Heartburn

1.0. Definitions

Heartburn and regurgitation are two of the most common symptoms experienced by patients. Heartburn is usually characterized as a burning substernal or epigastric discomfort which radiates towards the mouth. Regurgitation refers to reflux of food or bitter-tasting gastric contents from the stomach into the mouth. In this review, gastroesophageal reflux disease (GERD) refers to any clinically-significant symptomatic condition or histopathologic alteration presumed to be secondary to reflux of gastric contents into more proximal structures. The more restricted term reflux esophagitis refers to the situation where histopathologic changes in the esophageal mucosa thought secondary to reflux are present. Since the correlation between gastroesophageal reflux symptoms and histologic changes or complications is imperfect, many patients will have gastroesophageal reflux disease without evidence of reflux esophagitis.

MESH search strings used in this review include GASTROESOPHAGEAL-REFLUX, HEARTBURN, and REGURGITATION. These search strings were used in a search of Medline for years 1990-2000. Certain articles prior to 1990, and other recent published guidelines concerning this topic have also been reviewed (1, 2, 3, 4, 5, 6, 7, 8). This chapter will review the current literature available relating to the evaluation and management of adult patients with heartburn and regurgitation and formulate a management strategy based on these available data.

2.0. Epidemiology

Heartburn and regurgitation clearly are exceedingly common symptoms. Approximately 7-10% of Americans experience heartburn every day and 40% every month (9, 10). Heartburn was listed as the primary or secondary reason for 2.5 million office visits in 1985 in the U.S, and the cost of treating patients with esophagitis (a subset of patients with GERD) was estimated to be 1 billion dollars in 1985 (11). Omeprazole and ranitidine, two of the medications used to treat GERD, are among the most prescribed drugs in the world. Although gastroesophageal reflux is a spectrum of disease and may occur in healthy adults without adverse clinical consequences, patients who complain of heartburn or regurgitation to their primary care provider or who take medication regularly for these symptoms generally are diagnosed with gastroesophageal reflux disease (GERD). Limited data are available concerning the percentage of patients seeking medical attention for their symptoms of heartburn or regurgitation. In a 1988 Gallup survey of Americans, more than 40% of adults took antacids for their heartburn, and 25% sought medical attention from their physicians (12). The percentage of patients with GERD referred to subspecialists is not known. In terms of mucosal damage to the esophagus, a 3-4% prevalence of esophagitis exists in the general population (13), with 65 to 97% of these patients having mild or moderate esophagitis (14). As heartburn and regurgitation rarely cause hospital admission or death, most epidemiologic data concerning these symptoms are obtained by questionnaire studies with their inherent inaccuracies.

3.0 Pathophysiology

The pathophysiology of GERD is complex and involves a number of factors. Failure of one or a combination of the mechanisms listed below will increase the risk for GERD:

3.1 Lower esophageal sphincter (LES)

Interest has concentrated on the relationship between GERD and the valve between the esophagus and stomach. This valve, called the lower esophageal sphincter (LES), is composed of a slight thickening in the distal circular muscle of the esophagus. This valve is also referred to as the gastroesophageal flap valve due to its anatomical appearance on retrograde view from the stomach. Some patients with GERD have a low baseline LES pressure (<10mmHg). However, most GERD patients have normal baseline LES pressure, and reflux is thought to occur because of transient lower esophageal sphincter relaxations (tLESRs) (15). These tLESRs occur normally as part of a mechanism enabling belching, especially after a meal. However, in patients with reflux, tLESRs occur more frequently and last longer, thus allowing gastric reflux to take place.

3.2. Hiatal hernia and diaphragm

The majority of patients with hiatal hernias are asymptomatic and do not complain of heartburn or regurgitation. It is clear, however, that the majority of patients with esophagitis (up to 94%) do have a hiatal hernia (16). Proximal displacement of the lower esophageal sphincter into the thorax is felt to preclude the normal contribution of surrounding structures (diaphragm and phrenoesophageal ligament) to the high pressure zone (HPZ) between the esophagus and the stomach. At times of increased abdominal stress, the diaphragm is instrumental in fortifying the high pressure zone (17). This mechanism allows one to cough, sneeze, or Valsalva without significant reflux. Hiatal hernias are also felt to hold gastric secretions in a location adjacent to the gastroesophageal junction so that reflux of gastric material into the esophagus can more easily occur. Thus, the presence of a hiatal hernia promotes reflux of gastric contents into the esophagus by altering the mechanics of the esophageal flap valve, preventing diaphragmatic butressing of the high pressure zone at times of increased abdominal pressure, and by providing a pool of gastric fluid in close proximity to the esophagus. It should be emphasized again, however, that most patients with a hiatal hernia do not have symptomatic gastroesophageal reflux.

3.3. Esophageal defenses

Normally, small amounts of reflux enter the esophagus. Acid clearance mechanisms are present to help minimize acid exposure time of the esophagus. Esophageal peristalsis sweeps refluxed material back into the stomach. Swallowed saliva containing bicarbonate aids in this process. The esophageal mucosa also provides some defense against acid reflux (18). Tight junctions between the squamous mucosal cells of the esophagus prevent easy entry of caustic refluxate into the tissue. Glands in the esophagus also produce bicarbonate to help neutralize refluxed acid.

3.4 Gastric emptying

Proper emptying of the stomach is important in preventing GERD. Normal gastric emptying is a complex, highly coordinated process which depends on proper gastric muscular function as well as input from several neural components including the myenteric nervous system, central nervous system, and gut hormones. Failure of proper gastric emptying promotes reflux by increasing gastric pressure and increasing the length of time during which reflux can easily occur.

