Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Combined anomalies of the esophagus and trachea are regular occurrences in most tertiary neonatal units caring for high-risk infants. The spectrum of recognized deformities comprises a variety of combinations involving esophageal atresia (EA) and tracheoesophageal fistula (TEF). These malformations are well-recognized entities, yet they continue to stimulate a great deal of interest for a variety of reasons, and almost all are lethal unless surgically corrected. Many have concurrent pulmonary complications, and most are associated with other congenital anomalies that require careful coordination of interdisciplinary care. There are few conditions, if any, that require a greater degree of cooperation between neonatologist and surgeon.
The reported incidence of EA/TEF is roughly 1-2 per 3500 live births. The incidence of EA/TEF is higher in white populations than in nonwhite. There is also a slight preponderance of males and a disproportionate rate of twinning among affected infants. Although the majority of cases are sporadic, there are numerous well-recognized genetic associations. Approximately 7% of affected infants have a chromosomal abnormality such as trisomy 13, 18, or 21. Additionally, EA/TEF has been reported in infants with the Pierre-Robin, DiGeorge, Fanconi, and polysplenia syndromes.
Although usually sporadic, most cases of EA/TEF are not isolated. Most affected infants, up to 70% in some series, have at least one other anomaly. The likelihood of coexisting anomalies is considered to be greatest with pure EA and least with pure TEF. The most widely recognized relationship between EA/TEF and other congenital anomalies is within the constellation of defects referred to as the VACTERL association—abnormalities involving v ertebral, a norectal, c ardiac, t racheal, e sophageal, r enal/genitourinary, and l imb structures. There is a lesser association with CHARGE s yndrome ( c oloboma, h eart defects, a tresia choanae, r etarded development, g enital hypoplasia, e ar defects/deafness). Gastrointestinal anomalies (other than anorectal defects) associated with EA/TEF include duodenal atresia, annular pancreas, jejunoileal atresia, and intestinal malrotation. Neurologic defects, especially hydrocephalus, may be associated with EA/TEF, as may diaphragmatic hernia and abdominal wall defects. In order of approximate frequency, the affected organ systems most commonly associated with EA/TEF are cardiovascular (35%-45%), gastrointestinal (25%), genitourinary (25%), skeletal (15%), and neurologic (10%).
The anatomy of EA/TEF includes five well-recognized variants. A number of anatomic classifications have been devised but have been largely discarded in favor of simple descriptive nomenclature. The most common variant is the combination of a proximal esophageal atresia and a distal tracheoesophageal fistula. The proximal atresia/distal fistula variant occurs in about 85% of affected patients. It manifests classically as a dilated, blind-ending proximal pouch that extends to the lower neck or upper mediastinum, and a distal esophageal segment that originates from the posterior membranous wall of the trachea, carina, or main stem bronchus and connects to the stomach in a normal fashion. The next two most common variants are isolated EA and isolated TEF, both of which occur in approximately 5%-10% of patients. An isolated TEF is often referred to as an H-type fistula; the fistula itself is usually angled downward, with the esophageal end inferior to the tracheal end. The rarest variants include the proximal fistula/distal atresia, and the double fistula, both occurring in no more than 1% or 2% of patients.
No single unified model of pathogenesis has been developed to explain a variety of anomalies seen with esophageal atresia. Normal development of the esophagus and trachea includes the formation of a primordial lung bud as a diverticulum of the ventral foregut during the 4th week of gestation. The appearance of this diverticulum is associated with a pair of lateral infoldings of the foregut—the laryngotracheal folds—which begin at the caudal end of the lung bud and fuse to form the tracheoesophageal septum. This process of lateral invagination and fusion was historically believed to proceed cranially, displacing the orifice of the lung bud in a cephalad direction and separating the developing trachea from the esophagus as the tracheal bifurcation moved caudally in a relative sense. Therefore, in the traditional model for the cause of tracheoesophageal malformations, the esophagus and trachea shared a common foregut precursor over a significant distance, becoming separated by the cephalad movement of the tracheoesophageal septum. Perturbations of this process would account for the observed variety of abnormal connections between these two adjacent structures. This theory of cephalad migration of a tracheoesophageal septum is now controversial. Evidence suggests that the position of the laryngeal orifice remains constant relative to the developing notochord and that the tracheal bifurcation descends simply as a result of linear tracheal growth without any cephalad movement of a tracheoesophageal septum.
Given this alternative scheme of tracheoesophageal development, a different etiologic theory is required to explain tracheoesophageal anomalies. One such theory has arisen from observations in the Adriamycin-treated rat model of VACTERL deformities. In this model, pregnant rats are exposed to Adriamycin during early gestation and give birth to offspring that express VACTERL phenotypes, including EA/TEF. Investigations have demonstrated that the distal “esophageal” segment in EA/TEF arises from a respiratory precursor, descends from the carina along with the two main stem bronchi, and elongates caudally to merge with the stomach. A respiratory origin for this third structure is supported by the finding of pseudostratified respiratory epithelium and respiratory-specific thyroid transcription factor-1 (TTF-1) in distal fistulas from Adriamycin-treated rat model sources, and of TTF-1 expression in distal fistula specimens from human sources. This structure does not branch as the developing bronchi do, because of abnormal epithelial-mesenchymal interactions caused by deficiencies of mesenchymal fibroblast growth factor and epithelial FGF receptors. In support of this hypothesis, findings document a brief stage in the Adriamycin-treated rat model during which the stomach is anatomically disconnected from the rest of the developing foregut.
