Non-Immune Hydrops Fetalis

Abnormal fluid accumulation in two or more fetal compartments without red cell alloimmunisation: the common final pathway of chromosomal, monogenic, cardiovascular, lymphatic, haematological, metabolic, infectious, structural and placental disorders, with a diagnostic work-up that now includes fetal exome or genome sequencing.

Literature and regulatory information updated: 17 October 2026.

0. Initial Reference Table

1. Introduction

1.1. From fetal erythroblastosis to non-immune hydrops

The term «hydrops» derives from the Greek ὕδωρ (water) and denotes generalised accumulation of fluid in the fetus [1]. The first systematic characterisation of hydrops fetalis arose in the context of haemolytic disease: in 1932, Louis Diamond, Kenneth Blackfan and James Baty described in the Journal of Pediatrics universal oedema of the fetus as one of the forms of fetal erythroblastosis, alongside severe jaundice of the newborn and anaemia of the newborn [31]. Eleven years later, Edith Potter published in the American Journal of Obstetrics and Gynecology the classic description of «universal oedema of the fetus unassociated with erythroblastosis», which for the first time separated hydrops without haemolytic disease from the picture described by Diamond [32]. In 1970, Macafee, Fortune and Beischer gathered a case series under the designation «non-immunological hydrops fetalis», which fixed the terminology used ever since [33].

The introduction of anti-D immunoglobulin markedly reduced hydrops of immune cause and reversed the proportion between the two groups: in the Western world, the non-immune form now accounts for about 90% of hydrops cases [1, 3]. In the United Kingdom, with about 620,000 live births per year, severe RhD haemolytic disease affects fewer than 100 cases annually [34]. The systematic reviews by Bellini and colleagues estimated the non-immune fraction at 76%–87% (225 articles, 6361 individuals, publications up to 2007) and at more than 85% in the 2007–2013 update (24 publications, 1338 individuals) [35, 36].

1.2. From aetiological classification to genomics

In 1984, Holzgreve and colleagues, in San Francisco, studied 50 pregnancies with fetal ascites and generalised oedema and identified a cause in 84%, with 16% of cases idiopathic; the most frequent causes were cardiac disease, chromosomal abnormalities, congenital malformations, alpha-thalassaemia and twin-to-twin transfusion syndrome. The group proposed a stepwise investigation from the least to the most invasive — ultrasound, echocardiography, full blood count, Kleihauer-Betke test, TORCH group serology (toxoplasmosis, other, rubella, cytomegalovirus and herpes), amniocentesis and fetoscopy — and noted that only arrhythmias and selected cases of urinary obstruction were treatable in utero [37]. Bellini's 2009 systematic review organised the aetiology into 14 categories and attributed 17.8% of cases to unknown cause [35]. In 2015, the Society for Maternal-Fetal Medicine (SMFM) published Clinical Guideline #7, which recommended karyotype and/or CMA even without a structural anomaly, middle cerebral artery (MCA) Doppler and fetal echocardiography, and noted that, even in euploid fetuses, neonatal survival was often below 50% [26].

The following decade shifted the focus of investigation to sequencing. In 2020, Sparks and colleagues published in the New England Journal of Medicine exome sequencing in 127 consecutive cases of unexplained NIHF, with a diagnosis in 29% and a predominance of RASopathies [12]. The International Society for Prenatal Diagnosis (ISPD), the SMFM and the Perinatal Quality Foundation (PQF) had issued in 2018 a joint statement restricting prenatal sequencing to research protocols or selected cases [38]; the American College of Medical Genetics and Genomics (ACMG) published in 2020 its points to consider on the fetal exome [28]; and the ISPD updated its position in 2022, already with yield data by phenotype, among them isolated hydrops [27]. In March 2026, SMFM Consult Series #75, by Sparks, Norton and the SMFM Publications Committee, replaced the 2015 guideline and began recommending exome or genome sequencing after non-diagnostic CMA (GRADE 1C) [1].

1.3. Nomenclature and definitions

Hydrops fetalis is defined by the presence of abnormal fluid in at least two fetal compartments: ascites, pleural effusion, pericardial effusion greater than 2–3 mm and generalised skin oedema with skin thickness greater than 5 mm [1]. The Oxford Desk Reference gives the same ultrasound definition, with skin oedema measured over the skull, and associates with it polyhydramnios and increased placental thickness (more than 6 cm) [3]. Placentomegaly (thickness ≥4 cm in the 2nd trimester or ≥6 cm in the 3rd) and polyhydramnios (maximum vertical pocket ≥8 cm or amniotic fluid index ≥24 cm) are frequent but do not form part of the criteria [1].

NIHF is hydrops that does not result from red cell alloimmunisation [1]. The 2026 document introduces the concept of the «NIHF spectrum» for a single effusion — pleural effusion, pericardial effusion, ascites or skin oedema — attributed to a systemic process; collections of local cause are excluded, such as ascites from a ruptured urinoma or bladder rupture and skin oedema from amniotic bands. The management and prognosis of an isolated effusion are not equivalent to those of hydrops with two or more effusions, and the management recommendations rest on the literature of two or more effusions; the diagnostic work-up, however, is the same [1]. Some genomic series use still broader definitions, which include cystic hygroma and increased nuchal translucency (NT ≥3.5 mm), which has to be taken into account when yields are compared [12, 39]. In the inborn errors of metabolism literature, the definition used by Saudubray (adapted from Collardeau and Guibaud) likewise includes hygroma and increased NT, and treats hydrops as a prenatal «functional sign», distinct from malformations and dysplasias [40].

The Orphanet classification separates the group «hydrops fetalis» (ORPHA:1041) into immune (ORPHA:364013) and non-immune (ORPHA:363999) forms; the entry «idiopathic hydrops fetalis» (ORPHA:89845) has been made obsolete [4, 5]. Monogenic entities with hydrops as the main presentation have their own codes: haemoglobin Bart's hydrops fetalis syndrome (ORPHA:163596), generalised lymphatic dysplasia with NIHF related to PIEZO1 (ORPHA:568062) and hydrops fetalis of lymphatic cause related to EPHB4 (ORPHA:568065) [6–8]. The «hydropic villus» of partial moles in diandric triploidy is a placental change and should not be confused with hydrops fetalis [41].

The clinical scope of this article comprises the definition, the aetiologies by group (cardiovascular, chromosomal, RASopathies, lymphatic dysplasias, haematological, inborn errors of metabolism and lysosomal disorders, infectious, skeletal dysplasias, thoracic, placental and twin-related), the diagnostic work-up with prenatal exome and genome sequencing, fetal and neonatal treatment and prognosis. Noonan syndrome and the other RASopathies, Turner syndrome, lethal skeletal dysplasias and congenital heart disease are described here in the aspects relevant to hydrops. Neonatal liver failure (TRMU, NBAS, neonatal haemochromatosis) enters only as a differential diagnosis.

2. Molecular Genetics and Pathophysiology

2.1. Mechanisms of effusion formation

The mechanism common to all causes is the imbalance between the passage of fluid from the vascular space into the interstitium and its return through the lymphatic circulation: either the production of interstitial fluid increases, or lymphatic dysplasia reduces drainage [1]. On this basis, SMFM Consult Series #75 groups the mechanisms into eight pathways. Right heart failure and volume overload raise central venous pressure and cause secondary hepatic dysfunction (structural heart disease); thoracic masses obstruct venous or arterial flow; tachyarrhythmias shorten ventricular diastolic filling; hepatic failure or hepatic venous congestion cause hypoalbuminaemia (inborn errors of metabolism); infections increase capillary permeability; fetal anaemia causes high-output cardiac failure and extramedullary haematopoiesis, often with hepatic dysfunction (hereditary and acquired anaemias); and congenital nephrosis reduces oncotic pressure through protein loss [1]. The document acknowledges that the pathophysiology is not elucidated for every cause [1].

The Oxford Desk Reference, drawing on Bellini and Hennekam, summarises the mechanisms in three pathways: fetal hypomobility, congestive heart failure and obstruction of lymphatic flow [3]. Hypomobility explains hydrops in the fetal akinesia sequences, in which the absence of movement impairs lymphatic propulsion; lymphatic obstruction explains cystic hygroma, particularly associated with monosomy X [3, 42]. In trisomy 21 and monosomy X, the effusions are attributed to abnormal communications between the lymphatic and venous systems [1]. Tissue distension by fetal lymphoedema may leave permanent marks even when the oedema resolves: webbed neck, redundant nuchal folds, hypertelorism, epicanthic folds, low-set ears and widely spaced nipples; some cardiac defects, such as coarctation of the aorta and hypoplastic left heart, have been interpreted as a sequel of fetal lymphoedema [3].

In inborn errors of metabolism, the proposed mechanisms are visceromegaly with obstruction of venous return, decreased erythropoiesis with anaemia and hypoproteinaemia [1]. In lysosomal disorders, storage in the placenta and liver is added: in Niemann-Pick disease type A, lipid storage in the liver, brain, kidney and placenta begins before birth, the enlarged placenta is visible on ultrasound and placental storage is thought to contribute to fetal loss [43]. In skeletal dysplasias the mechanism is not elucidated; one hypothesis is hepatomegaly from intrahepatic proliferation of haematopoietic precursors, compensating for a small marrow cavity, with compression of the great vessels [1]. Large placental chorioangiomas behave as high-volume arteriovenous shunts and cause high-output cardiac failure, with greater risk from 4–5 cm upwards [1]. Arteriovenous malformations (RASA1, EPHB4) and sacrococcygeal teratoma act through the same high-output mechanism [44, 45].

2.2. Aetiological architecture

The aetiological distribution depends on the era, the method of investigation and the population. The table brings together the two Bellini systematic reviews and the 2026 SMFM estimate; the categories do not fully coincide and overlap.

Sources: [1, 35, 36]. The Oxford Desk Reference reproduces the 2009 review with 5437 cases grouped into ten categories (cardiovascular 1181, 21.7%; chromosomal 727, 13.4%; haematological 564, 10.4%; infections 366, 6.7%; thoracic 327, 6.0%; lymphatic dysplasia 310, 5.7%; twinning 304, 5.6%; syndromic including inborn errors 297, 5.5%; miscellaneous 395, 7.2%; unknown 966, 17.8%), a denominator that differs from the 6361 individuals of the original abstract without the consulted sources explaining the difference [3, 35]. The authors of the 2015 update warn that the small numbers per category make firm conclusions risky [36].

The SMFM also estimates the contribution of genetic disorders within the NIHF spectrum, with values extrapolated largely from sequencing studies performed after non-diagnostic karyotype and CMA [1]:

Source: [1].

A 2021 systematic review identified 131 genes with strong evidence and 46 with emerging evidence of association with NIHF; several genes previously attributed to hydrops had no cases with a molecular diagnosis, and many genes with strong evidence were not included in the commercial panels of the time [13]. In exome series, the inheritance of the diagnoses varies with selection: in the Sparks cohort, 68% (25/37) were autosomal dominant (88% de novo and 12% inherited), 27% (10/37) autosomal recessive (95% inherited), one X-linked recessive inherited and one of uncertain inheritance [12]; in the Mone meta-analysis (89 diagnoses), 57.3% were autosomal dominant, 86.3% of them de novo [14]; in the 2023 Al-Kouatly meta-analysis, which weights consanguineous populations more heavily, recessive inheritance accounted for 55%, dominant for 41% and X-linked for 4% [11].

2.3. RASopathies

RASopathies result from variants that activate the RAS–MAPK (mitogen-activated protein kinase) pathway and constitute the most frequent monogenic group in exome series of NIHF: 30% of diagnoses in the Sparks cohort, 30.3% (27/89) in the Mone meta-analysis, with PTPN11 in 44.4% of these, and 24% in the 2023 meta-analysis [11, 12, 14]. The SMFM, citing these studies, estimates RASopathies at about one third of euploid pregnancies with NIHF that proceed to sequencing [1]. The genes described in hydropic fetuses include PTPN11, RAF1, RIT1, RRAS2, SOS1, BRAF and HRAS [15, 16, 46]. The lymphatic anomaly of Noonan syndrome manifests prenatally as increased NT or hygroma, chylothorax and hydrops, and after birth as oedema of the dorsum of the hands and feet; in the compilation by Brouillard, Boon and Vikkula reproduced by the Oxford Desk Reference, 54% of cases with a PTPN11 variant and 63% with a SOS1 variant had lymphoedema, whereas chylothorax and chylous ascites were of low penetrance in variants of KRAS, RAF1 and HRAS [47, 48]. Nuchal thickening with lymphatic anomaly is more frequent with PTPN11 variants [49]. In Costello syndrome (HRAS, variants at codons 12 and 13, NM_005343.4, such as c.34G>A p.(Gly12Ser)), the increased birth weight reflects hydrops and not true macrosomia; severe cases combine hydrops, postnatal pulmonary effusions and progressive hypertrophic cardiomyopathy [49, 50].

2.4. Lymphatic dysplasias

Non-RAS congenital lymphatic anomalies account for 2%–4% of pregnancies with effusions, and lymphoedema-distichiasis, Milroy disease and generalised lymphatic dysplasia may manifest in utero with one or more effusions, increased NT or hygroma in early gestation [1].

The PIEZO1 gene (16q24.3) encodes a mechanosensitive cation channel and underlies two pictures with opposite mechanisms. In hereditary dehydrated stomatocytosis type 1 (xerocytosis), of autosomal dominant inheritance, gain-of-function variants delay inactivation of the channel, prolong the efflux of potassium and water and dehydrate the red cell, with non-immune haemolytic anaemia, raised mean corpuscular haemoglobin concentration and decreased osmotic fragility; perinatal oedema, ascites and effusions may occur and bear no relation to the fetal haemoglobin value [9]. In generalised lymphatic dysplasia with NIHF (lymphatic malformation 6, OMIM 616843), of autosomal recessive inheritance, biallelic loss-of-function variants deprive the endothelium of the mechanosensor that responds to shear stress and directs vessel formation, and the lymphatic structures do not develop normally; the result is uniform lymphoedema of all segments, with intestinal and pulmonary lymphangiectasia, pleural effusions, chylothorax and pericardial effusions, prenatal presentation as NIHF in a high proportion of cases and facial and four-limb lymphoedema in survivors (Fotiou and colleagues, 2015) [7, 9]. The same channel carries the Er blood group system [9].

The EPHB4 gene (7q22.1) encodes a receptor tyrosine kinase of the ephrin receptor family, which binds mainly to ephrin-B2 and participates in defining the arteriovenous boundary and in angiogenesis [10]. Heterozygous variants that inactivate the kinase cause lymphatic malformation 7 (OMIM 617300, «non-immune hydrops fetalis and/or atrial septal defect»), described by Martin-Almedina and colleagues in 2016 in two families: some carriers had severe NIHF, lethal in the perinatal period or resolving after the neonatal period, and others only an atrial septal defect or varicose veins in adulthood, which reflects variable expressivity [8, 10]. EPHB4 variants also cause capillary malformation–arteriovenous malformation type 2 (CM-AVM2), with a penetrance of 93% (102/110), and have been described in a four-generation family with central conducting lymphatic anomaly [10, 44]. OMIM classifies entry 617300 as a susceptibility [10].

Hennekam syndrome (generalised lymphatic dysplasia with lymphangiectasia) is autosomal recessive and genetically heterogeneous. CCBE1 explains about a quarter of cases according to Smith's atlas and 13 of 29 in a series by Alders (type 1), and was not found in 158 patients with milder lymphatic dysplasia; FAT4 explains about 20% (5 of 24 families; type 2, allelic with Van Maldergem syndrome 2); ADAMTS3 has been described in at least two families (type 3) [51–53]. The CCBE1 protein is required for the protease ADAMTS3 to convert pro-VEGF-C (vascular endothelial growth factor C) into the mature form that activates the VEGFR3 receptor, which places the three genes in the same pathway [51, 52]. With CCBE1, lymphoedema is always present from birth; with FAT4 it may begin in childhood. Connell and colleagues concluded that CCBE1 variants may cause mild atypical forms of generalised lymphatic dysplasia, but are not a frequent cause of NIHF [51].

In lymphoedema-distichiasis, of autosomal dominant inheritance from variants in FOXC2 (16q24), NIHF and prenatal hydrothorax are rare complications; about 75% of patients have an affected parent and 25% a de novo variant [20]. In Milroy disease, autosomal dominant with a penetrance of about 90% and caused by inactivating variants in FLT4 (VEGFR3, 5q), ultrasound may show oedema of the dorsum of the feet in the 2nd or 3rd trimester and mild pleural effusions that regress; hydrops is very rare and generalised lymphatic anomalies are rare [21, 48]. Isolated fetal chylothorax is associated with ITGA9 variants (six index cases in Brouillard's compilation), and Orphanet associates with NIHF the genes ANGPT2 (loss of function), CALCRL and THSD1 [4, 48]. Recessive truncating variants in THSD1 cause hydrops, polyhydramnios and ascites, at times with haemangiomas and intracranial aneurysms (four patients from a consanguineous family described by Abdelrahman in 2018) [40].

2.5. Haemoglobinopathies and hereditary anaemias

Alpha-thalassaemia is the most frequent hereditary haematological cause of fetal effusions [1]. The HBA1 and HBA2 genes (16p13.3) form, on each chromosome 16, a pair of alpha-globin genes; the α⁰ carrier has both genes deleted on the same chromosome (αα/−−, deletion in cis) and the α⁺ carrier has only one gene deleted or inactivated (αα/−α or αα/αᵀα) [54]. Carriers of deletions in cis are frequent in the populations of South-East Asia, the Philippines and the Mediterranean [1]. Haemoglobin Bart's hydrops fetalis syndrome results from the functional absence of all four alpha genes (−−/−−), usually from large deletions on both chromosomes and, rarely, from a non-deletional variant (for example, −−/αᵀ−) [17]. Without α chains, the γ chains form γ4 tetramers (haemoglobin Bart's), with extremely high affinity for oxygen and unable to release it in the tissues; the fetus suffers severe hypoxia, anaemia, congestive heart failure and effusions, with extramedullary erythropoiesis, hepatosplenomegaly and a massive placenta [1, 17]. The Oxford Desk Reference describes haemoglobin Bart's as tetramers of β-globin; this is an error, because haemoglobin Bart's is γ4 and haemoglobin H is β4 [17, 54].

