Application of next-generation sequencing in nonimmune hydrops fetalis and its impact on pregnancy decisions Authors: Yu Qin, Wei Li, Sha Zhu, Shunran Zhang, Ruijie Fang, Qingling Kang, Lei Fan, Jing Liu, Shufang Li, Jianli Wu, Yuanyuan Wu, Xinwei Shi, Ling Feng, Suhua Chen, Juan Xiao Published: 2026 ABSTRACT BackgroundNonimmune hydrops fetalis (NIHF) has a highly heterogeneous etiology, with genetic factors contributing substantially to its pathogenesis. This study aimed to investigate the chromosomal and monogenic etiologies of NIHF.MethodsThis single-center retrospective cohort study included 121 pregnancies complicated by NIHF evaluated at a tertiary referral center between August 2020 and August 2023. All cases underwent copy number variation sequencing (CNV-seq), and 95 additionally underwent whole-exome sequencing (WES), both assays were based on next-generation sequencing (NGS) technology. Cases were stratified according to gestational age at diagnosis, maternal age, conception method, and whether NIHF was isolated or non-isolated. Genetic findings and pregnancy outcomes were analyzed and compared across subgroups, with longitudinal follow-up of pregnancy outcomes.ResultsAmong the 121 NIHF cases, pathogenic or likely pathogenic genetic findings were identified in 62 cases, yielding an overall positivity rate of 51.2% (62/121). CNV-seq detected pathogenic abnormalities in 49 cases (40.5%, 49/121), including 38 numerical chromosomal abnormalities and 11 structural chromosomal abnormalities. Among the 95 fetuses that underwent WES, 14 had pathogenic or likely pathogenic variants, corresponding to a positivity rate of 14.7% (14/95). RASopathies (RIT1, RAF1, and BRAF) and skeletal disorders (FGFR2, COL1A1, and HSPG2) were the most frequently identified monogenic conditions. The overall genetic positivity rate was significantly higher in cases diagnosed at ≤ 13+6 weeks than in those diagnosed at ≥ 14 weeks (61.4% vs. 28.9%, P = 0.001), and in non-isolated NIHF compared with isolated NIHF (58.8% vs. 10.5%, P < 0.001). However, after excluding aneuploidies, no statistically significant differences were observed across clinical subgroups in the detection rates of pathogenic structural chromosomal abnormalities or WES-detected pathogenic/likely pathogenic variants. Cases with positive genetic findings had significantly more adverse pregnancy outcomes than genetically negative cases.ConclusionNIHF is associated with a broad spectrum of genetic abnormalities, including aneuploidies, pathogenic structural chromosomal abnormalities, and single-gene disorders. After excluding aneuploidy, the diagnostic yields of pathogenic CNVs and WES were comparable across clinical subgroups. These findings suggest that expanded genomic testing may be valuable across the spectrum of NIHF and may help inform genetic counseling and future pregnancy planning.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12884-026-09428-z. Introduction Fetal hydrops is characterized by the accumulation of fluid in the fetal soft tissues and body cavities, as observed through ultrasonographic findings such as pleural effusion, abdominal effusion, pericardial effusion, skin edema, and so on [1, 2]. This condition is further categorized into immune hydrops fetalis (IHF) and NIHF. IHF is primarily caused by maternal and fetal blood group incompatibility, and its incidence has markedly declined with the widespread use of anti-D immunoglobulin prophylaxis and advances in intrauterine transfusion. In contrast, NIHF remains the predominant form of hydrops fetalis, representing over 90% of all cases, with an estimated incidence of 1 in 1700–3000 pregnancies [3–5]. Although NIHF is uncommon, it is clinically significant because of its poor overall prognosis and high fetal or perinatal mortality, which has been reported to range from 55 to 90% [6]. The etiology of NIHF is highly variable and may involve fetal, placental, or maternal factors or represent a common end-stage manifestation of a wide range of underlying disorders. Common contributing factors include cardiovascular anomalies, thoracic malformations, fetal tumors, lymphatic dysplasia, hematologic disorders, twin-twin transfusion syndrome (TTTS), intrauterine infections, skeletal dysplasia, inborn metabolic disorders, and chromosomal and single-gene anomalies [7–10]. Previous studies have suggested that approximately one-third of NIHF cases have a genetic etiology [11], and this proportion may increase with the application of advanced molecular genetic testing. Genetic evaluation of NIHF mainly focuses on two levels: chromosomal abnormalities and single-gene disorders. Conventional karyotyping can identify numerical chromosomal abnormalities and large structural rearrangements, typically at a resolution of approximately 5–10 Mb [12]. In recent years, molecular techniques such as chromosomal microarray analysis and CNV-seq have enabled genome-wide detection of copy number variants (CNVs), including CNVs larger than approximately 100 kb [13], which can detect the etiology of approximately 25% of NIHF cases [2]. For monogenic disorders, both targeted gene panels and WES have been used in clinical practice. However, because NIHF is associated with a large and continually expanding number of genes, targeted panels may have limited diagnostic coverage [14]. In contrast, WES, which utilizes NGS technology, offers a higher detection rate and is more cost-effective for diagnosing NIHF compared to gene panels [15]. Studies have shown that WES can detect 37% of diagnostic genetic variants for NIHF when chromosomal abnormalities, fetal anemia, and infections are excluded [16]. Despite these advances, it remains unclear whether the diagnostic yields of structural chromosomal abnormalities and WES differ across clinical subgroups of NIHF, such as those defined by gestational age at diagnosis, maternal age, conception method, or the presence of additional ultrasound abnormalities. This question is clinically important because some providers may preferentially offer expanded genomic testing only to selected NIHF subgroups. Therefore, in this retrospective cohort study, we analyzed a series of NIHF cases evaluated by CNV-seq alone or in combination with WES at our center over a three-year period. We aimed to characterize the genetic spectrum of NIHF at both the chromosomal and monogenic levels and to assess the relationships between genetic findings, clinical phenotype, and pregnancy outcomes. Materials and methods Study design and participants This retrospective cohort study was conducted at the Prenatal Diagnosis Center of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, a