.
consistently, and pregnancy losses at higher gesta- The number in the denominator could be all
tional ages are also classified as miscarriage in some women in the population, women of fertile age, or
studies instead of as stillbirth or preterm neonatal women who had attempted pregnancy at least two
death. Thus, from a definitional perspective, it is or three times. Indeed, the estimate of the incidence
important to characterize the population being stud- or prevalence of RPL is very uncertain, since in
ied so that comparisons across therapeutic trials can most countries there is no nationwide registration
be made more appropriately and reliably. of miscarriages or RPL, and many early miscar-
Recurrent miscarriage should be defined, riages are not treated in hospitals and thus are not
according to the above definition of miscarriage, as registered. There is no valid estimate of the incidence
1
0415421306-Ch01 3/29/07 5:25 PM Page 2
RECURRENT PREGNANCY LOSS: CAUSES, CONTROVERSIES AND TREATMENT
of RPL, whereas there are a few estimates of its preva- previous miscarriages increases. The chance of
lence. One of the most informative studies of the subsequent live birth in untreated RPL patients
prevalence of RPL was performed by Alberman,3 who with 3, 4, and 5 or more miscarriages has been
asked female doctors to report retrospectively about found to be 42–86%, 41–72%, and 23–51%, respec-
the outcome of their previous pregnancies. Seven tively7–10 (Figure 1.1). The significant variability in
hundred and forty-two women had had three previ- the estimate of the subsequent risk of miscarriage in
ous pregnancies, and 355 women had had four previ- RPL patients can probably be attributed to the time
ous pregnancies. Nine women (0.8%) had had three of ascertainment of the pregnancies (Figure 1.2),
or more consecutive pregnancy losses. This study is since the average age of the patients and the dura-
probably the best estimate of the prevalence of RPL, tion of follow-up in the various studies were not
since the cohort was restricted to women who had different. The information in Figure 1.2 is based on
attempted pregnancy at least three times. As the study the data given in the literature8,10,11 or data that can
consisted of physicians, delayed menstruation, unequivocally be deduced from the literature. In
induced abortions, ectopic pregnancies, and miscar- studies where the patients are urged to contact the
riages were unlikely to be misclassified. However, department for inclusion in a treatment trial as
since the study was carried out before 1980, many soon as menstruation is delayed by 2–3 days, and a
early miscarriages may not have been registered due highly sensitive pregnancy test is positive,10 almost
to a lack of highly sensitive human chorionic all preclinical losses (including biochemical preg-
gonadotropin (hCG) tests and ultrasound examina- nancies) are identified, and the patients will be reg-
tions at that time. Furthermore, female doctors may istered as having a high fetal loss rate (47.1%) but a
not reflect the background population: they may be low non-pregnancy rate (14.7%) during the obser-
healthier than other women, which may lower the vation period. In studies where the patients are told
miscarriage risk, but (due to their long education) to call the department in gestational week 6–7 to be
they may be older than other women when attempt- included in treatment trials,11 or patients are
ing pregnancy, which increases the miscarriage risk.
Other estimates of the prevalence of RPL are gen-
erally in accordance with that of Alberman. A RPL
prevalence of 2.3% was found in 432 randomly iden- 3 miscarriages 4 miscarriages ≥ 5 miscarriages
tified women in a multicenter study.4 In a group of
90
5901 Norwegian women with at least two pregnan- 80
cies screened for Toxoplasma antibodies, 1.4% had 70
Birth rate (%)
experienced RPL.5 Data from a Danish question- 60

50
naire-based study6 found, in a random sample of 493 40
women with at least two intrauterine pregnancies, 30
that 0.6% had had at least three consecutive miscar- 20
10
riages, 0.8% at least three consecutive pregnancy
0
losses during all trimesters, and 1.8% at least three, 9
al on
8
ith
7 10
al
et rs Sm et
not necessarily consecutive, losses at some time rd ha d en
iffo rq
u
an ns
during pregnancy. Overall, these studies thus find the Cl Fa k ia
nd h oc r ist
prevalence of RPL to be between 0.6% and 2.3%. ya wc Ch
en
b Co
Qu
NUMBER OF PREVIOUS MISCARRIAGES
Figure 1.1 Subsequent birth rate according to the number
Almost all prospective studies of RPL patients of previous miscarriages in patients with recurrent pregnancy
loss reported in four studies of untreated or placebo-treated
show remarkable consistency in finding an patients.
increasing risk of miscarriage as the number of
2
0415421306-Ch01 3/29/07 5:25 PM Page 3
EPIDEMIOLOGY OF RECURRENT PREGNANCY LOSS
Not pregnant Miscarriages Births
60
50
40
% 30
20
10
0
Jablonowska et al11 Quenby and Christiansen et al10
Farquharson8
Figure 1.2 Incidence of subsequent live births and miscarriages. Christiansen et al10 indicated the proportion of miscarriages as pre-
clinical and clinical (56.3% were clinical and 43.7% were preclinical). All miscarriages, except one in the series by Jablonowska et al11
series and all in the series by Quenby and Farquharson,8 were clinical. p = 0.001 for the differences between the not-pregnant rates
and p < 0.0001 for differences between the miscarriage rates in the three studies.
enrolled only after ultrasonographic demonstration to the fact that maternal age and the presence of
of fetal heart activity,8 most preclinical miscarriages age-related risk factors for miscarriages are posi-
are not ascertained, and therefore significantly tively correlated to gravidity. However, in multivari-
higher non-pregnancy rates (38.3–55.6%) and sig- ate analyses of clinical and paraclinical parameters
nificantly lower miscarriage rates (11.1–14.4%) are of potentially prognostic importance in RPL, the
registered (Figure 1.2). The subsequent probability number of previous miscarriages has without excep-
of a live birth in RPL can best be estimated from the tion remained the strongest prognostic parameter,
placebo arm of studies of RPL (Figure 1.2), because even after adjustment for other risk factors.7,13,14
in placebo-controlled trials the ascertainment of
pregnancies is generally better than in non-ran- MATERNAL AGE
domized studies since the patients are included
according to a strict protocol and are more closely In a register-based study of 634 272 Danish women
monitored in early pregnancy. Hence, more very achieving pregnancy between 1978 and 1992 and
early pregnancy losses are included in placebo-con- who attended a hospital during the pregnancy,15 the
trolled trials and the live birth rate in the placebo- miscarriage rates were almost identical in women of
arm is expected to be lower than in non-randomized age 30–34 years with RPL and those of age 35–39
studies. In accordance with this, Carp et al12 have (38–40%), but increased to 70% in women aged
shown that the live birth rate in untreated patients in 40–44 (Figure 1.3). It seems that the impact of age
randomized studies was 15–20% lower than that of on the miscarriage rate in RPL is quite modest until
non-randomized patients independent of the age 40, but beyond this age it is the strongest prog-
number of previous miscarriages. nostic factor. In concordance with this, several mul-
The prognostical negative effect of the number tivariate analyses7,13,14 of prognostic variables for
of previous miscarriages could, in theory, be attributed live births in RPL patients (almost all of whom were
3
0415421306-Ch01 3/29/07 5:25 PM Page 4
80
70
60 Maternal age
Birth rate (%)

50 < 30 years
40 31–35 years
30 36–39 years
40–44 years
20
10
0
Nybo Andersen et al15 Clifford et al9
Figure 1.3 Subsequent birth rate according to maternal age in patients with recurrent pregnancy loss.
younger than 40) found that maternal age was not a In some studies, secondary RPL is defined as RPL
significant predictor of miscarriage after adjust- after a live birth or a pregnancy that progressed
ment for other relevant independent variables, the beyond gestational week 28;12,16 however, in this
adjusted odds ratios (OR) for live birth being very survey, 20 weeks is taken as the cut-off point.
similar in two studies: 0.93 (95% confidence interval Unfortunately, few studies separate the patients
(CI) 0.8–1.1)7 and 0.94 (95% CI 0.9–1.0).13 into primary or secondary RPL, which may indicate
that the authors consider the two disorders as identi-
SUBGROUPS OF RPL cal entities. It is indeed possible that secondary RPL
is not a particular entity but just the clinical appear-
The pregnancy history in women with RPL may ance of the RPL syndrome among patients who, by
include pregnancies that have ended in live births. chance, instead of delivering a child after three or
Three different groups can be identified that should four miscarriages deliver an infant in the first preg-
be assessed separately: nancy and subsequently experience a series of mis-
carriages. However, if the chance of delivering a child
● The primary RPL group consists of women with is equal in all consecutive pregnancies of women des-
three or more consecutive pregnancy losses with tined to experience RPL, we would expect the preva-
no pregnancy progressing beyond 20 weeks lence of patients with secondary RPL and at least
gestation. four miscarriages who delivered a child in the first
● The secondary RPL group consists of women pregnancy to be equal to that of those who had deliv-
who have had three or more pregnancy losses ered a child in the second pregnancy. We studied a
following a pregnancy that progressed beyond 20 cohort of patients with secondary RPL and four mis-
weeks’ gestation (which may have ended in live carriages17,18 and found that four times as many had
birth, or less commonly a stillbirth or neonatal had a successful outcome in the first pregnancy com-
death). pared with the second pregnancy (Table 1.1). Since it
● The tertiary RPL group (which is a group that is unlikely that patients in the latter group have a
has not been well characterized or studied) lower likelihood of being admitted to our clinic than
consists of women who have had several patients in the former group, we conclude that the
pregnancy losses before a pregnancy that risk of suffering RPL is much higher after a birth in
progressed beyond 20 weeks’ gestation followed the first pregnancy than in the second pregnancy.
by at least three more pregnancy losses.12 This observation supports the theory that some risk
4
0415421306-Ch01 3/29/07 5:25 PM Page 5
Table 1.1 Distribution of patients with secondary recurrent pregnancy loss (RPL) and at at least four
miscarriages according to the order of the birth in the pregnancy sequence
Pregnancy outcomesa Type of RPL Observed prevalence Expected prevalence p-value
B, M, M, M, M,… Secondary RPL 82 (79.6%) 51.5b (50.0%) < 0.00005

M, B, M, M, M,… Secondary RPL 21 (20.4%) 51.5b (50.0%)
a
B, birth; M, miscarriage.
b
Expected distribution of patients with secondary RPL and at least four miscarriages.
factors for RPL have a high chance of developing patients had had at least one second-trimester loss.
during the first ongoing pregnancy and that second- Among 228 RPL patients admitted to our clinic in
ary RPL is not a chance phenomenon. 2000–2004, 39 (17.1%) had experienced a mixture
If primary and secondary RPL have different of first- and second-trimester miscarriages, but
pathophysiological backgrounds, we should expect only three had suffered exclusively second-trimester
different prognoses for the two conditions. losses. Since almost all patients with second-
Summarizing the placebo-treated patients included trimester miscarriages had experienced at least one
in our placebo-controlled trials of immunother- first-trimester miscarriage, early and late RPL must
apy,10,19 the live birth rate was 17/35 = 48.6% in the have pathogenetic factors that partially overlap –
first pregnancy in women with primary RPL, com- but the observation that the overwhelming majority
pared with 11/34 = 32.4% in women with second- of patients only suffer first-trimester miscarriages
ary RPL (not significantly different) matched for suggests that some pathogenetic factors are specific
number of previous miscarriages and age. Other for those with early miscarriages. Several prospec-
studies have reported success rates8,9 in the two sub- tive studies indicate that a history of one or more
sets that are not different, which must be considered second-trimester pregnancy losses is a strong risk
to be the commonly accepted view. factor for a poor prognosis,21,22 which also suggests
In the secondary RPL group, however, the sex of that some pathogenetic factors are specific for
the firstborn seems to affect the prognosis. In a patients with late losses.
prospective study of 182 patients with secondary
RPL followed up from 1986 to 2002,17 the cumula- FAMILIAL AGGREGATION
tive chance of giving birth to a child after a seriesof
miscarriages was 58% in patients with a male first- Few studies have investigated the genetics of RPL in
born, compared with 76% in those with a female families of RPL couples with normal chromosomes.
firstborn (p = 0.01). The hazard ratio for a live birth Results from the published family studies are shown
(the relative chance for live birth adjusted for the in Table 1.2. Johnson et al23 and Alexander et al24
follow-up period and a series of prognostic factors) compared the prevalence of three not necessarily
in patients with a male firstborn was 0.59 (95% consecutive miscarriages among blood relatives of
CI 0.41–0.86) compared with those with a female women with RPL with the corresponding preva-
firstborn. The previous birth of a boy is therefore a lence in relatives of fertile controls. Ho et al25 com-
prognostic negative factor in these patients, which pared the prevalence of RPL in relatives of couples
may be attributed to immunization against male- with RPL with that in relatives of fertile control
specific minor histocompatibility antigens (HY couples. The data concerning the relatives were
antigens) in the first ongoing pregnancy. those stated by the probands. Christiansen et al26
RPL patients with second-trimester losses also obtained information concerning relatives’ preg-
constitute a subset with particular characteristics. nancy outcomes from questionnaires completed by
Drakeley et al20 found that 25% of their RPL the relatives themselves, and the stated pregnancy
5
0415421306-Ch01 3/29/07 5:25 PM Page 6
Table 1.2 Studies of occurrence of recurrent pregnancy loss (RPL) in relatives of women with RPL
Reference and kind of relatives studied RPL rate in relatives (%) RPL rate in controls (%) p-value
Johnson et al23 12.2 7.3

Blood relatives
Alexander et al24 7.0 0.0 0.02
Mothers and sisters
Ho et al25 1.4 0.2 0.0001
First degree relatives
Christiansen et al26
Sisters 10.6 1.8 0.00005
Brothers’ wives 6.3 1.8 NS
NS, not significant.
loses were confirmed from hospitals’ and practi- study has really addressed the question of partner
tioners’ records. The prevalence of at least three, not specificity. Out of 228 RPL patients admitted to our
necessarily consecutive, pregnancy losses in rela- clinic in 2000–2004, 38 (16.7%) had experienced
tives with at least two pregnancies was compared miscarriages with two or more partners, and with
with an external control group.6 Table 1.2 shows regard to pregnancy prognosis they do not behave
that the risk of RPL in first-degree relatives of RPL differently from those who had all pregnancies with
patients is 2–7 times higher than in the background the same partner. In a multivariate analysis (Nielsen
population. The relative frequency λ (the frequency et al, unpublished work), we have found that, after
of RPL in relatives divided by the frequency in the adjustment for all relevant prognostic factors, the
general population) is a measure of the degree of chance of a subsequent live birth was not different
heritability of a disorder: the higher the value of λ, in patients with secondary RPL who have had all
the higher is the genetic component. For type 1 pregnancies with the same partner compared with
(‘insulin-dependent’) diabetes mellitus, λ for sisters those who have had two different partners. The
is 15, suggesting a high degree of heritability. In the observation that there is a clear familial predisposi-
Danish RPL study,26 λ can be calculated to be 5.9 for tion to RPL, at least in the female partner (Table 1.2),
sisters and 3.5 for brothers’s wives when compared also argues against RPL as being partner-specific.
with the prevalence in the general population,6 indi-
cating a moderate degree of heritability. If a condition CLINICAL ASSOCIATIONS
is determined by multifactorial or polygenic inheri-
tance, it is expected that λ equals the inverse square An association has been found between RPL and
root of the population prevalence, i.e., 1 / 0.018 = 7.5, various late obstetric complications. These are fully
which is very close to the observed λ in sisters, and described in Chapter 16.
this emphasizes the hypothesis that RPL is a disorder
following a polygenic mode of inheritance.27 LIFESTYLE FACTORS
PARTNER SPECIFICITY Lifestyle factors are rarely, if ever, major causes

of RPL; however, epidemiological studies have
It is commonly assumed that unexplained RPL is a indicated that a series of lifestyle factors can increase
partner-specific condition, and a criterion that all the risk of miscarriage. There is good evidence that
pregnancies should be with the same partner has obesity,29,30 high daily caffeine intake,31–34 alcohol
been included in the definition of primary and consumption,34 and use of non-steroidal anti-
secondary RPL by some authors.28 However, no inflammatory drugs35,36 increase the risk of
6
0415421306-Ch01 3/29/07 5:25 PM Page 7
miscarriage significantly. Social class and occupation study has not been undertaken. It is quite likely that
also increase the chance of miscarriage, with the great- by stratifying the sample by number of previous
est risk ocurring in women exposed to high physical or miscarriages, the effect of the experimental inter-
psychological stress during work.37,38 Several studies vention will become easier to demonstrate in those
now also indicate that previous subfertility or infertil- women with higher numbers of previous miscar-
ity treatment may increase the risk of miscarriage.14,39 riages than in those with fewer previous miscar-
riages, because the spontaneous success rate is so
much lower in the former group.12,40
INTEGRATION OF EPIDEMIOLOGICAL Unfortunately, in many studies, patients with only
FACTORS IN THE RESEARCH two previous miscarriages are included. The occur-
AND MANAGEMENT OF RPL rence of two miscarriages may in many cases be a
chance phenomenon caused by de novo fetal chro-
OCCURRENCE mosome abnormalities (in particular, autosomal tri-
somies) rather than a recurrent maternal factor.
Estimation of the prevalence of RPL has several Cytogenetic evaluations of specimens of sporadic
applications: it can be used for comparing risks of abortions have revealed an overall incidence of chro-
RPL between different populations or in subgroups mosomal abnormalities of 43%.41 Thus, in theory, in
within the same population, and it can be used for 0.43 × 0.43 = 18.5% of all women with two consecu-
comparing change in risk over time – which is necess- tive miscarriages, the cause is the occurrence of two
ary for identifying risk factors. Furthermore, the chromosomally abnormal conceptions. The inclusion
observation that the prevalence of RPL is greater of women with only two early miscarriages in a study
than 1% indicates that RPL is not a random event of RPL will ‘dilute’ the estimate of the risk factor (in
but rather a disorder affecting women who have an case–control and cohort studies) or the treatment
increased risk of pregnancy loss. In theory, a woman effect in controlled clinical trials. The proportion of
could have RPL, and each of the three consecutive RPL patients in whom the disorder can be explained
pregnancy losses may be caused by the same factors by a random accumulation of ‘sporadic’ miscarriages
as those causing ‘sporadic’ miscarriages, especially declines with the number of previous miscarriages.42
fetal chromosome abnormalities. However, if all Conversely, the proportion of cases that can be
RPL cases were caused by a random accumulation of explained by a factor increasing the risk of miscarriage
‘sporadic’ miscarriages, the prevalence of RPL should of euploid embryos may increase with the number of
be 0.143 = 0.27% (based on a prevalence of 14% for previous miscarriages. Hence, the frequency of many
single pregnancy loss6) rather than 1%. The preva- immunological risk factors43–45 and the possible effect
lence of 1% indicates that at least three out of four of immunotherapy,12,40 increase and the frequency
RPL cases are caused by non-random factors that of chromosomally abnormal abortions decreases,46
increase the risk of miscarriage in each pregnancy. with the number of previous pregnancy losses.
NUMBER OF MISCARRIAGES MATERNAL AGE
It is clear from the evidence showing that the Because increased maternal age increases the subse-
number of previous miscarriages is the most quent miscarriage rate, stratification for age should
important prognostic factor in RPL that this be taken into account in therapeutic trials.
parameter has to be taken into account when plan- However, in RPL, age seems to display a significant
ning therapeutic trials. The ideal trial should strat- impact on pregnancy outome only after age 4015
ify for the number of previous miscarriages, with (Figure 1.3). Consequently, it may be sufficient to
randomization between control and experimental stratify according to two age groups: below and
treatments within each stratum. To date, such a above 40. Advanced age is associated with several
7
0415421306-Ch01 3/29/07 5:25 PM Page 8
disorders such as uterine fibroids and endocrine towards a higher prevalence of thrombophilic fac-
and autoimmune abnormalities, so age is a con- tors, especially the factor V Leiden mutation in
founding factor that should be adjusted for when patients with second-trimester miscarriages com-
the real impact of these disorders on subsequent pared with those with only early losses.47,51
reproductive performance is assessed. It is unclear whether parental chromosome
abnormalities (e.g., balanced translocations) have a
SUBGROUPS OF RPL different prevalence in primary or secondary RPL.
In a review52 of 79 relevant studies on the preva-
If secondary and primary RPL and RPL with first- lence of parental chromosome abnormalities in
and second-trimester losses have different patho- couples with RPL, a higher frequency of aberrations
genetic backgrounds, the frequency of recognized (3.7%) was found in couples with multiple preg-
risk factors for RPL and the efficacy of treatments nancy losses and one or more live births compared
may differ between the groups. Indeed, a series with couples with exclusively repeated pregnancy
of studies have provided data suggesting that such losses, among whom it was 2.9%. Franssen et al53
differences exist (Table 1.3). found that the frequency of parental chromosome
The factor V Leiden mutation is the commonest abnormalities was not different between couples with
cause of activated protein C (APC) resistance, RPL including a live birth and RPL couples without a
which is a risk factor for thrombosis and is probably previous birth. The frequency of parental chromo-
also associated with RPL.47 Wramsby et al48 found some anomalies thus seems to be almost similar in
factor V Leiden to be significantly associated with primary and secondary RPL, and the consequence is
primary but not secondary RPL, and Rai et al49 that screening of parental chromosomes should be a
found that APC resistance was significantly associated part of the routine workup in both types of RPL.
with the absence of a previous live birth among A series of immunological parameters have been
patients with RPL. In a study of three congenital described as being important for RPL and would be
thrombophilic factors (including the factor V expected to show a different distribution between
Leiden mutation), 25.5% with primary RPL the subgroups of RPL patients if the pathogenetic
compared with 15.1% of those with secondary RPL backgrounds were different. Research in immuno-
were positive for at least one factor.50 The literature logical factors in RPL has concentrated on alloanti-
thus points towards a lower prevalence of factor V bodies, autoantibodies, natural killer (NK) cells,
Leiden/APC resistance in secondary rather than and HLA (human leukocyte antigen) molecules of
primary RPL patients. Most studies also point the major histocompatibility complex (MHC).
Table 1.3 The prevalence of risk factors or effect of treatments in patients with primary and secondary
recurrent pregnancy loss (RPL) and RPL with second-trimester losses (late RPL)
Prevalence/effect in secondary Prevalence/effect in late vs early

Risk factor or treatment vs primary RPL primary RPL
Parental chromosome abnormality Equal NA

Antipaternal antibodies Higher Higher
Antiphospholipid antibodies Lower or equal Higher
Thrombophilia factors Lower Higher
Natural killer cell activity Lower NA
HLA-DR3 Higher NA
Allogeneic lymphocyte immunization Lower NA
Treatment with intravenous immunoglobulin Higher NA
NA, cannot be estimated.
8
0415421306-Ch01 3/29/07 5:25 PM Page 9
There is much evidence that the maternal immune allele HLA-DR3 was found significantly more often
system recognizes and reacts to the trophoblast and in the total patient group than in controls (OR 1.4,
fetus in an ongoing pregnancy: alloantibodies p < 0.02). However, among the 250 patients with
directed against paternal/fetal HLA antigens are secondary RPL, the frequency of the HLA-DR3
produced with increased gestation54 due to traffic of phenotype was 32.4%, compared with 21.0% in
fetal cells into the mother’s circulation in the third controls (p < 0.006). In patients with primary RPL,
trimester and at delivery. Anti-HLA antibodies the frequency of the HLA-DR3 phenotype was
often persist for years and can therefore be found 21.8%, which was clearly similar to that of controls.
more often in women with secondary compared It is thus clear that HLA-DR3 is only associated with
with primary RPL,55 but seem not to exhibit any secondary RPL but not with primary RPL.
pathological action.19,56 The finding that increased NK cytotoxicity is
Most autoantibodies can be found with associated with primary RPL indicates that exces-
increased prevalence in patients with RPL, and their sive innate immunity may be associated with primary
presence is associated with a poor pregnancy prog- RPL. However, the association between particular
nosis;13 however, few studies of autoantibodies in HLA-DR alleles and secondary RPL, and the epidemi-
RPL have differentiated between primary and ological evidence that immunization against the
secondary RPL. In patients with primary RPL, the male-specific HY antigen plays a role in secondary
prevalence of positive anticardiolipin or antinuclear RPL,17 indicate that adaptive immunity may have a
antibody concentrations has been reported to be role in secondary RPL, since recognition of alloanti-
higher than in those with secondary RPL.13,57,58 gens by T lymphocytes and subsequent sensitization
None of the individual differences were statistically requires involvement of HLA-DR antigens.
significant, but the clear trend emphasizes the There is evidence that immunotherapy such as
importance that future studies of autoantibodies in allogeneic lymphocyte injections/infusions or intra-
RPL clearly distinguish between primary and venous immunoglobulin (IVIG) exhibit different
secondary RPL. There is, however, a consensus that effects in primary and secondary RPL, respectively
antiphospholipid antibodies (aPL) display a (Table 1.3).
stronger association with late miscarriages than
with early RPL51 – a fact that is integrated into the FAMILIAL AGGREGATION
definition of the antiphospholipid (APS) syndrome.59
APS is considered to be present in an aPL-positive As discussed above, family studies (Table 1.2) have
patient with a history of only one death of a normal shown that the RPL prevalence in siblings of RPL
fetus beyond week 10, whereas at least three miscar- probands is in accordance with a multifactorial
riages are needed for the diagnosis in an APL-positive model for the inheritance of RPL. In internal med-
patient with miscarriages before 10 weeks. icine and other disciplines, the development of
NK-cell cytotoxicity, an important factor in many common diseases (e.g., arterial hypertension,
innate immune defence, has been reported to be diabetes mellitus, and schizophrenia) is thought to
predictive for a poor prognosis in patients with be determined by a multifactorial threshold model.
RPL.60 Only one study has differentiated between One risk factor is not sufficient to cause disease, but
primary and secondary RPL.16 NK-cell activity in when several intrinsic and extrinsic factors come
peripheral blood has been reported to be signifi- together in the same individual (or couple), the risk
cantly increased in women with primary but not exceeds a threshold level and disease develops. In
secondary RPL when compared with controls.16 recent years, research has identified a series of pre-
Class II HLA alleles are associated with most disposing factors for RPL. So many risk factors have
immunological disorders. In the largest published now been identified that it is very common to find
case–control study of HLA-DR alleles in patients several in the same patient. Both thrombophilic61
with RPL,45 the immunological high-responder and immunogenetic62 risk factors seem to aggregate
9
0415421306-Ch01 3/29/07 5:25 PM Page 10
significantly more frequently than expected in RPL surveillance in late pregnancy (e.g., repeated ultra-
patients and the presence of several risk factors in sound examinations) to decrease perinatal mortality
the same couples affects the pregnancy prognosis and morbidity. A series of factors associated with
negatively,63,64 indicating that they may exhibit an RPL – aPL, hereditary thrombophilias, and mannose-
additive or multiplicative effect on the RPL risk. binding lectin deficiency – have also been associated
Traditionally, the causes of RPL have been divided with low birthweight.44,51 Since RPL per se seems to
into single sufficient factors as slices of a pie: uterine be associated with low birthweight, prospective studies
malformations 10%, endocrine factors 10%, APL of the effect of the mentioned factors on perinatal
15%, etc., which, together with the unexplained complications should be adjusted for the confounding
group, comprise 100%. This model is probably not effect of the number and type (midtrimester losses)
adequate due to the above arguments. We therefore of previous miscarriages.
encourage scientists and clinicians working in the
area of RPL to think in a threshold rather than a pie LIFESTYLE FACTORS
model.65 The clinical implication is that, in principle,
an RPL patient should be screened for all potential As mentioned above, a number of lifestyle factors,
risk factors and the investigation should not stop as including obesity, occupation, alcohol and caffeine
soon as the first risk factor has been identified. consumption, and subfertility, are important risk
The recognition that RPL exhibits a high degree of factors for miscarriage. RPL is a complex disorder
heritability paves the way for the identification where lifestyle factors are expected to modify the
of susceptibility genes for RPL through the per- effect of the non-lifestyle (intrinsic) factors discussed
formance of genetic linkage analyses in families previously. The prevalence of the most important
with several siblings experiencing miscarriage or lifestyle factors should be described in research
RPL. Such linkage analyses have already docu- publications for both patients and controls, in order
mented that genes in the HLA region are important to document that studies assessing non-lifestyle risk
pathogenetic factors in RPL.66 factors or pregnancy outcome are matched for
lifestyle factors. Since it is likely that smoking aggra-
PARTNER SPECIFICITY vates the effect of thrombophilic risk factors on
pregnancy loss, details about smoking habits should
Early studies on HLA in RPL were based on the be reported in all studies of RPL and thrombo-
hypothesis that increased HLA similarity between philia. Another example of the importance of
partners would lead to inadequate maternal protective adjusting for lifestyle factors is that of the relationship
immune responses and fetal loss. Although a con- between polycystic ovary syndrome (PCOS) and
siderable number of studies on HLA sharing in RPL. It is generally recognized that women with
couples with RPL have been performed, the evidence PCOS exhibit an increased rate of miscarriage and
did not support this hypothesis.67,68 If good-quality RPL. However, when adjustment for obesity is
epidemiological studies showing little evidence of undertaken in multivariate analyses, the miscar-
partner specificity in RPL had been performed riage rate in PCOS is not dependent on polycystic
(Nielsen et al, unpublished work) prior to the HLA- ovarian pathology or PCOS-associated endocrine
sharing studies, the theories of increased HLA-sharing abnormalities.30
between RPL spouses might not have developed.
CLINICAL ASSOCIATIONS CONCLUSIONS
Women with a history of RPL exhibit a significantly Throughout medical science, epidemiological studies
increased risk of late pregnancy complications. can provide indispensable knowledge when basic
Hence, all RPL patients should be offered increased laboratory research, case–control studies, or controlled
10
0415421306-Ch01 3/29/07 5:25 PM Page 11
treatment trials are planned and carried out. This is in late pregnancy: reduced birthweight caused by
also true in RPL. However, it seems that epidemio- increased risk for preterm birth and intrauterine
logical knowledge has only been integrated to a very growth retardation. It remains to be clarified from
limited degree into the current clinical research and multivariate analyses which clinical and paraclinical
management of RPL. factors among RPL patients determine the risks in
The population prevalence of RPL is much late pregnancy.
higher than would be expected if RPL were merely Lifestyle factors are unfortunately rarely men-
a random accumulation of ‘sporadic’ miscarriages, tioned or are only reported very superficially in
and this indicates that most RPL cases are caused by clinical studies in RPL. Since lifestyle factors can
factors increasing an individual couple’s miscar- cause miscarriage by themselves, or through inter-
riage risk. action with intrinsic factors, they should be
Estimates of the future miscarriage risk in RPL reported in more detail in future studies, and
patients vary significantly between studies, mainly appropriate stratification should be performed.
due to different methods of ascertainment and In conclusion, the multifactorial and polygenic
monitoring. Some studies have estimated the progno- background of RPL should be included in the models
sis too optimistically, as preclinical pregnancy losses used for the pathogenesis of RPL. Additionally, the
have been classified as non-pregnancies. To over- recognition of the different natures and pathogenic
come this potential source of error, in future treat- background of primary and secondary RPL, and the
ment trials the take-home baby rate per time unit different nature of RPL with few miscarriages as
may be a better outcome measure than the miscar- opposed to RPL with many miscarriages, should help
riage rate per registrated pregnancy eliminate the practice of combining data from too
Not only is the number of previous miscarriages heterogeneous studies for meta-analysis.
the strongest prognostic factor, but with an
increased number of previous miscarriages, fetal REFERENCES
aneuploidy seems to play a decreasing role and
maternal factors an increasing pathogenic role. 1. World Health Organization. Recommended definitions; terminology
and format for statistical tables related to the perinatal period. Acta
Therefore, stratification by the number of previous Obstet Gynecol Scand 1977; 56:247–53.
miscarriages is important both in association studies 2. Farquharson R, Jauniaux E, Exalto N, on behalf of the ESHRE Special
and in treatment trials. Primary and seconday RPL, Interest Group for Early Pregnancy (SIGEP). Updated and revised
nomenclature for the description of early pregnancy events. Hum
from an epidemiological point of view, also seem to Reprod 2005; 20:3008–11.
be distinct entities, and in many case–control and 3. Alberman E. The epidemiology of repeated abortion. In: Beard RW,
treatment studies these two subgroups have indeed Sharp F, eds. Early Pregnancy Loss: Mechanisms and Treatment.
London: Springer-Verlag; 1988:9–17.
been found to behave quite differently (Table 1.3). 4. Warburton D, Strobino B. Recurrent spontaneous abortion. In:
There are almost no studies addressing the issue Bennet MJ, Edmonds DK, eds. Spontaneous and Recurrent Abortion.
of partner specificity in RPL, and such studies are Oxford: Blackwell Scientific, 1987:193.
5. Stray-Pedersen B, Lorentzen-Styr AM. The prevalence of Toxoplasma
urgently needed. However, a series of family studies antibodies among 11,736 pregnant women in Norway. Scand J Infect
have unanimously found a familial aggregation of Dis 1979; 11:159–65.
miscarriage and RPL among first-degree relatives of 6. Fertility and Employment 1979. The Danish Data Archives No. 0363,
Odense University.
RPL patients – especially sisters – which indicates a 7. Cowchock FS, Smith JB. Predictors for live birth after unexplained
significant degree of heritability. The pattern of inher- spontaneous abortions: correlations between immunological test
itance is multifactorial, and this is also in accordance results, obstetric histories, and outcome of the next pregnancy with-
out treatment. Am J Obstet Gynecol 1992; 167:1208–12.
with clinical evidence: several risk factors for RPL can 8. Quenby SM, Farquharson RG. Predicting recurring miscarriage:
very often be found in a single patient, and an aggrega- What is important? Obstet Gynecol 1993; 82:132–8.
tion of risk factors aggravates the pregnancy prognosis. 9. Clifford K, Rai R, Regan L. Future pregnancy outcome in unexplained
recurrent first trimester miscarriage. Hum Reprod 1997; 12:387–9.
A series of studies have unanimously reported 10. Christiansen OB, Pedersen B, Rosgaard A, et al. A randomized,
that RPL is associated with a series of complications double-blind, placebo-controlled trial of intravenous immunoglobulin
11
0415421306-Ch01 3/29/07 5:25 PM Page 12
in the prevention of recurrent miscarriage: evidence for a therapeutic 31. Infante-Rivard C, Fernandez A, Gauthier R et al. Fetal loss associated
effect in women with secondary recurrent miscarriage. Hum Reprod with caffeine intake before and during pregnancy. JAMA 1993;
2002; 17:809–16. 270:2940–3.
11. Jablonowska B, Selbing A, Palfi M, et al. Prevention of recurrent 32. Fenster L, Hubbard AE, Swan SH, et al. Caffeinated beverages, decaf-
spontaneous abortion by intravenous immunoglobulin: a double- feinated coffee, and spontaneous abortion. Epidemiology 1997;
blind placebo-controlled study. Hum Reprod 1999; 14:838–41. 8:515–23.
12. Carp HJ, Toder V, Torchinsky A, et al. Allogenic leukocyte immuniza- 33. Giannelli M, Doyle P, Roman E, et al. The effect of caffeine consump-
tion after five or more miscarriages. Recurrent Miscarriage tion and nausea on the risk of miscarriage. Paediatr Perinat Epidemiol
Immunotherapy Trialists Group. Hum Reprod 1997; 12:250–5. 2003; 17:316–23.
13. Nielsen HS, Christiansen OB. Prognostic impact of anticardiolipin 34. Rasch V. Cigarette, alcohol, and caffeine consumption: risk factors for
antibodies in women with recurrent miscarriages negative for the spontaneous abortion. Acta Obstet Gynecol Scand 2003; 82:182–8.
lupus anticoagulant. Hum Reprod 2005; 20:1720–8. 35. Nielsen GL, Sorensen HT, Larsen H, et al. Risk of adverse outcome
14. Cauchi MN, Coulam CB, Cowchock S, et al. Predictive factors in and miscarriage in pregnant users of non-steroidal anti-inflamma-
recurrent spontameous abortion – a multicenter study. Am J Reprod tory drugs: population based observational study and case-control
Immunol 1995; 33:165–70. study. BMJ 2001; 322:266–70.
15. Nybo Andersen AM, Wohlfahrt J, Christens P, et al. Maternal age and 36. Li DK, Liu L, Odouli R. Exposure to nonsteroidal anti-inflammarory
fetal loss: population based register study. BMJ 2000; 320:1708–12. drugs during pregnancy and risk of miscarriage: population based
16. Shakhar K, Ben-Eliyahu S, Loewenthal R, et al. Differences in number cohort study. BMJ 2003; 327:368–72.
and activity of peripheral natural killer cells in primary versus 37. Brandt LP, Nielsen CV. Job stress and adverse outcome of pregnancy:
secondary recurrent miscarriage. Fertil Steril 2003; 80:368–75. a causal link or recall bias? Am J Epidemiol 1992; 35:302–11.
17. Christiansen OB, Pedersen B, Nielsen HS, et al. Impact of the sex of 38. Florack EI, Zielhuis GA, Pellegrino JE, et al. Occupational physical
first child on the prognosis in secondary recurrent miscarriage. activity and the occurence of spontaneous abortion. Int J Epidemiol
Hum Reprod 2004; 19:2946–51. 1993; 22:878–84.
18. Christiansen OB, Kolte AM, Nielsen HS. Secondary recurrent miscar- 39. Wang JX, Norman RJ, Wilcox AJ. Incidence of spontaneous abortion
riage – unique entity with repect to etiology and treatment. among pregnancies produced by assisted reproductive technology.
Curr Women’s Health Rev 2006; 2:119–24. Hum Reprod 2004; 19:272–7.
19. Christiansen OB, Mathiesen O, Husth M, et al. Placebo-controlled 40. Daya S, Gunby J, and The Recurrent Miscarriage Trialists Group. The
trial of active immunization with third party leukocytes in recurrent effectiveness of allogeneic leukocyte immunization in unexplained
miscarriage. Acta Obstet Gynecol Scand 1994; 73:261–8. primary recurrent abortion. Am J Reprod Immunol 1994; 32:
20. Drakeley AJ, Quenby S, Farquharson RG. Mid-trimester loss; 294–302.
appraisal of a screening protocol. Hum Reprod 1998; 13:1471–9. 41. Creasy R. The cytogenetics of spontaneous abortion in humans. In:
21. Cowchock FS, Smith JB, David S, et al. Paternal mononuclear cell Beard RW, Sharp F, eds. Early Pregnancy Loss: Mechanisms and
immunization therapy for repeated miscarriage: predictive variables Treatment. London: Springer-Verlag, 1988:293–304.
for pregnancy success. Am J Reprod Immunol 1990; 22:12–17. 42. Christiansen OB. A fresh look at the causes and treatment of recur-
22. Goldenberg RL, Mayberry SK, Copper RL, et al. Pregnancy outcome rent miscarriage, especially its immunological aspects. Hum Reprod
following a second-trimester loss. Obstet Gynecol 1993; 81:444–6. Update 1996; 2:271–93.
23. Johnson PM, Chia KV, Risk JM, et al. Immunological and immuno- 43. Pfeiffer KA, Fimmers R, Engels G, et al. The HLA-G genotype is
genetic investigation of recurrent spontaneous abortion. Dis Mark potentially assoiated with idiopathic recurrent spontaneous abortion.
1988; 6:163–171. Mol Hum Reprod 2001; 7:373–8.
24. Alexander SA, Latinne D, Debruyere M, et al. Belgian experience with 44. Kruse C, Rosgaard A, Steffensen R, et al. Low serum level of mannan-bind-
repeat immunization in recurrent spontaneous abortion. In: Beard ing lectin is a determinant for pregnancy outcome in women with recur-
RW, Sharp F, eds. Early Pregnancy Loss: Mechanisms and Treatment. rent spontaneous abortion. Am J Obstet Gynecol 2002; 187:1313–20.
London: Springer-Verlag, 1988:355–63. 45. Kruse C, Steffensen R, Varming K, et al. A study of HLA-DR and -DQ
25. Ho H, Gill TJ, Hsieh C, et al. The prevalence of recurrent spontaneous alleles in 588 patients and 562 controls confirms that HLA-DRB1*03
abortion, cancer, and congenital anomalies in the families of couples is associated with recurrent miscarriage. Hum Reprod 2004; 19:
with recurrent spontaneous abortions or gestational trophoblastic 1215–21.
tumors. Am J Obstet Gynecol 1991; 165:461–6. 46. Ogasawara M, Aoki K, Okada S, et al. Embryonic karyotype of abor-
26. Christiansen OB, Mathiesen O, Lauritsen JG, et al. Idiopathic recur- tuses in relation to the number of previous miscarriages. Fertil Steril
rent spontaneous abortion. Evidence of a familial predisposition. 2000; 73:300.
Acta Obstet Gynecol Scand 1990; 69:597–601. 47. Rey E, Kahn SR, David M, Shrier I. Thrombophilic disorders and fetal
27. Emery AEH. Methodology in Medical Genetics. 2nd rev edn. loss: a meta-analysis. Lancet 2003; 361:901–8.
Edinburgh; Churchill Livingstone, 1986. 48. Wramsby ML, Sten-Linder M, Bremme K. Primary habitual abortions
28. Stephenson MD. Frequency of factors associated with habitual are associated with high frequency of factor V Leiden mutation. Fertil
abortion in 197 couples. Fertil Steril 1996; 66:124–9. Steril 2000; 74:987–91.
29. Fedorcsak P, Storeng R, Dale PO, et al. Obesity is a risk factor for early 49. Rai R, Shlebak A, Cohen H, et al. Factor V Leiden and acquired acti-
pregnancy loss after IVF or ICSI. Acta Obstet Gynecol Scand 2000; vated protein C resistance among 1000 women with recurrent miscar-
79:43–8. riage. Hum Reprod 2001; 16:961–5.
30. Wang JX, Davies MJ, Norman RJ. Polycystic ovarian syndrome and 50. Carp H, Salomon O, Seidman D, et al. Prevalence of genetic markers
the risk of spontaneous abortion following assisted reproductive tech- for thrombophilia in recurrent pregnancy loss. Hum Reprod 2002; 17:
nology treatment. Hum Reprod 2001; 16:2606–9. 1633–7.
12
0415421306-Ch01 3/29/07 5:25 PM Page 13
51. Roque H, Paidas MJ, Funai EF, et al. Maternal thrombophilias are not 61. Coulam CB, Jeyendran RS, Fishel LA, Roussev R. Multiple thro-
associated with early pregnancy loss. Thromb Haemost 2004; 91: 290–5. mobophilic gene mutations rather than specific gene mutations are
52. Tharapel AT, Tharapel SA, Bannerman RM. Recurrent pregnancy risk factors for recurrent miscarriage. Am J Reprod Immunol 2006;
losses and chromosome abnormalities: a review. Br J Obstet Gynaecol 55:360–8.
1985; 92:899–914. 62. Hviid TV, Christiansen OB. Linkage disequilibrium between human
53. Franssen MTM, Korevaar JC, Leschot NJ, et al. Selective chromosome leukocyte antigen (HLA) class II and HLA-G – possible implications
analysis in couples with two or more miscarriages: case-control study. for human reproduction and autoimmune disease. Hum Immunol
BMJ 2005; 331:137–41. 2005; 66:688–99.
54. Regan L. A prospective study of spontaneous abortion. In: Beard RW, 63. Jivraj S, Rai R, Underwood J, et al. Genetic thrombophilic mutations
Sharp F, eds. Early Pregnancy Loss. Mechanisms and Treatment. among couples with recurrent miscarriage. Hum Reprod 2006; 21:
London: Springer-Verlag, 1988:23–37. 1161–5.
55. Coulam CB. Immunological tests in the evaluation of reproductive 64. Christiansen OB, Kruse C, Steffenson R, Varming K. HLA Class II,
disorders: a critical review. Am J Obstet Gynecol 1992; 167:1844–51. mannan-binding lectin (MBL) and recurrent miscarriage. Presented at
56. Sargent IL, Wilkins T, Redman CWG. Maternal immune responses to the European Congress of Reproductive Immunology, Plzen, 2004,
the fetus in early pregnancy and recurrent miscarriage. Lancet 1988; Abstract I30 in special issue of Am J Reprod Immunol.
ii:1099–104. 65. Christiansen OB, Nybo-Andersen AM, Bosch E, et al. Evidence-based
57. Cowchock S, Bruce Smith J, Gocial B. Antibodies to phospholipids investigations and treatments of recurrent pregnancy loss. Fertil Steril
and nuclear antigens in patients with repeated abortions. Am J Obstet 2005; 83:821–39.
Gynecol 1986; 155:1002–10. 66. Christiansen OB, Andersen HH, Hojbjerre M, et al. Maternal HLA
58. Rai R, Regan L, Clifford K, et al. Antiphospholipid antibodies and Class II allogenotypes are markers for the predisposition to fetal losses
β2-glycoprotein-I in 500 women with recurrent miscarriage: results of in families of women with unexplained recurrent fetal loss. Eur J
a comprehensive screening approach. Hum Reprod 1995:10:2001–5. Immunogenetics 1995; 22:323–34.
59. Wilson WA, Gharavi AE, Koike T, et al. International concensus 67. Christiansen OB, Riisom K, Lauritsen JG, et al. No increased histo-
statement on preliminary classification criteria for definite antiphos- compatibility antigen sharing in couples with idiopathic habitual
pholipid syndrome: report of an international workshop. Arthritis abortions. Hum Reprod 1989; 4:160–2.
Rheum 1999; 42:309–11. 68. Ober C, van der Ven K. HLA and fertility. In: Hunt JB, ed. HLA and
60. Aoki K, Kajiura S, Matsumoto Y, et al. Preconceptional natural-killer the Maternal–Fetal Relationship. Austin, TX:. RG Landers,
activity as a predictor of miscarriage. Lancet 1995; 345:1340–2. 1996:133–56.
13
0415421306-Ch01 3/29/07 5:25 PM Page 14
0415421306-Ch02 3/29/07 5:26 PM Page 15
2. Signaling between embryo and mother

in early pregnancy: Basis for development
of tolerance
Eytan R Barnea
media or that produced by deciduomata (a non-

THE HYPOTHESIS pregnant environment). This strongly suggested
communication between embryo and mother
In 1978, Beer and Billingham,1 while working on before implantation, but specific factors were never
immunological recognition mechanisms in mam- identified.
malian pregnancy, published their view that the Later studies have recognized that there are mul-
maternal system is aware of the presence of the early tiple types of placenta in mammals. The hemocho-
embryo, and actively responds to it. This was sur- rial placenta (found in the human and the mouse)
prising, considering the differences in genetic is associated with intimate interaction, while in
make-up of the mother and fetus (semi or total), other species there is less invasiveness (an example
and was contrary to the prevailing opinion at that is the pig placenta, which communicates with the
time, which considered that the trophoblast was endometrium through the histiotroph). In addition,
hypoantigenic, as a protection from cellular immu- the secretory products of different types of placentas
nity. Beer and Billingham1 suggested that unique also differ: human chorionic gonadotropin (hCG)
HLA (human leukocyte antigen) molecules are pre- in humans, prolactin in rodents,5 etc.
sented to the maternal system, the responses to Despite such diversity at implantation, there are
which play a role in establishing and maintaining features that are common to the development of all
pregnancy. A decade later, Billingham and Head2 mammals before implantation: egg and sperm
suggested that local cell-based immunosuppressive fusion, and progressive development of the fertil-
and immunoprotective activity in the placenta was ized embryo up to the blastocyst stage. In a recent
mediated by suppressor and other unknown cells. review, Moffet and Loke6 concluded that pregnancy
Billingham and Head2 further suggested that HLA is not a classical acceptance/rejection phenomenon,
sharing in the parents leads to lack of maternal and the specific compounds derived from the con-
recognition and is therefore the basis for rejection, ceptus and the receptors present on immune cells
i.e., miscarriage. need to be identified to better understand the
Hansel and Hickey3 examined various com- unique interaction in pregnancy.
pounds that might be involved in the maternal The present review provides the rationale for
recognition of pregnancy, with an emphasis on early pregnancy recognition, with an emphasis on
domestic animals. They found several proteins, compounds that appear to be present prior to
including embryo-derived platelet-activating factor implantation. It discusses previous and recent data
(PAF), a trophoblastic protein with an antiluteolytic strongly suggesting that preimplantation factor
effect. Further progress regarding embryo–maternal (PIF) is a unique, universal compound that initiates
recognition was provided by Weitlauf,4 who pregnancy recognition (and tolerance) in mam-
reported that embryo-conditioned media has a spe- malian pregnancy. This recognition starts prior to
cific effect on the rat uterus compared with control direct embryo–maternal contact in the uterus.
15
0415421306-Ch02 3/29/07 5:26 PM Page 16
occurring. Once the sperm penetrates the egg at

RESCUE OF THE CORPUS LUTEUM
fertilization, it becomes ‘invisible’ to the maternal
immune system. As expected, following egg/sperm
Following ovulation, the corpus luteum (CL) is
fusion, there is no maternally induced immune
formed and secretes progesterone, which has a
rejection, for as long as the egg membrane does not
trophic effect on the endometrium. Studies have
change its characteristics (expressing foreign
shown that a variety of signals can rescue the CL.
antigens). Once foreign antigens are expressed, the
These include hCG in humans, prolactin in rodents,
fertilized egg rapidly becomes surrounded by the
and estrogen in pigs, indicating that the CL-rescuing
zona pellucida, a hard and impenetrable shell that
signals are species-specific. When cow and mare
wards off maternal immune cells. Further immune
uteruses are removed, prostaglandin F2α (PGF2α) is
protection is provided by maternal cumulus oopho-
not released and the CL persists long term; there-
rus cells, which further prevent direct access of
fore, the presence of the conceptus actually prevents
maternal immune cells to the embryo. However, the
luteolysis.7 But the presence of an embryo is not
cumulus cells persist only for a few days after fertil-
necessary, and hCG injections, for example, can
ization, as their primary role is to facilitate tubal
prolong the lifespan of the CL to a certain degree.
transport of the embryo towards the uterus. The
This contrasts with the uterus, in which a viable
cumulus has immune cells that secrete cytokines,
embryo must be present in order for the endo-
and may serve as a first relay system for propagating
metrium to become receptive. Thus, recognition of
embryo-derived signaling.8 Indeed, it has been
pregnancy and successful implantation take place
shown that within 8 hours after fertilization, there is
before the stage when rescue of the CL occurs,
emargination of platelets from the peripheral blood
strongly suggesting that there is no linkage between
in mice.9
tolerance and the CL.
Embryonic cell proliferation up to the 8-cell
stage is rather orderly. The blastomeres are totipo-
DOES THE EMBRYO–MATERNAL DIALOGUE tential (i.e., each of them could develop into a com-
START PRIOR TO IMPLANTATION? plete embryo). This process lasts approximately
3 days while the embryo travels within the fallopian
As we have seen, the CL initially does not need the tube. The speed of development is a good index to
conceptus, and can persist for a few weeks before it evaluate embryonic health with respect to likeli-
undergoes spontaneous regression, at menses. hood for implantation.
Therefore, the embryo–maternal dialogue required Evidence that the embryo may have an active role
for implantation likely does not involve the CL. In in immune recognition was suggested by studies
humans, for example, implantation takes place showing that embryo-conditioned medium has
1 week before the CL would undergo regression. If immune-suppressive properties.10,11 However, the
the CL were involved in embryo recognition, it compounds responsible for this immune effect have
could take place only when there was intimate not been fully characterized. Further data suggested
embryo–maternal contact. But this is clearly not the that a variety of compounds can be identified in
case, and the search for the elements involved in the the maternal circulation prior to implantation,
early interaction has been long going. First, here, we compared with non-pregnant subjects. However,
give a brief rationale of the need for maternal recog- whether the putative embryo-specific secreted prod-
nition of the embryo shortly after fertilization. ucts and the early-stage circulatory compounds are
the same remains unclear. If the embryo-secreted
THE FERTILIZATION PROCESS products and circulatory compounds are identical,
very low concentrations of embryo-secreted com-
The released mature egg reaches the ampular region pounds could reach the maternal circulation and
and survives for only 12–24 hours unless it is fertilized. cause changes in maternal immunity to initiate tol-
There is a one-in-three chance of fertilization erance. Obviously, this would mean that the embryo
16
0415421306-Ch02 3/29/07 5:26 PM Page 17
SIGNALING BETWEEN EMBRYO AND MOTHER IN EARLY PREGNANCY
plays a role in developing tolerance even prior to lead to successful progeny, unique embryo-derived
implantation. This signaling would also explain signals must be present, due to the absence of a
pathological pregnancies in which implantation host-versus-graft or graft-versus-host reaction.
occurs in sites outside the uterus, including the fal- Moreover, the maternal system must accommodate
lopian tube, ovary, or even (rarely) in the abdominal and nurture the conceptus until delivery, and any
cavity on the bowel. Ectopic pregnancies strongly immune tolerance is therefore conditional, because
suggest that maternal recognition of pregnancy rejection may take place at any moment until
must be systemic – not localized to the uterus. delivery. In addition, such a unique phenomenon
Moreover, experience with transfer of donor would have to be pregnancy-specific; for tolerance
(genetically dissimilar) embryos has shown high to be successful, the embryo must be viable and the
implantation and pregnancy success rates, further maternal system receptive. The signal must be pres-
implicating the role of the embryo in the recogni- ent early in embryo development, must be potent,
tion process. There is a 4- to 5-day delay between and must have specific sites of action both on the
fertilization and implantation, which is replicated maternal immune system and on the endometrium.
in embryo transfer following in vitro fertilization The signal must also be universally mammalian,
(IVF). The delay suggests that this time is required because the same early phenomenon takes place in
to establish tolerance and prime the endometrium, all mammals (and any diversity only occurs at the
making it both receptive and accommodating for implantation phase).
the incoming embryo. What properties would such a signal have? It
would modulate the maternal immune system with-
out suppressing it. This is essential, because during
GENOMIC ELEMENTS IN RECOGNITION pregnancy the mother is exposed to pathogens and
her ability to maintain an effective immune system
Recent data show that the embryo expresses its to combat disease is essential for survival – both for
genome as early as the 2-cell stage. Thus, in the earli- her and for the embryo. Therefore, the signal would
est stages of development, the embryo becomes a have to allow maternal immunity to function
partial or total ‘non-self’ from the perspective of the unimpeded, allowing it to fight bacteria, viruses,
mother. Thus, development of the zona pellucida as and parasites, while maintaining the tolerance
a protection against maternal adversity becomes nec- toward the embryo. The tolerance that this signal
essary. It has recently been been observed that there creates must not be excessive; otherwise the
is a major downregulation of genes in the preimplan- mother’s ability to reject a defective embryo or seri-
tation embryo compared with the unfertilized egg.12 ously infected fetus would be inhibited. Of course,
This downregulation may protect the embryo by most defective embryos are rejected early, and, in
minimizing its vulnerability, and in a mostly anero- case of infection, premature labor frequently
bic environment it may be advantageous to shut ensues. An additional role of this signal would be
down non-essential functions that are not necessary to prime the endometrium, and make the uterine
for survival. Additionally, the few genes that are environment hospitable to the embryo. Finally, it is
upregulated may have an important physiological clear that as the embryo–maternal interaction
role. Novel genes that are expressed very early may becomes intimate, the dynamics change, and there
lead to early maternal recognition of pregnancy.13 are complex events that take place that could be
labeled as a maintenance of tolerance rather than
the initiation that is the topic of this chapter.
UNIQUE PHENOMENA REQUIRE The following is a discussion of the compounds
UNIQUE SIGNALS implicated so far in early pregnancy events prior
to implantation. Unfortunately, most knowledge to
In order for a semi- or totally foreign embryo date about the embryo–maternal immune interaction
(or even a cross-species transfer) to implant and involves study of uterine milieu during implantation.
17
0415421306-Ch02 3/29/07 5:26 PM Page 18
Briefly, there is a tolerant (Th2) cytokine balance in can also fail to implant due to deficient expression
pregnancy: increased interleukin-4 (IL-4), IL-5, and of adhesion molecules (MMPs) as well as the lack of
IL-10 and reduced Th1-type cytokines, such as IL-2, secretory and cellular elements that aid in the
interferon-γ (IFN-γ) and tumor necrosis factor α immune maternal recognition of pregnancy.21 In
(TNF-α).14 However, excess Th1 cytokines are asso- addition, some embryos may only partially or tem-
ciated with reproductive failure.15 The preimplanta- porarily implant, later dislodging into the fallopian
tion embryo may protect itself from maternal tube, leading to chemical or ectopic pregnancy.
immune rejection by promoting a Th2 phenotype.3 Recent data show an imbalance toward stimulatory
Activated natural killer (NK) cells cause a Th1-type overinhibitory NK-cell receptors: CD158a and
response, while increased peripheral T lymphocytes CD158b inhibitory receptor expression by CD56dim
express progesterone receptors, and protect by CD16+ and CD56brightCD16− NK cells was
releasing IL-10 and transforming growth factor β decreased, while CD161-activating receptor expres-
(TGF-β).16 NK cells may also inhibit excessive tro- sion by CD56+CD3+ NKT cells was increased, in
phoblast invasiveness by recognizing unusual fetal patients with implantation failures.22
trophoblast major histocompatibility complex
(MHC) ligands.15 Other non-pregnancy-specific
compounds may also be involved: sex steroids, WHICH CURRENTLY KNOWN COMPOUND
integrins and IL-1b have no mRNA for receptors in COULD BE THE UNIVERSAL TOLERANCE
the embryo. While insulin- like growth factors BIOMARKER?
(IGFs) have receptors, the ligands expressed in the
early embryo have trophic effects on the embryo. The main diagnostic marker for human pregnancy
They are modulated by embryonic IGF-binding is hCG, but it does not reflect pregnancy viability, it
protein 3 (IGFBP-3). Leukemia inhibitory factor cannot be detected early in embryo culture media,
(LIF) and colony-stimulating factors that stimulate and its persistence in the circulation after preg-
matrix metalloproteinases (MMPs) are also nancy has terminated greatly limits its clinical use.
involved, and inhibition of mucin 1 (MUC-1) hCG has an important role in the maintenance of
expression on the endometrial surface facilitates the corpus luteum following implantation, and it
implantation.17,18 The presence of regulatory T cells has been shown to be involved in altering the bio-
(Treg, CD4+CD25+) increases prior to implantation, chemical behavior and morphology of endometrial
suggesting early embryo signaling.19 This cannot be cell types, by acting on a specific binding site
due to semen-induced factors, since implantation (CG/LH-R). A local immunological role has also
after embryo transfer following IVF without con- been ascribed to hCG.23 However, hCG is not preg-
tact with semen is also associated with upregulated nancy-specific, is unique to humans, and, signifi-
T cells. cantly, is also found in various cancers. It appears
However, none of these compounds are that most of the effect of hCG in supporting preg-
pregnancy-specific, and therefore cannot be the ini- nancy is at implantation and beyond.
tiating signal for tolerance. In contrast, failure to
implant is frequent and may be caused by any dis- PLATELET-ACTIVATING FACTOR
ruption of the delicate balance between the uterine
epithelial lining, which becomes the decidua, and PAF is an acetylated phosphoglyceride expressed by
the embryo. The endometrium can be hostile due to the embryo in both humans and rodents. Its role is
immune disruptors, such as high peripheral levels mostly local within the fallopian tube, aiding in the
of NK cells, altered hormonal priming, infection, transfer of the embryo into the uterus.24 However,
and deficient integrin expression. The role of in other species, other compounds play this role; for
antiphospholipid antibodies, for example, in failed example, in horses, prostaglandin E is secreted by the
implantation is still being debated.20 The embryo morula. PAF also has a trophic effect on the embryo.25
18
0415421306-Ch02 3/29/07 5:26 PM Page 19
PAF is not pregnancy-specific, and is present in trophoblast, but are cleaved from membrane-
platelets, leukocytes, and endothelial cells. Therefore, bound HLA-G1.35 Thus, HLA-G may be necessary,
it is clear that PAF could not be a unique signal but is certainly not sufficient, for initiating maternal
required for pregnancy tolerance. tolerance of pregnancy.
EARLY PREGNANCY FACTOR PREIMPLANTATION FACTOR
Early pregnancy factor (EPF) has been identified as Over the past several years, our team’s studies have
chaperonin 10, a 12 kDa protein. It can be detected focused in identifying and documenting the role of
prior to implantation in the maternal circulation.26 PIF in mammalian pregnancy. PIF is only found in
EPF has been shown to influence immune effects pregnancy, is similar in all mammals, and is found
mediating the suppressive effect by binding T cells, very early, shortly after fertilization – all of which
NK cells, and monocytes. The receptor for EPF is suggest that PIF is the factor initiating maternal
not a functional homologue of chaperonin 10.27 recognition of pregnancy.
EPF activity in the serum is determined by Earlier work had shown that viable human and
decreased rosette formation using a cumbersome rabbit human embryo culture media contain
bioassay. Similar activity in mare and cow serum is unidentified immune modulatory compounds.10,11
related to a 26 kDa protein that is different from the We developed a novel bioassay and reported that
chaperonin molecule.28 In addition, EPF is not viable human and mouse embryo-conditioned cul-
pregnancy-specific; it is also present is several non- ture media, and human and porcine pregnancy
pregnant tissues, including in the serum of patients serum, contain immune-modulatory compounds
with ovarian cancer.29 that increase rosette formation between donor lym-
phocytes and platelets in the presence of CD2MAb
HLA-G due to PIF, a low-molecular-weight peptide(s).36–42
A bioassay, unlike an immune assay, is a reflection
The embryo and trophoblast express non-classical of a biological phenomenon, which led us to study
forms of HLA-G, which may protect them against whether the compounds present in embryo culture
NK-mediated lysis, and lead to apoptosis of allo- media are also present in the maternal circulation.
geneic cytotoxic CD8+ T cells by Fas ligands.30 But The presence of PIF activity in maternal sera
HLA-G-negative embryos may implant, and there- 4 days after embryo transfer was followed in
fore HLA-G is not essential for implantation.31 27/38 (71%) live births, while only 3 pregnancies
Recent data have shown that NK cells, which are occurred from 114 embryo transfers (3%) with
dominant in the decidua, express a receptor for non-detectable PIF activity, due to delayed implan-
KIR2DL4, which interacts with HLA-G; however, a tation. In human pregnancy, detection of PIF activ-
multiparous woman who lacked the receptor still ity in maternal sera predicted viable pregnancy after
had normal pregnancies.32 Also, HLA-G polymor- IVF with excellent sensitivity, specificity, and posi-
phism has been investigated in recurrent sponta- tive and negative predictive values (88%, 95%, 94%,
neous abortion, but no difference has been found and 90%, respectively) in 65 patients beginning
between the fertile and abortion-prone popula- 4 days after embryo transfer. In a retrospective
tions.33 HLA-G can be detected in human embryo study, the presence of PIF was found to be highly
culture media by specific immunoassays. However, specific (100%) for pregnancy.36 The accuracy of
pregnancy can also occur in its absence. When pres- the PIF assay in predicting successful and non-
ent, there is a higher pregnancy rate, and therefore viable pregnancies has been confirmed in another
HLA-G testing has been used to determine which study.37 Furthermore, the premature disappearance
embryos should be transferred after IVF.34 of PIF activity led to embryonic demise, approxi-
However, the soluble forms are not secreted by the mately 3 weeks prior to decline in hCG levels.
19
0415421306-Ch02 3/29/07 5:26 PM Page 20
Chromosomal analysis of these spontaneous abor- of mated mice led to decreased rates of fetal absorp-
tuses revealed the presence of an abnormal kary- tion (Barnea et al, unpublished work). PIF is being
otype in over 60%. Only one woman who lost a examined in additional models in preparation for
euploid conceptus had PIF activity that was positive clinical trials. Thus, one aspect of the activity of PIF,
4 days after embryo transfer.38 namely its immunomodulatory properties, is a
The next stage focused on identifying PIF and necessity in pregnancy. Moreover, the effect is cross-
elucidating its biological role. The biochemical species, since the peptide originally derived from
nature of PIF was examined by isolating it from the mouse is effective on human cells.
viable mouse embryo-conditioned media using a The second characteristic of PIF, in order for it to
multistep process including affinity chromatogra- be relevant to the very early stages of embryo devel-
phy and high-performance liquid chromatography opment, is its ability to interact with the
(HPLC).42 The peptide responsible for the biologi- endometrium. Our data show that PIF has a clear
cal activity generated by the bioassay was isolated. effect on human endometrial cells, increasing
It was found to be a novel 15-amino-acid peptide receptivity molecules. The presence of PIF as deter-
sharing partial sequence homology with the cir- mined by bioassay36 was recently validated in pri-
cumsporozoite protein of the malaria parasite. The mate blood and was shown to be associated with
peptide that was initially identified in mouse subsequent endometrial pre-epithelial plaque reac-
embryo culture media was also found to be present tion, angiogenesis, and stromal compaction, as an
in human embryos. Due to the simple structure of index of impending implantation.37
PIF, a synthetic analog was designed that mimics The next task was to show whether PIF can be
the native peptide’s properties. It was now possible detected in the early stages of pregnancy and is
to examine the biological effects of PIF in vitro and capable of exerting its effects in low concentrations
in vivo and to generate highly specific antibodies at the relevant time. The initial observations using
(polyclonal and monoclonal) that could detect and the PIF bioassay in both normal pregnancy serum
measure the presence of PIF in both gestational tis- and viable embryo-conditioned media have
sues and peripheral blood. been confirmed. Using various enzyme-linked
Synthetic PIF was shown to have a potent dose- immunosorbent assay (ELISA) formats, we have
and time-dependent effect by mostly affecting measured PIF concentrations in maternal blood
mitogen-activated human immune cells. This was 8–10 days after embryo transfers that led to success-
shown by blocking peripheral blood mononuclear ful pregnancies. Very recent data have also shown
cell (PBMC) proliferation, modulating the secre- that a sensitive ELISA (using monoclonal anti-
tion of both Th1 and Th2 cytokines, favoring the bodies) can detect pregnancy within 5 days of
latter. These effects were exerted through apparently insemination in cows. In addition, we have meas-
novel binding sites, and a mechanism of action dis- ured PIF concentrations in both viable single
tinct from that of immunosuppressive agents human (4–8 cells) and mouse embryo culture
(Barnea et al, unpublished work). We have demon- media (Barnea, Roussev, and Coulam, unpublished
strated that PIF indeed has a beneficial effect on work). These data demonstrated the link between
various immune disorders including multiple scle- the presence of the peptide in peripheral blood
rosis, juvenile diabetes, and graft vs host disease in sufficient amounts to explain its observed biolog-
(due to transfer of foreign immune cells to a host) ical effects.
using relevant mouse models (Barnea et al, unpub-
lished work). These effects were achieved without
toxicity, and low-dose short-term therapy led to CONCLUSIONS
either prevention or long-term protection against
disease. The work on synthetic PIF concurred with In conclusion, three essential elements are required
our earlier studies where short-term PIF treatment for pregnancy to succeed: a viable embryo, immune
20
0415421306-Ch02 3/29/07 5:26 PM Page 21
tolerance, and a receptive uterus. Based on our data, 8. Piccinni MP. Scaletti C, Malvilia C, et al. Production of IL-4 and
leukemia inhibitory factor by T cells of the cumulus oophorus: a
PIF plays a major role in all three aspects: it is only favorable microenvironment for pre-implantation embryo develop-
secreted by viable embryos, it modulates the mater- ment. Eur J Immunol 2001; 24:31–7.
nal immune system towards tolerance prior to 9. O’Neill C. Partial characterization of the embryo-derived platelet-
activating factor in mice. J Reprod Fertil 1985; 75:285–290.
implantation, and it primes the endometrium for 10. Pinkas H, Fisch B, Tadir Y, et al. Immunesuppressive activity in culture
implantation. Thus, PIF is a true biomarker for media containing oocytes fertilized in vitro. Arch Androl 1992;
pregnancy that can be used for both diagnostic and 28:53–59.
11. Fortin M, Oulette MJ, Lambert RD. TGF-β and PGE2 in rabbit
therapeutic purposes in pregnancy and other blastocoelic fluid can modulate GM-CSF production by human
conditions. lymphocytes. Am J Reprod Immunol 1997; 38:129–39.
A PIF-ELISA could be used to investigate very 12. Alizadeh Z, Kageyama SI, Aoki F. Degradation of maternal mRNA in
mouse embryos: selective degradation of specific mRNAs after fertil-
early pregnancy events and help to improve various ization. Mol Reprod Dev 2005; 72:281–90.
aspects of reproduction. In IVF, PIF-ELISA could be 13. Sharma, S, Murphy S, Barnea ER. Genes regulating implantation and
used to detect the viability of fertilized embryos by fetal development: a focus on mouse knockout models. Front Biosci
2006; 20:2123–37.
testing the embryo culture media. Consequently, 14. Choudhury SR, Knapp LA. Human reproductive failure I: immuno-
the current low pregnancy rates (20–25%) associated logical factors. Hum Reprod Update 2001; 7:113–34.
with morphological analysis for embryo viability 15. Raghupathy R. Th1-type immunity is incompatible with successful
pregnancy. Immunol Today 1997; 18:478–82.
might be greatly improved, and the multiple- 16. Druckman R, Druckman MA. Progesterone and the immunology of
pregnancy rate associated with current IVF prac- pregnancy. J Steroid Biochem Mol Biol 2005; 97:389–96.
tices might be reduced. In addition, very early 17. Kralickova M, Sima P, Rokyta Z. Role of leukemia-inhibitory factor
gene mutations in infertile women: the embryo–endometrial cytokine
events in human pregnancy remain poorly cross talk during implantation – a delicate homeostatic equilibrium.
explored, especially those at the peri-implantation Folia Microbiol (Praha) 2005; 50:179–86.
period. Measuring PIF levels may help us to better 18. Aplin JD, Kimber SJ. Trophoblast-uterine interactions at implanta-
tion. Reprod Biol Endocrinol 2004; 2:48.
understand the events taking place during this 19. Somerset DA, Zheng Y, Kilby MD, Sansom DM, Drayson MT. Normal
period. Furthermore, PIF assays might be used to human pregnancy is associated with an elevation in the human sup-
monitor high-risk pregnancies, such as recurrent pressive CD25+CD4+ regulatory T-cell subset. Immunology 2004;
112:38–43.
pregnancy loss; immunocytochemistry of prema- 20. Francis J, Rai R, Sebire NJ, et al. Impaired expression of endometrial
ture-labor placentas has documented very low or differentiation markers and complement regulatory proteins in
absent expression of PIF. These potential applica- patients with recurrent pregnancy loss associated with antiphospho-
lipid syndrome. Mol Hum Reprod 2006; 12:435–42.
tions of PIF, and the possible treatment of various 21. Buckingham KL, Stone PR, Smith JF, Chamley LW. Antiphospholipid
immune disorders, remain to be fully explored. antibodies in serum and follicular fluid – Is there a correlation with
IVF implantation failure? Hum Reprod 2006; 21:728–34.
22. Ntrivalas EI, Bowser CR, Kwak-Kim J, Beaman KD, Gilman-Sachs A.
REFERENCES Expression of killer immunoglobulin-like receptors on peripheral
blood NK cell subsets of women with recurrent spontaneous abortions
1. Beer AE, Billingham RE. Maternal immunological recognition mecha- or implantation failures. Am J Reprod Immunol 2005; 53:215–21.
nisms during pregnancy. Ciba Found Symp 1978; 64:293–322. 23. Cameo P, Srisuparp S, Strakova S, Fazleabas AT. Chorionic gonadotropin
2. Billingham RE, Head JR. Recipient treatment to overcome the allograft and uterine dialogue in the primate. Reprod Biol Endocrinol 2004; 2:50.
reaction, with special reference to nature’s own solution. Prog Clin Biol 24. O’Niell C. The role of PAF in embryo physiology. Hum Reprod
Res 1986; 224:159–85. Update 2005; 11:215–28.
3. Hansel W, Hickey GJ. Early pregnancy signals in domestic animals. Ann 25. Roudboush WE, Wininger JD, Jones AE, et al. Embryonic platelet-
NY Acad Sci 1988; 541:472–84. activating factor: an indicator of embryo viability. Hum Reprod 2002;
4. Weitlauf HM. Embryonic signaling at implantation in the mouse. Prog 17:1306–10.
Clin Biol Res 1989; 294:359–76. 26. Cavanagh AC, Morton H. The purification of early-pregnancy factor
5. Soares JM. The prolactin and growth hormone families: pregnancy- to homogeneity from human platelets and identification as chaperonin
specific hormones/cytokines at the maternal–fetal interface. Reprod 10. Eur J Biochem 1994; 222:551–60.
Biol Endocrinol 2004; 2:51. 27. Ohnuma K, Ito K, Takahashi J, Nambo Y, Miyake Y. Partial purification
6. Moffett A, Loke YW. The immunological paradox of pregnancy: a of mare early pregnancy factor. Am J Reprod Immunol 2004; 51:95–101.
reappraisal. Placenta 2004; 25:1–8. 28. Athanasas-Platsis C, Somodevilla-Torres MJ, Morton H, Cavanagh
7. Wright JM, Kiracofe JH, Beeman KB. Factors associated with short- AC. Investigation of the immunocompetent cells that bind early preg-
ened estrous cycles after abortion in beef heifers. J Anim Sci 1988; nancy factor and preliminary studies of the early pregnancy factor
66:3185–9. target molecule. Immunol Cell Biol 2004; 82:361–9.
21
0415421306-Ch02 3/29/07 5:26 PM Page 22
29. Akyol S, Gercel-Taylor C, Reynolds HS, Taylor DD. HSP-10 in ovarian 36. Barnea ER, Lahijani KI, Roussev R, Barnea JD, Coulam CB. Use
cancer: expression and suppression of T-cell signaling. Gynecol Oncol of lymphocyte platelet binding assay for detecting a preimplantation
2006; 101:481–6. factor: a quantitative assay. Am J Reprod Immunol 1994; 32:133–8.
30. Fuzzi B, Rizzo R, Criscuoli L, et al. HLA-G expression in early 37. Rosario GX, Modi ND, Sachdeva G, et al. Morphological events in the pri-
embryos is a fundamental prerequisite for the obtainment of preg- mate endometrium in the presence of a preimplantation embryo, detected
nancy. Eur J Immunol 2002; 32:311–15. by the serum preimplantation assay. Hum Reprod 2005; 20:61–71.
31. Bainbridge D, Ellis S, Le Bouteiller P, Sargent I. HLA-G remains a 38. Coulam CB, Roussev RG, Thomasson EJ, Barnea ER. Preimplantation
mystery. Trends Immunol 2001; 22:548–52. factor (PIF) predicts subsequent pregnancy loss. Am J Reprod
32. Gomez-Lozano N, de Pablo R, Puente S, Vilches C. Recognition of Immunol 1995; 34:88–92.
HLA-G by the NK receptor KIR2DL4 is not essential for human 39. Roussev RG, Coulam CB, Kaider BD, Yarkoni M, et al. Embryonic
reproduction. Eur J Immunol 2003; 33:639–44. origin of preimplantation factor (PIF): biological activity and partial
33. Yan WH, Fan LA, Yang JQ, et al. HLA-G polymorphism in a Chinese characterization. Mol Hum Reprod 1996; 2:883–7.
Han population with recurrent spontaneous abortion. Int J 40. Roussev RG, Barnea ER, Thomason EJ, Coulam CB. A novel bioassay
Immunogenet 2006; 33:55–8. for detection of preimplantation factor (PIF). Am J Reprod Immunol
34. Criscuoli L, Rizzo R, Fuzzi B, et al. Lack of histocompatibility leuko- 1995; 33:68–73.
cyte antigen-G expression in early embryos is not related to germinal 41. Barnea ER, Simon J, Levine SP, et al. Progress in characterization of
defects or impairment of interleukin-10 production by embryos. pre-implantation factor in embryo cultures and in vivo. Am J Reprod
Gynecol Endocrinol 2005; 20:264–9. Immunol 1999; 42:95–9.
35. Sargent IL. Does ‘soluble’ HLA-G really exist? Another twist to the 42. Barnea ER. Insight into early pregnancy: emerging role of the embryo.
tale. Mol Human Reprod 2005; 11:695–8. Am J Reprod Immunol 2004; 51:319–22.
22
0415421306-Ch03 3/29/07 5:27 PM Page 23
3. Genetics of spontaneous abortions

Joe Leigh Simpson
INTRODUCTION 6% aneuploidy in sperm from ostensibly normal

males4,5 and in 20% in oocytes.6,7 Aneuploidy rates in
Genetic factors are the most common causes of embryos and oocytes increase as maternal age increases.
spontaneous abortions. From 50% to 80% of first- Chromosomal abnormalities are even more fre-
trimester abortions show chromosomal abnormalities. quent in morphologically abnormal embryos.
Additionally, there are other genetic causes distinct Using FISH with chromosome-specific probes,
from chromosomal abnormalities. Single-gene abnormality rates of 50–75% can be observed, even
abnormalities are almost unexplored in sponta- when not all chromosomes are tested.3
neous abortions, although single-gene defects are a
more common cause of birth defects in liveborns
than chromosomal abnormalities. Many causes of CHROMOSOMAL ABNORMALITIES: THE MOST
repeated abortions that are commonly classified as FREQUENT EXPLANATION FOR CLINICALLY
‘non-genetic’ are actually the result of perturbations RECOGNIZED SPONTANEOUS ABORTIONS
of gene products such as factor V Leiden, and other
genes associated with thromboembolism, and FREQUENCY
alloimmune pregnancy loss (which may involve
shared human leukocyte antigen (HLA) molecules). No less than 50% of clinically recognized pregnancy
In this chapter, we shall restrict the discussion to the losses show a chromosomal abnormality.8–10 This
frequency and most common genetic causes of spo- figure is based on analysis of spontaneously expelled
radic and recurrent abortions. products. If chorionic villus sampling (CVS) is per-
formed after ultrasound diagnosis of fetal demise,
the frequency is 75–90%.11,12 Comparative genomic
CHROMOSOMAL ABNORMALITIES IN hybridization (CGH; microarray analysis) also
PREIMPLANTATION EMBRYOS reveals abnormalities in abortuses that are not evi-
dent by karyotype.13 Using chromosomal microar-
The frequency of losses in human preimplantation rays, additional abnormalities can be detected.
embryos is very high.1,2 This is reflected by pregnancy Schaeffer et al14 performed CGH using microarrays
rates being no more than 25–35% per embryo trans- on 41 abortuses that had previously been analyzed
fer in assisted reproductive technology (ART), even by karyotype. Array analysis revealed heretofore
in experienced hands. Of morphologically normal unrecognized abnormalities in 4 of the 41 cases.
embryos, no less than 25% show chromosomal In the second trimester, chromosomal abnormali-
abnormalities (aneuploidy or polyploidy).3 This is ties are less frequent. The actual frequency is not cer-
based on studies using fluorescence in situ hybri- tain, because many abortuses recognized in the
dization (FISH) with chromosome-specific probes second trimester are really missed abortions that were
for only seven to nine chromosomes; rates would retained in utero after a first-trimester demise. It has
doubtless be higher if it were possible to routinely long been recognized that fetal demise may precede
perform a complete karyotype or microarray analysis spontaneous expulsion of the products of conception
on a blastomere. The 25% aneuploidy rate in mor- by several weeks.15 The anomalies seen in second-
phologically normal embryos is consistent with trimester abortions are similar to those observed in
23
0415421306-Ch03 3/29/07 5:27 PM Page 24
liveborn infants: trisomies 13, 18, and 21; monosomy X;

Table 3.1 Chromosomal completion in sponta-
and sex chromosomal polysomies. The frequency of neous abortions; recognized clinically in the first
these anomalies is estimated to be approximately 15%. trimestera
In third-trimester losses (stillborn infants), the
Completion Frequency Percentage
frequency of chromosomal abnormalities is 5%.16
This incidence is less than that observed in earlier Normal: 46,XX or 46,XY 54.1
abortuses, but greater than the 0.6% found in live- Triploidy: 7.7
borns. A major problem in assessing the frequency 69,XXX 2.7
69,XYX 0.2
of chromosomal abnormalities is that maceration 69,XXY 4.0
ensues soon after fetal death – usually days in Other 0.8
advance of delivery. Hefler et al17 found that 63% of Tetraploidy: 2.6
92,XXX 1.5
139 third-trimester losses were macerated, impeding 92,XXYY 0.55
accurate morphological assessment and cytogenetic Not stated 0.55
studies. Again, a large series of stillbirths studied by Monosomy X 18.6
Structural abnormalities 1.5
microarray analysis would be informative. Sex-chromosomal polysomy: 0.2
47,XXX 0.05
SPECTRUM OF CHROMOSOMAL ABNORMALITIES 47,XXY 0.15
Autosomal monosomy (G) 0.1
Autosomal trisomy for
AUTOSOMAL TRISOMY chromosomes: 22.3
1 0
2 1.11
Autosomal trisomies comprise approximately 50% 3 0.25
of cytogenetically abnormal spontaneous abor- 4 0.64
tions. Trisomy for every chromosome has now been 5 0.04
6 0.14
observed. Table 3.1 shows frequencies in one series. 7 0.89
The most common trisomies are 16, 22, 21, 15, 13, 8 0.79
and 14 (in descending order). Trisomy 16 is rarely, 9 0.72
10 0.36
if ever, observed in liveborns in non-mosaic form, 11 0.04
but is the most common aberration in the abortus. 12 0.18
These six chromosomes in aggregate account for 13 1.07
14 0.82
70% of trisomies – an important consideration in 15 1.68
selecting probes to exclude aneuploidy in preim- 16 7.27
plantation genetic diagnosis (PGD). 17 0.18
18 1.15
Correlations between placental morphological 19 0.01
abnormalities and specific trisomies have been 20 0.61
attempted, but remain imprecise. Attempts are 21 2.11
22 2.26
complicated by non-specific villous changes follow- Double trisomy 0.7
ing fetal demise in utero. Thus, low predictive value Mosaic trisomy 1.3
exists when placental histology is used to distin- Other abnormalities or
not specified 0.9
guish aneuploid from euploid abortuses. A few 100.0
correlations are valid. Fetuses with trisomies
a
Pooled data from several series, as referenced by Simpson and
incompatible with life grow more slowly than those Bombard.10
with trisomies compatible with life (e.g., trisomies
13, 18, and 21). The mean crown–rump length
(CRL) for the latter is 20.65mm, compared with
only 10.66 mm for the former.18 Either fetuses with
non-lethal trisomies live longer than those with lethal
24
0415421306-Ch03 3/29/07 5:27 PM Page 25
GENETICS OF SPONTANEOUS ABORTIONS
trisomies, or the fetuses with lethal trisomies also whether the abnormalities detected in polar bodies
exhibit greater intrauterine growth retardation are as directly applicable to clinically recognized
(IUGR), or both. Abortuses from non-lethal tri- abortuses as they clear are to preimplantation
somies (13, 18, and 21) tend to show anomalies con- embryos. Polar body studies have revealed that rates
sistent with those found in full-term liveborn of meiosis I errors are only marginally higher
trisomic infants.18,19 Malformations observed may (41.7% versus 35.2) than meiosis II errors; errors in
be more severe than those found in induced abor- both meiosis I and II are not uncommon.28 The rel-
tuses detected after prenatal diagnosis. ative distribution of errors thus differs from that
Most trisomies show a maternal age effect, but observed in trisomies recovered later in pregnancy.
the effect varies among chromosomes. The maternal
age correlates positively with errors at meiosis I, the DOUBLE TRISOMY
most common cytological explanation for trisomies.
The proportion of trisomies that arise at meiosis I The frequency of double trisomy in abortuses is more
versus meiosis II varies among aneuploidies. common than expected by chance. The frequency
Virtually all trisomy 16 cases are maternal in origin, varies more than for other chromosomal abnormali-
and arise in meiosis I.20 In trisomies 13 and 21, 90% ties, which may reflect vicissitudes of culture (failure)
are maternal, usually arising at meiosis I. In trisomy 18, or differences in sample characteristics (maternal or
however, two-thirds of the 90% of the maternal gestational age). Table 3.1 (based on series collected
origin cases arise at meiosis II.21,22 Maternal meiosis up to 1987) shows that double trisomies accounted
errors correlate not only with advanced maternal for 0.7% of abortuses. A similar prevalence was
age, but also with decreased or absent meiotic observed in pooled data tabulated in 1997 by Reddy.29
recombination.21–23 The mechanism invoked to However, a more recent report of 517 abortuses
explain this relationship is the product-line hypothesis. found double trisomies in 2.2% of 321 successfully
Oocytes ovulated earlier in life are believed to be karyotyped abortuses.30 Double trisomies most
more likely to have undergone genetic recombina- often involve the X chromosome, but may involve
tion and hence are less predisposed to non- the Y chromosme, or autosomes 21, 18, 16, 22, 13, 8,
disjunction.24 The location of the recombinant 2, and 15 in descending order (Table 3.2). Diego-
event on the chromosome and the exact nature of Alvarez et al30 have described the exact combinations
recombination are pivotal, as discussed elsewhere.25 of the 178 reported double trisomies. In liveborns,
Errors in paternal meiosis account for 10% of acro- approximately 50 double trisomies have been
centric (13, 14, 15, 21, and 22) trisomies.26 In non- reported.31 In liveborns, usually one of the additional
acrocentric trisomies paternal meiotic errors are chromosomes is an X and the other is 13, 18, or 21.
equally likely to arise at meiosis I or II.27 Paternal mei- The gestational age was 8.7 ± 2.2 weeks at abor-
otic errors account for 10% of trisomy 21 cases, and tion in double trisomies in Reddy’s29 series, com-
for some cases of trisomy 2 abortuses. A paternal con- pared with 10.1 ± 2.9 weeks for a single trisomy. In
tribution is uncommon in other abortus trisomies. the series of Diego-Alvarez et al,30 the gestational
The ability to analyze polar bodies (1st and 2nd) age was 8.2 ± 1.7 for double trisomies. The sex ratio
for PGD has generated a new body of information was approximately 1 in both series.
on maternal meiosis.28 This topic is beyond the Morphological examination usually reveals an
scope of this chapter, and it remains unclear empty sac29,30 and only occasionally an embryo.
Table 3.2 Chromosomes involved in double-trisomy abortuses30
Chromosome 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 X/Y Total

No. of cases 0 15 0 5 4 3 7 18 5 2 1 2 18 7 13 37 2 44 0 8 66 20 79 360
25
0415421306-Ch03 3/29/07 5:27 PM Page 26
In one study, 5 of 7 double trisomies showed no Triploid abortuses are usually 69,XXY or
morphological detail;30 one was anembryonic and 69,XXX. The origin has long been presumed to be
the other (48,XXX,+18) showed hydrops fetalis. due to dispermy, and this has been verified.33,36,37
Advanced maternal age is a striking feature.29–31 Triploidy may follow either fertilization by two haploid
In the series of Diego-Alvarez et al,30 the mean maternal sperm or fertilization by single diploid sperm.37,38
age was 39.7 ± 3.4 years. Almost all analyzed cases
originated at maternal meiosis. As expected, the stage TETRAPLOIDY
of meiotic error is consistent with that expected for
single trisomies. Thus, double trisomy involving Tetraploidy (4n = 92) is uncommon, rarely pro-
chromosome 18 is more likely to show meiosis II gressing beyond 2–3 weeks of embryonic life. This
errors than 48,XX,+16,+21. chromosomal abnormality can be associated with
persistent trophoblastic disease, and thus needs to
POLYPLOIDY be identified in order to provide appropriate
follow-up. Tetraploidy in embryonic tissue should
In polyploidy, more than two haploid chromosomal be distinguished from the not uncommon, and
complements exist. Non-mosaic triploidy (3n = 69) clinically insignificant, tetraploid cells found in
and tetraploidy (4n = 92) are not common in abor- amniotic fluid. Although uncommon, true fetal
tuses (Table 3.1). Diploid/triploid mosaicism is found tetraploidy does exist,39 and probably arises from
in approximately 30% of blastocysts. However, placen- failure of cytokinesis.40 Failure of cytokinesis has
tal mosaicism of this type is thought to involve been deduced on the basis of chromosomal com-
trophoectoderm rather than the embryo per se (inner plement (92,XXXX or 92,XXYY), and more recently
cell mass), and will therefore not be discussed here.32 confirmed by molecular studies.41
Of general interest, however, is the association between
diandric (paternally inherited) triploidy and hydatidi- MONOSOMY X
form mole. A ‘partial mole’ exists if molar tissue and
fetal parts coexist. Partial (triploid) moles must be dis- Monosomy X accounts for 15–20% of chromoso-
tinguished from the more common ‘complete’ hyda- mally abnormal specimens. Autosomal monosomy
tidiform moles. Complete moles are 46,XX, exclusively appears to be lethal prior to or just beyond implan-
of androgenetic origin, and exclusively villous tissue.33 tation, and thus seems not to persist to clinical
Placental findings in diandric triploid placentas recognition. Early monosomy X abortuses usually
include a disproportionately large gestational sac, consist of only an umbilical cord stump. If survival
focal (partial) hydropic degeneration of placental persists until later in gestation, anomalies charac-
villi, and trophoblast hyperplasia.34 Placental teristic of Turner syndrome may be seen. These
hydropic changes are progressive, and hence diffi- include cystic hygromas, generalized edema, and
cult to identify early in early pregnancy. Irrespective cardiac defects. Unlike liveborn 45,X individuals,
of chromosomal status, placental villi also undergo 45,X abortuses show germ cells; however, germ cells
non-specific hydropic degeneration following fetal rarely develop beyond the primordial stage. The
demise. This makes histological and cytogenetic pathogenesis of 45,X germ cell failure seems to be
corrections difficult. Embryonic/fetal malforma- rapid attrition of germ cells, rather than failure of
tions associated with triploid abortuses include germ cell development.42,43 Rapid attrition of germ
neural tube defects and omphaloceles – anomalies cells explains the rare but well-documented preg-
reminiscent of those in triploid conceptuses surviv- nancies occurring in 45,X individuals. Mosaicism
ing to term. Facial dysmorphia and limb abnormal- (45,X/46,XX) need not necessarily be invoked as the
ities have also been reported.35 There is no mechanism explaining pregnancies.44
correlation between embryonic morphology and Approximately 80% of monosomy X occurs
parental origin (diandry or digyny).35 as a result of paternal sex chromosome loss.45
26
0415421306-Ch03 3/29/07 5:27 PM Page 27
Consequently, there is a lack of a maternal age effect significant with respect to therapeutic management
in 45,X. An inverse age effect has been reported. (or lack thereof).
Some of the non-random distributions reflect an
SEX CHROMOSOMAL POLYSOMY (X OR Y) increasing incidence of aneuploidy as maternal age
increases. Adjustments for maternal age account for
The complements 47,XXY and 47,XYY each occur some of the ostensibly non-random distribution,
in about 1 per 800 liveborn male births; 47,XXX and, in the opinion of Warburton et al,46 precluded
occurs in 1 per 800 female births. X or Y polysomies a relationship. The study by Warburton et al46
are slightly (10%) more common in abortuses than pooled cases from a New York City sample and a
in liveborns. Hawaii sample.47 However, a confounder is that in
the New York City cases, the inclusion criteria
extended to 28 weeks’ gestation; these cases pre-
RECURRENT ANEUPLOIDY dictably had a lower overall aneuploidy rate than
the earlier-gestation sample from Hawaii. Hence,
DOES RECURRENT ANEUPLOIDY EXIST? recurrent aneuploidy that had previously seemed to
clearly exist in the Hawaii sample of Hassold et al,47
In first-trimester abortions, recurrent aneuploidy but was not statistically confirmed by Warburton et al.46
occurs more often than expected by chance. Studying recurrent aneuploidy in preimplantation
However, a lack of consensus exists on the extent to embryos has since seemingly convinced Warburton
which numerical chromosomal abnormalities (ane- of the concept of recurrent aneuploidy.48
uploidy) explain recurrent losses. In my view, recur- A different approach that also supports the con-
rent aneuploidy is a frequent explanation – at least cept of recurrent aneuploidy is calculation of aneu-
until the number of losses exceeds 4. This reasoning ploidy rates in prenatal diagnosis samples, in
is based on observations that the chromosomal comparison with prior pregnancy outcome. Bianco
complements of abortuses in a given family are et al48 studied 46 939 women undergoing prenatal
more likely to be either recurrently normal or genetic diagnosis (CVS or amniocentesis).
recurrently abnormal (Table 3.3). That is, if the The prevalence of aneuploidy increased progres-
complement of the first abortus is abnormal, sively as the number of prior spontaneous abor-
the likelihood is increased that the complement of the tuses increased (Table 3.4): 1.39% with no prior
second abortus will also be abnormal.46 Recurrence abortuses, 1.67% after one, 1.84% after two, and
usually involves trisomy. The ramifications become 2.18% after three abortions. After adjustments for
Table 3.3 Recurrent aneuploidy: relationship between karyotypes of successive abortuses46
Complement of second abortus
De novo
Complement of first abortus Normal Trisomy Monosomy Triploidy Tetraploidy rearrangement
Normal 142 18 5 7 3 2
Trisomy 31 30 1 4 3 1
Monosomy X 7 5 3 3 0 0
Triploidy 7 4 1 4 0 0
Tetraploidy 3 1 0 2 0 0
De novo 1 3 0 0 0 0
Rearrangement
27
0415421306-Ch03 3/29/07 5:27 PM Page 28
Table 3.4 Risk of aneuploidy by number of prior miscarriage; stratified by maternal age. Comparison is with
women with no spontaneous abortions, controlling for parity and indications for prenatal diagnosis49
Maternal age < 35 years
No. of prior spontaneous abortions Adjusted OR for trisomy 13, 18, or 21a Adjusted OR for all aneuploidiesa
0 1.00 1.00
1 1.27 (0.74–2.08) 1.19 (0.78–1.84)
2 1.31 (0.80–2.13) 1.21 (0.94–1.58)
≥3 1.36 (0.46–2.73) 1.41 (0.56–3.19)
Maternal age > 35 years
No. of prior spontaneous abortions Adjusted OR for trisomy 13, 18, or 21a Adjusted OR for all aneuploidiesa
0 1.00 1.00
1 1.23 (1.04–1.52) 1.23 (1.00–1.52)
2 1.34 (1.01–1.82) 1.30 (0.99–1.74)
≥3 1.56 (1.03–2.31) 1.68 (1.12–2.52)
a
OR, odds ratio. 95% confidence interval in parenthesis.
maternal age, ethnicity, and type of invasive proce- about 1%.52 Based on first-trimester trisomies,
dure (a surrogate indicator of gestational age), the which may or may not survive, Snijders and
odds ratios were 1.21 (95% confidence interval (CI) Nicholaides53 reported a recurrence rate of 0.7%
1.01–1.47), 1.26, and 1.51, respectively. These findings following trisomy 21 and 0.7% following trisomy 18.
thus confirmed an earlier study by Drugan et al.50 Bianco et al49 describe the consequences of a prior
Further supporting recurrent aneuploidy as a abortion of unknown karyotype. If abortions are
genuine phenomenon is the occurrence of trisomic recurrent but no information is available on the
preimplantation embryos in successive ART cycles. chromosomal status, the odds ratios provided by
Rubio et al51 showed increased aneuploid embryos Bianco et al49 can be applied to give a patient spe-
in couples with repeated abortions, compared with cific risk (Table 3.4). For example, if the a priori
couples undergoing PGD for mendelian indications. Down syndrome risk is 1 in 300, a woman’s calcu-
Frequencies of chromosomal abnormalities were lated risk after 3 abortions would be 1/300 × 1.5, or
71% versus 45%, respectively. In a similar study, 1/200.
Munne et al48 found rates to be 37% versus 21% in
women under age 35 years, and 34% versus 31.5% EXPECTATIONS OF THE KARYOTYPE IN RECURRENT
in women over 35 years. ABORTION
CONSEQUENCES FOR GENETIC COUNSELING The concept of recurrent aneuploidy has certain
corollaries, one of which has been the subject of
If couples are predisposed to recurrent aneuploidy, controversy. Given the existence of recurrent aneu-
they might logically be at increased risk not only for ploidy, and that 50% of all abortuses are abnormal
aneuploid abortuses but also for aneuploid live- cytogenetically, aneuploidy should be as likely to be
borns. The trisomic autosome in a subsequent detected in a recurrent abortus as in a sporadic
pregnancy might not always confer lethality, but abortus. This has proved to be true in most series.
might be compatible with life (e.g., trisomy 21). Stern et al54 found a 57% prevalence of chromoso-
Indeed, the risk of liveborn trisomy 21 following an mal abnormalities among abortuses of repetitively
aneuploid abortus has long been considered to be aborting women – a frequency coincidentally
28
0415421306-Ch03 3/29/07 5:27 PM Page 29
identical among abortuses of sporadically aborting CONSEQUENCES FOR CLINICAL MANAGEMENT

women. Among 420 abortuses obtained from
women with repeated losses, Stephenson et al55 That recurrent aneuploidy exists dictates that nei-
found 46% chromosomal abnormalities; 31% of ther a genetic nor a non-genetic etiology should be
the original sample was trisomic. Their comparison assumed on the basis of number of losses. Often,
was unselected pooled data, which showed 48% of there is no information concerning the chromoso-
abortuses to be abnormal; 27% of the original mal status of prior abortuses. However, paraffin
sample was trisomic. blocks of products of conception are suitable for
Other authors have concluded that a recurrent FISH analysis of chromosomes most likely to be tri-
abortus is likely to be cytogenetically normal somic (13, 14, 16, 18, 21, and 22). Chromosomal
whereas a sporadic abortuses will be cytogenetically microarrays may also yield information on paraffin
abnormal. Carp et al56 found that among women block specimens.58 If no information can be
having three or more abortuses, the likelihood that obtained, it is less clear whether prenatal genetic
the abortus would have an abnormal karyotype was diagnosis is appropriate. However, the risk of an
29%. After an aneuploid abortus, the likelihood of a aneuploid offspring is still increased, and indeed
subsequent live birth was 68% (13 of 19). If the can be calculated as discussed above.49 The small
abortus was euploid, the subsequent live birth rate but finite risk of amniocentesis or CVS is especially
was 41% (16 of 39). One explanation for the differ- troublesome to couples who have had difficulty
ence between this study and those cited above maintaining a pregnancy. Non-invasive approaches
might be a different referral pattern, for example may be the preferable initial option, but the
biased toward autoimmune causes. A second possi- sensitivity for detecting aneuploidy does not exceed
bility is simply the small numbers in each subgroup. 85–95%.59 PGD is another option. The selective
A third and more likely possibility is the increased transfer of euploid embryos clearly decreases
gestational age in this sample. That only 29% of clinical abortions in couples with RPL.60,61 Studies
abortuses in the series of Carp et al56 were chromo- are underway to verify the expected increase in
somally abnormal is consistent with inclusion liveborns.
criteria extending to 20 weeks’ gestation. There is
less reason to expect recurrent aneuploidy in the
second trimester, given the low (15%) frequency of STRUCTURAL CHROMOSOMAL
chromosomal abnormalities in the second REARRANGEMENTS: TRANSLOCATIONS
trimester. A fourth possibility is the higher mean
number of previous pregnancy losses (4.7) in the Structural chromosomal rearrangements are an
series of Carp et al.56 important cause of recurrent abortions, but account
for only 1.5% of all abortuses. The presence of a
RELATIONSHIP OF RECURRENT ANEUPLOIDY TO balanced rearrangement in one parent can result in
NUMBER OF LOSSES an unbalanced translocation in the offspring.62
Phenotypic consequences depend on the specific
Although recurrent aneuploidy appears to exist duplicated or deficient chromosomal segments.
with two or three losses, this does not necessarily
hold for higher-order losses. These seem more likely FREQUENCY
to be cytogenetically normal.57 Maternal factors
thus become plausible explanations when numbers A balanced translocation is found in 4–5% of couples
of losses exceed four. Consecutive losses of high experiencing repeated losses.63–65 These individuals
number also favor non-aneuploid explanations, are themselves phenotypically normal, but their off-
because one would not necessarily expect every spring (abortuses or abnormal liveborns) may show
single abortus to be aneuploid. chromosomal duplications or deficiencies as a result
29
0415421306-Ch03 3/29/07 5:27 PM Page 30
of normal meiotic segregation. The frequency of offspring of either female heterozygotes or male
balanced translocations is higher in females than heterozygotes.67,68
males,63 and if there is a family history of a stillborn The mode of ascertainment is important in
or abnormal liveborn.65 counseling. The frequency of unbalanced fetuses is
The likelihood of detecting a translocation het- lower if a parental balanced translocation was ascer-
erozygote does not necessarily reflect maternal age,66 tained through repetitive abortions (3%) than
nor does the likelihood of detecting a balanced through anomalous liveborns (nearly 20%).67
translocation substantially differ after 1, 2, or 3 mis- Presumably, the likelihood of severely unbalanced
carriages. In the tabulation by Simpson et al,63 detec- products (e.g., 3:1 segregation) is greater in the
tion rates in females after 2, 3, 4, and 5 losses were former. Detecting a chromosomal rearrangement in
0.8%, 1.7%, 2.3%, and 2.9%, respectively. For males, a parent obviously dictates that prenatal cytogenetic
the respective rates were 1.2%, 1.9%, 2.4%, and studies should be offered. Even if there is normal
0 (0/39). In the study by Goddijn et al,66 the odds transmission of chromosomes involved in the
ratios for finding a balanced translocation after 2, 3, translocation, a different chromosome could be
and 4 or more losses were 1.4 (95% CI 0.4–4.8), 2.2 aneuploid (interchromosomal effect).
(0.4–12.5), and 2.1 (0.3–15.4), respectively.
LIKELIHOOD OF SUBSEQUENT ABORTIONS
LIKELIHOOD OF ABNORMAL LIVEBORNS
Distinct from the likelihood of unbalanced segre-
There are two general types of translocations: gants is the quantitative likelihood of subsequent
Robertsonian and reciprocal. Robertsonian abortion. Does this differ from the expected
translocations involve centric fusion of an acro- 65–70% live birth rate observed in the general pop-
centric (13, 14, 15, 21, 22) chromosome. The theo- ulation with recurrent pregnancy loss (RPL)? A less
retical risk of a parent with t(14q;21q) having a favorable prognosis has been reported. In the study
liveborn child with Down syndrome is 33%. by Sugiura-Ogasawara et al,69 the loss rates were
However, the empirical risks are considerably less, 61% (11/18) for couples in which the male partner
given the lethality of certain complements. The had a translocation and 72.4% (21/29) if the female
risks are 2% if the father carries a translocation partner had the translocation. Of 1184 couples with
involving chromosome 21 and 10% if the mother two or more miscarriages who had normal kary-
carries such a translocation.67,68 Robertsonian otypes, the miscarriage rate, by contrast, was only
(centric fusion) translocations involving chromo- 28.3% (335/1184).69 In 2004, Carp et al70 reported
somes other than chromosome 21 show lower that 45.2% (33/73) pregnancies of couples with a
empirical risks. In t(13q;14q), the risk for liveborn translocation heterozygote resulted in a live birth,
trisomy 13 is 1% or less. This low risk presumably compared with 55.3% (325/588) without a translo-
reflects the lethality of many segregant products cation. The same group later found a similar per-
(trisomies and monosomies). centage of normal and balanced karyotypes (74%)
In reciprocal translocations, interchanges occur in embryos of translocation heterozygotes as well as
between two or more metacentric chromosomes. embryos of couples without a translocation
Empirical data for specific translocations are usu- (77%).71 Carp et al71 concluded that any decrease in
ally not available, and generalizations are typically the live birth rate was due to factors unrelated to the
made on the basis of pooled data derived from chromosomal imbalance. Different results have
many different translocations. As in Robertsonian been reported by others. Goddijn et al66 reported
translocations, the theoretical risks for abnormal only 26% miscarriages among 43 pregnancies in
offspring (unbalanced reciprocal translocations) 25 carrier couples. However, almost half of the patients
are much greater than the empirical risks. The sex in the series of Goddijn et al66 (55/115) had only
differences are less apparent. The risks are 12% for two miscarriages, which may account for the
30
0415421306-Ch03 3/29/07 5:27 PM Page 31
different results. Stephenson and Sierra72 studied

1893 couples, 40 of whom had a balanced translo- INVERSIONS
cation (28 reciprocal and 12 Robertsonian). Among
35 monitored pregnancies in the reciprocal trans- In an inversion, the order of genes is reversed.
location group, the live birth rate was 63% (22/35); The clinical consequence is analogous to that of a
in the Robertsonian translocation group, it was translocation, in that individuals heterozygous for
69% (9/13). These data are comparable to those in an inversion are normal but their genes are
the general RPL population. However, the propor- rearranged. Likewise, these individuals suffer unto-
tion of structurally unbalanced abortuses was ward reproductive consequences as result of normal
increased. Among abortuses of the translocation meiotic phenomena. However, crossing-over
heterozygote couples, 13 of 36 (36%) were unbal- involving the inverted segment may produce unbal-
anced, 11 of 36 (30%) aneuploid for another anced gametes. Duplication exists for some regions
chromosome (interchromosomal effect), and only and deficiency for others. There are two types of
12 of 36 (33%) normal. Among recurrent miscar- inversions. In pericentric inversions, breaks occur in
riage couples not having a translocation, the rates both arms. In paracentric inversions, breaks occur
were 2%, 44%, and 54%, respectively. These find- on the same arm. The frequency of inversions in
ings are at odds with those of Carp et al.71 However, couples having repetitive abortions is less than 1%.
the series of Stephenson and Sierra72 included
7 patients (14%) with two previous losses, whereas LIKELIHOOD OF ABNORMAL LIVEBORNS
that of Carp et al71 was restricted to patients with
three or more losses (mean 4.27). The different Females with a pericentric inversion have a 7% risk
number of previous losses may partially explain the of abnormal liveborns; males carry about a 5%
different results. risk.74 Pericentric inversions ascertained through
Rarely, a translocation precludes a normal live- phenotypically normal probands are less likely to
born infant. This occurs when a translocation result in abnormal live infants. The extent of origin
involves homologous acrocentric chromosomes of the crossing within the inverted segment influ-
(e.g., t(13q;13q) or t(21q;21q)). The only possibil- ences the likelihood of a fetal anomaly. The clinical
ity of normalcy is if trisomic rescue occurs; outcome is paradoxical. Inversions involving only
i.e., the ‘additional’ chromosome is ‘expelled’ from a small portion of the total chromosomal length
the nucleus to yield the normal chromosomal are usually lethal because, when recombinants
number with one homologous chromosome. If the arise, they yield large duplications or deficiencies.
father carries a homologous structural rearrange- By contrast, products of larger inversions, involving
ment, artificial insemination may be appropriate. 30–60% of the total chromosomal length, are rela-
If the mother carries the rearrangement, donor tively more likely to survive.74 On a molecular level,
oocytes or donor embryos should be considered. inversions less than 100 Mbp appear not to exert
In conclusion, when a balanced translocation is undue untoward outcomes.75 There were no recom-
detected in a couple experiencing recurrent binants in a tabulation when inversion was less than
abortions, the prognosis for a live birth remains 50 Mbp (40% of chromosome) length and only a few
uncertain compared with the situation if a around 50 Mbp (40–50%) length; a higher number
translocation had not been detected.70 In my occurred if the inversion was greater than100 Mbp.75
opinion, the increased frequency of loss dictates Data are limited on recurrence risks involving
offering the option of PGD, given that 80–100% paracentric inversions. Theoretically, there should
of embryos can be non-viable.28,73 The strategy is be a lower risk of unbalanced products of clinical
to identify and transfer only the (few) balanced consequence than with pericentric inversions,
embryos. Indeed, this decreases the likelihood of because nearly all paracentric recombinants should
abortion.28,60 be lethal. However, both abortions and abnormal
31
0415421306-Ch03 3/29/07 5:27 PM Page 32
liveborns have been observed within the same kin- complement may still point to a genetic etiology.
dred. The risk for unbalanced viable offspring has The lack of cytogenetic data on dissected specimens
been tabulated to be 4%.76 has made it difficult to determine the exact role that
non-cytogenetic mechanisms play in early embryonic
LIKELIHOOD OF SUBSEQUENT ABORTIONS maldevelopment. Philipp and Kalousek78 sought to
address this by correlating the cytogenetic status of
Few data exist on the likelihood of abortion follow- missed abortions with morphological abnormalities
ing the detection of an inversion in one of the par- observed at embryoscopy. Embryos with chromo-
ents. In the series of Stephenson and Sierra,72 there somal abnormalities usually showed one or more
were 7 inversion carriers with 35 pregnancies; 31% external anomalies, but some euploid embryos also
were livebirths and 69% miscarriages (24/35). showed anatomical anomalies.
In the series of Carp et al,70 8 of 15 pregnancies Indirect evidence further points to a mendelian
(53%) were live births. These outcomes are less etiology in human abortuses. Mosaicism may be
favorable than in the general population, but data restricted to the placenta, the embryo per se being
are sparse. normal. Termed ‘confined placental mosaicism,’ this
phenomenon has as a corollary uniparental disomy
(UPD). In UPD, both homologues for a given chro-
DEVELOPMENTAL GENES: MENDELIAN mosome are derived from a single parent. UPD is
AND POLYGENIC/MULTIFACTORIAL thought to occur as result of expulsion of a chro-
mosome from a trisomic zygote (‘trisomic rescue’).
From 50% to 80% of first-trimester abortuses show Although the karyotype would be euploid (46,XX
chromosomal abnormalities. Casual deductions or 46,XY), genes on the involved chromosome have
might lead one to conclude that the other 20–50% a 1 in 3 likelihood of having the genetic contribution
might not have a genetic etiology. However, this of only one single parent. Indeed, uniparental
would not be correct, because mendelian and poly- disomy for chromosome 21 has been detected in an
genic/multifactorial disorders show chromosomal embryonic abortus.79 Lethality would occur if that
abnormalities. Indeed, these conditions far more chromosome contains an imprinted gene that
commonly explain congenital anomalies in live- needed to be inherited from the parent whose chro-
borns than do chromosomal abnormalities. Thus, mosome was excluded. (The same problems can
it would be naive to assume that mendelian and occur in trisomic rescue involving a robertsonian
polygenic/multifactorial factors do not play pivotal translocation.) Another mechanism involves a het-
roles in embryonic mortality. The difficulty is that erozygous mutant that became homozygous (actually
few of the many genes required for differentiation doubly hemizygous) through trisomic rescue.
have been identified, despite there being a myriad of Another mechanism indirectly pointing to the
potential candidate genes. As one example, Baek77 existence of mutant genes is skewed X-inactivation.
enumerated over 30 highly plausible candidate Among 48 women having two prior losses with
genes. Many genes in animals are known to produce no obvious explanation, 7 (14.6%) had skewed
lethality, as demonstrated by null mutants (knock- X-inactivation as defined by 90% of their X chro-
outs) in mice. In embryonic humans, lethality is mosomes originating from one parent (expected
recognized for certain genes (e.g., OCT4), but stud- 50%); only 1 of 67 controls (1.5%) showed skewed
ies in embryos that survive until the first trimester X-inactivation.80 The non-random distribution
are limited. could be explained by lethality for an X-linked gene
Embryos that abort because of mendelian or on the X chromosome of a single parent. Thus, male
polygenic factors may or may not show structural offspring of a woman with skewed X-inactivation
anomalies. However, a structural anomaly found might preferentially be aborted. The loss could be
in an abortion having a normal chromosomal the specific result of UPD per se, or it could reflect
32
0415421306-Ch03 3/29/07 5:27 PM Page 33
the sole contribution by the single parent having a 20. Hassold T, Merrill M, Adkins K, et al. Recombination and maternal
lethal gene (heterozygosity). Pedigrees consistent age-dependent nondisjunction: molecular studies of trisomy 16. Am
J Hum Genet 1995; 57:867–74.
with this hypothesis have been reported.81 21. Fisher JM, Harvey JF, Morton NE, et al. Trisomy 18: studies of the
parent and cell division of origin and the effect of aberrant recombi-
nation on nondisjunction. Am J Hum Genet 1995; 56:669–75.
REFERENCES 22. Bugge M, Collins A, Petersen MB, et al. Non-disjunction of chromo-
some 18. Hum Mol Genet 1998; 7:661–9.
1. Plachot M, Junca AM, Mandelbaum J, et al. Chromosome investiga- 23. Hassold TJ. Nondisjunction in the human male. Curr Top Dev Biol
tions in early life. II. Human preimplantation embryos. Hum Reprod 1998; 37:383–406.
1987; 2:29–35. 24. Henderson SA, Edwards RG. Chiasma frequency and maternal age in
2. Papadopoulos G, Templeton AA, Fisk N, et al. The frequency of chro- mammals. Nature 1968; 217:22–8.
mosome anomalies in human preimplantation embryos after in-vitro 25. Lamb NE, Yu K, Shaffer J, et al. Association between maternal age and
fertilization. Hum Reprod 1989; 4:91–8. meiotic recombination for trisomy 21. Am J Hum Genet 2005; 76:91–9.
3. Munné S, Alikani M, Tomkin G, et al. Embryo morphology, develop- 26. Hassold T, Abruzzo M, Adkins K, et al. Human aneuploidy: incidence,
mental rates, and maternal age are correlated with chromosome origin, and etiology. Environ Mol Mutagen 1996; 28:167–75.
abnormalities. Fertil Steril 1995; 64:382–91. 27. Savage AR, Petersen MB, Pettay D, et al. Elucidating the mechanisms
4. Egozcue J, Blanco J, Vidal F. Chromosome studies in human sperm of paternal non-disjunction of chromosome 21 in humans. Hum Mol
nuclei using fluorescence in-situ hybridization (FISH). Hum Reprod Genet 1998; 7:1221–7.
Update 1997; 3:441–52. 28. Verlinsky Y, Kuliev A. Practical Preimplantation Genetic Diagnosis.
5. Pellestor F, Anahory T, Hamamah S. The chromosomal analysis of London: Springer-Verlag, 2005.
human oocytes. An overview of established procedures. Hum Reprod 29. Reddy KS. Double trisomy in spontaneous abortions. Hum Genet
Update 2005; 11:15–32. 1997; 101:339–45.
6. Plachot M. Genetics in human oocytes. In: Boutaleb Y, ed. New 30. Diego-Alvarez D, Ramos-Corrales C, Garcia-Hoyos M, et al. Double
Concepts in Reproduction. Lancater, UK: Parthenon, 1992:367. trisomy in spontaneous miscarriages: cytogenetic and molecular
7. Martin R. Chromosomal analysis of human spermatozoa. In: Verlinsky Y, approach. Hum Reprod 2006; 21:958–66.
Kuliev A, eds. Preimplantation Genetics. New York: Plenum Press, 31. Li S, Hassed S, Mulvihill JJ, et al. Double trisomy. Am J Med Genet
1991:91–102. 2004; 124:96–8.
8. Boue J, Boue A, Lazar P. Retrospective and prospective epidemiological 32. Clouston HJ, Herbert M, Fenwick J, et al. Cytogenetic analysis of
studies of 1500 karyotyped spontaneous human abortions. human blastocysts. Prenat Diagn 2002; 22:1143–52.
Teratology 1975; 12:11–26. 33. Beatty RA. The origin of human triploidy: an integration of qualita-
9. Hassold TJ. A cytogenetic study of repeated spontaneous abortions. tive and quantitative evidence. Ann Hum Genet 1978; 41:299–314.
Am J Hum Genet 1980; 32:723–30. 34. Jauniaux E, Burton GJ. Pathophysiology of histological changes in
10. Simpson JL, Bombard AT. Chromosomal abnormalities in spontaneous early pregnancy loss. Placenta 2005; 26:114–23.
abortion: frequency, pathology and genetic counseling. In: Edmonds KB, 35. McFadden DE, Robinson WP. Phenotype of triploid embryos. J Med
ed. Spontaneous Abortion. Oxford: Blackwell, 1987: 51–76. Genet 2006; 43:609–12.
11. Sorokin Y, Johnson MP, Uhlmann WR, et al. Postmortem chorionic 36. Jacobs PA, Angell RR, Buchanan IM, et al. The origin of human
villus sampling: correlation of cytogenetic and ultrasound findings. triploids. Ann Hum Genet 1978; 42:49–57.
Am J Med Genet 1991; 39:314–16. 37. Egozcue S, Blanco J, Vidal F, et al. Diploid sperm and the origin of
12. Strom CM, Ginsberg N, Applebaum M, et al. Analyses of 95 first- triploidy. Hum Reprod 2002; 17:5–7.
trimester spontaneous abortions by chorionic villus sampling and 38. McFadden DE, Langlois S. Parental and meiotic origin of triploidy in
karyotype. J Assist Reprod Genet 1992; 9:458–61. the embryonic and fetal periods. Clin Genet 2000; 58:192–200.
13. Ellett K, Buxton EJ, Luesley DM. The effect of ethnic origin on the rate 39. Schluth C, Doray B, Girard-Lemaire F, et al. Prenatal diagnosis of a
of spontaneous late mid-trimester abortion. Ethn Dis 1992; 2:84–6. true fetal tetraploidy in direct and cultured chorionic villi. Genet
14. Schaeffer AJ, Chung J, Heretis K, et al. Comparative genomic Couns 2004; 15:429–36.
hybridization-array analysis enhances the detection of aneuploidies 40. Rosenbusch B, Schneider M. A brief look at the origin of tetraploidy.
and submicroscopic imbalances in spontaneous miscarriages. Am J Cytogenet Genome Res 2004; 107:128–31.
Hum Genet 2004; 74:1168–74. 41. Baumer A, Dres D, Basaran S, et al. Parental origin of the two addi-
15. Simpson JL, Mills JL, Holmes LB, et al. Low fetal loss rates after ultra- tional haploid sets of chromosomes in an embryo with tetraploidy.
sound-proved viability in early pregnancy. JAMA 1987; 258:2555–7. Cytogenet Genome Res 2003; 101:5–7.
16. Kuleshov NP. Chromosome anomalies of infants dying during the 42. Singh RP, Carr DH. The anatomy and histology of XO human
perinatal period and premature newborn. Hum Genet 1976; embryos and fetuses. Anat Rec 1966; 155:369–83.
31:151–60. 43. Jirasek JE. Principles of reproductive embryology. In: Simpson JL, ed.
17. Hefler LA, Hersh DR, Moore PJ, et al. Clinical value of postnatal Disorders of Sex Differentiation: Etiology and Clinical Delineation.
autopsy and genetics consultation in fetal death. Am J Med Genet San Diego: Academic Press, 1976: 51–110.
2001; 104:165–8. 44. Simpson JL. Pregnancies in women with chromosomal abnormalities.
18. Warburton D, Byrne J, Canik N. Chromosome Anomalies and Prenatal In: Schulman JD, Simpson JL, eds. Genetic Diseases in Pregnancy.
Development: An Atlas. New York: Oxford University Press, 1991. New York: Academic Press, 1981: 439–71.
19. Kalousek DK. Pathology of abortion: chromosomal and genetic 45. Chandley AC. The origin of chromosomal aberrations in man and
correlations. In: Kraus FT, Damjanov IKN, eds. Pathology of their potential for survival and reproduction in the adult human
Reproductive Failure. Baltimore: Williams & Wilkins, 1991:228. population. Ann Genet 1981; 24:5–11.
33
0415421306-Ch03 3/29/07 5:27 PM Page 34
46. Warburton D, Kline J, Stein Z, et al. Does the karyotype of a sponta- 64. De Braekeleer M, Dao TN. Cytogenetic studies in couples experienc-
neous abortion predict the karyotype of a subsequent abortion? ing repeated pregnancy losses. Hum Reprod 1990; 5:519–28.
Evidence from 273 women with two karyotyped spontaneous abor- 65. Simpson JL, Elias S, Martin AO. Parental chromosomal rearrange-
tions. Am J Hum Genet 1987; 41:465–83. ments associated with repetitive spontaneous abortions. Fertil Steril
47. Hassold TJ, Matsuyama A, Newlands IM, et al. A cytogenetic study of 1981; 36:584–90.
spontaneous abortions in Hawaii. Ann Hum Genet 1978; 41:443–54. 66. Goddijn M, Joosten JH, Knegt AC, et al. Clinical relevance of diagnos-
48. Munné S, Sandalinas M, Magli C, et al. Increased rate of aneuploid ing structural chromosome abnormalities in couples with repeated
embryos in young women with previous aneuploid conceptions. miscarriage. Hum Reprod 2004; 19:1013–17.
Prenat Diagn 2004; 24:638–43. 67. Boue A, Gallano P. A collaborative study of the segregation of inher-
49. Bianco K, Caughey AB, Shaffer BL, et al. History of miscarriage and ited chromosome structural rearrangements in 1,356 prenatal diag-
increased incidence of fetal aneuploidy in subsequent pregnancy. noses. Prenat Diagn 1984; 4:45–67.
Obstet Gynecol 2006; 107:1098–102. 68. Daniel A, Hook EB, Wulf G. Risks of unbalanced progeny at amnio-
50. Drugan A, Koppitch FC, III, Williams JC III, et al. Prenatal genetic centesis to carriers of chromosome rearrangements: data from
diagnosis following recurrent early pregnancy loss. Obstet Gynecol European, USA, and Canadian laboratories. Am J Hum Genet 1988;
1990; 75:381–4. 43:918 (Abstr 230).
51. Rubio C, Simon C, Vidal F, et al. Chromosomal abnormalities and 69. Sugiura-Ogasawara M, Ozaki Y, Sato T, et al. Poor prognosis of
embryo development in recurrent miscarriage couples. Hum Reprod recurrent aborters with either maternal or paternal reciprocal translo-
2003; 18:182–8. cations. Fertil Steril 2004; 81:367–73.
52. Alberman ED. The abortus as a predictor of future trisomy 21. In: 70. Carp H, Feldman B, Oelsner G, Schiff E. Parental karyotype and sub-
Cruz DI, Gerald PS, eds. Trisomy 21 (Down syndrome). Baltimore: sequent live births in recurrent miscarriage. Fertil Steril 2004;
Raven Press, 1981:69. 81:1296–301.
53. Snijders RJ, Nicolaides KH. Ultrasound Markers for Fetal 71. Carp H, Guetta E, Dorf H, et al. Embryonic karyotype in recurrent
Chromosomal Defects. New York: Parthenon, 1996. miscarriage with parental karyotypic aberrations. Fertil Steril 2006;
54. Stern C, Chamley L, Hale L, et al. Antibodies to β2 glycoprotein I are 85:446–50.
associated with in vitro fertilization implantation failure as well as 72. Stephenson MD, Sierra S. Reproductive outcomes in recurrent preg-
recurrent miscarriage: results of a prevalence study. Fertil Steril 1998; nancy loss associated with a parental carrier of a structural chromo-
70:938–44. some rearrangement. Hum Reprod 2006; 21:1076–82.
55. Stephenson MD, Awartani KA, Robinson WP. Cytogenetic analysis of 73. Mackie OC, Scriven PN. Meiotic outcomes in reciprocal translocation
miscarriages from couples with recurrent miscarriage: a case–control carriers ascertained in 3-day human embryos. Eur J Hum Genet 2002;
study. Hum Reprod 2002; 17:446–51. 10:801–6.
56. Carp H, Toder V, Aviram A, et al. Karyotype of the abortus in recur- 74. Sutherland GR, Gardiner AJ, Carter RF. Familial pericentric inversion
rent miscarriage. Fertil Steril 2001; 75:678–82. of chromosome 19, inv(19)(p13q13) with a note on genetic counsel-
57. Ogasawara M, Aoki K, Okada S, et al. Embryonic karyotype of abor- ing of pericentric inversion carriers. Clin Genet 1976; 10:54–9.
tuses in relation to the number of previous miscarriages. Fertil Steril 75. Anton E, Vidal F, Egozcue J, et al. Genetic reproductive risk in inver-
2000; 73:300–4. sion carriers. Fertil Steril 2006; 85:661–6.
58. Benkhalifa M, Kasakyan S, Clement P, et al. Array comparative 76. Pettenati MJ, Rao PN, Phelan MC, et al. Paracentric inversions in
genomic hybridization profiling of first-trimester spontaneous abor- humans: a review of 446 paracentric inversions with presentation of
tions that fail to grow in vitro. Prenat Diagn 2005; 25:894–900. 120 new cases. Am J Med Genet 1995; 55:171–87.
59. Simpson JL. Choosing the best prenatal screening protocol. N Engl J 77. Baek KH. Aberrant gene expression associated with recurrent preg-
Med 2005; 353:2068–70. nancy loss. Mol Hum Reprod 2004; 10:291–7.
60. Munne S, Fischer J, Warner A, et al. Preimplantation genetic diagno- 78. Philipp T, Kalousek DK. Generalized abnormal embryonic develop-
sis significantly reduces pregnancy loss in infertile couples: a multi- ment in missed abortion: embryoscopic and cytogenetic findings. Am
center study. Fertil Steril 2006; 85:326–32. J Med Genet 2002; 111:43–7.
61. Munné S, Escudero T, Colls P, et al. Predictability of preimplantation 79. Henderson DJ, Sherman LS, Loughna SC, et al. Early embryonic fail-
genetic diagnosis of aneuploidy and translocations on prospective ure associated with uniparental disomy for human chromosome 21.
attempts. Reprod Biomed Online 2004; 9:645–51. Hum Mol Genet 1994; 3:1373–6.
62. Fortuny A, Carrio A, Soler A, et al. Detection of balanced chromosome 80. Lanasa MC, Hogge WA, Kubik C, et al. Highly skewed X-chromosome
rearrangements in 445 couples with repeated abortion and cytoge- inactivation is associated with idiopathic recurrent spontaneous
netic prenatal testing in carriers. Fertil Steril 1998; 49:774–779 abortion. Am J Hum Genet 1999; 65:252–4.
63. Simpson JL, Meyers CM, Martin AO, et al. Translocations are infre- 81. Pegoraro E, Whitaker J, Mowery-Rushton P, et al. Familial skewed X
quent among couples having repeated spontaneous abortions but no inactivation: a molecular trait associated with high spontaneous-
other abnormal pregnancies. Fertil Steril 1989; 51:811–14. abortion rate maps to Xq28. Am J Hum Genet 1997; 61:160–70.
34
0415421306-Ch03a 3/29/07 5:29 PM Page 35
DEBATE
3a. Should fetal karyotyping be performed in RPL?

