Pregnancy and homozygous beta thalassaemia major


Correspondence: Dr. G. Daskalakis, 4 Fidiou Street, 135 61 St. Anargiri, Athens, Greece.


Nine pregnant women with homozygous β-thalassaemia major followed a strict transfusion regimen to maintain their haemoglobin level > 10g/dl. One pregnancy was terminated because of concern about desferrioxamine teratogenicity and another ended in miscarriage at 11 weeks. All other women were delivered by elective caesarean section between 37 and 38 weeks. There were no obstetric complications or perinatal deaths.


The thalassaemias are a group of recessively inherited disorders. Failure of globin chain synthesis results in defective erythropoiesis and haemolysis. Homozygous α-thalassaemia is incompatible with life, but homozygous β-thalasssaemia, or Cooley's anaemia, has a more favorable prognosis. These patients develop severe anaemia and require repeated blood transfusions. This may lead to iron overload and tissue hemosiderosis, and increased erythropoiesis results in hepatosplenomegaly and skeletal abnormalities. Endocrine impairment is usually observed, and the onset and development of secondary sex characteristics is always delayed1. This has been attributed to pituitary and gonadal damage resulting from iron overload2,3. The introduction of intensive iron chelation by subcutaneous infusions of desferrioxamine, and improved haematological management, has prolonged life expectancy up to 40 years and has improved endocrine function. The desire for pregnancy has become an important issue for these patients4,5,6.

Reports of pregnancy in women with homozygous beta thalassaemia major are rare. An association with multiple maternal and fetal problems is recognized7,8. In this report experience of nine pregnancies in women with homozygous thalassaemia major is described, and their management and outcome discussed. To our knowledge this is the largest series reported from one Centre.


A total of nine pregnant women with homozygous beta thalassaemia major, managed in our Department between 1993 and 1996, were included in the study. The women were diagnosed as having beta thalassaemia major, during the first two years of their life, on the basis of the presence of severe anaemia, abnormal red cell indices on a stained film, family studies, haemoglobin electrophoresis and globin chain synthesis. In those years, routine screening for haemoglobinopathies was not available in Greece. Therefore, prenatal diagnosis of beta thalassaemia was unusual. Our patients were transfusion-dependent and had been on iron chelation therapy. The mean maternal age was 24 years (range 21–30 years). One was gravida two and eight were primigravidas. Information collected from medical record included mode of conception, antenatal events and outcome for both mother and baby. Gestational age at delivery was assessed by last menstrual period, or by early ultrasound scan if there was a discrepancy of more than seven days.

Antenatal management involved repeated blood transfusions to maintain the level of haemoglobin above 10g/dl and thereby suppress the synthesis of abnormal haemoglobin and defective cells susceptible to haemolysis. Increased folate supplementation was also given to prevent megaloblastic anaemia. Haemoglobin and ferritin levels were monitored throughout pregnancy. Cardiac function was assessed by electrocardiogram, echocardiogram and physical examination by a senior cardiologist every month. Screening for viral infections, undertaken at booking, included hepatitis B (surface antigen), antibodies to hepatitis C virus, and human immunodeficiency virus. Maternal weight gain, blood pressure and urine analysis were recorded at biweekly intervals. Ultrasonographic assessments were performed usually during the first trimester and then at 20th, 24th, 30th and 34th gestational week, while fetal biophysical profile was performed twice a week starting from the 30th week.


