Haemolytic Disease of the Newborn

Synonyms: HDN, Rhesus haemolytic disease, erythroblastosis fetalis.

Haemolytic Disease of the Newborn (HDN) was first described by a French midwife in a set of twins in 1609. It was later termed erythroblastosis fetalis when Louis K Diamond and his co-workers recognised the relationship between erythroblasts in the circulation, anaemia, fetal hydrops and jaundice. The rhesus blood group system was identified in 1940 and the link between rhesus haemolytic disease and alloimmunisation was recognised in 1953.

Rhesus alloimmunisation (Greek: allo = 'other' or 'different from') begins with red blood cells from a rhesus positive fetus crossing the placental barrier during pregnancy and delivery, and entering the maternal blood circulation. A rhesus positive father and a rhesus negative mother are required for this situation to develop. The incompatible antigens introduced result in a primary immune response and stimulate the production of maternal antibodies. A very small amount of fetal-maternal haemorrhage (FMH) needs to occur (less than 0.1 ml) and most goes unrecognised. Primary exposure can also be the result of amniocentesis, chorionic villus sampling and cordocentesis.

Several fetal rhesus antigens may cause alloimmunisation (c, C, d, D, e and E) and this can also occur with the Kell, Duffy, ABO and other blood group systems. The vast majority of haemolytic disease is caused by the rhesus D antigen.

There are rarely any problems during primary exposure but subsequent pregnancies result in large amounts of maternal anti-D antibodies being produced and the risk increases with each gestation. These are capable of crossing the placenta, where they affix to fetal red blood cells, which then become recognised as 'foreign' by the fetal immune system and haemolysed by fetal macrophages and lymphocytes. Significant haemolysis results in anaemia, hyperbilirubinaemia and the production of excessive erythropoid tissue in the liver, spleen, bone marrow, skin and placenta. In severe cases, multi-organ dysfunction and hypoproteinaemia can develop.

The incidence of HDN depends on the proportion of the population who are RhD negative. This varies within ethnic minorities. In the UK, this is highest in the white population (approximately 16%). In 2005, approximately 65,000 RhD positive babies were born to Rh negative women (approximately 10% of all births). With the use of anti-D immunoglobulin after a sensitising event but without rhesus antenatal anti-D prophylaxis (RAADP) about 1% of women would become sensitised (approximately 650). It is estimated that 550 of these women would go on to have a further pregnancy. Taking into account these further pregnancies, about 520 mothers would require close monitoring and 10-12% of babies would require intrauterine transfusions. Fetal anaemia or HDN would lead to about 37 fetal or neonatal deaths, 21 children would have minor developmental problems and 8 children would have major developmental problems.

Anti-D prophylaxis (mostly administered postnatally) and advances in neonatal care have reduced the frequency of HDN by almost a factor of 10 to 1 in 21,000 births.⁴ Deaths attributed to RhD alloimmunisation have fallen from 46/100,000 births before 1969, to 1.6/100,000 in 1990.This may not be entirely attributable to immunoglobulin; changes in abortion rates and racial composition may also play a part.⁵

One recent American study quoted a prevalence of RhD alloimmunisation of 6 in 1,000 births and suggests that this should now be considered a rare condition.⁶

Risk factors²

• Immunisation during first pregnancy.
• Immunisation during second or subsequent pregnancy.
• Failed prophylaxis.
• Over 99% of women have an FMH of less than 4 ml at delivery. 50% of women who have larger FMHs occur after normal deliveries. However, the following clinical circumstances are more likely to be associated with large FMH:
  • Traumatic deliveries including caesarean section.
  • Manual removal of the placenta.
  • Stillbirths and intrauterine deaths.
  • Abdominal trauma during the third trimester.
  • Multiple pregnancies (at delivery).
  • Unexplained hydrops fetalis.

Antenatally, the first indication of the condition is the presence of anti-D antibodies in the mother as detected by the indirect Coombs' test. All rhesus negative women have this test performed in the UK at the first antenatal visit.

Routine ultrasound screening may detect hydrops fetalis (see below) or polyhydramnios.

Infants born to alloimmunised mothers may appear clinically normal in mild cases. Diagnostic findings include jaundice (yellow amniotic fluid, yellow vermix, yellow skin), pallor and hepatosplenomegaly. Kernicterus (bilirubin encephalopathy) is a serious risk and hypoglycaemia is common. Hydrops fetalis may present antenatally as polyhydramnios (excessive amniotic fluid) or postnatally with subcutaneous oedema, pericardial effusion, pleural effusion, ascites and hepatosplenomegaly. The placenta may be thickened.