– For
Howard JA Carp
The objective of any investigation in clinical medicine genomic hybridization (CGH) or multiplex fluore-
is to reach a diagnosis, inform the patient of the scence in situ hydridization (M-FISH),12 may over-
subsequent prognosis, and offer effective treatment come these problems and allow additional genetic
(if available), based on these three parameters. diagnoses to be made, such as uniparental disomy
Recurrent pregnancy loss is no exception to this or skewed X-chromosome inactivation.13 It is not
basic plan of management. Sixty percent of spo- possible to reach an accurate diagnosis of cause or
radic miscarriages have been reported to have a recurrent miscarriage unless the chromosomal
chromosomal aberration as the underlying cause.1–3 status of the fetus is determined.
Although the incidence of chromosomal aberra- Karyotyping of the abortus allows the patient to
tions is lower in recurrent than in spontaneous be given prognostic information regarding subse-
miscarriages, incidences of 25–60% have been quent pregnancy outcomes. Warburton et al14 sum-
reported to be caused by embryonic chromosomal marized 273 women who had had abortuses
aberrations.4–8 In fact, embryonic chromosomal karyotyped. They concluded that after a previous
aberrations have been found in the presence of trisomic miscarriage, the prognosis is favorable.
other presumptive causes of pregnancy loss. In the Two subsequent studies5,6 have examined the out-
antiphospholipid syndrome (APS), two small series, come of the subsequent pregnancy according to the
both from Japan, have reported incidences of 20%9 karyotype of the miscarriage. In the series of
and 40%.5 Carp et al10 have reported four chromoso- Ogasawara et al5 (Figure 3a.1), there was a statisti-
mal aberrations in patients with hereditary throm- cally significant trend for a patient with an aneu-
bophilias. In the series of Carp et al,6 trisomies were ploidic abortion to have a better prognosis. The
the most common form of aberration, occurring in same trend was apparent in the series of Carp et al6
66.7% of chromosomally aberrant embryos, with tri- However, the smaller numbers in the series of Carp
somies 21, 16, and 18 being the most common. et al6 precluded the figures from reaching statistical
Monosomy X and triploidies were also seen. Since significance (95% confidence interval (CI)
the publication of that series, chromosomes 1 and 2 0.85–11.74). In women with three miscarriages and
trisomies have also been seen. Chromosome 1 and 2 an aneuploidic miscarriage, reassurance of a good
would not normally be sought in a pregestational prognosis may be sufficient, and may save the
screening (PGS) program. patient more extensive investigations and treatment
The standard banding technique for karyotyping of dubious value. This may not be the case in
can only assess structural and numerical rearrange- euploidic abortions. The better prognosis after an
ments, and is liable to fail due to contamination, aneuploid abortion is entirely logical, as fetal aneu-
culture failure, or overgrowth of maternal cells. ploidy is due to a fetal cause. Hence, there is a
Other more sophisticated tests such as comparative greater chance that in a subsequent pregnancy, with
35
0415421306-Ch03a 3/29/07 5:29 PM Page 36
Fetal karyotyping also directs treatment. If the

Carp et al6 fetus is karyotypically abnormal, a normal embryo
Ogasawara et al5 can be provided to the mother by PGS. However, if
67% the fetus is normal, the maternal environment
70%
62% needs treatment. As all previous maternal treatment
60% modalities have ignored the fetal karyotype, the
abnormal fetal karyotype has confounded the
50% results. Hence, there are debates in this book as to
37.8% 38% whether hormone supplementation, thrombopro-
40%
phylaxis, immunopotentiation, etc. should be used.
30% Additionally, the place of surgery for uterine anom-
alies and treatment of APS are equally debateable.
20% If these therapies had only been used in the patient
aborting chromosomally normal embryos, the
10%
above-mentioned therapies might be found to be
efficacious, and their role might need no debate.
0%
Euploidy Aneuploidy Additionally, due to the possibility of fetal chromo-
somal aberrations, PGS has come into use, in order
Figure 3a.1 Outcome of subsequent pregnancy according to
to provide the patient with a chromosomally
fetal karyotype. Odds ratio for a live birth after aneuploidic normal embryo. When this technique is used on
abortion: 3.28 (95% confidence interval 0.94–11.9);6 unselected patients with recurrent miscarriage, it
2.62 (1.21–5.67).5
has also been reported to be of little value.15 It is our
opinion that PGS has a role to play,16 but that role is
limited to patients with repeat aneuploidies, or
those in whom there is a greater likelihood of fetal
a new embryo, the next fetus will be euploid. chromosomal aberrations, such as the older patient.
However, a euploid abortus indicates that the cause of We have recently seen a 42-year-old patient with
miscarriage is more likely to be maternal; hence, the three consecutive aneuploid abortions. Her details
same problem is liable to return in a subsequent preg- are given in Table 3a.1. If fetal karyotyping had not
nancy, thus worsening the prognosis. Any prognosis is been performed, she would have been recom-
empirical if the karyotype of the abortus is unknown. mended paternal leukocyte immunization or
Fetal karyotyping has been assessed in a subse- immunoglobulin in view of her advanced age.
quent abortion in the study by Sullivan et al.8 Of However, as the karyotype was available, immun-
30 patients with an aneuploid abortion, only otherapy would probably not have increased her
3 (10%) had a subsequent aneuploid abortion. In the chance of a live birth. In fact, if she had been placed
present author’s series (unpublished), 43 abortuses in an immunotherapy trial, the trial might have
were found to be aneuploid, and a subsequent abor- concluded that immunotherapy was futile.
tion was karyotyped. Only 8 of the 43 abortuses In view of the advanced maternal age, which makes
were aneuploid (19%). Hence, only approximately chromosomal aberrations more likely, and as all
15% of aneuploid abortions will be followed by a 23 chromosomes cannot be assessed at PGS, she
subsequent aneuploid abortion. Hence, 85% of was recommended ovum donation. These poor-
patients with an aneuploid abortion can be assured prognosis patients are outside the normal guide-
that the prognosis is good, and that the aneuploid lines for treatment, and do not have the good
abortion may be a chance ocurrence. However, prognosis of 70–75% as other patients. However,
the other 15% may have a recurring cause of fetal they still require appropriate management, which
aneuploidy, and can be offered PGS. rests on accurate diagnosis.
36
0415421306-Ch03a 3/29/07 5:29 PM Page 37
DEBATE: SHOULD FETAL KARYOTYPING BE PERFORMED IN RPL? – FOR
parental chromosomal disorder (5 trisomies and

Table 3a.1 Patient with recurrent aneuploidy
2 monosomy X).
Age: 42 Parental karyotyping does not provide a progno-
Obstetric history: sis. Three papers have looked at the subsequent live
1st pregnancy: artificial abortion
2nd pregnancy: normal delivery, 3400 g birth rate in patients with recurrent miscarriage
3rd pregnancy: normal delivery, 3500 g and parental chromosomal rearrangements. These
4th pregnancy: blighted ovum are summarized in Table 3a.2. Ogasawara et al22
5th pregnancy: left ectopic pregnancy, treated conservatively with
methotrexate have reported a 32% subsequent live birth rate,
6th pregnancy: blighted ovum whereas Carp et al17 have reported a 44% rate and
7th pregnancy: terminated artificially at 20 weeks for 47,XXX fetal Goddijn et al23 a 70% rate. Taken together, the live
karyotype
8th pregnancy: biochemical pregnancy birth rate was 47.5% for patients with a mean of 3.7
9th pregnancy: missed abortion 10 weeks, after treatment with aspirin previous miscarriages. This live birth rate of 47.5%
and progesterone supplements; fetal karyotyping showed a 48,XX is the expected rate for patients with 3.7 abortions,
karyotype with both trisomies 14 and 15
10th pregnancy: missed abortion at 10 weeks; fetal karyotyping according to numerous series in the literature.24,25
showed trisomy 22 Parental karyotyping, in addition to not provid-
Previous investigations: showed no maternal cause for miscarriage ing a diagnosis and not providing a prognosis, does
Parental karyotype: 46,XX/ 46,XY not direct us towards treatment. Munné et al26 have
Treatment advised: ovum donation in view of advanced maternal age, reported 47 PGS cycles in 35 patients with chromo-
increasing chance of subsequent aneuploid abortion, impossibility
for screening all 23 chromosomes at PGS, and low possibility of somal rearrangements; 16 subsequent pregnancies
conceiving at IVF resulted in 14 live births. Simopoulou et al27 have
reported 11 PGS cycles in 8 patients with chromo-
somal rearrangements; euploid embryos were
replaced in all 11 cycles. There were four subsequent
There is no substitute for karyotyping the abortus – pregnancies, with three live births and one ‘bio-
two techniques have been attempted: karyotyping chemical’ pregnancy. Although these results look
of the parents and PGS as a diagnostic procedure. encouraging, both series only include six patients
Chromosomal aberrations are often suspected to with recurrent miscarriages. Eleven embryos were
have a recurring basis due to either a structural replaced, resulting in two pregnancies and one live
anomaly, such as a reciprocal or robertsonian birth. It is impossible to draw conclusions from one
translocation, or mosaicism for numerical aberra- live birth in two pregnancies.
tions. These conditions have been found in 10.8% PGS is also problematic as a diagnostic and ther-
of recurrently aborting women in the present apeutic procedure when it is unknown whether
author’s series,17 but the usually quoted prevalence the patient loses chromosomally normal or abnor-
is 3–5%.18–20 If a parental chromosomal aberration mal embryos. As it is impossible to screen for all
is found, this is usually assumed to be the cause of 23 chromosomes, an embryo can never be said to be
recurrent miscarriage. However, parental karyotyping
does not provide a diagnosis or prognosis, nor does
it direct treatment. Carp et al21 have karyotyped
Table 3a.2 Subsequent live birth rate with parental
39 abortuses from recurrently miscarrying couples chromosomal aberrations
with parental karyotypic aberrations. Of the 39,
17 (26%) were euploid. Another 10 (26%) had the Ogasawara Goddijn Carp
et al22 et al23 et al17
same balanced translocation as the parent. Hence,
69% were chromosomally normal. Only 5 (13%) Pregnancies 47 42 75
abortuses had unbalanced translocations, whereas Live births 15 30 33
Proportion 32% 70% 44%
7 (18%) of the abortuses had subsequent abortuses
Mean no. of abortions 2.9 3.9 4.23
with numerical aberrations unrelated to the
37
0415421306-Ch03a 3/29/07 5:29 PM Page 38
normal on PGS – it can only be stated that the most 12. Uhrig S, Schuffenhauer S, Fauth C, et al. Multiplex-FISH for pre
and postnatal diagnostic applications. Am J Hum Genet 1999;
common chromosomal aberrations are absent. The 65:448–62.
less commonly ocurring aberrations such as tri- 13. Lanasa MC, Hogge WA, Kubik C, et al. Highly skewed X chromosome
somy 2 are not usually screened for. Hence, PGS inactivation is associated with idiopathic recurrent spontaneous
abortion. Am J Hum Genet 1999; 65:252–4.
cannot provide a diagnosis, or select patients for 14. Warburton D, Kline J, Stein Z, et al. Does the karyotype of a sponta-
treatment of maternal factors. neous abortion predict the karyotype of a subsequent abortion?
Therefore, karyotyping of the abortus seems to be Evidence from 273 women with two karyotyped spontaneous abor-
tions. Am J Hum Genet 1987; 41:465–83.
the most important single investigation for the assess- 15. Platteau P, Staessen C, Michiels A, et al. Preimplantation genetic diag-
ment of recurrent miscarriage, as recommended by nosis for aneuploidy screening in patients with unexplained recurrent
the Royal College of Obstetricians and Gynae- miscarriages. Fertil Steril 2005; 83:393–7
16. Carp HJA, Dirnfeld M, Dor J, et al. ART in recurrent miscarriage: pre-
cologists in their 199828 guideline revised in 2003.29 implantation genetic diagnosis / screening or surrogacy? Hum Reprod
We find this test to be invaluable, and its absence in 2004; 19:1502–5.
cases where the culture fails or is contaminated leads 17. Carp HJA, Feldman B, Oelsner G, et al. Parental karyotype and subse-
quent live births in recurrent miscarriage. Fertil Steril 2004;
to an incomplete diagnosis, inaccurate prognosis, 81:1296–1301.
and possibly wrong advice as to management. 18. De Braekeleer, M. Dao TN. Cytogenetic studies in couples experienc-
ing repeated pregnancy losses. Hum Reprod 1990; 5:519–28.
19. Portnoi MF, Joye N, van den Akker J, et al. Karyotypes of 1142 couples
REFERENCES
with recurrent abortion. Obstet Gynecol 1988; 72:31–4.
20. Fortuny A, Carrio A, Soler A, et al. Detection of balanced chromo-
1. Boue J, Boue A, Lazar P. The epidemiology of human spontaneous some rearrangements in 445 couples with repeated abortion
abortions with chromosomal anomalies. In: Blandau RJ, Ed. Aging and cytogenetic prenatal testing in carriers. Fertil Steril 1988;
Gametes. Basel: Karger, 1975:330. 49:774–9.
2. Stein Z. Early fetal loss. Birth Defects 1981; 17:95–9. 21. Carp HJA, Guetta E, Dorf H, et al. Embryonic karyotype in recurrent
3. Sanchez JM, Franzi L, Collia F, et al. Cytogenetic study of spontaneous miscarriage with parental karyotypic aberrations. Fertil Steril 2006;
abortions by transabdominal villus sampling and direct analysis of 85:446–450.
villi. Prenat Diagn 1999; 19:601–3. 22. Sugiura-Ogasawara M, Ozaki Y, Sato T, et al. Poor prognosis of recur-
4. Stern JJ, Dorfman AD, Gutierez-Najar MD, et al. Frequency of abnor- rent aborters with either maternal or paternal reciprocal transloca-
mal karyotype among abortuses from women with and without a his- tions. Fertil Steril 2004; 81:367–73.
tory of recurrent spontaneous abortion. Fertil Steril 1996; 65:250–3. 23. Goddijn M, Joosten JH, Knegt AC, et al. Clinical relevance of diagnos-
5. Ogasawara M, Aoki K, Okada S, et al. Embryonic karyotype of abor- ing structural chromosome abnormalities in couples with repeated
tuses in relation to the number of previous miscarriages. Fertil Steril miscarriage. Hum Reprod 2004; 19:1013–17.
2000; 73:300–4. 24. Knudsen UB, Hansen V, Juul S, et al. Prognosis of a new pregnancy
6. Carp HJA, Toder V, Orgad S, et al. Karyotype of the abortus in recur- following previous spontaneous abortions. Eur J Obstet Gynecol
rent miscarriage. Fertil Steril 2001; 5:678–82. Reprod Biol 1991; 39:31–6.
7. Stephenson MD, Awartani KA, Robinson WP. Cytogenetic analysis of 25. Carp HJA. Investigation and treatment for recurrent pregnancy loss.
miscarriages from couples with recurrent miscarriage: a case–control In: Rainsbury P, Vinniker D, eds. A Practical Guide to Reproductive
study. Hum Reprod 2002; 17:446–51. Medicine. Carnforth, UK: Parthenon, 1997:337–62.
8. Sullivan AE, Silver RM, LaCoursiere DY, et al. Recurrent fetal aneu- 26. Munné S, Sandalinas M, Escudero T, et al. Outcome of preimplanta-
ploidy and recurrent miscarriage. Obstet Gynecol 2004; 104:784–8. tion genetic diagnosis of translocations. Fertil Steril 2000;
9. Takakuwa K, Asano K, Arakawa M, et al. Chromosome analysis of 73:1209–18.
aborted conceptuses of recurrent aborters positive for anticardiolipin 27. Simopoulou M, Harper JC, Fragouli E, et al. Preimplantation genetic
antibody. Fertil Steril 1997; 68:54–8. diagnosis of chromosome abnormalities: implications from the out-
10. Carp HJA, Dolitzky M, Inbal A. Thromboprophylaxis improves the come for couples with chromosomal rearrangements. Prenat Diagn
live birth rate in women with consecutive recurrent miscarriages and 2003; 23:652–62.
hereditary thrombophilia. J Thromb Hemost 2003; 1:433–8. 28. Royal College of Obstetricians and Gynaecologists, Guideline No. 17.
11. Daniely M, Aviram-Goldring A, Barkai G, et al. Detection of chromoso- The Management of Recurrent Miscarriage. London: RCOG, 1998.
mal abberation in fetuses arising from recurrent spontaneous abortion 29. Royal College of Obstetricians and Gynaecologists, Guideline No. 17.
by comparative genomic hybridization. Hum Reprod 1998; 13:805–9. The Management of Recurrent Miscarriage. London: RCOG, 2003.
38
0415421306-Ch03b 3/29/07 5:30 PM Page 39
DEBATE
3b. Should fetal karyotyping be performed

in RPL ?
– Against
Zvi Borochowitz
The term ‘spontaneous abortion’ refers to a preg-

INTRODUCTION nancy that ends spontaneously before the fetus has
reached a viable gestational age. The World Health
The first report of a chromosomal abnormality in Organization (WHO) defines it as expulsion or
aborted material was of a triploidy in spontaneous extraction of an embryo or fetus weighing 500 g or
abortion, four decades ago by Penrose and less from its mother. This typically corresponds to a
Delhanty.1 It took several years before cytogenetic gestational age of 20–22 weeks. In this chapter, the
analysis of miscarriage became an option in labora- term ‘miscarriage’ will be used to mean repeated,
tories, due to the difficulties of culturing fetal tissue. recurrent, multiple, or habitual abortion. The preva-
The development of techniques that allowed chori- lence of miscarriage has been estimated as 10–15% of
onic villi to be used for ‘long-term cultures’, and later all clinically recognized pregnancies. One of the most
direct preparation of metaphases from villi, revolu- remarkable, and as yet unexplained, aspects of the
tionized the cytogenetic analysis of products of con- first trimester of pregnancy is the fact that the major-
ception. There has since been a debate as to whether ity (90%) of karyotypically abnormal pregnancies
it is either clinically justifiable or psychologically miscarry in the first trimester, and the majority
essential to determine the cause of pregnancy loss for (93%) of karyotypically normal pregnancies con-
the task of counselling for a further pregnancy. The tinue.3 There are various risk factors for miscar-
crucial role of chromosomal imbalance in abnormal riage, including, among others, advanced maternal
early human development is well known. Most of the age, alcohol use, anesthetic gas use (i.e., nitrous
chromosomal abnormalities result in disordered oxide), heavy caffeine use, chronic maternal dis-
development incompatible with prolonged intrauter- eases, cigarette smoking, cocaine use, conception
ine survival and live birth. The mechanism by which 3–6 months after delivery, intrauterine device use,
a chromosomal abnormality could lead to regression maternal infections, medications, multiple previous
of the conceptus is unclear. Approximately 50–60% elective abortions, toxins (lead, arsenic, ethylene
of first-trimester spontaneous abortions have kary- glycol, etc.), and uterine abnormalities. Most chro-
otypic abnormalities – mainly numerical chromoso- mosomal abnormalities that result in spontaneous
mal changes such as autosomal trisomy, monosomy X, abortion are random events and are more likely to
and polyploidy. This conclusion is based on the results be associated with recurrent spontaneous abortion,
of cytogenetic studies conducted in laboratories but are uncommon even where one of the parents is
throughout the world.2 a carrier (4–6% of recurrent miscarriages4,5).
39
0415421306-Ch03b 3/29/07 5:30 PM Page 40
60–70% of women with unexplained RPL will have

CYTOGENETIC ABNORMALITIES a successful next pregnancy.8
RPL may be due to an abnormal embryo that is
Cytogenetic evaluation of sporadic spontaneous incompatible with life. As the number of miscar-
abortions has shown that 50–60% are chromoso- riages increases, the prevalence of chromosomal
mally abnormal. This means that about 5–10.5% of abnormality decreases, and the chance of recurring
all pregnancies result in sporadic abortions caused maternal cause increases.9 It is thus assumed that in
by chromosomal abnormalities. Pregnancy loss of recurrent miscarriage, there is more likely to be a
chromosomal origin, which is the scope of this recurring cause leading to RPL. Hence, RPL occurs
chapter, is uncommon after 15 weeks’ gestation; in up to 1% of the population – which is more than
therefore, this chapter is concerned mainly with would be expected if recurrent miscarriage occurs
first-trimester miscarriages. It is well documented as a result of repeated aneuploidy. Many couples
that fetal de novo chromosomal abnormalities are a will view these data with optimism.
major cause. These detection rates of chromosomal
abnormalities have remained constant over time,
and are independent of the culture method used or LABORATORY TECHNIQUE
the culture success rate, which is now reported to be
90%.2,4,6 The recurrence risk of another miscarriage Results of conventional cytogenetic analysis of spon-
after an aneuploidic miscarriage is not elevated, or taneous abortions depend strongly on tissue culture
is only slightly so (16%), when compared with the and are associated with a significant culture failure
initial risk of all women (10–15%). Thus, routine rate, which varies from 5% to 42% in different labo-
karyotyping of fetal material in miscarriages is not ratories. The banding technique for karyotyping can
thought to be worthwhile and is considered unnec- only assess structural and numerical rearrangements,
essary.4 Half of the structural abnormalities may be and is liable to fail as a result of contamination, cul-
inherited from a parent carrying a balanced chro- ture failure, or overgrowth of maternal cells. The
mosome translocation or inversion. This type of cytogenetic factors of cell death in vitro have not
chromosomal abnormality would be found by been sufficiently investigated. A possible disadvan-
parental karyotyping, which is now a routine proce- tage of the (semi)direct preparation is the discrep-
dure in most countries after two (or at most three) ancy that may occur between embryonic cells and
missed abortions. chorionic villi. Such a discrepancy might be due to
the fact that mosaicism is found only in placental
tissue (i.e., confined placental mosaicism). It is possi-
PROGNOSIS ble to assume that tissue culture failure is a marker of
particular genomic imbalances incompatible with
The presence of a cytogenetic abnormality in mis- normal cell proliferation. If this hypothesis is true,
carriages explains the reason for the loss. In most then the standard cytogenetic analysis of sponta-
couples with recurrent pregnancy loss (RPL), an neous abortions may underestimate the frequency
evaluation, including parental karyotype testing, and diversity of detected chromosomal abnormali-
will be negative. Therefore, a majority (approxi- ties. Thus, the fetal karyotype may not be represented
mately 50–75%) of couples with RPL will have no correctly by the villous karyotype. The estimated
certain diagnosis. Live birth rates of 35–85% are percentage of mosaicism is 1–2% for (semi)direct
commonly reported in couples with unexplained chromosome preparation in chorionic villus
RPL who undertake an untreated or placebo- sampling (CVS). In view of these difficulties, it has
treated subsequent pregnancy.7 Meta-analysis of been argued that other more sophisticated tests such
randomized prospective studies suggests that as comparative genomic hybridization (CGH) or
40
0415421306-Ch03b 3/29/07 5:30 PM Page 41
DEBATE: SHOULD FETAL KARYOTYPING BE PERFORMED IN RPL? – AGAINST
multiplex fluorescence in situ hybridization (M-

CONSENSUS REMARKS
FISH) may overcome these problems and allow addi-
tional genetic diagnoses to be made, such as
AMERICAN COLLEGE OF OBSTETRICIANS
uniparental disomy or skewed X-chromosome inac-
AND GYNECOLOGISTS (ACOG)7
tivation. However, it appears from these series that
there must be an alternative cause for the pregnancy
● ‘Many experts obtain a karyotype of the abortus
loss in most (71%) of patients. The results of FISH
tissue when a couple with recurrent pregnancy loss
analysis of uncultured cells from spontaneous abor-
experiences a subsequent spontaneous abortion.
tions reveal a high frequency of both intratissue and
The rationale is that if the abortus is aneuploid, the
confined placental mosaicism, and suggest that cul-
physician and patient may conclude that a
ture techniques underestimate the incidence of chro-
maternal cause of pregnancy loss is excluded. Also,
mosomal anomalies. This observation underlines the
an abnormal abortus karyotype is a legitimate
fact that early stages of human development are asso-
explanation for the loss that may provide a source
ciated with a high rate of mitotic errors, which has
of comfort to the couple. However, no published
been confirmed by data from molecular cytogenetic
evidence supports these hypotheses, and definite
analyses of cleavage-stage embryos. It is probable
recommendations for routinely obtaining abortus
that mitotic instability is an important factor in early
karyotypes cannot be made.’
prenatal selection.2,4,10,11
As current rates of chromosomal abnormalities ROYAL COLLEGE OF OBSTETRICIANS AND
are constant over time, and are independent of the GYNAECOLOGISTS (RCOG)12
current routine culture method used in most labo-
ratories, as well as the culture success rate, which is ● ‘If the karyotype of the miscarriage pregnancy is
now reported to be of 90%,6 it seems that FISH or abnormal, there is a better prognosis in the next
CGH techniques have not significantly changed the pregnancy. Cytogenetic testing is an expensive
clinical approach or the psychological benefits. The tool and may be reserved for patients who have
dubious value of FISH or CGH must be weighed undergone treatment in the index pregnancy or
against the highly sophisticated laboratory require- have been participating in a research trial. For
ments (still not routinely available in many countries them, karyotyping the products of conception
and laboratories) and the high costs involved. provides useful information for counselling and
It can be summarized that current laboratory future management.’
techniques are obstacles in making the karyotyping ● This statement is under Category C – level IV –
of abortus material a required routine test. which is based solely on expert opinion with no
Cytogenetic analysis of fetal tissue is expensive. clinical studies of good quality.
Formerly, it was thought that the histological fea-
tures of miscarriage could predict karyotype, and JOINT WORKING PARTY OF THE ROYAL COLLEGE OF
could possibly be an alternative to karyotyping. PATHOLOGISTS (RCPath) AND THE ROYAL COLLEGE
Examples of such histological features are villous OF OBSTETRICIANS AND GYNAECOLOGISTS (RCOG)13
contour, hydropic villi, trophoblastic hyperplasia,
trophoblastic lacunae, cisterns, inclusions, and ● ‘Because of the high failure rate of post-abortal
perivillous and intervillous fibrin deposits. So far, and post-stillbirth karyotyping, the Working
hydropic villi, trophoblastic lacunae, trophoblastic Party recommends that multiple samples be
hyperplasia, cisterns, and inclusions have shown collected, usually placenta and full-thickness
a significant association with triploidy in only a skin. Consideration should also be given to
very few studies. In general, histological features are collecting a specimen in utero before the
unhelpful for predicting karyotype.4 termination process begins.’
41
0415421306-Ch03b 3/29/07 5:30 PM Page 42
GYNECOLOGISTS, OBSTETRICIANS, AND FERTILITY

SPECIALISTS FROM 18 COUNTRIES PARTICIPATING IN
and selective growth of maternal cells, with
A THREE-DAY WORKSHOP HELD IN DENMARK IN
success rates varying from 5% to 42%.
200214
5. As rates of chromosomal abnormalities have
remained constant over time, and are independent
● ‘Improved techniques in cytogenetics have of the culture method used, FISH or CGH
permitted more accurate and reliable assessments techniques have not significantly changed the
of the products of conception. Given these clinical approach or the psychological benefits.
improvements in our diagnostic ability, it is even 6. The highly sophisticated laboratory requirements
more important that every effort be made to (unavailable in many laboratories), and high costs
study the products of conception in every case of should be considered.
miscarriage in therapeutic trials so that a more 7. Presently, current laboratory techniques are
valid evaluation can be made regarding the obstacles in allowing the karyotyping of the
efficacy of the experimental treatment.’ abortus to become a routine test.
● They do not recommend karyotyping of abortus 8. Summary of several major consensus papers:
material.
● ACOG: definite recommendations for
routinely obtaining abortus karyotypes
cannot be made.7
CONCLUDING REMARKS
● RCOG: cytogenetic testing is an expensive
tool and may be reserved for patients who
1. The recurrence risk of another miscarriage is not
have undergone treatment in the index
elevated, or is only slightly so (16%), when
pregnancy or have been participating in a
compared with the initial risk of all women
research trial.12
(10–15%), and thus routine karyotyping of fetal ● Joint Working Party of RCPath and RCOG:
material in miscarriages is not thought to be
because of the high failure rate of post-
worthwhile in daily practice, and is considered to
abortal karyotyping, the Working Party
be unnecessary.
recommends that consideration should also
2. More than half of abortuses have normal
be given to collecting a specimen in utero
chromosomes, while most of the abnormal
before the termination process begins.13
chromosomes are numerical abnormalities ● International workshop (Denmark, 2002): not
(86%), with trisomies of various chromosomes
as a routine procedure, but in therapeutic trials
occuring in more than two-thirds of these,
only.14
giving rise to a randomly occurring incident.
3. With so many possible causes for recurrent
Hence, one must conclude that there is no clinical
miscarriage, it would be tempting to think that
justification, nor any psychological benefit, for fetal
the prognosis for those women whose recurrent
karyotyping. This conclusion is well noted through-
miscarriages are unexplained is dire. But three-
out this in-depth current literature survey, as well in
quarters of these women will go on to have a
these consensus clinical guidelines of the more
successful pregnancy if offered nothing more,
important professional societies around the world.
and nothing less, than tender, love and care, and
reassurance through ultrasound that nothing is REFERENCES
abnormal.15
4. The current techniques of tissue culture used in 1. Penrose LS, Delhanty JD. Triploid cell cultures from a macerated
conventional cytogenetic analysis of abortal fetal foetus. Lancet 1961; i:1261–2.
2. Lebedev IN, Ostroverkhova NV, Nikitina TV, et al. Features of chromo-
material are laborious and subject to problems somal abnormalities in spontaneous abortion cell culture failures
such as external contamination, culture failure, detected by interphase FISH analysis. Eur J Hum Genet 2004; 12:513–20.
42
0415421306-Ch03b 3/29/07 5:30 PM Page 43
DEBATE: SHOULD FETAL KARYOTYPING BE PERFORMED IN RPL? – AGAINST
3. Quenby S, Vince G, Farquharson R, et al. Recurrent miscarriage: a 10. Carp H, Toder V, Aviram A, et al. Karyotype of the abortus in recur-
defect in nature’s quality control? Hum Reprod 2002; 17;1959–63. rent miscarriage. Fertil Steril 2001; 75:678–82.
4. Goddijn M, Leschot NJ. Genetic aspects of miscarriage. Baillieres Best 11. Stephenson MD, Awartani KA, Robinson WP. Cytogenetic analysis of
Pract Res Clin Obstet Gynaecol 2000; 14:855–65. miscarriages from couples with recurrent miscarriage: a case–control
5. Griebel CP, Halvorsen J, Golemon TB, et al. Management of sponta- study. Hum Reprod 2002; 17:446–51.
neous abortion. Am Fam Physician 2005; 72:1243–50. 12. Royal College of Obstetricians and Gynaecologists, Guideline No. 17.
6. Yusuf RZ, Naeem R. Cytogenetic abnormalities in products of concep- The Management of Recurrent Miscarriage. London: RCOG, 2003.
tion: a relationship revisited. Am J Reprod Immunol 2004; 52:88–96. 13. Fetal and Perinatal Pathology: Report of a Joint Working Party – 2005.
7. ACOG Practice Bulletin No. 24, Feb. 2001. Management of recurrent http://www.rcpath.org.
early pregnancy loss. Int J Gynecol Obstet 2002; 78:179–90. 14. Christiansen OB, Nybo Andersen AM, Bosch E, et al. Evidence-based
8. Jeng GT, Scott JR, Burmeister LF. A comparison of meta-analytic results investigations and treatments of recurrent pregnancy loss. Fertil Steril
using literature vs individual patient data. Paternal cell immunization 2005; 83:821–39.
for recurrent miscarriage. JAMA 1995; 274:830–6. 15. Kavalier F. Investigation of recurrent miscarriages. BMJ 2005;
9. Ogasawara M, Aoki K, Okada S, et al. Embryonic karyotype of abor- 331:121–2.
tuses in relation to the number of previous miscarriages. Fertil Steril
2000; 73:300–4.
43
0415421306-Ch03b 3/29/07 5:30 PM Page 44
0415421306-Ch03c 3/29/07 5:30 PM Page 45
DEBATE
3c. Should PGD be performed in RPL?

– For
Howard JA Carp and JG Grudzinskas
Pregnancy loss may be caused by a maternal factor 49 patients with recurrent miscarriage.7 Fifteen preg-
or a fetal factor. Until now, most investigations and nancies ensued from 31 cycles. These 15 pregnancies
treatment have concerned maternal factors causing resulted in 10 live births, 2 miscarriages and 3 bio-
pregnancy loss or recurrent pregnancy loss (RPL). chemical pregnancies. This 66% live birth rate
Preimplantation genetic diagnosis (PGD) is the (10 of 15 pregnancies) is often said to be no better
only strategy available to prevent loss due to than the natural rate of approximately 60%.
aneuploidy of the fetus. While it is difficult to gauge However, the mean number of miscarriages in the
the true incidence of aneuploid fetuses in RPL, var- series of Platteau et al7 was 4.7. In patients with 4.7
ious series in which the abortus has been analyzed miscarriages, a live birth rate of only 35–40% is
have quoted that 25–60% of recurrent miscarriages expected.8,9 Hence, a 66% live birth rate is an excel-
may be caused by embryonic chromosomal aberra- lent result for this cohort of patients, even when it is
tions.1–5 Rubio et al6 and Platteau et al7 have together considered that there was only a 24% implantation
summarized 559 and 413 embryos of RPL patients, rate (15 pregnancies of 63 embryos in 63 cycles).
respectively. Taken together, there was a 63.4% The results of the paper by Platteau et al7 indicate
prevalence of aneuploidy. In the series of Rubio the need for a larger trial, rather than arguing that
et al6 the prevalence was 395/559 embryos (70.7%), PGD is not worthwhile. If the figures from the series
compared with 97/218 (44.5%) of control embryos. of Platteau et al7 are pooled with those from Rubio
Thus, aneuploid miscarriages account for a signifi- et al6 then, there is a 76% live birth rate (29 of
cant number of pregnancy losses. Hence, the use of 38 pregancies) in women with 3.6 losses, which is
PGD may be appropriate for this significant subgroup higher than would be expected by chance. However,
of patients with RPL. it must be remembered that PGD requires assisted
Like most other modes of treatment, PGD is not a reproductive technology (ART). The implantation
panacea for all patients with RPL, and to argue the rate was only 24%, and patients with recurrent mis-
case for using PGD in all patients would show PGD to carriage do not per se have a problem conceiving.
be ineffective. Rubio et al6 published a trial of preges- Thus, the higher ongoing pregnancy rate may be a
tational screening (PGS) in 71 women with two or reflection of a reduced pregnancy rate.
more pregnancy losses (mean 2.9). There were 19 live It is our opinion that PGD is appropriate for a
births in 23 subsequent pregnancies. This 82% suc- subgroup of women with RPL. If the right patients
cess rate is higher than the 60% that would be are selected, further miscarriages can be prevented,
expected naturally. However, the series includes and the patient can be saved the heartbreak and
patients with two pregnancy losses, which makes the agony of having to terminate the pregnancy if ane-
figures difficult to interpret. Our opponents in this uploidy should be found by chorionic villus sam-
debate have published a prospective study of pling (CVS) or amniocentesis in the later stages
45
0415421306-Ch03c 3/29/07 5:30 PM Page 46
of pregnancy. We have published a guideline that parental chromosomal aberrations, as there have
gives the possible indications for using ART in been no case–control studies comparing miscar-
RPL.10 The indications included poor-prognosis riage rates between PGD and natural pregnancies.
patients, those not responding to simpler forms of There are three series describing PGD in the case of
treatment, repeated fetal aneuploidy, and the older parental chromosomal translocations. Chun et al11
patient. Judgment was reserved in cases of parental have described the results of 49 translocation
chromosomal rearrangements. Some of these con- carriers. Of 49 patients, 15 (30.6%) delivered after
ditions are explored in more detail below in order to 70 cycles. However, Chun et al11 do not state if any of
show that PGD has a definite role to play in RPL. their patients had recurrent miscarriage. Munné
et al12 have summarized 47 PGS cycles in 35 patients.
There were 16 pregnancies, with 14 live births.
REPEAT FETAL ANEUPLOIDY Simopoulou et al13 carried out 11 PGD cycles in
8 patients. There were three subsequent live births
Most fetal chromosomal aberrations are isolated and one ‘biochemical’ pregnancy. There were only
ocurrences. Hence, the patient who loses a chromo- six patients with recurrent miscarriage in both the
somally abnormal fetus has a greater chance of a series of Munné et al12 and of Simonopoulou et al13
live birth than the patient losing a euploid embryo together. Hence, at present, no conclusions can be
(see Figure 3a.1). However, there are a few patients drawn about the role of PGD in the presence
who repeatedly lose aneuploid embryos. Although of parental chromosomal aberrations. In the pres-
there is no large series assessing the incidence of ence of parental chromosome aberrations, the aber-
repeat aneuploidy in patients with recurrent mis- ration is passed on to the embryo in 18% of cases
carriage, there are two small series. Repeat fetal ane- (7 of 39 abortuses in the present authors’14 series. In
uploidy has been shown in 3 of 30 patients in the patients with parental chromosomal aberrations,
series of Sullivan et al5 and in 8 of 43 patients in that inherited in an unbalanced form by the embryo,
of Carp et al.10 If both series are pooled, this indicates PGD will prevent the subsequent abortion. One such
that approximately 15% of women with repeated patient is described as Patient No. 4 in Chapter 19.
pregnancy loss with fetal aneuploidy will have a
repeat aneuploidy. In cases of repeat aneuploidy
PGD/PGS should be used. If the repeat aneuploidy ADVANCED MATERNAL AGE
involves the same chromosome, PGD can be per-
formed, using the relevant probes. This is impor- With increasing female age, the incidence of chro-
tant, as the affected chromosome may be one that is mosomal abnormalities increases. The chromo-
not usually sought in standard five-, seven-, or even somal abnormalities seem to mainly originate from
nine-probe PGS. If different chromosomes are female meiosis. The higher incidence of aneuploidy
involved each time, it may be necessary to use PGS, compared with younger patients has been shown
screening for the affected chromosomes in past both when the abortus is studied3,5 and when
miscarriages. Even in this form of PGS, the screen- embryos obtained by in vitro fertilization (IVF)
ing will be more directed to affected chromosomes were studied.15,16 In the study by Rubio et al15 the
than in standard five-, seven-, or nine-probe PGS. incidence of chromosomal abnormalities was
63.59% in the group of women above 37 years of
age, and was significantly increased compared with
PARENTAL CHROMOSOMAL REARRANGEMENTS 33.1% in a control group of women with sex-linked
diseases. In this age group, there was a significant
A karyotypic aberration, in either partner, may lead increase (p < 0.05) in the prevalence of aberrations
to recurrent miscarriage. However, there is little for chromosomes 13, 16, 21, and 22. Although there
information on the role of PGD in RPL with was an increased prevalence of aberrations in
46
0415421306-Ch03c 3/29/07 5:30 PM Page 47
DEBATE: SHOULD PGD BE PERFORMED IN RPL? – FOR
chromosome 18 and the sex chromosomes, this dif- At that point, the rare chromosomal anomalies
ference did not reach statistical significance. Dailey which are found in RPL can be assessed. Great
et al17 used fluorescence in situ hybridization (FISH) strides have been made in this field with the intro-
studies on IVF patients and showed that non- duction of comparative genomic hybridization
disjunction seems to mainly affect the smaller chro- (CGH),21 multiplex FISH22 and isothermal whole-
mosomes such as 13, 16, 18, 21, 22, and X in women genome amplification.23 CGH has been used on
above 35 years of age. Other studies18 have found no the abortus24 and for PGS.25 However, if CGH is
anomalies regarding chromosomes 1, 9, or 12. used, whole-ploidy errors such as triploidy cannot
Hence, 7-probe PGS should exclude the common be excluded, and the time required in order to
aberrations leading to miscarriage, and, although obtain a result (up to 3 days) may preclude transfer
there are no direct figures, PGS should reduce the in the same cycle. The attendant cryopreservation
incidence of miscarriage, and the need to terminate and thawing is associated with loss of embryos.
the pregnancy if a chromosome aberration were to Hence, more work is required on the technique
be found at a later CVS or amniocentesis. before it can become standard clinical practice. At
There are few data available on the effect of PGS that point, it may be unnecessary to have a debate
in patients of advanced age. The only prospective such as this. The argument for using PGD/PGS may
study is that by Platteau et al.7 Only one pregnancy be obvious.
of five developed to a live birth after PGD in women Currently, we propose that PGS may be helpful to
aged above 35. Platteau et al7 use these figures to identify the proportion of euploid embryos in cou-
conclude that ‘The future obstetric outcome for the ples with RPL with a view to transferring only
older patients remains grim, whatever treatment is euploid embryos. Embryos that are deemed screened
done.’ We would contend that one out of five is too ‘normal’ or ‘no result’ should have follow-up FISH
small a sample to draw conclusions, as a larger analysis. This strategy should be considered in
sample is required. Other series have drawn differ- euploid couples, especially if the woman is 35 years of
ent conclusions to those of Platteau et al.7 Munné age or older, as well as couples in whom a balanced
et al19 have reported 20 infants born to 37 women translocation have been identified, particularly if the
from 44 PGS cycles. There was a 54% live birth rate aberrations have been transmitted to the embryo in
in women with a mean of 4.1 miscarriages. The an unbalanced form. The clinical history of the
study by Munné et al19 was retrospective, and the couple, together with knowledge of the karyotype of
results were compared with a calculated probability the abortus, should be used to determine whether
of a live birth according to the criteria of Brigham recourse to host surrogacy is indicated, as in the case
et al20 for the number of previous miscarriages and of the aborted euploid fetus, or PGS is incorporated
maternal age. However, their live birth rate was in the case of aneuploidy of the abortus.
above the rate expected for women with a similar
number of miscarriages and of the same age. REFERENCES
Munné et al19 concluded that ‘Our findings indicate
that PGD can be recommended to RM [repeated 1. Stern JJ, Dorfman AD, Gutierez-Najar MD, et al. Frequency of
miscarriage] patients who are 35 years and older abnormal karyotype among abortuses from women with and without a
history of recurrent spontaneous abortion. Fertil Steril 1996; 65:250–3.
and show no clear etiology of RM.’ 2. Ogasawara M, Aoki K, Okada S, et al. Embryonic karyotype of abor-
tuses in relation to the number of previous miscarriages. Fertil Steril
2000; 73:300–4.
3. Carp HJA, Toder V, Orgad S, et al. Karyotype of the abortus in recur-
CONCLUSIONS rent miscarriage. Fertil Steril 2001; 5:678–82.
4. Stephenson MD, Awartani KA, Robinson WP. Cytogenetic analysis of
Both PGD and PGS have a role in the management miscarriages from couples with recurrent miscarriage: a case-control
study. Hum Reprod 2002; 17:446–51.
of RPL. It now remains to develop a screening system 5. Sullivan AE, Silver RM, LaCoursiere DY, et al. Recurrent fetal aneu-
for the entire genome that can be used rapidly. ploidy and recurrent miscarriage. Obstet Gynecol 2004;104:784–8.
47
0415421306-Ch03c 3/29/07 5:30 PM Page 48
6. Rubio C, Simon C, Vidal F, et al. Chromosomal abnormalities and 16. Spandorter SD, Davis OK, Barmat LI, et al. Relationship between
embryo development in recurrent miscarriage couples. Hum Reprod aternal age and aneuploidy in in vitro fertilization pregnancy loss.
2003; 18:182–8. Fertil Steril 2004; 8:1265–9.
7. Platteau P, Staessen C, Michiels A, et al. Preimplantation genetic diag- 17. Dailey T, Dall B, Cohen J, et al. Association between nondisjunction
nosis for aneuploidy screening in patients with unexplained recurrent and maternal age in meiosis-II human oocytes. Am J Hum Genet
miscarriages. Fertil Steril 2005; 83:393–7. 1996; 59:176–84.
8. Knudsen UB, Hansen V, Juul S, et al. Prognosis of a new pregnancy 18. Cupisti S, Conn CM, Fragouli E, et al. Sequential FISH analysis of
following previous spontaneous abortions. Eur J Obstet Gynecol oocytes and polar bodies reveals aneuploidy mechanisms. Prenat
Reprod Biol 1991; 39:31–6. Diagn 2003; 23:663–8.
9. Carp HJA. Investigation and treatment for recurrent pregnancy loss. 19. Munné S, Chen S, Fischer J, et al. Preimplantation genetic diagnosis
In: Rainsbury P, Vinniker D, eds. A Practical Guide to Reproductive reduces pregnancy loss in women aged 35 years and older with a his-
Medicine. Carnforth, UK: Parthenon, 1997:337–62. tory of recurrent miscarriages. Fertil Steril 2005; 3 (Suppl 4):3:31–3.
10. Carp HJA, Dirnfeld M, Dor J, et al. ART in recurrent miscarriage: pre- 20. Brigham SA, Colon C, Farquharson G. A longitudinal study of preg-
implantation genetic diagnosis / screening or surrogacy? Hum Reprod nancy outcome following idiopathic recurrent miscarriage. Hum
2004; 19:1502–5. Reprod 1999;14:2868 –71.
11. Chun KL, Jin HJ, Dong MM, et al. Efficacy and clinical outcome of 21. Kallioniemi A, Kallioniemi OP, Sudar D, et al. Comparative genomic
preimplantation genetic diagnosis using FISH for couples of reciprocal hybridization for molecular cytogenetic analysis of solid tumors.
and robertsonian translocations: the Korean experience. Prenat Diagn Science 1992; 258:818–21.
2004; 24:556–61. 22. Uhrig S, Schuffenhauer S, Fauth C, et al. Multiplex-FISH for pre and
12. Munné S, Sandalinas M, Escudero T, et al. Outcome of preimplanta- postnatal diagnostic applications. Am J Hum Genet 1999; 65:448–62.
tion genetic diagnosis of translocations. Fertil Steril 2000; 73:1209–18. 23. Handyside AH, Robinson MD, Simpson RJ, et al. Isothermal whole
13. Simopoulou M, Harper JC, Fragouli E, et al. Preimplantation genetic genome amplification from single and small numbers of cells: a new
diagnosis of chromosome abnormalities: implications from the out- era for preimplantation genetic diagnosis of inherited disease. Mol
come for couples with chromosomal rearrangements. Prenat Diagn Hum Reprod 2004; 10:767–72.
2003; 23:652–62. 24. Daniely M, Aviram-Goldring A, Barkai G, et al. Detection of chromo-
14. Carp HJA, Guetta E, Dorf H, et al. Embryonic karyotype in recurrent somal abberation in fetuses arising from recurrent spontaneous
miscarriage with parental karyotypic aberrations Fertil Steril 2006; abortion by comparative genomic hybridization. Hum Reprod 1998;
85:446–50. 13:805–9.
15. Rubio C, Rodrigo L, Perez-Cano I, et al. FISH screening of aneuploi- 25. Wilton L, Williamson R, McBain J, et al. Birth of a healthy infant after
dies in preimplantation embryos to improve IVF outcome. RBM preimplantation confirmation of euploidy by comparative genomic
Online 2005; 11:497–506. hybridization. N Engl J Med 2001; 345:1537–41.
48
0415421306-Ch03d 3/29/07 5:31 PM Page 49
DEBATE
3d. Should PGD be performed in RPL?

– Against
Patricio Donoso, Andre Van Steirteghem, and Paul Devroey
INTRODUCTION difference in the chromosomal abnormality rate of

the products of conception from women with or
The management of recurrent pregnancy loss without recurrent miscarriage (57%). These dis-
(RPL) represents an important challenge for repro- crepancies have been attributed to different inclu-
ductive medicine specialists, as we are confronted sion criteria, since an inverse correlation between
with a distressed couple in need not only of medical the number of abortions and the frequency of
care but also of psychological support, especially chromosomal abnormalities has been described.7
when no etiology is found after a complete workup To date, therapeutic alternatives for unexplained
has been performed (50%).1 recurrent miscarriage remain empirical and with
Preimplantation genetic diagnosis (PGD) has unproven efficiency. Furthermore, a live birth rate
been proposed as a treatment for couples suffering of 70% can be achieved after three consecutive mis-
from recurrent miscarriage in two situations: unex- carriages merely by performing supportive care.8
plained recurrent miscarriage and in the presence of In this study, only age (over 40 years) had an
parental chromosome structural rearrangements. adverse impact on the pregnancy outcome. Similar
A separate analysis is presented for each group. results have been observed in other trials without
further treatment (68% live birth rate).5
PGD for aneuploidy screening (PGD-AS) has
UNEXPLAINED RECURRENT MISCARRIAGE been proposed as a therapeutic tool for the manage-
ment of unexplained recurrent miscarriage, since
Recurrent miscarriage defined as three or more an increased incidence of numerical chromosomal
consecutive pregnancy losses occurs in 0.5–3% of abnormalities (50–60%) has been reported in
women.2 embryos derived from these patients.9–13 The
Although a high rate of chromosomal abnor- hypothesis is that an enhanced embryo selection
malities (50–70%) has been demonstrated in spo- provided by PGD-AS may improve the chance of a
radic miscarriage,3,4 conflicting results are available live birth.
for unexplained recurrent miscarriage. Carp et al5 Only one randomized trial has included couples
evaluated the karyotypes of 167 abortuses in with recurrent miscarriage (n=11, PGD-AS group;
women with at least three consecutive miscarriages, n=8, control group).11 Even though there was a ten-
finding an abnormal result in only 29% of the dency towards a better outcome after PGD-AS
abortuses. In agreement with these results, a (7/11 vs 3/8 pregnancies per embryo transfer), the
case–control study6 showed an aneuploidy rate of limited number of patients does not allow definitive
only 25% in women with two or more consecutive conclusions to be drawn. Moreover, miscarriage
miscarriages. On the other hand, Stern et al3 found no rates were not reported, although they represent the
49
0415421306-Ch03d 3/29/07 5:31 PM Page 50
main outcome to be evaluated in these couples. significant reduction in the miscarriage rate in
In addition, this study was excluded from the meta- women over 35 years old after the performance of
analysis performed by the Cochrane database on PGD-AS compared with the expected probability
PGD-AS,13 since it contained insufficient data to of miscarriage (12% PGD-AS group vs 45%
assess the methodological quality. The largest expected). Table 3d.1 summarizes the available
observational trial (n=241 cycles) reported similar evidence from observational studies on the out-
miscarriage rates after PGD-AS compared with come of these patients after undergoing PGD-AS.
patients undergoing PGD for sex-linked diseases Although it has been suggested that PGD-AS
(12.3% PGD-AS vs 8.3% control) in couples with should be offered only to couples with documented
two or more miscarriages.14 However, patients in aneuploid miscarriages, two studies5,7 have shown
the control group did not have recurrent miscar- that these patients have a better prognosis than
riage, thus making it an inadequate group to com- those with euploid miscarriages. Therefore, results
pare the pregnancy outcome. A prospective study obtained from observational studies must be ana-
performed at our center including 49 patients with lyzed with caution, since most of these patients,
at least three consecutive recurrent pregnancy especially young women, already have a good
losses12 revealed an ongoing pregnancy rate per chance of success.
embryo transfer of 29% in young women (less than A major drawback to the effectiveness of PGD-AS
37 years old). In the case of older women, the results is the high incidence of mosaicism documented in
were extremely disappointing, as an ongoing preg- preimplantation embryos (up to 50%), since it leads
nancy rate of 5.5% was recorded. The main reason to the loss of suitable embryos for replacement.16
for this poor outcome is that most of these women Additionally, the incomplete assessment of the
also had infertility as a consequence of decreased embryo chromosomal constitution performed by
oocyte quality. Consequently, these patients had fluorescence in situ hybridization (FISH) does not
significantly more abnormal embryos than did enable the exclusion of aneuploidy.
patients under 37 years old (66.9% vs 43.8%, An alternative approach is to perform PGD-AS
respectively). Another study,15 however, reported a with diagnostic purposes, as it has been shown that
Table 3d.1 Summary of observational studies evaluating the pregnancy outcome of couples with
unexplained recurrent miscarriage after performing PGD-AS
Study Study group Control group Outcome
Pellicer et al9 ≥ 3 miscarriages: PGD sex-linked disease: Study group: 3 pregnant, 1 miscarriage
≤ 36 years (n=9) ≤ 36 years (n=10); Control group: ≤ 36 years: 3 pregnancies,
> 36 years (n=6) 1 miscarriage
> 36 years: 0 pregnancies
Rubio et al10 ≥ 2 miscarriages: PGD sex-linked disease: Study group: < 37 years, 10.5% miscarriage rate
< 37 years (n=51); < 37 years (n=15); ≥ 37 years, 25% miscarriage rate
≥ 37 years (n=20) ≥ 37 years (n=13) Control group: no miscarriage
Munné et al15 ≥ 3 miscarriages: Expected pregnancy loss Study group: < 35 years, 23% miscarriage rate
< 35 years (n=21) according to Brigham ≥ 35 years, 12% miscarriage rate
≥ 35 years (n=37) et al 1999 Control group: < 35 years, expected
pregnancy loss 29%;
≥ 35 years, expected pregnancy loss 44.5%
Platteau et al12 ≥ 3 miscarriages: No Ongoing pregnancy rate/embryo transfer:
< 37 years (35 cycles) < 37 years, 29%
≥ 37 years (34 cycles) ≥ 37 years, 5.5%
Rubio et al14 ≥ 2 miscarriages: PGD for sex-linked Study group: < 37 years, 10% miscarriage rate
< 37 years (163 cycles); disease n= 25 ≥ 37 years, 20% miscarriage rate
≥ 37 years (78 cycles) Control group: 8.3% miscarriage rate
50
0415421306-Ch03d 3/29/07 5:31 PM Page 51
DEBATE: SHOULD PGD BE PERFORMED IN RPL? – AGAINST
about 20% of these couples do not develop chro- having a healthy child.20 This was not the case, how-
mosomally normal embryos.10 In addition, a high ever, for the first and second pregnancies, where a
predictability rate of the first PGD cycle towards significantly lower success rate was encountered in
successive attempts has been described (73%).17 the carrier couples (62% vs 72% and 44% vs 55%,
This valuable information could help some of these respectively).
couples in their decision to undergo a second cycle Robertsonian translocations have been tradi-
of PGD-AS. tionally associated with a better outcome than
In conclusion, conflicting evidence currently reciprocal translocations, since a higher rate of
exists regarding the contribution of aneuploidy alternate segregation is present in the spermatozoa
screening in the outcome of these patients, (88% vs 46%, respectively).22 Additionally, up to
although it seems that PGD-AS does not perform 70% of the embryos from female carriers of recip-
better than expectant management. Therefore, con- rocal translocations show unbalanced rearrange-
sidering the need to undergo in vitro fertilization ments.23 Furthermore, in a series of 100 couples
(IVF) and the high cost associated with aneuploidy with recurrent miscarriages, only patients with
screening, this technique should not be performed reciprocal translocations revealed a higher miscar-
on a routine basis, unless informed consent is given. riage rate in the subsequent pregnancy (72% recip-
These couples should be well informed about the rocal translocations vs 36% robertsonian).24
lack of consistency in available evidence to ensure Nevertheless, after a long-term follow-up, no differ-
an appropriate decision is made. Only future well- ence in live births was observed between these
designed randomized trials will be able to establish two types of translocations (83% reciprocal vs
the definitive value of this technique in unexplained 82% robertsonian).20
recurrent miscarriage couples. PGD has been proposed to these couples as an
early form of prenatal diagnosis in order to avoid
the negative psychological effects of being con-
STRUCTURAL CHROMOSOME DISORDERS fronted with repeated miscarriages or the termina-
tion of a pregnancy when malformations are
The prevalence of structural rearrangement in cou- diagnosed. Unfortunately, the few data available for
ples with recurrent miscarriage is 3–5%, transloca- analysis are derived from retrospective studies only
tions (reciprocal and robertsonian) being the most (Table 3d.2). Munné et al23 found a significant
frequently observed.18,19 The diagnosis of a translo- decrease in the miscarriage rate after performing
cation is usually made after repeated miscarriages PGD (from 92% to 12.5%). A correlation between
have occurred or a child with a congenital defect has the pregnancy rate and the number of abnormal
been born as a consequence of an unbalanced embryos with a cutoff value of 50% was also found.
rearrangement. In the case of couples with recur- The ESHRE PGD Consortium Data Collection25
rent miscarriage, the chance of having an affected reported a clinical pregnancy rate per embryo
child is lower than that of couples with a sick child transfer of 24% for robertsonian translocations and
already (2–5% vs 20–22%).20 17% for reciprocal translocations. Chun et al26
Overall, considering translocations, inversions found a significant decrease in the miscarriage rate
and numeric mosaics, a live birth rate of 45% has after PGD (from 95.8% to 16.7%). Finally, a multi-
been reported after three consecutive miscarriages center report including 469 PGD cycles performed
without treatment.21 Moreover, a recent case–control for translocations showed a pregnancy rate per
study, including couples with two or more miscar- embryo transfer of 34%.27
riages with a follow-up of 5 years after the diagno- Overall, these results reveal that PGD does
sis of the structural chromosome defect, observed not perform better than expectant management.
no difference between the carrier and the non-carrier Furthermore, the risks and costs of IVF and
groups (83% vs 84%) in the global chance of PGD have to be considered in order to make a
51
0415421306-Ch03d 3/29/07 5:31 PM Page 52
Table 3d.2 Pregnancy outcome after performing PGD in translocation carrier patients with recurrent
miscarriage
Munné et al2 Harper et al25 Chun et al26 Verlinsky et al27
No. of 35 patients 574 cycles 49 patients 469 cycles

patients/cycles
Outcome Miscarriage rate decrease Pregnancy rate/embryo Delivery rate 32.6% Pregnancy rate/embryo
from 92% to 12.5% transfer: robertsonian transfer 34%
24%; reciprocal 17%
well-informed decision. When counselling these 10. Rubio C, Simon C, Vidal F, et al. Chromosomal abnormalities and
patients, multiple variables must be taken into con- embryo development in recurrent miscarriage couples. Hum Reprod
2003; 18:182–8.
sideration to enable the most appropriate choice to 11. Werlin L, Rodi I, DeCherney A, et al. Preimplantation genetic diagno-
be performed: woman’s age, number of previous sis as both a therapeutic and diagnostic tool in assisted reproductive
miscarriages, existence of an affected offspring, type technology. Fertil Steril 2003; 80:467–8.
12. Platteau P, Staessen C, Michiels A, et al. Preimplantation genetic diag-
of translocation, and presence of associated infertility. nosis for aneuploidy screening in patients with unexplained recurrent
It is also important to perform a complete workup, miscarriages. Fertil Steril 2005; 83:393–7.
since antiphospholipid syndrome has been diag- 13. Twisk M, Mastenbroeck S, van Wely, et al. Preimplantation genetic
screening for abnormal chromosomes (aneuploidies) in in vitro fer-
nosed in 29% of these patients.28 tilisation or intracytoplasmic sperm injection. Cochrane Database
In conclusion, on the basis of currently available Syst Rev 2006; (1):CD00S291.
evidence, it seems that PGD does not improve the 14. Rubio C, Rodrigo L, Perez-Cano I, et al. FISH screening of aneuploidies
in preimplantation embryos to improve IVF outcome. RBM Online
pregnancy outcome in couples who are carriers 2005; 11:497–506.
of structural chromosome rearrangements when 15. Munné S, Chen S, Fischer J, et al. Preimplantation genetic diagnosis
compared with expectant management. reduces pregnancy loss in women aged 35 years and older with a
history of recurrent miscarriages. Fertil Steril 2005; 84:331–5.
16. Bielanska M, Lin Tan S, Ao A. Chromosomal mosaicism throughout
REFERENCES human preimplantation development in vitro: incidence, type, and
relevance to embryo outcome. Hum Reprod 2002; 17:413–19.
1. Clifford K, Rai R, Watson H, et al. An informative protocol for the 17. Munné S, Escudero T, Colls P, et al. Predictability of preimplantation
investigation of recurrent miscarriage: Preliminary experience of 500 genetic diagnosis of aneuploidy and translocations on prospective
consecutive cases. Hum Reprod 1994; 9:1328–1332. attempts. RBM Online 2004; 9:645–51.
2. Li TC, Makris M, Tomsu M, et al. Recurrent miscarriage: aetiology, 18. Stephenson MD. Frequency of factors associated with habitual abortion
management and prognosis. Hum Reprod Update 2002; 8:463–81. in 197 couples. Fertil Steril 1996; 66:24–9.
3. Stern JJ, Dorfman AD, Gutierrez-Najar, et al. Frequency of abnormal 19. De Braekeleer M, Dao TN. Cytogenetic studies in couples experiencing
karyotype among abortuses from women with and without a history of repeated pregnancy losses. Hum Reprod 1990; 5:519–28.
recurrent spontaneous abortion. Fertil Steril 1996; 65:250–3. 20. Franssen M, Korevaar J, van der Veen F, Leschot NJ, et al.
4. Sanchez JM, Franzi L, Collia F, et al. Cytogenetic study of spontaneous Reproductive outcome after chromosome analysis in couples with
abortions by transabdominal villus sampling and direct analysis of two or more miscarriages: index–control study. BMJ 2006; 322:1012.
villi. Prenat Diagn 1999; 19:601–3. 21. Carp H, Feldman B, Oelsner G, et al. Parental karyotype and subsequent
5. Carp H, Toder V, Aviram A, et al. Karyotype of the abortus in recurrent live births in recurrent miscarriage. Fertil Steril 2004; 81:1296–301.
miscarriage. Fertil Steril 2001; 75:678–82. 22. Gardner RJM Sutherland GR. Chromosome Abnormalities and
6. Sullivan AE, Silver RM, LaCoursiere Y, et al. Recurrent fetal aneuploidy Genetic Counseling, 3rd edn. Oxford: Oxford University Press, 2004.
and recurrent miscarriage. Obstet Gynecol 2004; 104:784–8. 23. Munné S, Sandalinas M, Escudero T, et al. Outcome of preimplanta-
7. Ogasawara M, Aoki K, Okada S, et al. Embryonic karyotype of abortion genetic diagnosis of translocations. Fertil Steril 2000; 73:1209–18.
tuses in relation to the number of previous miscarriages. Fertil Steril 24. Sugiura-Ogasawara M, Ozaki Y, Sato T, et al. Poor prognosis of recur-
2000; 73:300–4. rent aborters with either maternal or paternal reciprocal translocations.
8. Clifford K, Rai R, Regan L. Future pregnancy outcome in unexplained Fertil Steril 2004; 81:367–73.
recurrent first trimester miscarriage. Hum Reprod 1997; 12:387–9. 25. Harper JC, Boelaert K, Geraedts J, et al. ESHRE PGD Consortium
9. Pellicer A, Rubio C, Vidal F, et al. In vitro fertilization plus preimplan- Data Collection V: cycles from January to December 2002 with preg-
tation genetic diagnosis in patients with recurrent miscarriage: an nancy follow-up to October 2003. Hum Reprod 2006; 21:3–21.
analysis of chromosome abnormalities in human preimplantation 26. Chun KL, Jin HJ, Dong MM, et al. Efficacy and clinical outcome of
embryos. Fertil Steril 1999; 71:1033–9. preimplantation genetic diagnosis using FISH for couples of reciprocal
52
0415421306-Ch03d 3/29/07 5:31 PM Page 53
DEBATE: SHOULD PGD BE PERFORMED IN RPL? – AGAINST
and robertsonian translocations: the Korean experience. Prenat Diagn 28. Stephenson M, Sierra S. Reproductive outcomes in recurrent pregnancy
2004; 24:556–61. loss associated with a parental carrier of a structural chromosome
27. Verlinsky Y, Cohen J, Munné S, et al. Over a decade of experience with rearrangement. Hum Reprod 2006; 21:1076–82.
preimplantation genetic diagnosis: a multicenter report. Fertil Steril
2004; 82:292–4.
53
0415421306-Ch03d 3/29/07 5:31 PM Page 54
0415421306-Ch03e 3/29/07 5:32 PM Page 55
OPINION
3e. Should CVS or amniocentesis be performed

in RPL without screening?
Howard Cuckle
Women with recurrent miscarriages are at increased The maternal age-specific birth prevalence of
risk of fetal chromosomal abnormalities in subse- Down syndrome increases to 0.11%, 0.26%, 0.98%,
quent pregnancies. I will argue that this risk is not and 3.5% by ages 30, 35, 40, and 45, respectively.5
sufficiently high to automatically justify invasive The estimated risk at term for all common autoso-
prenatal diagnosis. Instead, a policy is advocated of mal trisomies – Down, Edwards, and Patau syn-
continual risk reassessment using a sequential mul- dromes – is 0.48% and 1.6% at ages 35–39 and
tiple marker antenatal screening protocol. This con- 40–44, and for all chromosomal abnormalities, it is
clusion is reached in the context of the steady 0.81% and 2.4%, respectively.6 A family history of
improvement in the efficacy of antenatal screening chromosomal abnormality confers a much higher
methods over recent decades. risk than this when a maternal balanced transloca-
Screening for chromosomal abnormalities has the tion is found,7 while for paternal carriers and for
simple aim of identifying pregnancies at sufficiently non-carrier couples, there is only a modest excess
high risk of an affected birth to warrant the hazards over their maternal age-specific risk. With a Down
and costs of invasive testing. Chorionic villus sam- syndrome proband and non-carrier parents, the
pling (CVS) and amniocentesis lead to at least 0.5% excess at midtrimester is 0.54% for the same disor-
fetal losses;1 procedural and karyotyping costs exceed der and 0.24% for other aneuploidies.8
US$1000.2 Despite some progress, the occasional Testing these two high-risk groups can have little
fetal cells and abundant free fetal DNA fragments cir- impact on birth prevalence, as most chromosomal
culating in maternal blood cannot yet be reliably abnormalities occur in young women and are spo-
used for non-invasive inexpensive prenatal diagnosis radic. This consideration has led to the development
of the common chromosomal abnormalities.3 of newer methods of selection for invasive testing.
Local policy and national convention on what Beginning in the mid-1980s, a series of maternal
counts as sufficiently high risk to warrant invasive serum markers of aneuploidy was discovered:
testing have generally emerged from the push and human chorionic gonadotropin (hCG), the free-β
pull of healthcare providers, reimbursement tariffs, subunit of hCG, α-fetoprotein (AFP), unconjugated
and professional bodies. Chromosomal abnormali- estriol (uE3), inhibin A, and pregnancy-associated
ties are relatively rare at birth – about 0.6%, exclud- plasma protein (PAPP)-A. Meanwhile even more dis-
ing mosaics, in 70 000 consecutive newborns criminatory ultrasound markers were found – nuchal
karyotyped4 – so universal unselective invasive test- translucency (NT), nuchal skinfold (NF), nasal bone
ing has never been considered an option; rather, (NB), tricuspid regurgitation (TR), and ductus veno-
from the earliest days, there was selection based on sus (DV) – as well as the weaker so-called ‘soft’ mark-
advanced maternal age and family history of chro- ers determined by the late second-trimester anomaly
mosomal abnormality. scan or genetic sonogram.
55
0415421306-Ch03e 3/29/07 5:32 PM Page 56
Various marker combinations, determined con- notion of a ‘standard of care’ is important and may
currently, formed the basis for the first effective have implications for litigation. Now, most coun-
screening protocols. The efficacy of a given policy is tries have abandoned this hybrid policy, having
generally measured by applying a statistical model realized that universal screening would yield a very
to calculate the expected detection rate – the pro- high detection rate in the older women and sub-
portion of affected pregnancies selected for invasive stantially reduce the need for invasive testing.
testing – and the false-positive rate – the proportion For example, the first-trimester combined test
of unaffected pregnancies selected. When applied to described above with a 1-in-250 term cut-off risk
all women using a 1 in 250 term risk cut-off (the would detect 91% of the Down syndrome pregnan-
norm in the UK), the model-predicted Down syn- cies in women aged 35 or older, while selecting only
drome detection rate and false-positive rate are 69% 8.4% of older women for invasive testing. Similar
and 4.3% for the best second-trimester maternal considerations apply to those with a family history.
serum combination (AFP, free β-hCG, uE3, and Although women with recurrent miscarriages
inhibin), compared with 82% and 2.4%, respec- are at an increased risk of a fetal chromosomal
tively, for the most widely used first-trimester com- abnormality in subsequent pregnancies, this
bination (PAPP-A and free β-hCG at 10 weeks and increased risk is not very great. A recent study of
NT at 11 weeks).9 In the USA, where a 1-in-270 almost 47 000 women having invasive prenatal diag-
midtrimester risk cut-off is favored, equivalent to nosis found a steadily increasing trend in aneu-
about 1 in 350 at term, the corresponding rates are ploidy risk according to the number of previous
73%, 6.0%, 84%, and 3.3%, respectively. The same miscarriages.10 After adjustment for age, parity, and
markers can also detect a large proportion of other the indication for testing, the odds ratio compared
common autosomal trisomies; in the second with no miscarriages was 1.21, 1.26, and 1.51 for
trimester, this requires a separate risk cut-off for one, two, and three or more miscarriages, respec-
these disorders, but in the first-trimester, most are tively. For a woman aged 30 with recurrent miscar-
detected because of increased Down syndrome riages, this would barely increase the risk to that of
risks. Many of the remaining severe but non-lethal women aged 35, who are no longer considered
chromosomal abnormalities are also detected inci- automatic candidates for invasive testing.
dentally because of high trisomy risk or an extreme A large proportion of couples with recurrent
NT level. In Down syndrome (and possibly other miscarriages are carriers of structural chromosomal
chromosomal abnormalities), NF detects fewer rearrangements. And subsequent pregnancies are
affected pregnancies than does NT; nevertheless, with more likely to miscarry: in one study, the propor-
a 1-in-250 term cut-off risk, the model-predicted tion of pregnancies ending in a live birth among
detection rate and false-positive rate for NF as a 73 carrier and 588 non-carrier couples was 45%
single marker are 53% and 3.5%.9 The soft markers and 55% respectively;11 another study of 49 carriers
are generally much less powerful, but incorporating found 52% miscarriages, compared with 28% in a
them into the risk calculation could substantially large series of non-carriers.12 However, fetal unbal-
increase the efficacy. anced translocations neither account for the excess
Screening was not initially applied to all women of miscarriages nor contribute much to the overall
but only to those not already regarded as high risk chromosomal abnormality risk. In one study of sub-
based on age and history, which is inefficient, since sequent abortuses among 39 carrier couples, only
some with potentially low risks receive invasive test- 4 had unbalanced translocations;13 in a study of
ing. However, it was argued that women in the tra- pregnancies among 239 carriers, just 4 unbalanced
ditional high-risk groups expect to be provided karyotypes were found.14
with a diagnostic testing and the offer of a less Thus, purely in terms of aneuploidy risk, there is
definitive screening alternative was unfair. This case no compelling reason to offer invasive testing auto-
was made most forcefully in the USA, where the matically to women with recurrent miscarriages.
56
0415421306-Ch03e 3/29/07 5:32 PM Page 57
OPINION: SHOULD CVS OR AMNIOCENTESIS BE PERFORMED IN RPL WITHOUT SCREENING?
However, there may be an argument in favor of protocol that I have outlined here is a form of risk
more intensive antenatal screening than is currently reassessment, by turns reassuring many and focusing
provided routinely. Sequential screening protocols concerns on a few with extreme risks.
are being developed that considerably increase the
detection rate for Down syndrome and the other REFERENCES
common trisomies. Attendance for screening is
required on two or more occasions, which is easier 1. Evans MI, Wapner RJ. Invasive prenatal diagnostic procedures 2005.
Semin Perinatol 2005; 29:215–18.
to arrange for women with recurrent miscarriages, 2. Cusick W, Buchanan P, Hallahan TW, et al. Combined first-trimester
who already receive continual surveillance. The versus second-trimester serum screening for Down syndrome: a cost
simplest protocol starts with the first-trimester analysis. Am J Obstet Gynecol 2003; 188:745–51.
3. Hahn S, Huppertz B, Holzgreve W. Fetal cells and cell free fetal nucleic
combined test described above, but adopts an acids in maternal blood: new tools to study abnormal placentation?
extremely high cut-off risk, selecting a small Placenta 2005; 26:515–26.
number for immediate CVS. The remainder then 4. Hook EB, Hammerton JL. The frequency of chromosome abnormal-
ities detected in consecutive newborn studies; differences between
have the second-trimester test described above, studies; results by sex and severity of phenotypic involvement. In:
except that all seven first- and second-trimester Hook EB, Porter IH, eds. Population Cytogenetics: Studies in
marker levels are incorporated into the risk calcula- Humans. New York: Academic Press, 1977:63–79.
5. Cuckle HS, Wald NJ, Thompson SC. Estimating a women’s risk of
tion. The model-predicted Down syndrome detec- having a pregnancy associated with Down’s syndrome using her age
tion rate and false-positive rate with a 1 in 250 term and serum alpha-fetoprotein level. Br J Obstet Gynaecol 1987;
risk cut-off are 91% and 1.9%, respectively. 94:387–402.
6. Hook EBH. Chromosomal abnormalities: prevalence, risks and recur-
Incorporating the newer first-trimester ultrasound rence. In: Brock DJH, Rodeck CH, Ferguson-Smith MA, eds. Prenatal
markers (NB, TR, and DV) would substantially Diagnosis and Screening. Edinburgh: Churchill Livingstone, 1992:
enhance detection both of the common trisomies 351–92.
7. Boué A, Gallano P. A collaborative study of the segregation of inher-
and of other chromosomal abnormalities. For ited chromosome arrangements in 1356 prenatal diagnoses. Prenat
example, based on the latest parameters,15 routinely Diagn 1984; 4:45–67.
adding NB to the risk calculation would improve 8. Arbuzova S, Cuckle H, Mueller R, et al. Familial Down syndrome: evi-
dence supporting cytoplasmic inheritance. Clin Genet 2001;
the above rates to 95% and 1.1%, respectively. 60:456–62.
A further extension of the protocol could also be 9. Cuckle H, Benn P, Wright D. Down syndrome screening in the first
envisaged by incorporating into the risk calculation and/or second trimester: model predicted performance using meta-
analysis parameters. Seminars Perinatology 2005; 29:252–7.
NF and the soft markers, determined at the anomaly 10. Bianco K, Caughey B, Shaffer BL, et al. Spontaneous abortion and
scan, possibly after using a more extreme cut-off to aneuploidy Obstet Gynecol 2006; 107:1098–102.
select for amniocentesis at the second stage. As well 11. Carp H, Feldman B, Oelsner G, et al. Parental karyotype and subse-
quent live births in recurrent miscarriage. Fertil Steril 2004;
as increasing detection generally, it would, by defini- 81:1296–301.
tion, be more focused on the non-lethal disorders. 12. Sugiura-Ogasawara M, Ozaki Y, Sato T, et al. Poor prognosis of
Screening is a public health activity, and as such recurrent aborters with either maternal or paternal reciprocal transloca-
tions. Fertil Steril 2004; 81:367–73.
requires the definition of cut-off levels in order to 13. Carp H, Guetta E, Dorf H, et al. Embryonic karyotype in recurrent
predict use of resources. In practice, though, there miscarriage with parental karyotypic aberrations. Fertil Steril 2006;
is often less than strict adherence to the cut-off, 85:446–50.
14. Franssen MTM, Korevaar JC, van der Veen F, et al. Reproductive
which is merely taken to be a guide to action. Given outcome after chromosome analysis in couples with two or more
the high chance of pregnancy failure and the associ- miscarriages: case–control study. BMJ 2006; 332:759–63.
ated anxiety among women with recurrent miscar- 15. Cicero S, Rembouskos G, Vandecruys H, et al. Likelihood ratio for tri-
somy 21 in fetuses with absent nasal bone at the 11–14-week scan.
riages, it might be particularly appropriate to have a Ultrasound Obstet Gynecol 2004; 23:218–23.
loose interpretation. Seen in this way, the screening
57
0415421306-Ch03e 3/29/07 5:32 PM Page 58
0415421306-Ch04 3/29/07 5:33 PM Page 59
4. Does the maternal immune system regulate

the embryo’s response to teratogens?
Arkady Torchinsky and Vladimir Toder
INTRODUCTION allogeneic paternal strain lymphocytes.6 However, in

mice, immunization with syngeneic splenocytes prior
Maternal factors have long been known to determine to syngeneic mating results in perinatal and postnatal
the embryo’s resistance to teratogens. Research origi- mortality and an increased number of malformations
nally focused on the role of the maternal endocrinal, among the progeny.7 Additionally, it has been shown
cardiovascular, and nervous systems. The maternal that the sera of habitually aborting women who were
immune system was only investigated with regard immunized with paternal leucocytes improved
to the teratogenic potential of autoimmune phe- blastocyst development in culture and reversed the
nomena or vaccines. It is only since the beginning of embryotoxic effect of sera from non-immunized
the 1990s that the immune responses operating in the women with recurrent miscarriages.8
embryonic microenvironment have been recognized Finally, stimulation of the maternal immune
to be vital for pregnancy to develop successfully.1 response has been shown to improve the reproduc-
In addition to their role in regulating embryonic tive performance of mice with a high degree of
development, there is much evidence implicating spontaneous postimplantation embryonic loss. In
these immune responses in determining the tolerance the CBA/J × DBA/2J mouse mating combination,
of the embryo to environmental teratogens.2 This which is prone to resorption of pregnancies, alloim-
chapter summarizes the data dealing with the suscep- munization of the female with leukocytes of paternal
tibility of the embryo to teratogens, and the possible haplotype significantly decreased the proportion of
mechanisms whereby immune responses may affect resorbed pregnancies from approximately 40% to
the ability of the embryo to resist teratogenic 10–15%.9,10 The same effect has been achieved with
insults. non-specific immunostimulation of mice with
complete Freund’s adjuvant (CFA).11,12
FETOMATERNAL IMMUNOREACTIVITY
AND EMBRYONIC DEVELOPMENT FETOMATERNAL IMMUNOREACTIVITY
AND TERATOLOGIC SUSCEPTIBILITY
As early as in the mid-1960s, studies were performed
that demonstrated that immune responses had a The above studies, which have demonstrated that sur-
regulatory role in embryonic development. Mean vival of the embryo depends on immune responses
litter size and mean placental weight were found to be acting in the embryonic microenvironment, have
higher in allogeneic than in syngeneic pregnancies.3,4 initiated research into whether these responses are
The survival of transplanted embryos was also also involved in determining the susceptibility of
shown to be significantly higher when there was a the embryo to developmental toxicants. Torchinsky
difference in major histocompatibility complex et al13 have compared the effects of two teratogens,
(MHC) antigens between the parents.5 It has since cyclophosphamide (CP) and 2,3-quinoxalinedi-
been observed that the litter size and placental methanol-1,4-dioxide (CAS 17311-31-8)14 in syn-
weights were higher in mice preimmunized with geneically and allogeneically mated CBA/J and
59
0415421306-Ch04 3/29/07 5:33 PM Page 60
C57Bl/6 mice. Both strains had almost identical that injection of Pyran-activated macrophages to
sensitivities to these teratogens, and both strains CL/Fr mice, which have a high incidence of cleft lip
also showed a higher sensitivity to both teratogens and palate, decreased the incidence of these anom-
after syngeneic mating than allogeneic mating. alies. Torchinsky et al18 have also shown that
However, the design of these experiments precluded immunostimulation of ICR mice with a non-spe-
the authors from assessing the effect of genetic cific immune trigger (xenogeneic rat splenocytes)
differences between inbred and F1 (CBA/J × C57Bl/6) increases the tolerance of embryos to CP-induced
embryos on the different response to the teratogens. teratogenic effects (Figure 4.1). Furthermore, it has
Therefore, further experiments were performed in been found that immunization performed twice
C57Bl/6 females whose immune responses were (21 days before mating and on day 1 of pregnancy)
either depressed by removing the paraaortic lymph has a greater influence on the teratogenic response
nodes or activated by intrauterine immunization to CP than a single inoculation.18
with allogeneic paternal splenocytes.15,16 It has been The influence of the immune response on the sus-
observed15 that suppression of the maternal immune ceptibility to teratogens has also been investigated in
response significantly increases the sensitivity of F1 type 1 (‘insulin-dependent’) diabetes mellitus (IDM)
(C57Bl/6 × CBA/J) embryos to both teratogens and and heat shock. Meticulous metabolic control of dia-
almost eliminates the different responses between betes has significantly decreased the risk of gross
allogeneically and syngeneically mated females. structural malformations in newborn infants.
Thus, in mice undergoing extirpation of draining Nevertheless, the incidence of fetal malformations in
lymph nodes, CP produced a resorption rate of women with type 1 IDM (6–10%) is still three to five
approximately 20%, and a malformation rate of times higher than in non-diabetic women.19
77%, whereas in sham-operated females these In our studies,20 in laboratory animals, streptozocin
indices were 6% and 31%, respectively. In contrast, (STZ) was used to induce diabetes in ICR mice
females primed with allogeneic paternal spleno- treated with rat splenocytes 21 days before mating. In
cytes before allogeneic mating showed enhanced STZ-induced diabetic ICR mice, approximately
tolerance to both teratogens.16 9% of embryos show gross structural anomalies
The response to the above teratogens has also and the incidence of litters with malformed embryos
been tested in the second pregnancy of C57Bl/6 reaches 63%.21 Immunostimulation resulted in a
mice.16 It has been observed that the degree of decrease of both indices: only 18% of litters had
embryotoxicity induced by both teratogens depends malformed fetuses and the incidence of malformed
on the type of mating (allogeneic or syngeneic) in embryos was approximately 2%. Moreover, immuno-
the first and second pregnancy, and that embryos of stimulation was followed by in an increase in the preg-
females mated twice allogeneically demonstrate a nancy rate: approximately 70%, compared with
significantly higher resistance to both teratogens 44% in non-immunized diabetic females.
than do embryos of allogeneically mated primigravid Heat shock-induced teratogenic effects in
mice. These results suggested that the exposure of rodents are associated with the occurrence of
the maternal immune system to paternal antigens anomalies in the brain and eye.22 Our experiments
in the first pregnancy may modify the teratogenic in ICR mice23 have shown that immunization with
response of embryos in repeated pregnancies. rat splenocytes significantly decreases the propor-
Nomura et al17 have observed that embryos of tion of fetuses with exencephaly and open eyes.
ICR mice pretreated with synthetic (Pyran copoly- Also, the resorption rate in immunized mice was
mer) or biological (bacillus Calmette–Guerin similar to that seen in intact ICR mice (approxi-
(BCG) vaccine) agents that are known to activate mately 6%–10%), whereas in non-immunized mice
macrophages exhibit increased tolerance to terato- exposed to heat shock, it exceeded 20%.23
gens such as urethane, N-methyl-N-nitrosourea, Recently, a number of studies have been pub-
and ionizing radiation. Nomura et al17 also reported lished that concur with the above observations.
60
0415421306-Ch04 3/29/07 5:33 PM Page 61
DOES THE MATERNAL IMMUNE SYSTEM REGULATE THE EMBRYO’S RESPONSE TO TERATOGENS?
(a) (b) (c)
Figure 4.1 The teratogenic response of embryos of cyclophosphamide (CP)-treated intact and immunostimulated mice. CP induces
a specter of gross structural anomalies, such as open eyes, digit and limb reduction anomalies, exencephaly, gastroschisis, and growth
retardation, in a dose-dependent fashion. Immunostimulation of females with xenogeneic rat splenocytes is followed by a decrease
in the incidence and severity of these anomalies and an increase in fetal weight. (a) Fetus of an intact mouse. (b) Fetus of an
immunostimulated CP-treated mouse. (c) Fetus of a non-immunostimulated CP-treated mouse.
Holladay et al24 showed that immune stimulation of teratogenic effect but both of which do induce
pregnant mice with Pyran copolymer, attenuated BCG, postimplantation embryonic death.27 It has been
or CFA increased the resistance of embryos to terato- shown28 that immunization of C3H/HeJ female mice
gens such as 2,3,7,8-tetrachlorodibenzo-p-dioxin with allogeneic paternal splenocytes of DBA/2J
(TCDD, ‘dioxin’), urethane, N-methyl-N-nitrosourea, mice 7 days before mating reduces the number of
and valproic acid. This group also found that restraint stress-induced embryonic losses. Immune
maternal immune stimulation with CFA, granulo- stimulation of CBA/J female mice with paternal
cyte–macrophage colony-stimulating factor (GM-CSF), splenocytes of DBA/2J males 2 weeks before mating
or interferon-γ (IFN-γ) protects murine embryos decreases the number of ultrasonic stress-induced
against diabetes-induced teratogenic effects.25 In our resorptions. Finally, Hatta et al29 have reported that
studies,26 maternal immunostimulation with GM- stimulation of female mice with the biostimulators
CSF increased the resistance of murine embryos to CP. PSK and OK432 decreased the susceptibility of
The effect of maternal immunostimulation has embryos to the teratogen 5-azacytidine, whereas
also been demonstrated in mice exposed to ultra- injection of interleukin-1 (IL-1) decreased the tol-
sonic and restraint stresses, neither of which has a erance of embryos to this teratogen.
61
0415421306-Ch04 3/29/07 5:33 PM Page 62
The above studies provide evidence that immune cascades and prosurvival pathways acting to suppress
responses occurring between mother and fetus may the process of apoptosis.35 A number of molecules
influence the susceptibility of embryos to both envi- have been suggested to be key components of the
ronmental teratogens and detrimental stimuli gener- machinery of apoptosis and have also been impli-
ated by the mother. The mechanisms underlying cated as powerful determinants of teratogenic
this phenomenon remain largely undefined. susceptibility.30 It may be that teratogen-induced
However, research in reproductive immunology, alterations in the expression of these molecules may
developmental biology, and teratology may outline be normalized by maternal immune potentiation.
some of the mechanisms involved. Some possible The tumor suppressor protein p53, which is acti-
mechanisms are described below. vated by various cellular stresses that induce DNA
damage, is presently considered to be a key regulator
of apoptosis.36 It is thought that p53, which targets
POSSIBLE MECHANISMS OF INTERACTIONS several steps in the apoptotic process, increases
BETWEEN IMMUNE RESPONSES AND the probability that the process goes forward,
DEVELOPMENTAL TOXINS and ensures a well-coordinated program once
the process is initiated.36 Evidence is accumulating
MOLECULES REGULATING APOPTOSIS that suggests that p53 regulates the response of
IN THE EMBRYO embryos to teratogens such as benzo[a]pyrene,37
2-chloro-2-deoxyadenosine,38 ionizing radiation,39,40
Most teratogens act on the embryo itself. Therefore, diabetes,41 and CP.42 Our team43 has shown that a CP-
some of the mechanisms that determine the response induced teratogenic insult was followed by the accu-
of embryonic cells to teratogens must be affected by mulation of p53 protein in embryonic structures, and
maternal immune stimulation modifying teratological that maternal immune stimulation with xenogeneic
susceptibility. These mechanisms are mainly associ- rat splenocytes, or GM-CSF, while increasing the
ated with the mechanisms regulating programmed tolerance of murine embryos to the teratogen, par-
cell death (apoptosis) in embryos responding toter- tially normalized the expression of the protein.26
atogenic stress.30 A comprehensive review of the Sharova et al44 have shown that mice exposed to the
literature addressing this topic is clearly beyond the teratogen urethane (which induces cleft palate in
scope of this chapter. Suffice it to say that apoptosis mice), injection of CFA or IFN-γ was followed by a
plays a crucial role in normal embryogenesis. decreased incidence of malformed fetuses and that
Apoptosis is involved in eliminating abnormal, CFA also normalized the urethane-induced alter-
misplaced, non-functional, or harmful cells, sculpting ations in the expression of the gene TP53 encoding
structures, eliminating unwanted structures, and for p53. Sharova et al44 also reported that maternal
controlling cell numbers.31 Teratological studies have immune stimulation also normalized the expression
shown that many of the chemical and physical devel- of the BCL2 gene, which is thought to encode one of
opmental toxicants that induce structural anomalies the key antiapoptotic proteins, BCL-2.45
also induce excessive apoptosis in embryonic struc- Other proteins considered to be the main executors
tures, which are subsequently malformed.32,33 Toder of apoptosis are the caspases, which belong to the
et al34 investigated whether maternal immune stimu- family of cysteine proteases.46 Caspases are divided
lation affects the degree of teratogen-induced apop- into two groups: initiator and effector caspases. The
tosis, and reported that immune stimulation of activation of initiator caspases takes place after their
females with xenogeneic rat splenocytes did indeed binding to adapter molecules, and mature initiator
decrease the intensity of CP-induced excessive apop- caspases activate effector caspases. The initiator
tosis in embryonic structures. caspase 9 (and possibly caspase 2) operates in the
Apoptosis is a genetically regulated process that mitochondrial proapoptotic pathway, whereas the
is realized by the activation of both death signaling initiator caspases 8 and 10 act in the death-receptor
62
0415421306-Ch04 3/29/07 5:33 PM Page 63
proapoptotic pathway. Both pathways use effector condition for maternal–fetal immune tolerance.54–56
caspases (caspases 3, 6, and 7).47 It has been There are data indicating that cytokine imbalances
reported30 that at least one of the main initiator that precede or accompany embryonic death induce
caspases 8 or 9 and/or the main effector caspase 3 various stresses and are also involved in some of
are involved in the response to teratogens such as the mechanisms regulating the susceptibility of the
diabetes, ionizing radiation, heat shock, CP, sodium embryo to detrimental stimuli.57 Additionally,
arsenite, and retinoic acid.30 The possibility that teratogenic insults may also be accompanied by dys-
maternal immune stimulation may modify terato- regulation of the cytokines operating in the embryo.
logical susceptibility by affecting the process of We have observed that CP-induced teratogenesis is
teratogen-induced activation of caspases has been accompanied by an increase in TNF-α and decreases
supported by our recent study.48 The levels of active in transforming growth factor-β2 (TGF-β2) and
caspases 3, 8, and 9 were lower in the embryos of macrophage colony-stimulating factor (M-CSF;
immunostimulated CP-treated mice than in embryos colony-stimulating factor 1, CSF-1) expression at
of mice exposed to the teratogen alone.48 the fetomaternal interface.58–60 Increased TNF-α
The transcription factor, nuclear factor κB (NF- and decreased TGF-β2 expression have also been
κB) is also thought to be a key molecule preventing described in the uterus of diabetic mice.61–63
cell death via the activation of genes whose prod- Studies in TNF-α knockout mice have shown no
ucts function as antiapoptotic proteins.49 NF-κB is alterations in litter size, sex ratio, weight gain, or
transcriptionally active in embryos during organo- structural anomalies, indicating that TNF-α probably
genesis. One subunit of NF-κB, p65, has been shown does not play an essential role in regulating normal
to be indispensable for the protection of the embry- embryogenesis.64 Early studies addressing the func-
onic liver against the physiological apoptosis tional role of TNF-α in reproduction implied that
induced by tumor necrosis factor α (TNF-α).50 TNF-α may be a trigger of embryonic death caused
There are a number of studies implicating NF-κB as by developmental toxins, various stresses, and
a regulator of the response to teratogens such as maternal metabolic and immunological imbal-
thalidomide,51 phenytoin,52 and CP.53 Our recent ances.64 Subsequently, TNF-α was shown to activate
study48 has shown that NF-κB may be a target for both apoptotic and antiapoptotic signaling cas-
immune responses operating in the embryonic cades,65 which suggests that TNF-α may regulate
microenvironment. The results suggested that the response of the embryo to various stresses.
intrauterine immunostimulation with rat spleno- Indeed, in our experiments with CP, the incidence and
cytes attenuates the CP-induced suppression of NF- severity of CP-induced gross structural craniofacial
κB DNA-binding activity in mouse embryos. and limb anomalies were found to be higher in
The above data suggest some of the mechanisms by TNF-α-knockout fetuses than in their TNF-α-posi-
which maternal immune stimulation might alter tera- tive counterparts.66 TNF-α-knockout embryos have
tological susceptibility. However, these mechanisms also been found to be sensitive to diabetes-induced
operate in the embryo, and the pathways by which teratogenic stimuli.67
maternal immune stimulation affects these mecha- TGF-β, a multipotent growth factor, has been
nisms remain elusive. Therefore, we have only reported to be involved in regulating cell growth,
presented data that should be taken consideration differentiation, and migration, and extracellular
in research dealing with this topic. matrix deposition.68 TGF-β family isoforms such as
TGF-β1, TGF-β2, and TGF-β3 seem to be indispensa-
CYTOKINES AND GROWTH FACTORS OPERATING ble for normal embryogenesis. Indeed, TGF-β1-null
AT THE FETOMATERNAL INTERFACE embryos die before day 11 of pregnancy, whereas
25% of TGF-β2-knockout fetuses and 100% of
There is considerable evidence that the establish- TGF-β3-knockout fetuses exhibit cleft palate.69 A
ment of a balanced cytokine milieu is a necessary number of studies have reported that TGF-β may be
63
0415421306-Ch04 3/29/07 5:33 PM Page 64
involved in the mechanisms of induced teratogene-

sis. In experiments with the teratogen TCDD, which EMBRYO
induces cleft palate in mouse embryos, TGF-β3 p53 Caspases
was shown to counteract the effect of TCDD in Teratogenic Apoptosis machinery
stimuli
blocking palatal fusion.70 Additionally, TGF-β2- Bcl-2 NF-κB
knockout embryos have been found to be more sen-

sitive to retinoid-induced teratogenesis than their ? Maternal
?
TGF-β2-positive counterparts.71 immunostimulation
The above data imply that TNF-α and TGF-β are Cytokine balance
UTERUS
determinants of the teratological susceptibility of Mechanisms ensuring
maternal–fetal immune tolerance
embryos. Maternal immune stimulation, in addi-
tion to increasing the resistance of embryos to ter-
atogenic stress, also tends to normalize the
Figure 4.2 A simplified model depicting a possible pathway for
expression of these cytokines at the fetomaternal maternal immunostimulation-induced modification of terato-
interface,58,59,61,62 implicating maternally derived logic susceptibility. A teratogen affects the function of molecules
TNF-α and TGF-β in pathways through which regulating the teratogenic response (i.e., those regulating apop-
tosis) directly and, possibly, indirectly, via inducing an imbal-
maternal immune stimulation modifies the ance of cytokines operating in the embryonic vicinity. Maternal
responses of the embryo to teratogens. Although immunostimulation influences the teratological susceptibility
effective reciprocal signaling has been demon- via modifying the expression pattern of these cytokines.
strated between the uterus and preimplantation and
periimplantation embryos,72,73 the effectiveness of
reciprocal signaling during organogenesis (the immune system may modify the embryo’s sensitivity
period of greatest sensitivity to teratogens) remains not only to maternally derived immune abortifacient
undetermined. Nevertheless, the mechanisms stimuli, but also to environmental teratogens. These
thought to ensure maternal–fetal immune tolerance – mechanisms may also be relevant in interpreting
cytokines and growth factors acting in the embry- the mechanisms underlying ‘occult’ pregnancy
onic microenvironment – may be acting primarily loss74 and for developing therapy aimed at prevention
as mediator through which the maternal immune of pregnancy loss.
system regulates the response of the embryo to
environmental teratogens. The above data suggest a REFERENCES
model depicting a possible pathway by which
maternal immunostimulation may modify terato- 1. Hunt JS. Immunology of pregnancy. Curr Opin Immunol 1992; 4:1153–6.
2. Toder V, Torchinsky A. Immunoteratology: where we are and where to
logical susceptibility (Figure 4.2). Within the con- go. Am J Reprod Immunol 1996; 35;114–17.
text of this model, modification of teratological 3. Billingham RE. Transplantation immunity and the maternal–fetal
susceptibility by maternal immunostimulation relation. N Engl J Med 1964; 270:667–72.
4. Billington WD. Influence of immunologic dissimilarity of mother and
depends on both the type of teratogen and the type foetus on size placenta in mice. Nature 1964; 202:317–18.
of immune stimulator. 5. Kirby DKS. Transplantation and pregnancy. In: Rapoport FT,
Dausser J, eds. Human Transplantation. New York: Grune and Stratton,
1968: 565–8.
6. Beer AE, Scott JR, Billingham RE. Histocompatibility and maternal
CONCLUSIONS immune status as determinants of fetoplacental and litter weights in
rodents. J Exp Med 1975; 142:180–98.
7. Pechan PA. Syngeneic spleen immunization induces high mortality
This review has provided data indicating that among progeny in mice. Teratology 1986; 33:239–41.
maternal immune responses may be involved in 8. Zigril M, Fein A, Carp H, Toder V. Immunopotentiation reverses the
mechanisms determining the resistance of the embryotoxic effect of serum from women with pregnancy loss.
Fertil Steril 1991; 56:653–69.
embryo to teratogens. An important implication of 9. Chaouat G, Kiger N, Wegmann TG. Vaccination against spontaneous
this paradigm is that modulation of the maternal abortion in mice. J Reprod Immunol 1983; 5:389–92.
64
0415421306-Ch04 3/29/07 5:33 PM Page 65
10. Chaouat G, Menu E, Bonneton C, et al. Immunological manipulation OK432 on 5-azacytidine-induced birth defects. Congenit Anom
in animal pregnancy and models of pregnancy failure. Curr Opin (Kyoto) 2003; 43:46–56.
Immunol 1989; 1:1153–6. 30. Torchinsky A, Fein A, Toder V. Teratogen-induced apoptotic cell
11. Toder V, Strassburger D, Irlin Y, et al. Nonspecific immunopotentiators death: Does the apoptotic machinery act as a protector of embryos
and pregnancy loss: complete Freund adjuvant reverses high fetal exposed to teratogens? Birth Defects Res C Embryo Today 2005;
resorption rate in CBA/J × DBA/2 mouse combination. Am J Reprod 75:353–61.
Immunol 1990; 24:63–6. 31. Jacobson MD, Weil M, Raff MC. Programmed cell death in animal
12. Szekeres-Bartho J, Kinsky R, Kapovic M, et al. Complete Freund development. Cell 1997; 88:347–54.
adjuvant treatment of pregnant females influences resorption rates in 32. Knudsen TV. Cell death. In: Kavlock RJ, Daston GP, eds. Drug
CBA/J × DBA/2 matings via progesterone-mediated immunomodula- Toxicity in Embryonic Development I. Berlin; Springer-Verlag,
tion. Am J Reprod Immunol 1991; 26:82–3. 1997:211–44.
13. Torchinsky AM, Chirkova EM, Koppel MA, et al. Dependence of the 33. Mirkes PE. 2001 Warkany Lecture: To die or not to die, the role of
embryotoxic action of dioxidine and cyclophosphamide on the apoptosis in normal and abnormal mammalian development.
immunoreactivity of the maternal–fetal system in mice. Farmakol Teratology 2002; 65:228–39.
Toksikol 1985; 48:69–73. [in Russian] 34. Toder V, Savion S, Gorivodsky M, et al. Teratogen-induced apoptosis
14. Sweet DV. Registry of Toxic Effects of Chemical Substances. may be affected by immunopotentiation. J Reprod Immunol 1996;
1985–1986 Edition. Washington: US Government Printing Office, 30:173–85.
1986; 5:4305. 35. Danial NN, Korsmeyer SJ. Cell death: critical control points. Cell
15. Torchinsky A, Fein A, Toder V. Immunoteratology: I. MHC involvement 2004; 116:205–19.
in the embryo response to teratogens in mice. Am J Reprod Immunol 36. Fridman JS, Lowe SW. Control of apoptosis by p53. Oncogene 2003;
1995; 34:288–98. 22:9030–40.
16. Torchinsky A, Fein A, Carp H, et al. MHC-associated immunopoten- 37. Nicol CJ, Harrison ML, Laposa RR, et al. A teratologic suppresser role
tiation affects the embryo response to teratogen. Clin Exp Immunol for p53 in benzo[a]pyrene-treated transgenic p53-deficient mice. Nat
1994; 98:513–19. Genet 1995; 10:181–7.
17. Nomura T, Hata S, Kusafuka T. Suppression of developmental 38. Wubah JA, Ibrahim MM, Gao X, et al. Teratogen-induced eye defects
anomalies by maternal macrophages in mice. J Exp Med 1990; mediated by p53-dependent apoptosis. Curr Biol 1996; 6:60–9.
172:1325–30. 39. Norimura T, Nomoto S, Katsuki M, et al. p53-dependent apoptosis
18. Torchinsky A, Fein A, Toder V. Modulation of mouse sensitivity to suppresses radiation-induced teratogenesis. Nat Med 1996; 2:577–80.
cyclophosmamide-induced embryopathy by nonspecific intrauterine 40. Wang B, Ohyama H, Haginoya K, et al. Prenatal radiation-induced
immunopotentiation. Toxicol Meth 1995; 5:131–41. limb defects mediated by Trp53-dependent apoptosis in mice. Radiat
19. Reece EA, Homko CJ, Wu YK. Multifactorial basis of the syndrome of Res 2000; 154:673–9.
diabetic embryopathy. Teratology 1996; 54:171–83. 41. Pani L, Horal M, Loeken MR. Rescue of neural tube defects in Pax-3-
20. Torchinsky A, Toder V, Carp H, et al. In vivo evidence for the existence deficient embryos by p53 loss of function: implications for Pax-3-
of a threshold for hyperglycemia-induced major fetal malformations: dependent development and tumorigenesis. Genes Dev 2002;
relevance to the etiology of diabetic teratogenesis. Early Pregnancy 16:676–80.
1997; 3:27–33. 42. Moallem SA, Hales BF. The role of p53 and cell death by apoptosis and
21. Torchinsky A, Toder V, Savion S, et al. Immunopotentiation increases necrosis in 4-hydroperoxycyclophosphamide-induced limb malfor-
the resistance of mouse embryos to diabetes-induced teratogenic mations. Development 1998; 125:3225–34.
effect. Diabetologia 1997; 40:635–40. 43. Torchinsky A, Ivnitsky I, Savion S, et al. Cellular events and the pat-
22. Edwards MJ, Shiota K, Smith MRS, et al. Hyperthermia and birth tern of p53 protein expression following cyclophosphamide-initiated
defects. Reprod Toxicol 1995; 9:411–25. cell death in various organs of developing embryo. Teratog Carcinog
23. Yitzhakie D, Torchinsky A, Savion S, et al. Maternal immunopotentiation Mutagen 1999; 19:353–67.
affects the teratogenic response to hyperthermia. J Reprod Immunol 44. Sharova LV, Sura P, Smith BJ, et al. Non-specific stimulation of the
1999; 45:49–66. maternal immune system. I. Effects on fetal gene expression.
24. Holladay SD, Sharova L, Smith BJ, et al. Nonspecific stimulation of the Teratology 2000; 62:420–8.
maternal immune system. I. Effects on teratogen-induced fetal 45. Tsujimoto Y, Shimizu S. Bcl-2 family: life-or-death switch. FEBS Lett
malformations. Teratology 2000; 62:413–19. 2000; 466:6–10.
25. Punareewattana K, Holladay SD. Immunostimulation by complete 46. Degterev A, Boyce M, Yuan J. A decade of caspases. Oncogene 2003;
Freund’s adjuvant, granulocyte macrophage colony-stimulating 22:8543–67.
factor, or interferon-γ reduces severity of diabetic embryopathy in 47. Pommier Y, Antony S, Hayward RL, et al. Apoptosis defects and
ICR mice. Birth Defects Res A Clin Mol Teratol 2004; 70:20–7. chemotherapy resistance: molecular interaction maps and networks.
26. Savion S, Kamshitsky-Feldman A, Ivnitsky I, et al. Potentiation of the Oncogene 2004; 23:2934–49.
maternal immune system may modify the apoptotic process in 48. Torchinsky A, Gongadze M, Zaslavsky Z, et al. Maternal immunopo-
embryos exposed to developmental toxicants. Am J Reprod Immunol tentiation affects caspase activation and NF-κB DNA-binding activity
2003; 49:30–41. in embryos responding to an embryopathic stress. Am J Reprod
27. Scialli AR. Is stress a developmental toxin? Reprod Toxicol 1988; Immunol 2006; 55:36–44.
1:163–72. 49. Karin M, Lin A. NF-κB at the crossroad of life and death. Nat
28. Clark DA, Banwatt D, Chaouat G. Stress-triggered abortion in mice is Immunol 2002; 3:221–7.
prevented by alloimmunization. Am J Reprod Immunol 1993; 29:141–7. 50. Beg AA, Sha WC, Bronson RT, et al. Embryonic lethality and liver
29. Hatta A, Matsumoto A, Moriyama K, et al, Opposite effects of the degeneration in mice lacking the RelA component of NF-κB. Nature
maternal immune system activated by interleukin-1β vs. PSK and 1995; 376:167–70.
65
0415421306-Ch04 3/29/07 5:33 PM Page 66
51. Hansen JM, Harris C. A novel hypothesis for thalidomide-induced 63. Pampfer S. Dysregulation of the cytokine network in the uterus of the
limb teratogenesis: redox misregulation of the NF-κB pathway. diabetic rat. Am J Reprod Immunol 2001; 45:375–81.
Antioxid Redox Signal 2004; 6:1–14. 64. Toder V, Fein A, Carp H, et al. TNF-α in pregnancy loss and embryo
52. Kennedy JC, Memet S, Wells PG. Antisense evidence for nuclear maldevelopment: a mediator of detrimental stimuli or a protector of
factor-κB-dependent embryopathies initiated by phenytoin- the fetoplacental unit? J Assist Reprod Genet 2003; 20:73–81.
enhanced oxidative stress. Mol Pharmacol 2004; 66:404–12. 65. Baud V, Karin M. Signal transduction by tumor necrosis factor and its
53. Torchinsky A, Gongadze M, Savion S, et al. Differential teratogenic relatives. Trends Cell Biol 2001; 11:372–7.
response of TNFα+/+ and TNFα-/- mice to cyclophosphamide: the possi- 66. Torchinsky A, Shepshelovich J, Orenstein H, et al. TNF-α protects
ble role of NF-κB. Birth Defects Res A Clin Mol Teratol 2006; 76:437–44. embryos exposed to developmental toxicants. Am J Reprod Immunol
54. Raghupathy R. Pregnancy: success and failure within the Th1/Th2/Th3 2003; 49:159–68.
paradigm. Semin Immunol 2001; 13:219–27. 67. Torchinsky A, Gongadze M, Orenstein H, et al. TNF-α acts to prevent
55. Niederkorn JY. See no evil, hear no evil, do no evil: the lessons of occurrence of malformed fetuses in diabetic mice. Diabetologia 2004;
immune privilege. Nat Immunol 2006; 7:354–9. 47:132–9.
56. Trowsdale J, Betz AG. Mother’s little helpers: mechanisms of mater- 68. Massague J. How cells read TGF-β signals. Nat Rev Mol Cell Biol 2000;
nal–fetal tolerance. Nat Immunol 2006; 7:241–6. 1:169–78.
57. Arck PC. Stress and pregnancy loss: role of immune mediators, hor- 69. Nawshad A, LaGamba D, Hay ED. Transforming growth factor β
mones and neurotransmitters. Am J Reprod Immunol 2001; 46:117–23. (TGF-β) signalling in palatal growth, apoptosis and epithelial mes-
58. Gorivodsky M, Zemliak I, Orenstein H, et al. Tumor necrosis factor α enchymal transformation (EMT). Arch Oral Biol 2004; 49:675–89.
mRNA and protein expression in the uteroplacental unit of mice with 70. Thomae TL, Stevens EA, Bradfield CA. Transforming growth factor-β3
pregnancy loss. J Immunol 1998; 160:4280–8. restores fusion in palatal shelves exposed to 2,3,7,8-
59. Gorivodsky M, Torchinsky A, Zemliak I, et al. TGFβ2 mRNA expres- tetrachlorodibenzo-p-dioxin. J Biol Chem 2005; 280:12742–6.
sion and pregnancy failure in mice. Am J Reprod Immunol 1999: 71. Nugent P, Pisano MM, Weinrich MC, et al, Increased susceptibility to
42:124–33. retinoid-induced teratogenesis in TGF-β2 knockout mice. Reprod
60. Gorivodsky M, Torchinsky A, Shepshelovich J, et al. Colony-stimulating Toxicol 2002; 16:741–7.
factor-1 (CSF-1) expression in the uteroplacental unit of mice with 72. Diaz-Cueto L, Gerton GL. The influence of growth factors on the
spontaneous and induced pregnancy loss. Clin Exp Immunol 1999; development of preimplantation mammalian embryos. Arch Med Res
117:540–9. 2001; 32:619–26.
61. Fein A, Kostina E, Savion S, et al. Expression of tumor necrosis factor-α 73. Dominguez F, Pellicer A, Simon C. Paracrine dialogue in implanta-
in the uteroplacental unit of diabetic mice: effect of maternal tion. Mol Cell Endocrinol 2002; 186:175–81.
immunopotentiation. Am J Reprod Immunol 2001; 46:161–8. 74. Clark DA, Chaouat G, Gorczynski RM. Thinking outside the box:
62. Fein A, Magid N, Savion S, et al. Diabetes teratogenicity in mice is mechanisms of environmental selective pressures on the outcome of
accompanied with distorted expression of TGF-β2 in the uterus. the materno-fetal relationship. Am J Reprod Immunol 2002;
Teratog Carcinog Mutagen 2002; 22:59–71. 47:275–82.
66
0415421306-Ch05 3/29/07 5:34 PM Page 67
5. Fetal structural malformations

– embryoscopy
Thomas Philipp
INTRODUCTION Transcervical embryoscopy permits visualization

of the embryo in utero, without the damage caused
Most pregnancy losses are spontaneously aborted by instrumental evacuation or spontaneous passage.5
when the conceptus is undergoing embryonic Embryoscopy in early spontaneous abortions allows
development. Pregnancy loss is a significant health visualization of subtle morphological abnormalities
concern in economically advanced societies, where undetectable by ultrasound (Figure 5.1), and the
traditional early reproduction is replaced by a social diagnostic potential of transcervical embryoscopy in
trend towards establishing the mother’s career early failed pregnancies is just beginning to unfold.
before starting reproduction. As the whole repro-
ductive period may be shortened to 5–7 years, each
pregnancy becomes precious. Finding the cause of TECHNIQUE OF TRANSCERVICAL
pregnancy loss is essential for prognosis, recurrence EMBRYOSCOPY IN EARLY
risk counselling, and management of future preg- SPONTANEOUS OR MISSED ABORTIONS
nancies. Approximately 1% of fertile couples will
experience recurrent early pregnancy losses.1 Although Transcervical embryoscopy requires an average of
recurrent early pregnancy losses have been associ- 10 minutes (range 3–25 minutes) and is performed
ated with maternal factors such as maternal throm- by us under intravenous general anesthesia, it can
bophilic disorders, structural uterine anomalies, be organized into five different steps:
maternal immune dysfunction, endocrine abnor-
malities, and parental chromosomal anomalies (as 1. Insertion of hysteroscope and exploration of the
described in other chapters of this book), approxi- uterine cavity. The patient is placed in a dorsal
mately 50% of recurrent miscarriages are classified lithotomy position. A speculum is inserted into
as idiopathic following maternal investigation. For the vagina, which is cleaned with Betadine
affected couples, idiopathic pregnancy loss creates a solution. After careful dilatation of the cervix, a
great deal of grief and anxiety about the outcome of rigid hysteroscope (12° angle of view, with both
future pregnancies. It is currently unclear whether biopsy and irrigation working channels, Circon
embryonic maldevelopment is a contributiory factor Ch 25–8 mm) is passed through the cervix
in these cases. Investigations of the dead embryo are under direct vision. If vision is lost, the
rare.2,3 Demised embryos cannot be investigated for hysteroscope is withdrawn slightly, and
several practical reasons. Most losses occur when reinserted. A continuous normal saline flow is
the conceptus is undergoing embryonic develop- used throughout the procedure (pressure
ment. The small size of the embryo precludes 40–120 mmHg) to help distend and clean the
detailed examination, either by ultrasound (due to cervical canal and endometrial cavity, and thus
limitations of resolution) or by pathological tech- provide a clear view. In failed first-trimester
niques. Both instrumental evacuation and sponta- pregnancies, the decidua capsularis and parietalis
neous passage damage the embryo. It is rarely have not yet fused, so the uterine cavity can be
retrieved whole due to its minute size and fragility.4 assessed. (The uterine cavity is obliterated in
67
0415421306-Ch05 3/29/07 5:34 PM Page 68
Table 5.1 Incidence of acquired and congenital

(a) uterine abnormalities diagnosed by transcervical
embryo-hysteroscopy in missed abortion
Uterine pathology No. of cases Percentage
Acquired 33 64.7
Adhesions 26 51
Polyp 6 11.8
Fibroid 1 2
Congenital 18 35.3
Didelphys uterusa 1 2
Unicornuate uterusa 1 2
Septate uterusa 7 13.7
Arcuate uterus 4 7.8
Unclassified 5 9.8
(b) Total 51 100

a
Ascertained by laparoscopy/laparotomy investigating the external
uterine contour.
is opened with microscissors (CH 7–2 mm), due to

its opacity, and the embryo is first viewed through
the amnion. The small size of the embryo makes
high demands on image resolution. At the end of
the 8th week, it measures 30 mm, but already
possesses several thousand named structures.
Therefore, the embryoscope should be advanced
as close as possible to the embryo in order to
document the minute developing structures
Figure 5.1 (a) Endovaginal sonography prior to embryoscopy. such as the limbs (Figure 5.2). The amnion
The embryo, of 17 mm crown–rump length, showed no heart- usually obscures vision by reflecting light. In
beat. No abnormalities were identified sonographically. The
arrow marks the head of the embryo. U, umbilical cord. failed pregnancies, there is no need to avoid
(b) Embryoscopic lateral view of the upper portion revealed a amniotic rupture. Hence, the hysteroscope is
well-preserved embryo with anencephaly. The exposed brain inserted into the amniotic cavity after opening the
tissue (*) is still intact (exencephaly).The digital rays of the
hand (H) are notched. Parts of the external ear (E) are clearly membrane with microscissors. Documentation of
discernable. Remnants of the amnion are labeled (A). A normal the embryo’s details can be better achieved from
karyotype (46,XX) was diagnosed cytogenetically. within the amniotic cavity.
3. Morphological evaluation of the embryo. A com-
plete examination of the conceptus includes
midtrimester by fusion of the decidua capsularis visualization of the head, face, dorsal and ventral
with the decidua parietalis.) At this stage, walls, limbs, and umbilical cord. The incidence
congenital and acquired uterine defects can be of developmental defects is particularly high in
diagnosed. Table 5.1 shows the spectrum of early abortion specimens.6,7 The development of
uterine defects that we have been able to diagnose the human embryo is a dynamic process, with
by this technique in early failed pregnancies. constantly changing anatomy and hence appe-
2. Localization of the gestational sac and incision of arance. Early diagnosis of developmental defects
chorion and amnion. After inspection of the uterine by embryoscopy requires basic knowledge of the
cavity, the gestational sac is localized. The chorion anatomy of the developing human embryo.
68
0415421306-Ch05 3/29/07 5:34 PM Page 69
FETAL STRUCTURAL MALFORMATIONS – EMBRYOSCOPY
always a reliable result.12 Transcervical embryo-

scopy allows selective and reliable sampling of
chorionic tissues with minimal potential for
maternal contamination.13 Direct chorion biopsies
can be taken embryoscopically at the end of the
morphological examination13,14 (Figure 5.3). In
our service, direct chorionic villus sampling (CVS)
is performed under direct vision, through the
hysteroscope using a microforceps (CH 7–2 mm).
At the end of the procedure, chorionic villi are
placed in normal saline and carefully dissected. The
chorionic villi are then placed in culture medium
and immediately forwarded to the cytogenetic
laboratory for further processing. In our service,
the tissue is subsequently cultured and analyzed
cytogenetically, using standard G-banding
Figure 5.2 Embryoscopic lateral view of the upper portion of a
triploid embryo (69,XXY) 18 mm in length. The upper limb
cytogenetic techniques. Figure 5.4 shows the
(UL) shows characteristics of the 7th week of development. distribution of chromosome anomalies in our
The digital rays of the hands are notched. The elbow region is series of 359 specimens with an abnormal
appearing. The microcephalic embryo shows a poorly developed
cranium. The frontal area has lost the usual bulge expected in
karyotype.
embryos of this size. 5. Instrumental evacuation of the uterus. At the end
of the procedure, instrumental evacuation of the
uterus is performed.
Therefore, the investigator must develop the
ability to evaluate the developmental age of
embryos accurately, as the diagnosis of an
embryonic defect is dependent on precise
staging.8,9 The term ‘gestational age’, which is
used in clinical and ultrasound terminology,
should not used for studying missed abortions,
as most of these specimens are usually retained
in utero after embryonic demise. The actual
developmental age (DA) is derived from the
crown–rump length (CRL), measured by
ultrasonography, and from the developmental
stage assessed by embryoscopy.8
4. Tissue sampling. In couples with recurrent
miscarriage, and in cases of phenotypically
abnormal embryos (see ‘Etiology of devel-
opmental defects in early missed abortions’,
below), accurate cytogenetic analysis of
Figure 5.3 Direct chorionic villus sampling is performed
pregnancy tissue is essential.10,11 The value of under visual monitoring using a microforceps (M). Note the
karyotyping early abortion specimens is limited chorionic villi (V) at the tip of the microforceps. ‘A’ marks the
by frequent false-negative results, caused by remnants of the amnion. A microcephalic 45,XO embryo
(E) with a crown–rump length of 28 mm is visible in the
maternal tissue contamination. The finding of a background of the picture.
46,XX karyotype in the curettage material is not
69
0415421306-Ch05 3/29/07 5:34 PM Page 70
67
65
70
60
50
37
40
32
N
30
24 24
14
20
10 10 11
9 8 8
7 4 7
4 5 5
10 3 2 1
0 1 1 0
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Polyploidy
45,X
Structural
Other
Trisomy
Figure 5.4 Frequency of trisomy for each chromosome, polyploidy, monosomy X, and structural chromosome anomalies among
359 specimens with an abnormal karyotype.
COMMON MORPHOLOGICAL DEFECTS caudal pole can be differentiated. GD4 embryos have a
IN EARLY ABORTION SPECIMENS CRL > 10 mm, with a discernible head, trunk, and
DIAGNOSED EMBRYOSCOPICALLY limb buds. The limb buds show marked retardation in
development and the development of the facial struc-
This section provides an overview of developmental tures is highly abnormal.
defects that we have been able to diagnose using this In our experience, growth-disorganized embryos
technique of transcervical embryoscopy. Abnormal show a high frequency (92%) of autosomal tri-
embryonic development can be local or general. somies, trisomy 16 being the most common,
General embryonic maldevelopment is known as accounting for 46% of abnormal karyotypes.16
‘embryonic growth disorganization’. There are four
grades, which are based on the degree of abnormal LOCALIZED DEFECTS
embryonic development.15 An empty or anembry-
onic sac is known as grade 1 (GD1). The amnion, if Localized defects may be isolated or combined.
present, is usually closely adherent to the chorion Morphologically they are similar to developmen-
(fusion of the amnion to the chorion is abnormal tal defects seen in fetuses. Malformations that
prior to 10 weeks of gestation). GD2 conceptuses have external manifestation and that we have
show embryonic tissue of 3–5 mm in size, but with been able to diagnose embryoscopically include
no recognizable external embryonic landmarks and the following.
no retinal pigment. It is not possible to differentiate
caudal and cephalic poles (Figure 5.5). The embryo is HEAD DEFECTS
often directly attached to the chorionic plate. GD3
embryos are up to 10 mm long. They lack limb buds, Microcephaly, anencephaly, encephalocele, facial
but retinal pigment is often present. A cephalic and a dysplasia, cleft lip, cleft palate, fusions of the face to
70
0415421306-Ch05 3/29/07 5:34 PM Page 71
(a)
(c)
(b)
Figure 5.5 Endovaginal sonographic and embryoscopic examination (a) shows a monochorionic diamniotic twin pregnancy with
two intact amniotic sacs (A). The crown–rump lengths of the embryos (without cardiac activity) were 4 and 3 mm. Close-ups of
twin I (b) and twin II (c) showed two growth-disorganized embryos (GD2) with no recognizable external embryonic landmarks and
no retinal pigment after the amnion (A) was opened. Trisomy 16 (47,XY,+16) was diagnosed in this case.
the chest, absence of eyes, unfused eye globes, and covered by adherent discolored skin on embryoscopy.
proboscis are some of the defects that we have seen. Embryoscopy has identified encephaloceles in the
Microcephalic embryos may be seen on embry- frontal and parietal regions of the embryonic head,
oscopy with a poorly developed cranium with loss of unlike the situation in the fetus, where the defect usu-
normal vascular markings. In particular, the usual ally occurs in the occipital area. Encephaloceles may
bulge of the frontal area, which is expected in embryos range from small defects to large ones involving most
of this size, is absent (Figure 5.2). Embryos with a dys- of the cranium.11,17
plastic face show poorly developed branchial arches In anencephalic embryos, the brain tissue may still
and midface structures on embryoscopic examination. be present, and this condition is called exencephaly
Microcephaly and facial dysplasia are usually observed (Figure 5.1b). The developing cerebral structures
in combination. Chromosomal anomalies are the subsequently undergo varying degrees of destruction,
most common cause of these developmental defects. leaving a mass of vascular structures and degener-
Encephaloceles present as a bulge in the cranium, often ated neural tissue. Neural tube defects (anencephaly,
71
0415421306-Ch05 3/29/07 5:34 PM Page 72
encephalocele, and spina bifida) can be multifactorial TRUNK DEFECTS

in origin, caused by a lethal gene defect or non-genetic
mechanisms such as amniotic bands. Chromosomal Trunk defects include spina bifida, omphalocele, and
anomalies are the most common cause of embryonic gastroschisis. The phenotype of spina bifida is different
neural tube defects.11,17,18–20 The most common in the early developmental stages than the well-known
associations with chromosomal abnormalities are appearance in the fetus or neonate. In the embryo,
triploidy with spina bifida,21 and 45,XO and trisomies spina bifida is frequently observed as a plaque-like
9 and 14 with encephalocele.22 protrusion of neural tissue over the caudal spine.23 It is
Lateral and median cleft lip can be distinguished not clear whether the spina bifida seen in the embryo
embryoscopically. Lateral clefts may be unilateral or is due to a different cause to that seen in the fetus, or
bilateral. Cleft lip occurs when the maxillary promi- whether it is merely a precursor to the lesion observed
nence and the united medial nasal prominences fail in the fetus. Myeloceles vary in size and location.
to fuse. The midline cleft lip represents a fusion The most common site in the embryo is the lumbar
defect of the median nasal swellings (Figure 5.6). and sacral regions. Chromosomal aberrations are the
In the embryo, cleft lip cannot be diagnosed until most common cause of embryonic myeloceles.
after 7 weeks of development, since fusion does not Physiological midgut herniation is a macroscop-
occur until that time. Cleft lip may be part of a mal- ically visible process which starts in the 6th week
formation syndrome. Irregular clefting may be after fertilization. The midgut only fully returns to
caused by amniotic bands. In embryos, clefting the abdominal cavity at the end of the 10th week of
defects occur commonly with chromosomal aberra- development. Herniation is still physiological at
tions, especially trisomy 13. Cleft palate occurs if the 8 weeks of development; hence omphalocele can only
primary anterior palate, lateral palatine processes, be diagnosed in the fetal period. Gastroschisis differs
and nasal septum fail to unite. Cleft palate can only from the physiological herniation of the midgut, as
be diagnosed in the fetal period, since fusion is the umbilical cord is not involved and no sac is pres-
completed after the 10th week of development. ent. Gastroschisis is rarely observed in the embryo,
and occurs when the bowel protrudes from a defect
that is generally located to the right side of the
umbilicus. The pathogenesis of this defect is contro-
versial, with a variety of different theories having
been proposed.24–26 The theory of abdominal wall
disruption as a result of an ‘in utero’ vascular acci-
dent has gained most acceptance. Thus, gastroschisis
is considered to be a sporadic event with a negligible
risk of recurrence. Since the defect is usually not
associated with chromosome aberrations, it is rarely
observed in early spontaneous abortions.
LIMB DEFECTS
Polydactyly, syndactyly, split-hand/split-foot mal-

formation, and transverse limb reduction defects
are the most commonly observed malformations.
Figure 5.6 Close-up of the face of an embryo with a Polydactyly is one of the most common limb
crown–rump length of 27 mm. A median cleft lip (arrow) is abnormalities found in the embryo. It may be on the
present. ‘UL’ marks the right upper limb. Trisomy 9 (47,XY,+9)
was diagnosed.
radial (preaxial) or ulnar (postaxial) side of the limb.
Polydactyly may occur as an isolated malformation
72
0415421306-Ch05 3/29/07 5:34 PM Page 73
thin and/or short cords. The mechanical lesions of

the cord (knots, torsion, and stricture) are rarely
observed embryoscopically. Torsion of the umbilical
cord can often be found in macerated specimens,
but is usually a postmortem artifact. Umbilical cord
cysts and abnormal thin and/or short cords are usually
found in chromosomally abnormal embryos.
DUPLICATION ANOMALIES
Chorangiopagus parasiticus (CAPP), or acardiac

conjoined twins, and other conjoined twins have
been observed. The most severe defect in the acar-
diac conceptus is usually seen at the cranial pole.
The parasitic twin is usually seen as a markedly
edematous mass. The upper portion of the conceptus
Figure 5.7 An early fetus, of 60 mm crown–rump length, is has missing or highly abnormal facial structures.
shown with syndactyly of digits III and IV. The karyotype
showed triploidy (69,XXX). Usually, only remnants of the upper extremities, are
present, but the lower limbs are often well developed.
The ‘pump’ twin is also usually developmentally
or may be part of a malformation syndrome. abnormal.31,32 The circulation is through the
Postaxial polydactyly is common in trisomy 13.27 normal pump twin by a return reversed flow from
In syndactyly, two or more of the fingers or toes are artery directly to artery, or vein to vein, via anasto-
joined together. At the end of the 8th week of devel- moses of the cord, or via chorionic surface vessels.
opment fingers become free and syndactyly can be The observed anomalies of the parasitic twin are pre-
diagnosed embryoscopically (Figure 5.7). Syndactyly sumed to be caused by a combination of a primary
may be part of a malformation syndrome. Syndactyly developmental defects and decreased oxygenation of
of digits III and IV is common in triploidy.27,28 the recipient twin, with disruption of organogenesis.
The split-hand/split-foot malformation involves Conjoined twinning is the result of late and
ectrodactyly. The hand is divided into two parts, incomplete twin formation at the latest possible
which are opposed like a lobster claw. In the second moment when the embryonic axis is being laid
anatomical type, the radial rays are absent, with down (between 13 and 15 days postconception).
only the fifth digit remaining.29 Split hand can be a Most classifications are descriptive and based on the
part of numerous syndromes. In embryos with anatomical zones of coalescence. Fusion of the
split-hand malformation, trisomy 15 can often be thorax (thoracopagus) is most commonly (70%)
found. In the transverse limb reduction defect, distal reported.
structures of the limb are absent, with proximal The importance of identifying these rare dupli-
parts being more or less normal. These limb defects cation anomalies cannot be overemphasized; parents
are regarded as a disruption sequence that is pre- can be reassured that the anomalies are accidental
sumed to be a result of peripheral ischemia.30 The sequela of twinning, with no additional risk of
recurrence risk in future pregnancies is minimal.27 recurrence in future pregnancies.28
UMBILICAL CORD DEFECTS AMNION RUPTURE SEQUENCE
The following complications may affect the umbili- The pathogenesis of amniotic bands is still being
cal cord: knots, torsion, stricture, cysts, and abnormal debated. There are numerous theories.33 The theory
73
0415421306-Ch05 3/29/07 5:34 PM Page 74
of early amnion rupture, as proposed by Torpin,34 between the morphology and specific cytogenetic
has gained most acceptance. Amniotic rupture leads findings are shown in Table 5.3. It can be seen from
to subsequent amniotic band formation, which Table 5.2 that no external abnormalities were found
interferes with normal embryonic development by in 58 cases (11.3%), whereas abnormal develop-
causing malformations or disruptions. This sequence ment was seen in 456 (88.7%) cases of missed abor-
of events is known as the amnion rupture sequence tion. Among the abnormal cases, embryonic growth
(ARS).35 Although this sequence is uncommon in disorganization (GD1–4) was seen in 237 (46.1%)
liveborn infants, its frequency may be as high as 1 in cases. One hundred and ninety-eight cases (38.5%)
56 in previable fetuses. Bands that constrict the showed no disorganization of development, but
umbilical cord are recognized as the main cause of had severe combined localized defects. There were
death in this sequence.36 ARS may cause abnormali- isolated localized developmental defects in 21 spec-
ties that are detectable by embryoscopy, such as imens. Cytogenetic evaluation was successfully per-
encephaloceles, cleft lip, and amputations. When aber- formed in 495 (96.3%) of the 514 cases. Three
rant sheet or bands of tissue are seen on embryoscopy, hundred and fifty-nine (73%) specimens were
which are attached to the conceptus with character- abnormal, of which 230 (64.1%) were trisomic, 67
istic deformities in a non-embryological distribu- (18.7%) showed monosomy X, 37 (10.3%) were
tion, a diagnosis of amniotic band syndrome or polyploid, and 14 (3.9%) were structural chromo-
ARS can be made.37 Amniotic bands can occur as a somal anomalies. Trisomies were observed for
result of abdominal trauma,38 CVS,39 and connective all chromosomes except chromosomes 1 and 19
tissue abnormalities.40 However, in most cases of ARS, (Figure 5.4). The highest incidence of chromosomal
no such cause can be identified. Therefore, most anomalies was found in the 198 conceptuses with
authors consider ARS to be a sporadic event with a combined developmental defects. In this subgroup,
negligible risk of recurrence. cytogenetic evaluation was successful in 193 cases
(97.3%). Chromosomal abnormalities were found
in 166 cases (86%; Table 5.2). Of the 237 grossly
ETIOLOGY OF DEVELOPMENTAL DEFECTS disorganized embryos, 225 (95%) were analyzed
IN EARLY MISSED ABORTIONS cytogenetically. Of these, 156 (69.3%) were cytoge-
netically abnormal. The lowest incidence of chro-
Table 5.2 provides a general description of 514 cases mosomal abnormalies was found in phenotypically
studied by transcervical embryoscopy. The correlations normal specimens and in specimens with isolated
Table 5.2 Specimen morphology and karyotype of 514 missed abortions
Total specimens Specimens with abnormal

Total specimens successfully karyotyped karyoytype
Morphology No. %a No. %b No. %c
Normal 58 11.3 56 96.2 23 41.1

Growth disorganization 237 46.1 225 95 156 69.3
Combined defects 198 38.5 193 97.3 166 86.0
Isolated defects 21 4.1 21 100 14 66.7
Total 514 100 495 96.3 359 72.5
a
Percentage of total number of specimens with that morphology.
b
Percentage of each morphological category successfully karyotyped.
c
Percentage of each morphological category with an abnormal karyotype.
74
0415421306-Ch05 3/29/07 5:34 PM Page 75
Table 5.3 Summary of cytogenetic findings
No. of external embryonic Combined

Karyotype abnormalities Growth disorganization developmental defects Isolated
46,XY/46,XX 33 69 27 7
Trisomy 2 9
Trisomy 3 3
Trisomy 4 3 4
Trisomy 5 1
Trisomy 6 4
Trisomy 7 2 1 1
Trisomy 8 8 2
Trisomy 9 8
Trisomy 10 5
Trisomy 11 1
Trisomy 12 2
Trisomy 13 1 6 1
Trisomy 14 1 9
Trisomy 15 3 21
Trisomy 16 65
Trisomy 17 1
Trisomy 18 1 3 3
Trisomy 20 1 4
Trisomy 21 16 1 7
Trisomy 22 18 14
Triploidy 2 3 19 4
Tetraploidy 6 3
45,X 1 58 8
Structural defect 2 3 9
Other 10 1
No cytogenetic results 2 12 5
availabale
Total 58 237 198 21
This series comprises 58 embryos with normal external features, 237 growth-disorganized embryos, 198 embryos with combined localized
developmental defects, and 121 embryos with isolated localized developmental defects.
defects (Table 5.2). Of 58 cases with normal exter- pregnancies. They originate ‘de novo’ either in
nal features, 56 could be analyzed cytogenetically, gametogenesis (trisomy and monosomy) or from
with 23 cases (41.1%) showing cytogenetically polyspermic fertilization or failure of normal
abnormal results. Of 21 specimens with isolated cleavage (triploidy and tetraploidy). Therefore, all
defects, 14(66.7%) showed chromosomal abnor- embryoscopic findings should be supplemented by
malities. the results of cytogenetic analysis to distinguish
In summary, aneuploidy/polyploidy is the major between non-chromosomal and chromosomal
factor affecting normal embryonic development in causes of anomalies. Aneuploidy/polyploidy pro-
early intrauterine deaths, and may explain why vides a causal explanation for these developmental
spontaneous abortion is usually a sporadic event in defects in cases of phenotypically abnormal
a patient’s reproductive history although the inci- embryos, and also indicates that the recurrence
dence of developmental defects is high. Most (95%) risk for the observed developmental defect and
of the observed chromosomal mutations are not chromosomal abnormality in these couples is not
hereditary and carry no increased risk for future increased.41
75
0415421306-Ch05 3/29/07 5:34 PM Page 76
and may be caused be a single gene defect or

CLINICAL SIGNIFICANCE AND IMPLICATIONS
OF DETAILED MORPHOLOGICAL AND non-genetic mechanisms such as amniotic bands,
CYTOGENETIC EVALUATION OF EARLY duplication anomalies, vascular disruptions, etc. An
SPONTANEOUS ABORTION accurate description of these specimens is essential,
as it helps identify the specific mechanism leading
A detailed embryoscopic examination of the dead to the defect. This information would be completely
embryo is likely to be useful in couples who have lost if morphological examination of the demised
experienced recurrent abortion or have reproduc- embryo had not been carried out and the particular
tive loss after in vitro fertilization (IVF).14 Table 5.4 developmental defects remained undetected.
provides a general description of embryoscopic and The recurrence rate of these defects differs
cytogenetic findings of 53 patients with recurrent according to the etiology. If the observed defects are
miscarriages (three or more consecutive miscar- multifactorial in origin, then the risk of recurrence
riages). Of these cases, 32 of 50 (64%) had an is approximately 2–5%. The recurrence rate may be
abnormal embryonic karyotype. Fourteen embryos much higher for autosomal dominant or recessive
had a morphological defect with a normal kary- genes, or not significantly increased if non-genetic
otype, while no embryonic or chromosomal abnor- mechanisms (amniotic bands, duplication anom-
mality could be diagnosed in four cases. The Royal alies, or vascular disruptions) are responsible for
College of Obstetricians and Gynaecologists42 rec- abnormal embryonic development. Multiple local-
ommends fetal karyotyping in the investigation of ized developmental defects without a chromosomal
recurrent miscarriage. The value of karyotyping anomaly are rare, and may indicate a single-gene
early abortion specimens is limited by frequent defect.28 In these cases, a high recurrence rate
false-negative results caused by maternal contami- cannot be excluded. Diagnosis of a specific syn-
nation. A 46,XX karyotype in the curettage material drome is usually not possible at these early stages.
is therefore not a reliable result.12 In missed abor- Expert first-trimester ultrasonographic exami-
tions, transcervical embryoscopy allows selective nation has become standard management for
and reliable sampling of chorionic tissue with min- women at increased risk of hereditary conditions.
imal maternal contamination.13 In addition, uterine An accurate description of specific developmental
malformations can be diagnosed at the same time. defects by embryoscopy complements and aids
Isolated or combined localized developmental early prenatal ultrasonographic examination in
defects with an apparently normal karyotype might excluding a recurrence in subsequent pregnancies
be heterogeneous or multifactorial in origin, reaching the second trimester.
Table 5.4 Morphology and karyotype in 53 patients with recurrent miscarriage
Total specimens Specimens with abnormal

