The primary mechanism for GERD is transient LES relaxations (TLESR). The result of these inappropriate LES relaxations is the presence of gastric refluxate in direct contact with the esophageal mucosa. Although initially felt to be purely acidic, recent research has indicated that up to 40% of refluxate is not acidic. Studies have shown that the occurrence of nonacid (pH > 7) and/or weakly acidic reflux (4 < pH < 7) varies between 45% and 90% in children and infants. ^, Adult studies have implicated alkaline reflux as a causative factor in the development of Barrett esophageal metaplasia. The pathologic events that occur because of GERD are due to one or more failures of the normal physiologic barriers that help prevent gastric contents from entering the esophagus, limit injury to the esophagus as a result of gastric refluxate, or clear the refluxate that enters the esophagus.

In adults, the consequence of this refluxate in the esophagus is primarily limited to erosive esophagitis, esophageal stricture, and Barrett esophagitis. These are all related to irritation of the esophagus. In children, pathologic reflux is often associated with the problems of regurgitation like aspiration, failure to thrive, and BRUE. Many children with GERD have significant neurologic impairment. These children can have increased spasticity with retching and related increased abdominal pressures. Poor swallowing mechanisms lead to gagging and choking, which add to this intermittent increase in abdominal pressure. Sometimes, a hiatal hernia develops ( Fig. 26.1 ), further predisposing to GERD. Congenital anomalies such as esophageal atresia (EA) with or without tracheoesophageal fistula (TEF), duodenal and proximal small bowel atresias, congenital diaphragmatic hernia (CDH), and gastroschisis/omphalocele all predispose to the development of GERD.

The upper GI study (A) in this infant shows a portion of the stomach ( arrow ) herniated into the mediastinum. On the right (B), the operative photograph shows the congenital hiatal hernia with a large hiatal defect and herniation of the stomach into the mediastinum.

The most important factor for preventing reflux of gastric contents into the esophagus is the LES. Embryologically, the LES arises from the inner circular muscle layer of the esophagus, which is asymmetrically thickened in the distal esophagus. The muscle in combination with its position relative to the diaphragm creates the high-pressure zone. The phrenoesophageal membrane, arising from the septum transversum of the diaphragm and the collar of Helvetius, holds the LES in position. The result is an LES that lies partially in the chest and partially in the abdomen. This positioning is important for the normal barrier function against GER. Esophageal manometry can identify this transition from the thoracic to the abdominal esophagus.

The LES is an imperfect valve that creates a pressure gradient in the distal esophagus. The ability to prevent GER is directly proportional to the LES pressure and its length, provided that LES relaxation is normal. In an adult study, LES pressures >30 mmHg prevented GER, as documented by 24-hour pH study, whereas pressures between 0 and 5 mmHg correlated with abnormal pH studies in more than 80% of patients. Also, GER is statistically significantly more likely to develop in adults if the LES pressure falls below 6 mmHg at the respiratory inversion point or if the overall LES length is ≤2 cm. As noted previously, the LES is relatively fixed across the esophageal hiatus by its surrounding attachments. Malposition of the LES, which can occur with a hiatal hernia or abnormal development, results in loss of the protective function of the LES, resulting in GER. Finally, LES relaxation occurs with esophageal peristalsis initiated by the swallowing mechanism. This relaxation is normal and must occur. When children with symptoms of GER were studied with pH and manometry simultaneously, reflux episodes rarely correlated with decreased LES pressures. Rather, the majority of reflux episodes occurred during TLESRs, and no reflux episodes were identified during LES relaxation after swallowing with normal peristaltic movement. ^,

In summary, short LES length, abnormal smooth muscle function, increased frequency of TLESRs, and LES location within the chest can contribute individually (or in combination) to LES failure and GERD.

Another barrier to the development of symptomatic GERD is the intraabdominal length of the esophagus. Although no absolute effective intraabdominal esophageal length has been identified that prevents GER, correlation between several lengths and GER have been identified. In one report, an intraabdominal length of 3–4.5 cm in adults with normal abdominal pressure provided LES competency 100% of the time. A length of 3 cm was sufficient to prevent reflux in 64% of individuals, whereas a length of <1 cm of intraabdominal esophagus resulted in reflux in 81% of patients. It was thought that failure to mobilize adequate esophageal length for intraabdominal positioning during antireflux operations led to less than successful results or recurrent GER in adults. However, we now know, from the results of two multicenter, prospective, randomized trials, that these data are not applicable in infants and children, and that complete mobilization of the esophagus, in the absence of a hiatal hernia, is detrimental in infants and children. ^,

A third barrier to reflux is the angle of His, which is the angle at which the esophagus enters the stomach. The usual orientation is that of an acute angle, which creates a flap valve at the gastroesophageal junction. Although the actual functional component of the angle of His is not well known, it has been shown to provide resistance to GER. Experimentally, when this angle is more obtuse, GER is more prone to develop. Conversely, accentuation of the angle inhibits GER.