Intrauterine survival depends on preservation of the embryonic gene HBZ (zeta-globin). The −−SEA deletions spare HBZ, and the −−SEA/−−SEA fetus maintains 10%–20% haemoglobin Portland (ζ2γ2), which allows it to reach the 3rd trimester; deletions that include HBZ, such as −−FIL, lead to death from hypoxia and heart failure in utero or shortly after birth [17]. In the international registry of the University of California, San Francisco (UCSF), the 11 patients who survived without intrauterine transfusion had genotypes with zeta-globin preserved [55]. The designations −−SEA, −−FIL, −−THAI and −α3.7 are historical deletion nomenclature and were not normalised to HGVS in the sources consulted. Prenatal diagnostic errors from rare α⁰ alleles have been described: in Thailand, a −−SA/−−SEA fetus was initially interpreted as −−SEA/−−SEA because the father had the −−SA/−α genotype and the mother HbH-Queens Park disease (−−SEA/αα^QP), which shows that molecular panels should include the rare alleles of each population [56].

Other hereditary anaemias with hydrops act through the same high-output cardiac failure pathway. Diamond-Blackfan anaemia, from defects of ribosomal proteins, has a spectrum ranging from mild anaemia to severe fetal anaemia with NIHF; X-linked cytopenia from missense variants in GATA1 may cause hydrops requiring intrauterine transfusion, hypospadias and cryptorchidism; congenital dyserythropoietic anaemia type I occasionally causes severe fetal anaemia with hydrops [57–59]. In the KLF1 gene, the heterozygous variant NM_006563.5:c.973G>A p.(Glu325Lys) causes congenital dyserythropoietic anaemia type IVa, with severity ranging from moderate anaemia to fetal anaemia with hydrops and lifelong transfusion dependence; the recessive form (IVb) causes severe haemolysis from the first day, fetal haemoglobin of 18%–47% and small amounts of haemoglobin Bart's; loss-of-function heterozygotes have only the In(Lu) phenotype and hereditary persistence of fetal haemoglobin [60]. In a systematic review of 41 exome studies, hereditary anaemias explained 2.4% (5/207) of diagnosed NIHF, with the genes SEC23B, SPTA1, KLF1, RPL11, UNC13D and RFWD3 [61]. Congenital erythropoietic porphyria (UROS) may present as NIHF when uroporphyrinogen III synthase activity is below 1% [62].

2.6. Inborn errors of metabolism and lysosomal disorders

NIHF and severe growth restriction are frequent in inborn errors that disturb the synthesis or catabolism of complex molecules — lysosomal and peroxisomal disorders and defects of cholesterol biosynthesis — and in defects of intracellular trafficking, among them the RAS–MAPK pathway; «intoxication» disorders, of amino acid and organic acid catabolism, do not as a rule interfere with embryofetal development [40]. The accumulation of glycosaminoglycans (GAG) begins before birth, and in mucopolysaccharidosis type VII (MPS VII) hydrops is probably the most frequent presentation [63].

MPS VII (Sly disease) results from deficiency of β-glucuronidase, encoded by GUSB (7q21–q22; 12 exons; 651 amino acid polypeptide) [64]. Nonsense variants and deletions are associated with the severe phenotypes, and residual activity above 1.4% is associated with later onset and longer survival [18]. In a fetus with NIHF from the Sparks study, a homozygous deletion of exon 9 was identified [18]. GeneReviews describes a pseudodeficiency allele, c.454G>A p.(Asp152Asn), which reduces enzyme activity in vitro without causing disease and complicates enzymatic prenatal diagnosis and heterozygote detection; the same variant appears in Nyhan's Atlas in the historical nomenclature D152N, described in a mother with markedly reduced activity and no disease [18, 64]. There are several GUSB pseudogenes, which requires care in sequence analysis [18].

Galactosialidosis results from deficiency of the protective protein/cathepsin A (CTSA, 20q13.1), which stabilises β-galactosidase and neuraminidase; the early infantile form causes fetal and neonatal oedema, hepatosplenomegaly, dysostosis multiplex, heart failure and proteinuria, with death at a mean age of seven months [65]. Infantile free sialic acid storage disease (ISSD) and Salla disease are allelic and result from variants in the lysosomal sialic acid transporter, sialin, encoded by SLC17A5; the dysostosis multiplex table in Spranger's atlas records the gene as SLC7A5, a typographical error that should not be reproduced [63, 66, 67]. Sialidosis (NEU1), GM1 gangliosidosis (GLB1), mucolipidosis II (GNPTAB), Gaucher disease (GBA1), Niemann-Pick disease A (SMPD1) and C, Wolman disease (LIPA) and Farber disease (ASAH1) complete the lysosomal group [22, 63, 68]. In congenital disorders of glycosylation (CDG), especially PMM2-CDG, hypoproteinaemia and hepatic involvement explain the hydrops [69, 70].

Outside the lysosomal group, transaldolase deficiency (TALDO1, pentose phosphate pathway) causes hydrops with severe anaemia and hepatic involvement; S-adenosylhomocysteine hydrolase deficiency causes liver disease with coagulopathy, probably responsible for the hydrops in two siblings; sphingosine-1-phosphate lyase deficiency (SGPL1) causes steroid-resistant nephrotic syndrome, which may begin as NIHF, with adrenal insufficiency and adrenal calcifications; and respiratory chain defects (Pearson syndrome, LARS2, AARS2), intracellular cobalamin C deficiency, Barth syndrome (TAZ) and Greenberg dysplasia (LBR, 1q42.1, 3β-hydroxysterol Δ14-reductase; MIM 215140) complete the list [71–80].

2.7. Nomenclature of the cited variants

The following variants appear in the sources consulted. Only the PTPN11 variant was verified in VariantValidator and ClinVar; the remainder are reproduced with the transcript indicated by GeneReviews or with the nomenclature of the source, without manual conversion.

Nyhan's Atlas further describes, in a fetus with ascites and hydrops from galactosialidosis, a variant designated «Q49G» at position 146, an inconsistent notation that could not be normalised and that is not reproduced here as a validated variant [65].

2.8. Allelic disorders and modifiers

Several NIHF genes have allelic disorders with a very different postnatal phenotype. In PIEZO1, dominant gain of function causes xerocytosis and biallelic loss of function causes generalised lymphatic dysplasia [9]. In LBR, biallelic loss causes Greenberg dysplasia, lethal with hydrops, and heterozygosity causes the Pelger-Huët anomaly, benign, with nuclear hypolobulation of granulocytes [79]. In GLB1, the variants that cause GM1 gangliosidosis type I with hydrops are distinct from those that cause Morquio B disease [84]. In SLC17A5, ISSD (severe, infantile) and Salla disease (slowly progressive) are allelic [63]. In FAT4, Hennekam syndrome type 2 and Van Maldergem syndrome 2 are allelic, and on 20 February 2025 ClinGen classified «FAT4-related neurodevelopmental disorder» as definitive for autosomal recessive inheritance [53]. In EPHB4, lymphatic malformation with hydrops and CM-AVM2 coexist in the same families [10]. In galactosialidosis, the juvenile/adult form accounts for 70% of patients, and the phenotypic spectrum of GUSB, NEU1, GBA1 and PMM2 ranges from lethal hydrops to mild adult forms [18, 22, 63, 65, 70].

3. Clinical Spectrum and Manifestations by System

3.1. Fetal presentation and general natural history

NIHF presents as an ultrasound finding in any trimester. In the 1st trimester, increased NT and cystic hygroma are the early forms of the same process; in the 2nd and 3rd trimesters, ascites, pleural and pericardial effusions, skin oedema, placentomegaly and polyhydramnios appear [1, 3]. The natural history without intervention is unfavourable: 15%–21% of pregnancies with NIHF end in miscarriage or fetal death, and in series with effusions or hygroma in the 1st and early 2nd trimester spontaneous loss reached 62%–86%, partly because of the higher prevalence of aneuploidy in that group [1]. Polyhydramnios occurs in up to 59% and preterm birth in up to 86%, with a median gestational age at delivery of 32–36 weeks in contemporary cohorts [1]. In the prospective UCSF cohort, with 56 cases studied by exome sequencing, the median delivery was at 32.8 weeks (interquartile range 30.3–35.0) and 86% (48/56) were born before 37 weeks: 38% from spontaneous labour, 19% for maternal indication, chiefly pre-eclampsia, and 44% for fetal indication (non-reassuring testing or worsening effusions); neither the genetic diagnosis nor polyhydramnios was associated with preterm birth [24].

Neonatal data have a different denominator. In the Pediatrix national neonatal intensive care unit (NICU) database in the United States (162 NICUs, 253,651 discharges), 598 newborns had hydrops of any cause; the associated diagnoses were congenital heart disease (13.7%), rhythm disturbances (10.4%), twin-to-twin transfusion (9%), other congenital anomalies (8.7%), chromosomal abnormalities (7.5%), viral infections (6.7%), congenital anaemia (5%) and congenital chylothorax (3.2%); 215 died before discharge [85]. These figures describe the NIHF that reaches the NICU alive and not NIHF diagnosed in the fetus.

3.2. Chromosomal causes

Aneuploidies, chiefly monosomy X and trisomy 21, account for 13%–30% of all presentations in pregnancy; trisomy 13, trisomy 18 and triploidy complete the group [1]. In Turner syndrome, many fetuses are now identified by oedema or hydrops on the 12-week ultrasound; when detection is through hydrops, the probability of spontaneous loss is high, and when monosomy X is detected incidentally without hydrops survival is longer and the phenotype milder; webbed neck, ptosis and low posterior hairline are sequelae of fetal oedema [86]. In trisomy 21, transient abnormal myelopoiesis, which affects about 10% of infants, may cause fetal anaemia and effusions [1]. Sex chromosome mosaicism may be detected during the investigation of hydrops, and when the initial investigation was by array the laboratory generally recommends a karyotype [87]. Mosaic tetrasomy 12p (Pallister-Killian syndrome) combines diaphragmatic hernia, hydrops and relatively large fetal size; it is detected in amniocytes and fibroblasts, rarely in fetal blood [88]. In diandric triploidy, symmetrical growth restriction with structural anomalies (82%), a large placenta with partial molar changes and raised maternal hCG (80%) predominate [41].

3.3. RASopathies

Noonan syndrome, cardiofaciocutaneous (CFC) syndrome and Costello syndrome manifest in utero through increased NT, hygroma, pleural effusion, chylothorax, skin oedema, polyhydramnios, renal anomalies, heart disease and arrhythmias, up to generalised NIHF [1, 19]. Hygroma with scalp oedema, polyhydramnios, effusions, ascites or frank hydrops should suggest Noonan syndrome; heart disease is diagnosed prenatally in about 20% of cases, and pulmonary stenosis is difficult to detect on fetal echocardiography [19, 49]. Birth weight may be normal or increased because of oedema [49]. In CFC syndrome, polyhydramnios, fetal oedema and chylothorax are characteristic, and prenatally macrocephaly, macrosomia with relatively short femurs and increased nuchal fold are frequent [49, 89]. In Costello and CFC syndromes, supraventricular or chaotic atrial tachycardia is generally accompanied by hydrops [90]. The bleeding diathesis of Noonan syndrome (partial deficiencies of factors XI, XII and VIII, von Willebrand disease, platelet dysfunction and thrombocytopenia), present in 50%–89%, is relevant at delivery and during neonatal drainage procedures [47]. In low-risk populations, the study of 309 pregnancies with isolated NT above the 99th centile identified four cases of Noonan syndrome, and in the HYDROPS study Noonan syndrome was diagnosed by trio exome sequencing in four of 44 pregnancies with NIHF [19].

3.4. Lymphatic dysplasias

In generalised lymphatic dysplasia from PIEZO1, NIHF is the prenatal presentation in a high proportion of cases; survivors develop facial and four-limb lymphoedema, intestinal and pulmonary lymphangiectasia, pleural and pericardial effusions and chylothorax [9]. Orphanet records about 10 published cases, with onset between the prenatal period and childhood [7]. In EPHB4-related hydrops (two families), NIHF may be lethal in the perinatal period or resolve after the neonatal period, and other carriers in the same family have only an atrial septal defect or varicose veins [8, 10].

Hennekam syndrome combines distal, facial and genital lymphoedema with lymphangiectasias of the intestine and of other organs (pleura, pericardium, thyroid, skin, kidney), with chylothorax, ascites and pericardial effusion; severe hydrops in utero may cause prenatal or early postnatal death, and development ranges from normal to severe impairment, even within the same family [51]. The severity of CCBE1 disease ranges from NIHF to a mild form with segmental intestinal lymphangiectasia [52]. In lymphoedema-distichiasis, NIHF and prenatal hydrothorax are rare; if the fetus or newborn survives, the hydrops may resolve completely, and it has been proposed that hydrops and respiratory failure result from severe pulmonary lymphangiectasia; a five-generation family with lymphoedema has been described in which the youngest generation progressed to hydrops [20]. The lymphoedema of this syndrome generally appears between 5 and 20 years of age (80% by 30), with distichiasis in 95%, ptosis in 31%, congenital heart disease in 6.8% (10% in another compilation), cleft palate in 4% and early varicose veins in 49% [48, 91]. In Milroy disease, oedema of the dorsum of the feet may be visible in the 2nd or 3rd trimester, mild pleural effusions frequently regress and hydrops is very rare [21]. In Parkes Weber syndrome and in RASA1-related CM-AVM, NIHF results from an arteriovenous malformation and is an occasional manifestation [44, 92].

Examination of the surviving newborn may show the markers of intrauterine oedema — epicanthic folds, low-set ears, broad or webbed neck, redundant nuchal folds, widely spaced nipples — and Stemmer's sign (inability to pinch the skin at the base of the second toe), pathognomonic of lymphoedema [48].

3.5. Cardiovascular causes

Cardiovascular anomalies explain up to 20% of NIHF cases [1]. The most frequent are right heart lesions, cardiomyopathy and hypoplastic left heart syndrome; next come arrhythmias and large cardiac tumours [1]. In hypoplastic left heart syndrome, at least 21% of fetuses have extracardiac anomalies, including hydrops, against 10%–15% in postnatal series, and 12% have a chromosomal abnormality (6% when the defect is isolated) [93]. The most commonly associated tachyarrhythmias are supraventricular tachycardia (SVT) and atrial flutter; rarely, long QT causes ventricular tachycardia or torsades de pointes; heterotaxy may cause tachy- or bradyarrhythmias; the most frequent bradyarrhythmia is congenital atrioventricular block from maternal anti-Ro/La (SSA/SSB) antibodies; and conduction genes (SCN5A, GATA4) and RASopathies (HRAS) may underlie it [1]. Some clues point to a unifying cause: pulmonary stenosis suggests a RASopathy, rhabdomyomas suggest tuberous sclerosis and myocarditis suggests parvovirus [1]. For rhabdomyomas, the Toronto series showed rapid growth between 20 and 28 weeks (2.58 mm/week) and slow growth thereafter (0.44 mm/week); in the four cases with hydrops, the mean tumour diameter was 31.7 ± 6.2 mm, the tumour/femur ratio 0.68 ± 0.15 and growth 4.3 ± 1.7 mm/week [94]. In Barth syndrome, prenatal complications include growth restriction, oligohydramnios, ventricular dysfunction and hydrops [78]; in AARS2-related disease, hydrops occurs in 11% and hypertrophic cardiomyopathy dominates the picture [76]; in intracellular cobalamin C deficiency, the in utero presentation includes hydrops, cardiomyopathy (including non-compaction) and growth restriction [77]; and in early-onset generalised arterial calcification of infancy (GACI), 28% of patients had hydrops fetalis, 44% heart failure and 38% polyhydramnios [95].

3.6. Haematological causes

Alpha-thalassaemia accounts for 28%–55% of NIHF cases in South-East Asia, according to series from southern China and southern Thailand [1]. Haemoglobin Bart's hydrops fetalis syndrome is the most severe form: without treatment, it ends in stillbirth between 30 and 40 weeks or in death shortly after birth, with generalised oedema, pleural and pericardial effusions from anaemia-induced heart failure, hepatosplenomegaly, extramedullary erythropoiesis and a massive placenta [17]. At 13–14 weeks, the at-risk fetus may show increased nuchal thickness, a thickened placenta, increased MCA velocity and an increased cardiothoracic ratio; at 22–28 weeks, generalised oedema, ascites and effusions [17]. Deficient brain growth, hydrocephalus and cardiovascular and urogenital malformations have been described, and hypospadias is frequent in male fetuses [17, 30]. Pregnancy with an affected fetus carries a high maternal risk of pre-eclampsia and of other obstetric complications because of the enlarged placenta [54].

Diamond-Blackfan anaemia may cause severe fetal anaemia with NIHF and is associated with limb anomalies and orofacial clefts; haematological complications appear in 90% within the 1st year and malformations in up to 50%, with unpredictable expression within the same family [1, 57]. Dehydrated stomatocytosis (PIEZO1, gain of function) and the congenital dyserythropoietic anaemias complete the group of hereditary anaemias [1, 9, 59]. GATA1-related cytopenia combines hydrops, hypospadias, cryptorchidism and chronic haemolysis, findings described only with missense variants [58]. Acute or chronic isolated fetomaternal haemorrhage causes hypovolaemia, effusions and fetal death; the diagnosis has consequences because delivery or, in selected cases, intrauterine transfusion may be indicated [1]. Transaldolase deficiency and Pearson syndrome combine hydrops and anaemia [40, 80].

3.7. Inborn errors of metabolism and lysosomal disorders

Inborn errors of metabolism account for 1%–5% of NIHF cases, presenting as an isolated effusion, chiefly ascites, or as generalised hydrops; the associated findings include hepatomegaly, splenomegaly, cardiomyopathy, epiphyseal stippling and growth restriction [1]. In most cases there are other signs: facial dysmorphism, malformations, visceral or neurological manifestations, lactic acidosis or hepatic haemosiderosis [40]. In a systematic review of 22 series with 2678 NIHF cases (1979–2020), the frequency of lysosomal disorders was 6.6% (177/2663) among cases tested for any lysosomal disorder and 8.2% (177/2151) in idiopathic NIHF; the most frequent were MPS VII, galactosialidosis, ISSD, Gaucher disease, GM1 gangliosidosis and sialidosis, and more than 40% of the commonest lysosomal disorders had potential postnatal treatment [96]. In Makhamreh's meta-analysis of prenatal exome sequencing (41 studies, 2000–2024), lysosomal disorders accounted for 5% (27/558) of all sequenced NIHF cases and 13% (27/207) of those diagnosed; MPS VII was the most frequent (14/27, 52%), 68% of lysosomal NIHF cases were isolated, consanguinity was present in 57% (12/21) and recurrence of hydrops was 75% (15/20) where reported [97]. Whybra and colleagues described transient forms of NIHF in MPS IVA, in MPS VII and in Niemann-Pick disease C, often with a history of NIHF in an unstudied sibling [98].