Total specimens successfully karyotyped karyotype
Morphology No. %a No. %b No. %c
Normal 8 15.1 7 87.5 3 42.9

Growth disorganization 26 49.1 24 92.3 15 62.5
Combined defects 18 34 18 100 13 72.2
Isolated defects 1 1.9 1 100 1 100
Total 53 100 50 94.3 32 64
a
Percentage of embryos with the specific morphology.
b
Percentage of each morphological category successfully karyotyped.
c
Percentage of each morphological category with an abnormal karyotype.
76
0415421306-Ch05 3/29/07 5:34 PM Page 77
If single-gene defects exist in chromosomally REFERENCES

normal abortions with multiple localized develop-
mental defects, this would explain why a normal 1. Salat-Baroux J. Recurrent spontaneous abortions. Reprod Nutr Dev
1988; 28:1555–68.
karyotype is usually interpreted as a poor prognos- 2. Kalousek DK, PantzarT, Tsai M, et al. Early spontaneous abortion:
tic sign in early-abortion specimens.43,44 The morphologic and karyotypic findings in 3912 cases. Birth Defects
embryoscopic diagnosis of embryonic growth 1993; 29:53–61.
3. Warburton D, Byrne J, Canki N. Chromosome Anomalies and
disorganization is less informative. Chromosomal Prenatal Development: An Atlas. New York. Oxford University Press,
abnormalities can be found in 70% of cases, with 1991.
autosomal trisomies forming 92% of the 70%.16 4. Kalousek DK. Anatomical and chromosomal abnormalities in speci-
mens of early spontaneous abortions: seven years experience. Birth
Most of the chromosomal abnormalities found Defects 1987; 23:153–68.
in growth-disorganized embryos are non-viable 5. Philipp T, Kalousek DK. Transcervical embryoscopy in missed abor-
(Table 5.3), and their presence explains the minimal tion. J Assist Reprod Genet 2001; 18:285–90.
6. Philipp T, Philipp K, Reiner A, Beer F, Kalousek DK. Embryoscopic
embryonic development observed embryoscopically. and cytogenetic analysis of 233 missed abortions: factors involved in
Embryonic growth disorganization and a normal the pathogenesis of developmental defects of early failed pregnancies.
karyotype, which has a similar embryoscopic appear- Hum Reprod 2003; 18:1724–32.
7. Shiota K. Development and intrauterine fate of normal and abnormal
ance to growth disorganization resulting from an human conceptuses. Cong Anom 1991; 31:67–80.
aneuploidy/polyploidy, suggests that some cases of 8. Moore KL. The Developing Human – Clinically Orientated
growth disorganization may be genetic in origin, but Embryology, 5th edn. Philadelphia: WB Saunders Co., 1993.
9. Philipp T. Atlas der Embryologie. Embryoskopische Aufnahmen der
undetectable by current cytogenetic techniques. normalen und abnormen Embryonalentwicklung. Vienna: Facultas
Routine cytogenetic analysis of the abortus is ham- Verlag, 2004.
pered by contamination of the culture with maternal 10. Wolf GC, Horger EO. Indication for examination of spontaneous
abortion specimens: a reassessment. Am J Obstet Gynecol 1995;
tissues and by the limits of chromosome resolution. 5:1364–7.
Submicroscopic chromosomal rearrangements have 11. Philipp T, Kalousek DK. Neural tube defects in missed abortions –
only recently been considered to be etiologically related embryoscopic and cytogenetic findings. Am J Med Genet 2002;
107:52–7.
to pregnancy loss.45,46 The presence of submicroscopic 12. Bell KA, Van Deerlin PG, Haddad BR, et al. Cytogenetic diagnosis of
chromosomal rearrangements challenges the pre- ‘normal 46,XX’ karyotypes in spontaneous abortions frequently may
vailing assumption that if no routine laboratory test be misleading. Fertil Steril 1999; 71:334–41.
13. Ferro J, Martinez MC, Lara C, et al. Improved accuracy of hysteroem-
confirms the presence of a genetic disorder, then bryoscopic biopsies for karyotyping early missed abortions. Fertil
non-genetic causes should be sought. The advent of Steril 2003; 80:1260–4.
whole-genome screening technologies has introduced 14. Philipp T, Feichtinger W, Van Allen M, et al. Abnormal embryonic
development diagnosed embryoscopically in early intrauterine deaths
new opportunities for screening chromosomally after in vitro fertilization (IVF): a preliminary report of 23 cases.
normal pregnancy losses with developmental defects Fertil Steril 2004; 82:1337–42.
for previously undetectable submicroscopic chromo- 15. Poland BJ, Miller JR, Harris M, et al. Spontaneous abortion: a study of
1961 women and their conceptuses. Acta Obstet Gynecol Scand 1981;
somal abnormalities. Embryoscopy identifies subtle 102(Suppl):5–32.
morphological abnormalities undetectable by ultra- 16. Philipp T, Kalousek DK. Generalized abnormal embryonic develop-
sound, and may identify a highly characterized cohort ment in missed abortion: embryoscopic and cytogenetic findings. Am
J Med Genet 2002; 111:41–7.
of abortion specimens with apparently normal chro- 17. Mc Fadden DE, Kalousek DK. Survey of neural tube defects in spon-
mosomes as a starting point for further detailed genetic taneously aborted embryos. Am J Med Genet 1989; 32:356–8.
studies. Such studies are required in order to reach 18. Bell JE, Gosden CM. Central nervous system abnormalities –
contrasting patterns in early and late pregnancy. Clin Genet 1978;
a better understanding of embryopathogenesis, and 13:387–96.
consequently of early and recurrent pregnancy loss. 19. Coerdt W, Miller K, Holzgreve W, et al. Neural tube defects in chro-
mosomally normal and abnormal human embryos. Ultrasound
Obstet Gynecol 1997; 10:410–15.
ACKNOWLEDGMENT 20. Creasy MR, Alberman ED. Congenital malformations of the central
nervous system in spontaneous abortions. J Med Genet 1976;
I am grateful to a wonderful embryopathologist and 13:9–16.
21. Philipp T, Grillenberger K, Separovic ER, Philipp K, Kalousek DK.
teacher, who introduced me to embryopathology, Effects of triploidy on early human development. Prenat Diagn 2004;
Prof. Dr DK Kalousek. 242:276–81.
77
0415421306-Ch05 3/29/07 5:34 PM Page 78
22. Canki N, Warburton D, Byrne J. Morphological characteristics of 36. Hong CY, Simon MA. Amniotic bands knotted about umbilical cord.
monosomy X in spontaneous abortions. Ann Genet 1988; 31:4–13. A rare cause of fetal death. Obstet Gynecol 1963; 222:667–70.
23. Patten BM. Overgrowth of the neural tube in young human embryos. 37. Philipp T, Kalousek DK. Amnion rupture sequence in a first trimester
Anat Rec 1952; 113:381–93. missed abortion. Prenat Diagn 2001; 21:835–8.
24. Shaw A. The myth of gastroschisis. J Pediatr Surg 1975; 10:235–44. 38. Ossipoff V, Hall BD. Etiologic factors in the amniotic band syndrome.
25. De Vries PA. The pathogenesis of gastroschisis and omphalocele. A study of 24 patients. Birth Defects 1977; 13:117–32.
J Pediatr Surg 1980; 15:245–51. 39. Firth HV, Boyd PA, Chamberlain P, et al. Severe limb abnormalities
26. Hoyme H, Higginbottom MC, Jones KL. The vascular pathogenesis of after chorion villus sampling at 56–66 days gestation. Lancet 1991;
gastrochisis: intrauterine interruption of the omphalomesenteric 337:762–3.
artery. J Pediatr 1981; 98:228–3l. 40. Young ID, Lindenbaum RH, Thompsen EM, Pemburg ME.
27. Ramsing M, Duda V, Mehrain Y, et al. Hand malformations in the Amniotic bands in connective tissue disorders. Arch Dis Child 1985;
aborted embryo: An informative source of genetic information. Birth 60:1061–3.
Defects 1996; 30:79–94. 41. Warburton D, Kline J, Stein Z, et al. Does the karyotype of a sponta-
28. Dimmick JE, Kalousek DK. Developmental Pathology of the Embryo neous abortion predict the karyotype of a subsequent abortion?
and Fetus. Philadelphia: JB Lippincott, 1992. Evidence from 273 women with two karyotyped spontaneous abor-
29. Birch-Jensen A. Congenital Deformities of Upper Extremities. tions. Am J Hum Genet 1987; 41:465–83.
Copenhagen: Munksgaard, 1949. 42. Royal College of Obstetricians and Gynaecologists, Guideline No. 17.
30. Golden CM, Ryan LM, Holmes LB. Chorionic villus sampling: a dis- The Management of Recurrent Miscarriage. London: RCOG, 2003.
tinctive teratogenic effect on fingers. Birth Defects Res 2003; 67:557–62. 43. Osagawara M, Aoki K, Okada S, et al. Embryonic karyotype of abor-
31. Philipp T, Separovic ER, Philipp K, et al. Trancervical fetoscopic diag- tuses in relation to the number of previous miscarriages. Fertil Steril
nosis of structural defects in four first trimester monochorionic twin 2000; 73:300–4.
intrauterine deaths. Prenat Diagn 2003 ; 12:964–9. 44. Stephenson M, Awartani KA, Robinson WP. Cytogenetic analysis of
32. Napolitani FD, Schreiber I. The acardiac monster. A review of the miscarriages from couples with recurrent miscarriage: a case control
world literature and presentation of two cases. Am J Obstet Gynecol study. Hum Reprod 2002; 17:446–51.
1960; 82:708–11. 45. Schaeffer AJ, Chung J, Heretis K, et al. Comparative genomic
33. Evans MI. Amniotic bands. Ultrasound Obstet Gynecol 1997; 10:307–8. hybridization-array analysis enhances the detection of aneuploidies
34. Torpin R. Amniochorionic mesoblastic fibrous strings and amniotic and submicroscopic imbalances in spontaneous miscarriages. Am J
bands. Associated constricting fetal anomalies or fetal death. Am Med Genet 2004; 6:1168–74.
J Obstet Gynecol 1965; 91:65–75. 46. Le Caignec C, Boceno M, Saugier-Veber P, et al. Detection of genomic
35. Kalousek DK, Bamforth S. Amnion rupture sequence in previable imbalances by array based comparative genomic hybridisation in
fetuses. Am J Med Genet 1988; 3:63–73. fetuses with multiple malformations. J Med Genet 2005; 2:121–8.
78
0415421306-Ch06 3/29/07 5:35 PM Page 79
6. Endocrinology of pregnancy loss

Stefano Luisi, Lucia Lazzeri, and Andrea Riccardo Genazzani
INTRODUCTION secreted by the trophoblast, the corpus luteum

receives a signal to continue producing 17α-proges-
Following implantation, the maintenance of preg- terone along with estradiol, estrone, and relaxin.
nancy is dependent on a multitude of endocrinolog- Relaxin is a member of the insulin-like growth
ical events that will eventually aid in the successful factor family, and its increase parallels that of hCG,
growth and development of the fetus. Although the suggesting a synergistic effect with progesterone in
great majority of pregnant women have no preexist- maintaining myometrial quiescence.1,2
ing endocrine abnormalities, a small number can The corpus luteum maintains its capacity to
have certain endocrine alterations that could poten- synthesize progesterone almost throughout preg-
tially lead to recurrent pregnancy losses. nancy, but at approximately 7 weeks of gestation, its
It is estimated that approximately 8–12% of all functional ability decreases markedly, at the start
pregnancy losses are the result of endocrine factors. of the luteoplacental transition. Excission of the
During the preimplantation period, the uterus corpus luteum before the 8th week of gestation
undergoes important developmental changes stim- results in abortion, whereas after the 9th week it
ulated by estrogen and, more importantly, proges- does not.3
terone. Progesterone is essential for successful Abnormalities of the luteal phase have been
implantation and maintenance of pregnancy. reported to occur in up to 35% of women with
Therefore, disorders related to inadequate proges- recurrent pregnancy loss (RPL).4 There are several
terone secretion by the corpus luteum are likely to causes for LPD, including stress, exercise, weight
affect the outcome of the pregnancy. Luteal phase loss, hyperprolactinemia, and menstrual cycles at
deficiency (LPD), hyperprolactinemia, and polycys- the onset of puberty or perimenopause. The mech-
tic ovary syndrome (PCOS) are some examples. anisms by which LPD occurs are unclear, but
Several other endocrinological abnormalities, such could be associated with decreased progesterone
as thyroid disease, hypoparathyroidism, uncon- production by the corpus luteum, decreased follicle-
trolled diabetes, and decreased ovarian reserve, have stimulating hormone (FSH) levels in the follicular
been implicated as etiologic factors for recurrent phase, abnormal pattern of luteinizing hormone
pregnancy loss. Inhibins and activins are non- (LH) secretion, and a decreased response to proges-
steroidal glycoproteins thought to have important terone by the endometrium. Many studies suggest
roles in reproductive physiology, and have been that LPD originates as a preovulatory event. LPD
proposed as markers of fetal viability. may be secondary to abnormal follicle formation,
associated with poor oocyte quality.5,6
The methods used to diagnose LPD are contro-
LUTEAL PHASE DEFICIENCY AND versial and not universally accepted. For the diagnosis
PREGNANCY LOSS of LPD, both serum progesterone and the use of
endometrial biopsy are advocated.7,8 Serum proges-
Progesterone secreted by the corpus luteum plays a terone levels greater than 10 ng/ml in the midluteal
paramount role in the maintenance of early preg- phase are rarely associated with an abnormal
nancy. Immediately after implantation, under the luteal phase when a targeted endometrial biopsy is
influence of human chorionic gonadotropin (hCG) performed.9,10 The endometrium is considered to
79
0415421306-Ch06 3/29/07 5:35 PM Page 80
be out of phase when the histological dating lags evidence for a positive effect on the maintenance of
behind the menstrual dating by 2 days or more; as pregnancy.15
determined from the subsequent day of menses,
the diagnosis requires endometrial biopsies in a
minimum of two cycles. Endometrial biopsy, with HYPERPROLACTINEMIA AND PREGNANCY LOSS
evaluation of the morphological changes, has
been considered superior to serum progesterone The role of prolactin (PRL) in human ovarian
determinations because of the pulsatile nature of steroidogenesis remains unclear. PRL may partici-
progesterone secretion. More importantly, the pate in the regulation of ovarian steroidogenesis;
morphological changes in the endometrium better but, in pathological states with high PRL levels, it
represent the cumulative effect of cycle-specific may interfere with corpus luteum function. In vitro
patterns of corpus luteum function.11 Despite the studies have revealed that progesterone secretion by
above-mentioned rationale, there is considerable cultured granulosa cells obtained from human
inter- and intraobserver variation in the interpreta- ovarian follicles is almost completely inhibited by
tion of the endometrial biopsies. high PRL concentrations (100 ng/ml), but not by
Epidemiological studies of RPL appear to sup- lower concentrations (10–20 ng/ml).16 These
port the concept that LPD is in fact an etiological observations suggest the possibility that high PRL
factor. This is documented by studies demonstrat- concentrations in the early phase of follicular
ing that hormone treatment to enhance proges- growth may inhibit progesterone secretion, resulting
terone production or supplementation is associated in luteal phase defects. The precise cellular mecha-
with an increased chance of a term pregnancy in nism of PRL action in the human ovary remains
women with RPL.12 to be elucidated; furthermore, a recent study of
The treatment of LPD may include the use of 64 hyperprolactinemic women with RPL treated
ovulation-induction agents such as clomiphene with bromocriptine was associated with a higher
citrate, alone or in combination with gonadotropins. rate of successful pregnancy, and PRL levels were
These agents enhance progesterone secretion by significantly higher in women who miscarried.17
creating more than one follicle, resulting in multi- In conclusion, normal PRL levels may play an
ple sites for progesterone production; 5000 or important role in the growth and maintenance of
10 000 units of hCG injected at the time of expected early pregnancy.
ovulation have also been used. Progesterone supple-
mentation after ovulation, with or without the
use of ovulation-induction agents, can also be used THYROID ABNORMALITIES AND
2–3 days after the basal body temperature increases PREGNANCY LOSS
(or after a positive urinary LH test) and continued
up to 7–11 weeks of gestation.13 HYPERTHYROIDISM
Progesterone supplementation can be adminis-
tered by intravaginal suppositories (25–100 mg Excess production of thyroid hormone is not usu-
twice daily), by intramuscular injection of proges- ally correlated with infertility or RPL. Women with
terone in oil (50 mg daily), or as oral micronized subclinical or mild hyperthyroidism have evidence
progesterone (100 mg two or three times daily). of ovulation when endometrial sampling is per-
Bromocriptine or cabergoline are recommended formed. Although hyperthyroidism has been asso-
to treat luteal phase deficiency associated with ciated with poor pregnancy outcomes, including
hyperprolactinemia.14 The subject of LPD and its preterm delivery, abruptio placentae, maternal
association with RPL continues to be controversial. heart failure, fetal growth restriction, and still-
A meta-analysis of randomized trials of pregnancies birth,18,19 it has not been reported commonly as an
treated with progestational agents failed to find any independent cause of RPL. RPL has been associated
80
0415421306-Ch06 3/29/07 5:35 PM Page 81
ENDOCRINOLOGY OF PREGNANCY LOSS
with Graves disease in a case report;20 however, the exists between low thyroid function and pregnancy
patient was also diagnosed with antiphospholipid loss, direct evidence for a causal role is missing.25
syndrome, which is an independent cause of RPL. One postulated explanation for this relationship is
However, a recent retrospective study21 has sug- that luteal phase defects have been linked to thyroid
gested that excess exogenous thyroid hormone is hypofunction. Given that the production of proges-
associated with an elevated rate of fetal loss. This terone is a pivotal element of a successful preg-
study was performed in a unique population of nancy, it is conceivable that RPL could be related to
patients with a genotype (Arg243Gln mutation in deficient corpus luteum action. Given the method-
the thyroid hormone receptor β gene) showing ological flaws in the diagnosis of luteal phase
resistance to thyroid hormone, and a high serum defects,26,27 and the lack of evidence to recommend
concentration of free thyroxine and triiodothyro- the use of exogenous progesterone supplementation
nine without suppressed thyrotropin. These women in early gestation,28 their existence is controversial
maintain a euthyroid state despite high thyroid at best.
hormone levels. Patients were analyzed in three We believe that it is prudent to screen for thyroid
different groups: affected mothers (n = 9), affected disease and normalize thyroid function prior to
fathers (n = 9), and unaffected relatives (n = 18). conception when function is found to be abnormal.
The mean miscarriage rates were 22.9, 2.0, and 4.4%, Even if there is no clear cause–effect relationship
respectively (χ 2 = 8.66; p = 0.01). Affected mothers between hypothyroidism and RPL, there is some
had an increased rate of miscarriage (z = 3.10; evidence that subclinical hypothyroidism is corre-
p = 0.002, by Wilcoxon rank-sum test).21 lated with poor maternal outcome as well as prema-
turity and reduced intelligence quotient in the
HYPOTHYROIDISM offspring.24 There is disagreement as to the suitable
upper limit of normal serum thyroid-stimulating
The causes of thyroid dysfunction include autoim- hormone (TSH) in order to make the diagnosis of
munity, severe iodine deficiency, postpartum thy- subclinical hypothyroidism. There is a trend with
roiditis, and drug-induced hypothyroidism as a the new TSH assays to decrease the upper limit of
result of radical treatment of hyperthyroidism. the normal TSH range (4.5–5.0 mU/l) to 2.5 mU/l.
Autoimmune thyroid disease is the most prevalent This upper limit is recommended by the National
associated etiology in patients of reproductive age. Academy of Clinical Biochemistry guidelines, and is
There seems to be no doubt that hypoactive thyroid based on the fact that 2.5 mU/l represents more
hormone is associated with infertility. than two standard deviations above the level in
Thyroid hormones have an impact on oocytes at meticulously screened euthyroid volunteers.29
the level of the granulosa and luteal cells that inter- Clearly, this new upper limit will significantly
fere with normal ovulation.22 Low thyroxine levels increase the number of patients diagnosed with
have a positive feedback on thyrotropin-releasing subclinical hypothyroidism, and its clinical benefit
hormone (TRH). Elevations in TRH have been remains questionable.
associated with PRL elevation.23 It is believed that
elevated PRL alters the pulsatility of gonadotropin- ANTITHYROID ANTIBODIES
releasing hormone (GnRH) and interferes with
normal ovulation. Thyroid peroxidase and thyroglobulin are two key
Therefore, severe forms of hypothyroidism players in the biosynthesis of thyroid hormone.
rarely complicate pregnancy, because they are asso- The former is an enzyme involved in the coupling
ciated with anovulation and infertility. However, in of iodine residues and in the iodination of thyroid
mild hypothyroidism, pregnancies can occur, and residues. The role of thyroglobulin is in the stor-
are associated with higher rates of pregnancy loss and age and synthesis of thyroid hormones. Both can
maternal complications.24 Even if an association be involved in auto-antigenicity in thyroid
81
0415421306-Ch06 3/29/07 5:35 PM Page 82
immune disease.30 Although patients with RPL have ATA is an attempt to minimize the transmission of
a higher prevalence of antithyroid antibodies (ATA: autoimmune genes to the offspring.36
158 of 700 (22.5%)) when compared with controls A recent evidence-based review37 of investiga-
(29 of 200 (14.5%)), they are found with equal tions and treatments of RPL concluded that the
frequency in patients undergoing artificial repro- association between ATA and RPL is doubtful. This
ductive technology (ART: 162 of 688 (19.2%)).31 In report also found it doubtful that ATA is a cause
theory, it is feasible that patients with ATA have of RPL.
inferior reproductive outcomes; however, the relation- The above data may suggest that the association
ship between RPL and ATA remains specu1ative. In between ATA and RPL represents an immune disor-
a prospective cohort study, 134 RPL patients who der rather than an endocrinological abnormality.
were positive for ATA and had normal thyroid Nevertheless, patients with positive ATA and euthy-
functions were compared with 710 patients with roidism are at an increased risk for hypothyroidism
RPL but negative for ATA. No treatment was during and after pregnancy. These individuals should
offered to either group, and both groups yielded a have their TSH screened during each trimester of
live birth rate of 58%.32 The above findings suggest pregnancy and postpartum for thyroiditis.
that in the presence of normal thyroid function,
ATA have no impact on pregnancy outcome. How
can we explain the association between RPL DIABETES MELLITUS AND PREGNANCY LOSS
and ATA? One possibility is that positive ATAs coin-
cide with other autoimmune disorders, suggesting Patients with type 1 (‘insulin-dependent’) diabetes
an immunoregulatory dysfunction as the underly- mellitus (DM) with normal or near-normal
ing etiology of the reproductive losses. Up to glycemic control are not at higher risk for RPL.
45.5% of patients with active svstemic lupus erythe- A well-designed prospective cohort study by the
matosus have positive ATAs. Another possibility is National Institute of Child Health and Human
that ATA are markers of poor ovarian reserve and Development included 386 women with type 1 DM
signal a diminished oocyte cohort. Up to one-third and 432 women without diabetes before or within
of patients with premature ovarian failure have 21 days after conception and observed both groups
been reported to be ATA-positive,33 whereas only prospectively.38 The rate of spontaneous abortion
10% of the adult female population carry these was similar in both groups when the mothers had
antibodies.34 adequate glycemic control. In this same study, the
Given that reproductive failure cannot be efficacy of strict glycemic control was demon-
improved significantly by egg donation, another strated. For the subgroup of patients who had
interesting explanation may lie within the endo- elevated values for glycosylated hemoglobin in the
metrium of these patients. Activated T cells in the first trimester, each increase of one standard devia-
endometrium of patients with ATA may produce tion above the normal range was associated with an
cytokines disrupting implantation. Stewart-Akers increase of 3.1% in the rate of pregnancy loss (95%
et al35 found a significant increase in the endome- confidence interval (CI), 0.6–5.6). The importance
trial T-cell population in women with ATA com- of glycemic control in the period of conception and
pared with that of controls. The relative abundance early pregnancy cannot be overemphasized. There
of interleukin 4 (lL-4) and IL-10 was decreased in is a direct correlation between the glycosylated
women with ATA compared with controls, whereas hemoglobin (HbA1c) values and the incidence of
interferon-γ (IFN-γ) was increased. Stewart-Akers pregnancy loss and congenital malformations. In a
et al35 found that the source of cytokine production study of 303 pregnant type 1 DM patients, the
for IL-4, IL-10, and IFN-γ was endometrial leuko- risk of spontaneous abortion was found to be
cytes. Finally, a teleological explanation proposes 12.4% with first-trimester HbA1c ≤ 9.3%, and
that the RPL in women with elevated titers of 37.5% with HbA1c > 14.4% (risk ratio (RR) 3.0;
82
0415421306-Ch06 3/29/07 5:35 PM Page 83
95% CI 1.3–7.0). The risk for major malformations insulinemia and PAI-Fx are reversed to those of
was 3.0% with HbA1c ≤ 9.3% and 40% with normal controls.46 Others have suggested that the
HbAlc > 14.4% (RR 13.2; 95% CI 4.3–40.4).39 The association between PCOS and early pregnancy loss
improvements in pregnancy outcome due to may be due to increased LH, which stimulates a
glycemic control have been confirmed by a well- highly androgenic milieu and affects reproductive
conducted meta-analysis40 of 14 studies comparing outcome.47,48
the HbA1c of diabetic pregnant women receiving There are several studies confirming the findings
preconception care compared with diabetic patients of LH hypersecretion in RPL.49 One study found
without care. The risk of malformations was 2.1% that over 80% of patients with RPL had polycystic
and 6.5%, respectively (RR 0.36; 95% CI 0.22–0.59). ovaries, and that 81% of the PCOS women had
The increased rate of RPL in poorly controlled dia- elevated LH levels. The relevance of LH hypersecre-
betic patients may be due to several pathophysio- tion in RPL is controversial.50 Interestingly, extreme
logical processes, such as maternal vascular disease suppression of LH has been linked to a high miscar-
including uteroplacental deficiency and perhaps riage rate in ART, but no cause-and-effect relation-
immunological causes. ship has been studied.51 Moreover, a randomized
In summary, poorly controlled diabetes mellitus controlled trial52 in which 106 patients with RPL,
is associated with more frequent pregnancy losses, PCOS, and elevated LH were allocated preconcep-
and adequate control will normalize the rate of mis- tion to either receive GnRH to suppress LH, or
carriages. To prevent most of the pregnancy losses allowed to conceive naturally, failed to show any
in diabetic patients, it is recommended to have a advantage in the treatment arm in terms of either
prepregnancy level of HbAlc ≤ 7.5%.41 However, in pregnancy rates or miscarriage rates.
addition to miscarriages, other adverse pregnancy Hyperinsulinemia and insulin resistance have
outcomes can be avoided if the glycosylated hemo- been claimed to be a potential cause of the high
globin is less than 6.6%.42 rate of RPL in patients with PCOS and have been
linked to many of the metabolic and endocrine
abnormalities potentially associated with the phys-
POLYCYSTIC OVARY SYNDROME, INSULIN iopathology of RPL. The prevalence of insulin
RESISTANCE, AND PREGNANCY LOSS resistance, as diagnosed by fasting insulin and glu-
cose measurements using glucose-to-insulin ratios
Patients with PCOS have an increased rate of and HOMA (homeostasis model assessment) scores,
pregnancy loss (> 50%), predominantly during has been found approximately three times higher
early gestation.43 The underlying causes of RPL in in an unselected population of women with RPL
PCOS patients may involve several contributing when compared with an age-, race-, and body mass
and interrelated factors. These include obesity, index-matched group of controls.53 Metformin
hyperinsulinemia, hyperandrogenemia, insulin bas been the main treatment advocated to reduce
resistance, poor endometrial receptivity, and ele- RPL in PCOS patients. Although some studies
vated levels of LH. have demonstrated promising results with met-
One of the working hypotheses for the associa- formin, it is essential to have a well-randomized
tion between PCOS and pregnancy loss is that controlled trial with appropriate power before
PCOS causes hypofibrinolysis modulated by the metformin can be recommended as effective ther-
4G4G mutation of the plasminogen activator apy. Given that metformin is a pregnancy category
inhibitor 1 gene, as well as elevated levels of its B drug, and clinical practice is always ahead of
products, plasminogen activator inhibitor activity evidence-based practice, some investigators con-
(PAI-Fx).44 Furthermore, in uncontrolled trials, sider that the potential benefits of metformin out-
PCOS patients given metformin during pregnancy weigh its risks and advocate its use to reduce
exhibit a reduced rate of early losses45 and the high miscarriages.54
83
0415421306-Ch06 3/29/07 5:35 PM Page 84
to a highly homologous βA subunit. Inhibin B is a

ELEVATED FSH AND PREGNANCY LOSS similar dimeric glycoprotein with α and βB subunits.
Non-bioactive forms of the α subunit include the
The underlying challenge present in certain women amino-terminally extended product named inhibin
with unexplained RPL may reside in the quality and pro-αC. Inhibin A is the major circulating bioactive
quantity of their oocytes. In a retrospective com- inhibin found in early pregnancy. Inhibin B is not
parative analysis, Trout and Seifer55 measured FSH detectable in early pregnancy in the human.58
levels on day 3 of the cycle and estradiol (E2) in 36 Although the major function of inhibins is in the
patients with unexplained RPL and in 21 control negative feedback control of gonadotropin secre-
RPL patients with a known etiology. These finding tion, their function in pregnancy may possibly be
were reproduced subsequently in a similar analysis the promotion and modulation of placental secre-
of 58 patients with RPL and unknown etiology.56 tary activity and placental immune modulation.
Women with unexplained RPL were found to be Circulating levels of inhibin A and pro-αC have
more likely to have abnormal ovarian reserve. been implicated in the process of implantation and
Among the 36 patients with unexplained RPL, 11 early pregnancy development.59 Inhibins have also
(31%) had elevated FSH, 14 (39%) had elevated E2, been proposed as markers of fetal viability. In the
and 21 (58%) had abnormal results for one or both non-pregnant female, inhibins are secreted and syn-
tests, In the 20 patients of the control group, the thesized by both the developing graafian follicle and
findings were that 1 patient had elevated FSH (5%), corpus luteum.60,61 Inhibins are also involved in
3 (14%) had elevated E2, and 4 (19%) had abnor- the control of the fetomaternal communication
mal results for one or both tests. In a different study, required to maintain pregnancy. Human placenta,
Hofmann et al57 performed a clomiphene challenge decidua, and fetal membranes are the major sites of
test in 44 patients with RPL and found a similar production and secretion of inhibins A and B in
incidence of diminished ovarian reserve when com- maternal serum, amniotic fluid, and cord blood.62
pared with 648 general infertility patients. The corpus luteum has been shown to be the
Women with abnormal ovarian reserve testing major site of inhibin A production. Production of
and unexplained RPL may have a higher incidence inhibin A continues within the corpus luteum as
of chromosomal abnormalities in their miscar- pregnancy is established. During early pregnancy,
riages. Thus, we believe that it is prudent to incor- mRNA for inhibins α, βA and βB have been demon-
porate ovarian reserve testing in the workup of strated in the corpus luteum.63 However, inhibins
patients with RPL. However, we must acknowledge are also synthesized and secreted by the developing
that any assessment of ovarian reserve is not a diag- human placenta. Both α- and βA-subunit mRNAs
nostic test, but rather a screening tool. An abnormal and proteins64 have been localized in the human
test does not preclude the possibility of a live birth. placenta, the major expression being from the
Decreased ovarian reserve should not be presented syncytiotrophoblast.
to the patient as an absolute. Extensive counselling The local actions during placental growth and
is recommended, given that no treatment (other differentiation are mirrored by changes in the circu-
than egg donation) is available. lating levels of dimeric inhibins and inhibin pro-αC
as pregnancy progresses.65 Concentrations of
inhibin A in the circulation increase progressively in
INHIBINS AND PREGNANCY LOSS early pregnancy.66 There is a transient fall in circu-
lating concentration at approximately 12 weeks of
Inhibins are non-steroidal glycoproteins thought to gestation, followed by further increases in concen-
play important roles in reproductive physiology. tration from midgestation onwards.59 Studies
Inhibin A has a molecular weight of 34 kDa, and demonstrating lower levels of inhibin A in failing
comprises an α-subunit linked by a disulfide bond pregnancies have implicated inhibin A in the
84
0415421306-Ch06 3/29/07 5:35 PM Page 85
Serum inhibin A (MoM) 4
0
Healthy controls Ongoing Failing Incomplete Complete
pregnancy pregnancy
Threatened abortion Miscarriage
Figure 6.1 Maternal serum inhibin A levels in healthy pregnant women (control), patients with threatened abortion with ongoing
and failing pregnancy, and incomplete and complete miscarriage. Individual values are plotted (expressed as mean of mean) and
horizontal bars represent the group medians. *p < 0.05; **p < 0.001; ***p < 0.001 versus healthy controls and threatened abortion
with ongoing pregnancy. MoM is multiples of the median.
processes of successful implantation and early preg- significant correlation between serum concentrations
nancy development.67 of inhibin A and β-hCG (the degree of correlation
Recently, inhibin A concentrations have been varied according to the population group: normal
measured in the maternal circulation of healthy controls r = 0.55, sporadic miscarriage r = 0.79, and
spontaneously pregnant women progressing to recurrent miscarriage r = 0.66). In their study,
deliver a healthy term singleton baby, in patients Muttukrishna et al68 confirmed a statistically signif-
with missed abortion (either absence of fetal heart icant positive correlation between inhibin A and
activity or an anembryonic gestational sac) and β-hCG in women who had live births (r = 0.46,
with complete miscarriage (an empty uterus with a p = 0.4), but not in those who had a miscarriage.
history of passage of products of conception),62 in Given the small size of this and previous studies, it
order to ascertain whether inhibin A measurement is not possible at this stage to establish whether
might provide a rapid and useful marker of early serum inhibin A is a better marker than β-hCG, or
pregnancy viability, in comparison with hCG levels. whether combined inhibin A and β-hCG measure-
Patients with complete miscarriage had the lowest ment is superior to β-hCG alone. Further studies
hCG and inhibin A levels, followed by patients with are required to address these two questions.69
missed abortion, where the highest levels were seen
in ongoing pregnancies (Figure 6.1). REFERENCES
The potential value of inhibin A as a marker of
early pregnancy problems should be examined in 1. Szlachter N, O’Byrne E, Goldsmith L, et al. Myometrial inhibiting
conjunction with other established biochemical activity of relaxin containing extracts of human corpora lutea of
pregnancy. Am J Obstet Gynecol 1980; 136:584–6.
markers such as serum β-hCG, progesterone, and 2. Bigazzi M, Nardi E. Prolactin and relaxin: antagonism on the sponta-
glycodelin. Muttukrishna et al68 found a statistically neous motility of the uterus. J Clin Endocrinol Metab 1981; 53:665–7.
85
0415421306-Ch06 3/29/07 5:35 PM Page 86
3. Csapo AL, Pulkkinen MO, Wiest WG. Effects of lutectomy and prog- 25. Clifford K, Rai R, Watson H, et al. An informative protocol for the
esterone replacement in early pregnant patients. Am J Obstet Gynecol investigation of recurrent miscarriage: preliminary experience of 500
1973; 115:759–64. consecutive cases. Hum Reprod 1994; 9:1328–32.
4. Insler V. Corpus luteum defects. Curr Opin Obstet Gynecol 1992; 26. Coutifaris C, Myers ER, Guzick DS, et al. Histological dating of rimed
4:203–11. endometrial biopsy tissue is not related to fertility status. Fertil Steril
5. Tuckerman E, Laird SM, Stewart R, et al. Markers of endometrial 2004; 82:1264–72.
function in women with unexplained recurrent pregnancy loss. Hum 27. Kazer RR. Endometrial biopsy should be abandoned as a routine
Reprod 2004; 19:196–205. component of the infertility evaluation. Fertil Steril 2004; 82:1297–8.
6. Jacobs MH, Balash J, Gonzalez-Merlo JM. Endometrial cytosolic and 28. Oates-Whitehead RM, Haas DM, Carrier JA, et al. Progestogen for pre-
nuclear progesterone receptors in the luteal phase defect. J Clin venting miscarriage. Cochrane Database Syst Rev 2003; (4):CD003511.
Endocrinol Metab 1987; 64:472–8. 29. Baloch Z, Carayon P, Conte-Devolx, et al. Guidelines Committee,
7. Rosenfeld DL, Garcia CR. A comparison of endometrial histology National Academy of Clinical Biochemistry. Laboratory Medicine
with simultaneous plasma progesterone determinations in infertile Practice Guidelines. Laboratory support: for the diagnosis and moni-
women. Fertil Steril 1976; 27:1256–66. toring of thyroid disease. Thyroid 2003; 13:3–126.
8. Koninckx PR, Goddeeries PG, Lauweryns JM, et al. Accuracy of 30. Weetman AP. Autoimmune thyroid disease. Autoimmunity 2004;
endometrial biopsy dating in relation to the mid-cycle luteinizing 37:337–40.
hormone peak. Fertil Steril 1977; 28:443–5. 31. Kutteh WH, Yetman DL, Carr AC, et al. Increased prevalence of
9. Hensleigh PA, Fainstat T. Corpus luteum dysfunction: serum proges- antithyroid antibodies identified in women with recurrent pregnancy
terone levels in the diagnosis and assessment of therapy for recurrent loss but not in women undergoing assisted reproduction. Fertil Steril
and threatened abortion. Fertil Steril 1979; 32:396–400. 1999; 71:843–8.
10. Cumming DC, Honore LH, Scott JZ, et al. The late luteal phase in 32. Rushworth FH, Backos M, Rai R, et al. Prospective pregnancy out-
infertile women: comparison of simultaneous endometrial biopsy come in untreated recurrent miscarriers with thyroid autoantibodies.
and progesterone levels. Fertil Steril 1985; 43:715–19. Hum Reprod 2000; 15:1637–9.
11. Coutifaris C, Myers ER, Guzick DS, et al. Histological dating of timed 33. Santoro N. Research on the mechanisms of premature ovarian failure.
endometrial biopsy tissue is not related to fertility status. Fertil Steril J Soc Gynecol Investig 2001; 8:S10–12.
2004; 82:1264–72. 34. Hak AE, Pols HA, Visser TJ, et al. Subclinical hypothyroidism is an
12. Daya S. Efficacy of progesterone support for pregnancy in women independent risk factor for atherosclerosis and myocardial infarction
with recurrent miscarriage: a meta-analysis of controlled trials. Br J in elderly women: the Rotterdam Study. Ann Intern Med 2000;
Obstet Gynaecol 1989; 96:275–80. 132:270–8.
13. Karamardian LM, Grimes DA. Luteal phase deficiency effect of treat- 35. Stewart-Akers AM, Krasnow JS, Brekosky J, et al. Endometrial leuko-
ment on pregnancy rates. Am J Obstet Gynecol 1992; 167:1391–8. cytes are altered numerically and functionally in women with implan-
14. Roberts CP, Murphy AA. Endocrinopathies associated with recurrent tation defects. Am J Reprod Immunol 1998; 39:1–11.
pregnancy loss. Semin Reprod Med 2000; 18:357–62. 36. Matalon ST, Blank M, Ornoy A, et al. The association between anti-
15. Byrne JL, Ward K. Genetic factors in recurrent abortion. Clin Obstet thyroid antibodies and pregnancy loss. Am J Reprod Immunol 2001;
Gynecol 1994; 37:693–704. 45:72–7.
16. McNatty KP, Sawers RS. Relationship between the endocrine environ- 37. Christiansen OB, Nybo Andersen AM, Bosch E, et al. Evidence-based
ment within the graafian follicle and the subsequent secretion of investigations and treatments of recurrent pregnancy loss. Fertil Steril
progesterone by human granulosa cells in culture. J Endocrinol 1975; 2005; 83:821–39.
66:391–400. 38. Mills JL, Simpson JL, Driscoll SG, et al. Incidence of spontaneous
17. Hirahara F, Andoh N, Sawai K, et al. Hyperprolactinemic recurrent abortion among normal women and insulin-dependant diabetic
miscarriage and results of randomized bromocriptine treatment women whose pregnancies were identified within 21 days of conception.
trials. Fertil Steril 1998; 70:246–52. N Engl J Med 1988; 319:1617–23.
18. Kriplani A, Buckshee K, Bhargava VL, et al. Maternal and perinatal 39. Greene MF, Hare JW, Cloherty JP, et al. First-trimester hemoglobin Al
outcome in thyrotoxicosis complicating pregnancy. Eur J Obstet and risk for major malformation and spontaneous abortion in dia-
Gynecol Reprod Biol 1994; 54:159–63. betic pregnancy. Teratology 1989; 39:225–31.
19. Daniels GH. Thyroid disease and pregnancy: a clinical overview. 40. Ray JG, O’Brien TE, Chan WS, et al. Preconception care and the risk
Endocr Pract 1995; 1:287–301. of congenital anomalies in the offspring of women with diabetes mel-
20. Nakayama T, Yamamoto T, Kanmatsuse K, et al. Graves’ disease asso- litus: a meta-analysis. QJM 2001; 9:435–44.
ciated with anticardiolipin antibody positivity and acquired protein S 41. Temple R, Aldridge V, Greenwood R, et al. Association between out-
deficiency. Rheumatol Int 2003; 23:198–200. come of pregnancy and glycaemic control in early pregnancy in type
21. Anselmo J, Cao D, Karrison T, et al. Fetal loss associated with excess 1 diabetes: population based study. BMJ 2002; 325:1275–6.
thyroid hormone exposure. JAMA 2004; 292:691–5. 42. Nielsen GL, Sorensen HT, Nielsen PH, et al. Glycosylated hemoglobin
22. Wakim AN, Polizotto SL, Buffo MJ, et al. Thyroid hormones in as predictor of adverse fetal outcome in type 1 diabetic pregnancies.
human follicular fluid and thyroid hormone receptors in human Acta Diabetol 1997; 34:217–22.
granulosa cells. Fertil Steril 1993; 59:1187–90. 43. Glueck CJ, Wang P, Fontaine RN, et al. Pregnancy outcomes among
23. Steinberger E, Nader S, Rodriguez-Rigau L, et al. Prolactin response to women with polycystic ovary syndrome treated with metformin.
thyrotropin-releasing hormone in normoprolactinemic patients with Hum Reprod 2002; 17:2858–64.
ovulatory dysfunction and its use for selection of candidates for 44. Glueck CJ, Wang P, Fontaine RN, et al. Plasminogen activator
bromocriptine therapy. J Endocrinol Invest 1990; 13:637–42. inhibitor activity: an independent risk factor for the high miscarriage
24. Casey BM, Dashe JS, Wells CE, et al. Subclinical hypothyroidism and rate during pregnancy in women with polycystic ovary syndrome.
pregnancy outcomes. Obstet Gynecol 2005; 105:239–45. Metabolism 1999; 48:1589–95.
86
0415421306-Ch06 3/29/07 5:35 PM Page 87
45. Glueck CJ, Phillips H, Cameron D et al. Continuing metformin 57. Hofmann GE, Khoury J, Thie J. Recurrent pregnancy loss and dimin-
throughout pregnancy in women with polycystic ovary syndrome ished ovarian reserve. Fertil Steril 2000; 74:1192–5.
appears to safely reduce first-trimester spontaneous abortion: a pilot 58. Illingworth PJ, Groome NP, Duncan WC, et al. Measurement of circu-
study. Fertil Steril 200l; 75:46–52. lating inhibin forms during the establishment of pregnancy. J Clin
46. Velazquez EM, Mendoza SG, Wang P, Glueck CJ. Metformin therapy is Endocrinol Metab 1996; 81:1471–5.
associated with a decrease in plasma plasminogen activator inhibitor-1, 59. Muttukrishna S, George L, Fowler PA, et al. Measurement of serum
lipoprotein(a), and immunoreactive insulin levels in patients with the concentration of inhibin A (α-βA dimer) during human pregnancy.
polycystic ovary syndrome. Metabolism 1997; 46:454–7. Clin Endocrinol 1995; 42:391–7.
47. Wang JX, Davie MJ, Norman RJ, et al. Polycystic ovarian syndrome 60. Groome NP, Illingworth PJ, O’Brien M, et al. Detection of dimeric
and the risk of spontaneous abortion following assisted reproductive inhibin throughout the human menstrual cycle by two site enzyme
technology treatment. Hum Reprod 2001; 16:2606–9. immunoassay. Clin Endocrinol 1994; 40:717–23.
48. Tuckerman EM, Okon MA, Li T, et al. Do androgens have a direct 61. Muttukrishna S, Fowler P, Groome NP, et al. Serum concentrations of
effect on endometrial function? An in vitro study. Fertil Steril 2000; dimeric inhibin during the spontaneous human menstrual cycle and
74:771–9. after treatment with exogenous gonadotrophin. Hum Reprod 1994;
49. Regan L, Owen EJ, Jacobs HS, et al. Hypersecretion of luteinising hor- 9:1634–42.
mone, infertility, and miscarriage. Lancet 1990; 336:1141–4. 62. Luisi S, Florio P, Reis F, et al. Inhibins in female and male reproduc-
50. Nardo LG, Rai R, Backos M, et al. High serum luteinizing hormone tive physiology: role in gametogenesis, conception, implantation and
and testosterone concentrations do not predict pregnancy outcome in early pregnancy. Hum Reprod Update 2005; 11:123–35.
women with recurrent miscarriage. Fertil Steril 2002; 77:348–52. 63. Roberts VJ, Barth S, el-Roeiy A, et al. Expression of inhibin/activin
51. Westergaard LG, Laursen SB, Andersen CY. Increased risk of early subunits and follistatin messenger ribonucleic acids and proteins in
pregnancy loss by profound suppression of luteinizing hormone ovarian follicles and corpus luteum during the human menstrual
during ovarian stimulation in normogonadotrophic women under- cycle. J Clin Endocrinol Metab 1993; 77:1402–10.
going assisted reproduction. Hum Reprod 2000; 15:1003–8. 64. Petraglia F, Garuti C, Calza L, et al. Inhibin subunits in human pla-
52. Clifford K, Rai R, Watson H, et al. Does suppressing lureinising centa: localization and messenger ribonucleic acids during pregnancy.
hormone secretion reduce the miscarriage rare? Results of a ran- Am J Obstet Gynecol 1991; 165:750–8.
domised controlled trial. BMJ 1996; 312:1508–11. 65. Ledger W. Measurement of inhibin A and activin A in pregnancy – pos-
53. Craig LB, Ke RW, Kutteh WH. Increased prevalence of insulin resist- sible diagnostic applications. Mol Cell Endocrinol 2001; 180:117–21.
ance in women with a history of recurrent pregnancy loss. Fertil Steril 66. Lahiri S, Anobile CJ, Stewart P, et al. Changes in circulating concen-
2002; 78:487–90. trations of inhibins A and pro-a C during first trimestrer medical ter-
54. Stadtmauer LA, Wong BC, Oehninger S. Should patients with poly- mination of pregnancy. Hum Reprod 2003; 18:744–8.
cystic ovary syndrome be treated with metformin? Benefits of insulin 67. Florio P, Lombardo M et al. Activin A, corticotrophin-releasing factor
sensitizing drugs in polycystic ovary syndrome – beyond ovulation and prostaglandin F2α increase immunoreactive oxytocin release
induction. Hum Reprod 2002; 17:3016–26. from cultured human placental cells. Placenta 1995; 17:307–11.
55. Trout SW, Seifer DB. Do women with unexplained recurrent preg- 68. Muttukrishna S, Jauniaux E, Greenwold N, et al. Circulating levels of
nancy loss have higher day 3 serum FSH and estradiol values? Fertil inhibin A, activin A and follistatin in missed and recurrent miscar-
Steril 2000; 74:335–7. riages. Hum Reprod 2002; 17:3072–8.
56. Gurbuz B, Yalti S, Ozden S, et al. High basal estradiol level and FSH/ 69. Prakash A, Laird S, Tuckerman S, et al. Inhibin A and activin A may
LH ratio in unexplained recurrent pregnancy loss. Arch Gynecol be used to predict pregnancy outcome in women with recurrent mis-
Obstet 2004; 270:37–9. carriage Fertil Steril 2005; 83:1758–63.
87
0415421306-Ch06 3/29/07 5:35 PM Page 88
0415421306-Ch06a 3/29/07 5:35 PM Page 89
DEBATE
6a. Should progesterone supplements be

used?
– For
Jerome H Check
In pregnant women, a 34 kDa protein has been inhibits Th1 cytokines and favors Th2 cytokines,
identified that can block natural killer (NK)- thus inhibiting cellular immune response and pro-
cell-mediated lysis of K562 tumor cells.1,2 Because moting humoral responses. The suppression of the
the expression of this protein by CD8+ T lympho- cellular immune system is limited to the maternal–
cytes (γ/δ T cells) requires progesterone exposure for fetal interface, and this constitutes selective
expression, it was called the progesterone-induced immune tolerance.
blocking factor (PIBF).3 PIBF may induce a shift Much less PIBF expression was found in women
from Th1 to Th2 cytokines.2,4 Progesterone recep- who eventually lost their pregnancies than in
tors (PgR) have not been demonstrated in normal healthy pregnant women, thus supporting this
T lymphocytes, yet these receptors are found in hypothesis.11,12 However, no such difference was
normal pregnancy, but at a lower density than other found in aborters or non-aborters in patients
progesterone target receptor tissues.5–7 Liver trans- aggressively treated with progesterone.13 Surgical
plants and blood transfusions have been shown to removal of the ovary with the corpus luteum of
induce PgR on γ/δ T cells, even in males.8 Paternal pregnancy prior to 8 weeks will lead to spontaneous
lymphocyte injection prior to ovulation has been abortion.14 Donor oocytes do not work unless prog-
shown to increase PIBF expression in the mid to late esterone is supplemented for at least 8 weeks.
luteal phase in women exposed to embryo transfer.9 The difficulty in determining the role of proges-
These data support the following hypothesis as terone in preventing recurrent pregnancy loss is due
to one way the fetus escapes immune rejection by to difficulty in determining who has progesterone
NK cells. The fetal semi-allograft induces PgR in γ/δ deficiency and in establishing the diagnosis. Also,
T cells following trophoblast invasion. The interac- the question arises as to whether the problem starts
tion with high concentration of progesterone causes in the luteal phase or whether a need for more prog-
the expression of PIBF by γ/δ T cells with induced esterone can arise after the pregnancy is more
PgR. The PIBF is only made at the maternal–fetal advanced. Are some women more prone to have
interface, because that is where there is an adequate ovulatory dysfunction or corpus luteum of preg-
progesterone concentration. PgR are made in γ/δ nancy dysfunction – but not in every cycle – leading
T cells throughout the body, but the progesterone to a higher frequency of miscarriage in any given
level is insufficient to cause PIBF expression by γ/δ pregnancy but possibly without causing recurrent
T cells that are not situated at the maternal–fetal miscarriages? If one performs chromosome analysis
interface. The PIBF inhibits NK-cell cytologic activ- on an aborted fetus and aneuploidy is found, does
ity at least partially by inhibiting the release of per- that prove that the woman would not have lost the
forin from storage granules of NK cells.10 PIBF also pregnancy due to progesterone deficiency anyhow?
89
0415421306-Ch06a 3/29/07 5:35 PM Page 90
There is no question that the standard tests of drugs (but 4 of these 7 (57.1%) miscarried), versus 14
luteal phase defect (LPD) – for example the mid-luteal of 20 (70%) conceiving with the combination of folli-
phase serum progesterone level (or multiple cle-maturing drugs and progesterone (with only 1
integrated levels during the luteal phase) or an out- miscarriage (7.1%)).17 These data suggest that women
of-phase endometrial biopsy in mid or late luteal not producing mature follicles before egg release need
phase – lack sensitivity and specificity. Multiple flaws more than luteal phase support to achieve pregnan-
exist – such as variability of serum progesterone cies, but follicle-maturing drugs may not fully correct
assays, lack of standardization of what focal point the luteal phase, so that progesterone supplementation
should be used to obtain a serum progesterone level, is needed to inhibit miscarriage.17 These data are
and lack of standardization of the histological findings consistent with another prospective study where
on the endometrial biopsy, so that different patholo- 50 women conceiving after follicle-maturing drugs
gists may interpret the dating of the biopsy differently. who were given progesterone supplementation in the
If the need for extra progesterone supplementation luteal phase had a miscarriage rate of 6%, versus 28%
occurred in only one-third of pregnancies, would it in controls not given progesterone.18
not be worthwhile to use such supplementation in There were 58 women in the infertility study
every cycle to eliminate a 33% miscarriage rate? with LPD who did attain a mature follicle.17 They
I have considered that there may be a spectrum were randomized into a group receiving only prog-
of progesterone deficiency with effects ranging esterone supplementation in the luteal phase and a
from prevention of embryonic implantation, group receiving only follicle-maturing drugs. Of the
through successful implantation but with early loss women taking only progesterone, 24 (77.4%) con-
leading to a chemical pregnancy only, through suf- ceived within 6 months of therapy and 1 (4.1%) mis-
ficient progression to allow ultrasound evidence of carried.17 In comparison, only 3 of 27 (11.1%)
at least a gestational sac with either a non-viable conceived with clomiphene citrate or human meno-
fetal pole or an empty sac (so-called blighted pausal gonadotropin and 2 (66.7%) miscarried.17
ovum), through fetal viability but death before the Interestingly, 25 women failing to have successful
end of the first trimester, through second-trimester pregnancies with follicle-stimulating drugs were
loss, and finally pre-term delivery. treated exclusively with progesterone supplementa-
The question arises, however, as to whether there tion during the next 6 months, and 16 (64%)
are any data to confirm this ‘spectrum theory’. It is conceived, with only 1 (6.2%) miscarriage.17 These
known that estradiol (E2) induces PgR in the human data show that, at least in infertile women with
endometrium.15,16 Women with regular menses who apparent luteal phase defects, progesterone therapy
do exhibit a rise in progesterone during the luteal phase can reduce the risk of miscarriage. But what about
may actually not generate a sufficient peak serum E2 the group of women who can conceive without help
level to induce adequate PgR in the endometrium. We but then miscarry? A study was performed where
performed a study (unpublished) on fertile women women seeking help because of previous miscar-
seeking contraception, and found that the great major- riage(s) without delivery were offered progesterone
ity attained an average follicle diameter of a minimum therapy starting in the luteal phase and continued
18 mm and a serum E2 level of 200 pg/ml or more. through the first trimester, but were warned about the
We evaluated 100 women with a minimum of 1 year’s possible risk of birth defects. Of the 100 proges-
infertility with an out-of-phase late luteal phase terone-treated women, 10 (10%) had a first-trimester
endometrium, and found that 42 did not achieve a miscarriage, compared with 10 of 24 (41.6%) who
mature follicle. They were randomized into three treat- elected not to be treated with progesterone.19
ment regimens. Only 3 of 12 (25%) conceived within There is no question that low serum proges-
6 months following exclusive treatment with luteal terone levels are associated with poor pregnancy
phase progesterone (but there were no miscarriages), outcome. One study found that 17 of 18 women
versus 7 of 10 women (70%) with follicle-maturing with serum progesterone less than 15 ng/ml lost
90
0415421306-Ch06a 3/29/07 5:35 PM Page 91
DEBATE: SHOULD PROGESTERONE SUPPLEMENTS BE USED? – FOR
their pregnancy.20 The authors even suggested that, 2. Szekeres-Bartho J, Kilar F, Falkay G, et al. The mechanism of the
inhibitory effect of progesterone on lymphocyte cytotoxicity: I.
given the near inevitability of pregnancy loss,
Progesterone-treated lymphocytes release a substance inhibiting
women at risk for possible ectopic pregnancy should cytotoxicity and prostaglandin synthesis. Am J Reprod Immunol
have a dilatation and evacuation, and, in the event of Microbiol 1985; 9:15–18.
3. Szekeres-Bartho J, Autran B, Debre P, et al. Immunoregulatory effects
products of conception being found, evaluation for
of a suppressor factor from healthy pregnant women’s lymphocytes
ectopic pregnancy should be discontinued, while in after progesterone induction. Cell Immunol 1989; 122:281–94.
the absence of such products, serial ultrasound scans 4. Szekeres-Bartho J, Barakonyi A, Polgar B, et al. The role of γ/δ T cells
and β human chorionic gonadotropin (β-hCG) in progesterone-mediated immunomodulation during pregnancy: a
review. Am J Reprod Immunol 1999; 42:44–8.
determinations should be carried out.20 5. Szekeres-Bartho J, Csernus V, Hadnagy J. The blocking effect of prog-
However, it is not clear whether the low proges- esterone on lymphocyte responsiveness is receptor-mediated. Biol
Immunol Reprod 1989; 15:36.
terone is a result of a dying or dead fetus, reflecting
6. Szekeres-Bartho J, Szekeres Gy, Debre P, et al. Reactivity of lympho-
degeneration of the corpus luteum and/or placenta cytes to a progesterone receptor specific monoclonal antibody. Cell
or whether the low progesterone could be etiological. Immunol 1990; 125:273–83.
7. Chiu L, Nishimura M, Ishii Y, et al. Enhancement of the expression of
The fact that one study found that 70% (19/27) of
progesterone receptor on progesterone treated lymphocyte after
women with serum progesterone below 15 ng/ml immunotherapy in unexplained recurrent spontaneous abortion. Am
successfully completed the first trimester with J Reprod Immunol 1996; 35:552–7.
8. Szekeres-Bartho J, Weill BJ, Mike G, et al. Progesterone receptors in
aggressive progesterone therapy strongly implies
lymphocytes of liver-transplanted and transfused patients. Immunol
that low serum progesterone can be a correctable Lett 1989; 22:259–61.
cause of miscarriage.21 Another study found that 9. Check JH, Arwitz M, Gross J, et al. Lymphocyte immunotherapy (LI)
increases serum levels of progesterone induced blocking factor
60% of intrauterine pregnancies could be salvaged
(PIBF). Am J Reprod Immunol 1997; 37:17–20.
with a serum progesterone below 8 ng/ml.22 The 10. Faust Z, Laskarin G, Rukavina D, et al. Progesterone induced blocking
pure scientist may argue that many of the ‘controlled’ factor inhibits degranulation of NK cells. Am J Reprod Immunol
1999; 42:71–5.
studies in support of progesterone therapy were too
11. Szekeres-Bartho J, Varga P, Retjsik B. ELISA test for detecting a prog-
small or lacked placebo controls or were not double- esterone-induced immunological factor in pregnancy serum.
blinded. The data convince me that I would empiri- J Reprod Immunol 1989; 16:19–29.
12. Szekeres-Bartho J, Faust Z, Varga P. The expression of a progesterone-
cally treat any woman with a previous miscarriage
induced immunomodulatory protein in pregnancy lymphocytes. Am
with extra progesterone even if the aforementioned J Reprod Immunol 1995; 34:342–8.
studies of the efficiency of such treatment are not 13. Check JH, Ostrzenski A, Klimek R. Expression of an immunomodula-
tory protein known as progesterone induced blocking factor (PIBF)
scientifically convincing beyond a shadow of doubt.
does not correlate with first trimester spontaneous abortions in proges-
I would rather give a woman a benign treatment terone supplemented women. Am J Reprod Immunol 1997; 37:330–4.
that is later found not be effective than not treat her, 14. Csapo AI, Pulkkinen M. Indispensability of the human corpus luteum
in the maintenance of early pregnancy: lutectomy evidence. Obstet
have her miscarry, and later find studies showing
Gynecol Surv 1978; 3:69–81.
unequivocally that with such treatment definitely 15. Lessey BA, Killam AP, Metzger DA, et al. Immunohistochemical
reduces the risk of miscarriage. Possibly, sometimes analysis of human estrogen and progesterone receptors throughout
the menstrual cycle. J Clin Endocrinol Metab 1988; 67:334–40.
progesterone may be needed but may not be sufficient –
16. Bergquist A, Ferno M. Oestrogen and progesterone receptors in
for example if the fetal semi-allograft fails to induce endometriotic tissue and endometrium: comparison of different cycle
sufficient PgR in γ/δ T cells. Possibly, injection of phases and ages. Hum Reprod 1993; 8:2211–17.
17. Check JH, Nowroozi K, Wu CH, et al. Ovulation inducing drugs
lymphocytes may also be needed in some cases to
versus progesterone therapy for infertility in patients with luteal phase
maximize the benefit of progesterone therapy by defects. Int J Fertil 1988; 33:252–6.
allowing adequate production of immunomodula- 18. Check JH, Chase JS, Wu CH, et al. The efficacy of progesterone in
achieving successful pregnancy: I. Prophylactic use during luteal
tory proteins (e.g., PIBF).7,9,23–25
phase in anovulatory women. Int J Fertil 1987; 32:135–8.
19. Check JH, Chase JS, Nowroozi K, et al. Progesterone therapy to
decrease first-trimester spontaneous abortions in previous aborters.
REFERENCES Int J Fertil 1987; 32:197–9.
20. Yeko TR, Gorrill MJ, Hughes LH, et al. Timely diagnosis of early
1. Pence H, Petty WM, Rocklin RE. Suppression of maternal responsiveness ectopic pregnancy using a single blood progesterone measurement.
to paternal antigens by maternal plasma. J Immunol 1975; 114:525–8. Fertil Steril 1987; 48:1048–50.
91
0415421306-Ch06a 3/29/07 5:35 PM Page 92
21. Check JH, Winkel CA, Check ML. Abortion rate in progesterone immunotherapy and progesterone treatment versus progesterone alone
treated women presenting initially with low first trimester serum in primary habitual aborters. Gynecol Obstet Invest 1995; 39:257–61.
progesterone levels. Am J Gynecol Health 1990; 4:33–4. 24. Check JH, Liss JR, Check ML, et al. Lymphocyte immunotherapy can
22. Choe JK, Check JH, Nowroozi K, et al. Serum progesterone and improve pregnancy outcome following embryo transfer (ET) in
17-hydroxyprogesterone in the diagnosis of ectopic pregnancies and patients failing to conceive after two previous ET. Clin Exp Obstet
the value of progesterone replacement in intrauterine pregnancies Gynecol 2005; 32:21–2.
when serum progesterone levels are low. Gynecol Obstet Invest 1992; 25. Check JH, Liss J, Check ML, Diantonio A, Choe JK, Graziano V.
34:133–8. Leukocyte immunotherapy improves live delivery rates following
23. Check JH, Tarquini P, Gandy P, et al. A randomized study comparing the embryo transfer in women with at least two previous failures: A ret-
efficacy of reducing the spontaneous abortion rate following lymphocyte rospective review. Clin Exp Obst Gyn 2005; 32:85–8.
92
0415421306-Ch06b 3/29/07 5:36 PM Page 93
DEBATE
6b. Should progesterone supplements

be used?
– Against
Shazia Malik and Lesley Regan
INTRODUCTION suggestive of LPD in up to 50% of single menstrual

cycles and 25% of sequential cycles.4 A prevalence
The role of progesterone in the mammalian reproduc- study of out-of-phase endometrial biopsy speci-
tive cycle is well described and undisputed. mens5 failed to show any significant difference
Its pharmacodynamics have been extensively studied, between fertile and infertile patients and recurrent
and progesterone has been synthesized and pregnancy loss, which calls the role of this interven-
commercially available since 1935. However, despite tion into question. In a series of 74 women with
the putative role of progesterone in ameliorating recurrent miscarriage before 10 weeks of
unexplained recurrent pregnancy loss (RPL), the evi- gestation, there was no difference in pregnanediol
dence base for its use in this setting is lacking, despite excretion curves between those women who either mis-
decades of clinical use. With this in mind, we argue that carried or went on to have a successful pregnancy.6 In
the use of progesterone supplementation for women in fact estriol was a better prognostic indicator, showing
whom no identifiable cause for three or more suc- lower values in those destined to miscarry. In 1993,
cessive pregnancy losses prior to 20 weeks of ges- Quenby and Farquharson7 audited 203 consecutive
tation has been identified is currently unjustified. couples attending their clinic and found that, com-
pared with any other predictor, oligo-amenorrheic
women were most likely to have a further miscar-
SCIENTIFIC BASIS riages, and further, that they exhibited low luteal phase
estradiol levels but normal luteal progesterone and
LUTEAL PHASE DEFECT normal luteinizing hormone (LH) profiles through-
out the cycle. A more recent study found that a mid-
Removal of the corpus luteum before the end of the luteal progesterone level of less than 10 ng/ml (as a
7th week of amenorrhoea leads to miscarriage. marker of luteal phase deficiency) did not predict a
Rescue can be achieved with progesterone therapy, future pregnancy loss in women with two successive
but not with estrogen.1 Corpus luteum deficiency, unexplained first-trimester miscarriages.8
or luteal phase defect (LPD), has been cited as the
underlying pathology in 35–40% of unexplained PROGESTERONE AND IMMUNOMODULATION
recurrent pregnancy loss, manifesting in low serum IN EARLY PREGNANCY
progesterone levels and out-of-phase endometrial
biopsies.2,3 However, women with no history of recur- There is increasing evidence that progesterone is a
rent miscarriage exhibit an endometrial histology key modulator in the immune response required to
93
0415421306-Ch06b 3/29/07 5:36 PM Page 94
achieve a successful pregnancy outcome. The com- Murine experiments have shown a poor correlation
plexities of the adaptation between the maternal between Th1/Th2 cytokine ratios and abortion
immune system and the semi-allograft of the rates, implicating environmental selective pressures
fetoplacental unit are not clearly understood. in eliminating ‘genetically weaker’ embryos in early
The presence of progesterone and an upregulation pregnancy.17 While some rodent data are enticing,
of its receptors on decidual natural killer (NK) and PIBF data in human pregnancy are scanty, and the
placental lymphocytes appears to be required to mechanisms underlying immune-mediated preg-
defend the developing trophoblast from the mater- nancy loss remain incompletely elucidated.18
nal immune reaction.9 These activated cells then
synthesise progesterone-induced blocking factor
(PIBF), mediating both the immunomodulatory CLINICAL DATA
and antiabortive effects of progesterone.10 In addi-
tion to the shift towards Th2 cytokine production, The uterine decidual and systemic levels of proges-
NK cytolytic activity in human pregnancy is terone necessary to maintain an early pregnancy in
inversely related to the levels of PIBF-positive lym- humans are not known.19 Hence, clinical studies
phocytes,11 and neutralization in pregnant mice must, by definition, employ arbitrary doses/mode
results in NK-cell-mediated abortion.10 The cellular of delivery of drug. Hill20 has proposed study criteria
T-cell system, in particular the Th1 cells, modulates that should be fulfilled when designing a treatment
this immune response, releasing either Th1 trial for unexplained RPL (Table 6b.1). To date,
cytokines (e.g., tumor necrosis factor α, TNF-α) there are no trials that fulfil these criteria having
that induce cytotoxic and inflammatory reactions, systematically evaluated the role of progesterone
or Th2 cytokines (e.g., interleukin-10, IL-10) asso- treatment in recurrent pregnancy loss. Two meta-
ciated with B-cell production.12 Serum cytokine analyses published in the same journal reported
profiles demonstrate a shift towards Th2 in normal conflicting results regarding the value of proges-
pregnancy, whereas in recurrent miscarriage suffer- terone supplementation in miscarriage patients.
ers, the Th1 response predominates.13 A recent Goldstein et al21 included trials of women with a
study has reported that the administration of intra- ‘high-risk’ pregnancy, including a history of previous
muscular progesterone injections to recurrent mis-
carriage patients restored levels of soluble TNF
receptors to values seen in women with no such his-
tory.14 However, the treatment only commenced Table 6b.1 Study criteria for recurrent pregnancy
at 8 weeks of gestation, included women up to loss treatments20
40 years of age, and, furthermore, showed that in • Scientifically sound rationale
some of the cases no response in terms of receptor • Power calculation ensuring sufficient numbers under reasonable
levels was seen in pregnancies that then went on to assumptions (e.g., 60% success without and 80% success with
treatment)
miscarry. PIBF appears to be the main modulator of • Exclusion of patients with less than three unexplained clinical
the actions of progesterone, with significantly lower pregnancy losses
expression in recurrent miscarriage patients com- • Exclusion of patients with presumed causes for prior pregnancy
losses
pared with those with a healthy pregnancy.15 • Prospective study design
Conversely, Check et al16 treated women in the first • Prestratification of participants by age and number of prior losses
trimester aggressively with progesterone, but found (both of which are independent risk factors for subsequent loss)
• Effective randomization after prestratification
no differences in PIBF expression by lymphocytes. • Placebo-controlled
However Th1 and Th2 cytokines were not meas- • Double-blinded
ured in this study, and could not be correlated • No concomitant therapy
• Karyotype of subsequent losses
either with PIBF levels or with any given response to • Follow-up to ensure safety
progesterone supplementation in specific patients.
94
0415421306-Ch06b 3/29/07 5:36 PM Page 95
DEBATE: SHOULD PROGESTERONE SUPPLEMENTS BE USED? – AGAINST
miscarriage, stillbirth, or current preterm labour. dydrogesterone (10 mg orally) in early pregnancy
In addition, the authors used different preparations compared with those who remained untreated
commenced at varying gestations, and not surpris- (p<0.05). In this trial, women with an average of 3.3
ingly no benefit of treatment was identified. They previous unexplained recurrent abortions were ran-
subsequently recommended that randomized trials domized to receive either no treatment (n=30),
should be the only setting for the use of progesta- dydrogesterone (n=48), or human chorionic
tional agents in pregnancy. In contrast, Daya22 gonadotropin (hCG; 5000 IU intramuscularly every
presented a meta-analysis of controlled trials studying 4 days; n=36) from as soon as pregnancy was
the efficacy of progesterone support for pregnancy in confirmed until the 12th week of gestation. This
women with a history of recurrent miscarriage. trial does not, however, conform to the standards
Although the odds ratio (OR) for pregnancies reach- cited in Table 6b.1.
ing at least 20 weeks of gestation was 3.09 (95% con-
fidence interval (CI) 1.28–7.42), even he concluded
that recurrent miscarriage patients with LPD should CONCLUSIONS
have the efficacy of progesterone assessed in prospec-
tive double-blind randomized controlled trials. In the UK, three progestogenic products are licensed
Closer inspection reveals why this conclusion was for use in early pregnancy: intramuscular proges-
reached: only three studies met the inclusion/exclu- terone, vaginal progesterone, and oral dydroges-
sion criteria, and, as the differences between experi- terone, and have been authorized for between 10 and
mental groups were insignificant, they had to be 20 years. However, the number of studies examining
pooled to show a statistically significant power calcula- the efficacy of progesterone supplementation in early
tion. None of the three studies demonstrated a signif- pregnancy remains small, and they do not fulfil the
icant progesterone deficiency, each employed a criteria required for meaningful results. In addition,
different progestogen, and each had different inclu- the diversity of biological and pharmacological prop-
sion criteria recruiting patients only after they had erties does not allow extrapolation of results across
reached at least 8 weeks of gestation. In addition, only studies. Although there are no obvious adverse effects
a total of 50 treated and 45 control patients were iden- to mother or fetus, a low level of risk may as yet be
tified. In the study by Levine,23 patients were allocated unidentified. The observed frequency of another mis-
alternately, not randomly, and the series published by carriage after three is over 50%, and the wish to pre-
Swyer and Daley24 was not ‘blind’, as treated patients scribe an apparently safe and well-tolerated treatment
were administered an implant while some controls is appealing, especially in light of the emerging scientific
were offered a placebo tablet. A Cochrane meta- understanding of early pregnancy failure. As yet,
analysis looking at the same data also showed a statis- however, the evidence for ‘tender loving care’ shows a
tically significant reduction in miscarriage in favor of similar improvement in outcomes. The need ‘to do
those randomized to the progestogen group (OR something’ for a group of unfortunate patients often
0.37; 95% CI 0.17–0.91).25 Patients with primary recur- seems to over-ride the use of an evidence base. While
rent spontaneous abortion (RSA) were not differenti- treatment does not appear to do harm, the evidence
ated from those with secondary RSA. The for the use of progesterone supplementation in recur-
recommendations from the American College of rent pregnancy loss is contentious at best, dated, and
Obstetricians and Gynecologists26 conclude that it has poor at worst.
not been shown conclusively that progesterone treat-
ment or corpus luteum support influence outcome in REFERENCES
women with RPL, and that luteal phase progesterone
support is of unproven efficacy. 1. Csapo AI, Pulkkinen MO, Ruttner B, et al. The significance of the
human corpus luteum in pregnancy maintenance. I. Preliminary studies.
In a more recent trial,27 a significant reduction in Am J Obstet Gynecol 1972; 112:1061–7.
the miscarriage rate was observed in women receiving 2. Jones GS. The luteal phase defect. Fertil Steril 1976; 27:351–6.
95
0415421306-Ch06b 3/29/07 5:36 PM Page 96
16. Check JH, Ostrzenski A, Klimek R. Expression of an immunomodu-

3. Daya S, Ward S. Diagnostic test properties of serum progesterone
latory protein known as progesterone induced blocking factor (PIBF)
in the evaluation of luteal phase defects. Fertil Steril 1988;
does not correlate with first trimester spontaneous abortions in
49:168–70.
progesterone supplemented women. Am J Reprod Immunol 1997;
4. Davis OK, Berkeley AS, Naus GJ, et al. The incidence of luteal phase
37:330–4.
defect in normal, fertile women, determined by serial endometrial
biopsies. Fertil Steril 1989; 51:582–6. 17. Clark DA, Chaouat G, Gorczynski RM. Thinking outside the box:
5. Peters AJ, Lloyd RP, Coulam CB. Prevalence of out-of-phase endome- mechanisms of environmental selective pressures on the outcome of
trial biopsy specimens. Am J Obstet Gynecol 1992; 166:1738–45; the materno–fetal relationship. Am J Reprod Immunol 2002;
discussion 1745–6. 47:275–82.
6. Klopper A, Michie EA. The excretion of urinary pregnanediol after 18. Laird SM, Tuckerman EM, Cork BA, et al. A review of immune cells
the administration of progesterone. J Endocrinol 1956; 13:360–4. and molecules in women with recurrent miscarriage. Hum Reprod
7. Quenby SM, Farquharson RG. Predicting recurring miscarriage: Update 2003; 9:163–74.
What is important? Obstet Gynecol 1993; 82:132–8. 19. Azuma K, Calderon I, Besanko M, et al. Is the luteo-placental shift a
8. Ogasawara M, Kajiura S, Katano K, et al. Are serum progesterone myth? Analysis of low progesterone levels in successful ART pregnancies.
levels predictive of recurrent miscarriage in future pregnancies? Fertil J Clin Endocrinol Metab 1993; 77:195–8.
Steril 1997; 68:806–9. 20. Hill JA. Immunotherapy for recurrent pregnancy loss: ‘standard of
9. Roussev RG, Higgins NG, McIntyre JA. Phenotypic characterization care or buyer beware’. J Soc Gynecol Investig 1997; 4:267–73.
of normal human placental mononuclear cells. J Reprod Immunol 21. Goldstein P, Berrier J, Rosen S, et al. A meta-analysis of randomized
1993; 25:15–29. control trials of progestational agents in pregnancy. Br J Obstet
10. Szekeres-Bartho J, Barakonyi A, Polgar B, et al. The role of γ/δ T cells Gynaecol 1989; 96:265–74.
in progesterone-mediated immunomodulation during pregnancy: 22. Daya S. Efficacy of progesterone support for pregnancy in women
a review. Am J Reprod Immunol 1999; 42:44–8. with recurrent miscarriage. A meta-analysis of controlled trials. Br J
11. Szekeres-Bartho J, Faust Z, Varga P. The expression of a progesterone- Obstet Gynaecol 1989; 96:275–80.
induced immunomodulatory protein in pregnancy lymphocytes. 23. Levine L. Habitual abortion. A controlled study of progestational
Am J Reprod Immunol 1995; 34:342–8. therapy. West J Surg Obstet Gynecol 1964; 72:30–6.
12. Druckmann R, Druckmann MA. Progesterone and the immunology 24. Swyer GI, Daley D. Progesterone implantation in habitual abortion.
of pregnancy. J Steroid Biochem Mol Biol 2005;97:389–96. BMJ 1953; i:1073–7.
13. Raghupathy R, Makhseed M, Azizieh F, et al. Cytokine production by 25. Oates-Whitehead RM, Haas DM, Carrier JA. Progestogen for
maternal lymphocytes during normal human pregnancy and in unex- preventing miscarriage. Cochrane Database Syst Rev 2003;
plained recurrent spontaneous abortion. Hum Reprod 2000; 15:713–18. (4):CD003511.
14. Chernyshov VP, Vodyanik MA, Pisareva SP. Lack of soluble 26. American College of Obstetricians and Gynecologists. ACOG Practice
TNF-receptors in women with recurrent spontaneous abortion and Bulletin. Management of recurrent pregnancy loss. Number 24,
possibility for its correction. Am J Reprod Immunol 2005; 54:284–91. February 2001. Int J Gynaecol Obstet 2002; 78:179–90.
15. Szekeres-Bartho J, Barakonyi A, Miko E, et al, The role of γ/δ T cells 27. El-Zibdeh MY. Dydrogesterone in the reduction of recurrent sponta-
in the feto–maternal relationship. Semin Immunol 2001; 13:229–33. neous abortion. J Steroid Biochem Mol Biol 2005; 97:431–4.
96
0415421306-Ch06c 3/29/07 5:37 PM Page 97
DEBATE
6c. Should hCG supplementation be used?

– For
James Walker
INTRODUCTION support. This ‘handover’ time appeared to be

between 7 and 8 weeks’ gestation.
In the late 1920s, Willard Allen and colleagues1 Because progesterone was thought to be impor-
carried out experimental work demonstrating that tant for the maintenance of normal pregnancy,
preparations made from corpus luteum extracts the concept that a deficiency in progesterone might
could successfully support pregnancies in animals lead to miscarriage was a natural follow-on. By
where the ovaries had been ablated. Further work the late 1940s, it had been shown that functional
demonstrated that these preparations contained reproductive deficits sufficient to cause infertility or
progesterone. However, initial work on the use of recurrent abortion were present in women who
progesterone in maintaining early pregnancy in cas- appeared to be having regular menstrual cycles.5
trated rabbits failed. It was not until a study in 19382 These abnormalities were due to a deficit in proges-
that the maintenance of pregnancy in castrated terone secretion during the luteal phase of the cycle
animals was achieved using progesterone. This was (luteal phase deficiency, LPD). This disorder was
achieved when the castration was carried out after characterized by inadequate endometrial matura-
the 11th day following mating, when implantation tion, and was reported in up to 60% of women with
of the embryos had occurred. Other work suggested recurrent miscarriage. However, these early studies
a role for estrogen in supporting the early preg- are open to question, since there was no reliable
nancy.3 However, unlike in small animals, which method of dating the cycle. Since many of these
almost inevitably abort after ovariectomy, the studies presumed that the patient’s menstrual
results in pregnancies in larger animals and humans pattern is a normal 28-day cycle, these endometrial
was mixed, with many case being cited in which abnormalities could be related to prolonged cycles.
abortion was seen to follow removal of the corpus However, more recent studies on the hormonal
luteum, while in others pregnancy appeared to cycle have confirmed abnormalities of corpus luteal
continue successfully after ovariectomy in midpreg- function, with deficiency in progesterone levels in
nancy without hormonal support. It became clear the luteal phase and early pregnancy of those with a
that the effect was gestation-dependent.4 In human history of miscarriage, but in a lower percentage.6
pregnancy, exogenous progesterone support was In one of the few prospective studies evaluating
required only until around 8 weeks’ gestation, after women with three or more consecutive miscar-
which it was not required. The explanation of this riages, LPD was believed to be the cause in 17%.7
became obvious with the realization that the corpus Those found to have LPD are more likely to have
luteum is critical in early pregnancy, after which early losses (prior to the detection of fetal heart
time the trophoblast takes over the hormonal activity) than later loss.8
97
0415421306-Ch06c 3/29/07 5:37 PM Page 98
progesterone in a targeted population to improve

INTERVENTION – ESTROGEN the live birth rate.
If hormonal lack is associated with miscarriage,

hormonal support might be a possible therapy. The INTERVENTION – hCG
problem with steroid hormones is that they cannot
be taken by mouth, and synthetic substitutes are Human chorionic gonadotropin (hCG) is the hor-
required. Because of the early work showing the mone responsible for the corpus luteal support in
benefit of an estrogen,3 the first therapeutic trials early pregnancy. If the pregnancy is failing, levels of
were with diethylstilbestrol, a non-steroidal estro- hCG may well be low, resulting in low progesterone.
genic substance. This was given in large doses. None Rather than giving exogenous progesterone, hCG
of these studies showed benefit, and by the 1970s could be used directly, since this would stimulate
there was evidence of the effect of this medication the natural hormone and reduce the risks of abnor-
on the female offspring.9–11 Therefore, not only did mal effects on the fetus. Again, much work has
hormonal support not appear to be beneficial, it been done in the area of assisted conception. Here,
had major side-effects in the female offspring. hCG does appear to be of benefit in increasing birth
These included abnormalities of the cervix, recur- rates, but may not be any better than progesterone
rent pregnancy loss and, in extreme cases, clear cell on its own.12 In recurrent miscarriage, two early
carcinoma of the vagina. These studies had a major small studies were supportive of benefit,14,15 and
effect on the role of hormonal support in preg- this led to larger trials. From these, a meta-analysis
nancy, with the fear that hormones could have a of four studies14–17 showed a reduced risk of
significant effect on the unborn child. miscarriage for women with a history of recurrent
miscarriage (odds ratio 0.26, 95% confidence inter-
val 0.14–0.52) (Figure 6c.1).18 However, the authors
INTERVENTION – PROGESTERONE cautioned readers, as they felt that there was some
weakness in the larger trials. One of the studies in
Progesterone is the main supportive hormone of the analysis found particular benefit in women with
early pregnancy, and in recent years there has oligomenorrhea (Figure 6c.2).17,18
been an increased interest in its use in in vitro
fertilization (IVF). This treatment is given early to
support the luteal phase. The reasoning is as much THE PROBLEMS
for endometrial maturation as for pregnancy sup-
port, even though the early animal work did not The problems with all of these trials are the meth-
show any benefit of progesterone prior to implan- ods of recruitment and timing of intervention.
tation. In assisted conception, the randomized Most studies require the presence of an ultra-
trials of progesterone support in the luteal phase sound-proven fetal heart rate, even though, once
demonstrate an increased successful pregnancy this is present, it is unlikely that hormonal support
rate.12 However, in studies on the use of proges- will be beneficial, since successful implantation
terone purely to prevent miscarriage, there was no has occurred and the gestation is probably in
benefit for this therapy, irrespective of the route of advance of the time when most benefit would be
administration or type of progesterone used.13 It is achieved. All of the studies in human pregnancy
only in those with a history of recurrent miscar- have not shown any benefit after 8 weeks’ gesta-
riage that there was a trend towards an improved tion. Also by this time, the levels of endogenous
live birth rate. In all studies, there appear to be hCG are so high that it is unlikely that any exoge-
no obvious side-effects. Therefore, there does nous therapy will increase them. If an earlier
appear to be some possible benefit of the use of recruitment is used, trialists worry that they will
98
0415421306-Ch06c 3/29/07 5:37 PM Page 99
DEBATE: SHOULD hCG SUPPLEMENTATION BE USED? – FOR
Experiment:
Study n/N Control: n/N Peto OR (95% CI) Weight (%) Peto OR (95% CI)
Harrison et al15 0/10 7/10 13.8 0.05 [0.01–0.32]

Harrison et al16 6/36 8/31 32.2 0.58 [0.18–1.87]
Quenby and Farquharson17 6/36 10/29 34.9 0.39 [0.13–1.20]
Svigos14 1/13 9/15 19.1 0.11 [0.02–0.51]
Total (95% CI) 13/95 34/85 100.0 0.26 [0.14–0.52]

χ2=6.46 (df=3), z =3.91
0.1 0.2 1 5 10
Figure 6c.1 Meta-analysis of four trials of the use of hCG in the management of recurrent miscarriage.18 OR, odds ratio;
CI, confidence interval. Reproduced from Scott JR, Pattison N. Cochrane Database Syst Rev 2000(2): CD000101 with permission
from John Wiley & Sons, Ltd, UK.
include those who will inevitably miscarry. A CONCLUSIONS

recent medium-sized study, using early recruit-
ment, demonstrated a significant benefit from hCG therapy is not a panacea for all cases of
both an oral progesten and, less so, hCG, in recur- recurrent miscarriage, but it appears to be beneficial
rent miscarriage.19 Therefore, all the early work in a particular targeted group. The problem is
and recent trials suggest a potential benefit of how to diagnose these women. Certainly, those with
hormonal support in a targeted population. The a proven cause, such as antiphospholipid syn-
problem is what that target group is. Studies sug- drome and chromosomal abnormalities, will not
gest that up to 30% of women may benefit from benefit. Those with repeated early loss (prior to a
therapy,19 and these would be those with a history positive fetal heartbeat) and those with an irregular
of early loss20 and/or oligomenorrhea.17 cycle would appear to be more likely to benefit.
Experiment:
Study n/N Control: n/N Peto OR (95% CI) Weight (%) Peto OR (95% CI)
Quenby and Farquharson17 2/13 6/10 100.0 0.15 (0.03–0.83)
Total (95% Cl) 2/13 6/10 100.0 0.15 (0.03–0.83)

χ2=0.00 (df=0), z =2.18
0.1 0.2 1 5 10
Figure 6c.2 Meta-analysis of four trials of the use of hCG in the management of recurrent miscarriage showing a subanalysis of
those with oligomenorrhea.18 OR, odds ratio; CI, confidence interval. Reproduced from Scott JR, Pattison N. Cochrane Database
Syst Rev 2000(2): CD000101 with permission from John Wiley & Sons, Ltd, UK.
99
0415421306-Ch06c 3/29/07 5:37 PM Page 100
Using these criteria, around 30% of cases of recur- 8. Li TC, Iqbal T, Anstie B, et al. An analysis of the pattern of pregnancy
loss in women with recurrent miscarriage. Fertil Steril 2002;
rent miscarriage could be offered hormonal sup- 78:1100–6.
port, with a success rate of around 85%.19 Whether 9. Vessey MP, Fairweather DV, Norman-Smith B, et al. A randomized
to use a progestogen or hCG is a more difficult double-blind controlled trial of the value of stilboestrol therapy in
pregnancy: long-term follow-up of mothers and their offspring. Br J
argument. However, with the concerns over in utero Obstet Gynaecol 1983; 90:1007–17.
effects, hCG has the advantage of being more ‘natu- 10. Lanier AP, Noller KL, Decker DG, et al. Cancer and stilbestrol.
ral’, with evidence of similar benefits. A follow-up of 1,719 persons exposed to estrogens in utero and born
1943–1959. Mayo Clin Proc 1973; 48:793–9.
I therefore support the use of hCG in the treat- 11. Bamigboye AA, Morris J. Oestrogen supplementation, mainly diethyl-
ment of recurrent miscarriage in an appropriately stilbestrol, for preventing miscarriages and other adverse pregnancy
assessed population. outcomes. Cochrane Database Syst Rev 2003; (3):CD004271.
12. Daya S, Gunby J. Luteal phase support in assisted reproduction cycles.
Cochrane Database Syst Rev 2004; (3):CD004830.
REFERENCES 13. Oates-Whitehead RM, Haas DM, Carrier JA. Progestogen for prevent-
ing miscarriage. Cochrane Database Syst Rev 2003; (4):CD003511.
1. Allen WM, Corner GW. Physiology of the corpus luteum: III. Normal 14. Svigos J. Preliminary experience with the use of human chorionic
growth and implantation of embryos after very early ablation of the gonadotrophin therapy in women with repeated abortion. Clin
ovaries, under the influence of extracts of the corpus luteum. Am J Reprod Fertil 1982; 1:131–5.
Physiol 1929; 88:340–6. 15. Harrison RF. Treatment of habitual abortion with human chorionic
2. Allen WM, Heckel GP. Maintenance of pregnancy by progesterone in gonadotrophin: results of open and placebo-controlled studies. Eur J
rabbits castrated on the 11th day. Am J Physiol 1938; 125:31–5. Obstet Gynecol Reprod Biol 1985; 20:159–68.
3. Heckel GP, Allen WM. Maintenance of the corpus luteum and inhibi- 16. Harrison RF. Human chorionic gonadotrophin (hCG) in the
tion of parturition in the rabbit by injection of estrogenic hormone. management of recurrent abortion; results of a multi-centre placebo-
Endocrinology 1939; 24:137–9. controlled study. Eur J Obstet Gynecol Reprod Biol 1992; 47:175–9.
4. Russ W. The maintenance of pregnancy in the human after removal of 17. Quenby S, Farquharson RG. Human chorionic gonadotropin supple-
both ovaries – case report. Ann Surg 1940; 111:871–3. mentation in recurring pregnancy loss: a controlled trial. Fertil Steril
5. Jones GES. Some newer aspects of the management of infertility. JAMA 1994; 62:708–10.
1949; 141:1123–9. 18. Scott JR, Pattison N. Human chorionic gonadotrophin for recurrent
6. Miller H, Durant JA, Ross DM, et al. Corpus luteum deficiency as a miscarriage. Cochrane Database Syst Rev 2000; (2):CD000101.
cause of early recurrent abortion: a case history. Fertil Steril 1969; 19. El-Zibdeh MY. Dydrogesterone in the reduction of recurrent sponta-
20:433–8. neous abortion. J Steroid Biochem Mol Biol 2005; 97:431–4.
7. Tulppala M, Bjorses UM, Stenman UH, et al. Luteal phase defect in 20. Quenby SM, Farquharson RG. Predicting recurring miscarriage:
habitual abortion: progesterone in saliva. Fertil Steril 1991; 56:41–4. What is important? Obstet Gynecol 1993; 82:132–8.
100
0415421306-Ch06d 3/29/07 5:37 PM Page 101
DEBATE
6d. Should hCG supplementation be used?

– Against
Siobhan M Quenby and Roy G Farquharson
INTRODUCTION exogenous hCG in the absence of pregnancy has

been shown to prolong corpus luteum function,
Human chorionic gonadotropin (hCG) is a glyco- provided that exposure to hCG occurs before the
protein composed of two dissimilar subunits: onset of luteal regression, which begins on about
α and β. The α subunit is similar to pituitary the 10th postovulatory day. Secondly, ovariectomy
gonadotropins, while the β-hCG is very similar to or removal of the corpus luteum before the
luteinizing hormone (LH). hCG and LH bind to the 8th week of pregnancy causes miscarriage, whereas
same receptor in the corpus luteum. Therefore, both pregnancy continues independently of corpus
maternal LH and embryonic hCG can regulate hor- luteum function after the 9th week of gestation.3
mone secretion from the corpus luteum. During the Early pregnancy failure is associated with low
luteal phase of the menstrual cycle, either the serum levels of hCG. Low serum hCG may be due
corpus luteum undergoes spontaneous regression to the death of trophoblast several weeks before
within about 14 days or, if pregnancy occurs, hCG expulsion of the conceptus. Robust observational
binds to the LH/hCG receptor and extends its via- evidence indicates that hCG is critical to early preg-
bility. hCG stimulates the corpus luteum to produce nancy and that a lack of hCG is related to miscar-
a number of different hormones, including estro- riage. Understandably, the attractive theory emerges
gens, progestins, relaxins, and inhibin.1,2 These hor- that there may be a beneficial effect of hCG supple-
mones in turn maintain early pregnancy until the mentation in idiopathic recurring pregnancy loss
placenta can produce sufficient levels of steroids to (RPL) and in those women with an ‘endocrine
support the conceptus. Increasing levels of hCG are cause’ for their recurring pregnancy loss. Clearer
secreted by the syncytiotrophoblast of the early pla- attention to the type of pregnancy loss is needed, as
centa, and appear in serum. The rapidly increasing historical studies contain poor descriptive informa-
serum levels of hCG are sufficient to maintain the tion, which in turn can lead to ascertainment and
corpus luteum until its functional capacity dimin- reporting bias.4
ishes at about the 7th week of gestation, after which The exogenous administration of hCG is strate-
the corpus luteum regresses. The decline in corpus gically aimed at stimulating progesterone produc-
luteum function is coincident with the shift in prog- tion from the corpus luteum and thereafter the
esterone production from the corpus luteum to the fetoplacental unit. In recurring miscarriage
placenta, which then produces sufficient steroids to patients, a preconceptual, whole-cycle study of the
maintain the rest of the pregnancy.3 progesterone profile fails to provide proof that such
In the human, several sources of evidence demon- patients have a suboptimal progesterone rise in the
strate that hCG is necessary for the maintenance luteal phase. Interestingly, a suboptimal luteal estra-
of the corpus luteum. Firstly, the administration of diol rise was evident5 in the oligomenorrheic group.
101
0415421306-Ch06d 3/29/07 5:37 PM Page 102
prevention of recurrent miscarriage remains incon-

THE USE OF hCG SUPPLEMENTATION DURING
EARLY PREGNANCY clusive and that there was insufficient support for
its routine use in clinical practice.
THREATENED MISCARRIAGE
A double-blind, placebo-controlled trial exam-
ining hCG supplementation has been reported with
Recent evidence is available from a randomized cases of idiopathic recurring miscarriage.10 Eighty-
controlled trial in which hCG supplementation has one patients were recruited before 6 weeks’ gesta-
been used to treat threatened miscarriage and pre- tion, proven by ultrasonography, following their
vent pregnancy loss in 183 women.6 The study referral. All patients were given supportive therapy,
design was robust, based on a prospective, double- which consisted of review and a reassurance ultra-
blind, placebo-controlled trial. Compared with sound scan fortnightly between 6 and 14 weeks in
placebo, the authors clearly demonstrated no their pregnancy. Patient characteristics showed little
benefit from the use of hCG in early pregnancy, difference in the age or number of previous miscar-
with similar miscarriage rates in both treatment riages between each group of patients in the trial.
groups (11% vs 12%, risk ratio (RR) 1.1, 95% Unsuccessful pregnancies (total = 15) included
confidence interval (CI) 0.63–1.6). ectopic pregnancy (2) and first-trimester (12) and
second-trimester (1) miscarriage, and there were no
third-trimester intrauterine deaths. In the regular
PREGNANCY FOLLOWING OVULATION INDUCTION
menstrual cycle group, hCG had no beneficial effect
on the success rate above that obtained by reassurance
Single-center non-randomized observational stud- and placebo injections (86% vs 86%)(Table 6d.1).
ies have reported improved pregnancy outcome The high success rate in the placebo group (86%)
when hCG has been used in very early pregnancy.7 reflects the reported benefits of studies of support-
However, in the absence of a prospective random- ive therapy.11,12 In the subgroup of patients with
ized controlled trial design, there are clear concerns oligomenorrhea, there was a statistically significant
regarding selection and ascertainment bias that are advantage conferred by hCG. Of 13 patients treated
inherent in all studies of this nature. The authors with hCG, 11 had a favorable outcome (85%).
have focused on the possibility of ‘luteal dysfunc- However, of 10 patients treated with placebo, only 4
tion’ as a primary cause, but fail to define or meas- had a favorable outcome (40%) (Table 6d.1). A two-
ure this entity prior to conception. tailed Fisher exact test was used to give a p-value of
As with many studies in this area, the mecha- 0.039. hCG elevated the success rate of the
nism of hCG action in improving outcome or cor- oligomenorrheic group to that established in the
recting luteal dysfunction is not clearly identified. regular menstrual cycle group (85% vs 86%).
RECURRING MISCARRIAGE
In the absence of evidence-based practice, several Table 6d.1 Results of a randomized controlled trial
of hCG in women with recurrent miscarriage and
authors have examined the role of hCG supplemen- regular menstrual cycles or oligomenorrhea
tation during early pregnancy in differing clinical
scenarios. A well-recognized review has suggested Oligomenorrhea Regular cycle
that hCG is not recommended in women with
Placebo hCG Placebo hCG
recurring miscarriage, as it is of no proven value,8
but the same author states that an exception to this Miscarriages 6 2 4 4
rule applies when oligomenorrhea is present prior Live births 4 11 25 25
Success rate 40% 85% 86% 86%
to conception. A Cochrane review9 suggested p-value <0.039
that evidence for the use of hCG treatment in the
102
0415421306-Ch06d 3/29/07 5:37 PM Page 103
DEBATE: SHOULD hCG SUPPLEMENTATION BE USED? – AGAINST
PLACEBO EFFECT The first Dublin trial16 consisted of only 20 patients

and did not report on the method of blinding or
Prediction of success following idiopathic recurring randomization, and the difference in outcome
miscarriage has concentrated the minds of clini- between the two groups was large so as to seem
cians and empowered patients who have suffered implausible (100% vs 30% success rate). The second
pregnancy loss.12,13 Previous studies of supportive Dublin trial17 stated that some data was lost, as it was
therapy alone for RPL have observed success rates controlled by a drug company.
ranging from 33% to 77%.11,14 Subsequent preg- Meta-analysis is a methodological tool that
nancy success is driven mainly by maternal age and applies statistical methods and scientific strategies
number of losses. The average patient is approxi- to limit bias and to allow appraisal and synthesis of
mately 32 years old and has suffered three consecu- results from relevant studies on a specific topic. The
tive losses. Given these parameters, there is a 75% results of the first meta-analysis are held in the elec-
chance of success in the next pregnancy, irrespective tronic Cochrane database of perinatal trials, and
of treatment intervention and relying exclusively on this trial has been entered into that database.9
supportive care and ultrasound reassurance every Table 6d.2 shows the data from the original reports
2 weeks until 12 weeks. However, the figure in the Cochrane database.9 Hence, the figures are
decreases to 42% for a patient with five miscarriages, slightly different to those in the opposing debate in
aged 45 (see Chapter 19).12 For any given treatment favor of hCG supplementation. Both versions of the
intervention, this baseline success rate would need meta-analysis results indicate that hCG may have a
to be bettered, which in turn would require a power beneficial effect upon the risk of miscarriage for all
calculation invoving the largest patient recruitment women with a history of recurrent miscarriage, as
so far and which would need to outperform existing the confidence interval for the total result does not
reported randomized controlled trials. cross 1 (Table 6d.2). However the overall analysis is
greatly influenced by the two earliest studies,15,16
OTHER hCG TRIALS AND META-ANALYSIS
both of which had significant methodological
weaknesses. The confidence intervals for the later
The efficacy of hCG in early pregnancy had histori- two studies10,17 do cross 1, implying that any
cally been examined in three trials.15–17 Two found improvement could have occurred by chance.
hCG to improve pregnancy outcome15,16 and
one showed no statistically significant benefit.17 All OLIGOMENORRHEA AND RECURRENT MISCARRIAGE
trials had methodological difficulties. Svigos’
series15 did not use placebo for the control group When oligomenorrheic women with recurrent
and gave no details of randomization procedures. miscarriage are considered independently, hCG does
Table 6d.2 hCG versus placebo for recurrent miscarriagea
Miscarriages per treated pregnancy
Study Treatment Control Odds ratio (95% confidence interval)
Quenby and Farquharson10 6/41 10/39 0.39 (0.13–1.20)

Svigos15 1/13 9/15 0.11 (0.02–0.51)
Harrison16 0/10 7/10 0.05 (0.01–0.32)
Harrison17 6/36 8/31 0.58 (0.18–1.87)
Total 13/95 34/85 0.26 (0.14–0.52)
Between-trial test for heterogeneity; c (df =3) = 6.46.

2
a
Figures quoted are the original figures of the papers in the Cochrane database.9
103
0415421306-Ch06d 3/29/07 5:37 PM Page 104
confer a statistically significant reduced risk of with recurrent miscarriage and oligomenorrhea
miscarriage. However, the Liverpool study is the detected a difference in the hormone profile of
only trial to differentiate between regular cycles and these women compared with those with regular
oligomenorrhea, and the patient numbers (13 in menstrual cycles.5 Hasegawa et al18 demonstrated
the treatment arm and 10 in the placebo arm) that, in a population of 119 consecutive women
are too small to draw significant conclusions. with spontaneous first-trimester miscarriage, the
This observation is underpinned by the Cochrane incidence of oligomenorrhea was over-represented
review.9 at 11% (compared with 0.9% in the general female
Further observational data (1994–2000) from population19). In the oligomenorrheic women, a
the same unit (Figure 6d.1), based on a quasiran- normal fetal karyotype was shown in 34%, versus
domized patient preference basis, shows continued 12.5% (p < 0.01) in the women with normal men-
benefit for women with recurrent miscarriage and strual cycles. Further analysis revealed that, for
oligomenorrhea who receive hCG in early preg- those women with anembryonic pregnancy and
nancy. The data fail to show significant improve- normal karyotype, the incidence of oligomenorrhea
ment, but do indicate that some level of benefit may was 57%. These results suggest that oligomenorrhea
be present for this specific patient group. is associated with loss of normal rather than abnor-
We have performed a prospective cohort study mal pregnancies.
investigating the possibility of using elements of
the medical history and investigation of women with
recurrent miscarriage to predict future pregnancy OVERVIEW AND CONCLUSIONS
outcome.5 We found that oligomenorrhea (infre-
quent periods more than 35 days apart) was an At present, hCG support should not be offered to
important risk factor for predicting subsequent mis- those women with recurring miscarriage unless
carriage in this population of women suffering recur- oligomenorrhea is present. In the latter group, there
rent miscarriage. Further investigation of women is a need for a large prospective randomized
100%
90%
Percentage survival
80% 75%
70%
No hCG (n = 27)
50% 63%
hCG (n = 57)
40%
0 5 10 12 15 20 24
Weeks of gestation
Figure 6d.1 Quasirandomized study of hCG supplementation on pregnancy outcome in 84 patients with oligomenorrhea.
104
0415421306-Ch06d 3/29/07 5:37 PM Page 105
DEBATE: SHOULD hCG SUPPLEMENTATION BE USED? – AGAINST
controlled trial of hCG versus placebo in early 7. Blumenfeld Z, Ruach M. Early pregnancy wastage: the role of repeti-
tive human chorionic gonadotrophin supplementation during the
pregnancy support. A power calculation demands first 8 weeks of pregnancy. Fertil Steril 1992; 58:19–23.
recruitment of 300 women with oligomenorrhea 8. Li TC. Guide for practitioners: recurrent miscarriage: principles of
and recurrent miscarriage (150 in each arm) to have management. Hum Reprod 1998; 13: 478–82.
9. Scott JR, Pattison N. Human chorionic gonadotrophin for recurrent
an 80% chance of reaching statistical significance at miscarriage. Cochrane Database Syst Rev 2000; (2):CD000101.
the 5% level. As a result, the trial would need to be 10. Quenby SM, Farquharson RG. Human chorionic gonadotrophin sup-
multicentered, and is logistically possible, but plementation in recurring pregnancy loss: a controlled trial. Fertil
Steril 1994; 62:708–10.
financially demanding. The use of hCG supplemen- 11. Liddell HS, Pattison NS, Zanderigo A. Recurrent miscarriage outcome
tation should be confined exclusively to the realms after supportive care in early pregnancy. Aust NZ J Obstet Gynaecol
of a randomized controlled trial before advocating 1991; 31:3202.
12. Brigham SA, Conlon C, Farquharson RG. A longitudinal study of
a potentially beneficial treatment in the early anx- pregnancy outcome following idiopathic recurring miscarriage. Hum
ious stages of a subsequent pregnancy. Within the Reprod 1999; 14:2868–71.
parameters of such a trial, adherence to standard- 13. Kavalier F. Investigation of recurrent miscarriages. BMJ 2005; 331:121–2.
14. Stray Pedersen B, Stray Pedersen S. Etiological factors and subsequent
ized nomenclature4 and exclusion of abnormal obstetric performance in 195 couples with a prior history of habitual
karyotypes are essential.20,21 abortion. Am J Obstet Gynecol 1983; 148:140–6.
15. Svigos J. Preliminary experience with the use of human chorionic
gonadotrophin therapy in women with repeated abortion. Clin
REFERENCES
Reprod Fertil 1982; 1:131–5.
16. Harrison RF. Treatment of habitual abortion with human chorionic
1. Illingworth PJ, Reddi K, Smith K, et al. Pharmacologic rescue of gonadotropin: results of open and placebo-controlled studies. Eur J
the corpus luteum results in increased inhibin production. Obstet Gynecol Reprod Biol 1985; 20:159–68.
Clin Endocrinol 1990; 33:323–32. 17. Harrison RF. Human chorionic gonadotrophin (hCG) in the manage-
2. Duncan WC, McNeilly AS, Fraser HM, et al. Luteinising hormone ment of recurrent abortion: results of a multi-centre placebo con-
receptor in the human corpus luteum: lack of down-regulation during trolled study. Eur J Obstet Gynaecol 1992; 47:175–9.
maternal recognition of pregnancy. Hum Reprod 1996; 11:2291–7. 18. Hasegawa I, Takakuwa K, Tanaka K. The roles of oligomenorrhoea
3. Csapo AI, Pulkkinen MO, Rutner B, et al. The significance of human and fetal chromosomal abnormalities in spontaneous abortion. Hum
corpus luteum function in pregnancy maintenance. Am J Obstet Reprod 1996; 11:2304–5.
Gynecol 1972; 112:1061–7. 19. Munster K, Schmidt L, Helm P. Length and variation in the menstrual
4. Farquharson RG, Jauniaux E, Exalto N. Updated and revised nomen- cycle: a cross-sectional study from a Danish county. Br J Obstet
clature for description of early pregnancy events. Hum Reprod 2005; Gynaecol 1992; 99:422–9.
20:3008–11. 20. Morikawa M, Yamada H, Kato EH, et al. Embryo loss pattern is pre-
5. Quenby SM, Farquharson RG. Predicting recurring miscarriage: What dominant in miscarriages with normal chromosome karyotype among
is important? Obstet Gynaecol 1993; 82:132–8. women with repeated miscarriage. Hum Reprod 2004; 19:2644–7.
6. Qureshi NS, Edi-osagie EC, Ogbo V, et al. First trimester threatened 21. Stephenson MD, Awartini KA, Robinson WP. Cytogenetic analysis of
miscarriage treatment with human chorionic gonadotrophins: a ran- miscarriages from couples with recurring miscarriage: a case–control
domised controlled trial. Br J Obstet Gynaecol 2005; 112:1536–41. study. Hum Reprod 2002; 17: 446–51.
105
0415421306-Ch06d 3/29/07 5:37 PM Page 106
0415421306-Ch07 3/29/07 5:38 PM Page 107
7. Antiphospholipid syndrome –
pathophysiology
Gilad Twig, Yaniv Sherer, Miri Blank, and Yehuda Shoenfeld
INTRODUCTION ETIOLOGY OF APS
Antiphospholipid syndrome (APS) was first defined aPL have long been known to require a cofactor in
as a syndrome in 1983,1 consisting of a triad of order to have their effects. Today this co-factor
manifestations involving arterial and/or venous (apolipoprotein H or β2-glycoprotein I (β2GPI) is
thrombosis, recurrent fetal loss, accompanied by thought to be the antigen to which aPL bind.
mild to moderate thrombocytopenia and elevated Binding of aPL to β2GPI forms divalent IgG–β2GPI
titers of antiphospholipid (aPL) antibodies: lupus complexes that have increased affinity for mem-
anticoagulant (LA) and/or anticardiolipin anti- brane phospholipids.6 The physiological function of
bodies (aCL). Today, this syndrome is known to be β2GPI is unknown. β2GPI deficiency is not associ-
systemic and may affect almost every organ and ated with disease; homozygous β2GPI-null mice also
tissue in the body. The cause of APS is still consid- appear to suffer no pathological effects.7 The bind-
ered a mystery – yet, as in many other autoimmune ing of aPL–β2GPI to cell membranes, including
diseases, a combination of environmental and trophoblasts, results in injury and/or activation. Like
genetic factors has been proposed. Recent data many other autoimmune diseases, APS may have a
indicate that infectious agents may play a major role multifactorial etiology in which genetic susceptibil-
in the etiology of APS. The pathophysiology of APS ity is made apparent by environmental factors.
includes all arms of the coagulation system, as One environmental factor that has been inten-
well as other mechanisms not related to hyper- sively investigated in recent years is infection. In a
coagulability. In fact, aPL have been shown to be series of 100 APS patients,8 various infections
directly toxic to the developing fetus, as these anti- were shown to precede the development of APS,
bodies can be passively transferred from humans to including skin infections (18%), human immuno-
naive mice and will induce pregnancy loss in those deficiency virus (HIV) (17%), pneumonia (14%),
mice2 (Figure 7.1). Active immunization with hepatitis C virus (13%), and urinary tract infection
human pathogenic monoclonal anticardiolipin (10%). Helicobacter pylori, a common bacterial
antibody induces clinical manifestations of APS in pathogen that colonizes the gastric mucosa and
BALB/c mice.3 Additionally, the serum from women induces chronic gastric inflammation, has been
with APS is highly teratogenic to rat embryos in associated with APS. In pregnant women, H. pylori
culture and also affects embryonic growth.4 infection can cause intrauterine fetal growth retar-
Moreover, purification of the immunoglobulin G dation,9 and increases the risk of reproductive
(IgG) fraction of the sera of women with APS disorders.10 Molecular mimicry between (β2GPI)
directly affects the embryo and yolk sac, reducing and bacterial and viral epitopes is the principal
their growth.5 This chapter discusses the etiology mechanism that links infections to APS. In the case
and pathophysiology of APS, with special emphasis of H. pylori, 34% of patients were positive for anti-
on the reproductive system. β2GPI antibodies that showed high homology to the
107
0415421306-Ch07 3/29/07 5:38 PM Page 108
Figure 7.1 Experimental antiphospholipid syndrome (APS). (a) APS uterine horn showing resorbed pregnancies and live fetuses at
day 14 of pregnancy. (b) Control uterine horn showing developing pregnancies with no resorptions at day 14.
target epitopes of H. pylori structures.11,12 Indeed, presence of LA, a presentation characteristic of

β2GPI has been found to be immunogenic in experimental APS.13
vivo. Immunization of BALB/c or PL/J mice or Direct experimental evidence for molecular
New Zealand white rabbits with β2GPI resulted in mimicry of bacterial pathogens has emerged from
generation of anti-β2GPI antibodies.13 The high the effect of immunization with certain microbial
titers of mouse anti-β2GPI antibodies has been pathogens that share epitope homology with the
associated with an increased proportion of fetal β2GPI molecule. Pathogenic anti-β2GPI auto-
resorptions (the equivalent of fetal loss in humans), antibodies directed against the TLRVYK epitope
thrombocytopenia, and prolonged activated partial were formed in mice immunized with Haemophilus
thromboplastin time (aPTT), indicating the influenzae or Neisseria gonorrhoeae that exhibit
108
0415421306-Ch07 3/29/07 5:38 PM Page 109
ANTIPHOSPHOLIPID SYNDROME – PATHOPHYSIOLOGY
the TLRVYK sequence, or with tetanus toxoid that was shown to be altered at different levels. aPL
does not present linearly the sequence TLRVYK inhibit both protein C activation and the function
but can still serve as a mimotope. The formed anti- of activated protein C (APC), thereby preventing
β2GPI autoantibodies have been shown to be the inactivation of activated factors V and VIII.18
pathogenic and capable of inducing the clinical This inhibition is conditional upon the presence of
picture of experimental APS, manifested by a high β2GPI which is a pre-requisite for the binding of
percentage of fetal loss, thrombocytopenia, and a aPL to protein C. In addition, autoantibodies
prolonged aPTT.14 Moreover, the pathogenic effect directed against protein C, protein S, and thrombo-
of monoclonal antibodies to β2GPI is inhibited modulin have been detected in some APS patients.19
by the addition of synthetic peptides including the Tissue factor (TF), an initiator of the extrinsic
TLRVYK sequence. The latter prevented the devel- coagulation cascade, which is not normally
opment of APS in mice injected with monoclonal expressed by intravascular cells, has been shown to
antibodies to β2GPI, or decreased the degree of be altered in APS patients. It has been shown that
endothelial cell activation, monocyte adhesion, and TF-related procoagulant activity and TF mRNA
expression of adhesion molecules in vitro.15 levels in monocytes are increased in primary APS
Infection may be one of the mechanisms giving rise patients with thrombosis when compared with
to APS. In humans, infection with varicella zoster those without thrombosis.20 Injection of purified
virus (VZV) has been associated with APS.16 IgG aCL from APS patients with previous throm-
Although HIV, hepatitis A, hepatitis B, and hepatitis botic episodes induced a significant increase in
C are also associated with an increased prevalence both monocyte procoagulant activity and TF
of aCL antibodies, most of these are not β2GPI- expression, as compared with purified IgG or IgM
dependent.17 The difference between APS and the aCL from two systemic lupus erythematosus (SLE)
mere presence of aPL may be due to diseases such as patients without thrombosis.21 In addition, func-
syphilis and Lyme disease raising antibodies that tional anti-TF pathway inhibitor activity has
recognize phospholipids directly, whereas in APS been detected in the sera of a subset of APS patients,
the infections raise antibodies that recognize showing a correlation between the degree of
epitopes on phospholipid-binding proteins such as inhibition and associated occurrence of arterial
β2GPI. thrombosis and stroke.22
Endothelial cells are affected by aPL auto-
antibodies. Potentiation of human umbilical vein
MECHANISMS OF REPRODUCTIVE endothelial cells (HUVEC) procoagulant activity by
FAILURE IN APS aPL contained in sera from SLE patients is strongly
decreased after depleting IgG from the sera.23
THROMBOSIS Human anti-β2GPI IgM monoclonal antibodies
and polyclonal anti-β2GPI antibodies have been
As mentioned above, systemic thromboembolism is shown to induce tissue factor at both protein and
the principal manifestation of APS. Evidence for mRNA level in HUVEC monolayers in vitro.24 aPL
thrombi in the placental circulation and the benefi- can further up-regulate adhesion molecules
cial effect of antithrombotic therapy in APS patients (E-selectin, intercellular adhesion molecule 1
suffering from recurrent pregnancy loss (RPL) (ICAM-1) and vascular cell adhesion molecule 1
suggest a central role for this mechanism in repro- (VCAM-1)) expression and secretion of the proin-
ductive failure. The underlying basis for the hyper- flammatory cytokines interleukin-1b (IL-1b) and
coagulable state in APS is complex, and involves IL-6.25 Increased plasma levels of soluble VCAM-1
altered activity of all three major components that have been found in primary APS patients with
govern hemostasis: platelets, fibrinolysis, and the recurrent thrombotic events, and elevated levels of
coagulation cascade. The coagulation system in APS tissue plasminogen activator and von Willebrand
109
0415421306-Ch07 3/29/07 5:38 PM Page 110
factor (vWF) (as endothelial perturbation markers) clusters on exposed phospholipids, thereby forming
have been associated with aPL in SLE. a protective shield on the phospholipid surface.
Decreased endothelial cell prostacyclin (PGI2), Annexin V blocks phospholipids from becoming
the principal inhibitor of platelet aggregation, and available for coagulation reactions. The annexin V
increased thromboxane A2 (TXA2) production by protective shield could be damaged by either binding
platelets have both been implicated as mechanisms to anti-annexin V or preventing its binding to the
predisposing to thrombosis in patients with APS. phospholipid membrane, or by blocking autoanti-
aPL enhance platelet TXA2 production, and allow bodies against annexin V/phospholipids.33 Anti-
platelet activation to occur without a compensatory annexin V autoantibodies have been detected in
increment in the vascular biosynthesis of PGI2.26 patients with SLE and APS associated with preg-
Hypofibrinolysis can further aggravate the nancy loss, while reduced levels of annexin V have
prothrombotic state in APS. Endothelial cell dys- been observed on the placental villi of women having
function can increase plasma levels of plasminogen aPL and RPL and a thrombogenic background.34
activator inhibitor type-1 (PAI-1) and tissue-type Recently, complement activation was reported as
plasminogen activator (tPA) antigens.27 In addition essential and causative of APL-induced fetal injury.
the hypofibrinolytic state can be further aggravated Treatment with heparin prevented this activation
by the presence of autoantibodies against compo- both in-vivo and in-vitro.35
nents of the fibrinolytic system such as as anti-
plasmin/plasminogen28 and anti-tPA.29 ARACHIDONIC ACID AND PROSTACYCLIN
Platelets play a central role in primary hemosta-
sis and are involved in the prothrombotic state of aPL inhibit arachidonic acid release.36 Arachidonic
APS patients. Monoclonal aCL obtained from acid is an essential prerequisite for PGI2 production
patients with APS increased platelet interaction (PGI2 is a physiological inhibitor of thrombocyte
with the subendothelium.21 It has been proposed aggregation, and a potent vasodilatator). aPL have
that a minor degree of platelet activation can lead to been shown to increase the concentration of TXA2,
exposure of phospholipids, which can potentially thus altering the PGI2/ TXA2 balance.37 The alter-
be amplified to a much larger degree in the serum ation in the PGI2/TXA2 balance has two effects:
of APS patient than in controls.21 β2GPI initially vasoconstriction, which impedes the blood supply
binds to these phospholipids, then binds aPL to to the fetus, and platelet activation, with the proco-
form β2GPI–phospholipid complexes. The latter can agulant effects described above.
further activate platelet aggregation by allowing the In a mouse model of experimental APS,
interaction between the Fc portion and the platelet Shoenfeld and Blank38 infused aCL into pregnant
surface FcγRII receptors (the only FcγR molecules mice in order to induce APS. Mice that were
present on platelets).21,30 In addition to activation of cotreated with a thromboxane receptor antagonist
the FcγRII receptors, the β2GPI–phospholipid com- had a significant reduction in the fetal resorption
plexes can also exert their action through comple- rate from 45% to 19.8% and an increase in mean
ment activation, as complement generated in the placental and embryo weights. There was also an
presence of aPL binds to negatively charged phos- increased platelet count (from 597 100 to 1 075 000
pholipids and activates platelets.31 platelets/mm3) in treated mice, indicating the effect
In addition to the systemic prothrombotic effects, of thrombocyte aggegation in APS.
APS autoantibodies may alter the placental circula-
tion by attacking certain placental epitopes. Annexin ANTICYTOKINE EFFECT
A5, a potent anticoagulant protein that has a throm-
bomodulatory role in the placental circulation, is The anti-inflammatory cytokine IL-3 is important for
such a target.32 Annexin V is found on the apical the maintenance of normal pregnancy. IL-3 enhances
surface of placental syncytiotrophoblasts, and forms placental and fetal development while increasing the
110
0415421306-Ch07 3/29/07 5:38 PM Page 111
number of megakaryoctes. The serum level of IL-3 in also been shown to inhibit the rate of trophoblastic
pregnant patients with primary APS or APS second- cell growth and to accelerate the rate of apoptosis of
ary to SLE was found to be lower than in controls.39 cultured human placental cells.47,48
In vitro studies have revealed that low-dose aspirin aPL can alter the secretion of human chorionic
(10 mg/μl) stimulates IL-3 production through its gonadotropin (hCG). Our team has shown that the
ability to raise leukotriene production, while higher addition of human polyclonal purified aCL, which
doses of aspirin failed to induce IL-3 generation.39,40 were shown to induce experimental APS, suppressed
Furthermore, ciprofloxacin treatment significantly the pulsatile secretion of hCG.49 In vitro, mono-
decreased the rate of pregnancy loss in BALB/c mice clonal aPL halved trophoblastic hCG and human
with experimental APS. This effect correlated with an phospholipid production.50 These results may be
increases in the serum IL-3 level and in bone marrow secondary to the effect of aPL on trophoblast differ-
megakaryocytes.41 Other cytokines may also be entiation and invasion.
involved. The level of the proinflammatory and aPL do not seem to be teratogenic to the embryo,
prothrombotic cytokine tumor necrosis factor α however. The offspring of SLE/APS patients do not
(TNF-α), has been shown to be significantly higher in have a higher rate of anomalies, although the preg-
patients with APS than in healthy controls.42 nancies are often complicated by intrauterine
growth retardation.51
INDUCTION OF PLACENTAL CELL APOPTOSIS
Most the literature describes placental pathology in CONCLUSIONS

terms of aPL being directed against negatively
charged phospholipids, leading to placental infarc- APS is a systemic syndrome whose etiology involves
tion and eventually pregnancy loss. aPL may affect both environmental and genetic factors. Infections
the adhesion molecules between the elements of the may be highly important in the etiology of this
syncytiotrophoblast. Cellular activation increases syndrome, and molecular mimicry is probably the
the expression of cell adhesion molecules,43,44 which mechanism by which infectious agents induce aPL.
may promote leukocyte adhesion to the endothelial aPL exert their pathogenic effects via various mech-
surface. Although the thrombogenic effects of aPL anisms, including the induction of a hypercoagula-
are mediated by ICAM-1, VCAM-1, and P-selectin,45 ble state. Indeed, therapy of APS is usually directed
aPL may damage the trophoblast in a manner unre- towards eliminating the enhanced thrombosis. In
lated to thrombosis, as the cytotrophoblast cells RPL, the combination of low-molecular-weight
express phospholipid on their surface. This concept heparin and low-dose aspirin is considered to be the
is supported by histological evidence from patients treatment of choice. However, when therapy fails,
with aPL and fetal death. Women with aPL have other interventions aimed at controlling the levels
decreased vasculosyncitial membranes, increased of autoantibodies rather than their effects should be
syncytial knots, and substantially more fibrosis, considered, as other mechanisms involving autoan-
hypovascular villi, and infarcts than women without tibodies are also important in APS. Hence,
APS.46 The changes in syncytial membranes may be immunomodulation might theoretically block
secondary to thrombosis, but thrombosis could also some of the more detrimental effects of aPL rather
be secondary to placental damage that allows free than anticoagulant therapy alone.
transplacental passage of maternal aPL. Addition of
IgG purified from women with SLE/APS, positive ACKNOWLEDGMENTS
for aCL/anti-DNA antibodies, reduced yolk sac and
embryonic growth more than sera negative for This study was supported in part by the Federico
aPL but positive for anti-phosphatidylserine and Foundation Ernesto Hecht Research Grant (to Y
anti-laminin.47 The sera of SLE/APS patients has Sherer).
111
0415421306-Ch07 3/29/07 5:38 PM Page 112
REFERENCES 21. Reverter JC, Tassies D, Font J, et al. Effects of human monoclonal
anticardiolipin antibodies on platelet function and on tissue factor
1. Hughes GR. Thrombosis, abortion, cerebral disease, and the lupus expression on monocytes. Arthritis Rheum 1998; 41:1420–7.
anticoagulant. BMJ 1983; 287:1088–9. 22. Forastiero RR, Martinuzzo ME, Broze GJ. High titers of autoanti-
2. Blank M, Cohen J, Toder V, et al, Induction of anti-phospholipid bodies to tissue factor pathway inhibitor are associated with the
syndrome in naive mice with mouse lupus monoclonal and human antiphospholipid syndrome. J Thromb Haemost 2003: 1:718–24.
polyclonal anti-cardiolipin antibodies. Proc Natl Acad Sci USA. 1991; 23. Oosting JD, Derksen RH, Blokzijl L, et al. Antiphospholipid antibody
88:3069–73. positive sera enhance endothelial cell procoagulant activity – studies
3. Bakimer R, Fishman P, Blank M, et al. Induction of primary anti- in a thrombosis model. Thromb Haemost 1992; 68:278–84.
phospholipid syndrome in mice by immunization with a human 24. Kornberg A, Renaudineau Y, Blank M, et al. Anti-β2-glycoprotein I
monoclonal anticardiolipin antibody (H-3). J Clin Invest 1992; antibodies and anti-endothelial cell antibodies induce tissue factor in
89:1558–63. endothelial cells. Isr Med Assoc J 2000; 2(Suppl):27–31.
4. Ornoy A, Yacobi S, Avraham S, et al. The effect of sera from women 25. Meroni PL, Raschi E, Camera M, et al. Endothelial activation by aPL:
with SLE and/or antiphospholipid syndrome on rat embryos in cul- a potential pathogenetic mechanism for the clinical manifestations of
ture. Reprod Toxicol 1998; 12:185–91. the syndrome. J Autoimmun 2000: 15:237–40.
5. Ornoy A, Yacobi S, Tartakover Matalon S, et al. The effects of anti- 26. Lellouche F, Martinuzzo M, Said P, et al. Imbalance of thromboxane/
phospholipid antibodies obtained from women with SLE APS and prostacyclin biosynthesis in patients with lupus anticoagulant. Blood
associated pregnancy loss on rat embryos and placental explants in 1991; 78:2894–9.
culture. Lupus 2003; 12:573–8. 27. Jurado M, Paramo JA, Gutierrez-Pimentel M, et al. Fibrinolytic
6. Rand JH. The antiphospholipid syndrome. Annu Rev Med 2003; potential and antiphospholipid antibodies in systemic lupus erythe-
54:409–24. matosus and other connective tissue disorders. Thromb Haemost
7. Sheng Y, Reddel SW, Herzog H, et al. Impaired thrombin generation 1992; 68:516–20.
in β2-glycoprotein I null mice. J Biol Chem 2001; 276:13817–21. 28. Yang CD, Hwang KK, Yan W, et al. Identification of anti-plasmin anti-
8. Cervera R, Asherson RA, Acevedo ML, et al. Antiphospholipid bodies in the antiphospholipid syndrome that inhibit degradation of
syndrome associated with infections: clinical and microbiological fibrin. J Immunol 2004; 172:5765–73.
characteristics of 100 patients. Ann Rheum Dis 2004: 63:1312–17. 29. Cugno M, Cabibbe M, Galli M, et al. Antibodies to tissue-type plas-
9. Eslick GD, Yan P, Xia HH, et al. Foetal intrauterine growth restrictions minogen activator (tPA) in patients with antiphospholipid syndrome:
with Helicobacter pylori infection. Aliment Pharmacol Ther 2002: evidence of interaction between the antibodies and the catalytic
16:1677–82. domain of tPA in 2 patients. Blood 2004; 103:2121–6.
10. Figura N, Piomboni P, Ponzetto A, et al. Helicobacter pylori infection 30. Font J, Espinosa G, Tassies D, et al. Effects of β2-glycoprotein I and
and infertility. Eur J Gastroenterol Hepatol 2002: 14:663–9. monoclonal anticardiolipin antibodies in platelet interaction with suben-
11. Sorice M, Pittoni V, Griggi T, et al. Specificity of anti-phospholipid dothelium under flow conditions. Arthritis Rheum 2002; 46:3283–9.
antibodies in infectious mononucleosis: a role for anti-cofactor pro- 31. Shibata S, Sasaki T, Hirabayashi Y, et al. Risk factors in the pregnancy
tein antibodies. Clin Exp Immunol 2000; 120:301–6. of patients with systemic lupus erythematosus: association of
12. Blank M, Shoenfeld Y. β2-Glycoprotein-I, infections, antiphospho- hypocomplementaemia with poor prognosis. Ann Rheum Dis 1992;
lipid syndrome and therapeutic considerations. Clin Immunol 2004; 51:619–23.
112:190–9. 32. Wang X, Campos B, Kaetzel MA, et al. Annexin V is critical in the
13. Blank M, Faden D, Tincani A, et al. Immunization with anticardi- maintenance of murine placental integrity. Am J Obstet Gynecol
olipin cofactor (β2-glycoprotein I) induces experimental antiphos- 1999: 180:1008–16.
pholipid syndrome in naive mice. J Autoimmun 1994; 7:441–55. 33. Rand JH, Wu XX, Andree HA, et al. Pregnancy loss in the antiphos-
14. Blank M, Krause I, Fridkin M, et al. Bacterial induction of autoanti- pholipid-antibody syndrome – a possible thrombogenic mechanism.
bodies to β2-glycoprotein-I accounts for the infectious etiology of N Engl J Med 1997; 337:154–60.
antiphospholipid syndrome. J Clin Invest 2002; 109:797–804. 34. Matsubayashi H, Arai T, Izumi S, et al. Anti-annexin V antibodies in
15. Blank M, Shoenfeld Y, Cabilly S, et al. Prevention of experimental patients with early pregnancy loss or implantation failures. Fertil
antiphospholipid syndrome and endothelial cell activation by syn- Steril 2001; 76:694–9.
thetic peptides. Proc Natl Acad Sci USA 1999; 96:5164–8. 35. Girardi G, Redecha P, Salmon JE. Heparin prevents antiphospholipid
16. Manco Johnson MJ, Nuss R, Key N, et al. Lupus anticoagulant and antibody-induced fetal loss by inhibiting complement activation. Nat
protein S deficiency in children with postvaricella purpura fulminans Med 2004; 10:1222–6.
or thrombosis. J Pediatr 1996; 128:319–23. 36. Carreras LO, Vermylen JG. ‘Lupus’ anticoagulant and thrombosis –
17. Guglielmone H, Vitozzi S, Elbarcha O, et al. Cofactor dependence and possible role of inhibition of prostacyclin formation. Thromb
isotype distribution of anticardiolipin antibodies in viral infections. Haemost 1982; 48:38–40.
Ann Rheum Dis 2001; 60:500–4. 37. Robbins DL, Leung S, Miller-Blair DJ, et al. Effect of
18. de Groot PG, Horbach DA, Derksen RH. Protein C and other cofac- anticardiolipin/β2-glycoprotein I complexes on production of throm-
tors involved in the binding of antiphospholipid antibodies: relation boxane A2 by platelets from patients with the antiphospholipid syn-
to the pathogenesis of thrombosis. Lupus 1996; 5:488–93. drome. J Rheumatol 1998; 25:51–6.
19. Pengo V, Biasiolo A, Brocco T, et al. Autoantibodies to phospholipid- 38 Shoenfeld Y, Blank M. Effect of long-acting thromboxane receptor
binding plasma proteins in patients with thrombosis and phospho- antagonist (BMS 180,291) on experimental antiphospholipid syn-
lipid-reactive antibodies. Thromb Haemost 1996; 75:721–4. drome. Lupus 1994; 3:397–400.
20. Cuadrado MJ, Lopez-Pedrera C, Khamashta MA, et al. Thrombosis in 39. Fishman P, Falach-Vaknin E, Sredni B, et al. Aspirin–interleukin-3
primary antiphospholipid syndrome: a pivotal role for monocyte interrelationships in patients with anti-phospholipid syndrome. Am J
tissue factor expression. Arthritis Rheum 1997; 40:834–41. Reprod Immunol 1996; 35:80–4.
112
0415421306-Ch07 3/29/07 5:38 PM Page 113
40. Fishman P, Falach-Vaknin E, Sredni B, et al. Aspirin modulates inter- 46. Out HJ, Kooijman CD, Bruinse HW, et al. Histo-pathological findings
leukin-3 production: additional explanation for the preventive effects from patients with intrauterine fetal death and antiphospholipid
of aspirin in antiphospholipid antibody syndrome. J Rheumatol 1995; antibodies. Eur J Obstet Gynecol 1991; 41:179–86.
22:1086–90. 47. Ornoy A, Yacobi S, Matalon ST, et al. The effects of antiphospholipid
41. Blank M, George J, Fishman P, et al. Ciprofloxacin immunomo- antibodies obtained from women with SLE/APS and associated preg-
dulation of experimental antiphospholipid syndrome associated with nancy loss on rat embryos and placental explants in culture. Lupus
elevation of interleukin-3 and granulocyte–macrophage colony- 2003; 12:573–8.
stimulating factor expression. Arthritis Rheum 1998; 41:224–32. 48. Matalon ST, Shoenfeld Y, Blank M, et al. Antiphosphatidylserine
42. Bertolaccini ML, Atsumi T, Lanchbury JS, et al. Plasma tumor antibodies affect rat yolk sacs in culture: a mechanism for fetal loss
necrosis factor α levels and the 238A promoter polymorphism in in antiphospholipid syndrome. Am J Reprod Immunol 2004;
patients with antiphospholipid syndrome. Thromb Haemost 2001; 51:144–51.
85:198–203. 49. Shurtz-Swirski R, Inbar O, Blank M, et al. In vitro effect of anti-
43. Simantov R, LaSala JM, Lo SK, et al. Activation of cultured vascular cardiolipin autoantibodies upon total and pulsatile placental hCG
endothelial cells by antiphospholipid antibodies. J Clin Invest secretion during early pregnancy. Am J Reprod Immunol 1993;
1995; 96:2211–19. 29:206–10.
44. Meroni PL, Raschi E, Camera M, et al. Endothelial activation by aPL: 50. Katsuragawa H, Kanzaki H, Inoue T, et al. Monoclonal antibody
a potential pathogenetic mechanism for the clinical manifestations of against phosphatidylserine inhibits in vitro human trophoblastic
the syndrome. J Autoimmun 2000; 15:237–40. hormone production and invasion. Biol Reprod 1997; 56:50–8.
45. Pierangeli SS, Espinola RG, Liu X, et al. Thrombogenic effects of 51. Bernstein PS, Divon MY. Etiologies of fetal growth restriction. Clin
antiphospholipid antibodies are mediated by intercellular cell adhe- Obstet Gynecol 1997; 40:723–9.
sion molecule-1, vascular cell adhesion molecule-1, and P-selectin.
Circ Res 2001; 88:245–50.
113
0415421306-Ch07 3/29/07 5:38 PM Page 114
0415421306-Ch08 3/29/07 5:39 PM Page 115
8. Diagnosis of aPL-associated abortions

