It remains controversial as to the exact etiology for the development of EA/TEF. Around the fourth week of gestation, the foregut begins to differentiate into a dorsal esophageal part and a ventral respiratory segment. Originally, it was felt that the ventral respiratory segment separates from the esophagus by the formation of lateral tracheoesophageal folds which fuse in the midline to create the tracheoesophageal septum. This separation of the trachea and esophagus is usually complete by six to seven weeks of gestation. If there is incomplete fusion of these tracheoesophageal folds, a defective tracheoesophageal septum could develop with an abnormal connection between the trachea and esophagus. However, recently, chick embryo studies have not demonstrated these lateral tracheoesophageal folds. Instead, cranial and caudal folds were found in the region of the tracheoesophageal separation. According to the second theory, EA/TEF could develop due to an imbalance in the growth of these cranial and caudal folds. A third
theory is based on a defect in intracellular signaling due to a deficiency in the sonic hedgehog (Shh) genes. In culture experiments, the fistula tract has been induced to branch by exogenous administration of the Shh signaling molecule and its receptor. These studies support the fistula as arising from the respiratory tract because the normal lung, but not normal esophagus, branches in response to the Shh signaling. Finally, from rat studies, a fourth theory postulates that EA/TEF might result from disturbances in either apoptosis or epithelial proliferation.

EA is usually a sporadic occurrence. The incidence of EA/TEF is about 1 per 3,500 births. There appears to be a slight male predominance of 1.3. Among parents who have one affected child, the risk of another affected child is between 0.5% and 2%. In addition, the risk of a child with EA/TEF born to a mother or father with EA/TEF is 3% to 5%. A number of etiologic agents have been identified including medications such as progesterone and estrogen, maternal diabetes, thalidomide exposure, the use of methimazole in early pregnancy, and the prolonged use of contraceptive pills, among others. Also, chromosomal anomalies including Trisomy 18 and 21 have been identified in 5% to 10% of the cases. Finally, defects in three different genes have been found to be associated with EA/TEF.

As previously mentioned, the VACTERL syndrome is the strongest syndromic association with the entire EA spectrum of conditions. The frequencies of these associated conditions are as follows: vertebral (6% to 20%), atresias of the intestinal tract (10% to 15%), cardiac (15% to 35%), renal (5% to 15%), and limb (5% to 20%). EA has also been seen occasionally in patients with the CHARGE syndrome (Coloboma, Heart anomalies, Atresia of the choanae, Retardation, Genital hypoplasia, and Ear deformities).

In a 2007 paper from our institution, we reviewed the incidence of the VACTERL association in 112 patients with EA anomalies seen between 1985 and 2005. There were 62 males and 50 females. In this review, 24.1% of the babies had vertebral anomalies, 14.3% had intestinal atresias, and 32.1% had cardiac anomalies, which was defined as an intrinsic heart lesion other than a patent foramen ovale or patent ductus arteriosus. The most common cardiac condition was a ventriculoseptal defect (VSD) found in 22.3% cases. However, a VSD was found as an insolated condition in only 7.1% of the patients. Cyanotic heart disease was found in only 4.5% of the babies and all were tetralogy of Fallot. Of these 112 babies, renal anomalies occurred in 17% and skeletal in 16.1%. A distal TEF was found in 95.5% of the babies.

In a term baby, EA or EA/TEF is rarely an emergency and there is usually adequate time to complete the preoperative workup and plan the operative approach. However, in the preterm infant with respiratory distress syndrome requiring mechanical ventilation, the sudden inability to ventilate the baby can be encountered due to the preferential flow of the ventilated gas through the endotracheal tube, down the trachea, and out through the distal TEF. This flow pattern results in poor pulmonary inflation and abdominal distention, both of which inhibit adequate ventilation. Gastric and duodenal perforations have been described. In such cases, a variety of maneuvers have been successful, including immediate gastrostomy with placement of the tube to underwater drainage, immediate thoracotomy and ligation of the fistula, and attempted placement of a Fogarty balloon past the endotracheal tube (ETT) and into the fistula, thereby occluding it. Isolated case reports on the management of babies with this complication can be found in the literature.

While the infant is awaiting operation, it is important to prevent complications resulting from aspiration. The baby should be placed in a head-up position and tilted to one side. Also, it is important that the nasoesophageal catheter is on constant suction. Finally, the baby’s mouth and pharynx should be suctioned as necessary, as not all the oral and esophageal secretions will be removed via the nasoesophageal catheter. If the baby requires intubation and ventilation, gentle low pressure ventilation is essential. If the fistula has to be ligated emergently (but is not divided), a delayed primary anastomosis should be performed within 10 days, as recanalization of the distal fistula is likely to occur.

It should be noted that the most important aspect for a thoracoscopic repair is adequate visualization. Impaired visualization often occurs due to the lung being inflated with traditional ventilation. Also, the anesthesiologist (or resident in a teaching hospital) may have limited experience with these cases and fail to understand the importance of low pressure, low volume ventilation to reduce the lung inflation. Moreover, hypercarbia can easily develop. To counteract the problem of lung inflation, CO₂ insufflation has been utilized as a positive pressure maneuver to decompress the lung to allow an optimal view. However, this is not always successful in keeping the lung deflated. Over the past 3 years, our group has utilized the high-frequency ventilator (HFV) in the operating room for ventilation in the babies undergoing the thoracoscopic approach. With the use of the HFV and proper positioning of the ETT, the lung remains deflated and an adequate operating space is created. Also, the duration of the operation is not as important because the oscillating ventilator is able to remove the CO₂ very efficiently and hypercarbia is not common. Assuming a healthy 3 to 3.5 kg infant, the initial settings are a mean airway pressure of 13 to 15 cm H₂O, a δP of 30 to 35, and FiO₂ of 70%. This can usually be weaned to 40% or so after 10 to 15 minutes. The major problem with using the HFV is with the movement (shaking) of the baby during the ventilation, but this is something that the surgeon becomes accustomed after the first few cases. The Hertz is usually set initially at 11 to 13 but can be decreased to 8 to 9 if the baby is shaking too much (1 Hz is defined as 60 breaths/s). With this background, we rarely perform these operations on the weekend, as our respiratory therapists are routinely involved with management of the HFV in the operating room.

The baby is positioned supine on the operating table and the oropharynx is suctioned. It is my preference to perform a preoperative bronchoscopy for the purpose of identifying the site of the distal fistula. Others may perform a bronchoscopy to exclude a proximal fistula. By knowing the location of the distal TEF, the surgeon has a better understanding of the gap length between the two esophageal segments and where the distal fistula inserts into the trachea. For instance, if the fistula is identified entering the carina (Fig. 3), the gap length between the distal and proximal segments could be relatively large (up to 3 to 4 cm). On the other hand, if the fistula is seen to enter the mid-trachea, the gap length will likely be much shorter. The distance between the two esophageal segments will also depend on how caudal the proximal esophageal pouch extends. This bronchoscopy takes less than 5 minutes and provides important information to the surgeon performing the repair.