Combined Karyotyping, Copy Number Variation Sequencing, and Whole Exome Sequencing in the Prenatal Diagnosis of Congenital Heart Disease: Detection Rates and Phenotypic Associations

ABSTRACT

Objective

This study aimed to assess the value of genetic testing in the prenatal congenital heart disease (CHD) diagnosis and evaluate the relationship between genetic variants and prenatal phenotypes.

Methods

A total of 235 fetuses with CHD were classified into three groups: A: isolated simple CHD, B: isolated complex CHD, and C: non‐isolated CHD. All patients underwent karyotyping and copy number variation sequencing (CNV‐seq), with whole‐exome sequencing (WES) performed when indicated. Statistical analyses conducted using IBM SPSS assessed the diagnostic performance of the genetic testing and detection rates across CHD subtypes.

Results

Karyotyping and CNV‐seq revealed that chromosomal abnormality detection, particularly numerical abnormalities, was significantly associated with extracardiac anomalies (p < 0.05). In contrast, the detection rates of pathogenic/likely pathogenic (P/LP) CNVs (CNV‐seq, p = 0.384) and pathogenic gene variants (WES, p = 0.624) did not differ significantly among the three groups. The severity of intracardiac abnormalities, the presence of extracardiac anomalies, and the detection of genetic abnormalities significantly influenced parental decisions regarding pregnancy termination (p < 0.05). The combination of karyotyping and CNV‐seq increased the detection rate by 5.11%. In karyotyping/CNV‐seq‐negative fetuses who underwent WES, WES increased the detection rate by 37.21% (16/43).

Conclusion

Combined genetic testing significantly improved the detection rate and the detection of chromosomal numerical abnormalities was strongly associated with the presence of extracardiac anomalies. Parental decisions regarding pregnancy continuation were significantly influenced by the severity of intracardiac abnormalities, the presence of extracardiac anomalies, and the identification of genetic abnormalities.

This study combined karyotyping, copy number variation sequencing, and whole‐exome sequencing to assess the value of genetic testing in the prenatal diagnosis of congenital heart disease (CHD) and to evaluate the relationship between genetic variants and prenatal phenotypes.

1. Introduction

Congenital heart disease (CHD) refers to a group of conditions characterized by structural abnormalities of the heart or major blood vessels that arise during embryonic development. It is the most common congenital malformation, accounting for approximately 28% of all malformations [1]. The etiology of CHD is complex and includes genetic and environmental factors, as well as maternal age and diseases. Genetic factors are estimated to contribute approximately 40% to 90% of CHD etiology, including chromosomal aneuploidies, copy number variations (CNVs), and single nucleotide variants (SNVs) [2, 3]. However, genetic abnormalities remain undiagnosed in up to 55% of patients. Prenatal detection of clinically significant genetic abnormalities can facilitate informed decision‐making regarding pregnancy continuation, fetal interventions, and neonatal follow‐up. Therefore, prenatal diagnosis of genetic abnormalities in CHD is clinically important [4].

Fetal echocardiography provides a clear view of cardiac structure and function and defines the type, location, and severity of cardiac malformations, making it the preferred method for detecting CHD and other structural abnormalities [5, 6]. However, because most cardiac malformations develop gradually during fetal growth, they are often not detectable before the second trimester. The limitations of ultrasound highlight the importance of incorporating genetic testing for a more comprehensive assessment. Genetic testing can elucidate the underlying etiology of structural cardiac abnormalities, enabling more precise prediction of disease onset, progression, and prognosis, thereby providing a stronger basis for clinical decision‐making.

Chromosomal aneuploidy is the earliest established cause of CHD, accounting for 28%–45% of prenatally diagnosed CHD cases [7]. Karyotyping is primarily used to identify chromosomal numerical abnormalities and large structural variants; however, its resolution is limited, as it cannot detect abnormalities smaller than 5 Mb [8]. Genomic CNVs are important contributors to all CHD types [9]. Studies have shown that pathogenic CNVs are detected in 7%–36% of fetuses with isolated or non‐isolated CHD with a normal karyotype [10, 11]. Copy number variation sequencing (CNV‐seq) and microarray technologies, with high resolution and short turnaround times, can identify small CNVs in fetuses with CHD [12] and have become indispensable clinical diagnostic tools. In addition to chromosomal aneuploidies and CNVs, disruptive variants in pathogenic genes constitute a major genetic etiology of CHD [13, 14]. Whole‐exome sequencing (WES) enables analysis of SNVs and insertions/deletions within protein‐coding regions of genes, facilitating the discovery of novel pathogenic genes in patients with CHD [15, 16]. Ultrasound and genetic testing each have advantages and limitations, and a combined strategy helps to improve the comprehensiveness and accuracy of diagnosis, making the prenatal management of CHD more precise and efficient.

This study categorized 235 fetuses with CHD based on the anatomical complexity of cardiac malformations and the presence of extracardiac abnormalities. It aimed to evaluate the prenatal diagnostic value of karyotyping, CNV‐seq, and WES across CHD types, and to elucidate the relationship between genetic variants and CHD phenotypes, as well as their impact on pregnancy outcomes, thereby supporting optimization of prenatal diagnosis and genetic counseling.

2. Methods

2.1. Participants

This single‐center retrospective study included 235 pregnant women whose fetuses were diagnosed with CHD on prenatal ultrasound at Xiamen Maternal and Child Health Hospital between March 2020 and April 2025. The diagnosis was confirmed according to the National Health Ministry's birth defect map and the 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD‐11). Participants underwent concurrent amniotic fluid karyotyping and CNV‐seq testing. Whole‐exome sequencing (WES) was recommended for cases in which neither karyotyping nor CNV‐seq detected a pathogenic variant.

Fetuses with CHD were categorized into three groups: (A) Isolated simple CHD; (B) Isolated complex CHD; (C) Non‐isolated CHD. Isolated simple CHD was defined as a single cardiac malformation that typically does not cause hemodynamic compromise. Isolated complex CHD was defined as one or more structural abnormalities in the heart and/or great vessels that lead to significant hemodynamic compromise. All cases of CHD with extracardiac anomalies were classified as non‐isolated CHD, regardless of whether the underlying cardiac lesion was simple or complex. To better characterize the phenotypic composition of CHD, intracardiac abnormalities were classified according to the criteria described by Botto et al. [17]. Extracardiac anomalies were further subclassified into craniofacial and neck anomalies, urinary system anomalies, nervous system anomalies, skeletal system anomalies, digestive system anomalies, thoracic anomalies, abdominal anomalies, and situs anomalies.

