A 39-year-old, G4P2 (two natural vaginal births), patient with placenta praevia, unbooked, was brought in by an ambulance to a hospital with level 4 nursery, with major antepartum haemorrhage at 32+2/40, which was provoked by coitus. The estimated blood loss prior to arrival in hospital was 1.5 L. This was her second bleed following a small sentinel bleed prior to 30 weeks. She was fluid resuscitated and vaginal bleeding settled on arrival to hospital. On assessment, she had mild tightenings but had no active bleeding on speculum exam. She was hemodynamically stable, with a blood pressure of 108/72 and heart rate of 100, and the CTG showed a reassuring fetal trace.

Intravenous access was secured and fluid resuscitation with 3 L of crystalloid was commenced. A group and screen was requested on arrival, which detected the presence of red cell antibodies. The blood bank issued a warning that delays in obtaining crossmatched blood may occur. She was administered 11.4 mg betamethasone and commenced on a MgSO4 infusion. A plan was made for immediate transfer to a tertiary hospital prior to delivery; however, she developed recurrent moderate PV bleeding accompanied by mild tachycardia and borderline hypotension of 90/50. A plan was therefore made to undergo an emergency caesarean section (CS), with neonatal resuscitation and tertiary transfer following delivery. The Newborn Emergency Transport Service (NETS) was contacted preoperatively in anticipation of the neonate needing possible resuscitation and tertiary transfer.

The emergency CS was conducted under general anaesthesia, with a floppy neonate delivered after seven minutes. Intraoperatively, the placenta was found to be focally adherent with ongoing brisk bleeding after manual removal. There was significant uterine atony. The intra-operative blood loss was estimated to be 2.5 L, with a total estimated blood loss of 4 L. The postpartum haemorrhage did not respond to conservative measures including use of oxytocin, ergometrine, Dienoprost, placental bed sutures and use of Bakri balloon. The patient ultimately required a hysterectomy due to hemodynamic instability and ongoing uncontrolled blood loss.

A massive transfusion protocol was activated, and the patient was given 10 units of cryoprecipitate, two units of fresh frozen plasma and 1 g intravenous tranexamic acid. However, she was found to have a panagglutinin alloantibody, which rendered her incompatible to all red cells in the blood bank, causing a lengthy crossmatch delay of over 60 minutes. Her haemoglobin fell to 37 g/L with metabolic acidosis, and she required metaraminol and intravenous crystalloids to maintain her systolic blood pressure. She was eventually transfused four units of incompatible blood, with no immediate adverse reaction. Subsequent testing found she had a very rare blood group, Jk null, with anti-Jk3 antibodies. She had a mild delayed haemolytic transfusion reaction postoperatively, but did not develop life-threatening intravascular haemolysis.

A floppy male infant weighing 2116 grams was delivered with outlet forceps with Apgars of 2, 8 and 10. He initially required intermittent positive pressure ventilation and chest compressions for resuscitation and was transitioned to continuous positive airways pressure. Further desaturations prompted intubation and NETS transfer to tertiary hospital. The baby was transferred back to base hospital on day 9 of life and discharged home on day 27 of life. He did not have any features suggestive of haemolytic disease of the newborn and has had no significant complications since.


The Kidd blood group system was discovered in 1951 and is composed of two antithetical antigens, Jka and Jkb, along with a third high-incidence antigen, Jk3.1 Jk null, also known as Jk(a-b-), is a very rare blood group where patients lack all Kidd antigens. Jk is a urea transporter found within the renal medulla, and Jk null individuals are unable to maximally concentrate urine.2 If these individuals are exposed to the Jk antigen through transfusion or pregnancy, they are capable of producing antibodies directed against the high-incidence Jk3 antigen. This has serious clinical implications and renders transfusion extremely difficult, as Jk3 is expressed in over 99% of the Australian blood donor pool.

