Definition and classification
hiatal hernia is an extension of the peritoneal cavity into the chest through the esophageal hiatal of the diaphragm. The hernia consists of a peritoneal sac which usually contains part or all of the stomach but can rarely contain other abdominal viscera such as the colon or spleen. hiatal hernias are classified according to the position of the gastroesophageal junction in relation to the diaphragm. The type I or sliding hiatal hernia is characterized by proximal herniation of the gastroesophageal junction though the esophageal diaphragmatic hiatus into the posterior mediastinum. In a type II hernia, also known as a rolling or pure paraesophageal hernia, the gastroesophageal junction is anchored in its normal subdiaphragmatic position, and the gastric fundus herniates alongside the esophagus into the chest. A type III hernia, also known as a combined sliding/rolling, or mixed hernia, is usually a large hernia and is characterized by proximal herniation of both the gastroesophageal junction and the gastric fundus through a widely open diaphragmatic hiatus. Type III hernias are commonly but inappropriately classified as type II paraesophageal hernias, but a pure type II hernia is a rare finding. The endstage of a type I or II hernia is a large type III hernia, which occurs when the entire stomach migrates up into the chest by rotating 180° around its longitudinal axis, with the gastroesophageal junction and pylorus as fixed points. The greater curvature is thus uppermost in the chest. In this situation, the abnormality is usually referred to as an intrathoracic stomach with organoaxial volvulus. In a type IV hernia, the hiatal defect is very large and has allowed other organs such as the colon, spleen, small intestine, and pancreas to enter the chest. Type IV hernias develop from type III hernias, and are exceedingly uncommon.
Prevalence of hiatal hernia
The majority of hernias, probably more than 95 per cent, are type I sliding hernias. The exact prevalence of hiatal hernia, both in the general population and in patients with foregut symptoms, remains unclear. Reasons for this uncertainty are that most hernias are small sliding hernias which reduce in the upright position, most are asymptomatic, and the detection rate varies with the method used for diagnosis and with the age of the population studied. The reported prevalence of hiatal hernia detected by barium contrast studies in the general population ranges from 10 per cent to more than 70 per cent. It is likely that some of these studies, especially if they are based on static rather than videotaped imaging, classified some physiologic hernias as pathologic hernias. A reasonable estimate of the prevalence of pathologic herniation in the general adult population is 20 per cent, with the prevalence of hernias in patients with gastroesophageal reflux disease being 80 per cent. The mean age of patients with a type I hernia is approximately 50 years and for those with a type III hernia is approximately 60 years. Type III hernias are more common in females, with a female:male ratio of 4:1.
Pathophysiology
A pathologic hiatal hernia is an acquired condition. Anatomic alterations found in patients with hiatal hernia are widening of the hiatus and thinning and lengthening of the phrenoesophageal membrane. The phrenoesophageal membrane consists of an upper fascial layer of loose connective tissue and a lower fascial layer of thicker, stronger, elastic tissue. The upper layer is formed by a supradiaphragmatic continuation of the endothoracic fascia. The more important lower layer is formed by a subdiaphragmatic continuation of the transversalis fascia. The phrenoesophageal membrane functions as an elastic barrier. In comparison with the other diaphragmatic openings, the esophageal hiatal is relatively incompletely filled by the esophagus, and the barrier function of the phrenoesophageal membrane is needed to close the gap between esophagus and diaphragm. The elastic function of the phrenoesophageal membrane allows it to stretch and contract with swallowing and respiration. There are normally more than 1000 swallows per day, and the diaphragm moves up and down between 16 000 and 35 000 times per day. The lengthening and loss of elasticity of the phrenoesophageal membrane that is found in a hiatal hernia is thought to be a gradual age-related wear and tear effect, but obesity, pregnancy, and sudden traumatic elevations in the intra-abdominal pressure may all place additional pressure on the phrenoesophageal membrane.
Mixed or type III hernias are thought to be complications of type I hernias, and are more likely in patients who have an abnormally wide esophageal hiatus. Pure paraesophageal or type II hernias are thought to arise by the same mechanism as sliding hernias, with the difference that the gastroesophageal junction is fixed below the hiatus by firm posterior fixation to the preaortic fascia and the median arcuate ligament. Another difference is that the herniation may occur either through a widened esophageal hiatus or through a defect adjacent to the esophageal hiatus. The fundus of the stomach rolls up alongside the esophagus through the widened hiatus or the adjacent defect, but the gastroesophageal junction remains in its normal intra-abdominal position.