This study was approved by the Ethics Committee of Xiamen Maternal and Child Health Hospital (approval No. KY‐2026‐072‐K01), and written informed consent was obtained from all participants.

2.2. Ultrasound Examination

The examination was performed using a GE Voluson E8 color Doppler ultrasonography system with a corresponding probe (operating frequency 4–6 MHz). After localizing the fetus and assessing the heartbeat, a detailed fetal echocardiography was performed. Two physicians specializing in prenatal diagnosis performed independent assessments, and the final diagnosis was reached by consensus.

2.3. Karyotype Analysis

Amniotic fluid specimens were obtained via ultrasound‐guided transabdominal puncture and were subsequently inoculated, cultured, and prepared in accordance with standard protocols. Under the experimental requirements, 20–100 well‐spread metaphase cells were counted, and a minimum of five karyotypes were analyzed and designated according to the International System for Human Cytogenomic Nomenclature (ISCN 2020).

2.4. CNV‐Seq

Genomic DNA was extracted from cord blood cells or amniotic fluid for whole‐genome sequencing library preparation. Raw sequencing data were generated using the MGISEQ‐2000 platform (MGITech). The pathogenicity of CNVs was interpreted using commercial software from BGI, which follows the guidelines of the American College of Medical Genetics and Genomics (ACMG) [18].

2.5. WES

Genomic DNA extracted from umbilical cord blood or amniotic fluid was used for library preparation. The libraries were sequenced on the high‐throughput MGISEQ‐2000 platform. Sequencing reads were aligned to the UCSC hg19/GRCh37 human reference genome using BWA, and duplicate reads were removed. ExomeDepth, an R package based on a beta‐binomial model for read depth data, was employed for exome‐level CNV detection, enabling systematic calling of heterozygous deletions and duplications at the single‐ or multi‐exon level. All detected variants were interpreted for pathogenicity according to ACMG guidelines.

2.6. Definitions of Detection Rate and Diagnostic Rate

In this study, the detection rate was defined as the proportion of fetuses in whom a pathogenic (P) or likely pathogenic (LP) variant was identified by genetic testing, regardless of whether the variant was considered to explain the fetal phenotype. The diagnostic rate was defined as the proportion of fetuses in whom a P/LP variant considered to explain the fetal phenotype was identified by WES. In this study, a variant was considered phenotype‐explaining when all of the following criteria were met: (1) the gene–disease relationship was well established; (2) the phenotype associated with the variant was consistent with the fetal ultrasound findings (intracardiac and/or extracardiac); and (3) the zygosity and mode of inheritance were consistent with disease causation—that is, a heterozygous P/LP variant in a gene with autosomal dominant (AD) inheritance, or biallelic P/LP variants in a gene with autosomal recessive (AR) inheritance. A single heterozygous variant in a gene associated with an AR disorder was considered insufficient to explain the fetal phenotype, unless accompanied by a second P/LP variant or established evidence of a dominant effect; such variants were excluded from the diagnostic rate.

2.7. Follow‐Up

Pregnant women who did not undergo induced labor were followed up within 1 year postpartum. Those who refused to be interviewed or could not be contacted were considered lost to follow‐up.

2.8. Statistical Analysis

Statistical analyses were performed using IBM SPSS Statistics version 26. Parametric tests were used for age comparisons, while non‐parametric tests were applied for gestational age comparisons. Detection rates of karyotyping, CNV‐seq, and WES, as well as pregnancy outcomes, were compared using Pearson's chi‐square test or Fisher's exact test. A p < 0.05 was defined as statistically significant.

3. Results

3.1. Characteristics of the Participant Cohort and Overview of Genetic Abnormalities

Among the 235 participants enrolled in the study (Table 1), 44 were classified into group A, with a mean age of 29.8 years (range: 22–39) and a mean gestational age of 22.91 weeks (range: 14–28). Group B included 126 patients with a mean age of 30.62 years (range: 19–43) and a mean gestational age of 22.4 weeks (range: 13–39). The remaining 65 patients were assigned to group C, with a mean age of 31.03 years (range: 21–41) and a mean gestational age of 18.06 weeks (range: 12–32). There were no significant differences in the maternal age among the three groups (p = 0.221). The gestational age in group C was significantly different from those in group A and group B (p < 0.001 and p < 0.001, respectively).

TABLE 1.

Participant cohort characteristics in fetal congenital heart disease.

Groups	A (44)		B (126)		C (65)		Total (235)
Maternal age
Mean	29.8		30.62		31.03
Range	22–39		19–43		21–41
Gestational age
Mean	22.91		22.4		18.06
Range	14–28		13–39		12–32
CHD classification
CTD	0	0	57	45.24%	13	20.00%	70	29.79%
Septal defects	40	90.91%	7	5.56%	16	24.62%	63	26.81%
RVOTO	3	6.82%	19	15.08%	4	6.15%	26	11.06%
LVOTO	1	2.27%	17	13.49%	2	3.08%	20	8.51%
Heterotaxy	0	0	4	3.17%	12	18.46%	16	6.81%
VSD + LVOTO	0	0	9	7.14%	0	0	9	3.83%
AVSD	0	0	5	3.97%	4	6.15%	9	3.83%
SV/Complex SV	0	0	3	2.38%	2	3.08%	5	2.13%
APVR	0	0	2	1.59%	0	0	2	0.85%
SV + RVOTO	0	0	1	0.79%	0	0	1	0.43%
AVSD + LVOTO	0	0	1	0.79%	0	0	1	0.43%
Others	0	0	1	0.79%	0	0	4	1.70%
Lack of details	0	0	0	0	9	13.85%	9	0.09%

Abbreviations: A, isolated simple CHD; APVR, anomalous pulmonary venous return; AVSD, atrioventricular septal defect; B, isolated complex CHD; C, non‐isolated CHD; CTD, conotruncal defects; LVOTO, left ventricular outflow tract obstruction; n, number; NCA, numerical chromosomal abnormalities; RVOTO, right ventricular outflow tract obstruction; SV, single ventricle.