The ability of the angle of His to prevent GER may be diminished because of abnormal development or may be iatrogenic, as with gastrostomy placement. When a normal angle of His is present, there is a convoluted fold of mucosa present at the gastroesophageal junction. This mucosa creates a rosette-like configuration that collapses on itself with increases in intragastric pressure or negative pressure in the thoracic esophagus, thus acting as an additional weak antireflux valve. ^,

Patients with increased abdominal pressure as a result of neurologically related retching, physiologic effects (obesity, ascites, peritoneal dialysis), or anatomic abnormalities (gastroschisis, omphalocele, CDH) are at increased risk for developing GERD owing to the effects of chronic pressure from the abdomen into the thorax. Finally, certain congenital defects such as congenital short esophagus, congenital hiatal hernia, and EA/TEF predispose to GERD. In patients with EA/TEF, the esophagus has abnormal peristalsis and the LES is incompetent. It has been reported that up to 30% of these patients will require antireflux surgery after repair of their EA/TEF. CDH creates anatomic abnormalities of the esophageal hiatus and the esophagus which predisposes to GERD, with 15%–20% of surviving patients undergoing an antireflux operation for GERD.

Once the barrier to GER has been overcome, mechanisms for esophageal clearance become important in preventing damage associated with exposure of the esophageal mucosa to the gastric refluxate. The primary mechanism for esophageal clearance remains esophageal motility. However, gravity and saliva contribute to the ability of the esophagus to clear the refluxate. ^, There are three types of esophageal contractions: primary, secondary, and tertiary. Primary contraction waves are initiated with swallowing and are responsible for the clearance of refluxed contents in 80%–90% of reflux episodes. Secondary waves occur when material is refluxed into the esophagus and clearance is required, especially when the reflux occurs during sleep. ^, Tertiary waves have nothing to do with esophageal clearance and are sporadic, nonpropagating contractions. When impaired esophageal motility is present as a result of abnormal smooth muscle function, impaired vagal stimulation, or obstruction, refluxed gastric contents are not moved caudad into the stomach in a timely manner. This prolonged exposure can lead to esophageal mucosal injury and can potentiate the motility disturbance due to vagal and/or smooth muscle inflammation or injury. Saliva neutralizes refluxed material, and patients with GERD have been found to have decreased salivary function. Positional effects of GERD treatment may be related to gravity assisting in the clearance of esophageal refluxate.

The final element for prevention of esophageal injury related to GERD is the ability to limit injury once refluxed contents have reached the esophagus. In addition to functioning as a neutralizing agent, saliva also aids in lubricating the esophageal contents, thus making it easier to clear any retained refluxate. Acid exposure has traditionally been postulated to cause the most significant injury, but more recent data have also implicated alkaline bile reflux. ^, Some pediatric patients with documented GERD have been shown to have increased acid secretion. ^, To this end, the role of proton pump inhibitors (PPIs) in controlling GERD in this population may be important because they have the dual effect of increasing the gastric pH while simultaneously decreasing the acid volume. However, it is now recognized that many children with GERD have normal pH probe studies and acid reflux with esophageal injury is not as important an issue for this subset of patients. ^, 8-10 Other substances that increase esophageal mucosal injury include bile salts, pepsin, and trypsin. When combined with acid, bile salts are injurious to the esophageal mucosa by increasing its permeability to existing acid, thus further potentiating injury. ^, Pepsin and trypsin are both proteolytic enzymes that can injure the esophageal mucosa. Both of these enzymes are more toxic at lower pH levels and, hence, are more injurious in the presence of acid reflux. ^,

The presentation of GERD in infants and children is variable and depends on the patient’s age and overall medical condition. Although the symptoms of GERD can vary for each patient, the actual frequency of symptoms seen in infants who have required surgical intervention for GERD has been reported with regurgitation (81%) being the most common and 41% having some form of pulmonary manifestations.

When considering the symptoms associated with GERD, persistent regurgitation is the most common complaint reported by parents. However, in infants vomiting is often physiologic and can be normal. This type of vomiting is termed chalasia of infancy , often occurring after feeding or when the baby is placed in the recumbent position. Chalasia (origin: Greek for relaxation) does not interfere with normal growth or development, and rarely leads to other complications. It is a self-limited process, with most infants transitioning to being asymptomatic by 2 years of age or near the time of initiating solid foods. No treatment is necessary in patients who have chalasia, and no diagnostic evaluation should be pursued. However, when persistent regurgitation is the result of GER, it can lead to complications, including significant malnutrition and growth failure due to insufficient caloric intake.

