Surgery for Congenital Lesions of the Esophagus

Historical Aspects

The historical background relevant to EA is thoroughly reviewed by Harmon and Coran. Durston in 1670 and Gibson in 1697 described the first cases of EA. It took approximately 250 years before the first cases of survivors were reported by Leven and Ladd independently in 1939. Both were able to achieve success by performing a series of operations, including gastrostomy, ligation of fistula, marsupialization of the upper pouch, and final reconstruction with an antethoracic skin tube. The early attempts at primary repair were all unsuccessful. It was not until 1941 that Haight reported the first survivor of a primary repair. In the decade that followed, it became evident that the mortality in infants of lower birth weights, those with severe associated anomalies, and those critically ill from aspiration pneumonia was very high. There followed a shift toward staging the operation for sick babies, with a gastrostomy, followed by division of the TEF and the esophageal reconstruction performed as a third stage. In 1962, Waterston proposed a classification based on birth weight, presence of pneumonia and associated anomalies. The 1970s and 1980 witnessed major advances in respiratory, neonatal, anesthetic, and surgical care, as well as introduction of more effective antibiotics. These advances included endotracheal intubation, which made it easier to prevent aspiration from the esophageal pouch and to deal with its sequelae. As a result, multiple groups started recommending either direct primary anastomosis (anastomosis shortly after birth) or delayed primary anastomosis (anastomosis delayed for the treatment of other life-threatening anomalies or stabilization of the patient) regardless of the patient's weight but based on physiologic criteria. The end of the 20th century ushered in the application of thoracoscopy to the repair of EA/TEF and other congenital anomalies of the esophagus.

Epidemiology

The average rate of EA is reported to be approximately 2.4 per 10,000 births. There is no significant described sex predilection. Other congenital anomalies occur in patients with EA frequently, ranging from 30% to 76%. This might be because the malformation in EA occurs early in the first trimester when there is active organogenesis. As a result, the developmental cause of EA/TEF might also affect other organ systems at the same time. The number of associated anomalies occurring in each patient increases with decreasing birth weight. With the improvement in anesthetic, respiratory, and neonatal techniques over the past few decades, the associated anomalies are now the major contributor to mortality in patients with EA. The most common associated anomaly is congenital heart disease, present in some form in approximately one fifth of patients. Approximately 20% of patients will have some combination of the constellation of anomalies referred to as VATER or VACTERL association: v ertebral, a norectal, c ardiac, t racheo e sophageal, r enal or r adial, and l imb anomalies.

Babies born with EA often have low birth weight and are premature. In one study, 90% of the patients with EA were below the 50th percentile for gestational age, and 40% were below the 10th percentile or small for gestational age (SGA). The growth retardation might be secondary to decreased absorption of the amniotic fluid protein or from a mechanical factor. Severe intrauterine growth retardation increases the mortality rate of the SGA neonate by 5 to 20 times that of appropriate-for-gestational age neonates of the same gestational age.

Anatomy

There are five types of EA with or without TEF. The types of EA are numbered differently depending on the classification scheme, so it is preferable to describe the actual anomaly rather than assign it a number or a letter: EA with distal TEF, EA without TEF, EA with proximal TEF, EA with proximal and distal fistula, and isolated TEF (H-type TEF) ( Fig. 35-1 ). The distribution of the different types in large series has been relatively uniform across decades and countries with the most common being EA with distal TEF ( Table 35-1 ). The fistula is usually small and most of the time arises from the midline of the membranous portion of the trachea just above the bifurcation; however, significant variations exist.

Presentation

A significant number of cases of EA are now suspected on prenatal ultrasonography, with polyhydramnios, absent or small stomach bubble, and visualization of an esophageal pouch in the neck being the most prominent features. Suspecting the diagnosis prenatally is invaluable in preparing the family. Prenatal counseling with a pediatric surgeon and a neonatologist and planning for appropriate delivery arrangements are extremely helpful. Postnatally, most patients with EA are diagnosed in the first few hours after birth. Choking with feeding, regurgitation of saliva and feeds, respiratory distress from aspiration of saliva or gastric contents through the TEF are the most common signs and symptoms. Inability to pass a feeding tube confirms the diagnosis.

Patients with isolated TEF (H-type TEF) may not be diagnosed until later in life. Recurrent episodes of aspiration pneumonia and choking and coughing with feedings should raise the suspicion. Contrast esophagram and rigid bronchoscopy are complementary in making the diagnosis, which is often very difficult to make.

Patients with isolated EA often have a scaphoid abdomen because of the absence of gas in the intestines. If EA is suspected, one should always look for other physical signs of the VATER association: anorectal malformations, limb anomalies, and vertebral defects ( Fig. 35-2 ).