Nyhan's Atlas illustrates the two most frequent lysosomal presentations: Figure 82.4 shows, according to the legend, a six-week-old infant with MPS VII, fetal and neonatal ascites, macrocephaly, hepatosplenomegaly, bilateral inguinal hernias and gibbus, and Figures 97.1 and 97.2 a newborn with galactosialidosis and NIHF, with extreme vulval oedema, carrying the historical CTSA variant V132M [64, 65]. Storage disorders may mimic one another and the CDG: Pearson syndrome and the CDG may present as hydrops with hepatosplenomegaly and mimic a lysosomal disorder [40]. Carnitine transporter deficiency and neonatal haemochromatosis also appear in the hydrops differential diagnosis list of Nyhan's Atlas [103]. Multiple sulfatase deficiency was diagnosed by exome sequencing in the HYDROPS study [15]. Other clues from Saudubray's table are hepatic haemosiderosis (neonatal haemochromatosis, respiratory chain defects, transaldolase deficiency, Zellweger syndrome), coarse facies at birth (galactosialidosis, I-cell disease, GM1, sialidosis type 2, MPS VII) and variants in PLVAP (hydrops with facial dysmorphism and cardiac and renal anomalies) and in GDF2 [40].

3.8. Fetal akinesia and neuromuscular disorders

The fetal akinesia deformation sequence (Pena-Shokeir I) combines increased NT or hydrops, absence of movement, arthrogryposis, pterygia, pulmonary hypoplasia, cleft palate, micrognathia, growth restriction, polyhydramnios and breech presentation [42]. Lethal congenital contracture syndrome 1 and lethal arthrogryposis with anterior horn cell disease, allelic and caused by recessive variants in GLE1 in Finland, combine severe hydrops, micrognathia, pulmonary hypoplasia and loss of anterior horn neurons [42]. In lethal multiple pterygium syndrome, all cases end in stillbirth or immediate neonatal death from pulmonary hypoplasia; hydrops occurs in more than half and polyhydramnios in about a third; the «early» form causes intrauterine death in the 2nd trimester with hydrops or hygroma, and CHRNG explains about 8% of recessive cases [111]. On exome sequencing, neuromuscular disorders accounted for 21% of the diagnoses in the 2023 meta-analysis, and in the HYDROPS study variants in RYR1, AMER1 and BICD2 were identified; limb contractures were the most prevalent additional anomaly (17.3%, 19/110) in the Mone meta-analysis [11, 14, 15]. Where myopathy is suspected, muscle biopsy should be performed as early as possible after delivery and processed urgently [3].

3.9. Skeletal dysplasias

Skeletal dysplasias associated with hydrops include achondrogenesis, osteogenesis imperfecta, osteopetrosis, thanatophoric dysplasia, the short rib-polydactyly syndromes, asphyxiating thoracic dysplasia and, rarely, achondroplasia [1]. Severe oedema hampers measurement of the limbs, so skeletal radiography after delivery or termination is part of the work-up [3]. In short-limb dysplasias, hydrops and polyhydramnios are findings to be sought, and the placenta should be assessed because of the possibility of triploidy [112]. The association of osteopenia, abnormal dentition, hydrops fetalis and communicating hydrocephalus was described by Macdermot in 1995, in the context of the differential diagnosis of Cole-Carpenter syndrome [113].

The appendix of Smith's atlas indicates frequent hydrops in achondrogenesis IA and IB, in fibrochondrogenesis, in monosomy X, in monozygotic twinning with structural defects and in osteogenesis imperfecta type II, and occasional hydrops in Down syndrome, in lethal multiple pterygium syndrome, in Morquio disease, in MPS VII, in Noonan syndrome, in short rib-polydactyly type II and in achondrogenesis II [130].

3.10. Other syndromes

Simpson-Golabi-Behmel syndrome type 2 (Xp22; OFD1; also the recurrent PIGA variant NM_002641.4:c.1234C>T p.(Arg412Ter)) has a severe neonatal form with death in utero or neonatally from hydrops, heart disease, diaphragmatic hernia, sepsis and hypoglycaemia [131, 132]. IPEX syndrome (FOXP3, X-linked) may present in the fetus with hydrops, echogenic bowel, skin desquamation, growth restriction and akinesia, and is associated with recurrent loss of hydropic male fetuses [133]. In incontinentia pigmenti, affected male fetuses are generally lost in the 1st or early 2nd trimester, and hydrops may be seen at 8–13 weeks [134]. Hydrops is occasional in Mulibrey, Desbuquois and Yunis-Varón syndromes [135–137]. On exome sequencing, Kabuki syndrome (KMT2D) and PUF60-related syndromes have been identified in fetuses with NIHF [15, 16].

3.11. Thoracic, tumoral, placental, twin-related and other organ causes

NIHF occurs in about 6%–7% of fetuses with congenital pulmonary airway malformation (CPAM), and the prognosis is worse with larger lesions, especially with a CPAM volume ratio (CVR) ≥1.6 [1]. CPAM is diagnosed in about 1/10,000 pregnancies, frequently grows between 20 and 28 weeks and then improves spontaneously in many cases; with hydrops the prognosis is worse [45]. Bronchopulmonary sequestration and congenital high airway obstruction syndrome (CHAOS) are other thoracic causes [1]. NIHF complicates 4%–9% of congenital diaphragmatic hernias [1]. Among tumours, sacrococcygeal teratoma is one of the most frequent: incidence 1/35,000–1/40,000 live births, female to male ratio 4:1, intense vascularity, rapid growth and high-output cardiac failure with hydrops, of poor prognosis when present before 28 weeks [1, 45]. Tumours of the face and neck, rhabdomyomas, hepatoblastoma, vein of Galen aneurysm and neuroblastoma complete the group [1].

Placental and cord lesions include chorioangioma, which is the most reported (single or multiple; greater risk from 4–5 cm upwards), choriocarcinoma, angiomyxoma, umbilical vein aneurysm and umbilical vein thrombosis [1]. In monochorionic pregnancies, stage IV twin-to-twin transfusion syndrome is defined by NIHF in one or both twins, more often in the recipient, from hypervolaemia and raised central venous pressure; in the twin reversed arterial perfusion (TRAP) sequence, present in about 1/35,000 births, the normal twin perfuses the acardiac twin and is at risk of heart failure and hydrops [1, 138, 139].

Congenital nephrosis causes effusions through hypoproteinaemia, and survivors have massive proteinuria and renal failure; echogenic kidneys are an ultrasound clue [1, 140]. Gastrointestinal causes include volvulus, meconium peritonitis, intestinal atresia, malrotation, eventration and prune belly syndrome; hepatic fibrosis or cirrhosis occurs, for example, in Gaucher disease and in trisomy 21 [1].

3.12. Infectious causes

Infections account for about 7%–8% of NIHF cases, and parvovirus B19 is the most frequent infectious cause [1]. The virus has tropism for erythroid progenitors, inhibits erythropoiesis and causes anaemia, often transient; the outcome is worse when infection occurs before 20–24 weeks, and in infected fetuses the risk of NIHF reaches 12% and that of fetal death 5%–11% [1]. The Oxford Desk Reference considers parvovirus B19 probably the most frequent infectious cause in Europe and syphilis the most frequent worldwide, particularly where there is hydrops [90]. Cytomegalovirus (CMV), syphilis and toxoplasmosis are the other classic causes; less frequent infections include varicella-zoster, enteroviruses, herpes simplex virus type 1, human herpesvirus 6, respiratory syncytial virus, trypanosomiasis and lymphocytic choriomeningitis virus; the mechanisms are anaemia, anoxia, endothelial injury, increased capillary permeability and myocarditis [1].

3.13. Maternal complications

Mirror syndrome is a form of pre-eclampsia, with or without oedema, in which the pregnant woman develops a picture that «mirrors» that of the fetus; it occurs in 8%–38% of NIHF cases in small series, with oedema in about 90%, hypertension in 36%–60% and proteinuria in 40%–64%, in addition to headache, visual disturbances, oliguria and laboratory abnormalities of liver function, creatinine and platelets, and it may progress to severe pre-eclampsia, eclampsia or HELLP syndrome [1]. In small series, morbidity included pulmonary oedema in 8%–36%, postpartum haemorrhage in 44%–50%, transfusion in 19% and intensive care admission in 13%, whereas a large state-wide series recorded pulmonary oedema in 1.5%, postpartum haemorrhage in 8% and transfusion in 4% [1]. The syndrome resolves with delivery; there are cases of resolution after fetal intervention (arrhythmia, hydrothorax, parvovirus, bladder obstruction) or after spontaneous or induced fetal death, and the imbalance of angiogenic and antiangiogenic factors is similar to that of pre-eclampsia and corrects after treatment [1]. In the prospective UCSF cohort of 116 pregnancies with effusions after 20 weeks, hypertensive disease occurred in 22% (19/85) of NIHF cases and in 10% (3/31) of single effusions (p=0.12); of the 22 affected women, 36% had pre-eclampsia with severe features, 18% pre-eclampsia without severe features, 9% atypical pre-eclampsia and 36% gestational hypertension; 82% were diagnosed before delivery and fetal genetic disease did not increase the risk [141]. In alpha-thalassaemia major and in triploidy, the enlarged placenta raises the risk of pre-eclampsia [41, 54], and in LCHAD deficiency two pregnancies were complicated by HELLP syndrome [110].

4. Clinical Recognition and Warning Signs

4.1. First trimester: nuchal translucency and hygroma

NT is ideally assessed at 11–13+6 weeks; the value of 3 mm corresponds to the 95th centile [142]. In combined screening (free β fraction of human chorionic gonadotrophin, β-hCG; pregnancy-associated plasma protein A, PAPP-A; and NT at 12 weeks), detection of trisomy 21 is 88% for a screen-positive rate of 2% [3]. Fetal echocardiography is recommended in all cases with NT ≥3.5 mm: in six studies, the mean prevalence of major heart disease in euploid fetuses with NT ≥3.5 mm was 6.27% (3.1%–12.2%), and in another eight studies 5.88% (2.3%–9.7%); in one series, 11% of fetuses with NT above 3.5 mm, a normal karyotype and a negative ultrasound had anomalies diagnosed after birth [3]. The Bilardo series (451 euploid fetuses with increased NT, published in 2007 and re-presented in the 2010 review) shows that the proportion of adverse outcomes rises with thickness; the original table separates miscarriages and intrauterine deaths from terminations of pregnancy for severe hydrops, categories that the Oxford Desk Reference reproduces in a single column [3, 143]:

Source: Table 4 of Bilardo et al. 2010, modified from Bilardo et al. 2007 [143]. The absence of fetal deaths in the 5.5–6.4 mm class (0/19) reflects a small denominator and does not mean zero risk; terminations for severe hydrops cluster at ≥6.5 mm, where the observed spontaneous mortality (17%) underestimates lethality because some pregnancies with worsening hydrops were terminated before fetal death. In the combined data of three studies with 4991 euploid fetuses with increased NT, the prevalence of miscarriage or fetal death rose from 1.6% between the 95th and 99th centiles to about 20% at ≥6.5 mm, and up to 30% of couples with worsening fetal hydrops opted for termination before intrauterine death [143].

If the 20-week ultrasound is completely normal and the NT has regressed, the probability of an uneventful outcome is similar to that of the general population, a conclusion that the authors and the Oxford Desk Reference consider still dependent on larger series, given the small number of survivors with very markedly increased NT [3, 143]. In the series of 451 fetuses, adverse outcomes with a normal ultrasound occurred in 4% of cases (intrauterine deaths, undetected heart defects and genetic syndromes in equal proportions), and 7 of the 50 fetuses (14%) with suspicious ultrasound findings had an uneventful outcome [143]. In cystic hygroma, septated or not, aneuploidy is frequent: in a series of 72 cases, 52.7% had a chromosomal abnormality (37% trisomy 21) and, of the euploid fetuses, 52.9% were liveborn, with one case of Noonan syndrome; in another series of 42 cases, 47.6% were aneuploid (55% Turner syndrome and 20% trisomy 21), 29.4% of the euploid fetuses had major heart disease and regression of the hygroma (10% of aneuploid and 17.6% of euploid fetuses) had no diagnostic value; in a third series of 92 cases, 53% had a normal karyotype, additional anomalies and progression to hydrops or effusions were the high-risk factors, and 89% of continued, euploid pregnancies without those factors produced normal children between 1 and 5 years of age [3]. In Noonan syndrome, isolated NT above the 99th centile should prompt consideration of the condition even in low-risk populations [19]. Denmark has adopted NT ≥6.0 mm as a national indication for prenatal genome sequencing [144].

4.2. Second and third trimesters: when to suspect a specific cause

A single effusion — pleural effusion, ascites, pericardial effusion or skin oedema — should be assessed as possible NIHF spectrum, with the same work-up as complete hydrops [1]. The combination of findings is more informative than any one in isolation [40]. The clues that point towards a cause are summarised in the table.

4.3. Targeted history

The history should include a three-generation pedigree with previous fetal losses, a family history of anaemia, consanguinity and the parents' origin, because the latter informs the risk of alpha-thalassaemia or of glucose-6-phosphate dehydrogenase deficiency; it also includes infectious exposures, trauma and examination of the parents for signs of Noonan syndrome [1, 3]. A family history of lymphoedema may identify lymphoedema-distichiasis or Milroy disease; the five-generation family described in GeneReviews illustrates this situation [20].

4.4. Warning signs and emergencies

Pregnancy with NIHF has maternal and fetal warning signs that require immediate reassessment. On the maternal side, oedema, hypertension, proteinuria, headache, visual disturbances, oliguria and laboratory abnormalities (liver function, creatinine, platelets) suggest mirror syndrome or pre-eclampsia [1]. On the fetal side, new or worsening NIHF, non-reassuring tests of wellbeing, symptomatic polyhydramnios and rupture of membranes are reasons to discuss delivery [1, 24]. A sustained tachyarrhythmia or confirmed fetal anaemia are situations in which fetal treatment may alter the prognosis and justify rapid referral to a fetal medicine centre [1].

4.5. Neonatal period

Generalised oedema in the very sick newborn generally has no genetic basis; in the extremely preterm infant, the causes merge with those of fetal hydrops, and severe anaemia and hypoalbuminaemia are the first to exclude [48]. Oedema of the dorsum of the hands and feet, congenital chylothorax and the markers of intrauterine oedema point towards a lymphatic dysplasia or a RASopathy [47, 48]. Hepatosplenomegaly, coarse facies, inguinal hernias, cherry-red spot and vacuolated lymphocytes suggest a lysosomal disorder [40, 64]. High birth weight with oedema, hypertrophic cardiomyopathy and arrhythmias suggest Costello syndrome [50].

4.6. Newborn screening

There is no newborn screening for NIHF. In alpha-thalassaemia, haemoglobin Bart's in dried blood by high-performance liquid chromatography (HPLC) or isoelectric focusing above 15% at birth suggests HbH disease and values of 1%–8% indicate carrier states; in California, newborn screening estimated the prevalence of clinically significant alpha-thalassaemia at 9.6/100,000 births [17, 30].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Criteria

The diagnosis of NIHF is made by ultrasound and requires abnormal fluid in at least two compartments (ascites, pleural effusion, pericardial effusion >2–3 mm, skin oedema >5 mm) with exclusion of red cell alloimmunisation; the NIHF spectrum admits a single effusion when the process is presumed systemic [1]. The assessment is the same in both situations, and even a collection of local cause, such as a urinoma, warrants consideration of genetic assessment [1]. Identification of a structural anomaly does not close the investigation, because a genetic disorder may explain both the anomaly and the effusions [1]. Genetic counselling is an integral part of the assessment [1].

5.2. Assessment algorithm

SMFM Consult Series #75 sets out the assessment in Figure 2 of the document. The table follows the order of that figure and of the assessment section of the text.

If invasive testing is declined, cell-free DNA (cfDNA) aneuploidy screening and maternal serology for parvovirus, CMV and toxoplasmosis may be offered, bearing in mind that performance is variable and that a positive cfDNA result requires confirmation [1]. If there is a family history of a genetic disorder that increases the risk of effusions, the strategy should be directed at that disorder; if the parents are known carriers, the specific disorder is tested for [1].

The 2015 Clinical Guideline #7 recommended a similar sequence — antibody screening, detailed ultrasound with echocardiography and arrhythmia assessment, MCA Doppler, karyotype and/or CMA even without a structural anomaly — but did not include sequencing [26]. The Oxford Desk Reference adds biochemical analysis of amniotic fluid and cultured amniocytes for lysosomal disorders (the lysosomal «hydrops screen» offered by some centres) and, in consanguineous families or in recurrent hydrops, exome, genome or panel sequencing [3].

5.3. Maternal assessment for alpha-thalassaemia: MCV thresholds

SMFM Consult Series #75 is internally inconsistent as to the MCV threshold. The text of the document says: «If the red blood cell mean corpuscular volume is near or above 80 fL and haemoglobin electrophoresis results are normal, targeted testing with sequencing and deletion/duplication analysis for alpha thalassemia is indicated» (MCV «near or above 80 fL»). Figure 2 of the same document, in the box for reviewing the tests, says: «Testing for alpha thalassemia carrier status if MCV is near or ≤80 fL and haemoglobin electrophoresis is normal» (MCV «near or at or below 80 fL») [1]. The Figure 2 version is the one consistent with the physiology of alpha-thalassaemia trait, which causes microcytosis with normal electrophoresis, because electrophoresis and HPLC identify beta-thalassaemia carriers but not alpha-thalassaemia carriers [30]. The sources consulted do not include an SMFM erratum. In practice, reading the text (MCV ≥80 fL) does not identify the microcytic population in which molecular analysis has the highest yield, and the decision should rest on Figure 2 and on the physiological data, with the discrepancy recorded.