INTRODUCTION their immunoglobulin isotype (IgG, IgM, or IgA),

aPL units are defined as GPL, MPL, and APL units.
Antiphospholipid antibodies (aPL) are a heteroge- For the results to be reliable, the assays must be
neous group of autoantibodies directed against differ- properly standardized, standard calibrators must be
ent antigens, predominantly anionic phospholipids or used, and the range of normal values must derive
phospholipid-containing structures. aPL have been from measurements of aPL levels in a large number
associated with pregnancy disorders, including spon- of normal individuals.1
taneous miscarriage, recurrent miscarriage, preg- For the detection of aPL that prolong phospho-
nancy-induced hypertension, preeclampsia, and lipid-dependent clotting assays (lupus anticoagu-
intrauterine growth retardation. aPL can be detected lant, LA) clotting time prolongation assays are used,
using sensitive solid-phase immunoassays or coagula- in which prolongation of clotting is not corrected
tion tests. Both assays are easy to perform, but care is with normal plasma, but with the addition of phos-
required in the selection of the appropriate tests and in pholipids. These tests include the activated partial
interpretation of the results. In order to distinguish thromboplastin time (aPTT), the diluted Russell’s
between those women at risk of abortion and those viper venom time (dRVVT), and the kaolin clotting
not at risk, it is important to consider the following time (KCT).2
parameters: (a) the assays to be used, (b) the type and
(c) the isotype of aPL to be identified, (d) the antibody
level to be evaluated, (e) the interpretation of the TYPE OF aPL
results in relation to the heterogeneity of aPL and clin-
ical data, and (f) the timing for testing. Classically, the workup for the diagnosis of aPL-
associated abortions was limited to the detection of
LA (prolongation of at least one phospholipid-
ASSAYS FOR THE DETECTION OF aPL dependent clotting assay) and the detection of
β2GPI-dependent IgG and IgM aCL antibodies in the
aPL that do not prolong phospholipid-dependent serum (aCL-IgG >20 GPL units/ml or/and aCL-
clotting assays can be detected by immunoassays using IgM >20 MPL units/ml). LA and increased aCL-IgG
phospholipid-coated surfaces. Hence, antibodies antibodies are independently associated with recur-
against cardiolipin (aCL), phosphatidylethanolamine rent first- and second-trimester fetal loss and can be
(aPE), phosphatidylserine (aPS), phosphatodyl- used as prognostic and diagnostic markers.3 It has
choline (aPC), phosphatidylglycerol (aPG), phos- been reported that women positive for LA or aCL have
phatidylinositol (aPI), phosphatidic acid (aPA), and a 16% and 38% rate of pregnancy loss, respectively.4,5
β2-glycoprotein I (aβ2GPI) must be identified by It has recently become apparent that, although
standardized enzyme-linked immunosorbent assays extremely useful, the diagnostic tests for LA
(ELISA) using surfaces coated with the relevant and aCL may not be sufficient for diagnosing
phospholipid (usually complexed with cofactor all patients with aPL-associated pregnancy loss or
proteins). The results are expressed in aPL units, to elucidate the underlying pathology. A percentage
with 1 unit being equivalent to the binding capac- of women experiencing recurrent miscarriages may
ity of 1 μg/ml pure phospholipid. Depending on be negative for aCL but positive for other aPL.
115
0415421306-Ch08 3/29/07 5:39 PM Page 116
Several studies have shown that if only aCL are additional patients with RSA who had initial nega-
measured, 10–63% of positive APL are detected and tive tests for aCL (possibly because of the strong corre-
37–90% of women with reproductive autoimmune lation between the two autoantibodies).22 Finally,
failure syndrome (RAFS) will have the diagnosis the presence of more than one aPL appears to be a
missed.6–8 Although aPL other than LA and aCL more accurate variable than the presence of one aPL
have not been shown to be causative of pregnancy in predicting pregnancy loss.23
loss,9 it has been suggested that, at least for the diag-
nosis of RAFS-related abortions, a full aPL panel
should be measured (LA, aCL, aPE, aPS, aPC, aPG, ISOTYPE OF aPL
aPI, aPA, and aβ2GPI).10 The diagnostic and prog-
nostic value of some of these antibodies has been Most aPL are of the IgG or IgM class, but a small
well documented. It has been shown that the tro- proportion (10%) may be IgA. Some studies report
phoblastic layer directly in contact with the mater- a predominance of IgG antibodies in women with
nal circulation is more reactive with aPS than aCL,11 repeated miscarriages,18 while others have found
and that the aPS assay is more sensitive than the that the majority of aPL are of the IgM isotype.24
aCL assay for identifying women with autoimmune In a recent study, positivity for IgM aCL was found
abortions.12 The presence of IgG and IgM aPE anti- to show a stronger correlation to pregnancy outcome
bodies also appears to be a risk factor for early fetal than IgG aCL positivity.15 Abortion occurs more
losses (because of their effect on trophoblast forma- often in the presence of both IgG and IgM antibodies
tion) as well as for mid-to-late pregnancy loss (due to than in the presence of one isotype alone. Hence, in
binding to PE–kininogen complexes, which results order to avoid misdiagnosis and obtain helpful
in thrombin-induced platelet aggregation).13 information, testing should include at least the
Studies in women with recurrent miscarriage, detection of both IgG and IgM antibodies.
not receiving anticoagulation, have shown that LA
is mainly associated with second- rather than first-
trimester abortions.14 Moreover, the presence of TITER OF aPL
aCL in the absence of LA most likely reduces the
chance of live birth by 36–48% compared with the Since low levels of aPL may be found in otherwise-
absence of both aCL and LA,15 while the detection normal women, low titers detected in women with
of aCL before gestation or an increase during preg- repeated miscarriages cannot be considered as risk
nancy are considered as bad prognostic markers for factors for aPL-associated abortions (above the risk
the outcome of pregnancy.16 Measurement of aPS conferred by their medical history), and are of
and aPE is indicated in women with early recurrent doubtful clinical significance.25 In contrast,
pregnancy loss (RPL), since they represent those medium and high titers of aCL and/or other aPL
aPL that affect cell division during embryogenesis antibodies (>40 GPL units) identify the women
and the normal function of the trophoblast.17 The who would benefit from pharmacological prophylaxis
roles of other aPL (e.g., aPC, aPG, aPA, and aPI) in the next pregnancy. Eighty percent of women
and antibodies against cofactor proteins are less with very high levels of aCL (>80 GPL) and a history
clear, but they have been reported to have diagnostic of previous miscarriage(s) are expected to have fetal
significance. Some studies have shown a predomi- death in their next pregnancy.26
nance of aPC, aPG, aPA, and aPI in women with
recurrent spontaneous abortion (RSA),18–20 while
antibodies against cofactor proteins (prothrombin HETEROGENEITY OF aPL
and annexin V) have also been reported to be more
significant in reproductive failure than aCL alone.21 Depending on their type, aPL group members affect
Anti-β2GPI testing does not seem to identify pregnancy by acting through different pathways.
116
0415421306-Ch08 3/29/07 5:39 PM Page 117
DIAGNOSIS OF aPL-ASSOCIATED ABORTIONS
aCL, aPE, and aPS target relevant phospholipids on 4. Pattison NS, Chamley LW, McKay EJ, et al. Antiphospholipid antibodies
in pregnancy: prevalence and clinical associations. Br J Obstet
endothelial cells and lead to thrombosis in placental Gynaecol 1993; 100:909–13.
vessels. However, aPE and aPS also target relevant 5. Lockwood CJ, Romero R, Feinberg RF, et al. The prevalence and
phospholipids on the trophoblast and affect the biologic significance of lupus anticoagulant and anticardiolipin
antibodies in a general obstetric population. Am J Obstet Gynecol
formation of the syncytium. Furthermore, these 1989; 161:369–73.
antibodies and the aPC, aPG, and aPI target rele- 6. Yetman DL, Kutteh WH. Antiphospholipid antibody panels and recurrent
vant phospholipids in pre-embryonic tissues.27 The pregnancy loss: prevalence of anticardiolipin antibodies compared with
other antiphospholipid antibodies. Fertil Steril 1996; 66:540–6.
above data must be taken into account in the 7. Rote NS. Antiphospholipid antibodies other than lupus anticoagulant
workup. In second-trimester abortions where and anticardiolipin antibodies in women with recurrent pregnancy
thromboses have been found in the placental mate- loss, fertile controls, and antiphospholipid syndrome. Obstet Gynecol
1997; 90:642–4.
rial of previously missed abortions, it is advisable to 8. Coulam CB, Stern JJ, Kaider BD, et al. Comparison of frequency of
test for antibodies with anticoagulant effect on the anticardiolipin and other antiphospholipid antibodies among women
endothelium (aCL, aPS, and aPE), as there may be with reproductive autoimmune failure syndrome. Am J Reprod
Immunol 1997; 37:353–4.
classical antiphospholipid syndrome (APS), while 9. Branch DW, Silver R, Pierangeli S, et al. Antiphospholipid antibodies
in women with early pregnancy and pre-embryonic other than lupus anticoagulant and anticardiolipin antibodies in
losses, the work up should include aPL inhibiting women with recurrent pregnancy loss, fertile controls, and antiphos-
pholipid syndrome. Obstet Gynecol 1997; 89:549–55.
trophoblastic cell function (aPS and aPE) and 10. Coulam CB. The role of antiphospholipid antibodies in reproduction:
possibly those affecting implantation (aPS, aPE, questions answered and raised at the 18th Annual Meeting of the
aPC, aPI, aPG, and aPA). American Society of Reproductive Immunology. Am J Reprod
Immunol 1999; 41:1–4.
11. Lyden TW, Vogt E, Ng AK, et al. Monoclonal antiphospholipid anti-
body reactivity against human placental trophoblast. J Reprod
TIME OF TESTING Immunol 1992; 22:1–14.
12. Rote NS, Dostal-Johnson D, Branch DW. Antiphospholipid antibod-
ies and recurrent pregnancy loss: correlation between the activated
Finally, since aPL may appear transiently in normal partial thromboplastin time and antibodies against phosphatidylser-
individuals,28 it is recommended that in order to ine and cardiolipin. Am J Obstet Gynecol 1990; 163:575–84.
13. Sugi T, Matsubayashi H, Inomo A, et al. Antiphosphatidyletha-
establish the diagnosis of aPL-associated abortions, nolamine antibodies in recurrent early pregnancy loss and mid-to-late
increased titers of aPL must be present in the serum pregnancy loss. J Obstet Gynaecol Res 2004; 30:326–32.
in two measurements with an interval of 6 weeks 14. Carp HJ, Menashe Y, Frenkel Y, et al. Lupus anticoagulant. Significance
in habitual first-trimester abortion. J Reprod Med 1993; 38:549–52.
between the assays.29 Furthermore, it is important 15. Nielsen HS, Christiansen OB. Prognostic impact of anticardiolipin
to note that aPL that are increased during an unsuc- antibodies in women with recurrent miscarriage negative for the
cessful pregnancy may decrease afterwards. Hence, lupus anticoagulant. Hum Reprod 2005; 20:1720–8.
16. Kwak JY, Barini R, Gilman-Sachs A, et al. Down-regulation of maternal
aPL assays are of diagnostic value if performed antiphospholipid antibodies during early pregnancy and pregnancy
during pregnancy or at a time close to pregnancy outcome. Am J Obstet Gynecol 1994; 171:239–46.
loss. Occasionally, aCL seem to remain high outside 17. McIntyre JA. Antiphospholipid antibodies in implantation failures.
Am J Reprod Immunol 2003; 49:221–9.
pregnancy.30 18. Kaider AS, Kaider BD, Janowicz PB, et al. Immunodiagnostic evaluation
in women with reproductive failure. Am J Reprod Immunol 1999;
REFERENCES 42:335–46.
19. Ulcova-Gallova Z, Krauz V, Novakova P, et al. Anti-phospholipid
1. Tincani A, Allegri F, Balestrieri G, et al. Minimal requirements for antibodies against phosphatidylinositol, and phosphatidylserine are
antiphospholipid antibodies ELISAs proposed by the European Forum more significant in reproductive failure than antibodies against cardi-
on antiphospholipid antibodies. Thromb Res 2004; 114:553–8. olipin only. Am J Reprod Immunol 2005; 54:112–17.
2. Jennings I, Kitchen S, Woods TA, et al. Potentially clinically important 20. Franklin RD, Kutteh WH. Antiphospholipid antibodies (APA) and
inaccuracies in testing for the lupus anticoagulant: an analysis of results recurrent pregnancy loss: treating a unique APA positive population.
from three surveys of the UK National External Quality Assessment Hum Reprod 2002; 17:2981–5.
Scheme (NEQAS) for Blood Coagulation. Thromb Haemost 1997; 21. Bizzaro N, Tonutti E, Villalta D, et al. Prevalence and clinical correlation
77:934–7. of anti-phospholipid-binding protein antibodies in anticardiolipin-
3. Creagh MD, Malia RG, Cooper SM, et al. Screening for lupus anticoag- negative patients with systemic lupus erythematosus and women with
ulant and anticardiolipin antibodies in women with fetal loss. J Clin unexplained recurrent miscarriages. Arch Pathol Lab Med 2005;
Pathol 1991; 44:45–7. 129:61–8.
117
0415421306-Ch08 3/29/07 5:39 PM Page 118
22. Lee RM, Emlen W, Scott JR, et al. Anti-β2-glycoprotein I antibodies in 27. Coulam CB. Antiphospholipid antibody round table report. Am J
women with recurrent spontaneous abortion, unexplained fetal death, Reprod Immunol 2002; 48:262–5.
and antiphospholipid syndrome. Am J Obstet Gynecol 1999; 181:642–8. 28. Harris EN, Pierangeli S, Birch D. Anticardiolipin Wet Workshop
23. Aoki K, Hayashi Y, Hirao Y, et al. Specific antiphospholipid antibodies Report. Fifth International Symposium on Antiphospholipid
as a predictive variable in patients with recurrent pregnancy loss. Am Antibodies. Am J Clin Pathol 1994; 101:616–24.
J Reprod Immunol 1993; 29:82–7. 29. Coulam CB, Branch DW, Clark DA, et al. American Society for
24. Matzner W, Chong P, Xu G, et al. Characterization of antiphospho- Reproductive Immunology. Report of the Committee for Establishing
lipid antibodies in women with recurrent spontaneous abortions. Criteria for Diagnosis of Reproductive Autoimmune Syndrome. Am J
J Reprod Med 1994; 39:27–30. Reprod Immunol 1999; 41:121–32.
25. Silver RM, Porter TF, van Leeuween I, et al. Anticardiolipin antibodies: 30. Ruiz JE, Cubillos J, Mendoza JC, et al. Autoantibodies to phospho-
clinical consequences of ‘low titers’. Obstet Gynecol 1996; 87:494–500. lipids and nuclear antigens in non-pregnant and pregnant Colombian
26. Reece EA, Garofalo J, Zheng XZ, et al. Pregnancy outcome. Influence women with recurrent spontaneous abortions. J Reprod Immunol
of antiphospholipid antibody titer, prior pregnancy losses and treat- 1995; 28:41–51.
ment. J Reprod Med 1997; 42:49–55.
118
0415421306-Ch09 3/29/07 5:39 PM Page 119
9. Management of antiphospholipid syndrome

in pregnancy
Wendell A Wilson and Nigel Harris
there is room for debate about some aspects of cur-

INTRODUCTION rent therapy, because:
Appropriate management can greatly reduce the ● there have been relatively few controlled studies
fetal and maternal morbidity in antiphospholipid on which to base recommendations
syndrome (APS), an autoimmune disease that is now ● APS pregnancy is a somewhat heterogeneous
recognized as a leading cause of recurrent pregnancy clinical syndrome
loss (RPL). RPL occurs in about 1% of women.1,2 ● there was no general consensus on the definition
In about 15% of otherwise-healthy women, of APS pregnancy prior to publication of the
antiphospholipid antibodies (aPL) appear to be the Sapporo consensus criteria.13
sole explanation for their pregnancy loss.3,4 aPL
have also been associated with: Although there is consensus about the value of
heparin given concurrently with low-dose aspirin
● preterm birth, prior to 34 weeks’ gestation5,6 therapy, it is not clear what benefit, if any, is produced
● placental insufficiency and fetal growth by the aspirin component. Aspirin, given alone,
restriction5,6 did not improve the live birth rate compared with
● preeclampsia5,6 placebo,14,15 although the latter study of ‘low-risk
● venous, arterial, and small-vessel thrombophilia7,8 APS pregnancy’ included some pregnancies that fail
to satisfy current APS criteria. The study by Pattison
The aims of therapy of APS during pregnancy et al14 included aPL-positive women who had three
include the improvement of all these morbidities. or more prior pregnancy losses (predominantly in
the first trimester) and no history of thromboses
or systemic lupus erythematosus (SLE); it showed a
UNFRACTIONATED HEPARIN AND relatively high live birth rate of about 80% using
LOW-MOLECULAR-WEIGHT HEPARIN either aspirin or placebo, suggesting that there may
indeed be a subgroup of women with ‘low-risk APS
At present, maternally administered unfractionated pregnancy’ who may not need anticoagulant or
heparin (UH) or low-molecular-weight heparin antiplatelet therapy. Three studies have compared
(LMWH) given concurrently with low-dose aspirin the effect of aspirin alone with the effect of aspirin
are generally considered standard therapy to prevent with the addition of heparin. Two studies16,17
pregnancy loss in women with APS.9–12 This treat- showed a statistically significant benefit from adding
ment has been endorsed by an international panel heparin to the aspirin. An 80% live birth rate was
of experts, and it is usually started pre-pregnancy seen after heparin and aspirin, compared with 44%
or as early as possible in the first trimester, after after aspirin alone in Kutteh’s trial,16 and a 71% live
ultrasound demonstration of a live embryo. The live birth rate after heparin compared with 42% in
birth rate in APS is improved by about 50% with the trial by Rai et al.17 However, in the trial by
this therapy to about 80% of treated pregnancies. Farquharson et al,18 the live birth rate was similar
Notwithstanding the foregoing recommendations, whether aspirin was used alone or in combination
119
0415421306-Ch09 3/29/07 5:39 PM Page 120
with heparin. When the three trials were combined full therapeutic dose (e.g., enoxaparin 1mg/kg body
in a meta-analysis,19 there was a common odds ratio weight every 12 hours). Some authorities suggest
of 2.63 in favor of adding heparin to the regimen that women with prior late fetal loss (10 weeks ges-
(95% confidence interval (CI) 1.46–4.75). tation or later) and no prior history of thrombosis
The dose of heparin therapy needed for APS should also receive full or almost full therapeutic
pregnancy is also debatable.9,20 More aggressive, full- UH or LMWH doses, because these women may
dose heparin therapy may be appropriate for women also have a higher risk of thromboembolism during
who have a prior history of either venous, arterial, or or after pregnancy than women who experience
small-vessel thrombosis or pregnancy loss that early pregnancy losses.9,10,12,25 Our experience
occurred later than the 10th week of gestation, because with APS pregnancy has been in keeping with this
these groups of women appear to have an increased approach. It may therefore be useful to separate
risk of recurrence of thrombosis during pregnancy patients according to whether there is a history of
or the puerperium. However, there is some debate previous thrombosis, or whether the previous preg-
concerning the importance of the second factor. nancy losses were early or late.
The action of UH and LMWH in treating APS
pregnancy is thought to be mainly due to their ● 1. APS patients who have a prior history of
potent anticoagulant effects. APS is considered to be thrombosis. Most of these patients will be on
essentially an autoimmune thrombophilia that is long-term or lifelong thromboprophylaxis with
causally related to the persistent presence of hetero- warfarin. Before pregnancy is contemplated or
geneous antibodies directed against phospholipids attempted, these women should be informed of
and/or phospholipid-binding proteins.8,21 However, the teratogenic potential of warfarin during the
heparins may also bind directly to aPL,22,23 possibly first 14 weeks of pregnancy. Prior to conception,
inhibiting their activity in causing APS. In addition, warfarin should be replaced with a full therapeutic
recent studies in experimental models suggest that dose of UH or LMWH, supervised by a clinician
aPLs may in part cause APS by activating the com- with expertise in managing APS pregnancy.
plement system and that heparin may prevent fetal ● 2. APS patients who have at least one fetal
loss in APS by inhibiting complement activation, in loss after the 10th week of gestation, no prior
particular the formation of the activation products thromboses, and no other features of APS.
C3a and C5a;24 it is possible that C3a and C5a Although these patients have not had prior
amplify the procoagulant effects of aPL, thus caus- thromboses, some studies suggest that they may
ing thrombophilia. Detailed studies are needed to also be at significantly increased risk of
assess the relevance of these mechanisms to human thrombosis during pregnancy or the puerperium,
pregnancy. and should be managed similarly to group 1
Consensus guidelines for the management of (see above) patients during pregnancy and the
APS pregnancy were published in 2003 by a study puerperium.9,10,12,25
group from the International Symposium on aPL.11 ● 3. APS patients with three or more prior early
In general, patients with recurrent (three or more) pregnancy losses only (prior to 10 weeks’ gestation)
prior early pregnancy loss (before the 10th week of and no other features of APS. RPL that occurs prior
gestation) who have no prior venous, arterial, or to the 10th week of gestation is frequently due to
small-vessel thrombosis are managed with UH or other causes than APS and is less specific for APS
LMWH in low doses (e.g., enoxaparin 1mg/kg/day). than pregnancy loss that occurs after 10 weeks.
Outside of the context of APS, this dosage is gener- There is a paucity of controlled studies that
ally considered prophylactic rather than therapeutic address the management of recurrent pregnancy
in the prevention and management of venous loss prior to 10 weeks’ gestation. However, if
thrombosis. APS patients with prior venous, arte- other causes of fetal loss have been excluded,
rial, or small-vessel thrombosis are managed with a these patients should be managed with LMWH
120
0415421306-Ch09 3/29/07 5:39 PM Page 121
MANAGEMENT OF ANTIPHOSPHOLIPID SYNDROME IN PREGNANCY
in lower doses (1 mg/kg/day) and low-dose glucocorticoids waned after a small randomized
aspirin, begun prior to conception.9,10 trial showed heparin to be as effective as high-dose
prednisone.30 In addition, most subsequent con-
RISKS OF THERAPY trolled studies have not confirmed a beneficial effect
for glucocorticoids in APS pregnancy.31–33 Well-
The risks of heparin therapy appear to be significantly known prednisone toxicities such as gestational dia-
reduced with LMWH. Hemorrhage, osteoporosis, betes, hypertension, preeclampsia, infections, and
and heparin-induced thrombocytopenia in particular osteoporosis contributed to the abandonment of
appear to be substantially less frequent with LMWH prednisone therapy for APS pregnancy.12,25,29
than with UH. The increased bioavailability and Prednisone therapy does have a place in treating the
longer therapeutic half-life of LMWH also allow less SLE in those women who have APS that is second-
frequent injections. Additionally, there is no overlap ary to SLE. Among such women with SLE requiring
between the anticoagulant effect and the antithrom- prednisone therapy, the lowest dose of prednisone
botic effect – hence, there is no bleeding with LMWH, should be used that produces satisfactory SLE con-
and little need for monitoring. Because of these fac- trol. Recent studies in experimental models of APS
tors and recent studies that suggest LMWH to be as pregnancy, mentioned above, point to a role for
effective in preventing pregnancy loss in APS,26–29 complement activation in causing fetal loss. If this
LMWH has to a large extent supplanted UH. mechanism, which is potentially both proinflam-
matory and procoagulant, is validated in human
MONITORING OF UH AND LMWH IN APS PREGNANCY pregnancy, it may provide novel therapeutic targets
in APS pregnancy.
Full-dose UH therapy requires monitoring of par-
tial thromboplastin time (PTT) or factor Xa level. HYDROXYCHLOROQUINE
Patients in whom the baseline PTT is prolonged
due to the presence of lupus anticoagulant in APS In the past, it was customary to discontinue hydrox-
should be monitored by factor Xa level. Full antico- ychloroquine during SLE pregnancy because of its
agulation with LMWH may require monitoring by presumed potential for fetal toxicity.34 However,
Xa level, since LMWH does not prolong the PTT. women with SLE who are taking hydroxychloro-
Further studies are needed to address the potential quine (Plaquenil) have been shown to be at
value of monitoring factor Xa level in APS pregnancy. increased risk of SLE flare if the hydroxychloroquine
is discontinued.34 The treating physician may not
recognize the true cause of such flares, because
OTHER THERAPIES the flare may be delayed for weeks to months after
the hydroxychloroquine is stopped. In addition,
In meta-analyses of APS pregnancy,12,25 heparin recent studies have shown that, contrary to previous
given concurrently with aspirin improved the live dogma, hydroxychloroquine is not associated with
birth rate from about 25% to 75%. These and other fetal toxicity when used in conventional doses35,36
studies have shown that pregnancy losses continue (≤6.5mg/kg lean body weight). Most authorities
in 20–30% of women, in spite of therapy. Alternate therefore now consider it preferable to continue
therapies have included glucocorticoids or intra- hydroxychloroquine throughout SLE pregnancy, to
venous immunoglobulin (IVIG). reduce the risk of an SLE flare. Such flares may have
adverse effects on fetal and maternal morbidity and
GLUCOCORTICOIDS increase prednisone requirements.34–37 In APS
patients who also have SLE, this approach is addi-
Glucocorticoid therapy predated the use of heparin tionally supported by studies showing that hydrox-
in treating APS pregnancy. Early enthusiasm for ychloroquine reduces the thrombogenic potential
121
0415421306-Ch09 3/29/07 5:39 PM Page 122
of aPL in a mouse model, possibly because of its pregnancies. aPL titers of both immunoglobulin G
antiplatelet effects.38 Although there have been no (IgG) and IgM were significantly decreased after
trials of hydroxychloroquine in APS pregnancy, there each infusion. The development of IUGR correlated
is evidence that hydroxychloroquine may reduce the with rising antibody titers. Again, future trials are
thrombosis risk associated with aPL in non-pregnant necessary in order to confirm these results, and to
SLE patients.39 determine whether the possible improvement in
obstetric complications is real and justifies using
IVIG, with its prohibitive cost. At present, it seems
‘PROBABLE APS’ PREGNANCY that IVIG may have a place as a second line of treat-
ment in patients who are refractory to heparin or
The live birth rate does not appear to be improved who continue to suffer the late obstetric complica-
by heparin therapy in women who do not fulfil clin- tions of APS.
ical criteria for APS.14,15 In particular, heparin ther-
apy does not improve the live birth rate in women
who have only one or two prior early pregnancy TREATMENT OF REFRACTORY APS
losses (<10 weeks’ gestation) with a low titer of aPL
and no other features of APS.14,15 As stated above, pregnancy losses continue in
20–30% of women, despite conventional therapy
with heparin and aspirin. There are few definitive
INTRAVENOUS IMMUNOGLOBULIN answers regarding therapy in these patients. The
clinical features of the subsequent pregnancy loss
IVIG therapy has been of interest because of its effi- may give some insight into appropriate manage-
cacy in treating other autoimmune syndromes, and ment. If the treated pregnancy terminates in a
recent experimental studies that suggest it may first-trimester missed abortion, there may be a con-
improve thrombogenicity in experimental APS. founding factor such as unrelated structural mal-
However, clinical trials have found that it is no more formations46 or chromosome aberrations in the
effective in preventing pregnancy loss than heparin embryo. Ogasawara et al47 have reported that 40%
with aspirin in unselected APS pregnancy.40–42 The of abortuses had chromosomal aberrations in their
value of IVIG in the subgroup of women miscarrying series of 10 patients with recurrent miscarriage and
despite conventional therapy needs to be adequately APS. Takakuwa et al48 found 20% of abortuses to be
addressed by future studies. chromosomally abnormal in their series of 10 patients
Some preliminary observational data have with APS. If the patient aborts a chromosomally
reported that IVIG may lower the incidence of the aberrant fetus, it may be a chance event unrelated to
late obstetric complications described above com- APS. In these circumstances, the patient should be
pared with heparin and low-dose aspirin. These given the same regimen of LMWH again with or
complications include preeclampsia,41–44 intrauter- without aspirin.
ine growth restriction (IUGR),40,43–45 and prema- If the treated pregnancy terminates in a further
ture births,43,44 leading to fewer admissions to the midtrimester fetal death, the likelihood of chromo-
neonatal intensive care unit (14% of seven patients somal aberrations is far less, and there may be a true
in the IVIG group compared with 44% of nine in failure of treatment. In these cases, it must be remem-
the placebo group) in the trial by Branch et al.40 bered that the optimal doses of heparins are still to be
Kwak et al45 used IVIG in six patients who were worked out, and a higher dose may be necessary.
refractory to other forms of treatment and had However, the use of higher doses of LMWH should
complications in late pregnancy. Three pregnancies be weighed against the risk of side-effects. If the preg-
were complicated by IUGR, another three had twin nancy loss was accompanied by other autoimmune
122
0415421306-Ch09 3/29/07 5:39 PM Page 123
phenomena, including complications of SLE, or Obstetricians and Gynecologists (ACOG) guide-

vasculitis, steroids should be added to the regimen. lines29, 52 include the following:
Another alternative is IVIG. The addition of either
of these two drugs should be considered if artificial ● kidney and liver function baseline studies (e.g.,
termination of pregnancy was required due to other creatinine clearance, urine protein, and hepatic
autoimmune phenomena. However, if the preg- enzymes)
nancy required artificial termination for a maternal ● education about thrombosis, preeclampsia, and
indication, the patient should be carefully coun- heparin therapy
selled as to whether another pregnancy is warranted. ● monitoring of the platelet count in patients on
The past obstetric history will help the patient and heparin
physician arrive at an optimal decision. If the preg- ● monitoring for preeclampsia
nancy loss is due to obstetric complications such as ● after the late second trimester, sonograms every
preeclampsia or abruption, IVIG may be the appro- 3–4 weeks
priate choice for use in the next pregnancy.49 ● consideration of umbilical Doppler flow evalua-
Plasmapheresis has also been described as a method tions or more frequent fetal heart monitoring if
for washing out the antibodies.50,51 However, IUGR is suspected
plasmapheresis is considered to be a salvage therapy ● early delivery for fetal or maternal compromise
for pregnant women with secondary APS or cata-
strophic APS.
It is against this background of uncertainty that CONCLUSIONS
the results of treatment should be interpreted. There
are no clear guidelines regarding optimal therapy in It is clear from the above that the current manage-
failure of treatment; however, skill and clinical judgment of APS pregnancy improves the pregnancy
ment will be necessary to manage the following outcome, especially with regard to the live birth
pregnancy. rate. However, treatment is not uniformly effec-
tive, and heparin injections are inconvenient,
expensive, and generally painful. Occasional inter-
LABOR AND POSTPARTUM nal bleeding remains a problem that has been
reduced, but not eliminated, by using LMWH.
It has been reported that LMWH should be stopped However, it is gratifying that although the infants
at 36 weeks’ gestation to avoid the risk of epidural of APS mothers treated by current methods have
hematoma in patients needing epidural block, and an increased risk of prematurity, they appear to
UH should be substituted.25,52 However, LMWH is grow and develop quite normally when followed
often continued until 1 day before labor begins. If for up to 5 years of age.53
the patient has been free of LMWH, epidural anal-
gesia is probably quite safe. After labor has begun, REFERENCES
UH is stopped.29,52 Aspirin therapy is usually

1. Alberman E. The epidemiology of repeated abortion. In: Beard RW,
stopped at 34 weeks’ gestation to allow ductus arte- Sharp F, eds. Early Pregnancy Loss: Mechanisms and Treatment.
riosus closure. London: Springer-Verlag; 1988: 9–17.
2. Salat-Baroux J. Recurrent spontaneous abortions. Reprod Nutr Dev
1988; 28:1555–68. [in French]
MONITORING OF PREGNANCY 3. Stephenson MD. Frequency of factors associated with habitual abor-
tion in 197 couples. Fertil Steril 1996; 66:24–9.
Prompt management of potential problems is 4. Yettman DL, Kutteh WH. Antiphospholipid antibody panels and
recurrent pregnancy loss: prevalence of anticardiolipin antibodies
facilitated by patient education and monitoring as compared with other antiphospholipid antibodies. Fertil Steril 1996;
for high risk pregnancies. The American College of 66: 540–6.
123
0415421306-Ch09 3/29/07 5:39 PM Page 124
5. Branch WD, Silver RM, Blackwell JL, et al. Outcome of treated preg- 26. Triolo G, Ferrante A, Ciccia F, et al. Randomized study of subcuta-
nancies in women with antiphospholipid syndrome: an update of the neous low molecular weight heparin plus aspirin versus intravenous
Utah experience. Obstet Gynecol 1992; 80:614–20. immunoglobulin in the treatment of recurrent fetal loss associated
6. Lima F, Khamashta MA, Buchanan NM, et al. A study of sixty preg- with antiphospholipid antibodies. Arthritis Rheum 2003; 48:728–31.
nancies in patients with the antiphospholipid syndrome. Clin Exp 27. Stephenson MD, Ballem PJ, Tsang P, et al. Treatment of antiphospho-
Rheumatol 1996; 14:131–6. lipid antibody syndrome (APS) in pregnancy: a randomized pilot trial
7. Cervera R, Piette JC, Font J, et al. Antiphospholipid syndrome: clini- comparing low molecular weight heparin to unfractionated heparin.
cal and immunologic manifestations and patterns of disease expres- J Obstet Gynaecol Can 2004; 26:729–34.
sion in a cohort of 1000 patients. Arthritis Rheum 2002; 46:1019–27. 28. Gris JC, Balducchi JP, Quere I, et al. Enoxaparin sodium improves
8. Levine JS, Branch DW, Rauch J. The antiphospholipid syndrome. pregnancy outcome in aspirin-resistant antiphospholipid/antiprotein
N Engl J Med 2002; 346:752–63. antibody syndromes. Thromb Haemost 2002; 87:536–7.
9. Branch DW, Khamashta MA. Antiphospholipid syndrome: obstetric 29. American College of Obstetricians and Gynecologists. Anticoagulation
diagnosis, management, and controversies. Obstet Gynecol 2003; with low-molecular-weight heparin during pregnancy. ACOG
101:1333–44. Committee Opinion 211. Washington, DC: ACOG, 1998.
10. Derksen RH, Khamashta MA, Branch DW. Management of 30. Cowchock FS, Reece EA, Balaban D, et al. Repeated fetal losses
the obstetric antiphospholipid syndrome. Arthritis Rheum 2004; associated with antiphospholipid antibodies: a collaborative random-
50:1028–39. ized trial comparing prednisone with low-dose heparin treatment.
11. Tincani A, Branch W, Levy RA, et al. Treatment of pregnant patients Am J Obstet Gynecol 1992; 166:1318–23.
with antiphospholipid syndrome. Lupus 2003; 12:524–9. 31. Lockshin MD, Druzin ML, Qamar T. Prednisone does not prevent
12. Lassere M, Empson M. Treatment of antiphospholipid syndrome in recurrent fetal death in women with antiphospholipid antibody.
pregnancy – a systematic review of randomized therapeutic trials. Am J Obstet Gynecol 1989; 160:439–43.
Thromb Res 2004; 114:419–26. 32. Laskin CA, Bombardier C, Hannah ME, et al. Prednisone and aspirin
13. Wilson WA, Gharavi AE, Koike T, et al. International consensus in women with autoantibodies and unexplained recurrent fetal loss.
statement on preliminary classification for definite antiphospholipid N Engl J Med 1997; 337:148–53.
syndrome. Arthritis Rheum 1999; 42:1309–11. 33. Silver RK, MacGregor Sholl, JS, et al. Comparative trial of prednisone
14. Pattison NS, Chamley LW, Birdsall M, et al. Does aspirin have a role plus aspirin versus aspirin alone in the treatment of anticardiolipin
in improving pregnancy outcome for women with the antiphospho- antibody-positive obstetric patients. Am J Obstet Gynecol 1993;
lipid syndrome? A randomized controlled trial. Am J Obstet Gynecol 169:1411–17.
2000; 183:1008–12. 34. Canadian Consensus Conference on Hydroxychloroquine. J Rheumatol
15. Cowchock S, Reece EA, Balaban D, et al. Do low risk pregnant women 2000; 27:2919–21.
with antiphospholipid antiboidies need to be treated? Organising 35. Motta M, Tincani A, Faden D, et al. Follow-up of infants exposed to
Group of the Antiphospholipid Treatment Trial. Am J Obstet Gynecol hydroxychloroquine given to mothers during pregnancy and lactation.
1997; 176:1099–100. J Perinatol 2005; 25:86–9.
16. Kutteh WH. Antiphospholipid antibody-associated recurrent preg- 36. Khamashta MA, Buchanan NMM, Hughes GRV. The use of hydroxy-
nancy loss: treatment with heparin and low-dose aspirin is superior to chloroquine in lupus pregnancy the British experience. Lupus 1996;
low-dose aspirin alone. Am J Obstet Gynecol 1996; 174:1584–9. S1:65–6.
17. Rai R, Cohen H, Dave M, et al. Randomised controlled trial of aspirin 37. Parke A, West B. Hydroxychloroquine in pregnant patients with
and aspirin plus heparin in pregnant women with recurrent miscar- systemic lupus erythematosus. J Rheumatol 1996; 23:1715–18.
riage associated with phospholipid antibodies (or antiphospholipid 38. Espinola RG, Pierangeli SS, Ghara AE, et al. Hydroxychloroquine
antibodies). BMJ 1997; 314:253–7. reverses platelet activation induced by human IgG antiphospholipid
18. Farquharson RG, Quenby S, Greaves M. Antiphospholipid syndrome antibodies. Thromb Haemost 2002; 87:518–22.
in pregnancy: a randomized, controlled trial of treatment. Lupus 39. Petri M. Thrombosis and systemic lupus erythematosus: the Hopkins
2002; 100:408–13. Lupus Cohort perspective. Scand J Rheumatol 1996; 25: 191–3.
19. Carp HJA. Antiphospholipid syndrome in pregnancy. Curr Opin 40. Branch DW, Peaceman AM, Druzin M, et al. A multicenter, placebo-
Obstet Gynaecol 2004; 16:129–35. controlled pilot study of intravenous immune globulin treatment of
20. Erkan D, Merrill JT, Yazici Y, et al. High thrombosis rate after fetal loss antiphospholipid syndrome during pregnancy. Am J Obstet Gynecol
in antiphospholipid syndrome. Arthritis Rheum 2001; 44:1466–7. 2000; 182:122–7.
21. Gharavi AE, Pierangeli S, Wilson WA. The molecular basis of 41. Vaquero E, Lazzarin N, Valensise H, et al. Pregnancy outcome in
antiphospholipid syndrome. Lupus 2003; 12:579–83. recurrent spontaneous abortion associated with antiphospholipid
22. Ermel LD, Marshburn PB, Kutteh WH. Interaction of heparin with antibodies: a comparative study of intravenous immunoglobulin
antiphospholipid antibodies (APA) from the sera of women with recur- versus prednisone plus low-dose aspirin. Am J Reprod Immunol
rent pregnancy loss (RPL). Am J Reprod Immunol 1995; 33:14–20. 2001; 45:174–9.
23. Franklin RD, Kutteh WH. Effects of unfractionated and low molecu- 42. Spinnato JA, Clark AL, Pierangeli SS, et al. Intravenous immunoglob-
lar weight heparin on antiphospholipid antibody binding in vitro. ulin therapy for the antiphospholipid syndrome in pregnancy. Am J
Obstet Gynecol 2003; 101:455–62. Obstet Gynecol 1995; 172:690–4.
24. Girardi G, Redecha P, Salmon JE. Heparin prevents antiphospholipid 43. Harris EN, Pierangeli SS. Utilization of intravenous immunoglobulin
antibody-induced fetal loss by inhibiting complement system activa- therapy to treat recurrent pregnancy loss in the antiphospholipid
tion. Nat Med 2004; 10:1222–6. syndrome: a review. Scand J Rheumatol 1998; 107(Suppl):97–102.
25. Empson M, Lassere M, Craig JC, et al. Recurrent pregnancy loss with 44. Valensie H, Vaquero E, De Carolis C. Normal fetal growth in women
antiphospholipid antibody: a systematic review of therapeutic trials. with antiphospholipid syndrome treated with high dose intravenous
Obstet Gynecol 2002; 99:135–44. immunoglobulin (IVIG). Prenat Diagn 1995; 15:509–17.
124
0415421306-Ch09 3/29/07 5:39 PM Page 125
45. Kwak JY, Quilty EA, Gilman-Sachs A, et al. Intravenous immunoglobu- 50. Takeshita Y, Tsurumi Y, Touma S, et al. Successful delivery in a
lin infusion therapy in women with recurrent spontaneous abortions of pregnant woman with lupus anticoagulant positive systemic lupus
immune etiologies. J Reprod Immunol 1995; 28:175–88. erythematosus treated with double filtration plasmapheresis. Therap
46. Philipp T, Kalousek DK. Transcervical embryoscopy in missed abortion. Apher 2001; 5:22–4.
J Assist Reprod Genet 2001; 18:285–90. 51. Von Baeyer H. Plasmapheresis in immune hematology: review of
47. Ogasawara M, Aoki K, Okada S, et al. Embryonic karyotype of clinical outcome data with respect to evidence-based medicine and
abortuses in relation to the number of previous miscarriages. Fertil clinical experience. Therap Apher Dial 2003; 7:127–40.
Steril 2000; 73:300–4. 52. ACOG Practice Bulletin No. 68: Antiphospholipid Syndrome. Obstet
48. Takakuwa K, Asano K, Arakawa M, et al. Chromosome analysis of Gynecol 2005; 106:1113–21.
aborted conceptuses of recurrent aborters positive for anticardiolipin 53. Ruffatti A, Dalla Barba B, Del Ross T, et al. Outcome of fifty-five
antibody. Fertil Steril 1997; 68:54–8. newborns of antiphospholipid antibody-positive mothers treated
49. Galli M, Barbui T. Antiphospholipid antibodies and pregnancy. Best with calcium heparin during pregnancy. Clin Exp Rheumatol 1998;
Pract Res Clin Haematol 2003; 16:211–25. 16:605–10.
125
0415421306-Ch09 3/29/07 5:39 PM Page 126
0415421306-Ch10 3/29/07 5:40 PM Page 127
10. Defects in coagulation factors leading

to recurrent pregnancy loss
Aida Inbal and Howard JA Carp
INTRODUCTION and pregnancy loss. The various factors and their

association with the trophoblast are shown in
The etiology of pregnancy loss often remains an Figure 10.1.
enigma, even after exclusion of uterine abnormali-
ties and of genetic, immunological, infectious, or
endocrine disorders. Recently, thrombophilias, BLEEDING DIATHESES LEADING
whether hereditary or acquired, have been found in TO PREGNANCY LOSS
a significant number of women with recurrent
abortions without apparent cause. The evidence for HEREDITARY FACTOR XIII DEFICIENCY
pregnancy loss having a thrombotic basis is due to
the widely reported association between antiphos- Coagulation factor XIII (FXIII) is a plasma trans-
pholipid antibodies (aPL) and recurrent pregnancy glutaminase that participates in the final step of the
loss (RPL). aPL are thought to cause pregnancy loss coagulation cascade. Following activation by
by thrombosis in decidual vessels, impairing the thrombin, the active form (FXIIIa) crosslinks fibrin
blood supply to the fetus and leading to fetal death. chains through γ-glutamyl–ε-lysine bonds, creating
Due to the assumption that aPL induce thrombosis a stable clot resistant to fibrinolysis.1 In plasma,
causing pregnancy loss, it has been assumed that FXIII circulates as a heterotetramer (A2B2) com-
any prothrombotic state may also increase the posed of two catalytic A subunits (FXIII-A) and two
chance of pregnancy loss due to a thrombotic carrier B subunits (FXIII-B).2 FXIIIa is synthesized
mechanism, and that if this process recurs three by megakaryocytes, monocytes, and monocyte-
or more times, there is recurrent miscarriage. derived macrophages, whereas FXIII-B is synthe-
Hereditary thrombophilias that have been reported sized by hepatocytes. Platelets, monocytes, and
to be associated with recurrent pregnancy loss macrophages contain only the A subunits of FXIII
include antithrombin, protein C, and protein S defi- dimers.2
ciencies, factor V Leiden (FVL), the G20210A muta- In contrast to factors FVII, FVIII, FIX, FX, and
tion in the factor II (FII) gene, and homozygosity fibrinogen, which increase during normal preg-
for the thermolabile variant of methylenetetrahy- nancy, the concentration of plasma FXIII decreases
drofolate reductase (MTHFR C677T), which leads during pregnancy, reaching 50% of normal at term.
to hyperhomocysteinemia specifically in the pres- Likewise, the activity of FXIIIA is significantly
ence of low folate levels. In addition, deficiencies decreased at the time of abortion.3
of factor XIII (FXIII) and fibrinogen are associated FXIII deficiency is a hereditary bleeding disor-
with pregnancy loss. Both of these are bleeding der, characterized by severe bleeding manifesta-
diatheses that become apparent in childhood, and tions, delay in wound healing, and recurrent
are associated with impaired wound repair in addi- abortions in homozygous women.2 Women who are
tion to pregnancy loss and excessive bleeding. This homozygous for FXIII deficiency will not carry
chapter deals with the association between the pregnancy until term unless treated with FXIII
decreased or increased levels of coagulation factors concentrate throughout pregnancy.4 The minimal
127
0415421306-Ch10 3/29/07 5:40 PM Page 128
PROCOAGULANT EFFECT
FXIII HCY FNG ↑TG FVL FII CYT: IL-6, TNF-α MP
Syncytiotrophoblast
Cytotrophoblast
PC PS AT TFPI FS CYT: IL-4, IL-10
ANTICOAGULANT EFFECT
Figure 10.1 Procoagulant and anticoagulant balance of the trophoblast. FXIII, factor XIII; HCY, homocysteine; FNG, fibrinogen;
↑TG, increased thrombin generation; FVL, factor V Leiden; FII, prothrombin gene mutation (G20210A); CYT, cytokines; IL-6/4/10,
interleukin-6/4/10; TNF-α, tumor necrosis factor α; MP, microparticles; PC, protein C; PS, protein S; AT, antithrombin; TFPI, tissue
factor pathway inhibitor; FS, fibrinolytic system.
level of FXIII-A required for normal pregnancy is space of the extravillous cytotrophoblast shell
unknown; however, only 0.5–2% of FXIII-A is adjacent to Nitabuch’s layer. FXIII-A has also
required for normal hemostasis.2 been co-localized with fibrinogen and fibronectin
The mechanism by which FXIII supports normal at Nitabuch’s layer.9 FXIII-A has been reported
pregnancy is unknown. FXIII is essential for to be absent from the placental bed in women
implantation, placental attachment, and further with FXIII deficiency, leading to deficient cytotro-
placental development by crosslinking not only phoblastic shell formation.9 Thus, deficiency of
between fibrin chains but also between fibronectin FXIII-A at the site of implantation will adversely
and collagen, the major components of connective affect fibrin–fibronectin crosslinking, resulting in
tissue matrix.2,5 Hence, FXIII seems to play an detachment of the placenta from the uterus and
essential role in the interaction between the blasto- subsequent miscarriage.7,9 Recent studies have
cyst and the endometrium at implantation. FXIIIa shown FXIII-A to have proangiogenic activity both
also crosslinks fibrin(ogen) and fibronectin, both of in vitro and in vivo.10 Since embryo implantation
which are important for maintaining the attach- requires adequate angiogenesis, the supportive role
ment of the placenta to the uterus.6 FXIII deficiency of FXIII in implantation may be partly due to its
may result in periplacental hemorrhage and subse- proangiogenic activity.
quent spontaneous fetal loss. This hypothesis is Whatever the cause of pregnancy loss in FXIII-
supported by evidence from a mouse model of deficient women, administration of FXIII through-
FXIII deficiency: pregnant FXIII-A-subunit knockout pregnancy results in successful outcomes.2,4,5
out mice suffer excessive uterine bleeding followed A plasma-derived concentrate has been available
by embryonic demise.7 Kobayashi et al8 have since 1980. The FXIII concentrate seems to have a
reported that FXIII-A is present in the extracellular half-life of 10–12 days.11 Recently, a recombinant
128
0415421306-Ch10 3/29/07 5:40 PM Page 129
DEFECTS IN COAGULATION FACTORS LEADING TO RECURRENT PREGNANCY LOSS
FXIII-A-subunit protein has become available, with its half-life is 3–4.5 days.17 Thrombin cleaves
a half-life similar to that of the plasma-derived con- fibrinogen to fibrin monomer, which then poly-
centrate.12 merizes and is stabilized by FXIII. Fibrin(ogen) is
The timing and dose of FXIII replacement for also a target for fibrinolytic factors that dissolve
pregnant women and the optimal level of FXIII excess fibrin to maintain vascular patency and
remain unknown. The level of plasma FXIII gener- integrity. Fibrinogen is also a primary bridging
ally achieved for successful pregnancy is 10% in molecule, linking activated platelets together via
women with FXIII deficiency.11 We treat pregnant their glycoproteins (GP) IIb/IIIa.18
women prophylactically with 20 IU/kg of FXIII The three overlapping hereditary abnormalities
concentrate every 4 weeks to achieve a FXIII level of of fibrinogen – afibrinogenemia, dysfibrinogenemia,
above 3%. A booster dose of 1000 IU is also given and hypofibrinogenemia – have been associated
before amniocentesis or labor. with recurrent pregnancy loss. Afibrinogenemia – a
defect in hepatic fibrinogen secretion or release – is
OTHER ALTERATIONS IN FXIII inherited as an autosomal recessive trait and is asso-
ciated with bleeding diathesis, impaired wound
It is unknown if there is an association between repair, and recurrent pregnancy loss. A related form
normal or decreased levels of FXIII and RPL. of this disorder is hypofibrinogenemia. Hereditary
Whereas plasma FXIII-B concentrations increase dysfibrinogenemias are characterized by the biosyn-
during pregnancy, FXIII-A tends to decrease, result- thesis of structurally and functionally abnormal
ing in an overall steady reduction in plasma FXIII, fibrinogen.
reaching approximately 50% of normal at term.13 Brenner19 has reported that women with dysfib-
The A subunit rises with the onset of labor and falls rinogenemia are candidates for miscarriage. Of 64
postpartum.13 This is in contrast to the progressive pregnancies in women with dysfibrinogenemia,
increase in levels of fibrinogen, FVII, FVIII, FIX, 39% terminated in miscarriage. The mechanisms
and FX during pregnancy.14 Whether the reduction whereby dysfibrinigenemias are associated with a
of plasma FXIII during pregnancy represents tendency to thrombosis have been reviewed by
decreased synthesis of FXIII-A, increased utilization Mosesson.20
or destruction, or simple dilution by the expanded Hypofibrinogenemic women21 and experimental
plasma volume is not clear. In a cohort of non- afibrinogenemic mice22 show similar features with
FXIII-deficient women with a history of two or regard to bleeding tendency, miscarriage, and
more first-trimester miscarriages, plasma FXIII abnormal scar formation. Based on the mouse
levels were not found to be predictive for subse- model, absence or a significant decrease in maternal
quent pregnancy loss.15 A substitution of tyrosine fibrinogen is sufficient to cause rupture of the
by phenylalanine at position 204 in exon 5 of the maternal vasculature, thereby affecting embryonic
FXIII-A gene was found in one study to be more trophoblast infiltration leading to hemorrhage and
prevalent in women suffering three or more miscar- subsequent miscarriage.
riages.16 The Phe204 FXIII-A variant has been asso- Cryoprecipitate, fresh-frozen plasma and fibrino-
ciated with lower specific activity. However, in gen concentrate are the sources of fibrinogen that
subsequent studies, this association has not been are commercially available. Replacement therapy
confirmed. throughout pregnancy is feasible for patients with
pregnancy losses.23 It has been suggested that the
FIBRINOGEN DEFICIENCY minimal level of normal fibrinogen to maintain
pregnancy is about 60 mg/100 ml.24 A cryoprecipi-
Fibrinogen, a major blood glycoprotein, is a dimer tate infusion of 0.2 bags/kg body weight (approxi-
of three polypeptide chains: Aα, Bβ, and γ. It is mately 250 mg/bag) will raise the fibrinogen
synthesized by hepatic parenchymal cells and concentration to 100 mg/dl. Since the half-life of
129
0415421306-Ch10 3/29/07 5:40 PM Page 130
fibrinogen is approximately 4 days, two weekly This mutation leads to more efficient mRNA
infusions of cryoprecipitate during the gestational processing of the prothrombin gene, which in turn
period should be sufficient to keep the fibrinogen is associated with an increased level of prothrombin
level above 60 mg/dl and prevent pregnancy loss. and generation of thrombin. FVL and the G20210A
The benefits of substitution therapy should be mutation in the prothrombin gene are common
weighed against the possibility of inducing throm- among healthy whites (with prevalences of 5% and
bosis. Catastrophic thrombosis has been reported 1.5%, respectively), but are rare in Asians and
during fibrinogen replacement therapy in patients Africans. Homozygosity for MTHFR (C677T) may
with afibrinogenemia and dysfibrinogenemia.25 lead to hyperhomocysteinemia, mainly when folate
Prophylactic heparin or low-molecular-weight storage is decreased, which may also predispose to
heparin (LMWH) has been advocated for the thrombosis. The mechanism is multifactorial.
peripartum period in these patients.26
THROMBOSIS IN DECIDUAL VESSELS
THROMBOPHILIAS Thrombophilia has been suggested to be a cause

for microembolism in the placenta resulting in
The evidence for pregnancy loss having a throm- abortion or adverse outcome of pregnancy.28
botic mechanism rests on three pillars: increased However, it remains an assumption that hereditary
prevalence of thrombophilias in RPL, a higher inci- thrombophilias lead to thrombosis in placental ves-
dence of pregnancy loss in the presence of throm- sels in RPL, as no group has assessed the placenta
bophilias, and the demonstration of thrombosis in in recurrent miscarriage with hereditary throm-
decidual vessels. The hereditary thrombophilias bophilias. Genetic polymorphisms of the throm-
cause an increased tendency to venous thrombosis bophilic genes of the parents have a 50% likelihood
and comprise a number of conditions, such as of transmission to the fetus, potentially affecting
antithrombin, protein C, and protein S deficiencies, trophoblast function. Thus, to determine the true
FVL, prothrombin gene (FII) mutation G20210A, risk for adverse pregnancy outcome associated with
homozygosity for the MTHFR mutation C677T, genetic thrombophilias, it is necessary to test the
and increased FVIII. There are also various acquired offspring for these thrombophilias. In 1978,
hypercoagulable states, the most common of which Rushton29 reported histological findings in 1486
is antiphospholipid syndrome (APS), which is dis- abortuses. Thrombosis was found in the placenta of
cussed elsewhere in this book. Proteins C and S and 12.1% of abortuses examined. However, at that
antithrombin are physiological anticoagulants. time, there was no possibility of examining heredi-
Deficiencies of these anticoagulants are uncom- tary thrombophilias. Many et al28 have compared
mon.27 FVL is the most common cause of inherited the placental findings in women with severe preg-
thrombophilia.27 It results from the substitution of nancy complications with and without throm-
adenine for guanine at nucleotide 1691 of the factor bophilias. The number of women with villous and
V gene (G1691A), which causes the arginine in multiple infarcts was significantly higher in women
residue 506 of the factor V protein to be replaced by with thrombophilias. The number of placentas with
glutamine (Arg506Gln). The resulting protein is fibrinoid necrosis of the decidual vessels was also
called factor V Leiden. This mutation slows down significantly higher in women with thrombophilias.
the proteolytic inactivation of FVa, by activated However, a study by Mousa and Alfirevic30 could
protein C (termed activated protein C resistance, not confirm these results, but found a high inci-
APCR), which in turn leads to augmented generation dence of placental infarcts (50%) and thrombosis in
of thrombin. In the G20210A mutation of the both women with and women without thrombophil-
prothrombin gene, adenine is substituted for guanine ias. Arias et al31 evaluated 13 placentas of women
at the 3′ untranslated part of the prothrombin gene. with preeclampsia, preterm labor, intrauterine growth
130
0415421306-Ch10 3/29/07 5:40 PM Page 131
restriction (IUGR), or stillbirth. Of 13 women, 10 polymorphisms may contribute to thrombotic

(77%) had thrombophilias, including aPL, protein events in the placenta rather than to failure of
C, protein S, and antithrombin deficiencies, APCR, implantation. However, no study has assessed the
and FVL. However, rather than decidual thrombo- placenta in first-trimester pregnancy loss in the
sis, a fetal thrombotic vasculopathy was found with presence of thrombophilia, compared with second-
fibrotic villi, or stem villi obliterated by fibrous trimester losses, nor has any study assessed the
tissue. It is important to note that these histological placenta in the presence of genetic pregnancy loss
changes are on the fetal side of the placenta, not the compared with pregnancy losses with a normal
maternal side. There was also fetal stem vessel karyotype.
thrombosis, infarcts, hypoplasia, spiral artery
thrombosis, and perivillous fibrin deposition. The PREVALENCE OF THROMBOPHILIAS IN
fact that no specific placental lesion has been found FIRST-TRIMESTER MISCARRIAGE
in thrombophilia could have a number of explana-
tions. There may be other thrombophilias as yet Opinions are divided whether thrombophilias are
unknown, which could explain the high incidence more prevalent in women with any pregnancy loss
of placental pathology, or the lesions may be the or RPL. Prevalences have been reported to range
result of inflammatory changes in the placenta between 3% and 42%. Part of the confusion may
associated with the underlying pathology, and unre- stem from the fact that there are studies comparing
lated to thrombophilia. Even in APS, thrombosis the prevalence of thrombophilias in all forms of
has not been convincingly demonstrated in decid- pregnancy loss with parous patients. Other studies
ual vessels. On the contrary, after treatment with have sought differences in prevalences of throm-
monoclonal aPL, stained placental sections have bophilias in recurrent first-trimester losses
shown most reactivity to be localized to the cytotro- compared with late losses. Others have sought the
phoblast, suggesting that the trophoblast may be prevalence in patients with late obstetric complica-
directly damaged by mechanisms unrelated to tions. Some papers have found an increased preva-
thrombosis.32 As in hereditary thrombophilias, lence of certain thrombophilias, but not others.
these histological changes were on the fetal, rather The papers describing the prevalences in differ-
than maternal, side of the placenta. It seems that ent forms of RPL and pregnancy complications
cell surface-associated membrane receptors rather have been described by Kupferminc.34 When preg-
than soluble factors (e.g., thrombophilic factors) nancy loss is not broken down into subgroups,
are most relevant candidates to affect pregnancy thrombophilias seem to be more prevalent. Brenner
outcome.33 et al35 tested women with three or more first-
The maternal spiral arteries become remodeled trimester losses, two or more second-trimester
by pregnancy hormones and the trophoblast into losses, or one or more third-trimester loss. FVL was
uteroplacental arteries toward the end of the first more prevalent in the pregnancy loss group than in
trimester. In the uteroplacental arteries, the lumen controls; however, neither the MTHFR C677T nor
is larger, and the media is replaced by endovascular prothrombin mutations were more common in
trophoblast cells. If there is thrombosis of the women with pregnancy loss than controls. Forty-
maternal uteroplacental arteries, it is by no means nine percent of women with pregnancy loss had a
certain that thrombosis can also occur in first- thrombophilia, compared with 22% of controls.
trimester arteries. It is possible that first-trimester Thrombophilias were more common in second-
miscarriage may be due to failure in the mecha- and third-trimester losses, but first-trimester
nisms governing implantation or due to chromoso- recurrent miscarriage was not associated with
mal or other abnormalities in the fetus, whereas thrombophilia. Due to the controversy over
second-trimester losses may be a consequence of the prevalence of thrombophilia and pregnancy
thrombotic events in the placenta. Prothrombotic loss, Rey et al36 carried out a meta-analysis of
131
0415421306-Ch10 3/29/07 5:40 PM Page 132
31 studies in the literature. There was a significant deficiency, and 2.0 (95% CI 0.5–7.7) in FVL muta-
association between hereditary thrombophilias and tion. Sarig et al40 evaluated 145 patients with
pregnancy loss. recurrent miscarriage and 145 matched controls. At
Similar disagreement is found in the literature least one thrombophilic defect was found in 66% of
with regard to recurrent miscarriage. The disagree- study group patients, compared with 28% in controls.
ments in the literature have prompted the need for Late pregnancy wastage occurred more frequently
meta-analyses to determine whether the prevalence in women with thrombophilia compared with
is increased in recurrent miscarriage. Krabbendam women without thrombophilia. Grandone et al41
et al37 have reported a meta-analysis of 11 studies investigated the FVL mutation in 43 women with
regarding the association between thrombophilias two or more unexplained fetal losses and 118 con-
and recurrent miscarriage. There were significantly trols. The FVL mutation was more frequent in
higher serum homocysteine levels among women women with second-trimester loss, but the preva-
with a history of recurrent miscarriage, but no lence of the mutation in women with first-trimester
increased prevalence of the MTHFR C677T muta- loss and controls was similar.
tion. No relation was observed for the levels of
antithrombin, protein C, or protein S. Nelen et al38 PREVALENCE OF THROMBOPHILIAS IN LATE
have performed a meta-analysis to assess the rela- OBSTETRIC COMPLICATIONS
tionship between recurrent early pregnancy loss
and hyperhomocysteinemia. Overall, the pooled Kupferminc et al42 and a systematic review of 25
odds ratio (OR) for elevated homocysteine was 2.7 studies by Alfirevic et al43 have reported that various
(95% confidence interval (CI) 1.5–5.2), for after- hereditary thrombophilias are more prevalent in
load homocysteine it was 4.2 (95% CI 2.0–8.8), and pregnant women with IUGR, preeclampsia, abrup-
for MTHFR it was 1.4 (95% CI 1.0–2.0). These data tio placentae, or stillbirth. In a case–control study of
support hyperhomocysteinemia as a risk factor for 232 women with a history of one or more second-
recurrent early pregnancy loss. or third-trimester losses by Gris et al,44 21.1% of
There are some publications separating early patients and 3.9% of controls had at least one
and late pregnancy losses and the prevalence of thrombophilia (p < 0.00001). The OR for stillbirth
thrombophilias. Preston et al39 reported on hered- associated with any positive thrombophilia was 5.5
itary thrombophilias and fetal loss in a cohort (95% CI 3.4–9.0). Logistic regression analysis
of women with FVL or deficiencies of antithrom- showed four risk factors for stillbirth: protein S
bin, protein C, or protein S. Of 843 women with deficiency, positive anti-β2-glycoprotein I (β2GPI)
thrombophilia, 571 had 1524 pregnancies; of IgG antibodies, positive anticardiolipin IgG anti-
541 control women, 395 had 1019 pregnancies. bodies, and the FVL mutation. The conclusion was
The incidences of miscarriage (fetal loss at or that late fetal loss, through placenta thrombosis,
before 28 weeks of gestation) and stillbirth (fetal might sometimes be the consequence of a maternal
loss after 28 weeks of gestation) were assessed multifactorial prothrombotic state. Many et al45
jointly and separately. The risk of fetal loss was investigated women with intrauterine fetal death
increased in women with thrombophilia (OR 1.35; (IUFD) at 27 weeks of gestation or more. In 40
95% CI 1.01–1.82). The OR was higher for still- women with unexplained IUFD, the prevalence of
birth than for miscarriage: 3.6 (95% CI 1.4–9.4) inherited thrombophilias was 42.5% in the study
versus 1.27 (95% CI 0.94–1.71), respectively. The group, compared with 15% in controls (OR 2.8;
highest OR for stillbirth was in women with com- 95% CI 1.5–5.3; p = 0.001). However, this increased
bined defects, 14.3 (95% CI 2.4–86.0), compared prevalence has been disputed by Infante-Rivarde
with 5.2 (95% CI 1.5–18.1) in antithrombin et al.46 A systematic review by Alfirevic et al43 has
deficiency, 2.3 (95% CI 0.6–8.3) in protein C shown that placental abruption was more often
deficiency, 3.3 (95% CI 1.0–11.3) in protein S associated with homozygous and heterozygous
132
0415421306-Ch10 3/29/07 5:40 PM Page 133
FVL, heterozygous G20210A, and hyperhomocys- with FVL that were followed up prospectively, there
teinaemia. Women with preeclampsia/eclampsia was one case of HELLP syndrome (hemolysis, ele-
were more likely to have heterozygous FVL muta- vated liver enzymes, and low platelet count), but no
tion, heterozygous G20210A prothrombin gene other obstetric complications, and no deep vein
mutation, homozygous MTHFR C677T, protein C thrombosis or pulmonary embolus (unpublished
deficiency, protein S deficiency, or APCR. Stillbirth data). However, Sanson et al50 investigated women
was more often associated with FVL, protein S defi- with deficiencies of antithrombin, protein S, and
ciency, and activated protein C resistance. Women protein C. In the 60 deficient subjects, 22.3% of the
with IUGR had a higher prevalence of G20210A, 188 pregnancies resulted in miscarriage or stillbirth,
MTHFR C677T, or protein S deficiency. However, as compared with 11.4% of the 202 pregnancies in
Alfirevic et al43 concluded that ‘Women with the 69 non-deficient subjects. The relative risk of
adverse pregnancy outcome are more likely to have abortion and stillbirth per pregnancy for deficient
a positive thrombophilia screen but studies women as compared with non-deficient women
published so far are too small to adequately assess was 2.0 (95% CI 1.2–3.3). A longitudinal follow-up
the true size of this association.’ study is sorely needed to compare the incidence of
miscarriage, stillbirth, IUGR, and preeclampsia in
COHORT STUDIES women with hereditary thrombophilias.
Case–control studies can only show associations TREATMENT

between thrombophilias and pregnancy losses. In
order to infer cause and to come to conclusions This chapter only gives an outline of the treatment
about treatment, cohort studies are necessary. There options, as it is followed by a debate on the place of
are few cohort studies of patients with throm- treatment. Suffice it to say here that there are iso-
bophilias, in whom longitudinal studies have been lated reports that the presence of hereditary throm-
performed to assess the true incidence of pregnancy bophilias warrants thromboprophylaxis. However,
loss. Two studies have examined the subsequent live the role of treatment can only be determined in
birth rate in women with recurrent miscarriage and well-designed randomized trials where the effect
hereditary thrombophilias. In the report by of treatment is compared with untreated or
Ogasawara et al,15 the subsequent miscarriage rate placebo-treated patients. As yet, there are no
was not different for patients with decreased randomized placebo-controlled trials assessing
protein C or S activity, or antithrombin deficiency. prophylactic treatment with anticoagulants in
Carp et al47 found the live birth rate to be similar to hereditary thrombophilias.
that expected in recurrent miscarriage, whether In hereditary thrombophilias, a recent prospec-
the patient had FVL, G20210A, MTHFR C677T, tive study by Gris et al51 has compared enoxaparin
protein C, protein S, or antithrombin deficiencies. with aspirin in patients with thrombophilia and
Salomon et al48 have followed up 191 throm- one pregnancy loss. Enoxaparin was found to be
bophilic patients who attended an ultrasound clinic superior to low-dose aspirin. However, this study
to prospectively assess obstetric complications. The did not distinguish between early and late preg-
blood flow to the fetus was not compromised. No nancy losses, nor did it correct for early losses due
association was found between thrombophilias and to genetic or other factors known to affect the sub-
preeclampsia or IUGR. Lindqvist et al49 assessed sequent live birth rate. Carp et al52 have reported a
pregnancy complications in 270 patients with APCR. comparative cohort study comparing enoxaparin
This subgroup did not differ significantly from the with no treatment in women with hereditary
non-APCR patients in terms of pregnancy compli- thrombophilias and recurrent miscarriage. The pri-
cations, but was characterized by an eight-fold mary outcome measure was the incidence of subse-
higher risk of VTE. In our series of 21 pregnancies quent live births. Of the 37 pregnancies in treated
133
0415421306-Ch10 3/29/07 5:40 PM Page 134
patients 26 (70.2%) terminated in live births, com- apparent in patients with genetic predispositions to
pared with 21 of 48 (43.8%) in untreated patients thrombosis.
(OR 3.03; 95% CI 1.12–8.36). The beneficial effect
was mainly seen in primary aborters, i.e., women CYTOKINES
with no previous live births (OR 9.75; 95% CI 1.59–
52.48). This benefit was also found in patients with Cytokines are low-molecular-weight peptides or
a poor prognosis for a live birth (five or more mis- glycopeptides, produced by lymphocytes, mono-
carriages), where the live birth rate was increased cytes/macrophages, mast cells, eosinophils, and
from 18.2% to 61.6%. Although this trial was not blood vessel endothelial cells. Two cytokines have
randomized or blinded, it is the only trial in the lit- been associated with initiation of coagulation in
erature comparing the effect of treatment with a infections: TNF-α and interleukin-6 (IL-6) upregu-
cohort of untreated patients with hereditary throm- late the expression of tissue factor (TF), which ini-
bophilias. There has been no trial of anticoagulants tiates the extrinsic phase of the coagulation cascade
comparing the effects of treatment with untreated and subsequent thrombin generation.
patients regarding late obstetric complications. The Cytokine imbalances have been described in
optimal dose of anticoagulants has not yet been RPL,58 APS,59 preeclampsia,60 preterm births,61
determined. In a randomized prospective study, no and IUGR.62 The predominance of prothrombotic
difference was found between 40 mg and 80 mg of cytokines may well lead to placental thrombosis in
enoxaparin (Clexane, Sanofi Aventis Ltd, France) in genetically susceptible individuals.
women with thrombophilia and pregnancy losses.53
The mode of action of treatment also requires MICROPARTICLES
clarification. It has been assumed that throm-
bophilias act via thrombosis. However, anticoagu- Placental apoptosis has been described as a salient
lants also have anti-inflammatory effects. Heparin feature of pregnancy loss.63 Following apoptosis
increases serum tumor necrosis factor (TNF)-bind- and cell activation, the cell membrane is remodeled,
ing protein, protecting against systemic harmful with the release of microparticles. The microparti-
manifestations of TNF.54 Low-molecular-weight cles express procoagulant phospholipids, such as
heparins (LMWH) inhibit TNF-α production.55 phosphotidylserine, on their external surface. These
Thrombosis results in an inflammatory response in phospholipids are normally found inside the cell
the vein wall. Both heparin and LMWH limit the membrane. Microparticles lead to increased expres-
anti-inflammatory response,56 including neutrophil sion of adhesion molecules, thus amplifying the
extravasation and decreasing vein wall permeability. procoagulant and/or inflammatory response on the
Heparin also has direct effects on the trophoblast. endothelial cell surface. Microparticles have been
It has been reported to restore the invasive proper- found in increased numbers in normal pregnancy,
ties of the trophoblast in APS,57 and to enhance when there is constant deportation of trophoblast
placental human chorionic gonadotropin (hCG) pro- into the maternal circulation. Both Laude et al64 and
duction. The anti-inflammatory effects of heparins Carp et al65 have found increased levels of circulat-
and the direct placental effects may be as relevant, if ing microparticles in women with RPL. However,
not more so, than the anticoagulant effects. it has not been determined whether endothelial
microparticles may cause pregnancy loss through
subsequent thrombotic mechanisms, or may be a
OTHER PROTHROMBOTIC MECHANISMS OF consequence of embryonic death. Between 29% and
PREGNANCY LOSS 60% of recurrent first-trimester miscarriages are
due to chromosomal aberrations that are incompat-
There are other mechanisms that may induce ible with life and lead to miscarriage irrespective
thrombosis, or may allow thrombosis to become of other associations or causes of pregnancy loss,
134
0415421306-Ch10 3/29/07 5:40 PM Page 135
including the presence of microparticles. Even FETAL THROMBOPHILIA

in missed abortion due to chromosomal aberra-
tions, the trophoblast undergoes apoptosis with As placental histology usually shows a fetal vascu-
subsequent microparticle formation and thrombo- lopathy rather than maternal thrombosis, fetal
sis. However, in some patients, circulating endothe- thrombophilia may explain the pathological
lial microrparticles may themselves possibly induce changes. The hemostatic balance in the placenta
thrombosis and subsequent loss of a normal may be determined by both maternal and fetal fac-
pregnancy. tors cooperatively regulating coagulation at the
fetomaternal interface.74 Humans have an almost
HORMONES AND THROMBOSIS unique placentation in which trophoblast cells line
the maternal blood lakes rather than endothelial
The hormones of pregnancy – estrogen, proges- cells. Using genomewide expression analysis, Sood
terone, and hCG – all affect thrombosis. Estrogen et al33 identified a panel of genes that determine the
may alter the concentrations of clotting factors to a ability of fetal trophoblast cells to regulate hemosta-
prothrombotic profile, for example raising FVII66 sis at the fetomaternal interface. Additionally, the
and plasminogen activator inhibitor 1 (PAI-1)67 and trophoblast was shown to sense the presence of acti-
reducing antithrombin III.67 In mice, estrogen sul- vated coagulation factors via the expression of pro-
fotransferase (a cytosolic enzyme that catalyzes the tease-activated receptors. Engagement of these
sulfoconjugation of estrogens) plays a critical role receptors was reported to result in specific changes
in modulating estrogen activity in the mouse pla- in gene expression. Hence, fetal genes might modify
centa during midgestation.68 Inactivation of estro- the risks associated with maternal thrombophilia.
gen sulfotransferase caused local and systemic Additionally, coagulation activation at the fetoma-
estrogen excess and an increase in TF, leading to ternal interface might affect trophoblast physiology
placental thrombosis and fetal loss. Additionally, and alter placental function in the absence of frank
estrogen can either stimulate or inhibit the produc- thrombosis. We have seen fetal deaths in utero
tion of IL-1 and TNF.67 in which sonograms have shown complete occlu-
Progesterone, however, seems to have opposing sion of the umbilical blood vessels. However,
effects. It has prothrombotic effects, including upreg- it is impossible to say whether the thromboses
ulation of TF expression,70 but also induces the pro- caused fetal death or whether the changes occurred
duction of cytokines such as IL-4, which upregulates postmortem.
protein S, which inhibits coagulation.71 The
progestogen dydrogesterone inhibits production of REFERENCES
TNF-α (prothrombotic), but increases the levels of
IL-4 (antithrombotic) and IL-6 (prothrombotic).72 1. Lorand L, Losowsky MS, Miloszewski KJ. Human factor XIII fibrin sta-
In addition to its endocrine luteotrophic role, bilizing factor. Progr Thromb Haemost 1980; 5:245–90.
2. Muszbek L, Adany R, Mikkola H. Novel aspects of blood coagulation
hCG could also have a local role within the factor XIII. I. Structure, distribution, activation, and function. Crit Rev
uterine environment. Specific binding sites for Clin Lab Sci 1996; 33:357–421.
hCG have been shown in various cells of the 3. Schubring C, Grulich-Henn J, Burkhard PAT, et al. Fibrinolysis and
factor XIII in women with spontaneous abortion. Eur J Obstet Gynecol
endometrium and decidua. The local role of Reprod Biol 1990; 35:215–21.
hCG in the endometrium has not been fully 4. Burrows RF, Ray JG, Burrows EA. Bleeding risk and reproductive
elucidated. Uzumcu et al73 have assessed endome- capacity among patients with factor XIII deficiency: a case presentation
and review of the literature. Obstet Gynecol Surv 2000; 55:103–7.
trial production of cytokines when stimulated 5. Mosher DF, Schad PE, Kleinman HK. Cross-linking of fibronectin to
by hCG. Increasing doses of hCG caused a collagen by blood coagulation factor XIII. J Clin Invest 1979; 64:781–7.
dose-dependent increase in TNF-α and IL-6 6. Wartiovaara J, Leivo I, Virtanen I, et al. Cell surface and extracellu-
lar matrix glycoprotein fibronectin. Expression in embryogenesis
secretion, both of which have been reported to be and in teratocarcinoma differentiation. Ann NY Acad Sci 1978;
thrombogenic. 312:132–41.
135
0415421306-Ch10 3/29/07 5:40 PM Page 136
7. Koseki-Kuno S, Yamakawa M, Dickneite G, et al. Factor XIII A subunit 30. Mousa HA, Alfirevic Z. Do placental lesions reflect thrombophilia
deficient mice developed severe uterine bleeding events and subse- state in women with adverse pregnancy outcome? Hum Reprod 2000;
quent spontaneous miscarriages. Blood 2003; 102:4410–12. 15:1830–3.
8. Kobayashi T, Asahina T, Okada Y, et al. Studies on the localization of 31. Arias F, Romero R, Joist H, et al. Thrombophilia: a mechanism of dis-
adhesive proteins associated with the development of extravillous ease in women with adverse pregnancy outcome and thrombotic
cytotrophoblast. Trophoblast Res 1999; 13:35–53. lesions in the placenta. J Matern Fetal Med 1998; 7:277–86.
9. Asahina T, Kobayashi T, Okada Y, et al. Maternal blood coagulation 32. Lyden TW, Vogt E, Ng AK, et al. Monoclonal antiphospholipid anti-
factor XIII is associated with the development of cytotrophoblastic body reactivity against human placental trophoblast. J Reprod
shell. Placenta 2000; 21:388–93. Immunol 1992; 22:1–14.
10. Dardik R, Loscalzo J, Inbal A. Factor XIII (FXIII) and angiogenesis. 33. Sood R, Kalloway S, Mast AE, et al. Fetomaternal cross talk in the
J Thromb Haemost 2005; 4:19–25. placental vascular bed: control of coagulation by trophoblast cells.
11. Anwar T, Miloszewski K. Factor XIII deficiency. Br J Haematol 1999; Blood 2006; 107:3173–80.
107:468–84. 34. Kupferminc MJ. Thrombophilia and pregnancy. Reprod Biol
12. Reynolds TC, Butine MD, Visich JE, et al. Safety, pharmacokinetics, Endocrinol 2003; 1:111.
and immunogenicity of single-dose rFXIII administration J Thromb 35. Brenner B, Sarig G, Weiner Z, et al. Thrombophilic polymorphisms
Haemost 2005; 3:922–8. are common in women with fetal loss without apparent cause.
13. Hayano Y, Ima N, Kasaraura T. Studies on the physiologic changes of Thromb Haemost 1999; 82:6–9.
blood coagulation factor XIII during pregnancy and their signifi- 36. Rey E, Kahn SR, David M, et al. Thrombophilic disorders and fetal
cance. Acta Obstet Gynaecol Jpn 1982; 34:469–77. loss: a meta-analysis. Lancet 2003; 361:901–8.
14. Stirling Y, Woolf L, North WRS, et al. Haemostasis in normal preg- 37. Krabbendam I, Franx A, Bots ML, et al. Thrombophilias and recur-
nancy. Thromb Haemost 1984; 52:176. rent pregnancy loss: a critical appraisal of the literature. Eur J Obstet
15. Ogasawara MS, Aoki K, Katano K, et al. Factor XII but not protein C, Gynecol Reprod Biol 2005; 118:143–53.
protein S, antithrombin III, or factor XIII is a predictor of recurrent 38. Nelen WL, Blom HJ, Steegers EA, et al. Hyperhomocysteinemia and
miscarriage. Fertil Steril 2001; 75:916–19. recurrent early pregnancy loss: a meta-analysis. Fertil Steril 2000;
16. Anwar R, Gallivan L, Edmonds SD, et al. Genotype/phenotype corre- 74:1196–9.
lations for coagulation factor XIII: specific normal polymorphisms 39. Preston FE, Rosendaal FR, Walker ID, et al. Increased fetal loss in
are associated with high or low factor XIII specific activity. Blood women with heritable thrombophilia. Lancet 1996; 348:913–16.
1999; 93:897–905. 40. Sarig G, Younis JS, Hoffman R, et al. Thrombophilia is common in
17. Galanakis DK. Fibrinogen anomalies and disease. A clinical update. women with idiopathic pregnancy loss and is associated with late
Hematol Oncol Clin North Am 1992; 6:1171–87. pregnancy wastage. Fertil Steril 2002; 77:342–7.
18. Doolittle RF. The molecular biology of fibrin. In: Stamatoyannopoulos 41. Grandone E, Margaglione M, Colaizzo D, et al. Factor V Leiden is
GS, Nienhuis AW, Majerus PW, Harmus H, eds. The Molecular Basis associated with repeated and recurrent unexplained fetal losses.
of Blood Diseases. Philadelphia: WB Saunders, 1994:701–23. Thromb Haemost 1997; 77:822–4.
19. Brenner B. Inherited thrombophilia and fetal loss. Curr Opin 42. Kupferminc MJ, Eldor A, Steinman N, et al. Increased frequency of
Hematol 2000; 7:290–5. genetic thrombophilias in women with complications of pregnancy.
20. Mosesson MW. Dysfibrinogenemia and thrombosis. Semin Thromb N Engl J Med 1999; 340:9–13.
Hemost 1999; 25:311–19. 43. Alfirevic Z, Roberts D, Martlew V. How strong is the association
21. Ridgway HJ, Brennan SO, Faed JM, et al. Fibrinogen Otago: a major between maternal thrombophilia and adverse pregnancy outcome?
α chain truncation associated with severe hypofibrinogenaemia and A systematic review. Eur J Obstet Gynecol Reprod Biol 2002;
recurrent miscarriage. Br J Haematol 1997; 98:632–9. 101:6–14.
22. Suh TT, Holmback K, Jensen N, et al. Resolution of spontaneous 44. Gris JC, Quere I, Monpeyroux F, et al. Case–control study of the
bleeding events but failure of pregnancy in fibrinogen-deficient mice. frequency of thrombophilic disorders in couples with late fetal loss
Genes Dev 1995; 9:2020–33. and no thrombotic antecedent. The Nimes Obstetricians and
23. Inamoto Y, Terao T. First report of a case of congenital afibrinogene- Haematologists Study (NOHA). Thromb Haemost 1999; 81:891–9.
mia with successful delivery. Am J Obstet Gynecol 1985; 153:803–4. 45. Many A, Elad R, Yaron Y, et al. Third-trimester unexplained intrauter-
24. Gilabert J, Reganon E, Vila V, et al. Congenital hypofibrinogenemia ine fetal death is associated with inherited thrombophilia. Obstet
and pregnancy: obstetric and hematological management. Gynecol Gynecol 2002; 99:684–7.
Obstet Invest 1987; 24:271–6. 46. Infante-Rivard C, Rivard GE, Yotov WV, et al. Absence of association
25. MacKinnon HH, Fekete JF. Congenital afibrinogenemia: vascular of thrombophilia polymorphisms with intrauterine growth restric-
changes and multiple thromboses induced by fibrinogen infusions tion. N Engl J Med 2002; 347:19–25.
and contraceptive medication. CMAJ 1971; 140:597–9. 47. Carp HJA, Dolitzky M, Inbal A. Hereditary thrombophilias are not
26. Beck EA. Congenital abnormalities of fibrinogen. Clin Haematol associated with a decreased live birth rate in women with recurrent
1979; 8:169–81. miscarriage. Fertil Steril 2002; 78:58–62.
27. Seligsohn U, Lubetsky A. Genetic susceptibility to venous thrombosis. 48. Salomon O, Seligsohn U, Steinberg DM, et al. The common pro-
N Engl J Med 2001; 344:1222–31. thrombotic factors in nulliparous women do not compromise blood
28. Many A, Schreiber L, Rosner S, et al. Pathologic features of the pla- flow in the feto–maternal circulation and are not associated with
centa in women with severe pregnancy complications and throm- preeclampsia or intrauterine growth restriction. Am J Obstet Gynecol
bophilia. Obstet Gynecol 2001; 98:1041–4. 2004; 191:2002–9.
29. Rushton DI. Simplified classification of spontaneous abortions. 49. Lindqvist PG, Svensson PJ, Marsá K, et al. Activated protein C resist-
J Med Genet 1978; 15:1–9. ance (FV:Q506) and pregnancy. Thromb Haemost 1999; 81:532–7.
136
0415421306-Ch10 3/29/07 5:40 PM Page 137
50. Sanson BJ, Friederich PW, Simioni P, et al. The risk of abortion and 62. Hahn-Zoric M, Hagberg H, Kjellmer I, et al. Aberrations in placental
stillbirth in antithrombin-, protein C-, and protein S-deficient cytokine mRNA related to intrauterine growth retardation. Pediatr
women. Thromb Haemost 1996; 75:387–8. Res 2002; 51:201–6.
51. Gris JC, Mercier E, Quere I, et al. Low-molecular-weight heparin 63. Brill A, Torchinsky A., Carp HJA, et al. The role of apoptosis in normal
versus low-dose aspirin in women with one fetal loss and a constitu- and abnormal embryonic development. J Assist Reprod Genet 1999;
tional thrombophilic disorder. Blood 2004; 103:3695–9. 16:512–19.
52. Carp HJA, Dolitzky M, Inbal A. Thromboprophylaxis improves the 64. Laude I, Rongieres-Bertrand C, Boyer-Neumann C, et al. Circulating
live birth rate in women with consecutive recurrent miscarriages and procoagulant microparticles in women with unexplained pregnancy
hereditary thrombophilia. J Thromb Hemost 2003; 1:433–8. loss: a new insight. Thromb Haemost 2001; 85:18–21.
53. Brenner B, Hoffman R, Carp HJA, et al. The LIVE–ENOX Investigators. 65. Carp HJA, Dardik R, Lubetsky A, et al. Prevalence of circulating pro-
Efficacy and safety of two doses of enoxaparin in women with coagulant microparticles in women with recurrent miscarriage: a case
thrombophilia and recurrent pregnancy loss: the LIVE-ENOX study. controlled study. Hum Reprod 2004; 19:191–5.
J Thromb Haemost 2005; 3:227–9. 66. Meilahn EN, Kuller LH, Matthews KA, et al. Hemostatic factors
54. Lantz M, Thysell H, Nilsson E, et al. On the binding of tumor necro- according to menopausal status and use of hormone replacement
sis factor (TNF) to heparin and the release in vivo of the TNF-bind- therapy. Ann Epidemiol 1992; 2:445–55.
ing protein I by heparin. J Clin Invest 1991; 88:2026–31. 67. Cosman F, Baz-Hecht M, Cushman M, et al. Short-term effects
55. Baram D, Rashkovsky M, Hershkoviz R, et al. Inhibitory effects of low of estrogen, tamoxifen and raloxifene on hemostasis: a randomized-
molecular weight heparin on mediator release by mast cells: preferen- controlled study and review of the literature. Thromb Res 2005;
tial inhibition of cytokine production and mast cell-dependent 116:1–13.
cutaneous inflammation. Clin Exp Immunol 1997; 110:485–91. 68. Tong MH, Jiang H, Liu P, et al. Spontaneous fetal loss caused by
56. Downing LJ, Strieter RM, Kadell AM, et al. Low-dose low-molecular- placental thrombosis in estrogen sulfotransferase-deficient mice.
weight heparin is anti-inflammatory during venous thrombosis. Nat Med 2005; 11:153–9.
J Vasc Surg 1998; 28:848–54. 79. Polan ML, Daniele A, Kuo A. Gonadal steroids modulate human
57. Bose P, Black S, Kadyrov M, et al. Adverse effects of lupus anticoagu- monocyte interleukin-1 (IL-1) activity. Fertil Steril 1988; 49:964–8.
lant positive blood sera on placental viability can be prevented by 70. Schatz F, Krikun G, Caze R, et al. Progestin-regulated expression of
heparin in vitro. Am J Obstet Gynecol 2004; 191:2125–31. tissue factor in decidual cells: implications in endometrial hemostasis,
58. Carp HJA, Torchinsky A, Fein A, et al. Hormones, cytokines and menstruation and angiogenesis. Steroids 2003; 68:849–60.
fetal anomalies in habitual abortion. J Gynecol Endocrinol 2002; 71. Smiley ST, Boyer SN, Heeb MJ, et al. Protein S is inducible by inter-
15:472–83. leukin 4 in T cells and inhibits lymphoid cell procoagulant activity.
59. Krause I, Blank M, Levi Y, et al. Anti-idiotype mmunomodulation of Proc Natl Acad Sci USA 1997; 94:11484–9.
experimental antiphospholipid syndrome via effect on Th1/Th2 72. Raghupathy R, Al Mutawa E, Makhseed M, et al. Modulation of
expression. Clin Exp Immunol 1999; 117:190–7. cytokine prouction by dydrogesterone in lymphocytes from women
60. Darmochwal-Kolarz D, Rolinski J, Leszczynska-Goarzelak B, et al. The with recurrent miscarriage. Br J Obset Gynaecol 2005; 112:1096–101.
expressions of intracellular cytokines in the lymphocytes of 73. Uzumcu M, Coskun S, Jaroudi K, et al. Effect of human chorionic
preeclamptic patients. Am J Reprod Immunol 2002; 48:381–6. gonadotropin on cytokine production from human endometrial cells
61. Maymon E, Ghezzi F, Edwin SS, et al. The tumor necrosis factor alpha in vitro. Am J Reprod Immunol 1998; 40:83–8.
and its soluble receptor profile in term and preterm parturition. 74. Rosing J. Mechanisms of OC related thrombosis. Thromb Res 2005;
Am J Obstet Gynecol 1999; 181:1142–8. 115(Suppl 1):81–3.
137
0415421306-Ch10 3/29/07 5:40 PM Page 138
0415421306-Ch10a 3/29/07 5:41 PM Page 139
DEBATE
10a. Should thromboprophylaxis be used

in hereditary thrombophilias with RPL?
– For
Benjamin Brenner
(LMWH) in these settings.9,10 While interpretation of

THROMBOPHILIA AND FETAL LOSS the results of these studies has given rise to intensive
debate,11 it is clear that the field of thrombophilia
Thrombophilic risk factors are common and can and pregnancy complications continues to be at the
be found in 15–25% of Caucasian populations. focus of medical research and clinical practice.
Since pregnancy is an acquired hypercoagulable
state, women harboring thrombophilia may present
with clinical symptoms of vascular complications LMWH THERAPY IN PREGNANCY
for the first time during gestation.1 Recurrent
pregnancy loss affects 1–5% of women of reproduc- Until recently, studies on the treatment of women
tive age and bears a significant emotional, social, with inherited thrombophilia and pregnancy
and economical impact. A number of case–control loss were predominantly uncontrolled, and
and cohort studies have suggested an association included small series of patients treated mostly with
between inherited thrombophilia and recurrent LMWH. A collaborative study demonstrated the
pregnancy loss (RPL).2–4 Several recently reported safety of using LMWH during 486 gestations.12
meta-analyses support an association between A successful outcome was reported in 83 (89%) of
pregnancy loss and maternal factor V Leiden (FVL) 93 gestations in women with a history of RPL and
and factor II (FII) G20210A genotypes.5–7 Recently, in all 28 gestations in women who had experienced
Lissalde-Lavigne et al8 reported findings from the preeclampsia during a previous pregnancy. A retro-
‘NOHA first’ study, a large carefully designed case– spective French study on the use of the LMWH
control study nested in a cohort of nearly 32 700 enoxaparin during 624 pregnancies revealed a good
women of whom 18% had pregnancy loss in the first safety profile.13 More recently, a review of close to
gestation. The findings of the multivariate analysis 2800 treated pregnancies evaluated the safety and
clearly demonstrate an overall association between efficacy of LMWH in pregnancy.14 The main indi-
unexplained first pregnancy loss after 10 weeks of cations were prophylaxis of venous thromboem-
gestation and the two thrombophilic risks factors bolism (VTE) and prevention of pregnancy loss.
(odds ratio (OR) 3.46 and 2.60, respectively). The rate of bleeding complications was low (<2%)
Documentation of thrombophilic risk factors in and thrombocytopenia was rare, with no cases of
women with pregnancy complications may have sig- heparin-induced thrombocytopenia. Likewise, clin-
nificant therapeutic implications, since recent clinical ically significant osteoporosis was extremely rare.
studies have demonstrated the potential efficacy of The live birth rate was between 85% and 96%,
prophylaxis with low-molecular-weight heparin depending on the indication.
139
0415421306-Ch10a 3/29/07 5:41 PM Page 140
LMWH IN WOMEN WITH THROMBOPHILIA AND study recently conducted in Israel on women with
PREGNANCY LOSS thrombophilia and pregnancy loss, defined as three
or more first-trimester, two or more second-
Our group has treated 61 pregnancies in 50 women trimester, or at least one third-trimester loss; 180
with thrombophilia who presented with recurrent women were enrolled in the study. The study com-
fetal loss with enoxaparin throughout gestation and pared two doses of enoxaparin: 40 mg/day and
4–6 weeks into the postpartum period.15 The dose 40 mg twice daily throughout pregnancy, starting
was 40 mg/day, except for patients with combined at 5–10 weeks of gestation and for 6 weeks postpar-
thrombophilia or in the case of abnormal Doppler tum. The primary endpoint was the delivery of a
velocimetry suggesting decreased placental perfu- healthy infant. Secondary endpoints were duration
sion, where the dose was increased to 40 mg twice of gestation, birthweight, and incidence of gesta-
a day. Of the 61 pregnancies, 46 (75%) resulted in tional thrombosis and gestational vascular compli-
live birth, compared with a success rate of only 20% cations. The incidence of preeclampsia in the
in these 50 women in prior gestations without enoxaparin 40 and 80 mg/day groups was 6.7%
antithrombotic therapy. Carp et al16 have reported a and 14.3%, respectively, and the incidence of pla-
cohort study undertaken to assess the effect of cental abruption was 13.5% and 8.8%, respectively.
enoxaparin on the subsequent live birth rate in Approximately a quarter of the women in both
women with three or more consecutive pregnancy groups had had intrauterine growth restriction
losses and hereditary thrombophilia. The live birth (IUGR) in previous gestations (22.5% and 24.2%,
rate was higher in women treated with enoxaparin: respectively). The live birth rate prior to enroll-
26 (70.2%) of 37, compared with 21 (43.8%) of 48 ment in the study was only 28%. During the study,
in untreated patients. The beneficial effect was the live birth rate was 84% for the 40 mg/day
mainly in primary aborters and in those with five or group and 78% for the 80 mg/day group.9 Late ges-
more miscarriages. Gris et al17 have recently carried tational complications decreased after enoxaparin
out a randomized study to compare the effect of treatment. The incidence of preeclampsia in the
enoxaparin given throughout gestation at a dose of enoxaparin 40 and 80 mg/day groups was 3.4%
40 mg daily, compared with low-dose aspirin. The and 4.4%, respectively. Similarly, the incidence of
study comprised 160 women with thrombophilia placental abruption in the 40 and 80 mg/day
and one previous pregnancy loss after 10 weeks of groups was 4.5% and 3.3%, respectively.10 Both
gestation. The patients treated with enoxaparin had doses of enoxaparin appeared to be safe and
a significantly higher live birth rate than the well tolerated. The gestation period was longer
patients treated with low-dose aspirin (86% vs 29%, than 36 weeks in over 80% of patients in each
respectively). These dramatic differences were group. However, preterm delivery occurred in 10%
found in women with FVL and FII G20210A muta- and 18.5% of women in the enoxaparin 40 and
tions and in women with protein S deficiency. 80 mg/day groups, respectively. Postpartum bleed-
Moreover, thrombophilic women with the co- ing (1.1% of women in each group) and enoxa-
presence of protein Z deficiency or antibodies to parin-related allergic local skin reactions at the
protein Z had a reduced live birth rate following injection sites were observed in a small number of
enoxaparin prophylaxis.17 women (2.2% and 3.3% of those receiving 40 and
The optimal dosage of LMWH is yet unknown 80 mg/day, respectively. Thus, prophylaxis with
and should be determined by prospective random- enoxaparin (40 or 80 mg/day) is safe and effective
ized trials. Ideally, large placebo-controlled trials for improving pregnancy outcome and reducing
should be advocated. However, logistic and ethical late pregnancy complications in thrombophilic
difficulties limit this approach.11 The LIVE–ENOX women with a history of pregnancy loss.
study10 was a multicenter prospective randomized Prophylaxis with LMWH is indicated in women
140
0415421306-Ch10a 3/29/07 5:41 PM Page 141
DEBATE: SHOULD THROMBOPROPHYLAXIS BE USED IN RPL? – FOR
with thrombophilia and late or recurrent early In about 30–50% of women with pregnancy loss,
pregnancy loss. Women with severe thrombophilia no specific thrombophilic risk factor can be docu-
(i.e., antithrombin deficiency) or with combined mented. It is possible that in some of these cases,
thrombophilic risk factors may require higher doses subtle changes in a number of coagulation proteins
of LMWH. can lead to abnormal function of the protein C
system that can be detected by global assays.23
Presentation of membrane hemostatic proteins
MONITORING OF LMWH IN PREGNANCY such as TF, thrombomodulin (TM), and endothelial
protein C receptor (EPCR) on trophoblasts may
The need for monitoring of LMWH therapy in determine a local procoagulant profile of the pla-
pregnancy is debatable. However, consensus confer- centa.24 Abnormalities in these genes may serve as
ences such as the ACCP18 suggest that changes in risk modifiers in thrombophilia-associated gesta-
pharmacokinetic and pharmacodynamic properties tional vascular complications. Indeed, LMWH ther-
during pregnancy may require monitoring. Indeed, apy improves gestational outcome in obligatory
a recent study suggests that monitoring of anti-Xa carriers of EPCR knockout mice.25
and free tissue factor pathway inhibitor (TFPI)
levels is of value during pregnancy in throm- REFERENCES
bophilic women treated with enoxaparin.19 It is
debatable whether antithrombotic prophylaxis 1. Brenner B. Clinical management of thrombophilia-related placental
vascular complications. Blood 2004; 103:4003–9.
should be used for unexplained recurrent first- or
2. Grandone E, Margaglione M, Colaizzo D, et al. Factor V Leiden is
second-trimester pregnancy loss in women who do associated with repeated and recurrent unexplained fetal losses.
not harbor thrombophilia. A recent study by Thromb Haemost 1997; 77:822–4.
3. Ridker PM, Miletich JP, Buring JE, et al. Factor V Leiden mutation as
Dolitzky et al20 found that enoxaparin 40 mg/day
a risk factor for recurrent pregnancy loss. Ann Intern Med 1998;
and low-dose aspirin were equally effective in this 15:1000–3.
setting. However, selection bias cannot be excluded 4. Brenner B, Sarig G, Weiner Z, et al. Thrombophilic polymorphisms
are common in women with fetal loss without apparent cause.
in the study by Dolitzky et al,20 since women were
Thromb Haemost 1999; 82:6–9.
enrolled after demonstration of fetal heart beat. 5. Rey E, Kahn SR, David M, et al. Thrombophilic disorders and fetal
loss: a meta-analysis. Lancet 2003; 361:901–8.
6. Kovalevsky G, Gracia CR, Berlin JA, et al. Evaluation of the association
between hereditary thrombophilias and recurrent pregnancy loss: a
MECHANISM OF ACTION OF LMWH IN meta-analysis. Arch Intern Med 2004; 164:558–63.
PREGNANCY 7. Dudding TE, Attia J. The association between adverse pregnancy out-
comes and maternal factor V Leiden genotype: a meta-analysis.
Thromb Haemost 2004; 91:700–11.
The mechanisms of action of LMWH in women 8. Lissalde-Lavigne G, Fabbro-Peray P, Cochery-Nouvellon E, et al.
with placental vascular complications have not been Factor V Leiden and prothrombin G20210A polymorphisms as risk
factors for miscarriage during a first intended pregnancy: the matched
fully elucidated. While the systemic anticoagulant case–control ‘NOHA first’ study. J Thromb Haemost 2005; 3:2178–84.
effect in the maternal circulation is important, other 9. Brenner B, Bar J, Ellis M, et al. Effects of enoxaparin on late pregnancy
mechanisms have been suggested. These include complications and neonatal outcome in women with recurrent preg-
nancy loss and thrombophilia: results from the LIVE–ENOX study.
anti-inflammatory effects and modulation of local Fertil Steril 2005; 84:770–3.
hemostasis at the placental level. In particular, pla- 10. Brenner B, Hoffman R, Carp H, et al. Efficacy and safety of two doses
cental trophoblasts are characterized by a hemosta- of enoxaparin in women with thrombophilia and recurrent preg-
nancy loss: the LIVE–ENOX study. J Thromb Haemost 2005; 3:227–9.
tic balance between TF and TFPI.21 This balance is 11. Walker ID, Kujovich L, Greer IA, et al. The use of LMWH in pregnan-
hampered in placental vascular complications, as cies at risk: new evidence or perception? J Thromb Haemost 2005;
TFPI levels are reduced. Maternal treatment with 3:778–93.
12. Sanson BJ, Lensing AW, Prins MH, et al. Safety of low molecular
LMWH restores TFPI levels and improves the weight heparin in pregnancy: a systemic review. Thromb Haemost
TF/TFPI balance in the human placenta.22 1999; 81:668–72.
141
0415421306-Ch10a 3/29/07 5:41 PM Page 142
13. Lepercq J, Conard J, Borel-Derlon A, et al. Venous thromboembolism with thrombophilia and pregnancy loss. Thromb Haemost 2005;
during pregnancy: a retrospective study of enoxaparin safety in 94:980–5.
624 pregnancies. Br J Obstet Gynaecol 2001; 108:1134–40. 20. Dolitzky M, Inbal A, Weiss A, et al. Randomized study of thrombo-
14. Greer IA, Nelson-Piercy C. Low-molecular-weight heparins for throm- prophylaxis in women with unexplained consecutive recurrent
boprophylaxis and treatment of venous thromboembolism in preg- miscarriages. Fertil Steril 2006: 86:362–6.
nancy: a systematic review of safety and efficacy. Blood 2005; 106:401–7. 21. Aharon A, Brenner B, Katz T, et al. Tissue factor and tissue factor path-
15. Brenner B, Hoffman R, Blumenfeld Z, et al. Gestational outcome in way inhibitor levels in trophoblast cells: implications for placental
thrombophilic women with recurrent pregnancy loss treated by hemostasis. Thromb Haemost 2004; 92:776–86.
enoxaparin. Thromb Haemost 2000; 83:693–7. 22. Aharon A, Lanir N, Drugan A, et al. Placental TFPI is decreased in
16. Carp H, Dolitzky M, Inbal A. Thromboprophylaxis improves the live gestational vascular complications and can be restored by maternal
birth rate in women with consecutive recurrent miscarriages and enoxaparin treatment. J Thromb Haemost 2005; 3:2355–7.
hereditary thrombophilia. J Thromb Haemost 2003; 1:433–8. 23. Sarig G, Lanir N, Hoffman R, et al. Protein C global assay in the eval-
17. Gris JC, Mercier E, Quere I, et al. Low-molecular-weight heparin versus uation of women with idiopathic pregnancy loss. Thromb Haemost
low-dose aspirin in women with one fetal loss and a constitutional 2002; 88:32–6.
thrombophilic disorder. Blood 2004; 103:3695–9. 24. Sood R, Kalloway S, Mast AE, et al. Fetomaternal cross talk in the
18. Bates SM, Greer IA, Hirsh J, et al. Use of antithrombotic agents during placental vascular bed: control of coagulation by trophoblast cells.
pregnancy: the Seventh ACCP Conference on Antithrombotic and Blood 2006; 107:3173–80.
Thrombolytic Therapy. Chest 2004; 126(3 Suppl):627S-44S. 25. Gu JM, Crawley JT, Ferrell G, et al. Disruption of the endothelial cell
19. Sarig G, Blumenfeld Z, Leiba R, et al. Modulation of systemic protein C receptor gene in mice causes placental thrombosis and early
hemostatic parameters by enoxaparin during gestation in women embryonic lethality. J Biol Chem 2002; 277:43335–43.
142
0415421306-Ch10b 3/29/07 5:42 PM Page 143
DEBATE
10b. Should thromboprophylaxis be used

in hereditary thrombophilias with RPL?
– Against
Pelle G Lindqvist
will focus on these factors. First-trimester fetal loss

INTRODUCTION
has not been related to hereditary thrombophilia –
either in large cohorts4,5 or in the largest case–control
Habitual abortion is usually associated with up to a
studies.6,7 Studies of early RPL show no increased
75% live birth rate after three consecutive fetal
risk in the largest case–control studies.6,8 Although
losses.1,2 Recurrent pregnancy loss (RPL) is often
several meta-analyses and systematic reviews have
taken to have a broader definition of at least three
shown an increased prevalence of thrombophilias in
first-trimester fetal losses and/or two or more
RPL. Meta-analyses may not be a sound method of
second-trimester fetal losses. Factors that are major
amassing evidence from these studies, as the majority
determinants of the risk of miscarriage are maternal
of studies that have been surveyed are uncontrolled,
age, the number of prior fetal losses, and whether or
underpowered, and subject to several types of bias.
not cardiac activity had been detected.2 The patho-
In addition, the conclusions are mostly drawn from
genesis of recurrent fetal loss in the majority of cases
case–control studies, which tend to overestimate risk
is still unclear. Several causes have been suggested:
assessments when compared with cohort studies.
chromosome aberrations, thrombophilias, thyroid
With regard to second-trimester (or >10 gestational
abnormality, microparticles, and complement
weeks) or third-trimester fetal loss, there seems to be
activation.3
an increased risk.7
RELATIONSHIP BETWEEN FIRST-TRIMESTER SINGLE-GENE MUTATION AND HUMAN

FETAL LOSS AND THROMBOPHILIA EVOLUTION
Treatment of RPL with low-molecular-weight Both FVL and FII are single-gene mutations that
heparin (LMWH) is based on a supposed causative appear to have first occurred some 25 000 years
relationship between thrombophilias and recurrent ago.9,10 A strong relation between these throm-
fetal loss. Such a link has not been established and is bophilias and fetal loss would have had a strong
still disputed. Since the majority of recurrent fetal negative impact on the respective gene pool.
losses occur in the first trimester and more than Instead, the number of carriers of FVL has gone from
90% of women with heritable thrombophilias in one person 25 000 years ago to about 50 million car-
Caucasian populations are carriers of either coagu- riers of Caucasian descent today.11 Thus, from an
lation factor V Leiden (FVL) or the prothrombin evolutionary perspective, it is unlikely that an
gene mutation G20210A (FII), the analysis below increased risk of fetal loss in the general population
143
0415421306-Ch10b 3/29/07 5:42 PM Page 144
has been caused by these thrombophilias. However, causes, independent of medical intervention (see
this does not exclude the possibility of a relation- Chapter 18). Therefore, no conclusions may be
ship in small subgroups, i.e., second-trimester fetal drawn from the results.13,14 The study by Gris et al,17
losses or habitual abortions. The more prevalent a comparing enoxaparin treatment with low-dose
single-mutation thrombophilia is, the less likely is aspirin in nulliparous carriers of FVL with a single
an increased risk of fetal loss. Thus, a causative link pregnancy loss, has also been criticized for not
between fetal loss and rare mutations such as including untreated controls.18 The authors write
antithrombin, protein C, and protein S deficiencies that in the course of their investigation, the creation
is more likely, as compared with FVL and FII.12 of a control group ‘was tried out, but failed’.
Recently, we have compiled data on the magnitude
of ‘regression toward the mean’ (Table 10b.1).18 In
TREATMENT STUDIES fact, the untreated women with recurrent fetal loss
in our prospective cohort have a similar outcome to
There have been some treatment studies published those treated with enoxaparin.13,14,18 Moreover, our
using a ‘before and after’ design.13,14 This design untreated FVL carriers with a prior fetal loss had
lacks a control group, and the result is conditioned similar outcomes to FVL carriers with a single prior
by the phenomenon of ‘regression toward the fetal loss who were treated with enoxaparin17,18
mean’.15 ‘Regression toward the mean’ is a statistical (Table 10b.1). A discussion of the optimal dosage of
principle stating that, of related measurements, the LMWH is not relevant before a relation has been
expected value of the second is closer to the mean established.15
than the observed value of the first.16 Therefore, in Many of us remember when immunotherapy
a ‘before and after’ design, one should expect the was considered the optimal treatment for RPL, but
group of women with the highest rate of recurrent immunotherapy was widely introduced before con-
fetal loss to have a lower rate of RPL by natural clusive studies had been performed.19 There is only
Table 10b.1 Pregnancy outcome in different subgroups of women with prior fetal loss
Prior live birth rate (%) Present live birth rate (%)
Recurrent fetal lossa

Enoxaparin 40 or 80 mg/day (n = 50)13,b 20 75
Enoxaparin 40 mg/day (n = 89)14,b 28 84
Enoxaparin 80 mg/day (n = 91)14,b 28 78
No treatment (n = 37)18,c 28 89
Second-trimester fetal loss
1 prior (n = 43), no treatment18 49 98
≥ 2 prior (n = 10) prior, no treatment18 30 80
Nulliparous women with one prior fetal loss and carriers of
factor V leiden (FVL)
Low-dose aspirin (n = 36)17 0 29
Enoxaparin 40 mg/day (n = 36)17 0 94
No treatment (n = 20)18,d 0 95
No treatment (n = 52)18,e 40 98
a
Recurrent fetal loss: ≥ 3 first-trimester and/or ≥ 2 second-trimester fetal loss.
b
Includes or/and ≥ 1 stillbirth.
c
Includes women with and without thrombophilia.
d
Nulliparous carriers of FVL with at least 1 prior fetal loss.
e
FVL carriers with at least 1 prior fetal loss.
144
0415421306-Ch10b 3/29/07 5:42 PM Page 145
DEBATE: SHOULD THROMBOPROPHYLAXIS BE USED IN RPL? – AGAINST
one study designed to answer the question of 7. Lissalde-Lavigne G, Fabbro-Peray P, Cochery-Nouvellon E, et al.
Factor V Leiden and prothrombin G20210A polymorphisms as risk
whether it is of value to treat thrombophilic women factors for miscarriage during a first intended pregnancy: the matched
having habitual abortion20 with LMWH and no case–control ‘NOHA first’ study. J Thromb Haemost 2005; 3:2178–84.
study of women with recurrent second-trimester 8. Carp H, Salomon O, Seidman D, et al. Prevalence of genetic markers
for thrombophilia in recurrent pregnancy loss. Hum Reprod 2002;
losses. In 2003, Carp et al20 raised the possibility that 17:1633–7.
LMWH treatment might be beneficial for women 9. Zivelin A, Griffin JH, Xu X, et al. A single genetic origin for a common
with a history of habitual abortion. Most encourag- Caucasian risk factor for venous thrombosis. Blood 1997; 89:397–402.
10. Zivelin A, Rosenberg N, Faier S, et al. A single genetic origin for the
ing were the positive results in the small subgroup of common prothrombotic G20210A polymorphism in the prothrombin
primary aborters. However, the investigation needs gene. Blood 1998; 92:1119–24.
to be reproduced in a randomized manner. 11. Lindqvist PG, Zöller B, Dahlbäck B. Improved hemoglobin status and
reduced menstrual blood loss among female carriers of activated
Due to the lack of studies designed to address protein C resistance (FV Leiden). An evolutionary advantage?
the issue at hand, we have as yet no evidence to Thromb Haemost 2001; 86:1122–3.
recommend LMWH treatment for RPL due to 12. Preston FE, Rosendaal FR, Walker ID, et al. Increased fetal loss in
women with heritable thrombophilia. Lancet 1996; 348:913–16.
thrombophilia. 13. Brenner B, Hoffman R, Blumenfeld Z, et al. Gestational outcome in
thrombophilic women with recurrent pregnancy loss treated by
REFERENCES enoxaparin. Thromb Haemost 2000; 83:693–7.
14. Brenner B, Hoffman R, Carp H, et al. Efficacy and safety of two doses
of enoxaparin in women with thrombophilia and recurrent preg-
1. Warburton D, Fraser FC. On the probability that a woman who has had
nancy loss: the LIVE–ENOX study. J Thromb Haemost 2005; 3:227–9.
a spontaneous abortion will abort in subsequent pregnancies. J Obstet
15. Lindqvist PG, Merlo J. Low molecular weight heparin for repeated
Gynaecol Br Emp 1961; 68:784–8.
pregnancy loss: Is it based on solid evidence? J Thromb Haemost
2. Brigham SA, Conlon C, Farquharson RG. A longitudinal study of
2005; 3:221–3.
pregnancy outcome following idiopathic recurrent miscarriage.
16. Regression toward the mean. Wikipedia, the free encyclopedia, 2006.
Hum Reprod 1999; 14:2868–71.
http://en.wikipedia.org
3. Carp H, Dardik R, Lubetsky A, et al. Prevalence of circulating procoag-
17. Gris JC, Mercier E, Quere I, et al. Low-molecular-weight heparin
ulant microparticles in women with recurrent miscarriage: a
versus low-dose aspirin in women with one fetal loss and a constitu-
case-controlled study. Hum Reprod 2004; 19:191–5.
tional thrombophilic disorder. Blood 2004; 103:3695–9.
4. Lindqvist PG, Svensson P, Dahlback B. Activated protein C resistance –
18. Lindqvist PG, Merlo J. The natural course of women with recurrent
in the absence of factor V Leiden – and pregnancy. J Thromb Haemost
fetal loss. J Thromb Haemost 2006; 4:896–7.
2006; 4:361–6.
19. Scott JR. Immunotherapy for recurrent miscarriage. Cochrane
5. Roque H, Paidas MJ, Funai EF, et al. Maternal thrombophilias are not
Database Syst Rev 2003; (1):CD000112.
associated with early pregnancy loss. Thromb Haemost 2004; 91:290–5.
20. Carp H, Dolitzky M, Inbal A. Thromboprophylaxis improves the live
6. Rai R, Shlebak A, Cohen H, et al. Factor V Leiden and acquired activated
birth rate in women with consecutive recurrent miscarriages and
protein C resistance among 1000 women with recurrent miscarriage.
hereditary thrombophilia. J Thromb Haemost 2003; 1:433–8.
Hum Reprod 2001; 16:961–5.
145
0415421306-Ch10b 3/29/07 5:42 PM Page 146
0415421306-Ch11 3/30/07 9:40 AM Page 147
11. Uterine anomalies and recurrent

pregnancy loss
Daniel S Seidman and Mordechai Goldenberg
INTRODUCTION DEVELOPMENT AND CLASSIFICATION

OF MÜLLERIAN DUCT DEFECTS
Anatomical uterine defects have long been associated
with recurrent miscarriage. However, it is frustrating Uterine anatomical defects are commonly classified
to realize how little is actually known regarding the as congenital or acquired. The classification of
pathophysiology responsible for the proposed causal congenital uterine defects is largely based on the
association between uterine anomalies and fetal understanding of müllerian duct development.
wastage. The lack of a clear understanding of the The two paired müllerian ducts of the embryo
causative mechanisms may be due to the fact that the ultimately develop into the female reproductive tract.
prevalence and impact of uterine malformations has The cephalic ends of the müllerian ducts form the
so far not been conclusively determined.1 Even the fallopian tubes, and the caudal portions fuse to
true incidence of congenital uterine anomalies in form the uterus, cervix, and the upper two-thirds of
the general population is unknown. A review of the the vagina. The ovaries and lower one-third of the
available literature reveals a wide range of reported vagina have separate embryological origins. The
incidences, from 0.2% to 10.0%.2 müllerian ducts grow caudally and become enclosed
Using newer imaging modalities, it is currently in peritoneal folds that later develop into the round
estimated that the incidence of uterine anomalies in and ovarian ligaments. In the female embryo, sexual
the general population is approximately 1%, and is differentiation is marked by degeneration of the
about threefold higher in women with recurrent wolffian ducts in the absence of fetal testes
pregnancy loss (RPL) and poor reproductive out- and testosterone. Absence of müllerian-inhibiting
comes.2 Below, we will discuss in detail the new substance allows the müllerian ducts to fully
modes of imaging that have been introduced mature. At 9 weeks of gestation, the uterine cervix is
over the last two decades and which may modify the recognizable, and by 17 weeks the formation of
previously reported data on the incidence. the myometrium is complete. Vaginal development
In addition to pregnancy loss, uterine malforma- begins at approximately 9 weeks. The uterovaginal
tions predispose women to other reproductive plate forms between the caudal buds of the müllerian
difficulties, including infertility, preterm labor, and ducts and the dorsal wall of the urogenital sinus.
abnormal fetal presentation. These poor reproduc- These cells will degenerate, thereby increasing the
tive outcomes are often attributed to the presence distance between the uterus and urogenital sinus.
of a uterine septum, intrauterine adhesions, polyps, Hence, the upper two-thirds of the vagina derives
and fibroids, all of which are amenable to surgical from the müllerian ducts while the remainder
correction. Therefore, an accurate diagnosis is essen- derives from the urogenital sinus. Complete forma-
tial in order to offer appropriate treatment. tion and differentiation of the müllerian ducts into
In this chapter, we will review the common the segments of the female reproductive tract depend
congenital and acquired uterine anomalies associated on completion of the following three phases of devel-
with RPL, and discuss contemporary diagnosis and opment: organogenesis, fusion both laterally and
treatment options. vertically, and resorption.
147
0415421306-Ch11 3/30/07 9:40 AM Page 148
In failure of organogenesis, one or both müllerian The association between RPL and subseptate
ducts may not develop fully, resulting in abnormal- uterus has been attributed to the decreased amount
ities such as uterine agenesis or hypoplasia (bilat- of connective tissue in the relatively avascular
eral) or unicornuate uterus (unilateral). In lateral septum, resulting in poor decidualization and pla-
fusion defects, the process by which the lower seg- centation. In addition, the increased amount of
ments of the paired müllerian ducts fuse to form muscle tissue in the septum can cause miscarriage
the uterus, cervix, and upper vagina fails. Failure of by the production of local uncoordinated myome-
fusion results in anomalies such as bicornuate or trial contractility. The view that inadequate blood
didelphys uterus. Vertical fusion refers to fusion of supply to the developing embryo accounts for
the ascending sinovaginal bulb with the descending the fetal losses is supported by histological evalua-
müllerian system (i.e., fusion of the lower one-third tion of the septum showing a significantly reduced
and upper two-thirds of the vagina). Complete ver- vascular supply relative to the rest of the uterus.10,11
tical fusion forms a normal patent vagina, while If this theory is correct, then the likelihood of mis-
incomplete vertical fusion results in an imperforate carriage caused by septal implantation should
hymen. increase with the severity of the disruption of uterine
After the lower müllerian ducts fuse, a central morphology.6
septum is present, which subsequently must be Salim et al6 showed that the degree of distortion
resorbed to form a single uterine cavity and cervix. of the uterine cavity in subseptate uterus was higher
Failure of resorption results in septate uterus. in women with recurrent miscarriage, compared
Müllerian duct anomalies occur throughout with low-risk women. The uterine cavity was
development, although the etiology of these defects mainly distorted due to the reduced length of unaf-
remains poorly understood. The most commonly fected cavity, rather than increased septum length.
used classification of müllerian duct anomalies is The greater degree of uterine cavity distortion in
that of the American Fertility Society (now named RPL supports the hypothesis of septal implantation
the American Society for Reproductive Medicine),3 as a potential cause of miscarriage, since the
which is shown in Table 11.1. likelihood of septal implantation increases with an
increasing ratio of septal size to functional cavity.
SUBSEPTATE UTERUS
ARCUATE UTERUS
Subseptate uterus is considered to be the most
common major uterine anomaly in women with An arcuate uterus has, by definition, an intrauterine
RPL4 and recurrent first-trimester pregnancy loss.5 indentation less than 1 cm. Using three-dimensional
Indeed, the subseptate uterus accounted for 70–90% (3D) ultrasound, it has been found that the preva-
of major anomalies found in low-risk women with lence of arcuate uterus was 17% in women with
uterine anomalies.6–9 recurrent miscarriage,6 which is significantly higher
than the prevalence of 3.2% in low-risk women.8 In
addition, it has been shown that distortion of the
uterine cavity is greater in women with recurrent
Table 11.1 Classification of müllerian duct
anomalies
first-trimester loss, as with the subseptate uterus.
The diagnosis of arcuate uterus is difficult
1. Class I: uterine agenesis or hypoplasia when conventional diagnostic methods such as
2. Class II: unicornuate uterus
hysteroscopy or laparoscopy are used, as the diag-
3. Class III: didelphys uterus
4. Class IV: bicornuate uterus nostic criteria are far from clear.12 As a result, little
5. Class V: septate uterus is known about the prevalence and clinical signifi-
6. Class VI: arcuate uterus
cance. Although many believe that the arcuate
7. Class VII: diethystilbestrol (DES)-exposed uterus
uterus has little or no impact on reproduction and
148
0415421306-Ch11 3/30/07 9:40 AM Page 149
UTERINE ANOMALIES AND RECURRENT PREGNANCY LOSS
obstetric outcomes, some studies have reported an pregnancies result in miscarriage.9,17,18 The high
increase in adverse reproductive outcomes, mostly miscarriage rate is mostly attributed to abnormal
second-trimester loss.9,13,14 The pathophysiology of uterine vasculature and decreased muscle mass.
fetal loss in women with arcuate uterus remains Increased cesarean section rates are a result of fetal
obscure. malpresentation and irregular uterine contractions
during labor.
UNICORNUATE UTERUS There are no surgical procedures to correct the
unicornuate uterus. Prophylactic cervical cerclage
A unicornuate uterus is the result of complete, or has been suggested for the prevention of miscarriage
almost complete, arrest of development of one of in patients with unicornuate uterus, although there
the müllerian ducts (Figure 11.1). When the arrest is no clear evidence of cervical incompetence.19
is incomplete (in 90% of patients with unicornuate However, with little data to support the use of cer-
uterus), a rudimentary horn with or without func- clage, most clinicians prefer to use careful follow-
tioning endometrium is present. If the rudimentary up, with frequent clinical and sonographic
horn is obstructed, it may present as an enlarging evaluation of cervical length. Resection of the cavi-
pelvic mass, with unilateral cyclical pelvic pain sec- tated rudimentary horn is often recommended in
ondary to hematometra. Pregnancies can occur in symptomatic patients with unicornuate uterus suf-
the rudimentary horn, with an estimated incidence fering from dysmenorrhea and hematometra.
of 2%. These cases may be difficult to diagnose, and Laparoscopic excision of the rudimentary horn has
may result in rupture of the rudimentary horn. been shown to be an effective surgical approach.20
The incidence of unicornuate uterus has been
estimated to be 6.3% of uterine anomalies, and may UTERUS DIDELPHYS
be associated with urinary tract anomalies, espe-
cially renal. Urinary tract anomalies should be sus- A double uterus results from the complete failure
pected in all women with a unicornuate uterus.15 of the two müllerian ducts to fuse (Figures 11.2
Unicornuate uterus is associated with the worst and 11.3). Therefore, each duct develops into
reproductive outcome.16 About one-third of all a separate uterus, each of which is narrower than a
Figure 11.1 Three-dimensional transvaginal ultrasound of a Figure 11.2 Two-dimensional transvaginal ultrasound of a
unicornuate uterus using volume contrast imaging. (Courtesy didelphys uterus with obstructed right vagina (hematocolpos).
of Professor Yaron Zalael MD, Sheba Medical Center, Tel- (Courtesy of Professor Yaron Zalael MD, Sheba Medical Center,
Hashomer, Israel.) Tel-Hashomer, Israel.)
149
0415421306-Ch11 3/30/07 9:40 AM Page 150
Figure 11.3 Two- and three-dimensional transvaginal ultrasound of a didelphys uterus using volume contrast imaging in plane C.
(Courtesy of Professor Yaron Zalael MD, Sheba Medical Center, Tel-Hashomer, Israel.)
normal uterus and has only a single horn. The two However, endometriosis is more commonly associ-
uteri may each have a cervix or may share a cervix. ated with a didelphic uterus, possibly because of
In 67% of cases, a uterus didelphys is associated retrograde menstruation.21
with two vaginas separated by a thin wall. Didelphic
uteri are relatively uncommon, with an estimated BICORNUATE UTERUS
incidence of 6.3% of uterine anomalies.6 The
two uteri do not always function normally, and are A bicornuate uterus results from partial non-fusion
associated with a miscarriage rate of 20.9% and of the müllerian ducts (Figure 11.4). The central
a preterm delivery rate of 24.4%.6,21 A long-term myometrium may extend to the level of the internal
follow-up of 49 Finnish women with didelphic cervical os (bicornuate unicollis) or external cervical
uterus and a longitudinal vaginal septum reported os (bicornuate bicollis). The latter is distinguished
an obstructed hemivagina in nine women (18%). from uterus didelphys because it demonstrates
Eight of these nine women also had ipsilateral renal some degree of fusion between the two horns, while
agenesis.21 Cesarean section rates are higher, due to in uterus didelphys the two horns and cervices
uterine dystocia and malpresentation.22 In addition, are separated completely. In addition, the horns of
didelphic uterus is commonly associated with a the bicornuate uteri are not fully developed; typi-
patent or obstructed vaginal septum. Fertility in cally, they are smaller than those of didelphys uteri.
women with didelphic uterus, is not notably impaired. Bicornuate uterus is proably the most common
150
0415421306-Ch11 3/30/07 9:40 AM Page 151
spontaneous pregnancy losses, or ectopic pregnan-

cies.23 Women exposed to DES in utero are also at
increased risk for breast cancer and clear cell adeno-
carcinoma of the vagina and cervix.24
Goldberg and Falcone,25 in a meta-analysis
study of DES-exposed subjects, found a ninefold
increase in ectopic pregnancy, a twofold increase in
miscarriage rate, and a twofold increase in preterm
delivery compared with a matched control popu-
lation. Pregnancy rates were similar between
DES-exposed women and controls: 72% and 79%,
respectively. The poor obstetric outcomes are caused
not only by the uterine anomaly, but also by an anti-
estrogenic effect at the level of the endometrium.22
The clinical significance of DES exposure is rapidly
Figure 11.4 Three-dimensional transvaginal ultrasound of a
bicornuate uterus. (Courtesy of Professor Yaron Zalael MD, diminishing as those affected women pass their
Sheba Medical Center, Tel-Hashomer, Israel.) reproductive years.22
MYOMAS
uterine anomaly after septate and arcuate uterus.22
The reproductive outcome seems to be directly Myomas are considered the most common acquired
correlated with the severity of fundal indentation.9 anomaly of the uterus. It has been shown26 that
It is generally considered that bicornuate uterus infertile women with fibroids have a lower pregnancy
does not directly affect infertility, but may be linked rate when undergoing assisted reproduction than
with recurrent miscarriages. Bicornuate uterus can age-matched women with no fibroids. Submucous
be corrected surgically by metroplasty. myomas deform the uterine cavity, and the overly-
ing endometrium is usually thin and inadequate for
T-SHAPED UTERUS AND DIETHYLSTILBESTROL normal implantation. Submucous fibroids can also
EXPOSURE be associated with pregnancy loss.27 The situation is
less clear with intramural and subserosal fibroids.
Diethylstilbestrol (DES) is a synthetic estrogen that In these locations, the size and the number of
was used from 1948 up to its ban in 1971 to prevent fibroids may be significant. Significantly lower
further pregnancy losses in women with RPL. implantation and pregnancy rates have been found
However, approximately two-thirds of embryos in patients with intramural or submucosal fibroids
exposed in utero developed uterine abnormalities, undergoing in vitro fertilization and intracytoplas-
including a characteristic small, incompletely formed mic sperm injection, even when there was no uterine
uterus with a T-shaped cavity and a hypoplastic cavity deformation.28 Furthermore, the pregnancy
cervix. The spontaneous incidence of T-shaped rate observed within 1 year of myomectomy is higher
uterus is unknown in the general population. than that observed in couples with unexplained
In addition, approximately half of DES-exposed infertility and no treatment.29,30
women have structural cervical defects, including
an incompletely formed cervix. The mechanism by POLYPS
which DES disrupts normal uterine development is
not known. DES-exposed women are less likely Polyps are benign hyperplastic endometrial
than unexposed women to have a full-term live growths that have also been associated with adverse
birth, and are more likely to have premature births, pregnancy outcomes. It is postulated that polyps
151
0415421306-Ch11 3/30/07 9:40 AM Page 152
and fibroids with intracavitary extension may act as a duplicated cervix, duplicated uterus, uterine
like foreign bodies within the endometrial cavity.31 septum, or unicornuate uterus. TVS is also useful in
It has also been proposed that polyps and fibroids determining the size and location of uterine
might induce chronic inflammatory changes in myomas, as well as the presence of intrauterine
the endometrium that make it unfavorable for polyps and endometrial irregularities that might
pregnancy. Since the presence of polyps has been suggest adhesions.
associated with a worse prognosis for pregnancy, Recent reports on 2D and 3D TVS, as well as
hysteroscopic polypectomy is usually considered saline contrast sonohysterography, appear promis-
if no other explanation for the recurrent loss is ing for diagnosis and classification of congenital
found.32 uterine anomalies.36 The ability to visualize both
the uterine cavity and the fundal uterine contour on
INTRAUTERINE ADHESIONS a 3D scan facilitates the diagnosis of uterine anom-
alies and enables differentiation between septate
Intrauterine synechiae may not be a frequent cause and bicornuate uteri. The additional use of color
of recurrent abortion, but may lead to secondary Doppler ultrasound may also allow visualization of
infertility in these patients. Intrauterine adhesions intraseptal vascularity and may help in distinguishing
develop as a result of surgical procedures, typically the avascular from the vascular septum.
curettage, or endometritis. Intrauterine scars can Intravenous pyelography is recommended
probably interfere with the normal implantation during the workup of congenital anomalies. Defects
process, and may be responsible for pregnancy loss. in the urinary tract are commonly seen when a uterine
Intrauterine adhesions are expected to be more anomaly is diagnosed.15
common among patients with recurrent abortions,
since the formation of adhesions may even follow a HYSTEROSALPINGOGRAPHY
simple manual vacuum aspiration for early preg-
nancy loss.33 Among 23 patients with an otherwise Hysterosalpingography (HSG) has long been used
unexplained history of three or more first- or to evaluate the contour of the uterine cavity, cervi-
second-trimester miscarriages and no live births, cal canal, and fallopian tube.37 The radio-opaque
hysteroscopy showed that 5 (21.8%) had intrauterine contrast medium fills the cavity, allowing the accu-
adhesions.34 rate identification of filling defects, scarring, or a
septum. However, HSG alone cannot differentiate
between a septate uterus and a bicornuate uterus.
INVESTIGATION OF UTERINE INTEGRITY Furthermore, HSG cannot determine the myome-
trial extension or the size of intrauterine lesions.
In patients with RPL, imaging studies are important Therefore, HSG is primarily used to assess tubal
during the initial workup in order to assess the patency, and has a limited role in the imaging of
integrity of the uterus. The guidelines of the Royal uterine malformations.
College of Obstetricians and Gynaecologists35 for
investigating recurrent miscarriage recommend an THREE-DIMENSIONAL ULTRASOUND
ultrasound scan of the pelvis, but this recommen-
dation is based solely on the clinical experience of 3D ultrasound is now accepted as an accurate and
the guideline development group, rather than on reproducible means for the diagnosis of congenital
published evidence. Transvaginal sonography (TVS) uterine anomalies38 (Figures 11.1 and 11.5). It has
is usually the initial step, but is now enhanced using clear advantage over HSG, hysteroscopy, and
3D-mode ultrasound. TVS allows accurate and rapid laparoscopy for the diagnosis of congenital uterine
characterization of the uterus, including its size and anomalies, since it is a non-invasive method and is
position, as well as the presence of anomalies such currently available in most out patient settings.
152
0415421306-Ch11 3/30/07 9:40 AM Page 153
anomalies detected incidentally by 3D ultrasound.

