Exome sequencing in every pregnancy? Results of trio exome sequencing in structurally normal fetuses

Abstract

Objective

This study aimed to assess the detection rate of clinically significant results of prenatal exome sequencing (pES) in low‐risk pregnancies and apparently normal fetuses in non‐consanguineous couples.

Methods

A retrospective analysis of pES conducted at a single center from January 2020 to September 2023 was performed. Genetic counseling was provided, and detailed medical histories were obtained. High‐risk pregnancies were excluded due to major ultrasound anomalies, sonographic soft markers, abnormal maternal biochemical screening, or family history suggestive of monogenic diseases as well as cases with pathogenic and likely pathogenic (P/LP) chromosomal microarray results. Exome analysis focused on ∼2100 genes associated with Mendelian genetic disorders. Variant analysis and classification followed the American College of Medical Genetics and Genomics (ACMG) guidelines.

Results

Among 1825 pES conducted, 1020 low‐risk cases revealed 28 fetuses (2.7%) with potentially clinically significant variants indicating known monogenic diseases, primarily de novo dominant variants (64%). Among these 28 cases, 9 fetuses (0.9%) had the potential for severe phenotypes, including shortened lifespan and intellectual disability, and another 12 had the potential for milder phenotypes. Seven cases were reported with variants of uncertain significance (VUS) that, according to the ACMG criteria, leaned toward LP, constituting 0.7% of the entire cohort. Termination of pregnancy was elected in 13 out of 1020 cases (1.2%) in the cohort, including 7/9 in the severe phenotypes group, 2/12 in the milder phenotype group, and 4/7 in the VUS group.

Conclusion

The 2.7% detection rate highlights the significant contribution of pES in low‐risk pregnancies. However, it necessitates rigorous analysis, and comprehensive genetic counseling before and after testing.

Key points

What is already known about this topic?

• Prenatal Exome Sequencing (pES) Trios is known for its diagnostic yield, ranging from 8% to 47%, in cases of fetal structural anomalies detected by ultrasonography.

What does this study add?

• In 1020 low‐risk pES cases, 2.7% had potentially clinically significant variants (64% being dominant de‐novo) indicating monogenic diseases. Among these, 0.9% had severe phenotypes, 0.7% variants of uncertain significance (leaning toward likely pathogenic), and 1.2% opted for termination.

1. INTRODUCTION

Prenatal molecular testing encompasses various techniques such as chromosomal microarray analysis (CMA), exome sequencing, and targeted genetic testing. These tests are typically conducted when there are indications of fetal abnormalities on ultrasound, a family history of monogenic genetic disorders, or abnormal maternal biochemical screening results.

CMA is the first‐tier test used in the prenatal setting. It is offered upon identification of a structural abnormality, increased nuchal translucency (NT), or soft marker for aneuploidy. Additionally, it is recommended in cases of advanced maternal age and may be administered in response to parental requests, even in the absence of apparent indications. ¹

It is firmly established that pathogenic copy number variants (CNVs) can be detected in fetuses with no apparent structural abnormalities. Daum et al. summarized six papers and a total of 29,612 prenatal CMAs performed in structurally normal fetuses. The prevalence of highly penetrant pathogenic and likely pathogenic (P/LP) CNVs ranged from 0.4% to 1.4%. Considerable heterogeneity existed across the studies with regard to the timing of CMA implementation as well as the categorization of specific types of CNVs as pathogenic. ²

The growing use of CMA in the prenatal testing, even when fetal sonograms show no abnormalities and pregnancies are considered low‐risk, is driven by the belief that early detection of genetic diagnosis allows couples to better prepare and make informed decisions about the current and future pregnancies. Many genetic disorders do not present visible anatomical irregularities during gestation, necessitating genetic testing for early detection. ³

Exome sequencing is a reliable method for identifying disease‐causing SNVs and small insertions or deletions and is frequently used in prenatal diagnostics as a trio analysis involving parents and the fetus. The application of trio analysis offers notable advantages in time‐sensitive clinical scenarios by furnishing inheritance and segregation data to facilitate precise variant classification.