tertiary referral center and National Clinical Research Center for Obstetric and Gynecologic Diseases in Wuhan, China. We reviewed pregnancies complicated by fetal hydrops that underwent prenatal genetic counseling, prenatal diagnosis, and/or termination of pregnancy (TOP) between August 2020 and August 2023. Eligible cases were those with NIHF diagnosed by ultrasound and confirmed to be nonimmune after maternal blood typing and irregular antibody screening. After routine pretest genetic counseling, families provided informed consent and chose CNV-seq alone or CNV-seq combined with WES for fetal genetic evaluation. The study was approved by the Ethics Committee of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology (TJ-IRB202303142), and written informed consent was obtained from all participants. Inclusion and exclusion criteria Cases were included if ultrasound demonstrated abnormal fluid accumulation in at least two fetal compartments, including skin edema (skin thickness > 5 mm), pleural effusion, ascites, pericardial effusion, placentomegaly (placental thickness ≥ 4 cm in the second trimester or > 6 cm in the third trimester), or polyhydramnios. Polyhydramnios was defined as an amniotic fluid index ≥ 24 cm or a deepest vertical pocket ≥ 8 cm [17]. Cases were excluded if fetal hydrops was attributed to maternal–fetal blood group incompatibility or if the family declined fetal genetic testing. Clinical subgroup classification Cases were classified as isolated or non-isolated NIHF according to the presence or absence of additional ultrasound abnormalities or soft markers [18]. Maternal age was categorized as advanced or non-advanced according to local clinical criteria used in our center: ≥ 35 years versus < 35 years for singleton pregnancies and ≥ 32 years versus < 32 years for twin pregnancies. According to gestational age at diagnosis, cases were divided into early-gestation NIHF (≤ 13+6 weeks) and mid-to-late gestation NIHF (≥ 14 weeks). The conception method was categorized as natural conception or in vitro fertilization and embryo transfer (IVF-ET). Sample collection Depending on gestational age and pregnancy management, chorionic villi or amniotic fluid samples were collected by ultrasound-guided chorionic villus sampling (n = 2) or amniocentesis (n = 50), respectively. For pregnancies that underwent termination, fetal muscle tissue samples were collected after the procedure (n = 69). Copy number variation sequencing The process is shown in Fig. 1, created by Figdraw, and previous studies have elaborately detailed the specific methods [19]. In summary, amniotic fluid or fetal tissue was obtained to exclude maternal contamination using the short tandem repeat (STR) technique. Then, CNV-seq was performed with 10 ng of DNA extracted from the samples for library preparation. Following a successful quality control assessment, the library underwent high-throughput sequencing on a NextSeq CN500 sequencer. The resulting sequencing data were analyzed utilizing the Konoan data analysis software provided by Beijing Berry Hecom Genetic Diagnostics Co., Ltd. Copy number variants as small as 100 kb could be detected. Lastly, the identified CNVs were classified following the guidelines established by the American College of Medical Genetics and Genomics (ACMG) [20].Fig. 1Program for CNV-seq and WES detection. Created by Figdraw Program for CNV-seq and WES detection. Created by Figdraw Whole exome sequencing The genomic DNA of fetuses and their parents were used for library preparation using the Illumina Nextera Rapid Capture Exome Kit. High-throughput sequencing was performed using a Novaseq 6000 gene sequencer, and raw data were obtained after CASAVA v1.82 processing. Variant detection analysis was accomplished using the Verita Trekker® Variant Detection System developed by Berry Genetics and GATK (https://software.broadinstitute.org/gatk/). Variant annotation was done using ANNOVAR and the Enliven® Variant Annotation Interpretation System developed by Berry Genetics. Finally, the laboratory classified the variants according to the ACMG guidelines[21]. After identifying a pathogenic or likely pathogenic variant in a fetus, the corresponding genomic region was confirmed, and parental origin was assessed by Sanger sequencing. Statistical analysis Statistical analyses were performed using SPSS version 26.0 (IBM Corp., Armonk, NY, USA). Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. A two-sided P value < 0.05 was considered statistically significant. Results Basic information A total of 121 NIHF cases were included. The mean maternal age was 30 years (range, 20–44 years). Regarding conception methods, 109 cases were conceived naturally (90.1%), and 12 were conceived via IVF-ET (9.9%). There were 112 singleton pregnancies (92.6%) and 9 twin pregnancies (7.4%), including 4 dichorionic diamniotic and 5 monochorionic diamniotic twin pregnancies, and none of the monochorionic pregnancies met the diagnostic criteria for TTTS. The mean gestational age at diagnosis was 15 weeks (range, 11+4 to 31+1 weeks). According to the predefined criteria, 83 cases (68.6%) were diagnosed in the early gestational group (≤ 13+6 weeks) and 38 (31.4%) in the mid-to-late gestational group (≥ 14 weeks). Nineteen cases were classified as isolated NIHF (15.7%), whereas 102 were non-isolated NIHF (84.3%). The most common associated ultrasound findings were cystic hygroma, increased nuchal translucency, and cardiovascular abnormalities. All 121 cases underwent CNV-seq, and 95 of these additionally underwent WES (Fig. 2). Among the 95 cases that underwent WES, 56 underwent trio WES, 7 underwent proband-only WES followed by targeted parental validation, and 32 underwent proband-only WES without available parental validation.Fig. 2Results of CNV-seq and WES in 121 NIHF Results of CNV-seq and WES in 121 NIHF Results of copy number variation sequencing CNV-seq identified pathogenic chromosomal abnormalities in 49 cases, yielding a positivity rate of 40.5% (49/121). These included 38 numerical chromosomal abnormalities and 11 pathogenic structural chromosomal abnormalities. Turner syndrome (45,X) was the most prevalent numerical abnormality, observed in 19 cases (50.0%, 19/38), followed by trisomy 18 in 13 cases (34.2%, 13/38), trisomy 21 in 4 cases (10.5%, 4/38), and trisomy 13 in 2 cases (5.3%, 2/38), as illustrated in Fig. 3. Among the chromosomal CNVs identified in this study, microdeletion syndromes were present in 7 cases, microduplication syndromes were observed in 3 cases, and both microdeletion and microduplication syndromes were detected in 1 case, whose mother had a balanced translocation, 