In infants, another presenting symptom is irritability. Painful esophagitis can be the result of the reflux. Discomfort leads to crying despite consoling measures. ^, Occasionally, small volumes of feeds briefly assist in alleviating pain. However, this is generally not a lasting effect. ^, In contrast to infants, children with GERD more often present with complaints of pain. As with adults, the pain is retrosternal in nature and often described as heartburn. Long-standing GERD with esophagitis can lead to chronic inflammation or even ulcer formation with eventual scarring and stricture. Dysphagia develops as a result of a narrowed esophageal lumen, as well as possible esophageal dysmotility secondary to long-standing mucosal inflammation. Obstructive symptoms and pain are the two most common associated complaints when an esophageal stricture is present. ^,

Barrett esophagitis is a premalignant condition that is associated with prolonged GERD. It occurs when metaplasia develops in the esophageal squamous epithelium that is replaced with columnar epithelium. In adults, it is thought to be the result of chronic esophageal injury. Whether it develops from gastric acid injury or exposure to alkaline reflux is currently a controversial topic. Although uncommon in infants and children, when it does develop, serious complications often result. In addition to the increased risk for adenocarcinoma, approximately 50% of these patients will develop stricture and many patients will develop ulcers. ^, Aggressive GERD management, along with vigilant long-term surveillance via yearly esophagogastroscopy, should be pursued to minimize these often difficult and possibly fatal complications.

Respiratory symptoms are commonly seen in infants and children. Delineating the role of GER with pulmonary symptoms requires separating primary aspiration from the oropharynx from being secondary to GER. Chronic cough, choking, apnea, or near sudden infant death syndrome (SIDS) can all be symptoms attributable to GER. Recurrent bronchitis or pneumonia can occur from aspiration of the refluxate. Esophageal stimulation via acidification of the esophageal mucosa causes vagally mediated laryngospasm and bronchospasm, which clinically presents as apnea or choking, or mistakenly as asthma. ^, Esophageal inflammation, as seen with esophagitis, likely enhances this mechanism. ^, Although uncommon, hemorrhage can be a presenting symptom of GERD. Esophagitis, gastritis, and ulcer formation can lead to hematochezia or melena in a small percentage of infants and children.

Several articles have discussed the management of GER in infants and children. ^, These articles describe nonpharmacologic therapies for GER, which include parental reassurance, dietary modification, and positional adaptation. Reassurance, by showing compassion for the presumed impaired QOL for their infant, is important. ^{, ,} Dietary modifications have been suggested to reduce regurgitation to a greater extent and faster than natural evolution. Thickening the formula to help reduce regurgitation has been approved as a management strategy by both the European and North American Societies for Pediatric Gastroenterology. Positional adaptations have included the prone position, the immediate right side position with later left side after feeding, and supine 40° anti-Trendelenburg (supine with head in the air). ^, However, the prone position is no longer recommended owing to the increased risk of sudden infant death.

As previously noted, the clinical history is an invaluable asset when evaluating for the presence of GERD and determining the need for antireflux therapy. The clinical history and symptoms are often sufficient to proceed with management. In many instances, diagnostic evaluation is reserved for patients requiring advanced therapy such as fundoplication or in cases in which the diagnosis is unclear.

Medical management for GERD is centered primarily on pharmacologic therapies.

PPIs are considered the preferred option in children, as they are more effective than H ₂ receptor antagonists. However, PPIs do not reduce the incidence of reflux episodes; they only make the reflux less acidic.

The most commonly utilized test is the contrast upper gastrointestinal (GI) study. While this study may be specific for documenting reflux, it is not helpful in determining the severity of the patient’s symptoms. Furthermore, it is not very sensitive, as the absence of reflux on the UGI study is a poor indicator of GERD as the cause of the patient’s symptoms. In a study from our institution, we reviewed 843 patients from January 2000 to June 2007 who underwent fundoplication. An upper GI had been obtained in 656 of these patients, and a pH study was performed in 379 of the patients who had an upper GI. The sensitivity of the upper GI for reflux compared with the pH study was 30.8%. An abnormality on the upper GI study besides reflux that impacted the operative plan was found in 30 patients (4.5%), with the most common being malrotation. Malrotation was confirmed in 16 patients, but excluded in six of these patients at the time of fundoplication, and four patients had previously undergone a Ladd procedure. Therefore, the true incidence of the helpful finding of malrotation on the upper GI was 4.2%. In a similar study, there were only four unexpected findings in 572 cases in which an upper GI influenced the subsequent operation in patients undergoing either an antireflux procedure or gastrostomy placement. However, a survey was conducted of pediatric surgeons working in Child Health Corporation of America (CHCA) hospitals. Three hundred thirty-seven pediatric surgeons were contacted and 121 responded: 80% indicated they requested a preoperative upper GI study before an antireflux procedure.