4.0. Differential Diagnosis:

4.1. Typical Symptoms (Table 1)

In patients with classic symptoms of gastroesophageal reflux disease the diagnosis is often straightforward. Heartburn is most typically described as a burning substernal chest pain or epigastric pain which commonly radiates upward into the chest. Patients often will note a bitter or sour taste in their mouth or complain of regurgitation of undigested or partially digested food. These symptoms can be associated with nausea and sometimes with a chest pain that is clinically indistinguishable from cardiac chest pain. However, GERD symptoms are classically worse after eating, particularly after fatty, spicy, or acidic foods. The symptoms are made worse by bending over or lying down. The symptoms are often temporarily relieved by antacid tablets or liquid. Although the specificity of the above symptoms in diagnosing GERD is high, the sensitivity of these symptoms is relatively low. Using ambulatory pH monitoring as the gold standard, heartburn and regurgitation were found to have sensitivities of 38% and 6% and specificities of 89% and 95% respectively in 304 patients referred for ambulatory pH monitoring (19). This low sensitivity is because many patients with GERD (as defined by evidence of mucosal abnormalities or abnormal esophageal pH testing) are asymptomatic or present with extraesophageal symptoms (see below). Elderly patients or patients with Barrett's esophagus may be particularly prone to present initially with a GERD complication (see below) and no prior history of classic GERD symptoms. Why some patients with obvious GERD complications do not have typical symptoms is unknown, although it may relate to decreased pain perception or relative acid insensitivity of the esophagus. No current strategies allow identification of these asymptomatic patients before development of their reflux complications.

4.2 Chest Pain of Undetermined Origin

Patients with cardiac chest pain cannot reliably be distinguished from patients with chest pain related to GERD (20), thus, cardiac disease should be excluded in all patients with unexplained chest pain. As many as 30% of chest pain patients undergoing cardiac catheterization have findings which do not account for their chest discomfort (21), and are often defined as having "atypical chest pain" or chest pain of undetermined origin. Several studies using ambulatory pH and pressure monitoring have revealed that approximately one-quarter to one-half of these patients have evidence of abnormal GERD (2). In addition, patients with known coronary artery disease also have a high prevalence of GERD (as measured by ambulatory esophageal pH testing) (22). Recent data also provide evidence for the possibility that reflux into the esophagus may actually trigger decreased cardiac perfusion through a neural or humoral reflex arc, also known as "linked angina" (23). Although history and pH monitoring may suggest that a patient's chest pain is related to GERD, the only way to confirm this suspicion is positive response to aggressive anti-reflux therapy. In a double-blind, placebo-controlled trial of 36 consecutive patients with chest pain and positive ambulatory pH monitoring tests, patients randomized to omeprazole 20mg po bid had significantly more improvement in the fraction of chest pain days and severity as compared to the placebo group in 8 weeks follow-up (24). In an uncontrolled series, Bancewicz and coworkers found that 23 of 24 patients with GERD and chest pain had resolution of their pain after Nissen fundoplication with a 3-6 year follow-up period (25). The differential diagnosis in a patient with heartburn/regurgitation or chest pain of undetermined etiology is shown in (Table 1).

A subset of patients with GERD will not present with the typical symptoms such as heartburn and regurgitation, but will have extraesophageal symptoms related to GERD. These extraesophageal symptoms are listed in (Table 2). The relationship of GERD and these conditions should not be underestimated as up to 80% of patients with adult-onset asthma (26), 75% of patients with chronic hoarseness (27), and 20% of patients with chronic cough (28) have evidence for GERD. GERD has been linked to a number of other clinical conditions including posterior laryngitis, recurrent pneumonia, pharyngitis, subglottic stenosis, globus sensation (constant feeling of something caught in the throat), water brash (excessive salivation), laryngeal cancer, pulmonary fibrosis, chronic nausea/vomiting and dental disease.

Patients presenting with extraesophageal manifestations of GERD can be particularly difficult to identify because these patients often do not have the more common GERD symptoms such as heartburn and regurgitation. Thus, a high level of suspicion must be maintained in patients presenting with extraesophageal symptoms. Proving a causal relationship between GERD and each of these diseases is difficult because individuals may also have both conditions without a causal relationship. The differential diagnosis of each of these conditions is beyond the scope of this chapter. In general, however, infectious diseases, neoplasm, allergic and other exposure-related diseases, as well as chronic inflammatory conditions should be excluded.

5.0. Initial Diagnostic Evaluation and Therapeutic Approach

5.1 Are empiric treatment and reassurance sufficient?

Most patients with typical gastroesophageal reflux symptoms, do not require evaluation, and can be treated with an empiric course of medical therapy. In a study of 33 consecutive patients, the use of high does proton pump inhibitor therapy was 83% sensitive for diagnosing GERD as compared to ambulatory pH monitoring (see below)(91). However, if symptoms are refractory to treatment, prolonged in their course (> 5 years), warning symptoms are present, the diagnosis is in doubt, or surgical therapy is being considered, then further evaluation should be pursued (Table 3). Warning symptoms (Table 4) include dysphagia, odynophagia, weight loss, bleeding, anemia, atypical chest pain, or family history of gastric or esophageal cancer.

5.1.1 Approach to typical heartburn/regurgitation

If patients have mild typical symptoms (less than 3 times per week), then lifestyle changes and use of prn antacids or over-the-counter H₂RAs are the therapy of choice.

If patients have moderate to severe symptoms (more than 3 times per week), then lifestyle changes, and prescription H₂RAs should be the initial approach. If patients with typical symptoms do not respond to lifestyle changes or H₂RA therapy, then therapy with a proton pump inhibitor (PPI) should be initiated.

