What are the indications for surgical treatment of gastroesophageal reflux in children?
S.G. Jolley (Las Vegas)
Gastroesophageal reflux disease (GERD) in infants and children is often self-limiting. Repeated non-bilious emesis (with or without growth retardation), apnea, choking, respiratory arrest, apparent life-threatening events (ALTE), recurrent aspiration pneumonia, chronic coughing, stridor, recurrent wheezing, chronic lung disease, irritability, hematemesis, and esophagitis can be symptoms and signs of GERD in children. However, the disappearance of symptoms attributed to GERD may not reliably indicate that the GERD has resolved as well. Extended (18 to 24 hours) pH monitoring of the distal esophagus has provided the most accurate information to date regarding the characteristics of the defective gastroesophageal junction in children with GERD . Although promising for the future, manometric measurements of the gastroesophageal junction with a sleeve device or over a prolonged period of time have not provided clinically useful information for children .
Three pattern types of GERD (types I, II and III) can be detected from the esophageal pH record in prolonged recordings when children are fed apple juice . These reflux pattern types correlate with the clinical findings, radiographic findings, manometric findings, and the prognosis in childhood GERD (Table I). The type I and III reflux patterns have the lowest probability of resolving spontaneously with time and are the pattern types in which Barrett's esophagus or peptic esophageal stricture is most likely to be associated . Infants with the type II reflux pattern have the highest probability that the GERD will resolve spontaneously and has had no association with Barrett's esophagus or peptic esophageal stricture thus far.
The mean duration of acid relux episodes measured during sleep and after the first two hours postcibal (ZMD) is obtained easily from extended esophageal pH monitoring and correlates directly with the presence or absence of respiratory symptoms caused by GERD  A ZMD prolonged above the range for normal children indicates that respiratory symptoms are at least partially, if not completely, caused by the GERD. However, a ZMD within the normal range means that respiratory symptoms are unlikely to be caused by the GERD. Regardless of symptoms, a prolonged ZMD in infants and children indicates a significant risk for sudden death from uncontrolled GERD, particularly when associated with the type I or III reflux pattern  Children with a prolonged ZMD and no respiratory symptoms are more likely than those with a normal ZMD to subsequently develop serious respiratory complaints, even when parents or caretakers are unaware of the status of the ZMD.
The medical treatment for GERD in children has changed significantly over the past 50 years. Most clinicians still utilize the upright or prone-elevated (30-45 degrees) posture for infants supplemented by feedings thickened with cereal. However, the development and effectiveness of histamine-receptor blockers (cimetidine, ranitidine), prokinetic agents (domperidone, metoclopramide, cisapride), cholinergic agents (bethanechol), topical gels (sucralfate, alginic acid), and a proton-pump inhibitor (omeprazole) have led many clinicians to abandon or question the efficacy of postural therapy and thickened feedings. Except for cimetidine and bethanechol, the long-term effects of most antireflux medication mentioned in infants and children have not been determined. The best chance at achieving control of GERD in children with antireflux medications is by clinicians who are experienced in the diagnostic work-up, follow-up, and use of these medications in both infants and children.
The main antireflux operations used in children can be classified as either a fundoplication [7, 8] or a gastropexy . A fundoplication controls GERD by a nipple-valve effect and a gastropexy controls GERD by an anatomic «tightening» of the gastroesophageal junction. The side-effects and success of antireflux operations for controlling GERD in children have been well described . In the hands of surgeons who perform large numbers of major operations for the alimentary tract in infants and children each year, the operative mortality is less than 1% in otherwise healthy children and less than 5% even in children with severe associated disease. Although the safety of laparoscopic antireflux operations in children may be comparable to the open procedures, they require more operative time and the long term control of GERD by these operations is unknown.
Considerations in the decision for antireflux surgery
In general, an antireflux operation is considered in infants and children when GERD is uncontrolled by aggressive medical antireflux measures and likely to become life-threatening or debilitating, when the side-effects and risks of continued use of antireflux medications are higher than the risks of antireflux surgery, and when the child's parents or caretaker are incapable of administering medications properly. This author finds that both the clinical symptoms and the findings from extended esophageal pH monitoring are helpful for determining if and when an antireflux operation is needed in children .