Other causes of haemolytic disease of the newborn

• Rh system antibodies
• ABO system antibodies
• Kell system antibodies
• Duffy system antibodies (rare)
• MNS and s system antibodies (rare)

Other causes of neonatal jaundice:

• Intrauterine congenital infections (eg, syphilis, cytomegalovirus (CMV), Parvovirus)
• Erythrocyte membrane defects
• Red blood cell enzyme deficiencies
• Enclosed haemorrhages
• Hypothyroidism
• Gastrointestinal obstruction
• Metabolic diseases

Causes of non-immune hydrops fetalis:

• Anaemia
• Cardiac failure from dysrhythmia
• Congenital heart defects

• Indirect Coombs' test should be performed at the first antenatal visit for all rhesus negative mothers. If the test is positive, antibody titres should be monitored with serial samples.
• Antenatal ultrasound may detect signs of hydrops fetalis (see above). Doppler ultrasound of the middle cerebral artery has largely replaced fetal blood sampling as an initial test for the detection of fetal anaemia.⁷
• Fetal blood sampling: if the Doppler scan confirms anaemia, fetal blood sampling should be considered. The sample can be taken at the site of cord insertion or from the hepatic vein. The procedure is done under the guidance of ultrasound imaging. The intrahepatic site causes less fetal distress but is technically more difficult. Fetal loss varies from 1-20% depending on the site of sampling and the condition of the fetus.
• A full blood count shows:
  • Anaemia, increased nucleated red blood cells and other red blood cell abnormalities may be seen.
  • The reticulocyte count can be as high as 40% in severe cases.
  • If there is extensive disseminated intravascular coagulation, schistocytes and burr cells may be observed and neutropenia and thrombocytopenia may occur.
• Biochemical indices should be analysed: hypoglycaemia may be the results of islet cell hyperplasia and hyperinsulinism secondary to the release of metabolic byproducts from lysed red blood cells.

In utero

As soon as the blood samples confirm anaemia, transfusion should be commenced by with Group O negative packed cells cross-matched with maternal blood. This is best done at 18 weeks but samples can be taken 16 weeks if necessary. Intravenous transfusion under ultrasound guidance via the umbilical vein is to be preferred to the intraperitoneal route as the latter is more difficult in a hydropic fetus and causes more complications. Further transfusions should be dictated by serial Doppler scans. Following successful transfusion, delivery should be anticipated between 37-38 weeks. If complications arise, delivery at 32 weeks should be considered. The mode of delivery can be dictated by obstetric considerations.⁸

Future therapy will involve selective modulation of the maternal immune system, making the need for intrauterine transfusions a rarity.⁶

After delivery^1,8

• 50% of babies born to mothers with high maternal antibody titres have normal haemoglobin and bilirubin levels but should be monitored for the onset of late anaemia at 6-8 weeks.
• 25% have moderate disease and may require transfusion. Significant hyperbilirubinaemia may develop within the first 24 hours after birth which may require phototherapy to avoid kernicterus.
• The remaining 25% will have severe disease and either be stillborn or have hydrops fetalis. Live babies will require resuscitation, intensive support, transfusion and correction of acidosis.

The late sequelae of kernicterus (extrapyramidal, auditory and visual abnormalities and cognitive deficit) occasionally occur but are rarely seen with the success of modern treatment.⁸ Other potential complications include late-onset anaemia, graft-versus-host-disease, infections and various metabolic abnormalities. One meta-analysis showed a significant link between feto-maternal rhesus incompatibility and schizophrenia.⁹

Perinatal survival rates of more than 90% have been reported; hydrops fetalis reduces the chance for a viable outcome by up to 11%. Neonatal and infant outcomes are complicated by the need for repeated transfusions secondary to suppressed erythropoiesis. Long-term studies have revealed normal neurologic outcomes in more than 90% of cases.

Anti-D immunoglobulin prophylaxis should be given to all rhesus negative women who have not already been sensitised. It can be given soon after delivery and antenatally at 28 and 34 weeks. Treatment is also indicated after other sensitising events such as abortion, miscarriage, amniocentesis, ectopic pregnancy, abdominal trauma. For further details see Anti-D Immunoglobulin.