If current proposals of a respiratory-derived distal fistula are accurate, the proximal atresia must, therefore, have a separate etiology. One proposed explanation is an abnormal concentration gradient of the early embryonic morphogen known as sonic hedgehog (Shh), which is elaborated by the developing notochord and is believed to participate in early foregut differentiation. Notochord abnormalities have been documented in the Adriamycin-treated rat model and in specimens of human distal fistulas that had been shown to be specifically deficient in Shh. A unifying theory that links these newer models and provides a comprehensive explanation for tracheoesophageal maldevelopment is still lacking.
The clinical presentation of infants with EA/TEF depends on the specific anatomic variant encountered. Many cases of EA/TEF are not diagnosed prenatally, but current prenatal imaging techniques may raise a suspicion of EA/TEF in a significant number of infants. The combination of maternal polyhydramnios and a small or absent stomach has both sensitivity and a positive predictive value for EA/TEF of approximately 40%-60% in most series. The finding of a dilated upper pouch in the neck by high-resolution ultrasonography or magnetic resonance imaging can increase the diagnostic accuracy in this selected group of patients to nearly 100%.
In most affected infants, the diagnosis is made in the immediate postnatal period. The infant cannot swallow secretions and appears to drool excessively. As the upper pouch fills, antegrade aspiration may occur, leading to significant coughing, respiratory distress, and cyanosis. In patients with a distal fistula, the stomach may dilate with air, leading to the reflux of gastric secretions back into the lungs, resulting in significant and progressive respiratory distress caused by reactive bronchoconstriction and chemical pneumonitis.
Attempted passage of a gastric tube is usually diagnostic. The tube fails to pass distally and extrudes back through the mouth. A simple chest radiograph usually establishes the diagnosis by demonstrating the radiopaque catheter curled in the upper pouch ( Fig. 84.1 ). If the diagnosis is still in question, the upper pouch should be cautiously evaluated with water-soluble contrast by an experienced imager, because aspiration can cause parenchymal lung injury. Air itself makes an excellent contrast agent, and the catheter can be pulled back under fluoroscopic guidance until the tip is positioned in the proximal esophagus, and then 5-10 mL of air is injected slowly. A radiolucent dilated pouch is then easily documented. If there is no luminal gas below the diaphragm, the anomaly can be further defined as a pure or distal atresia ( Fig. 84.2 ). The clinical presentation of an H-type TEF is typically much more subtle and delayed. It usually includes a history of coughing during feedings and occasionally the development of pneumonia owing to aspiration through the fistula. An H-type TEF may be very difficult to confirm radiographically. Contrast must be injected through a tube positioned in the thoracic esophagus, with the patient prone and in the Trendelenburg position, allowing the contrast to flow back up the angled fistula. Often, an H-type fistula is documented only on rigid bronchoscopy during evaluation for respiratory distress.
Once the diagnosis is established, efforts are made to prevent aspiration-induced lung injury, to provide respiratory support if necessary, and to define associated anomalies while preparing the patient for surgical correction. Gastric acid blockade should be instituted while the infant is maintained in a slightly head-up position to reduce reflux of gastric secretions into the airway. A small-caliber sump tube is kept in the upper pouch and suctioned intermittently. The use of preoperative broad-spectrum antibiotics during the first several days should be considered. Spontaneous ventilation is preferred to minimize the introduction of air into the stomach, as gastric decompression is possible only by percutaneous needle aspiration. This is particularly important in the presence of distal intestinal obstruction. The use of continuous positive airway pressure is discouraged. If parenchymal lung disease mandates the use of positive-pressure ventilation, mean airway pressures are minimized to avoid shunting of tidal volume through the fistula. This may be difficult in patients with lung disease and poor compliance. Under these circumstances, the use of high-frequency oscillatory ventilation (HFOV) may allow adequate gas exchange to occur at lower peak airway pressures, preventing the loss of ventilation through the fistula. The failure or unavailability of HFOV in a patient requiring aggressive ventilatory support may be an indication for urgent surgical intervention to ligate the fistula, ligate the gastroesophageal junction, or occlude the fistula with a balloon catheter.
A thorough evaluation for other VACTERL-associated defects is mandatory. A preoperative radiograph provides much information regarding cardiopulmonary status, diaphragmatic integrity, and the presence of vertebral abnormalities. The only other studies necessary before surgical management of the EA/TEF are an echocardiogram in all patients and a renal ultrasound. Echocardiography not only identifies significant cardiac anomalies that may influence anesthetic management but also establishes the presence of aortic arch anomalies that may alter the surgical approach.