5.2 When is further evaluation necessary?

Patients with complicated GERD, i.e. those with warning symptoms (Table 4), prolonged history of GERD symptoms (over 5 years), uncertain diagnosis, refractory symptoms, possible extraesophageal symptoms, and those patients considering surgical therapy should be referred for further evaluation. Warning symptoms such as dysphagia, odynophagia, weight loss, bleeding, or family history of esophageal or gastric cancer increase the risk of finding GERD complications such as esophageal peptic stricture, Barrett's metaplasia (29), or adenocarcinoma (29, 30, 31, 32). Physical exam or laboratory findings such as adenopathy, cachexia, or anemia also suggest the need for referral. Patients with prolonged symptoms and more severe symptoms are at increased risk for the development of esophageal adenocarcinoma (92). Those with extraesophageal symptoms may also be at increased risk for the development of Barrett's metaplasia (29, 30). Patients with uncertain diagnosis or those considering surgery should have the diagnosis confirmed. Appropriate work-up of patients prior to referral is shown in (Table 10).

5.2.1 Patients with chest pain of undetermined origin

Patients with chest pain should have had a general history and physical examination performed (including careful chest wall and musculoskeletal exam) as well screening laboratory analysis including CBC, electrolytes, LFTs, amylase, and CXR (Table 10). These patients should also have undergone cardiac evaluation appropriate to their risk stratification for coronary artery disease. A cardiac evaluation should include EKG analysis at a minimum, and will include stress testing and/or cardiac catheterization in the majority of patients. An abdominal ultrasound may also be appropriate in selected patients. In a placebo-controlled crossover study of 39 patients with noncardiac chest pain, the use of omeprazole 40 mg q am and 20 mg q pm was 78% sensitive and 86% specific in identifying patients with GERD (93). The use of empiric PPI therapy resulted in an estimated 59% reduction in procedures (93).

Generally, an empiric trial of acid suppression therapy should be undertaken before referral for further esophageal evaluation.

5.2.2 Evaluation of Patients with Extraesophageal Symptoms

The appropriate pre-referral evaluation in patients with possible extraesophageal manifestations of GERD varies by the symptom. All patients should have undergone routine history and physical examination as well as screening laboratory analysis.

Patients with otolaryngologic symptoms must undergo an examination of the larynx and pharynx to exclude infectious or neoplastic etiology before referral for evaluation of possible GERD. The finding of posterior laryngitis on laryngoscopic examination may help solidify the diagnosis of GERD. In a study of 33 patients with hoarseness and a posterior laryngitis pattern on laryngoscopy, 75% had abnormal ambulatory esophageal pH testing (27).

Patients with pulmonary manifestations should undergo chest radiography. Pulmonary function testing prior to referral will also be appropriate in patients with asthma or other significant pulmonary dysfunction.

Patients with dental disease thought possibly secondary to gastroesophageal reflux should have a complete dental examination to help assess extent of disease and rule out other causes.

5.3 If necessary, how should I further evaluate the patient?

5.3.1 Esophagogastroduodenoscopy

Upper endoscopy is the initial test of choice in patients requiring further evaluation. It is the most sensitive and specific method to evaluate the mucosa of the esophagus. Esophagogastroduodenoscopy (EGD) allows direct visualization and the opportunity for biopsies. Endoscopic findings in patients with GERD can include distal esophageal erythema, linear or circumferential erosions/ulcerations, peptic strictures, and Barrett's esophagus.

However, it should be emphasized that only approximately 50-65% of patients with symptoms consistent with GERD will have endoscopic evidence of esophagitis, making endoscopy a relatively insensitive method of diagnosing GERD (13, 34). However, endoscopy is instrumental in ruling out other possible etiologies for the patient's symptoms including pill esophagitis, infectious esophagitis, malignancy, or peptic ulcer disease (PUD), and in finding GERD complications such as esophageal stricture or Barrett's metaplasia. Compared to other methods of evaluating GERD, endoscopy is expensive ($550-1100) and exposes the patient to small risks of the procedure (including bleeding, medication reaction, or perforation).

5.3.2 Ambulatory esophageal pH monitoring

Ambulatory esophageal pH monitoring is the current "gold-standard" for confirming the diagnosis of GERD. During ambulatory esophageal pH monitoring, a pH probe is placed through a patient's nose and into the esophagus in order to measure the pH in the esophagus over an extended period of time. The pH sensor is usually placed 5cm proximal to the lower esophageal sphincter (LES). Some centers use dual pH probes with the lower sensor 5cm proximal to the LES and the second pH sensor in the proximal esophagus. During the monitoring period, the patient is also asked to complete a symptom diary which carefully describes their symptoms and the time at which the symptoms occur. The external aspect of the probe is attached to a data storage device which stores the pH data until it can be downloaded into a computer. The downloaded pH data is then analyzed (usually by computer) for the number of reflux episodes (a reflux episode is generally defined as a pH drop to 4 or less), duration and time of episodes, percentage of total time that the pH is less than 4, and separate analysis for the supine and upright positions. A combined score based on the above parameters, the DeMeester score, is often calculated. The single best parameter in evaluating a pH monitoring study is the percentage of time that the pH is less than 4. In the distal esophagus, a percent time pH<4 of over 5% is usually considered to be abnormal. The data can also be analyzed for correlation between a symptom recorded in the diary and a reflux event. However, analysis of the correlation between reflux episodes and symptoms can be a problem, and the degree of correlation necessary to imply clinical significance is not known.