In their study, 1089 women with no history of infer-
tility or recurrent miscarriage, undergoing a trans-
vaginal ultrasound scan, were screened for uterine
abnormalities: 983 women had a normally shaped
uterine cavity, 72 an arcuate, 29 a subseptate, and
5 bicornuate uterus. Women with a subseptate
uterus had a significantly higher proportion of first-
trimester loss compared with women with a normal
uterus. Women with an arcuate uterus had a signif-
icantly greater proportion of second-trimester
loss and preterm labor. The study by Woelfer et al14
demonstrated the potential value of 3D ultrasound
and contributed evidence to the proposed associa-
Figure 11.5 Three-dimensional transvaginal ultrasound of a tion between congenital uterine anomalies and
septated uterus (3D rendering). (Courtesy of Professor Yaron adverse pregnancy outcomes.
Zalael MD, Sheba Medical Center, Tel-Hashomer, Israel.)
SONOHYSTEROGRAPHY
The results of 3D ultrasound have been shown to
concur with HSG in all cases of arcuate uterus and Valenzano et al39 have assessed the diagnostic value
major congenital anomalies.8 It has been suggested and usefulness of transvaginal sonohysterography
that the ability to visualize both the uterine cavity and (SHG) in the detection of uterine anomalies, com-
the myometrium on a 3D scan facilitates the diagno- pared with other diagnostic methods. SHG was able
sis of uterine anomalies and enables easy differ- to detect all uterine anomalies found in a study of
entiation between subseptate and bicornuate uteri. 54 patients with primary or secondary infertility or
Salim et al6 have examined the differences in repeated spontaneous abortion and with a clinically
the morphology of uterine anomalies found in or sonographically suspected abnormal uterus.
509 women with a history of unexplained recurrent SHG was carried out by the intrauterine injection
miscarriage and 1976 low-risk women who were of an isotonic saline solution. The sensitivity
examined for the presence of congenital uterine and specificity of SHG were the same as for
anomalies by 3D ultrasound. Salim et al6 detected hysteroscopy. However, there was no significant dif-
121 anomalies in the recurrent miscarriage group ference between the diagnostic capabilities of the
and 105 among low-risk women. Surprisingly, there methods analyzed. Valenzano et al39 therefore
was no significant difference in relative frequency of concluded that transvaginal SHG with saline solu-
various anomalies or the depth of fundal distortion tion is a low-cost, easy, and helpful examination
between the two groups. However, with both arcu- method for uterine malformations.
ate and subseptate uteri, the length of remaining It is now possible to combine 3D ultrasound
uterine cavity was significantly shorter and the with SHG. Sylvestre et al40 carried out a study of
distortion ratio was significantly higher in the 209 infertile patients suspected to have an intrauter-
recurrent miscarriage group. Salim et al6 therefore ine lesion on 3D SHG. Ninety-two patients with a
concluded that the distortion of uterine anatomy is lesion underwent hysteroscopy. In these 92 patients,
more severe in the congenital uterine anomalies, polyps were found in 48 women, submucous or
found in women with a history of recurrent first- intramural myomas in 35, both polyps and myomas
trimester miscarriage. in 3, müllerian anomalies in 4, thick endometrium
Woelfer et al14 tried to determine the reproduc- in 1, and intrauterine synechiae in 1. It was con-
tive outcomes in women with congenital uterine cluded that 3D SHG allowed precise recognition and
153
0415421306-Ch11 3/30/07 9:40 AM Page 154
localization of lesions. It was further suggested that and directed biopsies can be obtained when
if 2D and 3D SHG are normal, invasive diagnostic indicated. A retrospective study by Zuppi et al46
procedures such as hysteroscopy can be avoided. found an association between the hysteroscopic
Alborzi et al41 performed a prospective study to findings in 344 women with recurrent spontaneous
determine whether SHG can differentiate septate abortion and major (and even minor) uterine
from bicornuate uterus, in 20 patients with a anomalies. The anomalies were shown to correlate
history of RPL and an HSG diagnosis of septate or with an increased risk of recurrent miscarriage.46
bicornuate uterus. SHG was found to effectively Weiss et al47 performed hysteroscopy on
differentiate septate and bicornuate uterus, and 165 women referred for RPL: 67 after two and
may eliminate the need for laparoscopy in order to 98 after three or more consecutive miscarriages. The
differentiate between these uterine anomalies. prevalence of uterine anomalies did not differ signifi-
The diagnostic accuracy of SHG has been evalu- cantly: 32% and 28%, respectively. Weiss et al47 con-
ated prospectively compared with HSG and TVS in a cluded that hysteroscopy may be justified following
study comprising 65 infertile women.42 Hysteroscopy two spontaneous pregnancy losses.
was used as the gold standard. SHG was found to The intramyometrial extension of fibroids
have the same diagnostic accuracy, and sometimes cannot be assessed, however, and therefore the esti-
even to be markedly superior to hysteroscopy with mate of size remains imprecise. Hysteroscopy alone
respect to polypoid lesions and endometrial hyper- cannot differentiate between a septate uterus and a
plasia. In the diagnosis of intrauterine adhesions, bicornuate uterus; laparoscopy or SHG is required
SHG had limited accuracy, similar to that obtained to complete the evaluation. Hysteroscopic surgery is
by HSG, with a high false-positive diagnosis rate.42 currently considered the method of choice for cor-
recting the various types of intrauterine pathology.
MAGNETIC RESONANCE IMAGING
DIAGNOSTIC LAPAROSCOPY
Magnetic resonance imaging (MRI) is an accurate
non-invasive technique for the evaluation of uter- Laparoscopy allows the surgeon to assess the outer
ine anomalies. It has been shown to be a valuable surface of the uterus and other pelvic structures. It is
tool in the diagnosis of selected cases of müllerian used to establish the precise diagnosis of the various
duct anomalies.43 Although most anomalies will be congenital and acquired anomalies. Laparoscopy is
initially diagnosed with HSG and SHG, further also used for the removal of subserosal and intra-
imaging will often be required for definitive diagno- mural fibroids.48,49 Currently, laparoscopy is rarely
sis and elaboration of secondary findings.44 At this used just to clarify uterine anatomy, and is generally
time, MRI is justified only in special cases where its reserved for women in whom interventional therapy
high accuracy and detailed elaboration of uterovagi- is likely to be undertaken.
nal anatomy is needed.
The utility of MRI remains limited due to its CHOICE OF METHOD FOR IMAGING UTERINE
cost. However, in selected cases, careful use of MRI MORPHOLOGY
to delineate the pelvic soft tissues may greatly aid in
precise definition of the anomaly and in planning Ultrasonography is currently the most readily avail-
the most appropriate corrective surgery.45 able and least invasive mode of imaging in cases
of suspected uterine abnormalities (Table 11.2).
DIAGNOSTIC HYSTEROSCOPY 2D sonography allows excellent assessment of
myometrial morphology, and is especially useful for
Hysteroscopy offers the best and most direct assess- determining the number, size and location of
ment of the uterine cavity. During the procedure, myomas. Filling the uterine cavity with fluid facili-
the intracavitary structures can be directly visualized tates the use of SHG for accurate delineation of
154
0415421306-Ch11 3/30/07 9:40 AM Page 155
Table 11.2 Imaging modalities for assessing uterine anomalies in women with recurrent pregnancy loss
Imaging modalities Advantages Disadvantages Cost
Ultrasonography ● Readily available ● Poor demonstration of uterine contour Low

● Least invasive
● Excellent assessment of the myometrial ● Uterine cavity not clearly demonstrated
morphology
Hysterosalpingography ● Shows the contour of the uterine cavity, ● Exposure to radiation Moderate
cervical canal, and tubal lamina ● Iodine sensitivity risk
● Painful
● Pelvic inflammatory disease risk
● High false-positive rates
3D sonography ● Allows visualization of both the uterine ● Equipment not readily available Moderate
cavity and the myometrium ● Requires experienced operator
● Enables easy differentiation between
subseptate and bicornuate uteri
Sonohysterography ● Good evaluation of uterine ● Time-consuming Low

cavity ● High false-positive diagnosis rate
● Tubal patency assessed for intrauterine adhesions
Diagnostic hysteroscopy ● Most accurate assessment of the uterine ● Limited efficiency to differentiate Moderate
cavity between a uterine septum and a
● Simple outpatient procedure bicornuate uterus
● No information on tubal patency
● Invasive: risk of infection, perforation
Magnetic resonance ● Useful in clarifying details of ● No information on tubal patency High

imaging (MRI) soft tissue anatomy ● Not easy to interpret results
Diagnostic laparoscopy ● Accurate for differentiating between ● Invasive High

a uterine septum and a bicornuate ● Requires general anesthesia
uterus ● Low postoperative morbidity
intrauterine polyps, and improves the accuracy of However, hysteroscopy cannot fully differentiate
identifying submucous myomas encroaching on the between a uterine septum and a bicornuate uterus.
cavity and to assess the size of uterine septa. 3D The role of MRI is limited due to its cost.
sonography greatly enhances our ability to differen- However, in selected and complicated cases, MRI
tiate between a uterine septum and a bicornuate may help to clarify the details of soft tissue anatomy
uterus (Figures 11.2–11.5). HSG can help delineate and may be especially useful when planning surgical
the integrity of the uterine cavity, but, due to its correction. Laparoscopy used to be the gold standard
invasive nature and the associated exposure to radi- for differentiating between a uterine septum and a
ation, it is limited to infertility investigation where bicornuate uterus, but with modern imaging modal-
evaluation of tubal patency is required. ities, it is rarely needed for determination of uterine
Hysteroscopy can be performed nowadays with anatomy and is usually only used when a decision
2–3 mm scopes without the need for speculum, tenac- has been made to attempt surgical correction.
ulum, or anesthesia.50 This simple outpatient proce-
dure provides an accurate assessment of the uterine
cavity. It remains the method of choice for assessment TREATMENT
of the presence and extent of intrauterine adhesions.
It is also the optimal method to evaluate the size and As stated above, little evidence can be found in
extension of polyps and submucous myomas. the current literature demonstrating that uterine
155
0415421306-Ch11 3/30/07 9:40 AM Page 156
factors, including intrauterine adhesions, septa, by the ESHRE Special Interest Group for Early
myomas, and endometrial polyps, are causally linked Pregnancy (SIGEP) protocol for the investigation
with reproductive loss. However, there are reports and medical management of recurrent miscarriage
suggesting that treatment of these abnormalities concluded that the only interventions that do not
may improve fertility outcome.51 The published require more randomized controlled trials are
evidence includes several observational series that tender loving care and health advice.1
demonstrate successful fertility, with term preg- Surgery is the main treatment offered to patients
nancy rates ranging from 32% to 87% following with uterine anomalies (Table 11.3). However, not
hysteroscopic division of intrauterine adhesions.51 all anatomical defects can be surgically corrected,
The evidence supporting a direct link between a and not all anomalies require surgical intervention.
septate uterus and reproductive loss is derived from The most crucial step before making any treatment
the results of metroplasty. Several case series have decision is accurate imaging in order to determine
demonstrated a reduction in the spontaneous abor- the exact anomaly. Currently, endoscopic proce-
tion rate (from 91% to 17%), after hysteroscopic dures are the main approach used to correct most
metroplasty. Furthermore, following metroplasty, uterine defects. Operative hysteroscopy currently
the mean pregnancy rate in previously infertile allows a technically straightforward method of
patients is 47%. However, there are no prospective correcting intrauterine pathology such as septum,
controlled trials that have provided conclusive evi- fibroids, or polyps. Laparotomy currently has a very
dence that the correction of uterine anatomical limited role in the management of congenital uter-
abnormalities benefits the next pregnancy.52 ine anomalies in women with recurrent abortion.
Furthermore, the above data are mostly based on There are many questions regarding the optimal
observational, retrospective studies with small management of patients with recurrent abortions
sample sizes and heterogeneous patient popula- and uterine anomalies, such as the indications for
tions, and are therefore a far cry from the type of resection of a uterine septum and whether small
evidence required for current treatment guidelines. intrauterine polyps significantly influence repro-
A recent review of all published large randomized ductive performance. It is debatable whether surgi-
controlled trials and meta-analyses undertaken cal reconstruction such as Strassman’s metroplasty
Table 11.3 The role of surgical intervention in women with uterine anomalies and recurrent pregnancy loss
Study Postoperative morbidity Technical difficulty Likelihood of benefit Cost
Hysteroscopic polypecomy + + ++ +
Hysteroscopic adhesiolysis + + to ++ +++ +
Hysteroscopic myomectomy + to ++ ++ to +++ ++ + to ++
Hysteroscopic metroplasty + + ++ + to ++
for septate uterus
Hysteroscopic metroplasty + ++ + ++
for hypoplastic/DES-
exposed uterus
Abdominal metroplasty +++ +++ ++ +++
Cervical cerclage ++ ++ + ++
Interruption of a fallopian ++ ++ ++? ++

tube with hydrosalpinx
+ low; +++ high.
156
0415421306-Ch11 3/30/07 9:40 AM Page 157
should be performed for bicornuate uterus. When

should myomectomy be performed? What is the role
of non-surgical management of myomas? When
should cervical cerclage be offered? We will try to
discuss these questions in the light of the currently
available data.
SHOULD INTRAUTERINE POLYPS BE EXCISED?
Although the association between endometrial

polyps and pregnancy loss has not been proven,
polyps are more common in patients with recurrent
spontaneous abortion.53 Surgical excision is usually
recommended,2 since there are data suggesting that
hysteroscopic polypectomy can increase fertility.31,32 Figure 11.6 Two-dimensional transvaginal ultrasound of a sep-
A prospective randomized study in 215 infertile tated uterus. (Courtesy of Professor Yaron Zalael MD, Sheba
Medical Center, Tel-Hashomer, Israel.)
women scheduled to undergo intrauterine insemi-
nation (IUI) showed that hysteroscopic polypec-
tomy improved the likelihood of conception, with a
relative risk of 2.1 (95% confidence interval (CI) uterus and a septate uterus (Figures 11.6 and 11.7).
1.5–2.9).32 Pregnancies in the patients who under- In order to justify metroplasty, reliable diagnosis is
went polypectomy were obtained before the first required.
IUI in 65% of cases. Although no randomized controlled studies are
There is a consensus that hysteroscopy is the opti- available, observational studies have reported
mal method to perform polypectomy. Hysteroscopic impressive results following incision of a septum in
polypectomy can be performed by several tech- patients with recurrent abortion. Fedele et al57 stud-
niques, including excision with forceps or gentle ied the reproductive outcome in 102 patients with a
curettage. A recent study54 has assessed 240 cases of complete (n = 23) or partial (n = 79) septate uterus
hysteroscopic polypectomy using microscissors, and infertility or repeated abortion. Following hys-
grasping forceps, or electrosurgery either with a teroscopic metroplasty, the cumulative pregnancy
monopolar probe or a resectoscope. Resectoscopic and birth rates at 36 months were 89% and 75%,
polypectomy required more operating time, had respectively, in the septate uterus group and 80%
more glycine absorption and complications, but had and 67% in the subseptate uterus group. Fedele et al57
a lower recurrence rate than other hysteroscopic concluded that after hysteroscopic metroplasty, the
techniques. The resectoscope had a 0% recurrence reproductive prognosis was favorable and not influ-
rate and the grasping forceps had a 15% recurrence enced by the malformation subclass.57
rate.54 The introduction of bipolar electrodes may A meta-analysis of published retrospective data
increase the safety of hysteroscopic endometrial comparing pregnancy outcome before and after
polypectomy in an outpatient setting.55 hysteroscopic septoplasty indicated a marked
improvement after surgery.56 However, the signifi-
DOES THE RESECTION OF A UTERINE SEPTUM cance of this meta-analysis remains limited by the
IMPROVE PREGNANCY OUTCOME? nonrandomized observational methodology used
by the studies that were assessed. Grimbizis et al58
Septate uterus is more prevalent in women summarized the results of a a highly selected group
with repeated pregnancy loss.56 However, it may be of symptomatic patients drawn from a large
difficult to differentiate between a ‘normal’ arcuate number of reports. They had previously had term
157
0415421306-Ch11 3/30/07 9:40 AM Page 158
Figure 11.7 Two- and three-dimensional transvaginal ultrasound of a septated uterus of the same patient as in Figure 11.6 using
volume contrast imaging in plane C. (Courtesy of Professor Yaron Zalael MD, Sheba Medical Center, Tel-Hashomer, Israel.)
delivery and live birth rates of only 5%. After hys- uncomplicated hysteroscopic resection of submu-
teroscopic septum resection, the outcome was cous myomas or endometrial polyps.60 Uterine per-
remarkable, in that the subsequent term delivery foration and/or the use of electrosurgery increase
rate was approximately 75% and the live birth rate this risk, but are not considered independent risk
about 85%.58 However, this was not a randomized factors.60
trial. Homer et al56 have suggested that a septate
Transabdominal surgical techniques, such as the uterus per se is not an indication for surgical inter-
modified Tompkins metroplasty, are still occasion- vention, because it is not always associated with a
ally used to repair uterine septa.59 However, in light poor obstetric outcome. This approach is supported
of the low morbidity associated with hysteroscopic by a recent retrospective study of 67 patients who
resection and the possibility of performing the pro- had a complete septate uterus including the cervix
cedure on an ambulatory basis, abdominal surgery and a longitudinal vaginal septum.61 There was no
seems to be rarely, if ever, indicated.52 However, hys- association with primary infertility, and pregnancy
teroscopic metroplasty is associated with a substan- was reported to progress successfully without surgi-
tial, and as yet non-quantified, increased risk of cal treatment. The results did not support elective
uterine rupture during subsequent pregnancies.60 hysteroscopic incision of the septum in asympto-
This is especially significant when the risk of matic patients before the first pregnancy.61 In
uterine rupture after hysteroscopic metroplasty women with one miscarriage, the situation remains
is compared with that of women who underwent controversial, and a conservative approach has been
158
0415421306-Ch11 3/30/07 9:40 AM Page 159
suggested, since it is expected that after a single mis- higher pregnancy rates following hysteroscopic resec-
carriage 80–90% will have a live birth in the next tion. Since submucous myomas are easily treatable
pregnancy.56 However, a more liberalized approach in recurrent pregnancy loss, it has been suggested
to treatment is currently advocated by most author- that these patients should be identified early after
ities in light of the simplicity, minimal postopera- other potential causes of recurrent pregnancy loss
tive sequelae, and improved reproductive outcome have been eliminated.34
associated with hysteroscopic metroplasty.56,58 The location and size of the myomas are the two
parameters that influence the success of a future
SHOULD THE CERVICAL PORTION OF THE SEPTUM pregnancy.63 At present, it seems that subserosal
BE SPARED IN PATIENTS WITH A COMPLETE myomas have little, if any, effect on reproductive
SEPTATE UTERUS? outcome, especially if they are up to 5–7 cm in diam-
eter. The impact of intramural myomas on the out-
It was previously believed that the in patients with a come of pregnancy is still disputed.28,64 However,
complete septate uterus, the cervical portion of the intramural myomas that do not encroach upon the
septum should be spared and the dissection started endometrium also can be considered to be relatively
at the level of the internal os to avoid secondary cer- harmless to reproduction, if they are less than
vical incompetence.36 However, a recent multicenter 4–5 cm in diameter. Myomectomy is therefore cur-
randomized controlled clinical trial by Parsanezhad rently recommended for intramural myomas that
et al62 examined whether division of the cervical compress the uterine cavity and submucous myomas
portion of a uterine septum is associated with intra- significantly reduce pregnancy rates.63
operative bleeding, cervical incompetence, or sec- Hysteroscopic myomectomy is the gold standard
ondary infertility. Twenty-eight women with for the treatment of submucous myomas. Size and
complete uterine septum and a history of pregnancy intramural extension can limit its success, although
wastage or infertility were randomized to undergo this greatly depends on the operator’s experience.
metroplasty including division of the cervical por- The removal of larger fibroids may require two pro-
tion of the septum or the same procedure with cedures to avoid intraoperative complications.
preservation of the cervical portion. Resection of the Fibroids with significant intramural extension pres-
cervical portion was reported to make the procedure ent a challenge during the procedure.
safer, easier, and less complicated than preservation Laparoscopic myomectomy is gradually being
of the cervical septum.62 accepted as the gold standard for the removal
of most intramural and subserosal uterine myomas
MANAGEMENT OF MYOMAS IN RECURRENT in women who desire to preserve their uterus.48
PREGNANCY LOSS Traditionally, laparotomy used to be indicated
for the surgical management of fibroids in such
Myomas are frequently found in women of repro- locations, but laparoscopy can be used to manage
ductive age, and are more prevalent in women over most of these cases. Pregnancy rates following
35 years of age Although myomas are more prevalent myomectomy, both via laparoscopy and laparo-
in women with recurrent spontaneous abortion,34,53 tomy, are in the 50–60% range, with most having
the causal association remains poorly established. good outcomes.49 It is should be noted that sponta-
It is therefore still undetermined which women neous uterine rupture during pregnancy has been
will benefit most from surgical excision of uterine reported following laparoscopic myomectomy.65
myomas. Evidence – mostly from the in vitro fertil- Laparoscopic-assisted myomectomy (LAM) is
ization (IVF) literature – suggests that only those another new approach that is often a very conven-
myomas that distort the endometrial cavity impair ient and less invasive form of surgery.66 By decreas-
fertility.63 Patients with distorted uterine cavities ing the technical demands, and thereby the
due to submucous fibroids larger than 2 cm have operative time, LAM may be offered more widely
159
0415421306-Ch11 3/30/07 9:40 AM Page 160
to patients. In carefully selected cases, LAM is a 95% CI 1.4–27.7), malpresentations (OR 4.3; 95% CI
safe and efficient alternative to both laparoscopic 1.0–20.5), spontaneous abortion, abnormal placen-
myomectomy and myomectomy by laparotomy. tation, and postpartum hemorrhage. At present, it
These cases include patients with numerous large seems that although most pregnancies following
or deep intramural myomas. LAM allows easier uterine artery embolization are expected to have
repair of the uterus and rapid morcellation of the good outcomes, myomectomy should still be rec-
myomas. In women who desire a future pregnancy, ommended as the treatment of choice over uterine
LAM may be a better approach, because it allows artery embolization in most patients desiring future
meticulous suturing of the uterine defect in fertility.49,67
layers and thereby eliminates excessive electroco-
agulation.66 IS CERVICAL CERCLAGE INDICATED IN WOMEN
Laparoscopy is also being expanded to include WITH UTERINE ANOMALIES?
such techniques as laparoscopic uterine artery
ligation and directed laparoscopic cryomyolysis. Cervical incompetence has been associated with
However, many of these treatment options are uterine anomalies, as well as following in utero expo-
still associated with significant concerns regarding sure to DES.2 Furthermore, cervical incompetence
future reproductive performance. Additional is of special concern in women with RPL, as weak-
non-surgical techniques recently introduced to ening of the cervix may occasionally be due to
treat myomas include uterine artery embolization repeated trauma to the cervix, following overdilata-
and transabdominal interventional MRI-guided tion during repeated curettage.
cryoablation.48 Furthermore, the first MRI-guided Seidman et al68 have studied the effect of cervical
focused ultrasound surgery system used to treat cerclage on the survival rate of the fetus in 86 preg-
myomas was recently approved by the US Food and nancies in women with congenital uterine anom-
Drug Administration. It is apparent the physician’s alies and a random group of 106 pregnancies in
skills and experience, as well as local availability of women with normally shaped uteri.68 The uterine
these new techniques, will largely determine patient morphological factors were determined in all by
assignment to therapy.48 HSG, and, when necessary, hysteroscopy and
Uterine fibroid embolization is an increasingly laparoscopy were combined. The incidence of HSG-
popular, minimally invasive technique that has been proven cervical incompetence (23%) was similar in
successfully used in the management of sympto- the two groups. In the respective groups, 67 and
matic myomas.49,67 This procedure is not without 29 pregnancies were managed with cervical cerclage.
risk for women desiring to enhance their reproduc- The fetal outcome was stratified by cervical incom-
tive outcome. Following uterine fibroid emboliza- petence and obstetric history. The percentage of
tion, transient ovarian failure has been reported, viable newborns was significantly higher in women
as has permanent amenorrhea associated with with malformed uteri who underwent cerclage
endometrial atrophy. Amenorrhea seems to occur (88%) compared with those without cerclage
after the procedure in approximately 1% of patients (47%). No statistically significant beneficial effect of
and is highly age-dependent, with a reported inci- cerclage was found for normal uteri, even when
dence of 3% (range 1–7%) in women under 40 years only those patients with a history of recurrent fetal
of age and 41% (range 26–58%) in women over 50. loss were considered.68
The pregnancy rate has not been established fol- The precise indications for cervical cerclage
lowing uterine artery embolization. However, remain controversial. The Cervical Incompetence
higher rates of pregnancy complications have been Prevention Randomized Cerclage Trial (CIPRACT)
reported following uterine artery embolization found that therapeutic cerclage with bed rest
compared with myomectomy.49 These complica- reduces preterm delivery before 34 weeks of gesta-
tions include preterm delivery (odds ratio (OR) 6.2; tion and compound neonatal morbidity in women
160
0415421306-Ch11 3/30/07 9:40 AM Page 161
with risk factors and/or symptoms of cervical bicornuate uterus for pregnancy maintenance is
incompetence and a cervical length of less than poorly supported by data and rarely seems war-
25 mm before 27 weeks of gestation.69 Risk factors ranted. As a bicornuate uterus is usually associated
for cervical incompetence included in this major with problems during the third trimester of
study included, among others, DES exposure and pregnancy, the procedure should thus be limited to
uterine anomaly. very few well-selected cases with recurrent second-
Levine and Berkowitz70 studied the effect of con- and third-trimester problems. The development of a
servative management on pregnancy outcome in laparoscopic approach to metroplasty for bicornuate
120 DES-exposed women with and without gross uterus needs further study, since this new technique
structural lesions of the genital tract. Cerclage may be associated with less postoperative morbidity.
was limited to two women with a history of cervical
incompetence or acute cervical change in the DOES HYDROSALPINX AFFECT PREGNANCY
second trimester. Women with cervical change OUTCOME AFTER EARLY RECURRENT MISCARRIAGE?
occurring after 25 weeks of gestation were managed
with bed rest. It was found that the majority of It is well established that tubal disease, particularly
pregnancy losses in DES-exposed patients occurred hydrosalpinx, has a detrimental effect on the out-
in the first trimester. Patients exposed in utero to come of IVF. Although no randomized trial has
DES who had conservative management had good shown a significant benefit from surgical interven-
pregnancy outcomes.70 tion for tubal disease prior to IVF,73 a recent meta-
Cervical incompetence is a challenging clinical analysis concluded that laparoscopic salpingectomy
diagnosis, and is an infrequent cause of pregnancy should be considered for all women with hydrosalp-
loss even in patients with gross structural abnor- inges prior to IVF. It was also stated that further
malities of the genital tract. Prophylactic cerclage randomized trials are required to assess other treat-
for patients with uterine anomalies and DES expo- ment modalities for hydrosalpinx, such as salpin-
sure should be recommended only when other risk gostomy, tubal occlusion, or aspiration at oocyte
factors, such as three or more midtrimester preg- retrieval.73
nancy losses or preterm deliveries, are present.68,70 A recent prospective randomized controlled
trial74 enrolled 13 patients with a history of unex-
DOES STRASSMAN METROPLASTY STILL HAVE A plained recurrent early spontaneous abortion and a
ROLE IN PATIENTS WITH A BICORNUATE UTERUS? unilateral hydrosalpinx diagnosed by sonography
or HSG and in whom other causes of abortion had
The Strassman procedure involves the unification been excluded. The patients were randomized to
of the two uterine horns of a bicornuate uterus, and undergo laparoscopic unilateral tubal fulguration
is carried out via laparotomy. This procedure often or no surgical intervention. Six of the seven patients
leaves a small cavity with scarring, which makes in the treatment group and five of the six in the
implantation difficult, and may also cause pelvic control group conceived. Five patients in the treat-
adhesions resulting in secondary infertility. ment group and none in the control group had a
However, the postmetroplasty reproductive capac- pregnancy progress beyond the first trimester. The
ity of women with a bicornuate uterus has been progressing pregnancies in the treatment group
reported to be good.71,72 Furthermore, the role of reached 36–40 weeks of gestation – a statistically
abdominal metroplasty has been suggested as a significant difference. The authors concluded that
valid approach72 (using the Jones or Strassman laparoscopic tubal fulguration improves pregnancy
techniques) in patients with bicornuate, T-shaped, outcome in selected patients with previous recur-
or septate uteri, when associated with other pelvic rent early abortion and a unilateral hydrosalpinx.
lesions not amenable to the transcervical hystero- This study urgently needs confirmation in a larger
scopic approach. However, surgical correction of a patient sample.74
161
0415421306-Ch11 3/30/07 9:40 AM Page 162
IS HYSTEROSCOPIC METROPLASTY INDICATED clearly established.49 Consequently, the investiga-

IN DES-EXPOSED WOMEN? tion of women with recurrent abortion should be
limited in most cases to screening with ultrasonog-
Hysteroscopic metroplasty has been reported as a raphy, preferably utilizing 3D techniques and in
safe and feasible method to improve reproductive selected cases benefiting from the application of
performance in patients with DES-exposed and hydrosonography (Table 11.2). More invasive and
hypoplastic malformed uteri suffering from severe expensive imaging modalities, including hys-
infertility, recurrent pregnancy loss, or implanta- teroscopy, laparoscopy, and MRI, should be reserved
tion failures in an IVF programme.75–77 In one for inconclusive cases with a suspected uterine
series,75 eight patients referred for infertility, recur- deformity.
rent pregnancy loss, or both with an abnormal uter- Surgical intervention for uterine malformations
ine contour as seen by HSG underwent hysteroscopic remains poorly supported by randomized con-
metroplasty. Eeach patient served as their own controlled trials (Table 11.3). It is generally agreed that
trol. Three of five patients with secondary infertility adhesions, polyps, and protruding submucous
and RPL had live births, as did a patient with second- myomas should be hysteroscopically resected.
ary infertility.75 However, the need for hysteroscopic division of a
In a larger study76 with a similar design, 24 patients uterine septum remains debatable, but may be indi-
with hypoplastic uterus and/or uterine deformity as cated in a patient with two or more pregnancy
seen by HSG underwent hysteroscopic metroplasty. losses, as its associated morbidity is low. Abdominal
Postoperative HSG showed an improved uterine metroplasty for bicornuate uterus is even more dif-
cavity in 23 cases. The final result was considered ficult to support in the light of its significant associ-
to be excellent in terms of anatomical correction in ated morbidity and the lack of controlled data.
15 patients. Eleven pregnancies occurred, the abor- Abdominal metroplasty is currently recommended
tion rate decreased from 88% in previous pregnan- only in selected cases with recurrent severe prob-
cies to 12.5%, and the rate of term deliveries lems in the second and third trimesters. Cervical
increased from 3% to 87.5%.76 Ten patients deliv- cerclage is only indicated in women with uterine
ered healthy infants after 30 weeks of gestation; one anomalies in the presence of a clinical diagnosis of
patient delivered more prematurely. Six deliveries cervical incompetence or additional risk factors.
were normal and four required cesarean section.76 In women with hydrosalpinges and early recurrent
At present, it seems that hysteroscopic metro- abortion, laparoscopic salpingectomy or proximal
plasty, with its simplicity and minimal postoperative tubal occlusion should be considered.
sequelae, seems to be an operation of choice in Miscarriages – clinically detectable pregnancies
women with a hypoplastic malformed uterus and a that fail to progress – seem to be the inevitable
history of severe infertility and/or RPL.77 However, the byproduct of the limited efficiency of human repro-
previously quoted series used historical controls. duction, and do not always point to the presence of
Larger series with a better study design are necessary a correctable defect. Thus, surgical intervention
before hysteroscopic metroplasty can be recom- should be carefully considered based on the patient’s
mended for all women with DES-exposed or hypoplas- clinical history and not merely as an attempt to
tic malformed uterus and recurrent miscarriage. correct all anatomical uterine defects now more com-
monly diagnosed by modern imaging modalities.
CONCLUSIONS REFERENCES
A contemporary review of the current literature 1. Jauniaux E, Farquharson RG, Christiansen OB, Exalto N, On behalf of
ESHRE Special Interest Group for Early Pregnancy (SIGEP). Evidence-
shows that the prevalence and impact of uterine based guidelines for the investigation and medical treatment of recur-
malformations on reproduction are still not rent miscarriage. Hum Reprod 2006; 21:2216–22.
162
0415421306-Ch11 3/30/07 9:40 AM Page 163
2. Devi Wold AS, Pham N, Arici A. Anatomic factors in recurrent preg- 24. Veurink M, Koster M, Berg LT. The history of DES, lessons to be
nancy loss. Semin Reprod Med 2006; 24:25–32. learned. Pharm World Sci 2005; 27:139–43.
3. The American Fertility Society classifications of adnexal adhesions, 25. Goldberg JM, Falcone T. Effect of diethylstilbestrol on reproductive
distal tubal occlusion, tubal occlusion secondary to tubal ligation, function. Fertil Steril 1999; 72:1–7.
tubal pregnancies, müllerian anomalies and intrauterine adhesions. 26. Eldar-Geva T, Meagher S, Healy DL, et al. Effect of intramural, sub-
Fertil Steril 1988; 49:944–55. serosal, and submucosal uterine fibroids on the outcome of assisted
4. Homer HA, Li TC, Cooke ID. The septate uterus: a review of manage- reproductive technology treatment. Fertil Steril 1998; 70:687–91.
ment and reproductive outcome. Fertil Steril 2000; 73:1–4. 27. Casini ML, Rossi F, Agostini R, et al. Effects of the position of fibroids
5. Proctor JA, Haney AF. Recurrent first trimester pregnancy loss is on fertility. Gynecol Endocrinol 2006; 22:106–9.
associated with uterine septum but not with bicornuate uterus. Fertil 28. Oliveira FG, Abdelmassih VG, Diamond MP, et al. Impact of
Steril 2003; 80:1212–15. subserosal and intramural uterine fibroids that do not distort
6. Salim R, Regan L, Woelfer B, et al. A comparative study of the the endometrial cavity on the outcome of in vitro fertilization–
morphology of congenital uterine anomalies in women with and intracytoplasmic sperm injection. Fertil Steril 2004; 81:582–7.
without a history of recurrent first trimester miscarriage. Hum 29. Ribeiro SC, Reich H, Rosenberg J, et al. Laparoscopic myomectomy
Reprod 2003; 18:162–6. and pregnancy outcome in infertile patients. Fertil Steril 1999;
7. Simon C, Martinez L, Pardo F, et al. Müllerian defects in women with 71:571–4.
normal reproductive outcome. Fertil Steril 1991; 56:1192–3. 30. Rossetti A, Sizzi O, Soranna L, et al. Long-term results of laparoscopic
8. Jurkovic D, Geipel A, Gruboeck K, et al. Three-dimensional myomectomy: recurrence rate in comparison with abdominal
ultrasound for the assessment of uterine anatomy and detection of myomectomy. Hum Reprod 2001; 16:770–4.
congenital anomalies: a comparison with hysterosalpingography and 31. Neuwirth RS, Levin B, Keltz MD. Pregnancy rates after hysteroscopic
two-dimensional sonography. Ultrasound Obstet Gynecol 1995; polypectomy and myomectomy in infertile women. Obstet Gynecol
5:233–7. 1999; 94:168–71.
9. Raga F, Bauset C, Remohi J, et al. Reproductive impact of congenital 32. Perez-Medina T, Bajo-Arenas J, Salazar F, et al. Endometrial polyps
müllerian anomalies. Hum Reprod 1997; 12:2277–81. and their implication in the pregnancy rates of patients undergoing
10. Dabirashrafi H, Bahadori M, Mohammad K, et al. Septate uterus: new intrauterine insemination: a prospective, randomized study. Hum
idea on the histologic features of the septum in this abnormal uterus. Reprod 2005; 20:1632–5.
Am J Obstet Gynecol 1995; 172:105–7. 33. lton VK, Saunders NA, Harris LH, et al. Intrauterine adhesions after
11. Nakada K, Makino T, Tabuchi S, et al. Analysis of congenital uterine manual vacuum aspiration for early pregnancy failure. Fertil Steril
anomalies in habitual abortions, evaluation of metroplasty. Jpn J 2006; 85:1823.
Fertil Steril 1989; 34:842–7. 34. Ventolini G, Zhang M, Gruber J. Hysteroscopy in the evaluation of
12. Golan A, Ron-El R, Herman A, et al. Diagnostic hysteroscopy: its patients with recurrent pregnancy loss: a cohort study in a primary
value in an in-vitro fertilization/embryo transfer unit. Hum Reprod care population. Surg Endosc 2004; 18:1782–4.
1992; 7:1433–4. 35. Royal College of Obstetricians and Gynaecologists. The Investigation
13. Acien P. Reproductive performance of women with uterine malfor- and Treatment of Recurrent Miscarriage. Guideline No. 17. London:
mations. Hum Reprod 1993; 8:122–6. RCOG Press, 2003.
14. Woelfer B, Salim R, Banerjee S, et al. Reproductive outcomes in women 36. Kupesic S. Clinical implications of sonographic detection of uterine
with congenital uterine anomalies detected by three-dimensional anomalies for reproductive outcome. Ultrasound Obstet Gynecol
ultrasound screening. Obstet Gynecol 2001; 98:1099–103. 2001; 18:387–400.
15. Fedele L, Bianchi S, Agnoli B, et al. Urinary tract anomalies associated 37. Baramki TA. Hysterosalpingography. Fertil Steril 2005; 83:1595–606.
with unicornuate uterus. J Urol 1996; 155:847–8. 38. Salim R, Woelfer B, Backos M, et al. Reproducibility of three-
16. Heinonen PK, Saarikoski S, Pystynen P. Reproductive performance dimensional ultrasound diagnosis of congenital uterine anomalies.
of women with uterine anomalies. An evaluation of 182 cases. Acta Ultrasound Obstet Gynecol 2003; 21:578–82.
Obstet Gynecol Scand 1982; 61:157–62. 39. Valenzano MM, Mistrangelo E, Lijoi D, et al. Transvaginal sonohys-
17. Fedele L, Bianchi S, Tozzi L, et al. Fertility in women with unicornu- terographic evaluation of uterine malformations. Eur J Obstet
ate uterus. Br J Obstet Gynaecol 1995; 102:1007–9. Gynecol Reprod Biol 2006; 124:246–9.
18. Heinonen PK. Unicornuate uterus and rudimentary horn. Fertil Steril 40. Sylvestre C, Child TJ, Tulandi T, et al. A prospective study to evaluate
1997; 68:224–30. the efficacy of two- and three-dimensional sonohysterography in
19. Surico N, Ribaldone R, Arnulfo A, et al. Uterine malformations and women with intrauterine lesions. Fertil Steril 2003; 79:1222–5.
pregnancy losses: Is cervical cerclage effective? Clin Exp Obstet 41. Alborzi S, Dehbashi S, Parsanezhad ME. Differential diagnosis of
Gynecol 2000; 27:147–9. septate and bicornuate uterus by sonohysterography eliminates the
20. Fedele L, Bianchi S, Zanconato G, et al. Laparoscopic removal of the need for laparoscopy. Fertil Steril 2002; 78:176–8.
cavitated noncommunicating rudimentary uterine horn: surgical 42. Soares SR, Barbosa dos Reis MM, Camargos AF. Diagnostic accuracy
aspects in 10 cases. Fertil Steril 2005; 83:432–6. of sonohysterography, transvaginal sonography, and hysterosalpingo-
21. Heinonen P. Clinical implications of the didelphic uterus: long-term graphy in patients with uterine cavity diseases. Fertil Steril 2000;
follow-up of 49 cases. Eur J Obstet Gynecol Reprod Biol 2000; 73:406–11.
91:183–90. 43. Marten K, Vosshenrich R, Funke M, et al. MRI in the evaluation of
22. Lin PC. Reproductive outcomes in women with uterine anomalies. müllerian duct anomalies. Clin Imaging 2003; 27:346–50.
J Womens Health (Larchmt) 2004; 13:33–9. 44. Troiano RN, McCarthy SM. Müllerian duct anomalies: imaging and
23. Kaufman RH, Adam E, Hatch EE, et al. Continued follow-up of clinical issues. Radiology 2004; 233:19–34.
pregnancy outcomes in diethylstilbestrol-exposed offspring. Obstet 45. Pui MH. Imaging diagnosis of congenital uterine malformation.
Gynecol 2000; 96:483–9. Comput Med Imaging Graph 2004; 28:425–33.
163
0415421306-Ch11 3/30/07 9:40 AM Page 164
46. Zupi E, Marconi D, Vaquero E, et al. Hysteroscopic findings in 64. Benecke C, Kruger TF, Siebert TI, et al. Effect of fibroids on fertility in
344 women with recurrent spontaneous abortion. J Am Assoc patients undergoing assisted reproduction. A structured literature
Gynecol Laparosc 2001; 8:398–401. review. Gynecol Obstet Invest 2005; 59:225–30.
47. Weiss A, Shalev E, Romano S. Hysteroscopy may be justified after two 65. Seidman DS, Nezhat CH, Nezhat FR, Nezhat C. Spontaneous uterine
miscarriages. Hum Reprod 2005; 20:2628–31. rupture in pregnancy 8 years after laparoscopic myomectomy. J Am
48. Seidman DS, Nezhat CH, Nezhat F, et al. Minimally invasive surgery Assoc Gynecol Laparoscop 2001; 8:333–5.
for fibroids. Infert Reprod Med Clin North Am 2002; 13:375–91. 66. Seidman DS, Nezhat FR, Nezhat CH, Nezhat CR. The role of laparo-
49. Goldberg J, Pereira L. Pregnancy outcomes following treatment for scopic-assisted myomectomy (LAM). JSLS 2001; 5:299–303.
fibroids: uterine fibroid embolization versus laparoscopic myomec- 67. Mara M, Fucikova Z, Maskova J, et al. Uterine fibroid embolization
tomy. Curr Opin Obstet Gynecol 2006; 18:402–6. versus myomectomy in women wishing to preserve fertility: prelimi-
50. Sagiv R, Sadan O, Boaz M, et al. A new approach to office hysteroscopy nary results of a randomized controlled trial. Eur J Obstet Gynecol
compared with traditional hysteroscopy: a randomized controlled Reprod Biol 2006; 126:226–33.
trial. Obstet Gynecol 2006; 108:387–92. 68. Seidman DS, Ben-Rafael Z, Bider D, et al. The role of cervical cerclage
51. Sanders B. Uterine factors and infertility. J Reprod Med 2006; in the management of uterine anomalies. Surg Gynecol Obstet 1991;
51:169–76. 173:384–6.
52. Management of recurrent early pregnancy loss. ACOG Practice 69. Althuisius SM, Dekker GA, Hummel P, et al. Final results of the
Bulletin, No. 24, February 2001. Int J Gynecol Obstet 2002; 78:179–90. Cervical Incompetence Prevention Randomized Cerclage Trial
53. Valli E, Zupi E, Marconi D, et al. Hysteroscopic findings in 344 women (CIPRACT): therapeutic cerclage with bed rest versus bed rest alone.
with recurrent spontaneous abortion. J Am Assoc Gynecol Laparosc Am J Obstet Gynecol 2001; 185:1106–12.
2001; 8:398–401. 70. Levine RU, Berkowitz KM. Conservative management and
54. Preutthipan S, Herabutya Y. Hysteroscopic polypectomy in 240 pregnancy outcome in diethylstilbestrol-exposed women with and
premenopausal and postmenopausal women. Fertil Steril 2005; without gross genital tract abnormalities. Am J Obstet Gynecol 1993;
83:705–9. 169:1125–9.
55. Marsh F, Rogerson L, Duffy S. A randomised controlled trial compar- 71. Lolis DE, Paschopoulos M, Makrydimas G, et al. Reproductive out-
ing outpatient versus daycase endometrial polypectomy. BJOG 2006; come after Strassman metroplasty in women with a bicornuate
113:896–901. uterus. J Reprod Med 2005; 50:297–301.
56. Homer HA, Li TC, Cooke ID. The septate uterus: a review of manage- 72. Khalifa E, Toner JP, Jones HW Jr. The role of abdominal metroplasty
ment and reproductive outcome. Fertil Steril 2000; 73:1–14. in the era of operative hysteroscopy. Surg Gynecol Obstet 1993;
57. Fedele L, Arcaini L, Parazzini F, et al. Reproductive prognosis after 176:208–12.
hysteroscopic metroplasty in 102 women: life-table analysis. Fertil 73. Johnson NP, Mak W, Sowter MC. Surgical treatment for tubal disease
Steril 1993; 59:768–72. in women due to undergo in vitro fertilisation. Cochrane Database
58. Grimbizis GF, Camus M, Tarlatzis BC, et al. Clinical implications of Syst Rev 2004; (3):CD002125.
uterine malformations and hysteroscopic treatment results. Hum 74. Zolghadri J, Momtahan M, Alborzi S, et al. Pregnancy outcome in
Reprod Update 2001; 7:161–74. patients with early recurrent abortion following laparoscopic tubal
59. Patton PE, Novy MJ, Lee DM, et al. The diagnosis and reproductive corneal interruption of a fallopian tube with hydrosalpinx. Fertil
outcome after surgical treatment of the complete septate uterus, Steril 2006; 86:149–51.
duplicated cervix and vaginal septum. Am J Obstet Gynecol 2004; 75. Nagel TC, Malo JW. Hysteroscopic metroplasty in the diethylstilbestrol-
190:1669–75. exposed uterus and similar nonfusion anomalies: effects on subse-
60. Sentilhes L, Sergent F, Roman H, et al. Late complications of operative quent reproductive performance; a preliminary report. Fertil Steril
hysteroscopy: predicting patients at risk of uterine rupture during 1993; 59:502–6.
subsequent pregnancy. Eur J Obstet Gynecol Reprod Biol 2005; 76. Garbin O, Ohl J, Bettahar-Lebugle K, et al. Hysteroscopic metroplasty
120:134–8. in diethylstilboestrol-exposed and hypoplastic uterus: a report on
61. Heinonen PK. Complete septate uterus with longitudinal vaginal 24 cases. Hum Reprod 1998; 13:2751–5.
septum. Fertil Steril 2006; 85:700–5. 77. Barranger E, Gervaise A, Doumerc S, et al. Reproductive performance
62. Parsanezhad ME, Alborzi S, Zarei A, et al. Hysteroscopic metroplasty after hysteroscopic metroplasty in the hypoplastic uterus: a study of
of the complete uterine septum, duplicate cervix, and vaginal septum. 29 cases. BJOG 2002; 109:1331–4.
Fertil Steril 2006; 85:1473–7.
63. Kolankaya A, Arici A. Myomas and assisted reproductive technologies:
when and how to act? Obstet Gynecol Clin North Am 2006;
33:145–52.
164
0415421306-Ch12 3/30/07 9:41 AM Page 165
12. Immunobiology of recurrent miscarriage

INTRODUCTION autoantibodies, with antiphospholipid antibodies

(aPL) predominating.4 The observation that most
Initially, recurrent spontaneous abortions (RSA) of these women with abnormal autoimmune func-
were considered to be due either to chromosomal tion have no other symptoms except reproduction-
aberrations of the fetus that are incompatible with related symptoms led to the definition of a
its development or to maternal causes such as uter- separated diagnostic entity: reproductive autoim-
ine anatomical abnormalities, hormonal or meta- mune failure syndrome (RAFS).5 In 1999, the
bolic disturbances, hereditary thrombophilias, and American Society of Reproductive Immunology,
infectious agents. When all the above causes of mis- considering autoimmune abortions as one of the
carriage were excluded, the miscarriages were char- manifestations not only of RAFS but also of
acterized as ‘unexplained miscarriages’. During the antiphospholipid syndrome (APS), suggested that
last 20 years, it has become clear that a large propor- these two syndromes should be included in a
tion of unexplained RSA (possibly more than 80%) broader clinical entity: reproductive autoimmune
may be due to immunological causes.1 syndrome (RAS).6,7 Clinical and laboratory findings
In the 1980s, immune-mediated abortions of APS and RAFS are presented in Table 12.1.
were considered as a syndrome characterized by Given the clinical findings of RAS, an autoim-
(a) more than two miscarriages with the same part- mune cause of miscarriages should be suspected in
ner, (b) a higher frequency of ectopic pregnancies, women with a history of (1) three or more consec-
(c) a tendency to infertility (because of miscar- utive preembryonic or embryonic pregnancy losses,
riages), and (d) a higher frequency of fetal growth and (2) one or more unexplained fetal deaths above
retardation (in the case of live birth followed by 10 weeks of gestation. Testing for autoimmune dis-
recurrent losses).2 Today, immune-mediated abor- turbances should be also considered in women with
tions are known to be characterized by either fewer miscarriages if they have experienced throm-
autoimmune or alloimmune disturbances.3 In bosis or autoimmune thrombocytopenia (criteria
autoimmune abortions, the development of the pla- for APS), or they have a history of endometriosis or
centa and the embryo is affected by maternal autoan- unexplained difficulty in conceiving, or even a his-
tibodies and autoreactive cells, which target decidual tory of fetal growth retardation, severe preeclampsia
and trophoblastic molecules. In alloimmune abor- or other obstetric complications (i.e., abruptio pla-
tions, the maternal immune system reacts against centae, chorea gravidarum, herpes gestationis,
the embryo and damages trophoblast through HELLP syndrome (hemolysis, elevated liver enzymes,
allogeneic, rejection-type reactions. and low platelet count)) in previous successful preg-
nancies (criteria for RAFS). The workup in autoim-
mune abortions must include testing for aPL
AUTOIMMUNE ABORTIONS gammopathies (mainly immunoglobulin M (IgM)),
antinuclear antibodies (ANA), and organ-specific
Maternal autoimmune disturbances may be the autoantibodies. The most important among the
cause of a high percentage of hitherto-unexplained above disturbances is the increase in aPL, which are
miscarriages. Approximately 30% of women with recognized to have the strongest association with
‘unexplained’ RSA have increased serum levels of pregnancy loss. There is evidence that 2–20% of
165
0415421306-Ch12 3/30/07 9:41 AM Page 166
Table 12.1 Clinical and laboratory findings in reproductive autoimmune syndrome (RAS)
Antiphospholipid syndrome (APS) Reproductive autoimmune failure syndrome (RAFS)
CLINICAL ● Thrombosis (≥1 unexplained venous or arterial ● Fetal growth retardation (<34 weeks)
FEATURES thrombosis, including stroke) ● Severe preeclampsia
● Autoimmune thrombocytopenia ● Obstetric complications (abruptio placentae,
● Recurrent pregnancy loss: chorea gravidarum, herpes gestationis, HELLP
≥1 consecutive and otherwise unexplained fetal deaths syndromea)
(≥ 10 weeks) ● Unexplained infertility
≥ 3 consecutive and otherwise unexplained ● Endometriosis
preembryonic or embryonic pregnancy losses ● Recurrent pregnancy loss:
≥1 consecutive and otherwise unexplained fetal
deaths (≥ 10 weeks)
≥ 3 consecutive and otherwise unexplained
preembryonic or embryonic pregnancy losses
LABORATORY ● Anticardiolipin antibodies (aCL) (>20 GPL or ● Antiphospholipid antibodies (aPL)

FINDINGS MPL units) ● Lupus anticoagulant (LA)
● Lupus anticoagulant (LA) ● Gammopathy (usually polyclonal, mostly
immunoglobulin M (IgM))
● Antinuclear antibodies (ANA) (including
antibodies against histones)
● Organ-specific autoantibodies (antithyroid
antibodies (ATA), anti-smooth muscle antibodies
(ASMA))
a
Hemolysis, elevated liver enzymes, and low platelet count.
women with recurrent preembryonic or embryonic extracellular fluid or bloodstream.10 In situations of

pregnancy losses have increased titers of aPL and ischemia, cell injury or abnormal immunregulation
that these women have an 80–90% pregnancy loss (autoimmunity), negatively charged PL can be exteri-
rate, with half of their pregnancies being lost in the orized to the outer leaflet, while in the presence of
first trimester.8,9 excess calcium or low pH, cone-shaped, hexagonal-
phase phospholipid configurations can be formed.
AUTOIMMUNE ABORTIONS ASSOCIATED WITH These changes may either provide an antigenic stimulus
ANTIPHOSPHOLIPID ANTIBODIES for the production of aPL or permit a number of serum
proteins with procoagulant activity (β2-glycoprotein I
FORMATION AND CHARACTERIZATION OF (β2GPI), prothrombin, protein C, protein S, and
ANTIPHOSPHOLIPID ANTIBODIES annexin V) to bind PL epitopes and to be presented to
the immune system in unique ‘neoantigenic’ confor-
Phospholipids (PL) are the basic components of all mations, which are recognized and give rise to aPL.11
cell membranes, where they are present in two layers. In the latter case, aPL may recognize either only the
Each PL consists of a glycerol moiety attached to two PL region of the complex or an epitope consisting of
esterified fatty acid chains (one saturated and one the portion of the PL and neighboring amino acids
unsaturated) as well as a phosphodiester-linked alco- on the protein carrier or they clearly act with the pro-
hol side-chain. In normal situations, the inner leaflet tein alone. The most important among the proteins
of the phospholipid bilayers is composed of nega- that bind to PLs is β2GPI, the antibodies against
tively charged or anionic alcohol groups facing the which are usually enlisted in the group of aPL.3,4
cytoplasm, whereas the outer layer is composed of Recently, Kuwana et al12 were able to detect β2GPI-
neutral or zwitterionic alcohol groups facing the specific CD4+ and human leukocyte antigen
166
0415421306-Ch12 3/30/07 9:41 AM Page 167
IMMUNOBIOLOGY OF RECURRENT MISCARRIAGE
(HLA) class II restricted autoreactive T cells, which The main cause of fetal loss in the presence of
preferentially recognize the antigenic peptide con- aPL is hypoxia to the placenta because of uteropla-
taining the major pPL-binding site and have the cental blood supply insufficiency resulting from
capacity to stimulate B cells to produce pathogenic multiple intervillous thromboses, intravillous
anti-β2GPI antibodies.12 infractions, and decidual vasculopathy. Additionally
Because of the different methods whereby induc- to thrombosis, aPL directly target trophoblastic
tion can occur, aPL are a heterogeneous group of cells and may affect pregnancy by inhibiting normal
autoantibodies.11 aPL that bind to PL, present in PL functions related to trophoblastic cell division
unique hexagonal phases either alone or complexed (aPE), intertrophoblastic fusion, hormone secretion
with prothrombin or β2GPI, prolong pPL-dependent and trophoblast invasion (aPS).15
clotting assays and are known as lupus anticoagu-
lants (LA). The subgroup of aPL that bind to PL/pro- AUTOIMMUNE ABORTIONS NOT ASSOCIATED
tein complexes and may or may not prolong WITH aPL
PL-dependent clotting assays include antibodies
against cardiolipin (aCL), phosphatidylethanolamine The following autoantibodies other than aPL are
(aPE), phosphatidylserine (aPS), phosphatidyl- included in the diagnostic criteria for RAFS and are
choline (aPC), phosphatidylglycerol (aPG), phos- associated with pregnancy loss.
phatidylinositol (aPI), and phosphatidic acid (aPA).
These antibodies are detected by immunoassays ANTINUCLEAR ANTIBODIES AND ANTIBODIES TO
using PL-coated surfaces, and there is no agreement NUCLEAR ANTIGENS
if there is a correlation between them and LA or
discordance between their prevalence. ANA and antibodies against single- and double-
stranded DNA (ssDNA and dsDNA) appear to be
FORMATION OF aPL DURING PREGNANCY increased in about 35% of women with RSA, while
their percentage is less than 10% in fertile women
Pregnancy itself appears to be a triggering event with no abortion history. Antibodies against his-
that allows protein cofactors to bind PL and become tones or non-DNA nuclear components (Sm, RNP,
antigenic for aPL production. Placental tissues SSA, SSB, and Scl70) are also found in some of these
continuously change, and this major tissue women.16 The presence of ANA is associated with
remodeling results in externalization of inner inflammation around the placenta, so that the envi-
surface PL, which, when appearing on the outer ronment does not enhance the ‘acceptance’ of the
membrane, may either be a direct stimulus for aPL embryo. In these cases, anti-inflammatory preven-
production or permit plasma proteins to bind tion therapy with corticosteroids has been reported,
them so that neoantigens give rise to aPL. This but remains controversial.17
has been documented for phosphatidylserine (PS).
Despite the presence of an active membrane- ANTI-THYROID ANTIBODIES
associated adenosine triphosphate (ATP)-dependent
aminophospholipid translocase that normally relo- Anti-thyroid autoantibodies (ATA) have been sug-
cates PS from the outer to the inner monolayer, PS gested to be independent markers of ‘at-risk’ preg-
is exteriorized during trophoblast differentiation.13 nancy. Euthyroid women with RSA have increased
When exposed to the blood, PS allows β2GPI to levels of autoantibodies either against thyroglobulin
be immobilized and become antigenic for patho- (TG) or thyroid peroxidase (TPO) (22–37%, vs
genic aPL production. Infusion of aPS into preg- 7–19% in controls), while the probability of abor-
nant mice results in an increased fetal resorption tion in women with ATA has been shown to be
rate and lower mean weights of the placentas and 10–32% versus 3–16% in controls.18–21 The mecha-
fetuses.14 nism whereby ATA affect pregnancy is not known.
167
0415421306-Ch12 3/30/07 9:41 AM Page 168
It is possible that the high rate of miscarriage is and to be most prone to miscarriage in subsequent
related to a very mild thyroid ‘underfunction’, with pregnancies. According to our experience, the
the thyroid gland being less able to adapt to the preliminary results from the 13th International
increased requirements of pregnancy; thus, these Histocompatibility Workshop,25 and data from the
women would benefit from thyroid replacement 14th International Histocompatibility Workshop, an
therapy.20 Furthermore, it has been suggested that HLA-DQA1*0505 sharing is found in couples with
ATA may coexist with activated T cells in the uterus autoimmune RSA rather than those with alloim-
that secret abortogenic cytokines, or that their pres- mune abortions, and could possibly be used as a
ence may reflect an underlying immunological dys- marker for the autoimmune etiology of miscarriages.
function (possibly a T-lymphocyte defect).19 This
suggestion is supported by the coexistence of ATA
with non-organ-specific autoantibodies as well as ALLOIMMUNE ABORTIONS
with increased natural killer (NK) cells in habitual
aborters.21 In terms of this last explanation, treat- The observations that, in some cases of abortion,
ment with intravenous immunoglobulin (IVIG) is the embryo is infiltrated by lymphocytes and the
expected not only to neutralize the antibodies, but lesions found to the placenta resemble the allo-
also to provide the required modulation of immune geneic reactions found in transplanted grafts indi-
functioning. cate that in these cases the embryo is ‘rejected’ by
the mother.26 To assess the mechanisms causing
OTHER AUTOANTIBODIES AND IMMUNE DISTURBANCES such type of abortion (alloimmune abortion), it is
ASSOCIATED WITH RAS necessary to know the nature of the immune
response in normal pregnancy, since it is the distur-
Several ‘non-classical’ aPL (directed against pro- bances in normal pregnancy immunological mech-
thrombin or thromboplastin, or mitochondrial anisms that result in allogeneic antifetal reactions.
antibodies of M5 type) have been observed in
women with recurrent miscarriages, but their clinical THE IMMUNE RESPONSE IN NORMAL PREGNANCY
significance remains unclear.22
Another interesting finding in aborters with The conseptus is a semiallogeneic graft, because it is
ANA or ATA is the increased presence of peripheral produced by the contribution of both the mother
CD19+CD5+ cells,23 which are believed to produce and the father. Although fetal alloantigens encoded
polyvalent antibodies (mainly IgM) that are also by genes inherited from the father should provoke
directed against hormones (estradiol, progesterone, maternal responses and lead to fetal loss, normally
and human chorionic gonadotropin (hCG)) and this does not happen. This natural miracle, known
neurotransmitters (endorphins and serotonin), and as the immunological paradox of pregnancy, is con-
may be responsible for insufficient decidualization sidered to be the result of a particular immune
and decreased blood supply to the endometrium, response of the pregnant woman, and for more
respectively. than 50 years it has been a challenge for reproduc-
Finally, a genetic background may predispose to tive immunologists to attempt to elucidate the
fetal loss in women with autoimmune-mediated underlying immunological mechanisms.
abortions. Beer et al24 have reported an increase in
the HLA-DQA1*0501 allele (currently assigned as FACILITATION REACTION
0505) in women with recurrent pregnancy losses
who are aPL-positive. They have suggested that The first reliable explanation for fetal tolerance was
fetuses compatible with their mothers for this allele the suggestion that in allogeneic reactions such as
are autoimmune-unacceptable to the mother and transplantation and pregnancy, the immune
trigger her to develop aPL when the pregnancy fails response is a bipolar one that can be either harmful
168
0415421306-Ch12 3/30/07 9:41 AM Page 169
or favorable to the target cells expressing alloanti- and the exact factors modulating Th2 shift remain
gens. The harmful effect (rejection reaction observed unclarified. Several studies have examined the sig-
in transplantation) is characterized by cytotoxic nificance of molecules of several antigenic systems,
antibodies and cytotoxic cells that damage the anti- that are expressed on the trophoblast (molecules of
genic target. The enhancing effect (facilitation reac- the major histocompatibility complex (MHC), ery-
tion) is characterized by a predominance of throcyte antigens, complement regulatory proteins,
humoral responses, which may counteract the Fc receptors, various isoenzymes, adhesion mole-
rejection reaction and have a beneficial effect on the cule, R80K protein, etc.), but no specific antigenic-
antigenic target.27 Predominance of this facilitation ity has been proven.31 Nevertheless, specific
reaction over the rejection reaction appears in preg- trophoblastic molecules as well as various proteins
nancy, where enhancing non-complement-fixing produced by the trophoblast appear to modulate
antibodies and suppressor cells favor the acceptance the cytokine pattern towards preferential expression
of the embryo because they prevent complement- of Th2 cytokines. Heat-shock proteins (Hsp),
mediated cell lysis, while they block allogeneic reac- pregnancy-specific β1-glycoprotein, and increased
tions, either by covering the alloantigens or through expression of the non-classical MHC class I HLA-G
the function of an idiotype–anti-idiotype antibody molecule have been suggested to stimulate endome-
network.28 If the coexisting but suppressed rejection trial macrophages for IL-10 production, which
reaction is upregulated, the embryo is rejected. The enhances a Th2 shift.32,33 Decidual cells may also
suggestion that the facilitation reaction prevails over produce high levels of Th2-type cytokines after
the rejection reaction and succeeds in fetal tolerance interacting with trophoblastic CD1d molecules,
has been followed by a plethora of studies that have which present glycolipid antigens to specific cell
focused on the mechanisms mediating this specific populations bearing T- and NK-cell receptors.34
response. Moreover, binding of leukemia inhibitory factor
(LIF), which is produced by decidual cells, to its
Th2-TYPE IMMUNE RESPONSE receptor (LIF-R) on the syncytiotrophoblast may
enhance placental growth and differentiation and a
In 1987, Wegmann29 presented the ‘immunotrophic’ Th2 shift.35 Finally, hCG produced by the tro-
theory, according to which the normal development phoblast induces the production of progesterone by
of the placenta is the result of the influence of the corpus luteum. Through an immunoregulatory
cytokines (placenta immunotrophic cytokines, such protein known as progesterone-induced blocking
as granulocyte–macrophage colony-stimulating factor (PIBF), progesterone may induce the produc-
factor (GM-CSF), transforming growth factor β tion of IL-4 by γδ T lymphocytes and thus enhance
(TGF-β), and interleukin-3 (IL-3)). In 1993, a Th2 response.36
Wegman et al30 suggested that during pregnancy
there is a change of the T-helper 1(Th1)/Th2 equi- CYTOKINE AND HORMONE NETWORK
librium so that Th2-type cytokines (IL-4, IL-5, and
IL-10) predominate over Th1-type cytokines (IL-2 The acceptance of a Th2 shift alone for the mainte-
and interferon γ (IFN-γ)) and benefit the develop- nance of pregnancy must be considered as a simpli-
ing embryo by enhancing placental growth and fication of the cytokine-mediated mechanisms
function as well as by preventing inappropriate enhancing pregnancy at the fetomaternal interface.
anti-trophoblast cytotoxic reactions. It must not be ignored that in the first stages
The important role of cytokines in a fruitful of pregnancy, IFN-γ, a Th1 cytokine, contributes
maternal–fetal symbiosis has been well documented to the vascular development and remodeling of
through the years. However, the trophoblastic anti- uterine spiral arteries required for implantation
genic stimulus, the maternal cells that are stimu- and successful gestation.37 In addition, it must
lated for the initiation of the enhancing response, be remembered that the cytokine network at the
169
0415421306-Ch12 3/30/07 9:41 AM Page 170
fetomaternal interface is extremely complicated and response not only by producing cytokines and
the embryo has been successfully described as growth factors, but also by specific recognition of
‘bathing in a sea of cytokines’.38 Different cell popu- trophoblastic molecules, suppression of cytotoxic
lations are potentially involved in the production reactions, and control of trophoblast invention and
not only of Th2 cytokines but also of Th1 cytokines, NK-cell toxicity.
as well as other cytokines (i.e., IL-12, IL-15, and IL-18), Several specific immunosuppressive and cyto-
chemokines, and growth factors that control the toxicity-blocking mechanisms have been suggested
differentiation and the activation of immune cells to contribute to fetal tolerance.41 Sperm may pro-
locally. A cytokine that controls the shift to a Th1 mote local immunosuppression via prostaglandins,
response (i.e., IL-12) coexists with one enhancing while TGF-β contained in seminal plasma may play
the Th2 response(i.e., IL-10), and these are possibly a critical role in providing the necessary antigenic
controlled by primary regulatory factors on a com- and environmental signals for the production of
petitive basis. This regulatory and competitive role growth factors (GM-CSF) by the uterine epithelium
is attributed to homones (i.e., hCG, progesterone, and the initiation of an appropriate maternal
and relaxin), the secretion of which is induced by immune response to the conceptus if pregnancy is
cytokines at the same time that the hormones achieved.42 The maternal innate immune system,
themselves control the production of cytokines. For which is the first to confront the embryo, actively
example, Th2-type cytokines induce the secretion reacts to the ‘invader’ by developing an inflamma-
of hCG by the trophoblast, which stimulates the tory response, which may enhance conditions for
corpus luteum to produce progesterone. Progesterone tolerance.43 Decidual macrophages, apart from a
enhances the production of Th2 cytokines by com- tendency to express their activation by the produc-
peting with relaxin, which is also produced by the tion of anti-inflammatory rather than proinflamma-
corpus luteum and which enhances the production tory cytokines, appear to have an immunosuppressive
of Th1 cytokines.39 action and a limited antigen-presenting capacity.44
Another protective mechanism involves (at least in
OTHER MECHANISMS ENHANCING FETAL TOLERANCE animals) indoleamine 2,3-deoxygenase (IDO), an
enzyme for tryptophan catabolism. IDO expressed
Although Th2 cytokines characterize the immune by trophoblastic cells may catabolize tryptophan in
response in normal pregnancy, the Th2 shift is just placental immune cells (maternal T cells) and pre-
part of this particular immune response. Many dif- vent them from activating lethal antifetal immune
ferent mechanisms acting locally or at a distance responses.41 Apoptotic mechanisms may also con-
ensure tolerance to the semiallogeneic graft by the tribute to the protection of the embryo. For exam-
maternal natural and adaptive immune defences. ple, the presence of the CD95L (FasL) molecule
Thus, tolerance is modulated by the cumulative (CD95 (Fas) ligand) on trophoblastic cells seems to
effect of preimplantation factors, molecules protect them by inducing apoptosis of activated
expressed on the trophoblast as well as decidual CD95+ T lymphocytes.45 Finally, modulation of
immune cells. Changes in metabolic factors, hor- local placental immunity during pregnancy has
mones, and cytokines during ovulation, coitus, and been ascribed to HLA-G, whose distribution is
fertilization result in local immunosuppression mainly restricted to the placenta. It has been sug-
within the maternal genital tract and prepare the gested that HLA-G is an immunosuppressive mole-
uterus for the implantation of the blastocyst.40 cule inducing apoptosis of activated cytotoxic
Trophoblastic molecules may be specifically recog- T lymphocytes (CTL) and downregulating the prolif-
nized by maternal immune cells as alloantigens or eration of helper T cells. In addition, soluble HLA-G
may act as antigen-presenting molecules or have an molecules may block receptors on CTL and prevent
immunosuppressive/immunomodulatory function. their action on target cells expressing paternally
Decidual immune cells may regulate the immune derived alloantigens.46
170
0415421306-Ch12 3/30/07 9:41 AM Page 171
THE ROLE OF NK CELLS IN THE MAINTENANCE OF concept of an embryo recognition model through
PREGNANCY an ‘NK-cell allorecognition system’. High-affinity
interactions of NK-cell receptors with their ligands
In 1996, an important role was attributed to decid- may provide self-signals to either a cytotoxic NK-
ual NK (dNK) cells in the allorecognition mecha- cell activation (Th1 response) or inhibition of acti-
nisms acting in pregnancy, and it was suggested that vation and protection of the trophoblast (Th2
the concept of ‘the fetus as an allograft’ required response). Which one of the two responses will
redefinition to encompass these cells.47 NK-like predominate depends on the action of the
cells (CD3−CD16−CD56+bright) are the dominant inhibitory receptors, which prevails over the action
decidual cell population from the first stages of of the activating ones. So, if the inhibitory dNK
pregnancy through the first trimester. Due to their receptors recognize their specific ligands on the
increased presence and direct contact with invading trophoblast, they are expected to inhibit dNK
trophoblast, they have been considered as impor- cell activation for trophoblast damage, otherwise
tant for the establishment of normal pregnancy. dNK cells are allowed to develop anti-trophoblast
There is evidence that, coincident with blastocyst activity (reviewed by Varla-Leftherioti43).
implantation and decidualization, uterine NK cells Most studies that have investigated the effect of
become activated and produce IFN-γ, perforin, and dNK-cell receptors in the maintenance of preg-
other molecules, including angiogenetic factors. nancy have specifically focused on the interactions
In this way, they can control trophoblast invasion involving HLA-G molecules, because of their
through their cytotoxic activity; and they also initiate restricted distribution to placental tissues. HLA-G
vessel instability and remodeling of decidual arteries has been shown to be the ligand for at least three
to increase the blood supply to the fetoplacental inhibitory receptors, and the expression of some
unit.37 Furthermore, dNK cells may be involved in HLA-G isoforms has been shown to protect tro-
cytokine-mediated immunoregulation of the mater- phoblastic cells from lysis by activated cytotoxic
nal immune response, producing Th2-type cytokines cell clones.50 Nevertheless, the control of the anti-
and growth factors, which result in placental trophoblast activity of dNK cells is probably the
augmentation and local immunosuppression and result of the cumulative interaction of several
immunomodulation.48,49 receptors on maternal dNK cells, with different
As in any other population of NK cells, the self and non-self class I molecules appearing on
mode of action of dNK cells involves a repertoire of the HLA haplotypes expressed on the trophoblast.
activating and inhibitory receptors belonging to Among the different NK-cell receptor interactions
three main families: the KIR family (killer with their specific counterparts on the trophoblast,
immunoglobulin-like receptors), the C-type lectin the interactions between inhibitory receptors of
family (CD94/NKG) and the immunoglobulin-like the KIR family (inhKIR) and their ligands (HLA-C
transcripts (ILT or LIR). Through their receptors, molecules) appear to be those mainly involved
dNK cells may recognize selected epitopes on HLA in the function of an NK-cell-mediated allorecog-
class I molecules expressed on the invading tro- nition system in pregnancy.51 Given the differences
phoblast. It is interesting that the specific ligands in both the inhKIR repertoire and the HLA-C
for most of the receptors are the non-classical HLA allotypes among unrelated individuals, each preg-
class I molecules G and E as well the classical HLA nancy presents a different combination of maternal
class I antigen C, which are the only HLA mole- inhKIR receptors on dNK cells and self and
cules expressed on the extravillous trophoblast. non-self HLA-C allotypes on the trophoblast.
Moreover, some of the receptors recognizing HLA-G This combination is expected to ensure the appro-
and HLA-C epitopes are selectively expressed on priate receptor–ligand interactions to inhibit
dNK cells. The specific interaction of NK-cell dNK-cell anti-trophoblast activity, thus favoring
receptors with trophoblastic antigens led to the pregnancy.
171
0415421306-Ch12 3/30/07 9:41 AM Page 172
IMMUNOPATHOLOGY OF ALLOIMMUNE ABORTION endometrium to accept the semiallogeneic

In contrast to normal pregnancy, a predominant embryo.
Th1-type response or defective production of Th2-
● In an environment of hormone-dependent
type cytokines appears in spontaneous abortion.52 maternally and fetally derived immunosuppression,
In response to the conceptus or other antigens, specific decidual cells recognize specific molecules
decidual lymphocytes secrete proinflammatory on the trophoblast.
Th1-type cytokines such as IL-2, IFN-γ, and tumor
● Activated decidual cells secrete growth factors
necrosis factor α (TNF-α), which adversely affect (GM-CSF, TGF-β, and IL-3) that enhance placental
the development of the embryo. Fetal rejection growth (immunotrophism).
occurs through immune-induced inflammation
● Specific lymphocytes of the pregnant woman are
(delayed-type hypersensitivity reactions resulting in also activated and secrete anti-inflammatory
lymphocyte infiltration of the trophoblast), tissue cytokines (IL-4, IL-10, and IL-13), so that a Th2-
degradation (cytotoxic reactions resulting in type immune response is developed.
damage of the trophoblast by NK cells and cytotoxic
● Produced antibodies block cytotoxic reactions
antibodies produced by specific subpopulations of that would harm the embryo.
B lymphocytes), and coagulation (upregulation of a
novel prothrombinase fgl2, which results in vasculi- In abortion (Figure 12.2), tolerance enhancing
tis affecting the maternal blood supply to the preimplantation factors may be absent, trophoblas-
implanted embryo).33,48,53 In addition to the Th1- tic antigens may be inappropriately expressed, and
type response, other mechanisms thought to be the recognition of trophoblastic antigens and/or
involved in the response to normal pregnancy have immunoregulatory molecules may be defective.
also been found in abortions (namely, disturbances Th1 cells produce proinflammatory cytokines (IL-2,
in tryptophan catabolism and reduced apoptosis).53 IFN-γ, and TNF-a), which generate a Th1-type
Unfortunately, the specific mechanism causing response. In the absence of blocking factors, cyto-
fetal rejection is as yet undefined, since no relevant toxic antibodies and cytotoxic cells (mainly
single specific mechanism has been recognized as NK cells) damage the trophoblast.
essential for a successful pregnancy. We speculate
FACTORS INDUCING A Th1 RESPONSE IN ABORTION
that the disruption of one or more of the mecha-
nisms leading to tolerance in normal pregnancy
may occur in stress situations and can lead to rejec- Although the immunopathology underlying Th1 pre-
tion. These disturbances may include (a) absence of ponderance in abortion is unknown, it is widely
immunosuppressive proimplantation factors in the accepted that situations such as stress, infection, and
woman’s genital tract, (b) absence of immunode- autoimmunity may cause Th1 cytokine-triggered
pendent specific suppression at the fetomaternal abortions. Stress has been suggested to alter the
interface, and (c) inappropriate expression or defec- endocrine system (corticotrophin-releasing hor-
tive recognition of trophoblastic and immunoregu- mone, adrenocorticotropin, and progesterone),
latory molecules by decidual cells, including which triggers an immune bias towards an aborto-
disturbances in the NK allorecognition system. genic Th1 cytokine profile.54 Infections may also
Alone or in combination, the above disturbances cause Th1 cytokine-triggered abortions. Clark
seem to deregulate the sensitive balance of maternal et al55 have suggested that this kind of abortion
tolerance to the embryo and lead to its ‘rejection’. depends on the availability/presence of bacterial
Figure 12.1 is a diagram of the response in a endotoxins such as lipopolysacharide (LPS).
normal pregnancy: Infectious agents may trigger a Th1 response when
they are recognized by specific decidual T cells bear-
● Preimplantation factors cause local immuno- ing Vδ2 receptors, which when activated secrete
suppression in the genital tract and prepare the abortogenic cytokines. Barakonyi et al56 have shown
172
0415421306-Ch12 3/30/07 9:41 AM Page 173
Recognition
by specific decidual cells IL-4 Th2 > Th1
cytokine
Trophoblastic antigens ? IL-10
?
IL-13
Th2 RESPONSE
Local hormone-
dependent
immunosuppression TGF-β IL-3
GM-CSF
IMMUNOTROPHISM FACILITATION
Growth, maturation REACTION
of the placenta
Blocking
BLOCKAGE antibodies
of cytotoxic
reactions
Figure 12.1 Immunologic mechanisms in normal pregnancy. IL, interleukin; Th, T-helper; TGF-β, transforming growth factor β;
GM-CSF, granulocyte–macrophage colony-stimulating factor.
Defective recognition of trophoblastic

Th1 > Th2 Th1 cells
antigens and immunoregulatory molecules
cytokine
by decidual cells ?
TNF-α
Trophoblastic antigens IL-10 IL-2
IFN-γ
Inappropriate expression?
IL-13
Th1 RESPONSE
Absence of local
immunosuppressive
factors? TGF-β IL-3
GM-CSF
IMMUNOTROPHISM REJECTION
Growth, maturation REACTION
of the placenta
Blocking
antibodies
BLOCKAGE
of cytotoxic Cytotoxic
reactions antibodies
NK action:
and cells
damage of
trophoblast
Figure 12.2 Immunologic mechanisms in abortion.
173
0415421306-Ch12 3/30/07 9:41 AM Page 174
that peripheral blood γδ+ Τ cells from women attack the trophoblast (including activation signals
with RSA preferentially express the Vγ9Vδ2 T-cell provided through interactions of activating
receptor combination. Vγ9Vδ2 T cells are known to NKR–trophoblastic HLA-Cw pairs), are not inhib-
recognize non-peptide organophosphate and alkyl- ited and the embryo is not protected.63,64 Although
amine antigens and eliminate bacteria and para- these suggestions have not been confirmed, recent
sites. We have investigated the bias towards these data from the 14th International Histocompatibility
cells in women with unexplained pregnancy losses Workshop provide evidence on the role of KIR hap-
and have found that the majority had undiagnosed lotypes (activating KIR) in abortion. A high per-
genital tract or even systematic bacterial infections centage of women with RSA of alloimmune
(unpublished data). etiology were found to have a divergence of the
Genetic factors may also predispose to Th1- common AA KIR repertoire to a rather uncommon
triggered abortion. It has been suggested that the repertoire where A KIR haplotypes contain ‘extra’
cytokine balance is determined by maternal genes, activating KIR, while the ratio of inhibitory to
which regulate the response to stress, LPS, and activating KIR was slightly lower than that in
paternally inherited trophoblastic antigens.33,37 fertile women or women with autoimmune abor-
Furthermore, cytokine gene polymorphisms (i.e., tions. It is possible that this imbalance in favor of
TNF and IFN-γ) have been associated with recur- activating KIR gives to these women a high poten-
rent abortion in women with Th1 immunity to the tial for dNK-cell activation, which may damage
trophoblast.57 trophoblast and avert pregnancy.65
ROLE OF NK CELLS IN ABORTION WORKUP IN ALLOIMMUNE ABORTIONS
Decidual NK cells appear to be the main cell popula- The diagnosis of alloimmune abortions is complex, as
tion involved in alloimmune abortion. Under the pathophysiology is far from certain. To approach
the influence of Th1-type cytokines, they are stimu- it, one has to keep in mind the various different
lated to become classical NK cells expressing CD16 immunological mechanisms involved in normal
(CD3−CD16+CD56+), which can damage the tro- pregnancy, and to look for relevant disturbances that
phoblast either directly by releasing cytolytic sub- could damage the trophoblast. Immunological tests
stances or indirectly by producing inflammatory that have been used include the following:
cytokines.58 Clinical studies have demonstrated that
women who tend to abort have increased numbers of ● partner’s HLA typing, since increased HLA
NK cells of the conventional CD3−CD56+CD16+ type sharing between both spouses has been considered
in the uterus,59,60 as well as increased blood NK-cell as a marker of insufficient alloantigenic stimulus
subsets and NK-cell activity, all of which have been for the initiation of the appropriate immune
associated with abortion of chromosomally normal response
embryos.61,62 ● detection of lymphocytotoxic antibodies against
Considering the function of dNK-cell receptors, paternal cells (antipaternal antibodies), the absence
the rejection of the embryo may be the result of a of which may indirectly indicate a maternal under-
defect in the NK allorecognition system. Studies in responsiveness and lack of pregnancy enhancing
RSA couples as well as in cases of sporadic abortion humoral factors in women’s serum
have suggested that aborting women have a limited ● mixed lymphocyte cultures to evaluate whether
repertoire of inhKIR receptors and many of them the maternal antipaternal response is inhibited
lack the appropriate inhKIR to interact with tro- in the presence of maternal serum – an
phoblastic HLA-Cw molecules (lack of maternal indication that the woman’s serum does not
inhKIR–fetal HLA-C epitope matching). Hence, the contain satisfactory levels of antibodies that
triggering signals that the dNK cells may receive to would block antitrophoblast cytotoxic responses.
174
0415421306-Ch12 3/30/07 9:41 AM Page 175
However, none of the above tests has been shown to suppression of NK-cell activity67 and the
be diagnostic. enhancement of a Th2-type immune response.68
In the light of current knowledge, the workup in 2. Passive immunization in the form of intravenous
women suspected to have alloimmune abortion administration of immunoglobulin G (IVIG)
should include tests to detect the Th1 response (immunomodulation for the blockage of
(increased serum levels or intracellular predomi- allogeneic cytotoxic reactions). IVIG appears to
nance of IL-2, IFN-γ, and TNF-α), but the diagnos- suppress NK-cell activity and subsets,69 and it has
tic value of these measurements is also doubtful. In been suggested as the only therapeutic solution for
practice, alloimmune abortions are most often women having increased percentage and activity
diagnosed by the study of peripheral blood NK of peripheral NK cells. Recently, it was shown that,
cells, which, when occurring in increased numbers in women with alloimmune abortion, IVIG
or activity, have been associated with the abortion enhances a shift to a Th2-type response.70
of chromosomally normal embryos. NK-cell levels
have also been used as markers for the selection of In addition to immunotherapy, hormonal support
women for immunotherapy.61,62 Finally, diagnosis (progesterone or hCG supplements) has been used
can be supported by allogeneic-type reactions (vil- to improve the live birth rate in recurrently abort-
litis and increase in CD3−CD16+CD56+ dNK cells) ing women by modulating the balance between Th1
detected by histology and/or immunohistochem- and Th2 cytokines and possibly by preventing inap-
istry of previously aborted placental tissue.60 propriate apoptosis.39
PREVENTIVE IMMUNOTHERAPY IN ALLOIMMUNE

ABORTION EPILOGUE
For the prevention of recurrent alloimmune abor- It is conceivable that the two categories of autoim-
tions, two main types of immunotherapy have been mune- and alloimmune-mediated abortions may
used: not really represent distinct immunological
pathologies. Abortogenic cytokines known to medi-
1. Active immunization of women using paternal or ate alloimmune abortions may also mediate those
third-party lymphocytes. The application of this with autoimmune pathology. The paradigm of mis-
type of immunotherapy started in the early carriages associated with ATA is a good example
1980s, when it was considered that some of combined autoimmune and Th1 disturbances.
unexplained abortions are due to increased HLA As suggested by Gleicher,71 the disturbances in the
sharing between partners. In these cases, it was two types of abortion as well as of other immune-
expected that the administration of lymphocytes mediated reproductive diseases may reflect a mis-
expressing paternal antigens that are inherited direction of a more broadly based immune
by the embryo would reinforce the alloantigenic response (reproductive Immune Failure Syndrome,
stimulus for the initiation of the enhancing RIFS). Nevertheless, at present, the classification of
pregnancy response. According to a series of the immunological disturbances as autoimmune or
studies, including multicenter ones, this therapy alloimmune helps to explain a high percentage of
has given varying results, ranging from very abortions that were previously considered as ‘unex-
good to doubtful.66 Today, its application is plained’, to identify candidates for immune testing,
rather limited. Suggested mechanisms of action and to offer immunological treatments. In autoim-
include the induction of effective presentation of mune abortions – especially those associated with
paternally derived fetal antigens (amplification aPL – the immunopathology is better defined, the
of the alloeantigenic stimulus) and the regulation diagnosis is relatively simple, and the therapy has
of the maternal response mainly through the been widely described and shown to be beneficial.
175
0415421306-Ch12 3/30/07 9:41 AM Page 176
In alloimmune abortions, there are still many ques- 17. Kwak JY, Gilman-Sachs A, Beaman KD, et al. Reproductive outcome
in women with recurrent spontaneous abortions of alloimmune and
tions to be answered concerning the exact underly- autoimmune causes: preconception versus postconception treatment.
ing mechanisms, since, even in normal pregnancy, Am J Obstet Gynecol 1992; 166:1787–95.
many stages of the maternal immune response 18. Stagnaro-Green A, Roman SH, Cobin RH, et al. Detection of at-risk
pregnancy by means of highly sensitive assays for thyroid autoanti-
remain unclarified. Available diagnostic methods bodies. JAMA 1990; 264:1422–5.
can only provide indirect markers of the underlying 19. Pratt DE, Kaberlein G, Dudkiewicz A, et al. The association of
pathology – hence, the results must be interpreted antithyroid antibodies in euthyroid nonpregnant women with recur-
rent first trimester abortions in the next pregnancy. Fertil Steril 1993;
with caution, in order that the ‘diagnosis’ be estab- 60:1001–5.
lished and the appropriate immunointervention 20. Vaquero E, Lazzarin N, De Carolis C, et al. Mild thyroid abnormalities
administered. and recurrent spontaneous abortion: diagnostic and therapeutical
approach. Am J Reprod Immunol 2000; 43:204–8.
21. Marai I, Carp H, Shai S, et al. Autoantibody panel screening in recurrent
REFERENCES
miscarriages. Am J Reprod Immunol 2004; 51:235–40.
22. Sherer Y, Tartakover-Matalon S, Blank M, et al. Multiple autoantibod-
1. Mc Intyre JA, Coulam CB, Faulk WP. recurrent spontaneous abortion. ies associated with autoimmune reproductive failure. J Assist Reprod
Am J Reprod Immunol 1989; 21:100–4. Genet 2003; 20:53–7.
2. Clark DA, Daya S. Trials and tribulation in the treatment of recurrent 23. Beer AE, Kwak JY, Ruiz JE. Immunophenotypic profiles of peripheral
spontaneous abortion. Am J Reprod Immunol 1991; 25:18–24. blood lymphocytes in women with recurrent pregnancy losses and
3. Stern JJ, Coulam CB. Current status of immunologic recurrent in infertile women with multiple failed in vitro fertilization cycles.
pregnancy loss. Curr Opin Obstet Gynecol 1993; 5:252–9. Am J Reprod Immunol 1996; 35:376–82.
4. Matzner W, Chong P, Xu G, et al. Characterization of antiphospho- 24. Beer AE, Kwak JYH, Gilman-Sacks A, et al. New horizons in the eval-
lipid antibodies in women with recurrent spontaneous abortions. uation and treatment of recurrent pregnancy loss. In: Hunt JS, ed.
J Reprod Immunol 1994; 33:31–9. Immunobiology of Reproduction. New York: Serono Symposia USA,
5. Gleicher N, el-Roeiy A. The reproductive autoimmune failure syn- 1994: 316–34.
drome. Am J Obstet Gynecol 1988; 159:223–7. 25. Varla-Leftherioti M. Reproductive immunology component. In:
6. Coulam CB, Branch DW, Clark DA, et al. American Society for Hansen J, Dupont B, eds. HLA 2004: Immunobiology of the Human
Reproductive Immunology: report of the committee for establishing MHC: Proceedings of the 13th IHWC. Seattle: IHWC, 2006.
criteria for diagnosis of reproductive autoimmune syndrome. Am 26. Labarrere CA. Allogeneic recognition and rejection reactions in the
J Reprod Immunol 1999; 41:121–32. placenta. Am J Reprod Immunol 1989; 21:94–9.
7. Coulam CB. The role of antiphospholipid antibodies in reproduction: 27. Voisin GA. Immunological facilitation, a broadening of the concept of
questions answered and raised at the 18th Annual Meeting of the the enhancement phenomenon. Prog Allergy 1971; 15:328–485.
American Society of Reproductive Immunology. Am J Reprod 28. Chaouat G, Voisin GA, Escalier D, et al. Facilitation reaction (enhanc-
Immunol 1999; 41:1–4. ing antibodies and suppressor cells) and rejection reaction (sensitized
8. Kutteh WH. Antiphospholipid antibodies and reproduction. J Reprod cells) from the mother to the paternal antigens of the conceptus.
Immunol 1997; 35:151–71. Clin Exp Immunol 1979; 35:13–24.
9. Branch W. Antiphospholipid antibodies and reproductive outcome: 29. Wegmann TG. Placental immunotrophism: maternal T cells
the current status of affairs. J Reprod Immunol 1998; 38:75–87. enhance placental growth and function. Am J Reprod Immunol 1987;
10. Cullis PR, Hope MJ, Tilcockn CP. Lipid polymorphism and the roles 15: 67–9.
of lipids in membranes. Chem Phys Lipids 1986; 40:127–44. 30. Wegmann TG, Lin H, Guilbert L, et al. Bidirectional cytokine interac-
11. Lockwood CJ, Rand JH. The immunobiology and obstetrical conse- tions in the maternal–fetal relationship: Is successful pregnancy a
quences of antiphospholipid antibodies. Obstet Gynecol Surv 1994; TH2 phenomenon? Immunol Today 1993; 14:353–6.
49:432–41. 31. Clark DA. Immunobiological characterization of the trophoblast–
12. Kuwana M, Matsuura E, Kobayashi K, et al. Binding of β2-glycopro- decidual interface in human pregnancy. In: Kurpisz M, Fernandez N,
tein I to anionic phospholipids facilitates processing and presentation eds. Immunology of Human Reproduction. Oxford: BIOS, 1995:
of a cryptic epitope that activates pathogenic autoreactive T cells. 301–11.
Blood 2005;105:1552–7. 32. Clark D. Editorial. Signaling at the fetomaternal interface. Am J
13. Lyden TW, Vogt E, Ng AK, et al. Monoclonal antiphospholipid Reprod Immunol 1999; 41:169–73.
antibody reactivity against human placental trophoblast. J Reprod 33. Clark DA, Arck PC, Chaouat G. Why did your mother reject you?
Immunol 1992; 22:1–14. Immunogenetic determinants of the response to environmental selec-
14. Blank M, Tincani A, Shoenfeld Y. Induction of experimental tive pressure expressed at the uterine level. Am J Reprod Immunol
antiphospholipid syndrome in naive mice with purified IgG antiphos- 1999; 41:5–22.
phatidylserine antibodies. J Rheumatol 1994; 21:100–4. 34. Boyson JE, Rybalov B, Koopman LA, et al. CD1d and invariant NKT
15. Emoto K, Kobayashi T, Yamaji A, et al. Redistribution of phos- cells at the human maternal–fetal interface. Proc Natl Acad Sci 2002;
phatidylethanolamine at the cleavage furrow of dividing cells during 99:13741–6.
cytokinesis. Proc Natl Acad Sci USA 1996; 93:12867–72. 35. Kojima K, Kanzaki H, Iwai M, et al. Expression of leukaemia
16. Kaider AS, Kaider BD, Janowicz PB, et al. Immunodiagnostic evalua- inhibitory factor (LIF) receptor in human placenta: a possible role for
tion in women with reproductive failure. Am J Reprod Immunol LIF in the growth and differentiation of trophoblasts. Hum Reprod
1999; 42:335–46. 1995; 10:1907–11.
176
0415421306-Ch12 3/30/07 9:41 AM Page 177
36. Szekeres-Bartho J, Barakonyi A, Polgar B, et al. The role of γ/δ T cells 56. Barakonyi A, Polgar B, Szekeres-Bartho J. The role of γ/δ T-cell
in progesterone-mediated immunomodulation during pregnancy. receptor-positive cells in pregnancy: Part II. Am J Reprod Immunol
A review. Am J Reprod Immunol 1999; 42:44–8. 1999; 42:83–7.
37. Ashkar AA, Di Santo JP, Croy BA. Interferon gamma contributes to ini- 57. Daher S, Shulzhenko N, Morgun A, et al. Associations between
tiation of uterine vascular modification, decidual integrity, and uterine cytokine gene polymorohisms and recurrent pregnancy loss. J Reprod
natural killer cell maturation during normal murine pregnancy. Immunol 2003; 58:69–77.
J Exp Med 2000; 92:259–70. 58. King A, Wheeler R, Carter NP, et al. The response of human decidual
38. Chaouat G, Ledee-Bataille N, Dubanchet S, et al. TH1/TH2 paradigm leukocytes to IL-2. Cell Immunol 1992; 141:409–42.
in pregnancy: paradigm lost? Cytokines in pregnancy/early abortion: 59. Vassiliadou N, Bulmer JN. Immunohistochemical evidence for
reexamining the TH1/TH2 paradigm. Int Arch Allergy Immunol increased numbers of ‘classic’ CD57+ natural killer cells in the
2004; 134:93–119. endometrium of women suffering spontaneous early pregnancy loss.
39. Carp H, Torchinsky A, Fein A, et al. Hormones, cytokines and fetal Hum Reprod 1996; 11:1569–74.
anomalies in habitual abortion. Gynecol Endocrinol 2001; 15:472–83. 60. Kwak JY, Beer AE, Kim SH, et al. Immunopathology of the implantation
40. Mellor AL, Munn DH. Immunology at the maternal–fetal interface: site utilizing monoclonal antibodies to natural killer cells in
Lessons for T cell tolerance and suppression. Annu Rev Immunol women with recurrent pregnancy losses. Am J Reprod Immunol 1999;
2000; 18:367–91. 41:91–8.
41. Thellin O, Coumans B, Zorzi W, et al. Tolerance to the foeto–placental 61. Coulam CB, Goodman C, Roussev RG, et al. Systemic CD56+ cells
‘graft’: ten ways to support a child for nine months. Curr Opin can predict pregnancy outcome. Am J Reprod Immunol 1995;
Immunol 2000; 12:731–7. 33:40–6.
42. Robertson SA. Seminal plasma and male factor signalling in the 62. Aoki K, Kajiura S, Matsumoto Y, et al. Preconceptual natural killer cell
female reproductive tract. Cell Tissue Res 2005; 322:43–52. activity as a predictor of miscarriage. Lancet 1995; 345:1340–2.
43. Varla-Leftherioti M. The significance of women’s NK cell receptors’ 63. Varla-Leftherioti M, Spyropoulou-Vlachou M, Niokou D, et al.
repertoire in the maintenance of pregnancy. Chem Immunol Allergy Natural killer (NK) cell receptors’ repertoire in couples with recurrent
2005; 89:84–95. spontaneous abortions. Am J Reprod Immunol 2003; 49:183–91.
44. Mizuno M, Aoki K, Kimbara T. Functions of macrophages in human 64. Varla-Leftherioti M, Spyropoulou-Vlachou M, Keramitsoglou T,
decidual tissue in early pregnancy. Am J Reprod Immunol 1994; et al. Lack of the appropriate natural killer cell inhibitory receptors
3:180–8. in women with spontaneous abortion. Hum Immunol 2005;
45. Zorzi W, Thellin O, Coumans B, et al. Demonstration of the expres- 66:65–71.
sion of CD95 ligand transcript and protein in human placenta. 65. Varla-Leftherioti M, Keramitsoglou T, Spyropoulou-Vlachou M, et al.
Placenta 1998; 19:269–77. Report from the Reproductive Immunology Component at the 14th
46. Le Bouteiller P. HLA-G and local placental immunity. Gynecol Obstet International HLA and Immunogenetics Workshop. Tissue Antigens
Fertil 2003; 31:782–5. 2007; 69 suppl 1:1–7.
47. King A, Loke YW, Chaouat G. NK cells and reproduction. Immunol 66. Coulam CB, Clark DA, Collins J, et al. The Recurrent Miscarriage
Today 1997; 18:64–6. Immunotherapy Trialists. Worldwide collaborative observational study
48. Clark D, Arck PC, Jallili R, et al. Psycho–neuro–cytokine/endocrine and meta-analysis on allogenic leukocyte immunotherapy for recurrent
pathways in immunoregulation during pregnancy. Am J Reprod spontaneous abortion. Am J Reprod Immunol 1994; 32:55–72.
Immunol 1996; 35:330–7. 67. Gafter U, Sredni B, Segal J, et al. Suppressed cell-mediated immunity
49. Chaouat G, Tranchot Diallo J, Volumenie JL, et al. Immune suppres- and monocyte and natural killer cell activity following allogeneic
sion and Th1/Th2 balance in pregnancy revisited: a (very) personal immunization of women with spontaneous recurrent abortion.
tribute to Tom Wegmann. Am J Reprod Immunol 1997; 37:427–34. J Clin Immunol 1997; 17:408–19.
50. Menier C, Riteau B, Dausset J, et al. HLA-G truncated isoforms can 68. Hayakawa S, Karasaki, Suzuki M, et al. Effects of paternal lymphocyte
substitute for HLA-G1 in fetal survival. Hum Immunol 2000; immunization on peripheral TH1/TH2 balance and Tcr Vβ and Vγ
61:1118–25. repertoire usage of patients with recurrent spontaneous abortions.
51. Varla-Leftherioti M. Role of a KIR/HLA-C allorecognition system in Am J Reprod Immunol 2000; 43:107–15.
pregnancy. J Reprod Immunol 2004; 62:19–27. 69. Ruiz JE, Kwak JY, Baum L, et al. Intravenous immunoglobulin inhibits
52. Raghupathy R. TH1-type immunity is incompatible with successful natural killer cell activity in vivo in women with recurrent sponta-
pregnancy. Immunol Today 1997; 18:478–82. neous abortion. Am J Reprod Immunol 1996; 35:370–5.
53. Thellin O, Heinen E. Pregnancy and the immune system: between 70. Graphou O, Chioti A, Pantazi A, et al. Effect of intravenous
tolerance and rejection. Toxicology 2003; 185:179–84. immunoglobulin treatment on the Th1/Th2 balance in women
54. Arck PC. Stress and pregnancy loss: role of immune mediators, hor- with recurrent spontaneous abortions. Am J Reprod Immunol 2003;
mones and neurotransmitters. Am J Reprod Immunol 2001; 46:117–23. 49:21–9.
55. Clark DA, Chaouat G, Gorczynski RM. Thinking outside the box: mech- 71. Gleicher N. Some thoughts on the reproductive autoimmune failure
anisms of environmental selective pressures on the outcome of the syndrome (RAFS) and Th-1 versus Th-2 immune responses. Am J
materno–fetal relationship. Am J Reprod Immunol 2002; 47:275–82. Reprod Immunol 2002; 48:252–4.
177
0415421306-Ch12 3/30/07 9:41 AM Page 178
0415421306-Ch12a 3/30/07 9:41 AM Page 179
DEBATE
12a. Should paternal leukocyte immunization

be used in RPL?
–For
David A Clark
There are two types of arguments put forward to mononuclear leukocytes bearing paternal antigens
support offering paternal leukocyte immunization can counter loss. We know more now about how
to women with recurrent early pregnancy loss.1 The allogeneic leukocytes work. However, our under-
first relies on evidence-based medicine: if there is standing of the mechanism(s) by which a treatment
adequate evidence from controlled clinical trials for may work has no bearing on whether or not it does
efficacy, it is reasonable to offer treatment. The work.1 It follows that the empirical data from
second relies on rationale: if there is a credible clinical trials is crucial. Commonly, these form
mechanism, then treatment should work, and hence the basis for a systematic review, as exemplified
should be offered. Side effects and cost–benefit are by the Cochrane Database. A recent updated
usually also considered in making a positive recom- review by Porter et al6 concluded that none of the
mendation and in the patient’s decision. In 1994, the proposed immunological interventions for recur-
results of the Recurrent Miscarriage Immunotherapy rent pregnancy loss were effective. If one is unfamil-
Trialists Group (RMITG) meta-analysis were pub- iar with the flaws and pitfalls of a systematic review
lished.2 The meta-analysis was conducted by two (as described by Christiansen et al7) and is influ-
independent teams under the aegis of the ASRI enced by publication in a respected venue, one
Ethics Committee. John Collins and Robin Roberts might be persuaded to accept the negative point of
headed the first team; Jim Scott headed the second. view. However, the tenets of evidence-based medi-
Using an intention-to-treat analysis, the first team cine treat authority with justifiable suspicion, and
found that paternal leukocyte immunization there is an established tradition from Galileo and
increased the live birth rate by a statistically signifi- Luther to the effect that we can find the truth our-
cant 10%.2,3 The second team arrived at the same selves if we have access to the data and have a knack
conclusion, in an intention-to-treat analysis, that for the critical thinking required.8 Therefore, the
allogeneic leukocyte immunotherapy significantly paper in Human Reproduction Update5 carefully
improved the live birth rate.2,4 The effect of maternal evaluated the biological factors in the design and
age and number of previous abortions that impact outcome of individual clinical trials.
on probability of a successful next pregnancy were Table 12a.1 compares the meta-analysis of pater-
taken into account.2 nal leukocyte immunotherapy from the recent
As for the rationale behind paternal leuko- Cochrane Database review with my own updated
cyte immunization, a 2001 article5 in Human analysis, also using the Peto method. The data uti-
Reproduction Update5 explained the current lizing only published information (without having
knowledge of mechanisms of spontaneous abor- the raw data describing the individual patients) are
tion, and mechanisms whereby administration of shown in italics. For non-italicized authors, I have
179
0415421306-Ch12a 3/30/07 9:41 AM Page 180
Table 12a.1 Two meta-analyses of paternal leukocyte immunotherapy (LIT) for recurrent unexplained
recurrent miscarriages (Peto method)
Scott analysis (2006) Clark analysis (2006)
Study LIT Placebo OR LIT Placebo OR
Mowbray (London) 23/37 14/30 2.33 25/37 14/30 2.33

Cauchi (Melbourne) 13/20 16/22 0.70 13/20 16/22 0.70
Ho (Taiwan) 33/42 32/49 1.90 33/42 32/49 1.90
Gatenby (Newtown) 13/19 12/20 1.43 13/19 12/20 1.43
Christiansen (Aarlborg) 4/8 16/28 0.75 4/8 15/29 0.94
Daya (Hamilton) 5/10 1/6 3.75
Kilpatrick (Edinburgh) 8/12 6/10 1.32 8/12 6/10 1.32
Reznikoff (Paris) 17/26 14/26 1.60 17/26 14/26 1.60
Illeni (Milan) 10/16c 11/14c 0.48c 10/22 11/22 0.84
Stray Pedersen (Oslo) 24/33 22/31 1.09 24/33 22/31 1.09
Scott (Utah) 6/10 5/12 2.01 6/10 5/12 2.01
Ober (Chicago) 31/68 41/63 0.46 31/68 41/63 0.46
Pandey (Minnesota) 21/25c 6/20c 9.03c 21/32 6/47 10.17
Total patients 210/326 196/331 210/339 195/367
OR and 95% CI 1.27 (0.92–1.74)a 1.41 (1.04–1.92)b
Heterogeneity χ = 24.58, p < 0.01
2
χ = 32.24, p < 0.01
2
OR, odds ratio; CI, confidence interval.

a
As the 95% CI includes 1.0, an OR of 1.27 is not statistically different from 1.0 at 2p ≤ 0.05.
b
OR is significantly different from 1.0 at 2p ≤ 0.028. As the null hypothesis was that LIT does not increase the live birth rate in those to whom it is
given, p ≤ 0.014.
c
Figures in bold differ from intentions to treat figures in Clark’s meta-analysis.
used the numbers from the intention-to-treat Pandey and Agrawal14 has been published fairly
analysis by John Collins and Robin Roberts in the recently, and data have not yet been requested. It can
RMITG meta-analysis. In the dataset in Table 12a.1 be seen that the result of my analysis differs from that
the figures from Illeni et al have been updated, in the Cochrane database, and achieves statistical
based on their subsequently published paper.9 Two significance. I used intention-to-treat data from
of the trials are included even though they are Pandey and Agrawal as had been used for other data,
problematic. The trial by Illeni et al9 included many whereas only the patients who achieved pregnancy
women with a long interval between treatment and in the study by Pandey and Agrawal were included in
pregnancy.4 Mowbray and Underwood10,11 had the Cochrane meta-analysis6 as shown in Table 12a.1.
shown that women lost protection after 3 months Such post hoc selection can lead to misleading
unless they made antipaternal antibodies, and that conclusions.
boosting was required in most women (who do not A major problem with both analyses shown in
make antibodies) when pregnancy did occur. In the Table 12a.1 is a χ2 value for heterogeneity that is less
trial by Cauchi et al, a purportedly ineffective treat- than 0.05. Purists say one cannot combine the data
ment dose was given and the controls had had fewer in a meta-analysis when that happens. One either
prior abortions than the treated patients.2,12 Hence, has to omit the results of Ober et al13 or those of
the two groups were unmatched for the most signif- Pandey and Agrawal.14 There are many problems
icant prognostic factor, the number of previous concerning the trial of Ober et al13 that strongly sug-
miscarriages. For italicized entries, one is relying gest that it should be excluded from the meta-analy-
solely on published data. It was not possible to sis. Scott was involved in the trial by Ober et al.13
obtain the raw data from Stray Pederson or Scott His own data should not have been included in the
et al, or from the trial by Ober et al.13 The trial by Cochrane database meta-analysis, in order to avoid
180
0415421306-Ch12a 3/30/07 9:41 AM Page 181
DEBATE: SHOULD PATERNAL LEUKOCYTE IMMUNIZATION BE USED IN RPL? – FOR
a perception of bias.6 Additionally, the original miscarriage patients in randomized trials excluding
Ober et al13 database has been closed to outside autoantibody-positive patients showed a clear bene-
inspection. Hence, it has been impossible to deter- fit of treatment, and it has been argued that patients
mine if the patients were matched for the results of aborting after treatment are losing karyotype abnor-
autoantibody tests, or if all of the patients had the mal embryos that die from non-immunological
testing done. This is important, because a post mechanisms15,16 (Figure 12a.1). Other problems
hoc non-intention-to-treat analysis by Ober et al13 with the trial by Ober et al13 have been detailed else-
suggested that the abortion rate in patients who where.1,5,17 Ober et al13 stored the purified husband
achieved pregnancy was significantly greater than mononuclear leukocytes at 4°C overnight before
those who received control treatment, although by use, and storage can cause loss of effectiveness due
intention-to-treat analysis, these was no significant to shedding of surface CD200 molecules into the
effect on live birth rate.13 It was known from a logis- supernatant.5,18 Whole blood was used by Unander,19
tic regression analysis performed as part of the in an uncontrolled study of recurrent spontaneous
RMITG2 study that patients with secondary recur- abortion patients. This study also used 2 or more
rent abortion did not benefit from allogeneic leuko- units of blood (which contains more cells than used
cyte treatment, and those with antinuclear antibody by Ober et al13), and 2 or more units of whole blood
or anticardiolipin antibody had a reduced chance is sufficient to cause transfusion-related immuno-
of success.2,3 Indeed, analysis of primary recurrent modulation in humans, and CD200 molecules
1.0
0.9
0.8
0.7
Probability of live birth
0.6
0.5
Treatment
0.4
0.3
0.2
Control
0.1
0
3 4 5 6 7 8 9
Previous number of miscarriages (this partner)
Figure 12a.1 Expected chance of a live birth with leukocyte immunotherapy. Expected chance of a livebirth with (filled circles) or
without (open circles) leukocyte immunotherapy in relationship to previous number of spontaneous abortions with partner in
women with no autoimmunity (antinuclear antibodies or anticardiolipin antibodies) who achieved a pregnancy. Based on seven
randomized (and usually blinded) controlled trials. Total sample of 230 women: of 123 immunized, 111 became pregnant, of 135
controls, 119 became pregnant. The shaded area shows the proportion of losses in the untreated controls that were expected to fail
due to karyotype abnormalities in the embryo. This is an intention-to-treat analysis that included patients who were immunized but
did not achieve pregnancy. (Reproduced from Clark DA et al. Am J Reprod Immunol 1996; 35:495–8,15 with permission from
Blackwell Munksgaard.)
181
0415421306-Ch12a 3/30/07 9:42 AM Page 182
released into the plasma may exert suppression.18 efficacy and safety. The issue of safety was addressed
Pandey and Agrawal,14 however, stored purified in the RMITG2 meta-analysis, and more recently in
blood mononuclear leukocytes at 37°C and a long-term follow-up study on 2687 patients
obtained a strong anti-abortive effect. Small num- immunized in Germany.23 In both series, the inci-
bers of cells were given repeatedly until the women dence of side effects was minimal. However, the
developed antibodies, detected in a blocking assay. FDA does not accept trial data not performed on
Such antibodies, via immunological enhancement, US soil. There have been some attempts to start a
can prolong survival of donor cells that would oth- new US multicenter trial, but there have been a
erwise be rejected.5 We have recently shown20 that number of political roadblocks. The cost of the
human mononuclear leukocytes stored at 4°C lose study by Ober et al13 (estimated as $US2.2. million)
CD200, but that storage at 37°C greatly increases has reduced National Institutes of Health (NIH)
CD200 expression. Thus, there are important bio- funding for basic science research grants, and, given
logical explanations for the different outcomes current low NIH funding rates, the prospect of
shown in Table 12a.1 Focus on the rigor of random- mounting another large multicenter trial is vanish-
ization, blinding, concealment of allocation, analysis ingly small. In my opinion, there are sufficient data
by intention to treat, etc. (i.e., high quality) misses to argue that paternal leukocyte immunotherapy
key biological factors that impact on the result. should be offered to selected subgroups of patients:
Indeed, including biologically flawed data in a primary recurrent spontaneous abortion, no evi-
meta-analysis can swamp out significant effects. dence of autoimmunity, and with karyotyping of
Giving penicillin to 100 patients with fever and a any miscarriages to be sure that only karyotypically
sore throat will show little difference from placebo, normal embryos are being lost. By ‘spontaneous
whereas if 25 patients with such symptoms and recurrent miscarriage’, I refer to patients whose loss
a positive culture for a penicillin-sensitive pathogen process (as detected by serial β human chorionic
are treated, a positive result can be obtained. Of gonadotropin (β-hCG) values) does not begin
course, if the penicillin has been stored so as to render before 6 weeks of gestation.10 Losses at earlier time
it inactive, one should not be surprised if one’s null points raise the specter of autoimmune problems,
belief is confirmed. It should be noted that when where different treatments, (e.g., intravenous
the result of the double-blinded trials in the RMITG immunoglobulin, IVIG) merit consideration.5
register were compared with unblinded and cohort- The Cochrane review6 also concludes that IVIG
controlled trials (not included in Table 12a.1), there is ineffective, but again patient selection (i.e., hetero-
was actually a greater benefit of leukocyte immuno- geneity), timing of treatment, and choice of IVIG
therapy in the double-blinded trials.21 It has also product impact on the outcome – all important
been reported22 for a variety of different types of factors that render the Cochrane analysis suspect.24
treatments of non-fertility problems that the results IVIG, like leukocyte immunotherapy, appears to
of cohort-controlled trials and blinded randomized include a CD200-dependent activity that results in
trails are not significantly different. Tender loving suppression of natural killer (NK) cells.18 Elevated
care (TLC) has been touted as effective for recurrent blood NK-cell activity (even when in the upper end
miscarriage, and the mechanisms of stress-triggered of the so-called ‘normal’ range) is known to be a
loss (which can be prevented by leukocyte immu- predictor of subsequent miscarriage25,26 and ele-
nization) are well worked out in mice; none of the vated levels may reflect an abnormal response to
human studies of TLC has had appropriate controls stress in primary recurrent miscarriage patients.26
or has tested for non-inferiority to immunotherapy Furthermore, peripheral blood NK cells include a
in a randomized controlled trial. heterogeneous group of distinct types of cells.
In the USA the Food and Drug Administration Elevation of the NKT subset of NK cells has been
(FDA) has decided to regulate leukocyte immuno- reported to bias the local cytokine environment
therapy, and requires further clinical trials to show towards a pro-abortive T-helper (Th1) pattern.27
182
0415421306-Ch12a 3/30/07 9:42 AM Page 183
DEBATE: SHOULD PATERNAL LEUKOCYTE IMMUNIZATION BE USED IN RPL? – FOR
Some authors have focused on the problem recurrent unexplained abortion: Immunotherapy for recurrent spon-
of validity of the ‘blocking antibody’ theory as a taneous abortion: Analysis 1. Am J Reprod Immunol 1994; 32:275–80.
4. Scott JR. Immunotherapy for recurrent spontaneous abortion:
basis for disbelieving that paternal leukocyte Analysis 2. Am J Reprod Immunol 1994; 32:279–80.
immunotherapy can work.1 It has been established 5. Clark DA, Coulam CB, Daya S, et al. Unexplained sporadic and recur-
that antibodies can enhance the survival of grafts rent miscarriage in the new millennium: A critical analysis of immune
mechanisms and treatment. Hum Reprod Update 2001; 7:501–11.
of allogeneic cells by ‘immunological enhancement’, 6. Porter TF, LaCoursiere Y, Scott JR. Immunotherapy for recurrent
but Mowbray10 showed that an antibody response miscarriage. Cochrane Database Syst Rev 2006; (2):CD000112.
was not required for an anti-abortion effect – only for 7. Christiansen OB, Nybo Andersen AM, Bosch E, et al. Evidence-based
investigations and treatments of recurrent pregnancy loss. Fertil Steril
prolonged protection after paternal leukocyte 2005; 83:821–39.
immunotherapy. As previously mentioned, immuno- 8. Vamvakas EC, Blajchman MA, eds. Immunomodulatory Effects of
logical enhancement of survival of paternal cells Blood Transfusion. Bethesda, MD: AAAB Press, 1999.
9. Illeni MT, Marelli G, Parazzini F, et al. Immunotherapy and recurrent
bearing CD200 could explain Mowbray’s observa- abortion: a randomized clinical trial. Hum Reprod 1994; 9:1247–9.
tion.1,5 CD200 can have a variety of effects, includ- 10. Mowbray JF. Immunology of early pregnancy. Hum Reprod 1988;
ing stimulation of development of regulatory T cells 3:79–82.
11. Mowbray JF, Underwood JL. Effect of paternal lymphocyte immuniza-
of the type being recently proposed to play a role in tion on birthweight and pregnancy outcome. In: Chaouat G, Mowbray JF,
pregnancy success in mice and humans.28 eds. Cellular and Molecular Biology of the Materno–Fetal Relationship.
Immunization with allogeneic leukocytes can stim- Paris: INSERM/John Libby Eurotext, Colloque 212, 1991:295–302.
12. Clark DA, Daya S. Trials and tribulations in the treatment of recurrent
ulate the development of regulatory T cells.29 spontaneous abortion. Am J Reprod Immunol 1991; 25:18–24.
Antibodies are not required for these processes. 13. Ober C, Karrison T, Odem RR, et al. Mononuclear-cell immunization
A full understanding of the mechanism of action of in prevention of recurrent miscarriages: a randomized trial. Lancet
1999; 354:365–9.
paternal leukocytes and the underlying abortive 14. Pandey MJ, Agrawal S. Induction of MLR-Bf and protection of fetal
processes that may be antagonized5 may improve loss: a current double blind randomized trial of paternal lymphocyte
patient selection and methods of treatment – but immunization for women with recurrent spontaneous abortion. Int
Immunopharmacol 2004; 4:289–98.
this is not a prerequisite for the demonstration of 15. Daya S, Gunby J, and the Recurrent Miscarriage Immunotherapy
efficacy based on the existing evidence. Trialists Group. The effectiveness of allogeneic leukocyte immuniza-
Tissue therapies such as paternal leukocytes tion in unexplained primary recurrent spontaneous abortion. Am J
Reprod Immunol 1994; 32:294–303.
and IVIG may be replaced in the future by treat- 16. Clark DA, Daya S, Coulam CB, et al. Implications of abnormal human
ment with cytokines such as granulocyte– trophoblast karyotype of the evidence-based approach to the under-
macrophage colony-stimulating factor (GM-CSF) standing, investigation, and treatment of recurrent spontaneous
abortion. Am J Reprod Immunol 1996; 35:495–8.
and transforming growth factor β (TGF-β).24,30,31 17. Carp HJA. Immunization with refrigerated paternal cells did not
However, these must be tested in randomized con- prevent recurrent miscarriage. Evidence Based Obstet Gynecol 2000; 2:49.
trolled trials that pay proper attention to the biolog- 18. Clark DA, Chaouat G. Loss of surface CD200 on stored allogeneic
leukocytes may impair anti-abortive effect in vivo. Am J Reprod
ical as well as the methodological factors. Until Immunol 2004; 53:13–20.
then, it remains reasonable for cognoscenti to offer 19. Unander M. In: Beard RW, Sharp, F eds. Early Pregnancy Loss:
paternal leukocyte therapy to appropriately selected Mechanisms and Treatment. London: RCOG/Ashton-under-Lyne,
UK: Peacock Press, 1988: 400–3.
patients. 20. Clark DA, Banwatt D. Altered expression of cell surface CD200 tolerance-
signaling molecule on human PBL stored at 4°C or 37°C correlates with
reduced or increased efficacy in controlled trials of treatment of recur-
REFERENCES rent spontaneous abortion. Am J Reprod Immunol 2006; 55:392–3.
21. Clark DA, Gunby J, Daya S. The use of allogeneic leukocytes or IVIg G
1. Clark DA. Shall we properly re-examine the status of allogeneic therapy for the treatment of patients with recurrent spontaneous abortion.
for recurrent early pregnancy failure? Am J Reprod Immunol 2004; Transfus Med Rev 1997; 11:65–94.
51:7–15. 22. Benson K, Hartz AJ. A comparison of observational studies and
2. Recurrent Miscarriage Immunotherapy Trialists Group. Worldwide randomized controlled trials. N Engl J Med 2000; 342:1878–86.
collaborative observational study and meta-analysis on allogeneic 23. Kling C, Steinmann J, Westphal E, et al. Adverse effects of intradermal
leukocyte immunotherapy for recurrent spontaneous abortion. allogeneic lymphocyte immunotherapy: acute reactions and role of
Am J Reprod Immunol 1994; 32:55–72. autoimmunity. Hum Reprod 2006; 21:429–35.
3. Collins JA, Roberts RM. Reports of independent analyses of data from 24. Clark DA, Coulam CB, Stricker RB. Is intravenous immunoglobulins
the worldwide prospective collaborative study on immunotherapy for (IVIG) efficacious in early pregnancy failure? A critical review and
183
0415421306-Ch12a 3/30/07 9:42 AM Page 184
meta-analysis for patients who fail in vitro fertilization and embryo 29. Kitade H, Kawai M, Rutgeerts O, et al. Early presence of regulatory
transfer (IVF). J Assist Reprod Genet 2006; 23:1–13. cells in transplanted rats rendered tolerant by donor-specific blood
25. Aoki K, Kajijura S, Matsumoto Y, et al. Preconceptional natural-killer-cell transfusion. J Immunol 2005; 175:4963–70.
activity as a predictor of miscarriage. Lancet 1995;345:1340–2. 30. Scarpellini F, Grasso JA, Sbracia M, et al. GM-CSF treatment of
26. Shakhar K, Rosenne E, Lowenthan R, et al. High NK cell activity in women with habitual abortion showing low expression of IL-10 in
recurrent miscarriage: what are we really measuring? Hum Reprod secretory endometria. Am J Reprod Immunol 2005; 53:307–8.
2006; 21:2421–5. 31. Clark DA, Fernandes J, Banwatt D. Prevention of abortion in the CBA
27. Tsuda H, Sakai M, Michimata T, et al. Characterization of NKT cells ×DBA/2 model by intravaginal TGF-β is associated with local recruit-
in human peripheral blood and decidual lymphocytes. Am J Reprod ment of predominantly CD8+ Foxp3+ T cells to the genital tract. Am J
Immunol 2001; 45:295–302. Reprod Immunol 2006; 55:392.
28. Sasaki Y, Sasaki M, Miyazaki S. Decidual and peripheral blood
CD4+CD25+ regulatory T cells in early pregnancy subjects and
spontaneous abortion cases. Mol Hum Reprod 2004; 10:347–53.
184
0415421306-Ch12b 3/30/07 9:44 AM Page 185
DEBATE
12b. Should IVIG treatment be used in RPL?

