Practical Obstetric Hematology PDF

Obstetric Prelims 20/10/05 3:09 pm Page i
Practical Obstetric
Hematology
Obstetric Prelims 20/10/05 3:09 pm Page ii
Obstetric Prelims 20/10/05 3:09 pm Page iii
Practical Obstetric
Hematology
Peter Clark BSc MD FRCP FRCPath

Consultant Haematologist
Department of Transfusion Medicine, Ninewells Hospital
and Medical School, Dundee, UK
Ian A Greer MD FRCP FRCOG MFFP

Regius Professor of Obstetrics & Gynaecology
University of Glasgow,
Glasgow, UK
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2005 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works

Version Date: 20130415
International Standard Book Number-13: 978-0-203-44879-3 (eBook - PDF)
This book contains information obtained from authentic and highly regarded sources. While all reasonable
efforts have been made to publish reliable data and information, neither the author[s] nor the publisher can
accept any legal responsibility or liability for any errors or omissions that may be made. The publishers wish
to make clear that any views or opinions expressed in this book by individual editors, authors or contributors
are personal to them and do not necessarily reflect the views/opinions of the publishers. The information or
guidance contained in this book is intended for use by medical, scientific or health-care professionals and is
provided strictly as a supplement to the medical or other professional’s own judgement, their knowledge of
the patient’s medical history, relevant manufacturer’s instructions and the appropriate best practice guide-
lines. Because of the rapid advances in medical science, any information or advice on dosages, procedures
or diagnoses should be independently verified. The reader is strongly urged to consult the drug companies’
printed instructions, and their websites, before administering any of the drugs recommended in this book.
This book does not indicate whether a particular treatment is appropriate or suitable for a particular indi-
vidual. Ultimately it is the sole responsibility of the medical professional to make his or her own professional
judgements, so as to advise and treat patients appropriately. The authors and publishers have also attempted
to trace the copyright holders of all material reproduced in this publication and apologize to copyright
holders if permission to publish in this form has not been obtained. If any copyright material has not been
acknowledged please write and let us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmit-
ted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented,
including photocopying, microfilming, and recording, or in any information storage or retrieval system,
without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.
com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood
Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and
registration for a variety of users. For organizations that have been granted a photocopy license by the CCC,
a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used
only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at

http://www.crcpress.com
Obstetric Prelims 20/10/05 3:09 pm Page v
Contents
Preface vii
Glossary ix
Chapter 1 Hematology measurements in pregnancy 1
Chapter 2 Blood group serology and blood products 15
Chapter 3 Maternal anemia and hemato-oncology in pregnancy 43
Chapter 4 Bleeding disorders in pregnancy 66
Chapter 5 Antithrombotic therapy in pregnancy 88
Chapter 6 Management of thrombotic disorders 102
Chapter 7 Thrombophilia and pregnancy outcome 122
Chapter 8 Thrombocytopenia in pregnancy 140
Chapter 9 Hereditary red cell disorders 163
Index 189
Obstetric Prelims 20/10/05 3:09 pm Page vi
Obstetric Prelims 20/10/05 3:09 pm Page vii
Preface
In the last 10 years there has been a rapid expansion in the clinical interaction
between hematologists and obstetricians. As a result there is an increasing
challenge to hematologists, obstetricians and midwifery staff to understand and
manage the clinical manifestations of a number of rapidly developing areas of
hemato-obstetric science.
This book is intended to serve practitioners involved in the management of
pregnancy, from trainees in obstetrics, hematology and vascular medicine to
general practitioners involved in the day-to-day management of maternity care, as
well as midwifery staff and specialists in hematology, obstetrics and vascular
medicine. The book is designed to assist the reader to diagnose and treat these
conditions, but also provides the reader with a user-friendly, but authoritative,
source of information on the pathophysiology of hemato-obstetric problems that
incorporates the best practice contained within internationally accepted
guidelines of obstetric and hematological care.
The book is intended to assist the reader rapidly assimilate the essential aspects
of the pathophysiology of these complex disorders and provide an aide memoir
to the management of these conditions. As well as a guide to the pitfalls associated
with investigation and treatment, each chapter has a similar format and employs
extensive use of tables and figures.
The first chapter covers the physiological changes in hematological indices
caused by normal pregnancy. The second provides a succinct review of the basic
science of transfusion medicine and appropriate blood component usage as it
relates to obstetric practice. The subsequent chapters deal with the impact,
diagnosis and management of adverse transfusion events, hemolytic disease of the
newborn, anemia, malignant hematology, bleeding disorders, antithrombotic
therapy, thrombophilia, thrombotic disorders and red cell disorders in
pregnancy.
To assist the reader, each of these chapters conforms to a similar style that
includes: pathophysiology; presentation; differential diagnosis; diagnostic tests;
diagnostic difficulties; maternal complications, fetal complications; and potential
management complications. Each chapter also includes tables summarizing the
key points in diagnosis and management, as well as information on the likely
impact of new therapies and a preview of the developing issues relating to each
condition.
Obstetric Prelims 20/10/05 3:09 pm Page viii
Obstetric Prelims 20/10/05 3:09 pm Page ix
Glossary
ACCP American College of Chest Physicians
ACE Angiotensin-converting enzyme
ADAMTS A disintegrin and metalloproteinase with thrombospondin domains
ADP Adenosine diphosphate
AF Atrial fibrillation
AFLP Acute fatty liver of pregnancy
AIHA Autoimmune hemolytic anemia
AIN Autoimmune neutropenia
AIP Acute intermittent porphyria
AML Acute myeloid leukemia
APAS Antiphospholipid antibody syndrome
aPC Activated protein C
aPCR Activated protein C resistance
APS Antiphospholipid syndrome
APTT Activated partial thromboplastin time
Art T Arterial thrombosis
AST Aspartate Transaminase
AT Antithrombin
ATP Adenosine triphosphate
BMI Body mass index
CD Cluster of differentiation
CEP Congenital erythropoietic porphyria
CI Confidence interval
CIAO Cerebral ischemia of arterial origin
CMV Cytomegalovirus
CNS Central nervous system
CT Computerized tomography
CTG Cardiotocography
CVP Chorionic villus sampling
DDAVP Demopressin analog
DIC Disseminated intravascular coagulation
Obstetric Prelims 20/10/05 3:09 pm Page x
Glossary
2,3-DPG 2,3-diphosphoglycerate
dRVVT Dilute Russell viper venom test
DVT Deep vein thrombosis
ECG Electrocardiogram
EDTA Ethylenediaminetetraacetic acid
ELISA Enzyme-linked immunosorbent assay
EPCR Endothelial protein C receptor
Epo Erythropoietin
EPP Erythropoietic protoporphyria
ET Essential thrombocythemia
F Factor
FBC Full blood count
FDP Fibrin degradation products
FFP Fresh frozen plasma
FMH Feto-maternal hemorrhage
FPA Fibrinopeptide A
G6PDH Glucose-6-phosphate dehydrogenase
G-CSF Granulocyte colony-stimulating factor
GI Gastrointestinal
GPI Glycophosphatidylinositol
GVHD Graph-versus host disease
Hb Hemoglobin
HCY Homocysteine
HDN Hemolytic disease of the newborn
HELLP Hemolysis, elevated liver enzymes, and low platelets
HHV Human herpes virus
HIT Heparin-induce thrombocytopenia
HITTS HIT thrombosis syndrome
HLA Human leukocyte antigen
HPA Human platelet antigen
HPFH Hereditary persistence of fetal hemoglobin
HPLC High-performance liquid chromatography
HTLV Human T-cell leukemia/lymphoma virus
HUS Hemolytic uremic syndrome
ICH Intracerebral hemorrhage
x
Obstetric Prelims 20/10/05 3:09 pm Page xi
Glossary
IDA Iron deficiency anemia

IgG Immunoglobulin
INR International normalized ratio
ISI International sensitivity index
ITP Immune thrombocytopenia/idiopathic thrombocytopenic purpura
IUGR Intrauterine growth restriction
IVIgG Intravenous human normal IgG
LDH Lactate dehydrogenase
LDL Low-density lipoprotein
LFT Liver function test
LMWH Low-molecular-weight heparin
MAHA Microangiopathic hemolytic anemia
MB Methylene blue
MCH Mean corpuscular (red cell) hemoglobin
MCV Mean corpuscular (red cell) volume
MMA Methylmalonic acid
MRI Magnetic resonance imaging
MTHFR Methylene tetrahydrofolate reductase
MTP Mitochondrial trifunctional protein
NAD Nicotinamide adenine dinucleotide
NADH Reduced form of NAD
NADP Nicotinamide adenine dinucleotide phosphate
NADPH Reduced form of NADP
NAIT Neonatal alloimmune thrombocytopenia
NHL Non-Hodgkin’s lymphoma
NNA Normochromic normocytic anemia
NSAID Non-steroidal anti-inflammatory drug
nvCJD New variant Creutzfeldt-Jakob disease
PA Pernicious anemia
PAI Plasminogen activator inhibitor
PCR Polymerase chain reaction
PCT Porphyria cutanea tarda
PCV Packed cell volume
PET Pre-eclampsia
PF Platelet factor
xi
Obstetric Prelims 20/10/05 3:09 pm Page xii
Glossary
PMH Past medical history

PNH Paroxysmal nocturnal hemoglobinuria
PMN Polymorphonuclear cell
PPH Post partum hemorrhage
PT Prothrombin time
PTE Pulmonary thromboembolism
PTP Post-transfusional purpura
PTT Partial thromboplastin time
RA Rheumatoid arthritis
rbc Red blood cell
RFLP Restriction Fragment Length Polymorphism
rFVIIa Recombinant FVIIIa
RiCoF Ristocetin cofactor
rtPA Recombinant tissue plasminogen activator
SD-FFP Solvent detergent-treated FFP
SLE Systemic lupus erythematosis
stfR Serum transferrin receptor
TAFI Thrombin-activatable fibrinolysis inhibitor
TA-GVHD Transfusion-associated graft versus host disease
TAT Thrombin–antithrombin complex
TCT Thrombin clotting time
TF Tissue factor
TFPI Tissue factor pathway inhibitor
tfR Transferrin receptor
TIBC Total iron-binding capacity
tPA Tissue plasminogen activator
TRALI Transfusion-related acute lung injury
TT Thrombin clotting time
TTP Thrombotic thrombocytopenic purpura
UFH Unfractionated heparin
U/Q Ventilation/Perfusion
VTE Venous thromboembolism
vWD von Willebrand’s disease
vWF von Willebrand Factor
xii
Obstetric CH01 18/10/05 10:20 am Page 1
CHAPTER 1
Hematology
measurements in
pregnancy
Pregnancy results in significant changes in metabolism, fluid balance, organ
function and circulation, which are driven, in part, by estrogen and the presence
of the feto-placental unit. These dramatic changes influence a wide variety of
hematological parameters. A knowledge of these changes is essential when inter-
preting the results of hematological investigations to diagnose, or monitor,
illness in pregnancy.
Blood count
The maternal plasma volume begins to increase from around 6 weeks gestation,
but this is not accompanied by an immediate increase in red cell mass. The effect
of this is a dilutional anemia, which may be evident from the 7–8th week of gesta-
tion. Although the mechanism of this increase in blood volume is not fully
understood, it may be more marked in multiple pregnancies and occurs even
when iron stores are poor. The physiological purpose of this change may be to
reduce maternal blood viscosity, and improve delivery of oxygen and nutrients to
the fetus. The increase in plasma volume may reach 140% of non-pregnant
values and usually becomes maximal in the late second trimester. By this stage in
pregnancy, a 15–25% increase in red cell mass, driven by an increase in maternal
erythropoietin production, will also be present. These changes should not,
however, reduce the hemoglobin to <11g/dl in the first trimester or <10g/dl in
the second and third trimesters, and should result in a normochromic, normo-
cytic red blood cell pattern on the blood film. This ‘physiological anemia’ of
pregnancy does not worsen during the third trimester, which may reflect a reduc-
tion in maternal plasma volume and a stable red cell mass. Consequently, a rela-
tive increase in hemoglobin and hematocrit may be observed in the late third
trimester.
With normal pregnancy there may also be a rise in the mean corpuscular
volume (MCV), and a normal MCV does not, therefore, exclude iron deficiency.
Similarly, a modestly elevated MCV does not diagnose folate (or vitamin B12) defi-
ciency in pregnancy (Table 1.1).
1
Practical Obstetric Hematology
Table 1.1 – Blood count and hematinics in normal pregnancy
Parameter Change in pregnancy
Hemoglobin Falls, but >11g/dl in the first trimester and >10g/dl in the second and third
trimesters
Mean corpuscular No change, rise or fall all possible
volume (MCV)
Platelets >80 × 109/l
Neutrophils Variable rise (earlier forms seen)
Vitamin B12 <50% fall with increasing gestation
Serum folate <30% fall with increasing gestation
Red cell folate <20% rise, or no change with increasing gestation
Homocysteine No change, or fall
Vitamin B6 <20% fall with increasing gestation
Fe Variable fall with increasing gestation
Ferritin Variable fall; can be reduced to near non-pregnant iron-deficient ranges
Total iron-binding Variable rise with increasing gestation
capacity (TIBC)
Transferrin receptors (tfR) Increase after 20/40 weeks gestation
Towards the end of pregnancy at least 5% of women have a platelet count

below the usual non-pregnant reference range (150–400 × 109/l). The majority
of these lie above 80–100 × 109/l, although platelet counts as low as 50 × 109/l
without significant disease have been reported. This gestational thrombocytope-
nia is caused by a combination of hemodilution and an increase in platelet activa-
tion and clearance, which is partly compensated for by an increase in platelet
production and an increased proportion of younger, larger, platelets. The evid-
ence that an increase in platelet activation and clearance occurs includes: an
increase in mean platelet volume; an increase in the plasma concentration of the
platelet-specific protein β-thromboglobulin in the second and third trimesters;
and an increase in the metabolites of the prostanoid thromboxane A2 (which is
released following platelet activation, and which itself is a potent stimulus to
platelet activation and aggregation). Following delivery, the platelet count
increases in reaction to, and in compensation for, platelet consumption.
Normal pregnancy is associated with a rise in total white blood cell count. This
usually results in a mild neutrophilia (<15 × 109/l), but earlier granulocyte forms,
such as myelocytes and band forms, and a white count of 15–20 × 109/l is not
uncommon even in normal pregnancies. The presence of these more ‘primitive’
white cells may result in such samples being ‘flagged’ as abnormal in automated
white cell analyzers. Furthermore, the use of corticosteroids to assist in lung mat-
uration may result in maternal neutrophilia, and the process of delivery itself will
often precipitate a brisk leukocytosis.
2
Hematology measurements in pregnancy
Hematinic assessment
Normal pregnancy is associated with a progressive fall in serum iron levels and an
increase in serum total iron-binding capacity (TIBC: see Chapter 3). The TIBC
reflects the amount of the iron transport protein (transferrin) that is saturated
with iron; the level of TIBC fluctuates widely and is also influenced by recent
iron ingestion. A fall in serum ferritin levels also occurs in normal pregnancy,
although very low ferritin levels (<12ug/l) will still reliably indicate iron defi-
ciency. A number of newer tests of iron deficiency have been evaluated out-with
pregnancy. These include the plasma level of free transferrin receptors (tfR) and
the plasma level of free protoporphyrin. In the latter case, iron deficiency leads
to a reduction in the incorporation of iron into the prophyrin ring that forms
the heme moiety of hemoglobin. The level of the ‘unused’ protoporphyrin is
therefore a measure of deficiency. However, an increase in free protoporphyrin
occurs in normal pregnancy. The tfR is a transmembrane protein that mediates
delivery of iron to the developing erythroblast and is shed from the maturing red
cell. The serum tfr (stfR) level relates to the degree of iron deficiency and
increasing values precede changes in red cell morphology. stfR has been used
successfully to discriminate iron deficiency from the anemia of chronic disease,
as it does not increase in response to inflammation. However, stfR levels have
been shown to increase with gestation, returning to non-pregnant values 3
months postpartum. Although this could relate to undiagnosed iron deficiency,
it seems more likely to be due to an increase in maternal erythropoiesis.
Vitamin B12 levels may fall by 30–50% during pregnancy, with evidence of
reduced levels in the first trimester in some individuals. This occurs despite a diet
adequate in vitamin B12 and is likely to be due to a combination of increasing mater-
nal plasma volume, hormonal changes, increasing maternal requirement and
vitamin B12 transfer to the fetus. Overall, the bulk of current evidence suggests that
serum or plasma homocysteine falls with increasing gestation in most subjects. The
impact of folate repletion on these changes, however, requires further investigation.
Although vitamin B6 indices seem to decrease during pregnancy, it is not clear
whether normal pregnancy is associated with a vitamin-B6-deficient state or not,
and, even if it is, whether this has any clinical significance or not. In non-supple-
mented pregnancies there is a progressive fall in serum folate with increasing
gestation and either a rise, or no change, in red cell folate can both occur.
Hemostasis
Pregnancy is associated with significant changes in hemostasis. An understanding of
the normal hemostatic system and these physiological changes is essential to the diag-
nosis and management of bleeding and thrombotic disorders. It is also necessary to
understand the limitations of the D-dimer algorithms that are used in the screening
of non-pregnant subjects suspected of venous thromboembolism (VTE). Further-
more, it may provide an insight into the mechanisms of not only gestational venous
3
thrombosis, but also pregnancy loss and complications such as pre-eclampsia, which
have recently been associated with both acquired and heritable thrombophilia.
Pregnancy results in an upregulation in the maternal coagulation cascade,
which leads to an overall increase in thrombin generation. This maternal alteration
in coagulation is clearly a physiological preparation for delivery. As with many bio-
logical factors, this change in coagulation factors may relate to alterations in:
hormone-influenced synthesis; the volume of distribution; lipid interaction; or
catabolism. The progressive increase in maternal estrogen levels may alter the
uptake of coagulation proteins into the uterus and may also result in an increase in
the uterine generation of tissue factor. It is possible that both placental and mater-
nal thrombin have an important influence on cellular growth, and thrombin gen-
eration results in fibrin formation, which is essential for placental implantation.
Normal hemostasis control

Coagulation
Upon damage to the endothelium, platelets adhere to the subendothelium and
activation of the coagulation cascade occurs (Figure 1.1). This results in the com-
bination of a platelet plug and fibrin mesh formation at the site of injury. There
are, of course, substantial controls to limit thrombosis to sites of injury and
prevent excessive clot formation. Two cascades are still described in blood coagu-
lation. These are the intrinsic and the extrinsic pathways. It is now apparent that
the intrinsic pathway (which is activated when blood comes into contact with a
negatively charged surface) acts to amplify the coagulation activation occurring
as a consequence of the extrinsic pathway.
In the extrinsic pathway, when vascular injury occurs, cells (including endothelial
cells and monocytes) express tissue factor (TF), a membrane-bound protein (which
is also known as thromboplastin). When this is exposed to plasma, a complex of TF
and the plasma serine protease, Factor (F) VII is formed. TF acts as a cofactor for
activated FVII and is the most potent known initiator of the coagulation pathway.
Ninety-nine per cent of FVII circulates in the zymogen or unactivated form (FVIIc)
Figure 1.1 The coagulation cascade
The ‘extrinsic’ initiation of coagulation with a combination of tissue factor (TF), Factor (F) VII and activated FVII
(FVIIa) is shown. Solid lines indicate activation and dotted lines indicate inhibition; the principal pathway is
shown in bold type. The FVII–FVlla–TF complex activates FX (X) to FXa (Xa), as well as FIX (IX) to FIXa (XIa):
this results in the activation of prothrombin (PT) in a prothrombinase complex of calcium (Ca2+), phospholipid
(Plipid) and activated FV (Va), resulting in a burst of thrombin. This burst of thrombin activates FV (V) to FVa
(Va), FVIII (VIII) to FVIIIa (VIIIa) and FXI (XI) to FXIa (XIa). The combination of thrombin’s action and activation of
the ‘intrinsic’ coagulation pathway by exposure to damaged endothelium (via activation of FXII (XII) to FXIIa
(XIIa), results in activation of FXI (XI); this leads to activation of FIX (IX). The further activation of FX (X) in the
Tenase complex of FIXa (IXa), FVIIIa (VIIIa), Plipid and Ca2+ amplifies the coagulation initiated by the thrombin
burst. Thrombin also activates FXIII (XIII), which cross-links and stabilizes formed fibrin.
4
INTRINSIC
XII XIIa
XI
Ca2+
IX XIa

Ca2+/Plipid VIII
IXa
V
5
X VIIIa
TF/FVII/FVIIa
EXTRINSIC
Ca2+/Plipid PT Va
Tenase Fibrinogen
Xa
THROMBIN
Ca2+/Plipid burst Fibrin
Prothrombinase
XIII XIIIa Cross-linked

fibrin
and around 1% circulates as FVIIa, the activated form. Both FVIIc and FVIIa are
bound to TF expressed on the vessel wall and inflammatory cells, thereby localizing
the hemostatic response to the site of injury. Once FVIIc binds to TF it can be acti-
vated by TF itself. This complex can convert further FIXc to FIXa, FXc to FXa, and
FVc to FVa. Formed FXa binds to FVa to form a complex (the prothrombinase
complex) with calcium on a phospholipid membrane surface. In the presence of
these phospholipids, further FVc is converted to FVa and prothrombin is activated
leading to a ‘thrombin burst’. As prothrombin is cleaved by FXa to form thrombin,
a fragment is released. This fragment is called prothrombin fragment 1+2 (F1+2)
and, measured in plasma, can be used as a measure of thrombin generation. The
thrombin burst can then convert fibrinogen to fibrin, and activate parts of the
intrinsic pathway (FVIIIc and FXIc), as well as FVc, FXIIIc (which leads to cross-
linking and stabilization of formed fibrin), the endothelium, platelets and mono-
cytes. As thrombin is also involved in anticoagulation and fibrinolysis (see below), it
has a central role in orchestrating the hemostatic response.
In the intrinsic pathway, the binding of FXIIc to a negatively charged surface
leads to a local increase in FXIIc concentration, which results in its autoactiva-
tion. FXIIa causes the conversion of pre-kallikrein to kallikrein, and activates
FXIc to FXIa. FXIa, as well as the TF–FVIIa complex, activates FIX to FIXa. Acti-
vated FIXa forms a complex (the Tenase complex) with FVIIIa, calcium and
phospholipids. This results in activation of FXc to FXa, which contributes to the
prothrombinase complex and further thrombin generation.
Thrombin converts soluble fibrinogen to an insoluble fibrin polymer, which
seals the site of injury and protects damaged tissue during wound healing. Fib-
rinogen activation generates soluble fibrin monomers, fibrinopeptide A and fib-
rinopeptide B. The resultant fibrin mesh is stabilized by cross-linking catalyzed by
thrombin-activated coagulation FXIIIa.
Coagulation control
There are a number of controls of coagulation (Figure 1.2). TF pathway
inhibitor (TFPI) is released constitutively from endothelial cells, and, in associ-
Figure 1.2 Coagulation control
The principal points of physiological inhibition and downregulation of the coagulation cascade are shown. Solid
lines indicate a positive effect and dotted lines indicate inhibition. The main inhibitors are antithrombin (AT –
previously called antithrombin III). AT inhibits almost all of the coagulation factors, but its principal targets are
Factor (F) Xa and thrombin. The interaction of AT with thrombin results in the formation of the
thrombin–antithrombin (TAT) complex. Thrombin, when it binds to endothelial thrombomodulin (TM), loses its
procoagulant activity and activates protein C (PC) to activated protein C (aPC). aPC, with its cofactor protein S
(PS), cleaves activated FV (Va) and activated FVIII (VIIIa). PC can also be activated independently by the endothe-
lial protein C receptor (EPCR). Other inhibitors of coagulation include tissue factor pathway inhibitor (TFPI),
which rapidly inactivates the FVII–FVIIa–tissue factor (TF) complex. In addition, antitrypsin and C1-esterase
inhibitor are capable of inhibiting FXIa and FXIIa respectively. Ca2+, Calcium; Plipid, phospholipid.
6
INTRINSIC
C 1-esterase Inhibitor
XII XIIa
XIIi
XI
2+
Ca
IX XIa Antitrypsin

XIi VIII
Ca2+/Plipid
TM
IXa aPC PC
(EPCR)
V +PS
7
X VIIIa VIIIi +PS
TF/FVII/FVIIA
EXTRINSIC
Ca2+/Plipid PT Va Vi
Tenase Fibrinogen
Xa
AT THROMBIN
TFPI Ca2+/Plipid burst Fibrin
Xi
Prothrombinase
TAT

fibrin
ation with FXa, is able to rapidly inactivate the TF–FVIIa complex. Other con-
trols include antithrombin, which can inhibit many activated clotting factors
including thrombin and FXa. By binding to antithrombin, thrombin forms a
stable thrombin–antithrombin (TAT) complex, which is rapidly cleared from the
circulation. Another regulatory mechanism is protein C, which circulates as an
inactive zymogen. To become activated it binds to the endothelial protein C
receptor (EPCR), or the transmembrane protein thrombomodulin, which is
expressed on endothelial cells. The formation of a thrombin–thrombomodulin
complex on the endothelial surface accelerates thrombin’s activation of the
natural anticoagulant protein C by around 20 000-fold. Formation of this
complex directly inhibits the capacity of thrombin to cleave fibrinogen and acti-
vate platelets. To function as an inhibitor, activated protein C (aPC) dissociates
from EPCR and binds to its cofactor protein S. Protein S has no enzymatic activ-
ity of its own, but acts as a cofactor to aPC in the inactivation of FVa and FVIIIa.
This then results in a marked reduction in activity of the prothrombinase
complex.
Additional inhibitors include the complement component C1-esterase
inhibitor (which inhibits FXIIa) and antitrypsin (which can inhibit factor FXIa).
In addition, α2-macroglobulin acts as a broad-spectrum proteinase inhibitor,
which has a secondary function to other proteinases and may also be involved in
the regulation of the immune system.
Fibrinolysis
Additional clotting control comes from lysis of the fibrin clot (Figure 1.3). Fibri-
nolysis results from a series of enzymes that control the cleavage of fibrin. These
include plasminogen, which when activated by plasminogen activators results in
plasmin. This lyses fibrin to fibrin degradation products (including D-dimers).
The major activator of plasminogen is tissue plasminogen activator (tPA) – a
serine protease produced by endothelial cells – but FXIIa, FXIa, kallikrein and
Figure 1.3 Fibrinolysis
The action and principal controls of fibrin breakdown (fibrinolysis) are shown. Solid lines indicate activation and
dotted lines indicate inhibition.Tissue plasminogen activator (t-PA) activates plasminogen to plasmin, which can
break down both fibrin and fibrinogen. This results in the generation of a number of fibrinogen and fibrin
degradation products, which can have an overall inhibitory effect on further fibrin clot generation. Products
released from fibrinogen include D and E fragments, and products released from formed, cross-linked fibrin
include D-dimer, Y-dimer and DY fragments. t-PA is, in turn, inhibited and controlled by plasminogen activator
inhibitor (PAI)-1. In pregnancy the placenta produces PAI-2, although it does not have a major function in the
control of fibrinolysis in the mother. Plasmin is principally inhibited by α-2 antiplasmin, although there is a
contribution from α-2 macroglobulin. Thrombin is also capable of the activation of thrombin activatable
fibrinolysis inhibitor (TAFI), which inhibits the action of plasmin on formed fibrin. In addition, urokinase (u-PA),
which is found in urine, can activate plasminogen. uPA itself is activated from pre-u-PA after limited hydrolysis
by kallikrien, which is derived from the intrinsic pathway of coagulation.
8
Pre-u-PA
PAI-1
Coagulation cascade Kallikrien
u-PA t-PA
Fragments D/E
PAI-2

α-2 macroglobulin
α-2 antiplasmin
FIBRINOGEN
9
Plasmin Plasminogen
THROMBIN
burst
FPA
Fibrin
XIII XIIIa
CROSS-LINKED
FIBRIN
D-DIMER
Thrombomodulin DY/ YY FRAGMENTS
TAFI
urokinase (uPA) are also capable of activating plasminogen. Release of tPA from
the endothelium near the site of injury may result from the action of fibrin, by
thrombin attached to the fibrin clot or by the effect of venous occlusion itself.
The primary inhibitor of plasmin is α2-antiplasmin. Other potential plasmin
inhibitors include α2-macroglobulin, antithrombin, α1-antitrypsin and C1-esterase
inhibitor. In addition, there are four plasminogen activator inhibitors of which
plasminogen activator inhibitor (PAI)-1 is the most important in vivo. PAI-1 is
released from the endothelium in response to agents such as thrombin and
endotoxin. PAI-2 is produced only in the placenta and so is only found in preg-
nant women. In addition to these controls, thrombin, when bound to thrombo-
modulin, can activate thrombin-activatable fibrinolysis inhibitor (TAFI). TAFI
cleaves the COOH terminal end of fibrin. This reduces the ability of fibrin to
facilitate plasminogen activation via tPA. TAFI also directly inhibits plasmin activ-
ity and, when cross-linked to fibrin, prevents premature lysis.
Pregnancy-associated changes in coagulation and fibrinolysis

Coagulation
Increasing gestation is associated with a progressive increase in the plasma con-
centrations of fibrinogen, FVIIc, FVIIIc, FXc and FXIIc, von Willebrand factor
(vWF) antigen and the functional measure of vWF, ristocetin cofactor activity (see
Chapter 4). No consistent change in FVc and FIXc is seen, but a reduction in
FXIc is likely (Table 1.2). Despite the large number of longitudinal and cross-
sectional studies of hemostasis in normal pregnancy, the mechanism and signific-
ance of these changes remain poorly understood. An increase in TAT complex for-
mation occurs in normal pregnancy, and TAT complex levels higher than
non-pregnant control values have been reported to be present in 50% of women
during the first trimester, with all subjects showing elevated levels in the second and
third trimesters. A significant positive correlation between gestational stage and
F1+2 has also been shown, with levels elevated beyond non-pregnant subjects seen
early in pregnancy. In addition, pregnant plasma, in in vitro experiments, is capable
of more rapid and elevated generation of thrombin than non-pregnant plasma.
This, in combination with evidence of an increase in prothrombin activation and
TAT complex formation (in the absence of any increase in antithrombin levels) sug-
gests that a heightened thrombin generation potential may be a feature of normal
pregnancy and, it has been suggested, may indicate a prothrombotic state.
Coagulation control in pregnancy

A number of studies have shown no difference in protein C antigen, or activity,
when pregnant subjects were compared with those who were not pregnant. In
most studies, no effect of increasing gestation on protein C (activity or antigen
levels) or antithrombin levels has been observed (Table 1.3).
10
Table 1.2 – Coagulation in pregnancy
Change from first to Change from first to

second trimester (%) third trimester (%)
Fibrinogen 11 31

Prothrombin 0 0
FVc 0 0
FVIIc 16 42
FVIIIc 28 62
FIXc 0 0
FXc 15 20
FXIc 5 5
FXIIc 16 14
vWF antigen 22 96
TAT 82 128
F1+2 80 140
FPA 83 80
D-dimer 170 180
Soluble fibrin 26 45
F, Factor; FPA, Fibrinopeptide A; TAT, Thrombin–antithrombin complex; vWF, von Willebrand Factor.
Pregnancy results in a fall in free and total protein S levels, with levels less
than those of non-pregnant subjects observed in the first trimester, and levels
<50% of those observed in the first trimester evident between 36 and 40 weeks
gestation.
Table 1.3 – Coagulation control in pregnancy
Change from first to Change from first to

second trimester (%) third trimester (%)
Protein C activity 0 0
Protein C antigen 0 0
Antithrombin activity 0 0
Protein S total 9 15
Protein S free 20 30
Plasma thrombomodulin 10 41
APTT–APCR 13 20
Modified APTT–APCR 0 0
aPCR, Activated protein C resistance; APTT, Activated partial thromboplastin time.

11
As is discussed in Chapter 7, the significance of inherited resistance to aPC

(aPCR) was recognized by Dahlback and co-workers in 1993 (Dahlback B, et al
1993). In the majority of subjects, inherited resistance is due to the Leiden muta-
tion in the gene coding for coagulation FV (FV Leiden). Pregnancy is associated
with a progressive increase in aPCR, which is independent of the Leiden muta-
tion. This resistance can be assessed by activated partial thromboplastin time
(APTT) or TF-based assays (see below). In APTT-based methods, the results are
reported as a sensitivity ratio (which is inversely related to the degree of resis-
tance). Modification of these assays by pre-dilution of the patient plasma with FV-
deficient plasma renders the tests highly specific for the mutations in the FV
gene, and insensitive to acquired aPC resistance or the effects of pregnancy.
Thrombomodulin, as noted above, is found on maternal vascular endothelial
cells and on the trophoblastic surface of the placenta. A soluble form of throm-
bomodulin exists in plasma and urine, and has been widely used as a marker of
endothelial damage. Normal pregnancy may be associated with an increase in
soluble thrombomodulin levels, but the relationship between this cleaved form
and the overall anticoagulant effect of thrombin (via aPC generation) is
unknown. aPC is known to be regulated by several inhibitors, including protein C
inhibitor-1, α1-antitrypsin, α2-antiplasmin, C1 esterase inhibitor and α2-macroglob-
ulin. It is possible to assess the activity of thrombin on protein C activation by
measuring aPC, by measuring the peptide released from protein C on activation,
or by measuring complexes of aPC with its inhibitor α1-antitrypsin. An increase in
aPC–α1-antitrypsin in the third trimester to near double the level of that observed
in the first trimester occurs in normal pregnancy. This does not correlate with α1-
antitrypsin, suggesting that increasing gestation is associated with an increase in
the activation of protein C and therefore an increase in thrombin-mediated
antithrombotic activity.
The overall balance of the prothrombotic and antithrombotic effects of
thrombin has not been clarified, but higher levels of fibrinopeptide A (a frag-
ment released on the activation of fibrinogen), are also a feature of pregnancy,
indicating that there is a parallel increase in fibrin generation.
Fibrinolysis
The activation and inhibition of fibrinolysis occurs at a local level, so the study of
plasma can only give a limited insight into the balance of fibrinolysis during preg-
nancy. Increasing gestation is associated with a complex alteration in plasma
markers of plasminogen activation and inhibition. Plasminogen activity, and PAI-
1 and PAI-2 levels increase throughout pregnancy. This is accompanied by a
probable increase in tPA antigen but a reduction in its activity and release after
venous occlusion. No alteration in α2-antiplasmin levels has been reported in
normal pregnant subjects. A number of studies have shown that increasing gesta-
tion is associated with an overall reduction in systemic fibrinolytic activity.
Despite this, pregnancy is associated with an increase in circulating fibrin degra-
12
dation products. The reason for this apparent disparity is not known, although it
may point to an increase in fibrin generation and degradation in the placenta
(despite a reduction in the fibrinolytic potential of the general circulation), or to
a reduction in the plasma clearance of these products.
Delivery and the early puerperium

There are very few studies that have systematically examined the longitudinal
changes in coagulation that occur during labor and delivery. Many coagulation
factors show little change during labor, but a fall in fibrinogen levels, with a nadir
evident at the delivery of the placenta occurs. Levels then rise, perhaps as early as
3 hours after delivery, toward antenatal values. Fibrin degradation products are
elevated in early labor, fall during placental delivery and rise immediately again
(along with soluble fibrin and D-dimer and fibrinopeptide A levels), indicating
an alteration in intravascular coagulation. Whether there is a detectable alter-
ation in the overall ex vivo fibrinolytic activity of maternal plasma evident in the
first few hours after delivery is unclear.
No change in total protein S levels during delivery has been reported and a
gradual return to non-pregnant levels of both free and total protein S becomes
evident in the first week of the puerperium. The effect of delivery on protein C is
not fully understood. Maternal plasma antigen levels may fall during the delivery
prior to the clamping of the placental cord, rising again after clamping. The
duration of any elevation in levels is unclear, as elevated levels of protein C
antigen and activity may occur in the first 3–5 days after delivery, but levels may
also fall, or remain unchanged. Further study is required to clarify the influence
of delivery on protein C levels. Higher acquired aPC resistance, when compared
with non-pregnant subjects, does occur at delivery and a return towards non-
pregnant values is seen in the first week after delivery.
In the first postpartum week, no reduction (and perhaps a further elevation)
of fibrinogen levels from third trimester values is seen in longitudinal investiga-
tions. This may be accompanied by a slight rise from antenatal levels in FVc and
FXIc. A fall from antenatal levels of FVIIIc and FVIIc is evident between 1 and 7
postpartum days. In the first week of the puerperium, thrombin generation,
soluble fibrin levels and D-dimer formation remain higher than in early preg-
nancy. The first postpartum week is also associated with a fall, from antenatal
values, in plasma plasminogen and (consistent with a substantial placental
origin) in PAI levels.
Changes in other coagulation assays during pregnancy

The partial thromboplastin time (PTT) measures the intrinsic pathway of coagu-
lation and is so called because the coagulation activator used is phospholipid. By
contrast, the prothrombin time (PT) is activated by ‘complete’ thromboplastin: a
mixture of TF and phospholipid. There is considerable variation in the lipid
13
composition of the phospholipid used in different partial thromboplastin tests.

This means that each test will vary slightly in its interaction with the clotting
factors in the sample being tested. In addition, an activator is used to improve
consistency of the assay. Such activators include substances such as kaolin, silica,
celite and ellagic acid. To perform this APTT, the patient blood sample is taken
into citrate anticoagulant (which chelates ionized calcium, preventing coagula-
tion). The partial thromboplastin and activator are added along with calcium
and the time taken for fibrin clot to form is recorded.
PT is a measure of the extrinsic coagulation pathway and is the time taken for
a citrated sample of blood to form visible fibrin clot after the addition of calcium
and animal-derived TF to the sample.
In most studies of pregnancy, no marked change in the PT occurs with
increasing gestation. However, the APTT may shorten (depending upon the test)
by up to 4 seconds towards the end of pregnancy, and is also shortened following
a recent VTE or surgery. In pregnancy this shortening is likely to be due to the
marked physiological increase in FVIIIc, which occurs with increasing gestation.
This foreshortening should also be considered when monitoring unfractionated
heparin.
Similarly, although low-molecular-weight heparin (LMWH) dosing is often
based on non-pregnant kinetics, recent work suggests that pregnancy may result
in a later peak of anti-Xa levels (using prophylactic doses) than that found in
non-pregnant subjects (4 hours as opposed to 2 hours). Increasing gestation
appears to be associated with a reduction in anti-Xa levels.
References
Dahlback B, Carlsson M, Svensson PJ, Proc Natl Acad Sci USA 1993;
90(3):1004–8.
14
CHAPTER 2
Blood group serology

and blood products
Blood transfusion
There are 23 blood group systems that are coded by single or closely linked
genes. The genes for the ABO and Lewis systems code for enzymes (glycosyltrans-
ferases), which transfer sugar molecules onto a carbohydrate chain on the red
cell and other tissues. On the red cell, these A and B enzymes add sugars to the
group O substance (more properly called the H antigen) to form blood groups A
and B, respectively, or, in combination, group AB. The expression of these anti-
gens on the surface of red cells and platelets is important, as they can lead to the
formation of destructive alloantibodies when presented to antigen-negative sub-
jects. The presence of ABO substances on endothelial and epithelial membranes
can also lead to problems with solid organ and stem cell transplants.
Forty-two per cent of Caucasians and 36% of Black subjects are group A, with
16 and 25% of Caucasians group B or AB, respectively; the remainder are group
O. In addition, there are a number of A, B and O antigen subtypes. For example,
individuals typed by standard serology as group A can be further genetically sub-
typed as A1A1, A1A2, A1O, A2A2 or A2O. Indeed, the majority of group A subjects
are in fact A1O. Rarely, some individuals cannot form the O (H) precursor (the
so-called Bombay phenotype), and as a result develop naturally occurring anti-A,
-B and -H (O) antibodies, making transfusion almost impossible.
In 80% of Caucasians the secretor gene (Se) on chromosome 19 also codes for
a glycosyltransferase. This leads to secretion of free A or B substances into plasma
and secretions.
The clinical significance of any particular red cell antigen depends upon: the
frequency of the antigen; the ease of antibody development to it; the amount of
antigen exposure in an antigen-negative subject; and the destructive capacity of the
antibody (which depends upon its type, its thermal range and its ability to bind
complement or cross the placenta). In general, antibodies of the immunoglobulin
(IgM) M class bind complement more readily than IgG. IgM antibodies to groups
A or B occur naturally in subjects who do not carry A or B, and are thought to arise
from natural exposure to an unknown substance that is antigenically similar to the
A or B antigens. IgM antibodies, unlike IgG, do not cross the placenta.
Antibodies formed after exposure to a previously ‘unseen’ antigen (usually as
a consequence of transfusion or pregnancy) are usually of the IgG class. When
15
there has been a previous exposure, re-exposure may result in an anamnestic

response (a rapid rise in antibody titer), which is the mechanism behind a
delayed transfusion reaction.
Some individuals, such as those with autoimmune disorders or sickle-cell
disease, appear more susceptible to the production of red cell antibodies. In con-
trast, individuals with acquired, physiological (i.e. neonates), or congenital
hypogammaglobulinemia, appear to have a reduced susceptibility.
Blood component therapy

To ensure the effective use of donated blood and to provide plasma for fractiona-
tion, blood is prescribed as its components: red cells; platelets; fresh frozen
plasma (FFP); cryoprecipitate; fractionated clotting factor concentrates and
immunoglobulins (Table 2.1). In addition, there are a number of human, or
recombinant, clotting factors available for the management of hemophilia, throm-
bophilia and sepsis. These include recombinant Factor (F) VIIa, recombinant
antithrombin, human-derived protein C and recombinant activated protein C.
Red cells
Although many transfusion triggers have been proposed, as a general rule, red
cells should be used to treat significant hypoxia due to anemia, i.e. anemia
causing cardiac or respiratory distress. In certain circumstances, red cell transfu-
sions may also be used for other purposes, such as the suppression of dyserythro-
poiesis in thalassemia, or the dilution of sickle cells in sickling disorders.
Table 2.1 – Blood component usage
Component Indications
Red cells Hypoxia secondary to anemia
Platelets Thrombocytopenia + bleeding
Fresh frozen plasma (FFP) PT/APTT prolonged + bleeding
Cryoprecipitate Hypofibrinogenemia + bleeding

Uremia + bleeding
Fibrinogen concentrate Hypofibrinogenemia + bleeding

Placental maintenance
Recombinant Factor VIIa (rFVIIa) Life-threatening bleeding + no response to conventional coagulation/

platelet/surgical therapy
APTT, Activated partial thromboplastin time; PT, Prothrombin time.

16
Blood group serology and blood products
Commonly, the transfusion laboratory will perform a ‘group-and-save’ pro-

cedure for uncomplicated deliveries or Cesarean sections. In this, the blood is
typed to determine the ABO and Rh groups, and the patient’s serum (or plasma)
is tested against a panel of red cells to determine the presence of significant allo-
or autoantibodies. If antibodies are detected, the laboratory should proceed to
identify the antibody and immediately cross-match the patient’s plasma so that
compatible blood is available. When no antibody is detected, a rapid cross-match
(taking around 10–15 minutes) is performed only when blood is required.
Although the majority of patients will have no significant antibodies, the pres-
ence of an antibody may delay the provision of compatible blood for several
hours, and in some cases for several days. In an extreme emergency, ABO and Rh
group-specific, or donor blood phenotyped to be compatible with the major
hemolytic antigens in the patient, may have to be used.
FFP
FFP contains all the labile clotting factors and is used to correct a coagulation
deficiency – identified by prolongation of the prothrombin or partial thrombo-
plastin times – when bleeding is ongoing, or is predictable, e.g. immediately
prior to surgery or closed procedures such as liver biopsy, lumbar puncture and
epidural anesthesia. As it is a human blood product (with the attendant risks of
immunological or infectious complications), it should not be used for fluid
replacement and, where possible, single clotting factor deficiencies should be
treated with specific virally inactivated fractionated products. It should be
remembered that provision of FFP to the patient will require at least 30
minutes to permit thawing of the product and transport from the transfusion
laboratory.
The exceptions to the indications noted above are the provision of the pro-
tease required in the treatment of thrombotic thrombocytopenic purpura
(TTP)/hemolytic uremic syndrome (HUS) (see Chapter 8), and the provision of
anticoagulant proteins such as protein S when a specific product is not available.
FFP has no role in the reversal of heparin, but can be used in the reversal of war-
farin. However, severe bleeding secondary to warfarin will require a prothrombin
concentrate, as FFP will not lead to an adequate restoration of plasma FIXc levels
in these circumstances.
Cryoprecipitate and Fibrinogen

Cryoprecipitate is used when the plasma fibrinogen is reduced and there is
bleeding, or bleeding is likely. In the non-pregnant state, transfusion should be
considered when the plasma fibrinogen is <1g/l. In pregnancy there is a physio-
logical rise in plasma fibrinogen with increasing gestation. The threshold for its
use in pregnancy may therefore require adjustment to account for these changes
(see Chapter 1). In subjects with congenital hypofibrinogenemia, fibrinogen has
17
also been used to maintain the placental integrity. Where a licensed virally inacti-
vated fibrinogen concentrate is available, this should be used when such fibrino-
gen treatment is required.
Platelets
Platelets are produced either by plasma exchange (where one adult dose is
obtained from one donor) or from the buffy coat of whole blood donations
(where four donations are required to provide an equivalent adult dose).
Platelets are stored at room temperature (specifically 22 +/–2°C) and have a shelf
life of 5 days if continually agitated. Platelet transfusions are indicated in the pre-
vention and treatment of bleeding due to thrombocytopenia (a reduced platelet
count) or thrombasthenia (a qualitative platelet defect). As with all plasma prod-
ucts, platelets carry the risks of transfusion outlined below. In addition, there is
the potential risk of maternal RhD alloimmunization.
Thrombocytopenia can occur for a number of reasons (see Chapter 8).
Although, in general, platelets should only be used where bleeding is attended by
thrombocytopenia, there is also a role, albeit more limited, for prophylactic use.
A threshold for prophylaxis is not evidence based, but there is consensus that a
threshold of 10 × 109/l is now acceptable if there are no additional risk factors
such as infection, vomiting or coagulopathy. This contrasts with the previous
accepted threshold of 20 ×109/l. An individual decision on dosing will be
required if there is a concern over platelet refractoriness, or if there is immune
thrombocytopenia (ITP: when donor platelets may be destroyed in the same way
as endogenous platelets), or when the patient has a stable chronic thrombocy-
topenia.
In non-pregnant subjects, a platelet count of 50 × 109/l has been adopted as
the safe minimum for surgical procedures, or when there is major bleeding. With
Cesarean sections, a level of at least 80 × 109/l has been recommended to allow
epidural anesthesia; for a vaginal delivery (without epidural anesthesia), a level
of 50 × 109/l is acceptable. Clearly, in the peripartum patient with thrombocy-
topenia, non-steroidal analgesics should be avoided. In each instance a check
should be made that the appropriate platelet count has been achieved before the
commencement of surgery.
In platelet function disorders, medication that interferes with platelet func-
tion should be avoided or withdrawn (such as non-steroidal analgesics and
aspirin), and the hematocrit should be optimized prior to platelet transfusion.
Desmopressin (trade name DDAVP: see Chapter 4) should be considered if there
is an inherited platelet defect. As DDAVP is very similar to oxytocin, there was
concern that DDAVP might induce uterine contractility. However, it is now
known that these agents operate through different receptors and that DDAVP
does not cause uterine contractions.
When RhD-positive platelets are transfused to an RhD-negative woman (who
is pregnant or in the reproductive age group), a dose of 250 IU of anti-D
18
should be given to cover between 3 and 5 adult doses of RhD-positive platelets

in any 6-week period. Where there is thrombocytopenia it can be given subcuta-
neously. Platelets intended for intrauterine transfusion should be, where pos-
sible: <24 hours old; leukodepleted; apheresis derived; cytomegalovirus (CMV)
sero-negative and gamma irradiated. These factors will reduce the infusion
volume and the risks of infection and transfusion-associated graft versus host
disease.
In non-pregnant subjects, dosing formulas can be used to attempt to predict
the dose required to achieve a particular platelet increment. In practice,
however, it is usual to give a single adult dose (as noted above, derived from one
apheresis or four whole blood donors) over a 30 minute period. As noted above,
if there is doubt as to efficacy, the platelet increment can be checked at either 1
or 24 hours after dosing.
Immunoglobulin
Intravenous human normal Ig (IVIgG) is used in the management of a number
of conditions which can occur in pregnancy. These include ITP, neonatal alloim-
mune thrombocytopenia (NAIT), post-transfusional purpura (PTP: see Chapter
8), Guillain-Barré syndrome, myesthenia gravis and some cases of hemolytic
disease of the newborn (HDN). Although immunoglobulin has also been used in
the management of recurrent fetal loss, recent systematic reviews and clinical
trials have shown no evidence for this use. However, further information on
patient selection and the timing of infusions may yet indicate a subset of patients
who may benefit from it.
As with any blood component there is a risk of pathogen transmission,
although modern IVIgG should undergo at least two distinct viral inactivation
procedures; and furthermore, there is some evidence that the process of fraction-
ation itself leads to a reduction in prion contamination. The most commonly
reported significant side effects with IVIgG are fever, acute renal failure, jaun-
dice, thrombosis and a positive direct antiglobulin test (+/– autoimmune hemol-
ysis). However, these problems may also relate to the underlying condition.
Acute tubular necrosis appears more common in products containing sucrose,
and occurs more frequently in subjects with pre-existing renal disorders or those
on nephrotoxic drugs. In such cases a low sucrose product should be used. In all
cases, renal function should be checked prior to infusion and monitored in the
days after completion. In addition, the maximum dose and dose rate should
never be exceeded. Mild hypersensitivity to IVIgG is not uncommon, but severe
reactions are rare and may be associated with severe IgA deficiency. There are a
number of reports of venous thrombosis, perhaps related to increased plasma vis-
cosity or enhanced platelet aggregation; and there are also reports of cerebral
ischemia of arterial origin (CIAO) in association with IVIgG use. The exact risks
are, however, hard to define, as no substantial comparative trials of different
IVIgG products have ever been carried out.
19
Recombinant factor VIIa (rFVIIa)

rFVIIa has been used extensively in the management of acquired inhibitors to
FVIII. rFVIIa (complexed with tissue factor) generates thrombin, via activation of
FX, on activated platelets. At the dose of rFVIIa given therapeutically this reac-
tion is independent of FVIII and FIX. The requirement for exposed tissue factor
and activated platelets should, however, limit rFVIIa action to the site of injury.
The full thrombin burst generated by therapeutic-dose rFVIIa appears to be suffi-
cient to cause hemostasis in a wide variety of clinical circumstances, including
thrombasthenia and thrombocytopenia, although there is, as yet, little clinical
trial evidence for its use. The usual dose is 90–120µg/kg given 2–3 hourly;
although a full hemostatic effect may be observed after one dose and a useful
therapeutic effect may even be achieved at half this dose. The appropriate
method of monitoring rFVIIa is not fully defined, but thrombin generation,
thromboelastography, or whole blood clotting may prove useful in the future.
At present, the risk of venous thrombosis, myocardial infarction, CIAO and
disseminated intravascular coagulation (DIC) – in particular in those with pre-
existing liver disease, cardiac disease, or pregnancy – can only be assessed from
case reports. However, rFVIIa has been successfully used in the management of
post- and antepartum hemorrhage. Although the role of rFVIIa in acquired
bleeding disorders is not defined, if it is used in subjects with a severe life-threat-
ening bleeding diathesis (when coagulation screens and bleeding sources have
been, or cannot be, corrected by conventional means) there is likely to be a
favorable clinical and cost–benefit ratio.
Transfusion in warm autoimmune hemolytic anemia

The provision of compatible blood for patients with autoimmune hemolytic
anemia can be significantly problematic. The indications for transfusion,
however, are the same as for those without an autoantibody, as the survival of red
cells in such patients should be sufficient to improve oxygenation. As with any
transfusion, however, red cell compatibility tests should be carried out to exclude
the presence of any alloantibody that could cause significant hemolysis of trans-
fused red cells. This is achieved by testing the patient plasma against a panel of
donor cells to determine if antibodies to any of these cells are present in the
plasma. Some autoantibodies will react against the majority of donor cells, and
such a non-specific reaction may mask an alloantibody and make the provision of
a compatible transfusion difficult and extremely time consuming.
The presence of an autoantibody is suspected when there is a positive direct
antiglobulin (Coombs’) test. This indicates the presence of an antibody coating
the patient’s cells. Although, as noted above, free autoantibodies in the patient
plasma often react to all of the cells in a donor panel, occasionally the plasma
will react more strongly to some cells than others. This may indicate that the
autoantibody has some specificity to one particular antigen, or that an alloanti-
20
body is also present. When, as is more often the case, the patient’s serum reacts
equally against all donor cells, there are a number of strategies to determine if
there is also an underlying alloantibody.
If the patient has not received a transfusion within the previous 3 months,
then it can be assumed that there will not be significant levels of any previously
transfused donor red cells in the patient. In these circumstances, using a whole
blood sample, any autoantibody bound to the patient’s red cells is removed with
an enzyme (‘ZZAP’). These ‘stripped’ cells are then used to adsorb the free
autoantibody from a sample of the patient’s plasma at 37°C. This process
(autoadsorption) is repeated several times to remove as much of any autoanti-
body from the patient plasma as possible. The resultant plasma is then used
against donor cells to determine if any alloantibody is present. Such an autoad-
sorption procedure requires a considerable sample of the patient’s blood.
If there has been a history of transfusion within the previous 3 months then
any donor cells still in the patient may bind to the alloantibody, making detec-
tion of an alloantibody unreliable. If autoadsorption cannot be carried out, then
an attempt to elute the autoantibody from the patient’s serum using allogenic
red cells with a variety of antigens can be attempted. A combination of reactions
will often assist in detecting a significant hemolytic red cell antibody.
One further strategy is to phenotype the patient and select donor blood that
carries the same antigens. Ideally, this phenotyping should include the RhD, C, E,
c, e, Kell, Jka, Jkb, Fya, Fyb, S and s antigens. However, the presence of an autoanti-
body that is adherent to the patient’s red cells may make such extended pheno-
typing impossible, and in such cases autoadsorption will be required.
If the patient’s autoantibody is only cold acting, then the compatibility test
can be carried out strictly at 37°C. In such circumstances, the autoantibody will
have no effect on the compatibility procedure. Such patients should, of course,
be given their transfusion through a blood warmer.
In extremely urgent circumstances, the appropriate blood for transfusion may
have to be selected upon the basis of the ‘least reaction’ of donor cells against the
patient’s plasma. This is an extremely unreliable approach and every effort should
be made to formally exclude a significant alloantibody. Further, whether such a
selection of ‘least incompatible’ blood (based upon a slight variability in reactivity in
testing of the plasma against a variety of donor cells) improves the survival of donor
cells in the patient, when an alloantibody has already been excluded, is unknown.
In urgent circumstances, communication between the laboratory and the clin-
ician is vital, as there may often be time to carry out a limited phenotype to
improve the safety of the transfusion when time does not permit an auto-adsorp-
tion procedure.
Transfusion reactions
Reactions to transfusions can occur for a number of reasons (Table 2.2). Of these,
hemolytic reactions are the most clinically important given their potential fre-
21
quency and severity. Other reactions, such as transfusion-associated graft versus

host disease (TA-GVHD: see below), although significant, occur very rarely.
Table 2.2 – Reactions to transfusions
Hemolysis Antibody in the recipient (acute or delayed)

Alteration of the transfused cells:
– Excessive heat/freezing
– Bacterial contamination
– Mixing with intravenous fluids/drugs
– Trauma by infusion devices
Antibody in a plasma transfusion
Febrile non-hemolytic Red cell reactions: from recipient white cell antibodies
Platelet reactions: from donor white cell cytokines
TRALI: Donor HLA antibodies
Allergy Urticaria/anaphylaxis
Transfusion-associated graft Cytotoxic donor lymphocytes

versus host disease (TA-GVHD)
Post-transfusion purpura Recipient-derived anti-HPA Abs
Infection Bacterial:
– Environmental: Staphylococci / Pseudomonas
– Syphilis (Treponema)
– Lyme disease (Borrelia)
– Yersinnia/Salmonella
– Q fever (Coxiella)
Viral
– Hepatitis A, B, C or D
– HIV
– HTLV
– EBV, CMV, HHV
– Parvovirus
Protozoal
– Malaria
– Chagas’ disease
– Toxoplasma
– Babesia
– Leishmania
Prion?
CMV, Cytomegalovirus; EBV, Epstein-Barr virus; HHV, Human herpes virus; HLA, Human leukocyte antigen; HPA,
Human platelet antigen; HTLV, Human T-cell leukemia/lymphoma virus; TRALI, Transfusion-related acute lung injury.
22
Hemolytic reactions
These result from the binding of an antibody to a red cell antigen. This leads to
opsonization, with or without complement activation, red cell destruction and
the activation of inflammation. In acute reactions this occurs within 24 hours
whereas, in delayed reactions, this usually occurs within 5–7 days. Acute hemoly-
sis is often mediated by IgM, which efficiently activates complement leading to
intravascular destruction, with release of hemoglobin into the circulation and
urine. In the delayed reaction, IgG is often the mediator. As a result there may be
no, or incomplete, complement activation. In such cases, red cell breakdown
usually occurs in the spleen, leading to a fall in blood hemoglobin levels rather
than frank intravascular hemolysis. Cytokines may be released as a consequence
of both complement and macrophage/monocyte activation.
Antibodies against ABO are the most common cause of acute hemolytic reac-
tions. This most often results from clerical error leading to the misidentification
of the unit of red cells for donation, or misidentification of the recipient. Anti-
bodies against A, B, H (O), K (one of the Kell group), Jk a (one of the Kidd group)
and Le a (one of the Lewis antigens) are often associated with intravascular hemol-
ysis. Antibodies against the Rh, Duffy or M, N, or S blood group antigens are
most often associated with extravascular hemolysis.
In the acute reaction, even a small volume of red cells may be sufficient to
trigger hemolysis. The first signs are often patient apprehension, agitation, fever
and diffuse abdominal pain. This may lead to breathlessness, shock, disseminated
intravascular coagulation and acute renal failure. Immediately upon suspicion of
a reaction, the infusion of blood should be stopped. If possible, the unit of blood
should be taken down with the administration set still attached, which allows
examination of the unit for evidence of bacteriological contamination. Manage-
ment should follow the principles shown in Table 2.3.
Delayed hemolytic reactions are usually the result of an anamnestic response
occurring upon re-exposure to an antigen. Such re-exposure leads to an increase in
the level of antibody in the recipient which can result in red cell destruction. This
usually leads to asymptomatic coating of the transfused cells with antibody (a positive
direct antiglobulin test), but can result in clinically evident hemolysis with fever, jaun-
dice, a falling hemoglobin level and spherocytosis. This is usually clinically evident
some 5–7 days after the transfusion. Although multiple antibodies can be involved,
Rh group, Kell, Duffy and Kidd antibodies are most often implicated. Repeat testing
some days after the transfusion may be required to detect the antibody. Severe reac-
tions are uncommon and specific treatment, other than the provision of suitable
antigen-negative transfusion for continuing anemia, is seldom required.
Non-hemolytic febrile reactions

