When a Mother's Antibodies Injure Her Baby's Liver Before Birth — and How the Next Pregnancy Can Be Protected

Four studies published in 2025 on gestational alloimmune liver disease, and what they mean for families who have lost a baby to newborn liver failure

Gestational alloimmune liver disease is the commonest identified cause of liver failure in newborn babies: the mother's immune system makes antibodies that cross the placenta and injure the baby's developing liver before birth. It is rare and serious, but it can be treated after birth and largely prevented in the mother's next pregnancy. Four studies published in 2025 clarify how to recognise it and how well that prevention works.

None of this happens because anything is wrong with the mother's health, and none of it happens because of anything she did. Understanding that is the starting point, because the rest of this article is about a condition that is recognised too late far more often than it should be — and about what has changed in the past year.

A condition that was named after the wrong thing

For about thirty years this condition was called neonatal haemochromatosis, meaning iron overload in a newborn, because doctors examining affected babies found large amounts of iron in the liver, pancreas, heart and salivary glands [1]. The name stuck, and much of the testing still done today is built around looking for that iron.

What did not fit the name was the family pattern. Mothers often had a healthy baby first, and then lost one pregnancy after another. If the condition were inherited from the parents in the usual way, it would not behave like that — and it did not follow the father at all. When a woman who had lost babies had a new partner, the problem still recurred; when a man's partner changed, it usually did not [2]. That pointed away from an inherited gene and towards something in the mother's immune system. The proof arrived in 2004, when doctors gave women who had lost babies this way repeated infusions of immunoglobulin — purified antibodies given by a drip — throughout their next pregnancy, and healthy babies were born [3]. If blocking an antibody effect prevents the disease, the disease is caused by an antibody.

Later work filled in the biology. Antibodies of the type that normally cross the placenta to give a baby protection against infection are, in this condition, directed at the baby's own liver cells, and they trigger an immune cascade that destroys them [4]. As liver cells are lost, the liver tries to rebuild, forming abnormal small tubes that grow through the tissue and leave scarring behind [5]. The iron turned out to be a side effect rather than the cause: a healthy liver makes a hormone that controls how much iron crosses the placenta, and once liver cells are destroyed that control is lost, so iron floods in and settles in organs that can absorb it but not get rid of it [6]. So the condition was renamed after what actually causes it. One central mystery remains: nobody has yet identified exactly which part of the baby's liver the mother's antibodies recognise, despite years of searching [7].

Why the first affected baby is the hardest to help

For a family that already knows this has happened to them, prevention in the next pregnancy is well established [8], and starting the infusions at around 14 weeks of pregnancy is now standard [9]. Doctors also know what the condition does when it is not treated [10], and they have a treatment for an affected newborn — immunoglobulin plus an exchange transfusion, in which the baby's blood is gradually replaced to wash out the mother's antibodies [11], done early rather than late [12].

The difficulty is the first baby in a family. There is no history to raise suspicion, and the condition does not look the way liver damage usually looks. In older children and adults, a damaged liver releases enzymes into the blood, so a blood test shows high enzyme levels. Here those enzymes are often close to normal — because there are so few working liver cells left to release anything. What is abnormal instead is the baby's blood clotting, which fails badly, and the blood sugar, which falls dangerously low and will not stay up. Platelets are low, and the baby may look unwell from the very first hours.

For decades, standard guidance has been to look for iron deposits elsewhere in the body using a special magnetic resonance imaging scan or a tiny biopsy from inside the lip [13] — an idea first described in 1988 [14] and worked out in a French study of 56 babies who died before or shortly after birth [15]. The trouble is that these tests miss cases: the scan and the lip biopsy each find only about 60 out of every 100 affected babies on their own, and about 80 out of 100 when both are done [16]. A special stain used on liver tissue is also unreliable [17], and ultrasound scans during pregnancy show no consistent warning sign [18]. As a result, the first affected baby in a family is often diagnosed only after death. In one of the 2025 studies, from a hospital in Berlin, Germany, the diagnosis in the first affected babies was made at a median of 15 days of age, and two were diagnosed only at post-mortem examination [19]. Defining newborn liver failure properly, rather than borrowing definitions from older children, helps these babies be identified at all [20], and a review of three decades of cases shows how much survival depends on reaching a diagnosis [21].

