Prenatal diagnosis and perinatal outcomes of fetuses with congenital duodenal obstruction: A nine‐year retrospective study from China

Abstract

Introduction

Congenital duodenal obstruction (CDO) is one of the most common fetal gastrointestinal anomalies, but previous prenatal studies tend to investigate CDO as a portion of gastrointestinal obstruction. Few studies describe the genetic findings of CDO, especially copy number variants, in fetal cohorts with a relatively large sample size. The study aims to investigate the detection rate of genetic causes at different levels and to explore the potential influencing factors of perinatal outcomes of fetuses with CDO.

Material and Methods

This retrospective study analyzed karyotype, chromosomal microarray analysis, and trio‐whole exome sequencing (trio‐WES) results of singleton fetuses suspected of CDO in a tertiary center between January 2014 and September 2023. In addition, perinatal outcomes and postnatal medical records of enrolled cases were followed up and analyzed.

Results

A total of 98 fetuses were included in the study, of which 69 (70.4%) were classified as isolated CDO and 29 (29.6%) were classified as non‐isolated. The overall rate of genetic anomalies was 20.4% (20/98). Trisomy 21 (9/98 [9.2%]) and microduplication of the 17q12 region (3/98 [3.2%]) were the most common chromosomal numerical abnormalities and pathogenic copy number variants found in this cohort, respectively. Compared to chromosomal microarray analysis, no additional pathogenic or likely pathogenic variants were found in seven cases undergoing trio‐WES. The rate of chromosomal numerical and structural abnormalities was significantly higher in the non‐isolated group (8/29 [27.6%] vs. 6/69 [8.7%], p < 0.05). In terms of perinatal outcomes, the live birth rate was significantly higher in the isolated group (52/69 [75.4%] vs. 15/29 [51.7%], p < 0.05) largely due to fewer terminations of pregnancy. All cases with positive genetic results elected to terminate the pregnancy, and 82% of those with negative results opted to continue the pregnancy. Neonatal mortality was significantly higher in the non‐isolated group (2/15 [13.3%] vs. 0/52 [0.0%], p < 0.05). The overall neonatal survival rate was 97.0% (65/67).

Conclusions

The present study highlights the value of prenatal diagnostic testing for fetuses suspected of CDO, both in isolated and non‐isolated cases. Genetic diagnostic outcomes exert substantial influence on pregnancy decision‐making. Perinatal outcome and short‐term prognosis of affected fetuses are reasonably favorable when known genetic causes are excluded.

A retrospective study of 98 fetuses with congenital duodenal obstruction (CDO) revealed an overall genetic abnormality rate of 20.4%, with trisomy 21 being the predominant chromosomal anomaly. Comprehensive prenatal diagnostic testing is recommended for fetuses with suspected congenital duodenal obstruction, as the genetic findings exert substantial influence on pregnancy decision‐making. Perinatal outcome and short‐term prognosis are reasonably favorable when known genetic causes are excluded. Created with BioGDP.com.³²

Abbreviations

ACMG

American College of Medical Genetics

CDO

congenital duodenal obstruction

CMA

chromosomal microarray analysis

CNVs

copy number variations

MRI

magnetic resonance imaging

NIPT

non‐invasive prenatal testing

NT

nuchal translucency

P/LP

pathogenic or likely pathogenic

PLSVC

persistent left superior vena cava

TOP

termination of pregnancy

trio‐WES

trio‐whole exome sequencing

VOUS

variants of unknown significance

Key message.

This study highlights the necessity of chromosomal microarray analysis and the potential benefits of whole‐exome sequencing for fetal congenital duodenal obstructions. Furthermore, multidisciplinary team consultation is suggested for the parents‐to‐be.

1. INTRODUCTION

Congenital duodenal obstruction (CDO) is one of the most common fetal gastrointestinal anomalies, with a prevalence of 1 in 5000 to 10 000 live births. ¹ CDO is suggested to be caused by failure of bowel recanalization during post‐conception days 44 to 46. ² As a high incidence of anomalies of midline structures has been observed in CDO cases, some scholars have suspected disruptions in endoderm development to be the underlying pathogenesis. ³ , ⁴

The diagnosis of CDO prenatally is generally limited to ultrasonography and relies on the sign of a “double bubble” in the fetal upper abdomen (Figure 1). ⁵ , ⁶ , ⁷ Double bubbles are attributed to the dilated stomach and the proximal part of the duodenum, respectively, in which continuity between the two bubbles should be observed. Gao et al. determined cutoff values of 17.5 mm and 10.5 mm for “double bubble sign” lesions in the gastric and duodenal widths by ROC analysis. ⁸

FIGURE 1.

Presence of double‐bubble sign in fetal ultrasonographic scanning.

