Principles of Ambulatory Medicine, 7th Edition

Chapter 43

Peptic Ulcer Disease

Mack C. Mitchell Jr.

1. Franklin Herlong

Epidemiology and Natural History

Peptic ulcer disease remains a common problem, even though the incidence of duodenal ulcers has decreased over the last 40 years (1). The incidence of gastric ulceration has not changed over this time (1) nor has the incidence of complications, such as bleeding, perforation, obstruction, and penetration of peptic ulcers (2). The economic impact of peptic ulceration is over $5 billion a year in the United States (3).

The prevalence of duodenal and gastric ulcer is approximately the same in men and women (2) but the prevalence and incidence of duodenal and gastric ulcers are higher in older compared with younger age groups (4).

Ulcers are usually less than 1 cm in diameter, although giant ulcers (larger than 2.5 cm) occasionally occur. Duodenal ulcers are almost always located in the duodenal bulb or within 3 cm of the pyloric duodenal junction. Ulcers distal to the duodenal bulb should raise the suspicion of the Zollinger-Ellison syndrome (discussed later in this chapter) or Crohn disease of the duodenum. Gastric ulcers are most commonly located on the lesser curvature, at the junction of the body and antrum of the stomach. There is no risk of cancer in a duodenal ulcer, but 1% to 3% of gastric ulcers not attributable to nonsteroidal anti-inflammatory drugs (NSAIDs) are malignant.

Untreated duodenal ulcers can be chronic. Ulcers recur in 50% to 90% of patients within 1 year of diagnosis if curative therapy is not given (5). The rate of recurrence is highest within the first 5 years and then decreases over 10 to 20 years (6,7). The recurrence rate of gastric ulcers appears to be lower (5). Both duodenal and gastric ulcers tend to recur in the same place as the index ulcer. Long-term studies suggest that bleeding or perforation occurs at a rate of 1% to 3% a year, with a lifetime incidence of approximately 20% (8) if the ulcer is not adequately treated. Recurrent hemorrhage occurs in approximately 50% of patients who have had a previous bleed. The complication rate may be reduced by maintenance therapy (9) and can be eliminated if the bacterium Helicobacter pylori is eradicated (see below). The major risk of death from ulcer disease is on the initial presentation (10).

Pathophysiology

Duodenal Ulcer

Duodenal ulcer disease has traditionally been viewed as the result of an imbalance between the amount of acid delivered to the duodenum from the stomach and the normal duodenal defense mechanisms. A number of factors contribute to this imbalance including increased parietal cell mass, increased capacity of parietal cells to secrete acid, increased vagal drive to secrete acid, and defective inhibition of gastrin release and of gastric secretion. Alterations in normal duodenal defense mechanisms such as bicarbonate secretion, mucus production, vascular integrity, and endogenous prostaglandin production also promote ulcer formation. It is now clear that these imbalances can be caused by hypersecretory states, ingestion of NSAIDs, and most importantly by H. pylori infection.

P.660

Gastric Ulcer

The pathogenesis of gastric ulceration is unclear, although it is generally accepted that a disruption of the gastric mucosal barrier leads to gastric ulcer formation. This barrier can be affected by irritants such as bile, alcohol, and aspirin. These observations help explain the epidemiologic data associating alcohol with acute gastritis and aspirin with gastric ulcers and erosions. Furthermore, in patients with gastric ulcers, radiologic and manometric studies suggest increased duodenal gastric reflux. This reflux of bile across an incompetent pyloric sphincter results in the disruption of the mucosal barrier. Once the barrier is broken, hydrogen ion may diffuse back into the gastric cells, leading to ulceration via local histamine release, vasodilation, and tissue damage. According to this hypothesis, gastric ulcer formation requires injury to the gastric mucosal barrier and the presence of some, but not necessarily an excessive amount of, acid. H. pyloriinfection, bile acids, and NSAIDs may make the mucosa more susceptible to injury and result in a gastric ulcer.