Prenatal exome sequencing (pES) was introduced to the literature in 2017 ⁴ and has been clinically available since then. Until recently, the test was offered only for the diagnosis of fetuses with structural malformations identified through sonographic imaging or when family history suggested a monogenic disorder. ⁴ , ⁵ The diagnostic yield of pES in fetal structural anomalies detected by ultrasonography was reported to be 8%–47% depending on the inclusion criteria and the type of anomaly detected in the fetus. ⁴ , ⁶ , ⁷ However, data on the yield among fetuses without sonographic abnormalities is limited. Previous investigations by Daum et al. (482 cases) reported a diagnostic yield of 0.8%, while Vaknin et al. (160 cases) reported a diagnostic yield of 0.6% for significant findings. Despite these efforts, this area remains relatively unexplored. ⁸ , ⁹

Our study aimed to assess the utility of pES in detecting genetic abnormalities in low‐risk pregnancies without sonographic anomalies.

2. MATERIAL AND METHODS

This is a retrospective study. We reviewed the pretest genetic counseling and pregnancy tests of all pES conducted between January 2020 and September 2023 at the Maccabbi health services genetic institute.

Prior to amniocentesis, all patients underwent genetic counseling evaluation by a board‐certified medical geneticist, as is mandatory in Israel prior to fetal exome sequencing.

2.1. Study groups

The study included all pregnancies considered low risk for genetic diseases: non‐consanguineous couples, fetuses that were presented as normal by all tests performed by the time of amniocentesis and had normal CMA results.

As part of a nationwide screening program, complimentary carrier screening for diseases with a carrier frequency exceeding 1 in 100 is extended to all pregnant women in Israel. Approximately 85% of the individuals in our study cohort engaged in these screening assessments. In instances where both prospective parents were identified as carriers for the same disease, they were subsequently excluded from the analysis.

Participants provided an extensive family history and disclosed all results of previous genetic testing. Pregnancy tests in all cases included NT measurement between 11 and 13 weeks of gestation, followed by detailed early fetal anomaly scans at 14–16 weeks and late fetal anomaly scans at 20–24 weeks of gestation. Furthermore, first and second trimester (triple test) maternal biochemical screenings were conducted and meticulously documented.

Prior to amniocentesis, all patients underwent genetic counseling evaluation by a board‐certified medical geneticist.

The same fetal DNA that was extracted from amniocentesis for the CMA was used for the pES. DNA was also extracted from parental samples.

In cases involving variants of uncertain significance (VUSs), additional imaging, such as MRI, may be requested if the variant is predicted to be associated with a specific phenotype.

2.2. Exclusion criteria

Excluded were all fetuses with abnormal findings in sonographic examinations prior to pES: increased NT (>3 mm) at 11–13 weeks of gestational age, major fetal structural anomaly, intrauterine growth restriction, or polyhydramnios (AFI of ≥25 cm), minor ultrasound findings that mildly increased the risk for genetic disease, often considered as “soft markers” (e.g., mildly echogenic bowel, absent ductus venosus, cardiac echogenic focus, etc.), abnormality in first or second trimester biochemical maternal tests, those with any reported abnormality in the CMA, and those with family history of monogenic disease. In all cases, an anatomical scan was conducted, affirming the absence of any abnormalities. Fetuses with a family history of monogenic disease, including cases of both parents being carriers of the same autosomal recessive inherited disease, or when one parent has a known dominant or X linked monogenic disease.

All cases were analyzed as trios involving both parents and the fetus. Cases in which one or both parents were absent were excluded from the study.

2.3. Fetal exome sequencing trio

The fetal and parental DNA was sequenced on an Illumina NOVASEQ machine (Illumina) by the Maccabi health services genetic laboratory. We used targeted clinical pES, which included specific analysis of the data from a standard 100X ES.

RNA capture baits against approximately 60 Mb of the human exome (targeting >99% of regions in CCDS, RefSeq and Gencode databases) were used to enrich regions of interest from fragmented genomic DNA with Agilent's SureSelect Human All Exon V6 kit (Agilent Technologies). The generated library was sequenced on an Illumina platform to obtain an average coverage depth of ∼100×. Evaluation focused on coding exons along with flanking ± 20 intronic bases. All relevant inheritance patterns were considered.

2.4. Bioinformatics analysis

The analysis focused on a continuously updated list of genes (∼2100) that are associated with Mendelian genetic disorders (Supporting Information S1). The genes selected in the panel are known to cause Mendelian genetic disorders characterized by morbidity, lethality, neurodevelopmental disorders, or syndromes associated with structural anomalies leading to significant disabilities. The bioinformatic analysis was conducted using the FRANKLIN software platform developed by Genoox (https://franklin.genoox.com) (Genoox, 2023), supplemented with data from publicly available databases such as GnomAD (https://gnomad.broadinstitute.org/), ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/), and the PubMed database, which serves as a comprehensive repository of biomedical literature.