A hiatal hernia predisposes to the development of gastroesophageal reflux by four mechanisms. First, contraction of the crural components of the diaphragm augments the antireflux mechanism by buttressing the lower esophageal sphincter, adding the pressure produced by crural contraction to the underlying pressure in the lower esophageal sphincter. The presence of a hiatal hernia interferes with this buttressing because the normal alignment between the crura and the lower esophageal sphincter is lost. This encourages reflux during periods of increased intra-abdominal pressure. Second, the formation of a hiatal hernia results in loss of the acute angle of His. Third, the clearance of refluxed acid from the esophagus is also impaired in patients with hiatal hernias, especially non-reducing hernias, due to loss of anchoring of the distal esophagus. Fourth, the presence of a hiatal hernia interferes with transhiatal flow of gastric juice. This occurs in particular in patients with a non-reducing hernia, in whom the hernia remains above the diaphragm during inspiration. Fluid trapped within the supradiaphragmatic stomach is prevented from flowing into the subdiaphragmatic stomach by closure of the crura during inspiration, and thus refluxes freely into the negative-pressure thoracic esophagus.
Symptoms
Symptoms in patients with a sliding hernia are mostly those of the associated gastroesophageal reflux. The likelihood of reflux increases with increasing hernia size. Dysphagia and intermittent vomiting can occur in patients with a large sliding hernia, due to diaphragmatic contraction obstructing the passage of a bolus through the herniated stomach. Dysphagia is common in patients with type II hernias, due to compression and angulation of the esophagus by the herniated gastric fundus. In large type III hernias with gastric volvulus, dysphagia can be caused by twisting of the gastroesophageal junction. Recurrent bleeding or anemia may be caused by ulceration or erosive gastritis secondary to ischemia of the herniated gastric mucosa. Crural contractions may also cause ‘riding ulcers' at the level of the diaphragmatic indentations. Respiratory symptoms such as dyspnea may occur from compression of the lungs by a very large hernia, and cough, choking, and pneumonia may result from aspiration of regurgitated gastric juice into the airways.
Type II or III hernias can cause acute life-threatening events in up to 20 per cent of patients due to acute obstruction from gastric torsion or volvulus. Swallowed air causes gastric distention, which compresses the blood supply passing to the intrathoracic stomach along the margins of the diaphragmatic hiatus. This can result in gastric ischemia, perforation, and sepsis. The triad of epigastric pain, inability to vomit, and difficulty in the ability to pass a nasogastric tube is an indication for immediate endoscopic decompression or operative intervention.
Diagnosis
Radiology
Large hernias may be detected on an upright or lateral plain chest radiograph by the presence of a gas-filled viscus behind the heart. Smaller hernias require a barium swallow study for diagnosis. Barium studies are diagnostic in more than 90 per cent of patients. The barium swallow study should include examination in both upright and supine positions, to detect sliding hernias that reduce when upright. Video-esophagram studies demonstrate that most sliding hernias empty during expiration.
Endoscopy
A hiatal hernia is present endoscopically when the anatomic gastroesophageal junction, identified as the proximal extent of the gastric rugal folds, is more than 2 cm above the crura, identified by having the patient sniff. Small hernias are best detected endoscopically if the endoscope is retroflexed within the stomach, allowing inspection of the gastroesophageal junction from below. The geometry of the gastroesophageal junction seen on retroflexed view can be classified using the Hill grading system. This classification system emphasizes that development of a sliding hernia is accompanied by loss of the angle of His, signified by the loss of the normal frenulum over the endoscope on retroflexed viewing. A true paraesophageal hernia is identified on retroflexed view by noting a hernia orifice adjacent to the gastroesophageal junction. A type III hernia appears on retroflexed view as a sliding hernia with the gastroesophageal junction entering on the side of the sac, rather than at the apex of the sac as in a type I hernia.
Manometry
Stationary manometry studies are not used for the diagnosis of hiatal hernia, but an awareness of manometric features of hiatal hernia is helpful since many patients undergoing manometry for the investigation of reflux disease have a hiatal hernia. A ‘double hump' motility pattern is characteristic of a sliding hernia. As the manometry catheter is withdrawn from the stomach into the esophagus, pressurized areas or ‘humps' are recorded, with the distal hump corresponding to the crural compression of the stomach and the proximal hump corresponding to the lower esophageal sphincter. Between the two humps is a plateau region in which the pressure is slightly above the gastric baseline pressure. The plateau region corresponds to the hernia sac. The pressure fluctuations that occur with respiration indicate that the hernia sac may be exposed to both intra-abdominal and intrathoracic pressure conditions, depending upon whether the crura compartmentalize the herniated proximal stomach from the distal stomach and peritoneal cavity.