What the four 2025 studies found

The first study comes from CHU Sainte-Justine, a children's hospital in Montreal, Canada. Its doctors describe four babies they cared for over about ten years, and the four outcomes could hardly be more different [6]. One baby needed a liver transplant at three months of age and, at eight years old, has normal liver function and is growing and developing normally. A second recovered completely with two doses of immunoglobulin and no transplant. A third got better on her own without any treatment for the liver at all. The fourth, despite immunoglobulin and an exchange transfusion, died of a brain haemorrhage while waiting for a transplant. The mothers of the third and fourth babies each had a later pregnancy with immunoglobulin treatment, and each had a healthy child. The team's point is that this condition is not one fixed illness with one outcome; it ranges from mild and self-correcting to fatal, and thinking of it only as the severe form causes milder cases to be missed.

The second study, from two American children's hospitals — one in Chicago and one in Colorado — asked how to be sure of the diagnosis when liver tissue is available [16]. Four pathologists, working without knowing which case was which, examined liver tissue from 31 newborns who had had liver failure: 11 with this condition and 20 with other causes. They scored six features and then worked out which ones actually told the two groups apart. Three did: scarring through the body of the liver, the abnormal small tubes described above, and a shortage of healthy mature liver cells. Together those three separated the groups with good accuracy, and — importantly — the four pathologists agreed with each other. A low score reliably ruled the condition out; a high score reliably ruled it in. The authors also warn against a specific mistake: when tissue is examined after a baby has died, the abnormal tubes can be dismissed as a change that happens after death. They are not, and dismissing them means missing a diagnosis that could protect the mother's next pregnancy.

The third study is the one from Berlin [19]. Seven women had each already had a baby affected by this condition; five of those seven babies died. In ten later pregnancies, all seven women received immunoglobulin infusions once a week, at a dose based on body weight, starting at around 14 weeks of pregnancy — a median of 23 infusions per pregnancy. Every one of those ten pregnancies resulted in a living baby. Not one baby developed severe liver disease, not one needed an exchange transfusion, and not one died. Two had slightly abnormal blood tests that settled on their own without treatment. The mothers tolerated the infusions well: no woman had a serious side effect and none stopped treatment. The commonest complaints were a headache and tiredness with the first infusion. Four of the ten pregnancies involved some anaemia, mostly from iron deficiency, and one woman developed a rash; in both cases changing to a different immunoglobulin product solved the problem.

The fourth is not a set of results but a plan. Doctors at four hospitals in Japan — the National Center for Child Health and Development, Kanazawa University Hospital, Jichi Medical University Hospital and Kagoshima City Hospital — have published the design of the NH001 trial (immunoglobulin given to mothers during pregnancy to prevent this condition), which they state is the first clinical trial of this treatment anywhere in the world [2]. They deliberately chose not to divide women into a treated and an untreated group, because a treatment reported to work this well should not be withheld; instead, eight women will all be treated and their outcomes compared with records of untreated pregnancies from the past. The study measures whether babies are alive at three months without having needed immunoglobulin, an exchange transfusion, dialysis-like blood filtering, plasma exchange, or a liver transplant. It also builds in a safety step useful in itself: an antibody level is checked before every infusion and the next dose reduced if it is high, because very high levels can thicken the blood and raise the risk of a clot [22].

What families can take from this, and what is still unknown

If you have lost a baby to newborn liver failure, or a baby of yours has been diagnosed with this condition, the single most useful thing to know is that a future pregnancy can very probably be protected, and that this requires planning before pregnancy or very early in it — the infusions need to start at around 14 weeks. It is worth asking directly whether this condition was considered, and if a baby died without a clear cause, whether the liver tissue was examined with this diagnosis specifically in mind.

It is equally worth knowing what these studies do not settle, because honesty here is more useful than reassurance. The chance of the condition coming back in a later pregnancy without treatment is put at somewhere between 70% and 95% depending on which of the four papers you read [16] [6] [2]. How common the condition is overall is genuinely uncertain: two of the 2025 papers, citing the same earlier report, give figures that differ tenfold. A normal lip biopsy does not rule the condition out — all four Montreal babies had normal lip biopsies, including two in whom the diagnosis was later confirmed. A normal pregnancy scan does not rule it out either. And other conditions can look identical in a newborn, including certain viral infections [23] and some inherited metabolic disorders [24], which is why doctors send several sets of tests at once rather than one at a time.

