Red cell (D) allo-immunisation is a complication of pregnancy that can occur when a pregnant woman does not have the D antigen on her red blood cells (RhD negative) but her fetus does (RhD positive). When fetal blood cells pass into the maternal blood during pregnancy, an immune response can be triggered that produces anti-D antibodies, a process called sensitisation. Anti-D is an immunoglobulin G (IgG) antibody that passes across the placenta from mother to fetus and may result in haemolytic immune destruction of fetal red blood cells. Possible fetal complications from this process may include anaemia, hydrops fetalis and intrauterine death (haemolytic disease in the fetus and newborn (HDFN)) (Hadley and Soothill, 2005). A fetus of a sensitised primigravid woman may not be severely affected, but that becomes increasingly likely in subsequent pregnancies as the immune response becomes stronger. Therefore, women found to have RhD sensitisation should be referred to a fetal medicine unit without delay. The placenta is able to metabolise and transfer the bilirubin produced by fetal haemolysis to the mother during pregnancy, but after birth, haemolysis in the neonate may lead to the development of hyper-bilirubinaemia, and severe jaundice may result in kernicterus brain damage. After a neonate has been affected, the disease severity usually increases in each subsequent pregnancy with a RhD-positive fetus, related to the anti-D concentration in the mother's blood.
Since its introduction in the 1960s, anti-D immunoprophylaxis has greatly reduced the incidence of sensitisation of pregnant women and, therefore, the incidence of HDFN. The policy of antenatal anti-D prophylaxis for all RhD-negative women (National Institute for Health and Care Excellence, 2008) means that almost 40% of RhD-negative pregnant women receive anti-D unnecessarily as they are carrying a RhD-negative fetus; this equates to approximately 40 000 women per year in England and Wales (Soothill et al, 2015). In contrast, the opposite approach is taken at birth, when cord blood is sent for D-grouping and the mother offered postnatal anti-D only if the baby is RhD positive.
Maternal plasma DNA includes free-floating, non-cellular fragments of DNA from the mother and fetus, known as cell-free DNA (cfDNA), which is used in various aspects of prenatal diagnosis and screening (Lo et al, 1998; Soothill and Lo, 2014). Since the fetal cfDNA forms a significant proportion, it is possible to develop reliable testing for a fetal gene signal that is absent in the mother's genome. A test that uses cfDNA has been developed that can identify the fetal RhD status with great accuracy using a maternal peripheral blood sample (Lo et al, 1998; Finning et al, 2002) and this has proved readily acceptable to pregnant women and maternity professionals (Oxenford et al, 2013). This technology was initially used (from about 1999) to guide the management of women who were already allo-immunised and at risk of HDFN (Finning et al, 2002), but more recently a ‘scaled-up’ technique has been described aimed at providing this service to all RhD-negative women (Finning et al, 2008). A National Institute for Health Research-funded multi-centre study investigated the performance of this test and demonstrated the test is reliable after 11 weeks' gestation (Chitty et al, 2011). National fetal RhD testing programmes to direct antenatal anti-D prophylaxis have been introduced successfully in other countries (Clausen et al, 2012; de Haas et al, 2012; Wikman et al, 2012) and this has led to the suggestion that the continuing practice of giving anti-D (a blood product pooled from multiple donors) to healthy pregnant women is ethically unreasonable (Kent et al, 2014).
Avon project
Three NHS Trusts (University Hospitals Bristol NHS Foundation Trust, North Bristol NHS Trust and Weston Area Health NHS Trust) have been offering cfDNA RhD testing to all RhD-negative pregnant women, as an NHS clinical service, since April/May 2013 (Soothill et al, 2015). The maternity services’ ‘booking’ blood test process was unchanged. However, after the maternal blood group was identified as RhD-negative, a new leaflet about this service was added to the previous leaflet about being RhD-negative given to women. The new leaflet explained false positive results leading to administration of anti-D unnecessarily, as was the normal previous policy, and false negative results leading to antenatal anti-D not being given when it currently would have been. The leaflet quoted approximately 1 in 800 as the chance of a false negative result.
At the routine 15–17 week midwifery visit, the community midwives discussed the issues relating to RhD-negative blood group and the content of the information leaflet, answered any questions and then offered the cfDNA test. As with all clinical care, the midwives recorded that this counselling had occurred and boxes to help document the counselling and consent were added to the printed handheld notes. The Trusts' antenatal anti-D immunoprophylaxis guidelines were updated and a flowchart demonstrating the new policy steps was included.
Timing of test
It was decided to use the existing 15–17 week appointment for implementation cfDNA RhD testing, as no additional resources were required. Offering this test at either the 12-week or the 20-week scan appointments was considered. It was decided not to add this to the 12-week dating scan and Down syndrome screening appointment as the woman may not have received the written information prior to this appointment. The anomaly scan appointment was also not used because that would have required extra staff and/or resources. By using the existing 15–17 week appointment, no additional resources were required for this implementation, but the decision about the best timing of the test may be different according to the provision of local services.
Sample handling
When a woman chose to have the test, the blood sample was usually taken during the 15–17 week visit and when this had not been done (e.g. because of late booking or transfer of antenatal care into the service) midwives were asked to offer this until 26 weeks' gestation (stopping then because the result may not be available in time to guide routine prophylaxis). The sample was sent to the midwife's usual hospital pathology laboratory (three laboratories serve the area) with a form documenting the purpose of the sample and giving the estimated date of delivery (EDD) based on the dating scan. The EDD was essential and was reported on the result of the test, because it was used to identify the pregnancy in order to avoid the potential risk that a filed or stored result could be incorrectly ascribed to a possible future pregnancy. For example, if a result from a pregnancy that miscarried was on the computer system that could be misinterpreted.