– For
Carolyn B Coulam
INTRODUCTION Maternal causes of pregnancy loss have been classi-

fied as anatomical, hormonal, immunological, and
Recurrent pregnancy loss (RPL) is a significant clinical thrombophilic.
problem for which there is no known treatment. Immunological mechanisms have recently been
It includes recurrent spontaneous abortion and proposed as a major cause of RPL associated with
recurrent occult preclinical pregnancy or losses. the loss of chromosomally normal pregnancies.
The reason for this lack of effective treatment Forty-five percent of miscarriages and 95% of late
involves the multiple causes of RPL, in that one pregnancy losses from women experiencing RPL
treatment will not treat all causes. Thus, the key to are chromosomally normal. Literature is developing
successful treatment is to determine the cause of the that suggests a role of the immune system in the
RPL. If the cause can be modulated by intravenous majority of these losses.1–3 The end-result of
immunoglobulin (IVIG), then treatment with IVIG the immunological processes that leads to loss of
will be effective. This chapter will discuss the causes the pregnancy involves interference with the blood
of recurrent pregnancy loss, how to identify those supply to the pregnancy. Thrombophilias also
individuals most likely to respond to treatment with lead to vasculopathy of the vessels in and around
IVIG, and published success rates of IVIG therapy. the placenta and to secondary thrombosis, leading
to inadequate perfusion of the intervillous space.4
About 30% of obstetric complications, including
CAUSES OF RECURRENT SPONTANEOUS RPL, are associated with inherited throm-
ABORTIONS bophilia.5,6 An immune trigger may combine
with an underlying thrombophilic tendency
There are two major reasons for RPL. One is that that has been otherwise non-pathogenic to cause
there is something wrong with the pregnancy itself, pregnancy loss.
such as a chromosomal abnormality that prohibits
the pregnancy from implanting or growing properly.
The other reason is a problem in the maternal HOW TO IDENTIFY THOSE INDIVIDUALS
environment in which the pregnancy grows that MOST LIKELY TO RESPOND TO
does not allow an otherwise-normal embryo to TREATMENT WITH IVIG
implant or grow properly. However, if fetal chromo-
somal aberrations explained all of recurrent miscar- Of all of the causes of RPL discussed above, the ones
riage, the probability of three or more miscarriages that would be expected to respond to IVIG treatment
in a row resulting from (accidents) would account would be the etiologies that involve a mechanism
for 5% or less of the observed incidence of losses.1 that can be modulated by IVIG. The mechanisms
185
0415421306-Ch12b 3/30/07 9:44 AM Page 186
by which IVIG is believed to enhance live birth rates chromosomally abnormal pregnancies or anatomi-
include:7 cal, hormonal, or thrombotic risk factors contribut-
ing to their losses. Therefore, to select the person
● IVIG decreases killing activity of natural killer most likely to respond to IVIG treatment would
(NK) cells. require documentation of an immunological risk
● IVIG increases the activity of suppressor T cells. factor and the absence of non-immunological risk
● IVIG suppresses B-cell production of autoantibody. factors. An example of testing included in evalua-
● IVIG contains antibodies to antibodies or anti- tion for RPL would include:
idiotypic antibodies.
● IVIG act on Fc receptors, including binding ● chromosome analysis of previous pregnancy
of complement by the Fc component of losses or both partners
immunoglobulin G (IgG). ● hysterosonogram, hysterosalpingogram, or
hysteroscopy
Based upon these mechanisms, IVIG would be ● blood drawn for antiphospholipid antibodies
expected to enhance live birth rates in individuals (aPL), antinuclear antibodies (ANA), antithyroid
who had elevated circulating NK cells or elevated antibodies (ATA), lupus-like anticoagulant (LA),
NK cytotoxicity, activated T-cell activity, excess of reproductive immunophenotype, NK activation
proinflammatory T-helper 1 (Th1)-type cytokines, assay, and tests for embryotoxins (embryo-
elevated production of autoantibodies that can cause toxicity assay), including Th1 cytokines.
endothelial damage and clotting, and increased activa- ● thrombophilia panel.
tion of complement. Indeed, all of these findings have
been reported among women experiencing RPL.8–14
Proinflammatory cytokines at the maternal–fetal SUCCESS RATES OF IVIG THERAPY
surface can cause clotting of the placental vessels
and subsequent pregnancy loss. One source of these Originally, IVIG therapy was used to treat women
cytokines is the NK cell. Biopsies of the lining of the with postimplantation pregnancy losses who had
uterus from women experiencing RPL reveal an not been successful in pregnancies previously
increase in activated NK cells.15 Peripheral blood treated with aspirin and prednisone or heparin.
NK cells are also elevated in women with RPL com- The rationale for the use of IVIG in the original
pared with women without a history of pregnancy studies was the suppression of LA in a woman being
loss.16 Measurement of NK cells in peripheral blood treated for severe thrombocytopenia. IVIG was
of women with a history of RPL and a repeated fail- often given with prednisone or heparin plus aspirin.
ing pregnancy has shown a significant elevation More recently, IVIG therapy alone has been used to
associated with loss of a normal karyotypic preg- successfully treat women with aPL as well as women
nancy and a normal level associated with loss of who become refractory to conventional autoim-
embryos that are karyotypically abnormal.17,18 mune treatment with heparin or prednisone and
Furthermore, increased NK-cell activity in the aspirin.20 IVIG has been reported to successfully
blood of non-pregnant women is predictive of treat women with elevated circulating levels of NK
recurrence of pregnancy loss.9 T cells with a CD8+ cells, NK-cell killing activity, and embryotoxins,
marker are required for protection against NK with live birth rates between 70% and 80%.21
cytokine-dependent miscarriage.19 IVIG has been IVIG has also been used to treat women with
shown to decrease NK killing activity and enhance unexplained RPL. Ten controlled trials of IVIG for
CD8+ cell activity. Both of these events are necessary treatment of RPL have been published.21–30 Four of
for pregnancy to be successful. these reported significant enhancement in the live
IVIG would not be expected to be effective birth rate with IVIG treatment and six were unable
in enhancing live birth rates in women who had to show benefit of treatment. The number of
186
0415421306-Ch12b 3/30/07 9:44 AM Page 187
DEBATE: SHOULD IVIG TREATMENT BE USED IN RPL? – FOR
patients participating in each trial, the time of first been shown that some brands of IVIG can be as
IVIG administration (pre- or postconception), much as eight times more potent in suppressing NK
whether the patients were selected for treatment cells than others (that were used in ‘negative’
with IVIG based on obstetric history alone or trials).31 Indeed, IVIG can increase the success rate
obstetric history and immunological test results, in in vitro fertilization failure patients based on
and whether the trial showed benefit or no benefit meta-analysis of controlled clinical trials, provided
from treatment are summarized in Table 12b.1. Five treatment is given prior to conception/embryo
trials gave IVIG before conception, and four of transfer, and a value for immunological testing is
these five showed significant benefit in enhancing suggested by higher rates of efficacy in patients
live birth rates, whereas five trials delayed treatment selected for treatment using these tests.31 One
until pregnancy was established, and of these none would not expect IVIG to benefit patients lacking a
demonstrated benefit of treatment (p = 0.04, condition that IVIG could correct, especially if
Fisher’s exact test). Among the trials showing bene- given after the process causing the pregnancy loss
fit of treatment with IVIG, three out of four used had occurred or if the IVIG lacked the necessary
immune test results to select patients for IVIG treat- activity. Pooling flawed randomized clinical trials
ment, and among trials showing no benefit from with adequate randomized clinical trial in a meta-
treatment, none of six selected patients for treat- analysis (such as the Cochrane data32) would be
ment using immune testing (p = 0.03). By waiting expected to block detection of a beneficial effect for
until 5–8 weeks of pregnancy to begin treatment, patients.
women with pathology occurring earlier would
have been excluded and those pregnancies destined
to succeed would be included, leading to selection CONCLUSION
bias. Indeed, a negative correlation with delay in
treatment is significant. Only one study took into IVIG is effective treatment for women experiencing
account the pregnancies lost as a result of chromo- recurrent pregnancy loss if the patients treated
somal abnormalities.23 Approximately 60% of the actually have a condition that IVIG is expected to
pregnancies lost in the clinical trial would be correct and if the IVIG is given in a manner that
expected to have chromosomal abnormalities that would reasonably be expected to correct the
would not be corrected by IVIG. It has also recently underlying pathophysiology.
Table 12b.1 Classification of outcome of controlled trials of IVIG in RPL
Trial N IVIG started Selection Outcome benefit (p<0.05)
Coulam and 63 Preconception Immunological Yes

Goodman et al 21 testing
Coulam et al23 95 Preconception Obstetric history Yes
Kiprov et al26 35 Preconception Immunological Yes
testing
Stricker et al27 47 Preconception Immunological Yes
testing
Stevenson et al25 39 Preconception Obstetric history No
Mueller-Eckhart et al22 64 Postconception Obstetric history No
Christiansen et al24 34 Postconception Obstetric history No
Christiansen et al28 58 Postconception Obstetric history No
Perino et al29 46 Postconception Obstetric history No
Jablonowska et al30 41 Postconception Obstetric history No
187
0415421306-Ch12b 3/30/07 9:44 AM Page 188
REFERENCES 18. Clark DA, Daya S, Coulam CB, et al. Implications of abnormal human
trophoblast karyotype for the evidence-based approach to the under-
1. Clark DA. Is there any evidence for immunologically mediated or standing, investigation, and treatment of recurrent spontaneous
immunologic modifiable early pregnancy failure? J Assist Reprod abortion. Am J Reprod Immunol 1996; 35:495–8.
Genet 2003; 20:62–71. 19. Coulam CB, Roussev RG. Increasing circulating T-cell activation
2. Clark DA, Lea RG, Podor T, et al. Cytokines determining the success markers are linked to subsequent implantation failure after transfer of
or failure of pregnancy. Ann NY Acad Sci 1991; 626:524–36. in vitro fertilized embryos. Am J Reprod Immunol 2003; 50:340–5.
3. Raghupathy R. Th1 type immunity is incompatible with successful 20. Mac Lachlan NA, Letsky E, De Sweit M. The use of intravenous
pregnancy. Immunol Today 1997; 18:478–82. immunoglobulin therapy in the management of antiphospholipid
4. Many A, Schrieber L Rosner S, et al. Pathologic features of the pla- antibody associated pregnancies. Clin Exp Rheumatol 1990; 8:221–4.
centa in women with severe pregnancy complications and throm- 21. Coulam CB, Goodman C. Increased pregnancy rates after IVF/ET
bophilia. Obstet Gynecol 2001; 98:1041–4. with intravenous immunoglobulin treatment in women with elevated
5. Hiller E, Pihusch R. Thrombophilia caused by congenital disorders of circulating CD56+ cells. Early Pregnancy Biol Med 2000; 4:90–8.
blood coagulation. Semin Thromb Hemost 1998; 116:26–32. 22. Mueller-Eckhart G, Mallmann P, Neppert J, et al. Immunogenetic
6. Coulam CB, Jeyendran RS, Fishel LA, et al. Multiple thrombophilic and serological investigations of nonpregnancy and pregnant women
gene mutations rather than specific gene mutations are risk factors with a history of recurrent spontaneous abortion. German RSA/IVIG
for recurrent miscarriage. Am J Reprod Immunol 2006; 55:360–8. Trialist Group. J Reprod Immunol 1994; 27:95–109.
7. Sewell WAC, Jolles S. Immunomodulatory action of intravenous 23. Coulam CB, Krysa LW, Stern JJ, et al. Intravenous immunoglobulin
immunoglobulins. Immunology 2002; 107:387–93. for treatment of recurrent pregnancy loss. Am J Reprod Immunol
8. Coulam CB, Roussev RG. Correlation of NK cell activation and inhibition 1995; 34:333–7.
markers with NK cytotoxicity among women experiencing immuno- 24. Christiansen OB, Pedersen B, Rosgaard A, et al. A randomized,
logical implantation failure after in vitro fertilization and embryo double-blind, placebo controlled trial of intravenous immunoglobu-
transfer. J Assist Reprod Genet 2003; 20:58–62. lin in the prevention of recurrent miscarriage: evidence for a thera-
9. Aoki K, Kajijura S, Matsumoto Y, et al. Preconceptional natural killer peutic effect in women with secondary recurrent miscarriage. Hum
cell activity as a predictor of miscarriage. Lancet 1995; 135:1340–2. Reprod 2002; 17:809–16.
10. Yamada H, Morikawa M, Kato EH, et al. Preconceptional natural 25. Stephenson MD, Dreher K, Houlihan E, et al. Prevention of unex-
killer cell activity and percentage as predictors of biochemical preg- plained recurrent spontaneous abortion using intravenous
nancy and spontaneous abortion with a normal karyotype. Am J immunoglobulin: a prospective, randomized, double-blinded,
Reprod Immunol 2003; 50:351–4. placebo-controlled trial. Am J Reprod Immunol 1998; 39:82–8.
11. Ruiz JE, Kwak JY, Baum L, et al. Intravenous immunoglobulins 26. Kiprov DD, Nachtigall RD, Weaver RC, et al. The use of intravenous
inhibits natural killer activity in vivo in women with recurrent spon- immunoglobulin in recurrent pregnancy loss associated with com-
taneous abortion. Am J Reprod Immunol 1996; 35:370–5. bined alloimmune and autoimmune abnormalities. Am J Reprod
12. Kwak JY, Kwak FM, Ainbinder SW, et al. Elevated peripheral blood Immunol 1996; 36:228–34.
natural killer cells are effectively downregulated by immunoglobulin 27. Stricker RB, Steinleitner A, Bookoff CN, et al. Successful treatment of
G infusion in women with recurrent spontaneous abortions. Am J immunological abortion with low-dose intravenous immunoglobu-
Reprod Immunol 1996; 35:363–9. lin. Fertil Steril 2000; 73:536–40.
13. Ruiz JE, Kwak JY, Baum L, et al. Effects of intravenous immunoglob- 28. Christiansen OB, Mathiesen O, Husth M, et al. Placebo-controlled
ulin G on natural killer cell cytotoxicity in vitro in women with recurrent trial of treatment of unexplained secondary recurrent spontaneous
spontaneous abortion. J Reprod Immunol 1996; 31:125–41. abortions and recurrent late spontaneous abortions with i.v.
14. Graphou O, Chioti A, Pantazi A, et al. Effect of intravenous immunoglobulin. Hum Reprod 1995; 10:2690–5.
immunoglobulins treatment on the Th1/Th2 balance in women with 29. Perino A, Vassiliadis A, Vucetich A, et al. Short-term therapy for recur-
recurrent spontaneous abortions. Am J Reprod Immunol 2003; rent abortion using intravenous immunoglobulins: results of a
49:21–9. double-blind placebo-controlled Italian study. Hum Reprod 1997;
15. Lachapelle MH, Miron P, Hemmings R, et al. Endometrial T, B, and 12:2388–92.
NK cells in patients with recurrent spontaneous abortion. J Immunol 30. Jablonowska B, Selbing A, Palfi M, et al. Prevention of recurrent spon-
1996; 158:4886–91. taneous abortion by intravenous immunoglobulin: a double-blind
16. Yamada H, Morikawa M, Kato EH, et al. Preconceptional natural placebo-controlled study. Hum Reprod 1999; 14:838–41.
killer cell activity and percentage as predictors of biochemical preg- 31. Clark DA, Coulam CB, Stricker RB. Is intravenous immunoglobulins
nancy and spontaneous abortion with a normal karyotype. Am J (IVIG) efficacious in early pregnancy failure? A critical review and
Reprod Immunol 2003; 50:351–4. meta-analysis for patients who fail in vitro fertilization and embryo
17. Coulam CB, Stephenson M, Stern JJ, Clark DA. Immunotherapy for transfer (IVF). J Assist Reprod Genet 2006; 23:1–13.
recurrent pregnancy loss: analysis of results from clinical trials. Am J 32. Porter TF, LaCoursiere Y, Scott JR. Immunotherapy for recurrent
Reprod Immunol 1996; 35:352–9. miscarriage. Cochrane Database Syst Rev. 2006; (2):CD000112.
188
0415421306-Ch12c 3/30/07 9:45 AM Page 189
DEBATE
12c. Should immunotherapy be used in RPL?

– Against
Raj Rai
INTRODUCTION RATIONALE (OR NOT) FOR IMMUNOTHERAPY
The investigation and treatment of women with PATERNAL WHITE CELL IMMUNIZATION
recurrent pregnancy loss (RPL) has historically
been based on anecdotal evidence, personal bias of One concept of an alloimmune basis for RPL is
physicians, and the results of small uncontrolled based on an increased sharing of human leukocyte
studies.1 This has led to the situation where women antigens (HLA) between both partners that pre-
have been subjected to treatments of no proven vents the maternal production of a ‘blocking’ anti-
benefit, some of which have subsequently been body that protects the fetus against immunological
demonstrated to be harmful.2 This is unacceptable. attack.4 Women with successful pregnancies pro-
Indeed, in the current climate in which patient duce this ‘blocking’ antibody and those whose preg-
demands and expectations for a ‘treatment/cure’ of nancy ends in miscarriage do not. Leukocyte
their reproductive failure are ever-increasing, it is immunization has been reported to induce produc-
incumbent upon clinicians to reject previous prac- tion of the ‘blocking’ antibody.5 However, the
tice and embrace an evidence-based approach to ‘blocking antibody’ hypothesis has never been vali-
the management of RPL. dated, and an increased sharing of HLA class I alle-
The concept of immune dysfunction as a basis les between partners has been refuted in a recent
for miscarriage is an attractive one. However, meta-analysis.6 Further, (a) production of a ‘block-
while pregnancy has traditionally been viewed as a ing’ antibody is usually not evident until after
battle between the semiallogenic fetus and the 28 weeks’ gestation and may disappear between
mother, from an evolutionary viewpoint this is not pregnancies;7 (b) miscarriage occurs despite the
advantageous.3 Regardless, immunotherapy has presence of ‘blocking’ antibody;8,9 and (c) women
been introduced into clinical practice as a treat- who exhibit no production of ‘blocking’ antibodies
ment for RPL based on the hypotheses that either do experience successful pregnancies.
alloimunity or autoimmunity is responsible for
pregnancy failure. In order to critically evaluate IVIG
the use of paternal leukocyte immunization (active
immunization) and intravenous immunoglobulin Current concepts on the etiology of RPL focus on
(IVIG, passive immunization) as treatments for autoimmune-mediated pregnancy loss (e.g., antiph-
RPL, it is necessary to examine the rationale for ospholipid syndrome, APS), natural killer (NK)
their use. cells, and a disordered cytokine balance at the
189
0415421306-Ch12c 3/30/07 9:45 AM Page 190
fetomaternal interface. IVIG has a number of A recent fascinating study has cast further doubt
immunomodulatory effects on cytokine production, on the validity of PBNK cell testing in women with
antigen neutralization, Fc receptor blockade, RPL.24 The authors reported that immediately after
alteration in the distribution and function of T-cell insertion of an intravenous cannula for blood with-
subsets, antibodies, and autoantigens that may drawal, women with RPL show an increased pro-
potentially ameliorate a dysregulated immune portion of NK cells within lymphocytes, elevated
response causal of pregnancy loss. However, the role blood NK-cell concentrations, and augmented
of autoantibodies, apart from antiphospholipid NK activity per milliliter of blood compared with
antibodies (aPL), in the pathogenesis of RPL is control women who have no known fertility problems.
unproven.2 The relationship between peripheral However, these differences disappear after 20 minutes,
blood NK (PBNK) cells and reproductive failure is when blood is drawn again from the same cannula.
one of the most controversial fields in reproductive The authors concluded that the elevated NK indices
immunology. This relationship has been examined previously observed in women with RPL are due to
in several small observational studies.10–16 Although a transient increase in NK-cell numbers, rather than
the underlying etiology of reproductive failure a chronic state.
among women in individual studies may well be
different, the reports are consistent in associating
enhanced NK-cell activity with subsequent failure EFFICACY OF IMMUNOTHERAPY
to conceive or miscarriage. Hence, amid much pub-
licity, peripheral blood NK-cell testing is being pro- It is important when evaluating the effect of any
moted as a useful diagnostic test to guide the intervention proposed as a treatment for RPL to be
initiation of a variety of immunosuppressive therapies cognisant of the fact that the two most important
among patients with either RPL or infertility. determinants of the outcome of a particular preg-
There are fundamental flaws in the methodologies nancy are the mother’s age and the number of
used in the published studies. The level and activation miscarriages she has previously experienced. The
of NK cells are dependent on other variables, such rate of sporadic fetal aneuploidy is in the region of
as whether whole blood or fractionated mononu- 50% among women between 40 and 44 years of age,
clear cells are used in the assay, the time of day a rising to 75% among those older than 45 years. On
sample is taken, whether any physical exercise has the basis of a 15% clinical miscarriage rate, 35% of
been performed, the parity of the patient, and women with three consecutive miscarriages will
whether the samples have been previously frozen.17–21 have done so purely by chance alone. Such women
Different NK-cell assays have also been employed, have an excellent chance of a future successful preg-
and results may vary depending on whether the nancy. Among those less than 39 years, a live birth
chromium-51 (51Cr)-release cytotoxicity assay is used rate of 65–70% with supportive care alone can be
or CD69 expression is assayed. Importantly, it is not expected.25 It is against this high spontaneous reso-
known which in vitro assay most accurately reflects in lution rate that the efficacy of any putative treatment
vivo function, and indeed what biological relevance for RPL has to be judged.
such activity has. Furthermore, it is unclear what an
abnormal NK-cell number is. While traditionally a PATERNAL WHITE CELL IMMUNIZATION
peripheral NK-cell level greater than 12% of all
lymphocytes has been regarded as the cut-off A number of studies have examined the efficacy
between a raised and a normal level,22 this figure is of paternal leukocyte immunization as a treatment
well within the normal range (up to 29%) pub- for RPL. These studies, which have used differing
lished by others.23 Hence individuals with entirely methodologies, entry criteria, and analyses,
normal results are being labelled as having raised have reported conflicting results. The largest
NK-cell numbers. study (183 women), which was a double-blinded,
190
0415421306-Ch12c 3/30/07 9:45 AM Page 191
DEBATE: SHOULD IMMUNOTHERAPY BE USED IN RPL? – AGAINST
multicenter, randomized clinical trial, reported improve pregnancy outcome among women
that, on an intention-to-treat basis, the success rate with RPL.
was 36% in the treatment group versus 48% in
the control group (odds ratio (OR) 0.60; 95%
confidence interval (CI) 0.33–1.12).26 If analysis CONCLUSION
was restricted to only those who conceived, the
corresponding success rates were 46% with immu- The lack of scientific rationale for immunotherapy
nization but 65% with placebo saline injections has not stopped its introduction into clinical prac-
(OR 0.45; 95% CI 0.22–0.91), suggesting that tice. However, despite the limitations of meta-
immunization may increase the rate of clinically analyses, the use of either paternal leukocyte
recognized pregnancy loss. Partly on the basis of immunization or IVIG as a treatment for RPL has
this large study and the lack of scientific validity not been shown to be of benefit. The use of these
underlying paternal white cell immunization, the immunomodulatory agents should be resisted.
US Food and Drug Administration (FDA) issued
guidance in 2002 highlighting the lack of efficacy of REFERENCES
this treatment and reminding clinicians that it
should only be offered in the context of thera- 1. Rai R, Clifford K, Regan L. The modern preventative treatment of
recurrent miscarriage. Br J Obstet Gynaecol 1996; 103:106–10.
peutic studies and will require Investigational 2. Rai R, Regan L. Recurrent miscarriage. Lancet 2006; 368:601–11.
New Drug Approval (http://www.fda.gov/CBER/ 3. Parham P. NK cells and trophoblasts: partners in pregnancy. J Exp
ltr/lit013002.htm). Med 2004; 200:951–5.
4. Rocklin RE, Kitzmiller JL, Carpenter CB, et al. Maternal-fetal relation.
The conclusions of several published meta- Absence of an immunologic blocking factor from the serum of
analyses have also been conflicting. The largest and women with chronic abortions. N Engl J Med 1976; 295:1209–13.
most recent – a Cochrane review published in 2006, 5. Takakuwa K, Kanazawa K, Takeuchi S. Production of blocking anti-
bodies by vaccination with husband’s lymphocytes in unexplained
based on 12 trials (641 women) – reported an OR recurrent aborters: the role in successful pregnancy. Am J Reprod
of 1.23 (95% CI 0.89–1.70) among those adminis- Immunol Microbiol 1986; 10:1–9.
tered paternal leukocytes compared with controls.27 6. Beydoun H, Saftlas AF. Association of human leucocyte antigen sharing
with recurrent spontaneous abortions. Tissue Antigens 2005; 65:123–35.
Intention-to-treat analysis did not result in a 7. Regan L, Braude PR, Hill DP. A prospective study of the incidence,
significant difference between treatment and time of appearance and significance of anti-paternal lymphocytotoxic
controls (4 trials; 350 women; OR 1.35; 95% antibodies in human pregnancy. Hum Reprod 1991; 6:294–8.
8. Pena RB, Cadavid AP, Botero JH, et al. The production of MLR-block-
CI 0.89–2.05). ing factors after lymphocyte immunotherapy for RSA does not predict
the outcome of pregnancy. Am J Reprod Immunol 1998; 39:120–4.
IVIG 9. Jablonowska B, Palfi M, Ernerudh J, et al. Blocking antibodies in
blood from patients with recurrent spontaneous abortion in relation
to pregnancy outcome and intravenous immunoglobulin treatment.
Studies using IVIG have used different prepara- Am J Reprod Immunol 2001; 45:226–31.
tions, doses, starting times, frequency, and duration 10. Aoki K, Kajiura S, Matsumoto Y, et al. Preconceptional natural-killer-
cell activity as a predictor of miscarriage. Lancet 1995; 345:1340–2.
of administration. In addition, differing entry crite- 11. Beer AE, Kwak JY, Ruiz JE. Immunophenotypic profiles of peripheral
ria have been used. Some studies have included blood lymphocytes in women with recurrent pregnancy losses and
those with an autoimmune disturbance only, while in infertile women with multiple failed in vitro fertilization cycles.
Am J Reprod Immunol 1996; 35:376–82.
others have included those with ‘unexplained’ RPL. 12. Emmer PM, Nelen WL, Steegers EA, et al. Peripheral natural killer
Hence, at present, the only reasonable basis for cytotoxicity and CD56+CD16+ cells increase during early pregnancy
the assessment of the efficacy of IVIG as a treatment in women with a history of recurrent spontaneous abortion. Hum
Reprod 2000; 15:1163–9.
for RPL would be to examine the results of meta- 13. Fukui A, Fujii S, Yamaguchi E, et al. Natural killer cell subpopulations
analyses. The Cochrane review27 reports that, irre- and cytotoxicity for infertile patients undergoing in vitro fertilization.
spective of whether analysis is performed on an Am J Reprod Immunol 1999; 41:413–22.
14. Ntrivalas EI, Kwak-Kim JY, Gilman-Sachs A, et al. Status of peripheral
intention-to-treat basis (OR 1.18; 95% CI 0.72–1.93) blood natural killer cells in women with recurrent spontaneous abortions
or not (OR 0.98; 95% CI 0.61–1.58), IVIG does not and infertility of unknown aetiology. Hum Reprod 2001; 16:855–61.
191
0415421306-Ch12c 3/30/07 9:45 AM Page 192
15. Putowski L, Darmochwal-Kolarz D, Rolinski J, et al. The immunolog- Correlations in recovery of K- and NK-cell functions, surface markers,
ical profile of infertile women after repeated IVF failure (preliminary and morphology. J Clin Immunol 1982; 2:214–21.
study). Eur J Obstet Gynecol Reprod Biol 2004; 112:192–6. 22. Beer AE, Kwak JY, Ruiz JE. Immunophenotypic profiles of peripheral
16. Yamada H, Morikawa M, Kato EH, et al. Pre-conceptional natural blood lymphocytes in women with recurrent pregnancy losses and
killer cell activity and percentage as predictors of biochemical in infertile women with multiple failed in vitro fertilization cycles.
pregnancy and spontaneous abortion with normal chromosome Am J Reprod Immunol 1996; 35:376–82.
karyotype. Am J Reprod Immunol 2003; 50:351–4. 23. Eidukaite A, Siaurys A, Tamosiunas V. Differential expression of
17. Pross HF, Maroun JA. The standardization of NK cell assays for use in KIR/NKAT2 and CD94 molecules on decidual and peripheral blood
studies of biological response modifiers. J Immunol Meth 1984; CD56bright and CD56dim natural killer cell subsets. Fertil Steril 2004;
68:235–49. 81 (Suppl 1):863–8.
18. Plackett TP, Boehmer ED, Faunce DE, et al. Aging and innate immune 24. Shakhar K, Rosenne E, Loewenthal R, et al. High NK cell activity in
cells. J Leukoc Biol 2004; 76:291–9. recurrent miscarriage: What are we really measuring? Hum Reprod
19. Reichert T, DeBruyere M, Deneys V, et al. Lymphocyte subset reference 2006; 21:2421–5.
ranges in adult Caucasians. Clin Immunol Immunopathol 1991; 25. Clifford K, Rai R, Regan L. Future pregnancy outcome in unexplained
60:190–208. recurrent first trimester miscarriage. Hum Reprod 1997; 12:387–9.
20. Porzsolt F, Gaus W, Heimpel H. The evaluation of serial measure- 26. Ober C, Karrison T, Odem RR, et al. Mononuclear-cell immunisation
ments of the NK cell activity in man. Immunobiology 1983; in prevention of recurrent miscarriages: a randomised trial. Lancet
165:475–84. 1999; 354:365–9.
21. Strong DM, Ortaldo JR, Pandolfi F, et al. Cryopreservation of human 27. Porter TF, LaCoursiere Y, Scott JR. Immunotherapy for recurrent
mononuclear cells for quality control in clinical immunology. I. miscarriage. Cochrane Database Syst Rev 2006; (2):CD000112.
192
0415421306-Ch13 3/30/07 9:46 AM Page 193
13. Infections and recurrent pregnancy loss

David Alan Viniker
INTRODUCTION LISTERIOSIS
Any acute severe infection can be associated with Romana et al5 have put forward the case that latent
occasional pregnancy loss. The role of infection in listeriosis may cause recurrent miscarriage, as
recurrent pregnancy loss (RPL) has been unclear.1 anti-listeric antibodies have been detected by direct
In recent years, there has been increasing interest in immunofluorescence studies. Romana et al5 investi-
microorganisms as possible causes of pathology in gated 309 women: 207 had a total of 334 miscar-
previously unexplained medical conditions.2 In this riages, 67 delivered prematurely, 75 had stillbirths,
chapter, the evidence for and against infection as a and 43 had malformed living or stillborn infants.
cause of RPL is presented. Specific infections are Treatment resulted in the birth of 152 normal
discussed, and the recent developments in molec- babies, all negative on immunofluorescence for
ular biology are discussed as they relate to future anti-listeric antibodies. Manganiello and Yearke6
investigation. attempted to isolate Listeria monocytogenes from
the cervix and endometrium of patients present-
ing with a history of two or more fetal losses.
TUBERCULOSIS Endometrial tissue and endocervical swabs were
cultured. During a 10-year study period, none of
Chronic infection, notably tuberculosis, is more the patients with recurrent fetal loss were found to
commonly related to infertility, with only sporadic harbor the organism in their genital tract. Hence,
reports of pregnancy loss.3,4 Saracoglu et al3 diag- L. monocytogenes could account for occasional fetal
nosed 72 patients with pelvic tuberculosis from loss, but not on a recurring basis. Manganiello
1979 to 1989. The most common presentations and Yearke6 concluded that routinely culturing for
were infertility (47.2%), pelvic or abdominal pain L. monocytogenes in patients with recurrent miscar-
(32%), and abnormal bleeding (11%). There was riage is not warranted.
one case of RPL. Physical examination was normal
in 32% of the patients and chest X-ray was normal
in 81%. The most common site of infection was TORCH INFECTIONS
the fallopian tubes, with occlusion in 32 of the
34 patients having hysterosalpingography. There have been conflicting opinions on the role of
In a series of 25 cases of genital tuberculosis, cytomegalovirus (CMV) in recurrent miscarriage.
21 presented with infertility, 3 had postmenopausal Szkaradkiewicz et al7 found significantly elevated
bleeding, and 1 was admitted with an acute immunoglobulin G (IgG) in most of 11 women on
abdomen.4 Two women subsequently conceived, the first day after a second consecutive trimester
but both aborted. miscarriage. The control group were 15 women in
193
0415421306-Ch13 3/30/07 9:46 AM Page 194
the second trimester of a normal pregnancy. They and chromosome abnormalities were excluded from
concluded that in the majority of the studied women, the study by Kishore et al.11 Serum anti-Chlamydia
reactivation of chronic CMV infection occurred. trachomatis IgM positivity was found in 46.5% of
Cook et al8 used the polymerase chain reaction (PCR) 47 patients with RPL miscarriages, compared with
to detect CMV in gestational tissue of women with 13.8% of 29 age-matched controls of normal preg-
recurrent miscarriage. DNA was extracted from nant women (p <0.001). The relationship between
25 samples of gestational tissue from 21 women with high-titer IgG antibodies to C. trachomatis and
at least three unexplained spontaneous miscarriages. recurrent miscarriage has been investigated.12 It was
None of these specimens contained evidence of CMV found that 7 (41%) of 17 women with three miscar-
DNA, demonstrating that CMV is not a common riages and 6 (60%) of 10 women with four miscar-
direct cause of recurrent miscarriage. riages had anti-chlamydial antibodies, compared
Screening for TORCH infections (toxoplasmo- with 20 (14%) of 148 women with no miscarriages,
sis, rubella, CMV, and herpes simplex virus (HSV)) 6 (13%) of 47 women with one miscarriage, and
is unhelpful in the investigation of recurrent mis- 4 (12%) of 33 women with two miscarriages.
carriage. While these infections can be associated The incidence of three or more miscarriages was
with an individual pregnancy loss, they are illnesses 31.8% for women with high-titer IgG, compared
generally contracted once and therefore should not with 7.5% among women who were seronegative
result in RPL. The current recommendation is that (p <0.001). It was concluded that high-titer IgG
TORCH screening in the investigation of RPL to C. trachomatis was associated with recurrent mis-
should be abandoned.1 carriage and it was suggested that the mechanism
might involve reactivation of latent chlamydial
infection, endometrial damage from previous infec-
CHLAMYDIAE tion, or an immune response to an epitope shared
by chlamydial and fetal antigens.
Mezinova et al9 have reported that chlamydiae have In contrast, a study of 101 women with RPL
been found in 41.7% of 163 women with habitual did not find an association with chlamydiae.13
miscarriage in their series. The miscarriage rate was Screening involved direct examination, culture, and
59.1% in the presence of chlamydiae. However, all serological testing. The culture-positive and serol-
women treated for chlamydial infection went on to ogy-positive rates of 15% and 35% did not differ
deliver at term. It was concluded that women with from those of other unselected populations. The
habitual miscarriage and chlamydiae should receive time from last miscarriage or type of miscarriage
appropriate therapy. Endometrial, endocervical, was unrelated to C. trachomatis infection. The uns-
and urethral specimens have been obtained from elected population rates for chlamydiae in this
16 non-pregnant women with a history of RPL.10 study were noticeably higher than generally
Chlamydiae were isolated from the endometria of expected. Others have also failed to find an associa-
five women. No chlamydiae were isolated from the tion between IgG anti-chlamydial antibodies and
cervix or urethra of two patients with proven RPL.14–16 Rae et al14 looked at IgG to chlamydiae in
endometrial involvement. This study demonstrated 106 women with unexplained RPL and compared
that eradicating intrauterine chlamydial infection their findings with sera from a general antenatal
before pregnancy improved pregnancy outcome in population of 3890. Twenty-six (24.5%) women
women with RPL. It was suggested that asympto- with RPL had positive serology compared with
matic chlamydial infection might have an adverse 788 (20.3%) of controls. Anti-chlamydial antibody
effect on placentation. seropositivity did not correlate with subsequent
An association between positive chlamydia serol- pregnancy outcome. It was concluded that there is no
ogy and RPL has been reported by Kishore et al.11 association between IgG antibodies to C. trachomatis
Rhesus incompatibility and anatomical, endocrine, and recurrent spontaneous abortion.
194
0415421306-Ch13 3/30/07 9:46 AM Page 195
INFECTIONS AND RECURRENT PREGNANCY LOSS
In a prospective study,15 70 patients with RPL In gram-stained smears of vaginal fluid, BV is

attending a specialist recurrent miscarriage clinic diagnosed when three out of four of Amsel’s crite-
were selected. The controls were 40 normal pregnant ria18 are present: namely the presence of clue cells
women and 94 asymptomatic sexually active women. (vaginal epithelial cells heavily coated with bacilli
There was no statistical difference in the frequencies on wet preparation microscopy), vaginal pH>4.5;
of anti-chlamydial IgG or IgA antibodies between a homogenous discharge; and a strong fishy odor,
women with recurrent miscarriage and controls. In which may be amplified on adding alkali to the
another study16 of 504 patients with a history of two vaginal fluid. The organisms most often associated
or more consecutive first-trimester miscarriages, the with BV are Gardnerella vaginalis, Mycoplasma
presence of IgA and IgG antibodies to C. trachomatis hominis, Ureaplasma urealyticum, Mobiluncus
did not influence subsequent pregnancy outcome. spp., Prevotella, Porphyromonas, Bacteroides, and
There does not appear to be robust evidence to Peptostreptococcus spp. It has been suggested that
support serological investigations for chlamydiae as the clinical manifestations of BV depend on a
part of the routine investigation of RPL, but direct synergistic interaction of a variety of microorgan-
swab tests from the cervix may be taken and posi- isms. The gram-negative organisms, including
tive results treated appropriately. Bacteroides, are sensitive to metronidazole, whereas
Mycoplasma, Ureaplasma, and Mobiluncus are sen-
sitive to macrolides such as erythromycin and to the
SYPHILIS tetracyclines.
In a study by Wilson et al,19 749 consecutive
In some parts of Africa, the incidence of syphilis women undergoing in vitro fertilization had a
seroreactivity in pregnant women is at least 10%, vaginal smear taken at the time of egg collection in
and this is associated with spontaneous abortion, a study comparing the prevalence of BV according
perinatal mortality, or a viable infant with congeni- to causation of infertility.1 The smears were gram-
tal syphilis.17 Screening for syphilis should be stained and graded as normal, intermediate, or BV.
considered in at-risk populations. The smears were normal in 63.6%, intermediate in
12.1%, and BV in 24.3%. The rates of BV were
36.4% in tubal factor, 15.6% in male factor, 33.3%
BACTERIAL VAGINOSIS in anovulation, 12.5% in endometriosis, and 18.9%
in unexplained infertility. Women with tubal infer-
In women of reproductive age, lactobacilli are nor- tility were three times more likely to have BV than
mally the predominant bacteria in the vagina. women with male factor infertility, endometriosis,
Lactobacilli are responsible for reducing the vaginal or unexplained infertility. Women with anovulation
pH by metabolizing glycogen from squamous cells were also three times more likely to have BV com-
to lactic acid. The resulting acidic milieu provides pared with women with endometriosis or male factor
protection against infection. Bacterial vaginosis infertility, which the authors suggested supports the
(BV) has become the adopted nomenclature to theory that there is a hormonal influence on vaginal
describe a clinical condition characterized by an flora.
overgrowth of predominantly anaerobic bacteria The detrimental effects on pregnancy associated
within the vagina and a concomitant reduction or with BV may be due to the bacteria ascending
absence of lactobacilli. BV is recognized as the most into the uterus.20,21 One hypothesis suggests that
common cause of vaginal discharge. The discharge microorganisms, possibly those associated with BV,
tends to be malodorous, particularly after sexual may surreptitiously inhabit the uterine cavity (bac-
intercourse. A remarkable feature of BV is the absence teria endometrialis), where they are the culprits
of a host reaction – hence the suffix ‘osis’ rather than behind some common gynecological and obstetric
‘itis’, as signs of inflammation are absent. enigmas.21 Relatively little has been written about
195
0415421306-Ch13 3/30/07 9:46 AM Page 196
one or more microorganisms in the uterus, mostly

Table 13.1 A comparison of the bacteriology
of the vagina and the uterine cavity G. vaginalis, Enterobacter, or Streptococcus agalac-
tiae. The samples were not tested for mycoplasmas
Vagina Uterine cavity or Mobiluncus.
The prevalence of BV in pregnancy varies
• Rich in microorganisms • Relatively sterile
• High vaginal swabs: routine • Samples mainly confined from 9% to 23%.24 Coitus during pregnancy is not
clinical investigation to research centers related to BV or premature delivery. Pregnant
• BV: diagnosis by microscopy • Wet preparations and gram
women do not commonly develop BV after 16 weeks’
of wet preparation or gram stain stain not studied
• Culture unhelpful for BV • Cultures only. Specialist gestation. If present at 16 weeks, BV spontaneously
diagnosis centers required for remits in approximately 30–50% of those reaching
Mycoplasma hominis and
term.25,26
Ureaplasma urealyticum
• Possible marker for bacteria • Microorganisms can be
endometrialis present even with negative PREMATURE DELIVERY
cervical cultures
• BV: no clinical inflammation • Microorganisms can occur
with negative clinical There is a substantial body of evidence indicating
examination that BV is associated with premature delivery.24,27–32
BV, bacterial vaginosis. The association with premature delivery may have
implications for miscarriage, and raise the possibil-
ity that antibiotics may reduce this complication.
BV diagnosed in early pregnancy is particularly sig-
bacterial colonization of the endometrial cavity. nificant,27 as the presence of BV in early pregnancy
Bacteriological investigations of the vagina and is associated with a two- to threefold increased risk
endometrial cavity are compared in Table 13.1. The of preterm labor. Women who have abnormal
healthy vagina is rich in microorganisms, whereas vaginal flora that spontaneously return to normal
the endometrial cavity is considered to be relatively and who are not treated have as many abnormal
sterile. Many microorganisms colonize the vagina outcomes as those treated with placebo, suggesting
without necessarily being pathogenic. High vaginal that the damage occurs in early pregnancy28 or that
swabs are frequently obtained in routine clinical prac- the responsible microorganisms have ascended into
tice, whereas the bacteriology of the endometrium the uterine cavity. In order to determine whether
has been studied almost entirely in research proj- abnormal vaginal microflora are associated with pre-
ects. The bacteriological diagnosis of BV is depend- mature labor, a study was conducted by McDonald
ent on microscopic assessment of a wet preparation et al29 in Australia. The assessment included cultures
of the discharge or a gram stain rather than culture, for aerobic and anerobic bacteria, yeasts, genital
whereas investigation of the uterine cavity has mycoplasmas, and G. vaginalis. The results of 428
depended on culture alone. The clinical significance women in preterm labor were compared with those
of the varied patterns of bacteria found in the of 568 women in labour at term. Two distinct bac-
vagina remains controversial. Our knowledge about teriological groupings were associated with preterm
intrauterine microorganisms is comparatively spar- labor: the BV group of organisms and a group of
tan and more difficult to interpret. Pathogenic enteropharyngeal organisms. G. vaginalis was found
microorganisms can be found in the endometrial in 12% of women in preterm labour, compared
cavity without evidence of pelvic infection being with 6% at term. The prevalence of G. vaginalis was
visible at laparoscopy and with negative cervical even higher (17%) in women in preterm labour
cultures.22 The bacteriology of the endometrial at less than 34 weeks’ gestation. In an analysis
cavity has been investigated23 immediately after of 12 937 women screened for BV, the odds ratio
hysterectomy in 99 women. Nearly a quarter of all (OR) for preterm birth (<37 weeks’ gestation)
the patients in this study by Moller et al23 harbored for asymptomatic BV-positive versus BV-negative
196
0415421306-Ch13 3/30/07 9:46 AM Page 197
women ranged from 1.1 to 1.6 and did not vary the placebo group. Borisov et al34 compared the
significantly with the gestational age at the time of effect of intravaginal clindamycin with the effect
screening.30 of metronidazole in 128 pregnant women with BV.
Vaginal fluid was collected for gram staining BV was eradicated in 93% of the women using
from 354 women admitted in preterm labour with intravaginal clindamycin and in 87% of the group
intact membranes between 24 and 34 weeks’ gesta- receiving the metronidazole. Both treatments were
tion in a prospective blinded study in Paris.33 more effective than oral ampicillin for 7 days, which
Normal flora were found in 254 women of the had a cure rate of 62%.
354 women tested (72.3%). Intermediate changes Ugwumadu et al26 concluded that as previous
were found in 76 (21.7%) and BV in 24 (6.8%). research had shown that spontaneous resolution of
Women with normal, intermediate, and abnormal BV does not modify the risk of preterm birth, early
flora had 27 (10.6%), 14 (18.4%), and 6 (25.0%) screening and treatment should be advocated.
births before 33 weeks, respectively. A history of
spontaneous miscarriage after 14 weeks was the EFFECT OF ANTIBIOTICS IN WOMEN WITH PREVIOUS
only risk factor associated with BV. Preterm delivery PREMATURE LABOR
before 33 weeks was significantly associated with
the flora grade (p = 0.02). It was concluded that the The effect of antibiotics has been assessed in
frequency of BV and its association with preterm women at high risk of premature labor with BV.
delivery are variable and should be interpreted The results of two studies indicate that antibiotics
differently for different populations. Although may reduce the incidence of premature labor. Both
an association was found between BV and delivery metronidazole and metronidazole together with
before 33 weeks, the authors considered the predic- erythromycin have been assessed in double-blinded
tive value of BV to be disappointing, and the useful- placebo-controlled trials.
ness of testing for BV in women with premature In the study by Morales et al35 women with
labor was not demonstrated. premature labor or premature rupture of the mem-
branes in the preceding pregnancy were screened
TREATMENT AND PREMATURE DELIVERY for BV between 13 and 20 weeks’ gestation.
Patients with a positive screen were randomized to
EFFECT OF ANTIBIOTICS ON THE PREVALENCE OF BV receive metronidazole orally or placebo. Forty-four
patients received metronidazole and 36 received
Antibiotics have been shown to affect the presence placebo. The metronidazole group had fewer
of BV. Clindamycin has been shown to be effective hospital admissions for preterm labor (27% vs 78%),
in eradicating the bacteria, whether used intravagi- preterm births (18% vs 39%), low-birth-weight
nally34 or orally.26 Abnormal flora were found after infants (<2500 g: 14% vs 33%), and premature rup-
oral clindamycin in 10% of treated patients com- ture of the membranes (5% vs 33%).
pared with 93% of placebo patients (p <0.001) in Hauth et al36 performed BV testing at 23 weeks
the trail conducted by Ugwumadu et al26 involving in 624 pregnant women at risk of delivering prema-
462 women (231 in the clindamycin group and turely. Patients were randomized on a 2:1 basis to
231 in a placebo group). Normal flora were main- receive treatment with erythromycin and metron-
tained in two-thirds of women throughout preg- idazole (n = 426) or placebo (n = 190). A second
nancy. The results of four weekly smears were course of treatment was instituted for those women
compared in 135 women: 69 clindamycin-treated who still had BV at 28 weeks. In the antibiotic group,
and 66 placebo-treated. For the clindamycin 110 women delivered prematurely (26%), com-
group, the prevalence of abnormal flora was 15% at pared with 68 women in the placebo group (36%;
20 weeks’ and 17% at 36 weeks’ gestation, compared p = 0.01). The association between treatment and
with 69% at 20 weeks’ and 43% at 36 weeks’ in lower rates of prematurity was observed only
197
0415421306-Ch13 3/30/07 9:46 AM Page 198
among the 258 women who had BV (31% with (3/50) in the placebo group. Premature delivery
treatment vs 49% with placebo; p = 0.006). occurred in 20.7% (6/29) in those in whom BV
persisted, compared with 0% (0/26) where BV was
EFFECT OF ANTIBIOTICS WHEN BV successfully treated. Hence, it is not sufficient to
IS AN INCIDENTAL FINDING treat BV – the bacteria must be eradicated.
Intravaginal clindamycin was also assessed by
The effects of both metronidazole and clindamycin Rosenstein et al.28 Thirty-four women had normal
have been assessed on premature labour in random- vaginal flora at their first antenatal clinic visit, com-
ized placebo studies and meta-analyses in low-risk pared with 268 women who had abnormal vaginal
patients in whom BV was an incidental finding. flora. Follow-up assessed for pregnancy outcome,
McDonald et al37 reported a trial of metronida- vaginal flora, and detection of M. hominis and
zole in women with a heavy growth of G. vaginalis U. urealyticum after treatment. There were no sig-
or a gram stain indicative of BV at 19 weeks in nificantly different outcomes in pregnancy between
879 women. Metronidazole was administered at the treated and placebo groups. Women with grade
24 weeks and at 29 weeks if G. vaginalis persisted. III flora responded better to clindamycin than
There was no difference in overall preterm births women with grade II flora by number of abnormal
between metronidazole and placebo groups. In a outcomes (p = 0.03) and return to normal vaginal
subset of 46 women with a previous preterm flora. Women whose abnormal vaginal flora had
birth, metronidazole showed a significant reduction spontaneously returned to normal and who were
in spontaneous preterm birth: 2/22 (9.1%), versus therefore not treated had as many abnormal out-
10/24 (41.7%) in placebo-treated patients. In this comes as those receiving placebo, suggesting that
study, antibiotics were most effective when there damage by abnormal bacterial species occurred
was a history of previous premature labor. early in pregnancy.
In the study by Camargo et al38 of 785 low-risk The results of the study by Lamont et al31 do
Brazilian pregnant women, 134 women with BV not concur with those of Kurkinen-Raty et al39
were treated with metronidazole, tinidazole, or sec- or Rosenstein et al.28 In the randomized double-
nidazole; 71 women with BV received no treatment. blinded study by Lamont et al,31 409 women
Premature delivery occurred in 5.5% of women with abnormal genital tract flora on gram stain at
without BV, in 22.5% of women with untreated 13–20 weeks’ gestation received clindamycin
BV, and in 3.7% of treated women. Perinatal com- vaginal cream or placebo. Those who still had
plications were significantly higher in those abnormal vaginal flora 3 weeks later received a
women with untreated BV. The risk ratios were subsequent course of the original treatment. There
7.5 for premature rupture of the membranes was a statistically significant reduction in the inci-
3.4 for preterm labor 3, 6.0 for preterm birth, and dence of preterm birth in the clindamycin group
4.2 for low birthweight. (4%) compared with placebo (10%) (p < 0.03).
Mothers with singleton pregnancies and no It was concluded that clindamycin vaginal cream
history of preterm delivery in whom BV was diag- administered to women with abnormal vaginal flora
nosed by gram stain at 12 weeks’ gestation were ran- before 20 weeks’ gestation can decrease preterm
domized to receive vaginal clindamycin or placebo birth by 60% and reduce the need for neonatal
in the study by Kurkinen-Raty et al39 of 101 women intensive care.
with BV. Of 51 women, 17 (33%) were cured after Three meta-analyses have evaluated the poten-
clindamycin treatment, compared with 17 out of tial benefit of treating BV in pregnancy. That by
50 (34%) of the placebo-treated group. The failure Brocklehurst et al40 included 1504 women.
rate of clindamycin to cure BV was particularly high Antibiotics were highly effective in eradicating
in this study. The preterm birth rate was 13.7% infection. The effect of treating BV resulted in a
(7/51) in the clindamycin-treated patients and 6.0% trend to fewer births before 37 weeks’ gestation,
198
0415421306-Ch13 3/30/07 9:46 AM Page 199
which was most marked in women with a previous Some have found reduction of premature delivery
preterm birth. only in those with a history of preterm birth.24,35,37,40
In the meta-analysis by Guise et al24 seven Preterm delivery is the major cause of perinatal
randomized controlled trials of BV treatment were mortality and morbidity in the developed world.
included. BV treatment was found to be of no bene- According to Lamont and Sawant,44 in up to 40% of
fit for the average-risk woman. In women with pre- cases, infection is a significant cause of spontaneous
vious preterm delivery, three of the studies showed a preterm labour. They recommend clindamycin as
benefit of BV treatment for preterm delivery before the antibiotic of choice.
37 weeks. Two trials of high-risk women found an
increase in preterm delivery less than 34 weeks in MISCARRIAGE
women who did not have BV but received BV treat-
ment. Both meta-analyses concluded that there is ASSOCIATION BETWEEN BV AND EARLY MISCARRIAGE
no evidence in favor of screening all pregnant
women for BV. For women with a history of previ- There have been a few studies linking BV with early
ous preterm birth, there is support for diagnosing first-trimester miscarriage.45–48 There is stronger
and treating BV early in pregnancy to prevent a pro- evidence, however, that BV is related to late first-
portion of these women having a further preterm trimester and second-trimester loss.49,50 The rela-
birth. tionship between BV and early miscarriage has been
However, in contrast, screening and treating BV assessed mainly in in vitro fertilization (IVF)
in low-risk pregnancies produced a statistically patients, or threatened miscarriage, rather than
significant reduction in premature deliveries (rela- recurrent miscarriage.
tive risk (RR) 0.73) in the meta-analysis by Varma Miscarriage rates were assessed in 867 consecu-
and Gupta,41 but there was no benefit in high-risk tive women undergoing IVF.45 BV was found in
groups. It was hypothesised that premature 24.6% of the women before egg collection. There
deliveries in high- and low-risk pregnant women were no differences in the conception rates between
are different entities and not linear extremes of the those women with BV and those with normal
same syndrome. vaginal flora. Twenty-two women (31.6%) with BV
There are significant clinical and methodological who conceived had a significantly increased risk
differences between the above studies that may of miscarriage in the first trimester compared with
account for the variation of the results and conclu- 27 women (18.5%) with normal vaginal flora. The
sions. Hay et al25 recommended that as BV is increased rate of miscarriage remained significant
associated with second-trimester miscarriage and after adjusting for factors known to increase the risk
preterm labor, treatment should be given no later of miscarriage: maternal age, smoking, a history of
than the beginning of the second trimester. recurrent miscarriage, no previous live birth, and
Rosenstein et al28 concluded that earlier diagnosis polycystic ovary syndrome (PCOS).
and treatment may be more effective in preventing In a further study to investigate the effect of vagi-
abnormal outcome, and they suggested that screen- nal flora and vaginal inflammation on conception
ing and treating before pregnancy might be advan- and early pregnancy loss, 91 women undergoing
tageous. Some have observed that treatment with IVF were recruited.46 At the time of embryo trans-
topical vaginal antibiotics has proven to be less fer, samples were taken for BV. The overall live birth
effective for the prevention of premature delivery rate was 30% and the rate of early pregnancy loss
than oral antibiotics.42,43 This would indicate that was 34%. Women with BV, intermediate flora, and
the microorganisms responsible for premature labor normal flora, had early pregnancy loss rates of 33%
have ascended out of reach of topically administered (1/3), 42% (5/12), and 30% (3/10) (p = 0.06), respec-
antibiotics, and the endometrial cavity would be the tively. It was concluded that IVF patients with BV
most likely place for them to initiate contractions. may have increased rates of early pregnancy loss.
199
0415421306-Ch13 3/30/07 9:46 AM Page 200
French et al51 reported that in a prospective analy- and preterm labor in the presence of BV. In the
sis of 1100 pregnant women, 60% of women with study by French et al51 of 1100 pregnant women,
first-trimester bleeding had one or more infections systemic antibiotics reduced the rate of preterm
detected, such as BV (RR 1.5), Trichomonas vaginalis birth for women with BV without first-trimester
(RR 2.3) and C. trachomatis (RR 2.7). Each of these bleeding (RR 0.37), and treatment of women with
infections heightened the risk for preterm delivery in BV and first-trimester bleeding reduced preterm
women with BV and first-trimester bleeding: RR 4.4 birth (RR 0.52). Clindamycin treatment was associ-
for BV; RR 3.0 for BV with T. vaginalis. ated with a reduction in the number of late miscar-
riages and premature births in Berger and Kane’s53
ASSOCIATION BETWEEN BV AND LATE MISCARRIAGE study of women with asymptomatic BV between 12
and 22 weeks’ gestation.
The association with later pregnancy losses has McGregor et al54 analyzed the effect of systemic
been reported in a number of studies. Llahi-Camp treatment to reduce pregnancy loss (<22 weeks),
et al49 found a history of one late miscarriage more preterm premature rupture of the membranes, and
than twice as commonly (27/130; 21%) compared preterm delivery in a prospective controlled treat-
with women who had only early miscarriages ment trial. The overall presence of BV was 32.5%.
(31/370; 8%) (p <0.001). In this study, BV did not BV was associated with pregnancy loss at less than
appear to be related to recurrent early miscarriage. 22 weeks (RR 3.1). The relative risk of preterm pre-
Hay et al,32 in a prospective study, screened mature rupture of the membranes was 3.5 and the
783 women for BV at their first antenatal clinic visit. relative risk of preterm birth was 1.9. In the treat-
There were 12 late miscarriages (16–24 weeks’ ment phase of the study, women with BV received
gestation) and a significant association with BV clindamycin orally. After treatment, there were fever
(p < 0.001). Oakeshott et al50 prospectively assessed preterm births (RR 0.5) and fever preterm prema-
1201 women presenting before 10 weeks’ gestation. ture ruptures of the membranes (RR 0.5).
The relative risk of miscarriage associated with BV Ugwumadu et al55 prospectively screened
compared with women who were negative for BV 6120 asymptomatic women at the first antenatal
before 16 weeks was 1.2. BV was associated with visit between 12 and 22 weeks’ gestation. The
miscarriage at 13–15 weeks at a relative risk of 485 women with abnormal smears were randomly
3.5. BV was therefore not strongly associated with allocated to receive oral clindamycin or placebo.
early miscarriage, but may be a factor for pregnancy There were significantly fewer midtrimester miscar-
loss after 13 weeks’ gestation. riages or preterm deliveries in the clindamycin
Donders et al52 assessed 228 women at 14 weeks’ group (13/244) compared with the placebo group
gestation by culture for BV-associated bacteria, in (38/241) (p = 0.0003).
order to determine whether there is a relationship In a multicenter prospective randomized
between BV and pregnancy loss up to 20 weeks. controlled trial,56 4429 low-risk asymptomatic
As screening was performed at 14 weeks, only women were screened for BV at their first routine
second-trimester losses could be assessed. The antenatal visit early in the second trimester. In the
relative risk for pregnancy loss between 14 and intervention group, the women received standard
20 weeks was 5.4 in the presence of BV. M. hominis antibiotic treatment and follow-up for any
and U. urealyticum were also associated with an detected infection; the number of preterm deliver-
increased risk of late miscarriage. ies was significantly lower (3.0%) than in the
control group (5.3%) (p = 0.0001). There were
EFFECT OF TREATMENT OF BV ON MISCARRIAGE 8 late miscarriages in the intervention group and
15 in the control group. It was concluded that
There is a general consensus in the literature that introducing a simple infection screening program
antibiotics reduce the incidence of late miscarriages, into routine antenatal care can significantly reduce
200
0415421306-Ch13 3/30/07 9:46 AM Page 201
late miscarriages and preterm births in a low risk- studied in 633 women.59 Cervical colonization with
group of pregnant women. U. urealyticum was found in 42.6% of 310 normal
pregnant women, in 41.6% of 84 patients undergo-
ing pregnancy termination, in 41.5% of normal
MYCOPLASMAS fertile patients, in 53% of 122 patients with sponta-
neous miscarriage, and in 64.5% of 76 women with
Di Bartolomeo et al57 established the prevalence of recurrent miscarriage. The cervical colonization
microorganisms in 198 pregnant women with rate was significantly higher in patients with
vaginal discharge. Endocervical and vaginal samples spontaneous miscarriage (p <0.05) and recurrent
were assessed using direct methods, culture, immu- spontaneous miscarriage (p <0.005) compared with
nodetection, and PCR looking for C. trachomatis, normal pregnant women. Endometrial colonization
Neisseria gonorrhoeae, St. agalactiae, T. vaginalis, was more frequent in patients with recurrent mis-
Candida, M. hominis, U. urealyticum, and BV. carriage (27.6%) than in normal fertile women
In 51 cases (26%), one of the above was detected. (9.7%) (p <0.05). U. urealyticum was isolated in five
BV was diagnosed in 30 cases (15%); U. urealyticum of six women with intact membranes and uncon-
was found in 49%, Candida in 34%, M. hominis trollable preterm labor between 20 and 28 weeks’
in 14.1%, St. agalactiae in 5%, T. vaginalis in 4%, gestation. Ureaplasma was also isolated from
and C. trachomatis in 2.5%; N. gonorrhoeae was not the placenta in four patients and the amniotic fluid
detected. in two of four patients. It was concluded that
As the evidence suggested that vaginal colo- U. urealyticum is a common commensal of the
nization with genital mycoplasmas plays a role in lower genital tract, but it may play a role in miscar-
complications of pregnancy, a study was set up to riage and in uncontrollable preterm labor.
determine whether antibiotics would reduce spon- However, the role of U. urealyticum in adverse
taneous pregnancy loss.58 The loss of a pregnancy pregnancy outcomes is disputed. There was no
included spontaneous miscarriage, stillbirths, pre- difference in the incidence of premature rupture of
mature infants who died, or term infants who died the membranes, preterm labor or low-birth weight
from congenital pneumonia due to U. urealyticum. infants between women carrying U. urealyticum
Women with spontaneous pregnancy wastage and and those who did not in the study by Carey et al.60
who were mycoplasma-positive in the genital tract This study assessed whether genital colonization
were treated prospectively in 71 pregnancies. There with U. urealyticum was associated with adverse preg-
was a significant reduction in pregnancy loss rate nancy outcome in 4934 women evaluated between
among those treated with doxycycline before preg- 23 and 26 weeks’ gestation.
nancy or erythromycin during pregnancy. The The prevalence of infection certainly seems
pregnancy loss rate in the untreated group was to be higher when the abortus is cultured at
remarkably high, with 22 of the 24 pregnancies midtrimester abortion or preterm labor. McDonald
being lost. There were 18 out of 37 pregnancies lost et al61 performed a prospective study of the changes
in the doxycycline-only group, 3 lost out of 20 preg- in vaginal flora between midtrimester and labour
nancies in the erythromycin group, and 2 of 12 lost in 560 women. Forty-five women delivered
after doxycycline and erythromycin. The benefit prematurely. U. urealyticum and G. vaginalis were
was independent of maternal age, number of previ- both associated with preterm birth when present in
ous miscarriages, or gestational age at miscarriage. the midtrimester. Light and immunofluorescence
It was concluded that antibiotics prescribed for microscopy were used to investigate 118 late mis-
women colonized with mycoplasmas could prevent carriages at 18–28 weeks’ gestation.62 Intrauterine
recurrent spontaneous miscarriage. infections were found in 86 cases, with mycoplasmas
The role of U. urealyticum in spontaneous being found in 44 (37%). One hundred and twenty-
and recurrent spontaneous miscarriage has been nine spontaneously delivered, non-macerated
201
0415421306-Ch13 3/30/07 9:46 AM Page 202
midgestation placentae and fetuses, between 16 and from both vagina and semen were found in 77 cou-
26 weeks’ gestation, were examined and cultured for ples with a spontaneous miscarriage rate of 46.7%,
aerobic and anaerobic bacteria, genital mycoplas- compared with 17.6% with vaginal infection
mas, and yeasts.63 Microorganisms were recovered alone (p <0.01). It was concluded that endocervical
in 85 (66%). Group B streptococcus was the most microorganisms, even when treated, may affect
significant pathogen, being recovered in 21 cases. implantation, and this is enhanced when the semen
Escherichia coli (22 cases) and U. urealyticum has shown infection.
(24 cases) were present mostly as mixed infections.
Specimens from 51 spontaneous early miscarriages
and 56 pregnancy terminations were investigated by ANTIBIOTICS IN UNEXPLAINED
culture for yeasts, gram-positive and gram-negative PREGNANCY LOSSES
bacteria, and genital mycoplasmas.64 Molecular
diagnostic tests for DNA sequences were performed Antibiotics have been prescribed in some studies
for C. trachomatis, HSV, adenovirus, and human without bacteriological confirmation. The maternal
papilloma virus (HPV). None of these were and fetal outcomes of the next pregnancy were
detected in the pregnancy terminations, whereas recorded in 254 couples attending an infertility
spontaneous miscarriage tissues were positive for at clinic following one or more spontaneous miscar-
least one microorganism in 31.5% of cases. riages20. One hundred couples requested antibi-
In the case of first-trimester abortion, an associ- otics: 96 received doxycycline 100 mg twice daily for
ation has not been found with mycoplasmas or ure- 4 weeks or tetracycline 500 mg four times daily for
aplasmas when placental specimens from aborted 4 weeks to cover C. trachomatis and mycoplasmas.
material were subjected to PCR for karyotyping and In addition, 49 patients received erythromycin
detection of bacterial and viral DNA.65 No evidence 500 mg four times daily for 2 weeks. Four patients
of M. hominis, U. urealyticum, human CMV or received ampicillin or cephalexin. There was a sig-
adeno-associated virus was found. C. trachomatis nificantly lower chance of miscarriage in the antibi-
DNA was detected once. However, Ye et al 66 took otic-treated group (10%) compared with the
endocervical swabs for mycoplasma in 58 women untreated group (38%) (p <0.01). Premature rupture
with spontaneous abortion and compared the out- of the membranes occurred in 4% of the treated
come of pregnancy with a control group of 50 normal group, compared with 46% in the control group.
pregnant women. In the index cases, positive results The antibiotic group had a higher vaginal delivery
for U. urealyticum and M. hominis were found in rate (69% vs 56%) (p <0.01), lower incidences of
74.1% (43/58) and 27.6% (16/58), respectively. fetal distress (6% vs 26%), respiratory distress
These results were significantly different to those of syndrome, and neonatal infection, a higher birth
the controls, the corresponding results being 48% weight, and better Apgar scores. It was postulated
(24/50) (p <0.01) and 10% (5/50) (p <0.05). It was that some spontaneous miscarriages may be caused
concluded that mycoplasma infection could be one by bacteria present in the genital tract at the time of
of the causes of early embryonic death. conception and that these bacteria may have an
Microbiological screening of vaginal flora adverse effect on the pregnancy.
and semen was performed 4 weeks before IVF for Antibiotic therapy has been assessed for first-
951 couples.67 Infections were found in 218 women trimester threatened miscarriage in women with
(22.9%) and appropriate treatment was prescribed. previous spontaneous miscarriage.68 Only those at a
There were 69 with Candida albicans, 49 with gestational age of less than 9 weeks were included.
U. urealyticum, 43 with G. vaginalis, 24 with strepto- Women with mild abdominal cramping received
coccus B or D, and 22 with E. coli. The implantation amoxicillin and erythromycin for 7 days. Severe
rate was significantly reduced in patients with infec- abdominal pain was treated with amoxicillin
tion: 14.6% versus 19.3% (p <0.02). Positive cultures and clindamycin for 7 days. Of the 23 pregnancies,
202
0415421306-Ch13 3/30/07 9:46 AM Page 203
22 were carried to term. It was concluded that antibi- demonstrated in a variety of clinical conditions. In
otics might prevent pregnancy loss in women with gynecology, the role of HPV in premalignant and
threatened miscarriage and that further clinical trials malignant cervical disease has been confirmed, and
are warranted. in obstetrics there appears to be a link between
A randomized placebo-controlled trial was set premature delivery and BV. Helicobacter pylori has
up to determine whether metronidazole reduces become established as the cause of peptic ulcera-
early preterm labour in asymptomatic women with tion. Fredricks and Relman2 have suggested that
positive vaginal fetal fibronectin in the second there are a number of chronic diseases whose etiol-
trimester of pregnancy.69 The women had at least one ogy remains obscure but that have characteristics
risk factor, including midtrimester loss or preterm indicating a microbial involvement. These diseases
delivery, uterine abnormality, cervical surgery, or include Crohn’s disease, rheumatoid arthritis, sys-
cervical cerclage. Nine hundred pregnancies were temic lupus erythematosis, atherosclerosis, multiple
screened for fetal fibronectin at 24 and 27 weeks’ sclerosis, and diabetes mellitus. Fredricks and
gestation, and the positive cases were randomized Relman2 have suggested that traditional technology
to receive a 7-day course of oral metronidazole or for detecting pathogens is not sufficiently sensitive
placebo. The primary outcome was delivery before to identify the microorganisms responsible.
30 weeks’ gestation, and the secondary outcomes Whipple’s disease illustrates the limitation of
included delivery before 37 weeks. Fetal fibronectin conventional bacteriology. Whipple described the
was a good predictor of early preterm birth, with a disease that bears his name in 1907. The syndrome
positive predictive value at 24 weeks’ gestation for consists of polyarthritis, weight loss, diarrhea,
delivery by 30 weeks of 26% and a negative predic- malabsorption, and lymphadenopathy. Whipple
tive value of 99%. The trial steering committee observed rod-like bacillary structures in mesenteric
stopped the study early; 21% (11/53) of the women lymph nodes, raising the possibility of a bacterial
receiving metronidazole delivered before 30 weeks, etiology. Although Whipple’s bacillus could be seen
compared with 11% (5/46) of those taking the by microscopy, it could not be grown in culture
placebo. Furthermore, there were significantly more or in animal hosts, and no successful serological
preterm deliveries (<37 weeks) in women receiving test could be devised. It was not until the arrival
the metronidazole (33/53; 62%) compared with of molecular biology that the bacillus could be
placebo (18/46; 39%). Treatment was initiated characterized. Fredericks and Relman2 concluded
relatively late, and damage would have preceded that failure to cultivate a microorganism does
the metronidazole, as all the patients studied had not prove that a disease is not due to a pathogen.
positive fibronectin tests. Bacteria may cause chronic systemic disease span-
ning decades. Furthermore, steroids may produce
temporary improvement without proving that
FUTURE DEVELOPMENTS the disease is inflammatory or autoimmune rather
than infectious. Finally, documented improvement
More than a century ago, Robert Koch devised a or cure associated with antimicrobial agents in a
scientific standard for determining whether a dis- chronic disease suggests a microbial origin.
ease is a result of a specific microorganism. Koch’s Bacteria have a remarkable propensity to survive
postulates stated that the pathogen should be iso- even in the most hostile environments.70 The vast
lated from the diseased host, grow in pure culture, majority of microorganisms are ‘unculturable’ or
and reproduce the disease when inoculated into ‘fastidious’, which means that they cannot be iden-
a susceptible host. Interestingly, Koch accepted tified by conventional culture techniques. Over the
that his postulates were not always useful (cited last few years, the development and application of
by Fredricks and Relman2). In recent years, a previ- molecular diagnostic techniques has revolutionized
ously unexpected infectious etiology has been the diagnosis and monitoring of infectious diseases.
203
0415421306-Ch13 3/30/07 9:46 AM Page 204
Molecular biological techniques are increasingly The antiphospholipid syndrome (APS) has been
being adopted in clinical laboratories. These molec- linked to recurrent miscarriage and other preg-
ular methods have made it possible to characterize nancy complications. It may respond to thrombo-
mixed microflora in their entirety, including uncul- prophylaxis, improving the live birth rate.74
turables. Molecular studies of the vaginal flora have Antiphospholipid antibodies (aPL) may be associ-
discovered many unculturable bacteria, including ated with infection, and one is left to contemplate
bacteria in the Clostridiales order, which are highly the possibility that some cases of recurrent miscar-
specific indicators of BV. A more complete under- riage could be related to underlying treatable infec-
standing of vaginal microbial populations resulting tion. In this context, it is of interest that APS
from molecular biological techniques may lead to disappears when H. pylori is eradicated.75
new strategies to maintain healthy vaginal floras, While microorganisms can be associated with
and will provide opportunities to explore the role of miscarriage, the question will always arise as to
novel bacteria in reproductive tract disease.71 whether they are pathogenic or opportunistic.
Biofilms develop by bacteria aggregating in a Ultimately, from a clinical point of view, what really
hydrated polymeric matrix of their own synthesis matters is whether treatment can reduce the occur-
on moist surfaces. They are inherently resistant to rence of spontaneous miscarriage. A few clinical
antimicrobial agents and are increasingly recognized studies so far have shown encouraging results, and
as being at the root of many persistent and chronic further research is warranted. It is recognized that
bacterial infections.72 Fredricks and Relman2 have screening for and treatment of BV in early pregnancy
observed that for more than a century bacteriolo- among high-risk women with a previous history
gists have attempted to culture Treponema pallidum of second-trimester miscarriage or spontaneous
and Mycobacterium leprae without success, although preterm labor may reduce the risk of recurrent late
the pathogenicity of these organisms is not in pregnancy loss and preterm birth. The fundamental
doubt. These authors argue that just as we cannot question of efficacy of antibiotic treatment for BV
cultivate known pathogens, we must accept the before pregnancy in women with recurrent early
possibility that other pathogens may exist that resist miscarriage has yet to be addressed in clinical stud-
cultivation. They have provided a set of guidelines ies. Developments in serological tests and molecular
to help prove microbial disease causation using biological techniques are enhancing our capability
molecular biological sequence-based evidence to detect evidence of infections in obstetrics and
rather than culture. gynecology. Ultimately, there is the option of a trial
Some have concluded that the best evidence of therapy with a presumptive diagnosis of genital
suggests that infection is an occasional cause of infection being related to recurrent miscarriage
sporadic spontaneous miscarriage and that recurrent without laboratory confirmation. The antibiotics
miscarriage occurs with a much lower frequency. of choice – metronidazole and the macrolides
At the other extreme, mycoplasmas have been such as erythromycin – are relatively innocuous.
found in 74% of spontaneous miscarriages with Nevertheless, antibiotics should be used with
embryonic death, compared with 48% of controls.66 caution, as there is the potential risk of the develop-
Recently, attention has focused on the relationship ment of bacterial resistance.
between periodontal infection and adverse preg-
nancy outcomes, including late miscarriage.73 REFERENCES
Periodontal disease is one of the most common
chronic infections, with a prevalence of 10–60% 1. Royal College of Obstetricians and Gynaecologists, Guideline No 17.
depending on diagnostic criteria. So far, there have The Investigation and Treatment of Couples with Recurrent
Miscarriage. London: RCOG, 2003.
been no reports on any association between first- 2. Fredricks DN, Relman DA. Infectious agents and the etiology of
trimester miscarriage and periodontitis. chronic idiopathic diseases. Curr Clin Top Infect Dis 1998; 18:180–200.
204
0415421306-Ch13 3/30/07 9:46 AM Page 205
3. Saracoglu OF, Mungan T, Tanzer F. Pelvic tuberculosis. Int J Gynaecol 24. Guise JM, Mahon SM, Aickin M, et al. Screening for bacterial
Obstet 1992; 37:115–20. vaginosis in pregnancy. Am J Prev Med 2001; 20(Suppl 3):62–72.
4. Figueroa-Damian R, Martinez-Velazco I, Villagrana-Zesati R, et al. 25. Hay PE, Morgan DJ, Ison CA, et al. A longitudinal study of bacterial
Tuberculosis of the female reproductive tract: effect on function. Int J vaginosis during pregnancy. Br J Obstet Gynaecol 1994; 101:1048–53.
Fertil Menopausal Stud 1996; 41:430–6. 26. Ugwumadu A, Reid F, Hay P, et al. Natural history of bacterial
5. Romana C, Salleras L, Sage M. Latent listerosis may cause habitual vaginosis and intermediate flora in pregnancy and effect of oral
abortion, intrauterine deaths, fetal malformations. When diagnosed clindamycin. Obstet Gynecol 2004; 104:114–19.
and treated adequately normal children will be born. Acta Microbiol 27. Riduan JM, Hillier SL, Utomo B, et al. Bacterial vaginosis and prema-
Hung 1989; 36:171–2. turity in Indonesia: asociation in early and late pregnancy. Am J
6. Manganiello PD, Yearke RR. A 10-year prospective study of women Obstet Gynecol 1993; 169:175–8.
with a history of recurrent fetal losses fails to identify Listeria mono- 28. Rosenstein IJ, Morgan DJ, Lamont RF, et al. Effect of intravaginal
cytogenes in the female genital tract. Fertil Steril 1991; 56:781–2. clindamycin cream on pregnancy outcome and on abnormal vaginal
7. Szkaradkiewicz A, Pieta P, Tulecka T, et al. The diagnostic value of microbial flora of pregnant women. Infect Dis Obstet Gynecol 2000;
anti-CMV and anti-HPV-B19 antiviral antibodies in studies on 8:158–65.
causes of recurrent abortions. Ginekol Pol 1997; 68:181–6. 29. McDonald HM, O’Loughlin JA, Jolley P, et al. Vaginal infection and
8. Cook SM, Himebaugh KS, Frank TS. Absence of cytomegalovirus in preterm labour. Br J Obstet Gynaecol 1991; 98:427–35.
gestational tissue in recurrent spontaneous abortion. Diagn Mol 30. Klebanoff MA, Hillier SL, Nugent RP, et al. Is bacterial vaginosis a
Pathol 1993; 2:116–19. stronger risk factor for preterm birth when it is diagnosed earlier in
9. Mezinova NN, Chuchupalov PD, Evdokimova NS, et al. Effect of anti- gestation? Am J Obstet Gynecol 2005; 192:470–7.
chlamydial drugs on the effectiveness of the treatment of habitual 31. Lamont RF, Duncan SL, Mandal D, et al. Intravaginal clindamycin to
abortion. Akush Ginekol (Mosk) 1991; 7:30–2. reduce preterm birth in women with abnormal genital tract flora.
10. Mezinova NN, Chuchupalov PD. Endometrial Chlamydia infection in Obstet Gynecol 2003; 101:516–22.
women with habitual abortion. Akush Ginekol (Mosk) 1992; 2:25–6. 32. Hay PE, Lamont RF, Taylor-Robinson D, et al. Abnormal bacterial
11. Kishore J, Agarwal J, Agarwal S, et al. Seroanalysis of Chlamydia tra- colonisation of the genital tract and subsequent preterm delivery and
chomatis and S-TORCH agents in women with recurrent spontaneous late miscarriage. BMJ 1994; 308:295–8.
abortions. Indian J Pathol Microbiol 2003; 46:684–7. 33. Goffinet F, Maillard F, Mihoubi N, et al. Bacterial vaginosis:
12. Witkin SS, Ledger WJ. Antibodies to Chlamydia trachomatis in sera of prevalence and predictive value for premature delivery and neonatal
women with recurrent spontaneous abortions. Am J Obstet Gynecol infection in women with preterm labour and intact membranes.
1992; 167:135–9. Eur J Obstet Gynecol Reprod Biol 2003; 108:146–51.
13. Olliaro P, Regazzetti A, Gorini G, et al. Chlamydia trachomatis 34. Borisov I, Dimitrova V, Mazneikova V, et al. Therapeutic regimens for
infection in ‘sine causa’ recurrent abortion. Boll Ist Sieroter Milan treating bacterial vaginosis in pregnant women. Akush Ginekol
1991; 70:467–70. (Sofiia) 1999; 38:14–16.
14. Rae R, Smith IW, Liston WA, et al. Chlamydial serologic studies and 35. Morales WJ, Schorr S, Albritton J. Effect of metronidazole in patients
recurrent spontaneous abortion. Am J Obstet Gynecol 1994; 170:782–5. with preterm birth in preceding pregnancy and bacterial vaginosis:
15. Paukku M, Tulppala M, Puolakkainen M, et al. Lack of association a placebo-controlled, double-blind study. Am J Obstet Gynecol 1994;
between serum antibodies to Chlamydia trachomatis and a history of 171:345–7.
recurrent pregnancy loss. Fertil Steril 1999; 72:427–30. 36. Hauth JC, Goldenberg RL, Andrews WW, et al. Reduced incidence of
16. Sugiura-Ogasawara M, Ozaki Y, Nakanishi T, et al, Pregnancy preterm delivery with metronidazole and erythromycin in women
outcome in recurrent aborters is not influenced by Chlamydiae IGA with bacterial vaginosis. N Engl J Med 1995; 333:1732–6.
and/or G. Am J Reprod Immunol 2005; 53:50–3. 37. McDonald HM, O’Loughlin JA, Vigneswaran R, et al. Impact of
17. Schulz KF, Cates W Jr, O’Mara PR. A synopsis of the problems metronidazole therapy on preterm birth in women with bacterial
in Africa in syphilis and gonorrhoea during pregnancy. Afr J Sex vaginosis flora (Gardnerella vaginalis): a randomised, placebo
Transm Dis 1986; 2:56–7. controlled trial. Br J Obstet Gynaecol 1997; 104:1391–7.
18. Amsel R, Totten PA, Spiegel CA, et al. Nonspecific vaginitis. 38. Camargo RP, Simoes JA, Cecatti JG, et al. Impact of treatment for
Diagnostic criteria and microbial and epidemiologic associations. bacterial vaginosis on prematurity among Brazilian pregnant women:
Am J Med 1983; 74:14–22. a retrospective cohort study. Sao Paulo Med J 2005; 123:108–12.
19. Wilson JD, Ralph SG, Rutherford AJ. Rates of BV in women 39. Kurkinen-Raty M, Vuopala S, Koskela M, et al. A randomised
undergoing in vitro fertilisation for different types of infertility. BJOG controlled trial of vaginal clindamycin for early pregnancy bacterial
2002; 109:714–17. vaginosis. BJOG 2000; 107:1427–32.
20. Toth A, Lesser ML, Brooks-Toth CW, et al. Outcome of subsequent 40. Brocklehurst P, Hanna M, McDonald H. Interventions for treating
pregnancies following antibiotic therapy after primary or multiple bacterial vaginosis in pregnancy. Cochrane Database Syst Rev 2000;
spontaneous abortions. Surg Gynecol Obstet 1986; 163:243–50. (2):CD00026.
21. Viniker DA. Hypothesis on the role of sub-clinical bacteria of the 41. Varma R, Gupta JK. Antibiotic treatment of bacterial vaginosis in
endometrium (bacteria endometrialis) in gynaecological and obstet- pregnancy: multiple meta-analyses and dilemmas in interpretation.
ric enigmas. Hum Reprod Update 1999; 5:373–85. Eur J Obstet Gynecol Reprod Biol 2006; 124:10–14.
22. Lucisano A, Morandotti G, Marana R, et al. Chlamydial genital infec- 42. Majeroni BA. Bacterial vaginosis: an update. Am Fam Physician 1998;
tions and laparoscopic findings in infertile women. Eur J Epidemiol 57:1285–9.
1992; 8:645–9. 43. McGregor JA. Evidence based prevention of preterm birth/PROM:
23. Moller BR, Kristiansen FV, Thorsen P, et al. Sterility of the uterine infection and inflammation. Paper presented at the Problem with
cavity. Acta Obstet Gynecol Scand 1995; 74:216–19. Prematurity II. St Thomas’ Hospital, London, 7–9 September, 1998.
205
0415421306-Ch13 3/30/07 9:46 AM Page 206
44. Lamont RF, Sawant SR. Infection in the prediction and antibiotics in 60. Carey JC, Blackwelder WC, Nugent RP, et al. Antepartum cultures for
the prevention of spontaneous preterm labour and preterm birth. Ureaplasma urealyticum are not useful in predicting pregnancy
Minerva Ginecol 2005; 57:423–33. outcome. The Vaginal Infections and Prematurity Study Group. Am J
45. Ralph SG, Rutherford AJ, Wilson JD. Influence of bacterial vaginosis Obstet Gynecol 1991; 164:728–33.
on conception and miscarriage in the first trimester: cohort study. 61. McDonald HM, O’Loughlin JA, Jolley, PT, et al. Changes in vaginal
BMJ 1999; 319:220–3. flora during pregnancy and association with preterm birth. J Infect
46. Eckert LO, Moore DE, Patton DL, et al. Relationship of vaginal Dis 1994; 170:724–8.
bacteria and inflammation with conception and early pregnancy loss 62. Fedotova EP, Shastina GV. Intrauterine mycoplasmosis in late miscar-
following in-vitro fertilization. Infect Dis Obstet Gynecol 2003; riage. Arkh Patol 1994; 56:61–5.
11:11–17. 63. McDonald HM, Chambers HM. Intrauterine infection and sponta-
47. Ugwumadu AH. Bacterial vaginosis in pregnancy. Curr Opin Obstet neous midgestation abortion: Is the spectrum of microorganisms
Gynecol 2002; 14:115–18. similar to that in preterm labor? Infect Dis Obstet Gynecol 2000;
48. Leitich H, Bodner-Adler B, Brunbauer M, et al. Bacterial vaginosis as 8:220–7.
a risk factor for preterm delivery: a meta-analysis. Am J Obstet 64. Penta M, Lukic A, Conte MP, et al. Infectious agents in tissues
Gynecol 2003; 189:139–47. from spontaneous abortions in the first trimester of pregnancy.
49. Llahi-Camp JM, Rai R, Ison C, et al. Association of bacterial vaginosis New Microbiol 2003; 26:329–37.
with a history of second trimester miscarriage. Hum Reprod 1996; 65. Matovina M, Husnjak K, Milutin N, et al. Possible role of bacterial
11:1575–8. and viral infections in miscarriages. Fertil Steril 2004; 81:662–9.
50. Oakeshott P, Hay P, Hay S, et al. Association between bacterial 66. Ye LL, Zhang BY, Cao WL. Relationship between the endocervical
vaginosis or chlamydial infection and miscarriage before 16 weeks’ mycoplasma infection and spontaneous abortion due to early
gestation: prospective community based cohort study. BMJ 2002; embryonic death. Zhonghua Fu Chan Ke Za Zhi 2004; 39:83–5.
325:1334–7. 67. Wittemer C, Bettahar-Lebugle K, Ohl J, et al. Abnormal bacterial
51. French JI, McGregor JA, Draper D, et al. Gestational bleeding, colonisation of the vagina and implantation during assisted repro-
bacterial vaginosis and common reproductive tract infections: risk duction. Gynecol Obstet Fertil 2004; 32:135–9.
for preterm birth and benefit of treatment. Obstet Gynecol 1999; 68. Ou MC, Pang CC, Chen FM, et al. Antibiotic treatment for threatened
93:715–24. abortion during the early first trimester in women with previous
52. Donders GG, Van Bulck B, Caudron J, et al. Relationship of bacterial spontaneous abortion. Acta Obstet Gynecol Scand 2001; 80:753–6.
vaginosis and mycoplasmas to the risk of spontaneous abortion. Am 69. Shennan A, Crawshaw S, Briley A. A randomised controlled trial of
J Obstet Gynecol 2000; 183:431–7. metronidazole for the prevention of preterm birth in women positive
53. Berger A, Kane KY. Clindamycin for vaginosis reduces prematurity for cervicovaginal fetal fibronectin: The PREMET Study. BJOG 2006;
and late miscarriage. J Fam Pract 2003; 52:603–4. 113:65–74.
54. McGregor JA, French JI, Parker R, et al. Prevention of premature birth 70. Nichols CA, Guezennec J, Bowman JP. Bacterial exopolysaccharides
by screening and treatment for common genital tract infections: from extreme marine environments with special consideration of the
results of a prospective controlled evaluation. Am J Obstet Gynecol southern ocean, sea ice, and deep-sea hydrothermal vents: a review.
1995; 173:157–67. Mar Biotechnol (NY) 2005; 7:253–71.
55. Ugwumadu A, Manyonda I, Reid F, et al. Effect of early oral 71. Fredricks DN, Marrazzo JM. Molecular methodology in determining
clindamycin on late miscarriage and preterm delivery in asympto- vaginal flora in health and disease: its time has come. Curr Infect Dis
matic women with abnormal vaginal flora and bacterial vaginosis: a Rep 2005; 7:463–70.
randomised controlled trial. Lancet 2003; 361:983–8. 72. Costerton JW, Stewart PS, Greenberg EP. Bacterial biofilms;
56. Kiss H, Petricevic L, Husslein P. Prospective randomised controlled A common cause of persistent infections. Science 1999; 284:1318–22.
trial of an infection screening programme to reduce the rate of 73. Farrell Nee Moore S, Ide M, Wilson RF. The relationship between
preterm delivery. BMJ 2004; 329:371. maternal periodontitis, adverse pregnancy outcome and miscarriage
57. Di Bartolomeo S, Rodriguez M, Sauka D, et al. Microbiologic profile in never smokers. J Clin Periodontol 2006; 33:115–20.
in symptomatic pregnant women’s genital secretions in Gran Buenos 74. Rai R, Regan L. Antiphospholipid syndrome and pregnancy loss.
Aires, Argentina. Enferm Infecc Microbiol Clin 2001; 19:99–102. Hosp Med 1998; 59:637–9.
58. Quinn PA, Shewchuk AB, Shuber J, et al. Efficacy of antibiotic therapy 75. Cicconi V, Carloni E, Franceschi F, et al. Disappearance of antiphos-
in preventing spontaneous pregnancy loss among couples colonized pholipid antibodies syndrome after Helicobacter pylori eradication.
with genital mycoplasmas. Am J Obstet Gynecol 1983; 145:239–44. Am J Med 2001; 111:163–4.
59. Naessens A, Foulon W, Cammu H, et al. Epidemiology and pathogen-
esis of U. urealyticum in spontaneous abortion and early preterm
labor. Acta Obstet Gynecol Scand 1987; 66:513–16.
206
0415421306-Ch14 3/30/07 9:46 AM Page 207
14. Midtrimester loss – the role

of cerclage
Israel Hendler and Howard JA Carp
length, the particular problems of cerclage in

INTRODUCTION
recurrent pregnancy loss (RPL), the role of trans-
vaginal and transabdominal cervical cerclage, and
Cervical incompetence or (preferably named)
the optimal method of performing the technique.
cervical insufficiency is defined as ‘the inability of
the uterine cervix to retain a pregnancy in the
absence of contractions or labor’. It is a clinical
diagnosis characterized by recurrent painless cer- PATHOPHYSIOLOGY
vical dilatation and spontaneous midtrimester
loss, generally in the absence of obvious predis- The pathophysiology of cervical insufficiency is still
posing conditions such as spontaneous membrane poorly understood. The cervix of a woman with
rupture, bleeding, or infection, which may indicate cervical incompetence contains a higher proportion
other causes of preterm birth rather than cervical of smooth muscle cells compared with the cervix
insufficiency. Although cervical incompetence was of a pregnant woman without cervical incompe-
first described in the English language literature tence.2,3 Deficiencies of cervical collagen,4,5 cervical
in 1678, and great strides have been made in the elastin,6 or other structural, mechanical compo-
understanding of this condition, the clinical diag- nents of cervical connective tissue have been postu-
nosis is usually made in retrospect, after a poor lated as etiological factors. These factors normally
obstetric outcome. The diagnosis is difficult to resist softening, effacement, and dilatation caused
make, and usually depends on a careful history and by the gravitational effect of the fetus and amniotic
review of the medical records, rather than accurate fluid. It has been difficult to confirm any theory of
diagnostic means or laboratory tests. True cervical pathophysiology due to the difficulty in obtaining
insufficiency is probably an uncommon diagnosis; biopsy samples from the human cervix before,
however, the lack of clear diagnostic criteria make during, and after term and preterm deliveries.
the incidence difficult to ascertain objectively. The A different model has been suggested by more
condition is important, as midtrimester pregnancy recent studies using serial endovaginal ultrasound
loss will ensue or the condition can be part of a measurements of cervical length, dilatation, and
spectrum leading to preterm delivery in all stages funneling of the membranes into the cervical canal.
of the third trimester, with its attendant risks to This tentative model is based on both the sequential
the infant. measurements and on individual observations that
Cervical cerclage has been the classical treatment the pregnant cervix is a dynamic structure, occa-
for cervical incompetence since its introduction by sionally opening and closing with no apparent
Shirodkar1 in 1955. However, the indications for relation to uterine contractions. Iams7 has proposed
cerclage are still far from clear, as are the optimal the model of a continuum of cervical compliance
method and timing. This chapter focuses on the (‘competence’) similar to the natural biological
diagnosis of cervical insufficiency, the obstetric variation in other physical traits, such as height,
management of women at high risk for preterm tendon strength, and long bone length. According
delivery by ultrasonographic follow-up of cervical to this model, cervical compliance and cervical
207
0415421306-Ch14 3/30/07 9:46 AM Page 208
length are qualities that vary from woman to woman, the above tests indicating that there is no cervical
and these qualities are just some of the components insufficiency in the non-pregnant state. The Hegar
of uterine function that affect the timing of delivery. test has been adapted by Fournil et al.12 In this
At present, we have empirical data about how technique, an 8 mm Hegar dilator is inserted into
cervical length and dilatation occur in advancing the endocervical canal twice weekly until 19 weeks
pregnancy, but not why those changes occur before of gestation. However, this technique is not really
term in some women and at term in most women. different to that of Shirodkar. Indeed, Shirodkar’s
Although the predictive value of these measure- method of digital examination has been used to
ments remains controversial, their progressive detect preterm delivery by many other authors.13,14
nature has been employed to develop the concept of We use digital examination on a routine basis
varying degrees of functional cervical insufficiency. when following patients up in pregnancy after RPL.
This method is simple and non-invasive, and has
picked up a number of patients with impending
DIAGNOSIS OF CERVICAL INSUFFICIENCY second-trimester loss and preterm labor. In some
cases, it has indicated the need for late cerclage.
There are essentially no proven objective criteria to However, digital examination leads to a diagnosis
diagnose cervical insufficiency (other than those few after cervical dilatation has reached the external os.
cases with gross cervical malformations). Shirodkar1 At this stage, it may be impossible to prevent second-
stated that 95% of cases of cervical insufficiency trimester loss or preterm delivery. Therefore, none
were due to a weak cervical sphincter and the other of the above methods is entirely satisfactory.
few to an underdeveloped or malformed uterus. Sarti et al15 introduced abdominal ultrasound
Shirodkar claimed to diagnose the condition by to detect cervical dilatation of the cervix. However,
‘repeated internal examinations’. Since 1955, many transabdominal ultrasonography requires a full
physicians have searched for diagnostic tools. The bladder in order to visualize the cervix. The full
Hegar test is still used to measure the diameter of bladder can press on the dilated cervix and cause
the endocervical canal. A cut-off of 6–8 mm is a false-negative result, as the bladder can cause
usually taken to make the diagnosis of cervical lengthening of the canal.16,17 This problem can be
insufficiency.8,9 However, in RPL, the patient usu- overcome if a vaginal probe is used. Funneling
ally presents in the interval between pregnancies, and shortening can be identified. The relationship
when the endocervical canal is not subject to the between cervical length and the risk of preterm
forces present in pregnancy, but in a grossly incom- delivery has been described.18,19 Figure 14.1(a)
petent cervix the Hegar test may give obvious infor- shows the sonogram of a normal cervix on ultra-
mation about the diagnosis. In order to overcome sound. As shortening of cervical length seems to
the disadvantage of the lack of pressure from inside be a continuous process, ultrasound can detect
the uterus in the Hegar test, the traction test was dilatation of the internal os before the external os
devised. A catheter is inserted into the uterine cavity, is affected. Figure 14.1(b) shows shortening of
and the balloon is filled with 1 ml of fluid. The the cervical canal. Transvaginal ultrasound can
catheter is then pulled through the cervical canal. If theoretically detect midterm loss and preterm
the balloon can be removed by a force of less than delivery earlier than other methods of detection.
600 g, then the diagnosis of cervical insufficiency is Figure 14.2 shows funneling of the internal os
made.10 Hysterosalpingography provides informa- and shortening of the cervical canal. However,
tion about the diameter and the shape of the cervi- transcervical ultrasonography has a number of
cal canal.9,11 A funnel-shaped isthmus is taken to be drawbacks. Figure 14.3 shows an apparently
diagnostic for cervical insufficiency. normal-looking cervix. However, the application
It may also be necessary to diagnose insuffi- of light fundal pressure allows the insufficiency
ciency in pregnancy, as the cervix may dilate despite to become apparent. It is impractical to screen the
208
0415421306-Ch14 3/30/07 9:46 AM Page 209
MIDTRIMESTER LOSS – THE ROLE OF CERCLAGE
A B
Figure 14.1 Ultrasound of cervical length: (a) normal cervix of 35 mm length; (b) shortened cervix of 14 mm length. These
sonograms show normal cervices. The cervix in (a) is completely closed, with a length of 35 mm (as seen between the calipers).
The cervix in (b) is 14 mm in length, but can still be competent.
entire obstetric population, and grand multipara to screen all patients. Although the incidences
can have wide open cervices without insufficiency. of cervical incompetence,20 midtrimester loss, and
Hence, transcervical ultrasound is not always preterm labor are higher after RPL,20,21 (see Figure
selective, as Figure 14.1(b) shows. In patients with 11.4 in Chapter 11), this higher incidence is
recurrent first-trimester losses, it is also not practical probably not high enough to justify screening
Figure 14.2 Sonogram of cervical incompetence. This shows funneling of the cervix, with a dilatation of the internal os.
The remaining cervical canal from the funneling to the external os is extremely shortened.
209
0415421306-Ch14 3/30/07 9:46 AM Page 210
A B
Figure 14.3 A dynamic cervix: (a) cervix with no fundal pressure;. (b) cervix with fundal pressure. The cervix was shortened
from 28 mm to 0 mm during the examination by light fundal pressure.
the entire population on a regular basis. However, risk of preterm labor.26 Patients with multifetal
transvaginal ultrasonography is warranted to detect pregnancies also have a higher risk of preterm labor,
short cervical length and funneling in women at possibly due to the additional uterine distension
high risk for preterm delivery due to cervical and additional pressure of the uterine contents
incompetence.22,23 In RPL, we tend to screen preg- on the cervix.27 Spontaneous membrane rupture,
nant patients by transvaginal ultrasound if there bleeding, or infection also lead the uterus to
is a high risk of midtrimester loss or preterm labor, contract, leading to midtrimester loss or preterm
as shown by a past history, cervical tears, a labor. As there are essentially no proven objective
previous hysteroscopic diagnosis, or when digital criteria for diagnosing cervical insufficiency (other
examination indicates cervical insufficiency. In than perhaps the rare gross cervical malformation),
women with midtrimester loss, transvaginal ultra- the diagnosis is usually made by a history of
sound measurement of cervical length can be used midtrimester loss of a live fetus, after painless
as a screening tool indicating the need for cerclage, dilatation of the cervix. There may be a history of
as will be shown later in this chapter. trauma to the cervix from surgical procedures
such as conization,28 or previous dilatation and
DIFFERENTIAL DIAGNOSIS curettage.29 Alternately there may be a traumatic
delivery, such as precipitated labor with forceful
There are various conditions predisposing to expulsion of the fetus before full dilatation, leading
midtrimester loss and preterm labor. These include to cervical tearing, or a traumatic instrumental
uterine anomalies (see Chapter 11) and exposure to delivery with cervical tearing. However, diagnosis of
diethylstilbestol (DES) in utero.24,25 There is also a cervical insufficiency based on exclusion of other
group of women with an inherently short cervix factors is flawed, as an appreciable proportion of
(below the 5th percentile) who have an increased women who have apparent cervical insufficiency
210
0415421306-Ch14 3/30/07 9:46 AM Page 211
have evidence of subclinical intrauterine infection and the free ends of the threads are tied to leave a
on amniotic fluid analysis obtained by transabdom- loop, in order to facilitate early removal. However,
inal amniocentesis. the disadvantage of the McDonald technique is
that in cases of high cervical tears, it may not be
possible to place the suture high enough. Ideally,
CERVICAL CERCLAGE the suture should be at the level of the internal os,
but ultrasound often shows the suture to lay
Cerclage was first introduced by Shirodkar1 in halfway along the cervical canal.
1955 for the treatment of cervical insufficiency. In the Shirodkar technique, a transverse incision
Shirodkar described a group of 30 women who is made on the anterior side of the cervix and the
had at least 4 (and up to 11) previous abortions. bladder is pushed up above the internal cervical os.
Shirodkar stated that, in his opinion, ‘95% of cases A vertical incision is then made in the posterior
of cervical incompetence were due to a weak vaginal wall. The stitch is placed in two bites,
cervical sphincter and the other few to an underde- penetrating the cervical tissue laterally. In this way,
veloped or malformed uterus, etc.’ Shirodkar the cervix is encircled without affecting the more
emphasized that his work was confined to women laterally placed uterine vessels. The mucosal inci-
in whom he could prove the existence of weakness sions are then closed. The Shirodkar technique has
of the internal os by ‘repeated internal examina- the advantage of reaching up to 3 cm higher than
tions’. Fifty years later, cerclage is performed in the McDonald technique. If there are cervical
1:54–220 deliveries worldwide, although there is tears, the reflection of the bladder and the pouch
still confusion about the diagnostic criteria and of Douglas can be continued until above the level
uncertainty about its benefits. Before describing of the tear. However, there is the possibility of
its benefits and indications, the next section will injury to surrounding tissues. If the uterus con-
describe the techniques for performing cerclage. tracts and dilates, there may be tearing of the
uterus on the stitch. We tend to use a modified
TECHNIQUES OF CERCLAGE Shirodkar technique, in which the bladder is
reflected from the cervix, but not the pouch of
The McDonald technique30 is the simplest proce- Douglas. Additionally, we tend to use the modified
dure to perform technically. The procedure consists Shirodkar technique after failure of the McDonald
of suturing of the cervix as high as possible, by a procedure.
pursestring suture in five or six bites. The knot is If the suture needs to be placed still higher, or
positioned on the anterior aspect of the cervix. in cases with severe cervical scarring after multi-
The advantage of McDonald’s technique is that it ple failures of vaginal cerclage,32 an abdominal
can be performed by relatively junior staff, and approach can be used. This technique can be
there are few complications. In McDonald’s original performed by laparotomy, or more recently by
report, 70 women presented with second-trimester laparoscopy.33 In the technique of Anthony et al,34
dilatation of the cervix and bulging membranes the bladder is mobilized, the uterine arteries are
between 20 and 24 weeks of gestation. Thirty-three identified, and tunnels are created medial to
women (47%) gave birth to infants that survived. the uterine arteries. A 5 mm mersilene tape is
In a subsequent publication,31 McDonald reported then passed through the tunnels, and the stitch
that the proportion of success had increased, due is tied anteriorly. In the method of Topper and
to better selection of patients and the use of a Farquharson,35 a 2-gauge ethilon suture is passed
prophylactic cerclage at 14 weeks of gestation before through the muscle of the uterus medial to the
cervical dilatation. We often use a modified uterine vessels at the height of the isthmus above
McDonald technique in which the cervix is grasped the cardinal ligaments, and the stitch is tied anteri-
in three or four bites, the knot tied at 10–11 o’ clock, orly. The problem with this technique is that
211
0415421306-Ch14 3/30/07 9:46 AM Page 212
laparotomy or laparoscopy are required for This was an international multicenter study of
insertion. If the technique fails, laparotomy or women who were considered to be at risk for
laparoscopy may be required for removal. However, second-trimester pregnancy loss or preterm delivery
techniques have been developed for leaving the by history or previous cervical surgery. However,
knot so that it can be accessed without invasive the treating obstetrician was uncertain of the
techniques. In patients with badly scarred cervices, diagnosis of cervical insufficiency. A total of
the abdominal approach may be the only approach 1292 women were randomized to cerclage or no
that is possible. However, the vast majority of cerclage. This was a heterogeneous group of women
patients can still be treated with the less invasive with one or more risk factors for preterm birth. The
vaginal approach. 28% overall rate of spontaneous preterm birth
prior to 37 weeks was high. The intention-to-treat
VALUE OF CERCLAGE analysis showed a trend to a lower incidence of
preterm labor in the cerclage group compared with
The clinical value of cervical cerclage has been the non-cerclage group (26% and 31%, respec-
subject of many observational and randomized tively; p = 0.07), but this trend did not quite reach
clinical trials, and the studies have been subject to statistical significance. However, the trend to a
several systematic reviews. In this section, we lower preterm labor rate after cerclage did reach
will try to summarize the evidence regarding the statistical significance if 33 weeks was taken as the
benefit of cerclage for midtrimester loss. The cut-off point (13% after cerclage vs 17% after no
studies should be classified into those performed cerclage; p = 0.03). In a secondary analysis, stratified
prior to or subsequent to the era of transvaginal by primary risk factor for inclusion in the study,
ultrasound measurement of the cervical length. cerclage offered a significant benefit only to women
Beginning in 1982, three randomized clinical with a history of three or more spontaneous
trials of elective cerclage were performed, based on preterm births or second-trimester miscarriages
a history of midtrimester loss. Rush et al36 recruited (15% after cerclage vs 32% with no cerclage;
women who were referred to a reproductive failure p = 0.015). There was no advantage from cerclage
clinic, and evaluated whether the policy of per- in women with a history of one or two second-
forming cerclage prolongs gestation in women trimester losses. Women who were assigned to
with a history of late miscarriage. A total of cerclage received more tocolytic medications (34%
194 women were randomized to cerclage or no vs to 27% in women without cerclage), and had
cerclage. The study population had a very high rate a higher rate of antepartum hospital admissions
of preterm birth, as expected (33% delivered prior (37% vs 29%). Puerperal fever was also more
to 37 weeks). The proportion of preterm delivery common after cerclage (6% vs 3%; p = 0.03).
was 34% in the cerclage group and 32% in the However, it must be noted that women with a
patients who did not undergo cerclage. Women definite indication for cerclage (in the opinion of
with cerclage had more hospitalizations during their healthcare provider) were excluded from the
pregnancy, a lower birthweight and a higher rate of study – and this is a major caveat.
maternal puerperal fever. Due to the small number A meta-analysis by Drakley et al39 pooled the
of patients in the trial, a subanalysis restricted to results of four trials that together summarized
women with a history of second-trimester loss 2062 women. There was no difference in the total
was not performed. Lazar et al37 also failed to show pregnancy loss and early (<24 weeks) pregnancy
any benefit from cerclage in a larger cohort of loss rates (relative risk (RR) 0.86; 95% confidence
506 women. interval (CI) 0.59–1.25). Two trials reported on
The largest randomized study was performed delivery rates prior to 28 weeks of gestation, and
by the Medical Research Council/Royal College of three trials reported on deliveries before 32 weeks.
Obstetricians and Gynaecologists (MRC/RCOG).38 There was no beneficial effect of cerclage (RR 1.29;
212
0415421306-Ch14 3/30/07 9:46 AM Page 213
95% CI 0.67–2.49). There was also no difference cerclage group received preoperative and postoper-
in perinatal deaths (RR 0.8; 95% CI 0.48–1.36) or ative indomethacin. Rust et al41 randomly assigned
mean gestational age between the two groups. 138 at-risk women who were found to have a
All four studies reported on preterm delivery at sonographic cervical length of less than 25 mm or
less than 37 weeks of gestation, with no overall more than 25% funneling at 16–24 weeks of gesta-
significant difference between the two groups tion to cerclage or to a control group. There was
(RR 1.04; 95% CI 0.99–1.10). In summary, it no statistical difference between the cerclage and
appears that the use of cerclage in women with non-cerclage groups with regard to previous
various risk factors for spontaneous preterm birth second-trimester loss (12.1% and 27.3%, respec-
has little clinical benefit, but is associated with tively; p = 0.07). The eligible patient population
more medical complications and interventions. (12% of whom had multifetal gestations) under-
Women with recurrent preterm birth, including went transabdominal amniocentesis before cerclage
midtrimester losses, represent a population in need to rule out intrauterine infection, and both groups
of further study. were admitted to hospital to receive a 24–48-hour
In 1996, Iams et al26 described the length of the course of antibiotics and indomethacin before
cervix as measured by transvaginal ultrasound discharge home to modified bed rest. Preterm
in almost 3000 low-risk women, and found a bell- births occurred in 35% of the cerclage group and
shaped distribution curve of cervical length among 36% of the controls. Berghella et al42 studied
the population, with an inverse relationship between women with one or more high-risk factors for
cervical length and the risk of preterm birth. This preterm birth (one or more preterm births at less
led to the assumption that cervical length can be than 35 weeks, two or more curettages, DES
used as an indicator of cervical insufficiency. exposure, cone biopsy, müllerian anomaly, or twin
Subsequently, the demonstration of a short cervical gestation). These women were screened with
length on ultrasound was widely used as an indica- transvaginal ultrasonography every 2 weeks from
tion for prophylactic cerclage. There have been 14 weeks up to 24 weeks of gestation. Both asymp-
four recently published randomized clinical trials tomatic women who were at high risk and those
analyzing the benefit of cerclage in high- and who were identified to have a short cervix (< 25 mm)
low-risk women based on cervical length. These or significant funneling (>25%) were enrolled, but
trials have produced conflicting results. Althuisius no screened women who were at low risk and who
et al40 enrolled 35 women who had risk factors were identified incidentally. Sixty-one women were
for or symptoms of cervical incompetence and a randomly assigned to receive either McDonald
sonographic cervical length less than 25 mm. This cerclage or bed rest only. Forty-seven pregnan-
population included women with a clinical history cies (77%) were high-risk singleton gestations.
of painless dilatation or early chorioamnion rup- Thirty-one women (51%) were allocated to cerclage
ture in their previous pregnancies. If painless dilata- and 30 women (49%) were allocated to bed rest.
tion or early chorioamnion rupture had occurred in Preterm birth (< 35 weeks of gestation) occurred in
a prior, but not the most recent, gestation, the 14 women (45%) in the cerclage group and in
patient remained eligible for enrollment in the trial 14 women (47%) in the bed rest group (RR 0.94;
if the subsequent pregnancies were managed with 95% CI 0.34–2.58). There was no difference in
cerclage, bed rest or both. Of the 35 women enrolled any obstetric or neonatal outcome. The largest
in the study, the 19 women who were randomly randomized trial to date has been performed
assigned to cerclage and modified bed rest had no by To et al.43 Cervical length was measured in
preterm births prior to 34 weeks, compared with a 47 123 women. The cervix was 15 mm or less in
44% preterm birth rate in the 16 women treated by 470 women, 253 (54%) of whom participated in the
rest alone. Both groups received a 6-day course of study. One hundred and twenty-seven women
broad-spectrum antibiotics, whereas only the were randomized to undergo cervical cerclage and
213
0415421306-Ch14 3/30/07 9:46 AM Page 214
126 received expectant management Twenty-two OPTIMAL TIME FOR PLACEMENT OF CERCLAGE
percent (28 of 127 women) in the cerclage group
and 26% (33 of 126) of women in the control group It is debatable whether cerclage should be performed
had preterm birth prior to 33 weeks (RR 0.84; 95% prophylactically at 13–15 weeks of gestation, or
CI 0.54–1.31; p = 0.44). There were no significant whether the procedure should be performed on an
differences in perinatal or maternal morbidity or emergent basis later in pregnancy when it is found
mortality. to be definitely indicated by changes in the cervix.
Berghella et al44 have performed a meta-analysis Prophylactic cerclage has usually been considered
of the four randomized controlled trials described to be preferable, as it is performed when the cervix
above. The meta-analysis had the support of the is closed, before dilatation, before vaginal bacteria
authors of each of the trials, and additional data can ascend to the membranes, and before the
were added for patients whose results were not release of cytokines and prostaglandins, all of
available at the time of completion of the original which occur with cervical dilatation. However, if
reports. In the total population, preterm birth (< 35 prophylactic cerclage is performed, a large number
weeks) occurred in 29.2% (89/305) of the cerclage of unnecessary procedures will be carried out.
group, compared with 34.8% (105/302) of the Cerclage is not an innocent procedure: it is associ-
non-cerclage group (RR 0.84; 95% CI 0.67–1.06). ated with an increased risk of premature preterm
However, there was a significant reduction in rupture of the membranes, bleeding, and intrauter-
preterm births (<35 weeks) in the cerclage group ine infection. Cerclage may also cause severe pain
compared with the non-cerclage group in singleton and inconvenience throughout pregnancy. If late
gestations (RR 0.74, CI 0.57–0.96), singleton gesta- cerclage is performed, fewer procedures will be
tions with prior preterm birth (RR 0.61; 95% performed, thus reducing the incidence of side-
CI 0.40–0.92), and singleton gestations with prior effects. However, effective placement of an ‘emer-
second-trimester loss (RR 0.57; 95% CI 0.33–0.99). gent’ cerclage depends on an accurate diagnosis,
Berghella et al44 concluded that cerclage does not so that the cerclage is inserted prior to the mem-
prevent preterm birth in all women with a short branes becoming exposed to the vagina, and prior
cervical length on transvaginal ultrasonography, to dilatation of the external os, and subsequent
but may reduce the number of preterm births in contractions. Cerclage becomes much less effective
singleton gestations with a short cervical length, if the diagnosis of funneling is missed, and the
especially in those patients with a prior preterm membranes are exposed to the vagina (leading to
birth or prior second-trimester loss. Berghella infection) and if the external os is dilated.
et al,44 recommended that a sufficiently powered In order to accept the conclusion of Berghella
trial should be carried out in this group of et al44 that cerclage may reduce the number of
patients. preterm births in singleton gestations with a short
In summary, current evidence suggests the cervical length, it is necessary to validate the efficacy
following: of a cerclage that is introduced after the cervix
has shortened. Berghella et al45 have summarized
● Cervical cerclage does not reduce the rate of the results of 177 American women with prior
spontaneous preterm birth in women with a second-trimester spontaneous losses between 14 and
sonographically short cervix (≤15 mm ) and at 24 weeks of gestation. Patients were managed with
low risk for preterm delivery by their history. prophylactic cerclage or serial transvaginal sono-
● However, recent evidence suggests that cerclage grams of the cervix, starting at 14 weeks at the
may reduce the incidence of preterm birth obstetricians’ discretion. Cerclage was only per-
prior to 35 weeks in women with a short cervix formed if the cervical length was less than 25 mm or
(≤25 mm) and a history of prior second- funneling was greater than 25% of the cervical
trimester loss. length prior to 24 weeks of gestation. All cerclages
214
0415421306-Ch14 3/30/07 9:46 AM Page 215
were performed by the McDonald method. Of Both of these studies reported a beneficial
the 177 patients, 66 received prophylactic cerclage effect from cerclage. In the CIPRACT trial,47
and 111 were followed up with transvaginal sonog- 23 women were randomized to have an emergency
raphy, of whom 36% (40/111) had a therapeutic cerclage or bed rest. The inclusion criteria were
cerclage due to cervical changes. The obstetric out- prolapsed membranes through the dilated cervix.
come, including preterm delivery, was similar Thirteen women were randomized to have a cer-
whether prophylactic cerclage or late emergent clage, and 10 women to bed rest alone. One woman
cerclage had been performed. The preterm delivery in the cerclage group had a known complication
rates (<35 weeks) were 23% and 21% after prophy- of cerclage in these conditions – rupture of the
lactic and late cerclage, respectively (p = 0.3). membranes during the procedure. Cerclage was
If 33 weeks was taken as the cut-off point, the therefore abandoned. Delivery was delayed by a
preterm delivery rate was 21% and 26%, respec- mean of 54 days in the cerclage group, compared
tively (p = 0.5). Nor was there any difference in the with 24 days in the bed rest group (p = 0.046). The
gestational age at delivery whether prophylactic prevalence of preterm delivery before 34 weeks of
(34.6 ± 6.8 weeks) or emergent (34.4 ± 6.8 weeks; gestation was significantly lower in the cerclage
p = 0.8) cerclage had been performed. Groom et al46 group (7 of 13 women), compared with all of the
found similar results in a British population. 10 women in the bed rest group (p = 0.02).
Thirty-nine women undergoing elective cervical The inclusion criteria for the study by
cerclage were matched with women undergoing Olatunbosun et al48 were slightly different, requir-
serial ultrasound surveillance of cervical length ing a cervical dilatation of at least 4 cm at
with regard to maternal age, ethnic group, previous 20–27 weeks of gestation. The study was prospec-
cervical surgery, previous second-trimester loss and tive and included 43 women. However, the trial was
early preterm delivery. Cerclage was performed in not randomized, and management was chosen by
14 (36%) of the ‘ultrasound surveillance’ patients the attending obstetrician. Twenty-three women
due to cervical changes. There was no significant received a cerclage, whereas 20 women were treated
difference in median gestation at delivery (266 days with bed rest. The cerclage failed in one patient,
after elective cerclage and 260 days after emergent and five of the women due to have bed rest were
cerclage, p = 0.9), number of women delivering eventually given a cerclage. The mean gestational
before 24 weeks (15% and 13%, respectively; age at delivery was 33 ± 4.4 weeks after cerclage,
p = 0.9), number at 24–32 weeks (7.5% and 15%, which was significantly later than the 28.8 ± 4.4
respectively; p = 0.6), and number at 32–37 weeks weeks in the bed rest group (p = 0.001).
(15% and 13%, respectively; p = 0.9). Based on The results of both these studies suggest that
these two studies, it appears that in women at emergency cerclage should be performed, if possi-
high risk of preterm delivery, serial transvaginal ble, in an attempt to delay labor. However, clinical
ultrasound surveillance of cervical length appears judgment is necessary in order to select patients
to reduce the number of cerclage procedures who may benefit. The presence of infection,
performed, without compromising pregnancy frequent contractions, etc. make the procedure less
outcome. likely to succeed, and increases the possibility of
side-effects such as tearing of the cervix if cervical
EMERGENCY CERCLAGE dilatation continues. The addition of tocolytic
agents and prophylactic antibiotics at the time of
As stated above, cerclage becomes less effective cerclage may increase the efficacy of the procedure.
once the external os is dilated, and the membranes In cases of emergency cerclage, a modified McDonald
are exposed to the vagina (leading to infection). procedure is usually performed, as there is often
Two trials have compared cerclage with expectant little tissue to grasp and this is often the only proce-
management with bed rest after cervical dilatation.47,48 dure possible.
215
0415421306-Ch14 3/30/07 9:46 AM Page 216
SUMMARY 3. Roddick JW, Buckingham JC, Danforth DN. The muscular cervix – a
cause of incompetency in pregnancy. Obstet Gynecol 1961; 17:562–5.
4. Rechberger T, Uldbjerg N, Oxlund H. Connective tissue changes in
Patients with RPL are at a high risk for midtrimester
the cervix during normal pregnancy and pregnancy complicated by
loss and preterm labor. The increased risk is present cervical incompetence. Obstet Gynecol 1988; 71:563–7.
even if all previous pregnancy losses have been in 5. Petersen LK, Uldbjerg N. Cervical collagen in non-pregnant women
with previous cervical incompetence. Eur J Obstet Gynecol Reprod
the first trimester. Cervical incompetence may be
Biol 1996; 67:41–5.
more common in women with RPL due to the large 6. Leppert PC, Yu SY, Keller S, et al. Decreased elastic fibers and
number of previous curettages. Additionally, there desmosine content in incompetent cervix. Am J Obstet Gynecol 1987;
157:1134–9.
may be cervical tears from previous labors that
7. Iams JD. Prediction and early detection of preterm labor. Obstet
may cause recurrent second-trimester losses in a Gynecol 2003; 101:402–12.
secondary aborter (one or more labors followed by 8. Page EW. Incompetent internal os of the cervix causing late abortion
and premature labor. Technique for surgical repair. Obstet Gynecol
a string of pregnancy losses). If there are obvious
1958; 12:509–15.
tears of the cervix and a typical history of painless 9. Toaff R, Toaff ME, Ballas S, et al. Cervical incompetence: diagnostic
dilatation of the cervix in the second trimester, and therapeutic aspects. Isr J Med Sci 1977; 13:39–49.
10. Bergman P, Svennerud S. Traction test for demonstrating incompe-
prophylactic cerclage is indicated, and the suture
tence of the internal os of the cervix. Int J Fertil 1957; 2:163–7.
should be placed high enough to prevent cervical 11. Jeffcoate TNA, Wilson JK. Uterine causes of abortion and premature
dilatation. This may require the use of the labor. NY State J Med 1956; 56:680–90.
12. Fournil C, Hidden J, Lajoux P. Evaluation du calibre de l’isthme utérin
Shirodkar technique or an abdominal approach.
en début de grossesse. Intérêt dans l’indication du cerclage du col.
If there are no apparent tears but the patient is at Nouv Presse Med 1977; 6:523–4, 531–3.
a high risk for second-trimester loss or preterm 13. Bouyer J, Papiernik E, Dreyfus J, et al. Maturation signs of the cervix
and prediction of preterm birth. Obstet Gynecol 1986; 68:209–14.
labor, ultrasound should be performed serially,
14. Catalano PM, Ashikaga T, Mann LI. Cervical change and uterine
probably on a weekly basis, in order to detect cer- activity as predictors of preterm delivery. Am J Perinatol 1989;
vical shortening. Cervical shortening is not in itself 6:185–90.
15. Sarti DA, Sample WF, Hobel CJ, et al. Ultrasonic visualization of a
an indication for cerclage. However, women with
dilated cervix during pregnancy. Radiology 1979; 130:417–20.
the combination of prior second-trimester loss 16. Andersen HF. Transvaginal and transabdominal ultrasonography
and a cervical length of less than 25 mm or of the uterine cervix during pregnancy. J Clin Ultrasound 1991;
19:77–83.
funneling of greater than 25% of the cervical length 17. Confino E, Mayden KL, Giglia RV, et al. Pitfalls in sonographic imag-
before 24 weeks may have a reduced rate of preterm ing of the incompetent uterine cervix. Acta Obstet Gynecol Scand
birth with cerclage. There is no benefit to be gained 1986; 65:593–7.
18. Andersen HF, Nugent CE, Wanty SD, et al. Prediction of risk for
from prophylactic cerclage compared with cerclage preterm delivery by ultrasonographic measurement of cervical length.
performed due to cervical changes. In the low-risk Am J Obstet Gynecol 1990; 163:859–67.
patient with RPL, clinical follow-up may be suffi- 19. Berghella V, Kuhlman K, Weiner S, et al. Cervical funneling: sono-
graphic criteria predictive of preterm delivery. Ultrasound Obstet
cient, and ultrasound can be reserved for those Gynecol 1997; 10:161–6.
patients in whom it is clinically indicated. 20. Sheiner E, Levy A, Katz M, et al. Pregnancy outcome following recur-
In imminent preterm delivery with a dilated rent spontaneous abortions. Eur J Obst Gynecol Reprod Biol 2005;
118:61–5.
external os and prolapse of the membranes, emer- 21. Hughes N, Hamilton EF, Tulandi T. Obstetric outcome in women
gency cerclage seems to be beneficial if technically after multiple spontaneous abortions. J Reprod Med 1991; 36:165–6.
possible, and if there are no complicating factors 22. Guzman ER, Vintzileos AM, McLean DA, et al. The natural history of
a positive response to transfundal pressure in women at risk for
such as infection or uterine contractions. cervical incompetence. Am J Obstet Gynecol 1997; 176:634–8.
23. Althuisius SM, Dekker GA, van Geijn HP, et al. Cervical Incompetence
Prevention Randomized Cerclage Trial (CIPRACT): study design and
REFERENCES preliminary results. Am J Obstet Gynecol 2000; 183:823–9.
24. Ludmir J, Landon MB, Gabbe SG, et al. Management of the diethyl-
1. Shirodkar VN. A new method of operative treatment for habitual abor- stilbestrol-exposed pregnant patient: a prospective study. Am J Obstet
tions in the second trimester of pregnancy. Antiseptic 1955; 52:299–300. Gynecol 1987; 157:665–9.
2. Buckingham JC, Buethe RA, Danforth DN. Collagen–muscle ratio in 25. Mangan CE, Borow L, Burtnett-Rubin MM, et al. Pregnancy outcome
clinically normal and clinically incompetent cervices. Am J Obstet in 98 women exposed to diethylstilbestrol in utero, their mothers, and
Gynecol 1965; 92:232–7. unexposed siblings. Obstet Gynecol 1982; 59:315–19.
216
0415421306-Ch14 3/30/07 9:46 AM Page 217
26. Iams JD, Goldenberg RL, Meis PJ, et al. The length of the cervix and 39. Drakeley AJ, Roberts D, Alfirevic Z. Cervical cerclage for prevention of
the risk of spontaneous premature delivery. N Engl J Med 1996; 334: preterm delivery: meta-analysis of randomized trials. Obstet Gynecol
567–72. 2003; 102:621–7.
27. Michaels WH, Schreiber FR, Padgett RJ, et al. Ultrasound surveillance 40. Althuisius SM, Dekker GA, Hummel P, et al. Final results of the
of the cervix in twin gestations: management of cervical incompe- Cervical Incompetence Prevention Randomized Cerclage Trial
tency. Obstet Gynecol 1991; 78:739–44. (CIPRACT): therapeutic cerclage with bed rest versus bed rest alone.
28. Kristensen J, Langhoff-Roos J, Wittrup M, et al. Cervical conization Am J Obstet Gynecol 2001; 185:1106–12.
and preterm delivery/low birth weight. A systematic review of the 41. Rust OA, Atlas RO, Reed J, et al. Revisiting the short cervix detected by
literature. Acta Obstet Gynecol Scand 1993; 72:640–4. transvaginal ultrasound in the second trimester: why cerclage therapy
29. Wright CSW, Campbell S, Beazley J. Second-trimester abortion after may not help. Am J Obstet Gynecol 2001; 185:1098–105.
vaginal termination of pregnancy. Lancet 1972; i:1278–9. 42. Berghella V, Odibo AO, Tolosa JE. Cerclage for prevention of preterm
30. McDonald IA. Suture of the cervix for inevitable miscarriage. J Obstet birth in women with a short cervix found on transvaginal examina-
Gynaecol Br Emp 1957; 146:346–50. tion: a randomized trial. Am J Obstet Gynecol 2004; 191:1311–17.
31. McDonald IA. Incompetent cervix as a cause of recurrent abortion. 43. To MS, Alfirevic Z, Heath VCF, et al. Cervical cerclage for prevention
J Obstet Gynaecol Br Commonw 1963; 70:105–9. of preterm delivery in women with short cervix: randomized
32. Davis G, Berghella V, Talucci M, et al. Patients with a prior failed controlled trial. Lancet 2004; 363:1849–53.
transvaginal cerclage: a comparison of obstetric outcomes with either 44. Berghella V, Odibo OA, To MS, et al. Cervix on ultrasonography:
transabdominal or transvaginal cerclage. Am J Obstet Gynecol 2000; meta-analysis of trials using individual patient-level data. Obstet
183:836–9. Gynecol 2005; 106:181–9.
33. Scibetta JJ, Sanko SR, Phipps WR. Laparoscopic transabdominal cer- 45. Berghella V, Haas S, Chervoneva I, et al. Patients with prior second-
vicoisthmic cerclage. Fertil Steril 1998; 69:161–3. trimester loss: prophylactic cerclage or serial transvaginal sonograms?
34. Anthony GS, Walker RG, Cameron AD, et al. Transabdominal cervico- Am J Obstet Gynecol 2002; 187:747–51.
isthmic cerclage in the management of cervical incompetence. 46. Groom KM, Bennett PR, Golara M, et al. Elective cervical cerclage
Eur J Obstet Gynecol Reprod Biol 1997; 72:127–30. versus serial ultrasound surveillance of cervical length in a population
35. Topping J, Farquharson RG. Transabdominal cervical cerclage. at high risk for preterm delivery. Eur J Obstet Gynecol Reprod Biol
Br J Hosp Med 1995; 54:510–12. 2004; 112:158–61.
36. Rush RW, Isaacs S, McPherson K, et al. A randomized controlled trial 47. Althuisius SM, Dekker GA, Hummel P, et al. Cervical Incompetence
of cervical cerclage in women at high risk of spontaneous preterm Prevention Randomized Cerclage Trial (CIPRACT): emergency
delivery. Br J Obstet Gynaecol 1984; 91:724–30. cerclage with bed rest versus bed rest alone. Am J Obstet Gynecol
37. Lazar P, Gueguen S, Dreyfus J, et al. Multicentred controlled trial 2003; 189:907–10.
of cervical cerclage in women at moderate risk of preterm delivery. 48. Olatunbosun OA, al-Nuaim L, Turnell RW. Emergency cerclage
Br J Obstet Gynaecol 1984; 91:731–5. compared with bed rest for advanced cervical dilatation in pregnancy.
38. Quinn M. Final report of the MRC/RCOG randomised controlled trial Int Surg 1995; 80:170–4.
of cervical cerclage. Br J Obstet Gynaecol 1993; 100:1154–5.
217
0415421306-Ch14 3/30/07 9:46 AM Page 218
0415421306-Ch15 3/30/07 9:47 AM Page 219
15. Midtrimester loss and viability

Victor YH Yu
INTRODUCTION CONTRIBUTION OF EXTREME PREMATURITY

TO PERINATAL MORTALITY
In Chapter 16 it is shown that women with recur-
rent miscarriage have a higher incidence of preterm The State of Victoria in Australia has a population
labor. In addition, uterine anomalies and cervical of about 6 million and a birth rate of about 15 per
incompetence are two causes of recurrent pregnancy 1000. The legal requirements for birth registrations
loss (RPL) that predispose to second-trimester in the State are that a stillbirth must be registered if
fetal loss. Women with recurrent second-trimester the gestation was 20 weeks or above or, if the period
fetal loss contribute disproportionately to the of gestation was not known, the birthweight was
stillbirth rate, and second-trimester delivery of live 400 g or more. Any infant, regardless of maturity or
births contributes disproportionately to the neona- birthweight, who shows any sign of life after being
tal mortality rate, thus significantly increasing the born must be registered as a live birth (and if death
overall perinatal mortality rate. However, a proactive subsequently occurred within 28 days, as a neonatal
policy of transfer in utero of high-risk pregnancies death). Regional statewide perinatal mortality
in danger of extremely preterm delivery to a tertiary figures are generated from the information thus
perinatal centre for management by maternal–fetal collected.1 For the purpose of perinatal statistics, a
medicine specialists, together with competent stillbirth is defined as a stillborn infant weighing at
resuscitation at birth and prompt initiation of least 500 g or, if the weight is not known, born after
neonatal intensive care by neonatologists, has been at least 22 weeks’ gestation. A neonatal death is
found to improve survival and quality-adjusted sur- defined as a death occurring within 28 days of birth
vival for extremely low-birthweight (ELBW) infants in an infant whose birthweight is at least 500 g or, if
born under 1000 g, including those born in the the weight is not known, an infant born after at least
second trimester between 23 and 26 weeks’ gesta- 22 weeks.
tion. Clinical protocols have been established for For the year 2002, there were a total of 62 622
the management of those infants born alive at births (of which 267 were stillbirths) in Victoria.
borderline viability, but continued advances made The perinatal mortality rate, based on the above
in the knowledge and technology in neonatal inten- definitions, was 6.9 per 1000 births (stillbirth rate
sive care will result in ongoing revisions of current 4.2 per 1000, neonatal death rate 2.7 per 1000). For
medicolegal and ethical guidelines. Principles behind the purpose of international comparison, the World
decision-making on initiating and withdrawing Health Organization (WHO) recommends the
intensive care will, however, remain interpersonal publication of a standard mortality rate in which
and intimate, respectful to the infants’ lives and their the numerator and denominator are restricted to
parents’ autonomy, and sensitive to the emotional heavier and more mature infants. A stillbirth is thus
concerns of parents and staff. defined as a stillborn infant weighing at least 1000 g
219
0415421306-Ch15 3/30/07 9:47 AM Page 220
on the long-term outcome of a population-based

Table 15.1 Contribution of low-birthweight
and preterm births to perinatal deaths in the ELBW cohort born in the State of Victoria since
State of Victoria 1979–80 and up to 14 years of age.2–7 Within the
State, there are three level III perinatal centers, each
Percentage Percentage of
with its NICU, and a fourth stand-alone NICU in a
of births perinatal deaths
children’s hospital. There are 20 level II special-care
Birthweight units and 150 level I maternity units with small
<2500 g 5.9 66 neonatal nurseries attached.
<1500 g 1.1 44
<1000 g 0.5 35
Gestation PLACE OF BIRTH AND OUTCOME
<37 weeks 6.0 67
<32 weeks 1.1 43 The three perinatal centers deliver only about one-
<28 weeks 0.4 29
quarter of births in the State of Victoria. However,
70% of ELBW births were being delivered in the
three hospitals even during these early years, indicat-
ing that there was already an effective effort being
or, if the weight is not known, born after at least made to identify women with high-risk pregnancies
28 weeks’ gestation. A neonatal death is defined as a who might deliver an ELBW infant, and they were
death occurring within 7 days of birth in an infant being referred in utero by midwives and obstetri-
whose birthweight is at least 1000 g or, if the weight cians in the community for consultation by mater-
is not known, in an infant born after at least nal–fetal medicine specialists within our perinatal
28 weeks. In Victoria, the perinatal mortality rate, centers. For the remaining 30% ELBW infants
in accordance with these WHO definitions, was who were born outside the perinatal centers,
3.2 per 1000 births (stillbirth rate 2.4 per 1000, less than half (42%) were transferred for neonatal
neonatal death rate 0.8 per 1000). Table 15.1 shows intensive care after birth; those not referred, with
that ELBW infants have a major impact on perinatal very few exceptions, died. The perinatal mortality
mortality statistics that is disproportional to their rate of ELBW infants was significantly lower in
numbers. those born in the perinatal centers compared with
those born elsewhere (72% vs 93%), as were the
stillbirth rate (36% vs 59%) and neonatal death rate
CHANGING VIABILITY IN EXTREMELY (56% vs 82%).8,9
PRETERM INFANTS The VICS study defined long-term disability as
severe if the child had cerebral palsy and was unable
Population-based studies from a designated geo- to walk, low IQ defined as a psychological test score
graphical region, rather than from a single institu- of more than two standard deviations (SD) below
tion, are essential for the assessment of the true the mean, or bilateral blindness. Not only did our
impact of maternal–fetal and neonatal intensive inborn ELBW infants have a significantly higher
care practices on the survival and long-term survival rate compared with those who were outborn,
neurodevelopmental outcome of extremely preterm but the inborn survivors also had a significantly lower
live births. Significant numbers of preterm infants severe disability rate: 15% versus 50% at 2 years, 15%
born outside perinatal centers might not be trans- versus 38% at 5 years, 13% versus 39% at 8 years, and
ferred by a neonatal emergency transport service 10% versus 30% at 14 years. Their high disability
(NETS) to institutions with a neonatal intensive care rate was attributable to suboptimal perinatal care,
unit (NICU), and they would die at their hospital which was identified in 72% of outborn survivors,
of birth. Our research group, the Victorian Infant secondary to a failure or a delay in initiating inten-
Collaborative Study (VICS), has been reporting sive care among these outborn infants.
220
0415421306-Ch15 3/30/07 9:47 AM Page 221
MIDTRIMESTER LOSS AND VIABILITY
TRANSFER IN UTERO IMPROVES OUTCOME

