Stuart Jon Spechler M.D.1
1Chief, Division of Gastroenterology, Dallas VA Medical Center, Chief, Division of Gastroenterology, Dallas VA Medical Center
The author has received grants for clinical research or educational activities from AstraZeneca, Janssen Pharmaceutica Products LP, TAP Pharmaceuticals, Inc., and Wyeth-Ayerst.
Approach to the Patient with Dysphagia
The esophagus is susceptible to three types of diseases that cause dysphagia [see Table 1]: (1) mucosal (intrinsic) diseases that narrow the lumen of the esophagus through inflammation, fibrosis, or neoplasia; (2) mediastinal (extrinsic) diseases that encase and obstruct the esophagus by direct invasion or through lymph node enlargement; and (3) diseases affecting the esophageal muscle or its innervation (esophageal motility disorders) that disrupt peristalsis, interfere with sphincter relaxation, or both. The American Gastroenterological Association has endorsed algorithms summarizing the approach to the patient with dysphagia [see Figures 1, 2, 3, and 4].1
Table 1 Diseases of the Esophagus That Cause Dysphagia
Figure 1. Management of Dysphagia
Management of dysphagia.
Figure 2. Management of Peptic Esophageal Stricture
Management of peptic esophageal stricture.
Figure 3. Management of Schatzki Ring
Management of Schatzki ring.
Figure 4. Management of Achalasia
Management of achalasia.
On the basis of a careful history alone, the astute clinician can determine the cause of dysphagia in approximately 80% of patients.2 Eight questions form the key elements of the history.
Is the dysphagia for solid foods, liquids, or both?
Mucosal and mediastinal diseases that involve the esophagus cause dysphagia by narrowing the lumen. Such mechanical narrowing usually does not impede the passage of liquids, and consequently, such a disease causes dysphagia for solid foods only. Diseases that disrupt peristalsis, however, may cause dysphagia for both solids and liquids. Of the esophageal motility disorders, achalasia is the one most likely to cause dysphagia for liquids. In achalasia, chronic contraction of the lower esophageal sphincter (LES) causes complete mechanical obstruction of the esophagus that persists until either the sphincter relaxes or the hydrostatic pressure of the retained material exceeds the pressure generated by the sphincter muscle. Even in the absence of peristalsis, gravity often can empty the esophagus of liquid effectively, provided that the LES is relaxed. Therefore, patients who have disordered peristalsis with an LES that is profoundly hypotensive often experience no dysphagia or experience dysphagia only for solid foods.
Where does the patient perceive that ingested material sticks?
Patients with esophageal strictures often perceive that swallowed material sticks at a point that is either above or at the level of the stricture.3 It is uncommon for patients to perceive that swallowed material is stuck below the obstructing lesion. Thus, the history that a swallowed bolus sticks above the suprasternal notch is of little value in localizing the obstruction, because this sensation could be caused by a lesion located anywhere from the pharynx to the gastroesophageal junction. If the patient localizes the obstruction to a point below the suprasternal notch, however, then it is highly likely that the dysphagia is caused by an esophageal disorder.
Are there symptoms of oropharyngeal dysfunction?
Oropharyngeal dysfunction often results from diseases that affect the striated muscles of the oropharynx or their innervation. Examples include muscular dystrophies, dermatomyositis, myasthenia gravis, and cerebrovascular accidents.4 Patients with these neuromuscular diseases may experience difficulty in initiating a swallow, and swallowing may be accompanied by nasopharyngeal regurgitation, pulmonary aspiration, and a sensation that residual material remains in the pharynx. If any of these symptoms are prominent, evaluation for oropharyngeal dysfunction (e.g., with videofluoroscopy) is appropriate.
Is the dysphagia intermittent or progressive?
Patients who have a lower esophageal mucosal (Schatzki) ring (see below) typically complain of discrete episodes of dysphagia for solid foods that are intermittent and nonprogressive. The episodes often occur during meals in restaurants (hence the term steakhouse syndrome) or at social functions. Dysphagia episodes may be separated by a period of weeks, months, or years, and the patient typically experiences no swallowing difficulty between episodes. In contrast, esophageal strictures usually cause dysphagia that is progressive in frequency and severity. With benign strictures, the progression is typically slow and insidious (over a period of months to years), and weight loss is minimal. Malignant esophageal strictures usually cause dysphagia that progresses rapidly (over a period of weeks to months), and weight loss may be profound.
Is there a history of chronic heartburn?
Heartburn is the cardinal symptom of gastroesophageal reflux disease (GERD). Thus, a history of chronic heartburn supports the possibility that dysphagia may result from a peptic esophageal stricture. However, up to 25% of patients with peptic strictures have no antecedent history of heartburn. Furthermore, the majority of patients with dysphagia from adenocarcinoma in Barrett esophagus have a history of long-standing heartburn. Also, around 30% of patients with achalasia complain of heartburn. Therefore, conclusions regarding the etiology of dysphagia should not be based primarily on the presence or absence of heartburn.
Has the patient taken medications likely to cause pill esophagitis?
A number of medications taken in pill form are potentially caustic to the esophagus and can cause ulceration with stricture formation if they have prolonged contact with the esophageal mucosa. Although a large number of medications have been reported to cause pill esophagitis, most cases have been caused by antibiotics (e.g., doxycycline), potassium chloride preparations, nonsteroidal anti-inflammatory drugs (NSAIDs), and quinidine.
Is there a history of collagen vascular disease?
Collagen vascular diseases such as scleroderma, rheumatoid arthritis, and systemic lupus erythematosus can cause disordered esophageal motility that is often associated with Raynaud phenomenon.5 In scleroderma and related collagen vascular disorders, fibrosis and vascular obliteration in gut smooth muscle cause poor esophageal contractility and weakness of the lower esophageal sphincter that predisposes to severe GERD. Also, patients with collagen vascular disease often are treated with medications, such as NSAIDs, that can cause pill esophagitis. Consequently, dysphagia associated with collagen vascular disease may be the result of disordered esophageal motility, severe GERD, pill esophagitis, or some combination thereof.
Is the patient immunosuppressed?
Infectious esophagitis occurs most often in patients whose immune system has been compromised by infection with HIV, by advanced malignancy, or by organ transplantation with the administration of potent immunosuppressive drugs. Odynophagia is usually the predominant symptom in infectious esophagitis, but most patients with this disease also experience dysphagia.6 In rare cases, esophageal stricture can be a late complication of infectious esophagitis.
The physical examination of the patient with dysphagia is important primarily for assessing the patient's nutritional status and ability to tolerate the invasive procedures that may be considered to treat the esophageal disorder. Only infrequently does the physical examination provide specific clues to the etiology of dysphagia. For patients with dysphagia caused by collagen vascular disease, physical examination may reveal characteristic features such as joint abnormalities, calcinosis, telangiectasias, sclerodactyly, proximal muscle weakness, and rashes. A palpable left supraclavicular (Virchow) lymph node suggests dysphagia from a malignancy in the abdomen (e.g., adenocarcinoma of the esophagogastric junction). Also, the physical examination may reveal evidence of a neuromuscular disorder that can interfere with swallowing (e.g., Parkinson disease).
Testing in patients with dysphagia generally starts with barium swallow or endoscopy. If initial testing discloses an esophageal motility disorder, manometry may then be done. Videofluoroscopy, in which a motion recording is made while the patient swallows barium suspensions and barium-coated materials, is an excellent technique for assessing oropharyngeal function but generally is not needed for evaluating esophageal disorders.
There is an unresolved debate about whether to start the evaluation of dysphagia with a barium swallow or with esophageal endoscopy. Proponents of the latter approach argue that endoscopy is almost always required to evaluate dysphagia, for both diagnostic and therapeutic purposes, and that a barium swallow usually does not provide sufficient additional information to justify its expense and inconvenience. On the other side of the argument are those who contend that a barium swallow can provide valuable anatomic information about the esophagus that may help direct therapy and prevent procedural complications. In the absence of studies validating the cost-effectiveness of either approach, this debate continues.
A barium contrast examination can be more sensitive than endoscopy for detecting subtle narrowings of the esophagus (e.g., rings, peptic strictures greater than 10 mm in diameter) and for identifying esophageal dysmotility.7 A barium swallow may be especially helpful in suggesting the diagnoses of achalasia and diffuse esophageal spasm, conditions that may be difficult to identify endoscopically. The early radiographic demonstration of achalasia may spare the patient repeat endoscopy, a situation that can occur because the endoscopist either did not recognize the disorder on the initial evaluation or was not prepared to perform endoscopic therapy at that time. A barium swallow can identify lesions that may pose potential hazards or confuse the endoscopist, such as a large Zenker diverticulum or an epiphrenic diverticulum. For patients with an esophageal stricture, a barium swallow can provide information on the extent and severity of the lesion that may help in choosing the type of dilator to be used for treatment. Finally, an initial barium swallow provides an objective baseline record of the esophagus that can be useful in assessing the response to therapy or progression of disease.
For virtually all patients with dysphagia of esophageal origin, endoscopy is recommended to establish or confirm a diagnosis, seek evidence of esophagitis and malignancy, and implement therapy when appropriate. The endoscopist can obtain biopsy and brush cytology specimens of esophageal lesions that may establish the diagnosis of neoplasms or specific infections. Endoscopy also is more sensitive than radiology for identifying subtle mucosal lesions of the esophagus (e.g., mild esophagitis).