Treatment
Sliding hernias do not require treatment unless they produce symptoms specific to the presence of the hernia. In patients with gastroesophageal reflux disease, reduction of the hernia and closure of the hiatus is an essential part of operative treatment, and as a consequence the hernia is repaired. Patients with large hernias undergoing antireflux surgery are likely to have a shortened esophagus requiring special operative techniques. The high rate of serious or life-threatening complications in patients with type II or III hernias indicates that operative repair should be performed in all patients with these hernias, regardless of the presence or severity of symptoms or the size of the hernia. Type II or III hernias progressively worsen unless repaired by operation. There is an approximately 20 per cent mortality for emergency repair of an incarcerated paraesophageal hernia, compared with less than 1 per cent mortality for elective repair of a non-incarcerated hernia.
Repair of type II or III hernias by laparoscopic methods can be difficult and is associated with a higher rate of recurrence. We prefer an open approach to these large hernias. An open transthoracic approach allows full esophageal mobilization, resection of the sac, and proper closure of the wide hiatus. An open approach also facilitates performance of a Collis gastroplasty if this is needed to gain adequate esophageal length to reduce tension on the repair. The transabdominal open approach facilitates reduction of a volvulus, but it may be difficult to obtain an adequate view for dissection in the posterior mediastinum, the risk of vagus nerve injury is high, and stout closure of the widened hiatus may be compromised.
Lifestyle Modifications
Sensible changes in lifestyle, especially if their rationale is explained to the patient, should be part of the initial management of all subjects. These include head of the bed elevation, avoidance of tight-fitting clothes, weight loss, restriction of alcohol, elimination of smoking, dietary therapy, refraining from lying down after meals, and avoidance of evening snacks before bedtime. Physiological studies show that these maneuvers enhance esophageal acid clearance, minimize acid-reflux related events, or ease heartburn symptoms, but their therapeutic efficacy in controlled trials usually has not been evaluated. The head of the bed can be elevated either by putting 6- to 8-inch blocks under the legs of the bed or by using a Styrofoam wedge under the mattress to elevate the upper torso. Eating several hours before retiring and avoiding evening snacks keep the stomach empty at bedtime, thereby decreasing the number of nocturnal reflux episodes. These three lifestyle changes are recommended for patients with nocturnal GERD symptoms or laryngeal complaints. One study found that head of the bed elevation was nearly as effective as ranitidine therapy in healing esophagitis. Avoidance of tight-fitting clothes and weight loss are interventions aimed at reducing the incidence of reflux by the abdominal stress mechanism. The efficacy of weight reduction is especially controversial, but it may be helpful when discrete periods of weight gain can be associated with exacerbation of reflux symptoms. Cessation of smoking and elimination of alcohol are valuable because both agents lower LES pressure, reduce acid clearance, and impair intrinsic squamous epithelial protective functions. Dietary changes include reducing the size of the meal and intake of fats, carminatives, and chocolate, to reduce the frequency of reflux by decreasing gastric distention and by reducing the episodes of transient LESRs, and avoiding foods that lower basal LES pressure. Additionally, some patients complain of heartburn after consuming citrus drinks, spicy foods, tomato-based products, coffee, tea, or cola drinks. Stimulation of gastric acid or esophageal sensitivity to low pH or hyperosmolar liquid solutions may account for these symptoms. However, the indiscriminate prohibition of food products should be avoided, but rather tailored to those foods that bring on individual symptoms, to promote dietary compliance. Finally, patients should avoid, if possible, drugs that lower LES pressure or can promote localized esophagitis.
The rationale for GERD therapy depends on a careful definition of specific aims. In patients without esophagitis, the therapeutic goals are simply to relieve the acid-related symptoms and to prevent frequent symptomatic relapses. In patients with esophagitis, the goals are to relieve symptoms and to heal the esophagitis while attempting to prevent further relapses and the development of complications. These goals are set against a complex background: GERD is a chronic disease that may wax and wane in intensity, and relapses are common.