The 2021 international consensus on alpha-thalassaemia major uses a different threshold, for a different purpose: in preconception or early pregnancy screening with a full blood count, an MCV below 70 fL identifies women at risk of carrying a two-gene deletion in cis (−−/αα), the genotype that allows haemoglobin Bart's in the offspring [30]. The same consensus warns that red cell indices may miss carriers and that, in at-risk populations, molecular testing should be offered [30]. The two thresholds do not contradict one another when read as ≤80 fL: the SMFM's 80 fL serves to decide on molecular carrier testing in a pregnant woman with a hydropic fetus and normal electrophoresis; the consensus's 70 fL stratifies the risk of a deletion in cis in population screening.

5.4. Ultrasound, echocardiography and Doppler

The detailed anomaly scan at 18–22 weeks (19–21 weeks in the Oxford Desk Reference) assesses structural anomalies and skeletal dysplasias; detailed fetal echocardiography is mandatory in all NIHF cases [1, 3, 142]. MCA Doppler should be performed even where there are anatomical anomalies, because disorders such as Diamond-Blackfan anaemia combine anaemia with limb anomalies or clefts [1]. A high velocity warrants cordocentesis to confirm the anaemia and obtain a blood film, useful where a rare hereditary anaemia is suspected [1]. In alpha-thalassaemia major, however, MCA Doppler was frequently normal or only mildly elevated despite severe anaemia in the fetuses of the UCSF in utero transplantation trial, which limits its value for monitoring anaemia in this disease [145].

5.5. Cytogenetic and infectious testing

CMA detects copy number variations (CNV) not visible on the karyotype and is the standard in fetal anomaly; the karyotype remains necessary to demonstrate a structural rearrangement underlying unbalanced material, with consequences for the recurrence risk [1]. Two small retrospective NIHF series showed no significant incremental yield of CMA over the karyotype [1]. CMA detects the microdeletions that cause alpha-thalassaemia, an exception to the principle that monogenic disorders require sequencing [1]. Sex chromosome mosaicism may escape the array, and when the initial investigation was by array the laboratory generally recommends a karyotype [87]. PCR for parvovirus, CMV, toxoplasmosis and other agents should form part of the amniotic fluid study when infection is part of the differential [1].

5.6. Exome and genome sequencing

The yield of sequencing depends on the definition of NIHF, on prior cytogenetic screening, on the design (trio or proband only), on consanguinity and on sample size.

Sources: [11, 12, 14–16, 27, 39, 46, 146–148].

The Sparks cohort was reanalysed by Norton and colleagues to compare exome sequencing with panels: the ten commercial panels from seven laboratories (six RASopathy, three NIHF and one metabolic; median of 22 genes, 11–148) would have sequenced only 13–15 of the 29 diagnostic genes (45%–52%) and detected 19–24 of the pathogenic variants (51%–62%); the largest NIHF panel would have a yield of 18%, against 29% for the exome, even assuming 100% analytical sensitivity [149]. The SMFM summarises that panels miss 38%–89% of the diagnoses identifiable by exome sequencing [1]. When sequencing is not feasible, the available panels include hydrops panels, RASopathy panels, hereditary anaemia panels and metabolic panels, such as the «metabolic non-immune hydrops fetalis» panel referred to in the GeneReviews text on Niemann-Pick disease C [1, 150]. The GeneGuidance diagnostic library on NIHF refers to the Sparks study as the original reference research on exome yield and contains no dedicated guideline [151]. In a decision model with ten strategies and a threshold of 100,000 dollars per quality-adjusted life year (QALY), exome sequencing alone was dominant before 18 weeks and between 18 and 22 weeks, and after 22 weeks, without the option of termination, it had lower costs and more QALYs; it remained cost-effective up to 50,451 dollars per test before 18 weeks, 50,423 between 18 and 22 weeks and 9530 after 22 weeks [152]. The data are from the United States and do not transfer directly to European health systems.

Genome sequencing has a small incremental yield over the exome, with limited data: 7% (1/14) in the extended HYDROPS study and 2.5% (3/118) in the UCSF series, where the new diagnoses were due chiefly to updated tools and literature and not to genome technology itself [16, 39]. In a meta-analysis of congenital anomalies in general (18 studies, 902 cases), the yield of genome sequencing over quantitative fluorescent polymerase chain reaction (QF-PCR) and CMA was 26% (16% prenatally), but only 1% (95% CI 0%–4%) after exome sequencing; the VUS rate was 18%, the median turnaround time 18 days and the DNA quantity required 100 ng, against 350 ± 50 ng for QF-PCR, CMA and exome [153]. The SMFM considers it acceptable to choose exome or genome according to availability and preference; genome sequencing often has a shorter turnaround time and requires less DNA [1]. In Denmark, the Danish Fetal Medicine Society has included NIHF among the indications for routine prenatal genome sequencing nationally [144].

5.7. Limits of the tests

A negative prenatal exome result does not exclude a genetic cause and should not be used as a guarantee of normal development [28]. Prenatal exome sequencing is not validated for all CNV, which is why CMA is performed beforehand or in parallel [27, 29]. The exome is limited to the coding regions of more than 20,000 genes, about 1%–2% of the genome, so intronic and regulatory variants outside the exon-intron junctions are not analysed; even the genome has regions that are difficult to sequence and analyse [28]. In regions with pseudogenes, as in GUSB, sequence analysis requires careful confirmation [18]. Mosaicism is another limitation: mosaic tetrasomy 12p is detected in amniocytes and fibroblasts but rarely in fetal blood, and sex chromosome mosaicism may require a karyotype [87, 88]. Enzyme assay in chorionic villi or cultured amniocytes generally takes 14–21 days, and placental biopsy and amniocentesis in the 3rd trimester have a higher culture failure rate [154]. Pseudodeficiency alleles, such as GUSB c.454G>A p.(Asp152Asn), produce low enzyme activity without disease and may generate false positives in enzymatic diagnosis [18, 64]. Urine may be normal in MPS VII despite the disease [64]. Alpha-thalassaemia panels that do not include the rare α⁰ alleles of a population may misclassify the fetal genotype [56]. Maternal cell contamination should be excluded in all prenatal genetic studies [28]. The proportion of variants detectable by a method is not equivalent to the clinical sensitivity of the test in NIHF, which depends chiefly on the set of genes analysed, on the quality of the phenotype submitted and on the analysis strategy [27, 149].

5.8. Requesting, reporting and interpretation

The sequencing request should be accompanied by the detailed fetal phenotype (biometry, images, preferably in Human Phenotype Ontology terms), previous results and the family history, including consanguinity and origin, because exome interpretation is guided by the phenotype [1, 27, 28, 38]. Trio analysis (fetus and both parents) is preferable: it allows de novo variants to be identified immediately, the phase of biallelic variants to be determined and parental carrier status to be confirmed; if only the proband is sequenced, the inheritance of candidate variants should be confirmed by targeted parental testing [27, 28, 38]. When filtering by inheritance model yields no candidates, analysis should be performed without assuming inheritance [27].

The report should state the gene, the reference transcript with version, the variant in HGVS nomenclature, the zygosity, the inheritance and the class according to the ACMG/AMP (Association for Molecular Pathology) criteria. The ACMG recommends reporting pathogenic and likely pathogenic variants in disease genes compatible with the fetal phenotype; laboratories may report VUS in genes that fit the phenotype, especially in autosomal recessive disorders when the VUS is in trans with a pathogenic variant; a variant in a candidate gene should not be classified above VUS [28]. Incidental fetal findings in genes for moderate to severe childhood disease with high penetrance should be reported; fetal carrier status for recessive disorders unrelated to the indication should not be reported [28]. The Canadian College of Medical Geneticists (CCMG) reports only pathogenic and likely pathogenic variants with a high probability of relation to the phenotype, does not report secondary findings by default and reports incidental findings only where there is a significant risk of Mendelian disease in childhood [29]. The approaches diverge, and the laboratory's policy should be explained before testing.

Complementary studies help in interpreting VUS: in the meta-analysis of lysosomal disorders, enzyme studies combined with exome sequencing allowed variants of uncertain significance to be interpreted, and the four additional VUS were all in GUSB [97]. Reclassification of VUS yielded two additional diagnoses in the extended HYDROPS study [16]. Reanalysis of exome data should be considered if a new phenotype emerges after birth, if a new pregnancy is planned or if at least 12 months have elapsed [28].

5.9. Pre- and post-test counselling

Pre-test counselling should be provided by a professional trained in genetics and should discuss the types of result (pathogenic, likely pathogenic, VUS), the realistic probability of a useful result, the expected timeframe, the possibility of no result before delivery, incidental and secondary findings, adult-onset findings, the discovery of non-paternity or consanguinity, data sharing and reanalysis [27, 38]. Each parent gives separate consent for their own sample; only the pregnant woman consents to the invasive procedure [27, 38]. The ACMG adds the option of declining analysis of the secondary findings genes and recalls that the turnaround time has to be compatible with reproductive decisions [28]. Post-test counselling is recommended in all cases, including negative ones, and should be accompanied by written information in accessible language about the results and the reproductive options [28, 38].

The 2018 joint ISPD/SMFM/PQF statement considered that routine use of prenatal sequencing could not be supported for lack of validation data and that, ideally, it would be carried out within a research protocol or case by case under the guidance of multidisciplinary teams [38]. The 2022 ISPD position replaced it and now considers sequencing indicated in a single major or multisystem anomaly without a diagnosis after CMA, in patterns strongly suggestive of a monogenic disorder and in recurrences; it maintains that there is no evidence for its use in the absence of a fetal anomaly [27]. The ACMG (2020) considered that there were no data supporting exome sequencing in unexplained recurrent pregnancy loss without an anomaly [28]. The 2026 SMFM recommends exome or genome sequencing in NIHF and in the NIHF spectrum with GRADE 1C [1].

5.10. Diagnosis of haemoglobin Bart's hydrops fetalis syndrome

In a couple known to be carriers, hydrops fetalis should be considered alpha-thalassaemia major until proven otherwise [30]. Fetal diagnosis rests on severe hypochromic macrocytic anaemia without ABO or Rh incompatibility, with reticulocytosis that may exceed 60%, a blood film with large hypochromic red cells, severe anisopoikilocytosis and numerous erythroblasts, absent HbA and increased haemoglobin Bart's; the fetal MCV described is 136 ± 5.1 fL [17]. Fetal blood contains chiefly haemoglobin Bart's (γ4) and 10%–15% haemoglobin Portland (ζ2γ2) according to GeneReviews; the Oxford Desk Reference indicates 80% haemoglobin Bart's and 20% embryonic haemoglobins [17, 54]. Confirmation is by molecular demonstration of the deletion or inactivation of all four α alleles [17]. The 2021 consensus recommends rapid confirmation by cordocentesis with haemoglobin electrophoresis and accepts performing the first intrauterine transfusion at the same procedure, while the result is awaited [30]. The combination of hydrops with a very high proportion of haemoglobin Bart's occurs in no other condition [17].

5.11. Perinatal assessment and pathology

If the fetus or newborn dies, the investigation includes clinical photography, skeletal radiography, storage of tissue (a fibroblast line) and of DNA, and autopsy with examination of the placenta [3]. In recurrent hydrops without a diagnosis, a fibroblast line should be established and amniotic fluid frozen while the autopsy and radiographs are awaited [3]. Autopsy is essential where an inborn error of metabolism is suspected and allows fluids and tissues to be collected for biochemical confirmation, a concept termed the «metabolic autopsy» [40]. Histology of the placenta, liver, spleen, heart, blood and bone marrow and bone morphology guide the diagnosis, but definitive diagnosis requires DNA testing [1]. Thoracentesis fluid with increased lymphocytes indicates chylothorax, and placental pathology confirms infections and chorioangiomas [1]. The guide for perinatal pathologists published in 2026 organises the autopsy by phenotype — NIHF with bilateral cystic hygromas, the «pale hydropic baby» (anaemia), NIHF with cardiovascular anomalies, without evident structural anomaly and with prominent ascites — and insists on placental examination [155]. Studies on the incremental yield of pathology are limited and predate current genetics [1].

6. Differential Diagnosis

6.1. Comparative table

The differential diagnosis has two levels: distinguishing NIHF from the situations that mimic it (immune hydrops, local collections, acquired neonatal oedema, neonatal liver failure) and, within NIHF, distinguishing the aetiological groups from one another.

6.2. Immune hydrops

About 15% of Caucasian women are RhD-negative, but anti-D immunisation has become rare with prophylaxis; other alloantigens, such as Kell, are rare but can cause severe disease [34]. Haemolytic anaemia from IgG anti-D leads to heart failure and hydrops in severe cases without fetal transfusion [34]. Routine alloantibody screening and fetal RhD genotyping in cell-free DNA in maternal plasma, from the end of the 1st trimester when the partner is heterozygous, allow at-risk pregnancies to be identified; spectrophotometry of bilirubin in amniotic fluid (Liley index) has fallen out of use [34]. The SMFM insists on repeating the indirect Coombs test even when it was previously negative, because immune hydrops is treatable and changes the entire approach [1]. In immune haemolytic disease a phase 3 trial of nipocalimab is under way (NCT05912517, about 120 participants, primary completion expected August 2027), outside the scope of NIHF [157].

6.3. Differential diagnosis within NIHF

Distinguishing between the aetiological groups depends on the associated findings. The combination of hydrops with a very high proportion of haemoglobin Bart's is specific to alpha-thalassaemia major [17]. Fetal anaemia with common haemoglobinopathies excluded and non-genetic causes negative points to the rare hereditary anaemias, in which exome sequencing has a yield [61]. Prominent ascites with hepatosplenomegaly suggests a lysosomal disorder, but Pearson syndrome and the CDG may mimic a storage disorder [40]. Among the lysosomal disorders with a dysostosis multiplex phenotype, mucolipidosis II, sialidosis, GM1 gangliosidosis and galactosialidosis are morphologically similar and are distinguished by enzyme analysis [101]. Signs of Noonan syndrome in the parents point to an inherited RASopathy [3]. Contractures increase the probability of a monogenic diagnosis by exome sequencing [14].

6.4. Neonatal liver failure

Neonatal liver failure enters the differential because some of its causes present with fetal ascites or hydrops. Neonatal haemochromatosis appears in both the fetal hydrops and the acute liver failure lists of Nyhan's Atlas and is characterised by normal or disproportionately low transaminases [40, 103]. Transaldolase deficiency presents with hydrops and progresses to neonatal liver failure, cirrhosis and, in some patients, liver transplantation [72, 107]. TRMU deficiency, by contrast, typically manifests between 2 and 4 months with transient acute hepatic dysfunction, severe lactic acidosis, hypoglycaemia and hyperammonaemia, and improves with supportive care; it is not a typical cause of NIHF [156]. Disorders from NBAS, IARS, LARS and SCYL1 cause recurrent acute liver failure triggered by fever in early childhood, generally outside the neonatal period [40]. The acute liver failure list of Nyhan's Atlas further includes alpha-1-antitrypsin deficiency, bile acid defects, MPI-CDG, citrullinaemia, fatty acid oxidation disorders, galactosaemia, hepatorenal tyrosinaemia, hereditary fructose intolerance, mitochondrial disorders with mitochondrial DNA depletion (POLG, DGUOK, MPV17), Niemann-Pick disease B and C, Wilson disease and Wolcott-Rallison syndrome [103].

7. Treatments and Risk Management Strategies

7.1. General principles and counselling

Management decisions depend on the pregnant woman's preferences, maternal complications, the gestational age at onset, the aetiology and the availability of interventions; where fetal treatment is possible, the woman should be referred to a specialist centre [1]. Counselling is multidisciplinary (maternal-fetal medicine, neonatology, genetics) and addresses the mode of delivery, the gestational age from which resuscitation is offered, the degree of resuscitation, neonatal interventions and palliative care [1]. Because of the associated maternal risks, the SMFM recommends individualised counselling and the offer of all management options, including termination of pregnancy (recommendation 4, GRADE 1C) [1]. When the pregnancy continues with persistent NIHF, the woman should have serial blood pressure assessments, monitoring for symptoms of pre-eclampsia and periodic laboratory tests, and be informed of the risk of preterm birth and caesarean section [1]. In alpha-thalassaemia major, the 2021 consensus calls for non-directive counselling about expectant management, termination or intrauterine transfusion, and discourages expectant management because of the maternal risk of mirror syndrome; GeneReviews suggests that a clinical ethics consultation may be helpful [17, 30].

7.2. Fetal treatment by aetiology

Fetal treatment exists for only a fraction of the causes, and the evidence of efficacy varies widely between interventions [1]. The table reproduces the SMFM examples.

Source: [1]. Emerging therapies are being studied in trials at selected centres, and when the fetus has a known genetic disorder the decision about intervention is shared, for lack of evidence [1].

7.3. Intrauterine transfusion

Intrauterine transfusion (IUT) treats fetal anaemia confirmed by fetal blood sampling [1]. In parvovirus B19, outcomes after IUT are generally favourable, with survival of 85% in transfused fetuses and a worse course without transfusion; a small series raised concern about long-term neurodevelopmental delay, attributable to the hydrops, to the infection itself, to severe anaemia or to the transfusions [1]. In a retrospective European cohort from nine centres (France, Belgium and Czechia; January 2014 to May 2024) with 78 fetuses hydropic from parvovirus B19 and at least one IUT, perinatal survival without severe brain injury was 59.0% (46/78); the unfavourable outcome (41.0%) comprised fetal death in 25.6%, termination in 11.5%, continuation despite severe neurological findings in 1.3% and neonatal death in 2.6% [158]. A fetal haemoglobin above 9.3 g/dL after the 1st or 2nd IUT was associated with a lower risk of an unfavourable outcome (adjusted OR 0.6; 95% CI 0.4–0.8) and a femur length with a Z score below −2 with a worse outcome (adjusted OR 5.4; 95% CI 1.3–27.0); transplacental and transamniotic puncture did not differ [158]. The difference between the SMFM's 85% and the 59% of the European cohort reflects different denominators (overall survival versus survival without severe brain injury in already hydropic fetuses) and should not be read as a direct comparison.

