Genetic Diagnosis and Clinical Features of Fetuses With Congenital Diaphragmatic Hernia

ABSTRACT

Objective

Congenital diaphragmatic hernia (CDH) is a rare abnormality with highly heterogeneous genetic causes. This study investigated chromosomal and monogenic abnormalities in fetal CDH patients and evaluated the efficacy of chromosomal microarray analysis (CMA) and whole‐exome sequencing (WES) for genetic diagnosis. The clinical features of the patients were also evaluated.

Methods

We evaluated the genetic and clinical data of 51 prenatally diagnosed fetuses with CDH. CMA was performed for every patient. If CMA did not yield diagnostic results, the samples were subjected to WES.

Results

Compared with fetuses with isolated CDH (n = 42), those with non‐isolated CDH (n = 9) presented a higher genetic diagnostic rate (22.2% vs. 2.4%). The overall diagnostic yield was 5.9%, comprising 3.9% from chromosomal microarray analysis (CMA) and an additional 2.0% from whole exome sequencing (WES). CMA identified (1) mosaic trisomy 18 in a patient with isolated CDH; and (2) 4q terminal deletion syndrome in a patient with non‐isolated CDH. WES identified a novel missense mutation, PLS3 c.1763A > G, associated with X‐linked CDH in a patient with non‐isolated CDH and a family history of recurrent CDH.

Conclusion

Genetic testing should be offered for all fetuses with CDH, regardless of whether the cases are isolated or non‐isolated. WES should be considered if CMA fails to provide a diagnostic result, particularly in patients with non‐isolated CDH and a family history of recurrent CDH.

Summary.

• What’s already known about this topic?

  • ◦The genetic cause of congenital diaphragmatic hernia is highly heterogeneous.
  • ◦Genetic testing has a relatively low yield in individuals with isolated diaphragmatic hernia.

• What does this study add?

  • ◦We demonstrate a diagnostic yield of 3.9% from chromosomal microarray analysis and an additional 2.0% from whole exome sequencing for fetuses with diaphragmatic hernia.
  • ◦We provide evidence supporting the recommendation that genetic testing should be offered for all fetuses with diaphragmatic hernia, regardless of whether the cases are isolated or non‐isolated.
  • ◦We identify a new association of diaphragmatic hernia with 4q terminal syndrome, and discovered a novel missense mutation in PLS3 associated with X‐linked diaphragmatic hernia.

1. Introduction

Congenital diaphragmatic hernia (CDH) is a developmental defect in the diaphragm resulting in herniation of the abdominal organs into the thoracic cavity. The estimated incidence of CDH is approximately 3 per 10,000 live births [1, 2]. Isolated CDH without other major abnormalities constitutes 50%–60% of all CDH cases, and complex CDH constitutes the remaining 40%–50% [3]. Complex CDH may be syndromic, associated with a chromosomal abnormality or monogenic syndrome, or nonsyndromic without a recognizable syndromic phenotype [3]. The genetic cause of CDH is highly heterogeneous. Chromosomal abnormalities associated with CDH include trisomies, microdeletion syndromes, and unbalanced translocations, with the most prevalent being trisomy 18, trisomy 21, trisomy 13, deletions on chromosome 15q26, deletions on chromosome 8p23, and Pallister–Killian syndrome [3]. Prevalent monogenic disorders associated with CDH include Cornelia de Lange syndrome, Coffin‐Siris syndrome, and Fryns syndrome [4]. Recently, a new congenital anomaly syndrome characterized by an X‐linked trait of CDH caused by mutations in PLS3 was proposed as diaphragmatic hernia‐5 (DIH5, MIM: 306950) [5].

