The prenatal detection of distal tracheoesophageal fistulas in fetuses diagnosed with esophageal atresia

BACKGROUND

Esophageal atresia is a major anomaly of varying severity. The complexity of surgical correction depends on the presence of a distal fistula.

OBJECTIVE

This study aimed to determine the feasibility and accuracy of prenatal ultrasound detection of the distal fistula in fetuses diagnosed with esophageal atresia.

STUDY DESIGN

This was an observational study conducted at a single tertiary care center between 2019 and 2021. Included were pregnant patients carrying a fetus prenatally diagnosed with esophageal atresia that was confirmed postnatally during corrective surgery or at postmortem autopsy. During the scan, the performing investigator determined the presence or absence of a distal fistula by scanning the location of the lower esophagus during fetal breathing. Cases in which the lower esophagus was observed distending with amniotic fluid during breathing were deemed "fistula present," and the remaining cases "fistula absent." Test feasibility and performance indices, including sensitivity, specificity, and positive and negative predictive value were calculated. The offline clips and images were reviewed by 2 investigators for the assessment of interoperator agreement using Cohen's Kappa formula.

RESULTS

Included were 16 fetuses with esophageal atresia scanned between 2019 and 2021. All fetuses were successfully scanned with sufficient resolution of the area of interest during at least 3 cycles of breathing. It took a median of 8.5 minutes to determine the presence or absence of a distal fistula. The feasibility of the test was 100% (16/16). The test's sensitivity, specificity, and positive and negative predictive values were 80% (95% confidence interval, 55-100), 100% (95% confidence interval, 60-100), 100% (95% confidence interval, 65-100), and 75% (95% confidence interval, 45-100), respectively. The Cohen's Kappa for interoperator agreement was calculated to be 1, P<.001, corresponding to a "perfect" level of agreement.

CONCLUSION

Distal fistulas in esophageal atresia can be demonstrated prenatally by targeted scanning using appropriate technique. The method provided is feasible, reproducible, and has excellent performance indices. This novel technique and observations may improve the prenatal diagnosis and counseling of esophageal atresia.

Dynamic esophageal patency assessment: an effective method for prenatally diagnosing esophageal atresia

BACKGROUND

Esophageal atresia is a major anomaly with a low prenatal detection rate. We propose a sonographic method termed dynamic esophageal patency assessment.

OBJECTIVE

This study aimed to assess the feasibility and performance of the dynamic esophageal patency assessment in a high-risk population.

STUDY DESIGN

A prospective study was conducted in a single tertiary fetal ultrasound unit for 12 months. The study group included pregnant women referred for a targeted scan because of one or more of the following: (1) polyhydramnios; (2) small or absent stomach; (3) vertebral, anal atresia, cardiac, tracheoesophageal fistula, renal, and limb abnormalities; (4) first-degree relative with esophageal atresia; and (5) genetic mutation associated with esophageal atresia. In addition to dynamic esophageal patency assessment, a comprehensive anomaly scan was carried out. The fetal esophagus was observed during swallowing. Cases that demonstrated uninterrupted fluid propagation through the esophagus were classified as normal. Cases that demonstrated interrupted fluid propagation, with the formation of a pouch, were classified as abnormal. Cases with unclear visualization of the esophagus or cases that failed to demonstrate either fluid propagation or a pouch were classified as undetermined. Dynamic esophageal patency assessment results were compared with postnatal findings, considered "gold standard." Test performance indices and intra- and interobserver agreements were calculated.

RESULTS

For 12 months, 130 patients were recruited, and 132 fetuses were scanned. The median gestational age (interquartile range) at the time of scan was 31.4 weeks (29.0-35.3). Of 132 fetuses enrolled, 123 (93.2%) were normal, 8 (6%) were abnormal, and 1 (0.8%) was undetermined. Excluded from test performance analysis were 3 cases that were terminated without postmortem autopsy (1 was abnormal and 2 were normal), and a fourth case was excluded as it was classified as undetermined. The detection rate of esophageal atresia was 100%, with no false-positive or false-negative case. Sensitivity, specificity, and positive and negative predictive values of the dynamic esophageal patency assessment were 100%. The Kappa coefficient was 1 for both inter- and intraobserver agreements (P<.0001). The median time (interquartile range) required to complete the dynamic esophageal patency assessment was 6.00 minutes (3.00-13.25).