Non-hemolytic febrile reactions occurring with red cell transfusions are not
uncommon. They are caused by pre-formed anti human leukocyte antigen (anti-
23
Table 2.3 – Management of transfusion reactions
Severe Stop infusion

hemolysis? New venous access
Re-examine component for evidence of infection/hemolysis
Immediate Monitor Temperature

Blood pressure
Pulse
Urine output
Intravenous fluids Maintain blood pressure + urine output
Confirm identity Wrist band
Component label
Compatibility form
Alert transfusion lab Do not use other cross-matched units
until advised by transfusion lab
Repeat blood group and antibody screen
Lab to repeat pre-screening
Further tests Full blood count Anemia

Fragmentation
Spherocytes
Urinalysis Hemoglobinuria
Urea and electrolytes Renal function
Coagulation screening Disseminated intravascular coagulation
(DIC)
Blood cultures Sepsis
Evidence of Hemolysis Reticulocytes
Haptoglobin
Plasma hemoglobin
Other measures Dialysis

Disseminated intravascular
coagulation (DIC) management (if required)
Fever only Paracetamol

Monitor vital signs
Transfusion lab Advice on restarting
Urticarial Chlorpheniramine
HLA), or granulocyte-specific, antibodies in the recipient of the transfusion,

which react with white cells in the transfused component. Fever and tachycardia
can commence within 30 minutes of the beginning of the transfusion, which
usually settles within a few hours of cessation. Symptoms may be more severe
when transfusion is rapid and the incidence is much reduced by pre-transfusion
24
leukodepletion of the blood component. The principal importance of these reac-

tions is their correct distinction from a more severe hemolytic transfusion reac-
tion, or from bacterial contamination of the blood component. Accordingly,
such reactions will often necessitate cessation of the transfusion. Febrile non-
hemolytic reactions most often occur during platelet transfusions, in perhaps as
many as 40%, when they are likely to be mediated by leakage of inflammatory
cytokines from white cells in the platelet component. The universal leukodeple-
tion of blood components in the UK has resulted in a four-fold reduction in the
occurrence of such reactions.
Allergy
Mild allergic reactions, such as urticaria and wheeze, are not uncommon with the
transfusion of blood components, and may occur after single or multiple expo-
sure. Although often difficult to disentangle from a reaction to concomitant ther-
apies, their immediate management is akin to that of other allergic reactions. Of
note, is the rare occurrence of severe IgA deficiency (IgA <0.05mg/dl) in the
recipient. This may trigger a reaction on exposure to donor IgA. The detection
of specific anti-IgA antibodies in the recipient may assist in making the diagnosis,
and such individuals may require washed red cells (which removes the IgA) or
consideration for blood-sparing modalities such as cell salvage or autologous
donation.
Transfusion-related acute lung injury

Transfusion-related acute lung injury (TRALI) is a relatively rare (around 0.02%
of transfusion events), but potentially serious, consequence of transfusion. It is
caused by the transfusion of HLA antibodies (within a blood component), which
react with the recipient’s white cells. The donors are often multiparous women
who have formed these antibodies during previous pregnancies and any transfu-
sion component containing plasma can be the source. The antibodies react with
the patient’s white cells resulting in leukostasis and inflammation within the
lung. The patient classically presents a few hours after transfusion with respira-
tory distress, fever, hypoxia and evidence of bilateral infiltration on the chest X-
ray. It may be more common in subjects with sepsis and can be difficult to
distinguish from other causes of adult respiratory distress. In 50% of cases, HLA
or specific granulocyte antibodies can be detected in the implicated donor
plasma. Although mortality may be as high as 1 in 20, with adequate and prompt
ventilatory support, resolution occurs in the majority within 7 days of onset. In
many countries, strategies to reduce TRALI – plasma reduction of products,
screening of female donors for selected antibodies, or exclusion of parous
women from plasma-rich component donation – are being considered to reduce
the risk. The deferral of blood donors who have received a blood transfusion
could also be expected to reduce TRALI risk.
25
TA-GVHD
TA-GVHD is a very rare, but almost universally fatal, condition caused by the
infusion (in a transfusion component) of immune competent lymphocytes.
These cells are able to mount an immunological reaction against the recipient.
This most commonly occurs in those recipients who share an HLA haplotype
with the donor. However, immunodeficiency or immunosuppression in the
recipient also predispose to the condition. Patients often present with a rash,
diarrhea and abnormal liver function. This progresses to marrow failure and
sepsis within 4 weeks of diagnosis. Although there is no specific treatment for the
condition, it can be effectively prevented by gamma irradiation of cellular trans-
fusion components prior to transfusion.
PTP
Around 2% of Caucasians do not carry the platelet antigen HPA-1a on their
platelets and PTP usually occurs in parous women who do not carry this antigen
(see Chapter 8). It is assumed in such cases that the mother has carried an HPA-
1a positive fetus in a previous pregnancy that has resulted in immunization of the
mother, but has not led to the development of neonatal alloimmune thrombocy-
topenia (see Chapter 8). Subsequent exposure to the HPA-1a antigen, via a red
cell transfusion, results in an increase in the titer of anti-HPA-1a antibodies. This
leads, by an essentially unknown mechanism, to destruction of the women’s
platelets (despite the fact that they do not carry the HPA-1a antigen). This often
results in a rapid fall in platelet count presenting around 1 week after the trans-
fusion. The resultant thrombocytopenia results in widespread bruising and
bleeding from the gastrointestinal or renal tracts.
The condition is diagnosed by the history of recent transfusion and the detec-
tion of anti-HPA-1a antibodies in the recipient’s serum and exclusion of other
causes of thrombocytopenia (see Chapter 8).
Other HPA antibodies have also been reported in association with PTP,
including anti-HPA-1b, -3a, -3b, -4a or -5b. As further donor platelets (even if
antigen negative) will also be destroyed, platelet therapy may be ineffective.
Indeed, there is no evidence that the specific HPA antigen-negative platelets will
be more effective than random platelets. So therapy is geared towards suppres-
sion of the immune response with high-dose IVIgG therapy. A number of
successes have also been reported with plasma exchange. The provision of
blood component therapy after an episode can also be problematic, as a recur-
rence is unpredictable. It is usual practice to attempt to provide appropriate
HPA negative blood for future transfusions, although there may also be a role
for leukodepletion or autologous donation.
26
Bacterial contamination
Bacterial contamination may be present in as many as 1 in 3000 cellular blood
component transfusions and, when severe, presents as an acute hemolytic trans-
fusion reaction. The true incidence is difficult to judge, as minor contamination
may not lead to clinical problems. Around 60% of fatalities are associated with
microbial contamination of platelets.
The primary source of bacterial contamination of blood products is the
donor. This can be due to poor cleaning technique at the venesection, or asymp-
tomatic bacteremia in the donor (such as occurs after dental treatment). On
account of this, bacterial contamination is also a problem with pre-deposit auto-
logous donation. Problems can be reduced by: taking an adequate donor history;
meticulous venesection technique; specific systems to detect microorganisms in
blood components; and inspection of the blood component prior to transfusion.
As red cell components are stored at 4°C, Gram-negative organisms such as
Yersinnia and Pseudomonas are often implicated. By contrast, as platelets are
stored at room temperature, Gram-positive organisms such as staphlyococci and
streptococci are often the cause of such contamination.
New variant Creutzfeldt-Jakob disease (nvCJD)

nvCJD is one of a group of neurodegenerative disorders caused by infection with a
prion protein. The mechanism of the disease is not fully understood, but evidence
suggests that the infectious agent is an abnormal, protease-resistant form (desig-
nated PrPSc, where ‘Sc’ stands for ‘scrapie’) of the normal host prion glycoprotein
(designated PrPc), with a strong likelihood that the disease represents a human
form of bovine spongiform encephalopathy. Prion diseases are characterized by
deposition of abnormal PrPSc prion protein aggregates in the brain, leading to
spongiform degeneration and neuronal loss. The abnormal protein has the same
amino acid sequence as its physiological counterpart, but has a different sec-
ondary (folded) structure. It is this altered structure that allows the formation of
the protease-resistant aggregates. What is remarkable is that the abnormal protein
(PrPSc) interacts with the native protein (PrPc), leading it also to adopt this abnor-
mal secondary conformation and further the aggregate deposition.
nvCJD is characterized by a relatively rapid onset of neurological degenera-
tion and dementia. Although the prion’s full distribution is not known, there is
evidence that cellular interactions are essential in concentrating PrPc to allow its
transformation by PrPSc. PrPc is highly expressed on a variety of blood cells, lym-
phoid cells and cellular microparticles, where it may have a physiological role in
the protection from oxidative stress. As a consequence of this potential cellular
dependency, blood donations in the UK are routinely leukodepleted prior to
transfusion.
The scale of the risk to humans by nvCJD is unknown and whether there are a
substantial number of individuals incubating and able to transmit the disorder is
27
also not yet known. Although animals have a different tissue distribution of prion
proteins to humans, transmission of prion disease by transfusion has been
demonstrated in sheep. Furthermore, in 2004, a person who had donated blood
in the UK subsequently developed nvCJD. The recipient of his donation also
went on to develop nvCJD some time after the transfusion. Although a link
cannot be proven, there is the potential that direct introduction of the abnormal
prion by transfusion may bypass the protective effects of enzyme degradation that
take place if the prion is ingested in infected meat. It is likely, however, that the
absolute risk of prion transmission by transfusion is low. There is also likely to be
a reduction in blood product contamination with prions if the products have
undergone a fractionation process, such as is used in the manufacture of
immunoglobulin or clotting factor concentrates.
However, this theoretical risk of prion transmission by transfusion, allied to
the absence of a screening test (although many first generation tests are in devel-
opment) and evidence that the protein is not susceptible to standard viral inacti-
vation procedures, highlights the need to restrict blood product usage to those
indications where there is a clear, positive, benefit:risk ratio.
Cell saving
Cell saving encompasses a number of techniques which attempt to reduce expo-
sure to allogenic blood. These techniques include autologous pre-deposit, intra-
operative hemodilution and intra- or postoperative cell salvage. A number of
other therapies, such as recombinant erythropoietin and antifibrinolytic agents,
are also used to reduce allogenic exposure, and a further strategy is the develop-
ment of useful blood substitutes.
In view of the potential risks of allogenic transfusion there has been consider-
able interest in the use of cell-saving techniques in relation to delivery and, in
particular, Cesarean section. Although techniques such as pre-deposit of autolo-
gous blood and intraoperative cell salvage are well tolerated and considered safe
in the majority of instances, their use in pregnancy requires due consideration of
their potential effects on maternal and fetal wellbeing. In addition, the wide-
spread adoption of these techniques requires clarification of those in whom such
interventions would be clinically or cost-effective.
Autologous pre-deposit
There are many case series examining the feasibility of autologous pre-deposit in
relation to obstetrics. Most often pre-deposits (equivalent to 1–2 units of red
cells) are taken by standard venesection in the third trimester. Although mater-
nal side effects, such as faintness, are well tolerated, the effects of blood loss on
the feto-placental unit are not fully understood. Although it may be that there is
no effect on cord blood pressure and flow, it has been suggested that venesection
of 400ml of maternal blood in the third trimester is associated with a reduction
28
in fetal middle cerebral artery pulsatility in the first 24 hours after the event. In
general, to be suitable, women require a hematocrit sufficient (usually >33%) to
permit safe venesection. In addition, there has to be a high probability that col-
lected units will be used. Given this, autologous pre-deposit can be justified in
those with multiple or difficult red cell alloantibodies (e.g. thalassemia), those in
whom bleeding is likely (e.g. where there is a severe bleeding disorder or pla-
centa previa) and those subjects who cannot receive transfusions (e.g. Jehovah’s
Witnesses, Bombay phenotypes). In most obstetric patients without placenta
previa, the need for peripartum transfusion cannot be predicted with sufficient
accuracy to justify antepartum pre-deposit, and in studies of feasibility, only
14–34% of subjects with placenta previa were suitable for pre-deposit. Indeed, as
with many cell-saving techniques, the use of pre-deposit leads to an overall
increase in the likelihood of transfusion. This increase in the likelihood of use of
the autologous pre-deposit means that this modality of therapy actually carries a
similar risk of both bacterial contamination and of clerical error to that found in
allogenic transfusion.
Intraoperative cell salvage

Although intraoperative salvage has been used extensively in vascular, cardiac
and orthopedic surgery, its use in Cesarean section has been limited because of
the possibility of re-infusion of blood containing fetal red cells and amniotic fluid
components (such fetal cell debris, potassium and tissue factor). It has been sup-
posed that this risk will be minimized by commencement of salvage after removal
of the feto-placental unit, by washing and leukodepletion of the salvaged cells,
and by attention to the potential for maternal RhD immunization. Although
these measures should reduce the risk, and although salvage is likely to reduce
the risk of allogenic exposure, there have been no studies of adequate power to
determine if salvage increases the thrombotic and infection risk to the mother.
Given these potential problems, the lack of predictability of obstetric hemor-
rhage and the reliability of allogenic transfusion in developed countries, salvage
should be considered in cases where allogenic transfusion is unacceptable (or
technically problematic), and only when there is a high likelihood of transfusion.
Erythropoietin
Erythropoietin (Epo) is a glycoprotein, which is synthesized in the kidney in
response to hypoxia. Recombinant human Epo is derived from Chinese hamster
ovary cells, and has been used widely in anemia associated with renal failure,
malignancy and chemotherapy. It is given subcutaneously, at a dose of
50–200IU/kg two–three times per week along with oral or parenteral iron, with
the aim of increasing the hemoglobin by 1 g/dl/month. It has been used in
Jehovah’s Witnesses and in the management of renal failure in pregnant subjects,
and appears to be well tolerated. Although it does not cross the placenta, it
29
carries a risk of hypertension, and excessive elevation of hemoglobin may also be

associated with thrombosis. At present, Epo has a place in the management of
neonatal anemia of prematurity, but its use in maternal anemia, or as an adjunct
to autologous or salvage techniques in pregnancy, requires further clinical- and
cost-effectiveness evaluation.
Blood substitutes
There is a demand to produce red cell and platelet substitutes to address a com-
bination of concerns. These include shortages in blood supply, the risks of conta-
mination and immunomodulation, the costs associated with blood collection, the
cost of temperature-regulated storage, the need for cross-match procedures, and
the need for strict patient identification. There has been a substantial advance in
the development of solutions able to achieve volume expansion and, although a
number of blood substitutes able to carry oxygen are in development, the major-
ity are not yet in clinical use. Such products are required to take up oxygen in
the lungs and give it up in the tissues. They are also required to be non-toxic, to
have a useful half-life in the body and a useful shelf life for storage.
There has been much attention on hemoglobin-based solutions. Unfortu-
nately, hemoglobin readily participates in redox reactions leading to oxidative
damage. As a consequence, free hemoglobin has a short half-life and is a potent
vasoconstrictor. A number of attempts to improve the stability of hemoglobin by
cross-linking have been attempted. Of these, bovine polymerized hemoglobin
has a licence for use in veterinary practice. The different hemoglobin products in
current development, however, have differing oxygen affinities, hemoglobin con-
centrations and stability. The major side effect of these products is hypertension,
which results from vasoconstriction (by an as yet poorly defined mechanism,
perhaps related to a local reduction in nitric oxide or alteration in viscosity),
which may in fact hamper oxygen supply.
One other group of products are the perfluorocarbon emulsions, which take
up oxygen to a greater degree than human plasma, but do not have the flexibility
of uptake and supply associated with hemoglobin. Perfluorocarbon emulsions
may also modulate the inflammatory response, and their use can be associated
with the development of flu-like symptoms.
At present, none of the available red cell substitutes offer the immediate
prospect of a useful oxygen supply without the risks and costs associated with
current transfusion therapies, and there is no specific evidence of their safety in
pregnancy.
Hemolytic disease of the newborn

The Rh blood group antigens on red cells result from the action of two related
genes which code for RhD and RhC/RhE. These blood groups are inherited as
two haplotypes, consisting of a combination of c or C, D or no D, e or E. Of
30
these the most important Rh type in obstetric practice is RhD, which is around
50 times more immunogenic than other Rh types. Fifteen per cent of Cau-
casians, 5% of Black subjects and 2% of Asians are RhD negative. When an
RhD-negative mother is carrying an RhD-positive child, the transplacental
passage of both fetal blood and maternal IgG immunoglobulin may result in
the development of an anti-RhD antibody in the mother that is also active in
the fetus.
Whether the mother develops an antibody depends upon the amount of fetal
blood entering the maternal circulation and the presence of any feto-maternal
ABO mismatch, which reduces the development of an anti-RhD antibody (pre-
sumably by clearance of fetal cells by pre-formed maternal anti-A or anti-B anti-
bodies). The Rh haplotype carried by the fetus may also be important as the fetal
carriage of cDE/cde is more likely to induce antibodies than other RhD-positive
haplotypes. If an RhD-negative woman carries an ABO-compatible, RhD-positive
child there is a one in six chance of forming anti-RhD in the absence of prophy-
laxis. If an RhD-negative recipient receives an RhD-positive transfusion there is a
90% chance of forming anti-RhD. The majority of hemolytic disease cases occur
in the second pregnancy, but in some instances the fetal cells may result in a
brisk maternal response leading to significant fetal hemolysis in the first preg-
nancy.
The amount of RhD expressed on the red cell is dependent upon which
other members of the Rh haplotype are present. There are also many other
quantitative and qualitative variations in RhD expression. Where there is a
reduction in the expression of all RhD epitopes, this is termed ‘weak D’. The
definition of weak RhD is, however, dependent upon the sensitivity of the local
laboratory reagents used to detect RhD. On the other hand, the absence of
some of the RhD epitopes is termed ‘variant D’, and there are many classes of
variant D. As someone with a weak D has all the RhD epitopes present, exposure
to RhD-positive blood will not result in such an individual developing anti-RhD,
and such individuals should be classed as RhD-positive. The importance of
detecting weak RhD in blood donors is a matter of debate, but such weak D cells
could be capable of eliciting an antibody response in a recipient. On the other
hand, if someone has a variant D, they may be capable of developing anti-RhD
on exposure to RhD-positive cells (from a fetus or from transfusion). In particu-
lar, one type of variant RhD (called RhDVI) may, upon exposure to RhD, lead to
the development of both anti-RhD and HDN. On account of this, in obstetric
practice, the laboratory reagents used for antenatal typing should not detect the
RhDVI variant, and such mothers should be classed as RhD negative. As such,
they should receive RhD-negative blood products and prophylactic RhD
immunoglobulin if potential sensitizing events occur (Table 2.4). Whether a
transfusion of DVI donor blood carries a significant risk of sensitizing a recipient
is not clear. In any case, when a laboratory cannot reliably distinguish RhD vari-
ants from a weak RhD phenotype, it is safer for the mother to be considered to
be RhD-negative.
31
Table 2.4 – Potential RhD-sensitizing events
Antepartum hemorrhage (Including threatened abortions)

Abdominal trauma
Ectopic pregnancy
Fetal external version
Delivery
Invasive investigations Amniocentesis

Chorionic villous sampling
Fetal blood sampling
Embryo reduction
Shunt insertion
Fetal loss Intrauterine death

Stillbirth
Miscarriage with evacuation
Complete or incomplete miscarriage >12 weeks
Therapeutic termination
There are three general classes of IgG immunoglobulin, each with differing
characteristics. Of these, HDN only occurs with IgG of the subclasses IgG1 or IgG3,
which are capable of transplacental passage. The transplacental passage of IgG1 may
occur as early as 20 weeks gestation, with the peak of IgG3 evident at around 28
weeks. Despite this, IgG3 is associated with greater postnatal jaundice than IgG1. As
IgG1 rarely binds complement, the mechanism of red cell destruction is not clear,
but may involve increased presentation of the IgG Fc component to macrophages.
The majority of affected fetal red cells are destroyed in the fetal spleen.
Antenatal diagnosis and monitoring

Blood group screening of all women should take place when booking for antena-
tal care. To determine whether the RhD group is negative, a technique that is
sensitive for the presence of weak RhD phenotypes should be employed: circulat-
ing maternal anti-RhD is usually detected by an indirect antiglobulin test. Such
testing should be repeated routinely at least once between 28 and 32 weeks gesta-
tion. In the UK, an estimation between 34 and 36 weeks gestation is also recom-
mended. Where there is a routine antenatal prophylaxis program in place, the
testing at 28–32 weeks should be carried out prior to any routine prophylactic
dose of anti-RhD being given.
If there has been an antenatal sensitizing event in an RhD-negative woman
whether or not she has circulating anti-RhD should be determined, and an esti-
mation of feto-maternal hemorrhage (FMH) carried out (Table 2.5). The pres-
ence of any maternal anti-RhD should be determined again at delivery, and a
32
Table 2.5 – Feto-maternal hemorrhage (FMH) assessment
At delivery around 1 in 200 mothers have an FMH >4ml

< 20 weeks, FMH assessment not required
Assess FMH if mother RhD negative/fetus unknown
Pre-formed immune anti-RhD present, FMH not needed
If not sure if maternal anti-RhD is passive or immune, give standard dose anti-RhD
Maternal blood pattern Action

Fetal cell +, free RhD – Assess FMH and give anti-RhD
Repeat FMH at 48 hours
Fetal cell +, free RhD + Repeat FMH at 48 hours
Fetal cell –, free RhD + Do nothing
Fetal cell –, free RhD – Give standard anti-RhD
Repeat anti-RhD test at 48 hours
cord blood sample should be obtained to determine the ABO and RhD type of
the baby. At delivery, the presence of anti-RhD on the fetal cells should be exam-
ined by direct antiglobulin test, and the mother should have a blood sample
analyzed for an estimation of FMH.
There is a correlation between the gestation at which antibody is first detected
and the severity of any resultant HDN. Therefore, if antibodies are detected at
booking, they are likely to cause more problems than those detected for the first
time later in pregnancy. If an antibody associated with HDN is detected this
should be quantified. For most antibodies this will be achieved by the titer on an
indirect antiglobulin test. It has been recognized for some time that automated
quantification of anti-RhD antibody levels provides more reproducible and
meaningful values than manual titration, with the level of maternal anti-RhD
immunoglobulin used to plan further testing and the need for fetal blood sam-
pling or delivery. Indeed, in the case of RhD, it is recommended that the anti-
RhD titer should be reported in international units (IU). In general, the absolute
value is not as important as any rise in titer with increasing gestation. However,
when the levels remains at <4 IU/ml, HDN is unlikely and few fetuses suffer
major anemia. In such cases, serial anti-RhD monitoring and ultrasound determi-
nation of a combination of parameters (such as Doppler of umbilical vein or
middle cerebral artery flow, liver and spleen size, and the early detection of
ascites) may assist in management.
When the levels are 4–15IU/ml, there is a moderate risk of HDN and investi-
gation is intituted in some, but not all, areas of the UK. Early in pregnancy, geno-
typing techniques can be used to genotype the fetus (when the partner is
unknown, or it is an RhD heterozygote). Such polymerase chain reaction (PCR)
techniques do require due consideration of the ethnic origins of the partners,
and both false-positive and false-negative results can occur.
33
In addition, it is now possible to determine if the fetus is RhD positive (as well
as the Kell and c status: see below) from free fetal DNA obtained from the mater-
nal circulation. Although this is in its infancy, very high sensitivity has been
reported in some studies using fluorescence-based PCR techniques.
After 27 weeks gestation, amniocentesis can be used to determine the level of
bilirubin in the amniotic fluid by its absorption at 450 nm. The level obtained is
interpreted by a Lilley graph – essentially relating the level to a gestation-specific
reference range to assist in determining the necessity for intrauterine transfusion
or delivery. This technique reflects the degree of fetal hemolysis but not the fetal
hemopoietic response, and significant discrepancies between the fetal hemat-
ocrit and the amniotic fluid optical density can occur. Analysis of amniotic fluid
anti-D concentrations has also been investigated. Although these correlate well
with maternal serum anti-D levels, they have no added value in predicting the
severity of fetal hemolysis.
As yet, in such cases, fetal assessment by most non-invasive methods, including
fetal movement counting, ultrasound appearance and fetal electocardiogram
waveforms, have not proved helpful in determining whether the fetus will
become anemic. Abnormal fetal heart rate patterns have been described in
severe disease, and although cardiotocography (CTG) will identify the seriously
anemic fetus it is not sensitive enough to predict mild to moderate anemia. Com-
puterized fetal heart rate analysis has identified a consistent relationship between
baseline variability and fetal hematocrit, and warrants further study.
Weekly velocimetry of the fetal middle cerebral artery has also been used to
monitor affected fetuses. In such cases, peak systolic velocities >1.5 multiples of
the median for the specific gestation have been shown to be predictive of moder-
ate or severe fetal anemia (with 100% sensitivity and a false-positive rate of
around 12%), and ultrasound studies correlate well with the bilirubin level in the
amniotic fluid. This technique has begun to replace serial amniocentesis and is
now being used to predict when fetal blood sampling is required.
The only direct method of assessing the severity of RhD disease, however, is by
measuring the hematocrit of fetal blood samples. When levels exceed 15IU/ml
or if indicated by velocimetry, fetal blood sampling at >18 weeks gestation may be
required. This is not without complications (from <2% mortality in non-affected
fetuses to 5–15% fetal mortality in affected cases), and the procedure can result
in a marked increase in the maternal antibody levels. Other reported complica-
tions include fetal exsanguination, bradycardia, failure to obtain a sample and
cord constriction. Fetal blood sampling can be achieved by ultrasound-guided
sampling of the cord vessels, the hepatic vein, or the fetal heart, which will also
permit an assessment of the fetal RhD group. Fetal blood sampling should only
be carried out, however, if there is the expertise and facilities to carry out an
intrauterine transfusion, if required.
In summary, when the maternal anti-RhD level is <0.5IU/ml, four-weekly
monitoring of anti-RhD levels is all that is required. If anti-RhD levels are
>0.5IU/ml, then two-weekly anti-RhD assessment is required. If levels are
34
between 10 and 20IU/ml, fetal blood sampling will usually be required, but the
fetus does not often require an intrauterine transfusion. If the maternal anti-RhD
level is >20IU/ml, fetal blood sampling will be required and the fetus will prob-
ably require intrauterine transfusion. In general, if the anti-RhD level is >4IU/ml
but <10IU/ml, then early delivery may be required (however, usually at >36
weeks). For levels of >10IU/ml, 33–38 week delivery may be necessary.
Fetal presentation of HDN

The fetal presentation may be simply with a mild anemia, or in severe cases with a
severe anemia, marked jaundice, edema, cardiac failure, widespread effusions and
pulmonary hemorrhage. Neurological deficits and seizures (kernicterus) result
from passage of bilirubin across the fetal blood–brain barrier. The anemia results
in extramedullary hemopoiesis in the liver and spleen. This is associated with a
reduction in the production of plasma proteins, which further contributes to the
edema. The end result may be stillbirth, neonatal death, or complete resolution.
At birth, a low umbilical venous hemoglobin is seen, although this does not
directly equate with disease severity. There is also likely to be a marked reticulo-
cytosis, unless this has been suppressed by recent intrauterine transfusion. The
presence of a high concentration of nucleated red cells may result in a factitious
high white cell count on automated blood count analyzers.
Fetal management
If fetal maturity permits, delivery of the fetus is indicated when there is evidence
of a high level of bilirubin in the amniotic fluid. When fetal immaturity does not
make delivery feasible, fetal transfusion should be carried out. The introduction
of intrauterine fetal transfusion has reduced the perinatal mortality rate in severe
cases of HDN from 95 to 50%. Intrauterine transfusion is indicated if the fetal
hematocrit is <25% between 18 and 26 weeks gestation or <30% after 34 weeks
gestation. The blood used should be cross-matched against maternal serum and
should have a hematocrit of 75–90%, be sero-negative for cytomegalovirus and
gamma irradiated. The aim of the transfusion is to increase the fetal hematocrit
to around 45%, and the volume required is calculated from the pre-transfusion
hematocrit, the donor hematocrit and the estimated fetal blood volume. In some
units, pre-medication is given to the mother for maternal and fetal sedation,
whilst others have employed tocolytic agents, antibiotics and corticosteroids. The
transfusion may be given directly into the fetal vasculature (umbilical vein, umbil-
ical artery, or inferior vena cava), into the fetal heart, or into the fetal peritoneal
cavity. Further transfusions may be required at 1–3-weekly intervals. The compli-
cations of intrauterine transfusion include transfusion site trauma and
hematoma formation, spasm of the cord, umbilical arterial thrombosis, fetal
thrombocytopenia, an increase in maternal antibody levels, fetal leukoen-
cephalopathy and fetal bradycardia.
35
Other modalities of therapy have been used when there has been a history of
previous fetal loss, or where the likely father is homozygous for RhD. These
include repeated, maternal plasma exchange from early in pregnancy to remove
anti-RhD from the maternal circulation until fetal transfusion therapy can be
achieved. In such circumstances, maternal procedures, which could cause allo-
immunization, should be avoided. Similarly, regular maternal treatment with
high-dose human normal immunoglobulin (2g/kg on a two-weekly basis) has
also been used in this circumstance.
After delivery, neonatal treatment consists of phototherapy to oxidize biliru-
bin and, when required, exchange or top-up transfusion.
Prevention
In the 1960s around 46 deaths/100 000 live births were attributed to the RhD
immunization in the UK. The routine use of post-delivery anti-RhD immunoglob-
ulin immunoprophylaxis was introduced in 1969. By 1990, the number of fetal
deaths attributed to RhD immunization in the UK had fallen to 1.6/100 000.
Despite this, there continue to be cases of HDN. This may be due to maternal
sensitization in a previous pregnancy, previous transfusion of red cells or
platelets, inadequate treatment of potential sensitizing events, sensitization from
unrecognized events, or reactions to red cell antigens other than RhD.
Prevention is based upon the suppression of the maternal immune response
to the RhD antigen by the timely administration of a passive antibody, the
mechanism of which remains poorly understood. Intramuscular anti-RhD should
be administered into the deltoid muscle to all RhD-negative women as soon as is
practical, but certainly within 72 hours of delivery. If it is not given before 72
hours, every effort should be made to administer it within 9–10 (or even 28)
days, as this may still offer some protection. The dose should be determined by
the level of FMH. Given intramuscularly, 125IU of anti-RhD is considered suffi-
cient to protect against exposure to 1ml of RhD-positive red cells. In the UK,
250IU is routinely given for any potential sensitizing event that occurs before 20
weeks gestation and 500IU for any event after 20 weeks gestation (see Table 2.4).
If a further sensitizing event occurs >7 days after a prophylactic dose of anti-RhD,
it is common practice to give a further dose of anti-RhD. In addition, after 20
weeks gestation an FMH assessment should be carried out to determine if the
bleed exceeds 4 ml, when a further appropriate dose of anti-RhD may be
required (see Table 2.5). It is routine practice in the UK not to give anti-RhD to a
threatened miscarriage where the bleeding has stopped before 12 weeks gesta-
tion and there is still a continuing pregnancy.
As noted above, some at-risk women may become sensitized in the last trimester
by unrecognized events. In fact, given the success of the prophylaxis for sensitizing
events, this has become the most important cause of maternal sensitization in many
developed countries. One strategy designed to cope with these ‘silent’ FMH is the
routine administration of anti-RhD immunoglobulin to all RhD-negative women
36
who have no detectable anti-RhD antibodies in their blood. This has been shown to
further reduce the risk of antenatal immunization. At present, however, a variety of
dosages and dosage schedules for prophylaxis have been employed. Of these, two
options, which both seem effective, include a dose of 500IU at 28 and 34 weeks ges-
tation or a single larger dose (1500IU) early in the third trimester. Enthusiasm for
this method of HDN prevention, however, may be tempered by problems of anti-
RhD availability and, in Europe, the continuing concern about the possibility of
transmission of prions by blood products, although, as noted above, the process of
fractionation is thought to reduce this risk.
Non-RhD antibodies
In addition to RhD, at least 40 red cell antigens have been associated with HDN.
These include Rhc, RhC, RhE, Kell, Duffy, MNS, Lutheran, Kidd and U. After
anti-RhD, antibodies against c, Kell (K1) or E are the most frequently encoun-
tered antibodies requiring treatment.
Kell
The gene coding for Kell antigens is found on chromosome 7, although expres-
sion is also dependent upon genes on the X chromosome. Of the 22 Kell anti-
gens, antibodies to Kell (K1), k (cellano), Kpa (Penny), Kpb (Rautenberg), Jsa
(Sutter) and Jsb (Matthews) have all been implicated in HDN. Antibodies to K1
are the most commonly encountered antibodies after ABO or RhD, and are
usually IgG and occasionally complement binding. The K1 antigen is found in
9% of Caucasians (who are virtually all heterozygous). Kpb and Jsb antigens have a
very high frequency in Caucasians, and Jsb is almost universal in Black subjects. In
addition, a reduction in all Kell antigen expression can occur. This is called the
Mcleod phenotype and it arises from inheritance of Kx and leads to a chronic
mild hemolytic state.
Between 8 and 18% of pregnancies with detectable maternal anti-K1 result in a
K1-positive fetus, and hydrops has been reported in around 30% of such cases. In
recent years this incidence may have been reduced by the use of intrauterine
transfusion. In anti-K1 HDN, fetal anemia results from a combination of hemolysis
in the fetal blood as well as destruction in the marrow. This marrow suppression
leads to a lower bilirubin level in the amniotic fluid than would be expected by
the degree of anemia. The management of a sensitized women requires a combi-
nation of ultrasound and fetal blood sampling. One potential strategy in anti-K1-
positive pregnancies is to genotype the father (when paternity can be assured). If
the father is kk, no further action may be appropriate. If the father is K1k, then
anti-K1 titers <4 can be monitored monthly. If a level of 8 is reached, amniocente-
sis and fetal K1 typing may be appropriate, as 50% would be expected to be kk. If
the fetus is K1 positive, serial middle cerebral artery Doppler and cordocentesis
(with the potential for intrauterine transfusion) should be considered.
37
In HDN associated with antibodies to k (cellano), there may also be an associ-

ated reduction in effective red cell production in the marrow, as well as hemoly-
sis within the peripheral blood.
MNS and U
The MNS system is coded on chromosome 4 by two linked genes: GYP A (M and
N) and GYP B (S, s and U alleles). Anti-M is relatively common and can be IgG or
IgM. Anti M can react at 37°C and very occasionally has been associated with HDN.
There is insufficient data to give specific guidance, but the presence of IgG should
be determined, as IgM anti-M will not cross the placenta. High IgG titers should
lead to further investigation with amniocentesis, and possibly fetal and paternal
genotyping. Anti-N occurs rarely and is, most often, of no clinical significance.
Both anti-S and anti-s are usually IgG and have rarely been implicated in HTR
or HDN, with the majority of cases of HDN reported resulting in mild hemolysis
only. The U antigen is essentially only found in Black subjects, and anti U is very
rare and usually of immune origin. It can be associated, however, with HDN, and
geographically where this is a problem, provision of U-negative blood for fetal
transfusion can be problematic.
Duffy
The Duffy gene consists of co-dominant alleles including Fya and Fyb. The Duffy
antigen functions as a receptor for Plasmodium vivax and carriage of the Fy(a-b-)
phenotype has been associated with an increased risk of pre-eclampsia, perhaps
due to a reduction in Duffy-mediated clearance of interleukin-8.
Antibodies to Fya are more common than Fyb, are predominantly IgG and can
bind complement. Anti-Fyb is not associated with HDN; but HDN associated with
anti-Fya has been reported to carry significant morbidity (with one third of cases
requiring intrauterine transfusion) and fetal mortality (18% of cases), with
significant fetal disease possible at relatively low maternal antibody titers.
Kidd
The Kidd antigens are coded on chromosome 18 and consist of the co-dominant
Jka and Jkb. Antibodies to Jka are more common than Jkb. They usually bind com-
plement and there is often a high percentage of IgG3. Detection of the antibody
in maternal plasma can be problematic and often requires the use of homozy-
gous Jka or Jkb for detection (i.e. a cell that has homozygote expression of the
antigen is required for binding of the maternal antibody to occur). Detection of
antibodies may also be complicated by the rapid fall in titers which often occurs
soon after an immunization that has arisen as a consequence of transfusion. Anti-
Kidd antibodies have been occasionally associated with HDN, and fetal genotyp-
ing may assist in determining appropriate care.
38
Massive obstetric hemorrhage

There are a variety of definitions of major obstetric hemorrhage. These include a
requirement to replace 50% of the circulating blood volume (about 6–7l in later
pregnancy) in 3 hours, or blood loss requiring the replacement of the whole
blood volume (in practice indicated by transfusion of more than 10 units of red
cells). Practically, massive blood loss is likely when 1500 ml has been lost. This is
often associated with maternal hypotension and should be regarded as a sign of
impending massive loss.
It is to be expected that a normal vaginal delivery may involve blood loss of
<500 ml and Cesarean section blood loss of up to 1 litre. Although the predic-
tion of massive maternal blood loss is difficult, those with a past history of
abruption, postpartum hemorrhage (blood loss in the first 24 hours after deliv-
ery of at least 500 ml and more importantly >1 litre), placenta previa, heritable
or acquired coagulation defects (e.g. carriage of hemophilia, von Willebrand’s
disease, those receiving anticoagulation or with pre-existing DIC due to hemoly-
sis, elevated liver enzymes, and low platelets (HELLP) syndrome or pre-eclamp-
sia), as well as those with complicated or prolonged deliveries, are at increased
risk (Table 2.6).
Table 2.6 – Risk of massive obstetric hemorrhage
Complicates 1 in 1000 deliveries

Maternal death: 1 in 100 000 in developed and 1 in 1000 in developing countries
Abruptio placenta: fetal mortality <30%
Placenta previa: fetal mortality <5%
Loss of 15 000ml, or >25% blood volume, indicates massive loss is likely
High risk PMH Abruptio placenta

PMH PPH
Hemophilia carriage
von Willebrand’s disease (vWD)
Heparin
DIC
HELLP/PET
Therapy Crystalloid/colloid
Red cells
Early obstetric intervention
Fluid warmer
Maintain the PCV/platelets >50 × 109/l
1.5 × blood volume: consider platelet therapy
DIC, Disseminated intravascular coagulation; HELLP, Hemolysis, elevated liver enzymes, and low platelets;
PCV, Packed cell volume; PET, Pre-eclampsia; PMH, Past medical history; PPH, Post partum hemorrhage
39
Massive obstetric hemorrhage remains a major cause of maternal morbidity

and mortality, although in recent years the numbers of deaths in the UK have
been declining. Such hemorrhage complicates 1 in 1000 deliveries in developed
countries and around 1 in 100 000 results in maternal death. Furthermore, in
developing countries, maternal death may complicate 1 in 1000 of such events.
In approximately half of all women presenting with bleeding from the genital
tract after 24 weeks gestation, a diagnosis of placental abruption or placenta
previa will be made; no firm diagnosis will be made in the other half, although
most may be due to incidental causes such as cervical ectropion. When the blood
loss is caused by placental abruption, the fetal mortality is substantial and may
complicate up to 30% of cases. When it is due to a placenta previa, the fetal
death rate is still considerable at around 5% of cases.
Placental abruption
Placental abruption, or retroplacental hemorrhage, is the premature separation
of a normally implanted placenta from the uterine wall, resulting in hemorrhage
before the delivery of the fetus. It occurs in around 1 in 80 deliveries. Although
maternal mortality is rare, maternal morbidity from hemorrhage, shock, dissemi-
nated intravascular coagulation and renal failure is not uncommon. The risk to
the fetus depends on the severity of the abruption, the gestation, the birthweight
and the amount of concealed hemorrhage. The perinatal mortality with placen-
tal abruption is high, and may be >300/1000.
The bleeding, however, may be in whole, or in part, concealed if the
hematoma does not reach the margin of the placenta so that the blood loss is
revealed. In view of this, the amount of revealed blood loss reflects poorly the
degree of hemorrhage, as an enlarged hematoma may be concealed within the
uterus. The hematoma may also result in bleeding into the amniotic cavity, with
subsequent blood-stained liquor being noted when the membranes rupture. Fur-
thermore, the bleeding may infiltrate the myometrium resulting in so-called Cou-
velaire uterus. This infiltration of blood into the myometrium is associated with
sustained uterine contraction, making the uterus feel ‘solid’ on examination,
provoking labor and reducing utero-placental blood flow with serious compro-
mise to the fetus. Placental abruption may occur at any stage of pregnancy. In
severe abruption there may be heavy vaginal bleeding, or increasing abdominal
circumference if the bleeding is concealed and retained within the uterus. The
woman is usually in severe pain and may be in hypovolemic shock. The uterus
may be tender and irritable, with palpable contractions or uterine hypertonus,
often resulting in labor. Fetal distress or intrauterine fetal death may be present.
In less severe cases, the diagnosis of placental abruption may not be obvious,
particularly if the hemorrhage is largely concealed, and such cases may be mis-
diagnosed as idiopathic preterm labor. Thus, abruption should be considered in
any patient presenting with unexplained vaginal bleeding or possible preterm
labor, particularly if fetal distress is present.
40
Placental abruption is a common cause of coagulation failure in pregnancy,

and the degree of coagulation disturbance tends to correlate with size of abrup-
tion and blood loss. Absence of clotting in the blood within the vagina may be
apparent, or there may be bleeding from venepuncture sites or hematuria. FMH
may occur and can result in severe fetal anemia or fetal death. If the patient is
RhD negative, a Kleihauer (FMH) assessment should be performed, to estimate
the volume of FMH in order to provide sufficient anti-RhD immunoglobulin.
Management of massive blood loss

Both clinical assessment and blood pressure measurements are unreliable in esti-
mating the amount of blood loss, and fluid replacement should be guided by
other means. Central venous pressure monitoring is essential in serious hemor-
rhage, although central catheter placement may require correction of coagulopa-
Table 2.7 – Management of massive obstetric hemorrhage
Summon staff
(obstetrician, midwife, anesthetist, porter)
Inform
(blood transfusion and hematology)
Access
(<2 peripheral venous cannulae, chorionic villus sampling (CVP) – if possible, arterial line advised)
Samples
(cross-match ≥6 units, coagulation screen, full blood count)
Plasma expanders
(hemacel, albumin)
Consider need for uncross-matched blood
(O RhD negative – pressure cuff, blood warmer – if possible)
Monitor
(blood pressure, pulse, electrocardiogram, blood gas)
Monitor
(hemoglobin, platelets, coagulation – including fibrinogen)
Replace
(coagulation factors with fresh frozen plasma (FFP)/cryoprecipitate)
Early intervention
(artery ligation, Cesarean hysterectomy)
Assess
(need for intensive therapy unit)
41
thy. The immediate concern in the hemorrhaging patient is inadequate circula-

tion and priority should be given to restoring the circulating volume. Large-bore
cannulae should be sited, and blood-warming devices and pressure infusion bags
should be readily available (Table 2.7).
Blood products are not required for the immediate restoration of the circulat-
ing volume, and administration of crystalloid or colloid solutions may be valu-
able. The crystalloid solutions that may be given include normal saline or
Hartmann’s solution, but non-salt crystalloids such as 5% dextrose are ineffec-
tive. Large doses of colloids, such as gelatin, remain in the circulation longer, an
advantage if blood or blood components are not readily available, but a disadvan-
tage if fluid overload and pulmonary edema develop. Dextrans are contraindi-
cated in obstetric hemorrhage since they interfere with platelet function in vivo
and blood cross-matching in vitro, and carry a risk of allergic reaction and
uterine hypertonus. In all cases, uncross-matched group O, RhD-negative blood
should be available within 10 minutes and guidelines suggest that specific cross-
matched blood should be available within 30 minutes. In practice group specific
blood should be available in 20 minutes, but cross-matched blood may require
40 minutes. The packed cell volume should be maintained at >30 to prevent
platelet dysfunction and rapid transfusion should be by means of a fluid warmer
(although transfusion should not be delayed if this is not available). Co-infusion
of other solutions in the same intravenous access must be avoided, as this will
result in red cell lysis. Dilutional thrombocytopenia is likely when 1.5 times the
maternal blood volume has been transfused. Where RhD-positive platelets are
given to RhD-negative women, there is a requirement to give an appropriate dose
of anti-RhD prophylaxis.
Massive blood loss at delivery is rapidly complicated by the development of
DIC (see Chapter 8). This may be exacerbated by the hemodilution consequent
upon red cell and fluid replacement. In diagnosing DIC, it should be remem-
bered that the maternal plasma fibrinogen is often 4g/l (or more) in normal
women at term and that there may be foreshortening of the activated partial
thromboplastin time (APTT: usually no more than a decrease in the mean APTT
of 4 seconds) with increasing gestation. The level of fibrinogen required for
hemostasis should be at least the same as that of non-pregnant subjects at
>1.0g/l, using either cryoprecipitate or, if available, a double-virally inactivated
fibrinogen concentrate. The platelet count should also be maintained above
50 × 109/l when bleeding is continuing, and may need to be >80 × 109/l when an
operative intervention is required.
Early intervention to treat the underlying cause is essential with, in particular,
rapid delivery in cases of placental abruption or placenta previa. In postpartum
hemorrhage associated with uterine atony, rapid treatment with oxytocin or car-
boprost (a prostaglandin analog) and exclusion of retained placental tissue is
required. Further therapy with uterine packing, arterial embolization or hysterec-
tomy may also be required.
42
CHAPTER 3
Maternal anemia and

hemato-oncology in
pregnancy
Anemia in pregnancy
A fall in hemoglobin levels with increasing gestation occurs physiologically in
pregnancy as a result of hemodilution (see Chapter 1). However, this is distinct
from the development of significant anemia, which increases the risk of preg-
nancy complications. Anemia most commonly develops because of hematinic
deficiency, but may also develop in the context of a new red cell or marrow dis-
order, or because of a pre-existing condition affecting red cell production or sur-
vival (see Chapter 9).
As outlined in Chapter 1, a relative reduction in the hematocrit is evident,
especially in the first and second trimesters. This leads to a fall in the hemoglo-
bin concentration, which is seen from around 16 weeks gestation until the
middle of the third trimester. In subjects with no evidence of iron deficiency this
fall is often no more than 1g/dl and results in a normochromic normocytic
picture with a hemoglobin concentration of not <11g/dl in the first trimester, or
10g/dl in the second and third trimesters (Table 3.1).
Iron deficiency
Iron has a wide distribution in the body, with around 70% located in hemoglobin
and myoglobin. Most non-heme iron is found in ferritin or hemosiderin, in
either hepatocytes or macrophages. There is a small amount of iron associated
Table 3.1 – FBC changes in normal pregnancy
NNA from 16 to 34 weeks

Hemoglobin fall <1g/dl
Hemoglobin >11 g/dl in first trimester
Hemoglobin >10 g/dl in third trimester
FBC, Full blood count; NNA, Normochromic normocytic anemia

43
with the iron-transport protein transferrin; and iron is also required for the func-
tion of enzymes, such as catalases and peroxidases. When iron is required for
incorporation into these proteins, it is most often derived from recycled iron that
has been released from red cell breakdown.
Iron absorption takes place in the first part of the duodenum, and iron
balance is achieved by the regulation of this absorption. Iron is absorbed from
heme, or from ferrous (non-heme) iron by two distinct pathways. Non-heme iron
absorption can be enhanced by ascorbic acid, which assists in the conversion of
ferric iron to the ferrous form (the state that is required for absorption). By con-
trast, phytates (found in cereals) bind non-heme iron and inhibit its absorption,
and dietary phosphates also inhibit iron absorption from eggs and milk.
Iron is released from heme on the cell border, and inside the mucosal cell is
added to the pool of absorbed iron. Some of the iron is transported into the
plasma within a few hours of absorption, but much remains within the mucosal
cell. Loss of the mucosal cell (and its iron), by sloughing, appears to have a
major role in the regulation of iron balance.
It is likely that iron is transported from the mucosal cell by the action of a
transmembrane protein, whose mRNA synthesis is stimulated by iron deficiency.
Iron leaving the cell must be oxidized to the ferric state to allow binding to the
transport plasma protein (apo-) transferrin. Iron absorption is increased when
body iron stores are depleted and reduced when excessive body iron is present.
Iron absorption is also increased when there is an increase in erythropoiesis,
hypoxia or inflammation. Iron is then transported in the plasma by the glycopro-
tein transferrin. This is synthesized predominantly in the liver in response to iron
deficiency: the total amount of iron that the available plasma transferrin can
bind is called the total iron binding capacity (TIBC). In normal non-pregnant
subjects around one third of the TIBC is used. Transferrin carries iron to specific
transferrin receptors, which are prominent on the developing red cell but are
lost as the red cell nears maturity. When transferrin saturation falls to <16%
there is insufficient iron delivery to the bone marrow. This results in an increase
in free red-cell-derived protophorphyrin, due to the lack of iron leading to a
reduction of the incorporation of protophorphyrin into heme. As there is a
reduction in the red cell hemoglobin, the red cell continues to divide, resulting
in smaller red cells (i.e. microcytosis). Iron deficiency also results in a degree of
ineffective erythropoiesis, which results in a reduced red cell half-life (Table 3.2).
Iron deficiency in pregnancy

Pregnancy results in an average loss of maternal iron of around 700mg, associated
with the needs of the developing fetus and placenta, as well as the blood loss at
delivery. In addition, there is a requirement for around 450mg of iron to support
the increase in red cell mass associated with normal pregnancy. Although there
is an increase in absorption in the second and third trimesters, there is an iron
intake requirement of around 2.5mg/day throughout pregnancy, with perhaps
44
Maternal anemia and hemato-oncology in pregnancy
Table 3.2 – Detection of iron deficiency
Increasing iron deficiency
Pathophysiology Stores Plasma Marrow Free rbc Microcytosis Anemia Epithelial

iron iron protoporphyrin changes
Laboratory Marrow iron Percentage Mean corpuscular Hemoglobin

Serum ferritin saturation volume (MCV)
Protoporphyrin Mean corpuscular
Serum transferrin hemoglobin (MCH)
receptor
Clinical Lethargy Dyspnoea Tongue

Tiredness Chelitis
Insomnia Nail
Reduced exercise tolerance Dysphagia
Symptoms of underlying cause Infection
rbc, Red blood cell
3.0–7.5mg/day required in the third trimester. The support of lactation requires

<1mg/day, which is offset by the suppression of menstruation during lactation.
The effect of iron deficiency (prior to the development of anemia) on mater-
nal and fetal wellbeing is not fully understood. There has been speculation that
maternal iron deficiency may itself lead to increased blood loss at delivery
because of poor myometrial function, although this is by no means proven.
The fetus is ‘parasitic’ in its acquisition of iron from the mother, but there is
the potential that iron deficiency in the mother could result in poor fetal iron
stores at delivery, which may have a consequence on early childhood development
and behavior (especially if there is subsequent poor iron intake). Iron deficiency
has also been implicated in fetal programing, with an increased placental:fetal
weight ratio (relative placentomegaly) reported. This, in turn, may also have an
influence on maternal blood pressure, and perhaps even the blood pressure of
the child when it reaches adulthood. Severe maternal iron deficiency has been
associated with premature delivery and low birthweight, but this could also relate
to the underlying cause of the maternal anemia. Although repletion of iron leads
to an increase in hemoglobin and hematocrit, it may not be associated with any
improvement in pregnancy outcome. On account of this, both the importance of
iron deficiency and the benefits of iron replacement remain controversial.
Diagnosis of iron deficiency in pregnancy

The signs and symptoms of iron deficiency before the development of anemia
are non-specific, and include tiredness and reduced exercise tolerance. There
45
may also be symptoms and signs associated with the underlying cause. Severe
iron deficiency anemia (IDA) is associated with atrophy of the papillae, a painful
tongue, angular chelitis, nail ridging and, when severe, spooning and (with the
development of a post-cricoid web) dysphagia. There may also be an increased
risk of infection due to the effect of iron deficiency on cellular immunity and
phagocytosis. These changes, however, are extreme and are unlikely to be seen
in iron-deficient pregnancies in developed countries. Common causes of iron
deficiency in women of childbearing age are shown (Table 3.3).
As noted above, the physiological anemia of pregnancy is associated with a
normochromic normocytic picture. It should be noted, however, that a reduction
in hemoglobin concentration is a relatively late feature of iron deficiency, and is
preceded by depletion in iron stores and serum iron. Furthermore, the reduc-
tion in mean corpuscular volume (MCV), which is seen at an early stage with iron
deficiency in the non-pregnant, is not so reliable in pregnancy because of the
physiological changes in red cell mass and plasma volume. In particular, there
may be a rise in the MCV and, thus, a normal MCV does not exclude iron defi-
ciency. In pregnancy the initial appearance of IDA may therefore be nor-
mochromic and normocytic, and when there is a hypochromic microcytic
picture, the possibility of thalassemia should be considered and excluded.
Although serum ferritin contains little iron, its circulating concentration cor-
relates well with body iron stores, but the use of ferritin to detect iron deficiency
may be complicated, as it is an acute phase protein and the lower limit of the ref-
erence range is increased in the presence of inflammatory disorders. However, in
early pregnancy, reduced serum ferritin concentrations usually provide a reliable
indication of iron deficiency, but hemodilution in the second and third
trimesters of pregnancy reduces the concentrations of all measures of iron status,
and thus the ranges used in non-pregnant women are not always reliable.
Nonetheless, a low result still indicates iron deficiency in most pregnant subjects.
Table 3.3 – Causes of iron deficiency
Dietary Vegetarian/vegan
Blood loss Menorrhagia

Peptic ulceration
Inflammatory bowel disease
Hemorrhoids
Varices
Aspirin
Anticoagulants
Malabsorption Celiac
Gastrectomy
46
With regard to other tests of iron deficiency, normal pregnancy (as noted in
Chapter 1) is associated with a progressive fall in serum iron, and an increase in
both serum TIBC and free protoporphyrin levels. Serum transferrin receptor
levels may also increase with gestational age (especially after 20 weeks gestation),
and may not fall in response to iron given at a supplemental dose. As noted
above, it appears most likely that any increase in serum transferrin receptor levels
in later pregnancy relates to increased maternal erythropoiesis. From this it is dif-
ficult to know how reliable serum transferrin receptors are in diagnosing IDA in
pregnancy. Thus, a combination of assessments should be used to assess iron
status when the diagnosis is not clear cut.
Prevention of iron deficiency in pregnancy

The usual practice is to use the hemoglobin concentration as a screening test for
iron deficiency, with an assessment at booking and a further one in the early
third trimester. Whether routine supplementation to all women is warranted is
not resolved, as it is not clear (at least in developed countries) whether the fetus
benefits from routine maternal iron prophylaxis (see above). If required, supple-
mentation can be achieved with 60mg of iron/day and 400µg of folic acid given
for 6 months of pregnancy (where the prevalence of anemia is low), or contin-
ued to include the puerperium (in countries where there is a high incidence of
maternal anemia). In the USA, the American College of Obstetricians recom-
mends 30mg of elemental iron/day in the second and third trimesters to all
women, irrespective of iron status. This may be a reasonable strategy, as the main
side effects from iron are seen at replacement (200mg/day) rather than prophy-
lactic doses, and iron supplementation is safe, as it will not produce a supranor-
mal hemoglobin or hematocrit. There is, however, a consideration that 2–5 in
1000 subjects of European descent may be homozygous (or compound heterozy-
gous) for genes leading to iron overloading and hemochromatosis. Whether iron
supplementation in pregnancy in this group results in clinical disease is not clear,
as women (perhaps associated with menstrual iron loss) are at a lower risk of
organ damage in the reproductive years than men.
Treatment of iron deficiency in pregnancy