The overall survival rate of all prenatally diagnosed CDH individuals is 50% [3]. Whether CDH is isolated or complex and the severity of the hernia assessed by ultrasound and magnetic resonance imaging are the main prenatal prognostic indicators [6]. Our hospital is a maternal–fetal center and tertiary medical institution, and an increasing number of pregnant women with fetal CDH have been referred to our hospital for integrated prenatal and postnatal management. Genetic testing is an important component of prenatal practice. Chromosomal microarray analysis (CMA) is used as the first‐line standard genetic test for fetal structural abnormalities [7, 8]. Whole‐exome sequencing (WES) has an average incremental detection rate of approximately 10% [9]. The incremental yield of WES varies depending on the abnormality, from 1.8% for isolated increased nuchal translucency to 15.4% for multisystem abnormality [10, 11]. Weller et al. recently reported a diagnostic rate of 10% (5/50) in fetuses and neonates with isolated CDH and 17% (3/18) in cases with non‐isolated CDH [12]. However, the clinical application of WES for fetal CDH remains insufficiently researched and utilized. Our study aimed to investigate chromosomal and monogenic abnormalities in fetal CDH patients and evaluate the efficacy of CMA and WES for genetic diagnosis. In addition, the clinical features of fetal CDH patients were evaluated.

2. Materials and Methods

2.1. Population and Setting

A total of 51 prenatally diagnosed fetuses were enrolled in genetic diagnosis at Peking Union Medical College Hospital (PUMCH) from December 2021 to May 2024. When pregnant women with fetal CDH were referred to PUMCH, amniocentesis for CMA was offered after completing pretest counseling and providing informed consent for invasive prenatal diagnosis. For cases where the gestational age was too late for prenatal diagnosis at the time of referral, umbilical cord blood (UCB) was obtained for CMA upon delivery. If CMA did not yield diagnostic results, the samples were subjected to WES. Blood samples were collected from the parents for trio analysis when indicated. All patients met the following inclusion criteria: (1) the diagnosis of CDH was confirmed by a qualified sonographer at PUMCH who had credentials in prenatal ultrasound diagnosis, and (2) parents were willing to consent to the genetic testing procedure. Cases where CMA or WES were unsuccessful were excluded (n = 1). Clinical and genetic data were collected.

2.2. Collection and Analysis of Clinical Features

Clinical features were ascertained by reviewing medical records and included the following information: maternal age, previous pregnancy history, gestational age at which CDH was diagnosed, side of CDH, severity of CDH according to the observed‐to‐expected lung‐to‐head ratio (O/E LHR) obtained on ultrasound [13], other ultrasound abnormalities, fetal sex and perinatal outcome. Fetuses without additional ultrasound abnormalities were classified as having isolated CDH, whereas those with other abnormalities were classified as having non‐isolated CDH. The perinatal outcomes included pregnancy termination, intrauterine death, neonatal death and survival. Survival was defined as infant survival at discharge after surgical repair. Continuous data are expressed as the means and standard deviations (SDs), and categorical data are expressed as frequencies and percentages. Fisher's exact test was used to compare the differences between isolated and non‐isolated CDH cases. One‐sided p values < 0.05 were considered statistically significant. SPSS 29.0 (IBM, Armonk, NY, USA) was used for the statistical analysis.

2.3. Genetic Testing

Pathogenic variants, likely pathogenic variants and variants of uncertain significance (VOUS) were recorded and assessed by the clinical review panel to identify diagnostic variants for CDH. The pathogenicity of the variants was classified according to the standards and guidelines of the American College of Medical Genetics and Genomics [14, 15, 16].

2.3.1. Chromosomal Microarray Analysis With Single‐Nucleotide Polymorphism Arrays

The genomic DNA of fetuses was extracted using the QIAamp DNA Blood Mini Kit (Qiagen, Valencia, CA, USA). The DNA was digested, ligated with adapters, amplified, purified, and labeled with biotin. The labeled DNA was subsequently hybridized to the Affymetrix CytoScan 750K Array (Affymetrix Inc., CA, USA). The arrays were washed with Affymetrix GeneChip Fluidics Station 450 and scanned with an Affymetrix GeneChip Scanner 3000 according to the manufacturer's protocol. CEL files obtained by scanning the arrays were analyzed with Chromosome Analysis Suite v3.1 software. The GRCh37 (hg19) genome was used for annotation. Copy number variations (CNVs) that were larger than 200 kb or affected more than 50 markers were considered.