Prior to 2010, 24-hour pH probe monitoring had been considered the gold standard for diagnosing GERD over the previous 30 years, when DeMeester and colleagues established scores that related to the presence or absence of GERD. ^, Boix-Ochoa and colleagues later proposed a revised score that was applicable to pediatric patients aged 2 months to 3 years old and is still used today by some investigators. However, as previously mentioned, data from pH probe monitoring has taught us that not all reflux is acidic. Therefore, combined multichannel intraluminal impedance (MII) and pH measurements are now more commonly used for evaluating GERD in children. ^, MII is sensitive for evaluating GERD and is particularly good at detecting nonacid reflux episodes. It detects reflux episodes based on changes in resistance to flow from an electrical current between two electrodes in a probe when a liquid or a gas bolus moves between them. MII also distinguishes swallows (antegrade flow) from retrograde GER. It can also accurately detect the height of the refluxate while determining whether the refluxate is liquid, gas, or mixed. In one study, the authors found that 78% of GERD episodes that were temporarily associated with breathing irregularities were nonacidic (pH > 4). In another study, the authors found that respiratory symptoms occurred more frequently when GER was nonacidic (pH > 4). There are six channels on the probes, and correct positioning of the probe is important. The height reached by the refluxate is considered to be localized to the distal esophagus if it is confined to the two most distal impedance channels (channels 5 and 6). The refluxate is considered to be proximal if it reaches either or both of the most proximal channels (1 or 2).

MII has shown that GERD patients more commonly have liquid-type reflux events, whereas non-GERD patients generally have more gas-type reflux events. Also, MII data confirm that treatment with PPIs does not decrease the amount of reflux, but rather makes it less acidic. Endoscopic evaluation with biopsy is probably the most sensitive method for diagnosing GERD, but is also more invasive. The North American Society of Pediatric Gastroenterology, Hepatology and Nutrition (NASPGHN) has developed guidelines which state endoscopy is indicated in infants and children with GERD who fail to respond to pharmacologic therapy or is part of the initial management of symptoms for weight loss, unexplained anemia or fetal occult blood, recurrent pneumonia, or hematemesis. Also, in an adolescent in whom Barrett esophagus is suspected, mucosal biopsy should be performed to stage the severity of the esophagitis or to histologically exclude dysplasia or malignancy.

Gastric emptying is best evaluated using radionuclide scanning via technetium-99m-labeled meal. When documented preoperatively, a number of studies have shown that gastroparesis does not significantly improve when an emptying procedure is performed at the time of the antireflux procedure. One study evaluating patients with delayed emptying undergoing fundoplication showed significantly improved gastric emptying for both solids and liquids after fundoplication alone. Neurologically impaired (NI) children with GERD have been shown to have delayed emptying more often than neurologically normal children. Conflicting data regarding the benefit and complication rates for these patients undergoing emptying procedures at the time of their fundoplications have been reported as well. Currently, evaluation for delayed gastric emptying is not recommended prior to an initial fundoplication unless a second operative intervention would place the patient at significant morbidity or mortality. At the same time, evaluation for delayed gastric emptying may be prudent in patients undergoing a second or especially a third fundoplication, as delayed emptying may be part of the cause for failure of the fundoplication and/or transmigration of the fundoplication wrap.

At our institution, we have been moving away from preoperative imaging studies unless the information gained is helpful in planning or managing the patient’s operative care. The most useful maneuver in the patient who has problematic symptoms with gastric feeds is to feed post-pyloric with a nasojejunal tube. Resolution of symptoms with post-pyloric feedings may be an indication for fundoplication. In 114 patients requiring jejunal feeds preoperatively, 80% were able to tolerate gastric feeds immediately after fundoplication, and another 9% returned to gastric feeding over the first year.

Operative management usually follows failed medical management for growth failure, respiratory symptoms, and other symptoms such as pain and esophagitis. However, in some circumstances, it may be best to proceed with fundoplication without a trial of medical therapy.

Barrett esophagitis and esophageal stricture are two conditions in which initial operative therapy is recommended ( Fig. 26.2 ). The changes of Barrett esophagus will usually resolve in adolescents after fundoplication, although lifelong endoscopic surveillance is still needed. For children with a stricture, esophageal dilation can be performed at the time of fundoplication. Subsequent dilations may be needed in severe cases. Finally, children with a known hiatal hernia and symptomatic GER are not likely to respond to medical management. Initial fundoplication is a reasonable choice in these patients.