Workup

A chest radiograph showing a curved catheter in the proximal esophageal pouch is often all that is required to make the diagnosis ( Fig. 35-3 ). In patients with isolated EA, the radiograph reveals absence of intestinal air ( Fig. 35-4 ). If still in doubt, a small amount of air injected into the pouch accentuates it on a plain radiograph and confirms the presence of EA ( Fig. 35-5 ). The use of barium to look for an upper pouch fistula should be discouraged because it may potentially lead to aspiration. Upper pouch fistulae are rare and are usually found at the time of repair by either bronchoscopy or a careful dissection of the proximal pouch. An esophagram performed via a catheter being pulled up along the esophagus while the patient is prone is invaluable in making the diagnosis of H-type fistula ( Fig. 35-6 ).

An echocardiogram, renal ultrasound, and vertebral films are often performed to rule out major cardiac, renal, and vertebral anomalies as part of the VATER association. The echocardiogram is also helpful in determining the location of the aortic arch and of any aberrant central vessels, which might alter the surgical approach. Recently, the need for a preoperative echocardiogram in asymptomatic patients has been questioned.

Initial Management

Patients born with EA are at risk for aspiration of saliva or gastric contents into the tracheobronchial tree. A Replogle-type soft sump suction catheter with the holes close to the tip should be placed in the upper pouch and put on continuous suction. In the presence of a TEF, the patient should be placed in the reverse Trendelenburg position with the head up to minimize reflux of gastric contents into the trachea. Having the patient prone might also help to keep the gastroesophageal junction at a less dependent position and decrease gastric reflux. With isolated EA, the Trendelenburg position facilitates the passive drainage of the hypopharynx and complements the active suction of the catheter. Even with a drainage catheter in place, frequent suctioning of the hypopharynx helps in decreasing the risk of aspiration. If there is any evidence of aspiration pneumonitis on a radiograph, broad-spectrum antibiotics should be started.

If possible, positive-pressure ventilation should be avoided in a patient with a TEF to minimize shunting through the TEF and abdominal distention. Because most fistulae are located just proximal to the carina, the tip of the tube should be kept high in the trachea to prevent the tip of the tube from becoming lodged in the TEF. Sometimes, the TEF is significant enough that adequate ventilation cannot be maintained, especially in the presence of respiratory distress syndrome (RDS), in premature patients, with its attendant high intraparenchymal pressures. In these cases, emergent ligation of the TEF or division of the esophagus might be warranted. A more difficult way to control the TEF emergently is obliteration of the TEF with a Fogarty balloon introduced via bronchoscopy. A hazardous situation can occur when a patient with significant steal through a TEF also has a very high intestinal obstruction, such as duodenal atresia. The massive gastric distention exacerbates the respiratory compromise and could lead to perforation of the stomach. Emergent gastric decompression has to be performed, sometimes at the bedside with a needle.

The timing of surgery is of particular interest. The key issues affecting the timing include severity of the infant's condition, associated abnormalities, and birth weight. The Spitz classification identified low birth weight (<1500 g) and major congenital heart disease as two major risk factors associated with EA. This system stratifies patients into groups. Survival was estimated at 97% for group I (weight > 1500 g with no major congenital cardiac defect), 59% for group II (birth weight < 1500 g or major congenital cardiac defect), and 22% for group III (birth weight < 1500 g and major congenital cardiac defect). In 2006, Lopez validated the original Spitz classification in a series of 188 neonates. Weight-based classification systems such as the Spitz and Waterston classifications have traditionally been used to assess prognostic outcomes with surgical intervention. However, others have not found weight to be an independent predictor of survival.

Several groups investigated the issue of operating on extremely low–birth-weight (ELBW, weight < 1000g) infants and very low–birth-weight (VLBW, weight < 1500 g) infants and the ideal surgical management of these patients is still controversial. In 2006, Seitz and colleagues published their data of four children who underwent open primary repair of EA/TEF whose weight ranged from 780 to 1120 grams. Their postoperative complications included one patient with anastomotic leak and a separate patient with esophageal stenosis requiring dilation. They concluded that a one-stage approach is a good option in ELBW children. In 2009, Petrosyan and colleagues published a retrospective review of 25 VLBW infants with EA/TEF who underwent open operative repair between the years of 1987 and 2008. The researchers divided these patients into two groups: 16 of these patients underwent primary repair and 9 underwent staged repair. There was no difference in patient characteristics, including weight or mean gestational age. However, there was a significantly higher rate of anastomotic leak (50% versus 0%), strictures (81% versus 33%), and sepsis/pneumonia (81% versus 66%) in the primary repair versus staged repair groups, respectively. They concluded that there are increased risks of postoperative complications in premature VLBW infants with EA/TEF and recommended a staged surgical approach in these patients, which is consistent with the series of Chahine and Ricketts.