Patients with respiratory compromise and EA/TEF can present a particular challenge to the care team. Endotracheal intubation and the conversion to positive pressure ventilation can often lead to worsening respiratory status, because frequently the path of least resistance is through the fistula and into the GI tract. This leads to inadequate pulmonary ventilation. At the same time, the stomach becomes grossly distended, increasing intrathoracic pressure and further decreasing bronchial airflow. These children may require emergent bedside gastric decompression or fistula ligation to stabilize the pulmonary status. Any decision to place EA/TEF patients on positive pressure ventilation must be discussed among all team members.
The nature and timing of surgical intervention depend on the specific anatomic variant being treated. Surgical strategies include immediate primary repair, delayed primary repair, staged repair, and esophageal replacement. Immediate primary repair is appropriate for the majority of infants with EA/TEF. Rigid bronchoscopy immediately before surgical repair is often a useful adjunct to confirm the diagnosis before thoracotomy and to obtain useful information; the location of the fistula can be determined, the presence of a second upper pouch fistula can be detected, and the structural status of the trachea can be assessed ( Fig. 84.3 ).
An open repair is approached through a right extrapleural thoracotomy, or through the left side if preoperative echocardiography documents a right-sided aortic arch. An extrapleural approach is preferred to avoid a pleural empyema if an anastomotic leak subsequently develops, but this is more of a historic and not evidence-based decision. The division of the fistula is the key step of the procedure. Fistula division allows stabilization of the pulmonary mechanics. The fistula is divided and the tracheal side closed transversely to prevent narrowing of the trachea. A primary anastomosis between the proximal esophageal pouch and the distal esophageal segment is then performed.
Thorascopic repair of EA/TEF, although technically challenging, has become a standard mode of repair especially in full-term infants with relatively normal respiration. Thorascopic technique typically uses two 3-mm ports and one 5-mm port. Advocates of thorascopic repair have been able to demonstrate improved cosmoses, improved pain control, improved postoperative respiration, chest wall maldevelopment, and scoliosis, all known complications of neonatal thoracotomy. Leak and stricture rates are similar to open repair.
Compared with open repair (OR), a longer operative time has been associated with a thoracoscopic approach (TR), although the TR procedure could possibly reduce the length of hospital stay, time to extubation, and first oral feeding time. Meanwhile, the occurrence rates for leaks, strictures, and pulmonary complications; the fundoplication rate of GERD; and blood loss were similar between the OR and TR groups.
In most patients, a gap of some degree exists between the proximal and distal esophagus, requiring mobilization of both ends and still resulting in some degree of tension at the anastomosis. Anastomotic ischemia caused by overaggressive mobilization or excessive tension leads to fibrosis and stricture formation. The blood supply to the cervical esophagus originates proximally and extends distally through the submucosal plexus, allowing extensive mobilization of the upper pouch without causing ischemia at the distal tip. The blood supply to the normal thoracic esophagus derives from the intercostal vessels and is, therefore, more segmental. Whether or not this limits the ability to mobilize the distal fistula without causing ischemia is debated. If careful mobilization does not sufficiently reduce tension at the anastomosis, two options are available. One option is to lengthen the upper pouch by performing a circular or spiral myotomy. A second option is to employ a technique first described by Foker that obtains length by placing the two ends under tension for a prolonged period—up to 3 weeks in some patients—prior to primary anastamosis. Although met with great resistance when first described, it is now an accepted option and considered the first choice for repair of long-gap atresia by many surgeons.
The postoperative care of patients with primary esophageal repairs must be meticulous and conservative. Removal of the endotracheal tube should occur only after respiratory mechanics have been optimized, as failure of extubation may lead to aggressive bag-mask ventilation and reintubation—events that place the tracheal and esophageal repair at risk for disruption. The head of the bed is elevated and gastric acidity neutralized to minimize reflux injury to the healing anastomosis. After 5-7 days, a contrast esophagram is carefully performed with a water-soluble agent followed by barium. If no leak is demonstrated, oral feedings are initiated and the chest tube is removed. Small, asymptomatic leaks that are adequately controlled by the existing extrapleural chest tube may be managed nonoperatively and repeat studies performed until healing is documented. Large leaks and intrapleural leaks often mandate exploration for repair or diversion.
In many clinical situations, the described primary approach is neither feasible nor appropriate. In the premature infant with very low birth weight, or in any infant with a significant cardiac defect, respiratory compromise, or sepsis, primary anastomosis should be postponed until the patient is a better surgical candidate. If the anticipated surgical delay is brief, careful upper pouch suctioning and parenteral nutrition may be used until the delayed primary repair can be carried out. If a lengthy or indeterminate delay is expected, a staged approach may be required. This strategy provides temporizing measures that allow later esophageal repair to be undertaken electively. Typically, a gastrostomy tube is placed for decompression and feeding, and the fistula is divided via an extrapleural approach to protect the lungs from further injury. Later, a transpleural approach may be taken to restore esophageal continuity.