The treatment of established iron deficiency is with 200mg/day of elemental
iron. In pregnancy, there is no value in using 200mg three times per day, as is
often prescribed, as this will only increase the risk of side effects without enhanc-
ing the resolution of the anemia. Iron supplements may be associated with gas-
trointestinal (GI) upset (nausea, heartburn and diarrhea), but this can be dose
related and so is usually ameliorated by either a change in the type of oral prepa-
ration, or by a reduction in the dose to 100mg/day. In non-pregnant subjects,
200mg/day of ferrous sulphate is expected to result in a rise in hemoglobin of
around 1g/dl per week. However, there may be an improvement in wellbeing in
47
advance of any change in the red cell indices, which may reflect improvements in
the intracellular metabolic processes where iron is required.
Iron therapy may fail to replete iron stores when there is malabsorption, or
when iron lost in bleeding exceeds intake. Occasionally, no improvement is seen,
as the diagnosis of iron deficiency is incorrect. However, in the vast majority of
instances, failure to improve iron status results from poor compliance with the
iron therapy. Vitamin C can enhance absorption of ferrous iron, and this can be
achieved by taking the iron supplement with a glass of fresh orange juice or a
50mg tablet of ascorbic acid. It is the authors’ practice to routinely prescribe
vitamin C with iron supplements for treatment of IDA in pregnancy. For women
who have difficulty with swallowing tablets, liquid preparations are available, but
these can cause staining of the teeth.
There are a number of forms of parenteral iron therapy, which can be given
as a total dose intravenously (iron dextran and iron sucrose compounds) or, in
some instances (iron dextran), as multiple doses intramuscularly. The indica-
tions for parenteral therapy include malabsorption and failure of compliance
with oral therapy, and there may also be requirement when erythropoietin is
being used.
With the exception of renal failure requiring dialysis or malabsorption, par-
enteral iron probably does not produce a faster hemoglobin response than oral
iron. Despite this, parenteral iron therapy is often used when severe IDA is diag-
nosed at an advanced stage in pregnancy.
With intravenous dextran (and to some extent iron sucrose) compounds
there is a small risk of an anaphylactoid reaction (estimated at around 0.6% of
recipients), and it is recommended that a test dose of 25mg of iron should be
given over 15 minutes before a therapeutic dose is given. With intramuscular
preparations there is a risk of pain or abscess formation, skin pigmentation and
local sarcoma at the injection site. Overall, some side effects are reported in
around 25% of subjects receiving parenteral iron therapy. The majority of these
are mild, and include headache, urticaria and nausea, but serious adverse events
have been reported to occur in as many as 5% of subjects. In around 1% of recip-
ients of either preparation there is also the possibility of a delayed reaction with
myalgia, arthralgia, lymphadenopathy and fever.
Folic acid deficiency

Folates are found in a wide variety of foodstuffs, but liver, green vegetables
(such as brocolli, asparagus, sprouts and spinach), mushrooms and nuts
contain the highest quantities. Cooking, however, may result in a substantial
loss in the folic acid available from such produce. The average daily require-
ment in non-pregnant subjects is around 100µg/day, with an average Western
diet supplying around 250µg/day. Folic acid is predominantly absorbed in the
upper small intestine, with 50–90% of food folates presented to the upper
small bowel being absorbed. Folic acid and its metabolites are lost in the urine,
48
through the skin and there is also an enterohepatic circulation through the
bile. Folic acid deficiency is most commonly due to dietary insufficiency associ-
ated with alcoholism and poverty. Malabsorption, due to gluten-induced
enteropathy, partial gastrectomy or Crohn’s disease may also exacerbate dietary
insufficiency, and such patients should routinely receive folate supplements
throughout their pregnancy. Other subjects at risk include those with an
increased usage of folate. This includes those with chronic hemolytic con-
ditions (e.g. hemoglobinopathy, sickle-cell syndromes, red cell membrane and
enzyme disorders, myeloproliferative disorders, and autoimmune hemolytic
anemia), those taking antiepileptic drugs (e.g. carbamazepine), and those
requiring dialysis (where there may be an increase in loss through dialysis due
to loose plasma protein binding) (Table 3.4).
Polyglutamated folate is an essential coenzyme in the transfer of single carbon
units required for the manufacture of DNA and RNA. By this, and other means,
folic acid may function to reduce oxidative potential. Folic acid, along with
vitamin B12 and vitamin B6, assists in the homocysteine (HCY)–methionine
pathway (Figure 3.1), where the presence of folic acid drives the production of
methionine from HCY, leading to a reduction in the plasma levels of HCY.
Circulating HCY is readily oxidized, and as a result may increase the genera-
–
tion of superoxide (O 2) and hydrogen peroxide (H2O2) species. These free radi-
cals can induce oxidation of low-density lipoproteins (LDL) in the endothelium,
and it has been proposed that prolonged exposure of the endothelium to HCY
results in exhaustion of the mechanisms of endothelial protection against HCY.
HCY also has an effect on a number of aspects of coagulation, which includes
upregulation of Factor (F) V generation and downregulation of the anticoagu-
lants activated protein C (aPC) and thrombomodulin.
Although most subjects are diagnosed with deficiency because of an incidental
finding of a raised MCV, folate and vitamin B12 deficiency may present with
Table 3.4 – Cause of folic acid deficiency
Dietary Alcoholism
Poverty
Adolescence
Malabsorption Gluten-induced enteropathy (celiac)
Increased use Chronic hemolysis

Congenital red cell disorders
Hemoglobinopathy
Myeloproliferative disorders
Loss Dialysis
Miscellaneous Anticonvulsants
49
FOLATE CYSTEINE URINE
B6 CBS
5 METHYL
HCY
THF
MTHFR
SAH
5,10 B12 MET SAM

METHYLENE SYNTHASE
THF
THF METHIONINE
SERINE GLYCINE
B6
PROTEIN
Figure 3.1 Homocysteine (HCY)–methionine pathway

Methionine is converted to S-adenosyl methionine (SAM); this is converted via S-adenosyl homocysteine (SAH)
to HCY. HCY is converted via cystathionine beta-synthase (CBS) to cysteine, which requires vitamin B6 (B6) as a
cofactor; cysteine is excreted in the urine. HCY leads to the reformation of methionine, via methionine synthase
(MET SYNTHASE); this requires folate and vitamin B12 (B12) and methylene tetrahydrofolate reductase (MTHFR).
anemia (Table 3.5). Non-specific symptoms such as lethargy, breathlessness and,

occasionally, anginal symptoms may be accompanied by other features such as GI
upset, angular cheilitis and glossitis. In more severe cases there may be jaundice
(due to a rise in unconjugated bilirubin from ineffective erythropoiesis), and a
reduction in both white cell and platelet production, leading to infection and
bruising. The effect on platelets and leukocytes is often forgotten and clinicians
should be aware that incidental findings of thrombocytopenia on a blood film
may reflect folate deficiency.
Table 3.5 – Common causes of macrocytosis
Vitamin B12 deficiency

Folate deficiency
Alcohol
Hypothyroidism
Pregnancy
Liver disease
Myelodysplasia
Hemolysis
50
Folic acid deficiency in pregnancy

In pregnancy folic acid requirements increase to around 400µg/day, due to the
demands of the developing fetus and increased folate catabolism, which is
accompanied by a gradual fall in serum folate levels with increasing gestation.
Those with an inadequate dietary intake are at risk of the development of a
macrocytic anemia. The prevalence of deficiency is <5% of subjects in developed
countries and may be as low as 0.5% where there are routine prophylaxis pro-
grams. Folate deficiency is more common in multiple pregnancy, frequent child-
birth and when the mother is still an adolescent. In the first trimester, folic acid
deficiency may not often present with anemia but, linked with HCY generation, is
associated with fetal neural tube defects. Folate deficiency anemia often coexists
with iron deficiency (due to a generally poor diet), and more often presents at
the end of pregnancy or in the early puerperium. Very occasionally, and often in
the presence of infection, folate deficiency may present as an acute aplasia (with
pancytopenia accompanied by megaloblastic erythropoiesis in the bone marrow),
or come to light as a result of thrombocytopenia.
Diagnosis of folic acid deficiency in pregnancy

The diagnosis of megaloblastic anemia in pregnancy is suggested by the presence
of a macrocytic blood picture (MCV >100fl), the presence of right-shifted neu-
trophils (>5 segments) on the blood film and, in severe cases, a mild leukopenia
and thrombocytopenia (platelets 50–100 × 109/l). A reduction in serum folate
levels is a sensitive test of folic acid deficiency; however, it may also be reduced if
there has been only a recent folate lack in the diet and also reduces with increas-
ing gestation. Falsely normal serum folate results may also be seen in acute liver
failure, renal failure and vitamin B12 deficiency. Measurement of red cell folate is
more representative of body stores and is less affected by recent changes in
dietary intake, but can be falsely normal if there has been a recent red cell trans-
fusion, and has been reported to occasionally rise in an otherwise normal preg-
nancy (Table 3.6).
In detecting folate deficiency in pregnancy there may diagnostic difficulties,
as macrocytosis can be seen in pregnant subjects without megaloblastosis, and
is also seen in subjects with alcohol excess, hemolysis, hypothyroidism, vitamin
B12 deficiency, or renal failure. If, in such circumstances, folate assays are not
clear-cut, megalobastic erythropoiesis can be demonstrated by bone marrow
examination. The principal feature of megaloblastosis is asynchrony of cyto-
plasmic and nuclear development of red cell precursors (i.e. an immature
nuclear appearance with mature hemoglobinised cytoplasm). The diagnosis
can be complicated when there is also iron deficiency, as the MCV may be
normal. Usually, however, the blood film is dimorphic (i.e. showing evidence of
both the microcytes of iron deficiency and the macrocytes of megaloblastic ery-
thropoiesis).
51
Table 3.6 – Source of Hematinics
Hematinic Source Pregnancy Measure Measure Pregnancy Deficiency

requirement limitations prophylaxis treatment
Vitamin B12 Animals 3µg/day Serum vitamin B12 Gestation effect ?3µg/day 1mg
Urinary methyl- Poor correlation
malonic acid with vitamin B12
(MMA)
Folate Vegetables, 400µg/day Mean corpuscular Gestation effect or 400µg/day 15mg/day

liver, nuts volume (MCV) iron deficiency to 5mg/day
anemia (IDA)
Serum folate Recent diet/renal/
rbc folate vitamin B12
deficiency
Homocysteine Transfusion
(HCY) Poor correlation
with folate
Iron Meat, 2.5mg/day to MCV Gestation effect or 30–60µg/day 200mg/day

vegetables 3.0–7.5mg/day thalassemia
dairy Ferritin Acute phase
reactant and
gestation effect
Iron/total iron Gestation effect
binding capacity
(TIBC)
Serum transferrin Gestation effect
(sTF) receptors
Marrow Pain
rbc, Red blood cell
Prophylaxis and treatment of folic acid deficiency in pregnancy

Given the potential for mild folate deficiency to be associated with neural tube
defects, folic acid supplementation is routinely given in the first trimester at a
dose of 400µg/day. This also extends to those planning pregnancy with a recom-
mendation that such supplements be taken for 3 months prior to conception.
This routine prophylaxis results in at least a 70% reduction in the risk of neural
tube defects. There may also be potential benefits to other pregnancy complica-
tions, such as pre-eclampsia and intrauterine growth restriction, where hyper-
homocysteinemia may be a pathophysiological factor. Folic acid will reduce HCY
concentrations, but whether prolonged prophylaxis is effective in preventing or
ameliorationg these problems is not yet established. At any rate, there has been
no suggestion that such prolonged prophylaxis, at a dose of 400µg/day, is detri-
mental to the mother or fetus. In those who have had a previous child affected by
52
a neural tube defect, folic acid therapy (at a dose of at least 5mg/day) should be
given when pregnancy is planned and for the first trimester. As noted above, pro-
phylaxis is also required for subjects with chronic inherited or acquired red cell
disorders. In some such cases, there may also be a role for adding iron prophy-
laxis.
Proven folic acid deficiency anemia should be treated with folic acid therapy
at a dose of 5mg × 3/day. In all such cases of anemia, vitamin B12 deficiency must
also be excluded, as folate supplementation may improve the anemia of vitamin
B12 deficiency but exacerbate the neurological deterioration associated with it
(see below). In general, it is also good practice to check the vitamin B12 status of
all those presenting with significant folic acid deficiency. Folinic acid – a form of
folic acid which can bypass a block in the methylene tetrahydrofolate reductase
enzyme induced by methotrexate, or an excess dose of trimethoprim – can be
used to reverse folate inhibitory drugs, and is occasionally used when there is
severe folate deficient anemia from other causes.
In all women with evidence of folic acid deficiency it is important to take a
dietary history, and if the diet is lacking in folate-rich foods to provide dietary
advice.
Vitamin B12 deficiency

As shown in Figure 3.1, vitamin B12 is another essential cofactor in both the man-
ufacture of DNA and the disposal of HCY. Vitamin B12 is also essential in the con-
version of methylmalonyl-CoA to succinyl-CoA. In addition to similar symptoms
of anemia to those observed with folate deficiency, vitamin B12 deficiency also
results in a demyelinating neuropathy affecting the pyramidal and posterior
tracts of the spinal cord, the peripheral nerves and the optic tract.
Vitamin B12 is only synthesized by microorganisms and the only source avail-
able to humans is derived from animal foodstuffs. Of these, liver, kidney, meat,
fish, chicken and dairy products are the most important sources. Thus vegetari-
ans, and particularly vegans, are most at risk of nutritional deficiency. Unlike
dietary folates, cooking does not destroy vitamin B12. On average, there is a body
store of around 3mg, with a daily dietary requirement of around 3µg/day. As a
result, it may take 3–5 years after the onset of a dietary lack for clinical deficiency
to become apparent.
A very small amount of vitamin B12 is absorbed passively in the duodenum and
ileum, but the major site of absorption is by a specific mechanism in the ileum,
requiring the combination of vitamin B12 and intrinsic factor (produced in the
gastric parietal cells). There is also an enterohepatic circulation of up to 5µg of
vitamin B12, which is the reabsorbed in the terminal ileum.
The causes of vitamin B12 deficiency include a dietary lack, which is most often
seen in vegans who do not eat animal products. In addition to dietary deficiency,
other causes include pernicious anemia (due to a defect of intrinsic factor pro-
duction), partial gastric resection, ileal resection, intestinal stagnant loop syn-
53
drome, Crohn’s disease, tapeworms and tropical sprue. In addition, deficiency of

folate can result in small bowel malabsorption of vitamin B12.
As noted above, vitamin B12 deficiency should be excluded in all cases of
megaloblastic anemia, by assessment of serum vitamin B12 levels. Occasionally,
low levels, associated with neurological disturbance, can be seen even in the
absence of anemia.
In non-pregnant subjects, a diagnosis of pernicious anemia (PA) can be made
by demonstrating a lack of intrinsic factor secretion using vitamin B12 binding
studies (demonstrating the effect on urinary vitamin B12 excretion of adding
intrinsic factor to an oral dose of radiolabelled vitamin B12). Malabsorption is
indicated by a low urinary response to the radiolabelled oral vitamin B12 dose,
that is not sufficiently improved by the addition of intrinsic factor. Such a result
may require other intestinal absorption studies and bowel imaging to reach a
diagnosis.
In pregnancy, other tests, such as those to detect antibodies to intrinsic factor
or parietal cells, are useful to confirm the diagnosis of PA. Of these, intrinsic
factor antibodies (although seen in other autoimmune conditions) are more spe-
cific than parietal cell antibodies for pernicious anemia. Indeed, such antibodies
have been shown to cross the placenta and to lead to intrinsic factor deficiency in
the newborn.
The diagnosis of vitamin B12 deficiency can be confirmed by a therapeutic trial
of parenteral vitamin B12 (which will happen incidentally in non-pregnant sub-
jects when vitamin B12 binding studies are performed, as a non-radiolabelled par-
enteral vitamin B12 injection is required to complete the study). After such a trial,
a reticulocyte response should be evident between 3 and 7 days. As there may be
difficulty in interpreting borderline vitamin B12 levels, other assessments to
demonstrate tissue deficiency, such as plasma HCY levels (which correlate well
with folate but not with vitamin B12) and the urinary methylmalonic acid
(MMA):creatinine ratio (which can be determined by a single urinary measure),
have been proposed. The level of serum MMA, however, is less sensitive and spe-
cific than urinary estimates, as it is also raised in renal insufficiency, thyroid
disease, pregnancy and when there is dehydration.
Vitamin B12 deficiency in pregnancy

Vitamin B12 levels may fall by 30–50% during pregnancy, with evidence of
reduced levels in the first trimester in some individuals. This occurs despite a diet
adequate in vitamin B12, and is likely to be due to a combination of increasing
maternal plasma volume, hormonal changes, increasing maternal requirement
and transfer to the fetus. It appears that there is a poor correlation between
maternal serum vitamin B12 and serum HCY levels in pregnancy, and, in non-
anemic pregnant subjects, no correlation between urinary MMA levels and serum
vitamin B12 has been reported. This suggests that the physiological reduction in
serum vitamin B12 in pregnancy does not represent a true tissue deficiency.
54
A correlation between maternal and neonatal vitamin B12 levels has, however,
been reported. Thus, although mild vitamin B12 deficiency appears compatible with
a normal pregnancy outcome, it could result in a low level of vitamin B12 in the off-
spring (particularly if the baby is subsequently breastfed). Lower levels of vitamin
B12 have also been found in women who have had a child with a neural tube defect
and in those with early recurrent abortions. Indeed, it is thought that persisting
vitamin B12 deficiency can result in infertility, but the role of lesser deficiency in
fetal cognitive development or birthweight is not yet resolved. In countries where a
vegetarian diet is the norm, folate deficiency is less common and vitamin B12 defi-
ciency may play a more important role in the development of neural tube defects.
Autoimmune hemolytic anemia (AIHA)

On average, red cells survive for 120 days in the circulation. If an autoantibody
develops which can result in red cell lysis, the destruction occurs either within
the circulation and/or within the liver and spleen. Such autoantibodies may be
of IgG, IgM or IgA class, and this phenomenon is often classified, by the temper-
ature range of the antibody, into cold or warm acting.
In general, warm antibodies have their maximal effect at 37°C and cold-acting
ones have increasing activity with decreasing body temperature. Warm antibodies
are typically: of IgG type; may or may not result in complement binding to the
red cells; and most often result in loss of red cells in the spleen or liver. Cold
antibodies are: most often of IgM type; and are often capable of fixing comple-
ment and leading to intravascular hemolysis. There are, of course, exceptions to
this and AIHA can be associated with a mixed antibody pattern.
The mechanism whereby these antibodies develop is essentially unknown, but
may be related to a genetic predisposition through the patient’s human leuko-
cyte antigen (HLA) type, molecular mimicry (where there is a similarity to a
microorganism) and/or involve a change in immune tolerance.
Red cell destruction leads to an increase in red cell production in the bone
marrow (erythroid hyperplasia), with the release of early cells (called reticulo-
cytes) into the peripheral circulation (Table 3.7). Reticulocytes, unlike mature
Table 3.7 – Diagnosis of Intravascular Hemolysis
Macrocytosis
Reticulocytosis
Spherocytosis
Serum haptoglobin
Plasma hemoglobin
Hemosiderinuria/hemoglobinuria
Lactate dehydrogenase
Positive direct antiglobulin (Coombs’) test
55
red cells, contain detectable RNA. If there is brisk intravascular breakdown there
will be a reduction in the level of the serum protein haptoglobin, which com-
plexes with hemoglobin to remove it from the circulation. There may also be
detectable plasma hemoglobin and evidence of hemoglobinuria. When hemoly-
sis is less marked, free hemoglobin is catabolized by the renal tubular cells to
hemosiderin, which is detectable in the urine. The enzyme lactate dehydroge-
nase is also released from damaged red cells and the hemoglobin metabolite
bilirubin is released from the reticulo-endothelial system to be conjugated in the
liver and excreted in bile into the gut. This fecal urobilinogen is reabsorbed from
the gut into the circulation and is then excreted in the kidneys, and is also
detectable in the urine.
The presence of an antibody bound to a red cell antigen can be demonstrated
by the direct antiglobulin test (the Coombs’ test) if it is IgG. If it is IgM or IgA, its
presence can be inferred by a positive antiglobulin test that has detected the
presence of complement on the red cell surface. The presence of bound anti-
body or complement does not itself indicate hemolysis, but simply that an
autoantibody is present. Antibody-mediated damage results in the loss of red cell
membrane and this leads to the formation of detectable spherocytes within the
circulation.
Warm-acting autoantibodies are most often IgG and, although they occasion-
ally have a specificity for RhC, e, Kell or RhD, they most often react equally with a
number of donor red cells in cross-matching procedures (see Chapter 2). The
consequences of such antibodies can vary from an incidental positive direct
antiglobulin test to fulminant hemolysis. The majority of clinically affected sub-
jects present with a chronic stable anemia.
More than 90% of cold antibodies have the I red cell antigen as the target,
with the remainder aimed at the i antigen. In the case of the condition known as
paroxysmal cold hemoglobinuria (classically associated with syphilis, but also
associated with viral infections), the antibodies are most likely to be of polyclonal
IgG type and specific for the P red cell antigen. In this disorder the IgG binds in
the patient’s periphery (at cooler temperatures), but then results in intravascular
lysis when it binds complement in the warmer central circulation.
AIHA in pregnancy
In women of childbearing age, cold-acting IgM antibodies are now most often
idiopathic, or are found in association with the presence of autoimmune dis-
orders or infections, such as infectious mononucleosis, measles or mumps. If
these antibodies are only active at <30°C, they are unlikely to cause significant
hemolysis.
Warm antibodies in this age group are often idiopathic, but can also be associ-
ated with autoimmune disorders, some therapeutic drugs (classically penicillin,
quinidine or methydopa), viral infections, immunodeficiency (including HIV),
lymphoproliferative disorders and other malignancies (Table 3.8). The rate of
56
Table 3.8 – Causes of autoimmune hemolytic anemia (AIHA)
Predominantly cold Autoimmune disorders

Infectious mononucleosis
Measles
Mumps
Idiopathic
Predominantly warm Idiopathic

Autoimmune disorders
Drugs
HIV
Other viral infections
Pregnancy
Lymphoma/malignancy
AIHA in pregnancy appears to be higher than that of subjects of a similar age (1

in 50 000 versus 0.2 in 50 000), with the possibility of increasing hemolysis with
increasing gestation and remission after delivery in some cases. As with non-preg-
nant subjects, there is the possibility of concomitant immune thrombocytopenia
(Evan’s syndrome). In addition, when there is a maternal IgG antibody, there is
the possibility of transplacental passage of the antibody resulting in an asympto-
matic positive direct antiglobulin test in the baby, or self-limiting infant hemoly-
sis. In fact, a pregnancy-specific AIHA can occur. This remits after pregnancy but
can return in subsequent pregnancies.
Treatment of AIHA
Drug triggers for hemolysis should be sought and any potential implicated drugs
excluded, if possible. No treatment is required for asymptomatic direct antiglob-
ulin test positivity. When treatment is required for an idiopathic AIHA, the first-
line therapy should be corticosteroids. Most often this is given as oral
prednisolone at a dose of 1–2mg/kg body weight. When a stable hemoglobin is
achieved, the steroids are reduced at a rate of 5–10mg/week until a maintenance
dose is achieved. With high doses, the mother will be at risk of steroid-related
side effects, but at maintenance doses this is not usually a problem. However, any
mother on steroid supplements should receive 150mg of hydrocortisone 8 hourly
at the time of delivery. Fetal problems are not associated with steroid therapy,
although with very prolonged use of high doses (>30mg/day of prednisolone),
fetal adrenal suppression can occur, and the pediatricians should be alerted to
this possibility.
Where steroids cannot be used, or fail, regular infusions of intravenous
human normal immunoglobulin may be used to maintain hemoglobin levels. In
57
all cases, folic acid supplementation is required to prevent hemolysis-induced

folic acid deficiency. When transfusion is required, the principles detailed in
Chapter 2 should be followed. Other therapies such as azathioprine have been
used during pregnancy, but splenectomy should be reserved for life-threatening,
treatment-resistant anemia whilst the woman is still pregnant.
Hemato-oncology and pregnancy

Although a number of different hematological malignancies do occur in women
of childbearing age, it is rare for an acute disorder to be diagnosed for the first
time during pregnancy, but it is increasingly likely that someone who has been
previously treated for a malignant hematological condition will present for ante-
natal care. The specific care of acute hematological malignancy is a rapidly
changing area which is beyond the scope of this book, but the general principles
of management of malignant hematological conditions that may be found in
pregnant subjects will be outlined.
Essential thrombocythemia (ET)

ET is a myeloproliferative disorder, characterized by a persistent elevation in
platelet count (i.e. beyond the local reference range), occurring in the absence
of iron deficiency, bleeding, malignancy or other myeloproliferative and connec-
tive tissue disorders. Although it is a disease of late, middle and old age, an
excess of cases around 30 years of age has been reported. In practice, it should
be remembered, however, that around 85% of cases of thrombocytosis are reac-
tive rather than a primary phenomenon (Table 3.9).
Bleeding is the most frequent symptom in the majority of ET cases, and in
young women the clinical course is most often benign. Subjects are, however, at
increased risk of myocardial ischemia, transient neurological dysfunction (char-
acterized by headache, lightheadedness or parasthesia), peripheral arterial occlu-
sion (and erythromelalgia) and venous thrombosis. These complications are
more likely in older subjects, although those with a platelet count of 1000–2000 ×
Table 3.9 – Causes of thrombocytosis
Reactive 85% Bleeding

Iron deficiency
Malignancy
Connective tissue disorders
Splenectomy
Primary 15% Essential thrombocythemia (ET)

Other myeloproliferative disorders
58
109/l, and especially those with evidence of a platelet function defect (or
acquired von Willebrand’s disease), may be at particular risk of bleeding. Those
with a past history of thrombosis or bleeding should be considered at high risk of
another thrombotic or bleeding event. Whether there is any impact of co-inheri-
tance of thrombophilia on the occurrence of thrombosis in ET is unknown.
ET in pregnancy
Although, as noted above, the clinical course of ET is most often benign in the
young, there are reports of an association with an adverse pregnancy outcome.
The incidence of adverse effects comes predominantly from case reports or small
case series. From these, the most common complication is miscarriage (from 25
to 40%), with around a third occurring in the first trimester. The occurrence of
late fetal loss may also be increased, with stillbirth/intrauterine death reported to
occur in 5–11% of pregnancies, and there may also be an increased risk of
growth restriction and pre-term delivery. The consequences to the mother
appears more favorable, with the incidence of thrombosis at <3.5%, and most
probably less <1%, for significant vascular occlusion (cerebral venous thrombosis
or transient cerebral ischemia of arterial origin). The presence of placental
thrombosis has been reported in association with late fetal loss and bleeding has
been reported in <10% of cases, with the majority of events of a minor nature.
As with normal subjects, pregnancy is associated with a reduction in the
platelet count in subjects with ET. Whether the degree of fall in platelet count is
related to pregnancy outcome is not clear, and, in general, experience in a previ-
ous pregnancy, or the maternal pre-morbid condition, do not clearly indicate the
prognosis in the current pregnancy.
A platelet level of 1000–1500 × 109/l has been suggested as the level at which
cytoreduction is required. Despite this, it has been suggested that chemothera-
peutic agents can be safely withheld in pregnancy in subjects <30 years of age. At
present there is insufficient evidence, however, to recommend that therapy can
be safely withheld in asymptomatic individuals when a diagnosis is first made
during pregnancy. However, cytoreduction should be avoided in the first
trimester if possible.
Hydroxycarbamide (otherwise known as hydroxyurea, a ribonucleotide
reductase inhibitor) is widely used in non-pregnant subjects and has been
shown to reduce the risk of thrombosis. There is, however, the potential for fetal
teratogenicity and its safety in pregnancy, especially in the first trimester, is not
known.
Although anagrelide has been used to treat ET, it appears to be less effective
in the prevention of thrombotic events and possibly carries an increased risk of
myelofibrosis. It therefore has a very limited role in the management of ET
generally. As it also crosses the placenta, as well as the breast, and there is
limited information on its safety in pregnancy, it should not be used in this
circumstance.
59
In the main, the decision for therapy is an individual one. In those where
chemotherapeutic agents are required to achieve cytoreduction, α-interferon
may be the drug of choice as it does not cross the placenta, although it can be
present in breast milk.
In those subjects taking hydroxycarbamide, when pregnancy is planned, α-
interferon should be substituted 3–6 months prior to a planned conception. It
has been claimed that α-interferon improves pregnancy outcome in those with a
previous poor pregnancy, but this is by no means proven.
The role of platelet-pheresis is poorly defined, as it is not clear whether it influ-
ences outcome, but it is well tolerated and may be useful when an acute reduction
in platelet numbers is required. However, multiple (often twice weekly) treat-
ments are usually required to sustain a reduction in the platelet count.
Antiplatelet agents may be used to reduce the risk of placental thrombosis
and may also assist in the prevention of maternal thrombosis. This may be
particularly useful when there are symptoms of microvascular thrombosis, but
may exacerbate any platelet defect or acquired von Willebrand’s deficiency.
Heparin may be employed as thromboprophylaxis in those subjects with a pre-
vious vascular occlusion, but its role in the prolongation of pregnancy is not
known. In addition, the necessity and timing of thromboprophylaxis in asympto-
matic subjects remains unknown.
Paroxysmal nocturnal hemoglobinuria (PNH)

PNH is rare, but can present in women of childbearing age. It is associated with
episodes of intravascular hemolysis resulting in hemoglobinuria as well as
anemia, thrombocytopenia and leukopenia (Table 3.10). It is also associated with
Table 3.10 – Paroxysmal nocturnal hemoglobinuria (PNH) features
Intravascular hemolysis
Thrombocytopenia
Leukopenia
Median survival 10–15 years
Thrombosis/bleeding/infection/leukemia/aplasia/remission
GPI anchor CD 59 on RBC/PMN
Sensitivity to complement
Washed cellular products for transfusion
Significant fetal and maternal morbidity/mortality
CD, Cluster of differentiation; GPI, Glycophosphortidylinositol; PMN, Polymorphonuclear cell; RBC; Red blood cell
60
an increased risk of venous thrombosis, which classically occurs in the portal or

hepatic tracts, and may result in a peripheral deep vein thrombosis. The median
survival from diagnosis is 10–15 years, with the potential for fatal thrombosis,
thrombocytopenia-associated bleeding or infection. Very occasionally, the con-
dition may present or terminate with aplastic anemia or acute leukemia. There
are also reports, however, of spontaneous resolution.
PNH is caused by a clonal defect in the body’s own defences against the action
of complement. This defence is mediated by a number of proteins that are
attached to cell membranes by a glycophosphatidylinositol (GPI) anchor. An
acquired mutation in the PIG-A gene on the X chromosome results in defective
synthesis of this anchor, with the resultant loss of function of the proteins that
should be attached to it.
The disorder was in the past diagnosed by the Ham’s acid lysis test, which
demonstrates increased red cell lysis by the action of endogenous complement.
However, as it is recognized that this mutation affects all cells of the granulocyte
line (i.e. red cells, platelets and white cells (excluding lymphocytes)), the diagno-
sis is now made by demonstrating (using flow cytometry) a significant reduction
in one of the GPI-anchored proteins. This is usually achieved by demonstrating a
reduction in the expression of the cluster of differentiation (CD) molecule 59 on
either the red cell or the neutrophil.
Although there has been a great increase in the understanding of the
mechanism of PNH, there has been little advance in its therapy, with allogenic
bone marrow transplant being the only curative treatment. The mainstays of
routine therapy remain red cell and platelet components, anticoagulation and
the treatment of infection.
PNH in pregnancy
The management of PNH in pregnancy requires consideration of a number of
aspects of the disease. The thrombotic effect is likely to be due to an interaction
between white cells or platelets with complement, but may also, at least in part,
be mediated through circulating pro-coagulant microparticles, which are also
found in antiphospholipid syndromes. Thus, there may be similar processes and
consequences in both disorders.
Therapeutic anticoagulation is needed in those who have previously experi-
enced thrombosis. The role of thromboprophylaxis in those who have not yet
had a thrombosis is not known, but we would recommend this in all cases. As
with other disorders requiring anticoagulation, this is best achieved in pregnancy
with therapeutic or prophylactic dose, low-molecular-weight heparin (LMWH)
antepartum, and with warfarin or continued LMWH (avoiding the need for inter-
national normalized ratio (INR) monitoring) in the puerperium.
PNH is also associated with thrombocytopenia, and there may be an exaggera-
tion of the normal thrombocytopenia of pregnancy. It has been suggested that
the platelet count should be maintained at >30 × 109/l throughout pregnancy.
61
This may be difficult to achieve and will require washed platelet component
therapy to reduce the risk of exposure to exogenous complement. There is little
evidence that this is required and in the absence of bleeding problems and with a
platelet count >10–20 × 109/l, the present authors would manage the patient con-
servatively. As with other conditions (see Chapter 8), an adequate platelet count
is, however, required for delivery or Cesarean section, thus washed platelet
preparations are best reserved for delivery.
Epidural analgesia may be problematic from both thrombocytopenia and anti-
coagulant viewpoints, and so may be unavailable to such women. In all events, a
planned delivery will most often be required to ensure adequate hemostasis.
Anemia is common, and also requires the provision of washed red cell com-
ponents. The development of severe anemia may also have a bearing on fetal
oxygenation and development. Following delivery, the combined oral contracep-
tive must be avoided because of the thrombotic risk.
The maternal mortality may be of the order of 20% in cases of PNH during
pregnancy. Pregnancy may be associated with a substantial requirement for
maternal platelet and red cell component therapy, and an increased risk of both
venous thrombosis and infection. Although successful pregnancies have been
reported, the fetal mortality may be as high as 10%, with 50% delivering pre-
term, and there is also an increased risk of fetal growth restriction in those
mothers with severe anemia.
Acute leukemia in pregnancy

Although relatively rare, acute leukemia may present for the first time in preg-
nancy. Of those that do, the majority will be acute myeloid leukemia (around
70%), with the remainder of acute lymphoblastic type (Table 3.11). Affected sub-
jects often present with symptoms of anemia or infection. In general, such
leukemias have a long-term survival of around 50% and initial remission is likely
in around 75%. Without treatment there will be, however, a rapid deterioration,
with death frequently occurring within a few months of diagnosis. Overall, the
pregnancy has no effect on the leukemia and there is no evidence that a termina-
tion will improve the maternal outcome from the disease. The presence of preg-
nancy may however delay treatment, as there is concern that chemo- or
radiotherapies will have a deleterious effect on the fetus. This is most likely to be
the case if treatment is administered during the first trimester, when the risk of
teratogenesis from acute leukemia therapy has been suggested to be of the order
of 20%. It seems unlikely that treatment after the first trimester will lead to
significant fetal abnormality, although a transient fetal myelosuppression may be
evident after treatment in later pregnancy. Overall, miscarriage, growth restric-
tion and stillbirth is probably higher in affected pregnancies, but whether this
results from the maternal condition or the treatment is unknown. In general, a
termination is often offered to those mothers presenting in the first trimester. If
the pregnancy is continuing, a delay in therapy should be avoided, but may be
62
Table 3.11 – Leukemia and lymphoma in pregnancy
Acute leukemia 70% of acute leukemias will be AML

In this group there is a 40–50% long-term survival
Patients present with pancytopenia
Chemotherapy in the first trimester may run the risk of teratogenesis (particularly
methotrexate)
There may be an increased risk of IUGR/stillbirth
If the patient has thrombocytopenia, bleeding at delivery is possible
Lymphoma Hodgkin’s disease is probably more likely to present than NHL

Hodgkin’s disease: treatment delay/modification may be possible
NHL: full treatment usually required
If chest radiotherapy required then fetal shielding is effective
Pregnancy possible if previous radiotherapy to the abdomen, unless it was radical
uterine radiotherapy
There may be few late effects on children exposed to chemo- or radiotherapy in utero
AML, Acute myeloid leukemia; IUGR, Intrauterine growth restriction; NHL, Non-Hodgkin’s lymphoma.
contemplated (dependent upon maternal health) until the beginning of the

second trimester. In all cases, treatment should be instigated as soon as is prac-
tical and methotrexate (a methylene tetrahydrofolate reductase inhibitor used in
acute lymphoblastic leukemia treatment) should be avoided, as it may have a pro-
found affect on fetal development. Presentation in the late third trimester may
also lead to a delay in treatment being considered until after delivery. In such
cases, delivery may have to take account of the risks of maternal hemorrhage and
infection, but there are a number of reports of a successful pregnancy outcome
in such cases.
Lymphoma in pregnancy
The co-occurrence of lymphoma and pregnancy is relatively rare (Table 3.11). As
with the treatment of acute leukemia, management should consider the maternal
wishes, the type of disease, the likelihood of cure, the gestation at diagnosis, the
possibility of fetal effects and the possibility of collection of ovarian material for
storage. In the vast majority of lymphomas there is no evidence that the lym-
phoma itself influences pregnancy, or that the pregnancy has an adverse influ-
ence on the lymphoma.
When Hodgkin’s disease presents in the first trimester, a number of options
should be considered. Where there is aggressive disease, therapy may be required
immediately and the possibility of termination should be discussed. When, as is
more common, there is a lower grade of disease there are a number of possible
63
therapeutic options. These include monitoring the disease without therapy, the
use of single agent therapy (such as vinblastine) to control the disease until
definitive treatment is considered appropriate, use of a modified chemothera-
peutic regimen, full therapeutic combinations, or radiotherapy (where appropri-
ate) with fetal shielding. Non-Hodgkin’s lymphoma (NHL) usually presents later
and often with more aggressive disease. In most cases, immediate standard-dose
therapy is required and should not be delayed. However, if either NHL or
Hodgkin’s disease presents in late pregnancy, there may be scope to withhold
treatment until safe delivery can be achieved.
There is no suggestion that diagnostic radiation or nuclear diagnostic tests
themselves should be an indication for termination. Indeed supradiaphragmatic
irradiation, including spiral computerized tomography (CT) scanning, does not
expose the fetus to a significant radiation dose and is considered safe for the
fetus. Indeed, despite the caution expressed about chemotherapy (especially
when given in the first trimester), long-term follow-up of offspring of those who
received treatment for acute leukemia, Hodgkin’s disease or NHL, does not show
any evidence of an adverse effect on intellectual development, growth or the
occurrence of malignancy.
In subjects who have been previously treated for lymphoma there is an
increased risk of ovarian failure and premature menopause. This is particularly
the case in women treated after the age of 25 years with a combination of
chemotherapy and infradiaphragmatic radiotherapy. However, even those who
have undergone high-dose regimens may conceive (with hormone replace-
ment) with a successful outcome. Abdominal radiation affecting only a part
of the uterus does not preclude pregnancy, but when radiotherapy has
involved the uterus, early delivery and growth restriction may occur in sub-
sequent preg-nancies. At present, the significance of these observations is not
clear. After radical radiotherapy the uterus is unlikely to support conception,
even if there is satisfactory ovarian function. In general, after treatment for
lymphoma, patients are advised to wait 1–2 years before attempting pregnancy.
Whether this reduces risk is not clear, but appears sensible in the context of
possible relapse.
Placental blood banking

It has been recognized for at least 30 years that hematopoietic transplantation
was possible using fetal blood (50–150ml) obtained from the placental cord. This
is possible due to the presence of mononuclear cells in the cord blood which can
restore hematopoiesis in the treatment of a number of hematological malig-
nancies, platelet and sickling disorders. Although many of the ethical and
consent issues surrounding this modality of therapy are not yet resolved, there is
an increasing demand for the provision of such cord stem cells for the manage-
ment of hematological and, in the future, non-hematological conditions
(Table 3.12).
64
Table 3.12 – Considerations in cord stem cell collection
Ethics/consent
Viral safety
Genetic disorder exclusion
Microbial safety
Traceability
Storage safety and records
Dose suitable for pediatric cases
? Graft rejection
? GVHD
GVHD, Graft-versus host disease
The criteria for the selection of maternal donors for allogenic donation are
similar to those employed by most blood banking organizations, but with addi-
tional information required regarding past obstetric history and the possibility of
transmission of genetic defects. A number of practical issues, such as the safety
and yield of cells obtained when the cord is clamped early, or when the collec-
tion is taken prior to delivery of the placenta, are yet to be resolved. In addition,
the effect of Cesarean, rather than vaginal, delivery on the quality of cells also
requires further study. Overall, the collection appears to have little impact on
staff time, with often only an additional 10 minutes required for such collections.
The testing of such donations covers similar areas to that of blood donation,
including statutory virological screening (including hepatitis A, B and C, and
HIV, human T-cell leukemia/lymphoma virus (HTLV) and cytomegalovirus
(CMV)) and microbiological culture. The problem of window period infections
and the absence of the safety net that is presumed to come from those who have
successfully donated before, may all have an impact on the implementation of
such placental banking programs.
In addition, there is the requirement (as with hematopoietic stem cells derived
from adults) to determine the viability (from progenitor cell culture) and stem cell
component of each collection. It has been supposed that the human leukocyte
antigens (HLA) match may be less stringent with cord stem cells, although further
work on this, and the exclusion of maternal cells from the collection, is required.
At present, cord banking remains an expensive source of hematopoietic stem
cells, which, like cells derived from adults, requires a substantial infrastructure to
maintain safety and traceability. To date, a substantial number of pediatric trans-
plants have been carried out using cord stem cells, although much work is
required to allow these collections to form the basis of transplants for adults.
Ultimately, the expense of the collection and storage procedures may be offset by
the greater amount of transplant material available from this source and (if con-
firmed) the requirement for less stringent HLA matching and lesser likelihood
of graft-versus host disease with such transplants.
65
CHAPTER 4
Bleeding disorders in
pregnancy
Hemophilia A
The gene coding for Factor (F) VIII is found on the X chromosome at position
q28, and heritable deficiency of FVIII occurs as an X-linked recessive disorder.
The frequency of hemophilia A is between 1 in 10 000 and 1 in 20 000 of the
population. Although it has a wide distribution amongst different ethnic groups,
it is rare in Chinese populations. The disorder results from a variety of genetic
mutations, although 40% of severely affected individuals have an inversion
involving intron 22 of the gene. By contrast, the majority of mild disease results
from point mutations.
Clearly, this is a disorder predominantly affecting males; females will usually
be heterozygous and therefore can only be carriers. For a karyotypically normal
woman to have a hemorrhagic problem with hemophilia A, she would have to be
either homozygous for the mutation (with inheritance of affected X chromo-
somes from both an affected father and a carrier mother – an exceptionally rare
occurrence), or demonstrate significant lyonization (see below), with resulting
low FVIII levels. The latter is not unusual in carriers, but virtually all carriers in
this situation have levels of FVIII above that which would result in spontaneous
bleeding problems. Moreover (see Chapter 1), the vast majority of such reduced
FVIIIc levels will normalize during pregnancy. Other mechanisms of clinical
FVIII deficiency in a female include an acquired antibody to FVIII, a hemophilia
carrier with Turner’s syndrome (XO), and autosomal dominant hemophilia A
(due to a somatic gene affecting FVIII synthesis).
Hemophilia A tends to follow a similar pattern in affected family members,
reflecting the particular genetic mutation associated with hemophilia A in their
family, although in 30% of cases there may be no known family history. Daugh-
ters of an affected male are obligate carriers of the disease. Most of these will be
unaffected unless there is extreme lyonization. Lyonization is the random inacti-
vation of the X chromosome that occurs in all cells. If, by chance, this inactiva-
tion affects more of the normal X chromosome in FVIII-producing cells, this will
produce a mild hemophilia state in the affected female carriers. A carrier female
has a 50:50 chance of passing hemophilia onto her son and a 50:50 chance of
passing the carrier state onto her daughter.
66
Bleeding disorders in pregnancy
Presentation and treatment of affected males

Hemophilia results in a lifelong tendency to bleeding. In a severely affected male,
symptoms are often first noted when he begins to crawl or walk. It is unusual to see
bleeding problems in the neonate, including bleeding from the umbilical stump.
Symptoms are directly related to the level of remaining FVIIIc activity. A level
of <1IU/dl leads to severe spontaneous bleeding, 1–5IU/dl to moderate bleed-
ing only after minor trauma and 6–40IU/dl leading to mild bleeding only after
major trauma.
The first episode of joint bleeding (hemarthrosis) often leads to future bleed-
ing into the same joint (a target joint), resulting in arthropathy and joint failure.
Any joint may be affected, although the knee is the most common site, with
ankles, elbows, hips, shoulders and wrists also affected. Recurrent intramuscular,
subcutaneous and intraperitoneal hematomas may occur. There may also be
bleeding from the gastrointestinal (GI) tract, and bleeding after major trauma
and dental extraction is a common problem. In such circumstances delayed
bleeding is also common.
The diagnosis may be suspected if there is a prolongation of the activated
partial thromboplastin time (APTT), although a normal time does not exclude
mild disease. The diagnosis can be confirmed by determining the FVIIIc level by
an APTT- or chromogenic-based test (a chromogenic test determines the level of
FVIIIc by its ability to generate FXa from FXc, with the amount of FXa released
detected by a quantifiable color change in the reaction mix). In all cases of
severe hemophilia the level of FVIII antigen (determined by enzyme-linked
immunosorbent assay (ELISA)) will also be reduced. In milder cases, there may
Table 4.1 – Diagnosis of hemophilia
Hemophilia A Hemophilia B Hemophilia C von Willebrand’s Lupus

disease inhibitor
PT √ √ √ √ √/
APTT /√ /√ /√ Mild /√
TCT √ √ √ √ √
APTT APTT APTT APTT APTT Poor/

50:50 mix corrects corrects corrects corrects no correction
FVIIIc √ √ Mild –
FIXc √ √ √ –
FXIc √ √ √ –
vWF √/ √ √ –
APTT, Activated partial thromboplastin time; F, Factor; PT, Prothrombin time; TCT, Thrombin clotting time; vWF,
von Willebrand’s Factor.
67
be a disparity between antigen and activity (i.e. activity that is less than antigen
level), indicating the presence of a functional genetic defect. In most severe cases
the only likely alternative diagnosis is either hemophilia B (FIX deficiency) or C
(FXI deficiency). Thus, the initial diagnostic work-up should include an assay of
FVIIIc, FIXc and FXIc. In milder cases, von Willebrand’s disease (vWD) may also
have to be excluded (Table 4.1). When the disease presents with the incidental
finding of a prolonged APTT, a specific factor inhibitor (see below) or a lupus
inhibitor (lupus anticoagulant) may have to be excluded.
General principles of treatment

In general, hemophilia treatments should be initiated under the supervision of a
physician experienced in the management of hemophilia. In the UK there is a
system of regional hemophilia centers, and patients with hemophilia will usually
attend these centers or associated hospitals. In all cases, if a patient presents for
care, the center responsible for the patient’s hemophilia care should be con-
tacted to determine the treatment product usually used and to ascertain if the
patient is known to have a FVIII inhibitor, as this may influence the dose
required to achieve the desired increment.
Factor replacement
Replacement with factor products in hemophilia A should be achieved with
recombinant or high purity FVIII preparations where possible, with a dosage
based upon the plasma concentration of FVIII required and upon the body
weight or plasma volume. To calculate the required dose by weight the following
formula can be used: (body weight [kg] × desired rise in FVIII [IU/100ml])/2.
The amount of FVIII administered is expressed in IU: 1IU of FVIIIc equals the
quantity of FVIIIc in 1ml of normal human plasma. Empirically, 1IU of FVIII/kg
of body weight increases the plasma FVIIIc level by 1.5–2.0IU/100ml. After each
dose a peak level should be checked and further doses given, if required, to
maintain the target level (Table 4.2). When there has been bleeding, dose calcu-
lation by plasma volume may be more appropriate. When required, the plasma
volume can be derived from the patient’s hematocrit and their blood volume
(which can be estimated as 7% of the body weight). For most clotting factors, the
decline of therapeutic levels occurs in two phases: a rapid loss followed by a
slower decline. For FVIII infusions, around 80% of the dose can be recovered ini-
tially in the plasma, leading to an initial and subsequent half-life of infused FVIII
of 6 and 12 hours, respectively. This leads to a requirement to dose every 8–12
hours.
Because of repeated exposure to allogenic FVIII, the development of acquired
inhibitors to FVIII is not uncommon in hemophilia A. In such circumstances,
cross-reactivity of the antibody with porcine FVIII should be determined. Where
there is little or no cross-reaction, porcine FVIII products can be used. This
68
Table 4.2 – Factor (F) VIII dosing in hemophilia A (HA) and hemophilia B (HB)
Target FVIIIc (IU/dl) Dosage

Bleeding or FIXc (U/dl) frequency Dose duration
Minor surgery 30–60 1/day ≥ 24 hours
Major surgery 80–100 1–3/day Pre- and post-operative until

adequate wound heal, then
30–60 IU/dl for <7 days
Early hemarthrosis 20–40 1–2/day HA ≥ 24 hours until

Early muscle bleed 1/day HB resolution
Oral mucosa bleed
Large hematoma 30–60 1–2/day HA 3–4 days until

Severe hemarthrosis 1/day HB resolution
Life threatening 60–100 1–3/day Bleeding stopped

(e.g. head injury)
product is itself, however, associated with the possibility of antibody formation

and mild thrombocytopenia is also not uncommon during dosing.
Desmopressin analog, DDAVP

In mild hemophilia A, all cases should have an assessment of the effect of
DDAVP. DDAVP, when infused intravenously over 20 minutes at a dose of
0.3µg/kg diluted in 50–100ml of saline, will increase plasma FVIIIc and von
Willebrand Factor (vWF) by 2–5 times above the basal level in around 30–60
minutes.
Patients with a basal level of FVIIIc or vWF of 10–20IU/dl are most likely to
have a useful therapeutic response to DDAVP. In such subjects, the infusion can
be repeated 12–24 hourly as required, although there is often a marked diminu-
tion of response after around 3 days (presumably due to depletion of vWF or
FVIIIc stores, i.e. tachyphylaxis). DDAVP is therefore suitable for minor bleeds or
minor surgical procedures where prolonged bleeding (>3 days) is unlikely. The
drug is also available for subcutaneous or intranasal use. The intranasal spray
gives a dose of 150µg/ml and should be given as a single dose of 300µg. Side
effects of intravenous DDAVP include hypotension, flushing (which is common
and appears to be due to prostacyclin release), tachycardia, headache, and,
rarely, volume overload and hyponatremia. On repeated use it may be worth
checking urea and electrolytes, and the product should be used with caution in
those at risk of cardiac overload, fluid retention or hypertension.
Whether the increase in vWF and FVIII induced by DDAVP carries a throm-
botic risk is not proven, but it should be used with caution in individuals with ath-
69
erosclerosis or a high risk of thrombosis. When using DDAVP a further sample

should be obtained 3–6 hours after dosing to determine that an adequate dura-
tion of response has occurred. Theoretically, DDAVP used at delivery could carry
an increased risk of fetal hyponatremia. If this is a concern, then restricting use
until after clamping the placental cord will avoid such a problem.
Antifibrinolytic agents
For minor surgery, antifibrinolytic agents (such as tranexamic acid) can be a
useful adjunct to other therapy, and may be particularly useful in dental surgery.
Tranexamic acid binds to plasminogen, inhibiting the binding of plasminogen to
fibrin. This leads to a reduction in fibrinolysis, and has a plasma half-life of
around 2 hours. It should be given in advance of any procedure and can be given
as an oral dose, or as a mouthwash for dental procedures.
Renal tract bleeding and/or renal failure are contraindications to such thera-
pies, as thrombosis in the renal tract may further impair renal function. Tranex-
amic acid should also be used with caution in pregnancy, as fibrinolysis is already
shut down by the physiological increase in plasminogen activator inhibitor (PAI)-
1 and PAI-2 (see Chapter 1).
Carrier detection in pregnancy

In an ideal world, potential carriers would be identified before pregnancy and
counseling offered. In general, the phenotype of hemophilia A remains constant
in families, and carriage of a mild defect should result in a mildly affected son. A
mother is an obligate carrier of hemophilia if she has an affected father. She is
also considered a carrier if she has had two affected sons, or if she has a family
history and one affected son. It is also likely if she has no family history but has
had an affected son.
A reduced level of FVIIIc is consistent with carriage of hemophilia, although,
especially if an individual is pregnant or taking hormone therapy, a normal level
does not exclude it. In suspected cases with a normal FVIIIc, it has been shown
that the ratio of plasma FVIIIc:vWF levels may help distinguish between carriers
and non-carriers, but this usually requires repeated examinations for confirma-
tion; and it is unlikely to give useful information in pregnant subjects due to the
marked physiological changes in both factors.
There are two types of genetic diagnosis that can be achieved: direct and indi-
rect evidence of a mutation. In severe hemophiliacs, direct detection of the intron
22 inversion, or of another common inversion in part of the intron 22 sequence
that is found telomeric to the gene, should be carried out. In mildly affected indi-
viduals, linkage analysis of polymorphisms can be carried out. This is the indirect
method, which requires information from several family members, one of whom
is required to carry the disease, and relies upon normally occurring variations in
the genetic nucleotide sequence that can be used to track the mutant gene. Track-
70
ing is achieved by identifying the effect of these variations on enzyme-restriction

sites in the gene in different family members. The polymorphisms often used are
those involving the CA-repeat in intron 13, or other intragenic polymorphisms
such as Bcl1, HindIII, Xba1, Bg1, Msp1 and Taq1. Which one is informative
depends upon its frequency in the ethnic population under study.
Alternatively, direct sequencing of the FVIIIc promoter, exons and splice junc-
tions will give diagnostic information in the majority of subjects.
Prenatal and pre-implantation diagnosis

In those many cases which are sporadic, elimination of hemophilia will not be
possible by prenatal diagnosis. However, prenatal diagnosis can be achieved by a
number of techniques. Of these, chorionic villus sampling (CVS) is commonly
used.
CVS can be performed between 11 and 13 weeks gestation and carries around
a 1% risk of fetal loss. Also, as it is carried out prior to the possibility of fetal
sexing by ultrasound, CVS may result in unnecessary risk to a female fetus. In
addition, CVS carried out earlier than 11 weeks can be associated with the devel-
opment of fetal limb abnormality. CVS is used to determine the fetal sex in the
first instance, with a result possible within 24 hours of sampling. Where appropri-
ate, a known mutation can then be detected by sequencing within 7–10 days.
When a number of other family members are available (including an affected
individual), restriction length fragment polymorphism can lead to a diagnosis of
carriage within a few days, as there is no need to culture cells before analysis. The
sample can be obtained by a transvaginal or transabdominal route. Genetic diag-
nosis by CVS is not, however, infallible, as there is always the possibility of conta-
mination of the sample with maternal tissue and the possibility of mosaicism.
In the second trimester, it is usual to offer fetal sexing by ultrasound at
around 18 weeks gestation. Amniocentesis can also be used to acquire fetal
genetic material. Cordocentesis to determine fetal levels of FVIIIc or FIXc can be
used between 18 and 20 weeks, but is often seen as a last resort, as it may lead to
the stress of a later termination. Cordocentesis also carries a 1–3% risk of fetal
loss. With such investigations, the possibility of immunization of RhD-negative
women should always be considered and anti-D prophylaxis offered. Interest-
ingly, the uptake of prenatal diagnosis and termination of affected pregnancies is
relatively low (35 and 50%, respectively), since many carriers of hemophilia, who
have family experience of the disorder, do not consider it to be sufficiently dis-
abling to justify termination.
In the future the diagnosis may be achieved by isolation of fetal cells from the
maternal circulation. At present this has been achieved by the sexing of fetal lym-
phocytes in the maternal circulation. As such lymphocytes are often long lasting
this technique may only be applicable to primigravida. Other techniques, such as
detection and genetic analysis of fetal normoblasts or trophoblast in the maternal
circulation, are still in development.
71
The technology of assisted conception now offers the opportunity for pre-
implantation genetic diagnosis, so avoiding the need for termination. Following
ovulation induction and transvaginal ultrasound-guided oocyte retrieval, in vitro
fertilization is performed, and single-cell biopsy at the 4–8 cell stage allows deter-
mination of the embryonic sex (usually by fluorescent in situ hybridization). This
allows the potentially affected developing male embryos from a carrier mother,
or the potential carrier female embryos of an affected male father, to be dis-
carded and only unaffected embryos replaced, so avoiding the birth of an
affected male or carrier female. The success rate for pregnancies following this
procedure is consistent with other assisted reproduction techniques, and cur-
rently is of the order of 25–30%. There appears to be no developmental problem
with the embryo when biopsied at this stage. This technique is now available in a
limited number of centers in Europe and North America. In the future, more
precise genetic testing on single cells may become more readily available, which
will allow determination of whether the embryo carries the particular hemophilia
gene for that family.
Antepartum carrier management

The levels of vWF and FVIIIc increase with gestation (see Chapter 1). On
account of this, replacement products are rarely required in hemophilia A carri-
ers antepartum. However, when early complications such as miscarriage or
ectopic pregnancy occurs, or a procedure such as amniocentesis or CVS is
required in a carrier, problems are occasionally seen. This is due to the lack of
time for the physiological increase in FVIIIc to have occurred. If there is the
potential for use of blood products (including recombinant products) during the
pregnancy, then the woman should have her immunity to hepatitis A (HAV) and
B (HBV) checked. If necessary, she should be immunized against HBV (and
probably HAV), subcutaneously. It is safe to immunize her against HBV during
the pregnancy. As the HAV vaccine is inactivated, it is also probably safe to use in
pregnancy, but an assessment of the risk:benefit ratio depends upon the risk of
contamination of HAV by locally available blood and blood products. FVIIIc
levels should be checked in all carriers at the first antenatal visit and again in the
third trimester. A level of 40IU/dl is usually sufficient to avoid hemorrhagic
problems in early pregnancy and for a vaginal delivery. A level of ≥50IU/dl is
required for diagnostic procedures, such as amniocentesis, Cesarean section or
epidural anesthesia. Despite a lack of significant oxytocic effects, DDAVP is often
avoided antepartum in mothers who have not demonstrated an adequate
increase in FVIIIc levels. This is due to its association, in repeated doses, with
maternal hyponatremia. However, it is usually preferable to blood products and
it is unusual for repeated doses to be required. Moreover, it should be remem-
bered that repeated doses are associated with tachyphylaxis and are less likely to
give a useful clinical response. In the small number of carrier cases, where factor
treatment is required, this should be with recombinant factors if at all possible.
72
Management of delivery and puerperium

On admission to the labor suite, a full blood count, coagulation screen, appropri-
ate factor level assessment (unless these have previously been normal) and a
cross-match sample for ‘group-and-save’ should be taken. Intramuscular analge-
sia should be avoided if the mother has a FVIII level <40IU/dl, but such therapy
can be given intravenously.
The management of delivery must consider the possibility of fetal hemophilia.
Even if the fetus is a female, it is important to remember that she may still be at
risk of bleeding because of the combination of lyonization and the immaturity of
the fetal liver with reduced coagulation factor production. In an attempt to
‘boost’ production of vitamin K-dependent coagulation factors from the fetal
liver, many authorities prescribe vitamin K supplements (20mg/day) to the
mother from 36 weeks gestation (or earlier if premature delivery is considered a
possibility). Generally, vaginal delivery is preferred for known hemophilia carri-
ers, with few obstetricians recommending Cesarean section purely on the basis of
suspected (or known) fetal hemophilia. However, the main concern is of fetal
intracerebral bleeding in the course of delivery. Where the fetus may be a carrier
female or an affected male, it is prudent to avoid procedures that would place
the fetus at increased risk of bleeding. ‘Difficult’ forceps deliveries, rotational
forceps, ventouse extraction, scalp electrodes or fetal scalp blood sampling
should be avoided. A straightforward mid- or low-cavity forceps delivery is usually
considered to carry a reasonable risk. If, however, there is a delay in labor and an
easy vaginal delivery seems less likely, the mother should be delivered by
Cesarean section.
As pregnancy FVIIIc levels fall quickly to pre-pregnancy levels post-partum
hemorrhage is possible, and FVIII levels should be maintained >40IU/dl for at
least 3–4 days after vaginal delivery and for 4–5 days following Cesarean section.
However, it is unusual for carriers to require postpartum management with
DDAVP or FVIII concentrates. DDAVP can be used, if required, in breastfeeding
mothers, as it does not cross the breast.
For the newborn, a cord determination of FVIIIc levels, on blood aspirated by
clean venepuncture form the umbilical vein, should be obtained in all suspected
or known cases if possible. Vaccines should be given by the subcutaneous route,
and intramuscular injections and heel stabs (such as for the Guthrie test)
avoided until the level of clotting factors is known. In such cases, vitamin K pro-
phylaxis is routinely given by the oral rather than intramuscular route. Follow-up
for determination and, if required, treatment must be organized with the appro-
priate pediatric hemophilia center.
Difficulties arise when a sporadic case of hemophilia unexpectedly presents
with an intracerebral hemorrhage following vaginal delivery. In this context it is
important to remember that approximately one third of cases of hemophilia are
due to new mutations. In sporadic cases, hemophilia in the child may not be
immediately considered and the mild prolongation of the APTT (which occurs
73
in moderately affected individuals) may be interpreted as being within normal

limits for the newborn. In all suspected cases, specific assays for FVIII should be
carried out. Early prophylaxis after birth is often carried out if there has been an
instrumental delivery. In some areas, serial brain ultrasounds are used to screen
for intracerebral hemorrhage; screening for congenital dislocation of the hip is
carried out with caution, or postponed until the results of factor levels are
known.
Hemophilia B
Hemophilia B is another X-linked bleeding disorder, which, although consider-
ably less common than hemophilia A, is more often associated with a family
history. Hemophilia B may result from mutations that lead to a reduction in FIX
function, accompanied by either detectable, or reduced, FIX antigen levels. The
wide variety of mutations of the FIX gene includes those that alter carboxyl
group formation (carboxyl groups on vitamin K-dependent factors lead to
binding of the coagulation factor to activated phospholipid at the site of coagula-
tion, accelerating the coagulation response) and those that alter the activation of
FIXc by FXIa. In addition, a mutation in the FIX promoter region may lead to a
marked reduction in antigen at birth, but increasing production occurs with
increasing age (hemophilia B Leyden).
Severe FIX deficiency is considerably less common than severe hemophilia A,
but presents with similar problems as hemophilia A subjects with comparable
factor levels. The diagnosis of hemophilia B can be achieved by determination of
FIXc in an APTT-based assay.