Intracardiac abnormalities were classified according to criteria described by Botto et al. [17]. As shown in Table 1, among all participants, conotruncal defects (CTD) accounted for the largest proportion (29.78%), followed by septal defects (26.81%) and right ventricular outflow tract obstruction (RVOTO) (11.06%). Group A predominantly consisted of patients with septal defects, whereas group B was mainly characterized by CTD, left ventricular outflow tract obstruction (LVOTO), and RVOTO. Group C primarily comprised patients with septal defects, CTD, and heterotaxy. As shown in Table 2, across all categories, the highest detection rates of aneuploidy were observed in atrioventricular septal defects (AVSD) (44%) and septal defects (19.05%), whereas no chromosomal abnormalities were identified in cases with heterotaxy. Among AVSD cases, only chromosomal aneuploidies were detected, and no CNVs or pathogenic gene variants were identified.

TABLE 2.

Overview of genetic abnormalities in fetal congenital heart disease.

CHD classification (n)	Karyotyping positive rate						CNV‐seq positive rate		WES positive rate		Total
	Positive rate of NCA n, rate (%)		Positive rate of structural CA n, rate (%)		Positive rate of chromosome chimerism n, rate (%)		Positive rate Of P/LP CNV n, rate (%)		Positive rate Of P/LP gene n, rate (%)
CTD (70)	7	10.00	0	0	2	2.86	5	7.14	3	4.29	17	24.29
Septal defects (63)	12	19.05	1	1.59	3	4.76	5	7.94	1	1.59	22	34.92
ROVOTO (26)	2	7.69	1	3.85	0	0	4	15.38	1	3.85	8	30.77
LOVOTO (20)	2	10.00	0	0	0	0	3	15.00	2	10.00	7	35.00
Heterotaxy (16)	0	0	0	0	0	0	1	6.25	0	0	1	6.26
VSD + LVOTO (9)	1	11.11	0	0	1	11.11	0	0	0	0	2	22.22
AVSD (9)	4	44.44	0	0	0	0	0	0	0	0	4	44.44
SV/Complex SV (5)	0	0	0	0	0	0	0	0	1	20.00	1	20.00
APVR (2)	0	0	0	0	0	0	0	0	0	0	0	0
SV + RVOTO (1)	0	0	0	0	0	0	0	0	0	0	0	0
AVSD + LVOTO (1)	0	0	0	0	0	0	0	0	0	0	0	0
Others (4)	2	50.00	0	0	0	0	0	0	0	0	2	50.00
Lack of details (9)	7	77.78	0	0	0	0	0	0	0	0	7	77.78
Total (235)	37	15.74	2	0.85	6	2.55	18	7.66	8	3.40	71	30.21

Note: With the exception of positive rate Of P/LP gene (defined as the diagnostic rate, i.e., restricted to variants considered phenotype‐explaining), all other rates in this table are reported as detection rates. Two cases with karyotypic structural abnormalities—46,XX,der(6)t(6;14)(p25;q31) and 46,XN,?dup(1)(q32q41)—were found to carry corresponding pathogenic variants by CNV‐seq; these two cases are counted solely in the CNV‐seq positive rate (and not in the karyotype positive rate).

Abbreviations: A, isolated simple CHD; APVR, anomalous‐pulmonary venous return; AVSD, atrioventricular septal defect; B, isolated complex CHD; C, non‐isolated CHD; CA, chromosomal abnormality; CNV, copy number variants; CTD, conotruncal defects; LP, likely pathogenic; LVOTO, left ventricular outflow tract obstruction; n, number; NCA, numerical chromosomal abnormalities; P, pathogenic; RVOTO, right ventricular outflow tract obstruction; SV, single ventricle.

In group C, the most common extracardiac abnormality was increased nuchal translucency (NT) (35.38%), followed by facial and neck anomalies (27.69%), skeletal system anomalies (20.00%), and nervous system anomalies and fetal hydrops (18.46%), which ranked fourth (Table 3). We also analyzed the detection rate of chromosomal abnormalities in relation to extracardiac anomalies. The highest detection rate of chromosomal numerical abnormalities was observed in cases of fetal hydrops and choroid plexus cysts (100%), followed by neck cystic hygroma (92.31%), absent nasal bone (87.5%), increased NT (76.19%), and postural abnormalities (80%). Notably, no chromosomal abnormalities were detected in cases of organ heterotaxy or abnormalities of the urinary or digestive systems.

TABLE 3.

Ultrasound findings and detection rates of chromosomal abnormalities in fetal extracardiac anomalies associated with non‐isolated congenital heart disease.

Category	Total		NCA		P/LP CNV		P/LP gene
	N	Rate (%)	N	Rate (%)	N	Rate (%)	N	Rate (%)
Facial and neck anomalies	25	38.46	21	84.00	2	8.00	0	0
Neck cystic hygroma	13		12	92.31	0	0	0	0
Absence nasal bone	8		7	87.50	1	12.50	0	0
Cleft lip and palate	2		1	50.00	0	0	0	0
Microtia	1		0	0	0	0	0	0
Low‐set ears	1		1	100.00	1	100.00	0	0
Skeletal system anomalies	12	18.46	8	66.67	2	16.67	1	8.33
Postural abnormalities	5		4	80.00	1	20.00	1	20.00
Metacarpal hypoplasia	5		2	40.00	0	0	0	0
Strawberry head	2		2	100.00	1	50.00	0	0
Nervous system anomalies	12	18.46	8	66.67	1	8.33	2	16.67
Choroid plexus cyst	6		6	100.00	1	16.67	0	0
Blake's pouch cyst	2		1	50.00	0	0	1	50.00
Meningocele	2		0	0	0	0	0	0
Abnormal cavum septum pellucidum	1		0	0	0	0	1	100.00
Cerebellar hypoplasia and vermis hypoplasia	1		1	100.00	0	0	0	0
Organ heterotaxy	10	15.38	0	0	1	10.00	2	20.00
Urinary system anomalies	6	9.23	0	0	1	16.67	1	16.67
Renal hypoplasia	1		0	0	0	0	0	0
Renal agenesis	1		0	0	0	0	1	100.00
Horseshoe kidney	1		1	100.00	0	0	0	0
Megacystis	1		1	100.00	0	0	0	0
Bilateral hydronephrosis	1		0	0	1	100.00	0	0
Right renal displacement	1		0	0	0	0	0	0
Abdominal anomalies	4	6.15	3	75.00	1	25.00	0	0
Omphalocele	4		3	75.00	1	25.00	0	0
Digestive system anomalies	4	6.15	0	0	0	0	0	0
Asplenia	2		0	0	0	0	0	0
Esophageal atresia with esophagotracheal fistula	1		0	0	0	0	0	0
Abnormal gallbladder	1		0	0	0	0	0	0
Small stomach vesicles	1		0	0	0	0	0	0
Thoracic anomalies	2	3.08	1	50.00	0	0	0	0
Pulmonary hypoplasia	1		0	0	0	0	0	0
Pleural effusion	1		1	100.00	0	0	0	0
Others								0
Increased nuchal translucency/nuchal fold thickening	23	35.38	17	73.91	2	8.70	0	0
Nuchal translucency thickening	21		16	76.19	1	4.76	0	0
Nuchal fold thickening	2		1	50.00	1	50.00	0	0
Fetal hydrops	12	18.46	12	100.00	0	0	0	0
Fetal accessory structure abnormalities	9	13.85	3	33.33	2	22.22	0	0
Single umbilical artery	7		2	28.57	1	14.29	0	0
Umbilical cord tightly coiled	1		0	0	0	0	0	0
Umbilical cord cyst	1		1	100.00	1	100.00	0	0