2.3.2. Whole‐Exome Sequencing

Libraries were prepared from the genomic DNA samples using an Agilent SureSelect^XT Human All Exon V6 Kit (Agilent Technologies, Santa Clara, CA, USA) and subsequently sequenced on a HiSeq2500 platform (Illumina, San Diego, CA, USA) in 150‐base paired‐end mode following the manufacturer's recommended protocols. The average coverage depth was 100×, with over 98% of the target regions having a minimum coverage of 20×. Data analysis included variant calling and annotation. In brief, the raw data were filtered to generate "clean reads" by removing adapters and reads with a base quality below Q20 using Trimmomatic v0.39. The sequencing reads were then aligned to the reference human genome version hg19 using BWA (v.0.7.17‐r1188) with the MEM align method. Nucleotide changes observed in the aligned reads were called and reviewed using GATK4 software. Sequence variants were annotated using ANNOVAR according to in silico prediction tools, population allele frequency databases, and disease‒phenotype association databases, such as SIFT, MutationTaster, GnomAD, ClinVar, and OMIM. Variants with a minor allele frequency (MAF) greater than 1% in the population were excluded. Moreover, NxClinical (BioDiscovery, Hawthorne, CA) was used to detect SNVs/insertions and deletions (Indels) and CNVs in parallel. Candidate variants were selected after combining the clinical information of the proband, phenotype‒genotype association and above‐annotated information, with a focus on de novo, compound heterozygous, homozygous, hemizygous, and pathogenic/likely pathogenic variants.

2.3.3. Sanger Sequencing

When indicated, Sanger sequencing was employed to test samples from the parents to ascertain whether the variant was de novo or inherited. The amplicons were sequenced using an Applied Biosystems 3730XL Genetic Analyzer, and analyses were performed using the BioEdit Sequence Alignment Editor (Applied Biosystems, USA).

2.3.4. Protein Structure Modeling

Structural modeling was performed using SWISS MODEL to address the potential effects of the identified missense variants (https://swissmodel.expasy.org). The PDB file (PDB: 1WJO) was downloaded from the Protein Data Bank (PDB, https://www.rcsb.org/). The mutation prediction model was compared with the original model using PyMOL 3.0 (DeLano Scientific).

3. Results

3.1. Clinical Features

Table 1 shows the clinical features of the 51 fetuses with CDH in our cohort. Among them, 42 fetuses had isolated CDH, whereas 9 had non‐isolated CDH. The accompanying ultrasound abnormalities in fetuses with non‐isolated CDH are detailed in Table S1. Genetic testing detected diagnostic variants in three patients (3/51, 5.9%). Table 2 shows that, compared with fetuses with isolated CDH, those with non‐isolated CDH had a lower survival rate (25.0% vs. 65.9%). Although fetuses with non‐isolated CDH also had a seemingly higher rate of moderate to severe CDH (55.6% vs. 26.2%) and a seemingly higher diagnostic rate for genetic testing (22.2% vs. 2.4%), the difference was not statistically significant.

TABLE 1.

Clinical features of fetuses with CDH in our cohort.

