Sickle cell disease and β-thalassaemia major in pregnancy

Authors

  • Asma A Eissa MBBS MRCOG,

    Clinical Research Fellow
    1. Department of Obstetrics & Gynaecology, Royal Free Hospital, London, UK
    Search for more papers by this author
  • Susan M Tuck MD MRCGP FRCOG

    Consultant Obstetrician and Gynaecologist, Corresponding author
    1. Royal Free Hospital, London, Greater London, UK
    • Department of Obstetrics & Gynaecology, Royal Free Hospital, London, UK
    Search for more papers by this author

Correspondence: Susan Tuck. Email: susan.tuck@nhs.net

Abstract

Key content

  • Pregnancy in women with sickle cell disease continues to show high rates of maternal and fetal mortality and morbidity. This applies to all forms of sickle cell disease.
  • The place of blood transfusions in helping women with sickle cell disease in pregnancy continues to be controversial.
  • Genetic counselling for women with haemoglobinopathies should be provided as part of the UK national Antenatal Haemoglobinopathy Screening Programme, which was introduced in 2001.
  • There is an increased incidence of venous thromboembolism in individuals with sickle cell disease, transfusion-dependent β-thalassaemia and sickle cell trait.
  • Individuals with β-thalassaemia major who are able to survive childhood by virtue of the availability of regular and frequent blood transfusions frequently acquire the organ damage caused by haemosiderosis. Among the organs involved are the pituitary, the thyroid, the islet cells of the pancreas, the heart and the liver.
  • The main concern for women with β-thalassaemia major contemplating pregnancy is the possibility of cardiac dysfunction, resulting from myocardial haemosiderosis, causing both ventricular pump failure and arrythmias.

Learning objectives

  • To understand the key care requirements of women with sickle cell disease and β-thalassaemia major, prior to pregnancy, during pregnancy and in the postnatal period.
  • To appreciate the importance of knowledgeable care for this relatively uncommon group of patients, and to be able to work collaboratively with haematologists and others who have such experience and expertise.
  • To understand the principles relevant to antenatal screening for haemoglobinopathies, and to be able to give appropriate advice to couples.

Ethical issues

  • Suboptimal care is sometimes given when patients are cared for in hospitals with little experience of the condition, and outcomes can be improved by arranging care in centres with specialist expertise.
  • Should pregnancy be encouraged, and fertility treatment given for women with a reduced life expectancy? How should this be balanced with respect for patient autonomy?

Key issues in sickle cell disease

Pregnancy in women with sickle cell disease continues to show high rates of maternal and fetal mortality and morbidity. There is some suggestion, from the USA and UK, of a worsening trend in pregnancy outcomes over time. This may reflect better survival of sickle cell patients with worse disease patterns, who are now able to undertake pregnancy. There are continuing problems with the inability to provide a consistently high level of care for patients with sickle cell disease, as was demonstrated in the National Confidential Enquiry into Patient Outcome and Death (NCEPOD)[1] into deaths in patients with sickle cell disease in the UK.

The place of blood transfusions in helping women with sickle cell disease in pregnancy continues to be controversial. This is partly because there are significant related complications to consider, and partly because the natural history of sickle cell disease shows such a varying path, that it is difficult to distinguish the extent to which any improvement in outcome may be correctly ascribed to blood transfusions alone.

A significant related issue is the practical difficulty for women with sickle cell disease in receiving appropriate advice on safe and effective contraception, so that their pregnancies can be planned, when their sickle cell disease and related health issues are in an optimal state.

