Fetal congenital gastrointestinal obstruction: prenatal diagnosis of chromosome microarray analysis and pregnancy outcomes

Congenital gastrointestinal obstruction is a common digestive tract malformation in neonates, with an incidence of 0.7–4.2 per 10,000 live births [1‐3]. Congenital gastrointestinal obstruction may have some ultrasound features during pregnancy and the most common is the “double bubble” sign. As fetal gastrointestinal tract(GIT) is the main rout to absorb amniotic fluid, polyhydramnios were often observed in fetus with gastrointestinal obstruction [4, 5]. These ultrasound phenotype of fetal gastrointestinal obstruction is usually detected in second or third trimester [6]. Genetic abnormalities, embryonic teratogens and infection are suspected as high risk factors associated with congenital gastrointestinal obstruction.

Previous studies find that gastrointestinal obstruction was noticed in 3–5% trisomy 21 [7, 8]. Other microdeletions, such as 17q12 deletion and 4q22.3 deletion, have been reported in fetus with gastrointestinal obstruction [4, 9]. However, the data on the incidence of genetic findings in fetus with gastrointestinal obstruction was limited. Moreover, researches on the outcome of these babies after birth were rare.

To promote data based genetic counseling, we investigated the incidence of chromosome anomalies in different types of congenital gastrointestinal obstruction, followed by evaluating the pregnancy outcomes of fetuses with congenital gastrointestinal obstruction.

Sixty-four singleton fetuses diagnosed with congenital gastrointestinal obstruction were enrolled between January 2014 and December 2020. The typical ultrasound images of congenital gastrointestinal obstruction were shown in Fig. 1. In our study, the gestational ages for diagnosing congenital gastrointestinal obstruction ranged from 23 to 36 weeks. The numbers of Group A, Group B and Group C were 37, 15 and 12 respectively.

Joint Society of Obstetricians and Gynaecologists of Canada (SOGC)-Canadian College of Medical Geneticists (CCMG) recommendations suggested that CMA should be applied to fetuses with structural abnormalities [12]. CMA is recommended as a first-tier technology for prenatal diagnosis of congenital gastrointestinal obstruction. In our group, 6.3%(4/64) fetuses were diagnosed as trisomy 21. This was slightly lower than previous studies [12, 13]. This may due to application of noninvasive prenatal testing (NIPT) and improvement in ultrasonography. Additionally, submicroscopic CNVs were detected in 7.8% cases. The incidence is similar to Zhang WW’s findings [13]. The CNV findings in our study included 17q12 duplication syndrome, Williams syndrome (WS), 4p16 deletion, both 4p16 deletion and 10q duplication, both 4p16.3 duplication and 12p13.33–13.31 deletion.

Case 1, a fetus with duodenal atresia and polyhydramnios, had a 1.42 Mb duplication in chromosome 17q12. Duplication of this segment can lead to 17q12 duplication syndrome. The 17q12 duplication syndrome has mainly been proposed to include autism, behavioral abnormalities, structural brain abnormalities, learning disability, epilepsy, renal disease, atresia, and endocrine abnormalities [14‐16]. Among the effected genes is HNF1B. Overexpression of HNF1B is associated with annular pancreas(AP) [17]. Although the overexpression of HNF1B has not been demonstrated to be directly associated with duodenal atresia, AP results in the pancreatic tissue completely or incompletely surrounding the duodenum leading to duodenal obstruction. Maria Rasmussen et al. [18] reported a patient with 17q12 duplication syndrome who was suspected “duodenal atresia” prenatally, but it was not mentioned whether the patient has been diagnosis with AP. It was a pity that autopsy was refused by the parents in our study. More researches are needed to clarify whether duodenal atresia is constitutional phenotype of 17q12 duplication or only an incidental finding.