Note: In the non‐isolated group, only one case showed a karyotype of 46,XN,?dup(1)(q32q41), and CNV identified a duplication of approximately 10.37 Mb in the 1q32.1q41 region, which was included in the CNV count. No other chromosomal structural abnormalities were detected in the remaining cases. Bold rows indicate system‐level categories; non‐bold rows indicate specific anomalies within each category. Bold‐row counts are the sum of the corresponding non‐bold rows.

Abbreviations: CNV, copy number variants; LP, likely pathogenic; n, number; NCA, numerical chromosomal abnormalities; P, pathogenic.

3.2. Karyotype Analysis Findings and Detection Rate

As shown in Table 4, the karyotypic abnormality detection rates were 11.36% in group A, 6.35% in group B, and 50.77% in group C, respectively. The detection rate of karyotypic abnormalities was not significantly different between groups A and B (p = 0.281). However, the detection rate in group C was significantly different from that in groups A (p < 0.001) and B (p < 0.001). Among all fetuses with CHD, the abnormality detection rate by karyotyping was 19.57% (46/235), with autosomal aneuploidy being the most common, followed by structural abnormalities.

TABLE 4.

Results of chromosomal karyotyping analysis of fetal CHD.

Karyotype/group (n)	A (44) n, rate (%)		B (126) n, rate (%)		C (65) n, rate (%)		Total (235) n, rate (%)
Non‐pathogenic karyotype	39	88.64	118	93.65	32	49.23	189	80.43
Abnormal karyotype	5	11.36	8	6.35	33	50.77	46	19.57
NCA
ANA	3	6.82	4	3.17	23	35.38	30	12.77
SCA	1	2.27	0	0	4	6.15	5	2.13
SCA+ ANA	0	0	0	0	1	1.54	1	0.43
Chromosome chimerism	0	0	2	1.59	4	6.15	6	2.55
Structural CA	1	2.27	2	1.59	1	1.54	4	1.70

Abbreviations: A, isolated simple CHD; ANA, autosomal numerical abnormality; B, isolated complex CHD; C, non‐isolated CHD; CA, chromosomal abnormality; n, number; NCA, numerical chromosomal abnormalities; SCA, sex chromosome aneuploidy.

3.3. CNV‐Seq Findings and Detection Rate

As shown in Table 5, among cases with non‐pathogenic karyotypes, CNV analysis identified an additional 11 abnormal cases, whereas among cases with abnormal karyotypes, CNV analysis detected fewer abnormalities than karyotyping in three cases, all of which were balanced translocations or low‐level mosaicism. The overall detection rate of CNV was 22.98% (54/235), and the combined detection rate increased by 5.11% (12/235). As shown in Table 6, among the pathogenic variants detected by CNV‐seq, 39 (16.60%) were aneuploidies, and 19 (8.09%) were P/LP CNVs. When aneuploidies were excluded, the detection rates for group A, group B, and group C were 2.27%, 10.32%, and 7.69%, respectively, with no significant difference in the detection rates of P/LP CNVs among the three groups (p = 0.384). When aneuploidies were included, the detection rates were 9.09% in group A, 12.70% in group B, and 52.31% in group C. Group C showed a significant difference compared with the other two groups (group A vs. C, p < 0.001; group B vs. C, p < 0.001), whereas no statistically significant difference was observed between groups A and B (p = 0.817). As shown in Table 6, a total of 19 P/LP CNVs were detected in 18 patients, among which 22q11.2 deletion was the most common pathogenic CNV, accounting for 26.31% (5/19), followed by 15q11.2 deletion in three cases, accounting for 15.79%. In all five cases with 22q11.2 deletion, a deletion of approximately 3 Mb was detected, all involving the TBX1 gene. Prenatal phenotypes revealed only intracardiac abnormalities, including two cases of tetralogy of Fallot (TOF), two cases of interrupted aortic arch (IAA), and two cases of ventricular septal defect (VSD), with no extracardiac abnormalities observed. Deletions on chromosome 15 ranged from 500 kb to 1 Mb and encompassed the TUBGCP5, NIPA1, NIPA2, and CYFIP1 genes. Their prenatal phenotypes were heterogeneous, ranging from VSD and pulmonary stenosis to heterotaxy syndrome, all of which involved extracardiac abnormalities (Table 7).

TABLE 5.

Detection rates of P/LP CNVs in non‐pathogenic and abnormal karyotypes.

Group	A	B	C	Total
P/LP CNV/Non‐pathogenic karyotype	1/39	9/118	1/32	11/189
P/LP CNV/Abnormal karyotype	4/5	7/8	32/33	43/46
Total	5/44	16/126	33/65	54/235

Abbreviations: A, isolated simple CHD; B, isolated complex CHD; C, non‐isolated CHD; LP, likely pathogenic; P, pathogenic.

TABLE 6.

Results of CNV‐seq analysis of fetal CHD.