Maternal mortality

In the 1970s and 1980s the maternal mortality rate in the UK was about 3%, 200 times higher than the contemporaneous overall maternal mortality rate, with about 30 pregnancies per year in women with sickle cell disease.[2, 3] A decade later, a second survey of sickle cell pregnancies in the UK from 1991 to 1993 indicates a mortality rate of about 3.3%, 330 times the overall maternal mortality in the UK at that time of 9.8 per 100 000 births.[4, 5]

The more recent Confidential Enquiries reports indicate a similar rate of maternal mortality for women with sickle cell disease in the UK, with three deaths in the 2000/02 triennium (two who had haemoglobin [Hb] SS, and one with Hb SC), two deaths in the 2003/05 triennium (one with Hb SS and one with Hb SC) and four deaths in the 2006/08 triennium (two from coliform infection after spontaneous preterm prelabour rupture of membranes at 17 and 23 weeks of gestation, one with Hb SS following collapse with chest pain, presumed to be due to the sickle chest syndrome, and one with Hb SC due to Streptococcus pneumoniae urinary tract infection, leading to septicaemia).[6, 7] One might speculate on a number of possible explanations for this. It may indicate that optimal care is not yet being consistently provided, and it is likely that this would be more of a risk in hospitals not used to looking after patients with sickle cell disease. Indeed this is a concern raised in the specific national Confidential Enquiry into deaths in sickle cell disease in the UK covering the years 2005 and 2006.[1] This phenomenon would be consistent with recent publications from experienced clinicians at three centres with well-established specialist services for patients with sickle cell disease; at Hammersmith & Queen Charlotte's Hospital, St Thomas’ Hospital and King's College Hospital, all in London, which reported no maternal deaths in series of 25, 62 and 71 pregnancies, respectively, between 2000 and 2007.[8-10] It is also possible that improved paediatric and haematological care has meant the better survival to reproductive age of individuals with mutations that give a worse phenotype and natural history for their disease. In addition, it may be that women themselves are choosing to undertake pregnancy when they might previously have been advised against this.

Perinatal mortality

Perinatal mortality data from the same series show similar variations, although the absolute numbers are not so much higher than the prevailing rates for the general population as are the maternal mortality rates. For example, the perinatal mortality rate (PMR) from the UK series covering years 1975–1982 was 48 per 1000 births (four times the national PMR at that time), and from the 1991–1993 series was 60 per 1000 (six times the contemporary national rate).

Again, recent data support the impression that outcomes can be better in hospitals where there is expertise and experience with the condition. There were no perinatal deaths in the 25 and 62 pregnancies referred to above, at Queen Charlotte's and St Thomas’ Hospitals, and one perinatal loss in the 71 pregnancies described at King's College Hospital.

Maternal morbidity

All the published series referred to above indicate broadly similar experience with regard to the features of maternal morbidity. Acute sickling crises (sufficiently severe to require hospital admission), occur in about 37% of pregnancies before delivery and in about 12% of pregnancies during the puerperium. Episodes of infection are often compounded by acute sickling, with a particular propensity to urinary tract infections (seen in about 13% of pregnancies). Chest syndrome is a term used to cover episodes of acute illness which involve features suggestive of pneumonia, pulmonary sickling and/or pulmonary embolism, and is seen in about 12% of pregnancies. Since both sickle cell disease and pregnancy are recognised risk factors for venous thromboembolism, it is not surprising that this also occurs with increased frequency, although the diagnosis may be difficult to distinguish from the features of acute red cell sickling. A propensity for severe pre-eclampsia (seen in about 9% of pregnancies) is widely reported, and we may speculate that the cause of this link is related to the endothelial damage caused by the altered shape and increased rigidity of sickled red blood cells.

Sickle cell disease has a notably fluctuating course, both between individuals, with the same and different genetic mutations, and in the same individual over time. Therefore some variation in complication rates during pregnancy may relate to coincidental variations unconnected to the pregnancy or its care.

Fetal and neonatal morbidity

The fetal and neonatal problems are non-specific, and are simply those expected from a suboptimal intrauterine environment, namely of placental insufficiency, fetal growth restriction (in 18–23%), preterm delivery (in 16–24%) and fetal distress during labour. These figures are from the recent London hospital data, already referred to, and are accompanied by caesarean section rates of 30–72%.