Case 2 was a fetus with “double bubble” sign, polyhydramnios and fetal growth restriction (FGR). CMA revealed a 1.8 Mb 7q11.23 deletion, also known as Williams Syndrome (WS). This is a multisystem disorder, including but not limited to cardiovascular disease, a distinctive craniofacial appearance, and a specific cognitive and behavioral profile. Gastrointestinal obstruction was not a commonly finding of this syndrome. Previous studies [19, 20] reported a fetus of duodenal atresia who had 7q11.23 deletion. Haploinsufficiency of ELN gene located in this region [21] is responsible for the vascular and connective tissue features of WS [22]. Elastin haploinsufficiency could also cause hyperplasia of sub-endothelial migration and vascular smooth muscle cell, leading to encroachment on the vascular lumen and arterial stenosis [23]. Pathogenic evidence is needed to be provided to prove the causative effect of elastin deficiency and duodenal atresia.

Case 3 and case 5 were found to be absent stomach bubble, FGR and absence of nasal bone by ultrasound. 4p16 deletion was detected in both fetus by CMA. Case 5 carried a segmental duplication of 10q25. Deletions of 4p16.3 region cause Wolf-Hirschhorn syndrome (WHS). “Greek warrior helmet” face, congenital heart disease, developmental delay, hypotonia, intellectual disability and seizures are often observed in patients with WHS [24, 25]. The absence stomach bubble suggested by ultrasonography may be associated with hypotonia in patients with WHS [26, 27]. This is the first time to report absence of stomach bubble as a prenatal phenotype of WHS, in two unrelated fetuses.

Case 4 was a fetus with a 3.9 Mb 4p16.3 duplication and a 5.23 Mb 12p13.33–13.31 deletion presenting with “double bubble” sign and gallbladder not shown. So far, information about duplications of the 4p16.3 region is limited. No literature has confirmed the association between gastrointestinal obstruction and 4p16.3 deletion. The reported clinical features in patients with 12p13.3 deletions varied considerably, presumably as a result of variation in deletion size. Recurrent clinical findings in these patients include intrauterine growth retardation, schizophrenic features, muscular hypotonia, microcephaly and other congenital abnormalities [28, 29]. The deletion of 12p13.33–13.31 encompasses the CACNA1C and ERC1 gene. Patients with genetic variations in the CACNA1C gene have been shown to have increased risk for psychiatric disorders [30]. A recent genotype–phenotype characterization proposed ERC1/ELSKS as a good candidate gene for childhood apraxia of speech (CAS). Isabela et al. [31] suggested that ERC1 is the best candidate for the neurodevelopmental delay and autism spectrum disorders. However, neither fragment was found to be associated with gastrointestinal obstruction.

By dividing our subjects in to three subgroups, we found that the prevalence of chromosomal anomalies in non-isolated upper gastrointestinal obstruction (Group C) was about twice as high as that in isolated congenital gastrointestinal obstruction (Group A and Group B). This finding is also in accordance with Wu XQ and Meng XY’ data [32, 33]. These results proved that a detailed evaluation of the fetus and a fetal echocardiography is essential after the suscition of a GIT obstruction [34]. On the other hand, we didn’t find any genetic abnormalities in Group B. This is also observed in Orgul G’s corhort [35]. As the lower gastrointestinal obstruction is usually manifested very late in pregnancy, this information would be a great comfort for those women with such fetuses.

Postnatal outcomes of 27 live born babies in our study were generally good. Only one case of intestinal atresia combined with biliary atresia died due to liver cirrhosis. The remaining 16 babies underwent surgery at different times according to their clinical conditions. They were all alive during the follow-up period. The survival rate after surgery was 94.1%. A British study [36] showed that congenital duodenal obstruction surgery has a high success rate and a low reoperation rate. Short-term outcomes are generally good. The overall survival rate was reasonably good at 88% [37]. Other findings suggest that the long-term survival rate after surgery is more than 80% [38, 39]. Despite the progress in postnatal surgery, we should keep in mind that genetic abnormalities must be ruled out.

There are several limitations in our study. It was a retrospective study and the amount of data was small compared with other studies. Fetuses with pathogenic genetic findings were terminated without autopsy. This prevented us from further investigating the genetic-phenotype relationships. In the future, we would collect more cases and multi-center cooperation will be considered.

It is crucial to understand whether the gastrointestinal tract abnormality is isolated or associated to other findings. The risk of chromosomal abnormalities in fetuses with isolated lower gastrointestinal obstruction is lower than upper gastrointestinal obstruction. While genetic abnormalities excluded, a promising prognosis is expected for fetuses with congenital gastrointestinal obstruction.