CA/Group (n)	A (44)	B (126)	C (65)	Total (235)
NCA (n, %)	4, 9.09	4,3.17	31, 47.69	39, 16.60
ANA	3*	4	26*#	33
SCA	1	0	5#	6
P/LP CNVs (n, %)	1, 2.27	13, 10.32	5, 7.69	19, 8.09
del22q11.21q11.21	0	5	0	5
dup1q32.1q41	0	0	1	1
dup1q21.2q44	0	1	0	1
dup5p13.3p12	0	1	0	1
del6p25.3p25.2	0	1*	0	1
del7q11.23q11.23	0	1	0	1
dup8p12q21.11	0	0	1	1
dup14q31.3q32.33	0	1*	0	1
del15q11.2q11.2	0	0	3*	3
dup16p11.2p11.2	0	2	0	2
del16p11.2p11.2	1*	0	0	1
delXp21.1p21.1	0	1	0	1
Total (n, %)	4, 9.09	16, 12.70	34, 52.31	54, 22.98

Note: The superscript asterisk (*) and hash (#) symbols indicate that two pathogenic variants were found in the same patient within the same group, and each variant was counted separately. Therefore, the duplicates have been removed from the final total.

Abbreviations: A, isolated simple CHD; ANA, autosomal numerical abnormality; B, isolated complex CHD; C, non‐isolated CHD; CA, chromosomal abnormality; CNV, copy number variations; del, deletion; dup, duplication; LP, likely pathogenic; n, number; NCA, numerical chromosomal abnormalities; P, pathogenic; SCA, sex chromosome aneuploidy.

TABLE 7.

Details of P/LP CNVs identified by CNV‐seq in 19 fetuses with CHD.

CHR	N	CNV type	Cytoband	Start	End	Size	CHD‐related genes	Known syndrome	P/LP	Group	CHD type
chr1	1	dup	1q21.2q44	149,177,777	249,250,621	100.07 Mb	49 genes	/	P	B	LVOTO
chr5	1	dup	5p13.3p12	30,004,454	45,873,340	15.87 Mb	/	/	P	B	septal defects
chr6	1	del	6p25.3p25.2	399,570	3,917,387	3.52 Mb	FOXC1 + 20 genes	Axenfeld‐Rieger syndrome 3	P	B	RVOTO
chr7	1	del	7q11.23q11.23	72,736,245	74,124,972	1.39 Mb	ELN	Williams‐Beuren syndrome	P	B	CTD
chr8	1	dup	8p12q21.11	34,592,479	78,126,043	43.53 Mb	more than 49 genes	/	P	C	CTD
chr14	1	dup	14q31.3q32.33	89,224,118	107,286,875	18.06 Mb	more than 49 genes	/	P	B	RVOTO
chr15	3	del	15q11.2q11.2	22,832,519	23,090,897	578.25 kb	NIPA1	15q11.2 deletion syndrome	P	C	RVOTO
		del	15q11.2q11.2	22,554,928	23,650,020	1.10 Mb	TUBGCP5, NIPA1, NIPA2 and CYFIP1	15q11.3 deletion syndrome	P	C	Heterotaxy
		del	15q11.2q11.2	2,770,421	23,276,833	506.4 kb	NIPA1 + 3 genes	15q11.4 deletion syndrome	P	C	RVOTO
chr16	1	dup	16p11.2p11.2	29,465,079	30,336,261	871.18 kb	TBX6	16p11.2 duplication syndrome	P	B	RVOTO
chr16	1	del	16p11.2p11.2	29,378,709	30,512,952	1.13 Mb	TBX6	16p11.2 deletion syndrome	P	A	septal defects
chr22	5	del	22q11.21q11.21	18,629,058	21,657,982	3.03 Mb	TBX1 + 50 genes	DiGeorge syndrome	P	B	CTD
		del	22q11.21q11.21	18,882,825	21,878,026	3.0 Mb	TBX1, etc	DiGeorge syndrome	P	B	LVOTO
		del	22q11.21q11.21	18,950,678	21,657,982	2.71 Mb	TBX1, etc	DiGeorge syndrome	P	B	CTD
		del	22q11.21q11.21	18,920,346	21,517,754	2.6 Mb	TBX1, etc	DiGeorge syndrome	P	B	CTD
		del	22q11.21q11.21	18,648,855	21,800,471	3.1 Mb	CLTCL1 + 43 genes	DiGeorge syndrome	P	B	LVOTO
chrX	1	del	Xp21.1p21.1	31,591,295	31,756,653	165.36Kb	DMD	DMD	P	A	septal defects
chr1	1	dup	1q32.1q41	206,767,478	217,135,257	10.37 Mb	63 genes	/	LP	C	septal defects
chr16	1	dup	16p13.11p13.11	15,116,275	16,494,250	1.38 Mb	MYH11 + 13 genes	/	LP	A	septal defects

Abbreviations: /, indicates unknown; A, isolated simple CHD; B, isolated complex CHD; C, non‐isolated CHD; CNV, copy number variations. CTD, conotruncal defects; DMD: Duchenne muscular dystrophy; LP, likely pathogenic; LVOTO, left ventricular outflow tract obstruction; n, case number; P, pathogenic; RVOTO, right ventricular outflow tract obstruction.

