When a Mother's Antibodies Attack Her Baby's Blood: New Research on Predicting Who Will Need the Most Intensive Treatment

Four recent studies on hemolytic disease of the fetus and newborn — how often it happens, what families report going through, and how doctors are learning to plan ahead

In a small number of pregnancies a mother's immune system makes antibodies against her baby's red blood cells, and those antibodies cross the placenta and destroy them. This is hemolytic disease of the fetus and newborn — now rare, at roughly 29 pregnancies per 100,000, thanks to a preventive injection introduced in 1968. Four recent studies examine how often it still happens, what families go through, and how doctors can tell in advance which babies will need the most intensive treatment.

The four studies behind this article

A Dutch national study asked whether a blood test done on the mother during pregnancy can predict whether her baby will need an exchange transfusion after birth — and, separately, counted how many Dutch hospitals can still perform one [1]. A Californian study looked inside one very large health system to see how often the condition occurs and how affected babies actually present [2]. An American survey asked women who had been through these pregnancies what monitoring and treatment they received and what happened [3]. And a review of 238 published studies asked whether researchers around the world are even measuring the same things, so that their results can be compared [4].

What this condition is, and why it became so much less common

Everyone's red blood cells carry markers on their surface, a bit like a set of labels. The best known set is the Rh system, and the most important label in it is called D. If a woman whose cells lack the D label becomes pregnant with a baby whose cells carry it, some of the baby's cells can leak into her bloodstream — usually during birth, sometimes earlier — and her immune system may learn to recognise the D label as foreign. It then makes antibodies against it. In that pregnancy the baby is often unaffected. In a later pregnancy, those antibodies cross the placenta and attack the baby's red blood cells [5].

Two things then follow. Before birth, the baby can become anaemic — short of red blood cells — and if the anaemia becomes severe the baby's heart can fail and fluid can build up in the body, a dangerous state called hydrops. After birth, the broken-down red blood cells release a yellow pigment called bilirubin, which causes jaundice; if bilirubin climbs high enough it can injure the brain [5].

The turning point came in 1968, when an injection of anti-D immune globulin was introduced. Given to women who lack the D label, it prevents their immune system from ever learning to make the antibody in the first place, and over the following thirty-five years it removed most of this disease from countries that could afford it [6]. In the United States, counts of affected births between 1996 and 2010 had already fallen to tens of cases per 100,000 [7]. But the injection only works for the D label. There are more than fifty other red-cell labels that can cause the same disease, and there is no preventive injection for any of them. Antibodies can also still develop when the injection is missed or does not work.

Monitoring and treatment improved alongside prevention. Since 1995, doctors have been able to detect fetal anaemia without putting a needle into the womb, by using ultrasound to measure how fast blood moves through an artery in the baby's brain — anaemic blood is thinner and flows faster [8]. When anaemia is found, the baby can be given a blood transfusion while still in the womb, through a needle guided by ultrasound into the umbilical cord. Together, this monitoring and this treatment have made hydrops almost disappear in expert hands [9]. After birth, most jaundice is now managed with intensive light therapy, which converts bilirubin into a form the baby can excrete [10]. The much more invasive alternative — an exchange transfusion, in which most of the baby's blood volume is slowly replaced with donor blood — has become steadily rarer as the threshold for using it has risen, most recently in a 2022 guideline [11], [12].

That progress created an unexpected problem, and it is the problem the four new studies are about.

The Dutch finding: the tests are better at ruling out than ruling in

The Netherlands is an unusually good place to study a rare disease, because all antibody testing goes through two national laboratories, all transfusions in the womb are done at one hospital, and all blood products come from one national service. That let researchers link every exchange transfusion product ordered in the country over eleven years to the pregnancies behind it — against a background of 1,901,734 births [1].

Two tests were examined. The first, the antibody titre, measures how much antibody is in the mother's blood: her serum is diluted step by step, and the titre is the last dilution at which the antibody can still be detected, written as 1:32, 1:256 and so on. The second is a laboratory test of how much damage her antibodies actually do to red blood cells. Both tracked risk. Among pregnancies with anti-D antibodies, the chance that the baby would need an exchange transfusion rose from about 1 in 100 when the titre was low, to roughly 1 in 5 at the highest titres. The damage test separated risk further: about 1 in 100 babies needed an exchange transfusion when it was low, compared with about 1 in 5 when it was high.