Prepregnancy care

Prior to pregnancy, up-to-date checks should be made of the woman's serology for hepatitis B and C, HIV and rubella, and relevant measures taken, accordingly. She should be up-to-date with pneumococcal and hepatitis B immunisations. The woman should be taking folic acid supplements (5 mg) daily, and appropriate penicillin prophylaxis, given in the context of either previous splenectomy or functional loss resulting from previous splenic infarctions. Any other medication should be reviewed in the context of possible teratogenicity, and stopped or changed to alternatives, as appropriate. Hydroxycarbamide (previously known as hydroxyurea), for example, is sometimes prescribed for patients with sickle cell disease in order to stimulate the production of fetal haemoglobin and mitigate the effects of Hb S. The British National Formulary lists this as contraindicated in pregnancy and breastfeeding. It has been found to be teratogenic in animal studies and should ideally be stopped 3 months prior to conception. Iron chelators, which are essential treatment to reduce the organ damage which would result from excessive iron deposition derived from multiple blood transfusions, should be stopped as soon as pregnancy is diagnosed, for the same reasons. Vitamin C is usually given with iron chelators to enhance their efficacy. However, this also needs to be stopped when the chelators are stopped, to avoid its action in increasing the intestinal absorption of dietary iron. Women should be reminded of general measures to reduce the occurrence of sickling crises, such as keeping warm and well hydrated.

Assessments should be made of cardiac function (with electrocardiography and echocardiography), retinal screening (for sickling infarctions), and renal function (with blood pressure and urine protein checks). There may be general health issues which need to be addressed before pregnancy for the individual woman, such as advising smoking cessation. She also needs to be made aware of what the arrangements will be for her pregnancy care, so that she is prepared for these practicalities.

Genetic counselling

Genetic counselling for women with sickle cell disease and trait should be provided as part of the UK national antenatal haemoglobinopathy screening programme, which was introduced in 2001. Ideally, the woman's carrier status and that of her partner would have been identified as part of general prepregnancy health advice prior to planning pregnancy. Failing that, she should have haemoglobinopathy screening at the time of booking for antenatal care, unless she lives in a ‘low prevalence’ area (with fewer than 1.5 per 10 000 pregnancies with sickle cell disease fetuses/babies per year), in which case initial screening is undertaken by ascertaining her ethnicity, by a standardised family origin questionnaire, followed by laboratory testing, when relevant. Currently 1 in 35 pregnant women in the UK are identified by the national screening programme as carrying a haemoglobinopathy.

If her partner also carries a significant haemoglobinopathy, the expected inheritance pattern is that of an autosomal recessive condition, and genetic counselling can be given accordingly. Fetal testing by chorionic villus sampling, amniocentesis and fetal blood sampling are all possible, with their attendant possible complications, and the possibility of selective termination of the pregnancy if the fetus is found to be affected. As in all such counselling, it is important to give the couple full and up-to-date information on the potential health issues and care for the expected child, all of which is available through the NHS Sickle Cell and Thalassaemia Screening Programme.

Pregnancy care

The outcome data summarised above indicate that the level of care and attention in pregnancy and the puerperium should be the same for women with all types of sickle cell disease (Hb SS, SC, SD or Sβ0 thalassaemia), including women with infrequent complications prior to their pregnancy, as none of these marks a consistent prediction of the rate of complications related to pregnancy. There should be a clearly coordinated plan of care, with effective communications between all those involved, including a clear system for enabling the patient to seek emergency help without delay, whenever necessary. This level of care needs to continue into the puerperium, when the patient is still at risk of sickle cell crises and other serious complications.

The patient should be encouraged to come to hospital promptly whenever she experiences symptoms suggesting an imminent sickling crisis, as early support may be able to avert a more severe episode. This would include enhanced hydration, infection screen and early recourse to antibiotics for any infection, which may have been the trigger for the crisis. This particularly includes urinary and chest infections, endometritis in the puerperium, and malaria in those with recent foreign travel. A careful assessment of her state of oxygenation should be made, with appropriate support, as relevant. Thromboprophylactic measures should be instituted. Analgesia should be tailored to the needs of the individual, and is likely to include the use of morphine, which may be more effectively administered through a patient-controlled system. Pethidine is generally avoided, because it is less effective, and its metabolites tend to have longer-lasting depressant effects, as well as a tendency to cause convulsions. The patient should be transferred to an intensive care unit sooner rather than later, if her condition does not respond satisfactorily.