This 14-year-old presented with dysphagia. Workup revealed him to have this marked distal esophageal stricture and Barrett metaplasia. His initial management was esophageal dilation and fundoplication. He responded nicely to the dilations, and the Barrett metaplasia resolved following the laparoscopic fundoplication. He continues to be closely monitored for recurrence of the Barrett esophagus.

Laparoscopic Nissen Fundoplication

We have utilized several positions for this operation. Babies can be turned sideways at the head of the table, as is common for a laparoscopic pyloromyotomy. The remainder can be positioned supine at the head of the table. Stirrups are usually not necessary. Although a single monitor placed over the patient’s head is usually sufficient, two monitors, placed to the right and the left of the patient’s head, can be used as well.

After prepping and draping, a 5-mm vertical incision is made in the center of the umbilicus to place a 5 mm port. The incision should be small enough so that a stitch is unnecessary to hold it into place. A pneumoperitoneum is created to a pressure of 12–15 mmHg, and diagnostic laparoscopy is performed with a 5-mm, 45° angled telescope. Four stab incisions are then placed in infants, and three stab incisions and a 5-mm port for the Ligasure are utilized in children older than 5 years of age. The arrangement of these cannulas is seen in Fig. 26.3 . A liver retractor is introduced through the lateral right incision, or a Nathanson-style retractor can be positioned subxiphoid. The two main working sites are the instruments situated on either side of the midline. The assistant’s instrument is in the patient’s left lateral abdomen.

There are several ways to orient the instruments when performing a laparoscopic fundoplication. With our technique, a 45° angled, 5-mm telescope is introduced after insertion of the 5-mm umbilical cannula. The liver retractor is introduced in the patient’s right subcostal region ( solid arrow ). The two main working ports are in the left and right epigastrium. The main working port for the surgeon is the one in the patient’s left epigastric region. It is through this incision that dissecting instruments, needle holder, and suture are introduced. The instrument utilized by the surgical assistant is in the patient’s left subcostal region ( dotted arrow ). The stab incision technique can be utilized for both (A) infants and (B) adolescents.

From Holcomb GW III. Laparoscopic Nissen fundoplication. In: Holcomb GW, Georgeson KE, Rothenberg SS, eds. Atlas of Pediatric Laparoscopy and Thoracoscopy . Elsevier; 2008:15–20.

We have standardized our technique and have utilized it for many years. ^{, ,} Initially, the superior short gastric vessels are ligated and divided. Electrocautery connected to a Maryland dissecting instrument is used in the younger patients. As mentioned, Ligasure is used for older patients. The retroesophageal window is initially made from the patient’s left side because it is easy to accomplish after ligation/division of the short gastric vessels. We do not violate the esosophagophrenic membrane, as will be discussed later. We keep the dissection posterior to the esophagus except the loose attachments that keep the fundus against the diaphragm. Once the left side of the patient’s gastroesophageal junction has been visualized, the stomach is flipped to the patient’s left and the gastrohepatic ligament is incised to expose the esophagus and stomach on the right side. The location of the left gastric artery is identified to be sure the wrap is positioned above the artery. The opening in the retroesophageal window is then completed from the right side so that the fundus can be brought posteriorly for the Nissen fundoplication. Again, as little esophageal mobilization as possible is performed at the gastroesophageal junction ( Fig. 26.4 ). Once two instruments can be put through posteriorly, one large spread superior-inferior and another anterior-posterior spread creates an adequate window.

If an adequate length of intraabdominal esophagus is present, then as little dissection as possible is performed to help prevent migration of the fundoplication wrap through an enlarged esophageal hiatus. The phrenoesophageal ligament is kept intact on both the patient’s (A) right side and (B) left side of the esophagus. Note creation of the retroesophageal window has been initiated ( arrows ).

From Holcomb GW III. Laparoscopic Nissen fundoplication. In: Holcomb GW, Georgeson KE, Rothenberg SS, eds. Atlas of Pediatric Laparoscopy and Thoracoscopy . Elsevier; 2008:15–20.

At this point, a single suture can be sometimes needed posterior to the esophagus to close a small hiatal hernia that may have either been present initially or was created during the minimal dissection. This is usually accomplished with a 2-0 silk suture. However, we have recently shown these are not necessary as long as the phrenoesophageal membrane is not disrupted. After a recently completed prospective randomized trial, we have found these sutures are not necessary, so we no longer utilize them routinely. The bougie is then introduced. A table describing the appropriate bougie size for neonates weighing <15 kg has been developed, validated, and used for >15 years ( Table 26.1 ). The fundoplication is then performed using a standard Nissen technique. Usually, three 2-0 sutures are utilized to perform the fundoplication. The most superior suture also incorporates a small portion of anterior right crus at the 11 o’clock position. Usually a length of approximately 2 cm is desired. For older children, 2.5–3.0 cm may be appropriate.