The most useful way to read those numbers is backwards. When both tests were low, the chance the baby would not need an exchange transfusion was around 97–99%. When both were high, roughly four out of five babies still did not need one. So these tests are good for reassurance and poor for prediction — which is exactly how doctors should use them, and exactly what a family should understand if they are told their numbers are high.

Why knowing in advance now matters so much

The second half of the Dutch study explains the urgency. The researchers asked paediatricians at all 72 Dutch hospitals whether they still perform exchange transfusions. Among the general hospitals that replied, only 18 of 49 (37%) still did — compared with 41 of 49 before 2010, a fall of 56%. The reasons were practical: the procedure had become too rare to stay skilled at, staff were short, materials were not stocked, and it takes a long time. Asked how much risk they would accept before delivering such a baby locally, hospitals disagreed sharply — some would deliver any baby and transfer if bilirubin rose, others only if the predicted risk was under 5%.

For a family, this translates into a concrete and reasonable question to ask their team: does the hospital where we are planning to give birth perform exchange transfusions, and if not, what is the plan? It is not a sign of poor care if the answer is no — concentrating a difficult procedure in fewer, more experienced centres is usually safer. But it is a plan worth having before labour rather than during it.

The Californian finding: most affected babies do well

Across 464,711 pregnancies in one Californian health system, 136 were affected by this condition — about 29 per 100,000 [2]. Only 17 of those 136 (about one in eight) needed a transfusion in the womb. What affected babies mostly had was jaundice and earlier birth: they were more likely to be born between 33 and 36 weeks, and about two thirds developed jaundice. Among all 138 babies born, there were no cases of brain injury from bilirubin, no cerebral palsy, and one stillbirth. That is the most important sentence in this article for a family who has just been told their pregnancy is affected: in a well-monitored system, the expected outcome is a jaundiced baby who needs light therapy and goes home well.

The survey finding: what women reported is harder to hear

The third study asked 127 American women about 200 pregnancies [3]. Their reports were less reassuring than the figures from large European centres. Of the 45 pregnancies that involved a transfusion in the womb, 17 had a complication — far above the 3–5% per procedure reported by high-volume centres [13]. Survival among babies transfused in the womb was 84%, compared with 93–97% in those centres. Ten babies died before birth from the disease; five of those were the mother's first affected pregnancy, and two had received no monitoring or treatment at all. The authors' interpretation is not that American medicine is worse, but that this procedure is spread too thinly across too many places for anyone to become expert at it.

Two practical lessons come out of this. One is that a high antibody level does not tell you whether the baby actually carries the label being attacked — in this survey, antibody levels were high in half of the pregnancies where the baby turned out not to be affected at all. The other is that a first affected pregnancy can be severe, so it should not be monitored casually on the assumption that trouble only comes later. There is also real hope in the data: of ten later pregnancies in women who had lost a baby to this disease, eight ended in a live birth, with monitoring usually starting before 16 weeks.

The review finding: researchers are not measuring the same things

The fourth study explains why the figures above are hard to compare with each other [4]. Across 238 published studies, researchers used 104 different measures. Bilirubin — the single most important number in the newborn — was reported in six different ways. Only 29 studies (12%) followed children beyond the newborn period, even though a Dutch follow-up study showed more than a decade ago that development after transfusion in the womb can be measured properly [14]. And only two studies out of 238 asked patients themselves about their experience. The authors are now working towards an agreed list of outcomes that every future study should report.

What researchers are working on next

Four things. The first is a simpler version of the Dutch laboratory test: at present it uses a radioactive material and is performed only in the Netherlands, so the rest of the world cannot use its best predictor. The second is to test the "reassurance" finding properly, by using it prospectively to decide where low-risk pregnancies should deliver. The third is the agreed list of measurements, so that a Dutch result and an American result finally mean the same thing. The fourth is training: as the procedure becomes rarer, keeping enough clinicians competent at it — and building good simulation models, which do not yet exist — becomes its own project.

Read together, these four studies describe a disease that medicine has pushed to the edge of disappearing, and the new set of questions that edge creates: not mainly whether an affected baby can be helped, but whether the right test, the right hospital and the right expertise are lined up before the baby arrives. For most families facing this diagnosis today, the honest summary is that the outlook is good, the monitoring is what makes it good, and the most useful thing they can do is ask early where their care will happen and who will be doing it.