Clinical features	Results
Maternal age, years	32.4 $\mathit{\pm }$ 4.8
Parity, n (%)	Primiparous: 30 (58.8%)
	Multiparous: 21 (41.2%)
Family history of recurrent CDH	Yes: 2 (3.9%)
	No: 49 (96.1%)
Gestational age at CDH diagnosis, weeks	23.0 $\mathit{\pm }$ 3.3
Side of CDH	Left: 38 (74.5)
	Right: 10 (19.6)
	Bilateral: 3 (5.9)
Severity of CDH	Mild: 35 (68.6%)
	Moderate: 10 (19.6%)
	Severe: 6 (11.8%)
Accompanying ultrasound abnormalities	Isolated CDH: 42 (82.4%)
	Non‐isolated CDH: 9 (17.6%)
Fetal sex	Male: 33 (64.7%)
	Female: 18 (35.3%)
Perinatal outcome	Survival: 29 (56.9%)
	Neonatal death: 19 (37.3%)
	Intrauterine demise: 1 (2.0%)
	Termination of pregnancy: 2 (3.9)

TABLE 2.

Comparison of clinical features between fetuses with non‐isolated and isolated CDH.

Clinical features	Non‐isolated CDH (%, n/N)	Isolated CDH (%, n/N)	p Value
Rate of moderate to severe CDH	55.6%, 5/9	26.2%, 11/42	0.095
Diagnostic rate of genetic testing	22.2%, 2/9	2.4%, 1/42	0.077
Survival rate a	25.0%, 2/8	65.9%, 27/41	0.040

Note: A p value < 0.05 was considered significant, and significant values are marked in bold.

^a Two terminated pregnancies were excluded from the analysis.

3.2. Diagnostic Variants Identified Through Genetic Testing

CMA identified diagnostic variants in two patients (3.9%, 2/51): one with mosaic trisomy 18 and the other with 4q terminal syndrome. WES detected a mutation in PLS3 associated with DIH5, indicating an incremental diagnostic rate of 2.0% (1/51). The clinical information of these three patients is shown in Table 3.

TABLE 3.

Clinical features of 3 patients with diagnostic genetic variants.

Clinical features	Case 11	Case 31	Case 47
Gestational week at diagnosis	17	34	24
Side of CDH	Left	Right	Left
Severity of CDH	Severe	Mild	Mild
Other ultrasound abnormalities	Bilateral adducted thumb	No	Micrognathia
Abnormalities detected postnatally	None	None	Facial dysmorphism, anal atresia
Fetal sex	Male	Female	Male
Result of genetic testing	PLS3 c.1763A > G	29% trisomy 18	4q32.1q35.2(161,417,526‐187,900,882) x1
Perinatal outcome	Neonatal death	Survival	Termination of pregnancy

3.2.1. Diagnostic Variants Detected Through CMA

Patient 31 was the 2nd child of her mother, who had a healthy 14‐year‐old daughter. Non‐invasive prenatal testing conducted during pregnancy yielded normal results. An isolated mild right‐sided CDH was detected at 34 weeks of gestation. A female infant was born at 37 weeks and underwent immediate surgery for CDH, which was uneventful. CMA of the UCB obtained upon delivery revealed 29% mosaicism for trisomy 18 (Figure S1). At the 6‐month follow‐up, a comprehensive evaluation of the baby revealed no other abnormalities, and her growth and development were comparable to those of her peers. Continuous monitoring will be maintained.