As with hemophilia A, treatment of hemophilia B should be supervised by a
physician experienced in hemophilia management and in collaboration with the
center responsible for the patient’s long-term care.
When FIX is infused, only 50% is recovered initially, and the initial and sub-
sequent half-lives are 3 and 24 hours, respectively. This means that FIX treatment
requires a loading dose for immediate effect (even in minor bleeds), but sub-
sequent doses can be given at 24-hour intervals. The treatment of choice is
recombinant FIX, although this may require a higher dose than human-derived
equivalents (see Table 4.2). Empirically, infusion of 1IU of human-derived FIX
per kg of body weight increases the plasma FIXc level by 0.8% of normal activity.
The required dose to be given can be calculated by: (body weight [kg] × the
desired rise in FIXc [%]) × 1.2. If not available, other products, such as pro-
thrombin concentrates (usually containing FIX, FX, FII +/-FVII), may have to be
used. The dose is again calculated from the level of FIXc in such products. Aside
from these factors, the principles of therapy and management in pregnancy are
the same as for hemophilia A (see Table 4.2).
74
In subjects with complete absence of the FIX antigen, allergic reactions to

factor products can be problematic. In such cases, therapy with recombinant
FVIIa is a possible alternative. In the future it is hoped than gene therapy may
have a role in the management of hemophilia B.
Pregnancy management
There is substantial genetic heterogeneity in cases of hemophilia B, with many
mutations being peculiar to an individual family. Confirmation of female car-
riage requires knowledge of the mutation within the family to confirm the diag-
nosis. In general, to be certain that an identified mutation causes hemophilia, it
is recommended that the clinician confirms that the mutation is already included
in one of the international databases of hemophilia mutations. This is necessary,
as a number of mis-sense mutations result in no observable change in the func-
tion of the FIX molecule. Furthermore, where there is a sporadic case, the
mother may have a mosaic (a mixture of native and mutated DNA in the same
germ cell) and carrier status cannot be confirmed by assessment of her somatic
cells.
Female carriage is likely if there is a reduced level of FIX activity (which often
results in a more clear-cut distinction from non-carriers than FVIII carriage in
hemophilia A), but assessment of FIX antigen levels is more problematic, as
there may be a considerable overlap with non-hemophilic subjects. Although FIX
activity is higher in women on the combined oral contraceptive pill, it does not
change substantially with increasing gestation, and those carriers with a low level
tend not to ‘nomalize’ in pregnancy, in contrast to hemophilia A carriers.
All carriers should have an assessment of FIX levels at their first visit in preg-
nancy and in the third trimester. In more markedly affected carriers, hemostatic
support is more likely to be required during pregnancy than with equivalent
hemophilia A carriers. Such support should be with recombinant factor prod-
ucts, as DDAVP is usually ineffective in FIX deficiency. As with FVIII carriers, a
FIX level of at least 40 U/dl is required for a vaginal delivery, and a level of at
least 50 U/dl for Cesarean section or epidural. Levels should be maintained for
the first 4–5 days after delivery.
The methods of prenatal diagnosis of FIX deficiency, and the management of
the delivery and the postpartum phase, follow the same principles and tech-
niques as those used for cases of FVIII deficiency. To facilitate restriction frag-
ment length polymorphism (RFLP) detection, Taq1, Xmn1, Dde1, Hha1, Mnl1
and Mse1 are informative in around 90% of cases in Caucasians.
Hemophilia C
Hemophilia C is characterized by a reduction in coagulation FXIc. Unlike hemo-
philia A and B, it is inherited as an autosomal recessive trait. In homozygotes, the
FXIc level may be <20 U/dl: in heterozygous subjects the levels range between 20
75
and 70 U/dl. The disorder has a prevalence in the general population of around
one in one million, but occurs at a higher frequency in the Ashkenazi and Iraqi
Jew populations, where up to 0.3% of subjects are homozygous for the deficiency
and there is a gene prevalence of up to 11%.
The gene coding for FXI is found on chromosome 4, with at least 28 known
mutations. These can result in an alteration of gene splicing (Type I mutations),
the production of a stop codon (which leads to FXIc of <1U/dl), or a minor
amino acid substitution (Type III mutations). Of these, mutations associated with
the production of a stop codon (so-called Type II mutations) are most often asso-
ciated with bleeding. In the Ashkenazi Jew population, two mutations are
responsible for most cases, and compound Type II/Type III heterozygotes are
the commonest cause of moderate to severe disease. In other populations there
is a greater variation in the mutations involved.
Unlike hemophilia A and B, bleeding is more difficult to predict in subjects
with hemophilia C. Generally, hemarthrosis is uncommon and spontaneous
bleeding is rare. Furthermore, bleeding does not correlate well with residual FXI
activity and around 50% of heterozygous subjects may have a bleeding tendency.
By contrast, subjects with severe deficiency may not have a severe bleeding tend-
ency. In many affected individuals bleeding only occurs following surgery or
trauma. The possibility of delayed hemorrhage should be considered, especially
in oral, ear, nose and throat (ENT), gynecological and urological surgery, where
there is high local fibrinolytic activity.
FXI deficiency may prolong the APTT, although this is only likely to detect
homozygous subjects. In all suspected cases, a FXIc assay should be performed.
Management in pregnancy
As with other heritable bleeding disorders, pregnant subjects should be managed
in a center with expertise in the assessment and treatment of hemophilia.
There are a variety of studies that show no change in FXIc levels with gesta-
tion in normal individuals, although many show a gradual reduction. As there is
no good correlation between FXIc levels and bleeding in hemophilia C, close
monitoring of FXIc levels in pregnancy in hemophilia C subjects is of doubtful
value in any case.
Postpartum hemorrhage occurs in 20% of homozygotes, but excessive bleed-
ing may also occur in heterozygotes. In general, the majority of subjects with a
non-pregnant FXIc level of <15U/dl are likely to suffer excessive bleeding after
trauma or surgery, and any subject who has a history of excessive blood loss will
require hemostatic support for delivery. Even if specific hemostatic support is
thought not to be required, products should be available.
When therapy is needed, daily use of a FXI concentrate is probably justified,
with the aim of achieving a target FXIc level of no >70IU/dl (maximum dose of
30IU/kg). This level should be achieved during labor and maintained for 3–4
days after vaginal delivery and 4–5 days after Cesarean section.
76
The target level of FXIc should not exceed 100U/dl, as this has been associ-
ated with thrombosis, although it is not known whether there is an excessive risk
of thrombosis whilst using FXI concentrates in late pregnancy and the puer-
perium. Although many manufacturers have added heparin to FXI concentrates,
since the early reports of thrombosis, it is not yet clear that this has reduced the
thrombotic potential of these products. In subjects with a history of thrombosis
or atherosclerosis, FXI concentrate should probably be avoided. If hemostatic
support is required, then fresh frozen plasma (FFP) should be used (FFP should
be virally inactivated if at all possible).
If there has been no history of bleeding, tranexamic acid could be used to
cover delivery in the first instance. However, as noted earlier, fibrinolysis is
already impaired in pregnancy and tranexamic acid should be used with
caution in view of the potential for thrombosis. There should be early recourse
to FXI concentrate, if available, when treatment is needed. While FFP is an
alternative, the use of FFP in such circumstances is complicated by the relat-
ively large volume required and the variable half-life of the coagulation factors
within it. Similarly, if, in an emergency, FXI products are not available, then
FFP should be used if hemostatic support is required. The role of recombinant
FVIIa is not yet certain, although it has been used effectively in subjects with
FXI inhibitors, and it may be preferable to the use of human blood products in
pregnancy.
If regional anesthesia is required, correction of the APTT has been advocated,
although this remains controversial. As with any other hemophilia, the possibility
of fetal carriage should be considered when planning fetal monitoring, forceps
and ventouse extraction at delivery.
Acquired hemophilia
Acquired hemophilia results from the development of an autoantibody, most
often against FVIII. Although classically associated with repeated treatment with
allogenic factor products in subjects with hemophilia, it can also occur in normal
healthy individuals, in subjects with autoimmune disorders (such as systemic
lupus erythematosis (SLE) and rheumatoid arthritis (RA)), in those with lympho-
proliferative and solid tumors, as well as in association with pregnancy. In non-
hemophilia subjects it is a rare and sporadic disorder, and the incidence has
been estimated at between 0.2 and 1 per million person-years. Of these, around
10% occur in relation to pregnancy.
Presentation in pregnancy
Pregnancy-related cases are often diagnosed within 3 months of delivery,
although presentation up to 1 year after delivery has been reported. Presentation
varies from bleeding after surgical procedures to spontaneous bruising, wide-
spread hematomas and GI bleeding. If untreated, acquired hemophilia carries a
77
substantial morbidity and mortality. Unlike inherited hemophilia, the bleeding

tendency does not equate well with the level of inhibitor, and although GI bleed-
ing is more common, joint hemarthrosis rarely occurs. In cases associated with
pregnancy, the presentation may often be with a postpartum hemorrhage,
although maternal fatality has been reported in association with retroperitoneal
and respiratory tract bleeding.
The diagnosis is suspected if there is unexpected bleeding and evidence of a
prolonged APTT which does not correct with the addition of normal plasma
(most easily achieved by testing a 1:1 mixture of the patient and normal plasma).
A prolonged APTT that corrects with normal plasma, indicates a clotting factor
deficiency. If the APTT fails to correct, presence of either a coagulation factor
inhibitor (antibody) or a lupus inhibitor (lupus anticoagulant) is indicated. The
degree of inhibitor level is assessed by the Bethesda assay. Acquired inhibitors
can be directed at any clotting factor, although FVIII inhibitors are most often
recognized.
Treatment in pregnancy
The management of these patients requires consideration of both the level of
inhibitor and the clinical problem. In those with a high level of inhibitor
(Bethesda >10U), the therapeutic options include porcine FVIII, recombinant
FVIIa (rFVIIa), or activated clotting factor concentrates (e.g. FEIBA®) (Table
4.3). The potential value of porcine FVIII can be assessed by laboratory determi-
nation of cross-reactivity of the inhibitor with porcine FVIII. With low-level
inhibitors, spontaneous resolution without immunosuppression often occurs. In
some cases there may be no response to high-dose FVIII, even when a low titer of
inhibitor is present: indeed, even high-level inhibitors can sometimes respond to
normal doses of FVIII.
The 2-year mortality has been estimated at 3%, with complete resolution of
the antibody in almost all survivors by this time. Indeed, commonly, resolution
occurs 6 weeks to 6 months after presentation. The use of immunosuppression
may shorten the duration of the disease, but it is not yet clear whether it influ-
ences mortality. From the little information that is available, it appears that there
is unlikely to be a recurrence in a future pregnancy. If a recurrence does occur,
antepartum bleeding appears to be uncommon. In the vast majority of such cases
the fetus is not affected, although sporadic case reports of low FVIII levels in the
infant of an affected mother and of intracranial hemorrhage have been
reported.

vWF is required to bind platelets to the subendothelium upon vessel injury. The
non-covalent binding of vWF to FVIII in the plasma also protects FVIII from
degradation. Indeed, without vWF, the half-life of FVIII is reduced to a few
78
Table 4.3 – Management of acquired hemophilia
Bethesda <5U High-dose human FVIII Spontaneous resolution Expensive

High-dose likely
recombinant FVIII
Bethesda >10U Porcine FVIII Good response likely, Risk of allergy/

little cross-reaction with anaphylaxis/
most inhibitors thrombocytopenia
Recombinant VIIa Good response likely Expensive

Risk of thrombosis
Activated clotting factor Response can be Risk of thrombosis

concentrate measured clinically only Risk of disseminated
intravascular coagulation
(DIC)
Additional therapy Corticosteroids

Cyclophosphamide
Human normal intravenous immunoglobulin (IVIgG)
Azathioprine
Plasma exchange
hours, leaving FVIIIc levels of between 1 and 10IU/dl. vWD results in a reduction
of vWF and is the most common heritable bleeding disorder, with a frequency
that has been estimated at around 1% of many populations. Many mild cases
remain undiagnosed, as subjects are often asymptomatic.
The gene coding for vWF is found on chromosome 12, and the vWF molecule
is produced both in the megakaryocyte (the precursor of platelets) and in the
endothelium. Upon manufacture, it is stored in the Weibel-Palade bodies of the
endothelium and in the α granules of platelets. Upon release into the circula-
tion, vWF is cleaved by a proteinase of the A disintegrin and metalloprotease with
thrombospondin domain (ADAMTS) family to achieve the correct balance of
vWF multimers (see Chapter 8).
A brief description of the various types of vWD is shown in Table 4.4. The vari-
ants range from a mild/partial deficiency to a severe deficiency; between these
two extremes are other variants. Some reduce the interaction of vWF with
platelets (associated with either reduced or normal levels of high-molecular-
weight vWF multimers). One produces a variant vWF that has an increased affin-
ity for platelets. This, somewhat paradoxically, results in bleeding associated with
thrombocytopenia and increased clearance of high-molecular-weight vWF multi-
mers. There is also a vWD variant that has a reduced interaction with FVIII: in
this variant the reduction in the half-life of FVIII contributes to the bleeding phe-
notype.
79
Table 4.4 – Classification of von Willebrand’s disease (vWD)
Defect Inheritance Bleeding Plasma Plasma RiCoF Multimer analysis

time vWF FVIIIc
Type 1 Mild deficiency Aut dom /√ /√ √/ √/ √

Type 2 platelet-dependent Aut dom /√ √/ High molecular weight
vWF function (Type IIa)
vWF affinity to platelet Aut dom √/ √/ √/ High molecular weight
glycoprotein 1b High- ( Platelet
80
molecular-weight multimers affinity)

(Type IIb)
Platelet-dependent Aut dom /√ √/ √/ √

function
vWF affinity to FVIII Aut dom /√ √/ √/ √
Type 3 Severe deficiency of vWF Aut rec Variable
Aut dom, Autosomal dominant; Aut rec, Autosomal recessive; F, Factor; RiCoF, Ristocetin cofactor; vWF, von Willebrand Factor.
Clinically, a reduction in vWF results in a defect of primary hemostasis. In

severe vWD, the accompanying reduction in circulating FVIII also results in a
mild–moderate hemophilia A phenotype. The problems of milder vWD can be
summarized as bleeding from surfaces. This includes mennorrhagia, bruising,
epistaxis, postpartum hemorrhage and bleeding after dental extraction. In such
individuals the diagnosis may not be made until the patient is an adult. In more
severe defects, the significant contribution from the deficiency of FVIII results in
hematoma and hemarthrosis formation. In such severely affected females, bleed-
ing may come to attention in childhood, with life-threatening epistaxis, and bruis-
ing and menstrual bleeding is often extremely problematic at the menarche.
Screening for defects in primary hemostasis often employs the ‘bleeding time’
in the first instance. This is the time taken for hemostasis (more properly, the time
when there is no further visible change in bleeding from the wound) after a cut
(ideally a pair of cuts, made in a uniform direction) has been made (by a standard
blade, usually a springloaded device that cuts at uniform length and depth such as
the Simplate II) on the forearm, where a blood pressure cuff is maintained at
40mmHg. The bleeding time is difficult to standardize and reference ranges are
not known for many clinical circumstances. In normal non-pregnant subjects the
bleeding time is <9 minutes, but it may not be prolonged in all subjects with vWD.
In all suspected cases, an assessment of the APTT, vWF antigen, vWF activity
and FVIIIc should be carried out. As FVIIIc and vWF increase in inflammation
and in response to estrogen, the diagnosis of mild vWD can be problematic. In
such cases, multiple evaluations are often required to confirm the diagnosis. vWF
antigen is usually determined by ELISA and function is most often assessed by
the ristocetin cofactor (RiCoF) activity of the patient’s vWF. Ristocetin activates
the patient’s vWF, which then binds to test platelets (not the patient’s) via
platelet receptor Glycoprotein 1b (GPIb). By contrast, ristocetin can also be used
to test the response of patient platelets to ristocetin. This test is used to diagnose
platelet GPIb deficiency (Bernard Soulier syndrome) and to detect the increased
reactivity of patient platelets to vWF, which is seen in Type IIb vWD (see Table
4.4), and pseudo-vWD (due to the high affinity of platelet GP1b for vWF). Other
tests of vWF function (such as FVIII binding) are limited to more specialized lab-
oratories. In all cases of reduced vWF, multimer analysis (by gel electrophoresis)
is required to allow subclassification of the disease.

Most Type 1 vWD, the commonest form of vWD, will respond to an infusion of
the vasopressin analogue DDAVP (see page 69). By contrast, Type 3 disease will
require infusion of a product containing vWF. Although such products also
contain FVIII they result in a longer survival of FVIII than happens in equivalent
hemophilia A patients, due to the stabilization of endogenous FVIII by the
infused vWF. In Type 2a vWD, DDAVP may occasionally be useful, although vWF
concentrates are often required. In Type 2b vWD there is often a mild thrombo-
81
cytopenia, which often worsens in pregnancy and may fall in response to DDAVP.
In such cases DDAVP should probably be avoided. In subjects who do not
respond to DDAVP, treatment with a plasma-derived factor concentrate that
contains both FVIII and vWF is required (an ‘intermediate purity’ FVIII concen-
trate). A rough guide to dosing is shown in Table 4.5. The adequacy of the dose
can be assessed by monitoring either FVIIIc or vWF RiCoF levels. In some Type 2
vWD, RiCoF assessment is more useful, as the baseline FVIIIc is within reference
limits. There is no practical role for monitoring the bleeding time, as it reflects
platelet vWF rather than vWF in the plasma.
As noted above, tranexamic acid may be a useful adjunct to other therapy,
and in particular has a role in dental procedures. Although cryoprecipitate con-
tains up to 10 times more FVIII and vWF than FFP, unlike concentrates it under-
goes no virucidal procedures, and the exact amounts of factor contained in each
dose is unknown. Correspondingly, it should only be used in an emergency when
a suitable intermediate FVIII concentrate cannot be obtained.
Where methylene blue (MB)-treated cryoprecipitate is available, this should
be used in preference to an untreated version. MB is a cross-linking agent that
destroys some of the known blood-borne viruses. In most recently developed
products, >95% of the added MB is removed during product manufacture. The
potential for MB treatment to reduce the level of FVIII and vWF in the product
should, however, be borne in mind when dosing this product. In cases of Type 3
vWD that do not respond to vWF concentrates, a combination of vWF concen-
trate and platelet component therapy may prove useful.
Antenatal management
All women with vWD should be managed in a unit with rapid access to specialized
hemophilia care. If the fetus is at risk of Type 3 vWD, prenatal diagnosis may be
Table 4.5 – Intermediate Factor (F) VIII dosing in von Willebrand’s disease (vWD)
Condition Required Dosage Dose

FVIII/ RiCoF (IU/dl) frequency duration
Surgery Minor >30 1/day ≥ 24 hours
Major >50 1/day On op day and post-op until

adequate wound healing
Dental extraction >30 One dose ≥ 12 hours

pre-procedure
Bleeding >30 1/day Until resolution
RiCoF, Ristocetin cofactor.

82
required. This requires planning to determine the mutation or RFLP, which may
be informative. Although vWF levels increase with gestation, it is important to
assess vWF and FVIIIc levels between 34 and 36 weeks gestation in all women with
vWD, to confirm that the levels have increased sufficiently to allow labor and deliv-
ery without the need for DDAVP or blood products. In subjects with mild–moder-
ate vWD (Types 1 and 2), vWF levels usually increase from the 6th to 10th weeks
of gestation, increasing 3–4-fold by the third trimester, making antenatal therapy
(as with hemophilia A carriers) unnecessary in most cases. Indeed, it appears that
significant antepartum hemorrhages and miscarriages are no more common in
such subjects. However, as with hemophilia A, early pregnancy complications may
require specific treatment to improve vWF levels.
As discussed above, DDAVP, if required, has no significant oxytocic effects, is
not associated with teratogenicity in animal studies and has been used in early
pregnancy to permit CVS or amniocentesis. In Type 3 vWD there is no rise in
vWF with gestation or DDAVP, and factor therapy may be required to cover inter-
ventional procedures or spontaneous bleeding. As noted above (page 70),
tranexamic acid is not absolutely contraindicated in pregnancy and is not associ-
ated with teratogenicity.
Management of delivery
A vWF or FVIII activity level of >40IU/dl or >50IU/dl, respectively, is con-
sidered safe for vaginal delivery or Cesarean section (or other procedures such
as amniocentesis or evacuation of the uterus). After delivery, the levels of vWF
rapidly return to pre-pregnancy levels and so, in general, vWD is associated
with a higher risk of primary and secondary postpartum hemorrhage. However,
in most Type 1 and 2 vWD, prophylactic therapy for vaginal delivery is not
needed, although treatment may be required for episiotomy or perineal tears
in Type 2 disease. In Type 1 vWD, epidural analgesia should not be undertaken
lightly, but with due regard to the risks of spinal hematoma. In those who are
known to respond to DDAVP, this can be used to cover Cesarean section.
Although there maybe a theoretical risk, it is not clear whether DDAVP use at
delivery carries any increased risk of fetal hyponatremia. If this is a concern,
then restricting use until after clamping the placental cord will avoid such a
problem. In Type 1 vWD there will be a fall in vWF after delivery, and discharge
planning in women with symptomatic disease should take account of the possi-
bility of secondary postpartum hemorrhage (PPH) at 4–5 days after delivery.
In any case, vWF and FVIIIc levels should be monitored in all cases in the first
days after delivery. After Cesarean section, DDAVP or factor products (if they
have been used) should be continued for at least the first 2–3 days of the
puerperium.
Type 3 vWD is, however, often associated with intrapartum and postpartum
hemorrhage. FVIIIc levels at delivery have been used as a guide to the necessity
for therapy, although the prediction of bleeding (as with any delivery) is
83
problematic. As a guide, if the patient’s FVIIIc level is <20IU/dl, therapy may be

required. In such cases, therapy should be given to cover the episiotomy, and
daily therapy after vaginal or Cesarian section delivery should be continued for
the first week of the puerperium. Simlarly, epidural anesthesis is not recom-
mended for Type 2 or 3 disease.
As with all bleeding disorders, where a fetus is at risk of inheriting vWD, deliv-
ery should be managed to minimize the risk of trauma to the fetus. In all sus-
pected cases, the umbilical cord venous level of vWF should be assessed with a
sample obtained by clean venepuncture: it should be noted that this is not reli-
able for the exclusion of Type 1 disease. For some forms of Type 2 disease an
assessment of cord FVIIIc is also required to establish or exclude disease.
Hypo- and Dysfibrinogenemia

Fibrinogen is synthesized by hepatocytes coded by three genes on chromosome
4. Fibrinogen functions as the precursor of fibrin clot, interacts with activated
platelets and acts as a plasma carrier for FXIII. Two types of congenital fibrino-
gen disorder are recognized: those where the fibrinogen antigen is produced
but is poorly functioning (dysfibrinogenemia), and those where there is a reduc-
tion (hypofibrinogenemia) or absence (afibrinogenemia) of the fibrinogen
antigen.
Afibrinogenemia
Although afibrinogenemia, usually an autosomal recessive trait, can be associated
with life-threatening hemorrhage, bleeding is usually less than that observed in
hemophilia A and B. Bleeding may occur, however, from the umbilical stump
and bleeding into joints is not uncommon, but bleeding from the GI tract and
the central nervous system (CNS) is infrequently seen.
In affected mothers, there may be an increase in the risk of first trimester mis-
carriage, placental abruption and postpartum hemorrhage. In such circum-
stances, maternal fibrinogen replacement is often required for the pregnancy to
proceed to term.
As fibrinogen is the final component of hemostasis, congenital afibrinogene-
mia can be suspected from a failure to clot in the prothrombin time (PT), APTT
and thrombin clotting time (TCT). This is accompanied by an absence of fib-
rinogen antigen and, in some cases, a prolonged bleeding time.
When treatment is required, this should be achieved by a virus-inactivated fib-
rinogen concentrate where it is available. Where it is not, cryoprecipitate may
have to be used (with the caveats noted above). Prophylaxis is indicated in preg-
nancy and fibrinogen levels should be restored to at least 1g/l. In afibrinogene-
mia, exposure to allogenic fibrinogen may result in allergy or anaphylaxis, and
restoration of fibrinogen levels has also been associated with the development of
thrombosis.
84
Hypo- and dysfibrinogenemia

Patients with hypofibrinogenemia and a fibrinogen level >0.5g/l are usually
symptom free. As noted above, patients with a disparity between fibrinogen
antigen and activity are described as dysfibrinogenemia. Such disparity may affect
any aspect of fibrinogen function and, as a result, around 40% present with
bleeding, around 40% are asymptomatic, and the remainder present with throm-
bosis or a combination of bleeding and thrombosis.
Dysfibrinogenemia may result in a prolongation of the TCT (and occasionally
the PT), or, occasionally, shortening of the TCT. In all cases, the diagnosis
requires the demonstration of a disparity between the fibrinogen activity and fib-
rinogen antigen levels. In many centers, a persistent ratio of >1.7 antigen:activity
is considered to be in keeping with the diagnosis.
Congenital dysfibrinogenemia is most often inherited as an autosomal domi-
nant trait, and should be distinguished from acquired dysfibrinogenemia associ-
ated with liver disease, malignancy and, disseminated intravascular coagulation
(DIC). The majority of congenital cases require no therapy, but cryoprecipitate
and antifibrinolytic agents have been used to prevent bleeding. Cryoprecipitate
has also been used as prophylaxis in subjects who have suffered previous recur-
rent miscarriages, although evidence for this use is lacking. Because of the associ-
ation of dysfibrinogenemia and thrombosis, many clinicians have a concern
about using fibrinogen concentrate. However, in view of the limitations and risks
of repeated cryoprecipitate exposure noted above, and the likelihood that more
modern fibrinogen concentrates carry a lower thrombotic risk than older prod-
ucts, this is probably not justified, particularly if the patient has a history of hem-
orrhage rather than thrombosis.
Factor (F) XIII deficiency

As detailed in Chapter 1, FXIII is required for the stabilization of the fibrin clot.
FXIII deficiency can occur as a very rare autosomally inherited disorder, which
has a prevalence of around one in one million of Caucasian populations, but
there is likely to be a higher prevalence where there is a large Asian population,
or where there is a high incidence of consanguinity. As with other factor defi-
ciencies, there is a relationship between the residual level of factor and bleeding
risk. Allowing for differences in assay techniques, a level of <1IU/dl is associated
with severe spontaneous bleeding and a level of 1–4IU/dl with a risk of moder-
ately severe bleeding: levels >5IU/dl are most often asymptomatic, but bleeding
episodes can occur. Affected individuals often present with bleeding from the
umbilical stump in the first few days after delivery. In general, inheritance of the
disorder produces a significant phenotype and, akin to more common hemophil-
ias, is associated with bruising, muscle bleeding and hemarthrosis. There is also a
lifelong increased risk of intracerebral hemorrhage, as well as bleeding after
delivery, trauma and operative procedures. FXIII deficiency can also result in a
85
delay in wound healing and has been associated with a high risk of recurrent mis-
carriage.
Diagnosis
In suspected cases, the PT, APTT and bleeding time will be normal. If suspected,
a test to detect an abnormal solubility of fibrin can be used as a screening test for
the disorder. This is carried out by clotting an anticoagulated patient sample with
thrombin and calcium. The clot is then exposed to a chemical such as urea or
acetic acid. Of these, acetic acid may have a greater sensitivity to milder defects
(FXIII <10IU/dl) than urea (FXIII <5IU/dl), and is probably the more useful
screening test. Such tests should be confirmed by a specific assay of FXIIIc levels.
There are a number of FXIIIc assays available, although there is a great variation
in their performance between different laboratories; an FXIII antigen assessment
by ELISA is also available in some areas. Despite these variations, such tests will
be of use in monitoring the response to therapy.

As it is generally recognized that significant FXIII deficiency carries a not insub-
stantial risk of intracerebral hemorrhage, all patients with severe FXIII deficiency
(<1IU/dl) should be offered prophylaxis with FXIII concentrate from diagnosis.
It should also be actively considered in those with milder deficiencies. FXIII has a
long half-life of 7–10 days and 4-weekly dosing, at 10IU/kg, should be sufficient
to maintain levels of between 3 and 10IU/dl: this level should be sufficient to
prevent spontaneous hemorrhage. An acute bleed should be treated by a virus-
inactivated concentrate (with a dose of 10–20IU/kg to maintain levels within
normal limits until the bleeding has resolved), although FFP and cryoprecipitate
have also been used successfully to manage hemorrhage. As platelets also contain
FXIII, platelet therapy may also be helpful in an extreme emergency, if an appro-
priate product is not available.
Management in pregnancy
It has been reported that up to 50% of pregnancies in those with severe FXIII
deficiency will result in miscarriage without therapy. It is, therefore, usual prac-
tice for all severely affected women who conceive to receive monthly infusions to
maintain a FXIII trough level at >3IU/dl from the diagnosis of pregnancy.
As affected neonates may present with persistent bleeding from the umbilical
cord, and are at risk of intracerebral bleeding, close collaboration is required
with neonatologists to rapidly diagnose and treat infants presenting with unex-
plained hemorrhage in the presence of a normal coagulation screen and platelet
count. This is particularly relevant if there is known to be consanguinity in the
parents.
86
Ehlers Danlos syndromes

Mutations in some of the many genes which control collagen synthesis and
metabolism can result in the variety of clinical patterns that are grouped together
as Ehlers Danlos syndromes. The diagnosis is based upon the combination of
clinical findings and family history, and six major clinical patterns are recog-
nized. Each results in a variable combination of joint hyperextensibility, skin
extensibility and skin fragility: the latter is associated with spontaneous bruising
and the characteristic wide, poorly healed, papery scars which can occur on pres-
sure areas. In Ehlers Danlos syndromes the clotting screen is normal and, despite
the bruising tendency, the bleeding time and platelet function testing are most
often normal. There are also associated cardiac abnormalities, which most often
result in mitral valve prolapse, although, uncommonly there may be aortic dilata-
tion. In addition, there may be associated retinal and dental problems. Such
patients may be at increased risk of bleeding associated with delivery. Although
there are no specific proven treatments, prolonged bleeding times can be
improved with DDAVP in some cases, and there are case reports of a useful effect
if DDAVP is given in such subjects prophylactically in the management of labor.
As with other inherited disorders, there is a need to provide appropriate counsel-
ing as part of pre-pregnancy planning.
87
CHAPTER 5
Antithrombotic therapy
in pregnancy
In pregnancy, it is critical to consider the effects of any drug, not only on the
mother but also on the developing baby, e.g. teratogenesis in the first trimester,
developmental problems in the second and third trimesters, delivery problems,
and also the potential for longer term effects on childhood growth and develop-
ment. In this regard, particular concerns exist in relation to anticoagulants and
antithrombotic therapies in pregnancy, and a sound understanding of the effi-
cacy and safety of these agents is essential in order to provide the optimal man-
agement of thrombotic problems in pregnancy.
Coumarins
Of the coumarin group of drugs, warfarin is commonly employed in a large
number of countries. Warfarin is an orally active 4-hydroxycoumarin, which
inhibits the synthesis of the vitamin K-dependent clotting and anticoagulant
factors (see Table 5.1). This includes the coagulation factors (F) II, VII, IX and
X, and the natural endogenous anticoagulants protein C and protein S. War-
Table 5.1 – The effect of warfarin
Clotting factor activity FII, FVII, FIX, FX

PS, PC
Prothrombotic in first 24 hours If initiated alone in VTE

No risk if used for AF
Target INR ([PT/PTc]ISI) 2–3 DVT/AF

3–4 ArtT/mechanical valve/recurrent VTE
Fetal risks <12 (6–9) weeks Embryopathy

12–36 weeks IQ and developmental defects
>36 weeks ICH/abruption
AF, Atrial fibrillation; ArtT, Arterial thrombisis; DVT, Deep vein thrombosis; F, Factor; ICH, Intracerebral
hemorrhage; INR, International normalized ratio; IQ, Intelligence quotient; ISI, International sensitivity index;
PC, Protein C; PS, Protein S; PT, Prothrombin time; PTc, Control PT; VTE, Venous thromboembolism.
88
Antithrombotic therapy in pregnancy
farin reduces the carboxylation of these factors, which renders them unable to
bind to catalytic phospholipid, which in turn is required for successful coagula-
tion (see Chapter 1). Warfarin has a long half-life, but its effect reflects the
half-life of the vitamin K factors. When warfarin therapy is commenced there is
an initial reduction in the levels of both FVIIc and protein S that is evident
within around 6 hours, reflecting the short half-life of these factors. Indeed, it
is this rapid fall in FVIIc that makes the prothrombin time (PT) useful in pre-
dicting the outcome of acute liver injury. However, on account of the rapid fall
in protein S, warfarin, when used alone, may induce a pro-coagulant imbalance
in the first days of therapy. This may lead to microvascular thrombosis and skin
necrosis. The risk, however, appears to apply only to those subjects who have
suffered thrombosis (when there may be consumption of coagulation and anti-
coagulant factors, and perhaps an underlying thrombophilia), but does not
extend to those receiving warfarin as prophylaxis for atrial fibrillation. For this
reason, heparin therapy should overlap with warfarin in the treatment of
thrombosis until the international normalized ratio (INR: see below) is >2. The
other vitamin K-dependent coagulation factors have considerably longer half-
lives.
Warfarin is highly bound to albumin and is itself metabolized in the liver.
The dose of warfarin given depends upon its measured effect on the PT (see
Chapter 1), which is often reported as an INR. The INR compares the patient’s
PT (in seconds) with that of a normal pool of plasma. These clotting times are
examined simultaneously by clotting the anticoagulated plasma samples with
tissue factor (TF) and calcium. A ratio of the patient:normal pool plasma clot-
ting time is generated. To allow inter-laboratory and international compar-
isons, the TF used should have a known sensitivity (called the international
sensitivity index or (ISI)). To generate the INR, the patient:control ratio is
then logged to the power of the ISI, e.g. for a thromboplastin with an ISI of
1.1, the INR would be equal to the (patient PT:control PT)1.1. This allows the
comparison of PT generated in different laboratories using different TF. In
general, the nearer the ISI is to one, the more precise and safe the INR assess-
ment will be.
A number of factors can interact with warfarin, resulting in an enhanced or
reduced effect (Table 5.2). As a general rule, an INR of 2–3 is the aim for treat-
ment of deep vein thrombosis (DVT), or for the prophylaxis of atrial fibrillation
(often with a target INR of 2.5), and an INR of 3–4 (often with a target of 3.5) is
the aim for the management of arterial thrombosis, mechanical heart valves or a
venous thrombosis that has occurred despite an INR of 2–3.
On account of the individual variability of warfarin dosage, long-term war-
farin carries a not insubstantial risk of bleeding, which after 6 months of treat-
ment for a first DVT can be equivalent to the risk of DVT recurrence. This is
why anticoagulation for DVT is restricted to 3–6 months of therapy, as after this
time, in most cases, the risk of anticoagulation outweighs the risk of recurrent
thrombosis.
89
Table 5.2 – Factors increasing anticoagulation and bleeding with warfarin
Patient factors Age

Liver dysfunction
Alcohol
Cardiac failure
Renal failure
Weight loss/acute illness
Drug interaction NSAID Clotting factor synthesis

Albumin binding
Antibiotics Vitamin K absorption
Warfarin metabolism
(e.g. erythromycin)
Cimetidine Warfarin metabolism
Phenytoin Clotting factor synthesis, but also
Warfarin metabolism
Sodium valporate Warfarin metabolism
Laxatives Vitamin K absorption
Additional anticoagulation Aspirin

Heparins
Thrombolysis
NSAID, Non-steroidal anti-inflammatory drug.
Warfarin in pregnancy
Warfarin crosses the placenta and inhibits coagulation factor production by the
fetal liver. It is also teratogenic, but is not found, to any significant degree, in
breast milk and is therefore safe to use in breastfeeding mothers. In the first
trimester, exposure to warfarin may result in warfarin embryopathy. The fetus is
vulnerable between 6 and 9 weeks gestation, and the incidence has been esti-
mated at around 6% of pregnancies exposed to warfarin in the first trimester.
The effect appears to be dose (not INR) dependent, with an increased risk
evident with does >5mg/day. Warfarin embryopathy can be avoided when war-
farin is substituted by heparin between 6 and 12 weeks gestation. There is also,
however, an association between warfarin usage and miscarriage, and around
25% of women taking warfarin throughout pregnancy will miscarry. Similar
figures are seen even when heparin is substituted in the first trimester. However,
if warfarin is discontinued and heparin employed at/or prior to 6 weeks gesta-
tion, the miscarriage rate drops to 15%. This emphasizes the critical nature of
switching to heparin before 6 weeks gestation to obtain the maximal benefit.
Warfarin embryopathy results in chondrodysplasia punctata (nasal and mid-
face hypoplasia, frontal bossing, short proximal limbs, scoliosis, short phalanges
90
and short stature). There is also a characteristic facial appearance with nasal
hypoplasia. This is due to warfarin-induced premature calcification of the nasal
cartilage, which results in failure of growth of the nose during childhood devel-
opment. This defect is, however, amenable to corrective plastic surgery. In addi-
tion, when warfarin has been used between 12 and 36 weeks gestation, an
association has been reported between its use and neurological and develop-
mental problems in the infant. It is supposed that this is due to small intracere-
bral bleeds in the fetus, consequent upon excessive anticoagulation (due to the
immature status of the fetal liver). Indeed, intracerebral bleeds have been
reported in the fetus of women treated with warfarin in the second and third
trimesters. Between 36 weeks gestation and delivery, there may be an increased
risk of abruption and of intracerebral bleeding in the fetus (particularly during
delivery), as well as a risk of severe maternal bleeding at delivery.
Thus, although warfarin may be preferred to heparin in the maternal interest for
some conditions (such as mechanical heart valves or severe recurrent thrombosis),
due to the better efficacy data for preventing maternal thrombosis, it carries signific-
ant risks for the fetus. In each case it is critical to establish the optimal anticoagula-
tion regime, balancing the risks between the mother and the fetus, and taking into
account the mother’s view of these risks. Such decisions are clearly best made prior
to pregnancy. All women on oral anticoagulants who could become pregnant must
be aware of the risks of conceiving whilst on these drugs, and of the need to report
to their hematologist and/or high-risk obstetrician immediately they become preg-
nant to ensure that optimal anticoagulation and pregnancy management can be
provided. In particular, as noted above, a switch to heparin should be made before
6 weeks gestation if this is the therapeutic strategy that is to be employed.
Heparin
The heparins are glycosaminoglycans with a high affinity for the endogenous anti-
coagulant antithrombin (see Chapter 1), altering both its structure and anticoagu-
Table 5.3 – Low-molecular-weight heparin (LMWH)
No effect on APTT Even in overdose
Reversal may need repeated Start with 25–50mg, then dose by

protamine sulphate clinical response
Therapeutic use in pregnancy ? Repeated anti-Xa assays required
Around 3% bone loss in normal pregnancy Around 5% bone loss with LMWH in pregnancy
APTT, Activated partial thromboplastin time.

91
lant function. Two classes of heparin are used clinically: unfractionated heparin
(UFH), which contains a mixture of molecular weights, and the fractionated low-
molecular-weight heparins (LMWH: see Table 5.3). Binding of UFH to antithrom-
bin changes the conformational structure of antithrombin, resulting in a 1000-fold
increase in its ability to inactivate factor (F) Xa and FIIa. The smaller fragments
contained in LMWH preferentially bind the antithrombin–FXa complex.
The heparins currently available are inactive orally and have to given by an
intravenous or subcutaneous route. The half-life of intravenous UFH is variable,
but can be as short as 30–45 minutes. The therapeutic level of UFH can be
assessed by an activated partial thromboplastin time (APTT or PTT: see Chapter
1). This is achieved by determining a ratio of the APTT in the patient sample to
the APTT of a normal pooled plasma: this gives a therapeutic range of 1.5–2.5.
Unlike warfarin, there is no internationally accepted standardized ratio. Indeed,
when used in pregnancy, it must be remembered that the APTT may be short-
ened in pregnancy (as a result of an increase in the levels of key proteins such as
FVIIIc, fibrinogen and the heparin-binding proteins), giving the impression of
heparin resistance. When given subcutaneously, UFHs have a half-life of around
1.5 hours, and dose monitoring can be achieved by mid-interval APTT assessment
(4–6 hours after dosing). The rate of major bleeding in pregnancies treated with
UFH is 2%, which is consistent with the bleeding risk associated with both UFH
and warfarin therapy in non-pregnant subjects treated for DVT. Furthermore it
should be noted that dose-adjusted subcutaneous UFH can cause a persistent anti-
coagulant effect up to 28 hours after the last injection, which could be problem-
atic at the time of delivery. However, the mechanism for this effect is unclear.
LMWHs are given subcutaneously and have more predictable kinetics and
>90% bioavailability. This results in a half-life of 4–6 hours, which allows once, or
twice, a day dosing, and monitoring is seldom required. The APTT is insensitive
to the presence of LMWH (even when present in excess) and is likely to remain
within reference limits. If monitoring of these drugs is needed, this can be
achieved by assessment of the effect of the LMWH on the inactivation of coagula-
tion FXa (an anti-Xa assay).
Reversal of intravenous UFH can be achieved by protamine sulfate injection:
the amount to be given can be titrated to the predicted residual amount of UFH
in the circulation, which routinely results in a protamine dose of 25–50mg. The
downside of the long half-life of LMWHs is the difficulty in reversing their effect
if bleeding occurs, or an excess dose is given accidentally. Protamine will inhibit
only the LMWH currently in the circulation. In these circumstances, if bleeding
is ongoing, it has been suggested that further protamine dosing should be deter-
mined by clinical response alone, although an initial anti-Xa assessment is often
carried out in these circumstances. The anti-Xa level does not, however, predict
bleeding, and the frequent monitoring which would be required to determine
protamine dosing is often impractical. This problem can be significant, as an
excess of protamine can also act as an anticoagulant.
92
Heparin-induced thrombocytopenia
There are two types of thrombocytopenia associated with heparin (see also
Chapter 8). The first is non-immune heparin-induced thrombocytopenia (HIT),
which is sometimes less accurately described as Type 1 HIT. This is a benign non-
immune condition that results in a mild degree of thrombocytopenia, which
appears to be due to a direct effect of heparin on the platelet membrane and is
not associated with clinical symptoms, or thrombosis.
The more important type of thrombocytopenia associated with heparin is
Type 2 HIT, also known as HIT thrombosis syndrome (HITTS), reflecting the
clinical expression of the syndrome with thrombosis as well as thrombocytopenia.
This immune condition arises from the development of specific heparin-depend-
ent immunoglobulin (IgG) antibodies against the platelet factor (PF) 4–heparin
complex. It commonly has its onset 5–14 days after the commencement of
therapy, although it can occur after a brief exposure if there has been prior
heparin exposure within the previous 3 months. This leads to platelet activation,
generation of pro-coagulant microparticles from the platelet membrane and
platelet consumption. Platelet levels can fall to a median of around 60 × 109/ l,
but levels <20 × 109/l are encountered in 10–15% of cases. Furthermore, a
significant fall which does not result in a level less than the lower limit of the ref-
erence range can still be consistent with the diagnosis of Type 2 HIT. In addition
to platelet consumption, Type 2 HIT is also associated with thrombin generation.
This occurs via the generation of TF on endothelial cells activated by HIT anti-
bodies binding to PF4 found on endothelial heparan sulfate, as well as increased
expression of TF on monocytes upon activation by HIT antibodies. Clinically
severe arterial or venous thrombotic complications can ensue. In 5–10% of sub-
jects there may be venous thromboembolism, but myocardial infarction, limb
ischemia, mesenteric artery thrombosis and cerebrovascular thrombosis also
occur. The occurrence of thrombosis in affected patients is associated with a
significant mortality. Interestingly, despite the thromboctyopenia, petechiae are
absent, but skin reactions at injection sites and systemic inflammatory reactions
can also occur. Thus, if a patient develops a skin reaction whilst on heparin it is
prudent to monitor the platelet count for HIT.
The incidence of HIT has been estimated at 2.7% with UFH, and there is
anecdotal evidence that HIT also occurs with LMWH, although the incidence is
less easy to define. The incidence of HIT with LMWHs could be as low as 0% and
the overall relative risk is likely to be at least 8-fold greater with UFH than
LMWH. This is likely to relate to the heparin polysaccharide chain length, which
needs to be at least 12–14 polysaccharides long to form an HIT antigen with PF4.
The risk of both HIT and thrombosis varies with the patient type, the duration of
use and the type of heparin. With UFH, cardiac surgery patients are most at risk
of developing antibodies (around 50%), but few (<5% of those who become anti-
body positive) develop HIT, while orthopedic surgical patients have a lower rate
of antibody development (15%), but as many as one third will go on to develop
93
HIT. It is important to note that HIT does not always lead to a clinical throm-
botic event. Indeed, in cardiac surgery with UFH, 1% or less will develop a
thrombotic problem.
There are occasional reports of thrombocytopenia or heparin-dependent
platelet-activating antibodies in pregnant patients treated with LMWH. However,
even in pregnant patients diagnosed with HIT associated with LMWH, there is
often a previous sensitization or HIT episode related to UFH exposure. In
general terms, medical and obstetric patients are considered least at risk, and
recent recommendations do not consider that routine monitoring of the platelet
count is required with LMWH use in pregnancy (the risk with LMWH in preg-
nancy is considered to be <0.1%). Thus, platelet count monitoring should be
restricted to special situations (see Table 5.4).
HIT is a clinicopathologic syndrome, although the diagnosis is principally a
clinical one, supported by the detection of serologic factors. Key factors to con-
Table 5.4 – Heparin-induced thrombocytopenia (HIT)
Those at risk UFH > LMWH

5+ days treatment
Previous HIT/previous UFH exposure
Consider platelet On LMWH with previous exposure to UFH within 3 months

monitoring if: UFH used prior to LMWH
UFH is used for treatment or prevention of thrombosis
A systemic reaction after intravenous UFH occurs
A cutaneous reaction to UFH/LMWH occurs
Consider HIT An unexpected fall in platelet count occurs

diagnosis if: A new thrombotic problem has occurred
A skin reaction is accompanied by thrombocytopenia
Thrombocytopenia cannot be attributed to an established/alternative illness (including
major VTE)
Heparin is used in intravenous flushes or dialysis
+/– HIT antibodies detected
Thrombosis types MI, VTE, ArtT
Cross-reaction LMWH may cross-react with UFH

LMWH may cross-react with each other
Danaparoid can also cross-react with heparin
Action Stop all exposure to heparin

Change to danaparoid/ hirudin/ ?warfarin
Avoid platelet transfusions
ArtT, Arterial thrombisis; LMWH, Low-moleclular-weight heparin; MI, Myocardial infarction;

UFH, Unfractionated heparin; VTE, Venous thromboembolism.
94
sider in making the diagnosis are shown in Table 5.4. HIT antibodies can be
detected by either platelet-activation (functional) assays or by antigen assays
directed against the heparin–PF4 complex. The diagnostic specificity of a sensi-
tive washed platelet activation functional assay is greater than that of an antigen
assay. The diagnosis of HIT is made when there is an otherwise unexplained fall
in the platelet count (usually >50%, even if the platelet count remains
>150 × 109/l), when there are skin lesions at heparin injection sites, or when an
acute systemic reaction such as cardiorespiratory distress after intravenous
heparin administration has occurred. The diagnosis is supported by the detec-
tion of HIT antibodies, but can be made even when there is a failure to confirm
the presence of such antibodies. HIT antibody seroconversion on its own (i.e.
without thrombocytopenia or other clinical sequelae) is not, however, con-
sidered sufficient to constitute a diagnosis of HIT.
Treatment centers on the discontinuation of any heparin preparation and the
use of alternative thrombin inhibitors, usually in therapeutic doses. These
alternatives include: the heparinoid, danaparoid; recombinant hirudin; and the
direct thrombin inhibitor argatroban. Fondaparinux has also been employed in
this context. While warfarin is best for chronic therapy, it should be avoided in the
acute stage, as (noted above) it may result in a precipitous fall in the endogenous
vitamin K-dependent anticoagulant protein S. This may result in skin necrosis and
venous limb gangrene: those patients with protein C, protein S or antithrombin
deficiency, or FV Leiden carriage are perhaps most at risk. Warfarin should there-
fore be delayed until the platelet count is >100 × 109/l and the HIT is resolving
with the use of other antithrombotic agents. Adjunctive treatments that have been
employed in the management of HIT include: embolectomy; plasmapheresis;
thrombolysis; and high-dose intravenous normal immunoglobulin (to inhibit the
antibody-mediated platelet activation). It should be noted that platelet compo-
nent therapy should be avoided in HIT patients, as it is thought that this may
increase platelet consumption and precipitate thrombosis.
UFH versus LMWH in pregnancy

Both UFH and LMWH have been used extensively in pregnancy. Heparins do not
cross the placenta or enter breast milk, and no significant association with fetal dis-
orders has been found. However, heparins have been linked to an increased mater-
nal risk of osteoporosis, allergic reactions, HIT and hemorrhagic complications.
Osteoporosis
Of all the potential complications of heparin, perhaps the side effect that con-
cerns obstetricians most is osteoporosis. Pregnancy itself is associated with a
reduction in bone mineral density of the order of 3%. With prolonged UFH use,
>30% of women on UFH will lose significant (>10%) bone mass. This is related
to the dose and duration of therapy, and is more common with prolonged high-
95
dose treatment. However, problems have been reported with doses as low as
15 000 day and with therapy given for only 7 weeks. In one study of long-term
prophylactic UFH therapy in pregnancy, a 2.2% incidence of vertebral fractures
was reported. Although, in another study, spinal fractures were reported in as
many as 15% of patients receiving UFH at a dose of 10 000IU twice daily for a
period of 3–6 months.
Experimental model and animal studies provide some insight into the
mechanism of heparin-induced osteoporosis. A dose-dependent decrease in can-
cellous bone in the femur (resulting from decreased bone formation and
increased bone resorption) occurs in animals treated with UFH for 1 month. In
one animal study, UFH resulted in a 30% loss in cancellous bone during the first
month of treatment. In this study, heparin was shown to accumulate in the bone
and was retained there for at least a further 1 month after the cessation of
therapy. This prolonged accumulation would suggest that heparin-induced osteo-
porosis may not be rapidly reversible.
LMWH carry a substantially lower risk of osteoporosis than UFH. In one study
of non-pregnant venous thromboembolism (VTE) patients treated for 3–6
months, only 2.5% of subjects receiving LMWH experienced a vertebral fracture
compared to 15% in the UFH group. One randomized trial of LMWH versus
UFH in pregnancy measured bone mineral density in the lumbar spine for up to
3 years after delivery. In this study, bone density did not differ between healthy
controls and the dalteparin group, but was significantly lower in the UFH group
when compared to both controls and dalteparin-treated women. Multiple logistic
regression found that the type of heparin therapy was the only independent
factor associated with reduced bone mass. This is consistent with several observa-
tional pregnancy studies that have shown no difference in lumbar spine density
in subjects on prolonged LMWH therapy and no greater bone loss than that
which occurs in pregnancy naturally. In a review of LMWH (largely enoxaparin
and dalteparin) therapy in over 2500 pregnancies, only one case of heparin-
induced ostepenia with fracture (vertebral) was reported in a woman who had
received a high dose (15 000IU/day) of dalteparin for a total of 36 weeks.
Recently, however, three cases of osteoporotic fractures in association with tinza-
parin use in pregnancy in one center have been reported. Whether this finding is
causally related to tinzaparin (as additional therapies and underlying problems
can also influence the development of osteoporosis) and whether this risk
applies to other LMWH is unclear. In any event, the risk associated with LMWH
remains very low in comparison to UFH. This difference between UFH and
LMWH is consistent with animal studies, which show greater cancellous bone loss
with UFH than LMWH.
Bleeding risk
LMWH appear to carry a lower risk of bleeding than UFH: severe bleeding prob-
lems have not been attributed to either treatment or prophylactic-dose LMWH
96
during pregnancy. Furthermore, LMWH have not been associated with an

increased risk of severe peripartum bleeding, with an observed rate of around
2%. This, in most cases, is linked to a primary obstetric cause, such as uterine
atony or vaginal laceration, although it is possible that any such blood loss may
be exacerbated by the presence of LMWH.
Antithrombotic efficacy
The effectiveness of LMWH in the management of thrombotic conditions in
pregnancy is supported by the low VTE recurrence rate (just over 1%) reported
during treatment in both large cohort studies and systematic reviews. This com-
pares favorably with the recurrence rates of 5–8% reported in trials of non-preg-
nant patients treated with either LMWH or UFH (followed by coumarin therapy)
after 3–6 month follow-up. LMWH thromboprophylaxis is also associated with a
VTE rate of <1%.
Fetal outcome
With regard to fetal outcome, in general terms, the evidence supports a benefi-
cial effect of LMWH on pregnancy loss rates. The successful pregnancy rate,
reported in women receiving LMWH for previous adverse pregnancy outcomes
(such as recurrent fetal loss), is >80%. This rate is consistent with that found in
randomized trials of antithrombotic therapy (UFH or LMWH) in women with
previous pregnancy loss associated with antiphospholipid syndrome or heritable
thrombophilia, where such intervention resulted in a significant improvement in
pregnancy outcome.
Kinetics and allergy

The half-life of LMWH is reduced in pregnancy due to the physiological increase
in renal blood flow and glomerular filtration rate. Although allergic skin reac-
tions occur in 1–2% of patients on LMWH, this risk is less than that with UFH
and, as noted above, LMWH also carry a lesser risk of HIT. Thus, given their
effectiveness, ease of use, safety for mother and fetus, and reliable kinetics,
LMWH are, at present, the anticoagulants of choice in pregnancy.
Danaparoid
Danaparoid is a heparinoid (a mixture of heparan, dermatan and chondroitin
sulfates), that inactivates FXa and FIIa. It has a relatively long half-life of 25 hours
in non-pregnant subjects. Danaparoid has been used successfully in pregnancy-
associated HIT and, out-with pregnancy, danaparoid is successful in >90% of
cases. However, a positive cross-reaction can occur in 15% of cases of HIT. Dana-
paroid is also useful in pregnant patients with an allergy to LMWH. The throm-
97
boprophylactic dose for an average pregnant patient is 750IU given twice daily by
subcutaneous injection. At a prophylactic dose, it has no effect on platelet func-
tion and should not be associated with excess blood loss at delivery. When a ther-
apeutic dose is required, danaparoid can be monitored by an anti-Xa assay using
a danaparoid control. The recommended dosing with danaparoid for HIT out-
with pregnancy is an intravenous bolus of 2250IU, followed by an intravenous
infusion at 400IU/hour for 4 hours, then 300IU/hour for 4 hours, then
200IU/hour adjusted by anti-Xa levels. It does not cross the placenta and
although it is not yet clear whether it crosses the breast, this appears unlikely, as
no anti-Xa activity occurs in breast milk. In any event, it should be destroyed in
the infant GI tract. It is exclusively excreted by the kidney and should be used
with extreme caution in subjects with renal failure.
Direct thrombin inhibitors

