Genetic Analysis of Isolated Fetal Growth Restriction

1. Introduction

Fetal Growth Restriction (FGR) is a common obstetric complication characterized by fetal growth below its genetic growth potential, typically indicated by an Estimated Fetal Weight (EFW) or abdominal circumference below the 10th percentile for gestational age[1,2,3]. FGR poses significant risks for perinatal morbidity and mortality, leading to short-term issues such as fetal distress and stillbirth, as well as long-term developmental delay and adult diseases like metabolic syndrome[4,5,6,7,8]. The etiology of FGR is complex, involving factors from the mother, placenta, umbilical cord, and fetus[5,9,10,11]. Genetic abnormalities play a crucial role in FGR occurrence and prognosis, with 15%-20% of cases linked to chromosomal issues such as aneuploidy, copy number variations (CNVs), and epigenetic changes [11,12,13,14,15].Common causes include aneuploidy such as trisomy 21, trisomy 13, and trisomy 18, where about 90% of affected fetuses show FGR[16].

Advances in genetic diagnostic techniques have improved the detection of chromosomal abnormalities in FGR. Traditional karyotype analysis detects chromosomal aneuploidy and large structural variations with low resolution for small imbalances. Array comparative genomic hybridization (aCGH) and single-nucleotide polymorphism (SNP) arrays enhance the detection of chromosomal abnormalities, with SNP arrays identifying CNVs and uniparental disomy (UPD). These technologies significantly increase detection rates compared to traditional karyotyping [16,17,18,19,20,21,22,23]. The American College of Obstetricians and Gynecologists (ACOG) and the Society of Maternal-Fetal Medicine (SMFM) recommend chromosomal microarray analysis (CMA) as a primary prenatal diagnostic tool for fetal abnormalities [20,24]. However, despite progress in detecting chromosomal abnormalities with CMA, its diagnostic rate for isolated FGR varies from 0% to 26.3%, averaging 6.4%, highlighting the need for further genetic exploration [15,25,26,27,28]. Whole exome sequencing (WES) has emerged as a valuable tool for diagnosing single-gene diseases by identifying single nucleotide variants (SNVs) and insertions/deletions (indels), potentially increasing diagnostic rates for CMA-negative fetuses with structural abnormalities by 10%-31% [29,30]. Emerging evidence suggests that WES can identify genetic abnormalities in 10% to 15% of cases with isolated FGR[31,32]. However, research elucidating the genetic etiology of isolated FGR in the absence of structural anomalies remains limited. This gap in knowledge complicates clinical genetic counseling and informed pregnancy decision-making.

This study aims to analyze the genetic causes of isolated FGR fetuses using G-banded karyotype analysis, CMA and WES, exploring the clinical application value of these techniques in isolated FGR fetuses, providing a basis for optimizing the genetic diagnosis pathway for isolated FGR, and offering more targeted and effective guidance for clinical practice.

2. Materials and Methods

2.1. Patient Enrollment

A retrospective analysis was conducted on 153 fetuses following invasive prenatal diagnosis because of isolated FGR diagnosed by ultrasound at the Prenatal Diagnosis Center of the University of Hong Kong - Shenzhen Hospital from February 2018 to July 2024 (Figure 1). In this period, a total of 3655 invasive prenatal diagnoses were performed in our center, among which 166 cases were diagnosed with FGR, and 153 cases of isolated FGR were included in this study. All cases met the following criteria: (1) Diagnostic criteria for FGR: According to the standards for Asian populations set by the National Institute of Child Health and Human Development (NICHD), after correcting gestational age with early pregnancy ultrasound, with ultrasound measurements of EFW or abdominal circumference less than 10th percentile for fetuses of the same gestational age; (2) No fetal structural abnormalities on ultrasound; (3) No severe maternal medical conditions such as chronic hypertension or insulin dependent diabetes mellitus; (4) Singleton pregnancy. All participants signed informed consent, and this study was approved by the Ethics Committee of the University of Hong Kong-Shenzhen Hospital ( hkuszh2025258).

All cases underwent amniocentesis after genetic counseling, with chromosome karyotyping and CMA testing performed. For 141 couples with results negative for karyotype and CMA, after genetic counseling, 50 chose to proceed with additional trio-WES testing, and couples’ peripheral blood samples were collected.