What researchers are working on next falls into three areas. The first is finding the target — the exact part of the baby's liver the mother's antibodies attack — which would make a simple blood test for at-risk mothers possible. The second is testing the two new scoring systems properly: neither the liver-tissue score nor the blood-test score built by the Berlin team [19] has yet been tried in real time on new patients, and the authors of both say so plainly. The third is newer medicines that block the transfer of maternal antibodies across the placenta more precisely than immunoglobulin does, which have already shown promise in a related condition where a mother's antibodies attack her baby's red blood cells [25], and which might one day replace more than twenty weekly infusions with something simpler. If this affects your family, our articles on haemolytic disease of the fetus and newborn and low platelets caused by maternal antibodies cover conditions that work the same way, and support for parents in the neonatal unit may help alongside them.

References

  1. Whitington PF. Gestational alloimmune liver disease and neonatal hemochromatosis. Seminars in Liver Disease. 2012;32(4):325–332. doi:10.1055/s-0032-1329901 ↩
  2. Sasaki A, Yachie A, Mizuta K, et al. Intrauterine high-dose intravenous immunoglobulin therapy during pregnancy for women with a history of pregnancy ending in documented neonatal haemochromatosis (NH001): study protocol. BMJ Open. 2025;15(9):e097145. doi:10.1136/bmjopen-2024-097145 ↩
  3. Whitington PF, Hibbard JU. High-dose immunoglobulin during pregnancy for recurrent neonatal haemochromatosis. Lancet. 2004;364(9446):1690–1698. doi:10.1016/S0140-6736(04)17356-X ↩
  4. Pan X, Kelly S, Melin-Aldana H, Malladi P, Whitington PF. Novel mechanism of fetal hepatocyte injury in congenital alloimmune hepatitis involves the terminal complement cascade. Hepatology. 2010;51(6):2061–2068. doi:10.1002/hep.23581 ↩
  5. Asai A, Malladi S, Misch J, et al. Elaboration of tubules with active hedgehog drives parenchymal fibrogenesis in gestational alloimmune liver disease. Human Pathology. 2015;46(1):84–93. doi:10.1016/j.humpath.2014.09.010 ↩
  6. Helali N, Gagnon H, Álvarez F. Gestational alloimmune liver disease reconsidered: Advocating for a new nomenclature and enhanced diagnosis accuracy. World Journal of Clinical Pediatrics. 2025;14(4):106219. doi:10.5409/wjcp.v14.i4.106219 ↩
  7. Rieneck K, Rasmussen KK, Schoof EM, et al. Hunting for the elusive target antigen in gestational alloimmune liver disease (GALD). PLoS One. 2023;18(10):e0286432. doi:10.1371/journal.pone.0286432 ↩
  8. Whitington PF, Kelly S. Outcome of pregnancies at risk for neonatal hemochromatosis is improved by treatment with high-dose intravenous immunoglobulin. Pediatrics. 2008;121(6):e1615–e1621. doi:10.1542/peds.2007-3107 ↩
  9. Whitington PF, Kelly S, Taylor SA, et al. Antenatal treatment with intravenous immunoglobulin to prevent gestational alloimmune liver disease: comparative effectiveness of 14-week versus 18-week initiation. Fetal Diagnosis and Therapy. 2018;43(3):218–225. doi:10.1159/000477616 ↩
  10. Taylor SA, Kelly S, Alonso EM, Whitington PF. The effects of gestational alloimmune liver disease on fetal and infant morbidity and mortality. Journal of Pediatrics. 2018;196:123–128.e1. doi:10.1016/j.jpeds.2017.12.054 ↩