Nonprescription Therapy
Although GERD is common in the
Over-the-Counter Medications Over-the-counter antacids, Gaviscon, and H 2RAs are useful in treating mild and infrequent heartburn symptoms, especially when symptoms are brought on by lifestyle indiscretions. Antacids increase LES pressure but work primarily by buffering gastric acid in the esophagus and stomach, albeit for relative short periods. Heartburn symptoms are rapidly relieved, but patients need to take antacids frequently, usually 1 to 3 hours after meals and at bedtime, depending on symptom severity. Gaviscon, containing alginic acid and antacids, mixes with saliva to form a highly viscous solution that floats on the surface of the gastric pool and acts as a mechanical barrier. Both antacids and Gaviscon are more effective than placebo in relieving symptoms induced by a heartburn-promoting meal. However, these agents do not heal esophagitis, and long-term trials suggest effective symptom relief in only 20% of patients using antacids. Side effects of antacids and Gaviscon are minimal but include diarrhea from magnesium-containing antacids, constipation from aluminium-containing antacids, salt overload, magnesium or aluminium toxicity in patients with renal disease, and the milk-alkali syndrome (hypercalcemia, alkalosis, renal failure) from long-term and excessive use of calcium-containing antacids. H 2RAs are available in an OTC form at doses that are usually one half of the standard prescription dose. Although there are some differences in potency, duration, and rapidity of action, these drugs may be used interchangeably. Although their onset of relief is not as rapid as that of antacids, the OTC H 2RAs have a longer duration of action, up to 6 to 10 hours. Therefore, they are particularly useful when taken before a potentially refluxogenic activity, such as a heavy meal or exercise. Like antacids, the OTC H 2RAs are ineffective in healing esophagitis and should not be used regularly for more than 2 weeks.
Patients with frequent heartburn, esophagitis, or complications of GERD usually see a physician and receive prescription medications for their disease. Although prokinetic drugs attempt to correct the motility disorder associated with GERD, the most clinically effective medications for short- and long-term reflux treatment are the acid suppressive drugs.
Prokinetic Drugs
Until recently, three prokinetic drugs were available for the treatment of GERD: bethanechol, a cholinergic agonist; metoclopramide, a dopamine antagonist; and cisapride, a serotonin (5-HT 4) receptor agonist, which increases acetylcholine release in the myenteric plexus. These drugs improve reflux symptoms by increasing LES pressure, acid clearance, or gastric emptying. However, none alter the frequency of transient LESRs, and their physiological activity decreases as the disease severity worsens. Therefore, all the current prokinetics provide modest benefit in controlling heartburn, but they have little efficacy in healing esophagitis unless they are combined with an acid inhibiting drug. The use of prokinetic drugs is limited by their side effect profiles. Bethanechol commonly causes flushing, blurred vision, headaches, abdominal cramps, and urinary frequency. Metoclopramide, which crosses the blood-brain barrier, has a 20% to 50% incidence of fatigue, lethargy, anxiety, and restlessness and rarely causes tremor, parkinsonism, or tardive dyskinesia. It is possible to decrease the frequency of these side effects by dose reduction, by increasing the dosing regimen to twice a day, by taking a larger single dose before dinner or at bedtime, or by using a sustained-release tablet. Domperidone, another dopamine antagonist and one that does not cross the blood-brain barrier, has fewer side effects, but it is not available in the
Histamine 2 Receptor Antagonists
Cimetidine, ranitidine, famotidine, and nizatidine reduce acid secretion by competing with histamine receptors on the parietal cell. They are most effective in controlling nocturnal, as compared with meal-related acid, secretion because the parietal cell may also be stimulated postprandially by acetylcholine and gastrin. All the H 2RAs are equally effective when used in proper doses, usually twice a day before meals. Clinical GERD trials show that heartburn, both day and night, can be significantly decreased by H 2RAs, when compared with placebo, although symptoms are rarely abolished. Trials and a metaanalysis found that the overall esophagitis healing rates with H 2RAs rarely exceeded 60% after up to 12 weeks of treatment, even when higher than standard doses were used. Healing rates differ in individual trials, depending primarily on the degree of esophagitis being treated: grade I and II esophagitis heals in 60% to 90% of patients, whereas grade III and IV heals in 30% to 50% of patients despite high-dose regimens.
Reflux symptoms associated with nocturnal gastric acid breakthrough during PPI therapy have been recognized. At bedtime, H 2RAs successfully eliminated this problem, suggesting a new indication for H 2RAs in the PPI era. However, this study used only a single evening dose and did not account for the tolerance that frequently develops to H 2RAs over weeks to months. This may impair the ability of chronic long-term nocturnal dosing of H 2RAs to eliminate acid breakthrough symptoms, but it suggests an important clinical role as medications used on an as-needed basis when lifestyle indiscretions may promote nocturnal symptoms. As a class of drugs, the H 2RAs are very safe, with a side effect rate (most of which are minor and reversible) of about 4%. There have been some concerns about drug interactions with these agents. Serum concentrations of phenytoin, procainamide, theophylline, and warfarin are altered after the administration of cimetidine and, to a lesser degree, ranitidine, whereas this interaction is not reported with the other two H 2RAs. The former concern that these agents could alter blood ethanol levels has been discounted.