Risk Factors

Helicobacter pylori

1. pyloriinfection is the most common and important risk factor for duodenal ulcer. The risk of gastric ulcer is probably multifactorial, although many gastric ulcers are also associated with H. pylori.
2. pyloriis a gram-negative spiral organism found exclusively in gastric epithelium, although it may be seen in the duodenal bulb in regions where there is gastric metaplasia. In the United States, the organism is found in 10% of healthy people under 30 years of age and in approximately 60% of healthy people over age 60 (11). The rate of infection varies by country and by region of the United States and is more common in lower socioeconomic areas and in developing countries (11). The organism is probably spread by person to person contact.
3. pyloriis the etiologic agent for chronic nonerosive antral gastritis (see below) and is present in the stomach in 70% of patients with a peptic ulcer, making it the most common cause of that disease (12). It is estimated that 15% to 20% of infected patients will develop an ulcer in their lifetime. Other diseases associated with H. pyloriinfection are B-cell mucosa-associated lymphoid tissue (MALT) lymphoma and gastric adenocarcinoma. As such, H. pylori has been recently classified as a class I biologic carcinogen.

Eradication of the organism accelerates duodenal ulcer healing (13) and significantly decreases ulcer recurrence 1 and 2 years after treatment (13,14). The rate of reinfection is unknown but appears to be less than 1% up to 5 years after treatment. Eradication ofH. pylorialso reduces recurrent bleeding in patients who have had ulcer-related gastrointestinal (GI) bleeding.

Although H. pylori infection is a major risk factor for peptic ulcer disease, 20% to 30% of gastroduodenal ulcers occur in the absence of H. pylori infection (15,16) with most of these caused by NSAID use. False-negative H. pylori screening tests may, in part, account for these findings, but the availability of more than one reliable test for H. pylori should improve detection. Prior use of antibiotics, particularly macrolides, or proton pump inhibitors (PPIs) within 6 weeks of H. pylori testing may also result in false-negative tests.

Nonsteroidal Anti-Inflammatory Drugs

There is good evidence that all NSAIDs can cause gastric and duodenal ulcers. The greatest risk of NSAID-induced gastric or duodenal injury occurs in people older than 60 years of age (17), in patients with a history of a GI “event” (peptic ulcer or hemorrhage), in patients who are taking corticosteroids or anticoagulant drugs concurrently, and in patients taking a dose of an NSAID that is two or more times the usual prescribed dose (18). Prospective and retrospective studies have concluded that chronic use of corticosteroids by patients who are not taking NSAIDs does not increase the risk of peptic ulcer disease or its complications (19). The cyclooxygenase (COX)-2 specific NSAIDs are associated with a significantly lower risk of gastric ulceration and major complications, including GI bleeding, compared with traditional nonselective NSAIDs (20,21). Until recently, these NSAIDs were often considered for patients who require anti-inflammatory drugs, especially those at risk for GI complications from this class of agents. However, some of the medications in this class have recently been removed from the market (as of this writing, celecoxib is the only product in this class still available) because use of these drugs has been associated with an increased risk of stroke and myocardial infarction. COX-2 inhibitors, if available, should therefore be reserved for patients who require an NSAID for a relatively short time and who have a low cardiovascular risk profile.

Other Factors

Genetic factors were once thought to play an important role in the development of peptic ulcers, but since the identification of H. pylori and NSAIDs as major risk factors, it has been recognized that the role of inheritance is minimal (22). The role of emotional stress in the pathogenesis of peptic ulcer disease seems clinically evident but has been difficult to quantitate. In fact, in a case control study, the number of stressors in the lives of patients with duodenal ulcer was no greater than in control subjects (23).

P.661

Cigarette smoking is definitely associated with an increased frequency of duodenal ulcer disease, with frequencies ranging between 33% and 100% above those found among nonsmokers (24,25). Cigarette smoking may increase the frequency of recurrence if H. pylori is not eradicated (26). There is no epidemiologic evidence that alcohol is ulcerogenic, although it is a known cause of acute gastritis. Coffee, although not ulcerogenic, may exacerbate symptoms of peptic disease.