Table 26.1

Recommended Bougie Size for Esophageal Calibration in Patients Weighing <15 kg

Weight (kg)	Bougie Size
2.5–4.0	20–24
4.0–5.5	24–28
5.5–7.0	28–32
7.0–8.5	32–34
8.5–10.0	34–36
10.0–15.0	36–40

Gastrostomy

If a gastrostomy is also needed, the stab incision or cannula site in the patient’s left mid-epigastric area is the one utilized for exteriorization of the gastrostomy button. If a fundoplication has not been performed, then this same site is used for locating the button. In either event, this site is marked before insufflation at one finger breadth from the left costal margin near the midclavicular line ( Fig. 26.5 ).

This photograph from the operating table shows the typical site for the placement of a gastrostomy button.

The anterior wall of the stomach is grasped with a locking grasper and brought toward the anterior abdominal wall. The technique for laparoscopic gastrostomy is seen in Fig. 26.6 . Two 2-0 PDS sutures (Ethicon, Inc., Somerville, NJ) are placed through the anterior abdominal wall cephalad to the grasper, through the stomach, and out through the anterior abdominal wall inferior to the instrument that has been used to grasp the stomach. A needle is placed in the stomach in the center of the square formed by the two PDS sutures followed by the guide wire and dilators. Dilators from a Cook Vascular Dilator Set (Cook, Inc., Bloomington, IN) are used to serially dilate the anterior abdominal wall and gastrotomy. If using a 12-French stem, a 16-French dilator is usually the largest needed. Longer stems usually require the 20-French dilator. The gastrostomy button is then placed over the guide wire and into the stomach. Under visualization, the balloon on the Mic-Key (Ballard Medical Products, Draper, UT) gastrostomy button is inflated. Attention must be paid to be sure that the button is, in fact, in the stomach and not external to the stomach ( Fig. 26.7 ). This can also be confirmed with the angled telescope by looking around each side of the stomach with the button in place. The PDS sutures are then secured over the button to prevent its dislodgement. Our protocol is to remove these sutures in 5 days. This technique was initially described by Georgeson and Owings, and details about complications have been published. ^,

(A) After approximation of the stomach to the anterior abdominal wall, two sutures of 2-0 or 0 PDS (depending on the patient’s age) are placed extracorporeally through the abdominal wall, through the stomach, and out through the abdominal wall inferior to the gastrostomy. (B) After placing the extracorporeal sutures, an 18-gauge needle is introduced through the left epigastric incision and into the stomach under direct visualization. Following a rush of air through the needle, a guide wire is inserted through the needle and the needle is removed. With the guide wire in place, the tract is serially dilated using the Cook Vascular Dilator Set (Cook, Inc., Bloomington, IN). These dilators come in 8-, 12-, 16-, and 20-French sizes. (C) After dilating the tract and gastrotomy with the 20-French dilator, the 8-French dilator is placed through the Mic-Key gastrostomy button and is introduced over the guide wire and into the stomach. (D) After placement of the button within the stomach, the balloon on the button is inflated, the guide wire and dilator are removed, and the extracorporeal sutures are tied over the button to secure it to the anterior abdominal wall.

From Holcomb GW III. Gastroesophageal reflux in infants and children. In: Fischer JE, ed. Mastery of Surgery , 5th ed. Lippincott Williams & Wilkins; 2007:650–651.

After the button is introduced into the stomach and inflated, it is very important to ensure that the button is, in fact, in the gastric lumen. Often, it is helpful to take an angled telescope (70°) to look around the portion of the stomach that is adherent to the anterior abdominal wall so that one can feel secure that the button is not outside the stomach. In this patient, the button was deflated and removed, and then reinserted correctly.

Our group has been interested in the efficacy of laparoscopic fundoplication for over 20 years. A number of articles have been published from our institution detailing our thoughts about indications, complications, the operative technique, and ways to improve our results. ^{, , ,} Also, there have been many articles from other centers evaluating long-term outcomes.