2.5. Variant classification and pES reports

This classification was in accordance with the current American College of Medical Genetics and Genomics (ACMG). ¹⁰ In accordance with the Clingen recommendation, we applied the PM2 criteria at a “support” level of significance rather than “moderate.” ¹¹ We did not align with the Clingen recommendation for PP3 and BP4. ¹²

We reported only potentially clinically significant variants to the couple (P/LP/Hot/warm VUSs). For the purpose of these studies, the potentially clinically significant cases were divided into three groups:

• Group 1 – De novo variants in genes causing severe phenotypes (Tier 1 according to Lazarin et al. ¹³ ).

• Group 2 – Variants in genes associated with milder phenotypes (not Tier 1 according to Lazarin et al. ¹³ ).

• Group 3 – Hot/warm VUSs—variants that, according to the ACMG criteria, lean toward LP in genes causing severe phenotypes (Tier 1 according to Lazarin et al. ¹³ ).

“Hot” VUSs are characterized by a substantial body of supporting evidence, and if further confirmation of their pathogenicity can be acquired, they might be reclassified as pathogenic. On the other hand, “warm” VUSs are supported by evidence suggestive of a specific diagnosis, but obtaining additional data to definitively substantiate reclassification may pose challenges.

2.6. Pre‐test and post‐test counseling

Pre‐test counseling for pES in our practice is conducted by a clinical geneticist.

The testing technology is clarified for the couples, highlighting that it may uncover changes with uncertain clinical significance, incomplete penetrance, or variable expression.

It is explained that there are instances in which different laboratories classify and interpret various genomic variants differently. Additionally, they receive structured remarks indicating that DNA testing might reveal alterations in inheritance patterns, which could have medical implications for the couple. VUSs are only reported if relevant for pregnancy management by a multidisciplinary team and may be reclassified as (likely) pathogenic variants over time, even during pregnancy. This issue is comprehensively addressed during pre‐test counseling, using easily understandable examples. Post‐test counseling is also conducted by a clinical geneticist, discussing results with couples. Options to continue or terminate the pregnancy are revisited, with explanations provided on the termination procedure and emotional implications, as needed.

2.7. Secondary findings

Secondary findings as outlined by the ACMG guidelines were only disclosed to parents after their explicit consent, and the option to report these secondary findings status regarding fetuses was not available. ¹⁴ , ¹⁵ Within the scope of this article, we have not expanded upon these secondary findings.

2.8. Turn‐around times

The turn‐around times (TAT) for the ES analyses were documented for each case, measured in calendar days from the date of the analysis request. Subsequently, the mean TAT was s calculated for the entire dataset to assess the average time required for the pES results to be generated.

2.9. Pregnancy outcome

In Israel, physicians are legally required to offer genetic tests to couples during pregnancy, regardless of specific indications. Moreover, there is no gestational age limit for pregnancy termination if the anticipated risk of a “severe medical condition” exceeds 30%. If a couple seeks to terminate the pregnancy for personal, medical, or genetic reasons, they can request any hospital to convene a committee of experts, typically including physicians and ethicists, to evaluate their request. However, the identification of genetic variants has added complexity to this process in recent years. Approval for termination by the committee only follows a thorough evaluation of disease risk. ¹⁶ , ¹⁷

We gathered data on pregnancy outcomes from all cases undergoing pES.

3. RESULTS

During the study period, a total of 1825 pES were conducted in our clinic following genetic counseling. Amniocentesis was performed during the gestational period ranging from 17 to 22 weeks.

After applying the exclusion criteria, 1020 pES cases in non‐consanguineous couples were included and 805 cases were excluded. The reasons for excluding 805 cases are given in Table S1.

The mean maternal age was 34.1 ± 4.1 years, and the mean paternal age was 38.3 ± 5.7 years.

No Uniparental disomy (UPD) events were identified.

UPD was ruled out by initially using CMA to detect loss of heterozygosity, followed by trio exome sequencing, which showed that the region did not originate from the same parent.

Potentially clinically significant variants were reported in 28 fetuses (2.7%) of the total cohort. Out of these 28 cases, 18 (64%) exhibited de novo dominant variants. Two cases presented compound heterozygous recessive variants, and two cases involved maternal X‐linked inheritance in a male fetus. Eight cases were associated with parental autosomal dominant (AD) inheritance of a variant of which the parents were presymptomatic.

The mean maternal and paternal ages among the 18 cases with de novo variants did not differ significantly from those observed in the overall cohort (maternal age: 34.1 ± 4.1 vs. 34.25 ± 3.4, p = 0.97; paternal age: 38.3 ± 5.7 vs. 38.9 ± 2.2, p = 0.65).