46,XX,t(15;22)(q15.q12.2). Among families who underwent parental CNV testing, four CNVs were confirmed to be de novo, and one was inherited from the father. The clinical characteristics of the 11 cases are outlined in Table 1.Fig. 3Chromosome number abnormalities were present in 38 cases of NIHFTable 1Pathogenic copy number variants (CNVs) were present in 11 cases of NIHFPatientMaternal ageGestational age (week)Methods of pregnancyAbnormal ultrasound findings other than fetal hydropsGenetic origin of pathogenic CNVsCopy number variationsSize(Mb)ACMG classificationOutcome of pregnancyC102426SP-De novoseq[hg19]dup(8)(p21.1q21.13)(mos) chr8:g.28760000_80980000dup52.22PTOPC112513+5SPCystic hygromaParental follow-up lostseq[hg19]del(18)(p11.32p11.21) chr18:g.120000_14980000del14.86PTOPC123022+2SPRenal pelvis separation, increased intestinal echogenicityParental follow-up lostseq[hg19]dup(13)(q21.33q34)(mos) chr13:g.71740000_115100000dup43.36PTOPC133112+4SPSingle atrium, single ventricleParental follow-up lostseq[hg19]del(1)(p36.33p36.22)chr1:g.820000_9440000del8.62PMAC143214+1SPCystic hygroma, enlargement of the cardiothoracic ratioMaternal, material karyotype: 46, XX, t(15;22)(q15.q12.2)seq[hg19]del(22)(q11.1q11.21)chr22:g.16840000_20320000del3.48PMAseq[hg19]dup(15)(q11.2q14) chr15:g.22740000_39300000dup16.56PMAW302912+5SPNT3.6 mm, Separation of the left renal collecting system and dilatation of the left ureter, pes varus, SUADe novoseq[hg19]del(2)(q21.2q22.3) chr2:g.134080000_147340000del13.26PTOPW313313+2SPNT 4.7 mm, choroid plexus cystParental follow-up lostseq[hg19]del(16)(q12.2q21) chr16:g.55420000_59940000del4.52PTOPW323213+1SPCystic hygroma, congenital talipes equinovarusParental follow-up lostseq[hg19]del(10)(q26.2q26.3) chr10:g.129960000_135440000del5.48PTOPW332312+6SPNT4.47 mm, hydramniosDe novoseq[hg19]del(15)(q11.2) chr15:g.22760000_23080000del0.32PTOPW343030+2SPShort femur lengthPaternalseq[hg19]del(15)(q11.2q11.2) chr15:g.22740000_23100000del0.36PLive-bornW352929+6SPCME, narrow internal rectal diameter, aberrant right sub-clavicular arteryDe novoseq[hg19]dup(15)(q26.1q26.3) chr15:g.92600000_102400000dup9.8PTOPCME Cisterna magna enlargement, Dup Duplication, Del Deletion, MA Missed abortion, NT Nuchal translucency, P pathogenic, Parental follow-up lost, parental CNV testing was not available for these cases. SP Spontaneous pregnancy, SUV Single umbilical artery, TOP Termination of pregnancy Chromosome number abnormalities were present in 38 cases of NIHF Table 1Pathogenic copy number variants (CNVs) were present in 11 cases of NIHFPatientMaternal ageGestational age (week)Methods of pregnancyAbnormal ultrasound findings other than fetal hydropsGenetic origin of pathogenic CNVsCopy number variationsSize(Mb)ACMG classificationOutcome of pregnancyC102426SP-De novoseq[hg19]dup(8)(p21.1q21.13)(mos) chr8:g.28760000_80980000dup52.22PTOPC112513+5SPCystic hygromaParental follow-up lostseq[hg19]del(18)(p11.32p11.21) chr18:g.120000_14980000del14.86PTOPC123022+2SPRenal pelvis separation, increased intestinal echogenicityParental follow-up lostseq[hg19]dup(13)(q21.33q34)(mos) chr13:g.71740000_115100000dup43.36PTOPC133112+4SPSingle atrium, single ventricleParental follow-up lostseq[hg19]del(1)(p36.33p36.22)chr1:g.820000_9440000del8.62PMAC143214+1SPCystic hygroma, enlargement of the cardiothoracic ratioMaternal, material karyotype: 46, XX, t(15;22)(q15.q12.2)seq[hg19]del(22)(q11.1q11.21)chr22:g.16840000_20320000del3.48PMAseq[hg19]dup(15)(q11.2q14) chr15:g.22740000_39300000dup16.56PMAW302912+5SPNT3.6 mm, Separation of the left renal collecting system and dilatation of the left ureter, pes varus, SUADe novoseq[hg19]del(2)(q21.2q22.3) chr2:g.134080000_147340000del13.26PTOPW313313+2SPNT 4.7 mm, choroid plexus cystParental follow-up lostseq[hg19]del(16)(q12.2q21) chr16:g.55420000_59940000del4.52PTOPW323213+1SPCystic hygroma, congenital talipes equinovarusParental follow-up lostseq[hg19]del(10)(q26.2q26.3) chr10:g.129960000_135440000del5.48PTOPW332312+6SPNT4.47 mm, hydramniosDe novoseq[hg19]del(15)(q11.2) chr15:g.22760000_23080000del0.32PTOPW343030+2SPShort femur lengthPaternalseq[hg19]del(15)(q11.2q11.2) chr15:g.22740000_23100000del0.36PLive-bornW352929+6SPCME, narrow internal rectal diameter, aberrant right sub-clavicular arteryDe novoseq[hg19]dup(15)(q26.1q26.3) chr15:g.92600000_102400000dup9.8PTOPCME Cisterna magna enlargement, Dup Duplication, Del Deletion, MA Missed abortion, NT Nuchal translucency, P pathogenic, Parental follow-up lost, parental CNV testing was not available for these cases. SP Spontaneous pregnancy, SUV Single umbilical artery, TOP Termination of pregnancy Pathogenic copy number variants (CNVs) were present in 11 cases of NIHF seq[hg19]del(1)(p36.33p36.22) chr1:g.820000_9440000del seq[hg19]del(22)(q11.1q11.21) chr22:g.16840000_20320000del CME Cisterna magna enlargement, Dup Duplication, Del Deletion, MA Missed abortion, NT Nuchal translucency, P pathogenic, Parental follow-up lost, parental CNV testing was not available for these cases. SP Spontaneous pregnancy, SUV Single umbilical artery, TOP Termination of pregnancy Results of whole exome sequencing WES identified 14 pathogenic or likely pathogenic variants in 95 fetuses, corresponding to a positivity rate of 14.7% (14/95), including 1 case with both pathogenic CNVs and gene variants (Table 2). The gene variants identified were associated with various disorders, including lymphatic (EPHB4), immune system (FOXP3), multisystemic malformations (NOTCH2), cardiovascular (MYH7), and hematologic (F8) disorders, and congenital overgrowth syndrome (NSD1), 2 cases of congenital enzyme metabolism defects (NEU1, SLC3A1), skeletal disorders and RASopathy-associated disorders, with three cases in each category. Pathogenic or likely pathogenic variants associated with the skeletal system, such as FGFR2, COL1A1, and HSPG2, were observed in individuals with skeletal abnormalities. Additionally, pathogenic or likely pathogenic variants related to the RAS-MAPK cell signaling pathway, known as RASopathies, including RIT1, RAF1, and BRAF, were identified.Table 2WES identified 14 cases of NIHF with pathogenic or suspected pathogenic gene variantsPatientMaternal ageGestational ageAbnormal ultrasound findings other than fetal hydropsGeneMutation positionGene subregionTranscriptVariantMutation typeHeterozygosityACMG classificationRelated diseaseCategories of genetic disordersOutcome of pregnancyW35***2929 + 6Short rectal internal diameterNOTCH2chr1:119,967,576–119,967,576exon8NM_024408.4:c.1310A > CNP_077719.2:p.(Asp437Ala)nonsynonymous SNVHeterozygous(de‐novo)LPHajdu-Cheney Syndrome(AD),Alagille Syndrome 2(AD)Musculoskeletal disordersTOPW36*+24012 + 4NT 3.4 mmF8chrX:154,929,256–154,929,256exon14NM_000132.3:c.4534G > TNP_000123.1:p.