Certain disease states have been associated with an increased risk of peptic ulcer disease. Evidence of such associations must be carefully evaluated in light of the relatively high prevalence of ulcer disease in the general population and the frequent use of ulcerogenic drugs. However, there is good evidence linking peptic ulcer disease with chronic obstructive pulmonary disease, alcoholic cirrhosis, and chronic renal failure (27, 28, 29). Conditions that lead to increased gastric acid secretion also predispose to ulcer disease. The Zollinger-Ellison syndrome, or gastrinoma (discussed later in this chapter), is the best example of such a condition. Extensive small bowel resection may also lead to hyperplasia of antral gastrin-containing cells and result in ulcer disease. Retained antrum after gastric surgery (rarely seen today) is another example of a situation where uninhibited acid production is associated with recurrent ulcerations.

Many drugs other than NSAIDs are reputed to be ulcerogenic, although the evidence is not well established for most of them. Aspirin is one drug that causes gastric and duodenal ulcers. Experimentally, aspirin disrupts the gastric mucosa both physiologically and anatomically, leading to ulcer formation. The risk of gastric ulcer increases sixfold when patients use more than three aspirins a day (30). Aspirin use also increases the risk of bleeding from peptic lesions (see Chapter 56). Enteric-coated or buffered aspirin has no advantage over regular aspirin with respect to these untoward effects.

Diagnosis

History

The most common symptom of peptic ulcer disease is epigastric distress—vague discomfort or a feeling of gnawing hunger—usually in the midline. If actual pain occurs, it is typically aching or burning.

Classically, the distress of duodenal ulcer occurs 1 to 3 hours after a meal and may awaken the patient from sleep, typically between 1 and 2 a.m.; it is relieved within minutes by food, antacids, or vomiting. Pain is minimal before breakfast. In patients with gastric ulcers, the history is more variable. In some, a similar distress–food–relief pattern exists, whereas in others there is no relationship with food. Occasionally, the distress is actually exacerbated by food.

Other less common symptoms of ulcer disease include nausea, vomiting, and dyspepsia. However, in some patients with ulcers, one or more of these symptoms may occur in the absence of typical ulcer pain. Weight loss occurs in up to 50% of patients with a benign gastric ulcer and is therefore not a helpful feature in distinguishing a benign from a malignant ulcer. Patients with duodenal ulcer often gain weight because they eat more in an attempt to control their pain.

The history may also suggest certain complications. Pyloric obstruction presents first with early satiety and then with persistent vomiting, often of undigested food. A change in the quality of the pain or radiation of the pain to the back or shoulder suggests penetration of the ulcer. A history of melena suggests bleeding. Occasionally, one of these complications is the first clinical manifestation of peptic ulceration.

Physical Examination

The physical examination may provide supportive, although nonspecific, information. Localized epigastric tenderness is common. The presence of a succussion splash 4 hours or more after a meal suggests gastric outlet obstruction.

Radiologic Studies

In the past barium studies were used to confirm the presence of peptic ulcer disease. However, endoscopy is now used almost exclusively because is more accurate and it allows biopsy of the mucosa for H. pylori. Currently, there are few, if any, indications for barium radiographs in the diagnosis of peptic ulcer disease.

Endoscopy

Endoscopy is the diagnostic procedure of choice in the diagnosis of peptic ulcers. Experienced endoscopists are able to diagnose up to 95% of gastroduodenal ulcers (the ulcer ultimately proved by a second endoscopy or by surgery) (31). No firm guidelines have been established for the use of endoscopy in a patient with suspected peptic ulcer disease. Some patients with typical symptoms of duodenal ulcer that respond to a therapeutic trial of medical therapy do not need endoscopy. However, patients with persistent abdominal pain refractory to medical treatment or with recurrent symptoms after “curative” therapy should undergo endoscopy. Other indications for endoscopy include suspected outlet obstruction, active or suspected GI bleeding, the need to evaluate equivocal or indeterminate findings on upper GI series, if performed, and the

P.662

need to confirm the presence or absence of H. pylori. No followup endoscopic evaluation of an uncomplicated duodenal ulcer is required because the risk for malignancy is nil.

All gastric ulcers should be biopsied to exclude malignancy. If a gastric ulcer has been diagnosed by radiograph, endoscopy should be performed so that it can be biopsied (32). The ulcers should then be followed until they have healed, as confirmed by endoscopy. For a review of patient experience during endoscopy, see Chapter 45.