With regard to her general care during pregnancy, clearly this needs to be tailored to the individual patient. As a guide, the Hb concentration, haematocrit, platelet count, bilirubin, transaminase and lactose dehydrogenase levels should be checked every 2 weeks. Blood pressure checks and urine checks for infection, haematuria and proteinuria should be made at least every 2 weeks. Fetal growth should be monitored with measurements by serial ultrasound scans. There should be a plan arranged for specific thromboprophylactic measures appropriate for each individual. These are likely, as a minimum, to include daily injections of low molecular weight heparin for 6 weeks following delivery. The timing and mode of delivery is advised according to the relevant details of each woman's pregnancy, along standard obstetric lines. Vigilance and care should be sustained into the postnatal period, when there is also a significant incidence of sickling crises, as well as thromboembolic events.

The use of blood transfusion

This is the aspect of care that probably causes the greatest difficulty for clinicians. The available data are difficult to interpret appropriately. As Mahomed's review[11] states, there is ‘not enough evidence to draw conclusions’ on the extent to which blood transfusions help the outcome of pregnancies in women with sickle cell disease. There are some specific circumstances for which most clinicians would agree that blood transfusions are indicated, namely for patients experiencing frequent severe sickling episodes, when the haematocrit has dropped below 0.26 and/or below the patient's normal steady state level, if she has a particularly poor obstetric history, if there are the increased demands of a twin pregnancy, and it is usually appropriate to have a greater readiness to give blood transfusions during the third trimester of pregnancy.

Those clinicians who do advocate transfusions on a prophylactic basis vary considerably in how this is conducted and what their targets are from this. The potential for oxygen delivery only approaches normal at HbS levels of less than 20%. Such a target requires either multiple transfusions at frequent intervals during the pregnancy, or at least one large volume exchange transfusion. Even so, it is difficult to attain and maintain normal haematological parameters throughout pregnancy, partly because the donor erythrocytes will have a limited lifespan (of an average of about 90 days), and partly because the patient is continuing to produce her own HbS-containing erythrocytes. It is important to maintain normal circulating blood volume for the patient during transfusions, and this may require simultaneous venesection. At the same time, the resulting haematocrit needs to be a compromise between sufficient red blood cells (and haemoglobin) to provide adequate oxygen delivery, without increasing the viscosity of the blood to a level at which the relatively inflexible Hb S-containing erythrocytes will be unable to negotiate the capillary circulation, thus provoking sickling and vascular obstruction. The donor erythrocytes from transfusion will dilute this potential problem, as well as temporarily suppressing the new production of the patient's own Hb S-containing erythrocytes. A number of other precautions are needed during these transfusions, such as near-patient warming to body temperature (cold being among the potential stimuli of sickling crises), and checking that the donor blood is free of viruses such as cytomegalovirus. As for all multiply transfused patients, the blood transfusion laboratory should be informed that the patient is in this category, so that a full phenotype match is undertaken.

Among the possible complications of blood transfusions for these women are the formation of atypical red cell antibodies, the transmission of infection, and the deposition in vulnerable tissues of the excess iron accumulated. Paradoxically, blood transfusion may provoke an acute sickling episode, or a hyperhaemolysis syndrome. It is also important to consider the costs and time involved for the patient, as well as the acceptability of the principle of blood transfusions to her.

Contraceptive problems

As many as two-thirds of pregnancies in women with sickle cell disease are unplanned, with the attendant disadvantages of the lost opportunity to optimise the health of the woman and the potential outcome of her pregnancy. Family planning advice is notoriously lacking for these women.