As ambulatory esophageal pH monitoring is currently considered the "gold standard" in diagnosing GERD, the true sensitivity and specificity of this procedure are difficult to determine. In examining patients with symptoms of classic GERD with and without esophagitis as well as control patients without any symptoms of GERD, ambulatory pH monitoring had a sensitivity of 91% (35) in patients with classic symptoms and esophagitis. Sixty-one percent of patients with classic symptoms of GERD but no esophagitis had abnormal ambulatory pH studies. In evaluating all subjects (asymptomatic controls and symptomatic patients), the overall sensitivity was up to 81% with a specificity of 89%. If the diagnosis of GERD is not clearly established, it is generally best to study the patient off acid suppressive and prokinetic agents. Performing an ambulatory pH study on antireflux medication can be helpful in the patient with known GERD who continues to have symptoms despite aggressive antireflux therapy. However, ambulatory esophageal pH monitoring is not always available, can be uncomfortable, has its lowest sensitivity in the patient group of most interest (i.e. symptomatic patients without esophagitis), and is relatively expensive.

5.3.3 Other methods of evaluation

5.3.3.1 Barium esophagram

Although air-contrast barium esophagram is less expensive than endoscopy, the higher sensitivity and specificity of endoscopy plus its ability to take biopsies makes endoscopy a more cost-effective procedure than esophagography in evaluating patients with GERD (10). UGI series is a useful method to define a patient's anatomy and is the most sensitive way to evaluate for the presence of hiatal hernia. Reflux of barium into the patient's esophagus can sometimes be observed or provoked by increasing abdominal pressure or changing the patient's position. Unfortunately, reflux seen on barium study is of uncertain significance being positive in 25-71% of symptomatic patients but also in up to 20% of normal controls (36, 37). In terms of diagnosing reflux esophagitis, the diagnostic accuracy of radiography compared to endoscopy with biopsy is 24.6% when esophagitis is mild, 82.6% when it is moderate, and 98.7% when it is severe (4). A reticular mucosal pattern on radiography is a suggestive but insensitive sign for the presence of Barrett's esophagus; biopsy and histopathologic examination are necessary to confirm the diagnosis and determine the presence or absence of dysplasia.

5.3.3.2 Esophageal manometry

Esophageal manometry measures occlusive pressures produced in the esophagus. The measurements are obtained by placing a small flexible tube into the esophagus through the mouth or nose. Information on LES and upper esophageal sphincter (UES) pressure and relaxation with swallowing as well as peristaltic properties of the esophagus can be examined. However, detection of manometric abnormalities has not been shown to predict the presence of clinical GERD, and thus esophageal manometry is not recommended as an initial test in patients with GERD (6). Esophageal manometry can be useful in localizing the LES for proper ambulatory pH probe placement. Currently, esophageal manometry is also recommended prior to antireflux surgery in order to exclude severe peristaltic abnormalities which might increase the risk of postoperative dysphagia. Although it makes intuitive sense that significant peristaltic abnormalities might increase the risk of postoperative dysphagia, available data do not support this relationship.(38, 39, 40).

Manometric evaluation of patients with chest pain of undetermined origin is currently of limited usefulness. Although a high percentage of patients with noncardiac chest pain will have esophageal manometric abnormalities (such as "nutcracker esophagus", nonspecific esophageal motility disorders, or diffuse esophageal spasm) (41), these findings are usually not clearly related to the patient's symptom of discomfort (i.e. The manometric abnormalities may be present when the symptom is not, and the symptom may be present without the manometric abnormality.). Abnormalities found on esophageal manometry tracings in patients with chest pain of undetermined origin are currently believed to be nonspecific markers suggesting that the esophagus may be involved in the production of the patient's symptoms.

5.3.3.3 Acid perfusion test (Bernstein test)

No longer commonly performed, an acid perfusion test consists of infusion normal saline alternating with 0.1N hydrochloric acid into the esophagus. A positive test occurs when the patient experiences their typical symptoms during acid infusion but not during saline infusion. The true sensitivity of this test is unknown but thought to be low. The specificity, however, is felt to be quite high. In a prospective study of 75 consecutive patients with chest pain during ambulatory pH monitoring, the Bernstein test was found to have a sensitivity of only 32-46%, but a specificity of 83-94% in replicating the patient's chest pain (42). The presence of esophagitis may make a positive test more likely (43), and the presence of Barrett's esophagus may make it less likely (44). Although insensitive, a Bernstein test can help correlate symptoms to acid sensitivity in the esophagus, but it will not quantify the degree of reflux or document the presence or absence of esophagitis.

5.3.3.4 Nuclear medicine studies - For the evaluation of reflux disease, nuclear medicine tests are usually only performed in special circumstances

5.3.3.4.1 Gastric emptying study

A gastric emptying study allows quantitation of solid and/or liquid emptying rates from the stomach by using a gamma camera to observing passage of nuclear-labeled material from the stomach into the small intestine (45). In patients with GERD and symptoms suggesting possible gastroparesis (early satiety, significant nausea/vomiting) or risk factors for delayed gastric emptying such as diabetes mellitus or surgery, a gastric emptying study can be useful to evaluate the adequacy of gastric emptying. Generally, however, outflow obstruction should be ruled out and a trial of prokinetic therapy undertaken before proceeding to a gastric emptying study.

5.3.3.4.2 Esophageal scintigraphy

Esophageal transit has been evaluated using radiolabeled liquids and solids (46). Although less detailed information on esophageal motility is obtained, this method can be used as an alternative to esophageal manometry, particularly in patients unable to undergo manometric study.

5.3.3.4.3 Gastroesophageal scintigraphy

In a gastroesophageal scintigraphy study, a radiolabeled technetium bolus is placed into the stomach. A gamma camera quantitates the amount of nuclear-labeled material entering the chest with or without provocative reflux maneuvers such as increased abdominal pressure. The sensitivity of this technique ranges from 14-90% (4). In a recent comparative trial, gastroesophageal scintigraphy was 36% sensitive in determining abnormal reflux as compared to ambulatory pH monitoring. Scintigraphic scanning was positive in only 50% of patients with esophagitis, but specificity was 88% (47). This technique may be useful if the diagnosis of GERD must be verified in a patient who cannot undergo ambulatory pH monitoring.