The mean age of menarche was 15 years (range 13–17 years). Four women had undergone splenectomy and one of them cholocystectomy at an early age. The clinical details of all patients are summarised in table 1. Eight of the women had conceived spontaneously. One had conceived after ovulation induction using menotrophin and chorionic gonadotrophin. None had diabetes mellitus or thyroid dysfunction. All patients had been followed up in a specialist Centre for thalassaemia prior to their pregnancies, and the mean level of haemoglobin had been maintained above 10g/dl, in an effort to suppress erythropoietin, thereby reducing the impact of excessive ineffective erythropoiesis on bone development. This target had been achieved before pregnancy, with regular blood transfusions of a mean of two units of washed packed red cells every three to five weeks (mean transfusion interval 25 days). The transfusion requirement during pregnancy increased to a mean of two units of washed packed red cells every two to three weeks (mean transfusion interval 18 days). The total amount of blood administration during pregnancy ranged from 1,200 to 9,600 ml. None of the women had evidence of current hepatitis B virus infection, while two were anti-hepatitis B virus and one anti-hepatitis C virus positive. The haemoglobin levels ranged from 10.2 to 12.3 g/dl. The ferritin levels ranged from 1,480 to 9,160 μg/l (normal range: 14–150 μg/l). The white cell count ranged from 9,000 to 19,200/μl and the thrombocyte count ranged from 172,000 to 587,00. All women who had undergone splenectomy had been given prophylactic vaccination against pneumococcus prior to the surgery, and were receiving prophylactic penicillin V 250 mg, intramuscularly twice daily. One pregnancy was terminated at 18 weeks because of concern about desferrioxamine teratogenicity, although no abnormalities were identified. Another ended in a miscarriage at 11 weeks. All other women were delivered by elective caesarean section between 37–38 weeks and the postoperative course was uneventful. The mean birth weight was 2,836 g (range 2,450–3,270 g). There were no significant obstetric antenatal complications, nor perinatal deaths.


Only a few pregnancies in women with homozygous thalassaemia have been reported so far, the majority of these referring to the intermediate variety. The present report describes nine pregnancies in women with homozygous beta thalassaemia major. All of our patients were transfusion dependent. No adverse effects of pregnancy on the course of the disease, or the thalassaemia on pregnancy outcome, were observed. In some previous reports a relatively high rate of pregnancy loss, intrauterine growth retardation and prematurity have been reported and this had been attributed to chronic placental hypoxia caused by the maternal anaemia9,10. The apparent discrepancy may be accounted for by the stringent transfusion regimen in our group and suggests that every effort should be made to maintain maternal haemoglobin at an adequate level. Our results are in agreement with a recent report of Jensen et al. who concluded that pregnancy in women with beta thalassaemia does not appear to be associated with increased obstetric complications8.

In thalassaemic patients there is a marked increase in erythroid activity, in order to overcome the increased rate of destruction, which results in an enormous expansion of the bone marrow. There are no randomized trials comparing pregnancy outcome according to different haemoglobin levels and the optimal pretransfusion haemoglobin is a controversial issue. Although maintenance of the haemoglobin level above 10 g/dl with a strict transfusion regimen is not universally accepted, it has been shown that this may provide enough suppression of ineffective erythropoiesis to reduce its impact on bone development11,13. Moreover the amelioration of severe anaemia reduces significantly the increased gastrointestinal iron absorption13,14. The increased plasma volume and cardiac output in pregnancy in association with the severe anaemia and the myocardial hemosiderosis from iron overload in thalassaemia predispose to congestive heart failure. It is therefore essential to maintain a reasonably high mean haemoglobin concentration in order to prevent cardiac compromise. Another problem, associated with maternal anaemia, is fetal hypoxia which leads to growth retardation9,15. As a response, fetal factors stimulate placental hypertrophy, which however, has been shown to be inadequate when the haemoglobin level falls below 10 g/dl15. The use of young washed packed red blood cells may also reduce the transfusion frequency and minimize the risk of overloading the circulation13. All our patients required increased transfusion frequency in order to maintain the haemoglobin level above 10 g/dl throughout pregnancy and none of them developed heart failure. There is always the risk of transfusion-related viral infections, despite blood donor screening. Moreover, isoimmunisation to minor group antigens can cause difficulties in cross-matching blood and may predispose to fetal isoimmunisation. Folate demands in pregnancy are also increased and become even higher in thalassaemia because of the marrow over-activity. Therefore folic acid supplementation is required to prevent superimposed megaloblastic anaemia.