Proton Pump Inhibitors
This class of drugs markedly diminishes gastric acid secretion by inhibiting the final common pathway of acid secretion, the H +, K +-ATPase pump. PPIs inhibit daytime, nocturnal, and meal-stimulated acid secretion to a significantly greater degree than H 2RAs, but they rarely make patients achlorhydric. Unlike H 2RAs, the degree of acid inhibition with PPIs does not correlate with plasma concentration, but it is related to the concentration and duration. After oral ingestion, acid inhibition is delayed because PPIs need to accumulate in the secretory canaliculus of the parietal cell to bind irreversibly to actively secreting proton pumps. Therefore, the slower a PPI is cleared from the plasma, the more of it is available for delivery to the proton pumps. PPIs are best taken before the first meal of the day, when most proton pumps are active. Because not all pumps are active at any given time, a single PPI dose does not inhibit all pumps. A second dose, if needed, can be taken before the evening meal. The five available PPIs are omeprazole, lansoprazole, rabeprazole, pantoprazole, and esomeprazole, the S-isomer of the racemic omeprazole. Their superior efficacy compared with H 2RAs is based on their ability to maintain an intragastric pH high than 4 between 15 to 21 hours daily compared with approximately 8 hours daily with the H 2RAs. Multiple studies show that the PPIs are superior to H 2RAs in completely relieving heartburn symptoms in most patients with severe GERD, usually within 1 to 2 weeks. Symptom relief is slightly better in patients with erosive as compared with nonerosive disease. Controlled studies and a large metaanalysis report complete healing of even severe ulcerative esophagitis after 8 weeks in more than 80% of patients taking PPIs compared with 51% of patients taking H 2RAs and 28% receiving placebo. In those patients not healing initially, prolonged therapy with the same dose or an increased dose usually resulted in 100% healing. Studies with PPIs consistently show that they are superior to both regular and high-dose H 2RA therapy, but therapeutic efficacy among PPIs is similar. However, a large study found that the newest PPI, esomeprazole, at a dose of 40 mg, was superior to the parent compound (omeprazole, 20 mg) for complete healing of esophagitis at 4 and 8 weeks. This superiority over omeprazole is related to higher systemic bioavailability and less interpatient variability with esomeprazole. One PPI, pantoprazole, is available in the
Surgical Treatment of Acid Reflux Disease
Antireflux surgery reduces GER by increasing basal LES pressure, decreasing episodes of transient LESRs, and inhibiting complete LESR. This is accomplished by reducing the hiatal hernia back into the abdomen and thereby restoring an adequate length of intra-abdominal sphincter, reconstructing the diaphragmatic hiatus, and reinforcing the LES. Before the explosion of laparoscopic surgery, the three most common operations were the Nissen fundoplication, the Belsey Mark IV repair, and the Hill posterior gastropexy repair. Since the advent of minimally invasive surgery, the two most popular procedures are the Nissen fundoplication and the Toupet partial fundoplication. The former is a superior operation with more long-term durability, but it has a higher frequency of postoperative dysphagia and gas bloat symptoms. Both are now routinely performed laparoscopically through the abdomen. The hospital stay is 1 to 2 days, and many patients return to normal activity in 7 to 10 days. Patients with more severe disease and a short esophagus manifested by a large nonreducible hernia, a tight stricture, or a long-segment Barrett esophagus require a Collis lengthening procedure creating a 3- to 5-cm neoesophagus, so the fundoplication can be placed in the abdomen under minimal tension.
The common indications for antireflux surgery have evolved with the availability of more powerful medications for treating GERD. In the PPI era, the resolution of symptoms on treatment helps to predicate the success of antireflux surgery for both classical and atypical symptoms. Antireflux surgery is a reasonable option in the following situations:
Extensive physiological testing should be done before antireflux surgery is performed. All patients need endoscopy to exclude stricture, Barrett esophagus, and dysplasia. A barium esophagram can help define a nonreducible hernia, shortened esophagus, and poor esophageal motility. Esophageal manometry will identify ineffective esophageal peristalsis and previously misdiagnosed achalasia or scleroderma. Twenty-four hour pH testing is needed in all patients with nonerosive GERD or in those with esophagitis not responding to PPI therapy. Gastric analysis and gastric emptying studies may be indicated in selected patients. Careful testing will result in modification of the original operation or an alternative diagnosis in approximately 25% of patients.