Esophageal manometry is the gold standard test for esophageal motility disorders. Esophageal manometry has been shown to be especially useful for establishing the diagnoses of achalasia and diffuse esophageal spasm and for detecting esophageal motor abnormalities associated with collagen vascular diseases.8
For patients with dysphagia, the history and the results of the barium swallow or endoscopy can be used to decide whether esophageal manometry is indicated. An esophageal motility study usually is not needed for patients with mechanical causes of dysphagia, such as peptic strictures or rings, unless their dysphagia persists despite appropriate treatment. For patients thought to have dysphagia caused by motility abnormalities associated with collagen vascular diseases, manometry need not be performed routinely if dysphagia responds to treatment of any associated reflux esophagitis and esophageal stenoses. If the dysphagia persists despite such treatment, manometry can establish the nature of the motility problem.
Dysphagia from Benign Esophageal Strictures and Rings
BENIGN ESOPHAGEAL STRICTURES
Strictures develop from severe esophageal inflammation, usually associated with ulceration, that stimulates fibrous tissue production and collagen deposition. Approximately two thirds of all cases of benign esophageal stricture in the United States are caused by reflux esophagitis (so-called peptic strictures).9 The remainder are the result of caustic ingestions (e.g., lye), pill esophagitis, infectious esophagitis, and radiation esophagitis.
Benign esophageal strictures usually are treated with dilation [see Figure 2]. Three major types of esophageal dilating devices are used commonly: (1) mercury-filled bougies that are passed blindly through the mouth (e.g., tapered-tipped Maloney dilators, blunt-tipped Hurst dilators); (2) polyvinyl bougies that can be passed over a fine guide wire that is positioned in the stricture, under either fluoroscopic or endoscopic guidance (e.g., Savary dilators), and (3) balloon dilators that are passed either over a guide wire or through the endoscope (so-called through-the-scope [TTS] balloons). Usually, the physician passes a series of dilators of increasing diameter to stretch the stricture gradually. No study has established the superiority of one type of dilator over another. Serious complications such as perforation and bleeding occur in approximately 0.5% of all esophageal dilation procedures.2
For patients with peptic strictures caused by GERD, aggressive treatment with proton pump inhibitors both improves dysphagia and decreases the need for subsequent esophageal dilations.10 Surgical therapy can be used for esophageal strictures that do not respond to dilation and antisecretory therapy. There are two major surgical approaches: (1) antireflux surgery with intraoperative stricture dilation for patients with peptic strictures or (2) resection of the stenotic esophagus with esophageal reconstruction (e.g., by interposing a loop of bowel between the remaining esophagus and the stomach).
LOWER ESOPHAGEAL (SCHATZKI) RINGS
The lower esophageal mucosal (Schatzki) ring is a thin, diaphragmlike, circumferential fold of mucosa that protrudes into the lumen of the distal esophagus, thereby posing a physical barrier to the passage of solid material.11 Mucosal rings usually are located at the squamocolumnar junction and have squamous epithelium lining their upper surface and columnar epithelium lining the lower aspect. With careful radiologic technique aimed at distending the distal esophagus, a lower esophageal ring can be found in approximately 15% of all patients who have barium swallows. Only a minority of these rings cause dysphagia, however.
The pathogenesis of lower esophageal mucosal rings is disputed. It is not clear whether they are congenital or acquired structures. Lower esophageal mucosal rings often are associated with hiatal hernias and GERD, and some authorities have suggested that the rings are in fact thin peptic strictures. Data on the role of GERD in the pathogenesis of Schatzki rings are inconclusive and contradictory, however.
Dilation therapy is recommended for patients who have dysphagia from Schatzki rings [see Figure 3]. Traditionally, this involves the passage of a single large bougie or balloon (45 to 60 French) aimed at fracturing (rather than merely stretching) the mucosal fold. This approach differs from that for peptic strictures, which are treated by gradual stretching for fear of rupturing the fibrotic esophagus with a single, abrupt dilation. Most patients experience immediate relief of dysphagia after dilation, but recurrence is common and many patients require repeated dilations.
Esophageal Motility Abnormalities
Spechler and Castell have recently proposed a classification system for esophageal motility disorders. This system categorizes such disorders according to four major patterns of esophageal manometric abnormalities: inadequate LES relaxation, uncoordinated contraction, hypercontraction, and hypocontraction [see Table 2].12 Most esophageal motility abnormalities fall predominantly into one of these four major categories, although there can be considerable overlap.
Table 2 Classification of Esophageal Motility Abnormalities12
Processes that affect the inhibitory innervation of the LES (e.g., achalasia) can interfere with LES relaxation and thereby delay esophageal clearance. In the body of the esophagus, abnormal motility is characterized by uncoordinated contraction, hypercontraction, and hypocontraction. Uncoordinated esophageal contractions (i.e., contractions that are not peristaltic and directed toward the stomach) can delay esophageal clearance. Such uncoordinated contractions are the hallmark of diffuse esophageal spasm. Hypercontraction abnormalities are those that are characterized by contractions that are of high amplitude, long duration, or both. The putative disorders of hypercontraction (e.g., nutcracker esophagus, isolated hypertensive LES) are perhaps the most controversial of the abnormal esophageal motility patterns because it is not clear whether esophageal hypercontraction has any pathophysiologic significance. In contrast, hypocontraction abnormalities that result from weak (low-amplitude) muscle contractions can cause ineffective esophageal motility that delays esophageal clearance, and LES hypotension can result in GERD.
Primary achalasia is the best characterized of all the esophageal motility disorders.13 In achalasia, there is degeneration of neurons in the wall of the esophagus, especially the nitric oxide-producing inhibitory neurons that effect the relaxation of esophageal smooth muscle necessary for opening of the LES and for coordinated esophageal contraction. Degenerative changes also may be found in brain stem ganglion cells and their efferent fibers, but the disordered motility appears to result primarily from the degeneration of intramural neurons. The loss of inhibitory innervation in the LES causes basal sphincter pressures to rise and interferes with sphincter relaxation. In the body of the esophagus, the loss of intramural neurons results in aperistalsis.
Primary achalasia has an annual incidence of approximately one case per 100,000 population. The disorder affects men and women equally and usually is diagnosed in patients who are between 25 and 60 years of age.
Secondary achalasia, or pseudoachalasia, which can be caused by certain diseases, exhibits esophageal motor abnormalities identical to those of primary achalasia. In Chagas disease, for example, esophageal infection with Trypanosoma cruzi can destroy intramural ganglion cells and cause aperistalsis with incomplete LES relaxation. Malignancies can cause pseudoachalasia by invading the esophageal neural plexuses directly or, very rarely, by releasing uncharacterized humoral factors that disrupt esophageal function as part of a paraneoplastic syndrome.
Dysphagia for both solid foods and liquids is the primary symptom of achalasia. Moderate weight loss, regurgitation, and chest pain also are common clinical features. For reasons that are not clear, approximately one third of patients complain of heartburn, and achalasia occasionally can be confused with GERD. The symptoms of achalasia can be insidious in onset and gradual in progression, and patients frequently experience symptoms for years before seeking medical attention.
Achalasia can be confirmed with radiographic, manometric, and endoscopic evaluation. Occasionally, the diagnosis is suggested on a plain radiograph of the chest that shows widening of the mediastinum from the dilated esophagus and reveals absence of the normal gastric air bubble, because LES contraction prevents swallowed air from entering the stomach. Barium swallow, which has a diagnostic accuracy for achalasia of 95%, typically shows a dilated esophagus that terminates in a beaklike narrowing caused by persistent contraction of the LES [see Figure 5].
Figure 5. Dilatation of the Esophagus in Achalasia
Barium swallow in a patient with achalasia shows dilatation of the esophagus with a beaklike narrowing at the esophagogastric junction.
Esophageal manometry is the gold standard for the diagnosis of achalasia. The requisite manometric features are (1) incomplete relaxation of the LES (defined as a mean swallow-induced fall in resting LES pressure to a nadir value more than 8 mm above gastric pressure) and (2) aperistalsis characterized either by simultaneous esophageal contractions with amplitudes less than 40 mm Hg or by no apparent esophageal contractions. Other common manometric features of classic achalasia include LES hypertension (resting pressure greater than 45 mm Hg) and resting pressure in the esophageal body that exceeds resting pressure in the stomach.
Diagnostic endoscopy is recommended for patients with achalasia, primarily to exclude malignancy at the esophagogastric junction. In primary achalasia, the endoscopist sees a dilated esophagus that often contains residual food. The LES does not open spontaneously, but in contrast to cases of obstruction caused by neoplasms or fibrotic strictures, the contracted LES of achalasia usually can be traversed easily with gentle pressure on the endoscope.
Treatments for achalasia are aimed at decreasing the resting pressure in the LES to the point that the LES no longer impedes the passage of swallowed material [see Figure 4]. There is no therapy that can halt or reverse the degeneration of enteric neurons, and no treatment reliably restores peristaltic function in the body of the esophagus.