Thrombin inhibitors are potent inhibitors of both free and clot-bound thrombin,
which do not cross-react with HIT antibodies and have little, or no, effect on
related serine proteases.
Lepirudin
Lepirudin, a recombinant hirudin (hirudin is derived from the medicinal leech
Hirudo medicinalis), binds tightly to thrombin with slow dissociation, making it
effectively an irreversible inhibitor. There is very little information on its use in
pregnancy, but it may have a role as an alternative to danaparoid in the manage-
ment of pregnancy-related HIT. The dose used for HIT is an initial infusion of
0.15mg/kg/hour intravenously, with the aim to achieve a target APTT of 1.5–2.0
times either the patient’s baseline APTT, or the mean of APTT laboratory refer-
ence range. It has a relatively short half-life of about 80 minutes and is renally
excreted. At higher doses the ecarin clotting time has a more linear correlation
with plasma hirudin levels and may be more useful for monitoring. In animal
studies, hirudin has low placental transit and it does not appear to cross the
breast. When considering the use of hirudin, it should be borne in mind that it
currently has no antidote and it should be used with caution in those with renal
impairment. In practice, the use of hirudin should be limited to those subjects
where the HIT antibody is shown to cross-react with danaparoid, or in those who
have a cutaneous allergy to danaparoid. There is a high rate of antibody develop-
ment to hirudin of 40–60% and anaphylaxis can occur, although this is rare.
Argatroban
Argatroban is a synthetic thrombin inhibitor derived from L-arginine. It binds
tightly, yet reversibly, to thrombin. Individuals receiving prolonged, or repeated,
administration show no generation of antibodies and no enhancement or sup-
98
pression of the anticoagulant response. It is metabolized in the liver and its effect
can be monitored with the APTT (aiming to achieve 1.5–3 times control values).
Its action is quickly reversed by stopping the infusion (<60 minutes), but clear-
ance is reduced 3–4-fold if there is moderate hepatic impairment. It is used in
the treatment of HIT where its effectiveness is clearly demonstrated. In this situ-
ation the recommended dose is an initial infusion rate of 2µg/kg/minute, with
no initial bolus required. There is no significant experience of the use of this
agent in pregnancy.
Aspirin
Aspirin irreversibly inhibits (through acetylation) cyclo-oxygenase, which is
pivotal in the generation of thromboxane A2 in the platelet. The effect of aspirin
lasts between 7 and 10 days, as restoration of platelet function requires new
platelet synthesis. Aspirin only partially inhibits adenosine diphosphate (ADP)
and collagen-induced platelet activation, and has no effect on thrombin-medi-
ated platelet activation. Aspirin may be as effective as an antithrombotic agent at
modest doses (75mg/day) as it is at higher doses (1200mg/day). The principal
side effects of aspirin are bleeding and idiosyncratic or allergic reactions. In chil-
dren, ingestion of aspirin has been associated with Reye’s syndrome.
Aspirin has been widely assessed for the prevention of pregnancy-related com-
plications such as intrauterine growth restriction and pre-eclampsia. The use of
low dose aspirin (60–75mg/day) to prevent pre-eclampsia is based upon the
rationale that pre-eclampsia is associated with alterations in the production of
prostacyclin and thromboxane, secondary to activation of the clotting system and
changes in platelet function. A recent meta-analysis reported on 42 randomized
trials involving over 32 000 women comparing antiplatelet agents to placebo or
no treatment. Trials were subgrouped by maternal risk. Taking the cohort as a
whole, there was a 15% reduction (relative risk 0.85, 95% confidence interval
(CI) 0.78–0.92) in the risk of pre-eclampsia associated with the use of antiplatelet
agents. These assessments have shown that low-dose aspirin (around 75mg/day)
appears safe for use in pregnant women. At higher doses (>150mg/day) its
safety, especially in the first trimester, is controversial.
Although aspirin has been shown to reduce the risk of venous thrombosis in
the context of orthopedic surgery, its role in the context of pregnancy-related
thrombosis is not yet defined. However, it does not appear to offer the same level
of protection as UFH or LMWH for thromboprophylaxis. When aspirin has been
used as a substitute to coumarins (in subjects with mechanical heart valves), as
would be expected, it confers insufficient anticoagulation and results in an unac-
ceptable (and often fatal) 25% incidence of systemic embolism and 4% inci-
dence of valve thrombosis, although it appears safe from a fetal perspective.
Aspirin has a role in combination with other therapies, particularly in combi-
nation with LMWH, in the antiphospholipid syndrome. It is also of value where
specific antiplatelet therapy is required in pregnancy, such as in women with a
99
previous cerebral ischemia of arterial origin, ischemic heart disease or essential

thrombocythemia, or when other anticoagulant therapies are unsuitable.
The salicylates, including aspirin, are excreted at low doses in breast milk,
with <10% reaching the nursing infant. Although this could cause drug accumu-
lation in the fetus, maternal doses of <100mg/day are considered safe as regards
infant bleeding. Overall, as there is a risk that high-dose maternal aspirin could
lead to bleeding in the infant, and because of a theoretical risk of Reye’s syn-
drome, aspirin is not advised for nursing mothers in a number of countries.
Dextran
Dextran has been used for peripartum thromboprophylaxis, particularly during
Cesarean section. It does, however, carry a significant risk of anaphylactic and
anaphylactoid reactions, which have been associated with uterine hypertonus,
profound fetal distress, and a high incidence of fetal death or profound neuro-
logical damage. Thus, dextran for thromboprophylaxis should be avoided prior
to delivery.
New anticoagulant therapies

There are a number of new anticoagulants in development. These include sub-
stances aimed at the onset of coagulation, such as tissue factor pathway
inhibitors, to drugs aimed at the downregulation of thrombin generation. In
addition, there is work ongoing to develop an orally active heparin derivative. Of
the drugs in development, two classes have completed a number of clinical trials
and may soon achieve licences for use in non-pregnant subjects. These are the
synthetic pentassaccharides fonduparinux and idraparinux (which, like heparin,
inactivate FXa by binding to antithrombin), and the direct thrombin inhibitor
melagatran, which is given orally (as its pro-drug, ximelagatran).
Fondaparinux and Idraparinux

Fondaparinux binds to antithrombin, which results in a 300-fold increase in
antithrombin activity against FXa. It has a half-life of around 17 hours after sub-
cutaneous injection and has been shown to be efficacious in the prevention of
VTE after orthopedic surgery. It has been used in pregnancy where women
developed skin reactions to LMWH. Fondaparinux appears to pass the placental
barrier in vivo (in contrast to in vitro where no transfer is found), resulting in
measurable anti-FXa activity in umbilical cord blood. However, the concentration
of fondaparinux in umbilical cord blood appears to be well below the concentra-
tion required for effective anticoagulation. Nonetheless, because of this, the
availability of alternative therapies, and limited safety data, fondaparinux should
be avoided at present, or its use limited to those for whom there are no therapeu-
tic alternatives. Idraparinux is another synthetic pentassaccharide which also
100
leads to FXa inhibition, but has a prolonged half-life (80 hours), which could
allow once weekly dosing. There are currently no data on its use in pregnancy.
Protamine sulfate has no action on these drugs, but two studies have shown that
recombinant FVIIa (rFVIIa) may be able to reverse the biochemical changes
induced by these pentassaccharides in healthy volunteers. In future, heparinases
may also have a role in their inactivation.
Ximelagatran
Ximelagatran is the prodrug of melagatran, which is an active site-directed
thrombin inhibitor. Ximelagatran is well absorbed from the gut and has a plasma
half-life of 3–4 hours, allowing it to be administered twice daily. Thus, it has the
potential of being an alternative oral anticoagulant to warfarin. It may require a
dose adjustment when there is renal impairment, and in 6% of subjects a mild
rise in transaminases has been noted, although more severe reactions have also
occurred. This has led to concerns for the safety of this agent and it remains unli-
censed in both the USA and Europe at present. It appears most likely that xime-
lagatran will have a role in the prophylaxis of atrial fibrillation in the first
instance. At present, however, there is insufficient information on the safety of
this class of compounds in those who are pregnant or breastfeeding.
101
CHAPTER 6
Management of
thrombotic disorders
Pregnancy places a substantial demand on the cardiovascular system. With
increasing gestation there is a 30–50% increase in cardiac output, which is
accompanied by an increase in prothrombotic coagulation factors, a decrease in
some natural anticoagulants and an alteration in systemic fibrinolysis. This prepa-
ration for the hemostatic challenge of delivery poses a particular threat to those
women at risk of thrombosis.
More and more, obstetricians and hematologists are faced with the manage-
ment of pregnant subjects who have risk factors for either arterial or venous
thrombosis. This most commonly includes those with a previous thrombosis or
those with prosthetic heart valves, but there is an evolving awareness of genetic
and acquired risk factors that may identify those at particular risk of vascular
occlusion, not only in pregnancy but also in later life.
Arterial thrombosis in pregnancy

Acute arterial occlusion during pregnancy is very rare and evidence for its man-
agement is based upon case reports and small case series, with most data extrapo-
lated from non-pregnant subjects. More relevant to routine care, is the
prevention of disease in those at risk and minimizing the potential harm of anti-
coagulants to the mother and the fetus. Recently, there has also been the sugges-
tion that women with risk factors for arterial thrombosis are also at higher risk of
other pregnancy complications, such as pre-eclampsia, growth restriction and
fetal loss.
Risk factors for arterial disease in pregnancy

Arterial occlusion is usually the end result of a prolonged process that terminates
in the rupture of an atherosclerotic plaque with intramural thrombotic occlu-
sion, and a number of common factors are recognized as risks for arterial disease
(see Table 6.1). Of these, those that are relevant in women of childbearing age
include hypercholesterolemia, smoking, hypertension, diabetes mellitus, obesity,
low socio-economic status and a family history of ischemic heart disease. A
number of rarer systemic disorders are also associated with an increased risk of
vascular occlusion during pregnancy: these include essential thrombocythemia,
102
Management of thrombotic disorders
Table 6.1 – Risk factors for arterial thrombosis in young women
Hypercholesterolemia
Family history of ischemic heart disease
Smoking
Hypertension
Diabetes mellitus
Obesity
Essential thrombocythemia
Polcythemia rubra vera
Sickle-cell disease
Prosthetic heart valve
Mitral stenosis and atrial fibrillation
Mitral bioprothesis and atrial fibrillation
Mechanical mitral valve
Antiphospholipid syndrome
? Smoking with Factor V Leiden or prothrombin 20210A
Hyperhomocysteinemia
Cocaine
Ergot
?Marijuana
Previous cerebral ischemia of arterial origin
Previous myocardial infarction
Hypotension
Pre-eclampsia
primary polycythemia and paroxysmal nocturnal hemoglobinuria (see Chapter 3).

In addition, those with prosthetic heart valves, or a prior history of arterial occlu-
sion, are also at higher risk of arterial occlusion during pregnancy. Other, more
intriguing, risks for arterial occlusion that may have a bearing in the pregnancy
age group include cocaine and marijuana use. There is also the historical
consideration that the ergot derivatives used in the management of postpartum
hemorrhage may also be associated with an increased risk of arterial occlusion.
Recently, there has been interest in the possible risk resulting from the combi-
nation of acquired risk factors and single gene defects (see Chapter 7). These
include: elevated plasma factor (F) VIIIc levels (resulting from the interaction of
blood groups with the acute phase response); hyperhomocysteinemia (due to the
combination of hematinic deficiency and enzyme defects in the
homocysteine–methionine pathway); acquired activated protein C resistance
(due to a combination of arterial and venous thrombotic risk factors) and; the
antiphospholipid syndrome. Indeed, a small number of studies have suggested
that there may also be a link between either the prothrombin G20210A or FV
Leiden mutations and arterial disease in young female smokers. This remains dis-
103
puted, but, if confirmed, will contrast with more severe heritable thrombophilias,
such as antithrombin, or protein C or S deficiency, where there has never been
convincing evidence of a link with arterial thrombosis.
Pregnancy with valvular heart disease or prosthetic heart valves

Valvular heart disease results from either congenital cardiac disorders or rheumatic
heart disease. Rheumatic disease most commonly affects the mitral valve and may
result in stenosis, or incompetence. Indeed, previously undiagnosed mitral stenosis
may present for the first time in pregnancy with thromboembolism, due to the
onset of atrial fibrillation. In such cases, atrial fibrillation may be induced by the
change in cardiac output and cardiac pressures that accompany normal gestation.
Overall, the risk of thromboembolism from a mitral stenosis has been estimated at
<4.5%/annum when there is sinus rhythm, although the risk is likely to be around
10 times higher at the onset of atrial fibrillation.
Pregnancy in a woman with a prosthetic mechanical heart valve is not a benign
process. Overall, mechanical valves carry a 1–4% risk of maternal mortality, mainly
arising from complications of valve thrombosis. Inadequate anticoagulation
increases the risk of thromboembolism, such that subjects with mechanical pros-
thetic valves require full anticoagulation throughout pregnancy. This may be espe-
cially the case when the patient has an older valve type (such as Starr-Edwards or
Bjork-Shiley) and atrial fibrillation, and where the valve is in the mitral position.
Patients with bioprosthetic valves do not usually require anticoagulation in
pregnancy, and those subjects with satisfactory cardiovascular adaptation have a
good fetal outcome. There is, however, concern that pregnancy may result in
accelerated calcification of bioprosthetic valves. In one retrospective audit,
around 35% of pregnancies in subjects with bioprosthetic valves had sufficient
deterioration in valve function during the pregnancy such that replacement was
required during, or soon after, pregnancy. However, this effect of pregnancy on
the valve has not been confirmed in a further study of the long-term experience
of such valves where a history of pregnancy was not associated with a higher risk
of valve failure.
Management of valvular heart disease or prosthetic heart

valves in pregnancy
Given the potential risk of thromboembolism associated with mitral valve steno-
sis, anticoagulant prophylaxis may be appropriate during pregnancy in those
women with additional risk factors such as atrial fibrillation or heart failure. As
noted above, there is not usually a requirement for anticoagulation in women
with bioprosthetic valves unless they have additional risk factors.
The management of pregnancy in women with mechanical prosthetic valves
remains controversial, as there are few controlled trials to guide optimal
antithrombotic therapy. In addition, the balance of risk between the potential
104
fetal and maternal complications with the currently available therapies – warfarin
and heparin (see Chapter 5) – is also unknown. In particular, in view of the con-
cerns with oral coumarin therapy, there is a need to find alternative strategies
that are will be safe for both the mother and the fetus.
A recent review of maternal and fetal outcomes in women with prosthetic
heart valves reported that commonly used anticoagulant regimes were: (1) use of
coumarin throughout pregnancy; (2) replacement of coumarin with unfraction-
ated heparin (UFH) from 6 to 12 weeks gestation; and (3) UFH use throughout
pregnancy. Overall, fetal wastage was 33.6% in the first group, 26.5% in the
second and 19.6% in the third. In the coumarin regimes, heparin was used at
term to avoid delivery with an anticoagulated mother and fetus. Coumarin
embryopathy occurred in 6.4% of live births where a coumarin was used through
pregnancy: this was eliminated if heparin was substituted by 6 weeks gestation.
In this review (see Table 6.2), coumarin use through pregnancy was associated
with a lower risk of thromboembolic problems (3.9%) when compared with preg-
nancies where heparin was substituted for the coumarin. UFH used between 6
and 12 weeks gestation was associated with a risk of valve thrombosis of 9.2%, and
UFH used throughout pregnancy was associated with a 25% rate of valve throm-
bosis with adjusted-dose heparin and a 60% rate with low-dose UFH. However,
the numbers of patients exposed to UFH included in the review was modest. In
addition, many thromboembolic events were associated with an inadequate
heparin dosage. Moreover, modern bi-leaflet valves are substantially less throm-
bogenic than earlier mechanical valves such as the Bjork-Shiley and Starr-
Edwards valves, which were predominant in most of the work reviewed. These
limitations make it difficult to extrapolate early work directly to contemporary
practice with modern valves and well-controlled anticoagulation with UFH (or
indeed low-molecular-weight heparin (LMWH)) with regard to thrombotic risk.
Table 6.2 – Maternal outcomes from women with mechanical valves receiving anticoagulation
Anticoagulation schedule Thromboembolism Death

(%) (%)
Coumarin/heparin 3.9* 1.8

Heparin/coumarin/heparin 9.2* 4.2
Adjusted-dose heparin 25.0* 6.7
Low-dose heparin 60.0* 40.0
No heparin or antiplatelet therapy alone 24.3* 4.7
After Chan et al 2000.
*Up to 50% of complications occurred on low-dose heparin and others with inadequate dosing based on the
activated partial thromboplastin time. There is a 2.5% incidence of severe bleeding at delivery and the rate with
low-molecular-weight heparin is unclear with an estimate of 13%.
105
Thus, on the basis of the limited data available, coumarin appears more effect-
ive than UFH for thromboembolic prophylaxis in women with mechanical heart
valves in pregnancy (Table 6.2), but UFH is associated with a better fetal
outcome (see Table 6.3).
The situation is changing with the increasing use of LMWH in pregnant
women with prosthetic heart valves, which has attractions over UFH in terms of
monitoring and maternal side effects, although concerns still exist over how
effective LMWH is for thrombrophylaxis in this situation. In a review of 67 preg-
nancies treated with LMWH in women with prosthetic valves, valve thrombosis
occurred in seven cases, giving an incidence of 10.4% (95% confidence interval
(CI), 3.1–17.7%) and a mortality rate of 1.49% (95% CI, 0.03–8.34%). The
overall thromboembolic complication rate was 13.4% (9 in 67; 95% CI,
5.4–21.4%). However, eight of the nine patients with a thromboembolic compli-
cation had received a fixed dose of LMWH, and in two of these patients, a fixed
low-dose of LMWH. In 42 pregnancies with anti-Xa levels monitored, only one
thromboembolic complication occurred. The live birth rate was 89.5% (95% CI,
82.5–96.5%), the fetal wastage was 8.9% (95% CI, 2.1–15.7%) and there were no
reported congenital anomalies. Thus, adjusted-dose LMWH (anti-Xa level at a
minimum of 1.0IU/ml) compares favorably with UFH in this situation.
Because of the limited data, management remains controversial and several
approaches remain acceptable, including substitution of LMWH for UFH.
However, it is clearly critical to ensure adequate doses of UFH or LMWH. There
is still a need for consensus among experts to systematically gather the best avail-
able evidence, identify unresolved issues and generate studies designed to answer
these questions. Meanwhile, the American College of Chest Physicians (ACCP)
recommends either adjusted-dose twice-daily LMWH throughout pregnancy, in
doses adjusted either to keep a 4 hourly post-injection anti-Xa level at around
1.0–1.2IU/ml (preferably) or in doses adjusted by weight. Alternatively, aggres-
Table 6.3 – Fetal outcome in women with mechanical valves receiving anticoagulation
Treatment Spontaneous Congenital Fetal

miscarriage anomaly wastage
(%) (%) (%)
Coumarin throughout, with/without heparin at term 24.8 6.4* 33.6
Heparin in first trimester, then coumarin 24.8 3.5* 26.5
Heparin throughout 9.8 3.4* 19.6
After Chan et al 2000
*None if heparin started before 6 weeks

106
Table 6.4 – Management of mechanical prosthetic valves
Option 1 Dose-adjust UFH, 12 hourly subcutaneous, throughout pregnancy

Mid-interval APTT ≥ 2 × control or
Anti-Xa 0.35–0.70IU/ml
Option 2 Dose-adjust (by weight) LMWH throughout pregnancy

4 hour post-injection anti-Xa around 1.0IU/ml
Option 3 UFH or LMWH till 13th week

Change to warfarin until mid-third trimester
Restart UFH or LMWH until delivery
Option 4 Warfarin throughout pregnancy

with substitution of UFH or LMWH in late third trimester for delivery
APTT, Activated partial thromboplastin time; LMWH, Low-molecular-weight heparin; UFH, Unfractionated heparin.
sive adjusted-dose UFH throughout pregnancy, i.e. administered subcutaneously,

12 hourly, in doses adjusted to keep the mid-interval activated partial thrombo-
plastin time (APTT) at least twice control, or to attain an anti-Xa heparin level of
0.35–0.70IU/ml, can be used. One further option proposed is UFH or LMWH
(as above) until the 13th week, with a change to warfarin until the middle of the
third trimester, and then restarting UFH or LMWH (see Table 6.4). The ACCP
also suggest the addition of low-dose aspirin.
Long-term oral anticoagulants should be resumed postpartum with all regi-
mens, but there appears to be a significant risk of secondary postpartum hemor-
rhage when switching from therapeutic heparin to therapeutic coumarin soon
after delivery, and so the present authors delay this switch for several days after
delivery to minimize the risk whenever possible.
Cerebral ischemia of arterial origin (CIAO) in pregnancy

CIAO may result from atherosclerosis, from emboli of cardiac origin or from
hypotension. There is also an association between arteritis or antiphospholipid
syndrome and small-vessel brain disease. In published literature, there is a great
variation in the estimate of risk of CIAO during pregnancy. This may be due to a
lack of computerized tomography (CT) or magnetic resonance imaging (MRI)
in early studies, or failure to distinguish venous from arterial cerebral thrombosis
in some. Other studies are also compromised by their small sample size or refer-
ral bias. This results in a quoted risk of CIAO in pregnancy which varies from 1 in
5000 to 1 in 20 000 pregnancies, with the latter figure suggesting that pregnancy
may be associated with only a marginal risk when compared with non-pregnant
subjects (at around 5 in 100 000/women-years).
107
Most CIAO occur in the puerperium or the third trimester and, in >50% of
cases there may be evidence of a major arterial occlusion. In a substantial
number of cases, however, no evidence of major occlusion is found either at
angiogram or post-mortem. A number of common risk factors relate to CIAO in
pregnancy as they do in the non-pregnant. These include hypertension, smoking,
pre-existing premature atherosclerosis and valvular heart disease. In pregnancy,
eclampsia is also associated with small-vessel cerebral complications. CIAO is
likely to be more common in older women, in those with sepsis and in those who
have undergone Cesarean section delivery. However, in many studies, a clear dis-
tinction between ischemic and other cerebral events is not made. In addition,
rarer causes – such as choriocarcinoma, paradoxical embolism, sickle-cell disease,
lupus inhibitor-related thrombosis, ergot alkaloid usage and homocystinuria –
have also been reported in the etiology of pregnancy-related CIAO.
Management of CIAO in pregnancy

The investigation and management of CIAO in pregnancy should follow national
guidelines on the management of CIAO in non-pregnant subjects (see Table 6.5).
This should include, as a minimum, CT or MRI imaging to locate the infarct and
to determine if there is evidence of an associated cerebral hemorrhage.
CIAO may result in a fatal maternal outcome in 0–26% of cases. In one study,
13 recurrences were reported in the 5-year follow of a cohort of 373 women with
a history of CIAO related to pregnancy. The majority of these were observed in
relation to a subsequent pregnancy, with around a 2-fold higher risk associated
with the puerperium when compared with the antepartum period. Interestingly,
the overall outcome of subsequent pregnancies was similar to that expected in
the general population.
Although there is no substantial evidence to support its use, low-dose aspirin
therapy may be beneficial in the prevention of CIAO recurrence during preg-
nancy. In addition, the delivery should be managed to minimize the risk of
venous and arterial thrombosis. In particular, a prolonged second stage of labor
should be avoided and, if there is reduced mobility due to hemiplegia, there may
be a requirement for LMWH thromboprophylaxis. It has also been suggested
that those with a previously identified thrombophilic disorder may benefit from
LMWH therapy in subsequent pregnancies, although the validity of this
approach is not known.
Acute myocardial infarction in pregnancy

Acute myocardial infarction occurring during pregnancy may result in significant
fetal and maternal morbidity. It is, however, a relatively rare event, which occurs
in only 1 in 10 000 women. In general, most myocardial infarctions occur in
multigravid women, and a cumulative maternal mortality rate of 21% has been
reported.
108
Table 6.5 – Management of those with cerebral ischemia of arterial origin (CIAO) or myocardial infarction (MI) in
pregnancy
CIAO Previous Low-dose aspirin Yes

Avoid long second stage Yes
VTE thromboprophylaxis Yes
MI Acute Delay delivery for ≥2 weeks Yes

Thrombolytic therapy ?
Low-dose aspirin Yes
High-dose aspirin ?
LMWH Yes
Nitrates Yes
Beta-blockers Yes
Calcium-channel blockers Yes
Opiates Yes
Magnesium sulfate Yes
ACE inhibitors No
Delivery stress ?Cesarian section
VTE thromboprophylaxis Yes
Maternal death Fetal delivery <15 minutes
Previous Assessment Cause

Current IHD
Evidence of LVF
ACE, Angiotensin-converting enzyme; IHD, Ischaemic heart disease; LMWH, Low-molecular-weight heparin;
LVF, Left ventricular function; VTE, Venous thromboembolism.
Although infarction may occur at any gestation, the majority of deaths occur
at, or within 2 weeks of, the infarction, and often during labor or delivery. In sub-
jects in whom a post-mortem was carried out, coronary atherosclerosis was
evident in 43% of subjects and probable coronary thrombosis, without evidence
of atherosclerotic disease, was evident in 21%. Atherosclerosis was more com-
monly found in subjects with an antipartum myocardial infarction than in those
with a postpartum event. Dissection of the coronary artery was found in 16% of
subjects and was more often associated with infarction in the puerperium. No
detectable evidence of coronary occlusion was found in around 30% of antepar-
tum and 75% of postpartum events. The absence of detectable occlusive disease
may indicate a significant role of transient coronary artery spasm in these sub-
jects.
It is likely that the prothrombotic changes in coagulation, which occur even in
normal pregnancy, increase the risk for arterial thrombosis. These include an
increase in coagulation factors and a reduction in the coagulation inhibitor
protein S (although heritable protein S deficiency is not associated with adult
109
arterial thrombosis). Furthermore, at delivery, an increase in coagulation may be

induced by separation of the placenta, which is the principal source of tissue plas-
minogen activator (tPA) inhibitor. In cases of coronary dissection, the etiology
remains unknown, but may relate to changes in arterial histology (such as alter-
ation in elastic and reticular fibers, and mucopolysaccharide composition) which
are a feature of normal gestation. Such changes may be compounded by pre-exist-
ing risk factors for coronary heart disease, as well as the increasing blood
volume/heart rate, myocardial oxygen requirements, anemia and the reduction in
diastolic blood pressure that are a feature of normal pregnancy.
Diagnosis and management of myocardial infarction in

pregnancy
The diagnosis of myocardial infarction in pregnant women should be achieved in
the same way as in non-pregnant subjects. The use of invasive or radiological
techniques, however, requires an assessment of the balance of risks and benefits
in each case. It is also recognized that electrocardiographic changes may be diffi-
cult to interpret during elective Cesarean section or delivery, and that tachycar-
dia, atrial and ventricular ectopics are a feature of normal pregnancy.
Although the management of myocardial infarction in pregnancy is compara-
ble with that of non-pregnant subjects, the presence of a viable fetus may influ-
ence the modalities of therapy used (Table 6.5). In particular, morphine sulfate
crosses the placenta and may cause neonatal respiratory depression when given
shortly before delivery, although in practice this does not usually cause a signific-
ant clinical problem. Currently, thrombolytic therapy is contraindicated in preg-
nancy due to the potential for teratogenesis. However, there is no evidence that
streptokinase crosses the placenta in late pregnancy or, to a significant degree,
during labor. Indeed, the majority of reports using streptokinase or tPA have
shown a favorable fetal outcome, although maternal hemorrhage and an
increased risk of pre-term delivery and fetal loss has been reported (see below).
As with the management of mechanical valves, heparin is the anticoagulant of
choice in pregnancy. Aspirin therapy, at a low dose of ∼150mg/day, appears safe
in the second and third trimesters of pregnancy, and although aspirin is secreted
in breast milk, this is in low concentrations. The safety of high-dose aspirin
during pregnancy is not established. In general, the use of drugs such as nitrates,
beta-blockers, calcium-channel blockers and magnesium sulfate are generally
considered to be acceptable during pregnancy. However, angiotensin-converting
enzyme (ACE) inhibitors are contraindicated, due to an increased risk of fetal
morbidity and mortality.
If possible, delivery should be postponed until 2–3 weeks after the acute event,
as labor and delivery may increase the risk of myocardial ischemia. There is, cur-
rently, no specific recommendation on the safest mode of delivery and an indi-
vidual approach is required to minimize maternal cardiac workload. However, if
the event occurs close to term, elective Cesarean section is likely to be the safest
110
option for the mother, as this avoids the cardiovascular stress of labor and the
possibility of an emergency Cesarean section, which carries a substantially greater
risk of complications. There may also be a role for continuing LMWH, due to the
increased risk of venous thrombosis that may occur as a consequence of cardiac
failure, immobility or concurrent inflammation. There may be a requirement to
avoid neuraxial anesthesia if a combination of aspirin and clopidogrel is used, as
it may be associated with a more marked increase in the bleeding time than
aspirin alone. When a maternal death has occurred there should be a rapid deliv-
ery if a viable fetus is to be secured.
The risk associated with future pregnancy is dependent on the etiology of the
event, the presence of persisting myocardial ischemia and the degree of residual
impairment of left ventricular function.
Venous thrombosis in pregnancy

As noted above, pregnancy is associated with hypercoagulability, venous stasis
(from pelvic vein compression by the gravid uterus) and the vascular trauma
induced by delivery. It is no surprise then, that in developed countries, venous
thromboembolism (VTE) remains a major cause of maternal morbidity and mor-
tality.
VTE occurs in <1 in 1000 deliveries, but may be 5–10 times more common in
pregnancy than in non-pregnant women of the same age. In addition, those with
a previous thrombosis have a 3.5-fold increased risk of recurrence in a future
pregnancy, with a risk that has been estimated as falling between 0 and 13%. Not
surprisingly, prospective studies produce the more conservative estimates of
recurrence than retrospective analyses. A lower risk of antenatal recurrence has
been observed when the first VTE is associated with a transient risk factor and
there is no evidence of a thrombophilia disorder (2.4%), when compared with
those with either an idiopathic VTE or evidence of thrombophilia (6%). The risk
per day is higher in the first 6 weeks after delivery (the puerperium) than it is
antepartum, with 40% of VTE occurring after discharge from hospital. However,
VTE can occur in any trimester, and occurs in the left leg in around 90% of preg-
nancy-associated cases: a feature that may result from compression of the left iliac
vein by the right iliac and ovarian arteries.
After a pregnancy-related deep-vein thrombosis (DVT) as many as 80% of sub-
jects will experience post-phlebitic symptoms (swelling, skin discoloration,
inflammation and skin ulceration), with around 20% of women required to use
compression stockings up to 10 years after the event. As might be anticipated, the
risk of post-phlebitic symptoms is 2–3 times more likely in those presenting with a
DVT than in those presenting with pulmonary thromboembolism (PTE).
DVT is accompanied by PTE in around 16% of cases, and this is the major
source of mortality. Although the maternal mortality associated with PTE has
reduced in recent years in developed countries, the rate is still of the order of
around 2/million deliveries, with a higher rate in older women.
111
Diagnosis of VTE in pregnancy

It is likely that a clinical suspicion of DVT in pregnancy will be confirmed radio-
logically on only 10% of occasions when compared with 25% of non-pregnant
subjects. Similarly, only 1.8% of suspected PTE in pregnancy have a high proba-
bility ventilation/perfusion (V/Q) scan, when compared with <35% of non-preg-
nant subjects where the clinical suspicion of PTE can be confirmed by objective
diagnostic techniques. This need for an objective diagnosis is particularly import-
ant in pregnancy, as features such as leg edema and breathlessness are not
uncommon. However, the requirement for a definitive diagnosis needs to be bal-
anced with the need to minimize the radiation exposure to the mother and the
fetus that may result from investigations such as CT, angiography, ventilation-per-
fusion scanning or venography.
Clinically, the patient with a VTE may observe unilateral leg swelling and pain,
or the unexpected onset of breathlessness or chest pain. The use of the pre-test
probability assessments (a history of VTE, D-dimer assessment etc), which are
used in non-pregnant subjects to guide investigations, are also important in preg-
nancy, but are unlikely to have the same sensitivity or specificity.
DVT
The absence of elevated D-dimer levels (detected by enzyme-linked immunosor-
bent assay (ELISA) or whole blood agglutination assay) is useful to exclude VTE
in non-pregnant subjects, where they have a high negative predictive value. As
discussed in Chapter 1, elevated D-dimers are seen in around 50% of females in
the first trimester, and may be universal in later pregnancy and the puerperium.
On account of this, their ability to exclude VTE in pregnancy is not yet reliable.
Although a ‘normal’ result may be helpful, D-dimer results in pregnancy do not
remove the need for an objective diagnosis.
Although contrast venography is the gold-standard investigation for DVT diag-
nosis, compression ultrasound assessment is more likely to be used as the initial
assessment. Ultrasound has a high sensitivity and specificity for proximal thrombo-
sis in non-pregnant subjects, but has its greatest utility when used between the
inguinal ligament and the bifurcation of the popliteal vein. If, on compression, the
venous lumen collapses then there is unlikely to be thrombus present. When there
is the possibility of an isolated iliac or calf DVT, compression ultrasound may be of
less value in the exclusion of disease. In general, if there is a clinical suspicion,
treatment should be started and ultrasound scanning of the lower limbs should be
performed. If the scan is negative, and there is a low clinical suspicion, then treat-
ment can be stopped. If, despite a negative scan, there is still a high clinical suspi-
cion then treatment should be continued for 7 days and the scan should be
repeated. When there is a suspicion of isolated calf disease, serial ultrasound assess-
ment over the next 7–14 days, without treatment, has been proposed by some. This
proposal is based upon the low risk of extension to popliteal and femoral vessels
112
(of the order of 20–25%), and the low risk of embolization of isolated calf DVT,
although the general applicability of this approach is not clear. There may also be a
problem in the diagnosis of an isolated iliac thrombosis in pregnancy, as there may
be difficulty in assessing the vein as it passes behind the pregnant uterus. This is,
thankfully, an uncommon diagnostic problem and requires an individual approach
to minimize the risk of fetal radiation exposure from venography or CT scanning.
As a guide, bilateral venography may result in fetal exposure of around 0.6rad
without abdominal shielding, although more limited venography will markedly
reduce exposure (perhaps to <0.05rad). It is thought that in utero exposure to
<5rad does not result in fetal loss or congenital abnormality, but may be associ-
ated with a 2-fold increase in the risk of childhood malignancy. An alternative
technique may be MRI, which appears to give no immediate problems, but its
long-term effects on the fetus are unknown.
PTE
Patients with acute PTE in pregnancy often present with dysponea, but wheeze,
unexplained fever, pleuritc pain, as well as central chest pain, cyanosis and col-
lapse (associated with more massive thrombosis) may all be presenting features.
In such patients, other potential diagnoses such as infection, effusion and pneu-
mothorax, should be excluded by X-ray. Electrocardiogram (ECG) analysis does
not produce a sufficiently specific pattern to diagnose PTE, and blood gas analy-
sis interpretation is complicated by the presence of pregnancy-related respiratory
alkalosis. The sampling procedure itself may also result in bleeding from the
puncture site. Overall, blood gas analysis is probably not warranted in the routine
diagnosis or exclusion of PTE in pregnancy. Indeed, X-ray examinations, blood
gas analysis and ECG are, on their own, insufficient to exclude or diagnose PTE.
To investigate PTE in pregnancy, one potential strategy is to investigate the
lower limbs for evidence of DVT by compression ultrasound. If the results are
positive, treat the patient with a presumptive diagnosis of PTE.
When there is no alternative diagnosis that accounts for the symptoms and
signs, and there is no evidence of a peripheral DVT, V/Q scanning should be
performed. As this may result in fetal exposure of <0.5rad, useful information
can be obtained from a perfusion scan alone in some cases. Avoidance of the
ventilation scan can reduce the fetal exposure to <0.01rad. A positive result will,
however, require a ventilation scan to exclude conditions such as pneumonia.
As in non-pregnant subjects, a V/Q scan will lead to three possible outcomes:
a high probability of PTE (a large ventilation/perfusion mismatch); an interme-
diate probability (small, multiple defects); or a normal result (no perfusion
defect seen). Although, as in non-pregnant subjects, a negative result does not
entirely exclude thrombosis, a normal result will often be considered as evidence
of no PTE. If, however, there is continuing clinical suspicion, or when an inter-
mediate result is obtained, spiral CT should be considered. It may lead to less
radiation to the fetus than standard angiography by the femoral route, and even
113
V/Q scanning, but results in a substantial dose to the maternal breasts of around
2–3rads. In addition, although spiral CT has a high utility for central thrombosis,
it may not reliably detect small peripheral events and has no proven track record
in pregnancy. If there is sufficient diagnostic doubt, pulmonary angiography may
be required. In such cases the brachial route should be preferred, as this results
in a substantial reduction in the potential radiation exposure to the fetus. Many
of the known side effects attributed to pulmonary angiography predate the intro-
duction of low-ionic, low-osmolar contrast medium. Furthermore, pregnant
women are less likely to have major cardiac, respiratory or renal disease, and are
probably less susceptible to such side effects.
Risk assessment of VTE in pregnancy

In determining those at risk of VTE in pregnancy, many of the general risk
factors in the population have limited utility. As noted above, this is due to the
general absence of co-morbid conditions in this population. A number of
factors increase the risk beyond that of pregnancy itself (Table 6.6), which
includes: age, with those over 35 years at higher risk than those below (antenatal
risk 0.06 versus 0.12% and postnatal risk 0.03 versus 0.07%); a family history of
thrombosis; a past history of thrombosis; high parity (>3); operative delivery;
other operative procedures; infection; nephrotic syndrome; pre-eclampsia;
Table 6.6 – Assessment of risks for venous thromboembolism (VTE) in pregnancy
Pre-existing Family history VTE

Past medical history VTE
Thrombophilia
Parity >3
Obesity (>80kg or >25kg/m2)
Paraplegia
Intravenous drug abuse
Sickle-cell disease
Essential thrombocythemia
Nephrotic syndrome
Inflammatory bowel disease
Acquired Ovarian hyperstimulation

Operative delivery
Other operative procedure
Infection
Pre-eclampsia
Abruption/major blood loss
Dehydration
Long-distance travel
114
obesity (weight >80kg or body mass index (BMI) >25kg/m2); dehydration;

inflammatory bowel disease; paraplegia; abruption; IV drug abuse; long-distance
travel; and ovarian hyperstimulation in assisted conception.
As noted above, for those with a previous VTE, one prospective follow-up sug-
gests that the risk of antepartum recurrence relates, predominantly, to those who
have a detectable thrombophilia and/or a previous idiopathic VTE (recurrence
rate around 6%). Whereas, those who have no thrombophilia, and where the
previous VTE occurred in association with a temporary risk factor, appear to be
at a very modest increased risk of antepartum recurrence.
A number of scenarios may present to the attending clinician. These include
pregnant women with:
A family history of VTE in a first-degree relative (+/– known thrombophilia),
but no personal history;
A family history of thrombophilia only;
A personal history of thrombophilia with no personal history of VTE (an
incidental finding);
Previous VTE +/– known thrombophilia.
The risks associated with asymptomatic heritable thrombophilia are difficult
to estimate with confidence from current studies, and any patient requires due
consideration of the above categories. However, a rough guide of the risks associ-
ated with thrombophilia found in someone with no previous VTE or strong
family history is shown (Table 6.7).
In general, those women with a history of VTE should undergo screening for
thrombophilia. This should, if possible, be carried out before pregnancy, to allow
detailed discussion and planning, as well as to avoid the difficulties in assessment
of labile clotting factors in pregnancy (see Chapter 1).
Table 6.7 – Rough guide to thrombophilia and venous thromboembolism (VTE) risk related to pregnancy
Defect Risk of VTE in pregnancy*
FV Leiden heterozygote 1 in 400

FV Leiden heterozygote + family history 1 in 100–200
FV Leiden homozygote 5 in 100
PT 20210A heterozygote 1 in 300
FV Leiden + PT 20210A compound heterozygote 4–5 in 100
Protein S deficiency 1 in 1000
Protein C deficiency 1 in 100
Antithrombin deficiency (mild – severe) 2–7 in 100
*Assumes 1 VTE in 1500 deliveries and an unselected population. AT, Antithrombin; F, Factor; PT, Prothrombin.
115
Prophylaxis of VTE in pregnancy

All pregnancies should be managed to minimize the risk of VTE (see Table 6.8),
and prophylaxis with LMWH should be offered when appropriate (see Tables 6.9
and 6.10). All women with a previous VTE or thrombophilia should be encour-
aged to wear Class II graduated below-knee stockings (18–21mmHg at the ankle)
for the whole of pregnancy and the puerperium. There is no evidence, however,
that there is benefit in serial ultrasound screening of at-risk individuals, as DVT
could occur between scans and, in pregnancy, serial ultrasound is likely to have
<10% positive predictive value for DVT.
When antenatal prophylaxis is used, it should be commenced as soon as is prac-
tical after pregnancy is confirmed. Simlarly, postpartum prophylaxis should be com-
menced as soon as practical after delivery; however, if there has been a postpartum
hemorrhage, therapy should be delayed. In circumstances of high thrombosis and
hemorrhagic risk, there may be a place for intravenous UFH, as it has a shorter half-
life and is more easily, and predictably, reversed with protamine sulfate.
If there has been regional analgesia, LMWH should be withheld until 4 hours
have elapsed from the insertion, or removal, of the epidural catheter. In addi-
tion, placement of an epidural catheter should be carried out at least 12 hours
after a prophylactic dose of LMWH and at least 24 hours after a therapeutic dose.
Furthermore, catheter removal should also be delayed until 12 hours after the
last LMWH dose.
Table 6.8 – Reduction of post-partum venous thromboembolism (VTE) risk
Monitor for clinical signs of VTE in the first week of the puerperium
Early mobilization and hydration
2+ risk factors: Compression stockings
3+ risk factors: LMWH/UFH for 3–5 days
LMWH, Low-molecular-weight heparin; UFH, Unfractionated heparin.
Table 6.9 – Low-molecular-weight heparin (LWMH) dosing*
Standard prophylaxis High prophylaxis Treatment

(24 hourly) (12 hourly) (12 hourly)
Tinzaparin 4500U 4500U 90U/kg

Dalteparin 5000U 5000U 90U/kg
Enoxaparin 40mg 40mg 1mg/kg
*Assumes normal body weight.

116
Table 6.10 – Low-molecular-weight heparin (LMWH) prophylaxis against venous thromboembolism (VTE) in
pregnancy
Scenario Antenatal Postnatal Dose Comment

LMWH LMWH
PMH VTE Yes Yes Treatment/high ≤6 weeks PN

On LT anticoagulant prophylactic
PMH VTE Yes Yes Prophylactic ≤6 weeks PN

No thrombophilia If: previous
• >1 VTE
• Estrogen-linked VTE
• Strong FH
• Idopathic/persisting
risk factor
• Unusual VTE
PMH VTE Yes Yes Prophylactic/ ≤6 weeks PN

+ thrombophilia high prophylactic
(incl APAS) for AT deficiency
Thrombophilia, No Yes Prophylactic ≤6 weeks PN

no VTE Especially if
excluding AT deficient other risk factors
or homozygote/
combined heterozygote
Thrombophilia, Yes Yes Prophylactic ≤6 weeks PN

no VTE
AT deficient or
homozygote/
combined heterozygote
APA only No No/Yes Prophylactic 3–5 days PN

No previous VTE if other risk
or recurrent factors
loss
Multiple other ? Yes Prophylactic 3–5 days PN

risk factors +/– AN
AN, Antenatal; APA, Antiphospholipid antibody; APAS, Antiphospholipid syndrome; AT, Antithrombin;
FH, Family history; LT, Long term; PMH, Past medical history; PN, Postnatal
For those on treatment or high prophylactic doses of heparin, the dose should
be reduced to a standard prophylactic dose the day before a planned delivery and
continued throughout delivery at this dose. If the woman presents in labor, the
treatment dose should be withheld and protamine sulfate used if required.
117
For those on prophylactic therapy, this can be withheld on the morning of an

elective Cesarean section and given 3–4 hours (depending on the time elapsed
from any LMWH injection) postoperatively. Each subject being considered for
Cesarean section should undergo a risk assessment for VTE. Graduated below-
knee stockings should be used and the presence of additional risk factors, such as
obesity, an emergency Cesarian section during labor, older age (>35 years) and
the presence of thrombophilia, should be considered when deciding if, in addi-
tion, heparin prophylaxis is indicated. Indeed, if there are three or more risk
factors present, heparin prophylaxis should be considered for a vaginal delivery
(see Table 6.8).
Travel and thrombosis

An association between immobilization and VTE was first reported in 1940, and a
link with travel was reported in 1954. Although commonly associated with air
travel, the risk appears to relate to any form of travel. It is becoming increasingly
apparent that the risk is limited to journeys that are longer than 4 hours, and
particularly those >8–11 hours. Overall, the increased risk is modest and is likely
to be of the order of 2-fold, with <5% of all symptomatic VTE associated with
travel. The majority of travel VTE is likely to present within 3 days of travel,
although there may be an association of up to 1 month thereafter. In addition to
lower limb immobility and stasis, a number of risk factors peculiar to air travel
have been postulated. These include hypobaric hypoxia and low humidity
(although there is likely to be <100ml of water loss in a long haul flight).
However, such risks could be compounded by excess alcohol or dehydration.
Overall, the risk of travel thrombosis appears to be most important in those with
pre-existing risk factors for VTE. This includes those with a history of previous
VTE, malignancy (within 2 years), a family history of VTE, a history of CIAO and
those who have had recent (<3 months) total hip or knee replacement surgery.
In addition, other factors, such as obesity, venous stasis and estrogen usage and,
of course, pregnancy, should be considered. In small studies of selected popula-
tions, between 20 and 30% of subjects with a symptomatic DVT were observed to
carry a mutation such as FV Leiden. The contribution of this mutation to travel
VTE in those with no other risk factors requires further assessment.
To reduce the risk in pregnant travelers, there are a number of conservative
measures that may be appropriate (Table 6.11). For flights >3–4 hours, the
mother should be advised to carry out leg exercises, to request an aisle seat (as
the majority of VTE related to air travel are seen in subjects sitting in window or
middle seats), to walk around the cabin whenever possible, and to avoid dehyrda-
tion or excess alcohol or caffeine. Pregnant women should also be advised to
inform their insurance companies of the potential risk.
A small number of studies of non-pregnant subjects using graduated elastic
stockings (predominantly 18–21mmHg at the ankle) have shown a reduction in
the occurrence of Doppler-detected DVT related to long haul flights. Although
118
Table 6.11 – Reduction of travel venous thromboembolism (VTE) in pregnancy
Minimize alcohol/caffeine consumption

Avoid dehydration
Regular walks around the cabin and regular stops in car journeys
Isometric leg exercises
Class I to Class II graduated elastic stockings
Consider risk:benefit ratio of aspirin or low-molecular-weight heparin
there is no direct evidence in pregnancy, pregnant women with additional risk

factors should wear such compression hosiery for flights >3–4 hours and, indeed,
such stockings may be useful for prophylaxis, as well as comfort, in all pregnant
women. For those with additional risk factors, chemical prophylaxis should also
be considered and, where required, this should be achieved with LMWH. In such
subjects a dose of, for example, 20–40mg of enoxaparin (or equivalent other
LMWH), given 2–4 hours before each flight, would be appropriate.
Although there is reasonable evidence that aspirin may prevent VTE in the
context of orthopedic surgery, there is no evidence, at present, that it will
prevent travel-associated VTE. One study of long-haul flights has shown a neutral
effect, with 13% of individuals reporting gastrointestinal (GI) side effects.
Another has reported that around 50% of DVT associated with flights (>10
hours) occurred despite aspirin prophylaxis.
VTE treatment in pregnancy

At present there is little direct evidence to inform the management of acute VTE
in pregnancy, with most information derived from non-pregnant cases. Con-
firmed VTE in pregnancy requires immediate therapy with intravenous, treat-
ment-dose, UFH to achieve an APTT of at least 1.5 times the control value. It is
recognized that, in non-pregnant subjects, treatment-dose LMWH is as effective
as UFH, and this is increasingly used as the primary treatment of VTE; this use is
now extending to pregnant subjects. However, given its shorter half-life and
more complete reversibility with protamine, UFH is more appropriate if a VTE
occurs within 2 weeks of the due delivery date. When UFH is used, it should be
started with a 5000IU intravenous bolus, followed by 1000–2000IU/hour, com-
mencing with 1000IU/hour. The APTT or anti-Xa assay should be assessed at
around 6 hours after the bolus and, if a therapeutic dose is required, an APTT
between 1.5 and 2.5, or an anti-Xa of 0.3–0.7IU/ml, should be achieved within
the first 24 hours of therapy.
If LMWH is used as primary treatment, monitoring of the anti-Xa level is of
value, given the changes in LMWH kinetics with increasing gestation. A target of
0.5–1.2 anti-Xa units is desirable at 3–6 hours after dosing, but dosing requires
due consideration of the relative IIa/Xa activity of an individual LMWH.
119
Treatment should be continued for 3–6 months after the event and should
include the puerperium in all cases. Whether treatment should be continued for
6 or 12 weeks after delivery is not clear, but may depend on whether the VTE
occurred early or late in pregnancy. In any case, it is sensible to review all
patients at 3 months to determine if there are persisting risk factors. Whether it is
safe to reduce the dose of heparin after the first 2–3 weeks of therapy is also not
clear, but a change to LMWH with, or without, a dose reduction is widely prac-
ticed, especially if there is a high risk of osteoporosis or hemorrhage. In all cases,
the platelet count should be measured at least after 7 days of therapy, and
perhaps more frequently if there is to be prolonged use of UFH. In those with an
antenatal VTE, an individual decision is required as to whether heparin or war-
farin should be used for continuing treatment in the puerperium.
Vena cava filters

There is limited experience of vena cava filters in the management of pregnancy-
related VTE. However, as in the non-pregnant, the general indications are:
emboli that recur despite adequate anticoagulation; a significant contraindica-
tion to full-dose anticoagulation after a VTE; or marked patient debility, where it
is perceived that a further embolic event may be fatal. The value and safety of a
filter to permit a reduction in the dose of heparin when VTE occurs close to
delivery, and the need for a filter solely on the basis of a VTE in someone with
patent foramen ovale, both require further study.
Although early filters carried a significant risk of caval thrombosis (around
85%) and persisting lower limb venous insufficiency, more recent devices are
conical in shape: this allows formed clot to be concentrated at the apex of the
device and permits continued flow across the filter, which should lead to physio-
logical lysis of any formed clot. Such newer devices have been reported to have a
long-term patency rate of 95% and a risk of recurrence of PTE of <4%. However,
the lifelong placement of such filters has been associated with a number of com-
plications, including: device failure; device migration into the right atrium; pene-
tration through the vessel wall; an increased risk of pelvic or leg thrombosis and;
the possibility of small emboli occurring through small or collateral vessels.
From the limited experience of newer devices in pregnancy, they are associ-
ated with a favorable maternal and fetal outcome, although, around 25% of sub-
jects may have evidence of persisting lower limb edema well after delivery. More
recently, there have been a number of case reports of retrievable filters being
used when VTE occurs at, or near, delivery. The overall role and safety of this
modality of therapy in pregnancy is not yet clear.
Thrombolytic therapy and embolectomy

In those patients with PTE accompanied by cardiac compromise (or perhaps
PTE with right ventricular dysfunction alone), treatment with thrombolytic
120
therapy, or embolectomy, should be considered.