Patient 47 was her mother's first pregnancy. A non‐isolated mild left‐sided CDH with micrognathia was detected at 24 weeks of gestation. Amniocentesis was performed, and CMA detected a de novo 26.5 Mb deletion at 4q32.1q35.2 (hg19: 161,417,526‐187,900,882). The variant is classified as pathogenic for two reasons. First, the TENM3 gene within the region is predicted to be haploinsufficient (2H, score 0.15). Second, there are more than 35 protein‐coding RefSeq genes in the region (3C, score 0.90). The variant is within the region associated with 4q terminal deletion syndrome (Figure S2a). The parents opted to terminate the pregnancy. Autopsy examination confirmed the presence of CDH and facial dysmorphism, characterized by a depressed nasal bridge, micrognathia, low‐set ears, and hypertelorism (Figure S2b). Additionally, anal atresia and digital dysmorphism (overlapping fingers and toes) were identified (Figure S2c, d). To rule out the possibility of a monogenic cause of CDH, WES and Sanger sequencing were performed, and interestingly, a compound heterozygous VOUS, LONP1 c.2702C > T (maternal origin, PM2_supporting + PP3)/c.1532G > C (paternal origin, PM2_supporting), was detected. LONP1 is known to be associated with autosomal recessive CODAS (cerebral, ocular, dental, auricular, and skeletal anomalies) syndrome. CDH is not a specific phenotype, and the evidence for the pathogenicity of this variant is weak; therefore, a diagnosis of CODAS cannot be established. However, the accumulation of sequence variants in LONP1 has been reported to be associated with CDH. In addition to de novo variants, inherited heterozygous LONP1 variants have been observed in patients with more severe CDH in multiple unrelated families, whereas their parents carrying the same variant are asymptomatic [17]. In this case, the 4q terminal deletion is presumed to be the primary genetic cause, whereas the compound heterozygous variant in LONP1 might exacerbate the adverse genetic burden contributing to the pathogenesis of CDH.

The woman became pregnant again one year later. The prenatal CMA yielded normal results. A healthy neonate was delivered uneventfully.

3.2.2. Diagnostic Variants Detected Through WES

Patient 11 was the 4th child of his mother, who had previously given birth to a healthy 10‐year‐old daughter and two male babies with left‐sided CDH, both dying shortly after delivery (Figure 1a). No genetic testing had been conducted previously, and other phenotypic information was not available. Non‐isolated severe left‐sided CDH with a bilateral adducted thumb was detected at 17 weeks of gestation. A male infant was born at 35 weeks after premature membrane rupture and died one day after birth due to respiratory failure and persistent pulmonary hypertension. The parents declined an autopsy. WES of the UCB obtained upon delivery revealed a hemizygous PLS3 c.1763A > G (p.Y588C) missense mutation (Figure 1b). The variant was inherited from his unaffected mother. This variant is classified as a VOUS for two reasons. First, it is absent from large population databases (PM2_supporting). Second, multiple lines of computational evidence (REVEL_score: 0.94, MetaSVM_score: 1.08, and GERP++_RS: 5.59) predict that it has a deleterious effect on the gene product (PP3_strong). The variant was determined to be diagnostic for four reasons. First, the strong in silico algorithm supports its pathogenicity, with a REVEL score of $\mathit{\ge }$ 0.932 [16]. Second, the variant was inherited from the mother, who had a total of three male neonates who died from CDH. The family history is consistent with the inheritance pattern of X‐linked DIH5. Third, the Tyr588 residue is located within the calponin homology (CH) domain within the actin‐binding domain 2 (ABD2) (Figure 1c), which is the region where the DIH5 variant is typically found [5]. Fourth, protein structure modeling indicated that the variant creates specific alterations in ABD2 without altering the major conformation (Figure 1d). This finding is consistent with the pattern of changes induced by the DIH5 variant in the protein structure [5].

FIGURE 1.

Clinical and genetic information of Patient 11. (a) Pedigree of the family. (b) WES identified a c.1763C > T (p.Y588C) mutation in PLS3 (NM_005032.7). (c) Location of the variant. The cDNA image shows that PLS3 has 16 exons. The protein image shows that plastin‐3 contains an N‐terminal calcium‐binding domain that is composed of two EF‐hand domains and two C‐terminal ABDs that are each composed of two CH domains. (d) Predicted structures of the actin‐binding domain of the PLS3 protein. The predicted structures of p.Y588C variant plastin‐3 proteins (blue) are overlaid on the structure of the native plastin‐3 (lavender). The tyrosine (green) is replaced by a cysteine (red). Abbreviations: ABD, actin‐binding domain, CH; calponin‐homology, WES; whole exome sequencing.