3.4. Detection and Diagnostic Rates of WES Findings

A total of 44 cases underwent WES testing due to cardiac phenotypes, and 18 P/LP variants were detected in 17 participants. No patients in group A underwent WES testing. In group B, among the 31 patients with negative findings on both karyotyping and CNV‐seq, 10 carried 11 P/LP variants (P: MYH7; LP: PNKP, NOTCH1, OSGEP, DNAH5, TELO2, UBA2, MEGF8, ATIC, MPDZ, SLC29A3), yielding a detection rate of 32.26% (10/31). In group C, among the 11 cases with negative findings on both karyotyping and CNV‐seq that underwent WES, five were found to carry P/LP variants (MAP3K7, NPHP3, CEP290, SMAD6, ANKRD11), yielding a detection rate of 45.45% (5/11). There was no significant difference in the WES detection rates between groups B and C (p = 0.624). Among the 43 fetuses in groups B and C who underwent WES and had negative results on both karyotyping and CNV‐seq, WES identified 17 P/LP variants in 16 fetuses, yielding a detection rate of 37.21% (16/43) within this subgroup; these 16 positive cases accounted for 6.81% (16/235) of the overall cohort. Additionally, one fetus in group B (X113), who presented with a cardiac phenotype, was referred for WES. Although the karyotype was normal, CNV‐seq revealed a mosaic duplication of approximately 100.07 Mb at 1q21.2q44 (mosaic proportion: ~8%), which was difficult to interpret definitively. WES subsequently identified a likely pathogenic variant in NOTCH1. This case was not included in the above 43‐case subgroup because the CNV‐seq finding was not negative; therefore, this positive finding was not counted in the detection rate of that subgroup. Because WES is a phenotype‐driven testing strategy, some pathogenic variants can be definitively diagnosed when interpreted in conjunction with inheritance patterns and clinical phenotypes; therefore, we further calculated the diagnostic rate of WES in this study. Among the 18 P/LP variants identified by WES (in 17 fetuses), eight variants (in eight fetuses) followed an autosomal dominant inheritance pattern and were consistent with the fetal clinical phenotypes; these were thus classified as phenotype‐explaining positive findings and were included in the diagnostic rate (Table 8). The remaining 10 variants (in nine fetuses) were single heterozygous variants in genes associated with AR disorders, and a single variant was insufficient to explain the fetal phenotype; these were therefore excluded from the diagnostic rate (Table 9). Accordingly, the overall diagnostic rate of WES was 18.18% (8/44). Stratified by group, the diagnostic rate was 16.13% (5/31) in group B and 27.27% (3/11) in group C; the difference between the two groups was not statistically significant (p = 0.718). Among the 43 fetuses who underwent WES and had negative results on both karyotyping and CNV‐seq, WES additionally identified 6 P/LP variants in six fetuses, yielding a diagnostic rate of 13.95% (6/43) within this subgroup; these positive findings accounted for 2.55% (6/235) of the entire cohort (n = 235).

TABLE 8.

Pathogenic and likely pathogenic variants identified by whole‐exome sequencing (WES) with genetic models concordant with the fetal cardiac phenotypes.

Sample ID	Gene	P/LP	Transcript: variant	Zygosity/parent of origin	Disease or Syndrome/Inheritance	Group	Cardiac type	Extracardiac abnormalities
X113	NOTCH1	LP	c.4699delG (p.Glu1567Argfs*1 3)	het/paternal	OMIM:109730/AD	B	LVOTO
X168	MYH7	P	c.427C>T (p.Arg143Trp)	het/maternal	OMIM:613426/AD	B	LVOTO
X218	PTPN11	P	c.1510A>G (p.Met504Val)	het	OMIM:163950/AD	B	CTD
X203	NOTCH1	LP	c.3082C>T (p.Gln1028*)	het	OMIM:109730/AD	B	SV/Complex SV
X166	MAP3K7	P	c.1535C>T (p.Pro512Leu)	het	OMIM:157800/AD	C	CTD	Nervous system anomalies, Skeletal system anomalies
X183	UBA2	LP	c.771 + 1G>A	het	OMIM:619959/AD	B	CTD
X112	SMAD6	LP	c.834dupC (p.T yr279Leufs*24)	het	OMIM:614823/AD	C	Septal defects	Urinary system anomalies
X116	ANKRD11	LP	2.1:c.988A>T (p. Lys330*)	het/maternal	OMIM:148050/AD	C	RVOTO	Nervous system anomalies

Abbreviations: A, isolated simple CHD; AVSD, atrioventricular septal defect; B, isolated complex CHD; C, non‐isolated CHD; CTD, conotruncal defects; LP, likely pathogenic; LVOTO, left ventricular outflow tract obstruction; P, pathogenic; RVOTO, right ventricular outflow tract obstruction; SV, single ventricle.

TABLE 9.

Pathogenic and likely pathogenic variants identified by WES with genetic models discordant with the fetal cardiac phenotypes.

Sample ID	Gene	P/LP	Transcript: variant	Zygosity/Parent of origin	Disease or Syndrome/Inheritance	Group	Cardiac type	Extracardiac abnormalities
X107	ATIC	LP	c.1064_1065insAAAA (p.Lys357Glnfs*20)	het	OMIM:608688/AR	B	Heterotaxy
X122	OSGEP	LP	c.636 + 1G>A	het	OMIM:617729/AR	B	RVOTO
X131	MPDZ	LP	c.2633_2634insC (p.Ser879Ilefs*5)	het	OMIM:615219/AR	B	Heterotaxy
X131	SLC29A3	LP	c.971C>T(p.Pro324Leu)	het	OMIM:602782/AR	B	Heterotaxy
X142	DNAH5	LP	c.9020_9021delAA (p.Lys3007 Argfs*10)	het	OMIM:608644/AR	B	CTD
X154	NPHP3	P	c.3402_3403delTG (p.Ala1135S erfs*5)	het	OMIM:267010/AR	C	Heterotaxy	Organ heterotaxy
X158	TELO2	LP	c.1514_1515delAG (p.Glu505Al afs*21)	het	OMIM:616954/AR	B	RVOTO
X160	CEP290	P	c.5445_5448delAACT (p.Thr18 16Ilefs*3)	het	OMIM:615991/AR	C	Heterotaxy	Organ heterotaxy
X150	PNKP	LP	c.199‐10_203delCTT TCTGCAGCT GGGinsTCTG AGGGGT	het/maternal	OMIM:605589/AR	B	CTD
X227	MEGF8	LP	c.4975C > T (p.Arg1659*)	het	OMIM:614976/AR	B	CTD

Note: Variants are listed for transparency but were not considered to explain the fetal phenotype and were therefore excluded from the diagnostic rate.

Abbreviation: Abbreviations are as in Table 8.

Among the 44 fetuses who underwent WES, 15 also underwent trio‐WES. P/LP variants were identified in four of these trio cases. In case X168, WES was performed because of a history of fetal left ventricular non‐compaction in a previous pregnancy and recurrence in the current fetus; a pathogenic variant in the MYH7 gene was identified. Case X150 carried a likely pathogenic variant in PNKP, and case X116 carried a likely pathogenic variant in ANKRD11; all three variants were of maternal origin. Case X113 carried a likely pathogenic variant in NOTCH1 of paternal origin.