Nitrates and calcium channel blockers relax the smooth muscle of the LES, and these agents have been used to treat achalasia, with limited success. The drugs usually are taken sublingually 10 to 30 minutes before meals. Unfortunately, pharmacotherapy for achalasia is inconvenient, often ineffective, and frequently associated with side effects (e.g., headache and hypotension) and tachyphylaxis. Consequently, pharmacotherapy is used primarily for patients who are unwilling or unable to tolerate invasive therapies.
Pneumatic dilation therapy
In pneumatic dilation therapy for achalasia, a large deflated balloon is passed through the mouth to the LES, where the balloon is inflated rapidly to tear the sphincter muscle and thereby weaken the LES.14 Most studies describe good short-term relief of dysphagia in 60% to 85% of patients who are treated with a single session of pneumatic dilation. Esophageal perforation is the most common serious complication of the procedure, occurring in 2% to 6% of cases in most large series. Approximately 50% of patients who are treated with pneumatic dilation will require further therapy within 5 years, and subsequent pneumatic dilations are progressively less likely to result in a sustained remission.
Achalasia also can be treated by surgical myotomy, in which the surgeon weakens the LES by cutting its muscle fibers. The procedure now can be performed laparoscopically, and good to excellent relief of symptoms has been described in 70% to 90% of patients. There are few serious complications, although reflux esophagitis develops postoperatively in approximately 10% of patients. Some surgeons recommend that an antireflux (fundoplication) procedure be performed at the time of myotomy to prevent this complication. The few long-term studies available suggest that surgical myotomy results in sustained remission rates of approximately 85% at 10 years and 65% at 20 years. In a prospective, randomized comparison of myotomy with pneumatic dilation, excellent results were found after median follow-up of approximately 5 years in 40 of 42 (95%) patients in the surgical group, compared with 24 of 37 (65%) patients who had pneumatic dilation.15 Currently, the decision between pneumatic dilation and myotomy as initial therapy for achalasia should take into consideration the patient's preferences and the experience of available personnel.
Botulinum toxin therapy
Endoscopic injection of botulinum toxin into the LES poisons the excitatory (acetylcholine-releasing) neurons that contribute to LES smooth muscle tone, thereby decreasing LES pressure and relieving achalasia.16 The procedure is safe, and most patients treated with botulinum toxin injection experience immediate symptomatic improvement. Unfortunately, this effect usually is short lived. With repeated injections, approximately two thirds of patients with achalasia can be maintained in remission for 6 months. Of those patients in remission at 6 months, however, only about two thirds are still in remission at 1 year, despite repeated injections. Botulinum toxin injection is used primarily to treat patients with serious comorbidities, for whom pneumatic dilation and surgical myotomy pose inordinate risks.
DIFFUSE ESOPHAGEAL SPASM
Diffuse esophageal spasm is a rare condition of unknown etiology characterized by uncoordinated (spastic) motor activity in the smooth muscle portion of the esophagus. The esophageal spasms are manifested clinically by episodes of dysphagia and chest pain and on radiography and manometry by tertiary (nonperistaltic) contractions of the esophagus. The requisite manometric features of diffuse esophageal spasm are (1) simultaneous contractions associated with more than 10% of wet swallows and (2) a mean simultaneous contraction amplitude greater than 30 mm Hg.12 A common manometric pattern is intermittent normal peristalsis alternating with periods of spontaneous, repetitive, and multiple-peaked contractions.
Diffuse esophageal spasm is treated with agents that relax esophageal smooth muscle, such as nitrates and calcium channel blockers, although few reports document the efficacy of this therapy. Psychotropic agents such as tricyclic antidepressants may help relieve the pain of diffuse esophageal spasm. Surgical treatment by long myotomy of the esophagus has been reported, but the efficacy of this operation is poorly documented.
In nutcracker esophagus, manometry reveals peristaltic waves in the distal esophagus that have mean amplitudes more than 2 standard deviations above normal values.17 Although high-amplitude peristalsis is the most common motility abnormality observed in patients with noncardiac chest pain,18 the clinical and physiologic importance of this manometric finding is disputed. It is not clear whether the hypercontraction of esophageal muscle is a cause of chest pain or is merely an epiphenomenon that is associated with the chest pain syndrome. The clinical response to smooth muscle-relaxing agents (nitrates and calcium channel blockers) is often disappointing. Treatment with psychotropic agents such as tricyclic antidepressants can be effective in controlling the chest pain.
ISOLATED HYPERTENSIVE LES
In patients with isolated hypertensive LES, esophageal motility is normal and the LES relaxes appropriately, but the mean resting LES pressure is abnormally high (i.e., > 45 mm Hg). Although this condition has been reported in patients with noncardiac chest pain, it is unlikely that the isolated hypertensive LES has any clinical consequences.
In 80% of patients with scleroderma (progressive systemic sclerosis), fibrosis and ischemia damage the esophageal muscles and nerves. This damage causes manometric abnormalities, including weak contractions in the body of the esophagus and hypotension of the LES.19 The LES hypotension can result in GERD. When the amplitude of peristaltic contractions falls below 30 mm Hg, esophageal clearance is compromised.
In addition to weak peristalsis, patients with scleroderma often exhibit abnormalities in the progression of peristalsis, including (1) failed peristalsis, in which the peristaltic wave progresses through the pharynx and proximal esophagus but fails to traverse the entire length of the distal esophagus; (2) simultaneous esophageal contractions; and (3) absent esophageal contractions.
The manometric features of scleroderma are not specific for the disorder. Identical manometric abnormalities can be found in patients with the CREST variant (limited scleroderma [calcinosis, Raynaud phenomenon, esophageal involvement, sclerodactyly, and telangiectasias]) and other collagen vascular disorders (e.g., mixed connective tissue disease, rheumatoid arthritis, and systemic lupus erythematosus) and certain nonrheumatic diseases (e.g., diabetes mellitus and amyloidosis). Furthermore, otherwise healthy patients who have GERD often exhibit scleroderma-like motility disturbances, yet such patients infrequently develop rheumatic diseases. For these reasons, some authorities discourage use of the term scleroderma esophagus and recommend use of the term ineffective esophageal motility to describe the constellation of manometric abnormalities typical of scleroderma.20
There is currently no effective treatment for ineffective esophageal motility. Proton pump inhibitors can treat and prevent complications from associated GERD. Prokinetic agents such as metoclopramide are often used but are rarely effective, especially when the disease is advanced.
Gastroesophageal Reflux Disease
In GERD, the reflux of gastric juice into the esophagus or oropharynx causes symptoms, tissue injury, or both. Approximately 20% of adults in the United States experience GERD symptoms such as heartburn and acid regurgitation at least once a week.21 Severe GERD can result in ulceration of the esophagus, which can lead to fibrosis and esophageal stricture formation. In some cases, the ulcerated squamous epithelium of the distal esophagus is replaced by a metaplastic, intestinal-type mucosa (a condition called Barrett esophagus) that predisposes to cancer. Consequently, GERD is a strong risk factor for esophageal adenocarcinoma,22 a tumor whose frequency has increased profoundly in western countries over the past 2 decades.
The development of GERD is a multifactorial process involving dysfunction of the antireflux mechanisms that normally prevent gastric juice from entering the esophagus and of the clearance mechanisms that normally rid the esophagus of refluxed material.23
The normal barrier to gastroesophageal reflux comprises three major components: (1) the LES, (2) the crural diaphragm, and (3) anatomic features of the gastroesophageal junction.
Lower esophageal sphincter
The LES normally prevents reflux by maintaining a resting pressure that is 10 to 30 mm Hg higher than ambient pressure in the stomach.24In the 1980s, Dodds showed that episodic collapse of LES pressure, a phenomenon called transient LES relaxation (TLESR), is the major mechanism for gastroesophageal reflux both in normal individuals and in patients with GERD.25 Unlike the brief (2- to 8- second) LES relaxations that normally accompany primary (swallow-induced) peristalsis, TLESRs are not preceded by swallowing and last from 10 to 45 seconds.26 When LES pressure becomes identical to gastric pressure during a TLESR, the sphincter can no longer function as a barrier to acid reflux.
TLESRs occur two to six times per hour in normal persons, especially after meals, and 40% to 50% of these TLESRs are accompanied by brief episodes of acid reflux. In patients with GERD, TLESRs occur three to eight times per hour, and 60% to 70% of these are associated with acid reflux. In severe GERD, approximately 70% of reflux episodes are the result of TLESRs; the remaining reflux episodes are from feeble basal LES pressure, swallow-induced LES relaxation, and sudden elevations in abdominal pressure.24
Gaseous distention of the stomach normally triggers a TLESR that allows the gas to escape into the esophagus (the belch reflex).26 The nucleus tractus solitarius in the medulla mediates the belch reflex by integrating sensory information from the stomach and by controlling the neural circuits that induce the TLESR. Activation of medullary neurons with Îł-aminobutyric acid B (GABAB) receptors inhibits TLESRs, as does cholinergic blockade with atropine. The sphincter relaxation of the TLESR is effected by the activation of cholecystokinin-A receptors in LES muscle.
The right crus of the diaphragm normally encircles the distal esophagus. During inspiration, contraction of the diaphragmatic crura pinches the distal esophagus to prevent reflux. Crural contraction also helps minimize reflux during the sudden increases in abdominal pressure that accompany events such as coughing, sneezing, and straining. Thus, the crural muscle functions as an external esophageal sphincter that buttresses the LES.27 TLESRs often are accompanied by relaxation of the crura, and studies in dogs have shown that gastroesophageal reflux does not occur during a TLESR unless the episode is attended by neural inhibition of the crural diaphragm.