The list of postoperative complications that occur in this complex group of patients is long and varied. Beyond the immediate postsurgical period, when anastomotic leak is the most feared problem, anastomotic strictures represent the most common complications of surgical treatment, occurring in up to 40% of patients in some series. Strictures result from fibrosis during healing, which in turn results from ischemia, excessive tension, leakage, or acid-peptic injury. An anastomotic stricture should be suspected whenever dysphagia or respiratory symptoms occur in a patient who had previously tolerated oral feedings. A contrast esophagram reliably demonstrates the stricture and often demonstrates distention of the upper pouch with posterior compression of the trachea—the so-called upper pouch syndrome ( Fig. 84.4 ). Strictures are best treated by tangential dilation under fluoroscopic guidance and control of acid reflux. Near-occlusive strictures require placement of an indwelling transanastomotic guide line that can be used to pull sequential dilators safely through the stricture on a frequent basis. If pharmacologic suppression of gastric acid secretion fails to prevent recurrent strictures, an antireflux procedure may be required. Finally, a stricture refractory to aggressive management may require segmental resection.
Recurrent fistulas occur in 5%-10% of cases and usually present with respiratory distress during feeding or with recurrent aspiration pneumonia. Diagnosis is made with a dilute barium esophagram in the prone position. Most recurrent fistulas result from small, contained anastomotic leaks that cause chronic inflammatory changes and gradually erode back through the tracheal repair. Surgical options range from primary closure to segmental esophageal resection, and they must include interposition of healthy, well-vascularized soft tissue such as a pleural or strap muscle rotation flap. Some repairs can be approached from a cervical incision, which limits the complications associated with secondary leaks.
Functional disturbances of the esophagus are nearly universal after repair of EA/TEF. Occasionally, dysphagia is associated with severe esophageal dysmotility in the absence of a stricture. Solid-phase esophagography demonstrates that solid foods have great difficulty traversing the anastomosis and lower esophagus because of a lack of peristaltic force. In these situations, dietary modification and adjustment of feeding behavior may be all that can be offered. Gastroesophageal reflux (GER) disease is present to some degree in most of these patients and may be clinically significant in up to 50% of patients with tracheoesophageal malformations. Some advocate lifelong follow-up care. Deficient autonomic and enteric innervation, intrinsic developmental abnormalities of the lower esophageal muscle, shortening of the intra-abdominal esophagus, and distortion of the angle of His may all contribute to lower esophageal sphincter dysfunction in varying degrees. Lower esophageal sphincter incompetence is exacerbated by esophageal dysmotility, reducing clearance of gastric acid from the esophagus. Some degree of tracheomalacia is also expected in all patients with EA/TEF. Most patients exhibit the typical raspy cough caused by vibration of the weak and flattened tracheal wall, and they outgrow these minor symptoms with age.
In patients with isolated EA, a long gap is anticipated, and primary repair, immediate or delayed, is not feasible. Historically, long-gap EA had been an absolute indication for esophageal substitution without attempts at staged primary repair. Over the past two decades, however, the proportion of patients with pure atresia who eventually undergo successful esophageal repair has increased dramatically, obviating esophageal substitution in a large number of children. There is no universal agreement regarding the maximal gap that will permit esophageal repair. However, a fluoroscopically measured gap of greater than four vertebral bodies with the segments in neutral position, or greater than two vertebral bodies ( Fig. 84.5 ) with the segments stretched toward each other, portends a low likelihood of successful primary anastomosis. Elongating the upper and lower pouches over a 1- to 3-week period, followed by primary anastomosis, has been very successful at keeping the native esophagus and avoiding esophageal replacement. Fig. 84.6 and Fig. 84.7 show preoperative and postoperative contrast images of a patient with a long-gap atresia of nearly seven vertebral bodies treated by elongation and primary repair. Reflux requiring fundoplication is common in these patients, as are dilations to relieve strictures. However, when compared with the multiple, significant complications associated with esophageal replacement using either small or large bowel, the Foker procedure requires strong consideration in all patients with long-gap esophageal atresia. Esophageal replacement is reserved for those cases of long-gap atresia unsuitable for immediate or delayed primary repair, and when attempted primary repair has failed irretrievably because of leakage or stricture.
Long-term outcomes for all variants of EA/TEF have improved steadily over the last three to four decades. Waterston and colleagues were the first to stratify patients on the basis of risk factors shown to affect prognosis and to recommend treatment strategy based on this stratification. In this analysis, prognosis was dependent on birth weight, associated anomalies, and the presence of pneumonia. Many stratification schemes subsequently evolved, and Spitz and associates refined this prognostic model to include only birth weight and the presence of significant congenital heart disease ( Table 84.1 ). In the current era, most infants born with tracheoesophageal anomalies ultimately experience a positive outcome owing equally to refinements in surgical technique and to neonatal support over the past several decades.
Group | Characteristics | Survival |
---|---|---|
I | Birth weight >1500 g without CHD | 97% |
II | Birth weight <1500 g or CHD | 59% |
III | Birth weight <1500 g and CHD | 22% |
Esophageal duplication cysts are uncommon causes of esophageal obstruction during infancy. These structures result either from abnormalities in the vacuolization process that re-establishes the esophageal lumen after obliterative epithelial proliferation during early embryonic development, or from the “budding” and separation of a portion of the developing foregut. These structures are usually cystic, but they may be tubular and may be located within the muscular wall of the esophagus or exist separately in the posterior mediastinum. They contain an epithelial lining derived from any foregut structure: squamous, columnar, or pseudostratified ciliated respiratory epithelium. They may also contain gastric epithelium and pancreatic or adrenal rests.