Table 15.2 Improving survival and quality-adjusted
(QA) survival in a population-based study of
The benefits of a more proactive transfer-in-utero extremely low-birthweight (ELBW) infants
policy to level III perinatal centers for management
Percentage of ELBW infants in the time period
were established when this early VICS regional cohort
was compared with later VICS regional cohorts born 1979–80 1985–87 1991–92 1997
in 1985–87. Not only was there a significant 50% Birthweight (n = 351) (n = 560) (n = 429) (n = 233)
improvement in ELBW survival rate in 1985–87
500–599 g:
compared with 1979–80, but there was also a signif- Survival 0 0 16 52
icant reduction in the disability rate among ELBW QA survival 0 0 13 35
survivors at our 2-year assessment: severe from 18% 600–699 g:
to 11%, moderate from 5% to 1%, and mild from Survival 7 12 38 56
QA survival 5 10 27 43
28% to 18%.10 In this study, severe disability was
700–799 g:
defined as cerebral palsy in children unlikely ever to Survival 22 33 48 83
walk, Bayley Mental Developmental Index (MDI) QA survival 14 29 43 64
below 69, or bilateral blindness; moderate disability 800–899 g:
was defined as cerebral palsy in non-ambulant chil- Survival 35 51 72 73
dren who were likely to walk or sensorineural deaf-
900-999 g:
ness requiring amplification; and mild disability Survival 43 65 79 87
was defined as cerebral palsy in ambulant children QA survival 33 53 70 76
and Bayley MDI of 69–84. Total:
The VICS study identified that the primary Survival 25 38 56 73
factor in the improved outcome was the significant
increase in the proportion of the State’s ELBW
infants born within the three perinatal centers improved adjusted survival rate, secondary to a
(from 70% to 77%). A secondary factor was a greater reduction in the proportion of survivors with
number of outborn ELBW live births who received disability.17 Another benefit that has been identified
resuscitation and prompt intensive care even at the is that, in spite of an increase in the consumption
level II hospitals prior to the arrival of NETS.11 At of hospital resources that inevitably results from
5 years of age, severe cerebral palsy was reduced from a proactive treatment policy, economic evaluation
3% to 1%, blindness from 7% to 3%, deafness from of efficiency in terms of cost–effectiveness and cost–
6% to 0.5%, IQ score <2 SD from 21% to 7%, and utility has remained unchanged.18–20
IQ score <3 SD from 9% to 4%.12 However, these
figures still show a large number of disabilities. RISK OF RECURRENT PREGNANCY LOSS AND
PREMATURITY
FURTHER IMPROVEMENT IN OUTCOME
IN RECENT ERAS Women with a history of pregnancy loss, when com-
pared with those who have delivered a live birth, are
The VICS study reported that the proportion of known to have a higher risk of pregnancy loss in
ELBW infants who were inborn at the level III previous and subsequent pregnancies. In our study
perinatal centers continued to increase to 84% of women who had delivered an ELBW infant, the
in 1991–92 and to 91% in 1997.13,14 As a result, frequency of pregnancy loss in previous pregnan-
a greater number had the benefit of proactive resus- cies was 41% and that in subsequent pregnancies
citation and prompt intensive care initiated after was 31%.21 These rates are higher than that of
birth.15,16 Table 15.2 shows that this has resulted in 10–20% reported for our general population. The
a progressively improved survival rate, as well as an perinatal mortality rate is also known to increase
221
0415421306-Ch15 3/30/07 9:47 AM Page 222
more than threefold among women with one prior The first VICS regional cohort based on a gesta-
preterm birth and at least one prior pregnancy loss. tional age cohort consisted of 316 infants consecu-
In our study of women who had delivered an ELBW tively born in the three years 1985–87, at 24–26
infant, the perinatal mortality rate of their subse- weeks’ gestation.26 Gestational age was calculated
quent pregnancies was 51.7 per 1000 births. This from dates obtained by menstrual history, usually
was four times higher than that reported for our confirmed by ultrasound before 20 weeks’ gesta-
general population in the same time period (12.8 tion. Of the 95 five-year-old survivors, one was
per 1000 births of at least 20 weeks’ gestation and untraced but was assessed at two years to be free of
400 g birthweight). We also know that significantly disability. The overall survival rate to 5 years was
more infants are born preterm when there is a 30% and the severe disability rate among survivors
previous history of perinatal loss or prematurity. was 11% (Table 15.3). There was no trend toward
In our study of women who had delivered an ELBW increasing disability with lower gestational age.
infant, the prematurity rate was 28% and the low- Cerebral palsy was diagnosed in 13%, bilateral blind-
birthweight rate was 34% in subsequent pregnan- ness in 5%, deafness requiring hearing aids in 2%,
cies, which was about six times higher than that of and IQ more than 2 SD below the mean in 7%. These
the general population. Those women who had a outcome data were mostly favorable, and better than
diagnosis of cervical incompetence were at the those reported in other contemporaneous regional
highest risk of a subsequent preterm birth. The low- cohorts born in other parts of the world.
birthweight rates among live births subsequent to Postnatal surfactant replacement therapy was
the birth of an ELBW infant in our study were 36% introduced for routine clinical use in 1991 in Victoria.
less than 2500 g, 11% less than 1500 g, and 5% less It has been proven to reduce mortality in randomized
than 1000 g. controlled trials within NICUs, but regional data
are vital to assess the impact of such a therapeutic
OUTCOME ACCORDING TO GESTATIONAL AGE innovation on a whole population. Therefore,
in our post-surfactant era (1991–92), VICS studied
The use of birthweight as a framework for the 401 infants consecutively born at 23–27 weeks’ ges-
reporting of outcome data is a convenient system tation in Victoria.27 Of the 225 two-year extremely
for neonatologists, who have an accurate measure- preterm survivors, 219 (97%) were assessed, in addi-
ment on which to base the study. However, gestation to 242 contemporaneous normal-birthweight
tional age, not birthweight, is the parameter used by
obstetricians as a guide to critical decisions on the
management of the mother and fetus. A proactive
Table 15.3 Extremely preterm infants: survival
attitude among physicians in recent years has
and disability rates among survivors in three eras
improved the survival prospects even among
extremely preterm births of less than 26 weeks’ ges- 1985–87 1991–92 1997
tation.22 There is a tendency to underestimate birth-
Survival rate (n = 316) (n = 401) (n = 208)
weight in preterm infants before birth, and the 23 weeks 0% 10% 41%
perinatal mortality of those with clinical underesti- 24 weeks 12% 33% 41%
mation of birthweight is known to be higher than 25 weeks 28% 58% 73%
26 weeks 45% 72% 88%
that of those with correctly estimated birthweight. 27 weeks — 77% 86%
Therefore, studies with gestation as an independent Degree of disability (n = 95) (n = 219) (n = 148)
variable in determining outcome are necessary to Severe 11% 8% 16%
assist obstetricians, neonatologists, and parents in Moderate 7% 13% 12%
Mild 25% 25% 24%
their decision-making process, especially prior to None 61% 54% 49%
an extreme preterm birth.23–25
222
0415421306-Ch15 3/30/07 9:47 AM Page 223
controls, in whom 2% were found to have severe surfactant therapy plus the other factors associ-
disability. Compared with our regional 1985–1987 ated with survival per se. On day 28 and at hospital
cohort from the pre-surfactant era, the survival rate discharge, variables significantly associated with a
had improved significantly (Table 15.3) with no sig- lower rate of survival free of major disability were
nificant change in their severe disability rate (20% at grade 3 or 4 cerebroventricular hemorrhage, cystic
23 weeks 14% at 24 weeks, 6% at 25 weeks, 9% at periventricular leukomalacia, postnatal dexametha-
26 weeks, and 1% at 27 weeks). The rate of blind- sone therapy, and surgery. Almost half (47%) of the
ness was, however, significantly lower in 1991–92 extremely preterm survivors had none of these
(from 5% to 2%). adverse prognostic variables, and their major dis-
When parents are counseled at different time ability rate was 7%, which was not significantly dif-
periods before and after the birth of their extremely ferent from the rate of 3% for normal-birthweight
preterm infant, they wish to know not only whether children. The risks of major disability increased to
their child will survive but also whether their child 17% with one adverse variable, 47% with two vari-
will survive with or without disability. The 1991–92 ables, and 67% with three variables.
VICS regional cohort of infants born at 23–27 weeks’ The most recent VICS regional cohort based
gestation was assessed again at 5 years of age to on gestational age consisted of 208 consecutive
allow a more certain estimate of disability, and to live births at 23–27 weeks’ gestation and 188 con-
determine how the prognosis offered to parents temporaneous normal-birthweight controls born in
changed with increasing postnatal age and when dif- Victoria during 1997.29 Compared with our regional
ferent perinatal variables were taken into account.28 1991–92 cohort, the survival rate to 2 years of age
Of the 401 extremely preterm infants 67 (17%) were had improved at each week of gestation, with no sig-
born outside level III perinatal centers. Their place nificant changes in the disability rate (Table 15.3).
of birth was a significant factor for survival: 62% for Since the gain from a significant increase in survival
inborn infants and 28% for outborn infants. The was greater than the loss from a marginal increase
attitude of the attending physician determined in disability among survivors, the rate of survival
whether or not intensive care was offered. Overall, free of disability was higher in 1997 compared with
16% of live births at 23–27 weeks were not offered 1991–92, both overall and in all gestational age sub-
intensive care and died on the first day: 69% at groups. There was no gestational age below which
23 weeks, 35% at 24 weeks, 6% at 25 weeks, 2% at most survivors were disabled.
26 weeks, and 1% at 27 weeks. Of inborn infants, 9%
were not offered intensive care, compared with 54% COMPARISON WITH OTHER REGIONAL STUDIES
of outborn infants. Variables that were associated
positively with survival on day 1 were increasing In a review of survival rates for extremely preterm
maturity, antenatal corticosteroid therapy, multiple infants born in North America, none of the studies
births, female sex, and not being small for gesta- had regional or population-based cohorts that
tional age, and on day 7, grade 3 or 4 cerebroven- could be directly compared with our VICS study.30
tricular hemorrhage. Outcome data were available Two regional cohorts reported from the UK had
for 221 (98%) survivors (and 245 contemporane- survival rates at individual weeks’ of gestation lower
ous normal-birthweight controls): non-ambulatory than that in our study. In the Trent region of England,
cerebral palsy was diagnosed in 7%, bilateral blind- with an annual birth rate similar to that Victoria
ness in 2%, deafness requiring hearing aids in 1%, (60 000), the survival rates to discharge home in a
and IQ score more than 2 SD below the mean for cohort from 1994 to 1997 were 14% at 23 weeks, 26%
the normal-birthweight controls in 15%. Variables at 24 weeks, 41% at 25 weeks, 61% at 25 weeks, 75%
that were associated positively with survival free at 27 weeks, and 85% at 28 weeks.31 For the same
of major disability on day 1 were postnatal time period, survival rates from our VICS cohort
223
0415421306-Ch15 3/30/07 9:47 AM Page 224
decisions have to be made after the birth of a live-

Table 15.4 Comparison between the population-
based VICS and EPICure data for infants born at born infant to either withhold or withdraw inten-
23–25 weeks’ gestation sive care.36 Studies have shown great variability in
doctors’ attitudes and their management policies
VICS 1991–92 EPICure 1995 VICS 1997
for extreme prematurity. There is a tendency for
Survival rate 38% 29% 69% both obstetricians and neonatologists to underesti-
23 weeks 10% 10% 41% mate the potential for survival and overestimate the
24 weeks 33% 26% 41% risks of disability for extremely preterm infants.37–39
25 weeks 58% 43% 73%
Many neonatologists continue to selectively resusci-
Disability rate 8% 15% 13%
tatd extremely preterm infants at birth, which means
that liveborn infants are left to die through with-
holding of intensive care. If doctors believe that the
were 25% at 23 weeks, 46% at 24 weeks, 79% at infant has little prospect for survival or survival
25 weeks, 85% at 26 weeks, 82% at 27 weeks, and without disability, it is probable that their clinical
91% at 28 weeks. The EPICure study reported the management would be delayed or less than opti-
outcome at 21/2 years of 23–25-week gestation live mal, and may in fact be creating a self-fulfilling
births born in the UK in 1995.32 Comparative data prophecy.40 An Australian survey in the 1980s had
from the VICS study of cohorts born both prior shown at that time that a great number of neonatol-
to the EPICure cohort (1991–92) and after the ogists selectively resuscitated extremely preterm
EPICure cohort (1997) are shown in Table 15.4. The infants at birth, suggesting that many of these
1991–92 Australian cohort, which was 3 years live births were left to die through withholding of
before the 1995 UK cohort, already had better neonatal intensive care.37 More recent national
survival, and the survival rate of the infants born at surveys conducted in 2000 showed a more proactive
23–25 weeks’ gestation in the 1997 Australian resuscitation policy among Australian obstetricians
cohort was more than double that of the UK cohort and neonatologists (Table 15.5).38,39
(69% vs 29%). Hospital survival rates reported
from a regional cohort from the Netherlands relat- DECISION TO WITHHOLD INTENSIVE CARE
ing to infants born in 1995 were extremely poor.
The majority of deaths in that study at 23 and In the majority of level III perinatal centers within
24 weeks occurred before admission to the NICU: developed countries, all infants with a birthweight
2% at 23 weeks, 3% at 24 weeks, 29% at 25 weeks, of more than 500 g or a gestation of 24 weeks or
and 54% at 26 weeks.33 Disability rates in regional more are offered intensive care. At Monash Medical
cohorts of infants born at 23–25 weeks’ gestation Centre (MMC) in Australia, we have reported that
were 15% in the EPICure study,32 35% in the Trent 10% of 442 extremely preterm live births born at
region,34 and 45% in the Netherlands.35 In compar- 23–28 weeks’ gestation over the 10-year period
ison, the rates reported in the VICS study were 8% 1977–86, were not offered intensive care, 4% had
in our 1991–92 cohort and 13% in our 1997 cohort,
despite a significantly higher survival rate, especially
in the 1997 cohort (Table 15.4).
Table 15.5 Percentage of Australian doctors who
would recommend to parents that their extremely
preterm infants be resuscitated at the time of birth
ETHICAL DILEMMAS FOLLOWING EXTREMELY
PRETERM BIRTH 22 weeks 23 weeks 24 weeks 25 weeks
Obstetricians 9% 33% 59% 79%

Ethical problems of selective non-treatment arise in Neonatologists 13% 48% 92% 100%
caring for extremely preterm infants when clinical
224
0415421306-Ch15 3/30/07 9:47 AM Page 225
obvious major malformations, and 6% were con- been advocated as an acceptable and preferred
sidered ‘non-viable’, for which resuscitation at birth mode of operation in the NICU, one that has been
was not offered or was not successful.41 The propor- endorsed by the Canadian Pediatric Society and the
tion of live births in which treatment was withheld American Academy of Pediatrics.45,46 The attending
at the time of delivery was 37% at 23 weeks, 17% neonatologist has the primary role as an advocate
at 24 weeks, 8% at 25 weeks, 1% at 26 weeks, 1% at for the infant and medical advisor to the parents,
27 weeks, and 0% at 28 weeks. This approach to whereas the parents act as surrogates for their
offering intensive care was considered ahead of its infant. The shift in emphasis from curative to pallia-
time, even in developed countries 20–30 years ago.42 tive treatment requires consensus among all those
During an identical period, 1977–86, in another involved in the care of the infant, both medical and
level III perinatal center within a few kilometers of nursing staff, as well as consent from the parents,
MMC, 42% of similar live births born at 24–26 weeks’ who should be closely involved in this widely shared
gestation were not offered intensive care, all of decision-making process. At MMC, over the 8-year
whom died.41 This accounted for a lower survival period 1981–87, intensive care was withdrawn prior
rate among their infants born at 23–28 weeks’ to death in 65% of 316 deaths.47 Among these
gestation compared with those in MMC (29% infants, death was considered to be inevitable in the
vs 44%), as the survival rate among those who were short term, even with the continuation of neonatal
offered intensive care was similar. Our practice intensive care in 70% of the cases. In the remainder,
during the 1990s was consistent with what the the risk of severe brain damage was considered to be
Royal College of Paediatrics and Child Health in so great that death was considered preferable to a
the UK published in 1997, which stated that it life with major disability. Therefore, in our NICU,
would not be unreasonable to consider withhold- full treatment until death is uncommon and
ing treatment in an infant born at 23 weeks’ weigh- occurred in only one-third of cases. This experience
ing little more than 500 g.43 There is a general was not unique, as studies from the UK and
consensus in developed countries, even to this day, New Zealand have shown that 30–80% of deaths in
that parents of a 22-week infant should be discour- their NICUs follow a deliberate withdrawal of life-
aged from seeking active treatment, whereas those sustaining treatment.48,49
of a 25–26-week infant should be encouraged to There are three clinical situations in which selec-
consent to intensive care.44 It is reasonable not to tive withdrawal of intensive care is appropriate.
offer resuscitation for all 23–24-week infants, who
should be assessed on an individual basis at the ● Firstly, there are few who would disagree that
time of birth. withdrawal of intensive care is morally and ethi-
cally acceptable when death is considered to be
DECISION TO WITHDRAW INTENSIVE CARE inevitable and the infant is in the process of
dying, whatever treatment is provided. Intensive
However, a proactive policy to initiate intensive care would be considered in these cases a futile
care must take into consideration that a decision exercise and not in the best interest of the infant.
to withdraw such care might have to be made in Examples in this category include those infants
selective infants at a later stage in the course of the with severe respiratory failure or fulminating
treatment. In the event that the infant’s subsequent sepsis who have persistent or worsening hypox-
clinical course indicates that further curative emia, acidosis, and hypotension unresponsive
efforts are futile or lack compensating benefit, to ventilatory and inotropic support. There is no
intensive care should be discontinued and pallia- obligation to provide futile medical care in such
tive care, which provides symptomatic relief and cases, as no infant with progressive multiple
comfort, should be introduced. This approach, organ failure survives even with the provision of
termed ‘individualized prognostic strategy’ has cardiopulmonary resuscitation.
225
0415421306-Ch15 3/30/07 9:47 AM Page 226
● Secondly, it is appropriate also to consider MEDICOLEGAL PERSPECTIVE

withdrawal of intensive care even when death is
not inevitable with continued treatment, but Very few cases of selective non-treatment have
where there is a significantly high risk of severe reached the courts. It is considered appropriate
physical and mental disability should the infant for these difficult decisions to be made within
survive. Such a decision should not raise too the context of the infant/neonatologist/parent rela-
many moral and ethical problems if the infant’s tionship, and experience has shown that there is no
development of self-awareness and intentional excessive abuse in such private decision-making
action is believed to be virtually impossible or processes. The legal position appears to recognize
there is no prospect of the infant ever being able the importance of respecting parental decisions, but
to act on his or her own behalf. One scenario is emphasizes that the law court has the right to inter-
that of an extremely preterm infant with large vene and overrule a decision if that is necessary to
bilateral parenchymal hemorrhages, infarcts, protect the best interests of the infant. The British
and/or leukomalacia in the brain. legal system, for example, had upheld selective
● Thirdly, a more controversial issue is when sur- non-treatment in the three categories of neonatal
vival with moderate disability is possible with conditions referred to previously. Firstly, selective
treatment, but the infant is likely to suffer per- non-treatment was ruled to be legally acceptable
sistent pain, require recurrent hospitalization when death was inevitable in the case of a hydro-
and invasive treatment throughout life, and to cephalic preterm infant on the verge of death.
experience early death in childhood or early Secondly, legal precedence for selective non-
adulthood. This situation may arise with a treatment for an infant with severe brain damage,
preterm infant with severe chronic lung disease who was neither dying nor in severe pain, was
non-responsive to dexamethasone and with no found in a case presenting to court with a high risk
prospect of being weaned from mechanical ven- of multiple sensorineural disabilities. Thirdly, selec-
tilation, but for whom lung transplant is still tive non-treatment was considered lawful in an
considered an experimental option. infant where the benefits of life with treatment
failed to outweigh the burdens of a ‘demonstrably
The one principle shared by all the guidelines awful life’ of pain and suffering.
proposed in the UK, Canada, USA, and Australia is
that if continued life for the infant with treatment is THE DECISION-MAKING PROCESS
a worse outcome than death, then the principle of
primum non nocere imposes a professional, moral, The importance of less medical paternalism and
and humanitarian duty upon neonatologists to more informed parental involvement in the deci-
withhold or withdraw life-sustaining treatment. sion-making process of selective non-treatment
Infants cannot benefit from such treatment, and must be emphasized. The neonatologist should
death is not the worst outcome for them if they never make unilateral decisions regarding the right
cannot be rescued from irreversible medical deteri- to die. Adequate and consistent parental communi-
oration and death, cannot have life prolonged cation carried out by medical and nursing staff
without major sensorineural sequelae, and cannot must begin with the admission of all infants into
be relieved of ongoing pain and suffering. When the the NICU so that trust can be developed between
process of dying is being artificially prolonged, most the parents and staff, irrespective of outcome.50 An
would agree that the harm of continued treatment open-visiting policy for families is essential to pro-
exceeds any potential benefit. However, decisions mote such parental contact.51 A realistic assessment
based on quality-of-life considerations are more dif- of the infant’s clinical condition should be given by
ficult, as there is inevitably imprecision in predicting the neonatologist to the parents as soon as possible.
the risk of intolerable disability or suffering. The medical facts should be presented with an
226
0415421306-Ch15 3/30/07 9:47 AM Page 227
honest, sympathetic, and caring attitude. Often, the and guidelines for palliative care demand that basic
information has to be repeated and reinforced by nursing care should continue, with the emphasis on
the entire staff. Otherwise, misunderstandings and providing comfort to the infant. Electronic moni-
unrealistic expectations can lead to confusion, sus- toring of physiological parameters, diagnostic
picions, bitterness, and frank hostility. As with most investigations (such as X-rays and blood tests),
medical decisions made by neonatologists that medications (including oxygen and antibiotics),
require parental informed consent, much of the dis- and therapeutic procedures (including resuscita-
cussion on selective non-treatment depends on tion, all forms of assisted ventilation, and intra-
trust in the knowledge, judgment, and integrity of venous infusion) that might prolong the dying
the doctor. When a consensus has been reached by process should be discontinued. Prolonged terminal
the NICU staff that selective non-treatment is an weaning, defined as a stepwise or gradual decreas-
appropriate option to raise with the parents, one or ing of ventilator support over a period of hours, is
more intense and intimate meetings will be considered inappropriate. Dragging out the with-
required so that the crucial set of discussions can drawal serves only to prolong the dying process and
take place and in which a decision can be reached any attendant suffering. The argument that the
on the matter. These meetings usually involve both sudden withdrawal of ventilator support resembles
of the parents, the attending neonatologist, a nurse an intentional killing does not hold merit, as in both
representative, and a non-medical staff member cases a treatment on which the infant depends for life
who can act as the parents’ advocate, such as a med- is being discontinued and death is the expected out-
ical social worker. Ways of minimizing the chances come. The infant should be nursed in a normal cot
of unresolved disagreements and of maximizing the and warmth provided by light clothing. If the infant
chances of a just and ethical conclusion have has apparent distress, symptomatic relief should be
recently been reviewed.52 provided, such as suctioning to remove oropharyn-
The principles underlining clinical practice and geal secretions and sedation with normal therapeu-
the decision-making process should be the same for tic doses of morphine, on a PRN basis, even if the
developed and developing countries, but there must pain relief measures may inadvertently shorten the
be less medical paternalism and more informed dying process.
parental involvement in developing countries. A controversial issue involves the withdrawal of
Compared with developed countries, communica- enteral nutrition and hydration during palliative
tions between the medical and nursing staff and the care. Preterm or sick infants require gavage feeding,
parents are less adequate in developing countries. In and although it has been advocated that this feeding
most developed countries, intensive care is routinely method is part of medical treatment and should
offered to all who have reached 24 weeks’ gestation. therefore be discontinued during palliative care,
Limited resources in developing countries, however, others consider it as basic nursing care that must not
necessitate a different intervention point, which be withheld under any circumstances.54 A number
may be 26 weeks or even 28 weeks.53 In RPL, addi- of court decisions have supported the withdrawal of
tional factors need to be taken into consideration, as nutrition, thus equating the administration of artifi-
the parents may have undergone numerous losses cial nutrition with other medical procedures.55
before reaching this stage. They may be subject to Precedence has been set in a British court on the
further early losses, and may not reach this stage legality of withholding gavage feeding. Nevertheless,
again. most neonatologists would be reluctant not to pro-
vide gavage feeding, even when it might be lawful
PALLIATIVE CARE and appears to be in the infant’s best interest.
There is an obvious perception of a moral differ-
The neonatologist’s duty does not end with the ence between withdrawing ventilatory support
decision for selective non-treatment. The principles and withholding fluids or nutrition with selective
227
0415421306-Ch15 3/30/07 9:47 AM Page 228
non-treatment. The underlying principle is that institution-based studies. Such data are vital for
naturally or artificially administered hydration and answering questions such as ‘How low should we
nutrition may be given or withheld, depending on go?’ Quality outcomes depend more on the com-
the infant’s comfort. prehensive organization of an effective system of
Parents need a quiet place to be with their infant networking perinatal/neonatal services within a
during the dying process. They may wish that other geographically determined region than on the
family members and religious advisors be present. introduction of expensive high-technology thera-
Hospice concepts have been applied to neonatal care pies within individual NICUs.
by providing a family room that is private yet close Among the many neonatal ethical problems, the
to the NICU and by training NICU staff in more one that neonatologists are faced with on a regular
supportive approaches towards the families.56,57 basis involves the issue of selective non-treatment,
Such a program allows the staff to cope better with that is, clinical decisions made after the birth of a
the dying infants offered selective non-treatment liveborn infant to either withhold or withdraw
and facilitates the grieving process in the parents. treatment in certain clinical situations. If medical
Under certain circumstances, withdrawal of inten- doctors believe that the infant has little prospect for
sive care may be arranged to take place in the home, intact survival, their management would be subop-
so that death can occur in a more comforting timal and they create a self-fulfilling prophecy.
environment for the family. A policy establishing criteria for initiating life-
sustaining treatment must be developed with
proper consideration of the cultural, social,and eco-
CONCLUSIONS nomic factors operating in the developed or devel-
oping country. There are infants whose subsequent
A proactive policy of resuscitation at birth and clinical course after initiation of neonatal intensive
prompt initiation of intensive care have been shown care will indicate that further curative efforts are futile
to be associated with an improvement in the sur- or lack compensating benefit. A policy establishing
vival of extremely preterm infants, including those criteria for withdrawing life-sustaining treatment
born in the second trimester, in regional popula- must also be developed, to allow the appropriate use
tion-based studies within the State of Victoria in of palliative care in these instances. The clinical
Australia. As a greater percentage of live births were situations in which selective non-treatment is
offered intensive care in our series of studies,which taking place in the neonatal intensive care unit are
spanned over 20 years, the survival rate rose pro- (1) when death is considered to be inevitable what-
gressively in all birthweight and gestation sub- ever treatment is provided, (2) even when death is
groups among ELBW infants, including those who not inevitable, there is a significantly high risk of
were born at borderline viability down to 23 weeks’ severe physical and mental disability should the infant
gestational age. Their quality-adjusted survival rate survive, and (3) when survival with moderate disabil-
also rose progressively, since the large gains in survival ity is possible, but the infant is likely to experience
over time had not been offset by significant increases ongoing pain and suffering, repeated hospitalization
in survival with disability. Cost–effectiveness and and invasive treatment, and early death in childhood.
cost–utility ratios remained stable overall, with
efficiency gains in the smaller infants over time,
REFERENCES
as more such infants were being transferred in utero
and were born in level III perinatal centers with 1. Consultative Council on Obstetric and Paediatric Mortality and
the regionalization of perinatal/neonatal healthcare Morbidity. Annual Report for the Year 2002. Melbourne: Health
programs. Multicentered collaboration to conduct Department of Victoria, 2004.
2. Kitchen WH, Campbell N, Drew JH, et al. Provision of perinatal serv-
long-term studies of geographically defined cohorts ices and survival of extremely low birth weight infants in Victoria. Med
provides unique information not available from J Aust 1983; 2:314–18.
228
0415421306-Ch15 3/30/07 9:47 AM Page 229
3. Kitchen WH, Ford F, Orgill A, et al. Outcome of extremely low birth- 25. Yu VYH, ed. The Extremely Tiny Baby. London: WB Saunders, 1996.
weight infants in relation to the hospital of birth. Aust NZ J Obstet 26. Victorian Infant Collaborative Study Group. Outcome to five years of
Gynaecol 1984; 24:1–5. age of children 24–26 weeks’ gestational age born in the State of
4. Kitchen WH, Ford F, Orgill A, et al. Outcome of infants with birth Victoria. Med J Aust 1995; 163:11–14.
weight 500 to 999 gm: a regional study of 1979 and 1980 births. 27. Victorian Infant Collaborative Study Group. Outcome at 2 years
J Paediatr 1984; 104:921–7. of children 23–27 weeks’ gestation born in Victoria in 1991–92.
5. Kitchen WH, Ford F, Orgill A, et al. Outcome of infants of birth J Paediatr Child Health 1997; 33:161–5.
weight 500 to 999 g: a continuing regional study of 5-year-old sur- 28. Victorian Infant Collaborative Study Group. Outcome at 5 years of
vivors. J Pediatr 1987; 111:761–6. age of children 23–27 weeks’ gestation: refining the prognosis.
6. Victorian Infant Collaborative Study Group. Eight-year outcome in Pediatrics 2001; 108:134–41.
infants with birth weight 500 to 999 grams: continued regional study 29. Victorian Infant Collaborative Study Group. Neonatal intensive
of 1979 and 1980 births. J Pediatr 1991; 118:761–7. care at borderline viability – Is it worth it? Early Hum Dev 2004;
7. Victorian Infant Collaborative Study Group. Outcome at 14 years 80:103–13.
and birth weight 1000 g – a regional study. Arch Dis Child 2001; 30. Lorenz JM. The outcome of extreme prematurity. Semin Perinatol
85:F159–64. 2001; 25:348–59.
8. Lumley J, Kitchen WH, Roy RND, et al. The survival of extremely low 31. Draper ES, Manktelow B, Field DJ, et al. Prediction of survival
birth weight infants in Victoria: 1982–85. Med J Aust 1988; 149:242–6. for preterm births by weight and gestational age: retrospective popu-
9. Lumley J, Kitchen WH, Roy RND, et al. Method of delivery and resus- lation based study. BMJ 1999; 319:1093–7.
citation of very low birth weight infants in Victoria: 1982–85. Med J 32. Wood NS, Marlow N, Costeloe K, et al. Neurologic and developmen-
Aust 1990; 152:143–6. tal disability after extremely preterm birth. EPICure Study Group.
10. Victorian Infant Collaborative Study Group. Impact of improved care N Engl J Med 2000; 343:378–84.
for infants of birth weight under 1000 g. Arch Dis Child 1991; 66:765–9. 33. den Ouden AL, van Baar AL, Dorrepaal CA, et al. Overlevingskans van
11. Victorian Infant Collaborative Study Group. Improving the quality zeer immature pasgeborenen in Nederland. Tijdschr Kindergeneeskd
of survival for infants of birth weight <1000 g born in non-level III 2000; 142:241–6.
centres in Victoria. Med J Aust 1993; 158:24–7. 34. Bohin S, Draper ES, Field DJ. Health status of a population of
12. Victorian Infant Collaborative Study Group. Neurosensory outcome infants born before 26 weeks’ gestation derived from routine data
at 5 years and extremely low birth weight. Arch Dis Child 1995; collected between 21 and 27 months post-delivery. Early Hum Dev
73:F143–6. 1999; 55:9–18.
13. Victorian Infant Collaborative Study Group. Improved outcome into 35. Rijken M, Stoelhorst GMSJ, Martens SE, et al. Mortality and neuro-
the 1990s for infants of birth weight 500–999 g. Arch Dis Child 1997; logic, mental, and psychomotor development at 2 years in infants
77:F91–4. born less than 27 weeks’ gestation: the Leiden Follow-Up Project on
14. Victorian Infant Collaborative Study Group. Changing outcome for Prematurity. Pediatrics 2003; 112:351–8.
infants of birth weight 500–999 g born outside level III centres in 36. Yu VYH. The extremely low birth weight infant: an ethical approach
Victoria. Aust NZ J Obstet Gynaecol 1997; 37:253–7. to treatment. Aust Paediatr J 1987; 23:97–103.
15. Victorian Infant Collaborative Study Group. Changing availability of 37. deGaris C, Kuhse H, Singer P, et al. Attitudes of Australian neonatal
neonatal intensive care for extremely low birth weight infants in paediatricians to the treatment of extremely preterm infants. Aust
Victoria over two decades. Med J Aust 2004; 81:136–9. Paediatr J 1987; 23:223–6.
16. Victorian Infant Collaborative Study Group. Evaluation of neonatal 38. Mulvey S, Partridge JC, Martinez AM, et al. The management of
intensive care for extremely low birth weight infants in Victoria over extremely premature infants and the perceptions of viability and
two decades: I. Effectiveness. Pediatrics 2004; 113:505–9. parental counselling practices of Australian paediatricians. Aust NZ J
17. Yu VYH, Doyle LW. Regionalised long-term follow-up. Semin Neonatol Obstet Gynaecol 2001; 41:269–73.
2004; 9:135–44. 39. Munro M, Yu VYH, Partridge JC, et al. Antenatal counselling, resusci-
18. Victorian Infant Collaborative Study Group. The cost of improving tation practices and attitudes among Australian neonatologists
the outcome for infants of birth weight 500 to 999 g in Victoria. towards life support in extreme prematurity. Aust NZ J Obstet
J Paediatr Child Health 1993; 29:56–62. Gynaecol 2001; 44:275–80.
19. Victorian Infant Collaborative Study Group. Economic outcome for 40. Martinez AM, Partridge JC, Yu VYH, et al. Physician counseling prac-
intensive care of infants of birth weight 500–999 g born in Victoria in tices and decision-making for extremely low birth weight infants in
the post-surfactant era. J Paediatr Child Health 1997; 33:202–8. the Pacific Rim. J Paediatr Child Health 2005; 41:209–14.
20. Victorian Infant Collaborative Study Group. Evaluation of neonatal 41. Yu VYH, Gomez JM, Shah V, et al. Survival prospects of extremely
intensive care for extremely low birth weight infants in Victoria over preterm infants: a 10-year experience in a single perinatal centre. Am
two decades: II. Efficiency. Pediatrics 2004; 113:510–14. J Perinatol 1992; 9:164–9.
21. Yu VYH, Davis NG, Mercado MF, et al. Subsequent pregnancy follow- 42. Yu VYH. Selective non-treatment of newborn infants. Med J Aust
ing the birth of an extremely low birth-weight infant. Aust NZ J 1994; 161:627–9.
Obstet Gynaecol 1986; 26:115–19. 43. RCPCH Ethic Advisory Committee. Withholding or Withdrawing
22. Yu VYH, Doyle LW. Survival and disabilities in extremely tiny babies. Life Saving Treatment in Children. A Framework for Practice.
Semin Neonatol 1996; 1:257–65. London: Royal College of Paediatrics and Child Health, 1997.
23. Yu VYH, Wood EC, eds. Prematurity. Edinburgh: Churchill 44. Rennie JM. Perinatal management at the lower margin of viability.
Livingstone, 1987. Arch Dis Child 1996; 74:F214–18.
24. Yu VYH, Carse EA, Charlton MP. Outcome of infants born at less 45. Fetus and Newborn Committee, Canadian Pediatric Society;
than 26 weeks’ gestation. In: McIntosh N, Hansen T, eds. Current Maternal–Fetal Medicine Committee, Society of Obstetricians and
Topics in Neonatal Care. Philadelphia: WB Saunders, 1996:67–84. Gynecologists of Canada. Management of the woman with threatened
229
0415421306-Ch15 3/30/07 9:47 AM Page 230
birth of an infant of extremely low gestational age. CMAJ 1994; 52. Tripp J, McGregor D. Withholding and withdrawing of life sustaining
151:547–53. treatment in the newborn. Arch Dis Child 2006; 91:F67–71.
46. AAP Committee on Bioethics. Ethics and the care of critically ill 53. Partridge JC, Martinez AM, Nishida H, et al. International compari-
infants and children. Pediatrics 1996; 98:149–52. son of care for very low birth weight infants: parents’ perception of
47. Carse EA, Yu VYH. Deaths following withdrawal of treatment in a counseling and decision-making. Pediatrics 2005; 116:263–71.
neonatal intensive care unit. In: Wiknjosastro GK, ed. Proceedings of 54. Doyal L, Wilsher D. Towards guidelines for withholding and with-
the 5th Congress of the Federation of Asia–Oceania Perinatal drawal of life prolonging treatment in neonatal medicine. Arch Dis
Societies. Denpasar: Perinatal Society of Indonesia, 1988:55. Child 1994; 70:F66–70.
48. Whitelaw A. Death as an option in neonatal intensive care. Lancet 55. Mirale ED, Mahowald MB. Withholding nutrition from seriously ill
1986; ii:328–31. newborn infants: a parent’s perspective. J Pediatr 1988; 113:262–5.
49. Kelly NP, Rowley SR, Harding JE. Death in neonatal intensive care. 56. Yu VYH. Death as an option in the neonatal intensive care unit. The
J Paediatr Child Health 1994; 30:419–22. ethics of withdrawal of life-support. In: Burrows GD, Petrucco OM,
50. Yu VYH. Caring for parents of high-risk infants. Med J Aust 1977; Llewellyn-Jones D, eds. Psychosomatic Aspects of Reproductive
2:534–7. Medicine and Family Planning. Melbourne: York Press, 1987:112–19.
51. Yu VYH, Jamieson J, Astbury J. Parents’ reactions to unrestricted 57. Whitfield JM, Siegel RE, Glicken AD, et al. The application of hospice
parental contact with infants in the intensive care nursery. Med J Aust concepts to neonatal care. Am J Dis Child 1982; 136:421–4.
1981; 1:294–6.
230
0415421306-Ch16 3/30/07 9:48 AM Page 231
16. Obstetric outcomes after

recurrent pregnancy loss
Howard JA Carp
obtained from one of the author’s own databases,

INTRODUCTION
which contains information on 1800 patients attend-
ing the Recurrent Miscarriage Clinic of the Sheba
Most work on recurrent pregnancy loss (RPL) has
Medical Center, Tel Hashomer, Israel. The figures
concentrated on the causes, prognosis, treatment,
were entered into a computerized database (SPSS,
and subsequent live birth rate. However, a question
Chicago, Ill) and analyzed.
arises as to whether this group of patients is at
a higher risk for obstetric complications such as
STATISTICAL ANALYSIS
bleeding, fetal anomalies, preeclampsia, intrauter-
ine growth restriction (IUGR), preterm labor,
Odds ratios (OR) with 95% confidence intervals
and perinatal mortality, and whether prenatal care
(CI) were calculated for developing obstetric com-
has to be modified to seek these complications.
plications such as vaginal bleeding, anomalies,
The question then arises as to whether these com-
preeclampsia, IUGR, and perinatal mortality. When
plications are associated with specific conditions
true incidences were available from cohort studies,
associated with RPL. Antiphospholipid syndrome
the relative risks (RR) were calculated. These figures
(APS) and hereditary thrombophilias are two such
were also compared in subgroups of patients and
conditions that have been described to be associated
after various treatment interventions.
with recurrent pregnancy loss and late obstetric
complications. Various interventions have also
been reported to affect the incidence of later
obstetric complications. These interventions include INCIDENCE OF OBSTETRIC
paternal leukocyte immunization and intravenous COMPLICATIONS AFTER RPL
immunoglobulin (IVIG) for unexplained RPL,
anticoagulants and aspirin or IVIG for APS, and Most of the literature on obstetric complications
anticoagulants for hereditary thrombophilias. This comes from an era after APS had been defined
chapter assesses some of the obstetric complications but before hereditary thrombophilias had been
associated with different forms of RPL and the defined. Reginald et al,1 in a retrospective observa-
treatment modalities that have been used. tional cohort study, assessed the results of 175 preg-
nancies in 97 recurrently miscarrying women
whose subsequent pregnancy progressed beyond
METHOD OF STUDY 28 weeks. However, the underlying causes of RPL
in this group of women were not documented.
Publications describing the obstetric complications The results were not compared with a control group
in RPL were sought by a thorough literature search attending the same hospital, but with standard
including online databases, MEDLINE and EMBASE. figures from Scotland between 1973 and 1979.
The original database of the Recurrent Miscarriage A significantly higher incidence of preterm deliv-
Trialists Group is held by the author as one of the eries, perinatal deaths, and IUGR was found.
data contributors to that database. Figures were also In contrast, Hughes et al2 examined the obstetric
231
0415421306-Ch16 3/30/07 9:48 AM Page 232
outcome in 88 women with a past history of three patients with RPL. The following complications were
or more consecutive pregnancy losses and com- found to be associated with RPL: advanced mater-
pared the results with a control group drawn from nal age, cervical incompetence, diabetes mellitus,
their local obstetric population. The incidences of hypertensive disorders, placenta praevia and abrup-
small-for-gestational-age infants (3.4%), preterm tio placentae, malpresentations, and premature
delivery (12.5%), and perinatal mortality (0%) rupture of the membranes (PROM). A higher rate
were no different to those in the control group. of cesarean section was also found in patients with
As in the study by Reginald et al,1 there was no previous RPL compared with controls: 15.9% and
mention of APS. However, an increased incidence 10.9%, respectively (OR 1.6; 95% CI 1.5–1.7).
of gestational diabetes and pregnancy-induced An attempt was made to determine the common
hypertension was found. Tulppala et al3 conducted OR for various late complications of pregnancy after
a prospective study of 32 deliveries in 63 women RPL compared with controls. The series of Reginald
with RPL and presented the results of a detailed et al1 could not be included as there was no relevant
investigative protocol, including APS. The inci- control group. The series of Seidman et al5 was
dences of IUGR (20%), preterm delivery (9.7%), not included as it only described patients with one
and impaired glucose tolerance (22.8%) appeared previous miscarriage. The series of Tulppala et al3
to be increased. Unfortunately, the results were not could not be included as there was no control
compared with any control population. Jivraj et al4 group. The other three publications were combined
studied a cohort of 162 women with RPL compared in a meta-analysis.
with local controls, and found an increased inci-
dence of the same complications as Reginald et al,1 VAGINAL BLEEDING
but also an increased incidence of cesarean sec-
tions, which were performed for the above obstetric The incidence of bleeding was only described in the
conditions. Although that study did define the series of Reginald et al quoted by Beard,7 and
causes of pregnancy loss in the control group, the in that of Seidman et al.5 Hence no common OR
figures were too small to allow comparisons to could be calculated. Vaginal bleeding seems to
be drawn between different groups of patients. be increased in pregnancies that develop. Vaginal
Seidman et al5 reported the same complications in bleeding is a common complication, occurring in
338 women with pregnancy losses compared with 50 of 162 women in the series of Reginald et al7
13 338 parous women in a multicenter study in four and in 50 of 102 patients in the author’s series.8 The
large hospitals in Jerusalem. In the survey by reason for this bleeding remains unclear. Seventy-
Seidman et al,5 there was a statistically significant five percent of habitual abortions are blighted ova.8
14% prevalence of preterm labor, compared with However, when the pregnancy succeeds and there is
7.5% in control women (OR 2.32; 95% CI a live embryo within the uterus, bleeding still occurs
1.67–3.22). The different prevalence of IUGR was in 40–50% of patients. Seidman et al5 also reported
even more significant: 14% in aborting women, a 13.7% incidence of first-trimester bleeding,
compared with 5% in control women (OR 3.02; compared with a 13.8% incidence in the standard
95% CI 2.17–4.18). However, the study by Seidman population.
et al5 included patients with one pregnancy loss and
was not restricted to patients with RPL. A more ANOMALIES
recent population-based study has been reported by
Sheiner et al,6 in which all singleton pregnancies Little information is available on anomalies. The
were assessed in women with and without two study by Sheiner et al6 reports 2 anomalies in
or more consecutive recurrent abortions. Between 29 patients. Although this is a very small series, the
1988 and 2002, 154 294 singleton deliveries figures are higher than expected. Analysis of the
occurred. Of these deliveries, 7503 occurred in figures in the RMITG trial9 showed an anomaly rate
232
0415421306-Ch16 3/30/07 9:48 AM Page 233
OBSTETRIC OUTCOMES AFTER RECURRENT PREGNANCY LOSS
Study Statistics for each study OR and 95% Cl
Lower Upper
OR limit limit p
Hughes et al2 7.001 3.977 12.323 0.000

Jivraj et al4 2.335 0.739 7.379 0.149
Sheiner et al6 2.233 2.073 2.406 0.000
Common OR 2.277 2.116 2.451 0.000
0.1 1 10 100
Figure 16.1 Odds ratio (OR) for gestational diabetes in recurrent pregnancy loss.
of 4%. In the author’s series, there were three anom- significant (95% CI 2.1–2.45). Tulppala et al3 also
alies in 99 developing pregnancies in non-treated found a prevalence of 22.6% (7 of 31 patients tested).
patients. However, in the RMITG series and the
author’s series, no control group is available. PREGNANCY-INDUCED HYPERTENSION
DIABETES Hughes et al,2 Jivraj et al,4 and Sheiner et al6 quoted

the incidence of pregnancy-induced hypertension
Three papers have described the prevalence of (PIH). The figures are summarized in Figure 16.2.
diabetes: Hughes et al,2 Jivraj et al,4 and Sheiner In order to obtain significant numbers, the figures
et al.6 As a control group was available, their for preeclampsia and other forms of PIH were
common OR could be determined for 7753 patients analyzed as a whole. The common OR was
with RPL and 172 490 control patients. The preva- 1.89 (95% CI 1.74–2.06), which was statistically sig-
lence of diabetes was 11.75% and 4.95% in recur- nificant. The total number of patients was identical
rently miscarrying and control patients, respectively. to those with gestational diabetes. The prevalence of
The OR are summarized in Figure 16.1. As can be PIH was 671 of 7753 patients (8.6%) of recurrently
seen, there was a common OR of 2.27 for gesta- miscarrying patients, compared with 5.7% of
tional diabetes in RPL. This figure was statistically control women (9921 of 172 490 control patients).

Lower Upper
OR limit limit p
Hughes et al2 1.941 0.783 4.814 0.152
Jivraj et al4 0.728 0.412 1.285 0.274
Sheiner et al6 1.933 1.778 2.102 0.000
Common OR 1.894 1.743 2.057 0.000
0.1 1 10 100
Figure 16.2 Odds ratio (OR) for pregnancy-induced hypertension in recurrent pregnancy loss.
233
0415421306-Ch16 3/30/07 9:48 AM Page 234

Lower Upper
OR limit limit p
Hughes et al2 2.433 0.762 7.768 0.133
Jivraj et al4 6.885 4.318 10.978 0.000
Sheiner et al6 1.000 0.847 1.181 0.998
Common OR 1.257 1.076 1.467 0.004
0.1 1 10 100
Figure 16.3 Odds ratio (OR) for intrauterine growth restriction in recurrent pregnancy loss.
Neither Reginald et al1 nor Tulppala et al3 quoted Figure 16.4 summarizes the results. Again, there
figures for PIH. was a statistically significant association between
preterm labor and RPL. The common OR was
INTRAUTERINE GROWTH RESTRICTION 2.84 (95% CI 1.96–4.1). In the series of Reginald
et al,1 there was a 28% incidence of preterm labor.
The same three papers as above give figures for The relative risk was 3.3 when compared with the
IUGR. The common OR for developing IUGR is Scottish data for the equivalent period. Tulppala
1.25 (95% CI 1.08–1.47) (Figure 16.3). Reginald et al3 quoted a 9.7% incidence in their series.
et al1 reported a 33% incidence in 344 pregnancies
prospectively followed up, and reported that this PERINATAL MORTALITY
showed a relative risk of 3 compared with the
standard Scottish population. Tulppala et al3 also Again, there was an increased tendency for morta-
reported a 20% incidence (6 out of 30 pregnancies). lity after RPL, loss with a common OR of 1.25 (95%
CI 1.05–1.49). This is summarized in Figure 16.5.
PRETERM LABOR However, the perinatal mortality may be artificially
low due to obstetric intervention for other compli-
Comparative figures are only available for preterm cations in pregnancy. Reginald et al1 reported
labor from the series of Hughes et al2 and Jivraj et al.4 19 perinatal deaths in 118 infants (16.1%), which
Lower Upper
OR limit limit p
Hughes et al2 1.530 0.811 2.887 0.189
Jivraj et al4 3.890 2.470 6.126 0.000
Common OR 2.837 1.961 4.104 0.000
0.1 1 10 100
Figure 16.4 Odds ratio (OR) for preterm labor in recurrent pregnancy loss.
234
0415421306-Ch16 3/30/07 9:48 AM Page 235
Lower Upper
OR limit limit p
Hughes et al2 1.103 0.060 20.305 0.947
Jivraj et al4 2.506 0.922 6.814 0.072
Sheiner et al6 1.222 1.021 1.464 0.029
Common OR 1.250 1.047 1.492 0.014
0.1 1 10 100
Figure 16.5 Odds ratio (OR) for perinatal mortality in recurrent pregnancy loss.
was considerably higher than the standard figures rate was only obtained by early obstetric interven-
for England and Wales in the same period of tion to prevent intrauterine fetal deaths. A number
time, when the perinatal mortality was 10.1/1000. of other publications have attested that the risk
Tulppala et al3 did not quote perinatal mortality. of obstetric complications is high in APS.11,12
It seems, therefore, that the currently available The incidence of preeclampsia is particularly high
literature on the obstetric and neonatal outcome in APS.13,14 IUGR has been reported with a fre-
of pregnancies from women with a history of RPL quency ranging from 30% to 12% in different
shows a consistently worse prognosis. However, it is series. Some series show a significant increase in
unclear whether the worse prognosis is only found the incidence of IUGR,15,16 whereas others have not
in patients with predisposing causes such as APS confirmed the increased incidence.17,18 The different
and hereditary thrombophilias or is also present in results on the association of obstetric complications
patients with unexplained pregnancy losses. with aPL might have various explanations. None of
the results has been correlated with β2-glycoprotein
I(β2GPI)-dependent antibodies. None has corrected
ANTIPHOSPHOLIPID SYNDROME for fetal chromosomal aberrations, in which the aPL
may just be an epiphenomenon. Alternatively, the
Most series on APS have concentrated on the obstetric complications may be dependent on other
subsequent live birth rate and ignored the late factors associated with RPL rather than aPL. No
obstetric complications. The present author pub- series has compared the incidence of obstetric com-
lished a series10 in which the outcome of 24 preg- plications in RPL due to APS and in RPL without
nancies in patients with lupus anticoagulant (LA) APS. The incidence of preterm labor is increased in
and five or more first-trimester abortions was APS patients,16 but again has not been compared
compared with 22 pregnancies in women with no with patients with RPL and no APS.
antiphospholipid antibodies (aPL), and 5 or more The currently accepted optimal treatment regi-
miscarriages. Although the subsequent number of men for APS is heparin or low-molecular-weight
first-trimester miscarriages and the live birth rate heparin (LMWH) with the addition of low-dose
were similar in both groups of patients, the inci- aspirin. However, anticoagulants do not seem to lower
dences of second- or third-trimester fetal deaths, the incidence of obstetric complications associated
IUGR, and need for premature induction of labor with this syndrome.19 IVIG may have a beneficial
or preterm cesarean section were significantly effect in APS, as the action and production of aPL
higher in the APS patients. The similar live birth are inhibited by IVIG. The F(ab′) fragment of IVIG
235
0415421306-Ch16 3/30/07 9:48 AM Page 236
inhibits binding of anticardiolipin antibody (aCL) thrombophilias and severe preeclampsia, placental
to cardiolipin in a dose-dependent manner.20 The abruption, IUGR, and stillbirth. However, theirs
F(ab′) fragment of IVIG inhibits LA activity.21 was a case–control study looking at the prevalence
IVIG lowers levels of aCL after each infusion.22 of hereditary thrombophilias in women with
IVIG may contain anti-idiotypic antibodies to obstetric complications. Additionally, these findings
aPL, or may inactivate B-cell clones, leading to have beeen disputed by Infante-Rivarde et al,32 who
decreased autoantibody production.23 In the late also performed a case–control study. Alfirevic et al33
1980s and early 1990s, IVIG was used to improve have published a systematic review of the various
the live birth rate in APS.24-27 However, these were reports and have described the strengths of asso-
small series, and showed IVIG to have no apparent ciation between hereditary thrombophilias and
benefit over anticoagulants in terms of live births. preeclampsia, placental abruption, stillbirth, and
However, when the obstetric complications are IUGR. Sheiner et al6 have drawn attention to the
considered, a different clinical picture emerges. fact that the other publications on obstetric com-
Vaquero et al28 compared IVIG with prednisone plications in RPL were written at a time when the
and aspirin. The prevalence of IUGR and pre- hereditary thrombophilias had not yet been
term labor was similar in both groups, but the recognized. In their series, higher rates of IUGR,
prevalences of PIH and gestational diabetes were cesarean section, low Apgar scores and perinatal
significantly lower (p<0.05) after IVIG (5% of mortality were found among the 22 patients with
41 patients and 14% of 22 patients, respectively). known thrombophilia as compared with the
Branch et al29 compared IVIG with placebo: there controls, although the differences did not reach
were fewer cases of IUGR after IVIG (14% of statistical significance. There are no cohort studies
7 patients, compared with 33% of 9 patients, assessing the true incidence of obstetric compli-
respectively), and fewer admissions to the neonatal cations in the presence of thrombophilia and
intensive care unit (14% of 7 patients after IVIG, RPL. In the present author’s series of 21 pregnan-
compared with 44% of 9, respectively). Harris and cies with factor V Leiden which were followed up
Pierangelli30 reported that preeclampsia, IUGR prospectively, there was one case of HELLP
and prematurity were reduced when IVIG was syndrome (hemolysis, elevated liver enzymes, and
compared with prednisone and aspirin or heparin low platelet count), but no other obstetric compli-
and aspirin. cations, and no deep vein thrombosis or pulmonary
embolus.
The LMWH enoxaparin has been shown to be
HEREDITARY THROMBOPHILIAS associated with a significantly increased chance of
a live birth, both overall and when corrected for
There have recently been numerous publications the number of miscarriages34 (RR 0.53; 95%
associating genetic predispositions (hereditary CI 1.07–3.28), in women with RPL and hereditary
thrombophilias) to thrombosis with pregnancy loss. thrombophilia. There are isolated reports35–37 of
The hereditary thrombophilias include protein C, the use of anticoagulants in the presence of throm-
protein S, and antithrombin III deficiencies, bophilias to reduce the late obstetric complications,
activated protein C resistance (APCR), factor V and this approach has been endorsed by the
Leiden mutation, methylenetetrahydrofolate reduc- American College of Chest Physicians.38 However,
tase (MTHFR) mutation, the prothrombin gene there is as yet insufficient evidence that antico-
mutation (G20210A) and excessive factor VIII. agulants actually reduce the incidence of late
The features of hereditary thrombophilias and obstetric complications. Further trials need to be
the effects on obstetric complications are discussed performed in order to determine whether anticoag-
in Chapter 10. Suffice it to say here that Kupferminc ulants do indeed reduce the incidence of obstetric
et al31 have reported an association between complications.
236
0415421306-Ch16 3/30/07 9:48 AM Page 237
Table 16.1 Preterm labor, IUGR, perinatal

OBSTETRIC COMPLICATIONS AFTER mortality, and anomalies with paternal leukocyte
ALLOIMMUNIZATION immunization compared with controlsa
Immunized Controls
Some 20 years ago, alloimmunization had become
(n = 979)b (n = 483)b Relative riskc
the treatment of choice for patients with RPL.
Two methods of alloimmunization had been used Preterm labor 39 (3.9%) 52 (10.8%) 0.63 (0.49–0.79)
in order to improve the subsequent live birth rate in IUGR 17 (1.7%) 59 (12.2%) 0.32 (0.21–0.49)
Perinatal mortality 9 (0.9%) 21 (4.3%) 0.44 (0.26–0.47)
recurrently aborting women: active immunization Anomalies 25 (2.6%) 19 (3.9%) 0.84 (0.65–1.10)
with paternal leukocytes and passive immuniza-
a
Figures include 92 immunized and 175 non-immunized patients from
tion using IVIG.39 Paternal leukocyte immunization Beard’s series7, 759 immunized patients and 279 non-immunized
became standard management after the double- patients from the RMITG register,9 and 128 immunized patients and 29
blinded randomized trial by Mowbray et al40 and non-immunized patients from the present author’s series.
b
Incidences in parentheses.
a subsequent meta-analysis by the Recurrent c
95% confidence intervals in parentheses.
Miscarriage Immunization Trialists Group,9 but fell
out of favor due to the paper by Ober et al,41 which
claimed lack of efficacy. As with treatment for APS These figures are summarised in Table 16.2. As can
and hereditary thrombophilias, most publications be seen, the incidence of preeclampsia was lower
on immunomodulation have assessed the subsequent in immunized women. However, there was no
live birth rate as the primary outcome measure, and significant difference in the incidence of vaginal
have ignored the late obstetric complications of bleeding, which remained high: 38% and 34% in
pregnancy occurring in women with a subsequent immunized and control women, respectively.
live birth. Table 16.3 shows the obstetric complications
In order to summarize the obstetric complica- after IVIG in the 136 women in the present author’s
tions with or without immunotherapy, the various series. There is no relevant control group, as some
series in the literature have been pooled to obtain of the patients were administered IVIG after failure
a sufficient number of patients available for mean- of paternal leukocyte immunization. The figures
ingful analysis. The databases of Beard et al,7 the were compared with those for non-immunized
RMITG meta-analysis9 and the present author’s patients in the registers above and those in control
series have been combined in order to compare the patients in other series on IVIG in the literature. The
obstetric complications after paternal leukocyte incidences of IUGR, perinatal mortality, bleeding,
immunization: 979 immunized patients were avail-
able for analysis, compared with 483 non-immunized
patients. The prevalence of IUGR, perinatal mortal-
ity, and the incidence of anomalies were assessed. Table 16.2 Bleeding and preeclampsia with
These figures are summarized in Table 16.1. As can paternal leukocyte immunization compared
be seen, there was a significantly lower incidence of with controlsa
preterm labor, perinatal mortality, and IUGR after
Immunized Controls
paternal leukocyte immunization than in controls. (n = 216)b (n = 191)b Relative riskc
The incidence of fetal anomalies was not sig-
nificantly different in the two groups of patients. Bleeding 83 (38%) 64 (34%) 1.10 (0.92–1.33)
Preeclampsia 27 (13%) 57 (30%) 0.55 (0.40–0.76)
As the RMITG meta-analysis9 did not assess bleed-
ing and preeclampsia, data could only be obtained a
Figures include 88 immunized and 162 non-immunized patients from
from Beard’s7 series and the present author’s series. Beard’s series7 and 128 immunized patients and 29 non-immunized
patients from the present author’s series.
Two hundred and sixteen immunized patients b
and 191 control patients are available for analysis. c
95% confidence intervals in parentheses.
237
0415421306-Ch16 3/30/07 9:48 AM Page 238
detrimental to pregnancy; for example, tumor

Table 16.3 Obstetric complications after
IVIG compared with controlsa necrosis factor α (TNF-α) can activate NK cells at
the fetomaternal interface,43 leading to apoptosis
IVIGb Controlsb Relative riskc and trophoblast cell death.44 Wegman23 suggested
that appropriate cytokines are necessary in order to
Preterm labor 15/136 (11%) 52/483 (10.8%) 1.02 (0.53–1.96)
IUGR 6/136 (4%) 59/483 (12.2%) 0.39 (0.18–0.86) confer benefits on the developing fetoplacental unit.
Perinatal 2/136 (1.5%) 21/483 (4.3%) 0.39 (0.10–0.47) In early pregnancy, the conceptus and placental
mortality
tissues produce a variety of Th1-type cytokines,
Bleeding 5/136 (3.7%) 64/191 (34%) 0.14 (0.06–0.33)
Preeclampsia 4/136 (2.9%) 57/191 (30%) 0.13 (0.05–0.34) including interferons, interleukins and TNF.45
Anomalies 1/136 (0.7%) 19/483 (3.9%) 0.22 (0.03–1.51) In normal pregnancy, the maternal immune system
a
Figures show the incidence of anomalies as a function of the total seems to modulate the cytokine pattern to prefe-
number of patients in the sample. IVIG figures are from the present rential expression of Th2 cytokines, including
author’s series. The control figures are pooled data from the literature. GM-CSF, IL-3, IL-4, IL-5, IL-10, and IL-13, that
b
c
95% confidence intervals in parentheses. may enhance fetal survival.46 Preferential expression
of Th1 cytokines (TNF-α, interferon-γ (IFN-γ),
and IL-2) may result in abnormal placental and
and preeclampsia were all lower after IVIG, but the embryonic growth and subsequent fetal demise.
incidences of preterm labor and anomalies did An inappropriate cytokine balance has been
not seem to be affected. However, these results reported to act in early pregnancy, causing NK-cell
should be interpreted with caution due to the activation,43 placental apoptosis,44 teratogenesis,47
nature of the control group. and excessive coagulation, particularly TNF-α
and IL-6.48 Hence, cytokine imbalance is among the
mechanisms that have been proposed to underlie
CYTOKINES AS MEDIATORS OF PREGNANCY RPL. The effect of cytokine imbalance on coagula-
LOSS AND OBSTETRIC COMPLICATIONS tion may explain some of the effects of hereditary
thrombophilias. Additionally, APS and most of
Cytokines are low-molecular-weight peptides or the late complications of pregnancy have also been
glycopeptides, which are produced by lymphocytes, shown to have an underlying cytokine basis.
monocytes/macrophages, mast cells, eosinophils, Although it is generally accepted that aPL may
and blood vessel endothelial cells. Cytokines seem act by causing blood coagulation in decidual
to influence all stages of pregnancy. Cytokines such vessels, cytokines may also be responsible for the
as interleukin-3 (IL-3), granulocyte-macrophage action of aPL early in the pathogenesis of the
colony-stimulating factor (GM-CSF), and epidermal condition. Serum TNF-α is increased in women
growth factor (EGF) stimulate placental cell prolife- with APS.49 IL-3 is decreased in APS,50 and fetal loss
ration42 in vitro, and may enable the trophoblast to can be prevented in experimental APS by in vivo
secrete human chorionic gonadotropin(hCG) and administration of recombinant IL-3.51 Alteration of
human placental lactogen (hPL). Cytokines such as the Th1/Th2 balance may also be involved in the
transforming growth factor β2 (TGF-β2) can inhibit effect of anti-idiotypic antibodies on APS.52
lymphokine activation of natural killer (NK) cells Late obstetric complications have also been
(which may attack the trophoblast). Although shown to be associated with altered cytokine levels.
cytokines can have pleiotropic actions, these actions Preeclampsia is associated with reduced IL-10 and
of the above cytokines are classified as T-helper 2 higher IL-2 production from peripheral blood
(Th2) or anti-inflammatory in nature, and they mononuclear cells (PBMC),53,54 and high serum
seem to benefit the development of pregnancy. IL-8 and TNF-α.55 In preterm births, cytokine
Other cytokines can have actions that are known involvement has been reported,56 particularly
as Th1 or proinflammatory, and these may be increased amniotic IL-6, IL-8, and TNF-α.57–60
238
0415421306-Ch16 3/30/07 9:48 AM Page 239
In IUGR, TGF-β in cord blood and mRNA for complications have been described in APS and
IL-10 are significantly reduced, whereas IL-8 mRNA hereditary thrombophilias. However, there is insuf-
is significantly higher61,62 and placental TNF-α ficient evidence at present to determine whether
secretion is enhanced.62 the late obstetric complications occur exclusively
The various interventions used for improving in these two conditions, or whether they are associ-
the live birth rate in RPL may exert their effects by ated with RPL per se. Immunomodulation seems to
modulating cytokine balance, and this modulation reduce the incidence of some of these complica-
may also influence the incidence of later obstetric tions. The role of anticoagulants in reducing obstet-
complications. As stated in Chapter 10 it is possible ric complications is more doubtful. However,
that heparin or enoxaparin may work by anti- careful surveillance is required in pregnancies
inflammatory mechanisms rather than antico- following RPL, in order to detect obstetric compli-
agulation. Heparin increases serum TNF-binding cations. Further prospective cohort studies are
protein, hence protecting against systemic harmful necessary in order more accurately define the
manifestations.63 LMWH inhibit TNF-α production.64 patient at risk, the role of APS and hereditary
Thrombosis results in an inflammatory response of thrombophilia, and the effect of treatment modalities.
the vein wall. Both heparin and LMWH limit the
anti-inflammatory response, but in the series of REFERENCES
Downing et al,65 LMWH was more effective than
heparin in limiting neutrophil extravasation and was 1. Reginald PW, Beard RW, Chapple J, et al. Outcome of pregnancies
the only intervention to decrease vein wall perme- progressing beyond 28 weeks gestation in women with a history of
recurrent miscarriage. Br J Obstet Gynaecol 1987; 94:643–8.
ability. In addition, heparin has been shown to pre- 2. Hughes N, Hamilton EF, Tulandi T. Obstetric outcome in women
vent the accumulation of proliferating vascular smooth after multiple spontaneous abortions. J Reprod Med 1991; 36:165–6.
muscle cells encroaching on the lumen of arteries 3. Tulppala M, Palosuo T, Ramsay T, et al. A prospective study of 63
couples with a history of recurrent spontaneous abortion: contribut-
injured at angioplasty, hence enlarging the lumen.66 ing factors and outcome of subsequent pregnancies. Hum Reprod
The mode of action of paternal leukocyte immu- 1993; 8:764–70.
nization (PLI) is unclear. The two most plausible 4. Jivraj S, Anstie B, Cheong YC, et al. Obstetric and neonatal outcome
in women with a history of recurrent miscarriage: a cohort study.
explanations seem to be that PLI may downregulate Hum Reprod 2001; 16:102–6.
the numbers and killing activity of NK cells,67 5. Seidman DS, Gale R, Ever-Hadani P, et al. Reproductive complications
which may attack the trophoblast when activated by after previous spontaneous abortions. Pediatr Rev Commun 1990;
5:1–10.
cytokines. PLI may also alter the balance between 6. Sheiner E, Levy A, Katz M, et al. Pregnancy outcome following
Th1 and Th2 cytokines. (PLI has been reported to recurrent spontaneous abortions. Eur J Obstet Gynecol Reprod Biol
reduce IFN-γ and increase the secretion of IL-10 2005; 118:61–5.
7. Beard RW. Clinical associations of recurrent miscarriage. In: Beard
and TGF-β,68 and has also been shown to lower the RW, Sharp F, eds. Early Pregnancy Loss: Mechanisms and Treatment.
serum IL-6 and soluble IL-6 receptor (sIL-6R) levels London, UK: RCOG, 1988:3–8.
to the values observed in normal pregnancy.69) 8. Carp HJA, Toder V, Mashiach S, et al. Recurrent miscarriage:
a review of current concepts, immune mechanisms, and results of
IVIG has numerous actions. The actions of treatment. Obstet Gynecol Surv 1990; 45:657–69.
IVIG in pregnancy, including its effects on APS, have 9. Recurrent Miscarriage Immunotherapy Trialists Group. Worldwide
been summarized elsewhere.70 In addition to various collaborative observational study and metaanalysis on allogenic
leucocyte immunotherapy for recurrent spontaneous abortion.
other mechanisms, IVIG, like PLI, depresses NK-cell Am J Reprod Immunol 1994; 32:55–72.
function71 and enhances the action of Th2 cytokines.72 10. Carp HJA, Menashe Y, Frenkel Y, et al. Lupus anticoagulant: significance
in first trimester habitual abortion. J Reprod Med 1993; 38:549–52.
11. Tincani A, Balestrieri G, Danieli E, et al. Pregnancy complications of
the antiphospholipid syndrome. Autoimmunity 2003; 36:27–32.
CONCLUSIONS 12. Brewster JA, Shaw NJ, Farquharson RG. Neonatal and pediatric
outcome of infants born to mothers with antiphospholipid
syndrome. J Perinat Med 1999; 27:183–7.
Patients with RPL seem to comprise a high-risk group 13. Branch DW, Dudley DJ, Scott JR, et al. Antiphospholipid antibodies
for later obstetric complications. Late obstetric and fetal loss. N Engl J Med 1992; 326:952.
239
0415421306-Ch16 3/30/07 9:48 AM Page 240
14. Lima F, Khamashta MA, Buchanan NM, et al. A study of sixty 34. Carp HJA, Dolitzky M, Inbal A. Thromboprophylaxis improves the
pregnancies in patients with the antiphospholipid syndrome. Clin live birth rate in women with consecutive recurrent miscarriages and
Exp Rheumatol 1996; 14:1316. hereditary thrombophilia. J Thromb Hemost 2003; 1:433–8.
15. Branch DW, Silver RM, Blackwell JL, et al. Outcome of treated 35. Younis JS, Ohel G, Brenner B, et al. Familial thrombophilia – the
pregnancies in women with antiphospholipid syndrome: an update scientific rationale for thrombophylaxis in recurrent pregnancy loss.
of the Utah experience. Obstet Gynecol 1992; 80:614–20. Hum Reprod 1997; 12:1389–90.
16. Kutteh WH, Ermel LD. A clinical trial for the treatment of antiphos- 36. Brenner B, Hoffman R, Blumenfeld Z, et al. Gestational outcome
pholipid antibody-associated recurrent pregnancy loss with lower in thrombophilic women with recurrent pregnancy loss treated by
dose heparin and aspirin. Am J Reprod Immunol 1996; 35:4027. enoxaparin. Thromb Hemost 2000; 83:693–7.
17. Pattison NS, Chamley LW, McKay EJ, et al. Antiphospholipid anti- 37. Riyazi N, Leeda M, de Vries JI, et al. Low-molecular-weight heparin
bodies in pregnancy: prevalence and clinical associations. Br J Obstet combined with aspirin in pregnant women with thrombophilia and a
Gynaecol 1993; 100: 909–13. history of preeclampsia or fetal growth restriction: a preliminary
18. Lynch A, Marlar R, Murphy J, et al. Antiphospholipid antibodies study. Eur J Obstet Gynecol Reprod Biol 1998; 80:49–54.
in predicting adverse pregnancy outcome. A prospective study. Ann 38. Bates SM, Greer IA, Hirsh J, et al. Use of antithrombotic agents during
Intern Med 1994; 120:470–5. pregnancy: the Seventh ACCP Conference on Antithrombotic and
19. Shehata HA, Nelson-Piercy C, Khamashta MA. Management of Thrombolytic Therapy. Chest 2004; 126:627S–44S.
pregnancy in antiphospholipid syndrome. Rheum Dis Clin North 39. Coulam CB, Krysa LW, Stern JJ, et al. Intravenous immunoglobulin
Am 2001; 27:643–59. for treatment of recurrent pregnancy loss. Am J Reprod Immunol
20. Caccavo D, Vaccaro F, Ferri GM, et al. Anti-idiotypes against 1995; 34:333–7.
antiphospholipid antibodies are present in normal polyspecific 40. Mowbray JF, Gibbings CR, Liddell H. et al. Controlled trial of
immunoglobulins for therapeutic use. J Autoimmun 1994; 7:537–48. treatment of recurrent spontaneous abortions by immunization
21. Galli M, Cortelazzo S, Barbui T. In vivo efficacy of intravenous with paternal cells. Lancet 1985; i:941–3.
gammaglobulins in patients with lupus anticoagulant is not mediated 41. Ober C, Karrison T, Odem RR, et al. Mononuclear-cell immunisation
by anti-idiotypic mechanisms. Am J Hematol 1991; 38:184–8. in prevention of recurrent miscarriages: a randomised trial. Lancet
22. Kwak JY, Quilty EA, Gilman-Sachs A, et al. Intravenous immunoglob- 1999; 354:365–9.
ulin infusion therapy in women with recurrent spontaneous abor- 42. Chaouat G, Menu E, Wegmann TG. Role of lymphokines of the CSF
tions of immune etiologies. J Reprod Immunol 1995; 28:175–88. family and of TNF, gamma interferon and IL-2 in placental growth
23. Wegmann TG. The cytokine basis for cross-talk between the maternal and fetal survival studied in two murine models of spontaneous
immune and reproductive systems. Curr Opin Immunol 1990; 2:765–9. resorptions. In: Chaouat G, Mowbray JF, eds. Cellular and Molecular
24. Carreras LD, Perez GN, Vega HR, et al. Lupus anticoagulant and Biology of the Maternal–Fetal Relationship. Paris: INSERM/John
recurrent fetal loss: successful treatment with gammaglobulin. Lancet Libbey Eurotext, 1991:91.
1988; ii:393–4. 43. King A, Jokhi PP, Burrows TD, et al. Functions of human decidual NK
25. Cowchock FS, Wapner RJ, Needleman L, et al. A comparison of preg- cells. Am J Reprod Immunol 1996; 35:258–60.
nancy outcome after two treatments for antibodies to cardiolipin 44. Baines MG, Duglos AJ, de Fougerolles AR, et al. Immunological
(ACA). Clin Exp Rheumatol 1988; 6:200–6. prevention of spontaneous early embryo resorption is mediated
26. Triplett DA. Antiphospholipid antibodies and recurrent pregnancy by non specific immunostimulation. Am J Reprod Immunol 1996;
loss. Am J Reprod Immunol 1989; 20:52–67. 35:34–42.
27. Wapner RJ, Cowchock FS, Shapiro SS. Successful treatment in two 45. Schäfer-Somi S. Cytokines during early pregnancy of mammals:
women with antiphospholipid antibodies and refractory pregnancy a review. Animal Reprod Sci 2003; 75:73–94.
losses with intravenous immunoglobulin infusions. Am J Obstet 46. Saito S. Cytokine network at the feto–maternal interface. J Reprod
Gynecol 1989; 161:1271–2. Immunol 2000; 47:87–103.
28. Vaquero E, Lazzarin N, Valensie H, et al. Pregnancy outcome in recur- 47. Savion S, Brengauz-Breitmann M, Torchinsky A, et al. A possible
rent spontaneous abortion associated with antiphospholipid antibodies: role for granulocyte macrophage-colony stimulating factor in modu-
a comparative study of intravenous immunoglobulin versus prednisone lating teratogen-induced effects. Teratog Carcinog Mutagen 1999;
plus low-dose aspirin. Am J Reprod Immunol 2001; 45:174–9. 19:171–82.
29. Branch DW, Peaceman AM, Druzin M, et al. A multicenter, placebo- 48. Levi M, Ten Cate H. Disseminated intravascular coagulation. N Engl
controlled pilot study of intravenous immune globulin treatment of J Med 1999; 341:586–92.
antiphospholipid syndrome during pregnancy. The Pregnancy Loss 49. Bertolaccini ML, Atsumi T, Lanchbury JS, et al. Plasma tumor
Study Group. Am J Obstet Gynecol 2000; 182:122–7. necrosis factor α levels and the 238*A promoter polymorphism in
30. Harris EN, Pierangeli SS. Utilization of intravenous immunoglobulin patients with antiphospholipid syndrome. Thromb Haemost 2001;
therapy to treat recurrent pregnancy loss in the antiphospholipid syn- 85:198–203.
drome: a review. Scand J Rheumatol 1998; 107 (Suppl):97–102. 50. Shoenfeld Y, Sherer Y, Fishman P. Interleukin-3 and pregnancy loss
31. Kupferminc MJ, Eldor A, Steinman N, et al. Increased frequency in antiphospholipid syndrome. Scand J Rheumatol 1998; 107 (Suppl):
of genetic thrombophilia in women with complications of pregnancy. 19–22.
N Engl J Med 1999; 340:9–13. 51. Fishman P, Falach-Vaknine E, Zigelman R, et al. Prevention of
32. Infante-Rivard C, Rivard GE, Yotov WV, et al. Absence of association fetal loss in experimental antiphospholipid syndrome by in vivo
of thrombophilia polymorphisms with intrauterine growth restric- administration of recombinant interleukin-3. J Clin Invest 1993;
tion. N Engl J Med 2002; 347:19–25. 91:1834–7.
33. Alfirevic Z, Roberts D, Martlew V. How strong is the association 52. Krause I, Blank M, Levi Y, et al. Anti-idiotype immunomodulation
between maternal thrombophilia and adverse pregnancy outcome? of experimental anti-phospholipid syndrome via effect on Th1/Th2
A systematic review. Eur J Obstet Gynecol Reprod Biol 2002; 101:6–14. expression. Clin Exp Immunol 1999; 117:190–7.
240
0415421306-Ch16 3/30/07 9:48 AM Page 241
53. Darmochwal-Kolarz D, Rolinski J, Leszczynska-Goarzelak B, et al. 64. Baram D, Rashkovsky M, Hershkoviz R, et al. Inhibitory effects of
The expressions of intracellular cytokines in the lymphocytes of low molecular weight heparin on mediator release by mast cells: pref-
preeclamptic patients. Am J Reprod Immunol 2002; 48:381. erential inhibition of cytokine production and mast cell-dependent
54. Orange S, Horvath J, Hennessy A. Preeclampsia is associated with a cutaneous inflammation. Clin Exp Immunol 1997; 110:485–91.
reduced interleukin-10 production from peripheral blood mononu- 65. Downing LJ, Strieter RM, Kadell AM, et al. Low-dose low-molecular-
clear cells. Hypertens Pregnancy 2003; 22:1–8. weight heparin is anti-inflammatory during venous thrombosis.
55. Velzing-Aarts FV, Muskiet FA, Van der Dijs FP, et al. High serum J Vasc Surg 1998; 28:848–54.
interleukin-8 levels in Afro-Caribbean women with pre-eclampsia. 66. San Antonio JD, Verrecchio A, Pukac LA. Heparin sensitive and
Relations with tumor necrosis factor-α, Duffy negative phenotype resistant vascular smooth muscle cells: biology and role in restenosis.
and von Willebrand factor. Am J Reprod Immunol 2002; 48:319–22. Conn Tiss Res 1998; 37:87–103.
56. Park JS, Park CW, Lockwood CJ, et al. Role of cytokines in preterm 67. Kwak JY, Gilman-Sachs A, Moretti M, et al. Natural killer cell
labor and birth. Minerva Ginecol 2005; 57:349–66. cytotoxicity and paternal lymphocyte immunization in women
57. Fortunato SJ, Menon R, Lombardi SJ. Role of tumor necrosis factor-α with recurrent spontaneous abortions. Am J Reprod Immunol 1998;
in the premature rupture of membranes and preterm labor pathways. 40:352–8.
Am J Obstet Gynecol 2002; 187:1159–62. 68. Gafter U, Sredni B, Segal J, et al. Suppressed cell-mediated immunity
58. Jacobsson B, Mattsby-Baltzer I, Andersch B, et al. Microbial invasion and monocyte and natural killer cell activity following allogeneic
and cytokine response in amniotic fluid in a Swedish population of immunization of women with spontaneous recurrent abortion. J Clin
women in preterm labor. Acta Obstet Gynecol Scand 2003; 82:120–8. Immunol 1997; 17:408–19.
59. Maymon E, Ghezzi F, Edwin SS, et al. The tumor necrosis factor-α 69. Zenclussen AC, Kortebani G, Mazzolli A, et al. Interleukin-6 and
and its soluble receptor profile in term and preterm parturition. soluble interleukin-6 receptor serum levels in recurrent spontaneous
Am J Obstet Gynecol 1999; 181:1142–8. abortion women immunized with paternal white cells. Am J Reprod
60. Ognjanovic S, Bryant-Greenwood GD. Pre-B-cell colony-enhancing Immunol 2000; 44:22–9.
factor, a novel cytokine of human fetal membranes. Am J Obstet 70. Carp HJA, Sapir T, Shoenfeld Y. Intravenous immunoglobulin
Gynecol 2002; 187:1051–8. and recurrent pregnancy loss. Clin Rev Allergy Immunol 2005;
61. Hahn-Zoric M, Hagberg H, Kjellmer I, et al. Aberrations in placental 29:327–32.
cytokine mRNA related to intrauterine growth retardation. Pediatr 71. Ruiz JE, Kwak JY, Baum L, et al. Effect of intravenous
Res 2002; 51:201–6. immunoglobulin G on natural killer cell cytotoxicity in vitro in
62. Holcberg G, Huleihel M, Sapir O, et al. Increased production of tumor women with recurrent spontaneous abortion. J Reprod Immunol
necrosis factor-α (TNF-α) by IUGR human placentae. Eur J Obstet 1996; 31:125–41.
Gynecol Reprod Biol 2001; 94:69–72. 72. Graphou O, Chioti A, Pantazi A, et al. Effect of intravenous
63. Lantz M, Thysell H, Nilsson E, et al. On the binding of tumor immunoglobulin treatment on the Th1/Th2 balance in women with
necrosis factor (TNF) to heparin and the release in vivo of the recurrent spontaneous abortions. Am J Reprod Immunol 2003;
TNF-binding protein I by heparin. J Clin Invest 1991; 88:2026–31. 49:21–9.
241
0415421306-Ch16 3/30/07 9:48 AM Page 242
0415421306-Ch17 3/30/07 9:49 AM Page 243
17. Coping with recurrent pregnancy loss:

Psychological mechanisms
Keren Shakhar
basic expectations about family life. What is

INTRODUCTION
expected to be a fulfilling experience is instead an
experience of loss and disappointment. These mis-
Recurrent pregnancy loss (RPL) is clearly a stressful
carriages usually occur at a very sensitive phase in
experience, but very little is known about what sets
the couple’s development: becoming parents is a
its emotional effects apart from isolated sponta-
transitional stage that requires reconstruction of
neous miscarriages and from other forms of infer-
identities and preparation for new roles.
tility. When studying the psychological effects of
Only a few studies have specifically addressed the
RPL, it is important not only to examine them
psychological difficulties of couples suffering from
through a pathological perspective, i.e., the induc-
more than one miscarriage, focusing, as a rule, on
tion of distress and depression, but also to appreci-
the women. These studies suggest that the second
ate how couples cope with this experience in their
miscarriage has harsher emotional impact than the
everyday life. A more general perspective would also
first.1–3 Although it seems logical, the question as to
examine the effect of RPL on self-esteem and mari-
whether the third and fourth miscarriages further
tal and social relations. The degree of emotional
aggravate distress has never been assessed.
anguish couples experience largely depends on the
Surprisingly, no differences in psychological distress
significance they ascribe to RPL. This meaning is
were found between women who have had a child
influenced not only by the couple’s views, but also
and those who have not (primary or secondary
by the perception of infertility and the view of pre-
aborters).3–5 Perhaps mothers feel guilty for failing
natal life in their specific society. Here I shall
to provide a sibling for their child, and fear that their
describe the psychological aspects of RPL based on
child feels lonely. It is estimated that around 30% of
a review of the literature and on my assessment of
women with RPL are depressed and that even a
such couples using a focus group, stress question-
higher proportion have high levels of state and trait
naires, and informal interactions in the setting of a
anxiety.4,5
clinical study.
When women with RPL conceive again, they
exhibit high levels of anxiety, having difficulty get-
ting through each day.6 This anxiety is manifested as
PSYCHOLOGICAL REACTIONS TO general tension, despondence, and premonitions of
RECURRENT MISCARRIAGE miscarriage, and may be exhibited by weeping, fear
of detecting bleeding when going to the toilet or
RPL is a type of infertility that confronts couples examining underwear, extreme anxiety over any
with repeated cycles of hope and despair. Many abdominal pain, checking continuously for signs
couples view parenthood as an indispensable com- of pregnancy, avoidance of other pregnant women,
ponent of their marriage and many cases of RPL and reluctance to discuss the pregnancy with
occur before they have had a child. Young couples anyone, including their husband.6,7 As a so-called
often take for granted their ability to conceive and defense mechanism, some women show less emo-
become parents, and are only concerned with the tional attachment to their subsequent pregnancies,
question when to have a child. RPL shatters their and avoid thinking about their future child.6,8
243
0415421306-Ch17 3/30/07 9:49 AM Page 244
Although this type of reaction may alleviate the couples generally do not receive social support, and
constant anxiety and may protect women emotion- may also face insensitive attitudes. Sometimes, mis-
ally if eventually they miscarry, it also diminishes carriage occurs before the couple had shared the
the pleasure women can derive from being preg- news of the pregnancy with anyone, leaving them
nant, and may prevent grief from being processed lonely in the grieving process. It is crucial to under-
and the experience from being integrated. In addi- stand that even if the embryo was lost at a very early
tion, it is unclear how deep into pregnancy women gestational week, many couples already regard their
are less attached to their embryo and whether it embryo as a baby, name or nickname him, talk to him,
complicates the transition to motherhood. ascribe a specific personality to him, and imagine his
Since the psychological literature on RPL is lim- future.14
ited, and since women with RPL must cope with Unfortunately, family members and friends may
both miscarriages and potential infertility, I will not know how to respond to the bereaved couple,
next examine what is known about these two enti- and may not grasp what the pregnancy meant to
ties. The reaction to a sudden loss of pregnancy them.15 The variability in couples’ attitudes may
varies greatly among different individuals: some make it hard for their friends and family to support
exhibit little or no reaction, whereas others demon- them. A break in communications sometimes
strate a significant decline in their coping ability.9,10 occurs because of lack of response or because the
Major themes that describe the experience of mis- couple consider the response inappropriate.16,17
carriage are emptiness and guilt.11 Increased anxiety Typical attempts at consolation include ‘At least you
and depressive symptoms are also very common.9,12,13 can get pregnant’, ‘Maybe it’s good you miscarried,
These depressive symptoms can include staying in the baby was probably abnormal anyway’, ‘How can
bed and doing nothing, difficulty to perform daily you grieve so much, you were barely pregnant’ and
tasks, and a feeling of a physical illness. There is dis- ‘You can always conceive again’. While these per-
agreement, however, as to when these symptoms spectives may help some couples, many others do
decline. Several studies have found that 4 weeks after not want to forget their miscarried child at this
miscarriage, about half of the women were time, and resist the possibility that someday they
still depressed, and 18% of the women feared would feel as if the loss has never happened.6,17
another miscarriage to the extent that they consid- Based on studies of general infertility, friends and
ered not conceiving again.3 Others have shown family may feel guilty of their pregnancies and may
increased levels of depression as long as 6 months sometimes try to hide their pregnancy or talk less
after miscarriage.12,13 about their children, resulting in the couple feeling
An isolated miscarriage has little prognostic distanced from their friends, which can result in
value. Hence, one should be cautious in drawing social withdrawal.15 In addition, the couple may feel
inferences from a single miscarriage and applying that family and friends expect that they will shortly
these conclusions to RPL. In RPL, each additional conceive again to quickly replace their loss.
miscarriage reduces the prospects of having chil- Couples may also avoid social gatherings, par-
dren. Consequently, the repeated nature of RPL ties, and family occasions to avoid interactions with
may exacerbate the experience or teach couples to pregnant women or children. Some of them cannot
cope with it. Although never studied, the prognos- bear being expected to hold someone else’s child or
tic meaning the couple associates with the miscar- to listen to stories about the pleasures and difficul-
riage can further damage their sense of well-being. ties others are experiencing when raising children.
Many couples experiencing miscarriage undergo These often remind them of their loss.
a process of grieving10 (to be described later). They One way to compensate for the lack of social
mourn the lost child, their failed hopes for the support from family and friends is to seek couples
child, and their unaccomplished parenthood. who share similar experiences. However, unlike the
Unlike the grief over the death of a relative, these experience of an isolated spontaneous miscarriage,
244
0415421306-Ch17 3/30/07 9:49 AM Page 245
COPING WITH RECURRENT PREGNANCY LOSS: PSYCHOLOGICAL MECHANISMS
where many women have had a similar experience, Loss of self-esteem, guilt, and self-blame may be
women suffering from RPL usually do not know even more evident in women suffering from RPL.
other women in their situation and may lack some- Unlike many fertility problems, where the cause is
one to truly share their feelings with. Some of their either unknown or is attributed to both partners, in
closest friends may be pregnant or already have RPL, women feel that they are to blame because it
children, making it difficult for them to feel their was their body that betrayed and could not support
experience can be shared. Support groups are hard the pregnancy. This feeling is reinforced by the
to find and there are hardly any internet forums medical examinations that couples undergo: most
that are specific for RPL. clinical examinations evaluate possible etiologies in
Apart from being emotionally traumatic, mis- the women.
carriage can be physically traumatic as well; it may An aspect that is unique to RPL among fertility
involve sudden pain, loss of blood, rapid hospital- problems is the period when women are most
ization, and curettage.16 Some women identify the stressed and anxious. In most fertility problems,
physical process of miscarriage as the most stressful getting pregnant is the aim, and, once achieved, the
aspect, and they may find it harder to cope with mission has largely been accomplished. In contrast,
each time. this period is usually the most stressful for women
There has been considerable research on vari- who have experienced RPL, and the anxiety level
ables that moderate the influence of miscarriage on may peak around gestational weeks when previous
well-being, some of which may vary with time since miscarriages occurred.6 This anxiety is reflected by
the loss. Identifying these moderators is essential to extreme sensitivity to body signals, and increased
understand the variability in response to the loss, fear that miscarriage will happen again. The deci-
and, more importantly, it points at potential targets sion to conceive again is often very hard, because
for psychological interventions. Some of these women have to consider whether they can bear
mediators are uncontrollable – for example, young another miscarriage. In my interviews with women
age is associated with lesser well-being,2 and a later with RPL, they have often spoken about times that
gestational week of miscarriage has harsher psycho- are problematic for them to conceive, such as major
logical consequences. However, other factors can be holidays when they have to face family members.
controlled and are associated with adverse well-
being: these include attributing high personal sig-
nificance to miscarriage, low investment in domains THE GRIEVING PROCESS
of life other than parenthood, and low satisfaction
in other aspects of life, such as work, lack of social Couples experiencing RPL will often grieve for their
support, lower emotional strength, and use of pas- lost children, their lost parenthood, their biological
sive coping strategies.10,18–20 In contrast, women failure, the loss of control over their life, and for the
who reported that the recurrent miscarriages taught possibility that they would not have biological chil-
them to place greater value on their relationship with dren.12,25 Unlike losing a child, the couples do not
their spouses and to change priorities or personal have memories of the baby, and their loss is often
goals scored higher on well-being. not acknowledged by society.10 There are no rituals
Coping with infertility has been much explored associated with mourning a miscarriage. Couples
over the last 50 years. Many researchers describe may feel reluctant to share the experience with
infertility as a crisis having psychological effects, others, often cannot take days off from work, and
including loss of self-esteem, increased anxiety, may lack the time they would like to grieve for the
sexual problems, anger, depression, and self- loss. Couples may also be torn between their hopes
blame.21–25 The uncertainty of having biological for a successful pregnancy and their grief.
children evokes a sense that life is unpredictable and This grief process is often characterized by
that significant events in life are not under control. intense fluctuations in emotions, ranging from
245
0415421306-Ch17 3/30/07 9:49 AM Page 246
crying to laughing to being angry. This grief process term, and the medical team. Social support and
may last for months and even years, and often respect can help abate this anger.
extends into the subsequent pregnancies that serve 3. Isolation.27 Many couples exhibit social
as reminders for previous losses and can trigger withdrawal. This often happens because couples
intense emotions. Many couples may be very sur- try to hide their pregnancies and miscarriages,
prised by their mood swings and the intensity of the do not want to be judged or pitied by others, or
emotions that they experience. They may not be avoid occasions where they might meet children
aware that this is a normal reaction to their loss. or pregnant women. They also feel that their
It is very important to reassure them that their experience is unique, and that others whose
reaction is normal and common. experience of being pregnant is joyous cannot
Although there is no single right way to grieve, comprehend what they are going through.
several stages of grief are commonly experienced Moreover, peers and family members often
by people. There is disagreement whether all avoid discussing the recurrent miscarriages
people pass through each of these stages, and with the couples – either because they are
people differ in the time they spend at each stage. embarrassed with the ease of their having
The following list of the stages is mostly based on children or because they do not want to disrupt
Menning’s experience in his work as a counselor the couple’s privacy.
with infertile couples.26 4. Guilt.8,27 Women sometimes feel that the
recurrent miscarriages represent punishment for
1. Denial, shock and numbness.27,28 This stage often something they did. They may regret actions
begins with the shock that another miscarriage they took or failed to take prior to the
has occurred and is characterized by the feeling miscarriages.
that ‘this can’t be happening to me’. Sometimes, 5. Depression.4,8 At this stage, there is full
the couples will not even admit to themselves penetration of the distress and facing the loss.
that something may be wrong. This reaction Thoughts such as ‘My life is over, I can’t go on’ or
serves as a defense mechanism, and will usually ‘I don’t care any more’ are very frequent. Some
diminish as couples begin to acknowledge their women may feel a sense of great loss, mood
loss, usually within hours to days. This fluctuations, and loneliness.
emotional numbness and denial should not be 6. Rebuilding and healing. There is disagreement
confused with ‘lack of caring’. whether complete healing can occur. Still, at this
2. Anger.8,27 During this stage, the couple is stage, the couple start to deal with the reality of
preoccupied with the miscarriages that they have the situation. They restructure the event, organize
had. A feeling of unfairness surrounds these their activities, and plan to move forward in life,
thoughts: ‘Why me? Why us?’ The couple also and become more energetic and social.
experiences an intense yearning for the lost
child, for the lost parenthood, pregnancy, and
dreams. The anger associated with the WHAT DOMAINS OF LIFE ARE AFFECTED?
unfairness of the entire experience can focus on
the pain and inconvenience associated with RPL can affect many domains in a couple’s life:
miscarriage, with the tests and treatments, with from self-esteem to relationship with others, and
the social pressure they feel from their family even to financial costs. Here is a list of the main
and friends, and on comments regarding their domains that are affected:
miscarriages and childlessness. The anger may
also include broader targets such as abortion ● Self-esteem.21,25 Most people view the ability to
rights advocates, people who easily carry to conceive and have children as central to their
246
0415421306-Ch17 3/30/07 9:49 AM Page 247
personal identity. Our socialization process else with whom to have children. Pressures to
teaches girls and women to view motherhood as have children from the husband’s family can
an integral part of their self-worth and further exacerbate this fear.
femininity.29 In several religions, including the ● Sexual life.30,32 RPL, like other fertility problems,
Jewish communities where I conducted my may increase sexual discontent. Couples may feel
research, ‘Reproduce and fill the earth’ is one of a pressure to quickly conceive again, and with it
the most important precepts. Consequently, not an increased demand to have sex at certain times.
reproducing is often perceived in traditional Not being in the mood or being absent due to
societies as a degrading failure – impinging on various reasons such as business trips may
self-esteem and putting in question the woman’s increase the tension.
femininity and worth as a spouse. ● Financial cost. In addition, RPL frequently taxes
● Loss of control.6,15,25 For many women, RPL is the couples with financial costs: visits to a specialist,
first experience of a major loss of control: they tests, treatment, and absence from work.
lose control over their life, their body, and their
ability to plan the future. Some of this planning
includes the time of conception (e.g., the best THE MALE PARTNER
time to be absent from work, when it fits well
with their and their spouse’s career plans), and Spouses are often very lonely in their experience of
plans for adequate housing for an expanding RPL. Women are, after all, considered the main
family. patients – they experience the physical miscarriage,
● Relationship with peers.15,25 Couples may feel their reproductive system is assumed to hold the
excluded from friends whose interests focuses on cause, and they are subject to most diagnostic tests.
children, and may seek new reference groups to The idea that the spouse may also experience
belong to. Difficulties in facing pregnant women intense grief is often forgotten by society, by the
and young children also lead couples to avoid couple’s acquaintances, and by the medical team.
peers who are pregnant or already have children. Compared with women, the grief of male partners
In addition, their friends may feel uncomfortable is less active and is expressed for a shorter dura-
in disclosing their pregnancies, and this may be tion.33 Men are often ready to carry on with their
misinterpreted by the couple as a sign of lives earlier than women, and are also less interested
alienation. In contrast to many other fertility in discussing the miscarriage repeatedly.9, 21
problems and despite the fact that RPL is not so Spouses frequently find themselves in a very
rare, many women with RPL do not know other delicate position: at the same time, they endure a
women in the same situation to whom they can crisis, grieve, and need support, they feel that they
relate. ought to be strong to emotionally support their
● Marital stress.4,15,30 While the experience of partners. As a result, spouses suppress their feelings
infertility can improve marital adjustment for of loss instead of sharing it with their partners.
some couples, it may damage the relationships of Moreover, if the woman is depressed, they often feel
others and increase marital stress.24,30,31 This may that they are not doing a good job of supporting
result from differences between the spouses in her.33 They may struggle to say the right words, and
the attitude toward the losses, in the grief fear that what they say would make their partners
response, and in their motivation to have feel worse. Many of them fail to realize that their
children. In addition, women may feel guilty for female partners want to know that their grief is
failing their spouse’s expectations, and may feel shared by others. Also, although the spouses may
responsible for his pain. Many women fear that have the best intentions of providing support, there
their partner would leave them to find someone are sex differences in coping strategies with life
247
0415421306-Ch17 3/30/07 9:49 AM Page 248
stressors,34 and males tend to give instrumental However, there is still no evidence to support this
rather than social support, leaving women feeling notion in RPL.36
unsupported and the male partners feeling guilty
and unappreciated.
VALUE OF PSYCHOLOGICAL SUPPORT
IN COUPLES WITH RPL
THE PHYSICIAN
Clearly, the experience of RPL increases levels of
Although the physician and the couple share the distress, depressive symptoms, and anxiety. To
desire for pregnancy to succeed, the cooperation lower the emotional burden, couples often with-
between them is complex and may be very vulnera- draw from friends. RPL can affect almost every
ble. The challenge with which the physician is con- aspect of life, and the emotional burden usually
fronted when first seeing couples with RPL is almost becomes heavier during pregnancy. Obviously,
impossible. Usually, the time that can be devoted to these couples could benefit from psychological sup-
each couple is very limited when the routine compo- port. Although there is no one path that fits the
nents of a medical consultation are considered: needs of all couples, the following are some options.
taking a history, sorting and interpreting the results
of previous investigations, explaining the problem ● Support groups. Support is viewed as most
and its possible causes, subsequent prognosis, sug- credible coming from someone who has previo-
gesting additional investigations, answering the usly experienced and successfully managed a
couple’s questions, and showing sympathy. Often, similar crisis. In contrast to many other medical
this is a time when the couple’s anxiety and stress are conditions, couples with RPL often do not know
very intense, and they are very attentive and sensitive similarly afflicted people with whom to openly
to every word and gesture. Their first visit to the spe- share their feelings, thoughts, and concerns.
cialist can evoke many emotions: frustration, anger, Internet support groups usually lump together
stress, and inadequacy. This visit reminds them of women with one miscarriage and women with
past miscarriages, confronts them with their lack of several. Forums for infertility are more focused
control, clarifies that they should prepare for more on fertility treatments than on anxieties of
miscarriages, and confirms that they have a medical women with previous miscarriages. Meeting
condition that might leave them childless. other couples with RPL (past or present) can
Physicians are often unaware that the high stress that decrease the sense of loneliness, and reassure
the couple experiences interferes with their ability to couples that their reactions and feelings are
process the information received at the visit. This is normal.
a very common experience for many patients under- ● Teaching couples about the grieving process. This
going diagnosis: they often cannot recall what the can help them realize that their reaction to the
physician said, and tend to misinterpret what has grief process is normal and is experienced by
been told to them. many other couples. It can also help couples
Couples hope to identify a cause for their miscar- accept their grief, and proceed with it in their
riages. Understanding the cause, from their point of own way and pace.
view, means that a treatment can be offered. It was ● Activities for reducing anxiety. Physical activity,
suggested that women assign a cause to the miscar- art, meditation, relaxation and, yoga can reduce
riage themselves, when one is not assigned by the general anxiety in a non-specific manner.
doctor.16 Self-diagnoses include stress, certain foods, ● Cognitive restructuring.37 The individual interp-
and too much or too little exercise. This may reflect an retation of RPL influences the emotions evoked
attempt to regain a sense of control. It is thought that by this experience. Some of the negative
when the cause is detected, there is less self-blame.35 thoughts invoked are automatic and erroneous.
248
0415421306-Ch17 3/30/07 9:49 AM Page 249
Challenging these thoughts and restructuring they need to deal with the pain and grief. They
them into more truthful and positive thinking often feel that everyone is expecting them to
can improve well-being. Such techniques have quickly continue and to try again. They feel that
been shown to diminish stress, anxiety, depr- others deny them the legitimacy to say ‘I don’t
ession, and self-blame, and to increase enjoy- want to try again’. Raising the option to take a
ments in everyday life, in having each other, in break or to stop conceiving by the medical team
work, etc. An example of a common automatic may help relieve some of such pressure from
thought in women with RPL may be ‘I’ll never some women.
have any children.’ This thought is definitely not
true, and should be challenged. Some examples Since emotional anxiety tends to peak during
may be ‘This process is very painful for me, pregnancy, therapy should also be targeted to that
but there is a chance that I will eventually period. Although many of the above strategies can
have children.’ In addition, the significance only be realistically offered between pregnancies,
attributed to having biological children can be many can help to cope better with emotional
reframed. difficulties in subsequent pregnancies. Relaxation
● Improving dialogue with spouse. Sometimes, techniques can be employed whenever a woman
spouses fail to recognize what their partners are recognizes an increase in her stress levels, and
going through; this may create a cycle of cognitive restructuring can help maintain posi-
disrupted communication that decreases a tive thoughts and avoid the loop of negative
couple’s enjoyment in doing things together and thoughts.
increases their marital stress. It is therefore
important to encourage couples to have a fruitful
dialogue, by learning to listen more to each CAN STRESS CONTRIBUTE TO
other, by acknowledging the feelings of each RECURRENT MISCARRIAGE?
other, by being aware that they may be using
different coping strategies, and by recognizing A common question that bothers couples is whether
each other’s needs. excessive stress can adversely affect pregnancy and
● Learning of other parenting options. Although not lead to miscarriage. A belief in such a relationship
all couples feel ready to explore other means to can increase feelings of guilt and self-blame and
achieve parenthood, many could benefit from further increase stress in a self-perpetuating circle.
meeting couples who have chosen to adopt or Although it is a sensitive matter, this is an important
use the aid of a surrogate mother. This not only question to study. This question is difficult to exam-
informs them of the procedures and the ine, since retrospective reports of stress are skewed
emotions associated with choosing other paths, by the already-experienced outcome, since many
but it also confronts them with ‘their worst women in the general population miscarry due to
nightmare’. Although they may not decide to abnormal chromosomes, and since women at a high
follow these paths, couples often realize this is risk for miscarriage often experience high levels of
not as bad an option as they have imagined, and stress during pregnancy.
some of the fear that is associated with infertility Prospective human studies on the effects of
may be relieved. stress on miscarriage or IVF success are rare: some
● Discussing legitimacy. Many women with RPL have suggested a causative relation, some have indi-
report that they feel it is illegitimate to stop trying cated correlations, while others have found no
to conceive or to choose alternative means for association.38–43 With regard to miscarriage, a dis-
parenthood. Many feel like that they invest a lot tinction is not usually made between unexplained
of their energy in conceiving and reconceiving, in miscarriage and miscarriage due to chromosomal
hurrying to become parents, but at the same time abnormalities. One study that attempted to separate
249
0415421306-Ch17 3/30/07 9:49 AM Page 250
the two groups found a correlation between stress

Table 17.1 Pregnancy outcome in women with
levels and miscarriage only in cases thought not recurrent pregnancy loss after supportive care
to involve chromosomal abnormalities.44 This,
together with the fact that miscarriage is a dichoto- Success rates
mous variable, would necessitate a very large
Study Control group Supportive care group
sample to detect a correlation.
Studies in rodents have suggested that stress Stray-Pedersen and 8/24 (33%) 32/37 (86%)
increases the rates of implantation and resorption. Stray-Pedersen52
Clifford et al53 20/41 (49%) 118/160 (74%)
Exposing pregnant rats or mice to stress can result in Liddel et al7 3/9 (33%) 38/44 (86%)
lower pregnancy rates, higher embryonic death, Tupper and Weil54 5/19 (26%) 16/19 (84%)
more resorptions, and smaller litters.45–48 Adreno- Brigham et al55 167/222 (75%)
corticotropic hormone (ACTH) treatment for the Total 36/93 (38%) 371/482 (77%)
first 8 days of pregnancy reduced the number of
implantation sites in naive/sham operated as well as
adrenalectomized mice,49 suggesting a direct role for During the second pregnancy, these women are
this hormone. more stressed, which boosts the risk of another
Overall, based on animal models and on correl- miscarriage. If another miscarriage occurs, this
ative studies in humans, there is some evidence that increases their stress, and their chances of another
stress can adversely affect fertility in general. miscarriage.
However, it is unclear whether this effect can be
extended to recurrent miscarriage. Two studies in
women with RPL found that depressive symptoms PSYCHONEUROIMMUNOLOGY
and low satisfaction with social support are predic- AND ITS RELEVANCE
tive of subsequent miscarriage.50,51 Another prosp-
ective study, though, has failed to find such an There are several potential neuroendocrinological
association with regard to perceived stress.42 pathways through which stress might promote mis-
The best support for the contribution of psycho- carriage. However, an interesting pathway that has
logical factors to RPL comes from studies that have captured the attention of several investigators is the
evaluated the effect of psychological support or psychoneuroimmunologic (PNI) path.56, 57 Over the
therapy in women suffering from RPL.7,52–55 past 30 years, it has become clear that the immune
Interventions ranged from basic ‘tender loving care’ system is not autonomous but has bidirectional
to relaxation workshops and audiocassettes, weekly connections with the central nervous system.58–60
ultrasound examinations (to assure the woman that It has been shown that the immune responses can
the embryo develops appropriately), and other psy- be behaviorally conditioned, that various emotional
chological interventions. Remarkably, all four stud- and cognitive states can influence both cellular and
ies reported that women who received psychological humoral immunity, and that cytokines can affect
support had two- to fourfold lower miscarriage rates neural function.
than those who did not (Table 17.1). Although these The effects of psychological stress on various
studies suffer from methodological problems, it is immune measures have been extensively studied.
doubtful whether these flaws can account for such a In most cases, stress interferes with the normal
marked reduction in miscarriage rates (on average function of the immune system rather than assist-
from 62% to 23%). ing it. It has also been demonstrated that this per-
If stress does indeed contributes to miscarriage turbation can result in actual consequences to
in women with RPL, it could lead couples into a health, for example lowering resistance to infections
vicious circle. The first miscarriage could be due to and slowing wound healing.60,61 This is especially
some biological cause such as abnormal karyotype. the case with chronic and severe stress.
250
0415421306-Ch17 3/30/07 9:49 AM Page 251
Interestingly, the subset of immune cells that seem Although management of physician’s emotions is
to be most affected by stress are natural killer (NK) not considered part of the physician’s role, I believe
cells,61 the cells thought to be involved in the etiology that adopting an inclusive psychosocial perspective
of RPL. NK cells seem to carry the greatest density of would greatly improve the treatment of couples with
adrenergic receptors, and are thus more susceptible RPL. The anxiety, depression, anger, and frustration
to the influence of the sympathetic nervous system.62 these couples experience are critically influenced by
These receptors contribute to direct suppression of how significant they regard their miscarriages, by
NK-cell activity, detachment from endothelial cells, how their family, friends, and society perceive these
and redistribution after exposure to stress. The miscarriages, and by how much emotional support
number and activity of circulating NK cells were they receive. A supportive and empathic approach
reported to be highly affected by stressors such as by the medical team can ease this suffering, and
academic examination, exposure to disastrous hurri- psychological interventions can be used to improve
canes, interpersonal stress, first parachuting jumps, couples’ coping and enhance their well-being.
and marital disputes. Such interventions may not only relieve the emo-
From the PNI perspective, stress might be pro- tional burden of RPL but also lower the risk of
moting miscarriage by interfering with the uterine another miscarriage. Although some clinicians may
immunological conditions that protect pregnancy. dismiss such effects, the evidence for such a possibil-
In several experiments in mice, Arck et al63 have ity exceeds the support for several medical interven-
shown that stress around the fifth day of pregnancy tions already employed in RPL. Larger randomized
more than tripled the resorption rates in miscar- studies should examine this possibility more care-
riage-prone mice; depletion of NK cells prevented fully. Until proven, the psychosocial hypothesis
this effect. Studying women, we have recently should be raised with caution, as it can lead women
shown that the number and activity of peripheral to blame themselves for the miscarriage.
NK cells in RPL, which have previously been shown
to predict the outcome of subsequent pregnancy, is
REFERENCES
a transient response to the blood withdrawal.64 A
cannula was inserted into the veins of women with 1. Aoki K, Furukawa T, Ogasawara M, et al. Psychosocial factors in
RPL and controls, and blood was drawn immedi- recurrent miscarriages. Acta Obstet Gynecol Scand 1998; 77:572–3.
ately and 20 minutes later. NK-cell activity and cell 2. Neugebauer R. Depressive symptoms at two months after miscar-
riage: interpreting study findings from an epidemiological versus clin-
number were increased in RPL patients in the first ical perspective. Depress Anxiety 2003; 17:152–61.
blood withdrawal, but declined to a level similar to 3. Friedman T, Gath D. The psychiatric consequences of spontaneous
that of the control in the second blood withdrawal. abortion. Br J Psychiatry 1989; 155:810–13.
4. Klock SC, Chang G, Hiley A, et al. Psychological distress among
These levels remained almost unchanged in the women with recurrent spontaneous abortion. Psychosomatics 1997;
control groups. This may suggest that the increased 38:503–7.
NK-cell activity and numbers often observed 5. Craig M, Tata P, Regan L. Psychiatric morbidity among patients with
recurrent miscarriage. J Psychosom Obstet Gynaecol 2002; 23:157–64.
in women with RPL reflect hypersensitivity to 6. Cote-Arsenault D, Morrison-Beedy D. Women’s voices reflecting
the stress of blood withdrawal rather than the changed expectations for pregnancy after perinatal loss. J Nurs
immunological steady state. It remains to be deter- Scholarsh 2001; 33:239–44.
7. Liddell HS, Pattison NS, Zanderigo A. Recurrent miscarriage – outcome
mined whether such hypersensitivity is also predic- after supportive care in early pregnancy. Aust NZ J Obstet Gynaecol
tive of pregnancy outcome. 1991; 31:320–2.
8. Madden ME. The variety of emotional reactions to miscarriage.
Women Health 1994; 21:85–104.
9. Athey J, Spielvogel AM. Risk factors and interventions for psycholog-
SUMMARY ical sequelae in women after miscarriage. Prim Care Update Ob/Gyns
2000; 7:64–9.
10. Lee C, Slade P. Miscarriage as a traumatic event: a review of the liter-
More than any other fertility problem, RPL submits ature and new implications for intervention. J Psychosom Res 1996;
patients to repeated cycles of hope and despair. 40:235–44.
251
0415421306-Ch17 3/30/07 9:49 AM Page 252
11. Adolfsson A, Larsson PG, Wijma B, et al. Guilt and emptiness: 34. Jordan C, Revenson TA. Gender differences in coping with infertility:
women’s experiences of miscarriage. Health Care Women Int 2004; a meta-analysis. J Behav Med 1999; 22:341–58.
25:543–60. 35. Nikcevic AV, Tunkel SA, Kuczmierczyk AR, et al. Investigation of the
12. Nikcevic AV, Tunkel SA, Nicolaides KH. Psychological outcomes fol- cause of miscarriage and its influence on women’s psychological dis-
lowing missed abortions and provision of follow-up care. Ultrasound tress. Br J Obstet Gynaecol 1999; 106:808–13.
Obstet Gynecol 1998; 11:123–8. 36. Rowsell E, Jongman G, Kilby M, et al. The psychological impact of
13. Janssen HJ, Cuisinier MC, Hoogduin KA, et al. Controlled prospective recurrent miscarriage, and the role of counselling at a pre-pregnancy
study on the mental health of women following pregnancy loss. Am J counselling clinic. J Reprod Infant Psychol 2001; 19:33–45.
Psychiatry 1996; 153:226–30. 37. Beck JS. Cognitive Therapy: Basic and Beyond. New York: Guilford
14. Klier CM, Geller PA, Ritsher JB. Affective disorders in the aftermath of Press, 1995.
miscarriage: a comprehensive review. Arch Womens Ment Health 38. Domar AD, Clapp D, Slawsby EA, et al. Impact of group psychologi-
2002; 5:129–49. cal interventions on pregnancy rates in infertile women. Fertil Steril
15. Imeson M, McMurray A. Couples’ experiences of infertility: a phe- 2000; 73:805–11.
nomenological study. J Adv Nurs 1996; 24:1014–22. 39. Merari D, Feldberg D, Elizur A, et al. Psychological and hormonal
16. Bansen SS, Stevens HA. Women’s experiences of miscarriage in early changes in the course of in vitro fertilization. J Assist Reprod Genet
pregnancy. J Nurse Midwifery 1992; 37:84–90. 1992; 9:161–9.
17. Cecil R. ‘I wouldn’t have minded a wee one running about’: 40. Boivin J, Takefman JE. Stress level across stages of in vitro fertilization
Miscarriage and the family. Soc Sci Med 1994; 38:1415–22. in subsequently pregnant and nonpregnant women. Fertil Steril 1995;
18. Mindes EJ, Ingram KM, Kliewer W, et al. Longitudinal analyses of the 64:802–10.
relationship between unsupportive social interactions and psycholog- 41. Facchinetti F, Matteo ML, Artini GP, et al. An increased vulnerability to
ical adjustment among women with fertility problems. Soc Sci Med stress is associated with a poor outcome of in vitro fertilization–embryo
2003; 56:2165–80. transfer treatment. Fertil Steril 1997; 67:309–14.
19. Brier N. Understanding and managing the emotional reactions to a 42. Bergant AM, Reinstadler K, Moncayo HE, et al. Spontaneous abortion
miscarriage. Obstet Gynecol 1999; 93:151–5. and psychosomatics. A prospective study on the impact of psycholog-
20. Swanson KM. Predicting depressive symptoms after miscarriage: a ical factors as a cause for recurrent spontaneous abortion. Hum
path analysis based on the Lazarus paradigm. J Womens Health Gend Reprod 1997; 12:1106–10.
Based Med 2000; 9:191–206. 43. Boivin J, Andersson L, Skoog-Svanberg A, et al. Psychological reac-
21. Pasch LA, Dunkel-Schetter C, Christensen A. Differences between tions during in-vitro fertilization: similar response pattern in hus-
husbands’ and wives’ approach to infertility affect marital communi- bands and wives. Hum Reprod 1998; 13:3262–7.
cation and adjustment. Fertil Steril 2002; 77:1241–7. 44. Boyles SH, Ness RB, Grisso JA, et al. Life event stress and the associa-
22. Greil AL. Infertility and psychological distress: a critical review of the tion with spontaneous abortion in gravid women at an urban emer-
literature. Soc Sci Med 1997; 45:1679–704. gency department. Health Psychol 2000; 19:510–14.
23. Alesi R. Infertility and its treatment – an emotional roller coaster. Aust 45. Wiebold JL, Stanfield PH, Becker WC, et al. The effect of restraint
Fam Physician 2005; 34:135–8. stress in early pregnancy in mice. J Reprod Fertil 1986; 78:185–92.
24. Schover LR. Recognizing the stress of infertility. Cleve Clin J Med 46. Euker JS, Riegle GD. Effects of stress on pregnancy in the rat. J Reprod
1997; 64:211–14. Fertil 1973; 34:343–6.
25. Gonzalez LO. Infertility as a transformational process: a framework 47. Arck PC, Merali F, Chaouat G, et al. Inhibition of immunoprotective
for psychotherapeutic support of infertile women. Issues Ment Health CD8+ T cells as a basis for stress-triggered substance P-mediated abor-
Nurs 2000; 21:619–33. tion in mice. Cell Immunol 1996; 171:226–30.
26. Menning BE. The emotional needs of infertile couples. Fertil Steril 48. Arck PC, Merali FS, Manuel J, et al. Stress-triggered abortion: inhibi-
1980; 34:313–19. tion of protective suppression and promotion of tumor necrosis
27. Matthews AM, Matthews R. Beyond the mechanics of infertility – factor-α (TNF-α) release as a mechanism triggering resorptions in
perspectives on the social-psychology of infertility and involuntary mice. Am J Reprod Immunol 1995; 33:74–80.
childlessness. Family Relations 1986; 35:479–87. 49. Kittinger JW, Gutierrez-Cernosek RM, Cernosek SF, Jr, et al. Effects of
28. Lee C, Slade P, Lygo V. The influence of psychological debriefing adrenocorticotropin on pregnancy and prolactin in mice.
on emotional adaptation in women following early miscarriage: a Endocrinology 1980; 107:616–21.
preliminary study. Br J Med Psychol 1996; 69:47–58. 50. Nakano Y, Oshima M, Sugiura-Ogasawara M, et al. Psychosocial
29. Becker G, Nachtigall RD. ‘Born to be a mother’: the cultural construc- predictors of successful delivery after unexplained recurrent
tion of risk in infertility treatment in the U.S. Soc Sci Med 1994; spontaneous abortions: a cohort study. Acta Psychiatr Scand 2004;
39:507–18. 109:440–6.
30. Monga M, Alexandrescu B, Katz SE, et al. Impact of infertility on 51. Sugiura-Ogasawara M, Furukawa TA, Nakano Y, et al. Depression as a
quality of life, marital adjustment, and sexual function. Urology 2004; potential causal factor in subsequent miscarriage in recurrent sponta-
63:126–30. neous aborters. Hum Reprod 2002; 17:2580–4.
31. Connolly KJ, Edelmann RJ, Cooke ID, et al. The impact of infertility 52. Stray-Pedersen B, Stray-Pedersen S. Etiologic factors and subsequent
on psychological functioning. J Psychosom Res 1992; 36:459–68. reproductive performance in 195 couples with a prior history of
32. Seibel MM, Taymor ML. Emotional aspects of infertility. Fertil Steril habitual abortion. Am J Obstet Gynecol 1984; 148:140–6.
1982; 37:137–45. 53. Clifford K, Rai R, Regan L. Future pregnancy outcome in unexplained
33. Beutel M, Willner H, Deckardt R, et al. Similarities and differences recurrent first trimester miscarriage. Hum Reprod 1997; 12:387–9.
in couples’ grief reactions following a miscarriage: results from a lon- 54. Tupper C, Weil RJ. The problem of spontaneous abortion. Am J
gitudinal study. J Psychosom Res 1996; 40:245–53. Obstet Gynecol 1962; 83:421–4.
252
0415421306-Ch17 3/30/07 9:49 AM Page 253
55. Brigham SA, Conlon C, Farquharson RG. A longitudinal study of 60. Glaser R, Kiecolt-Glaser JK. Stress-induced immune dysfunction:
pregnancy outcome following idiopathic recurrent miscarriage. implications for health. Nat Rev Immunol 2005; 5:243–51.
Hum Reprod 1999; 14:2868–71. 61. Cohen S, Herbert TB. Health psychology: psychological factors and
56. Clark DA, Arck PC, Jalali R, et al. Psycho–neuro–cytokine/endocrine physical disease from the perspective of human psychoneuroim-
pathways in immunoregulation during pregnancy. Am J Reprod munology. Annu Rev Psychol 1996; 47:113–42.
Immunol 1996; 35:330–7. 62. Landmann R. Beta-adrenergic receptors in human leukocyte subpop-
57. Arck PC. Stress and pregnancy loss: role of immune mediators, ulations. Eur J Clini Invest 1992; 1:30–6.
hormones and neurotransmitters. Am J Reprod Immunol 2001; 46: 63. Arck PC, Merali FS, Stanisz AM, et al. Stress-induced murine abortion
117–23. associated with substance P-dependent alteration in cytokines in
58. Ader R. On the development of psychoneuroimmunology. Eur J maternal uterine decidua. Biol Reprod 1995; 53:814–19.
Pharmacol 2000; 405:167–76. 64. Shakhar K, Rosenne E, Loewenthal R, et al. High NK activity in recur-
59. Maier SF, Watkins LR, Fleshner M. Psychoneuroimmunology. The rent miscarriage: What are we really measuring? Hum Reprod 2006;
interface between behavior, brain, and immunity. Am Psychol 1994; 21:2421–5.
49:1004–17.
253
0415421306-Ch17 3/30/07 9:49 AM Page 254
0415421306-Ch18 3/30/07 9:50 AM Page 255
18. Methodological issues in

evidence-based evaluation of
treatment for recurrent miscarriage
Salim Daya
INTRODUCTION some instances to erroneous inferences and misleading

recommendations for clinical care.
The philosophy of using evidence from valid and Recently, the Special Interest Group for Early
current studies to assist in the clinical decision- Pregnancy, under the auspices of the European
making process is now widely acknowledged as Society for Human Reproduction and Embryology,
desirable for improving the quality of care provided updated their guidelines for the investigation and
to patients. The underlying principle of this evi- medical treatment of recurrent miscarriage.1
dence-based approach to healthcare management is Unfortunately, the paucity of good-quality evidence
having access to reliable and valid evidence that is led to the conclusion that many of the proposed
obtained by either searching the literature for investigations require further evaluation within
papers that can be critically appraised or designing research programs. In addition, ‘tender loving care’
studies of high methodological rigor. The resulting and health advice were the only interventions that
best available evidence is then used to answer did not require further study; most of the other
clearly defined and focused questions generated proposed therapies either require more investiga-
from encounters with patients presenting with their tion of their efficacy with randomized trials or are
clinical problems. This approach is now guiding associated with more harm than benefit.1
management in obstetrics and gynecology, includ- Reliable inferences regarding therapeutic inter-
ing the subspecialty of infertility dealing with the ventions can only be drawn from trials that have
problem of recurrent miscarriage. addressed all the elements necessary for internal
Over the years, increased attention has focused validity. The important issues include the folowing:
on the evaluation and management of recurrent
miscarriage. Protocols have been developed for diag- ● definition of recurrent miscarriage and its
nostic evaluation in couples with recurrent miscar- subgroups
riage so that a plan of care can be outlined based on ● avoidance of including women with repeated
the findings. However, to date, there is no consensus implantation failure
on the optimal evaluation and management strategy ● establishing the baseline risk for miscarriage so
to effectively address the problem of recurrent mis- that a control event rate can be obtained
carriage. The situation is made more challenging by ● controlling for female age and for male partner
the fact that the published literature is generally of status
poor quality and often has contradictory findings, in ● formulation of an appropriate research question
part due to sampling variability, but largely due to ● use of randomization
studies of low validity. The approach of systemati- ● importance of concealment of treatment
cally gathering the evidence and pooling outcome allocation and blinding
data with meta-analyses is an attempt to bring some ● avoidance of co-intervention
order to this field, but it too has its pitfalls, leading in ● requirement for an adequate sample size
255
0415421306-Ch18 3/30/07 9:50 AM Page 256
● avoidance of using historical controls The term ‘recurrent miscarriage’ has now replaced
● importance of analyzing data using an intention- the original term ‘habitual abortion’ because it is
to-treat approach kinder and does not imply that the woman is delib-
● clear identification of the orientation of the study erately causing her pregnancies to be terminated.
from the perspectives of superiority, equivalence, Recurrent miscarriage defines a clinical condition in
or non-inferiority of the interventions being which a woman has had at least three miscarriages.
compared However, because the pregnancy history in women
● clarity on when treatment should be commenced with recurrent miscarriage may include pregnancies
● avoiding postrandomization withdrawals that have ended in live birth, three different groups
● submitting products of conception for karyotypic can be identified. The groups should be assessed sep-
analysis in cases of intervention failure. arately because the risk of subsequent miscarriage
within each group varies.2
These requirements will be discussed in this chapter
by highlighting the common pitfalls that are ● Primary recurrent miscarriage group. This group
encountered. Also, by addressing these issues, it is consists of women with three or more consecutive
hoped the reader will become more versed in miscarriages with no pregnancy progressing
reviewing the literature on management of recur- beyond 20 weeks’ gestation.
rent miscarriage so that the judicious and explicit ● Secondary recurrent miscarriage group. This
use of the best current evidence can guide clinical group consists of women who have had three or
management. more miscarriages after a pregnancy that, having
gone beyond 20 weeks’ gestation, may have
ended in live birth, stillbirth, or neonatal death.
METHODOLOGICAL ISSUES IN ASSESSMENT OF ● Tertiary recurrent miscarriage group. This is a
EVIDENCE FROM THERAPEUTIC TRIALS IN group that has not been well characterized or
RECURRENT MISCARRIAGE studied, and consists of women who have had at
least three miscarriages that are not consecutive
DEFINITION OF RECURRENT MISCARRIAGE but are interspersed with pregnancies that have
(DEFINING THE POPULATION) progressed beyond 20 weeks’ gestation (and may
have ended in live birth, stillbirth, or neonatal
The term ‘miscarriage’ is used to describe a pregnancy death).
that fails to progress, resulting in death and expulsion
of the embryo or fetus. The generally accepted defini- These three groups are mutually exclusive and dis-
tion stipulates that the fetus or embryo should weigh tinct, and should be evaluated separately because
500 g or less,2,3 a stage that corresponds to a gesta- the group being selected will undoubtedly influ-
tional age of up to 20 weeks. Unfortunately, this defi- ence the prognosis for a successful outcome. The
nition is not used consistently, and pregnancy losses current approach of combining all three groups
at higher gestational ages are also classified as miscar- together does not allow the effect of the experimen-
riage in some countries. Additionally, the literature is tal intervention to be detected easily, because the
replete with studies on women with pregnancy loss – prognosis is determined by the relative contribution
a term that includes miscarriage and pregnancies that of subjects from each of the three groups.
have ended in stillbirth or preterm neonatal death.
Thus, from a definitional perspective, it is important EXCLUSION OF IMPLANTATION FAILURES
to characterize the population being studied so that (AVOIDING CLINICAL HETEROGENEITY)
comparisons across therapeutic trials can be made
more appropriately and reliably. Consensus on this The widespread availability of treatment with
issue is urgently required. assisted reproduction has created a challenge for the
256
0415421306-Ch18 3/30/07 9:50 AM Page 257
EVIDENCE-BASED EVALUATION OF TREATMENT FOR RECURRENT MISCARRIAGE
management of women who repeatedly fail to con- increased; the chance of a fourth pregnancy going
ceive despite undergoing uterine transfer of good- to term in women with three previous miscarriages
quality embryos. Repeated implantation failure is considerably lower than that of a third pregnancy
(RIF) is now a recognized entity defined as failure going to term with two previous miscarriages.2 For
to achieve a pregnancy after at least three cycles of many years, these mathematical estimates of mis-
in vitro fertilization (IVF)4 in which at least 10 carriage rate were used as control rates against
high-grade embryos were transferred into the which the efficacy of therapeutic regimens for
uterus. It is now being suggested that RIF and recur- recurrent miscarriage was assessed. The reliability
rent miscarriage represent different ends of the of these rates was challenged after evidence from a
same disorder.5 This position is difficult to accept, number of clinical studies suggested that the mis-
because the former is a preimplantation failure that carriage rate after three consecutive miscarriages
results in no pregnancy whereas the latter is a was substantially lower than had been predicted by
postimplantation failure that results in no live birth. the earlier mathematical models.2
Although there may be some overlap in the two Despite the varied methods of ascertainment, the
conditions from the diagnostic protocol perspec- results of the studies showed remarkable consistency
tive, it is evident from the results of these tests that in finding a positive correlation between risk of mis-
the two entities are distinct and should not be com- carriage and the number of previous miscarriages.
bined. For example, studies of cytokine expression This effect of prior losses on subsequent probability
in the endometrium have produced conflicting and of live birth was confirmed using the data from the
sometimes contradictory findings in these two con- placebo arm of studies in unexplained recurrent
ditions.6 Similarly, there is no evidence that measur- miscarriage and provided a quantitative estimate of
ing serum levels of antiphospholipids is of benefit the risk.9,10 Similar risk estimates were obtained
in RIF, in contrast to measurement in women with from a longitudinal study of pregnancy outcome in
recurrent miscarriage.7,8 The results of studies such women with idiopathic recurrent miscarriage.11 It is
as these indicate that RIF and recurrent miscarriage clear from this evidence that the number of previous
are two distinct entities that should not be lumped miscarriages is an important covariate, which has to
together under the assumption that they represent be taken into account when planning therapeutic
different aspects of the spectrum of reproductive trials: women with a higher number of previous
failure. By investigating them separately and by miscarriages constitute a group with a more severe
conducting efficacy trials in each group separately, form of recurrent miscarriage than those with rela-
the problem of clinical heterogeneity is avoided and tively lower numbers of previous miscarriages.
the benefit (or lack thereof) of interventions can be Consequently, the magnitude of the treatment effect
evaluated more accurately. is expected to be much larger in these more severe
forms of the disorder (because the control event rate
BASELINE RISK OF MISCARRIAGE (ESTABLISHING is so much lower) and is likely to be more easily
THE CONTROL EVENT RATE) detected if the subjects are grouped by severity.12
Thus, the ideal trial should have stratification for the
Initial estimates of the likelihood of a successful number of previous miscarriages, with randomiza-
pregnancy in women with previous miscarriages tion of subjects to control or experimental interven-
were based on the assumption that the overall mis- tions being performed within each stratum.
carriage rate consists of the sum of two independ- To date, such a study with a priori stratification
ent rates: one resulting from a random factor and has not been undertaken. Instead, the general (and
the other from a recurrent factor in miscarriage incorrect) approach has been to select the study
sequences. Such mathematical calculations demon- sample from the population of women having three
strated a higher risk of miscarriage in a subsequent or more miscarriages and ignore the importance of
pregnancy as the number of previous miscarriages stratification for number of previous miscarriages.
257
0415421306-Ch18 3/30/07 9:50 AM Page 258
The consequence is a sample that is likely to consist inferences that are drawn from the results can be
of a higher proportion of women with lower num- communicated without ambiguity. To do so
bers of previous miscarriages, thereby reducing the requires formulating the research question that is
probability of detecting a significant treatment relevant to the problem at hand and is structured
effect. in four parts: the population being evaluated, the
experimental intervention being tested, the con-
CONTROLLING FOR FEMALE AGE trol intervention used as the comparator, and the
(REDUCING SELECTION BIAS) outcome that has clinical importance. A lack of
clarity in the objective formulation becomes evi-
The risk of miscarriage resulting from fetal chro- dent when the findings are discussed, because
mosomal anomalies increases with maternal often several different outcomes are considered.
age,13,14 especially after age 35. Additionally, women Attempts are then made at the end of the trial to
who have recurrent miscarriages tend to have more develop an explanation for the findings that has
pregnancies and have their pregnancies at a later strayed from the original idea for which the study
age than those who have successful outcomes. The was commissioned.
relationship of gravidity with female age and the
relationship of chromosomal anomalies and female RANDOMIZATION (ENSURING SIMILARITY
age suggest that the increased risk of miscarriage AMONG INTERVENTION GROUPS)
with gravidity, can in part, be ascribed to the effect
of maternal age. Thus, clinical trials of treatment The randomized controlled trial has become the
efficacy must take female age into consideration ‘gold standard’ in evaluating treatment efficacy.
during the design phase by using stratification for Randomization of subjects to receive either experi-
this covariate. This approach will avoid the possibil- mental or control intervention generates two
ity of bias that may show the treatment to be less groups that are generally similar in all respects
favorable if the experimental group has a higher except for the single factor (the intervention) being
proportion of older women than the control group. studied. This approach ensures that any significant
difference in the outcome between the two groups is
CONTROLLING FOR MALE PARTNER likely due only to this single factor. Also, by ensur-
(REDUCING SELECTION BIAS) ing their equal distribution in the two groups, it
guards against differences in factors not known to
Women who have recurrent miscarriages with one be important to the outcome of interest.
male partner may have successful pregnancies with There are many methods of randomization,
a different male partner. This issue of partner speci- including simple coin tossing, drawing straws, and
ficity is an important consideration in avoiding the use of computer-generated random number
selection bias when evaluating treatment efficacy. tables. The use of block randomization is an addi-
To ensure homogeneity of the sample and maxi- tional maneuver that produces equal numbers of
mize the probability of detecting a true treatment subjects in each group – a result not usually obtained
effect, couples should be chosen for the trial only if with the other methods of randomization. Another
the consecutive miscarriages experienced by the approach that is not infrequently used is that of qua-
subject have occurred with the same male partner. sirandomization, wherein subjects are allocated
using either the subject’s clinical chart number (even
CLEARLY DEFINED OBJECTIVE (ARTICULATING number for the experimental group and odd number
THE RESEARCH QUESTION) for the control group) or the subject’s date of birth
(first half of the year for the experimental group and
Before commencing a trial, it is important to artic- second half of the year for the control group), or the
ulate the objective clearly and concisely so that the day of the week when the subject is seen in the clinic
258
0415421306-Ch18 3/30/07 9:50 AM Page 259
(Monday, Wednesday, and Friday for the experimen- assignment on the random list with opaque tape
tal group, and Tuesday, Thursday, and Saturday for that is removed to reveal the group only when the
the control group). Additionally, alternation is often next eligible subject is enrolled, (ii) the use of con-
used to create two intervention groups by alternating secutively numbered opaque envelopes containing
the assignment between experimental and control the group assignment, and (iii) the use of an indi-
interventions for each successive subject enrolled in vidual not directly involved with the trial. Although
the trial (i.e., first subject allocated to the experimen- the first two methods are simple and pragmatic, they
tal group, second subject to the control group, third are not tamperproof and need to be policed to pre-
subject to the experimental group, and so on). When vent investigators from peeking ahead of time under
carried out properly, both quasirandomization and the tape or in the envelope to determine where their
alternation are fairly simple and effective methods preferred intervention is located in the random
for generating experimental and control groups. sequence of assignments. The use of a third party,
However, both methods have several pitfalls, including such as a telephone operator who is located at a site
the openness of the process and, in the case of quasi- distant from the study center and who can be con-
randomization, the allocation of unequal numbers of tacted at the time of enrolment, or a pharmacist who
subjects to each group. is responsible for dispensing the treatments, pro-
To improve validity of the trial and to minimize vides the highest level of security because it ensures
postrandomization withdrawals of subjects (for rea- that the randomization list is kept away from the
sons such as change of mind, relocation to another investigators.
city, and so on), it is important to perform random- The openness of the quasirandomization and
ization as late as possible, preferably just prior to the alternation methods makes them less reliable in
intervention being administered. preventing selection bias unless all eligible subjects
are enrolled sequentially. From a methodological
CONCEALMENT OF GROUP ALLOCATION perspective, the debate over the validity of using
(AVOIDING SELECTION BIAS) these methods in efficacy trials is still ongoing.
Selection bias is encountered when potentially eligi- BLINDING OR MASKING (AVOIDING

ble subjects are selectively excluded from the trial ASCERTAINMENT BIAS)
because of prior knowledge of the group to which
they would have been allocated had they partici- The response to an intervention may not be entirely
pated in the trial. Although randomization is gener- due to the active chemical compound administered
ally effective in creating equally balanced groups, it or the surgical procedure performed, but may be
does not guard against selection bias, because the influenced by other factors, such as the subject’s
investigator may have a notion of the efficacy (or expectations, the enthusiasm and reputation of the
lack thereof) of the experimental intervention and healthcare provider, and the nature of the interven-
may consciously or unknowingly steer subjects tion. Consequently, the outcome of a trial may be
towards or away from this intervention. An effective biased (ascertainment bias) if the subject, the inves-
strategy to avoid selection bias is to ensure informa- tigator, or the outcomes assessor has knowledge of
tion regarding the group allocation is concealed the intervention that the subject is receiving.
from the investigators and care providers until the Blinding (or masking) is a strategy that keeps those
subject is irreversibly committed to the trial. In the involved in the trial unaware of the identity of the
absence of concealment, it has been shown that the intervention and is used to prevent ascertainment
effect of an experimental intervention may be over- bias because it eliminates the influence (either pos-
estimated by as much as 40%.15,16 itive or negative) that any knowledge of the inter-
There are several methods to conceal group vention being administered may have on the
allocation, including (i) covering each consecutive estimation of the treatment effect.
259
0415421306-Ch18 3/30/07 9:50 AM Page 260
Blinding is not the same as allocation conceal- with recurrent miscarriage undergoing another
ment. The role of blinding is in safeguarding the ran- pregnancy19,20 suggests that in recurrent miscar-
domization sequence after allocation has been riage research the placebo effect is likely to be of sig-
performed. In addition, for subjects enrolled in the nificant magnitude for which appropriate measures
trial, blinding enhances their compliance with the should be taken when designing an efficacy trial.
treatment protocol and encourages them to com- In trials evaluating surgical procedures, the use
plete the trial. A subject who perceives the experi- of placebo poses a unique challenge.21 Placebo sur-
mental intervention to be better than the control gery (also known as a sham operation) requires the
intervention may be less willing to remain in the subject to undergo all preparations (including anes-
trial, comply with the treatment protocol, or adhere thesia and surgical incision) essential to the true
to the follow-up procedures if she is aware that she operation except the surgical procedure itself. The
has not received the experimental intervention. In beneficial effect of the sham operation has been
the absence of blinding, the treatment effect is over- attributed to the placebo effect, with estimates that
estimated, leading to incorrect inferences about the the placebo response in surgery may be of the same
value of the experimental intervention.16 The magni- magnitude (about 35%) as that observed in medical
tude of the overestimation is much larger in infertil- trials.22 The placebo effect in surgery may be
ity trials with pregnancy as the outcome measure.17 defined as the difference between the overall effect
The testing of subjects under conditions of inten- of surgery and that attributable to the procedure
tional ignorance may include the use of dummy itself.23 This realization has prompted researchers to
interventions, such as placebo and sham surgery. reintroduce the sham operation to evaluate surgical
These methods ensure that none of the subjects and, interventions so that the high standard required in
where possible, the trial personnel, is able to recog- efficacy studies can be maintained. The risks to sub-
nize whether the intervention administered is active jects undergoing a sham operation are not trivial,
or inert until the code is broken at the conclusion of and it is important to balance these risks against the
the trial. A placebo is designed to be indistinguish- potential benefits to society at large if the surgical
able in physical properties from the active interven- procedure is proven effective. It is also important
tion. However, when a standard treatment exists, it that future patients be spared from the risks and
should be used as the comparator for the new inter- cost of an ineffective surgical procedure. However, if
vention, and every effort should be taken to make there is no proven alternative therapy available,
the interventions indistinguishable from each other then the sham operation for surgical therapy trials
by the trial participant. To do so often requires the is a desirable and valid approach to evaluate the effi-
use of a double-dummy approach (i.e., two place- cacy of the intervention, provided an appropriate
bos), especially if the routes of administration of the risk assessment has been undertaken.24
two interventions are different – for example oral In management of recurrent miscarriage, there are
versus intravenous, in which case the subject receives very few instances when surgical treatment can be
both oral and intravenous agents, one of which will considered, except for uterine anomalies, as described
be a placebo in the experimental group and vice in Chapter 11. The use of the sham procedure has not
versa for the control group. yet been explored in recurrent miscarriage.
The magnitude of the placebo effect (the
response observed in the placebo group) is difficult CO-INTERVENTION (AVOIDING TREATMENT BIAS)
to quantify, unless the placebo is compared with no
treatment. Estimates of the benefit have ranged The appeal of the randomized trial is the assurance
from none to between 35% and 75% of trial partic- that random allocation of the subjects to experimen-
ipants showing improved outcome.18 The observa- tal or control groups will produce groups that have
tion that the use of ‘tender loving care’ was more similar characteristics at baseline so the efficacy of the
efficacious than when it was not used in women experimental intervention can be tested cleanly and
260
0415421306-Ch18 3/30/07 9:50 AM Page 261
quantified reliably without interference from any size just large enough to test the null hypothesis. The
extraneous factors. In this context, it is important to goal is to increase the signal-to-noise ratio by recog-
ensure that, except for the interventions being com- nizing that statistical ‘noise’ (i.e.,variability) is
pared, the management protocol is held the same for inversely proportional to the square root of the
both groups. Co-intervention occurs when one group sample size (i.e., noise decreases as the sample size
is provided with additional care (e.g., supplementary increases). When the variation within groups gets
treatment, more monitoring, easier access to health larger (the louder the noise) or when the difference in
care personnel, and so on) that is not offered to the outcomes between the groups gets smaller (the fainter
other group. The efficacy of the experimental inter- the signal), the larger is the sample size needed to detect
vention will be biased by co-intervention, leading to the signal.
incorrect estimation of the size of the treatment The size of the sample needed to adequately test
effect. Also, with co-intervention, the research ques- the hypothesis of treatment efficacy can be calculated
tion changes from the original question ‘Is the exper- using a standard formula or with the use of readily
imental intervention more efficacious than the available software programs. The size of the treat-
control interventions?’ to the new question ‘Is experiment effect (the ‘signal’) is the difference in magni-
mental treatment in addition to the co-intervening tude between the outcomes in the experimental and
care more efficacious than the control intervention?’ control groups, and is selected by the investigators
because it has clinical relevance and importance. This
SAMPLE SIZE ESTIMATION (ENSURING THE chosen, clinically important, difference is the smallest
ABILITY TO DETECT A DIFFERENCE IN OUTCOME) difference at which the experimental treatment
would be expected to alter current clinical manage-
In an efficacy trial with comparable groups, any dif- ment. In addition, an indication of the variability
ferences observed in the primary outcome event are (the ‘noise’) is obtained from the standard deviation
due either to chance or to the effect of the experi- for outcomes that are continuous variables; for pro-
mental intervention. The possibility of finding a portions, the difference in event rates is all that is
treatment effect of the magnitude observed in such needed. It is also necessary to select appropriate
a trial is expressed by a probability value (p-value) values for the probability of making errors of
that indicates how likely an ineffective treatment hypothesis testing (typically 0.05 for type I errors and
would have been expected to produce the result 0.2 for type II). Finally, it should be established
observed; the lower the p-value, the less likely is the whether the statistical test used to compare the dif-
effect due to chance and the more likely is it due to ference in outcomes is to be based on a one-tailed
the experimental intervention being evaluated. By (difference in outcomes in one direction; i.e., benefit
convention, the threshold of this likelihood is taken with experimental intervention) or two-tailed (dif-
to be a probability value of 0.05, so that when the ference in outcomes in either direction; i.e., benefit
p-value is less than 0.05, the observed data are or harm with the experimental intervention).
inconsistent with the experimental intervention In research into recurrent miscarriage, the out-
being ineffective (i.e., the experimental intervention come events of most clinical relevance are clinical
is more efficacious than the control intervention). pregnancy and live birth, the rates of which are gen-
In clinical trials, it is important to be able to detect erally high. The sample size required to test the effi-
with a high level of confidence a clinically meaningful cacy of most interventions purported to improve
difference between experimental and control interpregnancy rates is often small enough to permit the
ventions. To do so requires conducting a trial with trial to be undertaken. For example, the control event
sufficient numbers of subjects to avoid a chance find- rate (i.e., success rate with placebo or no treatment)
ing (type I error) and to avoid missing the detection after three miscarriages is expected to be 65%, and
of a true difference if one exists (type II error). The any experimental intervention that can improve the
ideal situation is to conduct the trial with a sample outcome to that expected in the normal population
261
0415421306-Ch18 3/30/07 9:50 AM Page 262
(i.e., 85%) would produce an absolute treatment clinical decisions are often made from evidence
effect of 20% – a difference that is clinically important, derived from non-randomized observational stud-
implying that for every five women with recurrent ies, such as cohort, case control, and historical-
miscarriage treated with the experimental interven- control. When randomized trials were compared
tion, one additional successful outcome would be with observational studies to answer the same
obtained compared with the control intervention. clinical question, between-study heterogeneity was
To detect this magnitude of difference in clinical observed more frequently among observational
pregnancy rates would require a sample size of 162 studies.27 Some of this variability was reduced after
(81 in each group) using a two-tailed hypothesis test historical-control studies were excluded from the
with probabilities for types I and II errors set at 0.05 analyses.
and 0.2, respectively. Accruing this number of sub- A historical-control study is one in which the
jects is not difficult in centers specializing in the eval- outcome of an intervention is compared with the
uation and management of recurrent miscarriage, outcome observed prior to the administration of
but may require several years to complete a trial in the experimental intervention. In research into
institutions with an average volume of clinical activ- recurrent miscarriage this design is used fairly
ity. Consequently, in everyday practice, smaller trials frequently to support claims of improved efficacy of
are usually conducted, because they are easier to com- new interventions. As an example, a historical-
plete in a shorter period of time. Unfortunately, control study was performed to determine whether
because they are insufficiently powered to test the null metformin administered to women with polycystic
hypothesis, the results obtained often lead to erro- ovary syndrome to achieve pregnancy and then
neous inferences being drawn unless the results from continued throughout the pregnancy would reduce
these trials can be pooled with meta-analysis to gen- the likelihood of first-trimester miscarriage.28
erate more precise estimates of the treatment effect. Among the ten women evaluated in the study, their
Reviewing the literature on recurrent miscarriage collective history of 22 previous pregnancies with-
demonstrates an urgent need for trials with ade- out metformin included 16 (73%) miscarriages.
quate power to be carried out so that conclusions In contrast, the current pregnancy on metformin
about treatment efficacy can be made more reliably. therapy ended in miscarriage in only one woman
For example, a systematic review of immunotherapy (10%). The authors statistically (and incorrectly)
for recurrent miscarriage included 12 trials with an compared the two rates of miscarriage and con-
average sample size of 53 (range 22–131).25 Only 1 of cluded that there was a significant benefit with met-
the 12 trials had a sufficiently large sample size, with formin use. This approach is invalid from both
131 subjects. Similarly, a systematic review of treat- design (inappropriate controls) and analytical (lack
ment for recurrent miscarriage in women with of independence) perspectives.
antiphospholipid antibody or lupus anticoagulant There is also evidence demonstrating that
included 13 trials with an average sample size of 66 historical-control studies are associated with an over-
(range 16–202).26 This review contained only one estimation of the effect of treatment.27 Thus, historical-
trial with a sufficiently large sample size of 202 sub- control studies should be avoided in research into
jects. Thus, if progress is to be made in research into recurrent miscarriage because treatment estimates
recurrent miscarriage, larger trials are needed to test derived from them are less reliable than those from
the efficacy of new (and existing) interventions. prospectively undertaken controlled studies.29
AVOIDING HISTORICAL CONTROLS (AVOIDING INTENTION-TO-TREAT ANALYSIS (AVOIDING POST-