Anatomic features of the gastroesophageal junction
The acute angle formed by the junction of the esophagus and the stomach (the angle of His) can function as a one-way flap valve that stops reflux. Also, a portion of the distal esophagus normally is located within the abdomen, where the high ambient pressure pushes the esophageal walls together to prevent reflux.27
Disruption of the Antireflux Barrier by Hiatal Hernia
Most patients with severe GERD have a hiatal hernia in which the proximal portion of the stomach herniates into the chest through the diaphragmatic hiatus formed by the right crus of the diaphragm. With a large hiatal hernia, the LES muscle is completely dissociated from the crural diaphragm. In this situation, contraction of the crural diaphragm does not pinch the esophagus; rather, it creates an intrathoracic pouch of stomach whose contents may reflux readily into the esophagus.28,29 With no buttressing of the internal sphincter by the crural diaphragm, sudden elevations in abdominal pressure caused by inspiration, coughing, and straining can far exceed LES pressure, resulting in reflux. Reduction of the angle of His and loss of the intra-abdominal portion of the esophagus also may compromise the antireflux barrier. Furthermore, patients with large hiatal hernias have an abnormally high frequency of TLESRs induced by gastric distention.30
Disruption of Esophageal Clearance
To injure the esophagus, caustic refluxed material must remain in contact with the mucosa for a sufficient period. The duration of this contact is determined by the efficacy of the esophageal clearance mechanisms, which include gravity, peristalsis, salivation, and bicarbonate secretion by the submucosal glands of the esophagus. When gastric juice refluxes into the esophagus, most of the material is cleared by the combined effects of gravity and peristalsis.31 The small quantity of residual acidic material is neutralized by alkaline saliva and, to a lesser extent, by bicarbonate secreted by the submucosal glands.
Peristaltic abnormalities that can interfere with esophageal emptying (e.g., failed peristalsis, hypotensive peristalsis) have been found in 25% to 48% of patients with reflux esophagitis.32 Patients with large hiatal hernias frequently have impaired esophageal clearance because of the retrograde flow of material from the hernia back into the esophagus. Cigarette smoking also has been shown to increase esophageal acid exposure both by increasing the frequency of acid reflux events and by decreasing salivary flow.33
Normal persons regularly experience brief episodes of acid reflux that do not cause esophageal injury. Most patients with reflux esophagitis have prolonged esophageal acid exposure that overwhelms the normal epithelial defenses. In some patients with reflux esophagitis, however, 24-hour pH monitoring studies demonstrate a normal daily duration of acid reflux. It is conceivable that these patients have uncharacterized defects that render the esophageal epithelium vulnerable to normal acid reflux.
There is evidence to suggest that aspirin and other NSAIDs may contribute to GERD.34,35 Some NSAID preparations are caustic to the esophageal mucosa, and esophagitis can result if the tablet lingers in the esophagus (pill esophagitis). Patients with esophageal strictures may be especially susceptible to NSAID-induced esophageal injury.
Some studies suggest that gastric infection with H. pylori protects the esophagus from GERD and its complications.36 In general endoscopy units, H. pylori infection has been found significantly less often in patients with reflux esophagitis than in control patients without GERD.37Reflux esophagitis has developed in some patients with duodenal ulcers after their H. pylori infections were eradicated with antibiotics.38There is a negative association between esophageal adenocarcinoma and H. pylori infections, particularly for infections with cagA-positive strains that may be especially likely to cause severe pangastritis.39,40 Graham and others have proposed that H. pylori infections that cause pangastritis also cause a decrease in gastric acid production that may protect against GERD.41 This issue remains highly controversial, however, and the role of H. pylori infection in GERD is not yet clear.
Heartburn, the cardinal symptom of GERD, is an uncomfortable, burning sensation that is located behind the sternum. The sensation often originates in the epigastrium and radiates up the chest. Patients who describe heartburn often wave an open hand vertically over the sternum, whereas patients with angina pectoris typically hold a clenched fist stationary over the chest while describing their pain. Heartburn may be associated with the ingestion of foods that predispose to reflux by decreasing pressure in the LES, such as chocolate, onions, and fat. Spicy foods, citrus products, and tomato products may cause the sensation of heartburn by irritating the inflamed esophageal mucosa directly. Some patients experience heartburn when they bend over or exercise, presumably because these activities induce acid reflux by increasing intra-abdominal pressure. Characteristically, heartburn caused by gastroesophageal reflux is relieved immediately by antacids and can be eliminated by the administration of potent acid-suppressing agents.
When refluxed gastric juice reaches the oropharynx, patients may complain of regurgitation of sour (acid) or bitter-tasting (bilious) material. Dysphagia may result from a peptic stricture or from reflux esophagitis alone. With ulcerative esophagitis, patients may complain of odynophagia. Some patients experience water brash, in which the mouth suddenly fills with saliva as a result of reflex salivation stimulated by acid in the esophagus.
In addition to these typical symptoms, GERD can have a variety of so-called atypical manifestations.42,43,44,45 Esophageal irritation by acid reflux may result in chest pain that can mimic ischemic heart disease. If gastric juice reaches the oropharynx, it can cause globus, sore throat, and burning tongue, and the acid can erode dental enamel. Laryngitis and pulmonary problems such as chronic cough and asthma can result from aspiration of material into the airway. Asthma may also be a consequence of acid in the esophagus that triggers reflex bronchoconstriction.
Patients who have a typical history of heartburn that disappears with antisecretory therapy can be assumed to have GERD, and diagnostic tests are not necessary merely to confirm the diagnosis. Moreover, by definition, patients with GERD can have symptoms without objective evidence of esophageal damage, so normal test results often cannot exclude GERD as a cause of symptoms. However, diagnostic tests may be needed to evaluate atypical symptoms or to look for complications of GERD. For such patients, several options are available, such as radiography, endoscopy, biopsy, acid perfusion test, and pH monitoring.
Radiography has a limited role in the evaluation of GERD. A barium swallow can show signs of reflux esophagitis (e.g., thickening of the esophageal folds, erosions, and ulcerations) that support the diagnosis of GERD. A barium swallow can also identify peptic strictures. Radiography is considerably less sensitive than endoscopy for demonstrating esophagitis, however; and endoscopic examination has the added advantage of permitting biopsy of specimens from any abnormal areas.
An endoscopic examination is the most sensitive test for establishing the diagnosis of reflux esophagitis and Barrett esophagus. However, the clinician should appreciate that a normal endoscopic examination does not eliminate GERD as a cause of symptoms. Gastroesophageal reflux can cause disabling symptoms without causing visible esophageal damage.46 Endoscopy shows reflux esophagitis in only 50% to 70% of patients who complain of frequent heartburn, and heartburn severity is not a reliable index for the severity of esophagitis.47 Furthermore, the esophagus typically appears normal on endoscopy in patients who have only extraesophageal symptoms of GERD, such as chronic cough and laryngitis.
The Practice Parameters Committee of the American College of Gastroenterology recommends that endoscopy be reserved for patients with uncomplicated GERD in whom empirical therapy is unsuccessful. Endoscopic evaluation without an empirical trial of therapy is appropriate in patients who have symptoms suggesting complicated disease, including fever, anorexia, weight loss, dysphagia, odynophagia, and bleeding.48 In patients with long-standing GERD symptoms, particularly those 50 years of age and older, endoscopy is indicated to look for Barrett esophagus.49 It is important to appreciate that these guidelines are merely committee recommendations whose efficacy has not been established by clinical studies.
Biopsy specimens of the squamous epithelium in the distal esophagus of patients with GERD frequently show the histologic abnormalities of reflux esophagitis, including lengthening of the papillae, hyperplasia of cells in the basal zone, and infiltration of the epithelium with eosinophils and polymorphonuclear cells. The importance of these histologic changes of GERD is disputed, however; and routine biopsy of the squamous epithelium to seek evidence of reflux esophagitis generally is not recommended for clinical purposes.
Acid perfusion test
The acid perfusion (Bernstein) test has been used in patients who have atypical chest pain. The esophagus is perfused with 0.1N hydrochloric acid; reproduction of the chest pain implicates GERD as the etiology. This test has limited sensitivity and specificity, however, and has largely been replaced by ambulatory esophageal pH monitoring.
Ambulatory monitoring of esophageal pH is used to document the pattern, frequency, and duration of acid reflux and to seek a correlation between reflux episodes and symptoms. In most ambulatory systems, an episode of acid reflux is defined as a drop in esophageal pH below 4. Esophageal pH monitoring records a number of different variables, such as the total number of reflux episodes, the number of episodes that last longer than 5 minutes, and the duration of the longest episode. The most useful variable appears to be the percentage of the monitoring period in which esophageal pH remains below 4 (in normal individuals, this is less than 4.5% of the 24-hour test). Ambulatory esophageal pH monitoring usually is not needed for patients with typical signs and symptoms of GERD. For patients with atypical or unresponsive symptoms, however, the test can be very useful in documenting an association between symptoms and acid reflux, as well as the efficacy of antisecretory therapy in reducing acid reflux.