(Glu1512Ter)stopgainHeterozygous(maternal)LPHemophilia A(XLR)Hematologic disordersLive-birthW37*+22912 + 1NT 2.8 mmNSD1chr5:177,244,249–177,244,249exon9NM_022455.5:c.4358dupNP_071900.2:p.(Tyr1453Ter)stopgainHeterozygous (de‐novo)LPSotos Syndrome(AD)Overgrowth disordersTOPW38*+23017 + 3-FOXP3chrX:49,251,440–49251440exon12NM_014009.4:c.1190G > ANP_054728.2:p.(Arg397Gln)missense_variantHemizygote(maternal)LPIpex Syndrome(XLR)Immunologic disordersMAW39***2729 + 6Permanent right-sided umbilical vein, increased S/D of umbilical arteryEPHB4chr7:100,807,468–100807468exon13NM_004444.5:c.2231G > ANP_004435.3:p.(Arg744His)nonsynonymous SNVHeterozygous(de‐novo)LPCapillary Malformation-Arteriovenous Malformation 2(AD)Lymphatic disordersTOPW40***2016Umbilical hernia, Strong echoes of the intestinal tract, FGRFGFR2chr10:121,565,500–121565500exon3NM_000141.5: c.314A > GNP_000132.3:p.(Tyr105Cys)missense_variantHeterozygous(maternal)LPAcrocephalosyndactyly Type V, Antley-B ixler Syndrome(AD), Apert Syndrome(AD), Beare-Stevenson Cutis Gyrata Syndrome(AD)Musculoskeletal disordersTOPW41*+13112 + 5Cystic hygroma, short limbsHSPG2chr1:21,859,905–21859905exon41NM_005529.7:c.5112delNP_005520.4:p.(His1706ThrfsTer68)frameshift_variantHeterozygous(maternal)LPDyssegmental Dysplasia Silverman-Handmaker Type(AR),Schwartz-Jampel Syndrome(AR)Musculoskeletal disordersTOPHSPG3chr1:21,937,173–21,937,184exon1NM_005529.7:c.34_45delNP_005520.4:p.(Ala12_Leu15del)inframe_ deletionHeterozygous(unknown origin)VUSDyssegmental Dysplasia Silverman-Handmaker Type(AR),Schwartz-Jampel Syndrome(AR)Musculoskeletal disordersW42*+22916 + 2Cystic hygroma,TOFBRAFchr7:140,754,187–140754187exon14NM_004333.6:c.1741A > GNP_004324.2:p.(Asn581Asp)nonsynonymous SNVHeterozygous (de‐novo)PCardiofaciocutaneous Syndrome(AD)RASopathiesTOPW43***2911 + 4Leftward deviation of fetal cardiac axisNEU1chr6:31,862,112–31,862,112exon2NM_000434.4:c.239C > TNP_000425.1:p.(Pro80Leu)missense_variantHomozygous(parental)PSialidosis(AR)Inborn errors of metabolismTOPW44***2622 + 1TOF,short limbsMYH7chr14:23,431,584–23,431,584intron8NM_000257.4:c.732 + 1G > ANP_000248.2:p.(?)splice_donor_variantHeterozygous (de‐novo)PMYH7-Related DCM/HCM(AD)Cardiovascular disordersTOPW45*+23011 + 5Cystic hygroma,thoracic cyst on the right, VSDCOL1A1chr17:50,195,288–50195288exon19NM_000088.4:c.1243C > TNP_000079.2:p.(Arg415Ter)stopgainHeterozygous (paternal)PCaffey Disease(AD); Osteogenesis imperfecta, type IIMusculoskeletal disordersTOPW46*+22613 + 2NT 3.6 mm, TOF, permanent right-sided umbilical vein,RAF1chr3:12,604,200–12604200exon7NM_002880.4:c.770C > TNP_002871.1:p.(Ser257Leu)nonsynonymous SNVHeterozygous (de‐novo)PNoonan Syndrome 5(AD)RASopathiesTOPW47***3312 + 6Cystic hygromaRIT1chr1:155,904,494–155904494exon5NM_006912.6:c.246 T > G:NP_008843.1:p.(Phe82Leu)nonsynonymous SNVHeterozygous (de‐novo)PNoonan Syndrome 8(AD)RASopathiesTOPW48***4011Cystic hygroma, abnormal kidney sizes, increased echoSLC3A1chr2:44,301,104–44301104exon6NM_000341.4:c.1113C > ANP_000332.2:p.(Tyr371Ter)stop_gainedHeterozygous (paternal)PSulfocysteinuria(AD/AR)Inborn errors of metabolismContinued pregnancyAD Autosomal dominant, AR Autosomal recessive, FGR Fetal growth restriction, LP Likely pathogenic, MA Missed abortion, NT Nuchal translucency, P Pathogenic, TOP Termination of pregnancy, TOF Tetralogy of Fallot, VSD Ventricular septal defect, XLR X-linked recessiveWES***, trio-exome sequencing. WES * + 2,fetus-only exome sequencing and gene validation of parental, W41* + 1, fetus-only exome sequencing and gene validation of maternal Table 2WES identified 14 cases of NIHF with pathogenic or suspected pathogenic gene variantsPatientMaternal ageGestational ageAbnormal ultrasound findings other than fetal hydropsGeneMutation positionGene subregionTranscriptVariantMutation typeHeterozygosityACMG classificationRelated diseaseCategories of genetic disordersOutcome of pregnancyW35***2929 + 6Short rectal internal diameterNOTCH2chr1:119,967,576–119,967,576exon8NM_024408.4:c.1310A > CNP_077719.2:p.(Asp437Ala)nonsynonymous SNVHeterozygous(de‐novo)LPHajdu-Cheney Syndrome(AD),Alagille Syndrome 2(AD)Musculoskeletal disordersTOPW36*+24012 + 4NT 3.4 mmF8chrX:154,929,256–154,929,256exon14NM_000132.3:c.4534G > TNP_000123.1:p.(Glu1512Ter)stopgainHeterozygous(maternal)LPHemophilia A(XLR)Hematologic disordersLive-birthW37*+22912 + 1NT 2.8 mmNSD1chr5:177,244,249–177,244,249exon9NM_022455.5:c.4358dupNP_071900.2:p.(Tyr1453Ter)stopgainHeterozygous (de‐novo)LPSotos Syndrome(AD)Overgrowth disordersTOPW38*+23017 + 3-FOXP3chrX:49,251,440–49251440exon12NM_014009.4:c.1190G > ANP_054728.2:p.(Arg397Gln)missense_variantHemizygote(maternal)LPIpex Syndrome(XLR)Immunologic disordersMAW39***2729 + 6Permanent right-sided umbilical vein, increased S/D of umbilical arteryEPHB4chr7:100,807,468–100807468exon13NM_004444.5:c.2231G > ANP_004435.3:p.(Arg744His)nonsynonymous SNVHeterozygous(de‐novo)LPCapillary Malformation-Arteriovenous Malformation 2(AD)Lymphatic disordersTOPW40***2016Umbilical hernia, Strong echoes of the intestinal tract, FGRFGFR2chr10:121,565,500–121565500exon3NM_000141.5: c.314A > GNP_000132.3:p.(Tyr105Cys)missense_variantHeterozygous(maternal)LPAcrocephalosyndactyly Type V, Antley-B ixler Syndrome(AD), Apert Syndrome(AD), Beare-Stevenson Cutis Gyrata Syndrome(AD)Musculoskeletal disordersTOPW41*+13112 + 5Cystic hygroma, short limbsHSPG2chr1:21,859,905–21859905exon41NM_005529.7:c.5112delNP_005520.4:p.(His1706ThrfsTer68)frameshift_variantHeterozygous(maternal)LPDyssegmental Dysplasia Silverman-Handmaker Type(AR),Schwartz-Jampel Syndrome(AR)Musculoskeletal disordersTOPHSPG3chr1:21,937,173–21,937,184exon1NM_005529.7:c.34_45delNP_005520.4:p.(Ala12_Leu15del)inframe_ deletionHeterozygous(unknown origin)VUSDyssegmental Dysplasia Silverman-Handmaker Type(AR),Schwartz-Jampel Syndrome(AR)Musculoskeletal disordersW42*+22916 + 2Cystic hygroma,TOFBRAFchr7:140,754,187–140754187exon14NM_004333.6:c.1741A > GNP_004324.2:p.(Asn581Asp)nonsynonymous SNVHeterozygous (de‐novo)PCardiofaciocutaneous Syndrome(AD)RASopathiesTOPW43***2911 + 4Leftward deviation of fetal cardiac axisNEU1chr6:31,862,112–31,862,112exon2NM_000434.4:c.239C > TNP_000425.1:p.(Pro80Leu)missense_variantHomozygous(parental)PSialidosis(AR)Inborn errors of metabolismTOPW44***2622 + 1TOF,short limbsMYH7chr14:23,431,584–23,431,584intron8NM_000257.4:c.732 + 1G > ANP_000248.2:p.(?)splice_donor_variantHeterozygous (de‐novo)PMYH7-Related DCM/HCM(AD)Cardiovascular disordersTOPW45*+23011 + 5Cystic hygroma,thoracic cyst on the right, VSDCOL1A1chr17:50,195,288–50195288exon19NM_000088.4:c.1243C > TNP_000079.2:p.