Gastric Analysis

Gastric secretory studies have enhanced the understanding of the pathophysiology of ulcer disease but have very little clinical usefulness. Although many patients with duodenal ulcer disease secrete more acid than normal, more than half of duodenal ulcer patients actually have normal levels of acid production. Furthermore, acid hypersecretion does not predict the development of ulcer disease in an individual patient.

In patients with gastric ulcers, gastric analysis is not helpful since no reliable criteria exist to differentiate benign from malignant gastric ulcers. The main indications for gastric analysis are for preoperative evaluation of patients with suspected Zollinger-Ellison syndrome (see below) or in postgastrectomy patients with recurrent ulcer disease. However, gastric analysis is not useful as an indicator of the type of surgery needed. For example, more acid production does not necessarily mean that a more extensive gastric resection is required (33).

Serum Gastrin Measurements

Measurements of the level of gastrin in the blood are readily available and are useful in the evaluation of patients with suspected hypersecretory conditions. The fasting basal serum gastrin level is usually normal (less than 150 pg/mL) in patients with duodenal and gastric ulcer disease but is elevated, often to very high levels (usually 1,000 pg/mL or more), in patients with the Zollinger-Ellison syndrome (see below). In patients with recurrent ulcers and in patients refractory to conventional therapy—features suggestive of the syndrome—two or three fasting serum gastrin determinations are required to confirm this condition since gastrin levels may fluctuate. Concurrent use of a PPI can increase basal gastrin levels, and thus gastrin levels should ideally be measured several weeks or months after cessation of acid-suppression therapy. If acid suppression cannot safely be discontinued prior to gastrin measurement, then only markedly elevated values (i.e., greater than 1,000 pg/mL) should be considered as suggestive of a gastrinoma.

Diagnosis of Helicobacter pylori

Several diagnostic tests are available for the detection of H. pylori infection (34, 35, 36) (Table 43.1). The standard criterion is histologic testing (37), using at least two biopsies obtained endoscopically from different parts of the antrum and/or proximal body to ensure that the infection is not missed. This method has a sensitivity and specificity of 85% to 100%. It is time consuming, requires analysis by an experienced pathologist, and is expensive. H. pylori is identified in the mucosa with the use of hematoxylin and eosin, Warthin-Starry, or Giemsa stains. The sensitivity of cultures (essentially done for research only) is 50% to 95%, but the specificity approaches 100%. For patients undergoing endoscopy, one of the most cost-effective tests is the rapid urease (CLO) test. In this commercially available test, the biopsy is placed in a gel containing urea, phenol red, and a bacteriostatic agent. H. pylori causes a color change from yellow to red if the test is positive. This test is approximately 95% sensitive and 95% specific.

Several noninvasive tests are available for use in ambulatory patients (38). Serologic tests for the detection of immunoglobulin G (IgG) antibody, including a fingerstick method, have sensitivities and specificities of approximately 90%. Although a decline in IgG concentration correlates well with eradication of H. pylori 12 to 24 months after treatment (35), IgG testing is used less often to confirm eradication than to determine if a person has been exposed to H. pylori. Such a determination may be helpful if a biopsy shows changes suspicious for H. pylori but no organisms are identified. In this case, a positive IgG would be a compelling reason for treatment.

The urea breath test is another noninvasive test. It entails giving the patient ¹³C-labeled urea by mouth and then measuring the patient's breath for the labeled carbon dioxide generated when H. pylori splits the urea. This test is 96% sensitive and 98% specific. The ¹³C test is read by a mass spectrometer, so the expired air must be sent out of the office for reading. This is a reliable noninvasive test for evaluation of eradication of the organism (34), but may be too cumbersome and expensive for routine clinical use.

1. pyloristool antigen testing is the preferred method of diagnosing persistent H. pyloriinfection. This test utilizes a polyclonal anti-H. pylori antibody to detect bacterial antigens found in stool. Multiple studies have shown that stool antigen testing has similar posttreatment sensitivity and specificity as urea breath testing, and is less expensive and easier to obtain.