In alpha-thalassaemia major, IUT should start as soon as technically feasible, at 18 weeks in most fetal therapy centres, with a protocol similar to that for alloimmunisation [17, 30]. In the UCSF international registry (49 prenatal and 11 postnatal diagnoses), the group with a prenatal diagnosis alive at discharge (n=14) had better resolution of hydrops, delivery closer to term, shorter admissions and average or above-average neurodevelopment more often; earlier IUT correlated with better Vineland-3 scores (r=−0.72; P=0.02), and preterm birth after IUT occurred in 19% (3/16) [55]. In the data combined with two series, two or more IUTs were associated with better resolution of hydrops, delivery at ≥34 weeks and a 5-minute Apgar ≥7, with normal neurodevelopment in 17/18 against 5/13 with zero or one IUT (OR 2.74; P=0.01) [55]. IUT improved delivery outcomes and reduced maternal complications, but did not alter the incidence of structural anomalies [17]. In the UCSF in utero transplantation trial, fetuses received serial IUT every 2–3 weeks and delivery was planned within three weeks of the last IUT [145]. In congenital dyserythropoietic anaemia type I, prenatal management includes monitoring of fetal haemoglobin by Doppler and fetal transfusions to prevent hydrops where anaemia is severe; GATA1-related cytopenia and KLF1-related dyserythropoietic anaemia may also require IUT [58–60].

7.4. Maternally administered antiarrhythmics

Treatment of fetal tachyarrhythmias is by maternal administration of antiarrhythmics that cross the placenta. None of the drugs used has a fetal indication in the Food and Drug Administration (FDA) label: digoxin (Lanoxin®) is authorised for heart failure and rate control in atrial fibrillation, flecainide for the prevention of paroxysmal SVT and of paroxysmal atrial fibrillation or flutter without structural heart disease, and sotalol (Betapace®) for serious ventricular arrhythmias and maintenance of sinus rhythm in atrial fibrillation or flutter [159–161]. Fetal use is therefore off-label. The 2014 American Heart Association (AHA) scientific statement on fetal cardiac disease could not be consulted beyond the abstract, and no doses from that source were used [162]. The doses below are those of the protocol of the randomised FAST trial (NCT02624765), registered on ClinicalTrials.gov.

The available results come from studies without a randomised control group. In the single-arm Japanese multicentre trial (15 institutions, 2010–2017; 22 to <37 weeks; SVT or flutter ≥180 beats per minute), the protocol of digoxin, sotalol and flecainide resolved the tachyarrhythmia in 89.8% (44/49) overall and in 75.0% (3/4) of hydropic fetuses; maternal adverse effects occurred in 78.0% (39/50), with serious effects in one mother and four fetuses, two fetal deaths chiefly from heart failure and neonatal tachyarrhythmia in 31.9% (15/47) in the first two weeks [167]. At three-year follow-up, mortality was 2.2% (1/45) and neurodevelopmental impairment 9.3% (4/43), but 66.7% (2/3) in fetuses with hydrops and subcutaneous oedema, with cerebral palsy in two cases with severe oedema or ascites at an early gestational age; tachyarrhythmia fell from 31.9% in the neonatal period to 8.9% at 18 months and 4.5% at 36 months [168]. In Rotterdam (2007–2023), in fetal tachycardia without hydrops, oral digoxin alone converted 32% (9/28) at a median of 4.5 days, and the addition of oral flecainide raised total conversion to 93% (26/28) [164]. In the amiodarone meta-analysis (107 pregnancies, 27 studies), cardioversion was 69.9% (95% CI 57.6%–81.2%) and, in hydropic fetuses, 53.8%–87.5% (median 63.4%) [166]. The randomised FAST trial (phase 3, 105 participants) compared digoxin with sotalol in flutter without hydrops, digoxin with flecainide in SVT without hydrops and, in arm C, digoxin plus sotalol against digoxin plus flecainide in SVT with hydrops, with a term live birth (≥37+0 weeks) in normal rhythm as the primary outcome; it ended on 31 March 2024 and no results were found in the registry or in publication up to October 2026 [163]. The associated observational registry (330 participants, 14 countries, including Brazil, France, the Netherlands, the United Kingdom and Spain) ended in March 2025, also with no published results found [169].

In the European Union, digoxin, flecainide and sotalol are nationally authorised medicines, without a European public assessment report (EPAR); the national summaries of product characteristics and the trade names in Portugal were not consulted, and fetal use is likewise off-label.

7.5. Pleural effusions and shunts

Needle drainage or a thoracoamniotic shunt is considered where the aim is to prolong the pregnancy or where drainage may be of benefit before delivery [1]. Thoracentesis in primary chylothorax is one of the examples in which in utero intervention may improve outcomes [1]. Before delivery, drainage of large pleural effusions or ascites may avoid dystocia or facilitate resuscitation [1]. At King's College Hospital, in London (2014–2023), 20 newborns with NIHF were admitted to the NICU; the eight with a prenatal shunt (six pleuroamniotic and two abdominoamniotic) were born later (median 36 versus 32.5 weeks; p=0.025) and had shorter admissions (23 versus 95 days; p=0.019), without greater oxygen dependence at 36 weeks post-menstrual age, despite a more severe presentation [170]. The criteria for shunting were hydrops, polyhydramnios from oesophageal compression by a pleural effusion, and a large pleural effusion with probable compression of the inferior vena cava [170]. The series is small and retrospective.

7.6. CPAM, TRAP sequence and tumours

In larger CPAM, especially microcystic, the SMFM indicates maternal corticosteroids, with the possibility of repeating the course; in large cysts, aspiration or a thoracoamniotic shunt [1]. The document does not specify drug or dose. A phase 1/2 trial of antenatal corticosteroids in CPAM (NCT00670956) was terminated for insufficient recruitment, with one participant, because the drug was already being offered as standard treatment [171]. In the TRAP sequence, percutaneous ablation of the acardiac twin's cord by radiofrequency or microwave is offered [1].

In cardiac rhabdomyomas with obstruction or hydrops, maternal sirolimus is an off-label use described in small series. In the Toronto series (27 pregnancies, seven treated), treatment for more than seven days (n=3) produced tumour regression or resolution of the obstruction, prenatal discontinuation led to rebound growth (n=2) and treatment for seven days or less (n=4) had no effect; cerebral tubers appeared unchanged [94]. One report described reversal of hydrops from a massive rhabdomyoma in a twin pregnancy with tuberous sclerosis [172]. The sources consulted do not give a sirolimus dose, and no dosing is presented here.

7.7. Antepartum surveillance, timing and mode of delivery

Antepartum surveillance has not been shown to improve outcomes in NIHF, and there are no trials or series assessing it; it is reasonable to begin it from the gestational age at which neonatal interventions would be wanted, and deterioration or ultrasound worsening may prompt delivery [1]. There are no trials on the timing of delivery; up to 86% of births are preterm, elective preterm delivery has not been shown to be beneficial and prematurity worsens the prognosis [1]. By expert opinion, with new or worsening NIHF, delivery at 34 weeks is reasonable, and in stable situations the pregnancy may be prolonged [1]. The SMFM recommends individualising the timing of delivery and reserving preterm birth for obstetric indications — pre-eclampsia or mirror syndrome, preterm labour, preterm prelabour rupture of membranes, new or worsening NIHF, or risks of continuing that exceed those of delivery (recommendation 5, GRADE 1C); when the prognosis is poor and the woman opts for termination or palliative care, preterm induction should be offered [1].

There are no studies dedicated to antenatal corticosteroids in NIHF; one retrospective series showed a reduction in neonatal mortality and several others showed no benefit [1]. The SMFM recommends antenatal corticosteroids in continuing pregnancies when preterm birth is expected within the next seven days, if neonatal resuscitation is wanted and offered (recommendation 6, GRADE 1C), without specifying drug or dose [1]. The FDA label for betamethasone (Celestone Soluspan®) does not include the indication of fetal lung maturation, so that use is off-label in the United States and follows obstetric lung maturation protocols [173]. Tocolytics have not been studied in NIHF and their use is limited to completing the corticosteroid course or to the periprocedural period, when the aim is to prolong the pregnancy [1].

Up to 85% of NIHF cases are delivered by caesarean section [1]. Vaginal delivery is appropriate in the absence of acute deterioration or where palliative care has been chosen, and a large Californian retrospective study found no difference in neonatal mortality by mode of delivery [1, 23]. The SMFM recommends caesarean section for the usual obstetric indications when resuscitation and life support are planned (recommendation 7, GRADE 1C) [1]. Delivery should take place in a centre with a level III or IV NICU if the cause is treatable after birth or unknown, with antenatal transfer where necessary; transfer is unnecessary if termination or palliative care has been chosen and local services allow [1]. In alpha-thalassaemia major, delivery should take place in a tertiary centre with perinatology, paediatric haematology and neonatology; in Barth syndrome, pregnancies with male fetuses are managed in high-risk maternal-fetal medicine and delivery takes place in a tertiary centre with intensive care [30, 78].

7.8. Mirror syndrome

The SMFM recommends that all pregnant women who develop mirror syndrome in the context of NIHF receive individualised counselling about delivery or termination of pregnancy; expectant management should be reserved for rare circumstances, after counselling about the maternal risks and shared decision-making (recommendation 3, GRADE 1C) [1]. The 2015 guideline already considered mirror syndrome a form of severe pre-eclampsia that, in most cases, mandates delivery [26]. The syndrome resolves with delivery and, in described cases, after effective fetal treatment (arrhythmia, hydrothorax, parvovirus, bladder obstruction) or after fetal death [1].

7.9. Neonatal care

Stabilisation of the hydropic newborn requires a team prepared for immediate ventilation, drainage of the pleural and peritoneal cavities and correction of anaemia and coagulopathy; the case in section 10 illustrates immediate drainage of haemorrhagic effusions, prolonged mechanical ventilation and repeated chest drains for chylothorax [83]. Extracorporeal membrane oxygenation (ECMO) was used in 171 newborns with NIHF recorded in the Extracorporeal Life Support Organization (ELSO) registry between 2000 and 2024, with survival to discharge of 52% (88/171) [174]. Survival was higher with a weight above 2.5 kg, gestational age ≥35 weeks and female sex, and did not differ between venoarterial and venovenous ECMO nor change over the 25 years [174]. Neonatal intensive care is the supportive treatment of NIHF from MPS VII [18]. In PIEZO1-related lymphatic dysplasia and in CCBE1-related Hennekam syndrome there is no approved targeted therapy; treatment is supportive, with parenteral nutrition and a special formula in protein-losing enteropathy [9, 52]. A 2026 series describes the use of sirolimus in preterm infants with NIHF from vascular or lymphatic anomalies, but only the title was accessible and its data are not used here [175]. In Noonan syndrome, the bleeding diathesis should be considered before drainage procedures and surgery, and treatment of complications is the same as in the general population [19, 47].

7.10. Postnatal treatments directed at the cause

In alpha-thalassaemia major, the newborn is managed as a transfusion-dependent thalassaemia, with monthly transfusions and iron chelation; haematopoietic stem cell transplantation (HSCT) is curative and should be offered as early as possible if a donor is available, because outcomes are better before organ dysfunction from iron overload [17, 30]. After the switch from the γ to the β chain, the red cells come to contain chiefly HbH (β4), which is non-functional, requiring a more aggressive transfusion regimen and leading to early and extensive iron overload [17]. In the UCSF in utero transplantation trial, all infants became dependent on postnatal transfusions every 2–4 weeks, with a target functional haemoglobin, corrected for haemoglobin Bart's, above 10 g/dL; chelation began in all those transplanted in utero at a median of 15 months (14–23) and in 6/10 of those who received IUT only at a median of 17 months (9–24); hepatic iron was raised on magnetic resonance imaging at 9–18 months and cardiac iron was normal; three children from the IUT-only group underwent successful bone marrow transplantation [145]. Hypospadias, frequent in male fetuses with alpha-thalassaemia major, is corrected surgically [30].

In MPS VII, vestronidase alfa (Mepsevii®), recombinant human lysosomal β-glucuronidase, is the only disease-modifying therapy [18, 25]. The recommended dose, in paediatric and adult patients, is 4 mg/kg by intravenous infusion every two weeks, with no maximum dose defined in the source, administered over about 4 hours, with 2.5% of the total volume in the first hour and the remainder over the following three hours as tolerated; GeneReviews indicates rounding the dose up to the next whole vial [18, 25]. Premedication with a non-sedating antihistamine, with or without an antipyretic, 30 to 60 minutes before the infusion is recommended [25]. The FDA label carries a boxed warning for anaphylaxis, which may occur from the first dose: administration requires medical support to be available, observation during the infusion and for 60 minutes afterwards, and immediate discontinuation if anaphylaxis occurs [25]. In the phase III trial, with 12 patients, the mean reduction in urinary dermatan sulfate was 64.8% and in chondroitin sulfate 70.6% at 24 weeks compared with placebo (P <0.0001); 10 of the 12 patients had clinically relevant improvement in at least one domain of a multidomain responder index; 0.9% (2/215) of infusions caused hypersensitivity reactions and 7 of the 12 patients developed anti-drug antibodies without impact on GAG reduction [18]. The long-term extension and an open-label phase II study in eight children under 5 years showed similar reductions [18]. Vestronidase does not cross the blood-brain barrier, and the effect on central nervous system manifestations has not been determined [18, 25]. An implanted central venous catheter facilitates prolonged treatment [18]. Haematopoietic stem cell transplantation has had variable results in MPS VII, and in one Brazilian patient bone marrow transplantation had fatal complications [18, 64]. More than 40% of the lysosomal disorders most frequent in NIHF have potential postnatal treatment, which is an argument for aetiological diagnosis even when the pregnancy continues [96].

In Gaucher disease type 2 and in patients with hydrops fetalis, enzyme replacement therapy, substrate reduction therapy and HSCT are not indicated: they may prolong life, but do not alter the severe neurological manifestations [22]. In PIEZO1-related dehydrated stomatocytosis, splenectomy should be avoided because it aggravates the risk of thrombosis, and iron overload, almost universal even without transfusions, requires monitoring [9]. In transaldolase deficiency, liver transplantation was necessary in four of 16 patients in a multicentre cohort, and two died weighing less than 4 kg without access to transplantation [107].

7.11. Comparative regulatory status

7.12. Trials and intervention studies

Sources: [18, 55, 145, 158, 163, 167, 168, 171, 177–181].

Studies of fetal therapy are almost entirely case series, registries, retrospective cohorts and single-arm trials; there are no published randomised trials in NIHF, and the only relevant randomised trial (FAST) has not published results. Comparing rates between studies of different design and population is not valid: survival of 85% after IUT in parvovirus and survival without severe injury of 59% in the European cohort measure different outcomes in different fetuses. The comparison group for in utero transplantation is contemporaneous but not randomised, and the number of IUTs was higher in the transplanted group (6.8 versus 4.1; P=0.004), which precludes attributing differences to the transplantation [145]. MPS VII and galactosialidosis present with NIHF in utero and are theoretical targets of fetal enzyme therapy, but current replacement enzymes do not cross the blood-brain barrier [182]. The relative reduction of GAG with vestronidase is a biochemical outcome in 12 patients and not a clinical survival outcome [18]. The in utero transplantation trial and the amniotic fluid spectrometry trial are negative results, and the absence of publication of FAST does not prove the absence of results.

7.13. Developments of the last 12 months

The GeneReviews text on alpha-thalassaemia describes luspatercept in the context of HbH disease and adult alpha-thalassaemia; the datum was not confirmed in a primary source in the preparation of this article and does not apply to the hydropic fetus [17].

7.14. Pregnancy, lactation and agents to avoid

The drugs of fetal therapy are administered to the pregnant woman, who is the one who bears the adverse effects: in the Japanese trial, 78.0% of mothers had adverse effects [167]. Sotalol requires attention to maternal QT prolongation, and pharmacokinetic modelling suggests that doses of 480 mg/day exceed the maternal toxicity threshold in 16% of modelled pregnant women, above the maximum of 320 mg/day of the adult FDA label outside refractory ventricular arrhythmias [161, 165]. Amiodarone is associated with thyroid dysfunction, mostly transient and neonatal [166]. The FDA label for sotalol advises against breastfeeding, flecainide passes into milk at concentrations above plasma levels and digoxin requires monitoring for toxicity in the newborn [159–161]. There are no human data on vestronidase alfa in pregnancy; in rats an increase in stillbirths was observed at exposures below the human dose, of uncertain clinical relevance [25]. In PIEZO1-related dehydrated stomatocytosis, splenectomy is the procedure to be avoided [9].

7.15. Guideline recommendations and divergences

Sources: [1, 26–30, 38, 144].

8. Reproductive Options and Family Counselling

8.1. Principle: the recurrence risk depends on the diagnosis

NIHF has no recurrence risk of its own; the risk is that of the cause, and every effort should be directed at the specific diagnosis so that counselling can be based on it [3]. When no laboratory modality identifies the aetiology, parents should be counselled that the recurrence risk may be low if the cause is a de novo event, 25% if it is an autosomal recessive disorder and up to 50% if it is an autosomal dominant disorder with reduced penetrance or variable expressivity [28]. In exome series, most dominant variants are de novo (88% in the Sparks cohort; 86.3% in the Mone meta-analysis) and almost all recessive ones are inherited (95%) [12, 14]. In the meta-analysis of lysosomal disorders in NIHF, all variants were inherited, with a recurrence risk of 25%, and recurrence of hydrops was 75% (15/20) in the families where it was reported [97].

8.2. Cascade testing and carrier screening

Identification of a pathogenic or likely pathogenic variant in the fetus allows other at-risk relatives to be tested and future pregnancies to be planned [28]. In alpha-thalassaemia, couple screening in at-risk populations should be carried out before or in early pregnancy, and the ACMG includes alpha-thalassaemia in the carrier screening to be offered to all pregnant women or to those planning pregnancy; identification of silent carriers is not strongly indicated [17]. Alpha-thalassaemia carriers account for about 5% of the world population and about 40% in the endemic regions of South-East Asia, India and the Mediterranean; red cell indices may miss carriers, and in at-risk populations molecular testing should be offered [30]. The risk is indeterminate when one parent is −−/αα and the other has a thalassaemic haematological picture without an identified variant, or when the father is unknown and belongs to a high-risk population; these situations require formal counselling [17]. When the parents of a previous undiagnosed fetus present for preconception counselling and no sample from the affected fetus exists, sequencing both parents to look for a shared carrier state for a recessive disorder is appropriate, although tissue from the previous fetus is preferable [27, 38]. In relatives of patients with dominant disorders of variable expressivity (FOXC2, FLT4, EPHB4, RASopathies), clinical examination and targeted testing identify carriers with a minimal phenotype [3, 10, 20].