References

  1. de Winter DP, de Haas M, Hulzebos CV, Lukens MV, Prins SA, Been JV, et al.; for the WISE investigators. Antenatal risk estimation for exchange transfusions in neonates with hemolytic disease of the fetus and newborn in a changing treatment landscape: A multicenter, retrospective cohort study. Transfusion. 2026;66(2):296–305. doi:10.1111/trf.70056 ↩
  2. Fassett MJ, Khadka N, Xie F, Shi J, Chiu VY, Im TM, et al. Hemolytic Disease of the Fetus and Newborn in an Integrated Health Care System. Am J Perinatol. 2025;42(15):2038–2048. doi:10.1055/a-2558-7891 ↩
  3. Sherwood MR, Weathersby BM, Granger Howard ME, Markham KB. Patient-Reported Disease Presentation, Interventions, and Outcomes in United States Pregnancies Affected by Alloimmunization at Risk of Hemolytic Disease of the Fetus and Newborn. Children (Basel). 2025;12(7):822. doi:10.3390/children12070822 ↩
  4. Verweij EJT, Lopriore E, Fitzgibbon M, Noack J, Wilke T, Noel W, van Klink J. Reporting of management and outcomes in the hemolytic disease of the fetus and newborn: a systematic literature review. J Perinatol. 2025;45:1513–1519. doi:10.1038/s41372-025-02366-2 ↩
  5. de Haas M, Thurik FF, Koelewijn JM, van der Schoot CE. Haemolytic disease of the fetus and newborn. Vox Sang. 2015;109(2):99–113. doi:10.1111/vox.12265 ↩
  6. Bowman J. Thirty-five years of Rh prophylaxis. Transfusion. 2003;43(12):1661–1666. doi:10.1111/j.0041-1132.2003.00632.x ↩
  7. Yu D, Ling LE, Krumme AA, Tjoa ML, Moise KJ Jr. Live birth prevalence of hemolytic disease of the fetus and newborn in the United States from 1996 to 2010. AJOG Glob Rep. 2023;3(2):100203. doi:10.1016/j.xagr.2023.100203 ↩
  8. Mari G, Adrignolo A, Abuhamad AZ, Pirhonen J, Jones DC, Ludomirsky A, Copel JA. Diagnosis of fetal anemia with Doppler ultrasound in the pregnancy complicated by maternal blood group immunization. Ultrasound Obstet Gynecol. 1995;5(6):400–405. doi:10.1046/j.1469-0705.1995.05060400.x ↩
  9. Zwiers C, Oepkes D, Lopriore E, Klumper FJ, de Haas M, van Kamp IL. The near disappearance of fetal hydrops in relation to current state-of-the-art management of red cell alloimmunization. Prenat Diagn. 2018;38(12):943–950. doi:10.1002/pd.5355 ↩
  10. de Winter DP, Hulzebos C, Van 't Oever RM, de Haas M, Verweij EJT, Lopriore E. History and current standard of postnatal management in hemolytic disease of the fetus and newborn. Eur J Pediatr. 2023;182(2):489–500. doi:10.1007/s00431-022-04724-0 ↩
  11. Kemper AR, Newman TB, Slaughter JL, Maisels MJ, Watchko JF, Downs SM, et al. Clinical practice guideline revision: management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics. 2022;150(3):e2022058859. doi:10.1542/peds.2022-058859 ↩
  12. Ree IMC, Besuden CFJ, Wintjens V, Verweij JEJT, Oepkes D, de Haas M, et al. Exchange transfusions in severe Rh-mediated alloimmune haemolytic disease of the foetus and newborn: a 20-year overview on the incidence, associated risks and outcome. Vox Sang. 2021;116(9):990–997. doi:10.1111/vox.13090 ↩
  13. Zwiers C, Lindenburg ITM, Klumper FJ, de Haas M, Oepkes D, van Kamp IL. Complications of intrauterine intravascular blood transfusion: lessons learned after 1678 procedures. Ultrasound Obstet Gynecol. 2017;50(2):180–186. doi:10.1002/uog.17319 ↩
  14. Lindenburg IT, Smits-Wintjens VEHJ, van Klink JM, et al.; LOTUS study group. Long-term neurodevelopmental outcome after intrauterine transfusion for hemolytic disease of the fetus/newborn: the LOTUS study. Am J Obstet Gynecol. 2012;206(2):141.e1–141.e8. doi:10.1016/j.ajog.2011.09.024 ↩