A potential confounding factor in the noninvasive diagnosis of H. pylori is the fact that concurrent acid suppression can cause false-negative results. Given that H. pylori requires an acidic environment to replicate, acid suppression may decrease the number of organisms and the amount of antigen or urease available for the respective

P.663

tests. Consequently, tests to confirm eradication of H. pylori should only be done after a patient has discontinued PPIs for at least several weeks.

TABLE 43.1 Diagnostic Tests for H. pylori

Method Specimen Cautions Sensitivity (%) Specificity (%) Comments For Patients Who Do Not Require Endoscopy Serology test (in office) Serum or whole blood Lipemia 90–95 90 Performed in 5–20 min; detects IgG (elevated in patients with H. pylori). Multiwell ELISA (in laboratory) Serum Recent antibiotics make interpretation of a single sample difficult; 1%–5% of patients do not produce detectable antibody 90–95 90 Decrease in paired antibody titers can be assessed. Patients with rapid antibody decrease are cured. Persistent antibody after therapy is common and may slowly decrease (12–18 mo), which may represent cure or continued infection. Urea breath test Breath sample Antibiotics and PPIs can cause false-negative results; should be done 4 weeks or more after last antibacterial therapy; not affected by H2 receptor antagonists 95 95 Urease of H. pylorigenerates labeled CO2 from breakdown of ingested urea. Breath sample taken 20–40 min after fasting. Patient ingests isotope in liquid (13C) or capsule (14C. For Patients Undergoing Endoscopy and Biopsy Rapid urease test 1–2 mucosal biopsies Same as for breath test: do not take samples from ulcers or lesions 90 98 Simple biopsy test: Urease of H. pylorigenerates ammonia and causes pH change. One-step test, no reagents, result in 1–24 hr. Least expensive method of endoscopic diagnosis. Histology 2 mucosal biopsies Still regarded by many as the gold standard for H. pylori detection 95 98 Simple and very accurate: Specimen can be reexamined in equivocal cases. Provides a permanent record. Special stains occasionally necessary to detect organism. Stool H. pyloriantigen Stool specimen ELISA assay 90–95 90–95 Most useful test to document eradication Culture of biopsy 1–2 mucosal biopsies 3–6 days on moist, brain-heart blood agar in Campy pac or 10% CO2, 37°C 80 100 Provides antibiotic sensitivity information (if necessary). ELISA, enzyme-linked immunosorbent assay; PPI, proton pump inhibitors.

Medical Therapy

The treatment of peptic ulcer disease is directed at healing the ulcer and preventing its recurrence. Both can be accomplished with successful eradication of H. pylori, when present. Treatment also includes relief of pain with pharmacologic agents. Rarely, a patient is refractory to medical treatment or has complications that require surgery.

Nonpharmacologic Therapy

There is no evidence that dietary modification affects either the symptoms or the course of peptic ulcer disease. Frequent feedings, bland diets, increased milk consumption, and decreased consumption of spices and fruit juices do not affect healing. Although alcohol and caffeine increase acid secretion, no evidence exists that discontinuing either of these substances enhances healing, so it is reasonable to allow patients to continue to consume them in moderation. Dietary restrictions should be limited to substances that cause symptoms; otherwise, patients should be allowed to eat as they wish.

P.664

The most important nonpharmacologic intervention is to discontinue cigarette smoking (see Chapter 27) (24,26). Cigarette smoking increases the risk and delays the healing of duodenal ulcers. Controlled trials with H₂ antagonists have demonstrated impaired healing of duodenal ulcers in smokers compared to nonsmokers (26). However, there is no difference in peptic ulcer recurrence rates in smokers and nonsmokers following eradication of H. pylori (39). Smoking also predisposes patients to perforation and bleeding. Cigarette smoking also affects gastric ulceration, although it has not been studied as extensively as it has been in patients with duodenal ulceration. Even though duodenal ulcers do not recur if H. pylori is eradicated, it seems prudent to recommend smoking cessation to all patients with peptic ulcer disease.