3.1. Group 1—De novo variant in genes causing severe phenotypes

Nine of the 28 fetuses (comprising 0.9% of the entire cohort) harbored pathogenic variants in genes associated with severe phenotypes, including shortened lifespan and intellectual disability, all of which were AD and found to be de novo.

Detailed information of the variants causing severe phenotypes among the clinically significant cases is presented in Table 1.

TABLE 1.

Detailed information of the de‐novo variant in genes causing severe phenotypes—Group 1.

Case no.	Gene	Condition/OMIM	Mode of inheritance	NM	HGVSc.	HGVSp.	ACMG classification	Classification criteria	Mode of inheritance	Zygosity	Pregnancy follow‐up
1	ADNP	#615873 Helsmoortel‐van der Aa syndrome; ADNP‐related multiple congenital anomalies—Intellectual disability—Autism spectrum disorder	AD	NM_001282531.3	c.1223dupA	p.Ser409fs	LP	PM2, PVS1	De‐novo	Heterozygous	Termination of pregnancy
2	SETD1A	#619056 neurodevelopmental disorders with speech impairment and dysmorphic facies	AD	NM_014712.3	c.757C>T	p.Arg253*	LP	PVS1, PM2	De‐novo	Heterozygous	Termination of pregnancy
3	PUM1	#617931 spinocerebellar ataxia 47	AD	NM_001020658.2	c.2173delA	p.Thr725fs	LP	PVS1, PM2	De‐novo	Heterozygous	Termination of pregnancy
4	ANKRD11	# 148050 KBG syndrome	AD	NM_013275.6	c.2408_2412 delAAAAA	p.Lys803fs	P	PVS1, PS4, PM2	De‐novo	Heterozygous	Termination of pregnancy
5	TSC2	#613254 tuberous sclerosis‐2	AD	NM_000548.5	c.3922_3923delGT	p.Val1308fs	LP	PVS1, PM2	De‐novo	Heterozygous	Termination of pregnancy
6	CUL3	#619239 neurodevelopmental disorder with or without autism or seizures	AD	NM_003590.5	c.1664dupG	p.Ser556fs	LP	PVS1, PM2	De‐novo	Heterozygous	Termination of pregnancy
7	SCN8A	# 614558 neurodevelopmental disorder and structural brain anomalies with or without seizures and spasticity	AD	NM_001330260.1	c.1089A>T	p.Leu363Phe	LP	PM1, PM2, PM5, PP2, PP3	De‐novo	Heterozygous	Termination of pregnancy
8	WAC	#616708 Desanto‐Shinawi syndrome	AD	NM_016628.5	c.1477_1480del	p.Ser493ProfsTer6	LP	PVS1, PM2	De‐novo	Heterozygous	N/A
9	KAT6A	#616268 Arboleda‐Tham syndrome	AD	NM_006766.5	c.4460_4463delinsGTG	p.Gln1487ArgfsTer46	LP	PVS1, PM2	De‐novo	Heterozygous	N/A

Note: In accordance with the Clingen recommendation, we applied the PM2 criteria at a “support” level of significance rather than “moderate.”

Abbreviations: ACMG, American College of Medical Genetics and Genomics; AD, autosomal dominant; N/A, not available; PM, pathogenic moderate; PP, pathogenic supporting; PS, pathogenic strong; PVS, pathogenic very strong; TOP, termination of pregnancy.

3.2. Group 2—Variants in genes associated with milder phenotypes

Twelve of the 28 fetuses (comprising 1.1% of the entire cohort) harbored pathogenic variants in genes associated with mild phenotypic presentations. These conditions necessitate careful prenatal and postnatal monitoring and intervention to mitigate potential complications. Among these cases, the inheritance pattern in 10 out of 12 cases was AD, one case (no. 11) was found to be compound heterozygous for pathogenic variants in the PAH gene, predicted to cause AR Hyperphenylalaninemia (OMIM #612349), and one was X‐linked recessive in a male fetus (no. 18). In four out of the 10 AD genes, the variants were identified as de novo.

Parental diagnoses were made in six of the 12 cases, comprising 0.6% of the entire patient cohort.

Detailed information of the variants causing mild phenotypes among the clinically significant cases is presented in Table 2.

TABLE 2.

Detailed information of the variants in genes associated with milder phenotypes—Group 2.