Sickle cell disease is listed by the British National Formulary and by manufacturers’ literature as requiring caution with the prescription of combined oral contraceptive pills, although this is without basis in any research data. These authorities give no explicit advice with regard to the progestogen-only forms of contraceptive. Those women who do take oral contraceptive pills may have pill failures when their taking of the pill is interrupted by an emergency admission to hospital. Also the efficacy of the contraceptive pill may be compromised by the use of broad spectrum antibiotics, which are frequently prescribed if a sickling episode is thought to have been provoked by infection. To date the only randomised controlled trial published on this issue studied 23 women with homozygous sickle cell anaemia, over a 2-year period, in a single-blind cross-over study of depot medroxyprogesterone acetate, and found a reduced occurrence of acute sickling and haemolysis while the women were receiving the active injections.[12]

Clinicians are often instinctively reluctant to advise use of the intrauterine contraceptive device, because of the potential complications of menorrhagia, exacerbating the anaemia, and infection, provoking acute sickling crises. (Both these concerns would, in principle, be avoided by use of the progestogen-releasing intrauterine system.)

The outcome of this paucity of guidance is that the majority of women with sickle cell disease receive either confused advice or none at all. A survey of 149 sexually active women with sickle cell disease in north London in 1993 found that 36% had been specifically advised to avoid pregnancy.[13] Concerns of the significant mortality risks both for the mother and the baby, as well as a consideration of the generally reduced ability of the woman to care for her child subsequently, are less likely to prevail nowadays over a respect for the woman's autonomy to make her own judgements in this.

Sickle cell trait

Pregnancy in women with sickle cell trait has few additional complication rates compared with other women of the same ethnic and obstetric background, the only issue of significance being a susceptibility to urinary infections. Recurrent urinary tract infections are seen in 6% of women during pregnancy, with 16% showing microscopic haematuria.[14, 15] The latter finding is a reflection of micro-infarcts from localised sickling in the peculiarly challenging renal environment, which may be sufficient to provoke sickling of red blood cells, even when only half the haemoglobin they contain is Hb S.

It is important to diagnose concurrent iron deficiency correctly, and serum ferritin measurements should therefore be made in each trimester, in order to guide the appropriate prescription of iron supplements.

Recent evidence has emerged suggesting an increased incidence of venous thromboembolism in individuals with sickle cell trait, as is already well recognised in sickle cell disease. The rates found are of a doubled incidence of venous thrombosis and four-fold greater incidence of pulmonary embolism, compared with similar individuals without sickle haemoglobin.[16] This increase is similar to those of individuals carrying the Factor V Leiden mutation, for example. The published studies have not included pregnant women, who naturally have an increased likelihood of venous thromboembolism by virtue of the changes related to pregnancy itself. Whether this is sufficient justification for specific thromboprophylaxis with heparin preparations is open for debate, but certainly this, and other standard risk reduction precautions, should be considered if any other complication were to be added to the pregnancy or delivery.

Key issues in β-thalassaemia major

Most doctors in the UK will encounter far fewer women with β-thalassaemia major than with sickle cell disease, so that unfamiliarity with appropriate care and advice is inevitably even more of a difficulty than it is in relation to sickle cell disease. The total number of young adults with β-thalassaemia major in the UK is about 800, with 20–30 babies born per year with the condition. (The number of babies born in the UK with sickle cell disease is about 300 per year, with the total number of sufferers being about 12 000.)

Fertility and other endocrine problems

Individuals with β-thalassaemia major who are able to survive childhood by virtue of the availability of regular and frequent (usually monthly) blood transfusions, frequently acquire the organ damage caused by haemosiderosis, which results from the deposition of the excess iron derived from the breakdown of the donor erythrocytes received. Among the organs involved are the heart, liver, pituitary, thyroid, and the islet cells of the pancreas.

A study of 97 individuals in London in 1996 found 66% of young adult thalassaemics with hypogonadotrophic hypogonadism, 20% with diabetes, and 10% with hypothyroidism.[17] Osteoporosis is also a significant problem, affecting 40% in this study, caused by a combination of endocrine, genetic factors and direct effects of the dysfunctional bone marrow on the skeleton. There is some dispute about whether the gonads are also directly affected by haemosiderosis, although this is unlikely to be widespread, as both males and females usually respond well to appropriate gonadotrophin therapy. However, the genetic causes of β-thalassaemia major are complex, involving a multiplicity of mutations, with subtle differences in the consequent natural history of the disease. Thus the rates of subfertility vary in different populations, being notably less evident in Greece, for example.