The association between CDO and trisomy 21 has been well described in multiple studies. Down syndrome has been noted in approximately 16–50% of patients with CDO. ⁵ , ⁹ , ¹⁰ , ¹¹ , ¹² Since chromosomal microarray analysis (CMA) has gradually become the first‐tier genetic analysis in prenatal diagnostic testing, pathogenic and likely pathogenic (P/LP) copy number variants (CNVs) in CDO cases have been reported in a few studies. ¹⁰ , ¹³ However, both the number of studies and the corresponding sample size were limited. Here, we presented our experience on prenatal diagnosis of CDO mainly based on karyotyping and CMA, as well as postnatal outcomes of live births to provide more evidence for genetic consultation and pregnancy management.

2. MATERIAL AND METHODS

2.1. Participants and clinical data

This is a retrospective database cohort of fetuses who had undergone prenatal diagnostic testing for CDO from January 2014 to September 2023 in a single tertiary center.

A standardized protocol for aneuploidy screening was applied to all pregnancies with first‐trimester registration (<13⁺⁶ weeks). For those with normal nuchal translucency ultrasound findings, combined first‐trimester screening (integrating serum markers, maternal age, and nuchal translucency measurement) was recommended for women under 35 years, while non‐invasive prenatal testing (NIPT) was recommended for those aged 35 or older. Conversely, for cases with abnormal nuchal translucency findings, further evaluation with either NIPT or invasive prenatal diagnostic testing was recommended based on the specific clinical context and severity of the findings. All pregnant women were suggested to commit to regular prenatal care and receive sequential structural ultrasound scans.

Fetuses presenting with “double bubble” signs at any given gestational age were considered CDO and detailed anatomic scanning was performed by specialized sonographers. Cases were excluded when the “double bubble” sign could not be observed on repeat prenatal ultrasound.

Consecutive singleton pregnancies referred for prenatal diagnostic testing with CMA and (or) WES following ultrasonographic suspicion of CDO were enrolled, irrespective of prior aneuploidy screening results. All the cases enrolled were offered multi‐disciplinary team consultation. The multi‐disciplinary team was composed of experts in obstetrics, fetal medicine, and pediatric gastrointestinal surgery, which was intended to help couples understand medical and clinical information about the disease in detail.

The enrolled cases were divided into isolated and non‐isolated CDO according to the absence or presence of other ultrasonographic abnormalities including soft sonographic markers. Cases combined with isolated polyhydramnios were classified as isolated CDO because polyhydramnios mostly developed secondary to gastrointestinal obstruction.

Demographic as well as perinatal and genetic outcome data were collected by reviewing electronic medical records and making telephone follow‐ups. Operative reports and medical records were followed up to confirm the diagnosis of duodenal obstructions for liveborn cases. Genetic anomalies included chromosomal numerical and structural abnormalities, P/LP CNVs, P/LP single nucleotide variants (SNVs). Preterm deliveries are defined as those occurring at less than 37 weeks' gestational age. ¹⁴ Gestational complications include gestational diabetes mellitus, hypertension in pregnancy, hyperthyroidism, hypothyroidism and autoimmune diseases. Adverse outcomes include fetal demise, neonatal death, and gastrointestinal complications requiring reoperation during the follow‐up period.

2.2. Genetic analysis

For each case, chorionic villus sampling, amniocentesis, or cordocentesis was performed according to the established gestational age windows. The genomic DNA of the fetuses was extracted using the QIAamp DNA Blood Midi/Mini kit (QIAGEN GmbH; Hilden, Germany).

The quantitative fluorescent polymerase chain reaction (QF‐PCR) was performed to identify maternal cell contamination and abnormalities of chromosomes 21, 18, 13, and sex chromosomes. Karyotype analysis at a resolution of 400–500 band level was performed according to the standard G‐banded karyotyping protocol.

CMA was conducted using Affymetrix CytoScan HD/750K array (Affymetrix Inc., Santa Clara, CA, USA) with a single‐nucleotide polymorphism array (SNP array) and array‐based comparative genomic hybridization (aCGH) platforms at resolutions of 10 and 100 kb, respectively. Human genome version GRCh37 (hg19) was used. All detected CNVs were compared to in‐house and international public CNV databases for classification according to joint consensus recommendations of the American College of Medical Genetics and Genomics (ACMG). All the P/LP CNVs, and variants of unknown significance (VOUS) were documented.

Some of the cases were subjected to trio‐WES when showing negative findings of CMA. Libraries of paired‐end 150‐bp reads were sequenced using Hiseq XTen (Illumina, Inc., San Diego, CA, USA). The sequencing reads were aligned to the human reference genome (GRCh37/hg19). NextGENe v2.4.1.2 software (SoftGenetics, State College, PA, USA) was employed for variant calling. Variant annotation and interpretation were confirmed through literature and international public databases. Multiple computational algorithms were employed to determine the pathogenicity of the candidate variants. The variants are classified into five categories including pathogenic (P), likely pathogenic (LP), VOUS, likely benign (LB) and benign (B) according to the ACMG guidelines.