2.2. Aminotic Fluid Sampling and DNA Extraction

Under ultrasound guidance, amniocentesis was performed to extract 40ml of amniotic fluid. Of this, 10ml was used for CMA testing, Methylation-Specific Multiplex Ligation-dependent Probe Amplification (MS-MLPA) and maternal cell contamination testing, 20ml was equally divided into 2 independent cultures for standard G-banding karyotyping analysis, and the remaining 10ml was stored for further genetic testing (for instance, Trio-WES testing).

Peripheral blood (PB) and amniocytes DNA was extracted using the QIAamp DSP DNA Blood Mini Kit and Puregene cell kit (QIAGEN, Hilden, Germany) respectively, from 200ul of PB and 10ml of amniotic fluid, according to standard protocol (www.qiagen.com), and 50ul of DNA was eluted in TE buffer.

Maternal cell contamination was investigated using polymerase chain reaction (PCR)-based short tandem repeat (STR) analysis. Cultured amniocytes were used for downstream analysis if maternal contamination was detected.

2.3. CMA and Trio-WES

CMA was sent out to Shenzhen Maternity and Child Healthcare Hospital before 2019 using the CytoScan 750K chip kit (Affymetrix, USA) containing CNV and SNPs probes. From 2019 onward CMA was performed at HKU-SZH prenatal diagnostic laboratory using the SurePrint G3 Human CGH Microarray 8x60K (Agilent, USA) with/without ME034 MS-MLPA (MRC Holland, Amsterdam, The Netherlands). Data analysis was performed using CHAS software version. 4.2.80 (Affymetrix, USA) or the CytoGenomics Software v5.4 (Agilent, USA) in conjunction with relevant databases, including DGV, Decipher, UCSC and ClinGen databases.

Samples for WES testing were sent to the commercial laboratory BGI. The DNA library was sequenced on the MGISEQ-2000 platform with average sequencing depth >300X, the average sequencing depth of >20X was 98%. The analysis was conducted using an internal bioinformatics pipeline, which included aligning reads to the GRCh37/hg19 genome assembly, variant calling, annotation, and comprehensive variant filtering. The annotation databases primarily included: gnomAD (http://gnomad.broadinstitute.org/); the 1000 Genomes Project (http://browser.1000genomes.org); BGI’s database: dbSNP (http://www.ncbi.nlm.nih.gov/snp), SIFT (http://sift.jcvi.org), MutationAssessor (http://mutationassessor.org), HPO (https://hpo.jax.org/app), OMIM, ClinVar databases etc. Variants were classified into five categories according to the ACMG guidelines[33]for the interpretation of genetic variants: "pathogenic (P)", "likely pathogenic (LP)", "benign (B)", "likely benign (LB)", and "variant of uncertain significance (VUS)".

3. Results

3.1. Chromosomal Karyotype and Chromosomal Microarray Analysis Results.

A total of 3 abnormalities were detected in the chromosomal karyotype analysis (Table 1), with a detection rate of 2.0% (3/153). Among them, 1 case was a Turner syndrome mosaic (6%), 1 case was trisomy 21, and 1 case had a partial deletion on chromosome 11.

A total of 12 pathogenic or likely pathogenic copy number variations (CNVs) and uniparental disomy (UPD) were detected by CMA (Table 2), with a detection rate of 7.8% (12/153). These included 3 cases with chromosomal karyotype abnormalities, and the results of CMA were consistent with the karyotype results (Table 1). Among the 150 fetuses with normal karyotypes, CMA detected 9 cases of chromosomal abnormalities, including 6 cases of de novo chromosomal microdeletions (cases 5,6, 8,10,11,12), 1 case of 17p12 microduplication inherited from mother (case 7), and 2 cases of UPD (cases 4,9), with a detection rate of 6.0% (9/150). These chromosomal abnormalities involved 1q43-q44 microdeletion syndrome (case 6), Xp22.33p22.31 microdeletion associated with Leri-Weill dyschondrosteosis (case 8), 15q13.3 microdeletion syndrome (case 10), microdeletion of 19q13.11q13.12 (case 11), UPD(6) (case 9), segmental UPD(16) (case 4), microdeletion of 14q12 (case 5), 17p12 microduplication associated with Charcot-Marie-Tooth disease type 1A（CMT1A）(case 7), and 15q11.2 microdeletion syndrome (case 12).