Among the different categories of CHD (Table 1), the types with the highest detection rates of P/LP variants were: heterotaxy, 25.00%; SV/Complex SV, 20%; and RVOTO, 11.54%. Notably, among the 16 patients with heterotaxy syndrome in this study, six underwent WES, and four were found to carry P/LP variants. In the two cases presenting with heterotaxy syndrome confined to cardiac anomalies, variants were identified in ATIC, SLC29A3, and MPDZ, whereas the other two cases with heterotaxy involving other organs carried variants in NPHP3 and CEP290.

3.5. Pregnancy Outcomes

The live birth, stillbirth, and loss to follow‐up rates were 77.27%, 6.82%, and 15.91%, respectively, in group A; 23.81%, 68.25%, and 10.94%, respectively, in group B; and 10.77%, 86.15%, and 3.08%, respectively, in group C. The differences in pregnancy outcomes among the three groups were statistically significant (p < 0.001).

Of the 235 participants, 19 were lost to follow‐up, resulting in an overall loss rate of 8.09%. Pathogenic variants were detected using karyotyping, CNV‐seq, and WES in 78 patients. Among these, one was lost to follow‐up, four resulted in live births, and 68 resulted in pregnancy termination, corresponding to a termination rate of 93.15% (69/73) among cases with detected abnormalities. Among the 162 cases with no abnormalities detected by any of the three methods, 18 were lost to follow‐up, 67 resulted in live births, and 77 resulted in termination, corresponding to a termination rate of 47.53% (77/162). The presence or absence of detected genetic abnormalities significantly affected fetal survival (p < 0.001).

4. Discussion

Family history of CHD, maternal diseases, advanced maternal age, and maternal obesity have all been associated with an increased risk of fetal CHD. Maternal age was considered an independent risk factor with considerable clinical significance [18, 19]. Although no significant difference in maternal age was observed among the three groups in our study (p = 0.221), the age level in our cohort remained relatively high compared with the broader population of pregnant women in Xiamen, where the majority were concentrated in the 25–29‐year age group [20]. The mean gestational age in group C was significantly lower than that in groups A and B, which may be attributed to the more complex manifestations of non‐isolated CHD. Such cases often present with severe abnormalities at an earlier gestational age, prompting a greater tendency for pregnancy termination.

In our cohort, 16 patients with CHD presented with heterotaxy syndrome, with a notably higher incidence than that reported in previous large‐scale studies [21]. Among all patients with heterotaxy syndrome, including 10 with extracardiac anomalies involving organ heterotaxy, no chromosomal numerical abnormalities were detected. CNVs are recognized as an important genetic cause of heterotaxy syndrome [22, 23]; however, the detection rate of pathogenic CNVs is extremely low, and up to 90% of cases remain genetically unexplained [24, 25]. In our study, only one patient carried the del 15q11.2q11.2. The genetic etiology of heterotaxy syndrome is highly heterogeneous, involving not only CNVs but also SNVs, oligogenic combinations, and environmental factors [26]. Among these cases, only six underwent WES, and four were found to carry P/LP variants, with two genes clearly associated with heterotaxy syndrome. In these two cases, both cardiac and extracardiac manifestations included organ heterotaxy. Mutations in the NPHP3 gene are known to cause a range of early embryonic patterning defects, including organ heterotaxy [27], whereas mutations in CEP290 disrupt normal ciliary function, interfering with the establishment of the left–right body axis during early embryogenesis [28]. Another case presented with only intracardiac abnormalities, specifically left atrial isomerism, and was found to carry a variant in the MPDZ gene. To date, no study has directly linked MPDZ to heterotaxy syndromes. However, in an animal model, Mpdz knockout mice exhibit multisystem abnormalities [29], highlighting the important role of MPDZ in development. Since the MPDZ protein can bind to DAPLE, a protein involved in left–right axis defects, and cooperate to maintain normal cell polarity and development [30], we hypothesized that MPDZ may influence left–right body axis development through the same pathway, potentially serving as a susceptibility locus for heterotaxy syndrome.

In our study, the detection rate of chromosomal abnormalities by karyotyping was 19.57% (46/235), whereas the overall detection rate by CNV was 22.98% (54/235). Although CNV analysis cannot detect low‐level mosaicism or balanced rearrangements such as inversions and translocations, its unique advantages enabled the identification of pathogenic variants in 12 cases that were negative on karyotyping, resulting in an increase of 5.11% (12/235) in the combined detection rate. The detection rate of chromosomal abnormalities was significantly higher in group C, whereas no significant difference was observed between groups A and B, suggesting that the severity of intracardiac abnormalities was not significantly correlated with the incidence of chromosomal abnormalities [31]. Among the pathogenic variants detected by CNV‐seq, aneuploidy accounted for 39 cases (16.60%) and P/LP CNVs for 19 cases (8.09%), which is generally consistent with the report by Lu et al. [8]. Most current studies have focused solely on the utility of CMA/CNV‐seq in patients with CHD; however, karyotyping retains unique value and complements CNV analysis. In our study, one patient in group B with a karyotype of 46,XX,inv.(6)(p21.3p23) presented with intracardiac structural abnormalities. Subsequent CNV‐seq and WES analyses revealed no pathogenic variants, and the pregnancy was ultimately terminated. The inversion of chromosome 6 is a rare karyotypic finding, and its association with CHD has rarely been reported. To date, only one study has reported the same karyotype, inv.(6) (p21.3p23), in a family in which multiple members presented with VSD complicated by pulmonary arterial hypertension [32]. Wessels et al. identified a linkage between functional candidate genes associated with left ventricular non‐compaction, pulmonary valve stenosis, and left ventricular isomerism and the chromosomal region 6p24.3–24.4 in a family study [33]. Xie et al. identified a 6p21.33 duplication (TNXB) as a potential pathogenic variant in patients with pulmonary atresia and VSD [34]. Our patient presented with pulmonary atresia. Further studies are needed to clarify whether such genetic variants are associated with this phenotype. This case highlights that conventional karyotyping remains essential for identifying balanced chromosomal rearrangements, which are invisible to CNV‐seq.