2.3. Statistical analysis

Statistical analysis was performed with SPSS Statistics version 25 software (IBM, Armonk, NY, USA). The quantitative data of normal distribution were documented as the mean and the standard deviation; otherwise, the median and quantile were used. Categorical variables were documented as percentages. Student t‐test, Pearson's chi‐squared test, Yates' continuity corrected chi‐square test and Fisher's exact test were performed under appropriate conditions. A p‐value <0.05 was considered statistically significant.

3. RESULTS

A total of 98 fetuses diagnosed with CDO and having had prenatal diagnostic testing were included in our cohort, of which 69 (70.4%) were classified as isolated CDO and 29 (29.6%) were classified as non‐isolated (Figure 2).

FIGURE 2.

Flowchart of the study population.

The accompanying abnormalities in non‐isolated CDO are shown in Table S1, including aberrant right subclavian artery (N = 5), single umbilical artery (N = 5), fetal growth restriction (N = 5), mild cerebral ventriculomegaly (N = 4), short femur/humerus (N = 4), and ventricular septal defect (N = 4), etc. Demographic and outcome data for the isolated and non‐isolated CDO groups are presented in Table 1. The proportion of nulliparous women was significantly higher in the non‐isolated group.

TABLE 1.

Clinical characteristics of fetuses with isolated DO and non‐isolated DO.

	Isolated (n = 69)	Non‐isolated (n = 29)	t/Χ 2 /Z	p
Maternal age (years)	30.1 ± 4.5	29.6 ± 5.5	0.5	0.62
Nulliparous (%)	35 (50.7%)	21 (72.4%)	3.9	<0.05
Previous abortion or miscarriage (%)	27 (39.1%)	9 (31.0%)	0.6	0.45
Polyhydramnios (%)	38 (55.1%)	15 (51.7%)	0.1	0.76
Gestational complications (%)	9 (13.0%)	4 (13.8%)	0.0	>0.99 a
Gestational age at invasive testing (weeks)	26.1 (23.3–29.1)	27.1 (24.5–32.3)	−1.9	0.20
Specimen			2.5	0.30 b
Chorionic villi (%)	0 (0.0%)	1 (3.5%)
Amniotic fluid (%)	55 (79.7%)	21 (72.4%)
Cord blood (%)	14 (20.3%)	7 (24.1%)
Genetic outcomes			5.6	<0.05 b
Negative (%)	58 (84.1%)	20 (69.0%)	2.9	0.09
Chromosomal numerical and structural abnormalities (%)	6 (8.7%)	8 (27.6%)	4.5	<0.05 a
P/LP CNVs (%)	5 (7.2%)	1 (3.4%)	0.1	0.80 a
Pregnancy outcomes			7.4	<0.05 b
Live births (%)	52 (75.4%)	15 (51.7%)	5.3	<0.05
TOP (%)	14 (20.3%)	12 (41.4%)	4.7	<0.05
Fetal demises (%)	2 (2.9%)	0 (0.0%)	/	>0.99 b
Lost to follow‐up (%)	1 (1.4%)	2 (6.9%)	/	0.21 b

Abbreviations: P/LP CNVs, pathogenic or likely pathogenic copy number variants; TOP, termination of pregnancy.

^a Yates' continuity corrected chi‐squared test.

^b Fisher exact test.

The first‐trimester aneuploidy screening results were available for 92.9% (91/98) of the fetuses. Among the subset of 30 cases that underwent NIPT, two were identified as high‐risk, and both were subsequently confirmed by prenatal diagnostic testing. In terms of diagnostic testing, CMA results were available for most of the fetuses (96/98 [98.0%]). Of the remaining two cases, one was diagnosed with trisomy 21 by karyotype analysis, for which CMA was canceled, and the other preferred trio‐WES instead of CMA. Karyotype results and trio‐WES results were available for 33 (33/98 [33.7%]) and seven (7/98 [7.1%]) cases, respectively.

Genetic anomalies were noted in 11 (11/69 [15.9%]) and 9 (9/29 [31.0%]) cases in the isolated and non‐isolated groups, respectively. Chromosomal numerical and structural abnormalities were more likely to be detected in the non‐isolated group (27.6% [8/29] vs. 8.7% [6/69], p < 0.05).

Trisomy 21 was the most common chromosomal numerical abnormality in the cohort (9/98 [9.2%]) and was the most common abnormality in both the isolated (3/69 [4.3%]) and non‐isolated (6/29 [20.7%]) CDO cases (Table 2). Apparently, trisomy 21 was more likely to be detected in the non‐isolated group (p < 0.05). In addition, five cases in total of chromosomal inversion or deletion were detected by CMA.