Antireflux surgery relieves reflux symptoms and reduces the need for stricture dilation in more than 90% of patients, but Barrett esophagus rarely regresses, and the influence of surgery on the development of esophageal adenocarcinoma is controversial. Comparison studies have found antireflux surgery superior to lifestyle changes, antacids, and H 2RA and prokinetic therapy, but not PPI therapy, especially when dose titration is permitted. Mortality is rare (lesser than 1%) after antireflux surgery, but new postoperative complaints can occur in up to 25% of patients including dysphagia, gas bloat, diarrhea, and increased flatus. Most symptoms improve over 1 year, but persistent complaints suggest too tight a wrap, a displaced fundoplication, or inadvertent damage to the vagus nerve. Successful antireflux surgery, however, does not guarantee a permanent cure. The best surgical results are obtained by experienced surgeons in high-volume centers who report long-term symptom recurrence in only 10% to 15% of patients. However, most operations are performed by community or Veterans Affairs hospital surgeons. Here the results are not as good with studies finding relapse of esophagitis in 30% of patients and return to regular use of antireflux medications in 62% of patients 10 to 15 years after fundoplication. Potential factors contributing to these high relapse rates include inexperienced surgeons, low numbers of operations yearly per surgeon, and persistence of abdominal stressors (i.e., obesity, heavy isometric exercise or work) that gradually weaken the fundoplication. A suboptimal operation or severe symptom relapse may necessitate a second operation, which has less likelihood of a successful outcome. Because optimal medical therapy is available to all patients with GERD, the risk and benefits of both long-term medical treatment and antireflux surgery must be carefully discussed with patients, so they can take part in this important decision.
Acid Reflux Treatment in Elderly Patients
Older patients often complain of less severe reflux symptoms than their younger cohorts, but because of prolonged acid exposure over years, the elderly may have more complicated disease. Treatment of the older patient with GERD follows the same principles as in other adults, although they may require more aggressive acid suppression therapy. Pill-induced esophagitis may complicate their treatment. Metoclopramide must be used with caution because of frequent side effects in the elderly. H 2RAs can be associated with mental changes in older patients, and doses need to be decreased in patients with renal insufficiency. Fewer drug interactions are seen with famotidine and nizatidine. Alternative methods of administering PPIs may be necessary in debilitated older patients who cannot swallow intact omeprazole or lansoprazole capsules. Both capsules can be opened and the granules taken with water, an HCO 3 --based suspension, or apple or orange juice, or the granules can be sprinkled on applesauce or yogurt.
Teratogenicity or fetal harm from absorption of medications across the placenta is the foremost consideration in the treatment of GERD during pregnancy. Lifestyle modifications and antacids or Gaviscon remain the cornerstones of treatment, providing adequate relief to the majority of women with mild symptoms. Although rarely used in adults, sucralfate is a nonabsorbable mucosal binder that has been found superior to lifestyle changes in a controlled study in pregnant women. Metoclopramide, H 2RAs, and most PPIs (except omeprazole) have a Category B FDA safety profile for use during pregnancy, based on animal studies showing no risk, as well as on small case series and anecdotal human reports. Ranitidine is the only one of these drugs shown to be effective during pregnancy. PPIs may be safe for aspiration prophylaxis before anesthesia for elective cesarean sections. Antacids, sucralfate, and most H 2RAs (except nizatidine) are safe to use during lactation, even though the latter group of drugs is excreted in breast milk. PPIs are not recommended during breast-feeding, based on safety concerns in animal studies.
Extraesophageal Presentations
Acid reflux–related chest pain is easily treated by H 2RAs or PPIs, with efficacy substantiated by placebo-controlled studies. The efficacy of acid suppression therapy in asthma, cough, and other pulmonary complications of GERD is more mixed. Medical antireflux therapy improves asthma symptoms and reduces the need for asthma medications in more than 60% of patients, but objective improvement of peak expiratory flow rates is observed in only 25% of patients. Best results are found with higher doses of PPIs (usually twice-daily administration) given for 2 to 3 months. Potential positive predictors of PPI response include asthma that is difficult to control, associated acid regurgitation, proximal reflux on pH testing, and healing of esophagitis with antireflux therapy. Case studies report that 60% to 96% of patients with suspected acid-related ear, nose, and throat symptoms and signs improve with acid suppression. Here again, PPIs are more effective than H 2RAs, and extended therapy for up to 3 months may be required. Predictors of response have not been identified, although patients with milder laryngeal signs show better symptom improvement. In all these possible extraesophageal presentations of GERD, failure to respond to aggressive PPI therapy, confirmed by adequate acid control by pH testing, suggests a cause of these complaints other than acid.