Children who do not improve with at least 6 weeks of aggressive medical antireflux therapy are not likely to improve with continued nonoperative treatment. Persistent weight loss or poor weight gain from repeated emesis during the medical therapy is an indication for an antireflux operation prior to completing the trial of medical therapy. If life-threatening respiratory symptoms are present and caused by GERD, then the child is at risk for respiratory arrest with potential sudden death or disability during the arbitrary 6-week trial of medical antireflux therapy. Such children are selected for antireflux surgery, particularly if respiratory symptoms persist despite aggressive medical antireflux therapy or if the child has severe associated disease (i.e. congenital heart disease, central nervous system disease, respiratory tract obstruction, chronic lung disease). The presence of advanced esophagitis, peptic esophageal stricture, or Barrett's esophagus is an indication for an antireflux operation in children because aggressive medical antireflux therapy is unlikely to succeed in controlling GERD.
Reflux pattern type
The reflux pattern type is an important variable in the decision for an antireflux operation in children. Children with the type II reflux pattern will be continued on medical antireflux therapy longer because of the high probability for spontaneous resolution of the GERD by one year of age. Even in children with respiratory symptoms, the type II reflux pattern and a prolonged ZMD can be treated safely with aggressive medical antireflux therapy and home cardiac and apnea monitoring. On the other hand, children with a type I or III reflux pattern usually have more severe symptoms and a lower probability of spontaneous resolution. These children have the highest risk of sudden death from GERD when a prolonged ZMD is present. An antireflux operation is usually the preferred therapy at any point in the course of their treatment.
Children under two years of age are generally felt to have the ability to have their GERD improve with time. As mentioned previously, this improvement seems to correlate with the reflux pattern type. Such is not the situation for symptomatic children over two years of age. These older children usually require an antireflux operation because the GERD is very unlikely to resolve spontaneously with time. Even the unusual older patient with a type II reflux pattern doesn't seem to have long-term improvement and is strongly considered for an antireflux operation.
Placement of a feeding gastrostomy
Infants and children referred for a feeding gastrostomy are given special consideration for a concomitant protective antireflux operation. This is an area of some controversy in the pediatric surgical literature. Most patients already have GERD even if asymptomatic . In at least two thirds of the patients who do not have GERD initially, that condition will be induced acutely by placement of a feeding gastrostomy operatively or endoscopically [13-15]. The problems in these children include:
1) GERD can be present without symptoms, and these patients can die suddenly from the "asymptomatic" GERD;
2) patients without GERD prior to gastrostomy may develop GERD long term from the progression of their central nervous system (CNS) disease rather than from feeding gastrostomy;
3) with time, the GERD resulting acutely from feeding gastrostomy can resolve as the stomach pulls away from its attachment to the anterior abdominal wall and if CNS disease is either not present or improves;
4) caretakers for children with a feeding gastrostomy often do not administer gastrostomy feedings in a physiologic manner or as instructed, thereby resulting in the regurgitation of feedings even in the absence of GERD. Because most patients have GERD according to extended esophageal pH monitoring following gastrostomy placement alone, and because one cannot predict which patients will further develop GERD or have that condition resolve long term as a result of changes in their underlying disease process, this author routinely prefers a protective antireflux operation in infants and children referred for placement of a gastrostomy for feeding.
The indications for an antireflux operation in infants and children are summarized in Table II. Optimal treatment of children with GERD and appropriate selection of patients for an antireflux oepration requires the use of physicians who are familiar with the diseases of childhood and experienced in the medical or surgical treatment of GERD in children.
5. Halpern LM, Jolley SG, Tunell WP, et al. The mean duration of gastroesophageal reflux during sleep as an indicator of respiratory symptoms from gastroesophageal reflux in children. J Pediatr Surg 1991;26:686-690.