TABLE 43.2 Therapeutic Options for Eradication of H. pylori

Regimen Drugs Dosage Duration Dual therapy Omeprazole 40 mg q.d. 2 wk Clarithromycin 500 mg t.i.d. or RBC 400 mg b.i.d. 4 wk Clarithromycin 500 mg t.i.d. 2 wk Bismuth-based triple therapy Bismuth subsalicylate 2 tablets q.i.d. 2 wk Metronidazole 250 mg q.i.d. Tetracycline or 500 mg q.i.d. amoxicillin 500 mg q.i.d. PPI-based triple therapy Metronidazole or 500 mg b.i.d. 10–14 days amoxicillin 1,000 mg b.i.d. PPTa Amoxicillin 1,000 mg b.i.d. 10–14 days PPIa plus two antibiotics Clarithromycin 500 mg b.i.d. Quadruple therapy Bismuth-based triple therapy plus PPTa Triple therapy plus PPTa 14 days aSee Table 43.3. RBC, ranitidine and bismuth citrate; PPI, proton pump inhibitor.

Aspirin and other NSAIDs are well-known risk factors for gastric erosions and ulcers. There is also evidence of an increased risk of duodenal ulcers and erosions as well. Consequently, all such agents should be discontinued, if possible, in patients with active peptic ulcer disease.

Pharmacologic Therapy

Duodenal Ulcer

Patients with a proven diagnosis of duodenal (or gastric) ulcer associated with H. pylori infection should receive treatment aimed at eradication of H. pylori and pharmacologic therapy to decrease duodenal ulcer pain. Appropriate treatment results in healing of 85% to 95% of duodenal ulcers and reduction of duodenal ulcer recurrence to less than 1% per year. There are several important treatment principles: therapy with a single antibiotic is not effective, no single regimen has proven effective in all patients, and substitution of an antibiotic of the same class for a recommended antibiotic (e.g., ampicillin for amoxicillin or doxycycline for tetracycline) results in a dramatic decrease in healing rates. Each regimen discussed in this section should be evaluated based on efficacy, antimicrobial sensitivity, side effects, compliance, cost, and previous treatment.

Several regimens have received Food and Drug Administration (FDA) approval (40) (Table 43.2). Dual therapy and regimens that do not include a PPI, though FDA-approved, are no longer considered to be as effective as PPI-based triple therapy. PPI-based triple therapy is the preferred treatment. Eradication rates of up to 95% have been reported after 7 to 14 days of treatment with a PPI twice a day, amoxicillin (1,000 mg twice a day), and clarithromycin (500 mg twice a day). Optimal responses are achieved with at least 10 days of treatment. Metronidazole (500 mg twice daily) may be substituted for amoxicillin, usually with similar healing rates except in geographic areas where metronidazole resistance is high. Metronidazole resistance rates are approximately 15% and clarithromycin resistance rates are approximately 7%. Amoxicillin and tetracycline resistance is uncommon. All currently available PPIs can be used to treat H. pylori. An alternative regimen is triple therapy with bismuth subsalicylate (two tablets four times a day), metronidazole (250 mg four times a day), and tetracycline (500 mg four times a day) taken for 14 days, prepared as a blister pack for each day's dosage. More than 60% of the doses must be completed to ensure optimal eradication rates. Side effects have been variable with this

P.665

regimen, ranging from 10% to 30%, and may be reduced by the addition of a PPI. A single daily dose of a PPI will increase compliance, which is the key to a successful regimen. Patients who are intolerant of tetracycline can be given amoxicillin (500 mg four times a day) in its place.

If H. pylori is successfully eradicated, there is no need for maintenance therapy. If symptoms recur after adequate treatment, the patient should be tested for the presence of the organism (urea breath test, stool antigen testing, or repeat endoscopy, depending on availability and cost). If H. pylori is present, it is probably not due to reinfection but to ineffective treatment, and another course of treatment should be prescribed. If H. pylori is not found, evaluation for another cause of symptoms should be pursued.

Though the combination of PPIs with antibiotics is the preferred treatment of H. pylori-associated ulcers, antisecretory agents such as antacids and H₂ receptor antagonists are effective in relief of symptoms and may still be used in selected patients with intolerance to PPIs or with H. pylori-negative ulcers. lists the currently available drugs used in the treatment of peptic ulcer disease.