There are the same issues to consider as were touched on in relation to women with sickle cell disease, about the general advisability of pregnancy in women with a condition carrying a reduced life expectancy and a significant mortality (usually from cardiac causes) in pregnancy. However, in contrast with maternal sickle cell disease, the direct prospects for the fetus, with maternal β-thalassaemia major, are good.

Prepregnancy assessment

In the population and health service relevant to the UK, the main concern for women with β-thalassaemia contemplating pregnancy is the possibility of cardiac dysfunction, resulting from myocardial haemosiderosis, and causing both ventricular pump failure and arrythmias. Given the key importance of healthy cardiac function for the increased demands of pregnancy, therefore, a detailed assessment of cardiac function is the first prerequisite in the assessment of a woman before pregnancy. In addition to the standard methods of examining cardiac function and rhythm, magnetic resonance imaging of the heart provides a useful semiquantitative assessment of the iron deposition in the individual woman. A course of intensive chelation therapy may be able to achieve some reversal of these problems, sufficient to enable a woman to contemplate pregnancy appropriately.

Apart from the heart, the other key organ significantly susceptible to damage by haemosiderosis is the liver. Again this can be assessed in a semiquantitative manner by use of MRI, as well as by standard liver function tests, and histological assessment of a liver biopsy. The woman can then be advised of the relevant possibilities of deterioration in her liver function, and adverse fetal outcome, if she undertakes pregnancy.

The treatment of diabetes and hypothyroidism should be optimised before pregnancy, in the same way as in the general population, with the same issues being relevant to the risks of fetal anomalies and miscarriage, if not closely controlled.

Given their lifelong dependence on transfusions with donated blood, women with β-thalassaemia major are inevitably at risk of transfusion-transmitted infections, particularly hepatitis B, hepatitis C and HIV. These should be checked for before pregnancy, and appropriate antiviral treatment given, both for the benefit of the woman herself, and to reduce the risk of transmission to the fetus. The presence of atypical red cell antibodies, which might also have resulted from previous blood transfusions, should also be checked before pregnancy. In this context it may be helpful to check the blood group phenotype of her partner, in order to be able to anticipate the likelihood of haemolytic disease in the fetus, and give the couple information accordingly.

Osteoporosis is a widespread problem in adult patients with transfusion-dependent thalassaemia. This is likely to deteriorate in pregnancy, both because of the fetal demands for calcium, and the fact that the woman will need to be advised to stop bisphosphonate therapy during pregnancy, because of potential adverse effects on fetal skeletal development. Other supportive measures, such as weight-bearing exercise, and appropriate calcium and vitamin D supplements, can, however, be continued.

As for women with sickle cell disease, the genetic possibility for haemoglobinopathy in their offspring can be specifically predicted for the couple when the haemoglobinopathy status of the male partner is known. The couple can be advised accordingly about the availability of fetal testing, and also pre-implantation testing, if in vitro fertilisation techniques are used to achieve pregnancy.[18]

Pregnancy care

When a woman with β-thalassaemia major is actively trying for pregnancy or receiving fertility treatment, her usual iron chelation therapy will need to be discontinued, because of possible teratogenic risks to the fetus from this medication. Given that chelation therapy is to be stopped, vitamin C medication also needs to be stopped, and folic acid supplements (5 mg per day) should be given, to sustain the efforts for haemopoeisis in chronic anaemia, in conjunction with the fetal requirements.

A number of patients with β-thalassaemia major will have undergone splenectomy, in an attempt to prolong the life of their circulating erythrocytes. This renders them at risk of bacterial infections, particularly streptococcal infections. As after splenectomy for any reason, pneumococcal vaccination is advised, and any booster dose given prior to planned pregnancy. Frequently patients who have had splenectomy are maintained on prophylactic doses of penicillin, and this should be maintained prior to and during pregnancy, with a continuing awareness of the patient's susceptibility to infections.