Infants with esophageal duplication cysts are usually asymptomatic, and the diagnosis is made when chest radiography unexpectedly demonstrates a mediastinal mass. Some infants, however, experience feeding difficulties or respiratory compromise if the cystic structure compresses the adjacent esophagus or trachea. Cysts containing gastric mucosa may present with complications of acid-induced injury: upper gastrointestinal hemorrhage, ulceration, perforation, or erosion into the bronchial tree.
Although the diagnosis may be suggested by a plain chest radiograph or by contrast esophagram, computed tomography is definitive and helpful in planning the operative approach ( Fig. 84.8 ). Magnetic resonance imaging provides additional information about the status of the spinal cord, which may be abnormal; this information should be obtained in all patients with a vertebral abnormality or a cystic structure in close proximity to the spine. Esophagoscopy is unnecessary and potentially harmful.
Surgical treatment involves resection of the cyst and repair of any esophageal defect. This may require a thoracotomy, but many lesions lend themselves to thoracoscopic resection. Cysts complicated by infection with abscess formation or by internal hemorrhage may expand rapidly and may require urgent drainage before resection. Duplications that are deemed unresectable because of their extensive nature, which may include extension into the abdomen, should be treated by stripping the mucosa from the duplication and leaving the muscular remnant to heal to the existing esophageal wall. Excellent results can be expected in most patients.
See Chapter 66 .
The most common abdominal wall defects seen in neonates are omphalocele and gastroschisis, occurring in approximately 4 per 10,000 live births. These conditions result from different developmental miscues and manifest as distinct clinical entities. In the case of omphalocele, a central abdominal wall defect of variable size is covered by a domelike mesenchymal membrane composed of amnion. The umbilical cord connects to the central portion of this membrane ( Fig. 84.9 ). The underlying abdominal organs are protected from exposure to amniotic fluid. In gastroschisis, the defect is usually smaller and located to the right of the umbilical attachment. There is no protective membranous covering. The abdominal contents, therefore, are eviscerated and suspended in amniotic fluid during gestation ( Fig. 84.10 ).
Omphalocele results from a failure in the folding mechanism that converts the flat trilaminar germ disc into a complex “tubular” structure starting at about 5 weeks’ gestation. The lateral body folds and craniocaudal folds converge at the umbilical ring, which contracts, closing the ventral abdominal wall. In patients with omphalocele, this ring fails to contract and leaves a round defect of variable size and a corresponding sac composed of amnion. The liver and small intestine usually occupy a portion of the sac, along with a variable amount of other abdominal contents. The underlying failure of umbilical ring closure may be related to aberrant development and migration of abdominal wall muscular components, or to failure of the developing midgut to return to the abdominal cavity after a period of herniation into the umbilical stalk. Rarely, the defect is cephalad to the umbilicus, producing a complex deformity referred to as the pentalogy of Cantrell: sternal, diaphragmatic, and pericardial defects; upper abdominal omphalocele; and ectopia cordis ( Fig. 84.11 ). When centered below the umbilicus, bladder or cloacal exstrophy may occur. The cause of gastroschisis is equally unclear, but it may involve a rupture of the umbilical stalk during the period of midgut herniation. Gastroschisis is usually described as an abdominal wall defect, to the right of a normally inserted umbilical cord, without membranous covering of the extruded organs. Etiologically it has been suggested that gastroschisis represents a failure in the normal attachment between umbilical cord and umbilical ring. Competing theories involving thromboembolic infarction of the developing abdominal wall or a failure of mesenchymal migration are less convincing.
The factors influencing morbidity and survival in these two distinct conditions are very different. Omphalocele, with a stable incidence of about 2-2.5 per 10,000 live births, is associated with other structural or genetic defects in 50%-75% of affected infants. These associated anomalies, which may involve the cardiovascular, gastrointestinal, genitourinary, or central nervous systems, account for most of the morbidity in these patients. Because the abdominal contents are protected throughout gestation, little morbidity accrues from injury to the intestinal tract. The discrepancy between the volume of eviscerated abdominal organs and the size of the abdominal cavity—the “loss of domain”—accounts for the other major source of morbidity in these patients. In a giant omphalocele ( Fig. 84.12 ), eventual closure of the abdominal wall by any means may be challenging, leading to a variety of problems related to structural integrity of the torso, chronic ventral hernias, and posture and gait development. Additionally, infants with giant omphaloceles may have a high incidence of pulmonary hypoplasia, resulting in respiratory compromise and pulmonary hypertension.