Antacids are rapidly acting and are effective for transient relief of symptoms. A large selection of calcium carbonate, aluminum, or magnesium-based antacids is available. They should not be used as the sole treatment to heal peptic ulcers, but they may be used as adjunctive agents to control symptoms. Histamine receptor antagonists were the mainstay of therapy before the development of combination PPI-antibiotic regimens to treat peptic ulcer disease. Four H₂ receptor antagonists are available: cimetidine, ranitidine, famotidine, and nizatidine (Table 43.3). All are available as generic compounds, and all are available in over-the-counter (OTC) dosage formulations. They are used to treat ulcers in full dose given twice daily or as a nocturnal dose. When used in FDA-approved dosing schedules, H₂ receptor antagonists reduce acid production by 30% to 50%, keep intragastric pH above 3 (the critical value for ulcer healing) for 8 to 12 hours, and effectively heal more than 80% of peptic ulcers after 6 to 8 weeks of therapy (41). Side effects are infrequent (rare abdominal discomfort, nausea, and vomiting). Of the available H₂ receptor antagonists, cimetidine has the greatest potential for interaction with drugs that effect the cytochrome

P.666

P450 system such as diazepam, phenytoin, warfarin, and theophylline.

TABLE 43.3 Drugs Approved for the Treatment of Peptic Ulcer Disease

Drug Available Strength OTC Generic Initial Dose Principal Side Effectsa H2 Receptor Blockers Cimetidine (Tagamet) 100, 200, 300, 400, 800 mg + + 400 mg b.i.d. or 800 mg at bedtime Gynecomastia, confusion, impotence, blood dyscrasia, drug interaction Ranitidine (Zantac) 75, 150, 300 mg + + 150 b.i.d. or 300 mg at bedtime Gynecomastia, impotence, hepatitis (rare) Famotidine (Pepcid) 10, 20, 40 mg + – 20 mg b.i.d. or 40 mg at bedtime Headache, decreased libido, depression, mild increase in liver enzymes Nizatidine (Axid) 75, 150, 300 mg + – 150 mg b.i.d. or 300 mg at bedtime Sweating, urticaria (<1%), somnolence, elevated liver enzymes Proton Pump Inhibitorsb Omeprazole (Prilosec) 10, 20, 40 mg – – 20 mg/day (DU) 40 mg/day (GU) Headache, dizziness, rash, diarrhea, abdominal pain Lansoprazole (Prevacid)c 15, 30 mg – – 30 mg/day Diarrhea, abdominal pain Rabeprazole (Aciphex)c 20 mg – – 20 mg/day Headache Pantoprazole (Protonix)c 40 mg – – 40 mg/day Chest pain, headache, diarrhea Esomeprazole (Nexium)c 20, 40 mg – – 20–40 mg/day Headache, diarrhea, nausea, abdominal pain Sucralfate (Carafate) 1 g – – 1 g q.i.d. or 2 g q.i.d. Constipation a1%–10% of patients. bFor eradication of H. pylori infection, all PPIs are given b.i.d. (i.e., twice the initial dose listed in the table) for 10–14 days except esomeprazole, 40 mg, which is given once a day. cNot approved by FDA for treatment of gastric ulcer. OTC, over-the-counter; DU, duodenal ulcer; GU, gastric ulcer.

Sucralfate forms a protective coating on the mucosa of the stomach and duodenum and is as effective as H₂ blockers in the healing of peptic ulcers. It has been replaced by PPIs as initial and maintenance therapy for peptic ulcer disease.

Most H. pylori-negative duodenal ulcers are caused by NSAIDs and resolve with discontinuation of the offending agent. A PPI is often added to promote healing. Maintenance therapy is unnecessary unless the patient has had severe GI bleeding. When used, maintenance therapy consists of one-half the daily dose of whatever drug was given for acute healing.