8.3. Germline mosaicism

In de novo RASopathies, the recurrence risk is low but not zero, because of the theoretical possibility of gonadal mosaicism [49]. The sources consulted do not quantify this risk for NIHF in general.

8.4. Prenatal diagnosis and preimplantation genetic testing

With a known familial variant, molecular genetic testing is the method of choice for prenatal diagnosis and carrier detection; in lysosomal disorders, enzyme assay in chorionic villi or cultured amniocytes remains available, with the limitation of pseudodeficiency alleles [18, 64]. In MPS VII, the pseudodeficiency p.(Asp152Asn) may mimic disease on enzyme assay and in heterozygote detection [18, 64]. In alpha-thalassaemia, prenatal diagnosis and preimplantation genetic testing for a monogenic disorder (PGT-M) are possible once the parents' variants have been identified; PGT-M has limitations and requires prenatal confirmation [17, 30]. An increased NT in the 1st trimester in an at-risk pregnancy should prompt assessment [17]. In Morquio disease with hydrops, enzyme analysis in chorionic villi was used in the subsequent pregnancy [104]. In Noonan syndrome, prenatal diagnosis is by chorionic villus sampling or amniocentesis when the familial variant has been defined; without it, nuchal oedema and polyhydramnios are monitored [49]. In lymphatic dysplasias, prenatal diagnosis is only possible with a variant or anomaly defined in the proband [48]. In lymphoedema-distichiasis, fetal echocardiography and an additional ultrasound in the 3rd trimester are recommended because of the risk of hydrothorax or hydrops [20]. When the couple's variant is a VUS, targeted prenatal diagnosis and preimplantation testing may be refused under laboratory policy, which should be explained to couples [28].

8.5. Content of counselling after a pregnancy with NIHF

Counselling after a pregnancy with NIHF should include the diagnosis (or its absence), the recurrence risk according to the mechanism, reanalysis of the exome before a new pregnancy, the reproductive options (prenatal diagnosis, PGT-M, donor gametes) and written information, in accessible language, that the parents can use in future decisions [28, 38]. Reanalysis of inconclusive results from a fetus or another relative should be carried out, with consent, when a pregnancy is being planned and sufficient time has elapsed since the last report [38]. Storage of DNA, of a fibroblast line and of frozen amniotic fluid is the condition for any future investigation to be possible [3].

8.6. Minors and relatives without symptoms

Adult-onset findings in fetal samples, incidental parental findings and fetal carrier status for recessive disorders unrelated to the indication raise questions about the future child's right not to know; the ACMG recommends not reporting unrelated fetal carrier status and reporting only fetal findings of childhood onset with high penetrance, and the ISPD asks that the policy on adult-onset findings be discussed before testing [28, 38]. In the minor siblings of a patient with a dominant disorder of variable expressivity, testing is justified when the diagnosis alters surveillance in childhood, as in Noonan syndrome or lymphoedema-distichiasis; the sources consulted do not establish specific rules for other situations.

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence and what each figure measures

Estimates of the frequency of NIHF differ because they measure different things: pregnancies with an ultrasound diagnosis, live births with a diagnostic code, NICU admissions or birth prevalence registers.

The SMFM notes that there is little information on the current prevalence of NIHF and on variation between populations [1]. The proportion of NIHF among all hydrops cases (about 90% in the West) is a relative proportion and not a prevalence [1, 3].

9.2. United States

US data dominate the genomic literature: the Sparks cohort and the University of California Consortium, the HYDROPS study and the prospective UCSF cohort are the main sources on exome yield, timing of delivery and maternal complications [12, 15, 16, 24, 141, 148]. In California, neonatal mortality of NIHF was 35.1% (364/1037) and one-year mortality 43.2% (448/1037) [23]. The international alpha-thalassaemia major registry and the in utero fetal therapy trials are also centred on UCSF [55, 145, 177].

9.3. Europe

European data include the UK FIND/PAGE cohort (trio exome yield of 25%), the shunt series from King's College Hospital in London, the multicentre parvovirus B19 cohort from France, Belgium and Czechia and the fetal tachycardia series from Rotterdam [14, 158, 164, 170]. Denmark is the first country among the sources consulted to include NIHF among the national indications for routine prenatal genome sequencing [144]. In the United Kingdom, severe RhD haemolytic disease affects fewer than 100 cases per year among about 620,000 live births [34]. In the Mediterranean, the −−/αα genotype is very rare (frequency 0.002) and haemoglobin Bart's is rarely described; the −α3.7 allele reaches its maximum allele frequency in Sardinia (0.18) and its minimum in Spain, where the prevalence of alpha-thalassaemia in Musallam's review was 0.03/100,000 [17]. Greece and Cyprus are, in the Oxford Desk Reference, other regions in which homozygous alpha-thalassaemia is a cause of fetal loss [54].

9.4. Populations with a founder effect or high consanguinity

In South-East Asia, the −−SEA, −−THAI and −−FIL deletions are very frequent, haemoglobin Constant Spring is common and the expected incidence of haemoglobin Bart's is 0.5–5/1000 births, with HbH disease in 4–20/1000; alpha-thalassaemia explains 28%–55% of NIHF cases in this region [1, 17]. In equatorial Africa, the frequency of −α3.7 is 0.30–0.40, but −−/αα is very rare in North Africa and in African Americans; in the Arabian Peninsula, −α3.7 has a frequency of 0.01–0.67 (maximum in Oman); in tribes of Andhra Pradesh, in India, 0.35–0.92; in Vanuatu a regulatory variant that creates a GATA-1 binding site has been described [17]. In MPS VII, the founder variants are c.1244C>T p.(Pro415Leu) and the complex allele c.[1244C>T;1222C>T] in Mexico, c.1856C>T p.(Ala619Val) in Japan and c.526C>T p.(Leu176Phe) in Brazil [18]. In British Columbia, the estimated incidence of MPS VII was 1/300,000 live births, and the worldwide total of identified patients is fewer than 200 [18, 64]. In Finland, lethal congenital contracture syndrome 1 (GLE1) causes severe hydrops [42]. Cholestasis-lymphoedema syndrome (Aagenaes) is very rare outside Scandinavia [48].

Consanguinity increases exome yield and the fraction of recessive disorders: in the 2023 meta-analysis, the yield was higher in consanguineous families, and in Western Rajasthan consanguinity was present in 69.2% of cases with a genetic diagnosis, against 21.4% (OR 8.25; p=0.005) [11, 147]. In Jerusalem, exome sequencing diagnosed 47% of cases with a negative CMA, including four couples with recurrence [146]. In China, CNV-seq and exome sequencing gave an overall positivity of 51.2% [46].

9.5. Portugal and Brazil

The sources consulted include no epidemiological data on NIHF in Portugal. For Brazil, the available data are the founder variant GUSB c.526C>T p.(Leu176Phe), which accounts for 96% of alleles in a Brazilian MPS VII cohort, almost all in homozygosity, the description of a Brazilian patient with that variant who was attending a special school at 18 years of age and of another in whom bone marrow transplantation had fatal complications, and the participation of Brazilian centres in the FAST observational registry [18, 64, 169]. The absence of data in the sources consulted does not demonstrate the absence of published data.

9.6. Special populations

The pregnant woman with NIHF is herself a patient at risk, from mirror syndrome and pre-eclampsia [1, 141]. Inequality in access to exome and genome sequencing, in insurance coverage and in costs is acknowledged by the SMFM, which accepts panels when sequencing is not feasible [1]. The SMFM also anticipates that recorded spontaneous loss will increase with reduced access to termination of pregnancy [1]. Monochorionic pregnancies have causes of their own (twin-to-twin transfusion, TRAP sequence) [1].

9.7. European Reference Networks and patient associations

There is no ERN dedicated to fetal medicine among those consulted. The relevant networks are ERN ITHACA (congenital malformations, dysmorphic syndromes and intellectual disability; https://ern-ithaca.eu/), ERN-EuroBloodNet (rare haematological diseases, including thalassaemias and congenital anaemias; https://eurobloodnet.eu/), MetabERN (hereditary metabolic diseases, including lysosomal disorders; https://metab.ern-net.eu/), VASCERN (rare vascular diseases and vascular and lymphatic anomalies; https://vascern.eu/) and ERN GUARD-Heart (rare cardiac diseases and inherited arrhythmias; https://guardheart.ern-net.eu/). Patient associations and support organisations include the Thalassaemia International Federation (https://thalassaemia.org.cy/) and the Cooley's Anemia Foundation in the United States (https://www.thalassemia.org/), indicated by the 2021 consensus [30], the Lymphatic Education & Research Network (https://lymphaticnetwork.org/), the RASopathies Network (https://rasopathiesnet.org/), the MPS Society in the United Kingdom (https://www.mpssociety.org.uk/) and the Fetal Medicine Foundation (https://fetalmedicine.org/). The Oxford Desk Reference further indicates, in the United Kingdom, Antenatal Results and Choices (arc-uk.org), for support after a prenatal diagnosis, the Lymphoedema Support Network (lymphoedema.org), the British Lymphology Society (thebls.com) and the Turner Syndrome Support Society (tss.org.uk) [48, 86, 90].

10. Illustrative Clinical Case

Schönfeld and colleagues, from Mainz, Germany, published in 2020 in Molecular Genetics & Genomic Medicine the case of a newborn girl with NIHF in whom a virtual hydrops panel identified Noonan syndrome [83]. The mother was healthy and there was no consanguinity. During the pregnancy there was a prenatal suspicion of heart disease, without increased NT, polyhydramnios or short femur. The hydrops developed rapidly and was diagnosed at 30+0 weeks; the following day, at 30+1 weeks, an emergency caesarean section was performed [83].

The newborn weighed 1400 g (50th centile), measured 40 cm and had a head circumference of 28.5 cm; the Apgar scores were 1, 3 and 4 and the umbilical arterial pH 7.33 [83]. At birth she had NIHF, hypovolaemic shock, severe anaemia (haemoglobin 7.7 g/dL), severe thrombocytopenia (8/nL) and disseminated intravascular coagulation; the pleural and peritoneal cavities were drained immediately and the effusions were haemorrhagic [83]. She had hypotonia, a short webbed neck, and ultrasound showed a unilateral infarct and bilateral grade II intraventricular haemorrhage. Echocardiography revealed double outlet right ventricle and an atrial septal defect with a left-to-right shunt [83].

The neonatal course included mechanical ventilation until day 22, bilateral chest drains for chylothorax, repeatedly replaced up to 197 days, hypertrophic cardiomyopathy treated with propranolol, stenting of the arterial duct and a subocclusion of the bowel caused by a giant Meckel's diverticulum [83]. Infections, coagulopathies, alloimmune thrombocytopenia and familial thrombocytopenia were excluded. Analysis of a virtual hydrops panel of 119 genes, applied to exome data, identified the de novo variant NM_002834.5(PTPN11):c.218C>T, NP_002825.3:p.(Thr73Ile), absent in the parents [81, 83]. The variant is classified in ClinVar as pathogenic, with criteria provided by multiple submitters and no conflicts (VCV000013334) [82]. The panel included lysosomal genes (ARSA, ARSB, CTSA, GAA, GBA, GLB1, GUSB, GALNS, IDS, IDUA, NEU1, among others), RASopathy genes (BRAF, CBL, HRAS, PTPN11 and others), lymphatic genes (CALCRL, CCBE1, EPHB4, FAT4, FLT4, FOXC2, PIEZO1) and HBA1/HBA2, each with the reference transcript indicated in Table 1 of the article [83].

The thrombocytopenia remained severe and required weekly platelet transfusions. Figure 1 of the article (platelet count per nL against day of life) shows a «saw-tooth» pattern, with sharp peaks after each transfusion and a rapid fall, within a few days, to values frequently below 25/nL and close to zero, over about 180 days; the peaks lie approximately between 170 and 225/nL and, between days 71 and 127, are generally lower, with deeper nadirs [83]. The numerical values are approximate visual readings from the figure and not data tabulated by the authors. No blasts suggestive of transient myeloproliferative disorder or of juvenile myelomonocytic leukaemia (JMML) were found in peripheral blood; monocytes, which were 37% at birth, fell to 18%, with a slightly progressive trend [83]. The parents declined bone marrow aspiration and treatment with 6-mercaptopurine, mentioned in the literature cited by the authors. The child was discharged on day 264, without oxygen [83].

The authors discuss that p.(Thr73Ile) is associated with heart disease, persistent thrombocytopenia, a risk of myeloproliferative disease and JMML in the first five years, solid tumours such as neuroblastoma and amegakaryocytosis, and present the case as the first with this variant and NIHF [83]. The case illustrates several points of this article: the RASopathy as the most frequent cause of NIHF with a monogenic diagnosis, the late and rapid onset of the hydrops, the chylous nature of the effusions, the usefulness of a virtual panel applied to exome data for a rapid diagnosis and the need for haematological surveillance in Noonan syndrome with PTPN11 variants associated with myeloproliferation. The management in the report (for example, propranolol for the cardiomyopathy and weekly transfusions) represents individual decisions and not a treatment protocol; investigation by virtual panel, rather than full exome, is one of the strategies accepted by the SMFM where speed is wanted, but the 2026 document prefers exome or genome sequencing to panels [1].

11. Prognosis and Long-Term Follow-up

11.1. Overall prognosis

The prognosis of NIHF is that of its cause, modulated by the severity of the hydrops and by gestational age. Overall, 15%–21% of pregnancies end in miscarriage or fetal death, and survival of live-born infants ranges between 36% and 68% [1]. Among live-born infants, the risk of neurological impairment reaches 48% and that of major morbidity 39% [1]. In California, neonatal mortality was 35.1% and first-year mortality 43.2% [23]. In the Pediatrix NICU database, 215 of the 598 newborns with hydrops died before discharge, 267 were discharged and 115 were transferred; mortality was highest in congenital anomalies (57.7%) and lowest in congenital chylothorax (5.9%), and lower gestational age, a low 5-minute Apgar and greater support on the first day (inspired oxygen fraction, high-frequency ventilation) were independent factors for death [85]. The 2015 guideline already noted that aneuploidy has a poor prognosis and that, even in euploid fetuses, neonatal survival is often below 50% [26]. These figures mix different eras, definitions and populations, and none of them is an individual estimate for a particular pregnancy.

11.2. Prognostic factors

The prognostic factors described by the SMFM are the aetiology, the presence of other anomalies or of a genetic diagnosis, the number and type of effusions, spontaneous improvement, the response to intervention, the gestational age at diagnosis and at delivery, birth weight, the need for resuscitation, low serum albumin and severe acidaemia [1]. Lymphatic dysplasia, tachyarrhythmias, parvovirus and lung masses have a better prognosis, including for neurodevelopment; trisomies 13 and 18, inborn errors of metabolism and heart disease have a worse one [1]. In utero intervention may improve outcomes, as in thoracentesis for primary chylothorax and IUT in alpha-thalassaemia major [1]. In fetal SVT, persistence of the arrhythmia after birth is more likely with a late diagnosis and not with hydrops alone [1]. In the Japanese follow-up, neurodevelopmental impairment affected two of the three fetuses with hydrops and subcutaneous oedema [168]. In the Qin series, diagnosis by 13+6 weeks was associated with greater genetic positivity (61.4% versus 28.9%), largely because of aneuploidy [46].

11.3. Prognosis by aetiology

11.4. Surveillance

Antepartum surveillance has not been shown to improve outcomes, and the assessment intervals in pregnancy are not fixed by trials; the table brings together the recommendations of the sources consulted, with the action threshold of each.

11.5. Prognosis with treatment and transition

The prognosis has changed in treatable causes: alpha-thalassaemia major, previously universally fatal, now has transfused survivors without clinically significant learning difficulties, and HSCT may be curative [17, 55]. Survivors of lymphatic dysplasia need follow-up of the lymphoedema and the chylous effusions; those with a RASopathy, cardiological, haematological and developmental follow-up; and those with a lysosomal disorder, metabolic follow-up with the available therapies [9, 19, 96]. In transaldolase deficiency, follow-up to 20 years showed universal hepatic involvement, chronic kidney disease in 5 of 16, chronic cytopenia in 12, hypergonadotrophic hypogonadism in 7 and cardiomyopathy or ventricular dysfunction in 5, with a 2026 review bringing together 47 published cases and three adults [107]. Transition to adult care follows the rules of each entity.

12. Summary and Final Considerations

Non-immune hydrops fetalis is defined by the presence of abnormal fluid in at least two fetal compartments without red cell alloimmunisation, and the concept of the NIHF spectrum extends the same assessment to a single effusion of presumed systemic cause. It is not a disease, but the common final pathway of different mechanisms: heart failure from structural heart disease, arrhythmia, anaemia or a high-output shunt, lymphatic obstruction or dysplasia, hepatic or renal hypoproteinaemia, increased capillary permeability and compression by masses. The reported frequency is 1/1700–1/3000 pregnancies, NIHF accounts for about 90% of hydrops cases in the West and, among live births in California, the incidence was 2.5/10,000, with one-year mortality of 43.2%.

The work-up begins with exclusion of immune hydrops, with a repeated indirect Coombs test, and continues with the maternal full blood count, syphilis serology, the Kleihauer-Betke test when the MCA is elevated, detailed ultrasound, MCA Doppler and fetal echocardiography. Invasive testing with CMA, with or without karyotype, and PCR for infections is the first genetic test; when it is not diagnostic, the SMFM has recommended since 2026 exome or genome sequencing, which identifies a monogenic disorder in about a third of unexplained cases, with the RASopathies foremost, followed by neuromuscular, metabolic and lymphatic disorders. Panels miss a substantial part of the diagnoses that the exome achieves, and the genome adds little after a negative exome. The SMFM's inconsistency in the MCV threshold for alpha-thalassaemia testing (text «near or above 80 fL», Figure 2 «near or ≤80 fL») should be read in the light of the physiology of alpha-thalassaemia trait and of the 2021 consensus, which uses <70 fL for a different purpose.