The efficacy of antireflux therapy is inversely related to the severity of the underlying reflux esophagitis.50 A treatment that is highly effective for mild esophagitis may be virtually useless for patients with severe disease. Ulcerative esophagitis does not respond reliably to medical therapy with agents other than proton pump inhibitors (PPIs), the most effective antisecretory medications. For patients with mild or moderate esophagitis, some authorities advocate a step-up approach to therapy that begins with antireflux lifestyle modifications and progresses eventually to the most potent medications (i.e., PPIs) only when the disease does not respond to lesser treatments.51 Others advocate a step-down approach that begins with the most effective therapy (PPIs).52 The optimal approach remains disputed. However, for patients who are known to have severe, ulcerative reflux esophagitis, it is appropriate to begin therapy immediately with PPIs rather than proceeding stepwise through trials of agents unlikely to effect healing.
A number of lifestyle modifications have been proposed to decrease esophageal acid exposure [see Table 3]. Few published data support the efficacy of these lifestyle modifications in controlling GERD, however, and it is unclear how many patients for whom such modifications are prescribed actually comply with them.
Table 3 Lifestyle Modifications for Gastroesophageal Reflux Disease
Antacids and Alginic Acid
Antacids and alginic acid can provide temporary relief of episodic heartburn. Despite the wide use of these over-the-counter products, surprisingly few data are available on their utility for healing reflux esophagitis or for the long-term management of GERD symptoms.
H2 Receptor Blocking Agents
H2 receptor antagonists are safe medications that relieve GERD symptoms and heal esophagitis within 12 weeks in approximately one half to two thirds of all patients.50,53 The H2 blockers are most useful for patients with GERD of mild to moderate severity, in whom the highest rates of healing can be anticipated. However, healing rates with these agents are poor in patients who have severe reflux esophagitis. High doses of H2 receptor blockers (up to eight times the conventional dose) have been used effectively to treat esophagitis in severe cases of GERD, but this approach generally is not recommended. Few data document the long-term efficacy of H2 receptor blockers used in any dosage, and many patients develop tolerance to the antisecretory effects of these agents. For patients with severe GERD, most authorities prescribe PPIs rather than high-dose H2 receptor blocker therapy.
In theory, prokinetic agents may decrease gastroesophageal reflux by increasing LES pressure and by enhancing esophageal and gastric clearance. Currently, metoclopramide is the only prokinetic agent available in the United States for the treatment of GERD. Metoclopramide is a dopamine antagonist, and its use is limited by side effects such as agitation, restlessness, somnolence, and extrapyramidal symptoms, which occur in up to 30% of patients. Cisapride, a serotonin-4 (5-HT4) receptor agonist, demonstrated efficacy in mild GERD, but this agent was withdrawn when it was found to cause lethal cardiac arrhythmias in patients with a number of predisposing conditions.
Sucralfate, a complex metal salt of sulfated sucrose, is an exceptionally safe medication that has some demonstrated efficacy in the treatment of mild reflux esophagitis. Few published data are available on the use of sucralfate in GERD, however, and the drug has never achieved popularity as an antireflux therapy.
Proton Pump Inhibitors
The PPIs are substituted benzimidazoles that decrease gastric acid secretion through inhibition of H+,K+-ATPase, the proton pump of the parietal cell. These agents are clearly the most effective inhibitors of gastric acid secretion available. Five PPIs are used widely for the treatment of GERD: omeprazole, esomeprazole (the S-optical isomer of omeprazole), lansoprazole, pantoprazole, and rabeprazole. All of these preparations are similar in efficacy and side-effect profiles, although when used in conventional dosages, esomeprazole may effect marginally higher rates of healing of reflux esophagitis than lansoprazole.54 Patients with mild to moderately severe reflux esophagitis who are treated with PPIs in conventional dosages achieve healing rates of 80% to 100% within 8 to 12 weeks.50,55 Very severe reflux esophagitis may persist despite conventional-dose PPI therapy in up to 40% of cases, however.56 In most of these resistant cases, the esophagitis usually can be healed by increasing the dose of the PPI.57 Recent studies also have shown that aggressive acid suppression with PPIs improves dysphagia and decreases the need for esophageal dilation in patients who have peptic esophageal strictures.10
Most patients with severe GERD who respond to PPIs require maintenance therapy, because GERD recurs shortly after stopping the drug.58For most patients, the dose of PPI necessary to maintain remission is at least the dose required to heal the acute esophagitis. In some patients with severe GERD, furthermore, PPI maintenance-dose requirements may increase with time. One long-term study of patients with severe GERD who were given a maintenance dose of 20 mg of omeprazole daily found that relapses occurred at the rate of 1 per 9.4 treatment-years and that patients often required increasing doses of omeprazole (up to 120 mg/day) to maintain GERD in remission.57
The profound acid suppression that can be achieved with PPIs has raised theoretical concerns regarding their long-term safety.53Nevertheless, there are no reports of tumors or nutritional deficiencies clearly attributable to the use of PPIs after more than a decade of extensive clinical experience with these agents.
Antireflux operations share four fundamental features: (1) reduction of the hiatal hernia, (2) restoration of an intra-abdominal segment of esophagus, (3) approximation of the diaphragmatic crura, and (4) fundoplication, in which the surgeon wraps a portion of the gastric fundus around the distal esophagus. The operations differ primarily in the approach (e.g., transthoracic versus transabdominal) and in the extent of fundoplication.
The precise mechanisms whereby these operations prevent reflux are not clear, but a number of potential ones have been proposed.59 The fundoplication may prevent the distention of the gastric fundus that ordinarily triggers TLESRs. Restoration of the distal esophagus to the positive pressure environment of the abdomen may prevent reflux, and the anatomic rearrangement of the gastroesophageal junction may create an antireflux flap-valve effect. Also, reduction of the hiatal hernia and approximation of the diaphragmatic crura may restore the normal antireflux function of the crural diaphragm.
Antireflux surgery can be performed laparoscopically.60 The laparoscopic approach has become popular not because it is safer or produces a better functional result than the open procedure, but because it supposedly offers the advantages of less postoperative discomfort, shorter hospital stay, and better cosmetic outcome. Two recent randomized trials of laparoscopic versus open Nissen fundoplication found no significant differences in the functional results of the two procedures (i.e., relief of GERD symptoms and reduction in esophageal acid exposure).61,62 However, one of the studies was terminated prematurely because an interim analysis showed an excess of adverse outcomes (primarily dysphagia) in the group treated laparoscopically.62 At least one study has shown that the primary factor involved in overall patient satisfaction with antireflux surgery is the relief of GERD symptoms. These observations suggest that the primary decision for the clinician to make is whether the patient should have an antireflux operation, not how the operation should be performed.
Uncontrolled retrospective studies of antireflux surgery generally have described excellent results, with success rates exceeding 80%.60 Few randomized trials have done direct comparisons of medical and surgical antireflux therapies, however. In the late 1980s (before the release of PPIs for clinical use in the United States), the Department of Veterans Affairs conducted a large cooperative study that prospectively compared the efficacy of medical therapy with that of surgical therapy in 247 patients with complicated GERD.63 Patients were prescribed antireflux lifestyle modifications and randomly assigned to receive one of three types of treatment: (1) continuous medical therapy (antacid tablets and ranitidine taken on a daily basis regardless of symptoms, with metoclopramide and sucralfate added if necessary to control symptoms); (2) symptomatic medical therapy (the same drugs as in the continuous-medical-therapy group but used only when necessary for control of symptoms); or (3) surgical therapy (open Nissen fundoplication). All three groups showed significant improvements in the symptoms and endoscopic signs of GERD for up to 2 years. However, surgical therapy was significantly better than both medical therapies during that period, and surgical patients had higher overall satisfaction.
In the 1990s, a Scandinavian group conducted a randomized trial of omeprazole versus open antireflux surgery in 310 patients with erosive esophagitis.64 In patients who received a fixed dose of 20 mg of omeprazole a day, antireflux surgery was superior in maintaining GERD in remission for the 3-year duration of the study. In clinical practice, however, the dose of a PPI is typically titrated to control symptoms. In this study, when the physician was permitted to titrate the PPI dose, there was no statistically significant difference between the medical group and the surgical group in maintaining remission for 3 years.
Relatively few reports have described the long-term outcome of antireflux surgery. Some uncontrolled studies have found success rates that exceed 90% at 10 to 20 years after open fundoplication,65 whereas others have described breakdown of the operation and the return of reflux esophagitis in more than 50% of cases.66 A follow-up study was conducted on the patients who participated in the VA cooperative study (see above).67 During the follow-up period of 10 to 13 years, surgical patients were significantly less likely to have taken antireflux medications regularly, and when antireflux medications were discontinued, the GERD symptoms in these patients were significantly less severe than those in the medical group. However, 62% of the surgical patients took antireflux medications on a regular basis, and there were no significant differences between the groups in the rates of neoplastic and peptic complications of GERD, overall physical and mental well-being scores, and overall satisfaction with antireflux therapy. There were 79 deaths, involving 33 (40%) of the 82 surgical patients and 46 (28%) of the 165 medical patients (P = 0.047). For reasons that are not clear, the excess deaths in the surgical group were from heart disease. This and other studies suggest that antireflux surgery does not effect a permanent cure for GERD in the majority of patients and that surgery is no better than medication for preventing the peptic and neoplastic complications of GERD.