Putative three‐dimensional models of wild‐type PLS3 and the prebound model of mutant PLS3 were constructed to interpret the association between structural changes in the protein and the pathogenesis of CDH (Figure 2) [18]. As previously reported [19, 20], the prebound PLS3 adopts a closed horseshoe conformation with the C‐terminal CH4 in close contact with the N‐terminal CH1. ABD1 and ABD2 adopt a quasi‐antiparallel orientation. Residue Y588 is located in CH4 of the ABD2 domain and is precisely located at the interface between the CH1 and CH4 motif (Figure 2a). As a result, Y588 contributes to the CH1‐CH4 contact and is thus important for the interaction between ABD1 and ABD2. Therefore, substituting cysteine for tyrosine will affect the hydrophobic properties of this position (Figure 2b) and influence the pre‐bundling state of PLS3. The missense variation at M592, which is in close proximity to Y588, is also reportedly associated with CDH, underscoring the critical role of this region in sustaining the normal function of PLS3 [5]. In general, the Y588C mutation may cause disease by affecting the interaction interface between CH1 and CH4, thereby impacting the stability of the protein in the prebound state.

FIGURE 2.

Putative three‐dimensional models of wild‐type and mutant PLS3. (a) Wild type PLS3. Residue Y588 is located in CH4 of the ABD2 domain and is precisely located at the interface between the CH1 and CH4 motif. (b) Mutant PLS3 model A is templated on the AlphaFold‐v2 predicted structure (T‐Plastin monomer, pre‐bundling state). Substituting cysteine for tyrosine will affect the hydrophobic properties of this position. Abbreviations: ABD, actin‐binding domain, CH; calponin‐homology.

4. Discussion

This retrospective, single‐center study presents the data of a Chinese cohort that encompasses both phenotypic and genotypic features of fetal CDH. Our study demonstrated that fetuses with non‐isolated CDH exhibited greater severity, a lower survival rate, and a greater genetic detection rate than those with isolated CDH. The diagnostic rate of genetic testing in our cohort was 5.9%. CMA identified (1) mosaic trisomy 18 in a patient with isolated CDH; and (2) 4q terminal deletion syndrome in a patient with non‐isolated CDH. WES additionally detected a novel missense mutation in PLS3 associated with DIH5 in a patient with non‐isolated CDH and a family history of recurrent CDH.

It was previously believed that genetic testing has a relatively low yield in individuals with isolated CDH and a negative family history. In our study, two of three patients with a diagnostic genetic result had a negative family history. Furthermore, although two patients with the diagnostic variant presented with non‐isolated CDH prenatally, the accompanying ultrasound abnormalities were physical dysmorphism without internal organ involvement. Given the challenge of detecting physical dysmorphism on prenatal imaging and the potential for some phenotypes to manifest later in life, it is imperative to offer genetic testing for all fetuses with CDH, regardless of whether it is isolated or non‐isolated. The incremental diagnostic rate of WES in cases with negative CMA was 2.0% (1/51), which is comparable to the detection rate (2.7%–3.1%) of prenatal WES in structurally normal fetuses [21, 22]. However, it increases to 11.1% (1/9) for non‐isolated CDH patients. Therefore, WES is recommended for prenatal diagnosis of fetal CDH when CMA fails to provide a diagnostic result, particularly in cases with non‐isolated CDH and a family history of recurrent CDH.

In our study, WES identified a new variant in PLS3. PLS3 encodes an actin‐binding protein, plastin‐3, which plays an important role in multiple cellular processes. Plastin‐3 contains an N‐terminal calcium‐binding domain and two C‐terminal ABDs. Loss‐of‐function variants and deletions in PLS3 are known to be responsible for X‐linked osteoporosis (MIM: 300910). Variants associated with osteoporosis can occur in any domain of plastin‐3 and cause major alterations in protein structure. In contrast, all reported variants associated with DIH5 are missense mutations located within the ABD. These variants only affect actin cytoskeletal dynamics without altering the overall protein conformation. In addition to exhibiting diaphragmatic features, DIH5 patients can also present with abdominal wall defects, facial dysmorphism, neurodevelopmental retardation, and abnormalities in multiple systems. The inheritance mode of DIH5 is X‐linked recessive, meaning that males are severely affected and most die in early childhood. In our study, a missense mutation in PLS3 was identified in a male fetus with non‐isolated CDH and a family history of recurrent CDH. Unfortunately, genotypic information was not available for the previous two male neonates. Further functional experiments are warranted to validate the mechanism of this new variant.