(Arg415Ter)stopgainHeterozygous (paternal)PCaffey Disease(AD); Osteogenesis imperfecta, type IIMusculoskeletal disordersTOPW46*+22613 + 2NT 3.6 mm, TOF, permanent right-sided umbilical vein,RAF1chr3:12,604,200–12604200exon7NM_002880.4:c.770C > TNP_002871.1:p.(Ser257Leu)nonsynonymous SNVHeterozygous (de‐novo)PNoonan Syndrome 5(AD)RASopathiesTOPW47***3312 + 6Cystic hygromaRIT1chr1:155,904,494–155904494exon5NM_006912.6:c.246 T > G:NP_008843.1:p.(Phe82Leu)nonsynonymous SNVHeterozygous (de‐novo)PNoonan Syndrome 8(AD)RASopathiesTOPW48***4011Cystic hygroma, abnormal kidney sizes, increased echoSLC3A1chr2:44,301,104–44301104exon6NM_000341.4:c.1113C > ANP_000332.2:p.(Tyr371Ter)stop_gainedHeterozygous (paternal)PSulfocysteinuria(AD/AR)Inborn errors of metabolismContinued pregnancyAD Autosomal dominant, AR Autosomal recessive, FGR Fetal growth restriction, LP Likely pathogenic, MA Missed abortion, NT Nuchal translucency, P Pathogenic, TOP Termination of pregnancy, TOF Tetralogy of Fallot, VSD Ventricular septal defect, XLR X-linked recessiveWES***, trio-exome sequencing. WES * + 2,fetus-only exome sequencing and gene validation of parental, W41* + 1, fetus-only exome sequencing and gene validation of maternal WES identified 14 cases of NIHF with pathogenic or suspected pathogenic gene variants Heterozygous (de‐novo) Hajdu-Cheney Syndrome(AD), Alagille Syndrome 2(AD) Heterozygous (maternal) Hemizygote (maternal) Heterozygous (de‐novo) Heterozygous (maternal) Heterozygous (maternal) Dyssegmental Dysplasia Silverman-Handmaker Type(AR), Schwartz-Jampel Syndrome(AR) Heterozygous (unknown origin) Dyssegmental Dysplasia Silverman-Handmaker Type(AR), Schwartz-Jampel Syndrome(AR) Homozygous (parental) AD Autosomal dominant, AR Autosomal recessive, FGR Fetal growth restriction, LP Likely pathogenic, MA Missed abortion, NT Nuchal translucency, P Pathogenic, TOP Termination of pregnancy, TOF Tetralogy of Fallot, VSD Ventricular septal defect, XLR X-linked recessive WES***, trio-exome sequencing. WES * + 2,fetus-only exome sequencing and gene validation of parental, W41* + 1, fetus-only exome sequencing and gene validation of maternal Parental validation by Sanger sequencing was performed whenever feasible. Seven pathogenic or likely pathogenic variants were confirmed to be de novo, including heterozygous variants in BRAF, RIT1, NSD1, EPHB4, RAF1, MYH7, and NOTCH2. Seven variants were inherited: 3 maternally inherited variants (F8, FGFR2, and FOXP3), 2 paternally inherited variants (COL1A1 and SLC3A1), 1 homozygous NEU1 variant inherited from two heterozygous carrier parents, and 1 compound heterozygous HSPG2 finding in which one variant was confirmed to be maternally inherited, whereas the second variant could not be validated because it was located in a high-GC region and the father declined testing. Clinical characteristics of cases with genetic variants Overall, pathogenic or likely pathogenic genetic findings were identified in 62/121 NIHF cases (51.2%, Table 3). The overall genetic positivity rate was significantly higher in the early-gestation group than in the mid-to-late-gestation group (61.4% vs. 28.9%, P = 0.001). This difference was mainly driven by numerical chromosomal abnormalities, which were significantly more frequent in the early-gestation group (44.6% vs. 2.6%, P < 0.001). In contrast, the detection rates of pathogenic CNVs (7.2% vs. 13.2%, P = 0.318) and WES-detected pathogenic/likely pathogenic variants (12.7% vs. 18.8%, P = 0.542) did not differ significantly between the two groups.Table 3Comparison of CNV-seq and WES findings according to clinical characteristics in NIHFGroupsOverall genetic positivity rate, n (%)CNV-seq positivity rate, n (%)Numerical chromosomal abnormality rate, n (%)CNVs rate, n (%)WES positivity rate among cases undergoing WES, n (%)Early gestational group (≤ 13+6W) (CNV n = 83, WES n = 63)51 (61.4%)43 (51.8%)37 (44.6%)6 (7.2%)8 (12.7%)Mid to late gestational group (≥ 14 W) (CNV n = 38, WES n = 32)11 (28.9%)6 (15.8%)1 (2.6%)5 (13.2%)6 (18.8%)P value0.001 < 0.001 < 0.0010.3180.542Non-advanced maternal age group (singleton < 35 years; twin < 32 years) (CNV n = 96, WES n = 77)49 (51%)38 (39.6%)27 (28.1%)11 (11.5%)12 (15.6%)Advanced maternal age group (singleton ≥ 35 years; twin ≥ 32 years) (CNV n = 25, WES n = 18)13 (52%)11 (44%)11 (44%)02 (11.1%)P value0.9320.6890.1280.1181IVF-ET (CNV n = 12, WES n = 11)7 (58.3%)5 (41.7%)5 (41.7%)02(18.2%)Natural conception (CNV n = 109, WES n = 84)55 (50.5%)44 (40.4%)33 (30.3%)11 (10.1%)12 (14.3%)P value0.60410.5140.5990.663isolated NIHF (CNV n = 19, WES n = 14)2 (10.5%)1 (5.3%)01 (5.3%)1 (7.1%)non-isolated NIHF (CNV n = 102, WES n = 81)60 (58.8%)48 (47.1%)38 (37.3%)10 (9.8%)13 (16%)P value < 0.0010.0010.00110.685Overall genetic positivity and CNV-seq-related rates were calculated using the total number of cases in each subgroup as the denominator. WES positivity rates were calculated using the number of cases undergoing WES in each subgroup as the denominator. Italicized P values indicate comparisons performed using Fisher’s exact test; all other P values were calculated using the chi-square testAbbreviations: CNV-seq Copy number variation sequencing, WES Whole-exome sequencing, IVF-ET In vitro fertilization and embryo transfer, NIHF Nonimmune hydrops fetalis Table 3Comparison of CNV-seq and WES findings according to clinical characteristics in NIHFGroupsOverall genetic positivity rate, n (%)CNV-seq positivity rate, n (%)Numerical chromosomal abnormality rate, n (%)CNVs rate, n (%)WES positivity rate among cases undergoing WES, n (%)Early gestational group (≤ 13+6W) (CNV n = 83, WES n = 63)51 (61.4%)43 (51.8%)37 (44.6%)6 (7.2%)8 (12.7%)Mid to late gestational group (≥ 14 W) (CNV n = 38, WES n = 32)11 (28.9%)6 (15.8%)1 (2.6%)5 (13.2%)6 (18.8%)P value0.001 < 0.001 < 0.0010.3180.542Non-advanced maternal age group (singleton < 35 years; twin < 32 years) (CNV n = 96, WES n = 77)49 (51%)38 (39.6%)27 (28.1%)11 (11.5%)12 (15.6%)Advanced maternal age group (singleton ≥ 35 years; twin ≥ 32 years) (CNV n = 25, WES n = 18)13 (52%)11 (44%)11 (44%)02 (11.1%)P value0.9320.6890.1280.1181IVF-ET (CNV n = 12, WES n = 11)7 (58.3%)5 (41.7%)5 (41.7%)02(18.2%)Natural conception (CNV n = 109, WES n = 84)55 (50.5%)44 (40.4%)33 (30.3%)11 (10.1%)12 (14.3%)P value0.60410.5140.5990.663isolated NIHF (CNV n = 19, WES n = 14)2 (10.5%)1 (5.3%)01 (5.3%)1 (7.1%)non-isolated NIHF (CNV n = 102, WES n = 81)60 (58.8%)48 (47.1%)38 (37.3%)10 (9.8%)13 (16%)P value < 0.0010.0010.00110.685Overall genetic positivity and CNV-seq-related rates were calculated using the total number of cases in each subgroup as the denominator. WES positivity rates were calculated using