Endoscopic Antireflux Procedures
Two endoscopic therapies for GERD have been approved by the Food and Drug Administration. One system uses an endoscopic sewing-machine device to plicate the gastroesophageal junction; the other system delivers radiofrequency (microwave) energy to create thermal lesions in the LES muscle, which may narrow the lumen and destroy nerves that mediate TLESRs. A number of other endoscopic antireflux procedures are under investigation. Small studies describe promising results,68,69 but the safety and efficacy of these procedures are not yet known, and their role in the treatment of GERD is not clear.
Barrett esophagus is a sequela of chronic GERD in which the stratified squamous epithelium that normally lines the distal esophagus is replaced by an abnormal columnar epithelium.70 The abnormal columnar epithelium typical of Barrett esophagus is an incomplete form of intestinal metaplasia called specialized intestinal metaplasia. This metaplastic epithelium predisposes to esophageal adenocarcinoma, which develops in patients with Barrett esophagus at the rate of approximately 0.5% a year.
The diagnosis of Barrett esophagus is established when the endoscopist sees columnar epithelium lining the distal esophagus. Columnar epithelium has a characteristic dull, reddish appearance that is readily distinguished from squamous epithelium, which is normally glossy and pale [see Figure 6]. The diagnosis is confirmed by esophageal biopsy specimens showing specialized intestinal metaplasia.
Figure 6. Barrett Esophagus
In Barrett esophagus, columnar epithelium extends proximal to the gastroesophageal junction (the imaginary line at which the esophagus ends and the stomach begins, which corresponds to the most proximal extent of the gastric folds).
Barrett esophagus can be further categorized according to the extent of esophageal involvement. In traditional, or long-segment, Barrett esophagus (LSBE), specialized intestinal metaplasia extends for 3 cm or more into the distal esophagus. LSBE is usually found in patients who have severe GERD. Less than 3 cm of metaplasia constitutes short-segment Barrett esophagus (SSBE), a condition often associated with only mild GERD. It is not clear whether these two types of Barrett esophagus have the same pathogenesis and natural history, nor is it clear whether SSBE progresses to LSBE. Currently, however, SSBE and LSBE are managed similarly.
Regular endoscopic surveillance for esophageal cancer has been recommended in patients with Barrett esophagus. Retrospective studies have documented that cancers discovered during surveillance endoscopies tend to be less advanced than those detected during endoscopies performed because of cancer symptoms (e.g., dysphagia and weight loss). There is no direct proof that surveillance reduces cancer mortality in Barrett esophagus, however.
Esophageal biopsy specimens are taken during surveillance endoscopy primarily to identify dysplasia, a histologic diagnosis suggesting that one or more clones of epithelial cells have acquired genetic alterations rendering them neoplastic and predisposed to malignancy. Unfortunately, dysplasia is an imperfect predictor of malignancy in Barrett esophagus. The histologic abnormalities of low-grade dysplasia are not specific for neoplasia, and interobserver agreement for the diagnosis of low-grade dysplasia in Barrett esophagus may be less than 50%. For high-grade dysplasia, in contrast, interobserver agreement is approximately 85%. Dysplasia has no distinctive gross features, so endoscopists must rely on random biopsy sampling techniques to find it. Consequently, biopsy sampling error is a major problem. Of patients with Barrett esophagus whose biopsy specimens show high-grade dysplasia, approximately one third already have an invasive cancer that was missed because of sampling error. Extensive biopsy sampling can reduce, but not eliminate, this problem. Finally, the natural history of dysplasia is not well defined.
Researchers have been searching for better alternatives to random biopsy sampling for dysplasia in Barrett esophagus. Other markers for malignant potential (e.g., flow cytometry and abnormalities in p53 expression) have been studied, as have endoscopic techniques to target dysplastic areas for biopsy. Despite some promising preliminary data, none of these tests and techniques has yet been shown to provide sufficient clinical information to justify its routine application in practice.
Fit patients with verified high-grade dysplasia in Barrett esophagus have three management options: esophagectomy, endoscopic ablative therapy, and intensive surveillance (withholding invasive therapy until biopsy specimens show adenocarcinoma). Esophagectomy, the only therapy that clearly can prevent the progression from dysplasia to invasive cancer, is associated with operative mortality of 3% to 12% and with a 30% to 50% rate of serious operative complications. Endoscopic ablative therapies (e.g., laser photoablation and photodynamic therapy) use thermal or photochemical energy to destroy the metaplastic esophageal epithelium.71 No study has shown that endoscopic ablation decreases the long-term risk for cancer development, so at present these therapies should be considered experimental. Although intensive surveillance for high-grade dysplasia (e.g., endoscopy every 3 months) has been endorsed as a management option by the American College of Gastroenterology, few published data directly support the safety and efficacy of the practice. The esophagus has an extensive lymphatic system and no confining serosa, features that can facilitate the spread of malignant cells and contribute to the dismal prognosis for patients with esophageal cancer. The concern about surveillance for high-grade dysplasia is that by the time biopsy specimens reveal adenocarcinoma, the tumor already may be incurable because of systemic metastases.
A management algorithm for patients with Barrett esophagus has been developed [see Figure 7].70 This strategy assumes that patients have had an initial endoscopic examination in which four-quadrant biopsy specimens are taken at intervals of 2 cm or less throughout the columnar-lined esophagus. If biopsy sampling during the initial endoscopic procedure is not adequate or if there is any question regarding the degree of dysplasia, endoscopy should be repeated to resolve these issues. If inflammation interferes with the histologic assessment of dysplasia, the patient should be treated with intensive antireflux therapy (e.g., a PPI administered at least twice daily) for 8 to 12 weeks before repeating the endoscopy. Data regarding the safety of intensive endoscopic surveillance for patients with high-grade dysplasia are limited, and available studies have involved primarily older patients. The clinician should be especially cautious in applying the results of these studies to the management of high-grade dysplasia in younger patients. Intensive endoscopic surveillance may be a valid alternative to immediate esophagectomy for older patients with high-grade dysplasia who can comply with the program. For patients who are too old, infirm, or unwilling to assume the risks of esophagectomy, endoscopic ablative therapy may be a reasonable alternative if the procedure is performed as part of a study protocol. Finally, the clinician should bear in mind that these recommendations have not been validated by studies demonstrating that this strategy prolongs survival or enhances quality of life.
Figure 7. Management of Barrett Esophagus
Management of patients with Barrett esophagus.73
Most esophageal infections are caused by Candida, herpes simplex virus (HSV), or cytomegalovirus (CMV), alone or in combination.6 These organisms rarely infect the esophagus of normal persons, but they often cause esophagitis in patients whose immune system has been compromised by AIDS, advanced malignancy, or the immunosuppressive drugs used to prevent rejection of organ transplants. Immune dysfunction that can accompany diabetes mellitus, alcoholism, and advanced age also may predispose to esophageal infection, especially byCandida. Antibiotic therapy that alters the normal microbial flora of the oropharynx and esophagus and corticosteroid therapy that suppresses immune function also can result in candidal esophagitis, as can abnormalities that delay the clearance of Candida from the esophagus, such as progressive systemic sclerosis (scleroderma), achalasia, and esophageal strictures.
The clinical manifestations of esophageal infections are similar, regardless of the pathogen. Dysphagia and odynophagia are the presenting symptoms in 60% to 95% of patients with infectious esophagitis, and weight loss is reported by 35%.6 CMV esophagitis often is only one component of a generalized CMV infection, and 20% to 40% of patients with CMV esophagitis have systemic symptoms [see Cytomegalovirus Esophagitis, below]. In contrast, Candida and HSV esophagitis usually are not associated with infection in other organs, and systemic symptoms are uncommon. However, oral lesions (e.g., thrush and focal ulcerations) are found frequently in patients who have Candida and HSV esophagitis, but they are not found in patients with CMV esophagitis.
A surprising number of esophageal infections are asymptomatic. In published series, approximately one quarter of all cases of candidal esophagitis were discovered incidentally during radiographic or endoscopic examinations performed for the evaluation of extraesophageal symptoms.
Candida, a yeast that is part of normal oropharyngeal flora, is the most frequent cause of esophageal infections in immunocompromised patients. Most cases of candidal esophagitis are caused by C. albicans, although other candidal species, such as C. tropicalis and C. glabrata, occasionally infect the esophagus. Approximately 85% of patients with candidal esophagitis have oral thrush, and the combination of oral thrush and esophageal symptoms has a high positive predictive value for candidal esophagitis.
Typically, endoscopic evaluation of patients with candidal esophagitis reveals raised, white plaques that resemble cottage cheese clinging to the esophageal mucosa. On barium swallow, coating of the raised plaques and their interstices with barium gives the esophageal mucosa a characteristic irregular, shaggy appearance. Confirmation of the diagnosis requires the demonstration of budding yeast cells, hyphae, or pseudohyphae in brush cytology or biopsy specimens of the esophagus.
Several antifungal agents are available for the treatment of Candida infections. The decision regarding which agent to choose is influenced principally by the severity of the infection and the severity of the patient's immunocompromise. A patient who has a mild esophageal infection and minimal immunocompromise (e.g., a young patient who develops mild candidal esophagitis during a limited course of steroid therapy for asthma) often can be treated effectively with a topical antifungal agent such as clotrimazole. In contrast, a patient with moderately severe candidal esophagitis and substantial immunocompromise (e.g., a patient with troublesome odynophagia and AIDS) usually requires the oral administration of a systemic antifungal agent such as fluconazole. Patients who have severe infection and profound immunocompromise (e.g., disseminated candidiasis or candidal esophagitis in the setting of severe granulocytopenia) generally require treatment with intravenous amphotericin B [see 7:XXXVIII Mycotic Infections in the Compromised Host].