TABLE 2.

Details of genetic outcomes of fetuses with isolated DO and non‐isolated DO.

	Isolated (N = 69)	Non‐isolated (N = 29)	t/Χ 2 /Z	p
Trisomy 21	3 (4.3%)	6 (20.7%)	4.7	<0.05 a
Chromosomal deletion	2 (2.9%)	2 (6.9%)	0.1	0.72 a
Chromosomal inversion	1 (1.4%)	2 (6.9%)	/	0.21 b
P/LP CNVs	5 (7.2%)	1 (3.4%)	0.1	0.80

Abbreviation: P/LP CNVs, pathogenic or likely pathogenic copy number variants.

^a Yates' continuity corrected chi‐squared test.

^b Fisher exact test.

In terms of CNVs, microduplication of the 17q12 region was the most common pathogenic CNV in the cohort, noted in three (3/98 [3.1%]) cases in total. Pathogenic microdeletion of the 7q11.23 region, 13q21.33q31.1 region, and 15q25.2 region was noted in the other three cases, respectively. Of note, microdeletion of the 16q23.3q24.1 region, which was classified as VOUS, was found in one case of fetal demise.

Trio‐WES was applied to four isolated CDO cases and three non‐isolated CDO cases. Two P/LP CNVs were detected in two cases, which were consistent with their CMA results. In addition, one case was found to carry a hemizygous variant of filamin A (FLNA) gene c.7000C>T (p.Arg2334Cys) inherited from his mother. The variant was classified as VOUS and may be associated with congenital short‐bowel syndrome. The case ended in a male live birth with no additional structural anomalies found. The patient had normal development after standard surgical treatment. Details of genetic results are presented in Tables 3 and 4.

TABLE 3.

Details of cases of chromosomal numerical and structural abnormalities.

Case number	Maternal age (years)	Specimens	Ultrasound findings	Karyotype	CMA results	Pathogenicity of CNVs	Pregnancy outcome
Isolated group
1	30	AF	CDO	46,X?,+21,der(21;21)(q10;q10)	arr(21) × 3	P	TOP
2	30	AF	CDO	47,XN,+21	/		TOP
3	37	AF	CDO	/	arr21q11.2q22.3(15 006 457–48 097 372) × 2 ~ 3	P	TOP
4	28	AF	CDO	/	arr13q21.32q22.3(67 343 829–78 820 282) × 1	P	TOP
5	27	AF	CDO, polyhydramnios	46,XN,del (13)(q21q31)	arr13q21.2q31.1(60 952 932–84 894 325) × 1	P	TOP
6	37	AF	CDO, polyhydramnios	/	arr1q25.3q44(182 081 876–249 206 020) × 3	P	TOP
Non‐isolated group
7	30	AF	CDO, polyhydramnios, short femur/humerus, VSD, umbilical vein varix	/	arr(21) × 3	P	TOP
8	40	AF	CDO, polyhydramnios, VSD	/	arr21q11.2q22.3(15 023 401–48 097 372) × 3	P	TOP
9	38	AF	CDO, right subclavian vagus	/	arr21q11.2q22.3(15 006 457–48 097 372) × 3	P	TOP
10	34	AF	CDO, polyhydramnios, right subclavian vagus	/	arr21q11.2q22.3(15 023 401–48 097 372) × 3	P	TOP
11	24	AF	CDO, absent/hypoplastic nasal bone	/	arr21q11.2q22.3(15 023 401–48 097 372) × 3	P	TOP
12	24	CB	CDO, polyhydramnios, mild tricuspid regurgitation	/	arr21q11.2q22.3(15 006 457‐48 097 372) × 3	P	TOP
13 a	28	AF	CDO, right subclavian vagus, mild ventriculomegaly	/	arr1q41q44(215 272 546‐249 224 684) × 3	P	TOP
					arr10q26.12q26.3(122 567 637‐135 427 143) × 1	P
14 a	27	AF	CDO, right subclavian vagus, coarctation of the aorta, dysgenesis of the corpus callosum	/	arr1q41q44(215 269 409‐249 224 684) × 3	P	TOP
					arr10q26.12q26.3(122 562 370‐135 426 386) × 1	P

Abbreviations: AF, Amnionic fluid; CB, cord blood; CDO, congenital duodenal obstruction; CMA, chromosomal microarray analysis; CNVs, copy number variations; P, pathogenic; TOP, termination of pregnancy.

^a Case 13 and 14 were the two pregnancies of the same woman.

TABLE 4.

Details of cases of CNVs classified as P/LP and VOUS.