Thrombolysis has been achieved with streptokinase and recombinant tPA
(rtPA) in pregnancy. Both of these activate plasminogen to plasmin, which
cleaves fibrin (as well as fibrinogen, FVIII and FV).
As noted in Chapter 1, tPA preferentially activates plasminogen in the pres-
ence of fibrin and this should result in a more local lytic action. rtPA does not
cross the placenta and is not antigenic. Streptokinase does cross the placenta, but
not to a degree that results in lysis in the developing fetus. However, streptoki-
nase is antigenic and at least 6 months should elapse between doses. One further
lytic agent is urokinase (see Chapter 1), which does cross the placenta, but it is
not clear whether this causes a significant problem in the fetus.
From the limited case-report evidence of thrombolysis in pregnancy
(employed for a variety of indications), a maternal hemorrhage rate of between 1
and 8% is to be expected. There is low maternal mortality (<1%), which is
broadly comparable with that in the non-pregnant. However, the use of any such
therapy is associated with a 2–6% fetal loss rate and a comparable risk of prema-
ture delivery. Such data must be interpreted with caution, however, as they may
be influenced by the severity of the indication for such therapy. From the limited
reports, there is no clear evidence of superiority of any one particular agent in
the management of VTE.
One further advance is the use of catheter-directed thrombolysis. This
requires expertise in pulmonary artery catheterization and there is a need for
maternal radiation exposure to permit catheter placement. However, at present,
there is insufficient evidence to suggest that this approach is superior to standard
therapy in the management of either DVT or PTE in pregnancy.
There is very little information on the use of embolectomy in pregnancy. Its
use has been limited to severe maternal disease and is associated with a signific-
ant fetal loss rate of 20–40%, related to the need for cardiac bypass. However,
this may be of secondary importance given the severity of the maternal illness.
References
Chan et al. 2000. Anticoagulation of pregnant women with mechanical heart
valves: a systematic review of the literature. Arch Int Med, 160, 191–6.
121
CHAPTER 7
Thrombophilia and
pregnancy outcome
Serious pregnancy complications, including pre-eclampsia (PET), fetal loss,
intrauterine growth restriction (IUGR) and placental abruption, occur in up to
5% of women of reproductive age, and may be related to alterations in placental
perfusion and development. In the past few years maternal carriage of a number
of heritable thrombophilias, or thrombophilias with a mixture of heritable and
environmental influences, have been linked with such pregnancy complications.
Although a link between the thrombophilias and pregnancy-associated venous
thromboembolism (VTE: Table 7.1) is consistent with their potential to generate
an excess of thrombin and fibrin. The mechanism for their role in other
Table 7.1 – Relative risk of venous thromboembolism (VTE)
Event Relative risk
Pregnancy 4
Puerperium 14
Cancer 10
Medical immobilization 11
Postoperative 6–10
COCP 4–6
HRT 2–4
FVIIIC 4–6
HCY 2–4
PC/PS/AT deficiency Around 10
Family history Around 10
Heterozygous FV Leiden 5–8
Homozygous FV Leiden 50–80
Heterozygous PT 202010A 2–4
Homozygous PT 202010A 10
Antitphospholipid syndrome 7–9
AT, Antithrombin; COCP, Combined oral contraceptive pill; F, Factor; HCY, Homocysteine; HRT, Hormone
replacement therapy; PC, Protein C; PS, Protein S; PT, Prothrombin.
122
Thrombophilia and pregnancy outcome
pregnancy disorders is not so obvious, but could also be linked with alterations in
both thrombin generation and endothelial function.
As detailed in Chapter 1, thrombin is the pivotal protein in the coagulation
process. In addition to a wide variety of pro-coagulant functions (including
platelet activation, conversion of fibrinogen to fibrin and activation of Factors
(F) Vc and VIIIc), it also has an anticoagulant function via thrombomodulin and
the generation of activated protein C (aPC). Thus, via this negative feedback,
thrombin ultimately regulates its own generation. Adequate anticoagulant func-
tioning of thrombin via aPC depends upon intact functioning of the endothe-
lium, adequate levels of the components of the protein C/protein S system and a
normal sensitivity to the effects of aPC. Thrombin also participates directly in
tissue remodeling, wound repair, leukocyte chemotaxis, leukocyte adhesion, vas-
cular contraction and vascular permeability, via interactions with specific
endothelial and leukocyte thrombin receptors.
Although endothelial activation is a feature of normal pregnancy, heightened
activation has been linked with disorders such as PET. Any condition, such as a
thrombophilia, that has the potential to increase thrombin generation, or alter
endothelial function, may be capable, therefore, of influencing placental func-
tion and thereby fetal development. Maternal thrombophilia may, by excessive
thrombin generation, also alter the maternal response to the feto-placental unit.
This may, for example, contribute to the maternal presentation of PET. Whether
fetal inheritance of paternal thrombophilia contributes to these conditions has
not been fully examined, but could also be important in the generation of these
fetal and maternal disorders.
Heritable thrombophilias
Protein S deficiency
Protein S is a vitamin K-dependent, single-chain glycoprotein, which is synthe-
sized in the liver and vascular endothelium, and acts mainly as a cofactor to aPC
in the inactivation of FVIIIa and FVa (Figure 7.1). The plasma level of protein S
depends upon age, sex, lipid levels, estrogen, oral anticoagulant usage and the
presence of acute thrombosis (Table 7.2). In the plasma, around 60% of circulat-
ing protein S is bound to C4b binding protein, and only free protein S can func-
tion as a cofactor to aPC. The binding protein itself may also function as a link
between thrombosis and complement activation. Indeed, as C4b increases in
pregnancy, the binding protein also contributes to the physiological reduction in
protein S levels seen with increasing gestation.
Heritable protein S deficiency is transmitted as an autosomal dominant trait.
Those with heterozygous deficiency are at increased risk of VTE, as well as war-
farin-induced skin necrosis. Homozygote deficiency is extremely rare, as it is
usually associated with neonatal purpura fulminans and fetal or perinatal death.
However, there are reports of homozygous individuals who have survived long
123
enough until life-long anticoagulation can be established. Although there is no

agreed classification of inherited protein S deficiency, three subclasses have been
proposed. These result from mutations which lead to either:
A reduction in total and free protein S antigen levels;
A reduction in free protein S antigen and activity, without a reduction in the
total protein S level;
A reduction in protein S activity only, without a reduction in antigen.
However, as some functional protein S assays give falsely low protein S activity
assessments when FV Leiden is present, there is no universally accepted (or prac-
tical) definition of functional protein S deficiency.
With allowance for this, the prevalence of heterozygous protein S deficiency
in the general population is of the order of 1 in 300–400. Heterozygous defi-
ciency carries around a 10-fold risk of VTE, but is found in <5% of those with a
history of venous thrombosis. At present, however, there is no substantive evid-
ence which links protein S deficiency and arterial thrombosis in adults. As there
are a number of mutations which can result in protein S deficiency, the diagnosis
(unless a specific mutation is known in other family members) is made by assess-
ment of plasma protein S levels (Table 7.2).
Protein C deficiency
Like protein S, protein C is also a vitamin K-dependent glycoprotein. It is synthe-
sized in the liver and, when activated, inactivates FVa and FVIIIa in the presence
of protein S. Its plasma levels are influenced by age, sex, lipid levels, and the
presence of liver disease, renal disease, acute thrombosis, disseminated intravas-
cular coagulation (DIC) or warfarin use. Higher levels can be seen in the puer-
perium and in subjects on oral contraceptives (see Table 7.2).
As with protein S, heritable deficiency is transmitted as an autosomal trait and
heterozygous deficiency increases the risk of thrombosis around 10-fold. The
Figure 7.1 Thrombophilia
The principal points of the coagulation cascade that are associated with thrombophilia are shown. The actions of
antithrombin (AT – previously called antithrombin III), which are reduced by AT deficiency are shown. Protein C
(PC), via activated protein C (aPC) and its cofactor protein S (PS), acts to inactivate Factor (F) Va (Va) and FVIIIa
(VIIIa). The action on FVa is impaired when there is the FV Leiden (FVL) mutation, and the action on FVIIIa is
impaired when there is an elevation of FVIIIc. This occurs in pregnancy, during the acute phase reaction and also
in association with the combined oral contraceptive or hormone replacement therapy. Elevated FIXc levels are
also an independent risk factor for venous thromboembolism (VTE). Elevated fibrinogen levels and antibodies to
phospholipid are associated with an increased risk of VTE and arterial thrombosis. Ca2+, Calcium; Plipid,
phospholipid; PTG2021OA, prothrombin G20210A mutation; TAT, thrombin–antithrombin complex; TF, tissue
factor.
124
INTRINSIC
XII XIIa
XI
Ca2+
IXc XIa

XIi VIIIc
Ca2+/Plipid
TM
IXa aPC PC
(EPCR)
FVL +PS
125
X VIIIa VIIIi +PS
TF–FVII–FVIIA
EXTRINSIC
Ca2+/Plipid PTG20210A Va Vi
Tenase Fibrinogen
Xa
AT THROMBIN
Ca2+/Plipid burst Fibrin
Xi
Prothrombinase
TAT

fibrin
Table 7.2 – Factors influencing plasma thrombophilia assessments
Thrombophilia Factor Effect

Hyperhomocysteine Sample anticoagulant Different HCY levels with different
(hyperHCY) anticoagulants (citrate or EDTA) or
anticoagulant concentrations
Time to plasma separation Separate plasma <1 hour to avoid higher
HCY from ongoing wbc metabolism
Fasting versus non-fasting Role not defined in predicting risk
Methionine loading versus Role not defined in predicting risk
no methionine loading
Pregnancy Reduces HCY level
Renal impairment Increases HCY level
Hypothyroidism Increases HCY level
Haematinic deficiency Increases HCY level
Contraceptives Reduces HCY level
Protein C (PC) activity Acute thrombosis Reduces PC level
Coumarins Reduces PC level
Pregnancy No gestation effect but non-pregnant
ranges may not be appropriate
Levels may increase in early puerperium
Hyperlipidemia Increases PC level
Protein S (PS) antigen Acute thrombosis Reduces PS level
Coumarins Reduces PS level
Pregnancy Reduces PS level
AT activity Acute thrombosis Reduces AT level
Pregnancy No gestation effect but non-pregnant
ranges may not be appropriate
Heparin Reduces AT level
Lupus inhibitor Anticoagulants Unreliable result (?except low-molecular-
weight heparin)
aPCR Anticoagulants Unreliable result
(APTT-based) Blood group non-O Increases aPCR
Pregnancy Increases aPCR
Oral contraceptives Increases aPCR
Hormone replacement therapy Increases aPCR
Body mass index Increases aPCR
Blood pressure Increases aPCR
aPCR Anticoagulants Unreliable result
(thrombin-generation) Oral contraceptives Increases aPCR
Factor (F) VIIIc Blood group non-O Increases FVIIIc
Pregnancy Increases FVIIIc
Contraceptives Increases FVIIIc
Acute phase response Increases FVIIIc
Warfarin No effect
126
homozygous state is also associated with neonatal purpura fulminans. In such cir-
cumstances the use of human-derived protein C concentrate may rescue such
individuals until lifelong anticoagulation can be established. Heterozygous defi-
ciency has been estimated to occur in 1 in 300–500 of the population, but is also
found in <5% of those with venous thrombosis. As with protein S, there is no
evidence that protein C deficiency is linked with arterial thrombosis in adults. In
studies investigating blood donor populations, the majority of individuals identi-
fied (either biochemically or genetically) are symptom free. Protein C deficiency
has been classified into two types:
Type 1: a quantitative reduction in functionally normal protein C;
Type 2: a level of protein C activity that is less than the antigen level; a defi-
ciency that results from the production of a functionally abnormal protein C
molecule.
These deficiencies can result from a number of genetic mutations and, as a
result, the diagnosis of protein C deficiency is made from the detection of
reduced plasma levels, unless there is a known mutation in other family members
(see Table 7.2).
Antithrombin deficiency
Although originally termed antithrombin III, the use of III is now considered
redundant. This glycoprotein is capable of inactivating most of the activated clot-
ting factors (including FVIIa bound to tissue factor (TF)). The presence of
heparin increases its ability to inactivate FXa and FIIa (thrombin) around 1000-
fold. Functional plasma levels are reduced in acute thrombosis, DIC, and in
patients receiving heparin therapy. Otherwise, there is little variation in plasma
levels with age and sex, and levels often remain within the non-pregnant range
during pregnancy.
Antithrombin deficiency is inherited as an autosomal trait, which, in its severe
heterozygous form, increases the risk of VTE around 20-fold and in its less severe
form around 10-fold. A homozygous severe deficiency is probably incompatible
with survival, but milder homozygous deficiencies may be amenable to replace-
ment with human-derived or recombinant antithrombin concentrate until life-
long anticoagulation can be established. A number of genetic mutations, which
alter either the clotting or heparin binding sites, are described. More severe
defects reduce antigen and activity equally, and lesser defects result in less activity
than antigen. Akin to the situation with protein C and S, the number of potential
mutations means that deficiency is routinely diagnosed on the basis of reduced
plasma levels. Again, in some instances, it may be possible to detect, or exclude, a
mutation that is known to have occurred in other family members. The preva-
lence of defects in the general population leading to a severe deficiency is of the
order of 1 in 4000–5000, although milder defects are more common and may be
as prevalent as 1 in 400. As with protein C and S deficiency, antithrombin is
127
found infrequently in those with VTE (<3%), but is more often associated with
more severe disease, or disease onset at an early age. Indeed, around 50% of
severe heterozygotes experience thrombosis before the age of 30 years. Akin to
the situation with protein S and protein C, there is no evidence that antithrom-
bin deficiency is associated with arterial thrombosis.
FV Leiden
FV Leiden occurs as a result of a G→A mutation at position 1691 in the gene
coding for coagulation FV. It is the commonest heritable cause of resistance to
aPC (see below) and results in a change in one of the parts of the FV molecule
that is the target for cleavage by aPC. It is inherited as an autosomal dominant
condition with both heterozygotes and homozygotes surviving into adulthood.
Heterozygotes are reasonably common in European populations, with a preva-
lence of between 2 and 15%, but are less common, or even absent, in other pop-
ulations. In those of European extraction, FV Leiden is found in its homozygous
form in around 1 in 1000. Heterozygotes are also found in around 50% of those
with a history of VTE. In heterozygotes it carries around a 5-8-fold risk of VTE,
with a higher risk (around 50–80-fold) in homozygotes, and an increasing risk
associated with increasing age (Table 7.3). Given its high prevalence, it is not
unusual to find it in combination with other thrombophilic disorders. Whether,
in combination with smoking, it has a causal role in myocardial infarction is a
matter of topical debate, but there is no evidence linking FV Leiden with arterial
thrombosis in any other circumstances. Other heritable causes of aPC resistance
Table 7.3 – Age and absolute venous thromboembolism (VTE) risk
Age Subgroups VTE risk
Child All 1/million/year
<45 years All 1/10000/year

heterozygous Factor (F) V Leiden 1/1500/year
on COCP 1/2500/year
heterozygous FV Leiden on COCP 1/350/year
45–75 years All 1/1000/year

Menopausal 1/500/year
on hormone replacement therapy (HRT)
Menopausal + heterozygous FV Leiden 1/150/year
Menopausal + heterozygous FV Leiden on HRT 1/70/year
>75 years All 1/100–400/year
COCP, Combined oral contraceptive pill

128
include two other point mutations in the FV gene (FV Cambridge and FV Hong
Kong) and a group of FV polymorphisms, known as the HR2 haplotype. At
present, the clinical impact of these alterations is unknown.
Prothrombin G20210A mutation

A mutation in the gene coding for the prothrombin molecule was described in
1996 (Poort SR, et al, 1996). This results from a single base substitution (G→A) at
position 20210, which leads, by an as yet unknown mechanism, to higher levels of
circulating prothrombin. It is found in 1–2% of Caucasians, but is more common
in those of Mediterranean descent. It confers a 2–4- and around 10-fold risk of VTE
in heterozygotes and homozygotes, respectively. There is also likely to be an even
higher risk in those who are compound heterozygotes for the prothrombin
G20210A and FV Leiden mutations. Like FV Leiden, any potential link between
the mutation, smoking and myocardial infarction in young women is not yet
resolved, and there is no link with arterial thrombosis in other circumstances.
Thermolabile methylene tetrahydrofolate reductase (MTHFR

C677T)
As noted in Chapter 3, the manufacture of DNA results in the generation of
homocysteine (HCY). The blood level of HCY is under the influence of the
enzymes and hematinic cofactors of the HCY – methionine pathway. One of the
enzymes of this pathway is methylene tetrahydrofolate reductase (MTHFR), and
a specific–point mutation in its gene results in relative thermolability of the
enzyme called MTHFR C677T. This mutation occurs in the homozygous form in
5–15% and in its heterozygous form in up to 35% of Caucasians. At present,
there is insufficient evidence to suggest that this mutation has a role, independ-
ent from HCY, on the occurrence of venous or arterial thrombosis. For relation-
ship to pregnancy complications see ‘Hyperhomocysteinemia’ (p. 135).
Dysfibrinogenemia
See Chapter 4.
Heritable thrombophilia and pre-eclampsia (PET)

PET is a multisystem disorder that is characterized by hypertension, renal dys-
function and fluid retention during pregnancy in someone who has no pre-exist-
ing hypertension or renal dysfunction. It is defined as the development of
hypertension (defined as a blood pressure of ≥140/90mmHg, or a blood pres-
sure that is ≥30/15mmHg over the subject’s baseline level on at least two occa-
sions, 6 hours apart) accompanied by the development of proteinuria (defined
as >300mg/24 hours, or 30mg/dl on a confirmed random sample), indicating
129
renal damage. Formal quantification of the level of proteinuria is most often

assessed by obtaining a 24-hour collection. In practice, however, the diagnosis of
significant proteinuria is often accepted as the presence of >2+ proteinuria on
repeated urinary dipstick testing. Once established, PET progresses at a variable
and unpredictable pace until delivery. Both the hypertension and the protein-
uria resolve postnatally.
Although hypertension without proteinuria (i.e. pregnancy-induced hyperten-
sion) can lead to significant disease, it carries a much better prognosis for the
mother and the fetus than PET. At the other extreme, around one third of sub-
jects with PET present with seizures. This, termed eclampsia, can be complicated
by coma, neurological deficit and, in a small percentage, with intracerebral hem-
orrhage. It is due to cerebral involvement of the disease and is thought to involve
vasospasm leading to ischemia, disruption of the blood–brain barrier and cereb-
ral edema. Forty-four per cent of cases of eclampsia occur postnatally, 38% in the
antepartum period and 18% intrapartum.
Hemolysis, elevated liver enzymes, (HELLP) syndrome and low platelets is a
serious manifestation occurring in 4–12% of women with PET. Hemolysis reflects
microangiopathic hemolytic anemia (MAHA) and the elevated transaminases
reflect liver dysfunction occuring through vascular damage. Thrombocytopenia
reflects a coagulation disturbance, although in HELLP syndrome the main coag-
ulation parameters (activated partial thromboplastin time (APTT), prothrombin
time (PT) and thrombin clotting time (TT) remain within normal limits. Since
elevated blood pressure is not always present at the onset of this condition, it may
be confused with other conditions causing thrombocytopenia or abnormal liver
function tests. HELLP syndrome is more likely to occur in multiparous women.
The infants of HELLP syndrome mothers also have an increased risk of thrombo-
cytopenia, although this most likely results from a combination of fetal stress and
ineffective platelet development.
The presence of hyperuricemia can be useful diagnostically, as it can help dis-
tinguish between these pregnancy-induced hypertensive disorders and pre-exist-
ing hypertension. An elevated plasma uric acid level also occurs before the onset
of proteinuria and is a useful marker of disease severity, since it increases with
disease progression. The normal range for serum uric acid levels is, however,
lower during pregnancy than in non-pregnant women. Hence, at 32 weeks’ gesta-
tion, the upper limit of normal is 0.34 µmol/l and at 36 weeks gestation is
0.39µmol/l. Raised serum uric acid levels during pregnancy are best regarded as
an indicator of impaired renal function and renal blood flow. In women with
PET, a rising plasma urea or creatinine indicates a worsening of the disease. A
low platelet count occurs in normal pregnancy (see Chapter 8) and a falling
platelet count is seen in PET, but this is an inconsistent feature of the disease.
PET remains a poorly understood disorder that complicates around 3% of
pregnancies. A number of maternal constitutional risk factors have been
described (Table 7.4). In addition, a multiple pregnancy, maternal age (<20 or
>35 years), a history of previous PET, a maternal family history of PET, the pres-
130
ence of diabetes mellitus, migraine, as well as some fetal factors, such as hydrops
fetalis, hydatidiform mole, triploidy and trisomy 13, are also linked with a higher
risk of PET. In recent years a link with various inherited and/or acquired throm-
bophilias has been made (see Table 7.4). The etiology of the condition remains
unknown, but is generally accepted to be related to abnormal placentation, and
probably represents a failure of the normal immune tolerance that the mother
should have to the presence of the fetus.
In addition to the similarities between the placental pathology of PET (acute
atheroma – a necrotizing arteriopathy with fibrinoid necrosis and lipid-laden
macrophage accumulation in the placental bed) and cardiovascular pathology,
PET is accompanied by acute changes in coagulation, including a supraphysio-
logical increase in von Willebrand’s Factor (vWF), FVIIIc and TF expression, as
well as platelet activation and a reduction in antithrombin activity. These changes
can progress to encompass the syndrome of disseminated intravascular coagula-
tion. PET also carries an increased risk of IUGR, and women with a past history
of PET appear more likely to have higher circulating levels of coagulation
factors, lipids and insulin resistance. This may be reflected in a higher risk of car-
diovascular disease in later life for the mother, as well as a higher risk of vascular
Table 7.4 – Pre-eclampsia (PET) risk
Risk factor Relative risk
Thrombophilia AT/protein C (PC) deficiency ?

Factor (F) V Leiden 2–5*
Hyperhomocysteine 5†
MTHFR C677T 2*
Protein S 10*
APA 4*
Prothrombin G20210A 2–7*
Acquired activated PC resistance 7†
Constitutional PMH chronic HBP 10

Primigravida/primipaternity 5
Body mass index (BMI)/waist circumference 2.2–2.7‡
PMH pre-eclampsia 8
Smoking 0.5
*The risks associated with individual thrombophilias relate predominantly to retrospective studies, and in many
cases these have not been confirmed in all studies and particularly in prospective series.
†
The relative risk relates to the risk in the highest quartile compared with the lowest.
‡
The relative risk of PET with a waist circumference (at pregnancy presentation) of 80cm compared with a waist
circumference of <80cm, or of a BMI >26kg/m2 (at pregnancy presentation) when compared with a BMI <26kg/m2.
APA, Antiphospholipid; AT, Antithrombin; HBP, High blood pressure; PMH, Past medical history
131
disease and Type 2 diabetes mellitus in later life in those infants with a history of
IUGR.
A link between a history of previous VTE and the risk of PET has been
reported in one study, and several case-control studies have found at least one
inherited thrombophilia in anything up to 70% of women with PET (when com-
pared with <20% in control subjects). A number of studies have specifically
examined aPC resistance (both FV Leiden and acquired) and the risk of PET. In
those studies reporting a positive link, the magnitude of the risk attributed to
heterozygous FV Leiden is between 2- and 5-fold. In some of these studies it has
also been suggested that thrombophilia may be associated with more severe PET.
In the main, however, a positive link between thrombophilia has only been seen
in retrospective studies, and prospective studies have not confirmed any link.
Consequently, further work is required to determine whether this disparity in
evidence relates to differences in the prevalence of FV Leiden in the different
populations, differences in the severity of PET in different studies, or to a more
accurate diagnosis in prospective investigations.
The prothrombin G20210A mutation has also been linked to an increased risk
of PET (between 2- and 7-fold) in a small number of studies, although the major-
ity of reports have not found any association. Protein S deficiency has also been
found in up to 25% of subjects with a history of PET. However, as there is a physi-
ological fall in protein S in normal pregnancy, the exact contribution of abnor-
mal levels to the development of PET requires further study. At present, due to
the relative rarity of the disorders, there is insufficient evidence to give an estim-
ate of risk for those with antithrombin (AT) or protein C deficiency. Although
MTHFR C677T has also been linked with around 2-fold risk of PET in some
reports, the majority of studies have not found an independent role for this
genetic variant. These studies suggest that levels of the hematinic cofactors of the
HCY–methionine pathway (such as folic acid) may be considerably more import-
ant than the C677T in the causation of disease.
Although this potential involvement of heritable thrombophilia in the causa-
tion of PET is intriguing, and may shed light on the pathogenesis of this disease,
there is no evidence that identification of thrombophilia carriers would reliably
identify those at risk of developing PET. Furthermore, there is no substantial
evidence that reduction of thrombin generation in such individuals, by the use of
anticoagulants, would result in a beneficial pregnancy outcome. At present, then,
there is no case that routine, or even selected, thrombophilia screening of indi-
viduals (based upon past medical or obstetric history) would be of benefit in the
management of this condition.
Heritable thrombophilia and intrauterine growth restriction (IUGR)

A few studies have specifically addressed the impact of heritable thrombophilia
on IUGR. In some studies, carriage of a thrombophilia has been associated with
at least a 2-fold risk of IUGR (most commonly defined as growth to less than the
132
10th centile), with a relative 4–7-fold risk for carriage of either FV Leiden or pro-
thrombin G20210A. However, as with other pregnancy complications, a number
of studies have not shown any link between IUGR and heritable thrombophilia.
Further work, perhaps concentrating on births of <5th centile, are required
before a definite link with thrombophilia can be proven.
Heritable thrombophilia and fetal loss

Pregnancy loss occurs in 12–20% of first pregnancies, with two losses affecting
around 5% of women, and three or more losses (recurrent pregnancy loss)
affecting around 2% of women of reproductive age. It is likely that the majority
of losses related to immune, karyotypic, endocrine or other fetal/maternal
defects and are not directly related to thrombosis. However, the placenta in fetal
loss may show evidence of infarction, thrombosis, and/or perivillous fibrin depo-
sition. Indeed, these changes can be seen in fetal loss associated with throm-
bophilia, but they are not always evident and are reasonably common even in the
absence of any known thrombophilia.
On the basis of limited evidence protein C, protein S and AT deficiency have
all been associated with around a 2-fold risk of recurrent pregnancy loss. Fetal
loss also appears to be more common in those members of a thrombophilic
family who have the deficiency when compared with those family members who
do not. Although there is likely to be a link with fetal loss, because of the relative
rarity of these deficiencies, the magnitude of the risk and whether the risk relates
exclusively to late or early loss is not yet resolved.
In the majority of studies, a 2–7-fold increased risk of recurrent loss has been
associated with carriage of the FV Leiden mutation, suggesting that FV Leiden is
an independent risk factor for fetal loss. Overall, it appears that this risk is more
likely to be related to pregnancy loss occurring after the first trimester (in
particular after 24 weeks). The risk of loss is also likely to be higher in subjects
who are homozygote for FV Leiden (perhaps two times greater than heterozy-
gotes), or those who have a combination of defects (such as FV Leiden and
antiphospholipid syndrome, where the risk may be as high as 14-fold). Despite
this, the vast majority of subjects who are heterozygous for FV Leiden have a
normal pregnancy outcome. This suggests that other factors, perhaps including
fetal FV Leiden inheritance, may play a part in fetal loss associated with FV
Leiden.
So far, the role of the prothrombin G20210A mutation in fetal loss has not yet
been confirmed, although some studies suggest that it too may be associated with
a 2–3-fold risk of all recurrent pregnancy losses, including both single late and
recurrent early loss. This may be particularly so if the prothrombin G20210A
mutation is combined with other defects.
It is unlikely that heterozygous carriage of the MTHFR C677T variant results
in any increase in risk of pregnancy complications. Although a few studies have
reported that homozygous carriage of the mutation increases the risk of placen-
133
tal abruption and recurrent pregnancy loss, the vast majority of studies and met-
analyses have found that, despite a 3–4 increased risk of recurrent fetal loss asso-
ciated with hyperhomocysteinemia (see below), there appears to be little
independent contribution from the MTHFR C677T variant.
Thrombophilia with genetic and acquired components

Activated protein C resistance (aPCR)
Resistance to the anticoagulant effects of aPC (aPCR) was originally described in
association with antiphospholipid antibodies. Subsequently, an examination of
the effect of exogenously added aPC on the APTT of subjects with a personal or
family history of VTE demonstrated that aPCR could be inherited. The addition
of aPC (by inactivating FVa and FVIIIa in the sample) should lead to a prolonga-
tion of the APTT. In subjects with FV Leiden, addition of aPC does prolong the
APTT (as FV Leiden produces a relative, not absolute, resistance to aPC), but to
a much lesser degree than in those not carrying the mutation. Indeed, this forms
the basis of an APTT-based assay to detect the presence of FV Leiden. This test is,
however, also influenced by other factors. In particular, blood groups other than
O (non-O), higher plasma FVIIIc levels, combined oral contraceptives, preg-
nancy, hormone replacement therapy, antiphospholipid antibodies and the pres-
ence of the prothrombin G20210A mutation all lead to greater aPCR when
assessed by this test. The test is also unreliable in the presence of anticoagulants.
To render the APTT test more specific for heritable mutations in the FV gene,
the patient’s plasma sample can be pre-diluted with FV-deficient plasma. This
modification excludes these other influences, including anticoagulants, render-
ing it highly sensitive and specific for mutations in FV such as FV Leiden.
There are other methods for detecting aPCR. One method is an assay essentially
based upon the PT rather than the APTT. In this test the PT is modified so that the
end point is the amount of thrombin (not fibrin) generated with time. To detect
FV Leiden, the amount of thrombin generated after the sample is activation by
FVIIa/TF is assessed. The test is repeated with exogenously added aPC, and the dif-
ference in the amount of thrombin generated leads to an aPCR index. This test is
particularly sensitive (more so than the APTT-based test) to the different levels of
aPCR generated by the second and third generation oral contraceptive pills.
Both the unmodified APTT- and thrombin generation-based tests may be
capable of giving more information than simply the detection of genetic defects
such as FV Leiden, and the term ‘acquired aPCR’ has been coined for resistance
which is found without FV Leiden (or other mutations in FV). Indeed, these tests
may have, ultimately, a value in assessing overall arterial and venous thrombotic
potential. However, further work is required to standardize these tests before
they can become useful for the prediction of disease. Preliminary evidence does
suggest, however, that higher aPCR is associated with an increased risk of both
arterial and venous thrombosis.
134
aPCR and pregnancy outcome

As noted in Chapter 1, in around 80% of normal subjects, increasing gestation
leads to increasing acquired aPCR (as detected by the APTT method). In preg-
nancy, higher levels of acquired aPCR correlate with higher FVIIIc and FVc, and
lower protein S levels. In one study, those with the greatest resistance in early
pregnancy had a 7-fold increased risk of PET developing later in that pregnancy.
There may also be a link between the degree of acquired aPCR and fetal weight,
although no relationship between aPCR (when assessed during pregnancy) and
fetal loss has yet been described. However one study has reported that acquired
aPCR (assessed after pregnancy) is more common in those with a history of PET
or fetal loss. Although these observations are interesting, standardization of the
assay techniques is required before any clinical use can be made of the APTT-
based aPC resistance in the detection or prevention of disease.
At present there is also very little information on the relationship between
acquired aPCR (assessed by thrombin generation) and pregnancy outcome,
although preliminary results suggest that higher resistance in early pregnancy
does not predict an increased risk of PET.
Hyperhomocysteinemia
Mild–moderate hyperhomocysteinemia (hyperHCY) is found in around 5% of
North European populations, and may be twice as common in those with a
history of VTE and 2–5 times more common in those with atherosclerotic
disease. Although this is not a substantial risk, it is of interest, as hyperHCY may
be amenable to treatment with folic acid. However, the evidence that such inter-
vention reduces the risk of coronary artery disease is disputed and, at present,
there is no evidence at all that folic acid will substantially reduce the risk of recur-
rence of VTE. As noted in Chapter 3, the level of HCY is dependent on several
key enzymes in the HCY-methionine pathway, as well as the level of the vitamin
cofactors of these reactions (vitamins B12, vitamin B6 and folic acid).
HyperHCY and pregnancy outcome

As detailed in Chapter 3, hyperHCY is associated with an increased risk of
spina bifida and around a 2-fold risk has also been observed in homozygotes
for the MTHFR C677T variant, with a similar size of risk observed with other
polymorphisms in the HCY-methionine pathway. As the use of routine peri-
conceptual folic acid reduces the risk of neural tube defects by around 50%, it
would seem likely that the risk of neural tube defects depends not only on the
presence of these polymorphisms but also on the mother’s hematinic cofactor
levels.
There is some evidence that links hyperHCY with, predominantly, early fetal
loss. It is difficult to determine the exact magnitude of the risk as: studies used
135
differing definitions of hyperHCY; in some, the subjects were receiving routine

folic acid prophylaxis; and whilst some studies assessed HCY out-with pregnancy,
others made the diagnosis during the affected pregnancy. The evidence that
exists, however, suggests that the actual degree of risk may be directly correlated
with the plasma HCY level, with higher levels leading to greater risk. In those
studies showing a positive association, the risk of first loss varied from 3- to 7-fold,
with the recurrent loss risk some 3–4-fold increased. Some studies have also
shown a link between hyperHCY and later pregnancy loss, although this has not
been seen in all positive studies.
In subjects with hyperHCY (diagnosed out-with pregnancy), there is also an
increased risk of PET, with the highest risk seen in those with the highest levels
of HCY. In particular, those with levels >9–11µmol/l are likely to have around a
5-fold higher risk of PET when compared with those with lower levels. The great-
est risk is found in primigravid subjects, where the risk may be as high as 20-fold.
In one study, which assessed HCY levels in early pregnancy, a link between the
HCY level and the development of PET in the same pregnancy has also been
reported, although this finding has not been confirmed.
At present there is no convincing evidence that folate supplementation will
reduce the risk of recurrence of PET or the occurrence of fetal loss. However, as
folic acid supplementation is routinely carried out in most developed countries,
the main issue may be what is the appropriate and safe dose (400µg versus 5mg)
and what the duration of such therapy should be, rather than to whether it
should be prescribed or not. If HCY does influence pregnancy outcome, it would
appear likely that alterations in fetal/placental DNA replication, as well as throm-
bosis, are involved.
Elevated FVIIIc, FIXc and fibrinogen levels

In non-pregnant subjects, plasma FVIIIc levels in excess of 150IU/dl are associ-
ated with a 5-fold increased risk of VTE, and elevated levels are found in around
25% of subjects at least 6 months after VTE. There is also an association between
high FVIIIc levels and an increased risk of ischemic heart disease and cerebral
ischemia of arterial origin. Elevated FVIIIc levels are found in around 10% of the
general population and are more common in subjects of blood groups other
than O (i.e. groups A and B, and perhaps particularly those with genotypes AA,
AB or BB). This relation with blood group is consistent with over 40 years of data
suggesting that blood groups have a bearing on the risk of both VTE and arterial
thrombosis. The level of FVIIIc also relates to APTT-assessed aPCR (see above).
However, an independent role for FVIIIc in pregnancy complications may be dif-
ficult to define, as pregnancy is associated with a marked physiological rise in
FVIIIc with increasing gestation.
Akin to the situation with FVIIIc, both elevated FIXc and elevated fibrinogen
levels are also associated with VTE. In the case of fibrinogen there is also a link
with the development of arterial thrombosis.
136
Elevated FVIIIc, FIXc and fibrinogen levels and pregnancy outcome

As noted in Chapter 1, FVIIIc rises progressively with increasing gestation in normal
pregnancies. Given this change, locally derived gestation-specific reference ranges
are required before the impact of FVIIIc on pregnancy outcome can be assessed.
Limited information suggests that FVIIIc levels (perhaps through the acquired
aPCR phenotype) can have an influence on the development of PET. Further work
is, however, required to determine if elevated early pregnancy FVIIIc, FIXc or fib-
rinogen levels carry any independent role in predicting pregnancy outcome.
Antiphospholipid syndrome (APS)

APS is defined as a combination of thrombosis and/or vascular complications of
pregnancy that have occurred in the presence of persisting antibodies to nega-
tively charged membrane phospholipids. It may occur as a primary syndrome or
be a manifestation of systemic lupus erythematosus (SLE). However, it is becom-
ing increasingly apparent that APS is a misnomer, as, in the majority of cases, it is
not the binding of these antibodies to phospholipid but rather their binding to
phospholipid-binding proteins (such as β2-glycoprotein-1, prothrombin, FV or
protein C) that is important in their link with thrombosis.
Although individual assays, such as anti-β2-glycoprotein-1 or anti-prothrombin
antibody tests are available, most centers currently test for two forms of antiphos-
pholipid activity: anticardiolipin antibodies (which can be detected by enzyme-
linked immunosorbent assay (ELISA) and are of IgG, IgM or IgA class) and the
presence of a lupus inhibitor (widely known as a lupus anticoagulant). It should
be noted that the ELISA tests for anticardiolipin antibodies require β2-glycopro-
tein-1 to be present in the test kit for the great majority of antiphospholipid anti-
bodies to be detected.
Although it appears increasingly likely that the actions of the lupus inhibitor
are related to activity against a combination of phospholipid with either β2-glyco-
protein-1 or prothrombin, the lupus inhibitor is currently detected by the effect
it has on clotting tests. To diagnose a lupus inhibitor the patient plasma (which
has been spun to reduce the number of platelets present) should prolong a phos-
pholipid-dependent clotting time. This is usually demonstrated on an APTT (see
Chapter 1), with confirmation on another clotting test called the dilute Russell
viper venom test (dRVVT). An inhibitor will increase both the APTT and the
dRVVT. The diagnosis requires there to be a failure to correct this prolongation
when the test is repeated, using a mixture of patient plasma and normal plasma.
Finally, the assessment is complete if the test is corrected by the addition of
frozen, thawed platelets (the freeze–thaw process exposes phospholipid which is
normally expressed on the inner aspect of the platelet membrane) or the addi-
tion of another source of excess phospholipid. In all cases, the antibody activity
should be shown to be persistent, with a positive repeat assessment seen 6–8
weeks after the initial evaluation.
137
The prevalence of antiphospholipid antibodies (either a lupus inhibitor or

elevated levels of anticardiolipin antibodies) can be detected in 2–12% of women
of childbearing age, but may be of higher prevalence in those with a history of
arterial thrombosis, venous thrombosis or fetal loss. The exact relationship of
anticardiolipin antibodies to vascular disease is clouded by a number of issues:
the large number of assays available; the ongoing debate regarding the relevance
of antibodies of IgM or IgA class; the variations in the cut-off used for reporting
abnormal results and; the difference in the quantity of β2-glycoprotein-1 in each
ELISA assay.
The physiological function of β2-glycoprotein-1 is not clear, although β2-glyco-
protein-1 antibodies appear capable of counteracting the effects of TF pathway
inhibitor and can also affect the effect of aPC on FVa. Thus, by a variety of
actions, the mixture of detectable antiphospholipid antibodies can act to cause
either direct endothelial activation and oxidative injury, or to activate coagula-
tion through proteins such as prothrombin, FV and protein C or β2-glycoprotein-
1. In addition, binding of these antibodies to placental trophoblast may
contribute to alteration in implantation and pregnancy failure.
APS and pregnancy outcome

The presence of persisting antiphospholipid antibodies is associated with placen-
tal infarction, and such antibodies are more often found in those with fetal loss
occurring after 10 weeks gestation (i.e. fetal rather than embryonic or pre-embry-
onic loss), PET, IUGR, VTE and arterial thrombosis. Indeed, it has been esti-
mated that up to 20% of women with antiphospholipid antibodies will have
Table 7.5 – Management options for antiphospholipid syndrome (APS) in pregnancy
Consider past pregnancy outcome

Confirm ACA/ lupus inhibitor status
Manage systemic lupus erythematosis (SLE), (If present)
Use low dose aspirin, when pregnancy confirmed (benefits may be limited to those with multiple fetal losses)
If previous venous thromboembolism (VTE), consider thromboprophylaxis (see Chapter 6)
Uterine artery Doppler 20–24 weeks may have value in risk of PET IUGR
If high PET risk, monitor urine and blood pressure
Corticosteroids, no evidence of benefit
Intravenous immunoglobulin, no evidence for overall benefit
Vitamin C/E supplementation, consider only if high risk PET
Low-molecular-weight heparin (LMWH) + aspirin, antenatal: consider risk:benefit ratio if ≥3 losses,
Puerperium: only LMWH if required for thromboprophylaxis
ACA, Anticardiolipin antibody; IUGR, Intrauterine growth restriction; PET, Pre-eclampsia.
138
obstetrics complications. Of these, the most precise information comes from

those with a lupus inhibitor, which carries between a 3.0- and 4.8-fold risk of
obstetric complications, although it is more likely to be associated with fetal loss
in subjects who have SLE.
There is, however, a great variation in risk, which seems more dependent on
the maternal obstetric history rather than on the presence, or absence, of
detectable antiphospholipid activity. Given this variation in risk, the ability to use
these tests to usefully predict pregnancy-associated disorders, especially in primi-
gravid subjects, or in those in whom the antibody is an incidental finding, is far
from certain. The management of APS in pregnancy requires an individual
assessment, but the potential options are considered (Table 7.5).
References
Poort SR, Rosendaal FR, Reitsma PH, Bertina RM. A common genetic variation
in the 3´–untranslated region of the prothrombin gene is associated with ele-
vated plasma prothrombin levels and an increase in venous thrombosis. Blood
1996; 88(10): 3698–703.
139
CHAPTER 8
Thrombocytopenia in
pregnancy
Maternal thrombocytopenia is commonly found in pregnancy. This is usually due
to physiological changes with increasing gestation (termed gestational thrombo-
cytopenia), or is caused by idiopathic thrombocytopenic purpura (ITP). Throm-
bocytopenia also occurs in association with a wide variety of both
pregnancy-associated and non-pregnancy-associated conditions (Table 8.1).
Gestational thrombocytopenia
Towards the end of pregnancy at least 5% of women have a platelet count below
the usual non-pregnant reference range (150–400 × 109/l). Although this gesta-
tional thrombocytopenia can be difficult to distinguish from ITP, the platelet
count is seldom <80 × 109/l (although counts as low as 50 × 109/l without
significant disease have occasionally been reported). Thus, the diagnosis of ges-
tational thrombocytopenia is most often made in the third trimester. It is con-
Table 8.1 – Thrombocytopenia in pregnancy
Spurious (i.e. clumping or poor sampling)

Gestational
Immune thrombocytopenic purpura (ITP)
Heparin-induced thrombocytopenia (HIT)
Post-transfusion purpura (PTP)
Acute Fatty Liver of Pregnancy
Pre-eclampsia (PET)/ hemolysis, elevated liver enzymes, and low platelets (HELLP) syndrome
Thrombotic thrombocytopenic purpura (TTP)/ Hemolytic uremic syndrome (HUS)
Disseminated intravascular coagulation (DIC)
Drug-induced thrombocytopenia
Systemic lupus erythematosis (SLE)/antiphospholipid syndrome (APAS)
Viral (HIV, Epstein-Barr virus (EBV), cytomegalovirus (CMV))
Congenital thrombocythemias/thrombocytopenia
Hypersplenism
Type IIb von Willebrand’s disease (vWD)
Marrow dysfunction/hematinic deficiency
140
Thrombocytopenia in pregnancy
sidered that gestational thrombocytopenia is not the result of an immune dis-

order, but is due to a combination of hemodilution and a physiological increase
in platelet clearance, with gestational thrombocytopenia representing the
extreme end of the physiological reduction in platelet count which occurs in
most normal pregnancies. Gestational thrombocytopenia does not cause mater-
nal or fetal problems. In particular, fetal thrombocytopenia is no more common
in such pregnancies than it is in women with normal platelet counts (in whom
around 4% will have a baby with mild thrombocytopenia). If the platelet count
is >80 × 109/l, there are no implications for pregnancy management and it will
resolve after delivery.
Diagnosis of gestational thrombocytopenia

A diagnosis of gestational thrombocytopenia requires exclusion of the other
causes of thrombocytopenia (see Table 8.1), but this condition accounts for
almost 75% of cases of thrombocytopenia diagnosed in pregnancy.
All patients presenting with significant thrombocytopenia (<100 × 109/l) should
have a blood film examined for evidence of platelet clumping, which indicates a
factitious thrombocytopenia. In particular, in around 0.1% of adults, EDTA (which
is used as a sample anticoagulant) can lead to an EDTA-dependent platelet aggluti-
nation causing a spuriously low platelet count. The blood film should also be exam-
ined for red cell fragmentation (indicating microangiopathic hemolytic anemia –
MAHA), leukemia, myelodysplasia and megaloblastic anemia.
The platelet count in early pregnancy should be reviewed, as this will be
normal in gestational thrombocytopenia. Such individuals also require investiga-
tion of folic acid status, renal and liver function, a coagulation screen and exclu-
sion of autoimmune disorders such as antiphospholipid antibody syndrome
(APAS), systemic lupus erythematosis (SLE) and HIV infection (Table 8.2).
In addition, pregnancy complications, particularly pre-eclampsia (including
hemolysis, elevated liver enzymes, and low platelets (HELLP) syndrome), should
be considered and, rarely, a coincidental diagnosis of thrombotic thrombocy-
topenic purpura (TTP) and hemolytic uremic syndrome (HUS) may be made
Table 8.2 – Investigation of thrombocytopenia in pregnancy
Blood film to exclude platelet clumps, microangiopathic hemolytic anemia (MAHA) or other hematological
disorders
Coagulation screen (to include fibrinogen and D-dimer levels)
Renal and liver function tests
Antiphospholipid antibodies
Anti-DNA antibodies to exclude systemic lupus erythematosis (SLE) (ANA is sufficient as a screening test)
ANA, Ante-nuclear antigen

141
(Table 8.3). However, the main differential diagnosis is usually ITP or pre-
eclampsia, which account for around 4 and 21%, respectively, of cases of preg-
nancy thrombocytopenia. Clearly, it may be impossible to be certain that the
diagnosis is gestational thrombocytopenia rather than ITP, and in this situation it
is best to err on the side of caution and use the same precautions as in ITP to
minimize the risk of hemorrhagic complications in the baby (see Table 8.4).
Table 8.3 – Differential diagnosis of thrombocytopenia in pregnancy
Incidence Typical gestation Associated features Worst platelet

at diagnosis count (×109/l)
Gestational 5–7 in 100 Second to third Blood pressure (BP) √ <100

trimester Blood film √
Pre-eclampsia (PET) 2 in 100 Late second to BP Variable

hemolysis, elevated liver third trimester Proteinuria (with disease
enzymes, and low MAHA severity)
platelets (HELLP) Uric acid
LFT
±Antithrombin
Idiopathic <5 in 10 000 First trimester BP √ <50

thrombocytopenic Blood film √
purpura (ITP) ±Known ITP/AI history
Acute fatty liver 1 in 10 000– Third trimester Cholestatic LFT Variable

15000 Glucose (with disease
Fibrinogen severity)
Antithrombin
Thrombotic 1 in 25 000 Second to third MAHA <20
thrombocytopenic trimester/ Renal or
purpura (TTP)/hemolytic postpartum Neurological
uremic syndrome (HUS) Coagulation / √
antithrombin √
Disseminated – Second to third Coagulation √ Variable

intravascular trimester/ Fibrinogen + underlying (with disease
coagulation (DIC) postpartum disorder (PET/abruption or severity)
major obstetric hemorrhage/
amniotic fluid embolus/
uterine rupture/sepsis/
retained fetus)
AI, Auto immune; LFT, Liver function test

142
Table 8.4 – Gestational thrombocytopenia
Mild and asymptomatic occurs in 7% of pregnancies

Normal platelet count at booking
Occurs in third trimester
No fetal problem
Resolves after delivery
If platelets >80×109/l: no implications for pregnancy management
If unable to distinguish from idiopathic thrombocytopenic purpura (ITP), avoid fetal scalp electrode, fetal blood
sampling, ventouse delivery and rotational forceps
Immune thrombocytopenic purpura (ITP)

ITP results in a platelet count of <150 × 109/l and is due to auto-antibody-medi-
ated binding and destruction of platelets by the reticuloendothelial system. The
autoantibodies in ITP are generally directed against the platelet surface glycopro-
tein complexes IIb/IIIa or Ib/IX. Immune-mediated thrombocytopenia can also
be caused by allogenic antiplatelet antibodies which lead to post-transfusion
purpura (PTP), neonatal alloimmune thrombocytopenia (NAIT) and heparin-
induced thrombocytopenia (HIT) (Table 8.5).
Autoantiplatelet antibodies can occur in association with drugs, SLE, rheuma-
toid arthritis, HIV and lymphoma. Recently, eradication of Helicobacter pylori has
been shown to be of benefit in some subjects with ITP, although the causal
mechanism, or the role of such therapy in pregnancy, is not known (Table 8.6).
Nonetheless, a breath test or serological assessment to detect H pylori may be
worthwhile, especially in refractory cases.
As the antiplatelet antibodies cross the placenta the baby may become throm-
bocytopenic, with around 9–15% having thrombocyopenia and up to 1.5% devel-
oping an intracranial hemorrhage.
Table 8.5 – Disorders associated with immune-mediated thrombocytopenia
Helicobacter pylori
Systemic lupus erythematosis (SLE)
Lymphoma chronic lymphatic leukemia (CLL)
HIV
Drugs
143
Table 8.6 – Experimental therapies in idiopathic thrombocytopenic purpura (ITP)
Intravenous Anti-RhD treatment
• Anti-RhD is usually used to prevent RhD alloimmunization

• In RhD-positive subjects, 50–70mg/kg anti-RhD increases platelets in <70%
• Side effects include headache, fever and a risk of hemolysis
• Anti-RhD results in less donor exposure than IVIgG
• Repeat dosing may be required
• The effect on the fetus requires more study
• Anti-RhD may have a role as a secondline therapy
Helicobacter eradication
• Gastric Helicobacter may be more common in ITP

• In refractory ITP consider screening for Helicobacter
• The benefits of eradication of Helicobacter in pregnancy is unknown
Diagnosis of ITP in pregnancy

Adult chronic ITP often affects women of childbearing age and is found in 1–5 in
10 000 pregnancies. In fact, it is not infrequently diagnosed for the first time
during pregnancy. This is particularly so since the inclusion of routine platelet
counting in the mechanical assessment of the blood count. The most likely pre-
sentation is as asymptomatic thrombocytopenia on routine blood count testing
(Table 8.7). In addition, women known to have ITP may suffer an exacerbation
during pregnancy, and occasionally a new diagnosis is heralded by severe sympto-
matic thrombocytopenia in pregnancy.
As in non-pregnant subjects, the diagnosis of ITP in pregnancy is one of exclu-
sion of other disorders (see Tables 8.1 and 8.2). In general, if there are no addi-
tional clinical features, and the blood count and film are otherwise normal, bone
Table 8.7 – Features of thrombocytopenic purpura (ITP) in pregnancy
A low platelet count out-with pregnancy is invaluable in the diagnosis of ITP

ITP may be impossible to distinguish from gestational thrombocytopenia
Platelets <100 × 109/l: screen for pre-eclampsia (PET), coagulopathy, autoimmune disease, HIV
No bone marrow examination unless a lymphoma is suspected, or a splenectomy is required
The bleeding time does not predict hemorrhage
Corticosteroids, intravenous human normal immunoglobulin (IVIgG) or splenectomy do not effect the fetal
platelet count
The mode of delivery should be determined by obstetric indications
Neonatal alloimmune thrombocytopenia (NAIT) should be excluded in all neonates with thrombocytopenia and
hemorrhage
144
marrow examination is not required to make the diagnosis of ITP in pregnancy. In

addition, the measurement of maternal antiplatelet immunoglobulin or blood
thrombopoietin levels are not valuable in the routine diagnosis of ITP in pregnancy.
Treatment of ITP
There have been few randomized clinical trials on the management of ITP in
pregnancy. As it is very likely that the platelet count will continue to fall as preg-
nancy progresses (reaching a nadir in the third trimester), and as spontaneous
bleeding problems are unlikely if the platelet count is >20 × 109/l, the manage-
ment of ITP in pregnancy is usually focused on achieving an adequate platelet
count for delivery, in particular to allow neuraxial anesthesia, and to minimize
the risk of fetal bleeding during labor and delivery.
The treatment options during pregnancy are similar to those used out-with
pregnancy, with due regard to the presence of the fetus. In general, if the
platelet count is >20 × 109/l and the patient is asymptomatic, then monitoring of
the platelet count and of the patient for bleeding is all that is required. If, at
delivery, the patient is asymptomatic then a spontaneous vaginal delivery or
Cesarean section could take place if the platelet count is >50 × 109/l (uncompli-
cated vaginal delivery is probably safe at levels >30–40 × 109/l, but as the risk of
Cesarean section is present in every labor it is best to aim for at least a level of
50 × 109/l). If the patient requires epidural anesthesia then a platelet count of
>80 × 109/l is recommended (Table 8.8). Following delivery, non-steroidal anti-
inflammatory drugs (NSAID) should be avoided for postoperative analgesia
unless the platelet count is >100 × 109 /l. It is also important to consider the
woman’s risk of venous thromboembolism (VTE), and for those considered at
increased risk, graduated elastic compression stockings and/or low-molecular-
weight heparin (LMWH) can be used. The latter is considered satisfactory if the
platelet count is >50 × 109/l, the patient has no hemorrhagic problems and has a
significant risk of VTE.
Firstline therapies
There are two firstline options currently used. These are oral corticosteroids and
intravenous human normal immunoglobulin (IVIgG). Both appear to have
Table 8.8 –Treatment of idiopathic thrombocytopenic purpura (ITP)
Platelet count >20 × 109/l and asymptomatic Monitor until delivery

Platelet count >40–50×109/l and asymptomatic Safe for normal vaginal delivery
Platelet count >50 × 109/l and asymptomatic Safe for Cesarean section
Platelet count >80 × 109/l and asymptomatic Safe for epidural anesthesia
145
similar response rates – around 70% of patients will respond to IVIgG, with a
variable response rate for steroids, which is, at best, between 50 and 66%.
Clinicians vary in their choice, with some preferring IVIgG. This is most often
due to the perceived side effects of oral corticosteroids, although such side
effects are not usually critical with a short duration of high-dose steroids used in
anticipation of delivery. The cost of IVIgG and, more importantly, the fact that it
is a human blood product must also be considered. (Table 8.9). In general, if
treatment is to be required for a short time, then corticosteroids are a useful
option at a dose of prednisolone of 1mg/kg/day in the first instance, continued
for 2–4 weeks then tailing off the dose. If the duration of therapy is likely to be
longer (treatment starting in the first or second trimester), then IVIgG, at a dose
of 0.4g/kg/day for 5 days or 1g/kg/day for 2 days, has been used with an equiva-
lent response rate evident in at least one randomized trial. However, depending
on the product (see Chapter 2), there may be a higher risk of renal impairment
with the more intense schedule, and the first schedule is to be preferred if there
are pre-existing fluid balance problems or renal impairment. IVIgG most often
results in a response of 2–3 weeks duration and repeat dosing may be required if
a continuing hemostatic challenge is anticipated.
The mechanism of action of IVIg in ITP is largely unknown, but it may involve
the blockade of Fc receptors on macrophages and other effectors of antibody-
dependent cytotoxicity, or be due to the presence of anti-idiotype antibodies in
IVIgG which block autoantibody binding to circulating platelets. IVIgG may also
block Fc receptors on the placenta and so may prevent transfer of maternal
antiplatelet antibody but this is unproven in practice and remains a theoretical
benefit of IVIgG.
Secondline therapies
This includes high-dose methylprednisolone (1000mg) or azathioprine, which is
satisfactory for use in pregnancy, or a combination of these therapies with IVIgG.
Obviously, danazol, vinca alkaloids and cyclophosphamide are not suitable for
the pregnant patient. For patients who do not respond to conventional second-
line therapies, anti-RhD immunoglobulin (see Table 8.6), and even interferon
and mycophenolate mofetil or retuximab could be considered, but experience,
even in the non-pregnant, is very limited.
Splenectomy is best avoided in pregnancy, but, if considered essential, is best
carried out between 13 and 20 weeks gestation. At present, the role of lapro-
scopic splenectomy in pregnancy is not known. In 66% of non-pregnant sub-
jects, splenectomy will result in a normal platelet count. A number of studies
have suggested that the patient’s response to IVIgG may predict the response to
splenectomy. However, it is not known if this has any predictive value in preg-
nancy. The use of indium-labeled autologous platelet scanning to identify
whether the platelet destruction is splenic, hepatic or diffuse would not be a
favored option in pregnant subjects. This test is, however, the most sensitive pre-
146
Table 8.9 – Treatment and outcome of thrombocytopenia in pregnancy
Mechanism Treatment Treatment complications Maternal Fetal Fetal Recur

outcome testing outcomes
Gestational Hemodilution Not required – √ None Platelet (Plat) Yes
Clearance <150, 4%
Idiopathic Autoantibodies Corticosteroids Weight , blood pressure √ or Full blood Well 90% Yes
thrombocytopenic (BP) , glucose Hemorrhage count (FBC) Plat<50, <20%
purpura (ITP) Immunoglobulin Allergy. fluid , urea , Day 1–5 Intra cerebral
Azothioprine Infection transmission hemorrhage
Splenectomy ? Miscarriage/thrombosis/ (ICH) <1.5%
infection Mortality 0.6%
Pre-eclampsia (PET) Platelet/endothelial Supportive Allergy, fever, FBC None Yes
hemolysis, elevated activation (±platelets/fresh transmissible agents Day 1–5 Platelets
liver enzymes, and frozen plasma Intrauterine
low platelets (HELLP) (FFP)) growth restriction
147
(IUGR)
Deliver Abortion
Thrombotic thrombo- ADAMTS–13 Plasma exchange Fluid imbalance, BP Mortality – Yes
cytopenic purpura ±FFP Allergy, fever, 10–30%
(TTP)/hemolytic transmissible agents
uremic syndrome (HUS)
Acute fatty liver ? Long chain fatty Glucose Mortality Mortality <15% ?Yes
acid deficiency Fluid balance 5–20%
(LCHAD) General support
Liver transplant
±platelets/FFP/ Allergy, fever,

Cryoprecipitate (cryo) transmissible agents
Disseminated Consumption Treat cause Depends on – Variable No
intravascular Allergy, fever, etiology
coagulation (DIC) ±platelets/FFP/cryo transmissible agents
Table 8.10 – Preventing post-splenectomy infection
At least 2 weeks prior to splenectomy
• Polyvalent Pneumococcal vaccine (Pneumovax II)

• Hemophilus influenza B vaccine (Hib)
• Meningococcal C vaccine (Men C)
Following splenectomy
• Phenoxymethylpenicillin (250–500mg bd) or erythromycin (500mg od)*

• Annual influenza vaccine
• Antimalarial prophylaxis when appropriate
• Anticoagulant prophylaxis when appropriate
*The required duration is not yet known but may be lifelong.
dictor of response to splenectomy, by identifying those patients where splenic

destruction is present.
As with non-pregnant subjects, the removal of the spleen leads to a lifelong
increased risk of infection, predominantly with capsulated organisms (Table
8.10). Splenectomy also results in a lifelong thrombotic tendency. Although the
mechanism of this is not understood, it may relate to circulating nucleated red
cells (which are seen in the blood after splenectomy), or be due to the loss of
splenic regulation of white cell and platelet counts. In any event, this must be
considered when assessing the patient risk factors for thrombosis and thrombo-
prophylaxis.
Management of delivery in ITP

The baby’s platelet count cannot be reliably predicted from either the mother’s
platelet count, maternal antiplatelet immunoglobulin levels, or by the presence
(or absence) of the spleen. Fetal blood sampling by cordocentesis carries a mor-
tality similar to that of the risk of intracerebral hemorrhage (see Table 8.9), and
scalp blood sampling is prone to spurious low results (due to clotting of the
blood sample) and may itself cause significant hemorrhage. Thus, as the status of
the baby cannot be determined antenatally, procedures at delivery that would
pose an additional bleeding risk (particularly of intracranial hemorrhage) should
be avoided. There is no evidence that Cesarean section is safer for the thrombo-
cytopenic fetus than an uncomplicated vaginal delivery and, in any case, the
maximum fall in the platelet count in the newborn is likely to occur some 24–48
hours after delivery. Fetal scalp electrodes and fetal blood samples are best
avoided, and if an assisted vaginal delivery is required, ventouse should be
avoided as this can cause severe scalp hematoma. Rotational forceps are also best
148
avoided due to risk of intracranial bleeding if the baby is affected, but a straight-
forward mid- or low-cavity forceps delivery would be considered appropriate, if
required. At delivery, a cord platelet count should be determined in all babies
whose mother has been diagnosed with ITP. Close monitoring is required over
the next 2–5 days. In children with evidence of bleeding, or a platelet count
<20 × 109/l, IVIgG (at 1g/kg) produces a rapid response. Neonatal life-threaten-
ing hemorrhage may also require platelet transfusion.
Autoimmune neutropenia (AIN)

In parallel with ITP, AIN also occurs, and a number of cases of AIN in pregnancy
have also been reported. There is, however, very little information on pregnancy-
specific management. However, the genetically engineered granulocyte colony-
stimulating factor (G-CSF) has been used, with some success, to increase the
production and release of neutrophils in subjects who have an autoantibody to
their own neutrophils. In subjects on G-CSF who become pregnant, the G-CSF is
often witheld antepartum, as the effects on the fetus are not known. The passive
placental transfer of antibodies can result in neutropenia in the neonate, and
very limited evidence suggests that G-CSF, given in the final trimester, may have a
beneficial effect on the neonatal neutrophil count (either by increasing maternal
neutrophils and reducing free maternal antibody, or by stimulation of neutrophil
release in the fetus). However, whether this is a safe approach, and whether
maternal autoantibody titers, or the specificity of the antibody in the mother will
assist in the prediction of the neonatal blood count, requires further invesiga-
tion. G-CSF is, however, probably safe in nursing mothers. Although it does cross
the breast, limted evidence suggests that it causes a local, but not systemic effect,
in the neonate.
Thrombotic thrombocytopenic purpura (TTP)/hemolytic uremic

syndrome (HUS)
TTP and HUS are characterized by a low platelet count, red cell fragmentation
(resulting in MAHA: see above) and multi-organ failure. It is usually considered
that TTP is more often associated with neurological abnormalities and non-renal
organ ischemia, whilst patients with HUS have predominantly renal manifesta-
tions. However, as there is a significant cross-over in symptomatology, it is often
difficult to distinguish between these two conditions clinically.
TTP occurs most often as an idopathic syndrome and, interestingly, the inci-
dence of TTP is increasing in the Western world; this may be due to cases
precipitated by drug exposure. Although it can occur as a congenital form, which
may recur, around 66% of adult cases are non-recurrent. TTP also occurs sec-
ondary to pregnancy, drugs such as clopidogrel, ticlopidine, tacrolimus, the com-
bined contraceptive pill, bone marrow transplant, SLE, malignancy and infection
with HIV or Escherichia coli-0157. As with TTP, HUS may be precipitated by
149
pregnancy, follow infection with cytotoxin-producing serotypes of organisms

such as E Coli or Shigella, or follow administration of drugs such as cyclosporine,
quinine and chemotherapeutic agents.
The endothelium releases ultralarge multimers of von Willebrand Factor
(vWF), which are subsequently cleaved by the metalloprotease ADAMTS-13 (A
disintegrin and metalloproteinase with thrombospondin domains, number 13)
in the plasma. This results in the circulation of the correct balance of vWF mul-
timers and TTP/HUS is characterized by a failure of cleavage of these multi-
mers. As noted above, in TTP, this can be due to a congenital deficiency of the
cleavage enzyme, but is more commonly due to acquired autoantibody forma-
tion. However, in HUS, and in many cases of TTP, the vWF cleaving activity is
normal. In the rare patients with recurrent or familial HUS, there is an abnor-
mally low level of the complement control protein Factor H, with overactivity of
complement component 3 whenever the alternative complement pathway is
activated.
Just as insufficient vWF multimers lead to bleeding through a reduction in
platelet aggregation (see Chapter 4), an excess of circulating ultralarge multi-
mers results in an increase in platelet aggregation and platelet consumption,
which leads to microvascular platelet thrombi (with little fibrin, but an abun-
dance of vWF in the platelet thrombi formed) as well as endothelial apoptosis
and MAHA. This results in downstream ischemia and infarction in organ systems.
There are many reasons why pregnancy may pre-dispose to TTP/HUS. As
detailed in Chapter 1, normal pregnancy is associated with an increase in Factor
(F) VIIc, FVIIIc, FIXc, vWF and thrombin generation, as well as a reduction in
protein S. These physiological changes may result in hypercoagulability and exac-
erbate any predisposition to TTP or HUS. Indeed, thrombophilic risk factors
such as FV Leiden and obesity have also been linked with the occurrence of TTP.
More particularly, pregnancy is often the precipitant for recurrent episodes of
TTP that occur in subjects with the rare congenital deficiency of ADAMTS-13.
There is also evidence that there is a 30% reduction in ADAMTS-13 in normal
pregnancy and that the pregnancy level of ADAMTS-13 is inversely related to
vWF levels. If there is a congenital deficiency, a further reduction may be seen in
normal pregnancy. As noted above, TTP can, however, occur with no evident in
reduction in ADAMTS-13 activity, and reduction in ADAMTS-13 itself is not spe-
cific to TTP/HUS.
Diagnosis of TTP/HUS in pregnancy