Chelating treatment before pregnancy by subcutaneous infusion of deferrioxamine reduces the effects of iron overload and has increased the life expectancy and fertility rates. Whether the desferrioxamine administration is safe during pregnancy is not established. Teratogenic effects, such as retardation of bone ossification, vertebral aplasia and bifurcation or fusions of the ribs have been observed in some animal studies. Additionally, electroencephalogram abnormalities, ricket-like bone alterations, and cataract formation in adults on long term desferrioxamine, due to multiple transfusions for hypoplastic anaemia, have been reported16. Only eight women in the world literature received desferrioxamine for several weeks during their pregnancy6,8,17–21. There were no adverse fetal outcomes, but two patients had premature deliveries. Desferrioxamine therapy probably should be stopped as soon as pregnancy is diagnosed, or in mid-cycle when ovulation induction therapy is being used. Termination of pregnancy should be discussed with women who are worried about the risk of congenital malformations. One of our patients, who had taken desferrioxamine without realising that she was pregnant, had her pregnancy terminated at 18 weeks because of the fear of drug toxicity. The remaining patients showed no more than a 10% increase in their serum ferritin level following delivery, despite the fact that chelation therapy had been interrupted. This finding is in agreement with that of Jensen et al.8, and it strengthens the previous reported concept that pregnancy is by itself an efficient chelator of iron19. Breastfeeding during desferrioxamine treatment is disputable. Although there is a lack of clinical evidence, it is unlikely to be harmful to the infant, as it is little absorbed from the human gut. Vitamin C is usually administered to patients with homozygous beta-thalassaemia, because it increases the urinary excretion of iron by the chelating agent. However, in pregnancy it should be stopped, as it increases gastrointestinal absorption of dietary iron and might precipitate cardiac damage22.

Splenectomy due to hypersplenism is generally performed at a young age1. Hypersplenic crises seem to occur more frequently during pregnancy, especially in the last few weeks, so that anaemia, leukopenia and thrombocytopenia may occur either singly or in combination15. Splenectomy should be reversed for those with severe thrombocytopenia which causes uncontrolled bleeding. There is limited experience of pregnancy in splenectomised women8–10. Four of our patients had undergone splenectomy and no adverse effects were noted during pregnancy in any of them. Although a predisposition to venous thrombosis after splenectomy due to changes in the clotting mechanism has been assumed, none of our patients had any thrombotic event. This is in agreement with published data when only one out of 13 thalassaemic patients, who had undergone spelenectomy prior to their pregnancy, suffered saphenous thrombophlebitis6–9,18,19,23,24. Aspirin administration during pregnancy in this group of patients, in order to diminish the risk of thrombosis, has been suggested, but there are no clinical trials so far8,10.

The chronic anaemia of thalassaemic patients can have detrimental effects on the fetus by affecting placental gas exchange24. In our cases despite maintenance of a high haemoglobin level, intrauterine growth retardation occurred in two. Fetal biophysical profile twice a week starting from the 30th week was performed in all pregnancies. Although this may seem excessive, it was performed in order to detect early signs of placental insufficiency. The high reported incidence of this complication and the fact that the chance of a future pregnancy in these patients is uncertain, led to close observation for fetal assessment and maternal reassurance. All seven women were delivered by elective caesarean section at 37–38 weeks. Similar high rates of caesarean section have been mentioned in previous studies. However, all of the latter were performed for obstetric reasons8,10. Labor with a haemoglobin level below 10 g/dl can lead to significant maternal morbidity or even mortality. Our policy was to maintain the haemoglobin level above 10 g/dl throughout pregnancy and to deliver the patients under the best possible haematological condition. Thalassaemic patients in addition have small stature and cephalopelvic disproportion during labor. Emergency caesarean section is very common8,25. Therefore, in view of the low chance of spontaneous vaginal delivery, and because some of the fetuses were small for gestational age, elective caesarean section was performed near term to avoid fetal distress during labor. A planned caesarean section has also the advantage of readily available checked blood for a possible transfusion. Although this policy is not generally applied, it was accepted in our Unit as a safe way to manage these pregnancies. Vaginal delivery would be preferable for women with thrombocytopenia, in order to avoid bleeding during, or after the operation26.

Associated with intensive transfusions, splenectomy and deferrioxamine usage pregnancy in women with homozygous beta thalassaemia major is possible. Many women who have been on this therapy from their early life are now in their reproductive age. Pregnancy in these patients requires close monitoring and careful treatment with frequent blood transfusions and folate supplementation. This management can prevent fetal hypoxia and can lead to a successful pregnancy outcome.