3.2. Results of Trio-WES

In 50 cases with negative karyotype and CMA test results, pathogenic or likely pathogenic variants were identified in 12 cases (12/50, 24%) (Table 3). There were 10 pathogenic variants and 4 likely pathogenic variants found in 13 genes, and 1 case of UPD(6) which was not detected in the CMA (case 20). Among these, 7 (7/50,14%) harbored variants directly causative of FGR, including 5 cases of autosomal dominant diseases with mutations in the WHSC1, FGFR3, IHH, TUBB2B, and BRPF1 genes (cases 13,14,16,17, and 18), 1 case of autosomal recessive disease with a compound heterozygous mutation in the LIG4 gene (case 15) and 1 case of UPD(6) (case 20). Additionally, trio-WES identified 7 pathogenic or likely pathogenic variants not directly related to the FGR phenotype, involving the SMAD6, FLNB, MSH2, BRIP1, ENPP1, FGFR2, and PAH genes. Notably, two cases exhibited co-detection of incidental findings and FGR causative variants. Case 16 has a heterozygous IHH:c.84C>A (p.Cys28*) and a heterozygous ENPP1:c.783C>G (p.Tyr261*). The characteristics of this fetus include a short femur, with head circumference and abdominal circumference within normal ranges, and an estimated weight below the 10th percentile for gestational age, leading to a diagnosis of FGR. The heterozygous IHH:c.84C>A (p.Cys28*) inherited from father is a likely pathogenic variant. The father of the fetus is 170 cm tall and exhibits clinical features of short fingers, consistent with the phenotype of IHH related Brachydactyly, type A1 (OMIM 112500), suggesting a correlation between the IHH:c.84C>A (p.Cys28*) heterozygous mutation and the feature of short long bones in the fetus. This coexisted with an heterozygous ENPP1:c.783C>G (p.Tyr261*) which is inherited from mother, and the mother exhibits keratosis pilaris on her legs, consistent with the clinical manifestations of ENPP1 gene related Cole disease; this variant is not associated with FGR and is an incidental finding. Case 20, maternal UPD(6), a confirmed genetic etiology of FGR, was accompanied by an incidental finding FGFR2 variant (c.1032G>A, p.Ala344=) unrelated to growth restriction but associated with craniosynostosis syndromes. Additionally, in case 19, a heterozygous PAH:c.1197A>T was detected which was pathogenic variant and inherited from mother, and it was found that the mother had pathogenic PAH:c.1197A>T and c.1238G>C compound heterozygous variants. Further testing of the mother's blood phenylalanine level showed 757.42 µmo/L (reference range: 20-120 µmol/L), leading to a diagnosis of hyperphenylalaninemia in the mother. The mother had a history of one early pregnancy loss, one termination of pregnancy due to fetal heart defect, and one full-term delivery, but the child had intellectual disabilities. Prior to this prenatal diagnosis, genetic testing of family members had not been conducted. Although the heterozygous PAH:c.1197A>T of this fetus is not the cause of FGR and is considered incidental finding, trio-WES revealed the cause of FGR in this pregnancy was maternal hyperphenylalaninemia. The previous three adverse pregnancy outcomes were likely related to the mother's undiagnosed and uncontrolled hyperphenylalaninemia, similar to the adverse pregnancy outcomes reported in previous literature for pregnancies in women with uncontrolled hyperphenylalaninemia[34,35,36,37,38,39].

3.3. Pregnancy Outcome

In three cases of karyotype abnormalities, Case 1 had mosaic Turner syndrome and delivered at 38 weeks and 4 days with a neonatal weight of 2,690 g. Case 2, with trisomy 21, led to pregnancy termination. Case 3 involved a partial deletion of chromosome 11, resulting in intrauterine fetal demise.