The syndromes most often associated with omphalocele include the VACTERL association, Beckwith-Wiedemann syndrome (macrosomia, macroglossia, visceromegaly, hemihypertrophy, hypoglycemia, renal pathology), EEC syndrome (ectodermal dysplasia, ectrodactyly, cleft palate), and OEIS complex (omphalocele, exstrophy, imperforate anus, spinal defects). An oft-reported association of omphalocele with cryptorchidism is unsubstantiated. Chromosomal abnormalities, including trisomies 13, 18, and 21, occur in 25%-50% of affected patients. The presence of a small sac, the absence of liver in the sac, and the presence of other malformations strongly predict an abnormal karyotype.
Gastroschisis, in contrast, is sporadic in the vast majority of cases. The incidence of gastroschisis is approximately 1.5 per 10,000 live births and increasing. Risk factors for gastroschisis include young maternal age, lower socioeconomic status, and exposure to external agents, such as vasoconstricting decongestants, nonsteroidal anti-inflammatory agents, cocaine, and possibly pesticides/herbicides. The 5%-20% incidence of associated defects is lower than for omphalocele, and it represents mostly intestinal atresias directly associated with ischemic or mechanical injury to the eviscerated bowel during gestation. Abnormalities not directly related to the abdominal wall defect are uncommon.
Morbidity in infants with gastroschisis, as opposed to those with omphalocele, is almost entirely related to intestinal dysfunction caused by in utero injury to the eviscerated bowel. The spectrum of injury displayed by the eviscerated bowel in gastroschisis ranges from mild to catastrophic. Morphologically, the bowel appears edematous, matted, and foreshortened. Often, the mass appears to be contained within an inflammatory rind—the “peel” of gastroschisis ( Fig. 84.13 ). Histologically, the intestine is characterized by villous atrophy and blunting, submucosal fibrosis, muscular hypertrophy and hyperplasia, and serosal inflammation. The cause of injury is unclear, but it is probably related to a combination of two separate insults. Exposure to amniotic fluid appears to be a major contributing factor, as amniotic fluid exchange can prevent peel formation. The second cause of injury may be the partial closure of the defect around the base of the eviscerated intestinal mass. This causes progressive constriction around the intestinal mesentery, resulting in the obstruction of luminal, lymphatic, and venous outflow. Intra-abdominal bowel distension is associated with increased postnatal complications, including delay to full feeds and increased duration of hospital stay in infants with prenatally diagnosed gastroschisis; however, this association seems to be limited to those with multiple loops of dilated intra-abdominal bowel.
The functional consequences of these structural changes include impaired absorption, reduced brush border enzyme synthesis, and a prolonged motility disorder related to rigidity of the bowel wall and possible derangements in the production of enteric neurotransmitters such as nitric oxide. These changes concur with the clinical observation that infants with a dense peel suffer prolonged gastrointestinal dysfunction that requires precise nutritional management. This nutritional failure and the complications arising from prolonged enteral and parenteral nutritional therapy constitute a significant proportion of the adverse clinical outcomes in these patients.
The prenatal diagnosis of abdominal wall defects by fetal ultrasonography is well established ( Fig. 84.14 ). Any uncertainty in distinguishing omphalocele from gastroschisis may be eliminated by measuring amniotic fluid α-fetoprotein levels, which should be elevated in gastroschisis only. A prenatal diagnosis of omphalocele should prompt a thorough sonographic survey of the entire fetus to evaluate for associated anomalies. Chromosomal analysis may also be helpful in determining postnatal management and prognosis.
The obstetric decisions related to timing and route of delivery continue to engender considerable debate. In the case of omphalocele, if the membrane is intact, the pregnancy should be carried to term if possible, because early delivery has no theoretical benefit for the fetus. Cesarean delivery has historically been recommended to prevent rupture of the omphalocele membrane ( Fig. 84.15 ), which would necessitate emergent surgical intervention without the benefit of preoperative evaluation and stabilization. Studies, however, have suggested that the route of delivery has no effect on morbidity or prognosis in abdominal wall defects in general. Recommendations regarding the route of delivery for a fetus with a giant omphalocele or associated anomalies should be individualized.
Early delivery after lung maturation has been enthusiastically endorsed by some, but definitive evidence that this strategy confers any statistically verifiable benefit is lacking. A more selective approach in which early delivery is undertaken when sonographic surveillance suggests progressive bowel injury, as defined by bowel dilation and wall thickening, has yielded numerous conflicting reports. Until the efficacy of selective early delivery for gastroschisis has been studied in a prospective, randomized fashion, support for this strategy will remain anecdotal. Regardless of the timing of delivery, almost all infants with gastroschisis may be delivered vaginally without increased injury to the bowel.
The surgical goal of establishing complete fascial and skin closure without causing further injury to the underlying bowel is common to patients with both omphalocele and those with gastroschisis. The surgical strategies applied to these two conditions are, however, quite different. In patients with gastroschisis, urgent closure or coverage of the defect is of the highest priority to limit intestinal injury and reduce morbidity. The eviscerated intestine should be completely covered in the delivery room to prevent water and heat loss through evaporation, conduction, and convection. Temporary coverage can be provided by wrapping the torso with transparent plastic film or by placing the baby in a transparent surgical “bowel bag” and cinching the drawstring closed gently under the axillae. This arrangement effectively limits heat and water loss, and it allows the intestine to be visualized at all times so that inadvertent volvulus and ischemia can be detected and reversed. A lateral decubitus position is often better tolerated than supine. A gastric decompression tube is placed immediately to prevent intestinal dilation. Great care must be taken to keep the bowel directly above the belly and not draped to one side or the other while waiting for surgical repair. This keeps the vessels supplying the exposed bowel—the superior mesenteric artery and vein—from kinking and further compromising the bowel.