In early 2002, in reviewing our outcomes from January 2000 through March 2002, we believed that the need for repeat fundoplication was higher than desired. In 130 patients undergoing laparoscopic Nissen fundoplication during that time, the incidence of redo fundoplication was 12%. All patients who required a redo operation had transmigration of the fundoplication wrap. During that time period, the esophagus was being extensively mobilized to try to create at least a 2-cm length of intraabdominal esophagus. Moreover, there was no attempt to obliterate the space between the esophagus and the crura. These principles derived from prior adult training as well as literature reports in adults. ^,

In an attempt to reduce the incidence of postoperative transmigration of the wrap, two modifications were made in our operative technique. First, there was minimal mobilization of the esophagus. It was believed that the main reason for wrap transmigration was that the esophagus was being mobilized and a space was being created between the esophagus and crura to allow for the transmigration to occur. Therefore, the phrenoesophageal membrane was kept intact so as to not enlarge any space between the esophagus and crura (see Fig. 26.4 ). Second, to further obliterate this potential space, sutures were placed securing the esophagus to the crura. Initially, only two sutures were used, but eventually four sutures were placed ( Fig. 26.8 ). After making these modifications (from April 2002 through December 2004), the incidence of transmigration was reduced to 5%.

Following closure of the esophageal hiatus with a 2-0 silk suture placed posterior to the esophagus, esophagocrural sutures were placed at the 8, 11, 1, and 5 o’clock positions around the esophagus. These photographs show the (A) right and (B) left sides of the patient’s esophagus. The purpose of these sutures was to secure the esophagus in the intraabdominal position to reduce the incidence of postoperative reflux and also to obliterate the space between the esophagus and crura in an effort to prevent transmigration of the fundoplication wrap.

In 2005, we worked with Georgeson and colleagues at the University of Alabama–Birmingham to conduct a prospective, randomized trial to directly compare our modified technique to our earlier technique (extensive esophageal mobilization). The primary endpoint was transmigration of the wrap. A power analysis based on the difference between the 12% and 5% redo fundoplication rate generated a sample size of 360. One group was randomized to receive minimal esophageal mobilization with placement of the four esophagocrural sutures. The other group was randomized to extensive esophageal mobilization to create a 2-cm length of intraabdominal esophagus along with the four esophagocrural sutures. Patients were also randomized according to neurologic status. All patients received an upper GI contrast study at 1 year postoperatively to evaluate for transmigration of the fundoplication wrap.

The study was stopped early after 177 patients had been entered because the findings overwhelmingly favored minimal esophageal mobilization, with an 8% transmigration rate in the minimal dissection group compared to a 30% rate in the extensive dissection group ( P = .002). Neurologic status did not impact these outcomes. Furthermore, reoperation rates were higher in the extensive dissection group compared to the minimal dissection group (18% vs. 3%, P = .006). From this study, it was clear that minimal dissection in the pediatric patient without a hiatal hernia is important to prevent postoperative transmigration of the fundoplication wrap.

In a follow-up to the aforementioned study, our group followed patients from our center who participated in the previously discussed two-center prospective randomized trial. Of the 122 patients who enrolled from our institution, we were able to contact 70% at a median of 6.5 years from initial enrollment. There was no significant difference in age, gender, neurologic impairment, mortality, or percentage of patients who were successfully contacted in each group for this late follow-up. Since the initial report, only one additional patient in each group had required reoperation. However, there were many more dilations performed in the extensive dissection group.

Since the initial trial used four esophagocrural sutures in both groups, we subsequently conducted a randomized trial to compare the need for these sutures. There was minimal esophageal dissection in all patients. One group received the four esophagocrural sutures and the other group did not. The primary outcome was again postoperative wrap migration. A sample size of 120 was calculated using our previous data with a power of 0.82 accounting for some attrition. Again, a contrast study was scheduled at 1 year postoperatively. The mean follow-up in this study was conducted at a minimum of 1.5 years postoperatively. One hundred and twenty patients were enrolled between February 2010 and February 2014. In the 107 patients available for final analysis, the operating time was significantly longer in the group in whom the sutures were utilized by a mean of 20 minutes ( P < .01). However, no herniations were reported in either group. There has been one reoperation for wrap loosening in the nonesophagocrural suture group ( P = 1.0). Reflux symptoms and medications were no different at 1 month, 1 year, and final follow-up. Therefore, our group no longer utilizes these esophagocrural sutures and employs minimal esophageal dissection and mobilization in all patients to reduce the incidence of postoperative wrap transmigration.

There has been a randomized trial comparing open (OF) to laparoscopic (LF) fundoplication. This was a two-center study based in Norway with the main outcome measure being recurrence of GERD, which was defined as GERD combined with a reflux index >4 on pH monitoring and/or GER and/or herniated wrap on upper GI contrast study. Postoperative follow-up included 24-hour pH monitoring, upper GI contrast study, clinical examination at 6 months, and phone interviews after 1, 2, and 4 years. Eighty-seven children were randomized, with 44 undergoing LF and 43 undergoing OF. The median age was 4.7 years. Twenty-three patients in both groups were neurologically impaired (NI). The median follow-up was 4.0 years (0.3–8.9). Significantly more patients undergoing LF (37%) experienced recurrence of GERD compared to those undergoing OF (7%). The authors concluded that children operated with LF have a higher recurrence rate of GERD than those operated with OF. In an earlier study looking at 30-day outcome in this same group of children, the authors noted that 48 of the patients (55%) sustained complications within 30 days of the operation. Twenty-four of the patients were readmitted to the hospital because of complications after discharge. Also, the duration of the operation was documented as 150 + 34 minutes for LF and 89 + 25 minutes for OF ( P < .001). Furthermore, the median length of stay was 7.0 days. It was also noted that of the 88 patients, 75 of the patients came from one of two hospitals.