Case no.	Gene	Condition/OMIM	Mode of inheritance	NM	HGVSc.	HGVSp.	ACMG classification	Classification criteria	Mode of inheritance	Zygosity	Pregnancy follow‐up
10	EXT2	#133701 exostoses, multiple, type 2	AD	NM_207122.2	c.‐30‐2A>T	Splice acceptor	LP	PM2, PVS1	De‐novo	Heterozygous	Termination of pregnancy
11	PAH	#612349 hyperphenylalaninemia	AR	NM_000277.3	c.898G>T	p.Ala300Ser	P;	PS4, PS3, PM1, PM2;	Maternally and paternally inherited	Compound heterozygous	N/A
				NM_000277.3	c.1139C>T	p.Thr380Met	P	PS3, PM1, PP2, PM2, PP3, PP5
12	ENG	#187300 Telangiectasia, hereditary hemorrhagic, type 1 (HHT)	AD	NM_001114753.2	c.219G>A	p.Thr73Thr	LP	PS4, PM2, PP3	Maternally inherited	Heterozygous	Continuation of pregnancy
13	HARS	#616625 Charcot‐Marie‐Tooth disease, axonal, type 2W (highly variable age at onset [range childhood to late adult])	AD	NM_002109.6	c.464T>G	p.Val155Gly	LP	PS4, PM1, PM2, PP3	Maternally inherited	Heterozygous	N/A
14	ACVRL1	#600376 Telangiectasia, hereditary hemorrhagic, type 2	AD	NM_000020.3	c.1157G>A	p.Arg386His	LP	PS4_moderate, PM1, PM2, PP3	Maternally inherited	Heterozygous	Continuation of pregnancy
15	COL11A1	#154780 Stickler syndrome, type II	AD	NM_001854.4	c.1328dupC	p.Pro444fs	LP	PVS1, PM2	De‐novo	Heterozygous	Continuation of pregnancy
16	TSHR	#275200 hypothyroidism, congenital, nongoitrous, 1	AD	NM_000369.5	c.1349G>A	p.Arg450His	LP	PS4, PM1, PM2, PP3	Maternally inherited	Heterozygous	Continuation of pregnancy
17	FBN1	#154700 Marfan	AD	NM_000138.5	c.5660C>T	p.Thr1887Ile	LP	PS4_supp, PM1, PM2, PP2, PP3	Paternally inherited	Heterozygous	Continuation of pregnancy
18	CACNA1F	#300476 cone‐rod dystrophy, X‐linked	XL	NM_001256789.3	c.522‐13_526del	p.Leu175fs	LP	PVS1, PM2	Maternally inherited to a male fetus	Hemizygous	N/A
			Recessive		GCTCCCCCATCAGGCTGT
19	SOS1	# 610733 Noonan syndrome 4	AD	NM_005633.4	c.1310T>A	p.Ile437The	LP	PS4_supp, PM1, PM2, PM5, PP3	Paternally inherited	Heterozygous	N/A
20	COL4A1	#175780 brain small vessel disease with or without ocular anomalies	AD	NM_001845.6	c.359dupC	p.Gly121fs	LP	PVS1, PM2	De‐novo	Heterozygous	TOP, after IVH was documented in the 3rd trimester
21	NPR2	# 616255 short stature with nonspecific skeletal abnormalities	AD	NM_003995.3	c.2005C>T	p.Arg669*	LP	PVS1, PM2	De‐novo	Heterozygous	Continuation of pregnancy

Note: In accordance with the Clingen recommendation, we applied the PM2 criteria at a “support” level of significance rather than “moderate.”

Abbreviations: ACMG, American College of Medical Genetics and Genomics; AD, autosomal dominant; AR, autosomal recessive; IVH, intraventricular hemorrhage; N/A, not available; PM, pathogenic moderate; PP, pathogenic supporting; PS, pathogenic strong; PVS, pathogenic very strong; TOP, termination of pregnancy; XL, X linked.

3.3. Group 3—Hot/warm VUSs

Within our study cohort of 1020 cases, seven cases were reported with VUSs that, according to the ACMG criteria, lean toward LP, constituting 0.7% of the entire cohort. Notably, all reported VUSs are associated with genes linked to severe phenotypes, including conditions characterized by shortened lifespan and intellectual disability. Among these cases, five were AD de novo, one X‐linked dominant variant, maternally inherited to a male fetus, and one was compound heterozygous AR. Detailed information regarding these VUSs, which are linked to severe phenotypes, is detailed in Table 3.

TABLE 3.

Detailed information of the hot variants of unknown significance—Group 3.