Many infants with gastroschisis are born prematurely, and aggressive respiratory care, including supplemental oxygen, endotracheal intubation, and intratracheal surfactant, may be necessary. Intravenous access is established for fluid resuscitation and administration of broad-spectrum antibiotics. As soon as the baby is physiologically stable, surgical management is advisable. Unnecessary delay only compounds the problems of further bowel swelling and possible ischemia.
Gastroschisis can be managed by either primary or staged closure, depending on the size discrepancy and the infant's physiologic status. Although 60% to 70% of gastroschisis cases can be closed primarily, the decision to do so is individualized in each circumstance. As the intestine is returned to the abdominal cavity and the abdominal fascia approximated, increased abdominal pressure may prevent safe, complete closure. In infants with preexisting pulmonary compromise, the restriction of diaphragmatic excursion may decrease extrinsic compliance and cause ventilatory pressures to rise to unacceptable levels. Increased abdominal pressures may also impair mesenteric, hepatic, and renal perfusion. Trials of a “plastic closure” for gastroschisis have reported good success. In this technique, the bowel is returned to the abdomen in the NICU, and the umbilicus is placed over the defect with a dressing. The defect closes on its own without the need for sutures. A number of these patients have a resulting umbilical hernia, most of which close spontaneously with time.
When respiratory problems or abdominal pressures prevent safe primary closure, placement of a temporary prosthetic “silo” allows for more gradual reduction of the eviscerated intestine into the abdominal cavity and delayed primary closure at a later time ( Fig. 84.16 ). Previous retrospective studies had documented improved survival with primary closure, prompting an era marked by an aggressive approach to immediate closure. Appreciation of the fact that this survival advantage simply represented a selection bias, along with recognition of the deleterious effects of high intra-abdominal pressures, has led to a more liberal use of temporary silo coverage. Current commercially available presized silos are now available that can be placed in the NICU, obviating the need for a trip to the operating room. Some have suggested that prolonged silo placement leads to increased wound infection rates and fascial dehiscence, but these are small retrospective studies.
When a silo has been constructed, the intestinal contents are squeezed back into the abdominal cavity in daily increments. Abdominal wall cellulitis related to the open wound and presence of the prosthetic material limits the use of a silo to a period of approximately 1 week. When reduction of the silo contents is complete, the patient is returned to the operating room for final closure ( Fig. 84.17 ). Broad-spectrum antibiotics are given until the silo is removed. Placement of a silo does not preclude postoperative extubation, and spontaneous ventilation during staged closure is preferable to positive-pressure ventilation. Infants should be maintained on a ventilator to allow for neuromuscular paralysis in only the most severe cases of abdominovisceral disproportion requiring aggressive closure. When postoperative mechanical ventilation is required, increased levels of positive end-expiratory pressure may be necessary to maintain functional residual capacity and optimize compliance.
During staged closure, parenteral nutrition is administered through a peripherally inserted central venous catheter, or one placed at the time of silo placement. Complete bowel rest and gastric decompression are maintained during reduction of the silo. After abdominal wall closure, whether primary or delayed, enteral feedings should be initiated only after clinical resolution of the ileus is apparent—cessation of bilious gastric aspirates, presence of bowel sounds, and passage of meconium. Advancement of enteral feedings should be conservative, as infants with gastroschisis, especially those with a dense peel requiring silo closure, are extremely sensitive to changes in nutritional substrate load. Delayed enteral feedings and prolonged parenteral infusions are a principal source of morbidity in this group. Development of cholestatic jaundice is common, and hepatic dysfunction and fibrosis may occur in a small number of refractory patients. Early administration of partial enteral mini-feedings, meticulous avoidance of infection, and reduction of copper and manganese have all been advocated to reduce the incidence of cholestatic liver disease.
The coexistence of intestinal atresias with gastroschisis deserves special mention. Intestinal atresias occur in 5%-25% of patients with gastroschisis, and they are one of several independent variables that have a negative impact on prognosis in gastroschisis. In a patient who fails to exhibit intestinal patency within 1 month of abdominal wall closure, a water-soluble lower gastrointestinal contrast study should be obtained to exclude the presence of an unrecognized atresia. Introduction of exclusive human milk feedings after gastric repair has been shown to decrease the time to achieve full enteral feeds and time to discharge.
The mortality of infants with complex gastroschisis (gastroschisis with atresia, necrosis, perforation, or volvulus) present in 17% of cases with gastroschisis was more than double that of simple gastroschisis. Significantly different outcome was found for the following parameters: Infants with complex gastroschisis are started on enteral feedings later, and they take longer to full enteral feedings with a subsequent longer duration of parenteral nutrition. Their risk of sepsis, short bowel syndrome, and necrotizing enterocolitis is higher. They stay longer in hospital and are more likely to be sent home with enteral tube feedings and parenteral nutrition.