With regard to HUS in pregnancy, the presentation is typically a single episode
(almost always postpartum) with the classic triad of thrombocytopenia, hemoly-
sis, and renal failure. It can accompany HELLP syndrome, where the liver dys-
function is more marked (see Table 8.3). While TTP should present with the
classic pentad of fever, hemolysis, thrombocytopenia, central nervous system
(CNS) signs and renal dysfunction (proteinuria/hematuria), all five are only
150
Table 8.11 – Investigation of suspected TTP/HUS during pregnancy.
FBC/Film
Reticulocyte count
Coagulation screen
Urea and electrolytes
LFT
LDH
Coomb’s test
Urinalysis
FBC, Full blood count; LDH, Lactate dehydrogenase; LFT, Liver function tests
present in around 50% of cases. The CNS signs, thrombocytopenia and hemoly-
sis are, however, more marked than with PET or HUS, but the hypertension may
not be severe (Table 8.12). TTP, particularly recurrent TTP, usually presents
before 24 weeks of pregnancy and intrauterine growth restriction (IUGR) and
fetal death is usual. As noted above, isolated episodes of TTP can occur and
relapse is common in pregnancy.
Although routine clotting tests are often normal in the early stages of
TTP/HUS, as the disease progresses there may be activation of coagulation and
disseminated intravascular coagulation (DIC). However, the accompanying eleva-
tion of D-dimer levels may often be difficult to distinguish from the elevated D-
dimer levels of normal pregnancy (see Chapter 1).
A comparison of the symptoms and investigations in thrombocytopenic dis-
orders is shown (Table 8.3). It has been reported that plasma antithrombin levels
may be of assistance in the distinction of PET/HELLP (reduced levels) from
TTP/HUS (normal levels) (Table 8.3). However, reduced and normal levels may
be seen in either condition, and establishing a reduced level requires a local
pregnancy-related reference range.
Table 8.12 – TTP/HUS or PET/HELLP.
Significant anemia and microangiopathic hemolytic anemia (MAHA) points to TTP/HUS

Severe hypertension points more to PET/HELLP than TTP/HUS
Severe renal or central nervous system (CNS) disease points more to TTP/HUS than PET/HELLP
Fever is unusual in HELLP
Often delivery leads to resolution of PET/HELLP, but a transient deterioration may occur
TTP is not improved by delivery and often PET/HELLP and TTP/HUS are only distinguished after delivery
151
Treatment of TTP/HUS
With perhaps the exception of bacterial endotoxin-related HUS (where support-
ive care is the main aspect of therapy) and recurrent or congenital TTP (see
below), it is unlikely that a clear distinction between the two syndromes of TTP
and HUS will be possible in the majority of pregnancy-related cases. As a con-
sequence, both will be considered predominantly as a single syndrome when con-
sidering therapy, particularly as there may be benefit in plasma exchange in
non-toxin-related HUS (Table 8.13).
Firstline therapies
The mainstay of treatment of TTP/HUS is plasma exchange, which removes the
large vWF multimers and autoantibodies, and supplements the vWF cleaving
enzymes. This has produced a remarkable reduction in the mortality of this con-
dition. The treatment schedules used in pregnancy are derived from those used
in non-pregnant subjects. Plasma exchange should be instituted within 24 hours
of presentation, as this has a bearing on prognosis, and subjects should be
managed in a facility that can provide an uninterrupted service. The optimal
number of plasma exchange procedures required is not known, but most author-
ities recommend that there should be a further two procedures after complete
clinical remission. Indeed, many facilities would taper off the exchanges rather
than stopping them abruptly, even after these two further procedures. Similarly,
the optimal exchange regime has not been determined, and either daily
exchanges of 1 or 1.5 plasma volumes for the first 3 days followed by 1 plasma
volume thereafter have both been used to good effect.
In addition, the optimum fluid replacement has not yet been ascertained.
Fresh frozen plasma (FFP) is the standard therapy in most countries. However,
cryosupernatant may offer some advantages over FFP. Cryosupernatant is derived
from the manufacture of cryoprecipitate and contains lesser amounts of the
larger vWF multimers, but may not be widely available. In view of the potential
risk of viral contamination of FFP, solvent detergent-treated FFP (SD-FFP) can
Table 8.13 – Thrombotic thrombocytopenic purpura (TTP)/hemolytic uremic syndrome (HUS)
Daily plasma exchange as soon as possible

Fresh frozen plasma infusion if delay in plasma exchange
Consider corticosteroids and aspirin
Consider refractory if deterioration or no response in 7 days
If refractory, change replacement fluid or increase exchange frequency
Anaphylaxis to plasma exchange may occur in <1 in 400
If febrile, check for underlying infection
152
also be used as the fluid replacement in plasma exchange. SD-FFP contains lesser
amounts of the larger vWF multimers, and may also carry a lesser risk of adverse
reactions, than standard FFP.
When plasma exchange is not immediately available, FFP infusion (at
30ml/kg), with due attention to the risks of fluid overload, may be beneficial
whilst exchange is being organized. In the rare congenital TTP with deficiency of
the cleaving metalloproteinases, exchange may not be required and FFP infusion
alone may be sufficient to induce remission.
Additional therapies
Any agent that could have precipitated TTP, such as clopidogrel, should be
stopped. As an immunoglobulin-mediated reduction in ADAMTS-13 activity
occurs in association with TTP, intravenous methylprednisolone has been used as
an adjunct to plasma exchange, although the benefit:risk ratio of this approach
in pregnancy has not been formally assessed. However, there is no specific con-
traindication to corticosteroid use in this situation.
Similarly, as many of the effects of TTP may be mediated through platelet activa-
tion, the use of antiplatelet agents is an attractive option. In this regard, there may
be some benefit in the empirical addition of aspirin to the plasma exchange regime
when the platelet count is restored to >50 × 109/l. Use of aspirin when the patient is
severely thrombocytopenic can, however, precipitate hemorrhagic complications.
All patients should receive folic acid and be vaccinated (when the platelet
count allows) against hepatitis B. Red cell transfusion should be given according
to clinical need, but platelet transfusions should be avoided in TTP. As with
other conditions associated with thrombosis, there may be a role for thrombo-
prophylaxis with LMWH, although the use of this requires an individual assess-
ment of the relative bleeding/thrombosis risk.
Secondline therapies
The patient is regarded as being in remission if there is a resolution of neurologi-
cal symptoms, the platelet count has normalized, the lactate dehydrogenase
(LDH) has reduced to within reference limits and the hemoglobin is rising.
Refractory disease would be defined as a failure of this after seven exchanges. If
the patient deteriorates, or does not respond to the initial plasma exchange
regime, a change to a higher plasma volume exchange (1.5 plasma volumes), a
higher frequency (12 hourly) or a different replacement fluid (cryosupernatant
or SD-FFP rather than standard FFP) has been recommended. The use of other
immunosupressive therapies such as vincristine, cyclophosphamide and
cyclosporine would be problematic whilst there is still a viable fetus and their use
in pregnancy would be on a case-by-case basis. Generally, if the patient has
reached this stage, she will usually have been delivered in either the maternal or
fetal interest, or have suffered an intrauterine death.
153
Pre-eclampsia (PET)/hemolysis, elevated liver enzymes, and low

platelets (HELLP) syndrome
PET is a hypertensive disorder that occurs in 3–6% of pregnancies and is most
commonly seen in primigravid women, or more particularly in ‘first-partner’ preg-
nancies (see Chapter 7). The disease has varying severity from mild hypertension
to severe hypertension with seizures, and can occur in the context of HELLP syn-
drome (Table 8.14). As detailed in Chapter 7, the cause of this disorder remains
enigmatic, although it is undoubtedly related to abnormal placental trophoblast
development early in pregnancy. This may relate to immune maladaption, placen-
tal ischemia, lipid-related oxidative stress or excess thrombogenesis.
Thrombocytopenia occurs in around half of subjects with PET and is most
likely due to a combination of excess thrombin generation and endothelial acti-
vation. These prothrombotic changes result in an increase in activation, binding
and clearance of platelets. Certainly, in the majority of subjects, the degree of
thrombocytopenia is linked to the clinical severity of the PET.
Diagnosis of PET and HELLP

See Chapter 7.
Table 8.14 – Hemolysis, elevated liver enzymes, and low platelets (HELLP) summary
Hemolysis: schistocytes and LDH >600U/l

Elevated liver enzymes (AST >70U/l)
Low platelets (<100 × 109/l) with subclassification based upon count
Coagulation tests (activated partial thromboplastin time (APTT) and prothrombin time (PT)) remain normal,
but sensitive tests of coagulation consistent with low-grade disseminated intravascular coagulation (DIC))
Liver shows periportal hemorrhage, necrosis, fibrin deposition and steatosis (overlap with acute fatty liver of
pregnancy (AFLP))
Thirty-three per cent occur postpartum
In severe pre-eclampsia (PET), 6% have one of the HELLP criteria, 12% have two, and 10% have all criteria
HELLP presentation is non-specific with malaise, but RUQ pain and edema are common clinical features:
50% are normotensive and >6% do not have proteinuria at diagnosis
Management: delivery and supportive therapy with steroids
HELLP outcome in next pregnancy:
HELLP in 3%
PET in 19%
Intrauterine growth restriction (IUGR) in 12%
IU/l, International units/liter; LDH, Lactate dehydrogenase; RUQ, right upper quadrant; U/l, Units/liter
154
Treatment of PET/HELLP syndrome

The treatment of PET/HELLP is geared towards stabilization of the mother. This
requires the control of blood pressure, which allows the pregnancy to proceed
and reduces the need for medical intervention, although this does not appear to
alter the course of the disease. Indeed, delivery is the only cure for PET. To
control hypertension, the drug of choice is methyldopa, as there is a consider-
able experience of the use of this drug in pregnancy. A suitable alternative is
labetolol. More acute blood pressure reduction may require labetolol, nifedipine
or hydralazine. Fluid balance should be stabilized, but diuretics should be
avoided, unless required for cardiac failure. There is also the need to consider
seizure prophylaxis in fulminant disease and to achieve sufficient fetal matura-
tion for delivery. When blood pressure is uncontrolled, hematological and bio-
chemical parameters are deteriorating, or the patient is becoming symptomatic,
delivery is indicated. This is particularly so when the pregnancy is beyond 34
weeks, but when it is between 24 and 34 weeks, the risks to the fetus of early deliv-
ery and the risk to the mother of the continuing pregnancy have to be balanced
in determining the appropriate action. Following delivery, the disease process
will resolve, but the systemic manifestations may worsen in the immediate post-
partum period.
Symptomatic thrombocytopenia and the preparation for delivery may require
platelet transfusion. In such patients, platelets may fail to achieve the expected
increment due to platelet consumption, and the platelet count should be
checked after dosing to confirm an adequate response. The platelet count may
continue to fall after delivery, although in the majority of subjects delivery brings
about a resolution of the disorder. For those with persisting thrombocytopenia
and/or MAHA (indicating progressive disease after delivery), both plasma
exchange and corticosteroids have been used to assist recovery. Any associated
coagulopathy will require FFP. Hypofibrinogenemia is not common and, indeed,
may be difficult to define given the substantial rise in fibrinogen seen with
increasing gestation. If present, however, it may require specific treatment with
cryoprecipitate or fibrinogen concentrate.
Acute fatty liver of pregnancy (AFLP)

AFLP is an unusual condition (complicating 1 in 10 000–15 000 pregnancies) that
results from fatty infiltration of the liver and other organs during pregnancy. It is
more common in obese subjects, in those with a multiple pregnancy and in those
carrying a male fetus. It often presents with jaundice, which can rapidly progress
to coagulopathy, hepatic failure and both fetal and maternal death. In 50% of
subjects there may be associated features suggestive of PET, and in 20% of sub-
jects features suggestive of HELLP syndrome. The condition has been reported
to carry a maternal mortality risk of 5–20% and a fetal mortality risk of up to
15%.
155
Diagnosis of AFLP
The diagnosis of AFLP is suspected by a combination of clinical features (see
Table 8.3). In particular, the detection in the third trimester, of a cholestatic
pattern of liver function tests (LFTs), hypoglycemia and features of failure of syn-
thetic hepatic function (e.g. hypofibrinogenemia and a reduction in plasma
antithrombin levels) are suggestive of the disorder. The disease may be difficult
to distinguish from PET/HELLP and will require the exclusion of other causes of
hepatitis and confirmation of fatty infiltration in the liver by ultrasonography or
magnetic resonance imaging (MRI). In general, when compared to HELLP syn-
drome, AFLP is more likely to be associated with higher serum levels of aspartate
transaminase (AST) and hypoglycemia, and is more often accompanied by
marked hyperuricemia, DIC and leukocytosis. The combination of these features
will lead to a diagnosis in the majority of cases without the need for liver biopsy.
A potential causative mechanism is suggested by the link between AFLP and the
carriage of a fetus with an inborn error of metabolism (resulting in a failure to
oxidize long-chain fatty acids). Oxidation of fatty acids requires adequate function of
mictochondrial trifunctional protein (MTP). MTP is coded by chromosome 2 and
consists of three enzymes that carry out consecutive oxidation of free fatty acids. Free
fatty acid metabolism is used to generate glucose and eventually adenosine triphos-
phate (ATP). In the absence of free fatty acid metabolism and inadequate glucose, a
hypoglycemic lactic acidosis occurs, which is accompanied by infiltration (by abnor-
mal lipids) of the liver, skeletal muscle, myocardium, kidney, pancreas and lungs.
A number of cases of acute fatty liver of pregnancy have been associated with
the presence of fetal homozygous deficiency of one, or all three, of the com-
ponents of the MTP. However, how alterations in placental free fatty acid
metabolites lead to maternal disease is unknown, and further work is required to
determine the exact contribution of inborn errors of fatty acid oxidation to the
development of AFLP as well as PET/HELLP syndrome.
Treatment of AFLP
Treatment is aimed towards biochemical correction of hypoglycemia and fluid
balance, correction of any associated renal insufficiency, coagulopathy, hyperten-
sion, concomitant sepsis and preparation for delivery. Severe hepatic failure may
require the use of liver transplantation. It is known for AFLP to recur in sub-
sequent pregnancies, but the exact incidence is hard to estimate. This may be
due to the substantial mortality associated with the condition and the natural
reluctance of some survivors to undertake a further pregnancy.

HIT arises from the development of specific heparin-dependent IgG antibodies
and should be considered in the differential diagnosis of thrombocytopenia for
156
those receiving heparin in pregnancy. As detailed in Chapter 5, HIT commonly

occurs after 5 days of therapy and is associated with a high risk of severe throm-
bosis due to platelet activation.
Diagnosis of HIT
A variety of tests to determine the presence of antibodies against the combined
heparin-platelet Factor 4 antigen are employed. These are often based on
enzyme-linked iummunosorbent assay (ELISA) technology or platelet function
testing. The principal difficulty with these assays is the lack of standardization,
and often the lack of information about their sensitivity and specificity for HIT.
Because of this the diagnosis of HIT should remain principally a clinical one.
Treatment of HIT in pregnancy

For treatment options see Chapter 5.

PTP is a rare disorder (<1 in 5 million of Caucasian populations/year), although
around 2% of female Caucasians may be at risk. It is most often seen in multi-
parous females, and although acute PTP is most likely after transfusion for post-
partum hemorrhage, it should be considered in the differential diagnosis of
thrombocytopenia if there has been an antenatal transfusion (see Chapter 2).
PTP has similarities to NAIT, although there are no reports of maternal
thrombocytopenia in NAIT. Furthermore, the fetal thrombocytopenia in NAIT
occurs because the fetus is positive for the corresponding antigen target of the
IgG antibody transferred from the mother, whereas in PTP the platelet destruc-
tion occurs despite the mother being negative for the target platelet antigen. The
development of the antibody in PTP is not invariable and is more common in
subjects carrying human leukocyte antigen (HLA) DR3 (specifically DRw52).
Although after acute PTP recurrence is possible, this is difficult to predict as
there may be a period when further exposure to the human platelet antigen
(HPA) will not elicit an anamnestic response.
Prevention of PTP
Effective prevention in pregnancy would require universal screening to deter-
mine the HPA phenotype of pregnant women, as well as the presence of existing
maternal anti-HPA antibodies. It would also be necessary to supply appropriate
antigen-negative blood throughout pregnancy. At present, it would seem unlikely
that universal screening would be either cost or clinically effective. However, as
noted in Chapter 2, subjects already known to have such antibodies should be
transfused with appropriate HPA antigen-negative blood.
157
Treatment of PTP
For treatment options see Chapter 2.
Disseminated intravascular coagulation (DIC)

DIC has been defined as an ‘acquired’ syndrome, characterized by intravascular
activation of coagulation with loss of localization, arising from different causes. It
can originate from and cause damage to the microvasculature, which, if suffi-
ciently severe, can produce organ dysfunction. The activation of coagulation
leads to thrombin generation, consumption of clotting factors and anticoagu-
lants, and the development of MAHA. This can lead to widespread thrombosis,
bleeding and multi-organ failure.
DIC occurs in response to a wide variety of disorders (Table 8.15). In sepsis,
the trigger may be the activation of cytokines, which leads to coagulation activa-
tion. DIC may also occur in response to the release of pro-coagulant material,
such as fat, phospholipids, tissue factor and amniotic fluid. Such release is found
in association with mismatched blood transfusions, obstetric complications or
malignancy. DIC should be distinguished from other causes of MAHA, such as
HUS and TTP.
Diagnosis of DIC in pregnancy

DIC is diagnosed when there is consumption of coagulation factors, as seen by:
A prolonged prothrombin time (PT);
A prolonged activated partial thromboplastin time (APTT);
A reduction in fibrinogen;
Evidence of an increase in fibrinolysis as shown by an increase in measurable
fibrin degradation products, i.e. FDP, or more specifically D-dimers (see
Chapter 1).
Table 8.15 – Causes of disseminated intravascular coagulation (DIC)
Sepsis
Major trauma
Placental abruption
Amniotic fluid emboli
Mismatched transfusion
Hepatic failure
Malignancy
Pancreatitis
158
In addition, there may also be a reduction in plasma antithrombin (due to

consumption and destruction) and thrombomodulin expression (due to the
action of TNFα and interleukin 1β) (IL-1β). This leads to a reduction in protein C
activation (see Chapter 1). DIC may also be accompanied by an increase in plas-
minogen activator inhibitor-1 (PAI-1), leading to a reduction in fibrinolysis.
These features will often by accompanied by thrombocytopenia (due to con-
sumption of platelets in clot) and features of the causal disorder (see Table
8.15). Many of these altered biochemical parameters are sensitive to the presence
of DIC, but are not very specific. In particular, fibrinogen is helpful when
reduced, but may be maintained until late in the disease process. In such cases,
serial measurements may be most helpful. At present, other tests – such as
plasma levels of antithrombin, PAI-1, prothrombin fragment 1+2 (F1+2) and
soluble fibrin levels – remain of research value only.
Out-with pregnancy, when there is evidence of a predisposing condition, a
scoring system (using a combination of platelet count, fibrinogen, the prothrom-
bin time, D-dimer or FDP) may be useful in the diagnosis of DIC. However, DIC
may be more difficult to define in pregnancy, as D-dimer levels, FDP and fibrino-
gen all increase, and the APTT shortens even in normal pregnancy (see Chapter
1). Severe DIC may be obvious, but lesser forms will require reference to gesta-
tion-specific clotting factor ranges for diagnosis.
Treatment of DIC
The treatment, as in non-pregnant subjects, is directed to the cause of the
DIC. Supportive therapy with coagulation factor products and platelets may be
required if the patient is bleeding, or bleeding is likely. In amniotic fluid
embolism, and in disorders such as acute promyelocytic leukemia, the pre-
dominant manifestation may be ischemia and thrombosis. In these circum-
stances, anticoagulation may be required. Occasionally, there is both bleeding
and thrombosis, and a careful balancing act of therapeutic clotting factors
(avoiding concentrated products where possible) and anticoagulation is
required.
The coagulation and platelet product doses should be determined by fre-
quent laboratory monitoring, with the aim of restoring platelets to at least
50 × 109/l (perhaps even 80 × 109/l) and the coagulation screen to within refer-
ence limits, if the patient is bleeding. Traditionally, the fibrinogen is restored to
≥1g/l. Whether this is sufficient, given the higher reference range at the end of
pregnancy, is unknown. Other therapies, such as heparin, antithrombin concen-
trate, or tissue factor pathway inhibitor do not yet have a proven role in the man-
agement of DIC. Recent evidence suggests that recombinant activated protein C
therapy may have a beneficial effect on mortality from sepsis. However, a role for
such therapy in DIC in pregnancy is untested.
159
Obstetric CH08 7/11/05 2:13 pm Page 160
Major obstetric hemorrhage

See Chapter 3.
Neonatal alloimmune thrombocytopenia (NAIT)

Although NAIT is not associated with maternal thrombocytopenia, it is a signific-
ant cause of neonatal thrombocytopenia, occurring in 1 in 1000–2000 live births.
It can cause intracerebral hemorrhage (ICH), which most often occurs during or
after delivery, although up to 10% of cases may occur prior to the onset of labor.
The disease is caused by transplacental passage of fetal platelets, which initi-
ates a maternal reaction similar to that seen with red cells in RhD disease. This
leads to the maternal production of antiplatelet antibodies directed at antigens
on the fetal platelets (HPA inherited from the father). Five major HPA systems
have been identified as causing NAIT. Although HPA-1a is the most commonly
implicated antigen in the Caucasian population (>80% of cases), other HPA
platelet glycoprotein antigens, such as HPA-5b (5–15% of cases), can also result
in NAIT. In Asian populations, HPA-4a is commonly linked with the disorder.
Diagnosis of NAIT
The prevalence of the HPA-1a-negative platelet phenotype is about 2%. In most
countries, there is no routine screening procedure for anti-HPA-1a, or anti-HPA-
5b, so the first affected baby is usually diagnosed after birth. Ten per cent of ICH
occur antepartum (often between 30 and 35 weeks gestation), and occasionally
NAIT will be diagnosed in utero, when investigations have been carried out for
other reasons. Sixty per cent of identified cases occur in first pregnancies that are
otherwise uneventful, making it difficult to predict who may be at risk. The clini-
cal presentation of the affected infant may vary from incidental thrombocytope-
nia and/or a purpura rash to the presence of ICH. NAIT can lead to
developmental changes in 25% of affected subjects, with more severe disease
resulting in motor dysfunction and intellectual impairment. The diagnosis may
also come to light after investigations to exclude other causes of fetal thrombocy-
topenia, such as infection, birth defects, chromosomal abnormalities, intrauter-
ine growth restriction (IUGR), hemangioma or a maternal thrombocytopenia. It
may also be suspected where there has been a history of fetal ICH, recurrent mis-
carriage or fetal hydrocephalus. A diagnosis of NAIT should only be made after
exclusion of maternal history of an autoimmune disorder, thrombocytopenia or
drug abuse.
Predicting NAIT may be difficult, as antibody testing can produce conflicting
results. A mother with a positive antibody test may have an unaffected infant and
a mother with a negative result may have an affected infant. However, any woman
with a history of severe disease, such as resulting in neonatal death, should be
considered at risk and, consequently, tested for platelet alloimmunization. If
160
platelet-specific antibodies are identified, antigen phenotyping of parental speci-

mens is required to determine the platelet antigen system involved. Even in the
absence of positive serology, parental antigen incompatibility can still be identi-
fied with antigen phenotyping. In general, parents should be assessed prior to a
second pregnancy, so that the implicated antigen can be determined, and both
diagnostic and management strategies can be discussed. If the father is heterozy-
gous for the implicated antigen, it should be anticipated that the risk of recur-
rence, with the same parentage, will be 50%.
Investigations to determine the presence of a maternal alloantibody against a
paternal platelet antigen should use a combination of sensitive techniques, as a
maternal alloantibody may not be detectable in up to 20% of cases. In such cases,
the diagnosis will have to rely upon the clinical history and the presence of a rele-
vant HPA mismatch between the mother and father. When carrying out such inves-
tigations, it is important to exclude maternal platelet autoantibodies and HLA
antibodies. In addition, the maternal and paternal platelet antigens should be
determined, either by sensitive monoclonal antibody or genotyping techniques.
Treatment of NAIT
First pregnancy
As the platelet count may continue to fall in the first days after birth, infants
should be monitored daily. The presence of significant thrombocytopenia
requires ultrasound scanning for possible ICH. The passive thrombocytopenia
from NAIT persists for 1–3 weeks postpartum, depending on the rate of removal
of maternal platelet antibodies from the baby’s circulation. Compatible HPA-1a
negative or HPA-5b negative (as appropriate) allogenic platelets are most often
used for the neonatal treatment. In some instances, washed, irradiated (to
prevent graft versus host disease), maternal platelets may be used. If platelets are
unavailable, high-dose Ig therapy, at a dose of 1g/day for 1–3 days, has been
used. This requires 24–48 hours for an effect to be seen, but may be beneficial in
up to 75% of cases.
Second pregnancy
Two potential strategies for the management of a second pregnancy are:
the use of IVIgG ± steroids in the mother from around 24 weeks gestation to
inhibit antibody production and reduce placental transfer;
serial platelet transfusions for the fetus from around the same gestation.
At present, which management strategy is optimal is not yet clear. Indeed, a
combination of these strategies may also be useful.
IVIgG has been reported to improve the fetal platelet count in only 27% of
cases in some studies and in 66% in others, with an associated risk of ICH varying
161
from 0 to 7%. Furthermore, IVIgG therapy requires monitoring of fetal blood

count, which itself will require platelet cover. It has been suggested, however,
that where the previous sibling was not severely affected, fetal platelet count
monitoring may not be required.
The second strategy requires fetal full blood sampling from around 22–24
weeks gestation, with fetal platelet transfusions on a weekly basis into the intra-
hepatic vein or placental cord. This has been associated with two fatalities in 12
managed pregnancies. Where the fetus is, or is suspected to be, affected, delivery
will usually be by elective Cesarean section.
162
CHAPTER 9
Hereditary red cell

disorders
Hemoglobinopathy
The hemoglobin molecule is a tetramer made up of two α-like globin chains and
two β-like chains. The genes encoding the globin chains occur in two clusters on
chromsomes 16 and 11. At different times in embryonic and fetal development
the α-like chains can be either α or ζ-chains, and the β-like chains can be either
β-, γ-, δ-, or ε-chains. This results in a mixture of embryonic, fetal and adult hemo-
globins (Table 9.1). In late fetal development, the production of the fetal hemo-
globin F (HbF) (α2,γ2) reduces, leaving a predominance of adult hemoglobin A
(HbA) (α2β2) and <3.5% HbA2 (α2δ2).
Hemoglobin takes up oxygen at high oxygen saturation (in the lungs) and
rapidly releases oxygen in the relatively hypoxic peripheral tissues. Changes in
local pH, the red cell level of CO2 and the red cell level of the compound 2,3-
diphosphoglycerate (2,3-DPG) can all alter the affinity of hemoglobin for oxygen.
Hemoglobin’s affinity for oxygen is also important to the developing fetus, as
in general, HbF requires a higher affinity for oxygen than maternal HbA, to
allow transfer of oxygen from the maternal circulation.
The group of disorders which result in the production of a defective hemoglo-
bin are amongst the commonest maternal genetic disorders associated with an
adverse pregnancy outcome. This group ranges from the thalassemias, which are
characterized by a failure of production of a hemoglobin chain, to genetic
Table 9.1 – Hemoglobin (Hb) development
Stage Name α-chain type β-chain type Hb combination
Embryo Hb Gower 1 ζ ε ζ2ε2

Hb Portland ζ γ ζ2γ2
Hb Gower 2 α ε α2ε2
Fetal HbF α γ α2γ2
Adult HbA α β α2β2

Hb A2 α δ α2δ2
163
changes that result in an alteration of the conformational structure of the hemo-

globin (Table 9.2). Such conformational change can result in an alteration in
oxygen affinity, or an increased susceptibility of the hemoglobin molecule to
hypoxic damage. Principal amongst these are those disorders in which hypoxia
leads to the formation of a rigid hemoglobin structure: the sickling disorders.
The management of pregnancy in subjects known, or suspected, to have such
disorders can be divided into the management of an affected mother, the assess-
ment of the risks of fetal transmission from a maternal carrier and the manage-
ment of an affected fetus.
The thalassemias
The thalassemias are a heterogeneous group of genetic disorders of hemoglobin
synthesis: they are named after the hemoglobin that is deficient and can be
described as shown in Table 9.2. The mutation may result in a reduced rate of
production from the affected gene (for the β gene the notation used for this is
‘β+’ and for the α gene the notation is ‘α+’), or result in no globin chain synthesis
at all (again for the β gene the notation for this is ‘β0’ and for the α gene the
notation is ‘α0’). The majority of thalassemias are inherited in a Mendelian reces-
sive manner and there are known mutations that affect hemoglobin transcrip-
tion, mRNA processing, translation and post-translational modification.
Table 9.2 – Hemoglobinopathy types
Hemoglobinopathy type Subtypes
α Deletional
Non-deletional
β Deletional
Non-deletional
Normal HbA2 variant
Dominant variant
δβ Deletional
Non-deletional
γ –
δ Deletional
Non-deletional
εγδβ Large β gene deletion
Hereditary persistance of Deletional

HbF (HPFH) Non-deletional
164
Hereditary red cell disorders
Table 9.3 – Types of thalassemia intermedia
Beta thalassemias Homozygous mild beta thalassemia

Mild compound beta heterozygote
Dominant beta thalassemia trait
Compound heterozygote of beta and a beta variant (Hb C, E,
Lepore, δβ)
Alpha and beta combinations β+ with HbH disease

β+ with α+ or α0 mutations
Heterozygote β, co-inheritance of an extra α (ααα)
Beta mutations with increased γ-chains Homozygous β thalassemia with HPFH

Homozygous β thalassemia, with a γ promoter mutation
Given the diversity of genetic defects, and the possibility of genetic combina-
tions, thalassemias (irrespective of their molecular basis) are often classified by
their clinical effects into thalassemia minor, thalassemia intermedia and tha-
lassemia major. In general, thalassemia carriers are often symptomless and fall
into the minor category. Intermediate levels are more severely affected and may
often have anemia, although this does not require regular transfusion (Table 9.3).
In its major form, thalassemia presents with a lifelong transfusion dependency.
Alpha thalassemia
The α globin chain is common to both HbF and HbA, and deficiencies in the
gene (the alpha thalassemias), are found in subjects of Middle Eastern, Mediter-
ranean or African descent. As noted above, the human α globin gene cluster is
located on chromosome 16 and, normally, each individual has a total of four α
globin genes, with two on each chromosome. Although these genes are identical,
the α-2 globin gene location contributes significantly more α-globin chain pro-
duction to hemoglobin than the α-1 gene location. There are gene mutations
that can affect either one, or both, of the α globin genes on each chromosome.
Alpha thalassemia most often results from a deletion of these genes, although
non-deletional forms also occur (Table 9.4). The severity of the thalassemia phe-
Table 9.4 – Mechanism of alpha thalassemia
Hb Barts hydrops fetalis results from severe ‘(--/--)’

HbH commonly results from severe ‘(--/--α)’
Hb Bart and HbH have O2 affinity poor oxygen delivery
Hb Bart and HbH precipitate hemolysis
165
notype depends on the effect of the mutation on the production of α globin

chains (either complete deletion or reduced production) and the contribution
of the affected gene to the overall hemoglobin production.
The four genes are described as ‘(αα,αα)’ in a normal individual, with a ‘-’
used to describe a mutation. Mutations can then result in a number of combina-
tions. A single gene mutation is described as ‘(α-,αα)’. A two-gene mutation is
described as ‘(α-,α−)’ if one mutation is found on each chromosome and ‘(αα,-- -- )’
if both mutations are found on the same chromosome. If three mutations occur,
this is represented as ‘(-- --,-- α)’. Finally, if there are four mutations then this is
described as ‘(-- --,-- --)’. In addition, and as noted above, mutations which lead to
no gene output are also described as an ‘α0’, mutation and mutations which
result in some gene production are described as an ‘α+’ mutation.
In general, one- and two-gene mutations produce a symptomless carrier state,
but two mutations on the same chromosome, i.e. ‘(--,αα)’, produce a more severe
hematology pattern than when one mutation occurs on one chromosome and
one on the other, i.e. ‘(α-,α-)’.
When both the mother and father are carriers of an ‘α0’ ‘(-- --,αα)’, mutation,
then there are four possible outcomes:
The fetus is normal, i.e. ‘(αα,αα)’;
The fetus carries the paternal mutation, i.e. ‘(-- --,αα)’;
The fetus carries the maternal mutation, i.e. ‘(-- --,αα)’; or
The fetus inherits the ‘α0’ mutation from both parents and is therefore ‘(-- --,-- --)’.
In ‘α0’ mutations, when there is no α-chain production, the fetus attempts to
form tetramers of the available β-chain (γ), forming a γ-tetramer that is called Hb
Barts. Hb Barts has a high oxygen affinity and, if it is major hemoglobin, leads to
a severe anemia during fetal development, with the risk of cardiac failure. This
most often leads to stillbirth, or neonatal death in the third trimester (called Hb
Barts hydrops fetalis); the name comes from St Bartholomew’s Hospital in
London, where the condition was first described in a baby of Chinese origin.
Where an ‘α0’ carrier with ‘(-- --,αα)’ is partnered by an ‘α+’ carrier with
‘(α-,αα)’, this also results in four possible outcomes:
The fetus is normal, i.e. ‘(αα,αα)’;
The fetus carries the ‘α+’ mutation, i.e. ‘(α-,αα)’;
The fetus carries the ‘α0’ mutation, i.e. ‘(-- --,αα)’; or
The fetus inherits both the ‘α+’ and ‘α0’ mutations, i.e. ‘(-- --/-α)’.
In such circumstances, unlike Hb Barts hydrops fetalis, as there is some α-
chain production, the fetus is able to form some αβ hemoglobin, but also
attempts to form a tetramer of the available β-chains. This results in the forma-
tion of some β-chain tetramers and results in HbH disease. These HbH (β-chain)
tetramers can be seen in the cell and are called HbH inclusions. The abnormal β-
chain tetramers also result in a hemoglobin with a high oxygen affinity. Such
166
hemoglobins fail to deliver adequate oxygen to the tissues and these tetramers
also tend to precipitate within red cells, which shortens red cell survival by
hemolysis in the spleen (see Table 9.4).
Alpha thalassemia: problems in pregnancy

The problems in pregnancy can be considered from the viewpoint of maternal
carriers of one- or two-gene mutations, maternal carriers with an affected fetus,
the problems associated with maternal carriage of HbH disease and the features
of a fetus with Hb Barts.
Maternal alpha thalassemia carriers

As noted above, women with one or two ‘α0’ gene deletions,‘(α-,αα)’, ‘α-,α-)’ or
‘(-- --,αα)’, are usually symptom-free and have a normal pregnancy outcome.
Despite the fact that she is usually symptom-free, if a mother is carrying a fetus
affected by Hb Barts, she has a higher risk of pregnancy complications as a con-
sequence of placentomegaly. It has been estimated that at least 50% of subjects
carrying an Hb Barts fetus will have evidence of pregnancy-induced hypertension
and around 30% will have evidence of pre-eclampsia. It may be that other com-
plications, such as polyhydramnios, antepartum and postpartum hemorrhage,
disseminated intravascular coagulation (DIC), placental abruption, premature
labor and malpresentation, are also more common in such women (Table 9.5).
A mother with HbH disease

The majority of subjects are diagnosed in childhood, perhaps coming to light as a
result of acute hemolysis induced by infection (when hemoglobin levels may fall
by up to 3g/dl). The clinical features of an adult with HbH disease vary from mild
asymptomatic anemia (commonly with a hemoglobin of >8g/dl) to a severe trans-
Table 9.5 – Features of Hb Barts hydrops fetalis
Fetal
Third trimester stillbirth/neonatal death
Large placental volume
Pallor
Generalized edema
Marked hepatosplenomegaly
Other congenital anomalies
Hb 6–8g/dl
80% Hb Barts/20% Hb Portland
Maternal
Pre-eclampsia (PET)
APH and PPH due to placentomegaly
APH, Antepartum hemorrhage; PPH, Postpartum hemorrhage

167
fusion-dependent anemia, with jaundice, hepatosplenomegaly, growth restriction

and bone abnormalities. The symptoms of maternal HbH disease are variable,
with mild to moderate hemolysis being the predominant feature. The level of
HbH inclusions may be increased by pyrexial illness. Iron overload, even in the
absence of previous transfusion, is not uncommon in subjects from Far Eastern
countries. In addition, gallstones are not infrequent, and 20–70% of cases will
have hepatomegaly or splenomegaly. Infections, fever, oxidative drugs, parvovirus
infection, as well as pregnancy itself, may worsen the anemia (Table 9.6).
Fetal alpha thalssemia carriers

In newborn infants with ‘(-α/-α)’, a variable quantity of Hb Barts (0.5–2%) is
detectable at birth. In cases of ‘(-- --/αα)’ carriage, Hb Barts may constitute up to
10% of the hemoglobin at delivery, although this disappears by 6 months of age,
as there are sufficient α-chains for hemoglobin synthesis. Fetal carriers of mild α-
chain mutations are, as adults, unlikely to be affected during development, but
occasional HbH inclusions may be detected in the blood.
Fetal HbH disease

The more severe HbH disease, ‘(-- --/-α)’, most often presents in childhood as a
mild to moderate hemolytic anemia, although fetal HbH hydrops has rarely been
reported in the literature. There have also been isolated case reports of an associ-
ation between HbH disease and skeletal changes, facial dysmorphia and
retinopathy.
Fetal Hb Barts hydrops fetalis

In South East Asia, ‘α0’ mutations are common, and are the commonest cause of
hydrops fetalis in that region, where they may account for 1 in 4 perinatal deaths.
Hb Barts hydrops fetalis has also been reported in subjects of Mediterranean
descent.
In Hb Barts hydrops fetalis, γ-tetramers lead to hemoglobin instability, severe
anemia and extramedullary hemopoiesis (leading to hepatosplenomegaly). As
Table 9.6 – Clinical features of HbH disease
Survival to adulthood
Inheritance of ‘(--,-α)’ or homozygous non-deletion ‘α+α+’
Variable anemia
Variable splenomegaly
Periodic hemolysis with oxidant drugs or infection
Prominent red cell HbH inclusions
Hb 7–10g/dl
<40% HbH, variable HbA and HbA2
168
the change from ζ-chains to α-chains occurs by the 7th week of gestation, there is
a lack of fetal oxygenation from early in development. This results in fetal
hypoxia, abnormal organogenesis and placental enlargement. A number of con-
genital anomalies have also been reported in association with Hb Barts, which
include abnormal brain, cardiac, skeletal or urinary tract development.
The clinical signs on Hb Barts-delivered fetus may vary from signs of cardiac
failure to gross edema, hepatosplenomegaly and pallor (see Table 9.5). Although
the vast majority of fetuses with Hb Barts hydrops fetalis die, there are a number
of cases of severe alpha thalassemia which survive, to be maintained on transfu-
sion and chelation regimes, with or without, the use of subsequent bone marrow
transplantation.
Alpha thalassemia: diagnosis in pregnancy

Diagnosis of maternal carriers
Carriage of alpha thalassemia is associated with a mean red cell volume (MCV)
<80fl (often <70fl) with a mean red cell hemoglobin (MCH) <27pg, with, very
often, no evidence of anemia (Figure 9.1). In such cases, if iron deficiency is
excluded, then carriage of thalassemia should be suspected. If suspected in one
parent, the other parent should be examined to exclude a hypochromic/micro-
cytic blood picture. There should always be a high index of suspicion if there is a
previous family history of hydrops fetalis.
FBC +
Hb ELECTROPHORESIS
NORMAL β TRAIT IDA α TRAIT δ+β TRAITS HPFH

A 2 ok, β THAL
MCH >27 MCH <27 MCH <27 MCH <27 MCH <27 MCH <27
HbA √ HbA √ HbA √ HbA √ HbA √ HbA √

HbA 2 <3.5% HbA 2 >3.5% HbA 2 <3.5% HbA 2 <3.5% HbA 2 <3.5% HbA 2 <3.5%
Ferritin √ HbF 0.1–7% Ferritin Ferritin √ Ferritin √ Ferritin √
HbH + HbH - HbF 2–30%
β mutation Iron α mutation β mutation

by PCR repletion by PCR by PCR
Figure 9.1 Discrimination of hypochromia and microcytosis

FBC, Full blood count; Hb, Hemoglobin; HPFH, Hereditary persistence of HbF; IDA, Iron deficiency anemia;
MCH, Mean red cell hemoglobin; PCR, Polymerase chain reaction; Thal, Thalassemia.
169
Confirmation of carriage of alpha thalassemia trait requires a number of addi-

tional investigations. The search for HbH inclusions in the blood film is often
included in the screening for alpha thalassemia, although this can be problem-
atic given the infrequency of these inclusions in minor carrier states. A raised
level of HbA2 (>3.5%) in someone with hypochromia and microcytosis often sug-
gests carriage of a beta rather than alpha thalassemia. The presence of a normal
level of HbA2 in such individuals points towards alpha thalassemia carriage. A
normal level can occur, however, if the subject is a carrier of both alpha and beta
thalassemia traits. Carriers of some alpha thalassemia mutations, such as ‘-SEA’
may have detectable levels of ζ globin chains in red cells that can be detected by
immunophenotyping, although larger mutations may result in no detectable ζ-
chains. Although the diagnosis of alpha thalassemia is often made on the basis of
the absence of iron deficiency and/or other hemoglobinopathy in someone with
a hypochromic anemia (with or without HbH inclusions), for counseling it is
necessary to determine the exact parental mutation. This can be achieved by a
polymerase chain reaction (PCR) technique or Southern blot analysis.
Diagnosis of Hb Barts hydrops fetalis

In a case of suspected Hb Barts hydrops fetalis, ultrasonography may detect pla-
centomegaly as early as 10 weeks gestation, and hydropic changes in the fetus are
often evident at <20 weeks gestation. Fetal tissue (obtained by chorionic villus
sampling (CVS) in the first trimester or by amniocentesis in the second) can be
examined by PCR or Southern blot analysis for an alpha thalassemia mutation.
As with all such fetal sampling, contamination with maternal DNA, or difficulties
in establishing paternity, can be problematic. Examination of the fetal blood by
cordocentesis in the second trimester will also permit examination of the
expressed fetal hemoglobins by hemoglobin electrophoresis. A normal second
trimester fetus will have predominantly HbF and ≤10% HbA, but in a fetus with
Hb Barts the predominant hemoglobin will be Hb Barts (γ4), with <20% Hb Port-
land and possibly some HbH (β4).
Diagnosis of fetal carriage of alpha thalassemia traits

The detection of less severely affected infants, with either carriage of an α0
‘(-- --,αα)’ mutation, or HbH disease, can be achieved by detection of HbH or Hb
Barts in neonatal samples. Such detection may also indicate those parents at risk
of a future affected child.
Alpha thalassemia: treatment in pregnancy

The carriage of minor thalassemic traits is unlikely to require specific inter-
vention in pregnancy other than attendance to folic acid supplements and the
exclusion of fetal carriage of severe thalassemia.
In maternal HbH disease, pre-existing anemia will be exacerbated by the phys-
iological hemodilution associated with pregnancy (Table 9.7). Consequently, all
mothers should attend to folic acid supplementation, as folate deficiency often
170
Table 9.7 – Maternal HbH treatment in pregnancy
Folic acid
Avoid antioxidant drugs
If Hb <6g/dl consider transfusion
Treatment of pre-eclampsia (PET) and gallstones
Screening for infant carriage of thalassemia
accompanies chronic hemolysis. Antioxidant drugs should be avoided. From the

limited literature it appears that, in general, maternal HbH disease is consistent
with a normal pregnancy outcome. In carriers there is, of course, the potential
for transmission of either the ‘(-- --)’ or the ‘(- α)’ mutation to the fetus (see
above). If the hemoglobin falls to <6g/dl, there may be a case to transfuse the
mother to improve maternal and fetal oxygenation. There are some reports that
pre-eclampsia may be more common in mothers who have HbH disease. At
present, however, the role of maternal HbH disease in other pregnancy prob-
lems, such as miscarriage and cardiac failure, requires further study.
Beta thalassemia
The beta thalassemias are found in those of Baltic, Mediterranean, Middle
Eastern, African, Indian, Southern Chinese and South-East Asian extraction.
Akin to the situation in alpha thalassemia, the defect of β-chains in beta tha-
lassemia leads to the formation of α-chain tetramers, which interfere with the
maturation of the red cell and result in red cell destruction within the marrow
and the spleen (Table 9.8). In beta thalassemia, this reduction in β-chains is
accompanied by an increase in δ-chain hemoglobin production, which is identi-
fied as an increase in the level of HbA2 (α2δ2).
In the homozygous, or compound heterozygous, beta thalassemic states, there
is a lifelong chronic dyserythropoietic anemia, which results in splenomegaly and
Table 9.8 – Mechanism of beta thalassemia
α tetramers hemolysis in bone marrow and spleen

Reduced α-chains, or increased HbF improves symptoms
HbA2 is common
Beta thalassemia major presents in the first year

Insufficient transfusion deformity and infection
Insufficient chelation iron overload
171
skeletal deformity. This clinical picture is ameliorated by any mutation that

reduces the amount of α-chain production, as this reduces the imbalance in the
relative amounts of α- and β-chains, and reduces dyserythropoiesis. The disease is
also ameliorated in subjects who have also inherited hereditary persistence of
HbF (HPFH).
Subjects affected by severe beta thalassemia usually present within the first
year of life, with symptoms of anemia and evidence of splenomegaly. The pro-
gression of the disease is heavily dependent upon the adequacy of transfusion
and iron chelation. In the absence of adequate transfusion there may be pro-
found anemia, marked skeletal deformity of the long bones and skull, recurrent
infections, and death within the first few years. With adequate transfusion there
may be few symptoms and no development of a hyperdynamic state. By the
second decade, however, the effect of transfusion-related iron overload will result
in adrenal, pancreatic, hepatic and cardiac failure (Table 9.9). This may also
compromise sexual development and, in the absence of adequate chelation,
death is often due to cardiac failure or dysrythmia. With a new generation who
have received adequate and safe transfusion, and adequate chelation, there is the
hope of an improved life expectancy.
Beta thalassemia: problems in pregnancy

As with alpha thalassemia, carriers of beta thalassemia are usually symptom-free,
although a mild anemia may develop in pregnancy. More significant anemias
may occur in relation to dietary deficiency. In such cases there may be a require-
ment for folate and iron supplementation.
As noted above, females with homozygous beta thalassemia usually suffer from
endocrine abnormalities due to iron overload. This results in a failure of puber-
tal growth and severely delayed sexual development, and hypogonadotrophic
hypogonadism. On account of this, only a small number of successful pregnan-
cies are reported in the literature. Some of these have occurred spontaneously
and others have required stimulation of ovulation.
In severely affected individuals, a number of risks have to be considered
before embarking on pregnancy. These include the cardiovascular risk to the
mother and the impact of other iron-overload effects, such as diabetes mellitus
and hypothyroidism, that may be present. Cardiac complications are associated
Table 9.9 – Iron overload in beta thalassemia
Iron overload is evident in the second decade

Pituitary iron overload impaired sexual development
Hepatic iron overload cirrhosis
Cardiac iron overload dysrythmia and failure
Endocrine overload hypothyroidism and diabetes mellitus
172
with mortality in both transfused and un-transfused patients, which is important

in pregnancy where major changes in cardiac function occur, with a substantial
increase in cardiac output evident by 12 weeks gestation that may be com-
pounded by anemia.
Serum ferritin may provide a reasonable reflection of hepatic iron stores, but
does not relate well to pituitary or cardiac deposition. Myocardial iron deposition
can provoke left ventricular dysfunction, along with superimposed myocarditis.
Where significant left ventricular dysfunction is present, or when significant
arrthymias have occurred, then pregnancy may be best avoided. Most non-inva-
sive cardiac imaging techniques are relatively insensitive for detecting early
cardiac iron deposition. However, magnetic resonance imaging (MRI) tech-
niques can now quantify cardiac iron deposition and relate these to left ventricu-
lar dimensions.
The most common adverse event reported in successful pregnancies is the
need for Cesarean section, given that there is a high risk of cephalopelvic dispro-
portion. This is due to the small stature of the mother when the unaffected fetus
exhibits normal growth. The risk of pregnancy-related thrombosis in those
women who have previously undergone splenectomy is unknown, but there is a
risk of mechanical difficulties from an enlarged spleen if it is still present. In
those mothers who have undergone splenectomy, the need for antibiotic prophy-
laxis and vaccination should be remembered, as well as the need for thrombo-
prophylaxis (see Chapter 8, Table 8.10).
In mothers with well-treated (hemoglobin chronically maintained >10g/dl)
and well-chelated beta thalassemia, or in women with thalassemia intermedia
(with no evidence of bone abnormality and a hemoglobin level >7g/dl without
regular transfusion), there is a reasonable pregnancy success rate. In one study of
women who had undergone regular transfusion and chelation, and had normal
left ventricular function with no evidence of a red cell autoantibody at concep-
tion, a successful pregnancy was achieved in 21 of 23 pregnancies. In this study,
one miscarriage and one fatal fetal abnormality (perhaps related to maternal
iron overload) was reported. All mothers had their chelation stopped at least by
the confirmation of conception. Two mothers suffered biliary colic due to gall-
stones, but there was no evidence of significant cardiac, renal, hepatic or
endocrine deterioration in any of the subjects, and the fetal outcome was compa-
rable to that of the general population. This suggests that a careful pre-preg-
nancy medical assessment may result in appropriate counseling and the
possibility of a successful pregnancy in at least some women with beta thalassemia
major.
Beta thalassemia: diagnosis in pregnancy

Maternal
In beta thalassemia carriers, the full blood count reveals a reduction in the MCV
and MCH. Hemoglobin electrophoresis reveals the presence of normal HbA,
with elevated levels of HbA2 (4–6%), and perhaps an elevated level of HbF.
173
In homozygous, or compound heterozygous, beta thalassemia major, HbA

(α2β2) is absent and only HbF and HbA2 are evident. In less severe (β+) tha-
lassemia, HbF may constitute <90% of the hemoglobin. As the measured HbA2
level is an average of all cells, this may well be within normal limits. As with alpha
thalassemia, DNA analysis is needed for prenatal diagnosis.
Fetal
At birth, absence of HbA is a feature of beta thalassemia major. Although there is
little evidence that early detection is cost or clinically effective, early detection
will allow the commencement of surveillance and the early onset of transfusion
therapy. When heel stab cards are used for diagnosis, storage changes may occur
if there is a long delay in the analysis of the neonatal sample, making identifica-
tion of hemoglobins difficult. If a cord sample is used for neonatal diagnosis,
cross-contamination with maternal blood may be a problem. This can be sus-
pected if there is a high level of HbA, but careful cleaning of the cord may
markedly reduce the risk of such contamination. Hemoglobin identification can
be achieved by cellulose acetate electrophoresis, although immuno-electro-
phoresis and high-performance liquid chromatography (HPLC) are more sensi-
tive methods. The nature of the genetic mutation can be determined by a
number of different PCR-based techniques, often using a combination of probes
that relate to mutations that are common in the ethnic population of the patient
under study. The exact screening technique used depends upon the ethnic mix
and knowledge of predominant mutations.
Beta thalassemia major: treatment in pregnancy

Transfusion requirements increase with increasing gestation, and the aim should
be to maintain the hemoglobin concentration >10g/dl. When transfusion is
insufficient there is the risk of fetal hypoxia and pre-term delivery. As in the non-
pregnant, such blood should be screened for hepatitis A, B and C, and HIV, and
the cytomegalovirus (CMV) status of the individuals should also be taken into
account. All individuals undergoing regular transfusion programs should be vac-
cinated against hepatitis B and A. They should also have a red cell phenotype
determined prior to the onset of such transfusion programs (see Chapter 2) so
that, whenever possible, blood is used that is compatible with this phenotype to
prevent red cell alloimmunization. In the UK, blood is routinely leukodepleted,
but if this is not the case, consideration should be given to bedside leukocyte fil-
tration to reduce the risk of febrile non-hemolytic transfusion reactions as well as
CMV transmission. In such individuals there may also be a role for cell salvage at
delivery.
There is evidence, from animal studies, that the chelating agent desferrioxam-
ine may be teratogenic (Table 9.10). Also, as the newer chelating agents have no
safety track record in pregnancy, the role of continuing chelation therapy during
pregnancy is not yet defined. The increase in serum ferritin in such subjects is
likely to be <10% over the course of pregnancy. However, whether this represents
174
a reduction in iron loading from hemodilution or fetal usage of iron requires

further study. Currently, in the vast majority of planned pregnancies, chelating
agents will be stopped when conception is confirmed, but may be restarted after
delivery, even if breastfeeding. There is also a suggestion that vitamin C therapy
should be stopped in pregnancy, as it may increase intestinal absorption of iron.
Folic acid, at a dose of at least 5mg should be utilized in all subjects as the
chronic hemolytic state may result in folate deficiency, with the attendant risks
upon the maternal blood as well as fetal neural tube development. The mother
will require additional monitoring if there is evidence of significant maternal
endocrine dysfunction, particularly for thyroid and parathyroid dysfunction and
glucose intolerance. At term, consideration should be given to the possibility of
cephalopelvic disproportion (see above) when planning delivery. In those
women with hypersplenism, thrombocytopenia (<80 × 109/l) may contraindicate
neuraxial anesthesia. The presence of short stature, spinal abnormalities and
osteoporosis (and possible vertebral collapse), and disturbed linear growth asso-
ciated with spinal cartilaginous dysplasia due to childhood chelation therapy,
should also be taken into consideration with neuraxial anesthesia. Ideally, pre-
pregnancy spinal X–ray examination should be performed in women with these
complications.
Thalassemia intermedia
This term covers a variety of thalassemic disorders, in which a combination of
defects results in an anemia that is more severe than that found in carrier states
but does not result in lifelong transfusion dependence (see Table 9.3). This clini-
cal definition includes HbC or HbE thalassemia, Hb Lepore and thalassemic syn-
dromes that arise as a result of a variety of combinations of α, β and δβ
deficiencies. As noted above, the inheritance of an α-chain mutation, along with
a homozygous β-chain disorder, will ameliorate the clinical picture by reducing
the β/α-chain imbalance and the consequent ineffective erythropoiesis.
Table 9.10 – Management of beta thalassemia in pregnancy
Pre-pregnancy assessment if possible

Assess cardiac/endocrine reserve
Beta thalassemia carriers at risk of a mild anemia in pregnancy

Continue folic acid throughout pregnancy
Stop vitamin C
?Stop chelation
Consider infection risk if previous splenectomy
Consider risk of cephalopelvic disproportion
175
These combinations result in clinical syndromes which run from mild anemia
to the more obvious signs of a chronic hemolytic anemia, such as folate defi-
ciency, gallstones, leg ulceration, infection and extramedullary hemopoiesis
(with skeletal abnormality and splenomegaly).
Stem cell transplantation

It is now possible to treat the major hemoglobinopathies with hematopoietic
stem cell transplantation from a human leukocyte antigen – (HLA)-compatible
donor, and there is now a considerable body of literature reporting successful
treatment of thalassemia major, as well as, sickle-cell disorders.
Transplant procedures are not, however, an easy option, as they require, at
present, intensive chemotherapy to ablate the existing bone marrow. Such a pro-
cedure also necessitates a period of profound neutropenia and thrombocytope-
nia. This carries a significant procedure-related morbidity and mortality, and the
transplant also requires subsequent lifelong immunosuppression. There are also
the long-term problems of graft versus host disease and, in children, growth dis-
turbance (particularly after radiotherapy marrow conditioning). Such proce-
dures also carry a greater risk in older subjects, in those with iron overload and in
those with viral infections of the liver. Furthermore, a graft failure rate of around
10% has been reported in the transplantation of sickle-cell disorders.
At present, such procedures are therefore limited to younger, severely
affected individuals, where there is a suitable donor – preferably a sibling, as this
carries a lower risk of graft versus host disease and, in general, less procedural
mortality than an unrelated transplant – and where there is little evidence of iron
overload or organ failure. In developed countries, however, <30% of indivudals
have a suitable family donor and, indeed, such transplants are only pertinent to
countries where there are the considerable resources and facilities required for a
successful outcome.
A number of developments may increase the applicability of hematopoietic
stem cell transplantation for thalassemia and sickling disorders in the future. The
growth of bone marrow registries and cord stem cell banks will increase the avail-
ability of suitable donors, although, at present, the role for unrelated allogenic
transplants in hemoglobinopathies is not yet defined. As noted in Chapter 2,
cord stem cells may also offer the additional benefit of less stringent matching
requirements. It has also been reported that a number of transplants for tha-
lassemia have resulted in the formation of a stable chimer of donor and recipient
in the marrow. This appears to lead to a reduction in the severity of graft versus
host disease, whilst maintaining a useful amelioration of thalassemic symptoms.
This is one of the rationales for a reduction in the intensity of the bone marrow
conditioning regimens in thalassemia transplants from that required for trans-
plantation in malignant conditions. Indeed, successful transplants for hemoglo-
binopathies have been reported with purine analog conditioning, although a
reduction in the dose of cyclophsophamide used is associated with an increase in
176
graft rejection and graft failure. On account of this, when a transplant is being
considered for a young, otherwise fit, individual with a likelihood of a good
outcome, then full conditioning is probably to be recommended. Lesser induc-
tion regimens may be considered for the older, less fit, individual.
Sickle-cell disease
Sickle-cell disease is the most common hemoglobinopathy in the USA, with around
8% of African Americans carrying the sickle-cell gene. Disorders which result in
clinically significant sickling of the red cell result from the combination of a sickle-
cell gene on one β-chain combined with (on the other chromosome) either:
Another sickle-cell mutation (HbSS); or
An HbC mutation (HbSC); or
A beta thalassemia gene (HbSβ0).
Sickle-cell disease can result in a lifelong crippling hemolytic disorder (char-
acterized by crises caused by infection, aplasia, infarction and hemolysis) to a
diagnosis only detected on a routine blood film examination (Table 9.11). This
variation in symptoms can be due to the co-inheritance of other hemoglobin
changes, such as the HPFH or α thalassemia. The overall clinical expression of
the disease is also highly dependent on the availability of health care.
Table 9.11 – Sickle-cell disease features
Major sickling occurs in SS, Sβ0, SC, SD disease
Crisis management Hemolytic