The Jk null phenotype is extremely rare in most ethnic groups, although is reported to be slightly more frequent in Polynesians, with an estimated incidence of 0.1–1.4%.3 In Tonga, where our patient was born, the prevalence of the Jk null phenotype is 1.2% .4

Kidd antibodies, including those directed against Jk3, are considered to be clinically significant. They may cause acute and delayed haemolytic reactions as well as haemolytic disease of the fetus and newborn (HDFN).5 To further complicate matters, these antibodies are often present at very low levels and may be difficult to detect. Antibody identification can be delayed due to the need for specialised testing at a Red Cell Reference Service Laboratory.

Jk antigens have been detected on fetal red cells from 7–11 weeks gestation.6 Given the rarity of anti-Jk3 antibodies in pregnancy, there is limited published experience with HDFN. The largest case series of four maternity patients with anti-Jk3 antibodies described two neonates affected by mild to moderate HDFN, and two unaffected pregnancies.7 There are other case reports of mild HDFN reported in the literature due to maternal anti-Jk3 antibodies.8 All affected neonates were successfully managed with phototherapy, with no transfusion required.

In contrast, although rare, several life-threatening cases of intravascular haemolysis have occurred in patients with anti-Jk3 antibodies following transfusion with incompatible blood,9 and fatal cases have been described. Successful treatment of intravascular haemolysis with therapeutic plasma exchange has been reported.10 Mild delayed haemolytic transfusion reactions have also been reported. As there is no way of predicting the severity of the haemolytic transfusion reaction, giving compatible blood is strongly preferred. In our case, incompatible blood was administered as an immediate lifesaving measure due to the imminent risk of exsanguination.

Without prior knowledge of this antibody, finding compatible blood in an emergency transfusion setting is almost impossible. If the antibody is detected on antenatal screening, Australian Red Cross Lifeblood is able to source compatible units by contacting known Jk null donors, arranging directed donations or thawing frozen units. Close liaison between the blood bank, haematology and obstetrics is vital in order to ensure compatible blood products are available in the event of a life-threatening haemorrhage.

Lessons learned

Jk null is a rare blood group, although its prevalence is increased in Polynesians. Anti Jk-3 antibodies can cause substantial delay in obtaining compatible blood, have the potential to cause life-threatening transfusion reactions, and have also been linked to HDFN. This case highlights the critical importance of obtaining an antenatal antibody screen, particularly in women with risk factors for obstetric haemorrhage. Early detection of antibodies and close liaison with the blood bank are paramount to prevent potentially catastrophic situations.

Dr Lisa Clarke’s commentary on transfusing in the context of a rare red cell alloantibody in the obstetric setting

Red cell alloantibodies can cause haemolytic transfusion reactions and, in the obstetric setting, haemolytic disease of the fetus and newborn. These form in response to exposure to red cell antigens, through pregnancies or transfusion, which are not found on the woman’s red cells. The presence of alloantibodies is uncommon, occurring in less than 2%11 of the population, and are not always clinically significant; identification is therefore imperative to facilitate ‘safe’ transfusion should it be required. This process can be time consuming, particularly in the presence of a rare red cell alloantibody and challenging outside of a reference laboratory. So, while the provision of antigen negative blood is desirable for all patients with antibodies, the Australian New Zealand Society of Blood Transfusion recommends abandoning this at times of ‘massive blood loss if it will result in a delay in the provision of blood products’.12

As the majority of postpartum haemorrhages occur in the absence of identifiable risk factors, women who have alloantibodies identified prenatally should be optimised as per the principals of patient blood management. Those with rare alloantibodies require a transfusion strategy established by the obstetric and haematology teams in collaboration with The Australian Red Cross Lifeblood (Lifeblood).

In most instances when rare blood is required, Lifeblood can source a limited supply of fresh or frozen rare red cells from directed donations or national and/or international donors. There are several limitations to these options. Firstly, these units cannot be supplied in a time critical manner or an unlimited quantity and are therefore inadequate in the event of a massive haemorrhage. Should this occur, the rare red cells should be saved until bleeding has begun to slow.13 Secondly, thawing of frozen red cells is associated with red cell loss of more than 20%, resulting in a smaller haemoglobin increment at the time of transfusion.14 Finally, directed donations increase the risk of transfusion-associated graft-versus-host disease and must be irradiated prior to transfusion, which reduces their remaining lifespan to 14 days.15