The aetiology determines the treatment. Intrauterine transfusion treats the anaemia of parvovirus B19 and of alpha-thalassaemia major, in which two or more transfusions are associated with normal neurodevelopment in most survivors; maternally administered antiarrhythmics (digoxin, sotalol, flecainide, all off-label) treat tachyarrhythmias; drainage and thoracoamniotic shunts treat large effusions; maternal corticosteroids treat microcystic CPAM; and cord ablation treats the TRAP sequence. Vestronidase alfa is the postnatal enzyme therapy for MPS VII, the most frequent lysosomal disorder in NIHF series, and more than 40% of the lysosomal disorders found have potential postnatal treatment. In utero haematopoietic transplantation in alpha-thalassaemia major proved safe but was terminated for lack of efficacy, and in utero enzyme therapy is in a phase 1 trial.

The pregnant woman runs risks of her own — mirror syndrome, pre-eclampsia, preterm birth in up to 86% and caesarean section in up to 85% — and should receive individualised counselling with all the options, including termination. Preterm birth is reserved for obstetric indications, antenatal corticosteroids are given when delivery is expected within seven days and resuscitation is wanted, and delivery takes place in a centre with a level III or IV NICU when the cause is treatable or unknown. The recurrence risk is that of the cause: low in de novo variants and in non-genetic causes, 25% in recessive disorders, such as alpha-thalassaemia major and the lysosomal disorders, and up to 50% in inherited dominant disorders. Storage of DNA, fibroblasts and amniotic fluid, autopsy with placental examination and reanalysis of the exome before a new pregnancy are the measures that make diagnosis possible when the initial investigation does not achieve it.