HERPES SIMPLEX VIRUS ESOPHAGITIS
Primary HSV infections of the oropharynx are common in the general population, and they result when oral mucous membranes or breaks in the facial skin are exposed to secretions from a person with an active HSV infection. The virus enters the nerves that supply the infected epithelium, where it remains in latent form after healing of the primary infection. The latent virus in the neurons can be reactivated and spread to epithelial cells through the nerve fibers. In immunocompetent persons, HSV reactivation commonly causes cold sores of the lips (herpes labialis). When reactivation of latent virus occurs in the setting of immunodeficiency, however, HSV can spread to involve the squamous epithelium of the oropharynx and esophagus.
The endoscopic findings in persons with HSV esophagitis vary with the duration of infection.72 The earliest lesions are small (1 to 3 mm), rounded vesicles that usually involve the middle to distal esophagus. Sloughing of the vesicles results in small, sharply demarcated ulcers that have raised margins and a yellowish base. In severe cases, the small ulcers coalesce to form large ulcers that can be covered with dense exudates resembling candidal plaques.
Histologic diagnosis is best accomplished by examining biopsy and brush cytology specimens from the squamous epithelium at the edges of ulcerated areas. Specimens obtained from the ulcer base often contain only nonspecific granulation tissue and exudates. Typical histologic changes in HSV infection include multinucleated giant cells and intranuclear Cowdry type A inclusion bodies.
In immunocompetent patients, HSV esophagitis usually is a short-lived illness that may require no therapy other than supportive care and expectant management. HSV esophagitis often does not resolve spontaneously in immunocompromised patients, however, so such patients should receive systemic antiviral therapy. Acyclovir currently is the drug of choice for HSV infections of the esophagus.
CMV is a ubiquitous herpesvirus that usually is transmitted from person to person by exposure to infected secretions. The virus can also be transmitted through transfused blood that carries infected leukocytes or through transplanted infected organs. CMV can infect virtually any tissue in the body, and after recovery from the primary infection, evidence of latent CMV infection can be found in most organs. With the development of immunodeficiency, the latent virus can reactivate and cause esophagitis.73 Immunocompromised patients can also develop CMV esophagitis during primary CMV infections. CMV esophagitis is extremely uncommon in immunocompetent persons.
Patients who have CMV esophagitis often have widespread CMV infection, with systemic symptoms such as fever, weight loss, nausea, vomiting, and diarrhea. CMV tends to cause discrete, shallow esophageal ulcerations that are very elongated (up to 15 cm in length) and surrounded by normal-appearing esophageal mucosa. Tissue sampling for histologic examination and culture is necessary to distinguish these giant CMV ulcerations from the giant idiopathic esophageal ulcerations that can be associated with HIV infection.
Histologic examination of cells infected with CMV reveals distinctive abnormalities that include cellular enlargement and inclusion bodies in both the nucleus and the cytoplasm. Although the virus is found most often in fibroblasts and endothelial cells, biopsy specimens from granulation tissue in the base of the esophageal ulcer have a higher yield on histology and culture than specimens from squamous epithelial cells at the edges of the ulcer. Unfortunately, no single test for CMV infection is highly sensitive. In a study of 14 bone marrow transplant recipients who developed CMV disease, for example, conventional and centrifugation cultures of endoscopic biopsy specimens identified the organism in only 57% of patients, and conventional histologic examination of the specimens revealed characteristic findings in only 30%.74For patients with negative test results, therefore, repeated diagnostic testing may be necessary if the suspicion of CMV infection is high. However, evidence of CMV infection is not proof of the presence of CMV disease. The mere identification of CMV in an inflamed esophagus does not establish that CMV is the cause of the inflammation.
CMV disease can respond to treatment with ganciclovir. Maintenance therapy with ganciclovir may be indicated for patients who have recurrences of CMV disease or who have a high risk of recurrence (e.g., patients with advanced AIDS). Prophylactic antiviral therapy is commonly recommended for recipients of solid-organ and bone marrow transplants.
ESOPHAGEAL DISEASE IN HIV INFECTION
Within 2 to 3 weeks after primary exposure to HIV, some patients develop a self-limited, infectious mononucleosis-like illness with malaise, fever, myalgias, pharyngitis, and rash. This acute HIV seroconversion syndrome can be complicated by the development of esophageal ulcerations that cause odynophagia.75 Endoscopically, the ulcers are typically multiple, round, 3 to 15 mm in diameter, well demarcated, and surrounded by normal-appearing esophageal mucosa. Usually, the ulcers heal and the symptoms of the acute HIV seroconversion syndrome resolve spontaneously within 2 weeks. Patients then may remain asymptomatic for years until the development of AIDS.
Symptoms of esophageal disease occur in 30% to 40% of AIDS patients.76 Although the symptoms are usually from infections with Candida, HSV, or CMV, these patients can also have large esophageal ulcerations in which no pathogenic microorganism can be identified by culture or by histologic and immunohistochemical tests. Radiographically and endoscopically, HIV-associated idiopathic ulcerations of the esophagus closely resemble the large esophageal ulcerations caused by CMV. HIV-associated idiopathic esophageal ulcerations can be found in approximately 10% of patients with AIDS who complain of esophageal symptoms and in up to 40% of such patients who have discrete esophageal ulcerations on endoscopic examination.
HIV-associated idiopathic ulcerations generally do not respond to therapy with antimicrobial agents. Rather, patients with these lesions usually experience symptomatic relief and ulcer healing during treatment with systemic corticosteroids. Although corticosteroid therapy entails substantial risk for patients who already are profoundly immunosuppressed, the treatment is surprisingly well tolerated in most cases. The injection of methylprednisolone through the endoscope directly into idiopathic ulcerations also has resulted in relief of esophageal symptoms in some cases, but experience with this treatment is limited. Finally, thalidomide, which has immunomodulatory effects, has been used successfully to treat idiopathic ulcerations.
Candida is by far the most common cause of esophageal infection in AIDS, and candidal esophagitis is found in more than 50% of AIDS patients who have esophageal symptoms. Although CMV and HSV esophagitis also occur commonly in AIDS patients, these viruses often are not the sole pathogens that can be identified in the inflamed esophagus. CMV usually is discovered in biopsy specimens from an esophagus that is also infected by Candida, and most patients with coexistent CMV and candidal esophagitis respond well to antifungal therapy alone. Consequently, authorities have recommended that patients with AIDS who have esophageal symptoms should be treated empirically with antifungal therapy, usually fluconazole; endoscopy is reserved for patients who fail to respond to empirical treatment.
The two major histologic types of esophageal cancer, squamous cell carcinoma and adenocarcinoma, differ profoundly in their epidemiologic features.77 Squamous cell carcinoma of the esophagus has a strong predilection for blacks and Asians, whereas esophageal adenocarcinoma is predominantly a disease of whites. In the United States, the incidence of squamous cell carcinoma of the esophagus is six times greater in African Americans than in whites, whereas esophageal adenocarcinoma is at least four times more frequent in whites than in African Americans. Worldwide, more than 90% of all esophageal cancers are squamous cell carcinomas; this tumor ranks among the world's 10 most frequent malignancies. Exceptionally high incidence rates of squamous cell carcinoma are found in the Transkei region of South Africa, in southern Brazil, in parts of northern France and Italy, and throughout an esophageal cancer belt that extends from the shores of the Caspian Sea of Iran across northern China. In the Henan province of China, the incidence of esophageal squamous cell carcinoma exceeds 100 per 100,000. In the United States, in contrast, the incidence of this tumor in the general population is less than 4 per 100,000. In most countries, cancer of the esophagus affects men two to four times more often than women.
GERD and Barrett esophagus are the major risk factors for adenocarcinoma of the esophagus.70 For squamous cell carcinoma, cigarette smoking and alcoholism are the major risk factors.78 The combination of cigarette smoking and alcoholism appears to have a synergistic (rather than merely additive) effect in esophageal carcinogenesis, but the mechanism of this synergy is not known. Generalized malnutrition and a variety of specific nutritional deficiencies, including deficiencies in vitamin A, vitamin C, magnesium, selenium, and zinc, have been associated with squamous cell carcinoma. In contrast, obesity is a risk factor for adenocarcinoma. Carcinogens such as N-nitroso compounds can be formed from the nitrates and amines in pickled vegetables and cured meats, and ingestion of these foods has been linked to esophageal cancer. Regional practices such as opium smoking and the long-term ingestion of very hot foods and beverages may contribute to the pathogenesis of squamous cell cancers. Also, some high-incidence areas for squamous cell carcinoma have soils that are deficient in certain elements, such as molybdenum and zinc.