Among the additional 9 cases of chromosomal abnormalities detected by CMA, 5 chose to terminate the pregnancy. These cases include 1q43-q44 microdeletion syndrome (case 6), Xp22.33p22.31 microdeletion associated with Leri-Weill dyschondrosteosis (case 8), 15q13.3 microdeletion syndrome (case 10), 19q13.11q13.12 microdeletion (case 11), and UPD(6) (case 9). Case 5 with a microdeletion of 14q12 resulted in intrauterine fetal demise. Three cases continued the pregnancy: case 7 had a maternal 17p12 microduplication related to Charcot-Marie-Tooth disease, and delivered preterm at 34 weeks because of pre-eclampsia, with a birth weight of 1830 grams and hypospadias; case 12, with 15q11.2 microdeletion syndrome, was delivered at 39 weeks and 3 days with a birth weight of 2540 grams; case 4, with segmental UPD(16), was delivered at 37 weeks and 5 days with a birth weight of 1660 grams. All three live-born children had normal development during follow-up from 1 to 3 years of age.

Of the 7 WES-identified cases with pathogenic or likely pathogenic variants related to the FGR phenotype, 6 opted to terminate the pregnancy. These included cases with autosomal dominant genetic diseases caused by mutations in WHSC1, FGFR3, TUBB2B, and BRPF1 (cases 13, 14, 17, 18), an autosomal recessive genetic disease caused by compound heterozygous mutations in the LIG4 gene (case 15), and a case of maternal UPD(6) combined with a pathogenic mutation in the FGFR2 gene (case 20). Case 16, with a heterozygous IHH:c.84C>A(p.Cys28*) and a heterozygous ENPP1:c.783C>G (p.Tyr261*), continued the pregnancy and delivered at term with a birth weight of_2850 grams. Four cases with pathogenic or likely pathogenic variants not directly related to the FGR phenotype opted to continue the pregnancies. These cases involved variants in SMAD6, FLNB, MSH2, and BRIP1, and all resulted in full-term deliveries with birth weight ranging from 2330 gram to 2700 gram. All five live-born children had normal development during follow-up from 6 months to 2 years of age. Case 20, with a likely pathogenic variant in the FGFR2 gene and maternal UPD(6), led the couple to decide to terminate the pregnancy. Case 19, with maternal hyperphenylalaninemia, also terminated the pregnancy. Following the termination, Case 19 received comprehensive multidisciplinary care, including consultations with a nutritionist, pharmacist, genetic specialist, and obstetrician, to manage phenylalanine levels and prepare for future pregnancies. As of the last follow-up, the patient has not conceived again.

4. Discussion

The genetic causes of FGR are complex, involving chromosomal aneuploidy, CNVs, UPD, and single-gene mutations[40,41]. In this study, karyotype analysis identified three chromosomal abnormalities (2.0%, 3/153), highlighting large chromosomal abnormalities as a significant cause of FGR even in the absence of obvious structural anomalies. However, the aneuploidy rate in this study was lower than that reported in previous studies, which found detection rates of large chromosomal abnormalities ranging from 3.9% to 9.4%[9,11,15]. This discrepancy is likely attributable to the routine implementation of non-invasive prenatal testing (NIPT) in early pregnancy in the region, which effectively detects large chromosomal abnormalities and thereby reduces their prevalence in FGR cases.

CNV is an important genetic factor contributing to FGR. By altering gene dosage, CNVs can lead to gene dosage effects, affecting fetal growth and development. In this study, all karyotype abnormalities were detected by CMA. CMA increased diagnostic yield by 5.9%, detecting CNVs/UPD missed by karyotyping. This finding aligns with figures from literature reports[12,13,42,43]. Previous studies identified the Xp22.31 microdeletion[13,43] and the 17p12 duplication[12] in isolated FGR cases. Consistent with these earlier reports, this study identified one case of Xp22.31 microdeletion and one case of 17p12 microduplication. UPD refers to the situation where both homologous chromosomes come from the same parent, which may lead to abnormal expression of imprinted genes, thereby affecting fetal growth and development[44,45,46]. In this study, SNP-array detected one case of UPD(6) and one case of segmental UPD(16), in accordance with findings from previous literature[40,47,48]. Genetic counseling regarding pathogenic or likely pathogenic CNVs significantly influences couple's decisions regarding their pregnancies. For instance, in case 7, a 17p12 microduplication inherited from the mother and associated with Charcot-Marie-Tooth disease type 1A (CMT1A), and in case 12, a 15q11.2 microdeletion with low penetrance, detailed genetic counseling about the clinical implications and penetrance of these chromosomal variations[41,49] led the couples to decide to continue the pregnancies.

