Performance of prenatal diagnosis in esophageal atresia.

OBJECTIVE The aim of this study was to evaluate the performance of prenatal diagnosis of esophageal atresia (EA) and its associated abnormalities. METHODS We conducted a retrospective study from a pediatric database of EA managed postnatally in a single center. Prenatal data included ultrasound and magnetic resonance imaging parameters including amniotic fluid (AF) volume, stomach visualization, AF biochemistry, and associated malformations. Postnatal data included type of EA, mortality, and postnatal diagnosis of associated malformations. RESULTS One hundred twenty-two cases were included. The diagnosis was suspected prenatally in 39/122 (32%) cases. Polyhydramnios was noted in 64/122 (52.4%), and the stomach was not visualized or small in 39 (32%). There was 14 (11.5%), 2 (1.6%), 101 (82.8%), 5 (4.1%), and 0 (0%) types I, II, III, IV, and V, respectively. EA was suspected prenatally in 12/14 (85.7%) in type I and in 27/108 (25%) in cases with tracheoesophageal fistula (II + III + IV + V). Magnetic resonance imaging was performed in 28 cases, which confirmed EA in 19/28 (sensitivity 67.8%). AF biochemistry was performed in 17 cases, which confirmed EA in 15/17 (sensitivity 88.2%) cases. Of the 69 syndromic associations, 41/69 (59.4%) cases were detected prenatally. Associated malformation was a strong predictor of postnatal death [19/69 vs 3/53, odds ratio 6.33 (1.76; 22.75), p < 0.01]. CONCLUSION Prenatal diagnosis of EA remains challenging. MRI and AF biochemistry may prove useful in the diagnosis of EA. Prenatal ultrasound and MRI examination should also focus on associated anomalies. © 2015 John Wiley & Sons, Ltd.