Consultation with the Specialist: Gastroesophageal Reflux

Thirty years ago, gastroesophageal reflux disease was virtually unknown; today, it seems necessary to consider it a potential cause of an enormous range of symptoms. Not long ago, reflux was diagnosed simply by ordering an upper gastrointestinal barium study; now the complexity of differentiating between gastroesophageal reflux (GER)episodes and gastroesophageal reflux disease (GERD) mandates an understanding of the strengths and weaknesses of a variety of tests. Several decades ago, few options for therapy of reflux existed, and those that did were not very effective or even frankly detrimental; currently, the range of proven therapeutic options is greater, and the pediatrician must exercise finesse in prescribing these options.It has been useful to consider the symptoms, diagnostic complexity,and therapy of reflux as an“iceberg” (FigureF1). The largest section at the bottom represents common,easily diagnosed, and easily treated cases; the tip of the iceberg represents those cases that occur less frequently, provide considerable diagnostic and therapeutic challenges,and require subspecialist consultation. Exactly where the demarcation occurs between these sections of the iceberg depends on the individual pediatrician’s expertise and the local availability of subspecialty expertise. Each section that follows will consider both infantile GERD (which is more common) and childhood GERD.Regurgitation is the classic symptom of reflux in infancy. However, we now understand that infantile regurgitation is not necessarily GERD(reflux disease) and that GERD may occur without any regurgitation,even in infancy. Other symptoms also should prompt the pediatrician to consider the possibility of GERD affecting an infant (Table 1). During the second or third year of life, pain, particularly substernal or epigastric, gradually becomes more common than regurgitation as a presenting symptom for GERD. The pain, termed “heartburn,” suggests the presence of esophagitis, although the correlation between heartburn and esophagitis is incomplete. Other symptoms also suggest reflux in this older age group (Table 1).Mild reflux that does not represent reflux disease affects most infants in the form of regurgitation; 40% of healthy infants regurgitate more than once a day. This nonpathogenic regurgitation does not produce weight deficit, is not accompanied by abnormal irritability or other behaviors suggesting esophagitis, and is unassociated with respiratory disease or apnea. It generally begins after the first few weeks of life, peaks at about 4 months of age, and resolves by 8 to 12 months of age as the diet becomes more solid and the child’s muscle tone allows less provocative postures. Such normal regurgitation produces a wide range of reactions from parents, ranging from confident acceptance to extreme anxiety.For the infant whose parent complains of such nonpathogenic regurgitation, a number of simple measures can be useful. A complete reflux-related history can be obtained from the parent via an Infant Gastroesophageal Reflux Questionnaire (see Suggested Reading). The questionnaire, or similar careful verbal history-taking, allows identification of a number of potentially responsible caretaking practices,including overfeeding, provocative positioning, and exposure to tobacco smoke. Parents can be reassured by showing them their child’s normal growth graph and describing the(absent) symptoms that would indicate that the reflux was pathogenic and in need of diagnostic and therapeutic intervention.Finally, certain “life-style changes” can reduce problematic spitting up. This is designated as“conservative therapy,” in contrast to pharmacotherapy or surgical therapy. Such changes include the complete cessation of any smoke exposure, avoidance of unnecessary seated and supine positioning, and dietary modifications.Tobacco smoke induces lower esophageal sphincter relaxation (and increases the likelihood of apnea,sudden infant death syndrome,asthma, and pneumonia) and may provoke irritability in the infant.Seated and supine positions provoke reflux. The seated position should be avoided as much as possible, and the supine position should not be used for awake, observed times. Infants who have GERD benefit from prone position for sleep,and the American Academy of Pediatrics (AAP) exempts these infants from its recommendations against prone positioning. (Cautions about avoiding inappropriately soft bedding materials are indicated in conjunction with any recommendations for prone positioning.) In addition to decreasing reflux, prone positioning also improves gastric emptying,decreases aspiration, decreases energy expenditure, decreases crying time,and has numerous beneficial effects on respiratory disease.Thickening of the formula-fed infant’s feedings with 1 Tbsp of dry rice cereal per 1 oz of formula and using a cross-cut nipple that is slit slightly larger with a razor blade reduces spitting. The appropriate daily volume is about 4 oz of this thickened formula for each kilogram of the baby’s weight, which should be divided into at least five feedings. If as many as seven feedings are given, the interval between feedings can be compressed to as little as 3 hours, but feedings should not be given more frequently than this.It may be useful to observe the feeding and handling of