Terminal deletion of chromosome 4 is a rare structural chromosomal abnormality caused by a partial interstitial or terminal deletion of the long arm of chromosome 4 [23]. Patient 47 had a 26.5 Mb interstitial deletion on 4q32.1‐q35.2, retaining 3.3 Mb from q35.2 to 4qter, a region free of OMIM‐morbid genes. The clinical features of this syndrome include developmental delay, craniofacial dysmorphism, digital dysmorphism, and abnormalities in multiple systems [23, 24]. The facial and digital dysmorphism in Patient 47 is consistent with the typical phenotype of 4q terminal deletion syndrome. The manifestation and severity of the syndrome vary greatly, depending on the size of the deletion, the location of the chromosome breakpoint, and the genes involved [25]. There are sporadic reports of the 4q terminal deletion being involved in CDH [26, 27]. We reviewed the available original publications and summarized their features in Table S2. Excluding cases with complex chromosomal aberrations such as ring chromosomes and unbalanced translocations, all documented cases of isolated 4q terminal deletion primarily involve 4q31, with the affected region appearing larger than that in our study [28, 29, 30, 31, 32]. It is hypothesized that an accumulation of de novo sequence variants may also contribute to CDH [33, 34]. LONP1 has been suggested as one of the candidate risk genes. Further investigations into the role of LONP1 variants in CDH pathogenesis are warranted to test this hypothesis.

Trisomy 18 is the most prevalent aneuploidy associated with CDH. However, Patient 31 did not present with syndromic manifestations other than CDH. There might be two reasons for this. First, the mosaic level of trisomy 18 was low and might vary across the tissues. One prenatally diagnosed mosaic trisomy 18 patient with 70% mosaicism in peripheral blood and 6.4% mosaicism in buccal mucosal cells was reported to be phenotypically normal at 2.5 months of age [35]. Second, some phenotypes might manifest later in life.

There were several limitations to our study. First, our cohort had a relatively small sample size. The nonsignificant differences in severity and genetic detection rates between patients with non‐isolated and isolated CDH might be attributed to the small sample size. Second, the genotype‒phenotype correlation analysis was based mainly on clinical and genetic data. Unavailability of animal or functional experiments rendered the interpretation of the variant.

5. Conclusion

CDH is a rare abnormality with highly heterogeneous genetic causes. Our study revealed an overall diagnostic yield of prenatal genetic testing of 5.9%, comprising 3.9% from CMA and an additional 2.0% from WES. We identified an association between 4q terminal syndrome and CDH and detected a novel variant in PLS3 associated with DIH5. Both findings warrant further investigation for validation. On the basis of our findings, we recommend that genetic testing should be offered for all fetuses with CDH, regardless of whether the cases are isolated or non‐isolated. WES should be considered if CMA fails to provide a diagnostic result, particularly in patients with non‐isolated CDH and a family history of recurrent CDH. More data are warranted to further investigate the incremental diagnostic yield of WES in CDH.

Ethics Statement

This study adheres to the principles set out in the Declaration of Helsinki. Written informed consent was obtained from all participants, and the study was approved by the Institutional Review Board of PUMCH (IRB no. I‐24PJ1052).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting Information S1

Figure S1

Figure S2

Data Availability Statement

The datasets generated and/or analyzed during the current study are not publicly available in order to comply with hospital and IRB policy. According to the consent form, sequencing data cannot be accessed without patient's permission. But they are available from the corresponding author on reasonable request.