5.4 Therapeutic options

Therapy for GERD can be divided into three categories: lifestyle changes, medications, and surgery.

5.4.1 Lifestyle Changes (Table 5)

Although only a few placebo-controlled studies have been performed to confirm their effectiveness, a general review of the medical literature and practical clinical experience support recommendation of lifestyle changes to all patients with GERD (48 ,94). Available literature reveals that fat ingestion can increase esophageal acid exposure for up to 3 hours after eating (49), alcohol can impair normal acid clearance in the esophagus (50), and cigarette smoking can increase the frequency of reflux episodes (51). Elevation of the head of the bed or sleeping on a wedge has been shown to decrease distal esophageal acid exposure time (52, 53). Obesity may augment the gastroesophageal reflux gradient, and morbid obesity has been shown to correlate with reflux (54). Multiple medications are known to affect lower esophageal sphincter (LES) pressure (Table 6), however, data to confirm an increase in esophageal acid exposure in patients using these medications is limited (48).

5.4.2 Medications (Table 8)

5.4.2.1 Antacids

Antacids work by neutralizing acid and inactivating pepsin. The duration of action of these agents is short. Antacids have been shown to be more effective than placebo at relieving symptoms in patients with reflux esophagitis (55). However, documentation of the symptomatic efficacy of antacids in patients with GERD is limited (48). Furthermore, no well-designed studies reveal that antacids are beneficial in healing esophagitis. Combined antacid/ alginic acid therapy has been shown produce superior symptom relief compared to antacid therapy alone in a blinded, uncontrolled study (56). Over-the-counter (OTC) antacids with or without alginic acid may be useful in patients with mild, intermittent heartburn. In patients with significant symptoms or patients with documented esophagitis, these regimens are not effective.

5.4.2.2 Histamine-2 receptor antagonists (H₂RAs)

H₂RAs block production of acid by interfering with histamine binding on the non-luminal surface of parietal cells. Four H2RAs are currently available in the U.S: cimetidine, ranitidine, famotidine, and nizatidine. In a compilation of 32 studies (using various H₂RAs and doses), approximately 61% of patients (1318/2159) had relief of their reflux symptoms with H₂RA therapy (10). In another recent review of the literature, Kahrilas found that H₂RAs offer a therapeutic gain of 10-24% over placebo in healing esophagitis. After adjusting for potency, the agents were found to be interchangeable in terms of their efficacy (1). Generally, higher doses of H₂RA and split dosing schedules lead to higher esophagitis healing rates (57, 58, 59). These medications are available OTC at roughly a dose equal to approximately one-half the smallest prescription dose for use in patients with only mild or intermittent symptoms. Standard prescription dose H₂RA given twice a day is often a reasonable starting regimen for patients with mild GERD symptoms. If a patient has documented moderate to severe esophagitis and H₂RA therapy is prescribed, then high-dose H₂RA therapy (e.g. cimetidine 800mg bid, ranitidine 300mg bid, famotidine 40mg bid, or nizatidine 300mg bid) should be used. These high-dose H₂RA regimens can increase the response rate but also increase cost. Long-term use of H₂RAs is felt to be relatively safe with over 20 years of patient experience now available.

5.4.2.3 Prokinetic therapy

In theory, prokinetic drugs work by promoting gastric emptying, improving esophageal peristalsis, and increasing lower esophageal sphincter pressure. Bethanechol and metoclopramide have limited efficacy and CNS side effects. Erythromycin has not been studied in GERD (95). Cisapride is an agent which was previously FDA approved for the treatment of adult patients with GERD. Cisapride works by increasing the release of acetylcholine from the myenteric plexus and thus accelerates gastric emptying, increases LES pressure, increases esophageal peristalsis, and increases salivary volume. This agent is about as efficacious as standard dose H₂RA therapy in the healing of esophagitis and resolution of symptoms (1). However, due to the significant risk of cardiac arrythmias and even sudden death, cisapride is no longer available in North America.

5.4.2.4 Proton pump inhibitors (PPIs)

Proton pump inhibitors work by noncompetitive inhibition of the H^+/K+ ion pump located on the luminal surface of parietal cells. These medications are concentrated in parietal cells. Numerous studies have confirmed that PPIs have superior efficacy compared to H₂RAs in patients with reflux esophagitis. PPIs lead to more rapid and greater symptom relief and healing than H₂RAs (including high-dose H₂RAs), especially in patients with more severe disease. In a compilation of 10 clinical studies using omeprazole (20mg bid, 40mg qd, or 60mg qd) for 4-8 weeks, 83% (695/834) of patients with esophagitis had symptom relief and 78% had esophageal healing (10). In 3 clinical studies involving 653 patients with moderate esophagitis, lansoprazole 30mg qd led to healing rates of 80-84% at 4 weeks and 91-92% at 8 weeks (63). Similarly, rabeprazole (96) and and pantoprazole (97) have been shown in controlled studies to heal erosive esophagitis in approximately 90% of patients within 8 weeks. PPIs are 57-74% more effective than placebo at healing esophagitis (1). Data available thus far suggests that there is no significant difference in efficacy between the four PPIs currently marketed in the U.S., omeprazole, lansoprazole rabeprazole, and pantoprazole (1). Intravenous pantoprazole has recently become available in the United States. H₂RAs should not routinely be given in conjunction with PPIs, although some preliminary work suggests that an H₂RA taken at bedtime can further reduce nocturnal acid secretion in patients taking PPIs (98). Although PPIs are the most efficacious medical therapy for GERD and healing esophagitis, concerns about cost and long-term safety have sometimes limited their extended use. Use of once daily omeprazole, lansaprazole, rabeprazole, or pantoprazole is more expensive than H₂RA therapy. However, PPIs may be more cost effective than H₂RA in patients with esophagitis because of their increased efficacy (64). Esomeprazole, a purified isometric form of omeprazole, has recently been approved by the FDA for use in the U.S. This agen may be more effective in patients with severe esophagitis than standard PPI therapy (99).