13. References

1. Society for Maternal‐Fetal Medicine (SMFM), Sparks TN, Norton ME, SMFM Publications Committee. Society for Maternal-Fetal Medicine Consult Series #75: Evaluation and management of non-immune hydrops fetalis. Pregnancy (Hoboken). 2026;2(2):e70208. doi:10.1002/pmf2.70208. PMID: 42596935. PMC13344273.
2. Orphanet. Immune hydrops fetalis (ORPHA:364013) [Internet]. Paris: INSERM; [accessed 17 Oct 2026]. Available from: https://www.orpha.net/en/disease/detail/364013
3. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Oedema—increased nuchal translucency, cystic hygroma, and hydrops. p. 768–769.
4. Orphanet. Non-immune hydrops fetalis (ORPHA:363999) [Internet]. Paris: INSERM; [accessed 17 Oct 2026]. Available from: https://www.orpha.net/en/disease/detail/363999
5. Orphanet. Hydrops fetalis (ORPHA:1041) [Internet]. Paris: INSERM; [accessed 17 Oct 2026]. Available from: https://www.orpha.net/en/disease/detail/1041
6. Orphanet. Hemoglobin Bart's fetalis syndrome (ORPHA:163596) [Internet]. Paris: INSERM; [accessed 17 Oct 2026]. Available from: https://www.orpha.net/en/disease/detail/163596
7. Orphanet. PIEZO1-related generalized lymphatic dysplasia with non-immune hydrops fetalis (ORPHA:568062) [Internet]. Paris: INSERM; [accessed 17 Oct 2026]. Available from: https://www.orpha.net/en/disease/detail/568062
8. Orphanet. EPHB4-related lymphatic-related hydrops fetalis (ORPHA:568065) [Internet]. Paris: INSERM; [accessed 17 Oct 2026]. Available from: https://www.orpha.net/en/disease/detail/568065
9. GeneGuidance. PIEZO1 [curated database; version of 2026-09-26].
10. GeneGuidance. EPHB4 [curated database; version of 2026-09-26].
11. Al-Kouatly HB, Shivashankar K, Mossayebi MH, Makhamreh M, Critchlow E, Gao Z, et al. Diagnostic yield from prenatal exome sequencing for non-immune hydrops fetalis: A systematic review and meta-analysis. Clin Genet. 2023;103(5):503-512. doi:10.1111/cge.14309. PMID: 36757664.
12. Sparks TN, Lianoglou BR, Adami RR, Pluym ID, Holliman K, Duffy J, et al. Exome Sequencing for Prenatal Diagnosis in Nonimmune Hydrops Fetalis. N Engl J Med. 2020;383(18):1746-1756. doi:10.1056/nejmoa2023643. PMID: 33027564. PMC7650529.
13. Quinn AM, Valcarcel BN, Makhamreh MM, Al-Kouatly HB, Berger SI. A systematic review of monogenic etiologies of nonimmune hydrops fetalis. Genet Med. 2021;23(1):3-12. doi:10.1038/s41436-020-00967-0. PMID: 33082562. PMC7796968.
14. Mone F, Eberhardt RY, Hurles ME, Mcmullan DJ, Maher ER, Lord J, et al. Fetal hydrops and the Incremental yield of Next-generation sequencing over standard prenatal Diagnostic testing (FIND) study: prospective cohort study and meta-analysis. Ultrasound Obstet Gynecol. 2021;58(4):509-518. doi:10.1002/uog.23652. PMID: 33847422. PMC8487902.
15. Al-Kouatly HB, Makhamreh MM, Rice SM, Smith K, Harman C, Quinn A, et al. High diagnosis rate for nonimmune hydrops fetalis with prenatal clinical exome from the Hydrops-Yielding Diagnostic Results of Prenatal Sequencing (HYDROPS) Study. Genet Med. 2021;23(7):1325-1333. doi:10.1038/s41436-021-01121-0. PMID: 33686258.
16. Rice SM, Makhamreh MM, Wodoslawsky S, Toro M, Critchlow E, Firman B, et al. High Diagnosis Rate for Nonimmune Hydrops Fetalis With Prenatal Clinical Genome: Expanded Results From the Hydrops-Yielding Diagnostic Results of Prenatal Sequencing (HYDROPS) Study. Am J Med Genet A. 2026. doi:10.1002/ajmg.a.70265. PMID: 42624826.
17. Tamary H, Greenberg-Kushnir N, Dgany O. Alpha-Thalassemia. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2005 Nov 1 [updated 2026 Apr 23]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1435/
18. Sun A, Wang R. Mucopolysaccharidosis Type VII. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2024 Jan 4. Available from: https://www.ncbi.nlm.nih.gov/books/NBK598990/
19. Roberts AE. Noonan Syndrome. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2001 Nov 15 [updated 2021 Dec 16]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1124/
20. Mansour S, Brice GW, Jeffery S, Mortimer P. Lymphedema-Distichiasis Syndrome. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2005 Mar 29 [updated 2019 Apr 4]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1457/
21. Van Zanten M, Mansour S, Ostergaard P, Mortimer P, Gordon K. Milroy Disease. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2006 Apr 27 [updated 2021 Feb 18]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1239/
22. Hughes DA, Pastores GM. Gaucher Disease. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2000 Jul 27 [updated 2023 Dec 7]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1269/
23. Steurer MA, Peyvandi S, Baer RJ, MacKenzie T, Li BC, Norton ME, et al. Epidemiology of Live Born Infants with Nonimmune Hydrops Fetalis-Insights from a Population-Based Dataset. J Pediatr. 2017;187:182-188.e3. doi:10.1016/j.jpeds.2017.04.025. PMID: 28533037.
24. Swanson K, Norton ME, Downum SL, Gonzalez-Velez JM, Sparks TN. Understanding Preterm Birth in Pregnancies Complicated by Nonimmune Hydrops Fetalis. Am J Perinatol. 2023;40(9):917-922. doi:10.1055/a-2008-2495. PMID: 36603834. PMC10330491.
25. U.S. Food and Drug Administration. MEPSEVII (vestronidase alfa-vjbk) injection — prescribing information (BLA761047; initial approval 15/11/2017); revised 18/11/2025. Available from: https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=761047
26. Society for Maternal-Fetal Medicine (SMFM), Norton ME, Chauhan SP, Dashe JS. Society for maternal-fetal medicine (SMFM) clinical guideline #7: nonimmune hydrops fetalis. Am J Obstet Gynecol. 2015;212(2):127-139. doi:10.1016/j.ajog.2014.12.018. PMID: 25557883.
27. Van den Veyver IB, Chandler N, Wilkins-Haug LE, Wapner RJ, Chitty LS, ISPD Board of Directors. International Society for Prenatal Diagnosis Updated Position Statement on the use of genome-wide sequencing for prenatal diagnosis. Prenat Diagn. 2022;42(6):796-803. doi:10.1002/pd.6157. PMID: 35583085. PMC11220784.
28. Monaghan KG, Leach NT, Pekarek D, Prasad P, Rose NC, ACMG Professional Practice and Guidelines Committee. The use of fetal exome sequencing in prenatal diagnosis: a points to consider document of the American College of Medical Genetics and Genomics (ACMG). Genet Med. 2020;22(4):675-680. doi:10.1038/s41436-019-0731-7. PMID: 31911674.
29. Lazier J, Hartley T, Brock JA, Caluseriu O, Chitayat D, Laberge AM, et al. Clinical application of fetal genome-wide sequencing during pregnancy: position statement of the Canadian College of Medical Geneticists. J Med Genet. 2022;59(10):931-937. doi:10.1136/jmedgenet-2021-107897. PMID: 34544840. PMC9554053.
30. MacKenzie TC, Amid A, Angastiniotis M, Butler C, Gilbert S, Gonzalez J, et al. Consensus statement for the perinatal management of patients with α thalassemia major. Blood Adv. 2021;5(24):5636-5639. doi:10.1182/bloodadvances.2021005916. PMID: 34749399. PMC8714716.
31. Diamond LK, Blackfan KD, Baty JM. Erythroblastosis fetalis and its association with universal edema of the fetus, icterus gravis neonatorum and anemia of the newborn. J Pediatr. 1932;1(3):269-309. doi:10.1016/S0022-3476(32)80057-0.
32. Potter EL. Universal edema of the fetus unassociated with erythroblastosis. Am J Obstet Gynecol. 1943;46(1):130-134. doi:10.1016/S0002-9378(16)40458-8.
33. Macafee CA, Fortune DW, Beischer NA. Non-immunological hydrops fetalis. J Obstet Gynaecol Br Commonw. 1970;77(3):226-237. doi:10.1111/j.1471-0528.1970.tb03511.x. PMID: 4911699.
34. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Fetomaternal alloimmunization. p. 736–737.
35. Bellini C, Hennekam RC, Fulcheri E, Rutigliani M, Morcaldi G, Boccardo F, et al. Etiology of nonimmune hydrops fetalis: a systematic review. Am J Med Genet A. 2009;149A(5):844-851. doi:10.1002/ajmg.a.32655. PMID: 19334091.
36. Bellini C, Donarini G, Paladini D, Calevo MG, Bellini T, Ramenghi LA, et al. Etiology of non-immune hydrops fetalis: An update. Am J Med Genet A. 2015;167A(5):1082-1088. doi:10.1002/ajmg.a.36988. PMID: 25712632.
37. Holzgreve W, Curry CJ, Golbus MS, Callen PW, Filly RA, Smith JC. Investigation of nonimmune hydrops fetalis. Am J Obstet Gynecol. 1984;150(7):805-812. doi:10.1016/0002-9378(84)90453-8. PMID: 6391171.
38. International Society for Prenatal Diagnosis, Society for Maternal and Fetal Medicine, Perinatal Quality Foundation. Joint Position Statement from the International Society for Prenatal Diagnosis (ISPD), the Society for Maternal Fetal Medicine (SMFM), and the Perinatal Quality Foundation (PQF) on the use of genome-wide sequencing for fetal diagnosis. Prenat Diagn. 2018;38(1):6-9. doi:10.1002/pd.5195. PMID: 29315690.
39. Boutin RCT, Lianoglou BR, Sahin Hodoglugil NN, Tick K, Santoli CMA, Hodoglugil U, et al. Incremental yield of genome sequencing after exome sequencing for nonimmune hydrops fetalis spectrum. Am J Obstet Gynecol MFM. 2026;8(9):102067. doi:10.1016/j.ajogmf.2026.102067. PMID: 42471198. PMC13621888.
40. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, Clinical approach to inborn errors of metabolism in paediatrics, sections 1.1.1.2 (Table 1.1), 1.2.1, 1.2.2, 1.3.1, 1.3.5.2, 1.3.5.5, 1.3.5.6, 1.5.
41. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Triploidy. p. 694–695.
42. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Fetal akinesia. p. 724–727.
43. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Chapter 91, Niemann-Pick disease. p. 688–697.
44. Bayrak-Toydemir P, Stevenson DA. Capillary Malformation-Arteriovenous Malformation Syndrome. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2011 Feb 22 [updated 2019 Sep 12]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK52764/
45. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Congenital cystic lung lesions, Currarino syndrome, and sacrococcygeal teratoma. p. 712–713.
46. Qin Y, Li W, Zhu S, Zhang S, Fang R, Kang Q, et al. Application of next-generation sequencing in nonimmune hydrops fetalis and its impact on pregnancy decisions. BMC Pregnancy Childbirth. 2026;26(1):899. doi:10.1186/s12884-026-09428-z. PMID: 42265713. PMC13474967.
47. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Noonan syndrome. p. 148–151.
48. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Oedema—generalized or puffy extremities. p. 260–265.
49. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Noonan syndrome and the Ras/MAPK pathway syndromes. p. 512–515.
50. Gripp KW, Weaver KN. HRAS-Related Costello Syndrome. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2006 Aug 29 [updated 2023 Dec 21]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1507/
51. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Hennekam lymphedema-lymphangiectasia syndrome. p. 858–863.
52. GeneGuidance. CCBE1 [curated database; version of 2026-09-26].
53. GeneGuidance. FAT4 [curated database; version of 2026-09-26].
54. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Haemoglobinopathies. p. 436–441.
55. Schwab ME, Lianoglou BR, Gano D, Gonzalez Velez J, Allen IE, Arvon R, et al. The impact of in utero transfusions on perinatal outcomes in patients with alpha thalassemia major: the UCSF registry. Blood Adv. 2023;7(2):269-279. doi:10.1182/bloodadvances.2022007823. PMID: 36306387. PMC9860434.
56. Singha K, Yamsri S, Sanchaisuriya K, Fucharoen G, Fucharoen S. Prenatal diagnostic errors in hemoglobin Bart's hydrops fetalis caused by rare genetic interactions of α-thalassemia. Diagnosis (Berl). 2025;12(2):256-261. doi:10.1515/dx-2024-0114. PMID: 39286901.
57. Sieff C. DBA Syndrome. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2009 Jun 25 [updated 2025 Jul 31]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK7047/
58. Takasaki K, Kacena MA, Raskind WH, Weiss MJ, Chou ST. GATA1-Related Cytopenia. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2006 Nov 22 [updated 2023 Feb 16]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1364/
59. Tamary H, Dgany O. Congenital Dyserythropoietic Anemia Type I. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2009 Apr 21 [updated 2021 Jul 29]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK5313/
60. GeneGuidance. KLF1 [curated database; version of 2026-09-26].
61. Makhamreh MM, Rice SM, Shivashankar K, Brewer CJ, McLaren RA, Berger SI, et al. Hereditary Anemias as a Monogenic Etiology for Nonimmune Hydrops Fetalis. Clin Ther. 2026;48(7):592-595. doi:10.1016/j.clinthera.2026.01.002. PMID: 41638966.
62. Erwin A, Balwani M, Desnick RJ. Congenital Erythropoietic Porphyria. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2013 Sep 12 [updated 2021 Apr 15]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK154652/
63. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 41, sections 41.2.1 e 41.3.1.
64. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Chapter 82, Sly disease/β-glucuronidase deficiency/mucopolysaccharidosis VII. p. 608–612.
65. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Chapter 97, Galactosialidosis. p. 733–740.
66. Adams D, Huizing M, Wasserstein M. Free Sialic Acid Storage Disorder. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2003 Jun 13 [updated 2025 Jun 26]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1470/
67. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 5.1 Dysostosis multiplex. p. 109–116.
68. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 40, sections 40.2.1, 40.2.3, 40.3.4, 40.4, 40.5.
69. Lam C, Krasnewich DM. PMM2-CDG. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2005 Aug 15 [updated 2021 May 20]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1110/
70. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Chapter 101, PMM2-CDG. p. 771–777.
71. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Chondrodysplasia punctata. p. 98–101.
72. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 7, Disorders of glycolysis and the pentose phosphate pathway, section 7.10.1.
73. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 20, section 20.4.1 S-adenosylhomocysteine hydrolase deficiency.
74. Weaver KN, Sullivan B, Hildebrandt F, Strober J, Prasad R, Saba J, et al. Sphingosine Phosphate Lyase Insufficiency Syndrome. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2020 Oct 15 [updated 2026 Apr 9]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK562988/
75. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 39, Disorders of nucleic acid metabolism, tRNA metabolism and ribosomal biogenesis, section 39.1.
76. Chmiela T, Wszolek ZK. AARS2-Related Disorder. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2024 Oct 31. Available from: https://www.ncbi.nlm.nih.gov/books/NBK608563/
77. Sloan JL, Carrillo N, Adams D, Venditti CP. Disorders of Intracellular Cobalamin Metabolism. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2008 Feb 25 [updated 2018 Sep 6]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1328/
78. Ferreira C, Pierre G, Thompson R, Vernon H. Barth Syndrome. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2014 Oct 9 [updated 2026 Jan 8]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK247162/
79. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 37, section 37.6 Sterol Δ14-reductase deficiency (HEM/Greenberg).
80. Goldstein A, Falk MJ. Single Large-Scale Mitochondrial DNA Deletion Syndromes. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2003 Dec 17 [updated 2023 Sep 28]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1203/
81. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: Accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-68. doi:10.1002/humu.23348. Tool: https://variantvalidator.org/
82. National Center for Biotechnology Information. ClinVar; VCV000013334, NM_002834.5(PTPN11):c.218C>T (p.Thr73Ile). Available from: https://www.ncbi.nlm.nih.gov/clinvar/variation/13334/
83. Schönfeld M, Selig M, Russo A, Lindner C, Kampmann C, Mildenberger E, et al. Rapid detection by hydrops panel of Noonan syndrome with PTPN11 mutation (p.Thr73Ile) and persistent thrombocytopenia. Mol Genet Genomic Med. 2020;8(5):e1174. doi:10.1002/mgg3.1174. PMID: 32144894. PMC7216795.
84. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Morquio syndrome (MPS IV). p. 652–655.
85. Abrams ME, Meredith KS, Kinnard P, Clark RH. Hydrops fetalis: a retrospective review of cases reported to a large national database and identification of risk factors associated with death. Pediatrics. 2007;120(1):84-89. doi:10.1542/peds.2006-3680. PMID: 17606565.
86. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Turner syndrome, 45,X and variants. p. 696–699.
87. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Sex chromosome mosaicism. p. 688–689.
88. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Chromosomal mosaicism—prenatal. p. 654–659.
89. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Cardio-facio-cutaneous syndrome. p. 156–159.
90. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Polyhydramnios. p. 774–775.
91. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Distichiasis-lymphedema syndrome. p. 856–857.
92. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Parkes Weber syndrome. p. 728–731.
93. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Hypoplastic left heart. p. 740–741.
94. Vergote S, Van der Veeken L, Chitayat D, Jaeggi E, Ryan G, Miller E, et al. Impact of prenatal sirolimus on cardiac rhabdomyomas and brain tubers. Ultrasound Obstet Gynecol. 2026;67(3):350-359. doi:10.1002/uog.70185. PMID: 41674351. PMC12951259.
95. Ziegler SG, Ferreira CR. Generalized Arterial Calcification of Infancy. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2014 Nov 13 [updated 2026 May 19]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK253403/
96. Iyer NS, Gimovsky AC, Ferreira CR, Critchlow E, Al-Kouatly HB. Lysosomal storage disorders as an etiology of nonimmune hydrops fetalis: A systematic review. Clin Genet. 2021;100(5):493-503. doi:10.1111/cge.14005. PMID: 34057202. PMC12107385.
97. Makhamreh MM, Shivashankar K, Rice SM, Wodoslawsky S, Grant C, McLaren R, et al. Lysosomal storage disorders in nonimmune hydrops fetalis diagnosed by exome sequencing. Orphanet J Rare Dis. 2025;20(1):351. doi:10.1186/s13023-025-03851-9. PMID: 40640912. PMC12243238.
98. Whybra C, Mengel E, Russo A, Bahlmann F, Kampmann C, Beck M, et al. Lysosomal storage disorder in non-immunological hydrops fetalis (NIHF): more common than assumed? Report of four cases with transient NIHF and a review of the literature. Orphanet J Rare Dis. 2012;7:86. doi:10.1186/1750-1172-7-86. PMID: 23137060. PMC3552949.
99. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Chapter 90, Gaucher disease. p. 678–687.
100. Regier DS, Tifft CJ, Rothermel CE. GLB1-Related Disorders. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2013 Oct 17 [updated 2021 Apr 22]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK164500/
101. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 5.3 Mucolipidosis II. p. 124–128.
102. Erdem F, Canda E. GNPTAB-Related Disorders. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2008 Aug 26 [updated 2026 Aug 13]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1828/
103. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Appendix: Differential diagnosis of clinical phenotypes. p. 815–828 (p. 821 consulted).
104. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Chapter 80, Morquio syndrome. p. 590–599.
105. Magoulas PL, El-Hattab AW. Glycogen Storage Disease Type IV. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2013 Jan 3 [updated 2019 Aug 1]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK115333/
106. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 5, The glycogen storage diseases and related disorders, section 5.1.4.
107. GeneGuidance. TALDO1 [curated database; version of 2026-09-26].
108. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 33, Disorders of haem biosynthesis, section 33.2.3.
109. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 10.1 Greenberg dysplasia. p. 351–352.
110. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. Chapter 41, LCHAD (trifunctional protein) deficiency. p. 310–316.
111. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Lethal multiple pterygium syndrome. p. 240–241.
112. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Short limbs. p. 784–787.
113. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 14.10 Cole-Carpenter syndrome. p. 532–534.
114. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 7.1 Achondrogenesis type 1A. p. 157–159.
115. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Achondrogenesis types IA and IB. p. 460–461.
116. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 8.1 Achondrogenesis type IB. p. 245–246.
117. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 4.1 Achondrogenesis II, hypochondrogenesis. p. 65–67.
118. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Achondrogenesis type II-hypochondrogenesis. p. 462–463.
119. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Fibrochondrogenesis. p. 464–465.
120. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 7.3 Schneckenbecken dysplasia. p. 163–164.
121. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 11.3–11.5 Short rib (±polydactyly) syndromes. p. 390–397.
122. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 11.6 Cranioectodermal dysplasia. p. 398–402.
123. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 14.3 Osteogenesis imperfecta, type IIA. p. 506–508.
124. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Osteogenesis imperfecta. p. 682–687.
125. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 17.1 Blomstrand chondrodysplasia. p. 621–622.
126. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 17.2 Infantile cortical hyperostosis. p. 623–626.
127. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 17.3 Dysplastic cortical hyperostosis type Kozlowski-Tsuruta. p. 627–628.
128. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. 21.3 Osteocraniostenosis. p. 802–803.
129. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Chondrodysplasia punctata, X-linked dominant type. p. 530–533.
130. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Appendix (pattern of malformation as diagnostic aid). p. 1053.
131. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Simpson-Golabi-Behmel syndrome. p. 224–225.
132. Nisbet AF, Hathaway ER, Kalish JM. Simpson-Golabi-Behmel Syndrome Type 1. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2006 Dec 19 [updated 2023 Dec 7]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1219/
133. Tan QKG, Louie RJ, Sleasman JW. IPEX Syndrome. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2004 Oct 19 [updated 2024 Feb 1]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1118/
134. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Unusual hair, teeth, nails and skin. p. 330–332.
135. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Mulibrey nanism syndrome. p. 120–121.
136. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Desbuquois dysplasia. p. 632–633.
137. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Yunis-Varón syndrome. p. 630–631.
138. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Twins and twinning. p. 792–795.
139. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Monozygotic twinning and structural defects. p. 914–921.
140. Mardy AH, Chetty SP, Norton ME, Sparks TN. A system-based approach to the genetic etiologies of non-immune hydrops fetalis. Prenat Diagn. 2019;39(9):732-750. doi:10.1002/pd.5479. PMID: 31087399. PMC6699893.
141. Gulrajani NB, Robles DC, Gonzalez JM, Norton ME, Sparks TN. Risk of hypertensive disorders in pregnancies with non-immune hydrops fetalis and single fetal effusions. Pregnancy (Hoboken). 2025;1(4):e70072. doi:10.1002/pmf2.70072. PMID: 42597038. PMC13344758.
142. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Imaging in prenatal diagnosis. p. 742–743.
143. Bilardo CM, Timmerman E, Pajkrt E, van Maarle M. Increased nuchal translucency in euploid fetuses--what should we be telling the parents?. Prenat Diagn. 2010;30(2):93-102. doi:10.1002/pd.2396. PMID: 20077440.
144. Vogel I, Andreasen L, Balslev-Harder M, Becher N, Ernst A, Gadsbøll K, et al. Whole Genome Sequencing in Prenatal Diagnostics: The Danish Approach to Guideline Formation and Implementation Within Public Healthcare. Prenat Diagn. 2026;46(5-6):669-686. doi:10.1002/pd.6780. PMID: 40122702. PMC13170064.
145. MacKenzie TC, Lianoglou BR, Gonzalez Velez J, Dvorak CC, Kharbanda S, Herzeg A, et al. In utero hematopoietic cell transplantation in fetuses with α-thalassemia major: a phase 1 clinical trial. Blood Adv. 2026;10(14):5003-5013. doi:10.1182/bloodadvances.2026019655. PMID: 41984996. PMC13382725.
146. Wagner T, Fahham D, Frumkin A, Shaag A, Yagel S, Yanai N, et al. The many etiologies of nonimmune hydrops fetalis diagnosed by exome sequencing. Prenat Diagn. 2022;42(7):881-889. doi:10.1002/pd.5977. PMID: 34132406.
147. Sharma C, Gothwal M, Shekhawat DS, Yadav T, Jhirwal M, Tada N, et al. Enhancing Diagnostic Precision in Non-Immune Hydrops Fetalis: The Incremental Value of Whole Exome Sequencing-A Prospective Cohort Study. Clin Genet. 2026. doi:10.1111/cge.70236. PMID: 42637266.
148. Sparks TN, Thao K, Lianoglou BR, Boe NM, Bruce KG, Datkhaeva I, et al. Nonimmune hydrops fetalis: identifying the underlying genetic etiology. Genet Med. 2019;21(6):1339-1344. doi:10.1038/s41436-018-0352-6. PMID: 30410095. PMC6509016.
149. Norton ME, Ziffle JV, Lianoglou BR, Hodoglugil U, Devine WP, Sparks TN. Exome sequencing vs targeted gene panels for the evaluation of nonimmune hydrops fetalis. Am J Obstet Gynecol. 2022;226(1):128.e1-128.e11. doi:10.1016/j.ajog.2021.07.014. PMID: 34331894. PMC8748274.
150. Bremova-Ertl T, Patterson M. Niemann-Pick Disease Type C. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2000 Jan 26 [updated 2025 Nov 20]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1296/
151. GeneGuidance. Diagnostic library: Non-immune hydrops fetalis [curated database; version of 2026-09-26].
152. Avram CM, Caughey AB, Norton ME, Sparks TN. Cost-Effectiveness of Exome Sequencing versus Targeted Gene Panels for Prenatal Diagnosis of Fetal Effusions and Non-Immune Hydrops Fetalis. Am J Obstet Gynecol MFM. 2022;4(6):100724. doi:10.1016/j.ajogmf.2022.100724. PMID: 35995366. PMC9938838.
153. Shreeve N, Sproule C, Choy KW, Dong Z, Gajewska-Knapik K, Kilby MD, et al. Incremental yield of whole-genome sequencing over chromosomal microarray analysis and exome sequencing for congenital anomalies in prenatal period and infancy: systematic review and meta-analysis. Ultrasound Obstet Gynecol. 2024;63(1):15-23. doi:10.1002/uog.27491. PMID: 37725747.
154. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Invasive techniques and genetic tests (prenatal). p. 744–747.
155. Graf NS, Moghimi A, Yachmenikova V, Smet ME. Investigation of non-immune hydrops fetalis in singleton pregnancy: a practical guide for the perinatal pathologist. Pathology. 2026;58(6):653-663. doi:10.1016/j.pathol.2026.06.002. PMID: 42586875.
156. Reinhart M, Muraresku C, Ganetzky R. TRMU Deficiency. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. 2023 May 11. Available from: https://www.ncbi.nlm.nih.gov/books/NBK591557/
157. ClinicalTrials.gov. A Study of Nipocalimab in Pregnancies at Risk for Severe Hemolytic Disease of the Fetus and Newborn (HDFN). NCT05912517. Available from: https://clinicaltrials.gov/study/NCT05912517
158. Parent M, Maurice P, Van den Eede E, Mattern J, Sanchez M, Paret L, et al. Perinatal outcomes following in-utero transfusion for hydrops fetalis associated with parvovirus B19 infection. Ultrasound Obstet Gynecol. 2026;67(3):312-319. doi:10.1002/uog.70174. PMID: 41630156.
159. U.S. Food and Drug Administration. LANOXIN (digoxin) tablets — prescribing information (NDA009330; Covis Pharma). DailyMed setid 9ec84cac-6ecc-43dc-8006-6ccc029e2707; revised 17/10/2025. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=9ec84cac-6ecc-43dc-8006-6ccc029e2707
160. U.S. Food and Drug Administration. Flecainide acetate tablets USP — prescribing information (ANDA202821; Aurobindo Pharma). DailyMed setid 14e89448-3a18-42e3-8537-2589be1ef94e; revised 09/01/2024. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=14e89448-3a18-42e3-8537-2589be1ef94e
161. U.S. Food and Drug Administration. BETAPACE / BETAPACE AF (sotalol hydrochloride) — prescribing information (NDA019865; Legacy Pharma). DailyMed setid a6a7702f-36c9-45db-b640-453dc324baee; revised 16/12/2024. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=a6a7702f-36c9-45db-b640-453dc324baee
162. Donofrio MT, Moon-Grady AJ, Hornberger LK, Copel JA, Sklansky MS, Abuhamad A, et al. Diagnosis and treatment of fetal cardiac disease: a scientific statement from the American Heart Association. Circulation. 2014;129(21):2183-2242. doi:10.1161/01.cir.0000437597.44550.5d. PMID: 24763516.
163. ClinicalTrials.gov. Prospective Randomized Clinical Trial of Fetal Atrial Flutter & Supraventricular Tachycardia Therapy (FAST RCT). NCT02624765. Available from: https://clinicaltrials.gov/study/NCT02624765
164. Smeets RMA, van Beynum IM, van Kesteren C, Cornette JMJ, van der Zande JA, Roos-Hesselink JW, et al. The maternal exposure of digoxin and flecainide in relation to the safety and effectiveness in the treatment of non-hydropic fetal tachycardia. Heart Rhythm. 2025;22(11):3020-3027. doi:10.1016/j.hrthm.2025.05.039. PMID: 40412599.
165. van Hove H, van der Heijden JEM, van Uden A, Gijsen VMGJ, Greupink R, de Wildt SN, et al. Sotalol Dose Optimization for Fetal Tachycardia: A Pregnancy Physiologically Based Pharmacokinetic Model Study. Prenat Diagn. 2026;46(4):533-540. doi:10.1002/pd.70104. PMID: 41761048. PMC13070218.
166. Kong L, He Y, Li X, Lu Y, Zhang Y. Meta-analysis and systematic review of amiodarone-inclusive treatment for fetal tachyarrhythmia. Front Pharmacol. 2026;17:1912369. doi:10.3389/fphar.2026.1912369. PMID: 42769344. PMC13590743.
167. Miyoshi T, Maeno Y, Hamasaki T, Inamura N, Yasukochi S, Kawataki M, et al. Antenatal Therapy for Fetal Supraventricular Tachyarrhythmias: Multicenter Trial. J Am Coll Cardiol. 2019;74(7):874-885. doi:10.1016/j.jacc.2019.06.024. PMID: 31416531.
168. Miyoshi T, Maeno Y, Matsuda T, Ito Y, Inamura N, Kim KS, et al. Neurodevelopmental outcome after antenatal therapy for fetal supraventricular tachyarrhythmia: 3-year follow-up of multicenter trial. Ultrasound Obstet Gynecol. 2023;61(1):49-58. doi:10.1002/uog.26113. PMID: 36350016.
169. ClinicalTrials.gov. FAST Trial Registry: Prospective Observational Cohort Study of Fetal Atrial Flutter & Supraventricular Tachycardia. NCT03376438. Available from: https://clinicaltrials.gov/study/NCT03376438
170. Fragala V, Sabale S, Ashoor G, Harris C, Zorro C, Greenough A. Antenatal shunting and outcomes in fetuses with non-immune hydrops fetalis. J Perinat Med. 2025. doi:10.1515/jpm-2025-0198. PMID: 40851234.
171. ClinicalTrials.gov. Prenatal Steroids for Treatment of Congenital Cystic Adenomatoid Malformations (CCAM). NCT00670956. Available from: https://clinicaltrials.gov/study/NCT00670956
172. Ritter DM, Schneider K, Ezeakudo N, Krueger DA, Mizuno T, Knilans TK, et al. Maternal Sirolimus Treatment Reverses Cardiac Rhabdomyoma-Induced Hydrops Fetalis in a Twin Gestation With Tuberous Sclerosis Complex. Am J Med Genet A. 2025;197(11):e64175. doi:10.1002/ajmg.a.64175. PMID: 40605491.
173. U.S. Food and Drug Administration. CELESTONE SOLUSPAN (betamethasone sodium phosphate and betamethasone acetate) injectable suspension — prescribing information (NDA014602). DailyMed setid 95a19c2f-c329-4ce1-a0a7-174814aac3ad; revised 24/04/2026. Available from: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=95a19c2f-c329-4ce1-a0a7-174814aac3ad
174. Kuelshammer M, Matzik AS, Hansmann A, Moersdorf J, Strizek B, Dresbach T, et al. Extracorporeal membrane oxygenation in Non-Immune Hydrops fetalis - a registry-based data analysis over the last 25 years. Eur J Pediatr. 2026;185(7):531. doi:10.1007/s00431-026-07185-x. PMID: 42360500. PMC13309399.
175. Chawla V, Shashi KK, Niven ML, Devarapalli V, Shah P, Tran AD, et al. Sirolimus for non-immune hydrops due to vascular and lymphatic anomalies in preterm neonates: an imaging-anchored response phenotype from a level IV NICU. J Perinatol. 2026. doi:10.1038/s41372-026-02755-1. PMID: 42332051.
176. European Medicines Agency. Mepsevii (vestronidase alfa): EPAR — medicine overview. Authorisation 23/08/2018 (exceptional circumstances; orphan); last EC decision 08/07/2026. Available from: https://www.ema.europa.eu/en/medicines/human/EPAR/mepsevii
177. ClinicalTrials.gov. PEARL (PrEnAtal Enzyme Replacement Therapy for Lysosomal Storage Disorders). NCT04532047. Available from: https://clinicaltrials.gov/study/NCT04532047
178. Cohen JL, Chakraborty P, Fung-Kee-Fung K, Schwab ME, Bali D, Young SP, et al. In Utero Enzyme-Replacement Therapy for Infantile-Onset Pompe's Disease. N Engl J Med. 2022;387(23):2150-2158. doi:10.1056/nejmoa2200587. PMID: 36351280. PMC10794051.
179. ClinicalTrials.gov. In Utero Hematopoietic Stem Cell Transplantation for Alpha-thalassemia Major (ATM). NCT02986698. Available from: https://clinicaltrials.gov/study/NCT02986698
180. ClinicalTrials.gov. Feasibility of In-utero Hematopoietic Stem Cell Transplantation as a Conjunctive Treatment During In-utero Blood Transfusion for Hemoglobin Bart's Hydrops Fetalis Syndrome in Hong Kong. NCT05797272. Available from: https://clinicaltrials.gov/study/NCT05797272
181. ClinicalTrials.gov. Amniotic Fluid Tandem Mass Spectrometry for Pregnancies Complicated by NIH and Severe Symmetrical IUGR. NCT00143039. Available from: https://clinicaltrials.gov/study/NCT00143039
182. Borges B, Canepa E, Chang IJ, Herzeg A, Lianoglou B, Kishnani PS, et al. Prenatal Delivery of Enzyme Replacement Therapy to Fetuses Affected by Early-Onset Lysosomal Storage Diseases. Am J Med Genet C Semin Med Genet. 2025;199(3):203-217. doi:10.1002/ajmg.c.32132. PMID: 39891377. PMC12310989.
183. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Mucopolysaccharidosis VII (Sly syndrome). p. 656–657.