Esophageal Strictures
Dysphagia in patients with esophageal strictures is related to stricture diameter and severity of esophagitis. When the esophageal lumen diameter is less than 13 mm, dysphagia is a major complaint and esophageal dilation is required. Simple short strictures can be dilated by blind peroral passage of rubber Hurst (round ends) or Maloney (taper ends) mercury-filled dilators of increasing sizes (16 to 60 French, 3 French = 1 mm) to disrupt the fibrous bands producing the obstruction. Complicated longer, tighter, or more irregular strictures will require bougienage over a guidewire using hollow-centered Savary plastic-covered polyvinyl dilators or balloon (Gruentzig) dilators. Before and after dilation, medical therapy with PPI is indicated; it has been shown to be superior to H 2RAs in relieving symptoms and in reducing the frequency for repeat dilations. Maintenance PPI therapy for patients with strictures has dramatically reduced the incidence of repeat esophageal dilations and the cost of treating these patients. Recalcitrant strictures, requiring surgery, are now very uncommon and suggest another aggravating factor such as chronic pill injury.
Barrett Esophagus In general, esophagitis in the presence of Barrett esophagus can be easily healed with PPI therapy, but meaningful regression of Barrett epithelium, except for small squamous islands, is rarely reported even with high-dose PPI therapy. Recent ex vivo studies have suggested that intermittent “pulses” of acid result in enhanced Barrett epithelial cell proliferation, possibly increasing the risk of dysplasia and cancer. Some have suggested the need to eliminate all acid reflux in patients with Barrett esophagus, but this would require high and frequent doses of expensive medications and serial pH monitoring to document efficacy of therapy. Pending further clinical studies, patients with Barrett esophagus should be treated like others with chronic GERD. Esophageal resection of Barrett esophagus can prevent the progression to cancer, but this requires a total esophagectomy with high mortality, except in selected surgical centers. Therefore, ablation of Barrett epithelium in the setting of strict PPI anacidity has been proposed. Photodynamic therapy, laser, multipolar electrocoagulation, argon plasma coagulation, and endoscopic mucosal resection have been used for this purpose. In these studies, Barrett mucosa can be reversed completely in 70% to 80% of patients, but intestinal metaplasia underlying the new squamous mucosa is reported in almost all series, with the occasional residual foci of metaplasia developing adenocarcinoma. Adverse effects of ablation therapy have ranged from mild chest pain, sore throat, or odynophagia to esophageal perforation and death. The incidence of adenocarcinoma in patients with Barrett esophagus without dysplasia is probably so low that endoscopic ablation cannot be advocated outside of study protocols. Endoscopic therapy for patients with high-grade dysplasia or early cancer holds more promise, especially in the older patients with comorbid illnesses. Because there is currently no way of eliminating the malignant risk of Barrett esophagus, regular endoscopic surveillance is recommended. Biopsies should be taken from each quadrant every 2 cm axially within the metaplastic tissue. The rationale is that dysplasia within Barrett epithelium is often multifocal, and obtaining fewer tissue samples increases the risk of missing dysplastic areas. Brush cytology can compliment endoscopic biopsies. Biomarkers, such as p53, and flow cytometry may augment the yield of histological examination of biopsy specimens. Although prospective studies are not available, case series confirm that esophageal adenocarcinomas detected by endoscopic surveillance are at an earlier stage with a more favorable survival than cancers detected at the time of diagnosis of Barrett, typically when patients present with dysphagia. The appropriate surveillance interval for patients with Barrett esophagus has not been studied prospectively. However, current programs, such as proposed by the
Patients with long-segment Barrett esophagus have an estimated 30 to 125 times increased risk of developing esophageal cancer compared with the general population. Early studies suggested that the median cancer incidence was 1 per 100 patient-years of follow-up, but more recent studies with longer follow-up suggest a lower cancer rate of 1 per 200 to 250 patient-years. This is an annual incidence of approximately 0.5%, with about 500 cases of adenocarcinoma diagnosed annually. However, since the early 1980s, the incidence of squamous cell carcinoma has stayed constant, whereas the incidence of adenocarcinoma of the esophagus and esophagogastric junction has risen fivefold—a growth rate exceeding that of any other cancer. Currently, adenocarcinoma accounts for more than half of all esophageal cancers in the
Management of patients with Barretts metaplasia has two aspects: treating the underlying GERD and managing the risk of adenocarcinoma of the esophagus. The principles for treating peptic esophagitis and controlling symptoms in Barretts metaplasia are the same as for uncomplicated GERD with the proviso that because it is associated with extreme esophageal acid exposure, it will likely require more intensive treatment. In general, mucosal damage and symptoms can be controlled with proton pump inhibitor therapy, but surgery may be needed, or even desirable, in refractory cases.