5.4.3 Surgical therapy

5.4.3.1 Who should be considered for antireflux surgery?

Referral for surgical therapy is appropriate when the patient is a reasonable candidate for an operation from a cardiopulmonary standpoint, and the patient has:

5.4.3.3 Types of antireflux operations

Multiple surgical procedures have been used to treat GERD. Currently, the most commonly performed surgical procedures include the Nissen fundoplication, Toupet fundoplication, Belsey procedure, and the Hill gastropexy. These procedures return the hiatal hernia to the abdomen and help create a new gastroesophageal valve mechanism. There is no current data to strongly support one operation over another. Most surgeons perform the operation with which they are most comfortable.

Endoscopic methods of treating GERD are currently being developed. The short-term results of three new techniques have been reported and appear promising. Placing sutures near the gastroesophageal junction using a device attached to the endoscope has been shown to improve GERD symptoms, improve pH probe results and decrease medication use in short-term uncontrolled trials (100)(107). Similarly, endoscopically-controlled radiofrequency ablation (Stretta procedure) of the LES has been shown to improve typical GER symptoms, decrease esophageal acid exposure and decrease medication use in an uncontrolled trial with short-term follow-up (108). Endoscopic injection of Plexiglas microspheres into the submucosa at the gastroesophageal junction has also been shown to improve symptoms of GERD and decrease medication use at short-term follow-up (109). While each of these techniques appears promising in uncontrolled trials with short-term follow-up, larger randomized trials with longer follow-up are needed in order to determine their role in the treatment of GERD.

5.4.3.4 Surgical efficacy

In good surgical hands, these procedures have approximately a 90% chance of relieving heartburn and regurgitation in selected patients (65, 66, 67). These procedures are now being performed laparoscopically allowing shorter hospital stays while achieving short-term results similar to the open procedures. In a follow-up of 25 patients a median of 20 years after open Nissen fundiplication, approximately 80% still had relief of their reflux symptoms (68). Long-term outcome data following laparoscopic antireflux procedures are not yet available. It is difficult to compare the efficacy of antireflux surgery with that of optimal medical therapy as no published trial directly compares PPI therapy with laparoscopic or open surgical antireflux therapy. In a randomized study comparing intermittent or continuous H₂RA therapy (with the addition of metoclopramide and sucralfate if needed) to Nissen fundoplication in GERD patients with mucosal disease (presence of peptic esophageal ulcer, stricture, erosive esophagitis, or Barrett's esophagus), surgical therapy was significantly more effective than medical therapy in improving the symptoms and endoscopic signs of esophagitis in up to two years of follow-up (69). A recent 11-13 year long follow-up study of this cohort found that the surgically-treated group continued to have significantly better symptom control and less frequent use of the antireflux medication than the medically-treated patients (101). Unfortunately, no controlled trial comparing the efficacy of antireflux surgery to PPI therapy is currently available.

5.4.3.5 Surgical risks

As with any surgical procedure there are immediate surgical risks including anesthetic risks, as well as possible long-term complications. Long-term complications can include dysphagia, the "gas-bloat" syndrome (in which the patient feels epigastric bloating and discomfort, particularly after meals and may describe a decreased ability to belch), vagal nerve injury, diarrhea, and flatulence.

5.4.3.6 Cost effectiveness

In terms of cost-effectiveness, a study from the Netherlands, based on a meta-analysis of published articles, suggests that Nissen fundoplication might be more cost effective than PPI therapy in patients with refractory or severe GERD (70). In this study, a medical maintenance therapy of omeprazole 20-40mg per day for a period of more than 4 years was less cost effective than open Nissen fundoplication (or an estimated 1.4 years for laparoscopic fundoplication) in the Netherlands. In contrast, using a computer modeling strategy based on U.S. data, Heudebert found that PPI therapy was just as effective and less costly at 5 years than laparoscopic Nissen fundoplication (71). At 10 years the costs were estimated to be similar. As PPIs become generic, cost-effective strategies may change.

6.0 GERD complications

6.1 GERD Complications: Peptic stricture

Long-term reflux of gastric contents into the esophagus can cause development of an esophageal stricture. The reported prevalence of esophageal peptic strictures is quite variable ranging from 1-23% in patients with GERD seeking medical attention (72). Most patients with esophageal peptic strictures present with symptoms of slowly progressive solid dysphagia, although some will initially present with food impaction in the esophagus. The majority of patients with reflux-induced esophageal strictures will give a history of reflux symptoms; however, reflux symptoms may become less pronounced as the esophageal lumen narrows and reflux is limited by the stricture itself. Dysphagia in these patients can be effectively treated with dilation of the esophagus. Recent data confirms that aggressive treatment of GERD with PPI therapy will also improve dysphagia and reduce the need for repeat dilation in these patients (73). Thus, it is currently recommended that patients with peptic strictures be placed on PPI therapy indefinitely.

6.2 GERD Complications: Barrett's metaplasia

6.2.1 Definition and significance

Chronic reflux can result in a change of the normal stratified squamous epithelium of the esophagus to specialized columnar epithelium (i.e. columnar epithelium with the presence of goblet cells) known as Barrett's esophagus (or Barrett's metaplasia).