Case number	Maternal age (years)	Specimens	Ultrasound findings	Karyotype	CMA results	Pathogenicity of CNVs	Pregnancy outcome
Isolated group
15	32	AF	CDO	/	arr15q25.2(83 198 318–84 834 123) × 1	P	TOP
16	34	AF	CDO, polyhydramnios	N	arr17q12(34 822 466–36 397 279) × 3	P	TOP
17	29	AF	CDO	/	arr7q11.23(72 718 277–74 143 060) × 1	P	TOP
18	29	AF	CDO, polyhydramnios	/	arr13q21.33q31.1(70 986 515–80 427 919) × 1	P	TOP
19	26	CB	CDO, polyhydramnios	N	arr17q12(3 482 2465–36 410 559) × 3	P	TOP
21	34	AF	CDO	N	arr16q23.3q24.1(83 921 341‐86 412 783) × 1	VOUS	Fetal death
Non‐isolated group
20	28	AF	CDO, mild ventriculomegaly	/	arr17q12(34 822 466–36 378 678) × 3	P	TOP
22	26	CB	CDO, mild ventriculomegaly, enlarged cisterna magna	N	arr16q22.1(6 841 2069–69 535 869) × 3	VOUS	Live birth
23	31	AF	CDO, right aortic arch	/	arr3q23q24(139 523 132–142 861 530) × 1	VOUS	Live birth
24	29	CB	CDO, umbilical‐portal‐systemic venous shunt, absent gallbladder	N	arr8p22(14 261 319–15 427 340) × 3	VOUS	Live birth

Abbreviations: AF, Amnionic fluid; CB, cord blood; CDO, congenital duodenal obstruction; CMA, chromosomal microarray analysis; CNVs, copy number variations; P, pathogenic; TOP, termination of pregnancy; VOUS, variants of uncertain significance.

The pregnancy outcomes of the isolated group included 52 (75.4%) live births, 14 (20.3%) ended in termination of pregnancy (TOP), two (2.9%) fetal demises, and one (1.4%) lost to follow‐up. Whereas in the non‐isolated group, there were 15 (51.7%) live births, 12 (41.4%) ended in TOP, and two (6.9%) lost to follow‐up. Overall, the live birth rate was significantly higher in the isolated group (52/69 [75.4%] vs. 15/29 [51.7%], p < 0.05), largely due to fewer TOPs (14/69 [20.3%] vs. 12/29 [41.4%], p < 0.05).

Among 28 cases of non‐survivors, there were 20 cases of TOP after receiving positive genetic findings, three cases of TOP with negative genetic findings but major structural anomalies, three cases of TOP with negative genetic findings and isolated CDO, and two fetal demises with negative genetic findings. For live births, except for three cases delivered before the prenatal diagnostic results were available, the remaining 64 cases continued their pregnancies and delivered after receiving negative results. Therefore, pregnancy outcomes were strongly associated with the genetic findings.

Among 67 live births, two neonates did not undergo operative repair due to neonatal deaths. In neither case was the exact nature of the intestinal malformation determined (refusal of autopsy). Both neonatal deaths were from a non‐isolated group; one was accompanied by an absent nasal bone and a single umbilical artery, while the other was accompanied by fetal growth restriction. As a result, neonatal mortality was significantly higher in the non‐CDO group (2/15 [13.3%] vs. 0/52 [0.0%], p < 0.05). Both neonatal deaths were attributed to fetal distress and failure of neonatal rescue at gestational weeks of 31⁺³ and 35, respectively. The overall neonatal survival rate of fetuses with CDO was observed to be 97.0% (65/67) during the follow‐up period.

Duodenal obstructions were confirmed in all 65 neonates by postnatal surgery, whereas five of them presented with additional jejunal obstruction not detected prenatally. All five cases were from an isolated group.

Regarding novel genotypes identified postnatally (Table S2), in addition to jejunal obstruction identified in five cases of isolated CDO, ventricular septal defect, and persistent left superior vena cava (PLSVC) were each identified in one isolated CDO case. Among non‐isolated CDO cases, postnatal findings included partial anomalous pulmonary venous connection (N = 1), left‐sided cryptorchidism (N = 1), and concurrent ventricular septal defect (with PLSVC [N = 1]). None of the live‐born infants in this cohort underwent postnatal genetic testing.

A total of four cases in the isolated group and one case in the non‐isolated group underwent reoperation for gastrointestinal complications. Postnatal outcomes of live births are presented in Table 5. The non‐isolated cases were stratified by whether they had major structural anomalies or only ultrasonographic soft markers and were compared (Table 6). A total of 12 cases were found to combine with ultrasonographic soft markers only, of whom five (5/12 [41.7%]) were delivered. On the other hand, 17 cases were found to combine with major structural anomalies, of whom 10 (10/17 [58.8%]) were delivered. Overall, there was no considerable difference in maternal characteristics, fetal genetic anomalies, and postnatal outcomes of live births (Table 7).

TABLE 5.

Postnatal outcomes of live births with isolated DO and non‐isolated DO.