On the basis of in vitro and in vivo cell proliferation studies reporting increased cell proliferation and decreased differentiation in tissue exposed to acid, some authorities believe that complete acid suppression is desirable in therapy for Barretts metaplasia. However, there is no clinical evidence that aggressive antisecretory therapy or antireflux surgery prevents the occurrence of adenocarcinoma or causes regression of intestinal metaplasia. Proton pump inhibitor therapy has proved to be of no avail in reversing metaplasia or preventing adenocarcinoma. Thus, the available data support titrating therapy to control symptoms and treat esophagitis, irrespective of the presence of Barretts metaplasia.
Barrett esophagus is suspected at endoscopy and is confirmed by biopsy and histological examination. The columnar epithelium of the stomach is reddish pink, and the junction between the glossy white squamous mucosa and the columnar mucosa (Z-line) is normally found at the lower end of the tubular esophagus, just above the proximal folds of a hiatal hernia, if present. In Barrett esophagus, the distal esophagus is lined with columnar epithelium, extending upward for a variable distance, often 3 to 10 cm, but occasionally involving most of the esophagus.The proximal margin may be horizontal, or there may be irregular, often tongue-shaped upward extensions of columnar mucosa. Some patients have pale islands of regenerative or residual squamous epithelium, whereas others have punched-out benign ulcers in the columnar area. Strictures and esophagitis may be seen at the new squamocolumnar junction. The endoscopist should especially look for evidence of adenocarcinoma, such as nodularity or masses.
The characteristic histological finding in Barrett esophagus is a distinctive specialized intestinal epithelium. This is a glandular epithelium with mucin-type cells and the distinguishing presence of goblet cells. These are easily seen on hematoxylin and eosin–stained sections and can be demonstrated more prominently in sections stained with Alcian blue. It occupies most or all of the columnar-lined area and is the type of epithelium in which adenocarcinoma arises. Other types of epithelia seen with Barrett esophagus include gastric fundic and cardia-type epithelia, but these alone do not make the diagnosis of Barrett esophagus, nor are they associated with adenocarcinoma.
Currently, there is some controversy over the classification of Barrett esophagus. The classical or long-segment Barrett esophagus requires at least 3 cm of esophagus to be lined with columnar epithelium. This is the best-studied subset of Barrett esophagus, with traditional demographic features and a definite increased risk of becoming adenocarcinoma. Short-segment Barrett esophagus refers to shorter lengths or tongues of columnar epithelium, less than 3 cm, in the distal esophagus, with intestinal metaplasia on biopsy. This entity is three to five times more common than the long-segment variant, but, based on anecdotal reports, the risk of cancer appears to be lower. Intestinal metaplasia at the esophagogastric junction refers to microscopic findings on biopsy but no visible columnar epithelium in the esophagus at endoscopy. This finding has been reported in 10% to 32% of biopsies from unselected patients, many of whom have no reflux symptoms. The percentage of woman and African Americans is also higher with this lesion than with either long- or short-segment Barrett esophagus. The cause is controversial; some investigators suggest that this is the earliest form of GERD, whereas others believe these changes are secondary to H pylori infection. Cancer risk is minimal, if it exists at all.
In some patients with GERD, the squamous epithelium of the distal esophagus is replaced by specialized columnar epithelium, resembling that of the intestine and containing goblet cells. Although Dr. Norman Barrett thought this lesion was a congenitally shortened esophagus, studies consistently show that these patients have severe GERD with low LES pressures, poor esophageal motility, large hiatal hernias, and extensive acid and bile reflux. Furthermore, most patients have had chronic reflux symptoms for at least 10 years. Animal experiment show that, if the mucosal lining of the distal esophagus is excised in the setting of free acid reflux, columnar epithelium will regenerate in the area previously occupied by squamous epithelium. If reflux is controlled, the mucosal lining will regenerate with squamous epithelium. Pluripotential stem cells derived from the stratified squamous epithelium are the origin of the specialized columnar epithelium.
Barrett esophagus was once considered an uncommon condition, but estimates of its frequency at autopsy (1 in 57 to 1 in 105 cases), on general endoscopy survey (1 in 100 cases) and on endoscopic surveys of patients with GERD (10 in 100 to 15 in 100 cases), indicate that it is not uncommon, and it affects nearly 700,000 adults in the United States. An autopsy series from