Case no.	Gene	Condition/OMIM	Mode of inheritance	NM	HGVSc.	HGVSp.	Classification criteria	Mode of inheritance	Zygosity	Genetic counseling	Pregnancy follow‐up
22	ACTB	#243310 Baraitser‐Winter syndrome 1, dystonia, juvenile‐onset	AD	NM_001101.5	c.659C>T	p.Ala220Val	PM1, PM2, PP2, PP3	De‐novo	Heterozygous	Monitor signs of fluid accumulation, fetal movements, and corpus callosum development	Continuation of pregnancy
23	EFNB1	# 300035 Craniofrontonasal dysplasia	XL	NM_004429.5	c.466C>T	p.Arg156Cys	PS4_supp, PM1, PM2, PP3	Maternally inherited to a male fetus	Hemizygous	Developmental delays but without cognitive impairments. It is recommended to continue the pregnancy	Continuation of pregnancy
			Dominant
24	CHD4	# 617159 Sifrim‐Hitz‐Weiss syndrome	AD	NM_001273.5	c.2903C>T	p.Ser968Phe	PM1, PM2, PP2, PP3	De‐novo	Heterozygous	Continuously assess cerebral sonographic results and proceed with additional evaluation if any abnormalities are detected	TOP, 28w after ventriculomegaly on neuro‐sonography
25	KDM3B	Diets‐Jongmans syndrome	AD	NM_016604.4	c.2812C>T	p.Arg938Cys	PM1, PM2, PP2, PP3	De‐novo	Heterozygous	Severe condition not amenable to detection through sonographic imaging	Termination of pregnancy
26	SCN1A	# 607208 Dravet syndrome	AD	NM_001165963.4	c.451C>T	p.Pro151Ser	PM1, PM2, PP2, PP3	De‐novo	Heterozygous	Severe condition not amenable to detection through sonographic imaging	Termination of pregnancy
27	PTPN23	# 618890 neurodevelopmental disorder and structural brain anomalies with or without seizures and spasticity	AR	NM_015466.4	c.3697C>A c.3886_3888delAAG	p.Arg1233Ser; p.Lys1296del	PS4_mod, PM2, PM4;	Maternally and paternally inherited	Compound heterozygous	Continuously assess fetal growth, and cerebral sonographic results and proceed with additional evaluation if any abnormalities are detected	Termination of pregnancy
				NM_015466.4			PM2, PM3, PP3
28	DDX6	# 618653 intellectual developmental disorder with impaired language and dysmorphic facies	AD	NM_004397.6	c.1078T>G	p.Ser360Ala	PM1, PM2, PP3	De‐novo	Heterozygous	Severe condition not amenable to detection through sonographic imaging	Continuation of pregnancy

Note: In accordance with the Clingen recommendation, we applied the PM2 criteria at a “support” level of significance rather than “moderate.”

Abbreviations: ACMG, American College of Medical Genetics and Genomics; AD, autosomal dominant; AR, autosomal recessive; IVH intraventricular hemorrhage; N/A, not available; PM, pathogenic moderate; PP, pathogenic supporting; PS, pathogenic strong; PVS, pathogenic very strong; TOP, termination of pregnancy; XL, X linked.

3.4. Turn‐around times

Exome sequencing was conducted weekly, and the average TAT from the initial request was approximately 14 calendar days. In our study, all results were consistently available within a window of 9–28 days following sample collection, regardless of whether it followed a normal CMA or was conducted concurrently with it.

It is important to note that routine culture of the samples prior to analysis was not a standard practice. Instead, the primary approach involved direct DNA analysis. Only when an insufficient quantity of DNA was available for exome analysis did the process necessitate culture.

3.5. Pregnancy outcome

In our cohort, pregnancy termination was carried out in 13 out of the 28 reported cases (1.2% of the total cohort). This included seven cases from the severe phenotypes group (group 1), two from the milder phenotype group (group 2), and four cases from the Hot/warm VUS group (group 3). Pregnancy management decisions for each case are detailed in Tables 1, 2, 3.

4. DISCUSSION

We assessed the prevalence of potentially clinically significant variants indicative of known monogenic diseases in fetuses that displayed neither significant sonographic abnormalities nor a family history of monogenic disease. Within this meticulously curated cohort, we reported 21 (2.1%) cases with clinically significant variants (P/LP)—and an additional seven cases (0.7%) with potentially significant variants (Hot/warm VUSs), together constituting 2.7%.

The first group, consisting of nine cases (0.9%) known as the “Severe phenotypes group,” carried de novo variants associated with genes causing severe phenotypes with high penetrance.