CONCLUSION

The dynamic esophageal patency assessment is a feasible and highly effective method of ascertaining an intact esophagus and detecting esophageal atresia in suspected cases.

Esophageal atresia and tracheoesophageal fistula: prenatal sonographic manifestation from early to late pregnancy

OBJECTIVE

Esophageal atresia and/or tracheoesophageal fistula (EA/TEF) remains one of the most frequently missed congenital anomalies prenatally. The aim of our study was to elucidate the sonographic manifestation of EA/TEF throughout pregnancy.

METHODS

This was a retrospective study of data obtained from a tertiary center over a 12-year period. The prenatal ultrasound scans of fetuses with EA/TEF were assessed to determine the presence and timing of detection of three principal signs: small/absent stomach and worsening polyhydramnios, both of which were considered as 'suspected' EA/TEF, and esophageal pouch, which was considered as 'detected' EA/TEF. We assessed the yield of the early (14-16 weeks' gestation), routine mid-trimester (19-26 weeks) and third-trimester (≥ 27 weeks) anomaly scans in the prenatal diagnosis of EA/TEF.

RESULTS

Seventy-five cases of EA/TEF with available ultrasound images were included in the study. A small/absent stomach was detected on the early anomaly scan in 3.6% of fetuses scanned, without a definitive diagnosis. On the mid-trimester scan, 19.4% of scanned cases were suspected and 4.3% were detected. On the third-trimester anomaly scan, 43.9% of scanned cases were suspected and 33.9% were detected. An additional case with an esophageal pouch was detected on magnetic resonance imaging (MRI) in the mid-trimester and a further two were detected on MRI in the third trimester. In total, 44.0% of cases of EA/TEF in our cohort were suspected, 33.3% were detected and 10.7% were suspected but, eventually, not detected prenatally.

CONCLUSIONS

Prenatal diagnosis of EA/TEF on ultrasound is not feasible before the late second trimester. A small/absent stomach may be visualized as early as 15 weeks' gestation. Polyhydramnios does not develop before the mid-trimester. An esophageal pouch can be detected as early as 22 weeks on a targeted scan in suspected cases. The detection rates of all three signs increase with advancing pregnancy, peaking in the third trimester. The early and mid-trimester anomaly scans perform poorly as a screening and diagnostic test for EA/TEF. © 2020 International Society of Ultrasound in Obstetrics and Gynecology.

Sonographic demonstration of fetal esophagus using three-dimensional ultrasound imaging

OBJECTIVE

To evaluate the feasibility of antenatal direct visualization of normal and abnormal fetal esophagus using three-dimensional ultrasound (3D-US) with Crystal Vue rendering technology.

METHODS

Between February and April 2018, 3D-US volumes were collected from a non-consecutive series of singleton pregnancies, referred for clinically indicated detailed prenatal ultrasound at 19-28 weeks' gestation to one of two fetal medicine units in Italy. 3D volumes were acquired from a midsagittal section of the fetal thorax and upper abdomen with the fetus lying in supine position. Postprocessing with multiplanar mode was applied to orientate the volume and identify the esophagus. The region of interest was angled by approximately 30° to the spine and its thickness was adjusted in order to optimize visualization of the intrathoracic and intra-abdominal course of the esophagus. Crystal Vue software was used for image rendering of the fetal trunk in the coronal plane. Postnatal follow-up was available in all cases.

RESULTS

During the study period, 91 pregnancies met the inclusion criteria and were recruited. The study cohort included two pregnancies with suspicion of esophageal atresia due to suboptimal visualization of the stomach. Of the 89 cases with normal stomach on two-dimensional (2D) imaging, 3D-US with Crystal Vue rendering technology allowed direct evaluation of the whole course of the esophagus in 74 (83.1%). In the two cases with small or absent stomach bubble on 2D imaging, esophageal atresia was demonstrated antenatally on 3D Crystal Vue imaging and was confirmed postnatally. The mean time required for offline postprocessing and visualization of the esophageal anatomy was 4 min.

CONCLUSIONS

Using 3D-US with Crystal Vue rendering, it is possible to visualize antenatally the normal fetal esophagus and demonstrate presence of esophageal atresia. This should facilitate prenatal counseling and management of cases with suspected esophageal atresia. Copyright © 2019 ISUOG. Published by John Wiley & Sons Ltd.