Another study from the same group examined a prospective cohort of patients undergoing Nissen fundoplication between 2003 and 2009. Forty-six of these patients were NI and 41 were not. In this cohort review, the hospital stay was longer for the NI children (9 days [4–57] vs. non-NI: 4 days [2–16] P < .001). More than 90% of the parents in both groups reported that the fundoplication had improved the child’s overall condition. Recurrence of GERD was diagnosed in 12 NI and 7 non-NI patients ( P = .31). With this review, the authors concluded that early complications, recurrence, and parental satisfaction after fundoplication did not differ between NI and non-NI patients.

Another randomized trial comparing OF and LF in children reported 4-year results. In this review, recurrent GER was documented by upper GI contrast study and/or 24-hour pH study. Twenty patients had been randomized to OF and 19 to LF. The incidence of recurrent GERD was 12.5% in the OF group and 20% in the LF group ( P = NS). However, only one patient in each group required a redo fundoplication. Interestingly, the nutritional status and QOL improved in both groups ( P = NS). Also, LF was associated with a reduced incidence of retching ( P = .01).

In a monocentric retrospective study, chart review was performed on children who had undergone fundoplication between 2006 and 2013 ; 119 patients with a mean age of 4.8 years underwent fundoplication. At 6 months, 21 of these patients (17.6%) had required a second fundoplication, and 64 (53.8%) had been placed back on antireflux medications. The authors concluded that, although fundoplication has a role in the treatment of severe reflux disease in children, the majority of children needed to restart their antireflux medications within 6 months of surgery.

In another report reviewing parental satisfaction following fundoplication, with a median follow-up of 7.3 years, it was found that a high percentage of parents reported improved gastrointestinal reflux-related symptoms and a high level of satisfaction following fundoplication.

In a database review from 41 US hospitals in the Pediatric Health Information System database, it was found that 8.2% of patients admitted to these hospitals underwent a fundoplication in the 9 years between January 2002 and December 2010. Interestingly, more than half of these patients undergoing fundoplication (52.7%) were 6 months of age or younger.

Several studies have looked at the need for fundoplication after an initial gastrostomy alone. In one study of 684 patients undergoing gastrostomy alone, of which 124 were open, 282 laparoscopic, and 278 endoscopic (PEG technique), subsequent fundoplication was needed in 62 patients (9.1%) with the mean interval of fundoplication after gastrostomy of 20.7 months. Cerebral palsy and anoxic brain injury had the most significant correlation with the need for subsequent fundoplication and were independent predictors. Interestingly, the laparoscopic approach for gastrostomy had a negative correlation with the subsequent need for fundoplication. In a smaller study, in NI patients who underwent gastrostomy alone, a subsequent fundoplication was needed in 17% of the patients.

In another study, the authors looked at the need for fundoplication in patients undergoing open gastrostomy (OG) versus percutaneous endoscopic gastrostomy (PEG). Sixty-nine children were evaluated (PEG = 56, OG = 13). A higher percentage of patients who underwent OG (54%) needed fundoplication compared to 27% following the PEG gastrostomy. In an interesting study, 26 NI patients who underwent laparoscopic gastrostomy were evaluated to see if their GER improved or worsened postoperatively. After evaluating these patients with pH-MMI monitoring as well as gastric emptying studies, the authors found that the laparoscopic approach reduced GER in NI patients by improving gastric emptying.

Finally, the use of the anterior hemi-fundoplication, as opposed to the Nissen fundoplication, has also been evaluated. ^, The authors evaluated QOL from the parents of both neurologically healthy and neurologically delayed children after the anterior hemi-fundoplication. In both groups, the authors found a significant improvement in QOL for the children and their parents along with improvement in their key symptoms. ^,

In another paper from our institution evaluating 39 children undergoing open fundoplication for hypoplastic left heart syndrome, there was a high morbidity and mortality in this group of patients. The conclusion was a recommendation that routine fundoplication in this population should be performed only under prospective protocols, as there were 11 deaths (28%), with most of the deaths occurring during the first and second stages of cardiac repair. ^,