Blood transfusions are likely to be needed at more frequent intervals than before pregnancy, to avoid major fluctuations in the haemoglobin concentration and cardiac workload. The aim should be to maintain the haemoglobin concentration above 10 g/dL, and this is likely to require transfusion of two units of blood at about 3-week intervals. The same measures as are relevant in sickle cell disease, to minimise the risks of transfusion in a pregnant woman receiving multiple donations, are required, although in this instance there are no practical concerns about the patient's haematocrit becoming excessively elevated.

Potential problems from cardiac failure or arrythmias are a major concern in these women, and careful assessment of relevant symptoms and signs should be made throughout pregnancy. If detailed cardiac assessment shortly prior to pregnancy has been normal, then further formal assessment should be planned for the third trimester and in the postnatal period. More frequent review may be relevant for the individual patient. The care of diabetes and hypothyroidism are along the same principles as would be relevant to all women with these conditions.

Thalassaemic patients have a chronic hypercoagulable state, with an increased incidence of thromboembolic episodes. The mechanisms for this include the presence of abnormal red cells that shed prothombotic microvesicles,[19] a variety of changes in platelet function, elevated levels of endothelial adhesion proteins, elevated levels of plasma coagulation factors, and low plasma levels of the natural anticoagulants, protein C and protein S and heparin co-factor II.[20] Those who have had a splenectomy have a further increase in their risk of thrombosis,[22] and venous thromboembolism is more frequent in patients with associated organ dysfunction[23] such as diabetes, cardiopulmonary abnormalities, hypothyroidism, and liver function anomalies. Many of these effects can be partially abrogated by hypertransfusion which will suppress endogenous red cell production. Thus unless there is a history of venous thromboembolism or additional risk factors such as antiphospholid syndrome, heparin thromboprophylaxis is not routinely used during pregnancy in women with thalassaemia major, although it is routine practice to recommend subcutaneous heparin for 6 weeks post delivery.

In the absence of the confounding influence of maternal diabetes, fetal growth in maternal β-thalassaemia major is usually normal, and plans for timing and mode of delivery can be made along standard obstetric principles. Ironically, the fact that most patients with transfusion-dependent thalassaemia are themselves of lower than average stature, however, often requires caesarean delivery because of cephalo-pelvic disproportion. There is no contraindication to breastfeeding with the resumption of iron chelation therapy after delivery, if parenterally administered medication, such as desferrioxamine, is used, since this is not active by the oral route, and hence the baby is not affected.

Pregnancy outcomes

As already indicated, the genetic heterogeneity of β-thalassaemia major does result in varying natural histories in different genetic populations. Pregnancy outcomes in the mixed ethnic population of the UK are illustrated by a series of 29 pregnancies in 22 women cared for at the Royal Free Hospital, London, between 1989 and 2005.[21]

Thirteen of these pregnancies followed ovulation induction therapy. Four women had pre-existing diabetes. Five had hepatitis C and one had hepatitis B before pregnancy. Five women had significant cardiomyopathy prior to pregnancy, two of whom died during the postnatal period. Of the 29 pregnancies, two resulted in miscarriage and three had termination procedures. There were 26 live births, including two sets of twins (following gonadotrophin stimulation of ovulation). Only four of the pregnancies were delivered before 37 weeks of gestation, and the birthweights of the 21 term babies ranged from 2050 to 4100 g (mean: 3240 g). The rate of delivery by caesarean section was high, involving 18 of the 24 relevant pregnancies, and in 14 cases this was because of feto-maternal cephalo-pelvic disproportion.

Postnatal complications

During the 6 months following delivery, two women of this series developed diabetes, and two developed new significant cardiomyopathy. Five women developed secondary amenorrhoea. Obviously it is not possible to know whether these complications would have arisen at that time for these individuals, in any case, or whether any aspect of their pregnancy contributed to these problems. It may be that the relatively prolonged period without iron chelation therapy, during which time blood transfusions are given at more frequent intervals than usual, accelerates the development of organ damage from iron overload. Since the evidence for teratogenicity of desferrioxamine is weak (being demonstrated in animal studies only, although the potential for causing iron deficiency in the fetus needs to be borne in mind), it may be appropriate to administer limited courses of intravenous desferrioxamine to the mother during pregnancy, if her serum ferritin level rises to a concerning degree, in order to mitigate this.