Over the past few decades, several CNVs associated with CHD have been identified [35, 36]. In both our study and other reports, the 22q11.2 region represents the most common CNV in fetuses with CHD [8, 37, 38]. The most frequent cardiac malformations associated with 22q11.2 deletion syndrome are CTD, particularly TOF [39], and IAA [40], a pattern that is also reflected in our findings. Although many studies have described extracardiac phenotypes prenatally associated with the 22q11.2 region [41, 42], none of the five cases in our study presented with extracardiac anomalies, which may be attributable to the small sample size. The most common prenatal ultrasound structural abnormality in fetuses with 15q11.2 deletion is CHD [43, 44], and in our study, 15q11.2 deletion was the second most frequently detected CNV. Notably, the 15q11.2 microdeletion is considered a susceptibility factor for CHD with incomplete penetrance rather than a definitive pathogenic variant [45]. In the three cases reported here, the prenatal presentation of extracardiac findings was consistent with the phenotypic spectrum of 15q11.2 microdeletion, leading to their classification as pathogenic variants. Their prenatal phenotypes range from VSD and pulmonary stenosis to heterotaxy syndrome, reflecting the phenotypic diversity of this microdeletion [45, 46, 47].

In this study, one recurrent pathogenic gene, NOTCH1, was identified. The overall detection rate of NOTCH1 mutations in CHD is not high; however, they are frequently observed in patients with TOF and other CTD affecting the cardiac outflow tract [48]. Among the 44 patients who underwent WES in our study, two were found to carry NOTCH1 mutations, yielding a detection rate higher than the general prevalence (4.45% vs. 1%) [49], which may be associated with the high proportion of CTD (29.79%) in our cohort. NOTCH1 is a well‐established causative gene of aortic valve disease type 1 (AOVD1) [50]. Both affected patients presented with aortic atresia and hypoplasia of the ascending aorta and aortic arch, ultimately resulting in hypoplastic left‐sided heart syndrome. NOTCH1 mutations result in incomplete penetrance [51], and the type and severity of cardiac defects may vary considerably among family members. In case X113, the mutation was inherited from the father, who was suspected to have only mild cardiac abnormalities, whereas the proband had a stillbirth at 23 weeks of gestation. Notably, this patient also carried a duplication of approximately 100.07 Mb in the 1q21.2q44 region, which, despite being suggestive of pathogenicity, was difficult to definitively assess because of a mosaic ratio of 8%. Current reports on chromosome 1 primarily focus on the 1q21.1 region, with few reports on regions distal to 1q21.2. Studies have reported that microduplications involving 1q21.2 [52], 1q23.1q31.1 [53], and 1q32q44 [54] are associated with multisystem abnormalities, including cardiac malformations. In most studies, WES was performed only in cases where no pathogenic variants were detected by CNV and karyotyping. However, even in cases where pathogenic variants are identified by CNV‐seq or karyotyping, WES may yield additional pathogenic gene findings.

Among the four familial cases examined, three followed an autosomal dominant inheritance pattern. In case X113, the mutation was inherited from the father, who presented with only mild cardiac symptoms; in the other two cases (X116 and X168), the mutations were inherited from their respective mothers, both of whom were phenotypically normal. These observations can be explained by incomplete penetrance, a common feature of these gene mutations, which often leads to marked phenotypic variability among affected individuals. For case X150, the patient carried a PNKP gene mutation, a condition inherited in an AR pattern. The fetus inherited the mutation from the mother, while the father was wild‐type; both parents were phenotypically normal. This could be attributable to the father being a germline mosaic or having low‐level mosaicism rather than being truly wild‐type, or to the presence of a pathogenic variant undetectable by routine sequencing. Ultra‐deep sequencing would be required for definitive confirmation. For other cases of AR disorders in which only a single heterozygous pathogenic variant was identified, the genetic findings alone were insufficient to explain the clinical phenotypes. One plausible explanation for this observation is technical limitations, whereby the pathogenic variant may reside in regions not captured by WES, including deep intronic or regulatory sequences. WES demonstrated substantial incremental value over traditional methods in terms of variant identification. In the subgroup of fetuses with no abnormalities detected by both karyotyping and CNV‐seq who underwent WES testing, WES identified additional pathogenic variants, resulting in an increase in the detection rate of 37.21% (16/43) within this subgroup. However, given the potential selection bias, this rate may not be generalizable to all fetuses with negative findings on routine testing. Larger, unbiased studies are needed to accurately evaluate the added value of WES. CHD is characterized by high incidence, disability, and mortality rates, and poses a significant threat to maternal and infant health. Our results showed that fetal survival rates were highest in group A, followed by group B, and lowest in group C, with significant differences observed among all three groups. Moreover, a significant difference in survival rates was observed between patients with and without genetic abnormalities detected using any genetic testing method. These findings suggest that the severity of fetal malformations, the presence or absence of extracardiac abnormalities, and the detection of genetic anomalies are closely associated with fetal survival and serve as important factors influencing the decision to terminate a pregnancy. Therefore, accurate classification at the time of CHD diagnosis and prenatal identification of genetic abnormalities are crucial for providing more precise and appropriate counseling.

In the present study, the combined use of karyotyping, CNV‐seq, and WES significantly improved the detection rate of genetic abnormalities in patients with CHD. The highest aneuploidy detection rates were observed in AVSD and septal defects among intracardiac abnormalities, and in fetal hydrops, choroid plexus cysts, and cystic hygroma of the neck among extracardiac abnormalities. Chromosomal abnormalities, particularly numerical abnormalities, are significantly associated with extracardiac anomalies. In contrast, the detection of P/LP CNVs and pathogenic gene variants was not associated with the complexity of intracardiac abnormalities or the presence of extracardiac anomalies. Parental decisions regarding pregnancy continuation were significantly influenced by the severity of intracardiac abnormalities, the presence of extracardiac anomalies, and the identification of genetic abnormalities.

Author Contributions

Jieqiong Xie contributed to the conception and design, analysis, interpretation, and drafting of the initial manuscript. Yanru Huang, Wenbo Wang, and Jian Zhang performed the analysis and interpretation of CNV‐seq and WES data. Lutan Zhang interpreted the karyotyping results. Xingmei Yao and Yixi Zhou contributed to data acquisition and statistical analysis. Yunsheng Ge contributed to critical revision of the manuscript for important intellectual content, funding acquisition, administrative, technical, and material support, as well as supervision.

Funding

2024 Fujian Provincial Natural Science Foundation Jointly Funded Project (Grant No. 2024J011344).

Ethics Statement

The study was approved by the Ethics Committee of the Women and Children's Hospital, School of Medicine, Xiamen University.

Consent

Informed consent was obtained from all parents for invasive prenatal diagnosis.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data sets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.