Most couples undergoing pES would likely prefer to receive information about these variants during pregnancy for informed decision‐making. For example, in Case no. 4, a de novo frameshift variant in the ANKRD11 gene was identified, associated with KBG syndrome, featuring distinctive craniofacial and skeletal anomalies as well as neurologic impairments such as developmental delay and seizures.

Conversely, the second group, the “Milder phenotype group” consisted of 12 cases (1.1%). These cases may elicit diverse reactions from both genetic professionals and couples involved. In these cases, early detection lays the groundwork for prompt postnatal intervention, effective symptom management, and ultimately, enhanced patient prognosis. Although these cases still require thorough medical surveillance during the prenatal period, some couples might opt not to receive this information. In this subset, 70% of cases had inherited variants, carrying significant clinical implications for the fetus's health management as well as implications for the transmitting parent and future pregnancies. In case no. 17, a paternally inherited heterozygous missense variant in the FBN1 gene was found to be associated with AD Marfan syndrome. The father exhibited typical features of Marfan syndrome (aortic root dilatation, myopia, normal IQ). The couple opted to continue the pregnancy.

In the third group, the “Hot/warm VUS group,’’ we identified seven cases (0.7%). The third group demonstrates one of the challenges associated with pES—the complexity of variant classification. Evolving methodologies with rapid updates further compound this challenge. ¹¹ , ¹² , ¹⁸ This is particularly pronounced in fetal variant classification, particularly when the fetus exhibits no abnormal ultrasound findings. For instance, the PS2 criteria, which indicate a de novo variant when both parents are tested, ¹⁹ is not applicable in the absence of an identifiable phenotype, ¹¹ as in our cohort. Under these circumstances, achieving a “LP” classification is often exceedingly difficult. Theoretically, if the PS2 criteria had been applicable in all cases with reported hot/warm VUSs, they could have been classified as LP. For example, in Case No. 24, a VUS was identified in the CHD4 gene, which is associated with Sifrim‐Hitz‐Weiss syndrome. Subsequent ultrasound follow‐up revealed enlarged ventricles in this case. While this variant theoretically could be reclassified as LP due to a compatible phenotype and the application of the PS2 criterion it's important to note that the observed phenotypes might not be specific enough to definitively support this reclassification. Nevertheless, with each normal individual typically carrying 45–60 de novo mutations per genome, ²⁰ a de novo VUS may not necessarily lead to disease.

In case no. 26, pES analysis identified de novo missense heterozygous variant in SCN1A gene, associated with AD Dravet syndrome, a condition characterized by early onset seizures refractory to medical therapy and severe psychomotor delay. Notably, this variant has not been previously reported. Due to the absence of a discernible clinical phenotype, the PS2 criterion could not be applied, so the variant reached only VUS classification.

This latter case serves as an example of a common ethical dilemma: Is it justifiable to withhold the reporting of these variants in the context of non‐indicated pES? A similar ethical dilemma was previously discussed in the literature regarding the variants of unclear clinical significance in the CMA technology. The reported detection of VUSs in prenatal CMA ranges from 2% to 6% in prenatal cases without structural anomalies ²¹ to from 4% to 20% in cases involving fetuses with structural anomalies. ²² , ²³ , ²⁴

In case no. 23, a variant in the EFNB1 gene inherited from the mother to a male fetus highlights the complexity of variant classification and reporting thresholds. This condition usually follows X‐linked dominant inheritance and results in craniofrontonasal dysplasia, though it typically presents more severely in females. We reported this variant due to its documentation in one female and two affected males in the literature. ²⁵ , ²⁶ Despite the mother's reported health (with no measured intraorbital distance), variant expressivity and a mild phenotype cannot be entirely ruled out.

Lazier et al. highlighted the limited clinical utility of VUS in prenatal sequencing, noting its potential to cause patient anxiety and confusion. They also underscore the lack of evidence regarding the benefits of reporting secondary findings in fetuses. ²⁷ Meanwhile, Vears et al. proposed a decision‐making framework for reporting secondary and incidental findings in prenatal sequencing. ²⁸

The study by Tavtigian et al. reveals the probabilistic essence of ACMG criteria, setting a prior probability threshold at 10% for variant classification, predicated on the existence of a phenotype that can be correlated with genetic findings. ²⁹ This insight prompts critical considerations regarding the ACMG criteria's fit for screening contexts, where phenotypic details are typically lacking, and the focus shifts to gauging the incidence or prevalence of conditions at a population scale. In the absence of established ACMG criteria for fetuses lacking any phenotype and the absence of universally accepted guidelines delineating which variants to report and under what circumstances, the clinical laboratory should exercise discretion in making these determinations. As experience accumulates globally, it becomes imperative to develop and endorse specific guidelines for the prenatal setting.