Aplastic
Infarction: pain, bones, cerebral ischemia of arterial
origin (CIAO), myocardial infarction
Sequestration – liver, spleen (children), lung
Symptom amelioration Influence of HbF

Treatment with hydroxybutyrate ( HbF)
Influence of healthcare
Transfusion management Alloimmunization

Iron overload
Blood-borne infections
Infection risk Osteomyelitis

Capsulated organisms
Fetal risks of maternal sickle-cell disease Intrauterine growth restriction (IUGR)

Pre-term delivery
177
Sickle-cell disease most often presents with infarction of the fingers in child-
hood, called dactylitis. There may be also recurrent infection or sequestration of
blood into the spleen causing profound anemia. However, due to repeated
infarction, it is unusual for there to be a palpable spleen by the teenage years,
and indeed there is, more often, hematological evidence of hyposplenism. These
infarction crises are often precipitated by dehydration or infection, and present
with pain which can be localized or diffuse. Childhood infection with any organ-
ism, but particularly parvovirus, may trigger an aplastic crisis with profound
anemia. In childhood, infarction crises may also present with a cerebral ischemia
of arterial origin (CIAO). There is also the possibility of sequestration of sickle
cells into the liver, spleen and lungs. Such lung sequestration will result in a
rapid deterioration in oxygenation and is a common cause of death.
With repeated crises there may be the development of bone deformity,
osteomyelitis, renal failure, myocardial infarction, leg ulceration, gallstones and
cardiac failure. Indeed, almost 30% of young adults with sickle-cell disease
develop a widespread vasculopathy with glomerulosclerosis, retinopathy, restric-
tive lung disease and stroke. With repeated transfusions there may also be an
increased risk of blood-borne infections and iron overload.
In SS disease, but more severely in the case of SC disease, there is an increased
risk of a proliferation of blood vessels in the retina (due to hypoxia), ultimately
causing vitreous hemorrhage, retinal detachment and blindness. This lifelong
series of painful crises and hospital admission also results in a disturbance of
schooling and development.
Sickle-cell disease in pregnancy

It is very likely that maternal carriage of a sickle-cell trait is associated with an
uncomplicated maternal and fetal outcome. In such individuals, sickling only
occurs at PaO2 levels <15mmHg and is very unlikely to give rise to clinical prob-
lems in normal obstetric practice.
As noted above, the outcome of pregnancy in mothers with sickling dis-
orders is heavily dependent upon the adequacy of maternal health care. In the
USA, a maternal mortality rate is likely to be of the order of 0.25–0.5%, with
some 99% of pregnancies (which were viable after 28 weeks) resulting in a live
birth. Although the majority of pregnancies result in a healthy child, around
half of pregnancies are complicated by at least one painful crisis, and SS
patients are more likely to require hospital admissions. It is possible, however,
that these painful crises are no more frequent than would be expected in the
same women when not pregnant, and the rate of pyelonephritis has been
reported to be no greater than in women without a sickling disorder in some,
but not all, studies.
There is likely to be an increased risk of pre-eclampsia, which has been
reported to occur in 14% of subjects with a sickling disorders, including those
with only a sickle-cell trait. There is also an effect on fetal growth, with 70–85% of
178
pregnancies resulting in a baby below the 50th centile for the population and, in
up to 27%, the baby is below the 10th centile. This may relate to the known
association of sickling disorders and placental infarction, although the risk of a
small baby has also been linked with the occurrence of anemic events during
pregnancy and inversely related to maternal HbF levels.
Anemic events do not appear to be limited to any specific gestation, but they
are more common in sickle/thalassemia combinations than they are in pure
sickle-cell disease (Table 9.12). In African Americans, the incidence of acute
chest syndrome appears not to be increased in pregnancy. There is, however,
likely to be an increased risk of premature rupture of membranes, pre-term deliv-
ery and postpartum maternal infection. Multiple pregnancies, as a result of a
reduction in available nutrition and oxygen transfer to the fetus, appear to be at
more risk of poor fetal growth. Whether the use of narcotic analgesia affects fetal
development is not clear, although some may have a vasoconstrictive effect on
the placental vasculature.
An increased risk of a growth-restricted baby and postpartum infection has
also been reported in subjects with SC disease, although there is a lesser need for
transfusion and fewer painful crises in SC than in SS disease. In developing coun-
tries the outcome of pregnancy in subjects with a major sickling disorder may be
substantially worse, with maternal mortality approaching 25% and a fetal mortal-
ity approaching 40%.
Sickle-cell disorders: diagnosis in pregnancy

Sickle-cell trait results in no change to the hematological indices, i.e. the MCV
and MCH are normal unless another hemogloibin disorder is present. The trait
is diagnosed by a positive sickle test and the demonstration of both an HbA and
HbS band on gel electrophoresis, indicating the presence of both normal and
sickle hemoglobin.
Sickle-cell disease is diagnosed by the presence of anemia, sickled red cells on
the blood film, blood film appearances of hyposplenism, a positive sickle test: it is
confirmed by the hemoglobin electrophoresis pattern of HbS, and elevated HbF
with no HbA. The presence of a microcytosis may suggest the co-inheritance of
thalassemia, or the presence of iron deficiency. A higher than expected hemo-
Table 9.12 – Causes of anemia in sickle-cell disease in pregnancy
Blood loss
Infection
Vitamin deficiency
Aplastic crisis
Inflammation
179
globin level (11–13g/dl) may indicate the presence of HbC, or co-inheritance of

another hemoglobin variant.
To determine the presence of sickle-cell disease in the fetus, the father (where
paternity can be assured) should be screened for the presence of any hemoglo-
binopathy. The presence of sickle-cell disease in the fetus can be confirmed by
the analysis of fetal DNA for sickle-cell mutations using techniques outlined pre-
viously (see Chapter 4). The newborn is protected from sickling in the first few
months of life due to the high level of HbF at birth.
Sickle-cell disease: treatment in pregnancy

General management
In all subjects with a major sickling disorder, treatment must include the preven-
tion of infection. This is achieved with prophylactic penicillin and the use of
pneumococcal, meningococcal and Hemophilus influenzae vaccinations, with anti-
malarial prophylaxis where appropriate (Table 9.13). The management of a
painful crisis involves adequate pain control (with opiates if required), treatment
of any infection, and maintenance of oxygenation and hydration.
A worsening of anemia may be precipitated by a number of events (see Table
9.12). In general, regular blood transfusion is not required. If, however, there is
evidence of a falling hemoglobin level (indicating an increase in hemolysis), and
especially if there is evidence of a falling reticulocyte count (indicating an
impending aplastic phase), then transfusion should be given. When transfusion
is required and the hemoglobin level is already <5g/dl, it may be that a top-up
transfusion to 12–14g/dl will result in sufficient dilution of the sickle cells (to
reduce sickling) to the desired target level of <30% of the circulating red cells.
When transfusion is required at a higher hemoglobin level (8–10g/dl), then a
partial exchange transfusion should be carried out (removing 500ml by phle-
botomy whilst transfusing 2 red cell units).
Lung and sequestration crises require close monitoring. Sequestration into the
lung may present with fever, cough, pleuritic chest pain and evidence of lung infil-
trates on X-ray. In such cases, the patient is likely to require intensive-care therapy
with oxygen support, monitoring, hydration and exchange transfusion. In the ill
patient with a major sickling disorder, the possibility of sequestration into liver or
spleen may be assessed by regular measurement of the hemoglobin level and
assessment of the liver and spleen size. If acute sequestration occurs, the hemoglo-
bin level can drop within hours and urgent transfusion will be required.
It is recommended that a subject with a recent episode of CIAO should
receive regular red cell transfusions to keep the hemoglobin level >10g/dl and to
suppress the HbS level to <30% of the circulating red cells. A proliferative
retinopathy requires opthalmological assessment with laser therapy to prevent
bleeding.
In addition to SS disease, obstetricians must be aware of HbSC disease, where
there may be essentially normal levels of hemoglobin during the pregnancy, such
180
Table 9.13 – Management of sickle-cell disorders in pregnancy
Diagnosis Sickle test +ve,

HbS and HbF present, but no HbA
Sickle cells on blood film
Infection management Penicillin
Pneumovax
Prompt recognition of respiratory tract infection (RTI)/
urinary tract infection (UTI)
Malarial prophylaxis (if appropriate)
Pain management Multidisciplinary action
Controlled analgesia
Day unit attendance
Renal management Adequate hydration and urinary output
Prompt infection treatment
Anemia management Folate supplementation
Iron supplements (if appropriate)
Minimize red blood cell (rbc) alloimmunization
Consider transfusion (reduce sickle-cells to <30%) in:
Pre-eclampsia (PET)
Preoperative work-up
Acute anemia
Septicemia
Acute renal failure
Acute chest syndrome
Recent cerebral ischemia of arterial origin (CIAO)
Multiple pregnancy
Delivery management Planning, especially if placental insufficiency or poor maternal reserve
Neonatal screening Counseling
Remember Gallstone/cholecystitis risk
Retinopathy risk
that the obstetrician and the mother are unaware of the problem. These women
are, however, at risk of severe sickling crises during pregnancy and the peurperium.
Pregnancy-specific management
At the first antenatal visit the presence of any factors which could affect preg-
nancy outcome should be assessed. This includes the past obstetric history, evid-
ence of chronic organ damage and the use of, and need for, narcotic analgesia.
At all stages a multidisciplinary approach involving obstetrician, hematologist,
anesthetist and general/infectious disease physician is required for appropriate
management and planning of pregnancy cases.
181
The mainstays of management of a pregnancy in subjects with a severe sick-

ling disorder are: folic acid supplementation (at a dose of at least 5mg/day for
the whole pregnancy); regular hemoglobin level estimations; regular monitoring
of fetal growth; consideration of regular transfusion when there is a twin or
triplet pregnancy (see Table 9.13). The mother should be advised on the impor-
tance of the early recognition and treatment of infection, and the importance of
maintaining adequate hydration. Screening for HIV, and hepatitis A, B and C
should be carried out after suitable counseling. Routine urinalysis may assist in
the detection of infection. As with other higher risk pregnancies, there should be
regular monitoring of blood pressure and fetal development.
Randomized studies have shown no overall benefit in prophylactic blood trans-
fusions in pregnant women, although there may be a reduction in the frequency
of vaso-occlusive events when prophylactic transfusion has been used. Transfusion
should be considered, however, when there is an acute anemia (hemoglobin
level<5g/dl), pre-eclampsia, septicemia, acute renal failure, acute chest syndrome,
a recent episode of CIAO and when preparing for surgery. As noted above, mul-
tiple pregnancy will require assessment for transfusion on a more regular basis,
with the aim of maintaining HbS at <30% of circulating hemoglobin.
The management of an acute sickle-cell crisis in pregnancy should include
hydration with intravenous fluids, pain control and the treatment of any associ-
ated infection. Both the mother and fetus may require oxygen therapy and the
cardiotocography (CTG) may be non-reactive during a maternal crisis.
The timing of transfusion and the timing of delivery may be important factors
in limiting maternal and fetal mortality, especially if there is evidence of placen-
tal insufficiency, or if there is a limitation of maternal cardiac, renal or hepatic
reserve. Given the loss of fluids during delivery, close attention should be pro-
vided to maternal hydration, adequate pain control and oxygenation. After deliv-
ery, early mobilization and risk assessment for venous thrombosis and graduated
elastic compression stockings and low-molecular-weight heparin (LMWH),
should be used to prevent venous thromboembolism (VTE).
Although hydroxycarbamide (hydroxyurea) may ameliorate symptoms, by eleva-
tion of HbF levels and reduction of sickling episodes, it crosses the placenta and
may have a teratogenic effect on the human fetus. In animal studies it has been
associated with defects in the central nervous system, palate and skeleton, as well
as cardiovascular and ocular abnormalities. It is also possible that it may induce
myelosuppression in the fetus in the second and third trimesters. At present
there is little information on the safety of this drug when those maintained on it
prior to pregnancy conceive, although a number of normal pregnancy outcomes
have been reported. However, as there is no evidence that it is beneficial to the
mother or the fetus, as with the myeloproliferative disorders, it is probably best
avoided, especially in the first trimester.
The principles of anesthesia in pregnant patients with major sickling disorders
is similar to non-pregnant subjects, with the need to avoid dehydration and
hypothermia, and fluid replacement should be supplied through a fluid warmer.
182
Regional anesthesia, both spinal and epidural, appear safe for operative and non-
operative delivery, although epidural is often preferred. There must be, however,
adequate hydration to avoid any hypotension occurring secondary to sympathetic
blockade, but such blockade could, in theory, increase blood flow and reduce
the risk of vaso-occlusion. As noted above, there is a need to consider thrombo-
prophylaxis against VTE in such patients.
After delivery, there is a need to carefully attend to pain relief and hydration
to counteract the effects of increased metabolism and diuresis, and so avoid sick-
ling. In such circumstances, an acute chest crisis may be misdiagnosed as infec-
tion, fluid overload or a reaction to anesthesia.
Despite the improvements in feto-maternal outcomes in recent decades,
mothers with a major sickling disorder are still at risk of renal, cardiovascular and
cerebrovascular complications during pregnancy. In poorly developed countries
a number of factors are important in the success of pregnancies. These include:
adequate maternal nutrition; education on the nature of sickle-cell disorders; the
early recognition and treatment of urinary and respiratory infections; the mainte-
nance of adequate hydration and urinary output; the regular use of antimalarial
prophylaxis; systematic supplementation with multivitamins, including folate and
iron (depending on iron status); and the use of transfusion when the symptoms
of anemia are intolerable or appear likely to lead to cardiac failure. In these pop-
ulations, methods such as these may reduce the maternal and fetal mortality to a
level comparable with local non-sickle-cell patients and with sickle patients in
developed countries.
Other hemoglobinopathies
The combination of sickle-cell with HbC (HbSC) results in a lesser sickling dis-
order, although it carries a higher risk of retinopathy. The diagnosis can be sus-
pected by the presence of a large number of target cells and sickle cells on the
blood film and the presence of HbC on electrophoresis. However, both HbC and
HbE disease (without S) are associated with a relatively mild anemia and
splenomegaly, and often no specific treatment is required. Indeed, aside from
mild anemia, which may be exacerbated by iron or folic acid deficiency, both
HbCC and C/β0 syndromes are likely to be associated with a normal pregnancy
outcome.
Another Hb variant is HbD, which migrates to the same position as HbS on gel
electrophoresis. In homozygous DD there is a mild hemolytic anemia, and when it
occurs as a compound heterozygote with HbS it can result in a sickling disorder.
Heinz body anemias

A small deletion, or amino acid substitution, in the globin gene, affecting the
protein structure of the globin molecule around the heme pocket, can result in a
number of consequences. These include:
183
Instability of the hemoglobin; or

Easier oxidation of the heme moiety; or
Hemoglobin precipitation (with the formation of so-called Heinz bodies
within the red cell); or
An altered oxygen affinity of the hemoglobin.
These disorders are associated with a lifelong tendency to hemolysis, which is
exacerbated by infection or exposure to oxidative substances. The diagnosis can be
made by finding a combination of hemolysis, hemoglobin instability on exposure to
heat or isopropanol, exclusion of a more common hemoglobinopathy, and, in some
cases, alteration in the oxygen dissociation (affinity) of the hemoglobin molecule.
Such mutations are not always associated with an abnormal electrophoretic
pattern and may not always be associated with a family history. When such muta-
tions increase the oxygen affinity, they may come to light in the investigation of
erythrocytosis, and when they reduce the oxygen affinity they may result in the
investigation of congenital cyanosis. There is little information on their impact in
pregnancy.
Red cell membrane defects and enzymopathies

A variety of disorders of the red cell membrane or intracellular red cell enzymes
may result in congenital hemolysis. The majority, such as hereditary spherocyto-
sis, most often result in a mild hemolytic disorder which comes to light in child-
hood during an infection. When severe, these disorders are treated by
splenectomy. As they are associated with chronic hemolysis, pregnant mothers
are at increased risk of folic acid deficiency, leg ulcers and gallstones.
There are also a number of disorders of the red cell enzyme pathways which
can result in:
A reduction in adenosine triphosphate (ATP: a principal energy source for
the cell); or
A reduction in the reducing agents NADH or NADPH (the major antioxidant
potential of the red cell); or
A reduction in available 2,3-DPG, which may alter the hemoglobin’s oxygen
affinity.
A deficiency in the reducing agents NADH and NADPH will result in red cells
which are at increased risk of hemolysis under oxidative stress. The commonest
cause of this is a deficiency of the enzyme, glucose-6-phosphate dehydrogenase
deficiency (G6PDH). The gene coding for G6PDH is found on the X chromo-
some, and although most common in the tropics, it is found in many other areas.
The deficiency results in a lifelong tendency to hemolysis, often triggered by
oxidative foods (classically the fava bean) and oxidative drugs (such as antimalar-
ials and some vitamin K derivatives). Although it most often affects male carriers,
it can also affect female carriers if there is excessive lyonization in the normal
184
gene (see Chapter 4). Patients are most often well between the episodes of severe
hemolysis. G6PDH deficiency is also associated with neonatal jaundice in an
affected male. The mechanism for this is unclear, but the management is similar
to other causes of neonatal jaundice, and includes phototherapy and, where
necessary, exchange transfusion. There are a variety of methods for detection of
G6PDH levels in the red cell that can be used to distinguish this disorder from
other forms of Heinz-body-forming hemolysis.
The porphyrias
The porphyrias result from a number of inherited defects in the enzymes which
are required for the synthesis of heme. Although there are a wide variety of mole-
cular defects, seven clinical types of porphyria can be distinguished. These can be
considered by whether the defect results in the accumulation of porphyrins
(whose fluorescent properties result in the generation of free radicals and a
marked, often scarring, photosensitive skin rash), porphyrin precursors (which
result in neurological manifestations), or both. In some instances, the enzyme
defects are inducible, which can lead to an acute disorder. The exact diagnosis is
determined by the combination of detectable porphyrins and precursors in
plasma, erythrocytes, urine and stool. The effect on hemoglobin production may
be slight, as many of implicated enzymes are not rate limiting in heme biosynthe-
sis, although there may be a mild hemolytic anemia.
Table 9.14 – Porphyria in pregnancy
Porphyria cutanea tarda (PCT) First trimester has highest risk of symptoms
There may be glucose intolerance
There may be associated hepatitis C or B
Avoid alcohol
Avoid sun
No routine iron
If symptoms, venesect
If refractory symptoms, consider chloroquine
Acute intermittent porphyria (AIP) Avoid drugs/precipitants
If symptoms:
Remove precipitant
Hydrate patient
Carbohydrate supplements
Hematin
Pain control
Seizure treatment
185
Of the seven types, congenital erythropoietic porphyria (CEP), porphyria

cutanea tarda (PCT), erythropoietic protoporphyria (EPP) and acute intermittent
porphyria (AIP) have been reported to be influenced by pregnancy, or to affect
pregnancy management (Table 9.14) The prevalence in European populations is
estimated at around 1 in 1000 000 for CEP, 1 in 25 000–70 000 for PCT, 1 in
75 000–200 000 for EPP and 1–2 in 100 000 for clinically overt AIP. The actual
prevalence of AIP is, however, likely to be around 10 times higher, as 90% of sub-
jects have latent disease. Although a number of ‘bedside’ tests are available for por-
phyrias, their investigation is best carried out in a center with expertise in this area.
CEP
In CEP there is often severe, cutaneous, scarring, photosensitivity that is evident
from childhood. This is accompanied by a normochromic, normocytic anemia,
with basophilic stippling evident in the red cell and nucleated erythrocytes a
common feature.
Standard suncreams are often ineffective, as they do not screen out the wave-
lengths of around 400nm that lead to the photosensitivity. Reflective creams,
with zinc or titanium, may be effective. Other strategies such as activated char-
coal (binding bile porphyrin) and hypertransfusion (to suppress endogenous
erythropoiesis) have been used successfully in some cases. If anemia is a problem,
it may respond to splenectomy.
One of the CEP mutations is associated with severe disease, which can result
in non-immune hydrops fetalis, or transfusion-dependant anemia from birth. In
many cases, prenatal diagnosis from amniotic cells is possible.
PCT
PCT results from a reduction in activity of one of the heme synthesis enzymes in
the liver. This leads to an accumulation of uropophyrins in the plasma, which are
excreted in the urine. PCT results in marked skin fragility in light-exposed areas,
facial hypertrichosis and hyperpigmentation. It is not, however, associated with
neuropsychiatric disturbance. It presents late, often in association with alcohol
excess, iron overload and/or hepatitis B or C carriage.
Exogenous estrogen is associated with expression of PCT and it has also been
diagnosed for the first time in pregnancy, although there is debate as to whether
pregnancy exacerbates the disease or not. If exacerbation does occur, this is
likely to be in the first trimester, as there is a reduction in urine porphyrin and
placenta-derived estrogen and progestogen production in the second trimester.
In pregnancy, it would be sensible to monitor known cases (as there may be a
greater risk of glucose intolerance), to minimize exposure to alcohol and sun-
light, and to refrain from iron supplementation. In non-pregnant subjects, avoid-
ance of alcohol, removal of iron by venesection, recombinant erythopoietin and
treatment of hepatitis C, if present, with α-interferon have all been reported as
186
beneficial. In symptomatic and refractory cases in pregnancy, venesection and

chloroquine have both been used, although chloroquine is known to be terato-
genic in animals.
EPP
EPP results from a defect in ferrochelatase, which is the last enzyme in the heme
pathway. The level of the enzyme does not, however, correlate well with symp-
toms. EPP most often becomes evident in childhood, with exposure to sunlight
leading to itching and burning, and local erythema. It can also present in the dif-
ferential diagnosis of a sideroblastic anemia and, in later life, may lead to gall-
stones due to precipitated protoporphyrin. Rarely, it may manifest with a
neurological crisis resembling AIP.
EPP is diagnosed by the detection of elevated levels of erythrocyte-derived free
protophorphyrin. Although elevated levels also occur in iron deficiency and lead
poisoning, the levels in EPP are usually considerably higher than those in iron
deficiency (normal <50µg/dl, iron deficiency anemia (IDA) <300µg/dl, EPP and
lead poisoning 300–4500µg/dl), but unlike lead poisoning and iron deficiency,
the protophorphyrin of EPP is non-chelated.
Treatments include avoidance of sun exposure, zinc- or titanium-based sun-
screens, induction of hypercarotenemia (which can reduce photosensitivity),
hematin therapy (which reduces the production of porphyrins and their precur-
sors), cholestyramine (which binds bile protoporphyrins) and activated charcoal.
It is possible that there may be an improvement in symptoms in pregnancy,
although this requires confirmation and the mechanism of such an improvement
is unknown.
AIP
AIP results from a decreased conversion of porphobilinogen to prophyrins. It is
transmitted as an autosomal dominant condition, with the majority of cases due
to single base mutations. The mechanism of neurotoxicity is unknown, but could
be due to a rise in the porphyrin precursor 5-aminolevulinic acid (which may
inhibit neurotransmitters), or a depletion of cellular heme (leading to reduction
in energy production in neural tissues).
The condition is characterized by the acute onset of autonomic neuropathy,
which commonly presents with abdominal pain, constipation, tachycardia, hyper-
tension and vomiting. This may be accompanied by acute anxiety, confusion, psy-
chosis and seizures. In 50% of subjects, persisting hypertension and renal
impairment may also occur. The blood count is usually normal aside from a
raised white cell count in the acute phase. Both 5-aminolevulinic acid and por-
phobilinogen can be detected in the urine, especially during an acute attack.
A number of circumstances can lead to an acute attack. These include
gonadal hormones (particularly progestogens), decreased caloric intake, alcohol
187
and other drugs (usually by induction of cytochrome p450 leading to an

increased need for intracellular heme), such as barbiturates, valproate, nifedip-
ine and diclofenac. It is impossible to predict whether a drug will or will not
produce a reaction, but drugs such as opiates, corticosteroids, vitamin C, aspirin,
insulin, nitrous oxide and labetolol are generally considered to be safe. A fuller
list of drugs thought to be suitable or unsuitable is to be found in the British
National Formulary (www.bnf.org).
The management of an acute event requires removal of any precipitating
factors, appropriate hydration and carbohydrate supplementation (by nasogas-
tric feeding). If attacks do not improve within 24 hours, therapy with intravenous
hematin has been suggested. The dose is of the order of 3mg/kg, once daily, for
4 days. As too rapid an infusion may cause irritation at the injection site, it should
be given through central venous access. In addition, too high a dose may result
in renal failure. After treatment, the symptoms of AIP usually improve within 48
hours. The treatment of seizures can be problematic, as many antiepileptic med-
ications are also associated with AIP. Possible options include correction of any
hyponatremia and treatment with clonazepam or magnesium sulfate.
Reports of AIP in pregnancy are not common, with an attack rate, in those
known to have porphyria, of 16% in the antenatal and 8% in the postnatal
period. Although estrogen is capable of inducing heme synthesis, other features
of pregnancy, such as hyperemesis (leading to calorie loss) or metoclopramide
therapy, may precipitate an attack. Initial reports suggested a dire outcome for
those with AIP, but more recent work suggests that, with adequate maternal care,
a good outcome is possible. A recent study has suggested that miscarriage may be
more common in symptomatic AIP patients when compared with those with
latent disease, but this observation requires confirmation. In this study, around
20% of subjects reported an improvement, around 10% a decline and around
70% no change in symptoms during pregnancy. The risk of an acute episode,
however, cannot be predicted in a primigravid subject, and AIP may be confused
with symptoms of hyperemesis, eclampsia and even (when there is a progressive
muscle weakness) Guillain-Barré syndrome in pregnancy.
The management of an acute episode in pregnancy follows the same princi-
ples outlined above. Although caution is recommended with hematin (haem
arginate) in pregnancy, it has been used with success in pregnant subjects. When
the diagnosis is made for the first time, it should be remembered that other
family members should be screened for the disorder.
188
Obstetric Index 19/10/05 10:19 am Page 189
Index
ABO antigens 23 diagnosis in pregnancy to FVIII 66
ABO blood system 15 169–70 in hemolysis 23, 24
activated clotting factor fetal carriers 168 HIT 95
concentrates 78 diagnosis 170 warm 55, 56, 57
activated partial fetal HbH disease 168 see also autoantibodies
thromboplastin time genetics 166–7 anticoagulation, therapeutic
(APTT) 14 maternal carriers 167 61
and activated protein C maternal HbH disease long-term oral 107
resistance 134 167–8, 170–1 new anticoagulants 100–1
in afibrinogenemia 84 mechanism 165 and prosthetic heart valves
in dysfibrinogenemia 86 pregnancy, problems in 166–7 fetal outcomes 106
in fluid replacement therapy severity 166–7 maternal outcomes 105
42 treatment 170–1 see also specific agents
in hemophilia 77, 78 amniocentesis antiphospholipid syndrome
in pregnancy 10 bilirubin level determination 99, 137–9
prolongation in hemophilia A 34 definition 137
67 in hemophilia diagnosis 71 management options 138
in von Willebrand’s disease anagrelide 59 and pregnancy outcome
81 anamnestic response 16, 23 138–9
activated protein C (aPC) 6 anemia 43–8 antiplatelet agents 60, 153
therapy 159 autoimmune hemolytic 20–1 antithrombin (AT) 6, 8, 10
activated protein C resistance dilutional 1 and fetal loss 133
(aPCR) 13 fetal loss 34 in pregnancy 10
in pregnancy 10, 11, 134–5 fetal assessment 34 and thrombophilia
outcome 135 folate deficiency 51 assessments 126
and thrombophilia Heinz body 183–4 antithrombin deficiency 124,
assessments 126 iron deficiency 43–8 127–8
acute chest syndrome 179 diagnosis 45–7 antithrombotic therapy
acute fatty liver of pregnancy in pregnancy 44–8 88–101
(AFLP) 155–6 prevention 47 antitrypsin 6, 8, 10
diagnosis, differential 142 treatment 47–8 aplasia, acute 51
mortality 155 macrocytic 50–1 argatroban 98–9
treatment/outcome 147 megaloblastic 51, 54 arterial thrombosis 102–4
acute intermittent porphyria pernicious 54 in pregnancy 102–4
(AIP) 187–8 physiological 1 risk in young women 102
in pregnancy 185, 188 in sickle-cell disease 178, 179 Ashkenazi Jews 76, 132–3
acute myocardial infarction anesthesia in sickle-cell disease aspirin 99–100
108–11 182–3 in acute myocardial infarction
diagnosis/management angiotensin-converting enzyme 110
110–11 (ACE) inhibitors 110 in CIAO 108
afibrinogenemia 84 anti human leukocyte antigen and heparins 99
age and VTE risk 128 23, 24 in TTP/HUS 153
AIHA see autoimmune anti-RhD immunoglobulin autoadsorption 21
hemolytic anemia (AIHA) immunoprophylaxis 36–7, autoantibodies 20, 55
allergic reaction 144 removal 21
to heparins 97 antibodies autoimmune hemolytic anemia
to transfusion 22, 25 antiplatelet 143 (AIHA) 20–1, 55–8
alloantibodies 20–1 cold 56, 57 causes 57
alpha thalassemia formation 16 in pregnancy 56–7
Index
treatment 57–8 blood volume in pregnancy 1 and thrombophilia

autoimmune neutropenia Bombay phenotype 15 assessments 126
(AIN) 149 and von Willebrand factor
autologous donation 26 C1-esterase inhibitor 6, 8, 10 78
pre-deposit 27, 28–9 cardiac output in pregnancy in von Willebrand’s
azathioprine 58, 146 102 disease 81–2
cardiotocography (CTG) 34 FIX 4, 6, 10
bacterial contamination in catheter-directed thrombolysis FX 4, 6, 8, 10, 67
transfusion 22, 27 121 FXI 4, 6, 8, 10, 68, 75–7
Bernard Soulier syndrome 81 cell salvage, intraoperative FXII 4, 6, 8, 10
beta thalassemia 28–9 FXIII 4, 6, 84–6
diagnosis in pregnancy cell saving 28–30 coagulation 4–6
fetal 174 autologous pre-deposit 28–9 control 6–8
maternal 173–4 blood substitutes 30 pregnancy-associated
features 171–2 erythropoietin 29–30 changes 10, 11–12
iron overload 172 intraoperative cell salvage 29 failure in placental abruption
mechanism 170–1 management 108 41
pregnancy, problems in 172–3 central venous pressure pathways
treatment 174–5 monitoring 41 extrinsic 4, 6
bilirubin cerebral ischemia of arterial intrinsic 4
in AIHA 56 origin (CIAO) 19, 107–8, and placental separation 110
in HND 180 pregnancy-associated
effects on fetus 34–5 management 108, 109 changes 10–12
level determination 34 risk 107 coagulation cascade 4
bioprosthetic heart valves 104 Cesarean section colloid therapy 42
bleeding disorders 66–87 cell saving 28 congenital dysfibrinogenemia
bleeding risk with heparin 96–7 platelet transfusion 18 85
bleeding time 81, 86 chelating agents 174 congenital erythropoietic
blood banking, placental 64–5 childhood porphyria (CEP) 186
blood cell salvage see cell development/behavior and Coombs’ test see direct
salvage, intraoperative iron 45 antiglobulin (Coombs’) test
blood cell saving see cell saving chorionic villus sampling cord blood sampling 34
blood component therapy (CVS) 71 cord stem cells 64
16–20 CIAO see cerebral ischemia of cordocentesis 71
blood count 1–2 arterial origin (CIAO) coronary dissection 110
in pregnancy 2 clotting factors 16 corticosteroids
blood group screening, activated concentrates 78 in AIHA 57
antenatal 32 FIX in ITP 144–5
blood group systems 15 elevated 136–7 and neutrophilia 2
blood loss in hemophilia 68, 74, 75 coumarins 88–98
management 41–2 recombinant 74 and heparin 91–8
in normal vaginal delivery FV 4, 6, 8, 10 and prosthetic heart valves
39 FVII 4, 6, 8, 10 105–6
see also hemorrhage, massive and warfarin therapy 89 warfarin 88–91
obstetric see also recombinant factor Creutzfeldt–Jakob disease, new
blood transfusion 15–16 VIIa (rFVIIa) variant (nCJD) 27–8
allergic reaction to 22, 25 FVIII 4, 6, 10, 14 cryoprecipitate treatment 16,
bacterial contamination in acquired inhibitors 68 17–18
transfusion 22, 27 deficiency 66 cryosupernatant 152
‘group-and-save’ procedure and desmopressin 69 crystalloid therapy 42
17, 72 elevated 136–7 cytoreduction 59
in thalassemia 174 in hemophilia 67, 68, 70,
in warm autoimmune 73, 78 D-dimer 3, 8
hemolytic porcine 78 and DVT 112
anemia 20–1 in pregnancy 10 in pregnancy 10
see also intrauterine replacement 68 in puerperium 13
transfusion therapy 68–9 dalteparin 96, 116
190
Index
danaparoid 97–8 in VTE 116 absorption/excretion 48–9

DDAVP see desmopressin erythroid hyperplasia 55 malabsorption 49
analog (DDAVP) erythropoiesis 44 in pregnancy 2, 52
deep vein thrombosis (DVT) erythropoietic protoporphyria sources 48
111 (EPP) 187 supplementation 58, 153,
diagnosis 112–13 in pregnancy 186 170–1, 175, 182
warfarin therapy 89 erythropoietin 29–30 folic acid deficiency 48–53
delivery 13 essential thrombocythemia causes 49
blood loss in 39 (ET) 58–60 in pregnancy 51–3
in hemophilia 73 in pregnancy 59–60 diagnosis 51
in ITP 148–9 estrogen and coagulation 4 prophylaxis 52–3
protein S levels during 13 ethnicity prevalence 51
in von Willebrand’s disease and factor XIII deficiency 85 symptoms 50
83–4 hemolytic disease of newborn treatment 53
desferrioxamine 174 31, 33 fondaparinux 100
desmopressin analog (DDAVP) Evan’s syndrome see immune free fetal DNA 34
18, 69–70, 81–2, 83 thrombocytopenia (Evan’s fresh frozen plasma (FFP) 16,
in breastfeeding mothers 73 syndrome) 17
in Ehlers–Danlos syndrome in hemophilia 76
87 factors see clotting factors in TTP/HUS 152
in hemophilia 72 ferritin 43 FV Leiden mutation 124, 128–9
in hemophilia A 69 in iron deficiency detection and activated protein C
dextrans 100 46 resistance 134
contraindication 42 in pregnancy 2, 3 and fetal loss 133
iron dextran 48 fetal cell isolation 71 and IUGR 133
diabetes and pre-eclampsia fetal heart rate patterns 34 and pre-eclampsia 132
132 fetal loss 4
dilute Russell viper venom test in placental abruption 40 G6PDH deficiency 184–5
(dRVVT) 137 RhD-sensitization 32 gestational thrombocytopenia
direct antiglobulin (Coombs’) and thrombophilia 133–4 2, 140–3
test 20, 56 in TTP 151 diagnosis 141–2
direct thrombin inhibitors fetal platelet count 148 differential 142
98–9 fetal programming 45 treatment/outcome 147
disseminated intravascular feto-maternal hemorrhage glucose-6-phosphate
coagulation (DIC) 42, assessment 32–3, 36 dehydrogenase (G6PDH)
158–9 ‘silent’ 36–7 deficiency 184–5
causes 158 fibrin 10, 11 graft versus host disease
diagnosis, differential 142 in pregnancy 10 stem cell transplantation 176
treatment/outcome 147 in puerperium 13 transfusion-related (TA-
Duffy antibodies 38 fibrinogen GVHD) 22, 26
DVT see deep vein thrombosis in coagulation cascade 6 granulocyte colony-stimulating
(DVT) in DIC 159 factor (G-CSF) 149
DY fragments 8 elevated 136–7
dysfibrinogenemia 85 in fluid replacement therapy Ham’s acid lysis test 61
42 Hb Barts hydrops fetalis 166,
eclampsia 130 levels during birth 13 167, 168–9
Ehlers–Danlos syndrome 86 in pregnancy 11 delivered fetus 169
elastic stockings 116, 118–19 synthesis 84 diagnosis 170
embolectomy 120–1 fibrinogen treatment 16, 17–18 maternal effects 167
embryopathy 90–1 fibrinolysis 8–10 HbH disease
endothelial protein C receptor pregnancy-associated features 168
(EPCR) 8 changes 12–13 fetal 168
enoxaparin 116 fibrinopeptide A 10 maternal 167–8
enzymopathies 184–5 fluid replacement therapy Heinz body anemias 183–4
epidural analgesia 83 152–3 differential diagnosis 185
in PNH 62 folate deficiency anemia 51, 53 Helicobacter pylori eradication
in sickle-cell disease 183 folic acid 143, 144
191
Index
HELLP syndrome see fetal presentation 35 control, normal 4–10

hemolysis, elevated liver prevention 36–7 defects 81
enzymes and low platelets sampling regimen 34–5 and von Willebrand factor 81
(HELLP) syndrome hemolytic uremic syndrome heparin-induced
hemarthrosis 67, 76 (HUS) 17, 149–53 thrombocytopenia 93–5,
hematinics diagnosis 150–1 156–7
assessment 3 differential 142, 151 diagnosis 94–5
deficiency 43 investigation 151 incidence 93–4
in pregnancy 2 refractory 153 management 95
sources 52 treatment 152–3 argatroban 99
hemato-oncology 58–64 additional 153 danaparoid 97–8
hematocrit in pregnancy 1 firstline 152–3 lepirudin 98
heme disorders see porphyrias secondline 153 onset 157
hemodilution 1, 2 treatment/outcome 147 Type 1 (HIT) 93
hemoglobin 163 hemophilia A 66–74 Type 2 (HIT thrombosis
development 163 antepartum management of syndrome)(HITTS) 93
elevation with erythropoietin carriers 72 UHF vs LMWH 93–4
29–30 carrier detection in heparins 60, 91–8
Hb Barts 166 pregnancy 70–1 antithrombotic efficacy 97
in HND 34–5 delivery/puerperium fetal outcome 97
identification in thalassemia management 73–4 kinetics 97
172 diagnosis 67 route of administration 92
iron content 43 factor VIII dosing 69 UFH vs LMWH in pregnancy
in iron deficiency 46 joint bleeding 67 95–7
in pregnancy 1, 2 onset 67 vs aspirin 99
hemoglobin-based blood prenatal/pre-implantational hepatic failure 156
substitutes 30 diagnosis 71–2 hepatitis vaccination 72
hemoglobinopathy 163–84 presentation of females 66 hereditary persistence of HbF
sickle-cell disease 177–83 presentation/treatment of 172
stem cell transplantation affected males 67–8 hirudin 98
176–7 risk of thrombosis 69–70 HLA antibodies 25
thalassemias 164–76 treatment 68–70 HLA haplotypes 26
alpha thalassemia 165–71 antifibrinolytic agents 70 HND see hemolytic disease of
beta thalassemia 171–5 desmopressin analog newborn (HND)
thalassemia intermedia (DDAVP) 69–70 Hodgkin’s disease 63–4
175–6 factor replacement 68–9 homocysteine (HCY)
types 164 general principles 68 and folate 52
hemolysis hemophilia, acquired 77–8 in folate deficiency 49, 50
intravascular 55 management 79 in pregnancy 2, 3
lifelong tendency 184 hemophilia B 74–5 and vitamin B12 deficiency
hemolysis, elevated liver diagnosis 67 53
enzymes and low platelets factor VIII dosing 69 see also
(HELLP) syndrome 130, hemophilia B Leyden 74 hyperhomocysteinemia
154–5 hemophilia C 75–7 homocysteine– methionine
diagnosis, differential 151 diagnosis 67 pathway 49, 50
and HUS 150 hemorrhage HPA antibodies 26
treatment/outcome 147 delayed 76 parental mismatch 160–1
hemolytic disease of newborn effects on fetus 28–9 HUS see hemolytic uremic
(HND) 30–8 massive obstetric 39–42 syndrome (HUS)
antenatal diagnosis/ definitions 39 hydroxycarbamide
monitoring 32–5 management 41–2 (hydroxyurea) 59, 60, 182
antibodies mortality 40 hyperhomocysteinemia 52,
detection, timing of 33 placental abruption 40–1 103, 135–6
non-RhD 37–8 risk factors 39 and thrombophilia
RhD 31–7 postpartum 39, 76, 83 assessments 126
ethnicity 31, 33 hemosiderin 43 hypertension 130
fetal management 35–6 hemostasis 3–4 and cell saving 30
192
Index
in pre-eclampsia 129 diagnosis 45–7 MAHA see microangiopathic

hyperuricemia 130 in pregnancy 44–8 hemolytic anemia (MAHA)
hypochromia 169, 170 prevention 47 massive obstetric hemorrhage
hypofibrinogenemia 85, 155 treatment 47–8 see hemorrhage, massive
iron deficiency anemia (IDA), obstetric
idraparinux 100–1 severe 46, 48 mean corpuscular volume
imaging, radiation risk in 64 iron overload 172 (MCV)
immune (idiopathic) iron therapy in iron deficiency 46
thrombocytopenic purpura parenteral 48 in pregnancy 1, 2
(ITP) 143–9 side effects 47, 48 megaloblastic anemia 51, 54
associated disorders 143 supplements 47–8 melagatran 101
delivery management 148–9 IUGR see intrauterine growth methotrexate 63
diagnosis restriction (IUGR) methylprednisolone 146
differential 142 microangiopathic hemolytic
in pregnancy 144–5 Jehovah’s Witnesses 29 anemia (MAHA) 158
treatment 145–8 joint bleeding 67 in gestational
experimental therapies 144 thrombocytopenia 141
firstline 145–6 kallikrein 8 in PET/HELLP syndrome
outcomes 147 Kell antibodies 37–8 130
secondline 146, 148 maternal management 37 in TTP/HUS 149, 150, 151
immune thrombocytopenia Kidd antibodies 38 microcytosis 169, 170
(Evan’s syndrome) 57 miscarriage
immunoglobulin antibodies 15 lactation and iron deficiency in dysfibrinogenemia 86
immunoglobulin 45 and warfarin therapy 90
immunoprophylaxis 36–7 Leiden mutation 11 mitral valve prolapse 87
immunoglobulin therapy 19 lepirudin 98 MNS antibodies 38
in AIHA 57–8 leukemia, acute 62–3 MTHFR C677T see
in ITP 144–5 leukocytosis 2 thermolabile methylene
in NAIT 161–2 leukodepletion of blood tetrahydrofolate reductase
immunoglobulins components 25, 174 (MTHFR C677T)
in hemolysis 23 Lewis blood system 15 myocardial infarction, acute
in hemolytic disease of linkage analysis of see acute myocardial
newborn 32 polymorphisms 70–1 infarction
infection control, prophylactic LMWH see low-molecular myoglobin 43
180 weight heparin (LMWH)
α-interferon 60 low-molecular weight heparin neonatal alloimmune
international normalized ratio (LMWH) 61, 91, 92 thrombocytopenia (NAIT)
(INR) 89 in acute myocardial 160–2
intracerebral hemorrhage infarction 111 diagnosis 160–1
73–4, 143 bleeding risk 96–7 treatment
intrauterine growth restriction in CIAO 108 first pregnancy 161
(IUGR) 133 dosing 14 second pregnancy 161–2
in sickle-cell disease 179 and osteoporosis 96 neutrophilia 2
and thrombophilia 132–3 and prosthetic heart valves non-Hodgkin’s lymphoma 64
in TTP 151 105, 106, 107
intrauterine transfusion 35 vs UFH in pregnancy 95–7 osteoporosis 95–6
intravascular hemolysis 55 in VTE 119–20 ovarian failure 64
Iraqi Jews 76 dosing 116 oxidative stress 184
iron prophylaxis 117, 118
absorbtion/transport 44 lung injury, transfusion-related paroxysmal nocturnal
distribution 43–4 acute (TRALI) 25 hemoglobulinuria (PNH)
fetal acquisition 45 lupus inhibitor 126, 137, 138 60–2
functions 44 lymphoma 63–4 in pregnancy 61–2
malabsorption 48 lyonization 66, 184 partial thromboplastin time
in pregnancy 2, 3, 52 (PTT) 13
iron deficiency 43–8, 51 macrocytosis 50–1 perfluorocarbon emulsions 30
causes 46, 48 α2-macroglobulin 8, 10 pernicious anemia (PA) 54
193
Index
phenotyping, extended 21 preventing 148 protoporphyrin 3, 44

phototherapy 36 post-transfusion purpura PTE see pulmonary
placental abruption 40–1 (PTP) 22, 26, 157 thromboembolism (PTE)
placental blood banking 64–5 pre-eclampsia (PET) 4, 154–5 puerperium, early 13
placentomegaly, relative 45 and activated protein C pulmonary thromboembolism
plasma exchange 18, 26, 36 resistance 135 (PTE) 111, 113
in TTP/HUS 152, 153 in alpha thalassemia 171 diagnosis 113–14
plasma volume, maternal, in and aspirin 99 radiation risk 64, 113
pregnancy 1 definition 129
plasminogen 8 diagnosis, differential 142, recombinant factor IX 74
activation 8, 10 151 recombinant factor VIIa
in pregnancy 11 and hyperhomocysteinemia (rFVIIa) 16, 20
in puerperium 13 136 in hemophilia 74, 78
plasminogen activator placental pathology 131 red cell antibodies 15–16
inhibitor (PAI) 8, 10, 11 risk factors 130–1 red cell destruction 55
in puerperium 13 in sickle-cell disease 178 red cell disorders, hereditary
platelets and thrombin 123 163–88
activation 2 and thrombophilia 129–32 see also specific disorders
in coagulation 4–6 treatment/outcome 147 red cell mass in pregnancy 1
platelet antigens 26 prednisolone 57 red cell membrane defects
platelet count pregnancy loss see fetal loss 184–5
fetal 148 prion diseases 27 disorders caused 184
in fluid replacement therapy prosthetic heart valves in red cell transfusion 16–17
42 pregnancy 104–7 restriction length fragment
in NAIT 161 protamine 92 polymorphism (RFLP) 83
in pregnancy 2 protein binding in warfarin 89 in hemophilia diagnosis 71,
platelet function disorders 18 protein C (PC) 6, 8, 123 75
platelet-pheresis 60 activation 11 reticulocytes 55–6
platelet transfusion 16, 18–19 in pregnancy 10 retroplacental hemorrhage see
dosing formulas 19 and thrombophilia placental abruption
in NAIT 161–2 assessments 126 Rh haplotype 30–1
non-hemolytic reactions 24 protein C (PC) deficiency 124, RhD alloimmunization 31
in PET/HELLP syndrome 127 assessment of severity 34
155 and fetal loss 133 platelet transfusion 18–19
PNH see paroxysmal nocturnal protein S 8, 123 sensitizing events 32, 36
hemoglobulinuria (PNH) levels during delivery 13 RhD group screening 32–3
polyglutamated folate 49 in pregnancy 10 RhD status in massive
polymorphism linkage analysis and thrombophilia transfusion 42
70–1 assessments 126 rheumatic disease 104
porphyria cutanea tarda (PCT) and FFP 17 ristocetin 81
186–7 and warfarin therapy 89 ristocetin cofactor activity
in pregnancy 185 protein S deficiency 123–4 10, 81
porphyrias 185–8 and fetal loss 133
acute intermittent (AIP) proteinuria 130 sickle-cell crises 177, 178
187–8 prothrombin 6, 11 pain control 180
in pregnancy 185, 188 prothrombin fragment 1+2 sickle-cell disease 16, 177–83
classification 185 (F1+2) 6, 10 causes of anemia 179
congenital erythropoietic in pregnancy 10 diagnosis 179–80
(CEP) 186 prothrombin G20210A features 177–8
erythropoietic mutation 124, 128–9 maternal carriers 178
protoporphyria and fetal loss 133 in pregnancy 178–9
(EPP) 187 and IUGR 133 treatment
in pregnancy 186 and pre-eclampsia 132 general management
porphyria cutanea tarda prothrombin time (PT) 13, 14 180–1
(PCT) 186–7 in afibrinogenemia 84 pregnancy-specific 181–3
in pregnancy 185 in dysfibrinogenemia 85, 86 sickling disorder 183
post-splenectomy infection, and warfarin therapy 89 splenectomy 58, 146, 148
194
Index
post-splenectomy infection pregnancy 10 tetrahydrofolate reductase

148 in pregnancy 4, 11 (MTHFR C677T)
stem cell transplantation 176–7 in puerperium 13 128–9, 132
stem cells 64 thrombin activatable see also dysfibrinogenemia
streptokinase 110, 121 fibrinolysis inhibitor (TAFI) and pre-eclampsia 129–32
synthetic pentasaccharides 8, 10 and pregnancy complications
100–1 thrombin– antithrombin 122
systemic lupus erythematosus (TAT) complex 8, 10 and pregnancy outcome
(SLE) 137, 139 thrombin burst 6, 20 122–39
thrombin clotting time (TCT) thrombophilia assessments,
termination for leukemia 84, 85 factors influencing 126
patients 62 thrombin inhibitors, direct thromboplastin see tissue
thalassemia(s) 164–76 98–9 factor (TF)
alpha thalassemia 165–71 thrombin– thrombomodulin (thromboplastin)
diagnosis in pregnancy complex 8 thromboprophylaxis 61
169–70 thrombocytopenia 140–62 thrombotic thrombocytopenic
fetal carriers 168 diagnosis, differential 142 purpura (TTP) 17, 149–53
diagnosis 170 dilutional 42 diagnosis 150–1
fetal HbH disease 168 fetal 141 differential 142, 151
genetics 166–7 heparin-induced 93–5 investigation 151
maternal carriers 167 immune (Evan’s syndrome) refractory 153
diagnosis 169–70 see immune treatment 152–3
maternal HbH disease thrombocytopenia additional 153
167–8, 170–1 (Evan’s syndrome) firstline 152–3
mechanism 165 investigation 141 secondline 153
pregnancy, problems in platelet transfusion 18, 19 treatment/outcome 147
166–7 in PNH 61 tinzaparin 96, 116
severity 166–7 in pregnancy 140 tissue factor pathway inhibitor
treatment 170–1 treatment/outcome 147 (TFPI) 6, 8
beta thalassemia 171–5 see also specific disorders tissue factor (TF)
diagnosis in pregnancy thrombocytosis 58 (thromboplastin) 4
fetal 174 thrombolytic therapy 120–1 tissue plasminogen activator
maternal 173–4 contraindication 110 (tPA) 8, 10
features 171–2 thrombomodulin 8 therapy 110, 121
iron overload 172 in pregnancy 10, 11 total iron-binding capacity
mechanism 170–1 thrombophilia(s) 4 (TIBC) 44, 47
pregnancy, problems in coagulation cascade in pregnancy 2, 3
172–3 associations 124 screening test 47
treatment 174–5 genetic/acquired tranexamic acid 70, 82
classification 165 components 134–9 transferrin receptors (tfR) 44,
inheritance 164 activated protein C 47
stem cell transplantation 176 resistance in pregnancy 2, 3
thalassemia intermedia (aPCR) 134–5 transfusion see blood
175–6 heritable component therapy; blood
types 165 antithrombin deficiency transfusion; platelet
treatment 16 127–8 transfusion; red cell
thermolabile methylene and fetal loss 133–4 transfusion
tetrahydrofolate reductase FV Leiden mutation transfusion reactions 21–8
(MTHFR C677T) 128–9 128–9, 132 allergy 22, 25
and fetal loss 133–4 and intrauterine growth bacterial contamination 22,
and hyperhomocysteinemia restriction 132–3 27
135 protein C deficiency 124, hemolytic 22, 23
and pre-eclampsia 132 127 infection 22
thrombin protein S deficiency 123–4 management 24
in coagulation cascade 4, 6, prothrombin G20210A new variant
8, 123 mutation 128–9, 132 Creutzfeldt–Jakob disease
generation potential in thermolabile methylene (nCJD) 27–8
195
Index
non-hemolytic febrile 22, velocimetry 34 sources 53

23–5 vena cava filters 120 C 48
post-transfusion purpura venous thromboembolism K 72
(PTP) 22, 26 (VTE) 112–21 von Willebrand factor (vWF)
prevention 19 deep vein thrombosis (DVT) antigen 10
transfusion-related acute lung 111, 112–13 and desmopressin 69
injury (TRALI) 25 and immobilization 118 in pregnancy 10
transfusion-related graft incidence/risks 111 in TTP 150
versus host disease (TA- prophylaxis 116–18 von Willebrand’s disease
GVHD) 22, 26 pulmonary 78–84
TTP see thrombotic thromboembolism antenatal management 82–3
thrombocytopenic purpura (PTE) 111, 113–14 classification 79
(TTP) risk delivery management 83–4
tubular necrosis, acute, post- age and 128 treatment 81–2
transfusion 19 assessment 114–15 VTE see venous
factors 118, 122 thromboembolism (VTE)
U antibodies 38 reduction, postpartum 116
UFH see unfractionated and thrombophilia 122 warfarin 61, 88–91
heparin (UFH) and travel 118–19 effects 88
unfractionated heparin (UFH) treatment in pregnancy factors affecting
92 119–21 anticoagulation 89–90
bleeding risk 96 venous thrombosis in pregnancy 90–1
and osteoporosis 95–6 gestational 3–4 and transfusion 17
and prosthetic heart valves after immunoglobulin white blood cell count in
105–7 therapy 19 pregnancy 2
reversal 92 in pregnancy 111–21
vs LMWH in pregnancy 95–7 see also venous ximelagatran 101
in VTE 119, 120 thromboembolism (VTE)
uric acid level elevation 130 vitamins Y-dimer 8
urobilinogen 56 B6 2, 3
urokinase (uPA) 8, 10, 121 B12 ‘ZZAP’ enzyme 21
deficiency 53–5
valvular heart disease in and folate deficiency 53
pregnancy 104–7 in pregnancy 2, 3, 52, 54–5
196

Practical Obstetric Hematology PDF

Uploaded by

Copyright:

Available Formats

Practical Obstetric Hematology PDF

Uploaded by

Document Information

Original Title

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

Practical Obstetric Hematology PDF

Uploaded by

Copyright:

Available Formats

Obstetric Prelims 20/10/05 3:09 pm Page i

Peter Clark BSc MD FRCP FRCPath

Ian A Greer MD FRCP FRCOG MFFP

© 2005 by Taylor & Francis Group, LLC

No claim to original U.S. Government works

International Standard Book Number-13: 978-0-203-44879-3 (eBook - PDF)

and the CRC Press Web site at

Chapter 1 Hematology measurements in pregnancy 1

Chapter 2 Blood group serology and blood products 15

Chapter 3 Maternal anemia and hemato-oncology in pregnancy 43

Chapter 4 Bleeding disorders in pregnancy 66

Chapter 5 Antithrombotic therapy in pregnancy 88

Chapter 6 Management of thrombotic disorders 102

Chapter 7 Thrombophilia and pregnancy outcome 122

Chapter 8 Thrombocytopenia in pregnancy 140

Chapter 9 Hereditary red cell disorders 163

IDA Iron deficiency anemia

PMH Past medical history

Practical Obstetric Hematology

Table 1.1 – Blood count and hematinics in normal pregnancy

Parameter Change in pregnancy

Towards the end of pregnancy at least 5% of women have a platelet count

Hematology measurements in pregnancy

Practical Obstetric Hematology

Normal hemostasis control

Figure 1.1 The coagulation cascade

Hematology measurements in pregnancy

XIII XIIIa Cross-linked

Practical Obstetric Hematology

Figure 1.2 Coagulation control

Hematology measurements in pregnancy

X VIIIa VIIIi +PS

XIII XIIIa Cross-linked

Practical Obstetric Hematology

Figure 1.3 Fibrinolysis

Hematology measurements in pregnancy

Practical Obstetric Hematology

Pregnancy-associated changes in coagulation and fibrinolysis

Coagulation control in pregnancy

Hematology measurements in pregnancy

Table 1.2 – Coagulation in pregnancy

Change from first to Change from first to

Fibrinogen 11 31

Table 1.3 – Coagulation control in pregnancy

Change from first to Change from first to

aPCR, Activated protein C resistance; APTT, Activated partial thromboplastin time.

Practical Obstetric Hematology

As is discussed in Chapter 7, the significance of inherited resistance to aPC

Hematology measurements in pregnancy

Delivery and the early puerperium

Changes in other coagulation assays during pregnancy

Practical Obstetric Hematology

composition of the phospholipid used in different partial thromboplastin tests.

Blood group serology

Practical Obstetric Hematology

there has been a previous exposure, re-exposure may result in an anamnestic

Blood component therapy

Table 2.1 – Blood component usage

Red cells Hypoxia secondary to anemia