The surgical options for abdominal wall closure for omphalocele are similar to those for gastroschisis. The main difference relates to the preoperative management of these patients. A careful assessment of physiologic status and a thorough search for syndromic and chromosomal anomalies are imperative to guide anesthetic management and to identify prognostic factors. The presence of a protective membrane allows a careful and unhurried preoperative evaluation.
As defined previously, the factors affecting prognosis for gastroschisis and omphalocele are quite distinct. Prematurity, degree of peel formation, and associated atresias account for most of the morbidity in gastroschisis, which has an overall survival rate of 90%-95%. Most of the deaths attributed to gastroschisis are related to perioperative complications, such as sepsis, necrotizing enterocolitis, and abdominal visceral ischemia, or to late hepatic failure caused by parenteral nutrition-related cholestatic disease. Surprisingly, midgut volvulus related to obligatory intestinal malrotation in these patients is virtually nonexistent, possibly because of the development of peritoneal adhesions that limit mobility of the intestine. The reported survival rate for infants with omphalocele ranges from 30%-80%. When mortality caused by associated malformations is excluded, the survival rates approach those for gastroschisis. Long-term tolerance of enteral feedings, as well as physical growth and development, are usually normal after 1 or 2 years, even in severe cases. With thoughtful management and attention to the prevention of parenteral nutrition–related hepatic complications, most infants born with abdominal wall defects should survive with an acceptable quality of life.
Congenital deficiencies of mesenteric fixation of the stomach to the surrounding structures predispose to gastric volvulus and can take two distinct forms. Absence or laxity of the gastrohepatic and gastrosplenic ligaments allows the stomach to rotate around its longitudinal axis, producing an organoaxial volvulus. Similar abnormalities of the gastrophrenic ligament and duodenal attachments allow rotation around the stomach's transverse axis, referred to as a mesentericoaxial volvulus . Organoaxial volvulus is the more common of the two types in infants and children. A strong association has been found between gastric volvulus and malrotation, asplenia, and congenital abnormalities of the diaphragm. Gastric volvulus may also be associated with conditions that result in gastric distention, such as aerophagia and hypertrophic pyloric stenosis. Because these entities all result in absence or stretching of stabilizing attachments, a causative role is assumed.
Although gastric volvulus can occur as either an acute or a chronic problem, the acute form is more common in children. The classic presentation of sudden epigastric pain, retching without emesis, and inability to advance a nasogastric tube into the stomach is rarely encountered in the actual clinical setting. Children may experience emesis, which can be bilious or nonbilious, and may not have abdominal distention. Intermittent gastric volvulus may be considered in the workup of infants presenting with apparent life-threatening events. Any combination of symptoms suggesting a partial or complete proximal mechanical obstruction may be present. Profound physiologic decompensation, hemodynamic instability, or unrelenting metabolic acidosis suggests strangulation, ischemic necrosis, and possibly perforation.
Radiologic assessment can reveal several characteristic findings. On plain abdominal radiographs, massive gastric dilation can usually be seen, often with a distinct incisura pointing toward the right upper quadrant. The spleen and small intestine may be displaced inferiorly. If a contrast study has been attempted, the contrast column may be confined to the esophagus, with a long, gradual tapering at the bottom. Occasionally, a paraesophageal hiatal hernia is detected. Classic findings on barium upper gastrointestinal study include transverse lie of the stomach and inversion of the greater curvature and pylorus ( Fig. 84.18 ). Operative treatment of acute gastric volvulus includes gastric decompression by nasogastric suction or needle aspiration and reduction of the volvulus. Coexisting anomalies, such as malrotation and diaphragmatic defects, should be corrected, and recurrence is rare.
Congenital microgastria is a rare anomaly (only 43 cases have been reported) in which the stomach is characterized by very small volume, a tubular shape, and abnormal fixation. Gastric volume does not undergo complete compensatory growth and remains relatively small as the child ages. Associated abnormalities include megaesophagus, GER, intrinsic duodenal obstruction, malrotation, biliary anomalies, situs inversus, asplenia, and skeletal defects.
Microgastria presents with vomiting and failure to thrive in the infant, often associated with persistent diarrhea. Vomiting may be caused by gastric insufficiency, GER, or duodenal obstruction. Diarrhea is presumed to be due to rapid gastric transit and dumping. The diagnosis is established by contrast upper gastrointestinal tract study, which reveals a small stomach with a transverse lie and frequently a large, patulous esophagus ( Fig. 84.19 ). Particular attention should be given to the anatomy of the duodenum and the ligament of Treitz.
The initial management of microgastria includes continuous drip enteral feeding and supplemental parenteral nutrition. When continuous feedings can be maintained for several weeks, the stomach may undergo some enlargement, allowing for a gradual transition to bolus and ad libitum feedings. Antireflux precautions, including small, frequent meals, may be required indefinitely. Although case reports of successful management by gastrojejunostomy exist, the favored surgical approach involves gastric augmentation with a Roux-en-Y jejunal reservoir.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here