the number of cases undergoing WES in each subgroup as the denominator. Italicized P values indicate comparisons performed using Fisher’s exact test; all other P values were calculated using the chi-square testAbbreviations: CNV-seq Copy number variation sequencing, WES Whole-exome sequencing, IVF-ET In vitro fertilization and embryo transfer, NIHF Nonimmune hydrops fetalis Comparison of CNV-seq and WES findings according to clinical characteristics in NIHF Overall genetic positivity and CNV-seq-related rates were calculated using the total number of cases in each subgroup as the denominator. WES positivity rates were calculated using the number of cases undergoing WES in each subgroup as the denominator. Italicized P values indicate comparisons performed using Fisher’s exact test; all other P values were calculated using the chi-square test Abbreviations: CNV-seq Copy number variation sequencing, WES Whole-exome sequencing, IVF-ET In vitro fertilization and embryo transfer, NIHF Nonimmune hydrops fetalis Similarly, non-isolated NIHF had a higher overall genetic positivity rate than isolated NIHF (58.8% vs. 10.5%, P < 0.001), largely because of the higher rate of numerical chromosomal abnormalities (37.3% vs. 0%, P = 0.001). However, no significant differences were observed in the detection rates of pathogenic CNVs (9.8% vs. 5.3%, P = 1.000) or WES-detected variants (16.0% vs. 7.1%, P = 0.685). Genetic positivity did not differ significantly by maternal age or conception method, including overall positivity, CNV-seq positivity, numerical chromosomal abnormality rate, pathogenic CNVs, or WES positivity rate, with all P values > 0.05. Pregnancy outcome As demonstrated in Table 4, pregnancy outcome data were available for all 121 cases. Seventy-five cases resulted in induced TOP (62.0%), 22 in miscarriage or stillbirth (18.2%), 2 in neonatal death (1.7%), and 22 in healthy live birth and survival during follow-up(18.2%). For the purpose of outcome analysis, adverse pregnancy outcomes were defined as induced TOP, miscarriage, stillbirth, or neonatal death, as shown in Table 5. Further stratification by gestational age at diagnosis showed that the adverse pregnancy outcome rate did not differ significantly between the early-gestation group and the mid-to-late-gestation group (84.3% vs. 76.3%, P = 0.288). Subsequently, as depicted in Table 6, among cases without induced TOP, spontaneous fetal or neonatal death occurred more frequently in genetically positive cases than in genetically negative cases (76.9% vs. 42.4%, P = 0.035). In addition, the rate of induced TOP was significantly higher among families with genetically positive results than among those with genetically negative results (94.2% vs. 57.8%, P < 0.001).Table 4Pregnancy outcomes of genetic variantsDetection of NIFH by CNV-seq and WESTermination of pregnancy (TOP)Stillbirth or missed abortionNeonatal deathLive birthThe total number of NIHFGenetic disorders or chromosome abnormalities49 (79.0%)10 (16.1%)0 (0.0%)3 (4.8%)62 (100.0%)Non-presence of pathogenic CNVs or genes26 (44.1%)12 (20.3%)2 (3.4%)19 (32.2%)59 (100.0%)The total number of NIHF75 (62.0%)22 (18.2%)2 (1.7%)22 (18.2%)121 (100.0%)Table 5Pregnancy outcomes according to gestational age at diagnosisPregnancy outcomeEarly-gestation group (≤ 13+6 weeks, n = 83)Mid-to-late-gestation group (≥ 14 weeks, n = 38)P valueAdverse pregnancy outcomes70 (84.3%)29 (76.3%)0.288Healthy survival13 (15.7%)9 (23.7%)Adverse pregnancy outcomes included termination of pregnancy (TOP), miscarriage, stillbirth, and neonatal deathTable 6Genetic positivity rate concerning adverse pregnancy outcomesGroupsPositive resultsNegative resultsP valueStillborns or missed abortion and neonatal death10 (76.9%)14 (42.4%)0.035Live birth3 (23.1%)19 (57.6%)Termination of Pregnancy49 (94.2%)26 (57.8%) < 0.001Live birth3 (5.8%)19 (42.2%) Table 4Pregnancy outcomes of genetic variantsDetection of NIFH by CNV-seq and WESTermination of pregnancy (TOP)Stillbirth or missed abortionNeonatal deathLive birthThe total number of NIHFGenetic disorders or chromosome abnormalities49 (79.0%)10 (16.1%)0 (0.0%)3 (4.8%)62 (100.0%)Non-presence of pathogenic CNVs or genes26 (44.1%)12 (20.3%)2 (3.4%)19 (32.2%)59 (100.0%)The total number of NIHF75 (62.0%)22 (18.2%)2 (1.7%)22 (18.2%)121 (100.0%) Pregnancy outcomes of genetic variants Table 5Pregnancy outcomes according to gestational age at diagnosisPregnancy outcomeEarly-gestation group (≤ 13+6 weeks, n = 83)Mid-to-late-gestation group (≥ 14 weeks, n = 38)P valueAdverse pregnancy outcomes70 (84.3%)29 (76.3%)0.288Healthy survival13 (15.7%)9 (23.7%)Adverse pregnancy outcomes included termination of pregnancy (TOP), miscarriage, stillbirth, and neonatal death Pregnancy outcomes according to gestational age at diagnosis Adverse pregnancy outcomes included termination of pregnancy (TOP), miscarriage, stillbirth, and neonatal death Table 6Genetic positivity rate concerning adverse pregnancy outcomesGroupsPositive resultsNegative resultsP valueStillborns or missed abortion and neonatal death10 (76.9%)14 (42.4%)0.035Live birth3 (23.1%)19 (57.6%)Termination of Pregnancy49 (94.2%)26 (57.8%) < 0.001Live birth3 (5.8%)19 (42.2%) Genetic positivity rate concerning adverse pregnancy outcomes As shown in Appendix 1, among the 22 live-born survivors, hydrops resolved between 16 and 28 weeks of gestation in 13 cases without accompanying fetal structural anomalies. Notably, three cases exhibited genetic abnormalities, including one with a heterozygous variant of the F8 gene of maternal origin, one with a heterozygous variation of the SLC3A1 gene of paternal origin, and one with 15q11.2 microdeletion syndrome of paternal origin. Postnatally, edema persisted in nine cases, with five cases resolving spontaneously within one week postpartum and one case resolving spontaneously within one month postpartum. Among the remaining cases, hydrops resolved after postnatal surgical intervention in one case, and two cases with chyloperitoneum improved after abdominal paracentesis. Subsequent pregnancies Follow-up data on subsequent pregnancies were available for 20 families. Among the 13 families in whom a genetic abnormality had been confirmed in the first pregnancy, all subsequent pregnancies resulted in the delivery of healthy full-term neonates (13/13). Of these subsequent pregnancies, 9 underwent amniocentesis-based genetic testing, all of which yielded negative results. Among the 7 families with negative genetic testing in the first pregnancy, 5 subsequent pregnancies resulted in the delivery of healthy full-term neonates, 1 was terminated because of fetal hydrops associated with multiple ultrasonographic