Barrett's esophagus is classified as a pre-malignant condition since it is associated with the development of esophageal adenocarcinoma and adenocarcinoma of the gastric cardia. The presence of Barrett's esophagus has been estimated to increase the risk of developing adenocarcinoma by 30- to 40-fold (29). Furthermore, the incidence of esophageal adenocarcinoma in white males is rapidly increasing in Western society (74, 75) Barrett's esophagus develops in the face of chronic GERD (29). The prevalence of Barrett's esophagus in patients with GERD undergoing endoscopy has been found to be between 8.2 and 12.6% (30); these numbers may have been biased since patients with more severe GERD tend to be selected for endoscopy. Perhaps more worrisome is the finding that up to forty percent of patients presenting with adenocarcinoma associated with Barrett's metaplasia have no prior symptoms of GERD (29, 92) . In fact, the presence of Barrett's esophagus may make the esophageal mucosa less sensitive to acid (44).

6.2.2 Diagnosis of Barrett's Esophagus

During endoscopic examination, the squamocolumnar junction (where the squamous mucosa of the esophagus meets the columnar mucosa of the stomach) is usually recognized as being displaced at least 2 cm. proximally in patients with Barrett's esophagus; however, "short-segment" Barrett's metaplasia (without a 2 cm. proximal progression of the squamocolumnar junction) can also occur and has been associated with the development of adenocarcinoma (76). Biopsies are necessary to confirm the diagnosis. Careful histologic examination and flow cytometry analysis of endoscopic biopsies can be performed to look for the presence of dysplasia and aneuploidy, factors which increase the risk of adenocarcinoma (102).

6.2.3 Who to screen for Barrett's Esophagus

The important question of which GERD patients to screen for the presence of Barrett's esophagus and esophageal adenocarcinoma remains unanswered. In data available at this time, the pattern and severity of heartburn cannot reliably distinguish which patients with GERD will have Barrett's esophagus (77). Several risk factors for the development of esophageal adenocarcinoma are known. These include age (>40 years), male gender, Caucasian race, esophageal peptic stricture, dysphagia, more severe symptoms, longer duration of symptoms, and possibly the presence of other GERD complications such as extraesophageal symptoms (29, 92) . Although definitive data is not available, clinically-appropriate patients with complicated GERD should undergo endoscopic screening for Barrett's esophagus.

6.2.4 Surveillance protocol

Optimal surveillance protocols for patients with documented Barrett's esophagus are also not yet established. Surveillance guidelines developed by the Working Party on Barrett's Oesophagus at the Ninth World Congresses of Gastroenterology (78) are shown in (Table 7).

Patients with high-grade dysplasia should consider esophagectomy or be placed into an intensive protocol biopsy surveillance program at a research center. Patients with esophageal adenocarcinoma should be considered for esophagectomy. A second opinion is appropriate before proceeding to esophagectomy in a patient with high-grade dysplasia or adenocarcinoma. Currently, there is no clear evidence that antireflux surgery or aggressive acid suppressive therapy with PPIs prevents Barrett's esophagus or decreases the risk of developing adenocarcinoma (8). Recently, photodynamic therapy has been used to ablate Barrett's metaplasia (79). Photodynamic therapy involves application of a light-sensitizing agent followed by exposure of the affected tissue to the appropriate wavelength of light. The desired effect is ablation of the Barrett's metaplastic tissue with subsequent regrowth of squamous mucosa. PDT has been successful in ablating Barrett's mucosa in some patients (103). However, recurrence of Barrett's epithelium underneath newly-formed squamous mucosa, persistent genetic abnormalities, and the subsequent development of malignancies in follow-up has been reported after PDT (104). However, the efficacy of these treatment strategies remains to be established.

6.3 Complications of GERD: Extraesophageal manifestations of GERD

7.0 Maintenance therapy

Regardless of the initial medical therapy, GERD symptoms will return in a substantial proportion of patients when the medical therapy is stopped (80, 81). Lifestyle changes should be emphasized in these patients (Table 9).

7.1 Predictors of the need for long-term therapy

A long duration of symptoms before initial therapy, recurrence of symptoms after healing, initial presence of esophagitis requiring PPI therapy, and the need for prolonged therapy to relieve symptoms or heal the initial episode of esophagitis all predict relapse and the need for long-term medical management (80, 81).

7.2 Taper approach

Most experts would try to taper a patient's medication over time. Usually this is done after a 2-3 month treatment course with PPI therapy to heal the patient's esophagitis, and to give the patient time to enact lifestyle changes. Often a PPI is tapered to once a day or every other day in a gradual manner over several weeks. If the patient tolerates these changes, then a switch to H2RA therapy may be appropriate given their lower cost and longer history of established safety. In one study using this strategy, over 50% of the patients one PPI therapy were able to be tapered off their PPI (105).However, in a randomized study comparing five different medical regimens for maintenance therapy in patients with reflux esophagitis, strategies employing a PPI were more effective than H2RA or H2RA plus a prokinetic agent in maintaining healing of esophagitis (82). Similarly, a long-term, randomized, double-blind study comparing omeprazole versus ranitidine in patients with reflux esophagitis found that omeprazole 20mg qday resulted in 89% remission at one year as compared to just 32% remission with three consecutive days a week of omeprazole and 25% with ranitidine 150mg bid (83). Rabeprazole and pantaprozole have been shown to maintain healing of erosive esophagitis in the majority of patients, similar to rates seen with comparable doses of omeprazole (96, 97). Lansoprazole has also been shown to be effective maintenance therapy in patients with reflux esophagitis. In a randomized, double-blind, placebo controlled trial, 90% of patients treated with 30mg lansoprazole q day, 79% of patients treated with 15mg of lansoprazole qday versus only 24% of patients treated with placebo remained in remission after 12 months of follow-up (84). Treatment with omeprazole has been shown to be effective for at least 5 years in patients with GERD resistant to H₂RA therapy (85).