an infant by her or his parents to detect other conservative measures that may be useful in individual cases. Parents who hold the bottle low so that air is ingested and those who jiggle their infants violently on their laps to soothe them often can benefit from simple instructions for refining their techniques.If these conservative measures do not produce a satisfactory response,particularly if there is a family history of allergy, a 2-week trial of an elemental formula will clarify whether cow milk allergy is playing any role. (The use of a soy formula as a diagnostic trial is not very helpful because of the frequency of concurrent sensitivity to both cow milk and soy.) For breastfed infants, the mother may try eliminating cow milk protein from her diet for a similar period, but diagnosing and treating food allergy in the breastfed infant is a much more complex problem and rarely worthwhile for nonpathogenic regurgitation.Infants who regurgitate and fail to thrive or who have other symptoms strongly suggestive of reflux disease also benefit from the measures previously noted. A number of them will improve so dramatically that further diagnostic or therapeutic measures are unnecessary.It is debatable whether pharmacotherapy should be employed in infants prior to diagnostic evaluation. To some extent, this depends on the local availability of effective diagnostic evaluation. The details of pharmacotherapy are discussed later in this article, but general principles in the untested infant include: 1) a prokinetic agent should not be used for regurgitation or vomiting without prior upper gastrointestinal radiography to assure normal anatomy,including the absence of malrotation with its potential for intermittent volvulus as the cause of the vomiting; 2) administration of a liquid antacid can be a useful trial therapy if doses are given to treat episodic irritability and succeed immediately after each dose.Aggressive empiric pharmacotherapy is warranted primarily for an infant who has an acute respiratory illness that is believed to be provoking or caused by reflux. In this setting, several weeks of an empiric combination of a prokinetic and an acid suppressor during the acute illness is likely to have more benefit than risk. Persistent or recurring episodes require diagnostic evaluation.Older children generally do not manifest nonpathogenic reflux;reflux symptoms usually suggest some degree of esophagitis that requires diagnostic evaluation. Nonetheless,very mild heartburn or an acute onset of more prominent classic heartburn may be treated empirically with acid suppression or antacid and with or without a prokinetic. It is important that such treatment not be continued for more than a few weeks if the symptoms are not improving or for more than a couple of months in any case without proceeding to more formal diagnostic evaluation.Dramatic life-style changes are less feasible in older children,although it may be useful to eliminate acidic foods, carbonated beverages,caffeine, peppermint, fatty meals,and snacking at bedtime. Tobacco smoke exposure and obesity should be addressed if present.Infants whose symptoms suggest probable GERD (ie, they are falling away from the growth curve,showing signs of esophagitis, or having apnea or respiratory disease) and who do not respond to conservative measures require diagnostic evaluation. Similarly, ambiguous symptoms possibly due to GERD are best evaluated to establish the diagnosis prior to instituting any prolonged pharmacotherapy, particularly because these ambiguous symptoms may be somewhat resistant to therapy. Diagnostic tests used depend on availability and exactly what questions need to be answered.Barium radiography evaluates whether anatomy is normal;scintigraphy defines the potential role of delayed gastric emptying or aspiration; pH probe (with concurrent pneumocardiography or symptom recording) assesses the temporal association of reflux episodes with frequent but discrete symptoms;esophageal biopsy evaluates for esophagitis; and endoscopy pinpoints a site of bleeding or establishes whether allergic (eosinophilic)esophagogastroduodenitis is present.Much of this testing is employed to evaluate diseases in the differential diagnosis (eg, malrotation or allergic enteropathy) or to establish a particular pathophysiologic mechanism(eg, delayed gastric emptying).The tests are only as effective as the consultants who interpret them,and it often is most cost-effective to undertake pediatric gastroenterologic consultation prior to any diagnostic testing. Many radiographs obtained outside of pediatric centers are taken by personnel unfamiliar both with handling infants effectively and with the most important differential diagnostic considerations, which could result in less than optimal findings. A pH probe study that is performed and evaluated strictly with a score,without assessment of the tracing for artifacts and without consideration of the effects of diet, positioning, or medications, is similarly wasteful. We require pediatric gastroenterologic consultation prior to pH probe testing, endoscopy, or biopsy to prevent such occurrences as infants undergoing a pH probe test who have incipient pertussis. Further, tailored feedings and medications allow the most information from a pH probe study if it is indicated.Older children require diagnostic evaluation for