	Isolated (N = 52)	Non‐isolated (N = 15)	t/Χ 2 /Z	p
Infant gender			0.2	0.62
Male	21 (40.4%)	5 (33.3%)
Female	31 (59.6%)	10 (66.7%)
Preterm (%)	22 (42.3%)	5 (33.3%)	0.4	0.53
Gestational age at delivery (weeks)	37.0 ± 2.2	37.4 ± 2.2	−0.6	0.59
Mode of delivery			0.9	0.64 a
Spontaneous vaginal delivery	22 (42.3%)	6 (40.0%)
Forceps delivery	4 (7.7%)	0
Cesarean section	26 (50.0%)	9 (60.0%)
Adverse outcomes (%)
Neonatal deaths (%)	0 (0.0%)	2 (13.3%)	/	<0.05 a
Reoperation (%)	4 (7.7%)	1 (6.7%)	0.0	>0.99 b

^a Fisher exact test.

^b Yates' continuity corrected chi‐squared test.

TABLE 6.

Clinical characteristics of fetuses with ultrasonographic soft markers and major structural anomalies.

	Ultrasonographic soft markers (N = 12)	Major structural anomalies (N = 17)	t/Χ 2 /Z	p
Maternal age (years)	29.8 ± 4.9	29.5 ± 6.0	0.2	0.89
Nulliparous (%)	7 (58.3%)	14 (82.4%)	/	0.22 a
Previous abortion or miscarriage (%)	3 (25.0%)	6 (35.3%)	/	0.69 a
Polyhydramnios (%)	7 (58.3%)	8 (47.1%)	/	0.71 a
Gestational complications (%)	1 (8.3%)	3 (17.6%)	/	0.62 a
Gestational age at invasive testing (weeks)	28.3 ± 3.5	27.5 ± 6.3	0.4	0.69
Specimen			1.5	0.66 a
Chorionic villi (%)	0 (0.0%)	1 (5.9%)
Amniotic fluid (%)	8 (66.7%)	13 (76.5%)
Cord blood (%)	4 (33.3%)	3 (17.7%)
Genetic outcomes			3.8	0.13 a
Negative (%)	6 (50.0%)	14 (82.4%)	/	0.11 a
Chromosomal numerical and structural abnormalities (%)	5 (41.7%)	3 (17.6%)	/	0.22 a
P/LP CNVs (%)	1 (8.3%)	0 (0.0%)	/	0.41 a
Pregnancy outcomes			1.1	0.72 a
Live births (%)	5 (41.7%)	10 (58.8%)	/	0.46 a
TOP (%)	6 (50.0%)	6 (35.3%)	/	0.47 a
Lost to follow‐up (%)	1 (8.3%)	1 (5.9%)	/	>0.99 a

Note: Details of soft markers and major structural anomalies are listed in Table S1.

Abbreviations: P/LP CNVs, pathogenic or likely pathogenic copy number variants; TOP, termination of pregnancy.

^a Fisher exact test.

TABLE 7.

Postnatal outcomes of live births with ultrasonographic soft markers and major structural anomalies.

	Ultrasonic soft markers (n = 5)	Major anomalies (n = 10)	t/Χ 2 /Z	p
Infant gender			/	>0.99 a
Male	2 (40.0%)	3 (30.0%)
Female	3 (60.0%)	7 (70.0%)
Preterm (%)	2 (40.0%)	3 (30.0%)	/	>0.99 a
Gestational age at delivery (weeks)	36.8 ± 3.2	37.6 ± 1.7	−6.0	0.55
Mode of delivery			/	>0.99 a
Spontaneous vaginal delivery	2 (40.0%)	4 (40.0%)
Cesarean section	3 (60.0%)	6 (60.0%)
Adverse outcomes
Neonatal deaths (%)	1 (20.0%)	1 (10.0%)	/	>0.99 a
Reoperation (%)	1 (20.0%)	0 (0.0%)	/	0.33 a

^a Fisher exact test.

4. DISCUSSION

Congenital duodenal obstruction, characterized by a connected “double bubble” sign in the fetal upper abdomen, is a common abnormality of fetal digestive tract malformation that accounts for approximately one‐half of all intestinal obstructions. ⁸ , ¹³ , ¹⁵

Previous prenatal studies tended to investigate CDO within the framework of gastrointestinal obstruction, in which the association of trisomy 21 and CDO has been well established. ¹⁰ , ¹² , ¹³ , ¹⁶ Few studies, however, describe the genetic findings associated with CDO in a more comprehensive manner, employing CMA and trio‐WES results in a fetal cohort with a relatively large sample size. This retrospective study investigated the detection rate of genetic causes with a more in‐depth approach using updated methodologies to broaden the disease spectrum of CDO and explored the influences that the genetic diagnoses and other factors, such as additional sonographic findings, might have on parental choices and perinatal outcomes of fetuses with CDO.