Local differences in endemic microflora have been proposed as underlying reasons for some of the regional variations in the incidence of esophageal squamous cell carcinoma. For example, the food and water in some high-incidence areas are contaminated with fungi and bacteria that promote the formation of N-nitroso compounds from dietary nitrates. The human papillomavirus (HPV) can infect squamous epithelial cells, and HPV infection has been implicated in the development of squamous cell carcinoma of the esophagus. In high-incidence regions for esophageal cancer, such as China and South Africa, researchers have found HPV DNA in more than 20% of squamous cell carcinomas. In low-incidence areas, such as the United States, however, esophageal tumors generally do not show evidence of HPV infection.
A number of medical conditions predispose to the development of esophageal squamous cell carcinoma. Patients with tylosis, a rare heritable disorder characterized by hyperkeratosis of the palms and soles, are at very high risk for development of the esophageal tumor. These patients have mutations in the tylosis esophageal cancer gene, a putative tumor suppressor gene located on the long arm of chromosome 17. Achalasia, lye stricture of the esophagus, and Plummer-Vinson, or Paterson-Kelly, syndrome also are risk factors for squamous cell cancers, perhaps because these conditions are associated with stasis of esophageal contents that leads to chronic inflammation of the mucosa. Squamous cell cancer of the esophagus is strongly associated with malignancies of the head, neck, and lungs, probably because these tumors share the strong risk factor of cigarette smoking. Finally, celiac sprue has been associated with esophageal cancer, for reasons that are not clear.
Most patients with cancer of the esophagus present with dysphagia and weight loss. The dysphagia usually involves solid foods only and progresses rapidly in severity (over a period of weeks to months). Approximately 60% of patients who have esophageal adenocarcinoma have a long-standing history of GERD symptoms. Proximal esophageal tumors can invade the recurrent laryngeal nerve, causing vocal cord paralysis with hoarseness. The development of coughing associated with swallowing may indicate that the tumor has invaded the airway and caused an esophagobronchial fistula. Ulcerated tumors can cause odynophagia, and tumor necrosis occasionally causes esophageal hemorrhage. Local tumor invasion can cause chest pain, and metastatic disease can cause bone pain. Symptoms of esophageal cancer generally develop only when the tumor has grown to the extent that it has narrowed the lumen of the esophagus substantially, has invaded local structures, or has metastasized. Therefore, the presence of symptoms usually indicates advanced disease and a poor prognosis.
Barium swallow and endoscopy
Both barium swallow and endoscopy are useful for the evaluation of patients with esophageal cancer. Radiographic features that suggest malignancy include irregular borders and sharp angles [see Figure 8]. Endoscopically, esophageal cancers typically appear as nodular lesions that protrude into the lumen of the esophagus [see Figure 9]. In Asian countries where there is a high incidence of esophageal cancer, endoscopists often recognize early esophageal cancers that cause either slight elevations or shallow depressions in the mucosal surface. Staining of the esophagus with vital dyes such as toluidine blue or Lugol iodine (chromoendoscopy) can be useful for finding such early lesions during endoscopic evaluation. These superficial esophageal cancers are diagnosed infrequently in western countries.
Figure 8. Barium Swallow Showing Esophgeal Cancer
Barium swallow showing an extensive cancer of the esophagus.
Figure 9. Esophageal Adenocarcinoma
Endoscopic photograph of an esophageal adenocarcinoma.
Computed tomography of the chest and abdomen generally is recommended to assess the extent of disease within the chest and to look for metastases. However, the sensitivity and specificity of CT for determining the depth of esophageal tumor penetration (the T level) and the presence of regional lymph node metastases (the N status) are poor. Endoscopic ultrasonography (EUS) is superior to CT in this regard, accurately predicting the T level and N status in 70% to 80% of patients.
In addition to EUS, invasive diagnostic modalities sometimes used for the staging of esophageal cancer include bronchoscopy, laparoscopy, thoracotomy, and thoracoscopy. There is little consensus regarding the need for these procedures in the routine evaluation of patients with esophageal cancer, and usage of the procedures varies widely among different institutions.
Cancer of the esophagus usually is disseminated at the time of diagnosis, and because there is no treatment that reliably eradicates metastatic disease, cure is not possible in most cases. Furthermore, patients are often elderly, and many have severe comorbidities (e.g., malnutrition or pulmonary, cardiac, or liver disease) that further limit their treatment options. Initial treatment usually involves a choice between surgery, radiation therapy, chemotherapy, and some combination of these three modalities.79,80 Squamous cell cancer and adenocarcinoma are treated similarly, with similarly poor survival rates. Despite recent advances in therapeutic options, overall cure rates for cancer of the esophagus remain below 10%.
Esophagectomy, with or without lymphadenectomy, can provide immediate palliation of symptoms and, arguably, the best potential for cure of esophageal cancer. Mortality for esophagectomy ranges from 3% to more than 12%, and serious complications of the operation (e.g., pneumonia, atelectasis, arrhythmias, myocardial infarction, heart failure, wound infections, and anastomotic leaks) can be expected in 30% to 50% of patients. Cure rates vary widely among institutions. Prognostic factors include tumor stage and the number of positive lymph nodes. Surgery generally is not recommended for patients who have metastatic disease.
The acute mortality of radiation therapy is low, and radiation can cover a wider treatment area than is practical with surgery (to eradicate local and regional disease). However, radiation therapy usually takes 2 to 8 weeks to complete; palliation can be delayed for weeks; there can be substantial radiation damage to surrounding normal tissues; and the overall cure rate is low. Trials of radiation therapy as the sole treatment modality for esophageal cancer have involved primarily patients with advanced squamous cell carcinomas that were deemed unresectable. Results appear to be comparable to surgery, but there are no randomized trials directly comparing radiation therapy alone with surgery alone.
Chemotherapy has the potential to reach the disseminated disease that usually is present in symptomatic patients. Unfortunately, chemotherapy is associated with substantial morbidity and considerable mortality; it is often ineffective; and the tumor response, if any, is often brief. Studies of chemotherapy as the sole treatment modality have included primarily patients with unresectable tumors. Modern studies have used cisplatin-based regimens, and response rates appear to be better with combination regimens than with single agents. Chemotherapy alone does not appear to improve survival, however.
A number of studies have explored the role of radiation therapy or chemotherapy used either before (neoadjuvant) or after (adjuvant) definitive surgery for squamous cell carcinoma of the esophagus. Unfortunately, most randomized, controlled trials have shown no convincing benefit for neoadjuvant or adjuvant treatment with either radiation therapy or chemotherapy for patients with potentially resectable tumors. One recent trial showed a modest survival benefit for patients who received preoperative chemotherapy in a relatively low dose, however.80
Much recent interest has focused on the role of combining chemotherapy with radiation therapy (chemoradiotherapy) for esophageal cancer.81 In some studies of patients treated with chemoradiotherapy followed by esophagectomy, complete histologic response (defined as no histologic evidence of tumor in the resected specimen) has been observed in almost 30% of cases. Complete histologic response is not tantamount to cure, however, and even complete responders frequently succumb to recurrent disease. Furthermore, chemoradiotherapy is associated with serious toxicity. Some randomized trials of chemoradiotherapy for patients with potentially resectable tumors have shown significant improvements in survival, whereas others have not. Consequently, the role of chemoradiotherapy remains unclear. Preliminary studies suggest that patients who have locally advanced tumors might benefit from preoperative chemoradiotherapy.
Purely palliative therapies include esophageal dilatation and the placement of intraluminal stents. There also are palliative techniques designed to ablate the portion of the neoplasm that obstructs the esophageal lumen. These ablative therapies include endoscopic laser irradiation, the application of tumor probes that burn the neoplasm directly, the injection of caustic chemicals directly into the tumor body, and photodynamic therapy that uses photochemical energy to destroy the tumor.
Given that the optimal treatment for cancer of the esophagus is not clear, patients should be treated according to well-designed, established research protocols whenever possible. If the initial use of research protocols is not feasible, management should be individualized [seeFigure 10]. After staging of the tumor that includes at least EUS and a CT scan of the chest and abdomen, the next step is to decide whether the patient is fit enough to undergo surgery. If surgery is not a viable option because of advanced age or comorbidity, primary therapy might include chemoradiation or supportive care alone. In general, surgery is not indicated for patients with metastatic disease. For tumors that do not invade beyond the muscularis propria and do not involve local lymph nodes, surgery appears to offer the best hope for cure. For lesions that are more advanced (i.e., with lymph node involvement or invasion to the esophageal adventitia and beyond), the choices for primary therapy include chemoradiation with or without surgery.
Figure 10. Management of Patients with Cancer of the Esophagus
Management of patients with cancer of the esophagus. (CRTâ€”chemoradiotherapy)
If these primary treatments fail or if the tumor recurs, there are a number of palliative treatment options. For patients who are severely debilitated and have advanced disease, the most humane option may be supportive care only, with careful attention to pain control. If there are no apparent metastases and complete excision of the tumor is possible, surgery can be considered for palliation. The other options are ablative therapies or stents. Stents may not provide good palliation for patients with proximal tumors, bleeding tumors, and necrotic tumors; ablative therapy may be preferable in these circumstances. Alternatively, ablative therapy has little to offer for a patient with an esophagobronchial fistula or for a patient with a tumor that is extrinsic or infiltrative. Also, ablative therapy may be difficult and time-consuming for patients with very long tumors. Stenting may be preferable in these circumstances.
Figures 4, 8, and 9 American Gastroenterological Association.
Figure 6 Seward Hung.
Editors: Dale, David C.; Federman, Daniel D.