Laboratory aspects
On arrival at the laboratory, the sample was recorded on the existing pathology computer system and then transferred as a ‘send-away’ sample from each of the hospital pathology laboratories to the NHS Blood and Transplant (NHSBT) laboratory in Filton, Bristol. The established NHSBT transport system was used, so no additional cost was incurred. Test results were available electronically to hospital laboratories using the NHSBT specialist services electronic reporting system (Sp-ICE). When the paper reports were received by the hospital pathology laboratory, the results were entered onto the local Trust's IT system (and so made available to the midwives) and a paper report was also sent to the bases recorded on the original form for filing in the woman's handheld records.
Management of results
There were three possible results reported:
The midwives informed the pregnant women of the result and the associated recommended management of either having or avoiding antenatal anti-D. The laboratory result was filed in the woman's handheld maternity records at her next midwifery appointment. The advice not to recommend anti-D applied both to routine antenatal prophylaxis and if any potentially sensitising events occurred (e.g. vaginal bleeding). However, the information leaflet and counselling were clear that if a woman wanted anti-D despite a RhD-negative result, it could be requested and would then be given.
Delivery
At the delivery of a RhD-negative woman, cord blood is routinely sent for blood grouping and a maternal blood sample is sent for a Kleihauer–Betke test so that if the fetus is RhD-positive and a large feto–maternal haemorrhage is found, extra anti-D can be given. As the Kleihauer–Betke test is a maternal test taken at the delivery, the result in the laboratory computer records allows the date of delivery to be identified and can provide a link to the name and NHS number of the baby; this information allows the cord blood group to be obtained. It was planned that repeat testing by the reference laboratory would be arranged for any cases in which a positive cord RhD group was found after a negative cfDNA result i.e. false negative, but that did not occur in this pilot.
Findings of the pilot and the future
There were no cases in which antenatal anti-D prophylaxis should have been given but were not (i.e. no false negative cfDNA results). However, this policy is likely to result in some women being given anti-D when the fetus is RhD negative, as a result of either false positive or ‘indeterminate—treat as positive’ results when the fetus was negative. About 10% of the test results were ‘indeterminate—treat as positive’, but with advances in the laboratory techniques it is possible that this proportion may fall. However, about 75% of the women with inconclusive results in fact gave birth to positive fetuses and so should have received anti-D. There was one case when the cfDNA result was positive and the cord blood result was negative, but that was as expected, because some people with an RHD gene do not express the RhD protein on the red cells so discrepancies can occur between genetic and serology tests. The consequence of a false positive result is that the woman is given antenatal anti-D unnecessarily, so this is the same as the previous recommended policy, but in much lower numbers.
The high accuracy of this test will raise the question of whether services need to continue cord blood testing of RhD-negative pregnancies when the cfDNA result is negative. Since this study was completed, there has been one case in which the cord blood was positive after a negative cfDNA result. Further investigation showed that the cfDNA RhD test was correct and the RhD-positive cord blood result was incorrect because the result was wrongly transposed from the laboratory records to the IT system. This supports the potential to stop cord blood testing after antenatal cfDNA results and, if this was implemented, there could be a further saving of midwifery and laboratory staff time as well as a cost saving. The three NHS Trusts have not yet taken that step because they require more data on false negative results and whether this policy would result in any such cases not receiving postnatal anti-D as well as antenatal. With cord blood testing, a false negative cfDNA result could be corrected and postnatal anti-D would be given (but this has not happened as yet). Despite the above, it seems likely that postnatal testing will not be considered necessary in the future.
Overall, the three NHS Trusts have reported a fall in the total use of anti-D of about 30%, which was a slightly smaller decrease than the 40% of RhD-negative women with RhD-negative fetuses. That was for the following reasons:
Financial aspects
In a British maternity service with 10 000 births per year, about 1500 women will be RhD negative. As a result of a reduced use of anti-D, the authors estimate a cost saving of about £60 000 per year. Without considering the midwifery time relating to anti-D administration (and so equating that to the time for counselling and blood-taking for cfDNA testing), this relates to a budget of about £60 000/1500 = £40 per test. Our pilot was implemented and continues to run without any additional clinical, laboratory or administrative staff and with the use of existing systems (such as the blood transfusion service sample transport). This could be scaled-up to national numbers, depending on the exact implementation chosen.
Ethics
Irrespective of the financial costs or benefits, it has been argued that it is ethically unacceptable to continue administering a blood product to all RhD-negative women when a fetal RHD genotyping test using maternal blood could identify those women who do not need this product (Kent et al, 2014). Each dose of anti-D is prepared from blood donations from many donors. Although anti-D has been an exceptionally safe product (which should be strongly recommended to those women with a RhD-positive fetus), conditions such as prion diseases should continue to make us ensure blood products are used only when needed. In addition, the availability of anti-D is limited and requires deliberate sensitisation of volunteers. Both the difficulties of availability and the theoretical risks mean it should be used only when required.
Conclusions
Our experience of using cfDNA to guide antenatal anti-D administration within British maternity services has been very positive, with the implementation of the new protocols being well accepted by both staff and women, without additional funding or resources. The procedure for offering cfDNA RhD testing to guide antenatal anti-D administration outlined in this article may be useful in developing protocols as the test becomes available on a wider scale in the NHS. We recommend that this service now be extended more widely in the NHS because it has allowed anti-D (a multi-donor blood product) to be used in a more precise and indicated way, and the cost of the tests can be resourced by the saving in anti-D.