7.3.1 Antacids

Long-term use of antacids is generally considered safe unless chronic high doses are used. Chronic high doses of aluminum containing antacids can lead to neurotoxicity, fecal impaction, osteomalacia, and phosphorous depletion; calcium containing antacids can lead to fecal impaction, metabolic alkalosis, hypercalcemia associated with milk alkali syndrome and renal calculi; magnesium containing antacids with renal calculi, hypermagnesemia; malgaldarate with neurotoxicity osteomalacia, osteoporosis, phosphorous depletion and hypermagnesemia; and sodium bicarbonate with lower extremity edema, metabolic alkalosis, and hypercalcemia associated with milk alkali syndrome.

7.3.2 H₂RAs

Long-term use of H₂RA is also generally considered safe with over 20 years of clinical experience with these medications. However, side effects such as agranulocytosis, allergic reactions, antiandrogenic effects, confusion, constipation, diarrhea, sexual dysfunction, loss of hair, nausea, and hepatitis have been reported.

7.3.3 Cisapride

The main toxicity of cisapride is induction of cardiac arrhythmias including ventricular tachycardia and torsades de pointes. Cisapride is no longer available in North America except on a compassionate use basis.

7.3.4 PPIs

PPIs have been used in humans successfully for more than 10 years, but proton pump inhibitor therapy is occasionally associated with , abdominal pain, headache, fatigue, bone marrow suppression and allergic reactions. However, long-term safety issues with PPIs have primarily been three-fold. The first concern is that of bacterial overgrowth in the face of prolonged gastric acid suppression. Although small intestinal bacterial overgrowth can occur in patients on acid suppressive regimens, significant bacterial overgrowth has only rarely been reported and is thought to be unlikely even with long-term PPI therapy (86). The second concern is the development of carcinoid tumors due to prolonged exposure of enterochromaffin-like (ECL) cells to elevated gastrin levels. Elevated gastrin levels occur secondary to the effective acid suppression of PPI therapy; however, the gastrin levels observed are far below those seen in patients with pernicious anemia. Although carcinoid tumors have been seen in rat models subjected to prolonged hypergastrinemia due to PPI therapy, they have not been observed in other animal models nor in humans thus far (86). More recently it has been suggested that long-term use of PPIs in patients with Helicobacter pylori (HP) infection may increase the development of gastric mucosal atrophy. Increased gastric mucosal atrophy may result in intestinal metaplasia and thus increase the risk of gastric cancer. In a study evaluating the long-term efficacy and safety of omeprazole maintenance therapy, treatment was associated with a persistent increase in serum gastrin levels and an increase of micronodular argyrophil cell hyperplasia and subatropic or atropic gastritis (85). In another relevant study, Kuipers et al. (87) found that there was a significant increase in gastric atrophy in Helicobacter pylori-positive reflux patients treated with long-term PPI therapy as compared to patients treated with surgical antireflux therapy. Kuipers study has been criticized for using a control group from a different country and of different mean age. It is known that the HP infection tends to produce different histologic patterns in different countries, probably due to environmental influences, and that the development of atrophic gastritis is age related. A more recent randomized study revealed no relationship of PPI therapy to the development of atrophic gastritis (88, 106) , and an FDA Gastrointestinal Drugs Advisory Committee concluded that current data do not show any increased prevalence of atrophic gastritis or intestinal metaplasia, nor any increased risk of gastric cancer in association with long-term antisecretory therapy.

8.0 Refractory symptoms

Patients with refractory symptoms or esophagitis should undergo a trial of high-dose PPI therapy (at least omeprazole 20mg. Bid, lansoprazole 30mg bid, rabeprazole 20mg bid, or pantoprazole 40mg bid). In patients with persistent symptoms despite twice daily PPI therapy, the addition of a H2RA at bedtime may be helpful in suppressing nocturnal acid production. In some patients with symptoms refractory to standard dose PPI therapy, the addition of metoclopramide may be helpful. Patients requiring high-dose medical regimens, those refractory to medical treatment, and those with non-healing esophagitis should be referred for further evaluation (EGD and ambulatory esophageal pH monitoring) and consideration of surgical therapy. Patients with refractory GERD should have fasting serum gastrin levels checked in order to rule out Zollinger-Ellison syndrome.

9.0 Heartburn and regurgitation in pregnancy

Daily heartburn can occur in up to 25% of pregnant women, and 30-80% will suffer from symptomatic gastroesophageal reflux during their pregnancy (9). Although the precise mechanism for increased heartburn during pregnancy is unknown, it likely relates to hormonal effects as well as increased intraabdominal pressure. Lifestyle changes (Table 5) are always safe to recommend. Unfortunately, drugs used to treat GERD have not been extensively tested in pregnant women. Magnesium, aluminum, and calcium antacids are generally thought to be safe. Alginic acid is also likely safe because it is not absorbed, but no studies are available to document safety. Although H₂RAs and PPIs are not recommended during pregnancy, sucralfate has been used safely and efficaciously in the treatment of reflux in pregnant woman (89). In terms of their risk during pregnancy, sucralfate, cimetidine, ranitidine, famotidine, and metoclopramide are classified as Category B drugs. Omeprazole, nizatidine, lansoprazole, rabeprazole, and cisapride are classified as Category C medications. In general, treatment strategies in pregnant women should begin with lifestyle changes (Table 5), supplemented with antacids if needed. If required, the next line of therapy should be sucralfate. If a patient fails to respond to this regimen, therapy with cimetidine should be considered (90). As with all medications taken during pregnancy, a careful analysis of the risks versus benefit to the patient and fetus must be taken into account. Endoscopy can be performed safely during pregnancy, but this procedure should be delayed until after childbirth if possible.