In this study, CDO was confirmed in all live births by postnatal surgery except for the two neonatal deaths, demonstrating the diagnostic value of the “double bubble” sign in prenatal ultrasonographic scanning. However, five out of the 67 live births were found in combination with jejunal obstruction after birth, and one had to undergo a secondary operation for stoma closure. None of these five cases had undergone magnetic resonance imaging (MRI), which was suggested to be able to identify the exact level of the gastrointestinal obstruction and specify the presence or absence of normal bowel loops below the identified obstruction. ¹² Therefore, complementary prenatal MRI is suggested to aid in diagnosis and concurrently assess for other associated anomalies, which provides additional evidence for pregnancy decision‐making and postnatal prognostic evaluation.

The overall rate of genetic anomalies in this cohort was 20.4% (20/98), consisting of chromosomal numerical and structural abnormalities (14.3% [14/98]) and pathogenic/likely pathogenic CNVs (6.1% [6/98]).

Trisomy 21 accounted for 9.2% of the enrolled cases, which was slightly lower than that in previous studies. ⁵ , ⁹ , ¹⁰ , ¹¹ , ¹² It is probably attributed to the wide application of NIPT and improvement of ultrasonography which enabled trisomy 21 to be detected earlier before the “double bubble” sign was noted in the late second or third trimester of pregnancy in most cases. ⁵ , ¹⁰ Among the nine cases of trisomy 21 identified, four were classified as low‐risk by combined first‐trimester screening, and two were identified as high‐risk by NIPT. These findings indicate a non‐negligible false‐negative risk associated with the combined serum screening, highlighting the necessity of sequential structural ultrasound scans throughout pregnancy. Prompt genetic testing is indicated when ultrasound scans identify fetal structural anomalies, particularly those strongly associated with aneuploidy, even in pregnancies previously assessed as low‐risk by serum screening.

Consistent with previous studies, the prevalence of chromosomal anomalies, specifically trisomy 21, was significantly higher in the non‐isolated CDO group in this study. ¹⁰ , ¹³ , ¹⁷ Accompanying structural anomalies of trisomy 21 cases include aberrant right subclavian artery (N = 2), ventricular septal defect (N = 2), short femur/humerus (N = 1), absent/hypoplastic nasal bone (N = 1), and mild tricuspid regurgitation (N = 1).

In terms of CNVs, pathogenic microduplications of the 17q12 region were detected in three cases from both isolated and non‐isolated CDO groups. The 17q12 microduplication syndrome is a recurrent chromosomal imbalance with an autosomal dominant transmission pattern, containing pathogenic genes, such as HNFIB and LHXI. It has been reported to be associated with autism, behavioral abnormalities, learning disability, epilepsy, and cardiac and renal diseases. ¹⁸ , ¹⁹ , ²⁰ Duodenal obstruction has also been reported in multiple studies of 17q12 microduplication syndrome. ¹⁷ , ²¹ , ²² , ²³ In Cai's study of 17q12 microduplication syndrome, two out of five fetuses presented with CDO, suggesting that 17q12 microduplication might be an underlying genetic cause of CDO. ²¹ Experiments on zebrafish demonstrated that overexpression of HNF1B was associated with annular pancreas. ²⁴ Similarly, pathogenic microdeletions of 13q were detected in two CDO cases in a study by Meng et al. ¹³ Of note, microdeletion of 13q was found in one case in this study.

Compared to CMA, no additional P/LP variants were found in cases undergoing trio‐WES, suggesting a possibly limited diagnostic value of trio‐WES in isolated CDO. Multi‐center studies of CDO with a large sample size are expected to clarify the additional diagnostic value of trio‐WES. CMA has identified several microduplications and microdeletions through which genes responsible for CDO and their roles in the developmental process might be elucidated; additional candidate genes may be forthcoming as more patients with CDO are studied using CMA, and possibly using WES as well.

In a word, comprehensive prenatal diagnostic testing should be considered for fetuses with suspected CDO to diagnose or rule out associated genetic anomalies.