OVERESTIMATION OF THE EFFECT OF TREATMENT) RANDOMIZATION EXCLUSION OF DATA)
Despite the acknowledgement that the randomized In trials of treatment efficacy, the purpose of ran-
trial is the ideal in evaluating therapeutic efficacy, domization is to avoid bias in the selection of subjects
262
0415421306-Ch18 3/30/07 9:50 AM Page 263
so that comparable experimental and control and minimizes the likelihood of making a type I
groups can be studied to provide a reliable estimate error in hypothesis testing.
of the size of the treatment effect. Once a subject A good strategy to reduce the numbers of post-
has been randomized into the trial, she needs to be randomization withdrawals is to perform the
included in the analyses, even if she never began random allocation as late as possible, preferably just
the treatment or stopped taking the treatment prior to the administration of the intervention.
partway through the trial. After randomization,
any changes in the composition of the groups LACK OF SUPERIORITY, EQUIVALENCE, AND
(e.g., withdrawing from treatment, being excluded NON-INFERIORITY (ENSURING APPROPRIATE
from the analysis for failing to follow protocol, TESTING OF NULL HYPOTHESIS)
crossing over to the alternative intervention group,
and so on) will disturb the balance between them The placebo-controlled trial is the optimal design for
and may affect their comparability. Therefore, it is evaluating the efficacy of new treatments. However,
important to ensure that subjects not only remain once efficacy of a treatment has been established,
in the groups to which they were allocated, but also newer treatments should be compared against these
complete the study. standard active treatments, because the use of
Unfortunately, despite diligent attention to detail placebo for such subsequent evaluations is consid-
and monitoring of trial progress, the ideal goal of ered unethical.
achieving perfect compliance is often not reached. In general, efficacy trials of active treatments are
The usual approach to dealing with such postran- designed to determine whether a new (experimen-
domization withdrawals is to analyze the data from tal) intervention is superior to the standard (con-
only those who completed the assigned treatment trol) intervention. The objective of such superiority
(i.e., per-protocol analysis) and ignore those who trials is to rule out equality of the interventions by
deviated from the protocol. Although this approach rejecting the null hypothesis that there is no differ-
seems sensible, it is not correct from a methodolog- ence between the two treatments. Sometimes, such
ical perspective because the power of the study to trials are undertaken with the expectation that the
detect a clinically meaningful treatment effect is new intervention will fare better, and the objective
reduced. Also, confining the analysis only to those is to demonstrate this fact unequivocally. More
who are compliant with the protocol will produce commonly, though, the new intervention is only
an estimate of the treatment effect that is biased expected to demonstrate similar efficacy to the
because non-compliance is not a random occur- standard intervention so that healthcare providers
rence and may be associated with a poorer (or can offer their patients a choice of treatment
better) outcome. The correct approach is to per- options. The objective in such a trial is to demon-
form the analysis according to the original random strate equivalent efficacy (equivalence) of the two
assignment using an intention-to-treat method interventions.
whereby all subjects allocated to the group at the A common mistake, when a superiority trial fails
time of randomization are analyzed together as to reject the null hypothesis of no difference, is to
having received the intervention originally assigned conclude that the two interventions are equivalent.
to that group. For example, consider a trial with a sample size of
In most clinical trials, because it is expected that 50 women with recurrent miscarriage in which the
there will be some degree of non-compliance, the standard intervention produced a live birth rate of
intention-to-treat analysis will tend to underesti- 65% and the experimental treatment intervention
mate the effect of the experimental intervention. produced a live birth rate of 85%. After performing
Maintaining group similarity and preserving the the c2 test, this difference in pregnancy rates of 20%
balance among prognostic factors in the study is found not to be statistically significant and may
groups produce a cautious method for evaluation lead one to conclude that the two interventions are
263
0415421306-Ch18 3/30/07 9:50 AM Page 264
the same (i.e., equivalent). This is an incorrect An alternative strategy that avoids the need for a
interpretation, because, although a lack of proof of very large sample size is to conduct a ‘non-inferiority’
superiority may be consistent with equivalence, it is trial. The objective of the active-controlled, non-
not proof that equivalence is present. If the same inferiority trial is not to demonstrate superiority, but
result were observed in a sample of 200 women, the rather to establish that the effect of the experimental
observed treatment effect would be statistically sig- intervention, when compared with the control inter-
nificant. The current practice of conducting small vention, is not below some prestated non-inferiority
comparative trials that fail to show superiority of margin. In other words, given that it is impossible
the new intervention should be avoided when the to prove the null hypothesis of no difference, an
objective is to demonstrate equivalence, because the operational definition must be considered that
‘lack of evidence of a difference’ is not synonymous allows the experimental intervention to be inferior
with ‘evidence of a lack of difference’. to the standard (control) intervention by a clini-
The goal in an equivalence trial is to rule out cally tolerable amount (i.e., the experimental inter-
differences of clinical importance in the primary vention is ‘not much inferior’ to the control
outcome between two interventions. To calculate intervention). Clinicians have to decide on the
a sample size for an equivalence trial requires amount of non-inferiority they are willing to
defining a priori a clinically important difference, accept as medically insignificant or tolerable as the
starting with the assumption that there is a zero basis for non-inferiority claims. Such an approach
difference in the outcome event rates between the is often used for newer drugs that may have lower
two interventions. The null hypothesis (in con- side-effects, better tolerance, or lower cost than the
trast to that in a superiority trial) is stated differ- standard intervention. For a non-inferiority evalu-
ently as a minimum difference that is acceptable ation, the null hypothesis states that the control
that would render the two interventions inter- intervention is superior to the experimental inter-
changeable.30 By rejecting this null hypothesis in vention. The alternative hypothesis is that the exper-
favor of the alternative hypothesis that the differ- imental intervention is not inferior to the control
ence in outcomes between the two interventions is intervention.
zero, one can conclude that the interventions are The design of a non-inferiority trial requires spec-
equivalent. ifying the non-inferiority margin, i.e., the extent to
It should be recognized that the outcome event which the outcome with the control intervention can
rate with the experimental intervention might be exceed that with the experimental intervention and
slightly larger or slightly smaller than that with the still render the experimental intervention non-inferior
control intervention. Thus, a range of possible out- to the control intervention. The null hypothesis
come event rates can be generated. By starting with states that the outcome with the control intervention
a zero percent difference in outcome event rates, a is at least as large as or exceeds this margin; if the null
confidence interval around this value is chosen to hypothesis cannot be rejected, then the control inter-
represent the clinically important range within vention is more efficacious than the experimental
which any differences in outcome rates can lie for intervention. Rejection of the null hypothesis is
interventions that are equivalent. If the observed dif- required to establish non-inferiority of the control
ference in outcome event rates lies entirely within intervention. It should be noted that, according
the range selected for clinical importance, when the to the alternative hypothesis, the experimental
trial is completed, equivalence can be claimed. The intervention might perform better than the control
smaller the selected range, the larger the sample size intervention, but not to an extent greater than the
required. Thus, the execution of an equivalence trial inferiority margin that has been established.
is challenging, because it requires a much larger Because the direction of effect being assessed is
sample size than that for a superiority trial. one-sided (i.e., the experimental intervention is not
264
0415421306-Ch18 3/30/07 9:50 AM Page 265
inferior to the control intervention), a one-sided in other studies.33 Sometimes treatment is instituted
hypothesis test is performed. Thus, by not requiring only after fetal cardiac activity has been demon-
a two-sided hypothesis test, the sample size required strated, as seen with the use of heparin in women
will be much lower. However, if the experimental with elevated antiphospholipid levels.26 In this situ-
treatment performs better than the control inter- ation, the likelihood of a successful outcome with-
vention (and the outcome event rate exceeds the out treatment once fetal cardiac activity has been
non-inferiority margin), one cannot conclude that demonstrated is relatively high, and will result in
it is superior, because the trial design was not set up efficacy studies failing to accurately quantify the
to test this hypothesis. magnitude of the treatment effect with the experi-
Before a non-inferiority trial can be undertaken, mental intervention.
it is necessary to confirm that the control interven- The optimal time of onset of therapy will vary
tion has been shown to be better than placebo. depending on the cause of the recurrent miscar-
Furthermore, it is important to select a non- riages, but it makes sense to commence treatment
inferiority margin that is small enough to not before conception for most causes that can be iden-
exceed that which has clinical relevance; choosing a tified using the current diagnostic protocol. Clearly,
large inferiority margin will risk the generalization consensus on this issue is urgently required so that
of the study findings, because it can be argued that treatment benefit can be maximized.
such a large difference is clinically meaningful, sug-
gesting that the control intervention is superior to CENSORING SUBJECTS WHO FAIL TO
the experimental intervention. Finally, it has to be CONCEIVE (AVOIDING TREATMENT BIAS)
assumed that the control intervention is superior
to placebo (had such a comparison taken place). To Another methodological concern in efficacy trials is
do so requires the inferiority margin to be smaller the practice of restricting the number of cycles of
than the smallest effect size that the experimental preconceptional treatment patients may undergo. If
intervention would be expected to produce had it these women do not conceive, they are withdrawn
been compared with placebo. In this context, for from the study and are replaced by other women,
the experimental intervention to be designated as who take their place in the trial. This strategy vio-
being ‘at least as good as’ the control intervention, lates the principle of randomization and introduces
it has to retain at least 50% of the superiority of the treatment bias by enrolling women with high
control intervention over placebo.31,32 fecundity rates. It becomes difficult to generalize
the results of such trials to the population of
ONSET OF TREATMENT (MAXIMIZING women with recurrent miscarriage.
MAGNITUDE OF TREATMENT EFFECT) Women with recurrent miscarriage have a longer
interpregnancy conception interval compared with
A major problem in the management of recurrent those with sporadic miscarriage (i.e., the length of
miscarriage is the assumption that diagnostic tests time taken for conception to occur after a previous
carried out in the non-pregnant state can identify miscarriage increases with the number of miscar-
potential causes of the miscarriages that have riages).34,35 The mechanism underlying this observa-
already occurred. Treatment is then offered to pre- tion is not clear. One possible hypothesis is that fear
vent miscarriage in a subsequent pregnancy. of miscarriage in a subsequent pregnancy induces
Unfortunately, there is no standardization in many significant stress that may adversely influence the
of the treatment protocols regarding the onset of hypothalamus and result in subtle ovulatory dys-
treatment. For example, intravenous immunoglobu- function.2 Thus, it is clear that for treatments com-
lin has been administered before conception in some menced before conception, a sufficient length of time
studies, and only after confirmation of the pregnancy will be required before pregnancy can be achieved.
265
0415421306-Ch18 3/30/07 9:50 AM Page 266
For this reason and for the methodological reasons study the products of conception in every case of
discussed, women enrolled into randomized trials of miscarriage in therapeutic trials so that a more valid
treatment for recurrent miscarriage should not be assessment of the efficacy of the experimental treat-
withdrawn just because pregnancy has not occurred ment can be performed.
in a short period of time.
KARYOTYPIC ANALYSIS OF THE ABORTUS SUMMARY

(IMPROVING ACCURACY OF TREATMENT
EFFECT ESTIMATION) Therapeutic decision-making relies on the avail-
ability of good-quality evidence generated from
The risk of aneuploidy arising de novo is present in all studies of high quality and high internal validity
pregnancies. Consequently, it is possible that women and without bias. In the area of efficacy evaluation
receiving experimental or control interventions may of therapy for recurrent miscarriage, the random-
experience another miscarriage due to a chromoso- ized trial is the gold standard and should be
mal aberration, unrelated to the intervention itself designed and executed with attention to the
(see the debates following Chapter 3). Hence, in effi- methodological details outlined in this chapter. The
cacy trials, it is prudent to submit all products of con- starting point requires defining the research objec-
ception for karyotypic analysis to exclude the tive that is articulated unambiguously as a research
presence of aneuploidy. Without this information, it question describing the patient population, experi-
is impossible to ascertain whether the miscarriage is mental and control interventions, and outcome of
the result of a failure of the administered intervention interest. The population should be clearly defined
or is due to a de novo chromosomal anomaly. The and women with repeated implantation failure
magnitude of the treatment effect will be less accurate should be excluded. The subjects selected for the
without adjusting for the miscarriages that were trial should be randomly allocated to the experi-
inevitable because of aneuploidy. However, such mental or control interventions to ensure similarity
adjustments are only reliable if abortus material from in the composition of the group. The allocation
all subjects who participated in the trial and had a sequence should be concealed from all investigators
miscarriage are successfully analyzed. and healthcare personnel involved with the trial, so
Improvement in ultrasonographic technology has that bias in selecting participants can be avoided.
resulted in images of early pregnancy having higher Whenever possible, attempts should be made to
resolution to permit earlier diagnosis of pregnancy blind subjects, investigators, healthcare personnel
failure – a process that is assisted with serial measure- providing care, and outcomes assessors so that
ment of serum levels of progesterone and human ascertainment bias can be avoided. Further bias of
chorionic gonadotropin. Thus, it is possible to collect treatment assessment can be avoided by ensuring
fetal and trophoblast tissue electively and prior to the that co-intervention does not occur and historical
commencement of the miscarriage so that it is in a controls are avoided.
sterile condition to enable karyotypic analyses to be Testing the null hypothesis of no difference in
performed without contamination of the cell culture – outcomes between the experimental and control
a problem that is frequently encountered when tissue interventions requires enrolling sufficiently large
is collected in a non-sterile manner after the miscar- numbers of participants so that a clinically impor-
riage process had commenced. Furthermore, tant treatment effect can be detected. Smaller
improved techniques in cytogenetics have permitted sample sizes will usually lead to erroneous infer-
more accurate and reliable assessments of the prod- ences about treatment efficacy. Wherever possible,
ucts of conception (see Chapter 3). stratification should be performed for the number
Given these improvements in diagnostic tech- of previous miscarriages. Also stratification, or sub-
niques, it is essential that every effort be made to grouping, by female age should be considered.
266
0415421306-Ch18 3/30/07 9:50 AM Page 267
Treatment should be commenced prior to preg- 7. Stern C, Chamley L. Antiphospholipid antibodies and coagulation
defects in women with implantation failure after IVF and recurrent
nancy and should be continued for a sufficient miscarriage. Reprod Biomed Online 2006; 13:29–37.
length of time to permit pregnancy to occur; post- 8. Mardesic T, Ulcova-Gallova Z, Huttelova R, et al, The influence of dif-
randomization withdrawals of subjects who fail to ferent types of antibodies on in vitro fertilization results. Am J Reprod
Immunol 2000; 43:1–5.
conceive after a short period of time should be 9. Daya S. Immunotherapy for unexplained recurrent spontaneous
avoided. abortion. Infertil Reprod Med Clinc North Am 1997; 8:65–77.
All analyses should be undertaken using an 10. Carp HJA. Investigation and treatment for recurrent pregnancy loss.
In: Rainsbury P, Vinniker D, eds. A Practical Guide to Reproductive
intention-to-treat approach so that postrandomiza- Medicine. Carnforth, UK. Parthenon, 1997:337–62.
tion exclusion of data does not occur. The approach 11. Brigham SA, Conlon C, Farquharson RG. A longitudinal study of
of the study from the perspectives of superiority, pregnancy outcome following idiopathic recurring miscarriage. Hum
Reprod 1999; 14:2868–71.
equivalence, or non-inferiority should be estab- 12. Carp HJ, Toder V, Torchinsky A, et al. Allogenic leukocyte immuniza-
lished before the trial begins, so that the required tion after five or more miscarriages. Recurrent Miscarriage
numbers of participants can be accrued and the Immunotherapy Trialists Group. Hum Reprod 1997; 12:250–5.
13. Carp HJA, Toder V, Orgad S, et al. Karyotype of the abortus in recur-
hypothesis testing can be directed appropriately. rent miscarriage. Fertil Steril 2001; 5:678–82.
The products of conception should be submitted 14. Sullivan AE, Silver RM, LaCoursiere DY, et al. Recurrent fetal aneu-
for karyotypic analysis for all intervention failures, ploidy and recurrent miscarriage. Obstet Gynecol 2004; 104:784–8.
15. Chalmers TC, Celano P, Sacks HS, et al. Bias in treatment assignment
so that more accurate estimates of the effect of in controlled clinical trials. N Engl J Med 1983; 309:1359–61.
treatment can be generated. 16. Schultz KF, Chalmers I, Hayes RJ, et al. Empirical evidence of bias:
Finally, the reporting of the results should follow dimensions of methodological quality associated with estimates of
treatment effects in controlled trials. JAMA 1995; 273:408–12.
the guidelines established by the Consolidated 17. Khan KS, Daya S, Collins JA, et al. Empirical evidence of bias in infer-
Standards of Reporting Trials (CONSORT).36 This tility research: overestimation of treatment effect in crossover trials
statement consists of a checklist and flow diagram using pregnancy as the outcome measure. Fertil Steril 1996; 65:939–45.
18. Daya S. The placebo effect. Evidence-Based Obstet Gynecol 2000; 2:1.
for reporting the results from randomized con- 19. Stray Pedersen B, Stray Pedersen S. Etiological factors and subsequent
trolled trials and prompts investigators to ensure obstetric performance in 195 couples with a prior history of habitual
that important elements in clinical trial design have abortion. Am J Obstet Gynecol 1983; 148:140–6.
20. Clifford K, Rai R, Regan L. Future pregnancy outcome in unexplained
been addressed. The issues raised in this chapter recurrent first trimester miscarriage. Hum Reprod 1997; 12:387–9.
are important for creating the basis for using an 21. Daya S. Issues in surgical therapy evaluation: the sham operation.
evidence-based approach to inform therapeutic Evidence-Based Obstet Gynecol 2000; 2:31–2.
22. Beecher HK. Surgery as placebo. A quantitative study of bias. JAMA
decisions in the clinical management of recurrent 1961; 176:1102–7.
miscarriage. 23. Johnson AG. Surgery as a placebo. Lancet 1994; 344:1140–2.
24. American Medical Association. Report of the AMA House of Delegates,
2000 Annual Meeting, Recommendation No. 5. www.ama-assn.org
REFERENCES 25. Porter TF, LaCoursiere Y, Scott JR. Immunotherapy for recurrent mis-
carriage. Cochrane Database Syst Rev 2006; (2):CD000112.
1. Jauniaux E, Farquharson RG, Christiansen OB, et al. Evidence-based 26. Empson M, Lassere M, Craig J, et al. Prevention of recurrent miscar-
guidelines for the investigation and medical treatment of recurrent riage for women with antiphospholipid antibody or lupus anticoagu-
miscarriage. Hum Reprod 2006; 21:2216–22. lant. Cochrane Database Syst Rev 2005; (2):CD002859.
2. Daya S. Habitual abortion. In: Copeland LJ, Jarrell JF, eds. Textbook of 27. Ioannidis JPA, Haidich A-B, Pappa M, et al. Comparison of evidence
Gynecology 2nd edn. Philadelphia: WB Saunders, 2000:227–71. of treatment effects in randomized and nonrandomized studies.
3. WHO recommended definitions, terminology and format for statisti- JAMA 2001; 286:821–30.
cal tables related to the perinatal period. Acta Obstet Gynecol Scand 28. Glueck CJ, Phillips H, Cameron D, et al. Continuing metformin
1977; 56:247–53. throughout pregnancy in women with polycystic ovary syndrome
4. Tan BK, Vandekerckhove P, Kennedy R, et al. Investigation and current appears to safely reduce first-trimester spontaneous abortion: a pilot
management of recurrent IVF treatment failure in the UK. Br J Obstet study. Fertil Steril 2001; 75:46–52.
Gynaecol 2005; 112:773–80. 29. Daya S. Evaluation of treatment efficacy – randomization or observa-
5. Nardo LG, Li TC, Edwards RG. Introduction: human embryo implantation? Evidence-Based Obstet Gynecol 2001; 3:111–13.
tion failure and recurrent miscarriage: basic science and clinical practice. 30. Daya S. Issues in assessing therapeutic equivalence. Evidence-Based
Reprod Biomed Online 2006; 13:11–12. Obstet Gynecol 2001; 3:167–8.
6. Laird S, Tuckerman EM, Li TC. Cytokine expression in the 31. D’Agostino Sr RB, Massaro JM, Sullivan JM. Non-inferiority trials:
endometrium of women with implantation failure and recurrent mis- design concepts and issues – the encounters of academic consultants
carriage. Reprod Biomed Online 2006; 13:13–23. in statistics. Statist Med 2003; 22:169–86.
267
0415421306-Ch18 3/30/07 9:50 AM Page 268
32. Jones B, Jarvis P, Lewis JA, et al. Trials to assess equivalence: the 35. Fitzsimmons J, Jackson D, Wapner R, et al. Subsequent reproductive
importance of rigorous methods. BMJ 1996: 313:36–9. outcome in couples with repeated pregnancy loss. Am J Med Genet
33. Daya S, Gunby J, Porter E, et al. Critical analysis of intravenous 1983; 16:583–7.
immunoglobulin therapy for recurrent miscarriage. Hum Reprod 36. Moher D, Schulz KF, Altman D. The CONSORT statement: revised
Update 1999; 5:475–82. recommendations for improving the quality of reports of parallel-
34. Strobino BR, Kline J, Shrout P, et al. Recurrent spontaneous abortion: group randomized trials. Lancet 2001; 357:1191–4.
definition of a syndrome. In: Porter IH, Hook EB, eds Human
Embryonic and Fetal Death. New York: Academic Press, 1980: 315.
268
0415421306-Ch19 3/30/07 9:51 AM Page 269
19. Investigation protocol for

recurrent pregnancy loss
Howard JA Carp
(LMWH) and aspirin may be the standard treat-

INTRODUCTION
ment, but a different approach is indicated in the
patient who continues losing pregnancies despite
The patient with recurrent pregnancy loss (RPL)
treatment. In this chapter, some of the standard
usually seeks a diagnosis of the cause, a prognosis
protocols will be discussed, and some other
for further pregnancies, and treatment if available.
approaches discussed that may be appropriate in
The purpose of an investigation protocol is to assist
particular patients.
physicians as to which investigations are worthwhile
in order to reach a diagnosis. Various protocols
have been published by leading professional organ-
izations such as the Royal College of Obstetricians,1 INCLUSION CRITERIA
the American College of Obstetricians and
Gynecologists,2 the European Society of Human The standard protocols listed above differ with
Reproduction and Embryology (ESHRE),3 and regard to who should be investigated, and with
numerous others. However, virtually all protocols regard to the criteria for investigation. The ACOG
tend to classify RPL as one homogeneous condition, protocol2 recommends investigation after two
and try to suggest a group of investigations or or more pregnancy losses, whereas the RCOG1 and
treatment either based on an evidence-based ESHRE3 protocols only recommend assessment
approach or the experience of the particular after three or more losses. However, no protocol
authors. However, treating RPL as one homoge- defines pregnancy loss. A problem arises with pre-
neous condition takes no account of individual clinical or biochemical pregnancy losses. In these
circumstances in different patients. The prognosis is cases, no pregnancy sac can be visualized on
different in different patients. We classify patients ultrasound. No investigation protocol says whether
into those with a good prognosis and those with these ‘biochemical pregnancies’ should be consid-
a poor prognosis. We have tended to use an ered pregnancy losses. A positive human chorionic
approach that differentiates between patients with gonadotropin (hCG) level may be due to ‘phantom’
a good or poor prognosis: primary versus second- hCG,7 an intrauterine pregnancy, or an extrauterine
ary aborters,4 those with late versus early pregnancy pregnancy. This problem has become especially
losses (as late losses have a worse prognosis5), and common since the wide use of in vitro fertiliza-
recently those losing karyotypically abnormal tion (IVF), where hCG testing is often performed
versus those losing karyotypically normal embryos 12 days after exogenous hCG administration.
(as euploid abortions are associated with a worse Although hCG should be cleared from the circula-
prognosis than aneuploid abortions6). Additionally, tion by 12 days, some may still be present in
treatment is often controversial, as demonstrated by certain patients, leading to a false-positive result.
the various debates in this book. We are of the We have previously defined a biochemical preg-
opinion that there may not be one approach to nancy as a β-hCG level between 10 and 1000 IU/l
treatment. For example, in antiphospholipid syn- in a cycle in which no hCG was administered, no
drome (APS), low-molecular-weight heparins pregnancy sac was demonstrated on ultrasound,
269
0415421306-Ch19 3/30/07 9:51 AM Page 270
and menstruation was delayed by no more than 1 Table 19.1 Levels of evidence
week.8 This definition has since been accepted by
ESHRE.9 However, we have tended to become more Ia. Evidence obtained from meta-analysis of randomized
restrictive, and only accept a biochemical pregnancy controlled trials
Ib. Evidence obtained from at least one randomized
as such if there are two readings that show a rising controlled trial
level. These pregnancies may be better defined as IIa. Evidence obtained from at least one well-designed
pregnancies of unknown location (PUL). If they controlled study without randomization
IIb. Evidence obtained from at least one other type of
recur three times, the author does consider these well-designed quasi-experimental study
events as early pregnancy losses. III. Evidence obtained from well-designed non-experimental
Similar confusion surrounds the upper level of descriptive studies, such as comparative studies,
correlation studies, and case studies
pregnancy loss. Traditionally, any pregnancy that IV. Evidence obtained from expert committee reports or
has been lost prior to viability was considered as opinions and/or clinical experience of respected authorities
an abortion. The more recent North American
definition includes pregnancy losses up to 20 weeks
as miscarriages. However, there are many excep- parental karyotyping, fetal karyotyping, ultra-
tions to this rule. Preston et al,10 in a leading paper sound, or hydrosonography for uterine anomalies,
on hereditary thrombophilias, assessed ‘miscar- APS testing and interpretation according to the
riages’ as up to 27 weeks. Ober et al, in the paper ‘Sapporo’ criteria,13 and treatment with heparin and
most often quoted to show that paternal leuko- aspirin. The protocol claims that there is insuffi-
cyte immunization is ineffective,11 included non- cient evidence to assess progesterone and hCG
consecutive abortions and pregnancies up to supplementation, bacterial vaginosis, factor V
29 weeks (C Ober, personal communication). Laskin Leiden, or the other hereditary thrombophilias.
et al,12 in a leading article usually quoted to show Assessment of thyroid function, the glucose chal-
that steroids have no place in APS, included patients lenge test, antithyroid antibodies, alloimmune testing
with pregnancy losses up to 31 weeks. It is difficult and immunotherapy, and assessment of TORCH
to believe that research on patients with two and other infective agents are not recommended.
losses at 27, 29, or 31 weeks has relevance to patients The RCOG protocol1 is the generally accepted norm
with five or more losses of blighted ova. I tend to in the UK. The guideline states that a significant
agree with the conclusions laid out by Farquharson proportion of cases of recurrent miscarriage remain
et al,9 that RPL needs to be much better defined unexplained, despite detailed investigation, and that
before any relevant investigation or treatment the prognosis for a successful future pregnancy with
protocols can be determined. supportive care alone is in the region of 75%.
However, the guideline takes no account of specific
types of pregnancy loss, and does not distinguish
STANDARD PROTOCOLS between different types of patient. There are no
suggestions regarding patients who subsequently
The RCOG protocol1 was originlly published in miscarry despite the reassurance of a 75% prognosis
1997, and updated in 2003. The protocol attempts for a live birth. The fact that the guideline states
to be evidence-based as far as possible. Evidence is ‘The use of empirical treatment in women with unex-
classified as in Table 19.1. The recommendations plained recurrent miscarriage is unnecessary and
are made for and against various causes of miscar- should be resisted’ has denied many British patients
riage, and methods of treatment are graded accord- with large numbers of miscarriages treatment that
ing to the level of evidence available. Areas lacking may be effective in certain subgroups of patient.
evidence are called ‘good practice points’. The evi- The ACOG guideline2 is much less dogmatic
dence is mainly taken from the Cochrane Register than the RCOG guideline. Two pregnancy losses are
of Controlled Trials. The guideline recommends recognized as warranting investigation. The ACOG
270
0415421306-Ch19 3/30/07 9:51 AM Page 271
INVESTIGATION PROTOCOL FOR RECURRENT PREGNANCY LOSS
guideline does not base its recommendations on a does recommend testing blood sugar levels and thy-
strictly evidence-based approach, and does not state roid function tests, antiphospholipid antibodies
that new and controversial etiologies should not be (aPL: lupus anticoagulant (LA) and anticardiolipin
investigated or treated, but that they should be dis- antibodies (aCL)), parental karyotyping, and
cussed between practitioner and patient. Although assessment of the uterine cavity by pelvic ultra-
the guideline does not take account of different sound or hysterosalpingography. Hysteroscopy
types of patient, or different prognoses, it does state and laparoscopy are reserved as ‘advanced investiga-
clearly that it should not be construed as dictating tions’, but the protocol does not make clear which
an exclusive course of treatment or procedure. The patients warrant such ‘advanced investigations’.
guideline also states that variations in practice may There is a new category of investigations – known
be warranted based on the needs of the individual as ‘investigations that should be used in the
patient, resources, and limitations unique to the framework of a clinical trial’. These include fetal
institution or type of practice. Like the RCOG karyotyping, testing of natural killer (NK) cells,
guideline,1 the ACOG guideline2 recommends luteal phase endometrial biopsy, and homocysteine
parental karyotyping, and suggests that the couple levels. Treatment is classified separately from inves-
should be offered prenatal diagnosis if one parent tigation in this protocol. Both ‘tender loving care’
has a chromosomal aberration. The guideline and health advice such as diet and abstention or
abstains from giving an opinion on karyotyping of reduction of coffee intake, smoking, and alcohol
the abortus, and reserves judgment on assessment are described as established treatments. However,
of the uterine cavity. The guideline claims that no evidence, results, or references are quoted to
assessment of the uterine cavity is based on concen- justify calling these treatment modalities estab-
sus alone, without good evidence. As in the RCOG lished treatment. The following are said to require
guideline,1 there is said to be insufficient evidence more randomized controlled trials before definite
to assess luteal phase defect or progesterone or hCG recommendations can be made: aspirin and
supplements. The ACOG2 does not recommend LMWH or unfractionated heparin for APS, antico-
assessment of antithyroid antibodies or of infec- agulants for inherited thrombophilia, progesterone
tions such as chlamydia, mycoplasma, and bacterial supplementation, IVIG, folic acid in women with
vaginosis. Alloimmune testing, paternal leukocyte hyperhomocysteinemia, and immunization with
immunization and intravenous immunoglobulin third-party donor leukocytes. However, immuniza-
(IVIG) are also not recommended. tion with paternal leukocytes is said to be of no
The ESHRE guideline,3 like the RCOG guide- proven benefit, nor is multivitamin supplementa-
line,1 restricts the definition of recurrent miscar- tion. Steroids are said to be associated with more
riage to three or more consecutive miscarriages. harm than benefit during the first half of pregnancy.
It does take account of different types of patient, Again, no evidence or references are provided.
as the introduction states ‘The number of previous Table 19.2 contrasts the recommendations for
miscarriages and maternal age are the most impor- various investigations and treatment modalities in
tant covariates and they have to be taken into the three protocols. Reliance on these guidelines will
account when planning therapeutic trials. The ideal leave the physician in a quandary as to which inves-
trial should have stratification for the number of tigations to perform and which treatment to offer.
previous miscarriages and maternal age, with
randomization between control and experimental
treatments within each stratum’. The protocol FACTORS AFFECTING SUBSEQUENT
discusses investigations of cause and treatment PROGNOSIS
interventions separately, and, unlike the RCOG1
and ACOG2 guidelines, does not quote the level of The chance of a third pregnancy loss after two mis-
evidence for its recommendations. The protocol carriages is usually quoted to be approximately 20%,
271
0415421306-Ch19 3/30/07 9:51 AM Page 272
Table 19.2 Comparison of three protocols for the investigation and treatment of recurrent pregnancy loss
Investigation or treatment RCOG protocol ACOG protocol ESHRE protocol
Parental karyotyping Recommended Recommended Recommended

APS assessment (aCL and LA) Recommended Recommended Recommended
Fetal karyotyping Recommended Insufficient evidence Trial required
Uterine cavity Recommended by ultrasound Insufficient evidence Recommended by ultrasound or
assessment or hydrosonography hysterosalpingography
Resection of uterine septum Insufficient evidence Recommended despite lack —
of evidence
APS assessment (aCL and LA) Recommended Recommended Recommended
Treatment of APS with heparin Recommended Recommended Insufficient evidence
and aspirin
Luteal phase investigation — Insufficient evidence Insufficient evidence: trials
required
Progesterone or hCG supplementation Insufficient evidence Insufficient evidence Insufficient evidence: more
RCTs required
Bacterial vaginosis Insufficient evidence Not recommended —
Hereditary thrombophilias Insufficient evidence Insufficient evidence Recommended as advanced
investigation
Anticoagulants for hereditary Insufficient evidence Insufficient evidence Insufficient evidence
thrombophilia
Thyroid function Not recommended Not recommended Recommended
Glucose challenge test Not recommended Not recommended Recommended
TORCH testing Not recommended Not recommended Not recommended
Alloimmune testing Not recommended Not recommended Insufficient evidence
Immunotherapy Not recommended Not recommended Insufficient evidence: RCT required
for IVIG and third-party
leukocytes; PLI no proven effect
‘Tender loving care’ Insufficient evidence — Recommended
Diet, smoking, alcohol reduction — — Recommended
Folic acid for hyperhomocysteinemia — — Insufficient evidence
Vitamin supplementation — — Not recommended
Steroids Not recommended — Not recommended
APS, antiphospholipid syndrome; aCL, anticardiolipin antibodies; LA, lupus anticoagulant; hCG, human chorionic gonadotrophin; RCT, randomized
controlled trial; IVIG, intravenous immunoglobulin; PLI, paternal leukocyte immunization.
and the chance of a fourth miscarriage after three of a live birth6 (Figure 3a.1 in Chapter 3a shows
previous miscarriages as approximately 40%. In prognosis according to fetal karyotype)
certain forms of RPL, the recurrence rate is ● concurrent infertility14,16
unknown – for example, in recurrent biochemical ● maternal age16,17
pregnancies, after IVF, in APS, and in the older ● antipaternal complement-dependent antibodies
woman. However, there are certain factors that help (APCA) – these have also been reported to be
to predict the prognosis: predictive of a successful pregnancy outcome14,18
● NK cells19,20
● number of previous pregnancy losses – as the ● early or late pregnancy losses – patients with late
number of previous losses increases, the chance losses tend to have a worse prognosis.5
of a live birth decreases14,15
● primary, secondary, or tertiary aborter status – The most important predictive factor is the
the secondary aborter has a better prognosis number of previous miscarriages. Figure 1.1 in
than the primary aborter4 Chapter 1 shows the decreasing live birth rate with
● karyotype of previous miscarriage – the patient increasing number of miscarriages. Carp et al4
with an aneuploid abortion has a better chance have previously published figures for their series.
272
0415421306-Ch19 3/30/07 9:51 AM Page 273
2 miscarriages 3 miscarriages ≥5 miscarriages

10% 10% 20% 20% 32.5% 32.5%
80% 60% 35%
Live births Genetic abortions Maternal factor

abortions
Figure 19.1 Number of previous abortions and effect of treatment for maternal factors. Patients with two miscarriages have
an 80% chance of a live birth if untreated. If 50% of subsequent miscarriages are chromosomally abnormal, any treatment aimed
at correcting a maternal cause of miscarriage can only raise the live birth rate from 80% to 90%. A mega-trial is necessary to show
a statistical significance between 80% and 90%. Hence, most treatment regimens used on patients with two miscarriages will be
ineffective. Treating patients with three miscarriages can only raise the live birth rate by 20%. However, if treatment is used on
patients with a poor prognosis, the live birth rate can be raised by 32%, making it relatively easy to show a statistically significant
effect of treatment.
After three miscarriages, there was a 55% live birth Ultrasound scans on a regular basis can reassure
rate in untreated patients with unexplained RPL the patient and their partner that the pregnancy is
(33 of 85 patients). The incidence of live births was progressing normally. The early pregnancy centres
45% after 4 miscarriages (17 of 38 patients), 41% in the UK are invaluable in this approach, especially
after 5 miscarriages (10 of 24 patients), 13% of if they allow the patient access on a ‘walk-in’ basis.
patients after 6 miscarriages (2 of 15 patients), and The patient should be reassured that in the event
23% after 7–12 miscarriages (4 of 17 patients). of another miscarriage, further investigations will
Figure 19.1 shows the effect of assessing treat- be carried out, including karyotyping of the abor-
ment on patients with two miscarriages. If there is a tus, and possibly embryoscopy. It is doubtful
subsequent 80% live birth rate, and 50% of subse- whether ‘good-prognosis’ patients need pharmaco-
quent miscarriages are chromosomally abnormal, logical support on an empirical basis. A question
any treatment aimed at correcting a maternal arises regarding patients who have undergone
cause of miscarriage can only raise the live birth partial investigations. For example, if a patient with
rate from 80% to 90%. In order to show a statistical
significance between 80% and 90%, a mega-trial will
be required. Hence, any trial that includes patients
with two miscarriages will show any treatment to Table 19.3 Relative prognoses according to
be ineffective. Table 19.3 shows a rough scale of clinical features
the prognosis according to the various prognostic
Good Medium Poor
factors, and should give physicians and patients a prognosis prognosis prognosis
rough idea as to the relative prognosis.
No. of miscarriages 2 3 4 5 6 7 8 9
Age 20s 30s 40s
Karyotype of abortus Aberrant Normal Normal
‘GOOD-PROGNOSIS’ PATIENTS Primary, secondary, Secondary Primary or Primary or
or tertiary aborter tertiary tertiary
Early or late losses Early Early Late
These patients include young patients with two or Infertility Normal Infertility
possibly three first-trimester miscarriages. ‘Good- fertility
prognosis’ patients probably require very little APCA Positive Negative Negative
NK cells Normal High
investigation. However, they do require reassurance
of their prognosis, and ‘tender loving care’. APCA, antipaternal complement-dependent antibodies; NK, natural killer.
273
0415421306-Ch19 3/30/07 9:51 AM Page 274
two blighted ova is found to have a septum, it is (although debatable), that these hormones may
questionable whether the septum is the cause, or improve the prognosis by approximately 25%. This
whether it should be resected. A septum has been treatment is empirical, as there is no investigation
described to cause abortions of live fetuses in the in the interval between pregnancies that can diag-
second or third trimesters after a ‘mini labor’.21 nose a hormonal defficiency. A problem may arise
Therefore, should the septum be left in situ, as there when the clinical presentation is at variance with
is no evidence that it is the cause, or should it be the laboratory investigations. For example, should a
resected, as it may cause late abortions and preterm patient with aPL and a chromosomally abnormal
labor? These questions should be discussed with the abortus in a previous pregnancy be treated by
patient and partner. It is important to remember anticoagulants? As with ‘good-prognosis’ patients,
that the patient’s views are as valid as those laid skill and experience may be necessary to interpret
down in official guidelines. In any recurrent miscar- the results.
riage clinic, the majority of patients will have a If there is a presumptive diagnosis, treatment
good prognosis. Their good prognosis should not should be prescribed accordingly. Some examples
influence the management of patients with a poor are given below:
prognosis.
● Opinions are divided as to whether patients with
parental chromosomal aberrations have a worse
‘MEDIUM-PROGNOSIS’ PATIENTS prognosis.24–26 Additionally, they seem to lose
eukaryotypic abortuses.27 Only a few abortuses
This group of patients will include women with inherit the aberration in an unbalanced form
three and possibly four miscarriages. The prognosis (5 of 39 abortuses in the series of Carp et al.27
for a live birth is approximately 60% after three mis- However, if the fetus does inherit the chromo-
carriages (40% after four miscarriages) (Figure 19.1). somal aberration in an unbalanced form, preim-
If these patients are included in a trial, Figure 19.1 plantation genetic diagnosis (PGD) may be
shows that treatment of maternal factors can appropriate treatment.
raise the live birth rate by approximately 25%. ● When fetal karyotypic aberrations are present,
Again, a trial of treatment for maternal factors there is usually a good prognosis. However, there
would need large numbers to achieve the power to are a few patients with repeat aneuploidy. This
show a statistically significant benefit from treat- was found in 19% of patients in the series of
ment. For example, paternal leukocyte immuniza- Carp et al,27 and 10% of patients in the series
tion was shown to have a statistically significant of Sullivan et al.28 PGD is appropriate in cases of
benefit in the RMITG trial of 419 patients,14 but repeat aneuploidy.
not in the trial by Ober et al11 with 200 patients. ● aPL are generally accepted as a cause of
I believe that these patients should be investigated, pregnancy loss. However, no trial has assessed
and the standard protocols assessed above give an β2-glycoprotein I (β2GPI)-dependent antibodies,
indication of the criteria for investigation. In which might be more relevant than assessing
this group of patients, investigation may vary, aCL and LA. With our present lack of knowledge,
depending on the clinical presentation. Various treatment seems to be indicated. However, the
clinical presentations and their likely causes are ‘Sapporo’ criteria of two readings at least 6 weeks
described below. In ‘medium-prognosis’ patients, apart should be observed before a definitive
treatment should be directed at the cause as far as diagnosis.13 In a questionaire29 sent to 16 experts
possible. However, despite extensive investigations, in obstetrics, rheumatology, immunology, and
the cause is often not apparent. In these cases, internal medicine in the USA, UK, France, Spain,
there may be a place for empirical hormone support Netherlands, Italy, Israel, Argentina, and Brazil
with progesterone or hCG, as there is evidence22,23 the general opinion was to treat with LMWH
274
0415421306-Ch19 3/30/07 9:51 AM Page 275
and low-dose aspirin from the moment that other intrauterine pathology such as polyps,
pregnancy was diagnosed. However, until now, fibroids, etc. Three-dimensional (3D) ultrasound
there has been no evidence that aspirin has a is probably the best procedure for distinguish-
therapeutic effect. On the contrary, a meta- ing between a septate and bicornuate uterus.
analysis of three trials of aspirin has failed to find This distinction is essential if hysteroscopic
any therapeutic effect.30 septotomy is considered. However, 3D ultra-
● Hereditary thrombophilias are controversial sound requires specialized equipment and highly
with regard to their role in pregnancy loss. trained staff.
They seem to be associated with late losses rather
than early losses.10 However, the literature is
divided on this issue. We investigate hereditary ‘POOR-PROGNOSIS’ PATIENTS
thrombophilias: protein C and antithrombin
activities are measured by chromogenic assays, These are the patients with five or more consecutive
and free protein S antigen is measured by miscarriages. They have been poorly described in
enzyme-linked immunosorbent assay (ELISA). the literature, and have formed the subjects of few
Patients are diagnosed as having protein C, protein trials. We have previously reported that these
S, or antithrombin deficiency if the value of patients constitute approximately 20% of the patients
the corresponding protein is below two standard in the Recurrent Miscarriage Immunotherapy
deviations (SD) of the mean level. Protein C Trialists Group register, and 30% of the patients in
resistance is assessed by clotting techniques. our service.32 The Sheba Medical Center acts as a
Factor V Leiden, the C677T substitution in the tertiary referral center for patients with RPL, which
methylenetetrahydrofolate reductase (MTHFR) may explain the higher number of patients with a
gene, and the G20210A substitution in the poor prognosis in our service. The feature that dis-
factor II gene are detected by polymerase chain tinguishes these patients is that they have usually
reaction (PCR) amplification. However, these had all the investigations and empirical treatments
tests are costly. Serum fasting homocysteine levels available. Hormone supplements, anticoagulants,
are possibly better indicators than MTHFR. hysteroscopic surgery, and often IVF have been
At present, we treat patients with hereditary tried. Additionally, there may be APS patients who
thrombophilias with anticoagulants, usually have failed treatment, patients who continue mis-
the LWMH enoxaparin. We have found this carrying after surgery for uterine anomalies, and, in
medication to raise the live birth rate by 25% in our service, patients who have been treated with
a comparative cohort study.31 Randomized trials anticoagulants for hereditary thrombophilias
are sorely needed in order to determine if this without success. However, most of these patients
approach is justified. have not had fetal karyotyping performed. After
● There is also a dearth of trials to determine five or more miscarriages, the chance of fetal chro-
the place of uterine malformations. Classically, mosomal aberrations is less than after three miscar-
hysterosalpingography was used to make the riages. Ogasawara et al6 have shown clearly that the
diagnosis of uterine anomalies, but the X-ray is incidence of chromosomal aberrations decreases
uncomfortable for the patient and can only diag- with the number of miscarriages. Our approach in
nose the uterine cavity. Hysterosalpingography these patients is to perform alloimmune testing.
cannot distinguish between a septate and bicor- This will include a cytotoxic crossmatch between
nuate uterus. Recently, hysteroscopy has tended maternal serum and paternal cells to detect APCA.
to replace X-ray. Hysteroscopy is associated with Although the absence of these antibodies may not
much less discomfort, but also cannot distin- be relevant after three miscarriages, after five or
guish between a septate and bicornuate uterus. more miscarriages, APCA are less prevalent than in
However, it is the best procedure for diagnosing the parous population.33 The absence of these
275
0415421306-Ch19 3/30/07 9:51 AM Page 276
antibodies indicates a poorer prognosis,14 and their these patients on the basis of their prognosis,
presence indicates a better prognosis.17 The num- as listed above, and to treat them according to kary-
bers and activity of NK cells can also be helpful. otypic findings, and to use immunotherapy if there
Increased numbers of NK cells have been found is a poor prognosis.
in the peripheral blood in RPL,34 particularly in
primary aborters,35 and in luteal phase endometrial RECURRENT SECOND-TRIMESTER FETAL DEATH
biopsies in RPL.36 Increased numbers and activity
of NK cells have also been associated with a poorer This group of patients has a poorer prognosis
prognosis in the subsequent pregnancy.37,38 than that after first-trimester losses.5 It is therefore
In ‘poor-prognosis’ patients in whom other forms justified to investigate and treat after two losses.
of treatment have failed, we do use immunotherapy The chance of a second-trimester loss being due to
with either paternal leukocyte immunization or chromosomal aberrations is less than in first-
IVIG. Paternal leukocyte immunization has been trimester miscarriages. However, there may be fetal
shown to confer a greater benefit after five or more structural anomalies. Hence, detailed ultrasound
abortions than after three abortions in two meta- may assist the diagnosis. Another possibility for
analyses.15,32 The effect is seen mainly in primary diagnosing fetal structural anomalies is embryofe-
aborters. Other trials and meta-analyses of paternal toscopy. Diabetes should be excluded, as it predis-
leukocyte immunization are not appropriate for poses to fetal anomalies.
judging the effect on ‘poor-prognosis’ patients, as the Thrombotic mechanisms, due either to APS or to
results have been obscured by the ‘good-prognosis’ hereditary thrombophilias, are more likely to cause
and ‘medium-prognosis’ patients. IVIG has also not fetal demise than first-trimester miscarriages.10,46 If
been found to be effective when all patients are either of these is found, in the presence of recurrent
judged as a homogeneous group.39,40 However, when second-trimester fetal deaths, treatment by antico-
‘poor-prognosis’ patients are selected, IVIG has been agulants is warranted. New thrombophilias are con-
found to improve the live birth rate.41–43 Unlike pater- stantly being identified. Microparticles and protein Z
nal leukocyte immunization, the effect has mainly deficiency are two such examples. These throm-
been seen in secondary aborters.44 Immunotherapy bophilias are not usually excluded in any investiga-
is probably more appropriate in patients losing tion protocol. Hence, there may be a place for using
karyotypically normal embryos.45 anticoagulants on an empirical basis in the absence
As with the ‘medium-prognosis’ patients, we of APS or hereditary thrombophilia. However, no
attempt to karyotype the embryo. If immunother- trial has assessed anticoagulants in unexplained
apy fails, and the embryo is karyotypically normal, recurrent second-trimester losses.
surrogacy may offer the only possibility of a live Drakeley et al47 have summarized a database
birth. If, however, the pregnancy is karyotypically analysis of 636 patients attending a UK miscarriage
abnormal, a second pregnancy can be attempted clinic. Second-trimester miscarriages accounted for
with immunotherapy, as immunotherapy cannot 25% of miscarriages in their series; 33% tested pos-
prevent chromosomal aberrations. If, however, itive for aPL, there was a 4% prevalence of uterine
the patient loses two karyotypically abnormal anomaly, 3% could be explained by infections, and
embryos, PGD should be offered. 2% of patients were hypothyroid. In 50% of patients,
no diagnosis was apparent. However, hereditary
thrombophilias were not investigated in this series.
SPECIFIC FORMS OF PREGNANCY LOSS
LOSSES OF LIVE EMBRYOS
The majority of RPL are losses of blighted ova,
in which no fetal heartbeat, or even a fetal shadow, Live embryos may be lost in the first or second
was ever detected on ultrasound. We tend to assess trimesters. The distinguishing feature of these losses
276
0415421306-Ch19 3/30/07 9:51 AM Page 277
is that the uterus starts to contract, and vaginal pregnancy loss, the cause is more likely to be due to
bleeding precedes fetal demise. There may be pla- chance, and the prognosis is good. Inclusion of
cental separation and retroplacental hematoma these patients in a trial of treatment may well
formation. These forms of pregnancy loss are confound the results, and raise the live birth rate
relatively rare, comprising approximately 11% of of a control group of patients. In our opinion,
RPL.48 Losses of live embryos or fetuses are less they probably do not require active treatment.
likely to be due to an embryonic or fetal factor, If included in any research protocol, they should be
and more likely to be due to a uterine or other considered as a separate group of patients.
maternal factor. However, patients with this clinical
presentation have not been investigated as a sepa-
rate group. Hence, there is no evidence to support CASE PRESENTATIONS
any conclusions about this group. In the first
trimester, there is a typical history. Embryonic This section illustrates certain difficult cases in
development is normal. The uterus suddenly starts order to show their different presentations and the
to contract, and abortion can ensue. Abortion may likely causes and methods of management.
be fast, within half an hour,48 or it may take longer.
In abortion of live embryos we recommend testing PATIENT 1
for uterine anomalies and infections. In patients
who are pregnant and present with a hematoma, This patient, age 22, para 0, presented after six
empirical prophylactic antibiotics may have a place miscarriages between 8 and 9 weeks. No fetal heart
in preventing the hematoma from becoming had ever been detected, except in the fourth preg-
infected. In the case of an infection, uterine contrac- nancy, when a fetal heart was said to be present at
tions follow rapidly, with expulsion of the uterine 6 weeks. However, the pregnancy showed no fetal
contents. In the case of second-trimester losses of shadow from 7 weeks onwards until curettage
live fetuses, uterine anomalies, infections, and pos- was performed for a blighted ovum at 9 weeks.
sibly diabetes (which predisposes to infections) The fourth pregnancy was found to have a normal
should be investigated. In the presence of contrac- 46,XX karyotype. The following features had been
tions in the second trimester, tocolytic agents may investigated and found to be normal. Parental
be appropriate. Again, the appropriate trials to karyotypes were 46,XX and 46,XY. LA, aCL, and
determine an optimal course of management have hereditary thrombophilias were normal. Hormone
not been carried out. levels (luteinizing hormone, follicle-stimulating
Unfortunately, most patients do not know the hormone, and prolactin) were normal. Midluteal
character of the miscarriage. They will only know progesterone levels were 18 ng/ml. Thyroid function
this if ultrasound has previously been performed in was normal. There was no diabetes. Hysteroscopy
order to detect a fetal heartbeat. showed a normal cavity. APCA were negative.
NK-cell testing was not performed at that time.
MIXED PATTERN OF PREGNANCY LOSSES The third pregnancy was treated with progesterone
supplements. The fourth and fifth pregnancies
In many cases, each pregnancy loss may have a were treated with enoxaparin and aspirin on an
different clinical presentation. For example, there empirical basis. The sixth pregnancy was untreated.
may be a blighted ovum followed by an abortion of The patient was treated by paternal leukocyte
a live fetus in the second trimester, followed by a immunization between the sixth and seventh preg-
missed abortion. These mixed patterns of pregnancies. Immunizations were boosted until sero-
nancy loss are relatively frequent in patients with conversion ocurred with the development of APCA
three losses, but rare in patients with five or directed towards paternal HLA antigens. This is
more losses. In patients with a mixed pattern of our current regimen.18 The seventh pregnancy
277
0415421306-Ch19 3/30/07 9:51 AM Page 278
was uneventful. No additional medications were of patients,51 and both doses have been found to
administered. The seventh pregnancy terminated in be equally effective, there may be individual
the delivery of a female infant, 3580 g, at 40 weeks. patients in whom the larger doses are required.
The eighth pregnancy is presently 7 weeks, with a
live embryo. PATIENT 4
PATIENT 2 This patient, age 38, was a secondary aborter with

two live births followed by six miscarriages, most of
This patient, age 24, para 0, presented after three which were missed abortions in which a previous
abortions between 12 and 16 weeks. From the fetal heart was lost between 10 and 12 weeks.
history, it was apparent that these were abortions Investigation showed APCA to be positive. There
of live fetuses. Hysteroscopy showed a large and was no APS, thrombophilia, or other cause appar-
thick septum that divided the uterus. The external ent for the miscarriages. The parental karyotypes
contour of the uterus was shown to be normal on were 46,XX and 46,XY with a balanced transloca-
laparoscopy. The septum was resected hysteroscop- tion: t(14;13)(p11;q12). The subsequent pregnancy
ically until the fundus of the uterus. The fourth was a missed abortion at 10 weeks. Again, a previ-
pregnancy terminated as a blighted ovum at 8 weeks. ously detected fetal heartbeat was lost. This preg-
Although a fourth miscarriage may sound like a nancy was found to be 46,XY, −4, tder(4;13), i.e.,
failure of treatment, this was not so, as the blighted monosomy 4. Instead of the second chromosome 4,
ovum was found to be triploid. The fifth pregnancy there was a chromosome with a small section of
terminated as induced labor at 42 weeks, and the chromosome 4 and the translocated sections of
sixth pregnancy in spontaneous labor at 40 weeks. chromosomes 13 and 14. The patient has been
advised to have PGD if she desires another child.
PATIENT 3 PGD will use probes for chromosomes 4, 13, and 14.
Meantime, the patient has decided to complete her
This patient, age 30, para 1, a secondary aborter, family with two children.
presented after four midtrimester losses. The first
pregnancy terminated as an uncomplicated delivery PATIENT 5
of a female infant 4050 g. The second pregnancy
terminated as a fetal death at 17 weeks, and the This patient, age 40, para 0, presented after four
third pregnancy as a fetal death at 19 weeks. pregnancy losses. There had been ruptured
Parental karyotyping was normal. Glucose chal- membranes at 20 weeks, and two intrauterine fetal
lenge tests and thyroid function were normal. deaths at 20 weeks, accompanied by hypertension
Hysteroscopy showed a normal uterine cavity. and gestational diabetes. The fourth pregnancy
Thrombophilia testing showed the patient to be was a missed abortion at 14 weeks. These four
homozygous for the MTHFR mutation.49,50 However, pregnancies were achieved from four cycles of
homocysteine levels were normal. The fourth zygote intrafallopian transfer (ZIFT). There were
and fifth pregnancies were treated with enoxaparin no apparent explanations for the pregnancy losses.
40 mg from detection of the fetal heartbeat. There was no aPL or hereditary thrombophilia.
However, these pregnancies terminated at 18 and Hysteroscopy was normal. The parental karyotypes
16 weeks, respectively, with intrauterine fetal deaths. were 46,XX and 46,XY. The fifth pregnancy was
The sixth pregnancy was treated with enoxaparin achieved by eight cycles of IVF following 22 months
80 mg daily. The pregnancy terminated as a of infertility. The fifth pregnancy was treated
cesarean section at 39 weeks. A live male infant of with aspirin 100 mg. However, the pregnancy
3240 g was delivered. Although the dose of 40 mg was terminated artificially at 22 weeks for severe
has been compared with 80 mg in a large cohort preeclampsia with HELLP syndrome.
278
0415421306-Ch19 3/30/07 9:51 AM Page 279
The patient was advised surrogacy. However, she 4. Carp HJA. Update on recurrent pregnancy loss. In: Ratnam SS, Ng SC,
Arulkumaran S, eds. Contributions to Obstetrics and Gynaecology.
conceived spontaneously while making arrange- Singapore: Oxford University Press, 2000.
ments for surrogacy. She was treated empirically with 5. Goldenberg RL, Mayberry SK, Copper RL, et al. Pregnancy outcome
enoxaparin 40 mg daily and aspirin 100 mg. At following a second-trimester loss. Obstet Gynecol 1993; 81:444–6.
6. Ogasawara M, Aoki K, Okada S, et al. Embryonic karyotype of abor-
12 weeks, nuchal translucency screening was normal. tuses in relation to the number of previous miscarriages. Fertil Steril
A Shirodkar suture was inserted at 13 weeks, due 2000; 73:300–4.
to the previous ruptured membranes at 20 weeks. 7. Storring PL, Gaines-Das RE, Bangham DR. International reference
preparation of human chorionic gonadotrophin for immunoassay:
However, preeclampsia and gestational diabetes potency estimates in various bioassay and protein binding assay
developed at 18 weeks, followed by fetal demise. systems; and international reference preparations of the α and
The patient used a surrogate mother who has deliv- β subunits of human chorionic gonadotrophin for immunoassay.
J Endocrinol 1980; 84:295–310.
ered her a healthy term male infant. 8. Carp HJA, Toder V, Mashiach S, Rabinovici J. The effect of paternal
leukocyte immunization on implantation after recurrent biochemical
pregnancies and repeated failure of embryo transfer. Am J Reprod
Immunol 1994; 31:112–15.
CONCLUSIONS
9. Farquharson RG, Jauniaux E, Exalto N. ESHRE Special Interest Group
for Early Pregnancy (SIGEP). Updated and revised nomenclature
RPL is not one homogeneous condition. Hence, for description of early pregnancy events. Hum Reprod 2005;
20:3008–11.
there is no one protocol that is applicable. The aim 10. Preston FE, Rosendaal FR, Walker ID, et al. Increased fetal loss in
of the standard protocols is entirely laudible – to women with heritable thrombophilia. Lancet 1996; 348:913–16.
advise physicians with little experience of RPL as to 11. Ober C, Karrison T, Odem RR, et al. Mononuclear-cell immunisation
in prevention of recurrent miscarriages: a randomised trial. Lancet
the optimal methods of diagnosis and treatment. 1999; 354:365–9.
Hence, the standard protocols try to guarantee that 12. Laskin CA, Bombardier C, Hannah ME, et al. Prednisone and aspirin
the patient receives effective treatment, and that in women with autoantibodies and unexplained recurrent fetal loss.
N Engl J Med 1997; 337:148–53.
ineffective treatment is not used. However, the 13. Wilson WA, Gharavi AE, Koike T, et al. International consensus
standard protocols listed above might have done statement on preliminary classification for definite antiphospholipid
more harm than good, as they treat RPL as one syndrome. Arthritis Rheum 1999; 42:1309–11.
14. Recurrent Miscarriage Immunotherapy Trialists Group. Worldwide
homogeneous group. Hence, their recommenda- collaborative observational study and metaanalysis on allogenic
tions preclude the treatment of subgroups of leukocyte immunotherapy for recurrent spontaneous abortion.
patients. The development of an optimal investi- Am J Reprod Immunol 1994; 32:55–72.
15. Daya S, Gunby J. The effectiveness of allogeneic leukocyte immu-
gation protocol depends on reaching an accurate nization in unexplained primary recurrent spontaneous abortion.
diagnosis of cause and directing treatment to that Am J Reprod Immunol 1994; 32:294-302.
diagnosis. Fetal karyotyping and embryoscopy hold 16. Cauchi MN, Pepperell R, Kloss M, et al. Predictors of pregnancy
success in repeated miscarriages. Am J Reprod Immunol 1991; 26:72–5.
out the possibility of more accurately diagnosing 17. Cowchock FS, Smith JB, David S, et al. Paternal mononuclear cell
embryonic or fetal causes of pregnancy loss. immunization. Therapy for repeated miscarriage: predictive values
Treatments that have not been shown to be effective for pregnancy success. Am J Reprod Immunol 1990; 22:12–17.
18. Orgad S, Gazit E, Lowenthal R, et al. The prognostic value of antipa-
when tried on a large cohort of patients may be ternal antibodies and leukocyte immunizations on the proportion of
found to be highly effective when only used on a live births in couples with consecutive recurrent miscarriages. Hum
subgroup of patients with an accurate diagnosis. Reprod 1999; 14:2974–9.
19. Aoki K, Kajiura S, Matsumoto Y, et al. Preconceptional natural-
killer-cell activity as a predictor of miscarriage. Lancet 1995; 345:1340–2.
REFERENCES 20. Shakhar K, Ben-Eliyahu S, Rosen E, et al. Primary versus secondary
recurrent miscarriage: differences in number and activity of periph-
1. Royal College of Obstetricians and Gynaecologists, Guideline No. 17. eral NK cells. Fertil Steril 2003; 80:368–75.
The Management of Recurrent Miscarriage. London: RCOG, 2003. 21. Rock JA, Jones HW. The clinical management of the double uterus.
2. American College of Obstetricians and Gynecologists. Management Fertil Steril 1977; 28:798–806.
of recurrent early pregnancy loss, ACOG Practice Bulletin No. 24, 22. Daya S. Efficacy of progesterone support for pregnancy in women
February 2001. Int J Gynecol Obstet 2002; 78:179–90. with recurrent miscarriage: a meta-analysis of controlled trials.
3. Jauniaux E, Farquharson RG, Christiansen OB, et al. Evidence-based Br J Obstet Gynaecol 1989; 96:275–80.
guidelines for the investigation and medical treatment of recurrent 23. Scott JR, Pattison N. Human chorionic gonadotrophin for recurrent
miscarriage. Hum Reprod 2006; 21:2216–22. miscarriage. Cochrane Database Syst Rev 2000; (2):CD000101.
279
0415421306-Ch19 3/30/07 9:51 AM Page 280
24. Carp HJA, Feldman B, Oelsner G, et al. Parental karyotype and subse- 40. Porter TF, LaCoursiere Y, Scott JR. Immunotherapy for recurrent
quent live births in recurrent miscarriage. Fertil Steril 2004; 81:1296–301. miscarriage. Cochrane Database Syst Rev 2006; (2):CD000112.
25. Goddijn M, Joosten JH, Knegt AC, et al. Clinical relevance of diagnos- 41. Carp HJA, Toder V, Gazit E. Further experience with intravenous
ing structural chromosome abnormalities in couples with repeated immunoglobulin in women with recurrent miscarriage and a poor
miscarriage. Hum Reprod 2004; 19:1013–17. prognosis. Am J Reprod Immunol 2001; 46:268–73.
26. Sugiura-Ogasawara M, Ozaki Y, Sato T, et al. Poor prognosis of recur- 42. Yamada H, Kishida T, Kobayashi N, et al. Massive immunoglobulin
rent aborters with either maternal or paternal reciprocal transloca- treatment in women with four or more recurrent spontaneous
tions. Fertil Steril 2004; 81:367–73. primary abortions of unexplained aetiology. Hum Reprod 1998;
27. Carp HJA, Guetta E, Dorf H, et al. Embryonic karyotype in recurrent 13:2620–3.
miscarriage with parental karyotypic aberrations. Fertil Steril 2006; 43. Christiansen OB. Intravenous immunoglobulin in the prevention of
85:446–50. recurrent spontaneous abortion: the European experience. Am J
28. Sullivan AE, Silver RM, LaCoursiere DY, et al. Recurrent fetal aneu- Reprod Immunol 1998; 39:77–81.
ploidy and recurrent miscarriage. Obstet Gynecol 2004; 104:784–8. 44. Christiansen OB, Pedersen B, Rosgaard A, et al. A randomized,
29. Tincani A, Branch DW, Levy RA, et al. Treatment of pregnant patients double-blind, placebo-controlled trial of intravenous immunoglobu-
with antiphospholipid syndrome. Lupus 2003; 12:524–9. lin in the prevention of recurrent miscarriage: evidence for a thera-
30. Empson M, Lassere M, Craig JC, et al. Recurrent pregnancy loss with peutic effect in women with secondary recurrent miscarriage. Hum
antiphospholipid antibody: a systematic review of therapeutic trials. Reprod 2002; 17:809–16.
Obstet Gynecol 2002; 99:135–44. 45. Clark DA, Daya S, Coulam CB, et al. Implications of abnormal human
31. Carp HJA, Dolitzky M, Inbal A. Thromboprophylaxis improves trophoblast karyotype of the evidence-based approach to the under-
the live birth rate in women with consecutive recurrent miscarriages standing, investigation, and treatment of recurrent spontaneous
and hereditary thrombophilia. J Thromb Haemost 2003; 1:433–8. abortion. Am J Reprod Immunol 1996; 35:495–8.
32. Carp HJA, Toder V, Torchinsky A, et al. Allogeneic leukocyte 46. Grandone E, Margaglione M, Colaizzo D, et al. Factor V Leiden is
immunization in women with five or more recurrent abortions. associated with repeated and recurrent unexplained fetal losses.
Hum Reprod 1997; 12:250–5. Thromb Haemost 1997; 77:822–4.
33. Carp HJA, Toder V, Mashiach S. Immunotherapy of habitual 47. Drakeley AJ, Quenby S, Farquharson RG. Mid-trimester loss –
abortion. Am J Reprod Immunol 1993; 28:281–4. appraisal of a screening protocol. Hum Reprod 1998; 13:1975–80.
34. Emmer PM, Nelen WL, Steegers EA, et al. Peripheral natural killer 48. Carp HJA, Toder V, Mashiach S, et al. Recurrent miscarriage: a review
cytotoxicity and CD56+CD16+ cells increase during early pregnancy of current concepts, immune mechanisms, and results of treatment.
in women with a history of recurrent spontaneous abortion. Hum Obstet Gynecol Surv 1990; 45:657–69.
Reprod 2000; 15:1163–9. 49. Arruda VR, Von zuben PM, Chiapurini LC, et al. The mutation
35. Shakhar K, Ben-Eliyahu S, Loewenthal R, et al. Differences in number Ala677-Val in the methylene tetrahydrofolate reductase gene: a risk
and activity of peripheral natural killer cells in primary versus factor for arterial disease and venous thrombosis. Thromb Haemost
secondary recurrent miscarriage. Fertil Steril 2003; 80:368–75. 1997; 77:818–21.
36. Clifford K, Flanagan AM, Regan L. Endometrial CD56+ natural killer 50. Nelen WL, Blom HJ, Thomas CM, et al. Methylenetetrahydrofolate
cells in women with recurrent miscarriage: a histomorphometric reductase polymorphism affects the change in homocysteine and
study. Hum Reprod 1999; 14:2727–30. folate concentrations resulting from low dose folic acid supplementa-
37. Aoki K, Kajiura S, Matsumoto Y, et al. Preconceptional natural-killer- tion in women with unexplained recurrent miscarriages. J Nutr 1998;
cell activity as a predictor of miscarriage. Lancet 1995; 345:1340–2. 128:1336–41.
38. Coulam CB, Beaman KD. Reciprocal alteration in circulating 51. Brenner B, Hoffman R, Carp HJA, et al. The LIVE-ENOX
TJ6+CD19+ and TJ6+CD56+ leukocytes in early pregnancy predicts Investigators. Efficacy and safety of two doses of enoxaparin in women
success or miscarriage. Am J Reprod Immunol 1995; 34:219–24. with thrombophilia and recurrent pregnancy loss: the LIVE–ENOX
39. Daya S, Gunby J, Clark DA. Intravenous immunoglobulin therapy study. J Thromb Haemost 2005; 3:227–9.
for recurrent spontaneous abortion: a meta-analysis. Am J Reprod
Immunol 1998; 39:69–76.
280
0415421306-Ch20 3/30/07 9:52 AM Page 281
20. A patient’s perspective

Mindy Gross
I had six miscarriages in less than three years. In ret- some towels and headed to the car to take me to the
rospect, this seems to be a physical impossibility. hospital.
Although each miscarriage stands starkly alone in I had felt pregnant. I had experienced the morn-
my mind, having six of them in such a short span of ing sickness that seemed to last the whole day. Now
time produced a cumulative effect. Each presented I wish I hadn’t. At least I would have been spared the
me with peculiar challenges and carried its own pain of feeling the symptoms disappear. I was filled
unique message. On each occasion, I was in a differ- to the brim with disappointment. Miscarriage is a
ent hospital either in America or in Israel, with dif- death, though perhaps not acknowledged as such by
ferent doctors and different walls to witness my the world outside of the family who have suffered
misery. It was natural to begin getting that deja vu the loss. At the time of my miscarriages, I could not
feeling, ‘Haven’t I been here before?’ Yet, when I articulate this feeling, but it was very real. I was
looked around, I was forced to admit that – ‘No, I frightened by the fragility of life. I knew that I had
had never been here before’. Looking back, it seems experienced a touch of death, for something had
as if each miscarriage demanded its own identity, its died within me.
own space in my brain. Each refused to be lumped The trauma of miscarriage stems from lost hope
together with the others. that was so briefly, and so very vividly, alive. For
In retrospect, I had undergone six different and all of the frustration of infertility, it is one-dimensional,
distinct losses. One might think that the loss monochromatic – negative, negative, negative, void
becomes easier or at least less painful with each suc- and nothingness. Miscarriage by contrast is multi-
cessive miscarriage. It doesn’t. On the contrary, the chromatic. You have hope and a life inside you, and
physical pain is as fresh and as potent each time. then it is lost – both the child and the hope.
The emotional suffering involved in the loss only Surprisingly, I found that I never became jaded. No
increases. It has been said that humans can accli- matter how many losses, each conception reawak-
mate themselves to the most horrendous of circum- ened in me the belief that this time it was going to
stances; the ‘getting used to it’ impulse seems to be be different.
very strong. This was not the case. I never got used The assumption that children will arrive soon or
to losing a nascent child. easily, or at least whenever the couple desires them, is
Although each time I was more prepared on a a normal assumption, but also a very dangerous one.
practical level, I was never prepared on an emo- My family and friends have children. They don’t have
tional level. Each miscarriage came as a shock, miscarriages. So, although I may have known intel-
though the physical symptoms often repeated lectually that miscarriages are not that uncommon,
themselves. The initial trouble always began sud- my unexpected complications were not part of my
denly; without warning, a stain would appear. everyday consciousness. Infertility and miscarriage
Foolishly, I would think ‘Could I just wish it away – are unfortunate things that happen to ‘other people’.
make believe it didn’t exist?’ As the sharp pains in Since it was too painful to imagine these difficulties
my back and cramping increased, I started to in our own lives, when they did occur, I found that I
ramble, ‘Could I have a nervous breakdown simulta- was at a total loss. As the effects of one miscarriage
neously with a miscarriage?’ The hemorrhaging was seemed to spill over into the next miscarriage, one
merciless and, as was now my custom, I grabbed truth pervaded my thoughts: I was still barren.
281
0415421306-Ch20 3/30/07 9:52 AM Page 282
Barren. What a horrible word. It conjures up Pregnancy Loss1 write: ‘When you lost your first preg-
images of the American southwest, of Arizona and nancy, everyone told you not to worry, it happens to
Georgia O’Keeffe’s paintings. One of O’Keeffe’s a lot of people. Remember, you are young and
favorite motifs is a dry horse’s skull sitting on the healthy and have lots of time to have babies.’ But the
desert’s sand. No signs of life there. Some of the most authors continue, ‘Other women have babies so
wretched terrain on earth, incapable of creating and easily; why not you? Lightening is not supposed to
sustaining life. Barren. What a pitiful way to hear strike twice in the same place, and certainly not
oneself described. The label ‘fruitless’ was so contrary three or four times.’
to the way I had perceived myself my whole life. I was My earlier difficulties with conception often left
always very fruitful, I was a producer. Now I was defi- me with a very frightening thought: ‘What if I never
cient. I didn’t have what it takes to ‘produce’ in the conceive?’ With the miscarriages, I knew my situa-
most valuable of all endeavors – conceiving a child tion was different. Somehow, I believed that in most
and sustaining a pregnancy. cases when a woman conceives again and again, the
Bitter. A word I never had an affinity for. likelihood is that sooner or later a pregnancy should
Bittersweet chocolate is, to me, a very poor substi- sustain itself. I would periodically remind myself
tute for the real thing: gooey milk chocolate. Bitter that there were women I knew who never even had
implies something unappealing and most certainly the good fortune of having a hope to cling to. Yet, a
not the ideal. Over the span of my years of infertility nagging doubt was lodged in my consciousness and
and often agonizing tests, I asked myself, ‘Was I a subtle fear accompanied me wherever I went. My
becoming bitter?’ I resolved with all my heart that fear was based on the lesson that the miscarriages
this was one sentiment that would have no place in had taught me. Human existence is very fragile.
my vocabulary. Bitterness was a particularly astrin- There are no guarantees.
gent emotion. One whose intensity I felt would be My body was out of control. First it refused my
better put to use on other needs and emotions, par- command to become pregnant and then it refused
ticularly at such a sensitive time in my life. Bitterness to hold on to the pregnancy. Since my infertility
twists one’s core personality and corrodes a person’s problems were unexpected, as until now I was a
resources and strengths. I was concerned that healthy female specimen, this bizarre turn of events
becoming bitter would have held me back from full caused me to lose my balance. I was young, athletic,
participation in and enjoyment of the births of my never smoked, and only had an occasional glass of
nieces and nephews as well as countless friends. It wine. Why was my body failing me? I felt that I was
would have constricted my own flow of love and twirling in an almost dizzy fashion. Not only was
giving when that is indeed what I needed to do most. my body out of control, but my life seemed to be
All I wanted was to have a normal life. After spiraling in a direction I could not identify. Once
having been married for seven years and wanting the possibility of childlessness entered my mind, it
children, ‘normal’ by definition would mean a baby never departed. It lurked in the shadows of my
in my arms. Other women I know who have suf- brain, pushing to the fore at the most unexpected
fered pregnancy losses have also remarked that they moments. Just when I was enjoying myself, actively
too just wanted their lives to be ‘normal’. I craved engaged in the world around me, the sinking feeling
the ordinary, the mundane tasks of motherhood, would come rushing back: I may never, ever give
but they continued to elude me. I also felt confused. birth to a child. I may never, ever be a natural biolog-
I hadn’t heard much about miscarriages before, and ical mother. Paralysis seeped in. It was as if my body
even if they were relatively commonplace, they were froze and my mind locked. Everything was now out
not part of my lexicon. I did read in a book about of control. I would try to push the ugly thoughts
women suffering from multiple miscarriages the out and concentrate on my life. I focused on my
following, which I could strongly relate to. career, my family, my community, and my friends.
Friedman and Gradstein in their book, Surviving But the thoughts were still there. The harsh fact of
282
0415421306-Ch20 3/30/07 9:52 AM Page 283
A PATIENT’S PERSPECTIVE
my being a ‘habitual aborter’ lingered on and on. It Hallmark store. Amid the baubles and balloons, the
aggressively invaded my consciousness and colored shelves of colorful cards announcing every sort of
my perception of the world. In fact, the cumulative happy occasion, I was occasionless. I wept and felt
experiences of years of infertility and failed preg- my precious lost souls wept with me. I stared in dis-
nancy had shaken me to my core. My world had belief at the dates and the memories they evoked. I
suddenly become a whirlwind of intense and felt weighted down as I held the small maroon diary
mighty emotions. Hope and despair would rage. I in my hand. The year 1986 was embossed in gold
found that in the midst of my mundane activities, I numbers on the front cover of the book. Could I
was now insecure and frightened. Some of these have imagined what occasion would yet occur
emotions I felt quite powerfully for the first time in during that year?
my life. I was distraught and disillusioned. The A well-meaning friend had clipped an article
world had turned bleary. from the local newspaper detailing an experimental
On a particularly overcast and gloomy day in treatment for multiple miscarriages. The treatment
New York, I entered Brooks Brothers Department was initiated in England and was now being offered
Store. Suddenly, I had an overwhelming need to buy by a pioneering doctor in Israel. This was actually a
a personal diary. The pocket diary I found was rather frequent occurrence; concerned friends
maroon and leather and was a present to myself, a would drop by to relay information about some
consolation prize for bearing circumstances that new and innovative medical technology they had
would crumble many a strong individual. The gold- learned of from the media. Of course, I very much
leafed diary cost more than the budget would nor- appreciated their support, but as time passed and
mally allow, even if it did bear the distinguished no experts had the answers I sought, and all the
Brooks Brothers insignia. I distinctly remember the latest treatments had apparently failed, these sug-
day that I bought it. I had been diagnosed with a gestions only served as a constant reminder of my
‘grapefruit-size’ ovarian cyst, the latest mishap in childlessness. I truly felt loved by all those who
my uncontrollable reproductive system. I was noth- called and cared enough to keep me in their
ing less than frenzied, with my cyst surgery loom- thoughts and prayers. No doubt it was my pain and
ing. With the diary in hand, I hurried to the corner my insecurity about the future that made each sug-
of an enormous Hallmark card store in Manhattan gestion so difficult. I was tired of being disap-
and opened it to the back page. With a burning pointed, sick of hanging my hope of becoming a
need, I wrote down the dates, locations and treat- mother on some new innovative medical treatment.
ments of the miscarriages. I also included the Looking back, I think I was also tired of disappoint-
upcoming appointments, the date of my cyst sur- ing everyone around me. This particular newspaper
gery, and all the various treatments I had undergone article was especially ill-timed; it arrived as the
since the onset of my infertility problems. What if hemorrhaging of my sixth consecutive miscarriage
this information would be important some day? worsened. As I lay on my bed, the only thing I could
How would I remember all the details if I didn’t be sure of was that I didn’t want to look at, much
record it somewhere? My response to all my suffer- less consider, any new ‘treatments’. I certainly could
ing at that time was the emotional equivalent of the not face another doctor. Each new doctor would
diary’s blank pages. I was silenced. No words could need my medical history, and with each retelling I
ever convey the emptiness I felt. But the precise found myself reliving. Six miscarriages, six different
week of each pregnancy loss, each doctor, and each doctors, six different hospitals. Could I tell this sor-
treatment would occupy the gold-leaf pages. Some rowful tale one more time? I crumpled the article
empty lines would be filled in, in the not too distant and placed it in the wastepaper basket next to my
future, with the last and most crucial treatment that bed amongst all of the tear-filled tissues. The
was yet to come. From time to time, I take out the cramping continued and I knew with certainty that
diary to remember the tears that were shed in that I couldn’t and wouldn’t ever have the energy to face
283
0415421306-Ch20 3/30/07 9:52 AM Page 284
another treatment protocol. In fact, the only thing I and powerful emotions, which need to be recog-
wanted was to survive this most recent hardship nized, identified, and dealt with. My own personal
and sit very still, by myself, for a long, long time. experience was that my spirit could be crushed or
As the bleeding intensified, I could fool myself no elated by the result of a blood test, because its
longer. I needed a doctor. I turned to the wastepaper results meant the difference between life and death
basket. The discarded ball of newspaper stared up at in pregnancy, or between a healthy organ and a dis-
me. I felt it was actually challenging me – daring me eased one. Each doctor’s visit became a focal point,
to try once again. Suddenly, that rejected article a touchstone, a painful reality check as to how real-
became the focus of my anger, frustration, and all istic it was to believe that motherhood was still
my hope. My decision to take an experimental treat- within my grasp. Now nearly twenty years later
ment was not an easy one. I was concerned with the from the birth of the first of my three children,
unclear repercussions, but the confidence and sin- writing this chapter still evokes such powerful emo-
cere concern of my doctor helped me move forward. tions that it is as if it is in ‘real time’. Tears flow
Treatments in general are not just the filling of pre- freely. Painful memories are now intertwined with
scriptions, scheduling doctors’ appointments, and joy that soars and knows no bounds. The years of
undergoing procedures. The word ‘treatment’ equals infertility and multiple pregnancy loss will always
the word ‘hope’. It was therefore very difficult for me be an integral part of my most essential self. That
to see any treatment as routine. self came to motherhood with profound blessings
Infertility and pregnancy loss presented me with from G-d and the pioneering and brave efforts of
challenges and choices that were often painful and dedicated doctors and hospital staff, all to whom I
difficult to make. My personal suffering had been am forever grateful.
well hidden behind the guise of a ‘normal life’. As a
result, the life crisis that evolved with my pregnancy REFERENCE
losses was often misunderstood. Pregnancy loss cre-
ates isolation, an overwhelming feeling of sadness. I 1. Friedman R, Gradstein B. Surviving Pregnancy Loss: A Complete
Sourcebook for Women and Their Families. Boston: Little, Brown, 1992.
felt vulnerable and out of control. These are intense
284
0415421306-IDX 4/3/07 12:01 PM Page 285
Index
N.B. Page numbers in italic denote figures and tables.
abnormal liveborns, chromosomal anticoagulant–procoagulant balance 128 APS see antiphospholipid syndrome
anomalies in 23 anticoagulants 272 arachidonic acid 110
inversions 31–2 physiological 130 arcuate uterus 148–9, 148
translocations 30 see also specific agents ascertainment bias 259–60
trisomies 28 anticytokine effect 110–11 aspirin
abnormal ovarian reserve 83 antinuclear antibodies (ANA) 167 for APS 119, 121, 122, 235, 272
abortion antipaternal antibodies 174 in labor/postpartum 123
definition 1 antiphospholipid (aPL) for thrombophilia/pregnancy loss 133, 141
see also miscarriage; spontaneous abortion and hCG secretion 111 autoantibodies 9
acardiac conjoined twins 73 heterogeneity 116–17 autoimmune abortions 165–8
adhesiolysis, hysteroscopic 156 immunoglobulin isotypes 116 aPL-antibody-associated 166–7
afibrinogenemia 129 testing for not aPL-antibody-associated 167–8
age assays 115 workup 165
controlling for, in studies 258 time of testing 117 autosomal monosomy 24
and RPL 3–4 titers 116 autosomal trisomy 24–5, 24
age-related factors 4, 7–8 types of aPL 115–16
PGD for 46–7 antiphospholipid (aPL) antibodies 9, 115–18 β2-glyprotein 1 (β2GP1) 107, 108
and trisomies 25, 26 characterization 166–7 antibodies 108
alloimmune abortions 168–75 formation 166 autoantibodies 108–9
immunopathology 172, 174 during pregnancy 167 bacterial identification 203–4
natural killer (NK) cells 174 and prognosis 274–5 bacterial vaginosis (BV) 195–201, 272
Th1 response-inducing factors 172 in RPL 127 antibiotics 197, 199, 200, 204
immunotherapy, preventative 175 types tested for 117 and prevalence 197
workup 174–5 antiphospholipid-associated abortions 115–18 and previous premature labor 197–8
alloimmune testing 272 antiphospholipid co-factor 107 when BV is incidental finding 198–9
alloimmunization, obstetric complications antiphospholipid syndrome (APS) 9 definition 195
after 236 chromosomal abnormalities 35 miscarriage 199–201
American Academy of Pediatrics on clinical associations 119 early 199–200
withdrawing intensive care 225 etiology 107–9 late 200
American College of Obstetricians and evaluation 272 therapy, effect of 200–1
Gynecologists (ACOG) experimental 108 organisms associated 195
on APS management 123 and Helicobacter pylori 204 premature delivery 196–9
on karyotyping 41 immunization 107 BCL-2 protein 62
on RPL management 269, 270–1, 272 management 119–25 bicornuate uterus 148, 148, 150–1, 151, 161
amniocentesis 55–7 consensus guidelines 119 bicornuate bicollis 150
amnion rupture sequence defects 73–4 glucocorticoids 121 bicornuate unicollis 150
amniotic bands 72, 73–4, 76 heparin 119–21, 122 biochemical pregnancies, definition of 269–70
amoxycillin 202 hydroxychloroquine 121–2 biofilms 204
anencephaly 68, 71 intravenous immunoglobulin 122 bleeding diatheses 127–30
aneuploidy labor/postpartum 123 blinding (masking) 259–60
and developmental defects 75 pregnancy monitoring 123
incidence/prevalence 45 of ‘probable APS’ 122 Canadian Paediatric Society on withdrawing
recurrent see recurrent aneuploidy of refractory APS 122–3 intensive care 225
serum markers 55 mechanisms of reproductive failure cardiolipin antibodies (aCL) 115–16
ultrasound markers 55–6 anticytokine effect 110–11 caspases 62–3
aneuploidy screening (PGD-AS) arachidonic acid/prostacyclin 110 causes of RPL 185–8
diagnostic 50–1 placental cell apoptosis 111 cervical cerclage 149, 156, 160–1, 207, 211–16
observational studies 50 thrombosis 109–10 clinical value 212–14
therapeutic 49–50 outcomes 235–6 complications 214
anger 246 pathophysiology 107–13 emergency 215–16
annexin V 109–10 treatment 235–6 patient selection 215
antibiotics antithyroid antibodies (ATA) 81–2, 167–8 techniques 211–12
and bacterial vaginosis 197, 199, 200, 204 anxiety 243–4 timing of placement 214–16
and mycoplasmas 201 coping activities 248 cervical incompetence/insufficiency 160–1
and premature delivery 196 apolipoprotein H 107 cervical competence continuum model
previous 197–8 apoptosis 62 207–8
and unexplained pregnancy loss 202–3 in fetal tolerance 170 cervical length 210, 213–14
see also specific agents placental cell 111 definition 207
285
0415421306-IDX 4/3/07 12:01 PM Page 286
INDEX
cervical incompetence/insufficiency decidual immune cells 173 endocrine factors in RPL (Continued)
(Continued) decision-making process in extreme elevated follicle-secreting hormone 84
diagnosis 207–11 prematurity 226–7 frequency 79
differential 210–11 denial 246 hyperprolactinemia 80
pathophysiology 207–8 depression 243, 244, 246 inhibins 79, 84–5
Cervical Incompetence Prevention DES see diethylstilbestrol insulin resistance 83
Randomized Cerclage Trial developmental defects polycystic ovary syndrome (PCOS) 83
(CIPRACT) 160, 215 etiology 74–5 thyroid abnormalities 80–2
cervical length and insufficiency see also fetal structural malformations see also luteal phase defect
210, 213–14 developmental gene abnormalities 32–3 endometriosis 150
cesarean section after RPL 232 developmental toxins 62–3 endothelial cells in APS 109–10
Chlamydia trachomatis antibodies susceptibility to 59 enoxaparin 133–4, 139
194–5 diabetes, gestational prophylaxis 140, 141, 144
chlamydial infections 194–5, 202 odds ratio in RPL 233 for thrombophilia 236
chorangiopagus parasiticus (CAPP) 73 after RPL 233 epidemiology 1–13
chorionic villus sampling (CVS) diabetes-induced teratogenicity 60 clinical associations 6, 10
errors in 40 diabetes mellitus 82–3 familial aggregation 5–6, 6, 9–10
and screening 55–7 didelphys uterus 148, 148, 149–50, 149, 150 lifestyle factors 6–7, 10
in transcervical embryoscopy dietary supplements 272 maternal age 3–4, 7–8
69, 69 diethylstilbestrol (DES) 151, 162 number of previous miscarriages 2–3, 3, 7
chromosomal abnormalities 7, 39–40 diethylstilbestrol (DES)-exposed uterus occurrence 1–2, 7
developmental genes 32–3 148, 151, 161 partner specificity 6, 10
frequency 23–4, 35–8 disability rates of ELBW infants 222, relevant factors 1–7
parental 37, 37, 39, 40 222, 223, 224, 224 research/management 7–10
and PGD 46 double trisomy 24, 25–6 subgroups of RPL 4–5, 8–9
prognosis 274 chromosomes 25 equivalence trials 264
in preimplantation embryos 23 Down syndrome 55–7 erythromycin 200, 202
prognosis 40, 274 detection rate 56, 57 estrogen and thrombosis 135
recurrent aneuploidy 27–9 doxycycline 200, 202 ethical dilemmas
and RPL subgroup 8 duplication anomalies 73 extreme prematurity 224–8
spectrum 24–7 dysfibrinogenemia 129 withdrawing intensive care 225–6
structural 29–32, 56 withholding intensive care 224–5
frequency 70 early missed abortions 74–5 European Society of Human Reproduction
and PGD 51–2 early pregnancy factor (EPF) 19 and Embryology (ESHRE)
cleft lip 72, 72 early pregnancy recognition 15 PGD Consortium Data Collection 51
and teratogens 60, 62, 63–4 early RPL 5 on RPL management 269, 271, 272
clindamycin 197, 198, 200, 202 hCG therapy 99 Special Interest Group for Early Pregnancy
clinical associations 6, 10 Edwards syndrome 55 (SIGEP) 156, 255
coagulation factor defects 127–37 ELBW (extreme low birthweight) infants exencephaly 68, 71
cytokine imbalances 134 see prematurity, extreme extreme low birthweight (ELBW) infants see
factor XIII deficiency 127–9 embryo–maternal signaling 15–22 prematurity, extreme
fibrinogen deficiency 129–30 genomic elements 17
hormones 135 hypothesis 15 facial dysplasia 71
microparticles 134–5 onset 16–17 factor II (FII) G20210A 127, 128, 130, 143
thrombophilias 130–4 uniqueness of signal 17 factor V Leiden (FVL) 127, 128, 130, 143
cognitive restructuring 248–9 universal tolerance biomarker 18–20 and thromboprophylaxis 144
comparative genomic hybridization embryonic growth disorganization factor XIII (FXIII) deficiency 127–9
(CGH) 40, 41 70, 74, 77 FXIII replacement 128–9
conjoined twinning 73 embryoscopy, transcervical mechanism of action 128
control, loss of 247 complementing ultrasound 76 familial aggregation 5–6, 6, 9–10
coping mechanisms 243–53 fetal structural malformations 67–78 fertilization process and signaling 16–17
cognitive restructuring 248–9 in spontaneous/missed abortions 67 fetal anomalies
corpus luteum triploid embryo 69 and IVIG 238
deficiency see luteal phase defect uterine abnormalities diagnosed 68 obstetric outcomes after RPL 232–3
rescue 15, 16 technique and paternal leukocyte immunization
CVS see chorionic villus sampling chorionic villus sampling (CVS) 69, 69 237, 237
cyclophosphamide 59–60, 62 embryo evaluation 68–9 see also chromosomal abnormalities; fetal
cytokines 134, 238 endovaginal sonography 68 structural malformations
in alloimmune abortions 169–70 gestational sac localization and entry 68 fetal structural malformations 67–78
anticytokine effect in APS 110–11 hysteroscope insertion 67 common defects 70–4
in embryo tolerance 18, 89, 94 tissue sampling 69 amnion rupture sequence 73–4
imbalances 134, 238 uterine cavity exploration 67–8 duplication anomalies 73
and obstetric complications 238–9 uterine evacuation 69 head 70–2
in teratogen susceptibility 63 encephaloceles 71 limbs 72–3
therapeutic 183 endocrine factors in RPL 79–87 trunk 72
cytomegalovirus 193–4 diabetes mellitus 82–3 umbilical cord 73
286
0415421306-IDX 4/3/07 12:01 PM Page 287
INDEX
fetal structural malformations (Continued) human chorionic gonadotropin (hCG) incidence

cytogenetic findings 75 supplementation 99, 272 aneuploidy 45
and diabetes 83 arguments against 101–5 RPL 1–2
embryonic growth disorganization 70, 74, 77 arguments for 97–100 ‘individualized prognostic strategy’ 225
embryoscopy technique, transcervical 67–70 in early pregnancy 102–4 indoleamine 2,3-deoxygenase (IDO) 170
etiology in early missed abortions 74–5 following ovulation induction 102 infections
karyotype and morphology 74, 75, 76 recurrent miscarriage 102–4 and APS 107–9
morphological/cytogenetic evaluation, oligomenorrhea 102, 102, 103–4 detection 203–4
clinical significance of 76–7 placebo effect 103, 103 future developments 203–4
fetal tolerance see tolerance of embryo trials/meta-analyses 103 and RPL 193–206
fetomaternal immunoreactivity 59–66 threatened miscarriage 102 infertility
and embryonic development 59 human leukocyte antigen (HLA) 9, 10 coping 245
mechanisms in embryo–maternal signaling 15, 19 patient’s perspective 281–2, 284
embryo apoptosis regulation 62–3 HLA-C 171 inhibins 79, 84–5
at fetomaternal interface 63–4 HLA-G 19, 170–1 insulin resistance 83
and teratologic susceptibility 59–62 in immune response initiation 174 intensive care
fibrinogen deficiency 127, 129–30 hydrosalpinx 161 withdrawing 225–6
fibroids 152 hydroxychloroquine 121–2 withholding 224–5
embolization 160 hyperinsulinemia 83 interleukin IL-3 110–11
financial loss 247 hyperprolactinemia 80 International Histocompatability
fluorescence in situ hybridization, hypertension, pregnancy-induced 233–4, 233 Workshops 168
multiplex (M-FISH) 41 hyperthyroidism 80–1 interpregnancy conception interval
follicle-secreting hormone (FSH), elevated 84 hypofibrinogenemia 129 264–5
hypothyroidism 81 interventions see therapy
Gardnerella vaginalis 196, 198 hysterosalpingography (HSG) 152, intrauterine adhesions (synechiae) 152
gastroschisis 72 154–5, 160 intrauterine growth restriction (IUGR)
gavage feeding, withdrawal of 227–8 advantages/disadvantages 155 and APS 235
genetic counseling 28 in cervical incompetence 208 and IVIG 237, 238
genetics hysteroscopy, diagnostic 154 odds ratio in RPL 234
chromosomal abnormalities 7 advantages/disadvantages 155 and paternal leukocyte immunization
factor XIII deficiency 129 237, 237
familial aggregation 5–6 immune response after RPL 234
single gene mutations 143–4 avoidance by fetus 89 intravenous immunoglobulin (IVIG)
spontaneous abortions 23–34 see also embryo–maternal signaling; 122, 123, 182, 276
Th1-triggered abortion 174 fetomaternal immunoreactivity for alloimmune abortion 175
glucocorticoids 121 cytokine/hormone network 169–70 for APS 235–6
glucose challenge test 272 facilitation reaction 168–9 arguments against 191
Graves disease 81 fetal tolerance, mechanisms enhancing 170 arguments for 185–8
grief/grieving process 244, 245–6 natural killer (NK) cells 171 indications 186
stages 246 in normal pregnancy 168–71 mechanisms of action 186
teaching couples 248 recognition of embryo 16–17 mode of action 239
growth factors in teratogen susceptibility and teratogen susceptibility 60 outcome classification 187
63–4 Th2-type response 169 patient selection 185–6
guilt 243, 244, 246 immune system and teratogens 59–66 rationale 189–90
immunization success rates 186–7
hCG see human chorionic gonadotropin alloimmune abortion 175 inversions, chromosomal 31–2
head defects 70–2 antiphospholipid syndrome 107 abnormal liveborns 31–2
healing 246 see also paternal leukocyte immunization subsequent losses 32
heat shock-induced teratogenicity 60 immunobiology of RPL 165–77 investigation protocol for RPL 269–80
Hegar test 208 immunoglobulin, intravenous (IVIG) see case presentations 277–9
Helicobacter pylori 107, 203, 204 intravenous immunoglobulin inclusion criteria 269–70
heparin immunological factors and RPL subgroup 8–9 prognosis
for APS 235, 272 immunological mechanisms 185 good 273–4
for fibrinogen deficiency 130 abortion 173 medium 274–5
mode of action 134, 239 normal pregnancy 173 poor 275–6
trophoblast, effects on 134 immunological paradox of pregnancy 168 subsequent 271–3
see also low-molecular-weight heparin; immunomodulation and progesterone 93–4 specific forms of RPL 276–7
unfractionated heparin for APS immunoreactivity, fetomaternal see loss of live embryos 276–7
hereditary thrombophilias see thrombophilias, fetomaternal immunoreactivity mixed pattern of pregnancy loss 277
hereditary immunostimulation, maternal 59, 60, 61 recurrent second-trimester fetal
heritability 6 simplified model 64 death 276
herpes simplex virus 194 immunotherapy 276 standard protocols 270–1
hormones and thrombosis 135 arguments against 189–92 isolation 246
human chorionic gonadotropin (hCG) efficacy 190–1 IVIG see intravenous immunoglobulin
79, 85, 101 ‘immunotrophic’ theory 169
in pregnancy maintenance 101 imperforate hymen 148 Jones metroplasty 161
287
0415421306-IDX 4/3/07 12:01 PM Page 288
INDEX
karyotype methodological issues in therapy myomectomy (Continued)

and fetal malformation morphology assessment 255–68 laparoscopic 159
75, 76 ascertainment bias 259–60 laparoscopy-assisted (LAM) 159–60
and prognosis 269 baseline risk of miscarriage 257–8
in RPL 27, 28–9 blinding (masking) 259–60 natural killer (NK) cells 9
karyotyping clinical heterogeneity 256–7 in alloimmune abortions 174
fetal 272 co-intervention 260–1 and embryo tolerance 18
arguments against 39–43 control event rate 257–8 inhibition by PIBF 89
arguments for 35–8 difference in outcome, detecting and IVIG 175, 186, 190
consensus 41–2 261–2 in maintenance of pregnancy 171
in early abortion 76 exclusion of implantation failures 256–7 in paternal leukocyte immunization 182
outcome of subsequent pregnancy female age, controlling for 258 neonatal death, definition of 220
35–6, 36 group allocation concealment 259 neural tube defects 71–2
techniques 35, 40 historical controls 262 non-inferiority trials 264
parental 37, 40, 272 intention-to-treat analysis 262–3 nuclear antigens, antibodies to 167
intervention groups, ensuring similarity number of previous miscarriages 2–3, 3, 7
laparoscopy, diagnostic 154 between 258–9 and prognoses 272–3, 273
advantages/disadvantages 155 karyotypic analysis of abortus 266
late RPL 5, 8 male partner, controlling for 258 obstetric outcomes after RPL 231–41
lifestyle factors 6–7, 10, 272 null hypothesis testing 263–5 diabetes mellitus 233
limb defects 72–3 onset of treatment 265 fetal anomalies 232–3
listeriosis 193 population 256 incidence of complications 231–5
liveborns, abnormal, chromosomal anomalies post-randomization exclusions 262–3 intrauterine growth restriction (IUGR) 234
in 23 randomization 258–9 perinatal mortality 234–5
inversions 31–2 recurrent miscarriage, definition of 256 pregnancy-induced hypertension 233–4
translocations 30 research question 258 preterm labor 234
trisomies 28 sample size estimation 261–2 statistical analyses 231
low-molecular-weight heparin selection bias 258, 259 study method 231
for APS 119–21, 122, 235 treatment bias 260–1, 265–6 vaginal bleeding 232
guidelines 120–1 treatment effect 262, 265 oligomenorrhea
in labor/postpartum 123 estimation accuracy 266 hCG therapy 99, 99, 102, 103–4
mechanism of action 120 magnitude 257 clinical trials 102–3, 102
monitoring 121 metronidazole 197, 198, 203 omphalocele 72
risks 121 metroplasty ovarian reserve, abnormal 83
complications 139 abdominal 156, 158, 161
dosage 140 hysteroscopic 156, 157–8, 162 p53 protein 62
for fibrinogen deficiency 130 septum-sparing 159 palliative care
indications 139 Strassman 156–7, 161 extreme prematurity 227–8
mechanism of action 134, 141 Tompkins, modified 158 gavage feeding 227–8
mode of action 239 microcephaly 69, 71 hospice concepts 228
monitoring 141 microparticles 134–5 prolonged terminal weaning 227
in pregnancy 139 midtrimester loss partial moles 26
for RPL 143 cerclage, role of 207–17 partner specificity 6, 10
for thrombophilia/pregnancy loss and viability 219–30 Patau syndrome 55
140–1, 236 miscarriage paternal leukocyte immunization 175, 276
luteal phase defect 79–80, 97 definitions 1, 270 arguments against 190–1
causes 79 patient’s perspective 281, 284 arguments for 179–84
diagnosis 79–80 missed abortions ‘blocking antibody theory’ 183, 189
investigation 272 early 74–5 expected chance of live birth 181
and progesterone supplementation 90, 93 transcervical embryoscopy 67–70 meta-analyses 179–80, 180
treatments 80, 98 monosomy, autosomal 24 mode of action 239
problems with trials 98–9 monosomy X 24, 26–7 obstetric outcomes after 237, 237
luteinizing hormone (LH) 83 frequency 24, 70 rationale 189
mosaic trisomy 24 patient’s perspective 281–4
magnetic resonance imaging (MRI) 154 mosaicism 26, 32 peer relationships 247
advantages/disadvantages 155, 155 and CVS 40 perinatal deaths
male partner, controlling for, in studies 258 and PGD-AS 50 contribution of extreme prematurity
masking (blinding) 259–60 Mullerian duct defects 147–52 219–20
maternal age classification 147–8, 148 in Victoria 220
PGD/PGS for 46–7 development 147–8 and IVIG 237, 238
recurrent aneuploidy 28 multiplex fluorescence in situ odds ratio in RPL 235
RPL 3–4, 4, 7–8 hybridization (M-FISH) 41 and paternal leukocyte immunization
McDonald cerclage technique 211, 215 mycoplasmas 201–2 237, 237
medicolegal perspective on extreme myomas 151, 159–60 after RPL 234–5
prematurity 226 myomectomy peripheral blood NK cell testing 190
metformin 83 hysteroscopic 156, 159 PGD see pregestational diagnosis
288
0415421306-IDX 4/3/07 12:01 PM Page 289
INDEX
phosphatidylethanolamine (aPE) prematurity (Continued) recurrent pregnancy loss (RPL)

antibodies 115, 116 survival rates of ELBW infants 221, 221, associated complications 232
phospholipids 166 222–4, 222, 224 causes 185
PIBF see progesterone-induced blocking factor withdrawing intensive care 225–6 classification 4–5, 8–9
placebo effect 260 withholding intensive care 224–5 definition 1
placebo surgery 260 and mycoplasmas 201 early 67
placenta preterm labor specific forms 276–7
in embryo– maternal signaling 15 and IVIG 238 therapeutic alternatives 49–51
infarcts/thrombosis 130–1 odds ratio in RPL 234 unexplained 49–51
placental apoptosis 134 and paternal leukocyte immunization 237, repeat fetal aneuploidy see recurrent
placental cell apoptosis 111 237 aneuploidy
plasmapheresis 123 after RPL 234 repeated implantation failure (RIF) 257
platelet-activating factor (PAF) 15, 18–19 prevalence reproductive autoimmune failure
polar body analysis 25 aneuploidy 45 syndrome (RAFS) 165
polycystic ovary syndrome (PCOS) 10, 83 RPL 1, 2, 7 reproductive autoimmune syndrome (RAS) 165
polydactyly 72–3 primary RPL 4, 5, 8, 8, 9, 256 associated autoantibodies/immune
polypectomy, hysteroscopic 156, 157 procoagulant–anticoagulant balance 128 disorders 168
polyploidy 24, 26 progesterone clinical/laboratory findings 166
and developmental defects 75 deficiency spectrum 90 resuscitation of ELBW infants 224–5
frequency 70 and immunomodulation 93–4 and gestational age 224
polyps 151–2 in pregnancy maintenance 79 Robertsonian translocations 30
polysomy of sex chromosomes 24, 27 and pregnancy outcome 90–1 and PGD 51
prednisone 121 and thrombosis 135 Royal College of Obstetricians and
preeclampsia progesterone-induced blocking factor Gynecologists (RCOG)
and APS 235 (PIBF) 89 on karyotyping 38, 41, 76
and IVIG 238, 238 progesterone receptors 89 on RPL management 269, 270, 272
and paternal leukocyte immunization progesterone supplementation 80, 97, Royal College of Pediatrics and Child Health
237, 237 98, 272 on resuscitation of ELBW infants 225
pregestational diagnosis (PGD) arguments against 93–6 Royal College of Pathologists (RCPath) on
aneuploidy screening (PGD-AS) clinical data 94–5 karyotyping 41
diagnostic 50–1 scientific basis 93–4 RPL see recurrent pregnancy loss
observational studies 50 arguments for 89–92 rubella 194
therapeutic 49–50 and donor oocytes 89
arguments against 49–53 prognostic factors 272–3 secondary RPL 4–5, 5, 8, 8, 9, 256
arguments for 45–8 prolactin 80, 81 selection bias 258, 259
indications 46 prolonged terminal weaning 227 selective resuscitation see intensive care,
and maternal age 47 prostacyclin 110 withholding
parental chromosomal abnormalities 46 protein C 109 self-esteem 246–7
patient selection 45–6 psychological reactions to RPL 243–5 semiallogenic graft, conceptus as 168
for recurrent aneuploidy 46 affected domains of life 246–7 septate uterus 148, 153, 157–60, 157
pregestational screening (PGS) 37, 41 grieving process 245–6 septoplasty 157–9, 272
chromosomes sought 35, 37, 46 male partner 247–8 sex chromasome polysomy 24, 27
and maternal age 47 physician 248 frequency 24
in recurrent aneuploidy 36, 37 psychoneuroimmunology 250–1 sexual life 247
for recurrent aneuploidy 46 stress 249–50 sham operation 260
pregnancy support for couples 248–9 Shirodkar cerclage technique 211
immune response in 168–71 psychological coping mechanisms 243–53 signaling 15–22
immunological mechanisms 173 cognitive restructuring 248–9 corpus luteum, rescue of 16
immunological paradox of 168 psychoneuroimmunology 250–1 see also embryo–maternal signaling
preimplantation embryos smoking 10
chromosomal abnormalities in 23 2,3-quinoxalinedimethanol-1,4-dioxide 59 social support 244
frequency of losses 23 sonohysterography 153–4
preimplantation factor (PIF) 15 randomization 258–9 advantages/disadvantages 155–6, 155
synthetic 18 reciprocal translocations 30 Special Interest Group for Early
as universal tolerance biomarker 19–20 and PGD 51 Pregnancy (SIGEP) 156, 255
prematurity recurrent aneuploidy 27–9 spina bifida 72
and bacterial vaginosis 196–9, 200 and clinical management 29 split-hand/-foot malformation 73
and DES exposure 151, 162 and karyotyping 36, 37 spontaneous abortion
extreme and maternal age 28 definition 39
decision-making process 226–7 and number of losses 28, 29 early 76–7
disability rates of ELBW infants 222, 222, PGD for 46 genetics 23–34
223, 224, 224 prevalence 27–8 transcervical embryoscopy 67–70
ethical dilemmas 224–8 recurrent miscarriage, definitions of 1, 256 steroids 121, 123, 272
medicolegal perspective 226 Recurrent Miscarriage Immunotherapy stillbirth, definition 219–20
palliative care 227–8 Trialists Group (RMITG) meta-analyses stress 249–50
and perinatal mortality 219–20 179, 181, 182, 232 animal models 250
289
0415421306-IDX 4/3/07 12:01 PM Page 290
INDEX
stress (Continued) torsion of umbilical cord 73 uterine anomalies (Continued)

marital 247 toxoplasmosis 194 three-dimensional ultrasound 152–3, 155
and psychoneuroimmunology 250–1 transforming growth factor β (TGF-β) 63, 64 transvaginal sonography (TVS) 152, 155
subseptate uterus 148 translocations, chromosomal 29–31 incidence 147
support for couples 244–5, 248–9 abnormal liveborns 30 Müllerian duct defects 147–52
support groups 244, 248 frequency 29–30 arcuate uterus 148–9, 148
supportive care 49, 250 and PGD 51, 52 bicornuate uterus 148, 148, 150–1,
survival rates of ELBW infants 221, 221, subsequent losses 30–1 151, 161
222–4, 222, 224 transvaginal sonography (TVS) 152 classification 147–8, 148
syndactyly 73, 73 cervical incompetence 210 development 147–8
syphilis 195 didelphys uterus 149, 150 didelphys uterus 148, 148, 149–50,
septate uterus 157, 158 149, 150
T-shaped uterus 151 see also ultrasound, three-dimensional imperforate hymen 148
‘tender loving care’ 156, 182, 271, 272, 273 transverse limb reduction defect 73 intrauterine adhesions (synechiae) 152
teratogens treatment see therapy myomas 151, 159–60
embryo’s responses 59–66 treatment bias 260–1, 265–6 polyps 151–2
regulating apoptosis in embryo 62–3 treatment effect 262, 265 septate uterus 148, 153, 157–60, 157
tertiary RPL 4, 256 estimation accuracy 266 subseptate uterus 148
tetracycline 202 magnitude 257 T-shaped uterus 151
tetraploidy 24, 26 triploidy 24, 26 unicornuate uterus 148, 148, 149, 149
therapy and spina bifida 72 uterine agenesis/hypoplasia 148, 148
estrogen 98 trisomies and prognosis 275
evidence-based evaluation 255–68 autosomal 24–5, 24 reproductive outcomes 147
human chorionic gonadotropin 98 double 24, 25–6 treatment 155–62
progesterone 98 chromosomes 25 abdominal metroplasty 156, 158, 161
study criteria for RPL treatments 94 and encephaloceles 72 cervical cerclage 149, 156, 160–1
for unexplained RPL 49–51 frequency 24, 70 hysteroscopic adhesiolysis 156
see also specific agents and maternal age 25, 26 hysteroscopic metroplasty 156, 157–8, 162
thrombophilias, hereditary 130–4, 185, 272 mosaic 24 hysteroscopic myomectomy 156, 159
autoimmune see antiphospholipid syndrome in recurrent aneuploidy 27 hysteroscopic polypectomy 156, 157
cohort studies 133 trunk defects 72 hysteroscopic septoplasty 157–9
fetal 135 tuberculosis 193 laparoscopic myomectomy 159
and fetal loss 139 tumor necrosis factor α (TNF-α) 63, 64 laparoscopic salpingectomy 161
early 131–2, 143 laparoscopy-assisted myomectomy
late 132–3 ultrasound 159–60
hereditary 127 advantages/disadvantages 154, 155 septum-sparing metroplasty 159
outcomes 236 in cervical incompetence 208–10, 209, 210 Strassman metroplasty 161
prevalence markers for aneuploidy 55–6 uterine cavity
first-trimester losses 131–2 three-dimensional 152–3 exploration 67–8, 272
late obstetric complications 132–3 advantages/disadvantages 155 flora 196, 196
and prognosis 275 bicornuate uterus 151 uterine integrity investigations 152–5
thromboprophylaxis didelphys uterus 150 uterine septum resection see septoplasty
arguments against 143–5 septate uterus 153, 158 uterus didelphys see didelphys uterus
arguments for 139–42 with sonohysterography 154
treatment 133–4 unicornuate uterus 149 vaginal bleeding
thrombophilic risk factors 139 umbilical cord defects 73 and IVIG 237, 238
thromboprophylaxis unfractionated heparin for APS 119–21 and paternal leukocyte immunization
arguments against 143–5 guidelines 120–1 237, 237
arguments for 139–42 in labor/postpartum 123 after RPL 232
clinical studies 144–5 mechanism of action 120 vaginal flora 195–7, 196
pregnancy outcomes 144 monitoring 121 viability
thrombosis risks 121 in extreme prematurity 220–4
in APS 109–10 unicornuate uterus 148, 148, 149, 149 and midtrimester loss 219–30
in fibrinogen replacement 130 universal tolerance biomarker 18–20 outcome
hormones and 135 Ureaplasma urealyticum 195, 196, 198, 200, and gestational age 222–3
in placenta 130–1 201, 202 and place of birth 220
thyroid abnormalities 80–2, 272 ureteroplacental arteries, thrombosis of 131 recent improvements 221
thyroid-stimulating hormone 81 uterine agenesis/hypoplasia 148, 148 and transfer in utero 221
tissue factor 109 uterine anomalies 68, 147–64 prematurity and RPL risk 221–2
tolerance of embryo 16–20 imaging 152–5 Victorian Infant Collaborative Study
mechanisms enhancing 170 diagnostic hysterography 154, 155 (VICS) 220–3
Tompkins metroplasty, modified 158 diagnostic laparoscopy 154, 155 vs other regional studies 223–4
Topper and Farquharson cerclage technique hysterosalpingography (HSG) 152, 155 Victorian Infant Collaborative Study
211–12 magnetic resonance imaging (MRI) 154, 155 (VICS) 220–3
TORCH infections 193–4 method choice 154–5, 155
testing for 272 sonohysterography 153–4, 155 Whipple’s disease 203
290

Abortus Berulang Trapi Dan Penatalaksanaan

Uploaded by

Copyright:

Available Formats

Abortus Berulang Trapi Dan Penatalaksanaan

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Abortus Berulang Trapi Dan Penatalaksanaan

Uploaded by

Copyright:

Available Formats

0415421306-FM 4/3/07 3:31 PM Page i

Recurrent Pregnancy Loss

SERIES IN MATERNAL-FETAL MEDICINE

Recurrent Pregnancy Loss

Howard JA Carp MB BS FRCOG

This edition published in the Taylor & Francis e-Library, 2008.

Tel: +44 (0)20 7017 5000

Library of Congress Cataloging-in-Publication Data

Data available on application

ISBN 0-203-93167-X Master e-book ISBN

ISBN-10: 0 415 42130 6 (Print Edition)

Distributed in North and South America by

Within Continental USA

Distributed in the rest of the world by

Composition by CEPHA IMAGING PVT LTD, Bangalore, India.

1. Epidemiology of recurrent pregnancy loss 1

3. Genetics of spontaneous abortions 23

Debate: Should fetal karyotyping be performed in RPL?

Debate: Should PGD be performed in RPL?

Debate: Should progesterone supplements be used?

Debate: Should hCG supplementation be used?

7. Antiphospholipid syndrome – pathophysiology 107

8. Diagnosis of aPL-associated abortions 115

9. Management of antiphospholipid syndrome in pregnancy 119

10. Defects in coagulation factors leading to recurrent pregnancy loss 127

Debate: Should thromboprophylaxis be used in hereditary thrombophilias with RPL?

11. Uterine anomalies and recurrent pregnancy loss 147

12. Immunobiology of recurrent miscarriage 165

Debate: Should immunotherapy be used in RPL?

13. Infections and recurrent pregnancy loss 193

14. Midtrimester loss – the role of cerclage 207

15. Midtrimester loss and viability 219

16. Obstetric outcomes after recurrent pregnancy loss 231

17. Coping with recurrent pregnancy loss: Psychological mechanisms 243

19. Investigation protocol for recurrent pregnancy loss 269

20. A patient’s perspective 281

Howard Cuckle MA MSc DPhil Mordechai Goldenberg MD

Andrea Riccardo Genazzani MD PhD Aida Inbal MD

Lucia Lazzeri MD Lesley Regan MD FRCOG

Pelle G Lindqvist MD PhD Daniel S Seidman MD MMSc

Siobhan M Quenby MD MRCOG Yehuda Shoenfeld MD

Joe Leigh Simpson MD Marighoula Varla-Leftherioti MD PhD

Professor Emeritus at Faculty of Life Sciences

Prof Howard JA Carp MB BS FRCOG

INTRODUCTION at least three consecutive miscarriages, whereas

RECURRENT PREGNANCY LOSS: CAUSES, CONTROVERSIES AND TREATMENT

experienced RPL.5 Data from a Danish question- 60

EPIDEMIOLOGY OF RECURRENT PREGNANCY LOSS

Not pregnant Miscarriages Births

RECURRENT PREGNANCY LOSS: CAUSES, CONTROVERSIES AND TREATMENT

Birth rate (%)

EPIDEMIOLOGY OF RECURRENT PREGNANCY LOSS

Pregnancy outcomesa Type of RPL Observed prevalence Expected prevalence p-value

B, M, M, M, M,… Secondary RPL 82 (79.6%) 51.5b (50.0%) < 0.00005

RECURRENT PREGNANCY LOSS: CAUSES, CONTROVERSIES AND TREATMENT

Johnson et al23 12.2 7.3

NS, not significant.

PARTNER SPECIFICITY Lifestyle factors are rarely, if ever, major causes

EPIDEMIOLOGY OF RECURRENT PREGNANCY LOSS

NUMBER OF MISCARRIAGES MATERNAL AGE