Trio-WES identified causal variants in 14% of CMA-negative cases, including UPD, which is undetectable by CMA. For example, case 20 was found to have maternal UPD(6) by trio-WES which was missed in CMA. In this case, oligo CMA failed to detect UPD(6), while MS-MLPA indicated an increased methylation signal in the 6q24.2 region with normal copy number, suggesting maternal UPD(6). It is worth noting that SNP array cannot detect heterozygous UPD (hetero UPD), which further underscores the limitations of CMA in identifying UPD. This finding is consistent with previous studies demonstrating the limitations of CMA in detecting UPD, with approximately one-third of all molecularly confirmed UPD cases being undetectable by CMA alone [45,50,51,52]. These findings highlight the value of trio-WES in identifying UPD events that may be missed by CMA. WES also identified 7 pathogenic or likely pathogenic variants that are not directly related to the FGR phenotype. Although these variants are not directly associated with FGR, they may have significant implications for the long-term health of the fetus. For instance, the UPD(6) identified in case 20 is considered a genetic cause of FGR, and the incidental finding of a heterozygous FGFR2:c.1032G>A(p.Ala344=) in the same case directly influenced the couple's pregnancy decision, because of FGFR2-related diseases. The MSH2 gene mutation detected in case 23 is associated with Lynch syndrome and may increase the risk of colorectal cancer for the fetus in adulthood. In case 19, the fetal heterozygous PAH:c.1197A>T(p.Val399=) is not the cause of FGR, but trio-WES revealed pathogenic PAH:c.1197A>T and c.1238G>C compound heterozygous variants in the mother, indicating maternal hyperphenylalaninemia, which is the genetic cause of FGR and previous adverse pregnancies. As highlighted in the literature[53], this case underscores the significance of WES in uncovering maternal genetic disorders that can impact fetal development. By identifying such disorders, WES facilitates a more comprehensive understanding of the genetic factors contributing to FGR and enables tailored management strategies to improve pregnancy outcomes. Thus, WES not only aids in the etiological diagnosis of FGR but also provides more comprehensive genetic counseling for the family.

This study highlights the critical role of CMA and WES in elucidating the genetic complexity of isolated FGR. Beyond the diagnostic superiority of CMA over karyotyping (5.9% higher yield) and the additive value of WES (which achieved a 14% diagnosis yield in CMA-negative cases), our findings include two cases of co-occurrence of causal FGR variants and incidental findings within the same individual, adding to the counseling complexity.

5. Conclusions

This study highlights the advantages of integrating CMA and WES into prenatal diagnostic protocols for isolated FGR. CMA demonstrates superior diagnostic capabilities compared to traditional karyotyping, with a notable increase in the detection rate of chromosomal abnormalities. WES enables the detection of single-gene disorders and enhances the identification of UPD that may be missed by CMA alone. The combined use of these genetic techniques provides a more comprehensive evaluation of the genetic etiology of isolated FGR. Moreover, WES can help identify potential maternal genetic disorders that may impact fetal development. By detecting these disorders, WES facilitates better management strategies, which can potentially improve outcomes in future pregnancies.

However, the identification of incidental findings through WES underscores the critical need for multidisciplinary genetic counseling to address both fetal growth concerns and potential long-term health implications, while alleviating the anxiety associated with incidental findings. Future research should prioritize cost-effectiveness analyses and the development of standardized guidelines for variant interpretation in prenatal settings to further optimize clinical practice and improve patient outcomes.

6. Limitation

The small cohort size (n=153) and the WES subgroup (n=50) may reduce the statistical power to identify rare genetic variants. All participants were recruited from one tertiary center, which may have introduced selection bias towards more severe phenotypes or particular demographic groups. The study was conducted exclusively on an Asian population, which restricts the applicability of the findings to other ethnic groups with different genetic backgrounds. The outcomes assessed were limited to pregnancy termination or the neonatal status at birth. The long-term developmental, metabolic, and oncologic risks associated with incidental findings have not been characterized.