Consistent with a higher prevalence of genetic anomalies, the live birth rate significantly decreased in the non‐isolated CDO group mainly due to more TOPs in this study. This finding has been demonstrated in multiple studies. ⁹ , ¹⁰ , ¹³ , ¹⁷ In total, 100% (20/20) of cases terminated their pregnancies following receiving positive prenatal diagnostic results, and 82% (64/78) of cases continued their pregnancy and delivered following receiving negative prenatal diagnostic results. It can be reasonably inferred that genetic diagnostic outcomes, whether positive or negative, exert substantial influence on pregnancy decision‐making, which is consistent with the conclusion of a previous study. ²⁵

Novel phenotypes were detected in live births, including primarily jejunal obstruction, ventricular septum defect and persistent left superior vena cava. As known, prenatal ultrasonographic visualization can be compromised by multiple factors (e.g., gestational age, fetal position, amniotic fluid volume, operator experience, and equipment resolution), resulting in an inevitable rate of missed diagnoses. Therefore, serial prenatal ultrasonography is recommended for monitoring fetal structural development, with supplementary MRI and fetal echocardiography when indicated. ²⁶

All live births received operations except for two neonatal deaths. Most of the fetuses had normal development after standard surgical treatments. The overall neonatal survival rate of prenatally diagnosed CDO is reported to be as high as 96%, and most patients do very well long‐term following operative repair. ⁹ , ¹⁵ , ²⁷ , ²⁸ It is suggested that causes for mortality in neonates with GI obstruction are usually unrelated to the gastrointestinal pathology, such as genetic abnormalities, extreme prematurity, and other coexisting anomalies. ²⁹ In this study, the overall neonatal survival rate was determined at 97.0% during the follow‐up period and both neonatal deaths were attributed to fetal distress, which was consistent with previous studies. In general, short‐term prognosis is reasonably favorable when known genetic causes are excluded, which would be a great comfort for clinicians and parents‐to‐be.

Studies on prenatally and postnatally diagnosed CDO cases revealed that the former ones had worse perinatal outcomes but better prognosis after surgeries, which may be attributed to prompt management for such patients. ³⁰ , ³¹ Therefore, prenatal diagnosis of CDO by imaging examinations and implementation of prenatal diagnostic testing was of great value for affected fetuses.

The limitation of the study is first the small sample size, especially for non‐isolated CDO cases. Cases included in this study were those suspected of CDO by serial prenatal ultrasound findings of a double‐bubble sign. However, inherent limitations of prenatal ultrasonography may lead to inevitable missed diagnoses, potentially introducing bias into the results. Additionally, selection bias may be introduced since only pregnancies that underwent prenatal diagnostic testing were included in the cohort, which may both under‐ and overestimate the association with specific genetic aberrations. Second, not all the CMA‐negative cases underwent subsequent WES, making it unavailable to study the monogenic disorder spectrum underlying CDO. Third, some of the enrolled fetuses were born and treated at other hospitals; therefore, detailed prenatal ultrasonographic findings, neonates' conditions, and operation details were not available. Finally, while short‐term outcomes are known to be very favorable in isolated CDO following repair, long‐term follow‐up data would be valuable as might be acquired through a large cohort study.

5. CONCLUSION

This study showed that the overall rate of genetic anomalies in CDO was 20.4%. The risk of chromosomal numerical and structural abnormalities increased in the non‐isolated CDO group compared to the isolated CDO group. Comprehensive prenatal diagnostic testing is recommended for fetuses with suspected CDO, as the genetic findings exert substantial influence on pregnancy decision‐making. In addition, serial prenatal ultrasonography is recommended for monitoring fetal structural development, with supplementary MRI and fetal echocardiography when indicated. Multi‐disciplinary team consultation should be offered to help couples understand medical and clinical information about the disease in detail. Perinatal outcome and short‐term prognosis are reasonably favorable when known genetic causes are excluded. Most of the affected fetuses had normal development after standard surgical treatments during the follow‐up period.

AUTHOR CONTRIBUTIONS

Conceptualization and writing: J.L., F.G., and F.F.; Karyotype analysis: S.Y.; Chromosomal microarray data analysis: R.L. and X.J. Exome sequence variant classification: R.L.; Bioinformatics analysis: Q.Y., M.Y. and X.Z.; Clinical data collection: H.C., C.M. and L.L.; Clinical data statistics: H.Z., R.H., J.H. and D.L.; Supervision and editing: C.L. All authors have read and agreed to the published version of the manuscript.

FUNDING INFORMATION

This research was funded by the sub‐project of the National Key R&D Program (2021YFC2701002), the National Natural Science Foundation of China (82101796, 82201884), the Guangdong Basic and Applied Basic Research Foundation (2023A1515010115), Guangzhou Science and Technology Project (202201020643, 202201010879, 20221A011029, 202201020604, 202102020061), and the Research Foundation of Guangzhou Women and Children's Medical Center for Clinical Doctors (2020BS030).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

ETHICS STATEMENT

This study was approved by the Ethics Committee of Guangzhou Women and Children's Medical Center (No. 2021‐356B01) on December 7, 2021. Informed consent was obtained from all individual participants included in the study.

Supporting information

Table S1.

Lu J, Fu F, Guo F, et al. Prenatal diagnosis and perinatal outcomes of fetuses with congenital duodenal obstruction: A nine‐year retrospective study from China. Acta Obstet Gynecol Scand. 2026;105:268‐279. doi: 10.1111/aogs.70117

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.