Based on our cohort's data, advanced parental age doesn't seem to significantly increase the risk of de novo disease‐related genetic alterations. While the literature suggests a higher prevalence of point variants with older fathers, ³⁰ our cohort's results may not align with this trend.

Existing literature on fetuses without structural abnormalities is currently limited, with only two published studies available. The first reports an ES yield of 0.6% (1 out of 160), ⁸ while the second reports a yield of 0.83% (4 out of 482) for moderate to severe diseases. ⁹ These yields closely resemble what we observed in our first group. Thus, the low diagnostic yield in these studies might be due to specific variant selection criteria.

It is estimated that severe pediatric monogenic disorders may affect approximately 1%–5% of live births overall. ³¹ , ³² , ³³ If we theoretically exclude the cases involving abnormal fetuses from these estimations, the prevalence would likely only be slightly lower since most of these diseases do not manifest prenatally. Thus, our observed prevalence of about 2% aligns well with the existing knowledge and expectations regarding the prevalence of these genetic disorders in the general population.

In considering pES, many disorders may not show detectable anatomical abnormalities in fetal development, emphasizing the diagnostic potential even without obvious sonographic anomalies. Sukenik‐Halevy et al. found that 52.5% of postnatally diagnosed neurocognitive monogenic disorders lacked prenatal sonographic findings, with 55.9% of diagnosed genes not associated with congenital structural anomaliess. ³⁴

pES and variant curation are primarily driven by the phenotype. However, when no phenotype is present, this process must be approached with meticulous consideration and utmost caution. Specifically, without a phenotype, the pre‐test probability is much lower (i.e., population incidence of the condition).

The disclosure of uncertain variants induces anxiety in expectant parents who worry about their child's future health implications. Conversely, the decision to withhold information about suspected disease‐causing variants in genes linked to severe phenotypes raises ethical dilemmas. These situations not only present challenges for the couple but also for the laboratory and/or clinician in deciding the implications of not reporting variants. This underscores the critical need for thorough bioinformatic analysis, on one hand, and pre‐test counseling, on the other, to assist parents in making informed decisions within this emotionally charged context.

Our study stands out due to its sample size and the stringent selection of a truly unified cohort of more than a thousand low risk pregnancies, thereby minimizing confounding variables. Further adding to its significance, it explores the scarcely researched realm of prenatal diagnosis of monogenic diseases in fetuses devoid of structural abnormalities. Our multidisciplinary team of bioinformaticians and clinicians, abiding by the most recent ACMG guidelines, ensured rigorous variant analysis and comprehensive genetic counseling. However, there are several limitations to consider in this study. A significant study limitation arises from the absence of outcome data, which can be attributed to the study's primary focus on clinical laboratory aspects. This focus has resulted in the lack of reported autopsy findings and the absence of data related to infant or child outcomes.

Another limitation is the deliberate exclusion of cases with abnormal biochemical screening and soft markers on ultrasound. While this exclusion was a well‐considered choice, it can also be viewed as a limitation, as these cases are not typically indicated for pES.

In conclusion, our study highlights the importance of pES as a diagnostic tool, even in the absence of indicators of monogenic disease such as significant ultrasound findings or a suggestive family history. pES can greatly aid in the prognosis of the fetus, guide potential in‐utero treatments, inform delivery planning, steer parental and neonatal care, assist in planning future pregnancies, and consider reproductive options. We advocate for the expanded use of pES as part of routine pregnancy surveillance and urge the establishment of dedicated clinical guidelines to optimize its utilization. Further research is warranted to evaluate the utility of pES and assess its cost‐effectiveness and optimal timing in low‐risk pregnancies.

CONFLICT OF INTEREST STATEMENT

The authors report no conflicts of interest.

ETHICS STATEMENT

This study was carried out according to the Declaration of Helsinki and approved by the Helsinki Ethics Committee of Maccabi Healthcare Services (one of the four Health Maintenance Organizations [HMO's] operating in Israel).

Supporting information

Supporting Information S1

Table S1

Levy M, Lifshitz S, Goldenberg‐Fumanov M, et al. Exome sequencing in every pregnancy? Results of trio exome sequencing in structurally normal fetuses. Prenat Diagn. 2025;45(3):276‐286. 10.1002/pd.6585

DATA AVAILABILITY STATEMENT

The data supporting the findings of this study are available upon request from the corresponding author. Data sharing is subject to ethical approval.