  11. Rand EB, Karpen SJ, Kelly S, et al. Treatment of neonatal hemochromatosis with exchange transfusion and intravenous immunoglobulin. Journal of Pediatrics. 2009;155(4):566–571.e1. doi:10.1016/j.jpeds.2009.04.012 ↩
  12. Fischer HS, Staufner C, Sallmon H, Henning S, Bührer C. Early exchange transfusion to treat neonates with gestational alloimmune liver disease: an 11-year cohort study. Journal of Pediatric Gastroenterology and Nutrition. 2020;70(4):444–449. doi:10.1097/MPG.0000000000002593 ↩
  13. Feldman AG, Whitington PF. Neonatal hemochromatosis. Journal of Clinical and Experimental Hepatology. 2013;3(4):313–320. doi:10.1016/j.jceh.2013.10.004 ↩
  14. Knisely AS, O'Shea PA, Stocks JF, Dimmick JE. Oropharyngeal and upper respiratory tract mucosal-gland siderosis in neonatal hemochromatosis: an approach to biopsy diagnosis. Journal of Pediatrics. 1988;113(5):871–874. doi:10.1016/S0022-3476(88)80022-2 ↩
  15. Heissat S, Collardeau-Frachon S, Baruteau J, et al. Neonatal hemochromatosis: diagnostic work-up based on a series of 56 cases of fetal death and neonatal liver failure. Journal of Pediatrics. 2015;166(1):66–73. doi:10.1016/j.jpeds.2014.09.030 ↩
  16. Conover KR, Saul S, Chapin CA, et al. Defining a Histologic Scoring System for Gestational Alloimmune Liver Disease. American Journal of Surgical Pathology. 2025;49(12):1326–1333. doi:10.1097/PAS.0000000000002454 ↩
  17. Dubruc E, Nadaud B, Ruchelli E, et al. Relevance of C5b9 immunostaining in the diagnosis of neonatal hemochromatosis. Pediatric Research. 2017;81(5):712–721. doi:10.1038/pr.2017.8 ↩
  18. Staicu A, Popa-Stanila R, Albu C, et al. Neonatal hemochromatosis: systematic review of prenatal ultrasound findings — is there a place for MRI in the diagnostic process? Journal of Clinical Medicine. 2023;12(7):2679. doi:10.3390/jcm12072679 ↩
  19. Gabrysch CH, Dressler-Steinbach I, Verlohren S, Bührer C, Mayer B, Dame C. Maternal Immunoglobulin Treatment of Gestational Alloimmune Liver Disease to Prevent Neonatal Liver Failure: A Single Center Case Series. Transfusion Medicine and Hemotherapy. 2025;52(6):368–378. doi:10.1159/000548476 ↩
  20. Borovsky K, Banc-Husu AM, Saul SA, et al. Applying an age-specific definition to better characterize etiologies and outcomes in neonatal acute liver failure. Journal of Pediatric Gastroenterology and Nutrition. 2021;73(1):80–85. doi:10.1097/MPG.0000000000003103 ↩
  21. Kelgeri C, Kanthimathinathan HK, Couper M, et al. Etiology, characteristics, and outcomes of neonatal liver failure: lessons learned over the last 3 decades. Journal of Pediatrics. 2024;275:114245. doi:10.1016/j.jpeds.2024.114245 ↩
  22. Okada N, Sasaki A, Saito J, et al. The Japanese experience and pharmacokinetics of antenatal maternal high-dose immunoglobulin treatment as a prophylaxis for neonatal hemochromatosis in siblings. Journal of Maternal-Fetal & Neonatal Medicine. 2020;33(1):142–148. doi:10.1080/14767058.2018.1487940 ↩
  23. Ikuse T, Matsui T, Shoji K, et al. Neonatal acute liver failure cases with echovirus 11 infections, Japan, August to November 2024. Eurosurveillance. 2025;30(1):2400822. doi:10.2807/1560-7917.ES.2025.30.1.2400822 ↩
  24. Chan H, Wong R, Fung C, et al. Neonatal acute liver failure due to citrin deficiency (NALFCD). JIMD Reports. 2026;67(4):e70110. doi:10.1002/jmd2.70110 ↩
  25. Moise KJ Jr, Ling LE, Oepkes D, et al. Nipocalimab in early-onset severe hemolytic disease of the fetus and newborn. New England Journal of Medicine. 2024;391(6):526–537. doi:10.1056/NEJMoa2314466 ↩