typical symptoms that persist despite brief empiric therapy or for symptoms whose etiology is ambiguous or obscure.Endoscopy with biopsies is required to evaluate for esophagitis in the child who has heartburn or odynophagia. Barium radiography is needed for the child who exhibits vomiting or dysphagia. Respiratory symptoms may require a pH probe with pneumography, a modified Bernstein test using esophageal acid infusion with pulmonary function testing, or bronchoscopy. Consultation with a pulmonologist,otolaryngologist, or allergist may be very useful in the diagnostic evaluation of these children. Aspiration may be assessed by nuclear medicine scintigraphy, salivagram, bronchoscopy for lipid-laden macrophages, or the feeding of methylene blue-labeled material to the child who has a tracheostomy.Assuring that items in the differential diagnosis are excluded often is the most important challenge for the clinician (Table 2). Infants who have GERD should receive pharmacotherapy when conservative therapy has not resolved the symptoms and when the diagnosis is reasonably certain, following the thoughtful discarding of other possibilities in the differential diagnosis. Usually this will have involved diagnostic testing and often subspecialist consultation.The recently introduced cisapride is more attractive to the general pediatrician than previous prokinetics,particularly metoclopramide, which has a narrow therapeutic range and potential for extrapyramidal reactions. However, even cisapride is associated with possible risks,including cardiac arrythmias that may be provoked by concomitant use of“-azole” or “-mycin” antibiotics. The potential for side effects is the rationale for generally avoiding pharmacotherapy for the infant who has GER but not GERD or the “happy spitter.”For the infant who has GERD, it is useful to begin with conservative measures, add a prokinetic (usually cisapride) when the diagnosis of GERD is established, and add a histamine-2 (H2) blocker when esophagitis is established (Table 3). Cimetidine currently is the optimal H2 blocker for pediatric GERD because of its inexpensive generic availability, identical dosing schedule to cisapride, and better-established pediatric dosing requirements than other H2 blockers. The infant should be reassessed at least every 2 months, doses increased to keep up with weight gain, and treatment withdrawn when symptoms resolve. Withdrawal of the acid-suppressing agent before the prokinetic agent is the most logical approach, depending on symptoms. Pharmacotherapy generally is discontinued by 12 months of age; the need to continue therapy beyond this age suggests the appropriateness of gastroenterologic consultation.GERD is far less prevalent among older children than infants. Because these older children are likely to have a chronically relapsing course,formal diagnostic evaluation with subspecialist involvement generally is indicated prior to embarking on chronic therapy. This evaluation will establish the diagnosis, address the criteria for stopping and restarting therapy, and make available to the family continually updated information on new developments regarding the disease and its therapy.Optimal initial pharmacotherapy involves both a prokinetic and an acid suppressor. Once symptoms have resolved, it is reasonable to attempt to withdraw pharmacotherapy, realizing that many children will relapse. Those who have relapsed but have been brought back under control may be able to be maintained on reduced therapy, such as a prokinetic alone or therapy limited to bedtime.Although proton pump inhibitors such as omeprazole are used increasingly in adults, even as initial therapy, they should be employed very cautiously in children. These medications suppress acid more completely than H2 blockers and may disturb the gastrointestinal ecology. They stimulate gastrin hypersecretion, and there is concern that long-term use might induce tumors. Additionally, withdrawal of these drugs is associated with a relapse of symptoms, mandating potentially very long-term usage. They should not be used without ongoing pediatric gastroenterologist consultation.Infants should be referred to a pediatric gastroenterologist when they require evaluation for GERD beyond barium radiography. Referral is appropriate for an infant whose symptoms fail to resolve after conservative therapy with or without a brief period of pharmacotherapy. Infants also should be referred if their vomiting, regurgitation, or other symptoms attributed to reflux persist beyond 1 year of age. Finally, they should be referred for evaluation if antireflux surgery is contemplated.Referral to a pediatric pulmonologist, otolaryngologist, or surgeon may be indicated for particularly challenging diagnoses or to optimize therapy of reflux-associated respiratory disease. When endoscopic evaluation of reflux-associated respiratory disease is contemplated, the potential merits of combining airway and upper gastrointestinal endoscopies should be considered.Most older children evaluated or treated for chronic or recurrent GERD should be followed by a pediatric gastroenterologist. This facilitates access to the most up-to-date information on continuously improving therapeutic options as well as monitoring of patients for complications of long-term GERD or its therapy, including strictures or premalignant Barrett esophagus.