abnormalities in the second pregnancy, and 1 was ongoing at the time of follow-up. In this subgroup, 4 pregnancies underwent amniocentesis-based genetic testing, with all results negative, whereas the remaining 3 had low-risk findings on non-invasive prenatal testing. Overall, except for one pregnancy that was terminated because of recurrent fetal hydrops, no major ultrasound abnormalities were detected in the other 19 subsequent pregnancies. Representative case One fetus presented with NIHF accompanied by hydronephrosis, a small rectal diameter, and an enlarged cisterna magna. CNV-seq revealed a de novo 9.8-Mb duplication involving 15q26, and WES identified a de novo likely pathogenic variant in NOTCH2 (Fig. 4). The parents elected termination of pregnancy. In a subsequent pregnancy, the fetus showed no major ultrasound abnormalities. Because of concern about recurrence, the couple chose amniocentesis with karyotyping, CNV-seq, and WES; all results were negative. The pregnancy continued and resulted in the birth of a healthy child who remained well at follow-up.Fig. 4a. Results of CNV-seq of case W35 and the parents and the second pregnant fetus. b Validation of NOTCH2 Sanger sequencing of case W35 and the parents and the second pregnant fetus a. Results of CNV-seq of case W35 and the parents and the second pregnant fetus. b Validation of NOTCH2 Sanger sequencing of case W35 and the parents and the second pregnant fetus Discussion In this retrospective cohort of 121 NIHF cases, pathogenic or likely pathogenic genetic findings were identified in 51.2% of cases, highlighting the substantial contribution of genetic disorders to the etiology of NIHF. CNV-seq identified abnormalities in 40.5% of the cohort, with numerical chromosomal abnormalities accounting for the majority of positive findings. Monosomy X and trisomy 18 were the most frequent abnormalities, consistent with previous reports [22, 23]. Among the 11 cases with pathogenic structural chromosomal abnormalities, nine involved genomic regions previously associated with abnormal organ development, whereas two involved regions mainly associated with neurodevelopmental abnormalities [24–30]. Existing research suggests variable detection rates of pathogenic CNVs in NIHF cases, with these CNVs predominantly associated with abnormalities in organ development, a trend observed within this specific cohort [18, 31, 32]. In the cohort, the detection rate of WES was 14.7% (14/95), similar to findings in other studies [33]. One case carried both a pathogenic structural chromosomal abnormality and a likely pathogenic sequence variant. The genes harboring pathogenic or likely pathogenic variants identified in this research are frequently linked to skeletal system abnormalities and RASopathies. The variants in FGFR2 and COL1A1, which were associated with skeletal abnormalities, were inherited from affected parents. The mother of case W40, who carried the FGFR2 variant, showed a navicular-shaped head and elbow valgus, whereas the father of case W45, who carried the COL1A1 variant, had mild osteoporosis. Previous studies have also noted discrepancies in clinical presentations between parents and their offspring carrying pathogenic variants [11]. Additional variants, including those in F8 and FOXP3, followed a maternally inherited X-linked recessive pattern. Hemizygous male fetuses may be affected, whereas heterozygous female carriers are usually unaffected. The aforementioned couples faced a 50% risk of transmitting the pathogenic variant in subsequent pregnancies. Furthermore, One fetus carried a homozygous pathogenic variant in NEU1, inherited from two heterozygous carrier parents, indicating a 25% recurrence risk in future pregnancies. In conclusion, parents carrying pathogenic variants are at an increased risk of recurrent fetal abnormalities in subsequent pregnancies. Preimplantation genetic testing, such as PGT-M when appropriate, together with targeted prenatal diagnosis in future pregnancies, may help identify and manage recurrence risks. RASopathies are a group of disorders caused by pathogenic variants in genes involved in the RAS/MAPK signaling pathway, including PTPN11, KRAS, RIT1, SHOC2, HRAS, BRAF, SOS1, and SOS2 [34, 35]. These disorders are recognized genetic causes of NIHF and are frequently associated with congenital heart defects [36, 37]. In this study, 2 of the 3 fetuses with RASopathy-associated variants also exhibited tetralogy of Fallot, further supporting this association. The variants identified in RASopathy-associated genes, including RAF1, RIT1, and BRAF, were de novo. Although the recurrence risk after a confirmed de novo variant is generally low, a residual risk of approximately 1–2% has been reported because of possible parental germline mosaicism. Therefore, spontaneous conception with prenatal diagnosis by mid-trimester amniocentesis may be considered in subsequent pregnancies. One case in this study had both a 15q26 duplication and a de novo likely pathogenic variant in NOTCH2, highlighting the potential value of combined chromosomal and sequence-level analysis in complex fetal presentations. The 15q26 duplication is associated with 15q26 overgrowth syndrome, which is characterized by overgrowth and cardiac and renal malformations [38, 39]. Likely pathogenic variants in NOTCH2 have been associated with malformations in multiple organs, including the liver, skeleton, and heart [40, 41]. An interesting finding of this study is that the observed differences in overall genetic diagnostic yield across clinical subgroups were largely attributable to aneuploidy. Cases diagnosed in early gestation and those with non-isolated NIHF had significantly higher overall positivity rates. However, these differences were primarily driven by the increased prevalence of numerical chromosomal abnormalities, particularly 45,X and trisomy 18, in those groups. After excluding aneuploidies, no statistically significant differences were observed in the detection rates of pathogenic CNVs or WES findings across gestational age groups, or isolated versus non-isolated NIHF. This distinction is clinically important. It suggests that while early presentation and additional ultrasound abnormalities may enrich for aneuploidy, the potential contribution of CNVs and single-gene disorders is not confined to specific NIHF subgroups. In this cohort, no statistically significant difference was identified in the overall genetic diagnostic yield among different maternal age groups. One possible explanation is that the CNV-seq-positive findings included a substantial proportion of 45,X and pathogenic structural chromosomal abnormalities. Unlike