α- and β-thalassaemia traits

Genetic counselling

In theory, genetic counselling for couples who are both carriers of α- or β-thalassaemia is relatively straightforward, in that both conditions are inherited in a simple autosomal recessive fashion, and fetal testing by chorionic villus sampling, amniocentesis or fetal blood sampling is available. Pre-implantation testing is also technically possible for these conditions, if conception is by in vitro fertilisation. However, different mutations and combinations of mutations do vary in the haematological phenotype produced, and it is important that couples are given correct information about the severity of clinical condition which is potentially relevant to their child.

In α-thalassaemia major, the individual is incapable of making any α-globin chains, and neither normal fetal nor normal adult haemoglobins (Hb F and Hb A) can be produced. Such a fetus would therefore have severe anaemia, resulting in cardiac failure (hydrops fetalis) and intrauterine death, with the pregnancy sometimes also complicated by severe pre-eclampsia. In contrast, a fetus with β-thalassaemia major will be healthy, with a normal production of fetal haemoglobin, and will not manifest clinical symptoms until 4–6 months of postnatal age, by which time the production of fetal haemoglobin will have substantially ceased, and the child is dependent on its defective production of adult haemoglobin. Such an individual will then have a lifelong dependence on repeated blood transfusions, as is described in the preceding section. However, modern haematological care for such children does now include the possibility of bone marrow transplantation, as well as better-tolerated regimens of iron chelation therapy, including oral preparations, which would reduce the possibility of organ damage from haemosiderosis.

Pregnancy care

β-thalassaemia trait is relatively easy to diagnose, on the basis of standard red cell indices (indicating hypochromia and microcytosis) and quantification of the different haemoglobin types in the individual (with an increased percentage of fetal haemoglobin and haemoglobin A2). Many women will already know, prior to pregnancy, that they are carriers of β-thalassaemia, and their partner's haemoglobinopathy status will also be known. Otherwise, testing in the national Antenatal Screening Programme should be undertaken as early as possible in the pregnancy. Reliable diagnosis of α-thalassaemia trait may be more problematic, as the haematological picture can be difficult to differentiate from iron deficiency anaemia. Knowledge of the couple's ethnic origin will help in judging the probability of significant α-thalassaemia mutations. Specific α-globin chain analysis will also permit the correct diagnosis.

Once these issues have been clarified, the only other practical aspect relevant to pregnancy care is the need to recognise the development of concurrent iron deficiency. This will require measurement of serum ferritin levels, which should be undertaken in each trimester of the pregnancy. Conversely, it is important to avoid giving unnecessary iron supplements to the patient, if the cause of her microcytic anaemia has been misunderstood.

Summary

Rates of maternal and fetal mortality and morbidity for pregnancies complicated by all forms of maternal sickle cell disease are high. Key points are highlighted for the assessment of women with sickle cell disease prior to pregnancy and their care during pregnancy and the postnatal period, including the controversial issue of the appropriate use of blood transfusions. The absence of clear evidence-based advice for women with sickle cell disease on the risks associated with effective contraceptive measures is highlighted. Genetic counselling for couples with haemoglobinopathies should ideally be provided before pregnancy, or through the national Antenatal Screening Programme. Young adults with β-thalassaemia major commonly have multiple organ damage from deposition of excessive iron derived from their lifelong dependence on frequent blood transfusions. This produces multiple endocrine disorders, including hypogonadotrophic hypogonadism and osteoporosis. Cardiac dysfunction is the critical issue for their health in pregnancy. Women with haemoglobinopathy traits have few complications in pregnancy, other than the difficulty of distinguishing concurrent iron deficiency, and the added incidence of urinary tract infections in women with sickle cell trait. All major haemoglobinopathies and sickle cell trait increase the incidence of venous thromboembolism.

Disclosure of interests

None to declare.

Ancillary