common autosomal trisomies, monosomy X is generally considered to have a weaker association with increasing maternal age. In addition, submicroscopic pathogenic CNVs are not typically regarded as strongly maternal-age dependent. As a result, the contribution of maternal age to the overall genetic detection rate may have been less apparent in our dataset. Moreover, the relatively small sample size in the advanced maternal age group may have reduced the statistical power of the analysis, thereby limiting our ability to detect modest group differences. Taken together, these factors may explain the lack of a significant association observed in the present study. Future investigations based on larger cohorts are warranted to further clarify the relationship between maternal age and the heterogeneous genetic causes of NIHF. These findings have practical implications for prenatal diagnosis. In some settings, clinicians may be more inclined to offer WES only for selected NIHF cases, such as those presenting early in gestation or with multiple anomalies. However, our data do not support restricting WES to these subgroups. Rather, once common aneuploidies have been excluded, broader genomic testing may remain informative across the spectrum of NIHF phenotypes. This may be particularly relevant when prior screening, such as NIPT or rapid aneuploidy testing has reduced the suspicion for common numerical chromosomal abnormalities. In such circumstances, progression to CNV-level analysis and/or WES may be considered, depending on local practice patterns, turnaround time, resource availability, and patient preferences. Our study highlights the important clinical value of NGS in the etiologic evaluation of NIHF. CNV-seq can effectively identify aneuploidies and submicroscopic copy number variants, whereas WES provides additional diagnostic yield for monogenic disorders, demonstrating clear complementarity between the two approaches. Given the marked etiologic heterogeneity of NIHF, an NGS-based tiered testing strategy may improve diagnostic yield, support genetic counseling, facilitate recurrence-risk assessment, and inform the management of subsequent pregnancies. At the same time, the prenatal application of next-generation sequencing still faces several challenges, including testing cost, variant interpretation, and the limited availability of fetal phenotypic information; therefore, genetic findings should be interpreted in conjunction with ultrasound features, familial validation, and the broader clinical context. Compared with the findings reported in Deng et al. [18], the overall diagnostic yield in our cohort was relatively high, which may be attributable to several factors. First, the majority of cases in our cohort were non-isolated NIHF, accounting for 84.3% (102/121), a proportion substantially higher than that reported in Deng et al. (35.8%, 39/109). Non-isolated NIHF is more likely to be associated with detectable chromosomal or monogenic abnormalities, which may have contributed to the higher diagnostic yield. Second, our institution is a national clinical research center for obstetric and gynecologic diseases and serves as a tertiary referral center; consequently, our cohort may have been enriched for more severe cases or fetuses with complex congenital anomalies, introducing a degree of referral bias. Therefore, differences in cohort composition and testing strategies should be taken into account when comparing diagnostic yields across studies. Pregnancy outcome analysis showed that the rates of intrauterine or postnatal death were significantly higher in genetically positive cases than in genetically negative cases, and the rate of termination of pregnancy was also markedly increased in the genetically positive group, indicating that genetic testing results have a substantial impact on parental decision-making. Nevertheless, it should also be noted that some fetuses with identified genetic abnormalities still achieved favorable outcomes during pregnancy or after birth. This finding suggests that prognostic assessment should not rely solely on genetic results, but should instead incorporate ultrasound phenotype, the severity and dynamic progression of hydrops, parental phenotypes, and familial validation findings in a comprehensive manner. This study has several limitations. First, it was a single-center retrospective study with a limited sample size and an inherent risk of selection bias. Second, some families who elected termination of pregnancy declined fetal autopsy for cultural, economic, or ethical reasons, which limited the assessment of terminal phenotypes in certain cases. Third, the long-term developmental outcomes of live-born infants remain to be evaluated through extended follow-up. Future multicenter studies with larger sample sizes, together with systematic postnatal follow-up, are needed to further refine genotype–phenotype correlations and prognostic assessment in NIHF. Looking ahead, fetal whole-genome sequencing (WGS) has the potential to provide a more comprehensive molecular evaluation of NIHF, as it may detect single-nucleotide variants, small insertions/deletions, CNVs, and certain structural variants within a single assay. With appropriate analytical pipelines, WGS may also allow assessment of mitochondrial DNA variants [42]. Although the prenatal application of WGS still faces challenges related to cost, turnaround time, data interpretation, incidental findings, and clinical implementation, it represents an important future direction for the genetic evaluation of NIHF. Conclusion NIHF is closely associated with a broad and heterogeneous spectrum of genetic abnormalities, including aneuploidies, pathogenic structural chromosomal abnormalities, and single-gene disorders. Cases diagnosed in early gestation and those with additional ultrasound abnormalities showed higher overall genetic positivity rates; however, these differences were mainly driven by numerical chromosomal abnormalities. After excluding aneuploidies, the detection rates of pathogenic or likely pathogenic structural chromosomal abnormalities and WES-detected pathogenic/likely pathogenic variants did not differ significantly across clinical subgroups. These findings suggest that expanded genetic testing may have value across the clinical spectrum of NIHF rather than being limited to selected subgroups. When interpreted together with ultrasound phenotypes, parental testing, and family-based analysis, genetic findings may inform prognostic assessment, recurrence-risk counseling, and the management of future pregnancies. Supplementary Information Supplementary Material 1. Supplementary Material 1.