Gastric Ulcer

In general, the same drugs used to treat duodenal ulcers are effective in the treatment of gastric ulcer (Table 43.3). If H. pylori is present, it should be eradicated (Table 43.2), regardless of whether NSAIDs are thought to be implicated.

Healing must be documented by endoscopy 4 to 8 weeks after the diagnosis of a gastric ulcer to exclude possible underlying malignancy. No dietary restrictions affect healing, and hospitalization is unnecessary in the absence of complications. NSAIDs, including aspirin, should be discontinued if possible. Smoking cessation should be strongly recommended. Maintenance therapy is necessary if the ulcer was due to H. pylori, until the infection is eradicated.

Chapter 77 discusses the prevention and treatment of NSAID-associated gastropathy.

Surgical Therapy

Although surgery is effective therapy for the relief of ulcer symptoms and for the prevention of ulcer recurrence, with current medical therapy it is rarely indicated. At present the only indications for surgical treatment of peptic ulcer disease are perforation, uncontrolled hemorrhage, gastric outlet obstruction, and, rarely, intractability. Because most operations are performed on an emergency basis, the type of surgery is generally decided upon in the operating room after the surgeon has evaluated the gastroduodenal area.

Duodenal Ulcer

Three operations are currently used for duodenal ulcer: vagotomy with pyloroplasty, vagotomy with antrectomy, and parietal cell (or highly selective) vagotomy. The first two procedures involve a selective vagotomy (gastric vagal fibers only) plus a drainage procedure to facilitate gastric emptying postoperatively. The third procedure involves cutting vagal fibers to the body (acid-secreting cells) of the stomach, preserving antral innervation, and eliminating the need for a drainage procedure. Vagotomy with pyloroplasty has the lowest mortality, shortest operation time, and a recurrence rate of 6% to 8%. Vagotomy with antrectomy is technically more difficult and may predispose patients to more postoperative morbidity (because an anastomosis to the remaining stomach is required) but has the lowest recurrence rate (less than 2%) (42). Postoperative complications (Table 43.4) are significantly more common with either procedure than with parietal cell vagotomy. A parietal cell vagotomy has a recurrence rate of 10% at 10 years (42), approximately four to five times that of vagotomy with antrectomy. Long-term complications (of dumping syndrome and diarrhea) are less than 60% those of vagotomy with antrectomy. In the hands of experienced surgeons, a parietal cell vagotomy is the operation of choice. The higher recurrence rate (which can be managed medically in 80%) is an acceptable trade-off for the decreased complication rate.

Gastric Ulcer

The surgical approach for gastric ulcers is not as well established since the causes of gastric ulceration are more complicated. In contrast to duodenal ulcer disease, a vagotomy may not be indicated in all patients with a gastric ulcer who undergo surgery. However, in patients who have evidence of concomitant duodenal ulcer disease (approximately 10% to 40% of patients with gastric ulcers) and in patients who have pyloric ulcers, which generally behave as duodenal ulcers, a vagotomy is clearly indicated. If the ulcer is within the antrum, an antrectomy or hemigastrectomy that includes the ulcer is often the preferred operation. When the ulcer cannot be included in the gastric resection, a full-thickness biopsy of the ulcer should be taken for frozen section to rule out malignancy. The recurrence rate for gastric ulcers after these types of operation is very low (1% to 2%).

After antrectomy or hemigastrectomy, the stomach may be anastomosed to the duodenum (Billroth I anastomosis) or to the jejunum (Billroth II). The type of anastomosis is determined by the surgeon, based on the degree of duodenal deformity and on technical considerations.

Postgastrectomy Syndromes

Many problems develop after gastrectomy (Table 43.4). In 10% of patients, postgastrectomy complications are severe. Many of these conditions result from the altered physiology created by the surgery.

Zollinger-Ellison Syndrome

The Zollinger-Ellison syndrome dramatically demonstrates the relationship between gastrin, acid secretion,

P.667

and ulcer formation. The syndrome results from a non–β islet cell tumor of the pancreas that autonomously secretes gastrin and is therefore called a gastrinoma. In most cases multiple tumors are present, most commonly found in the head of the pancreas. These tumors vary considerably in size from several millimeters, often undetectable at surgery, to huge masses that may even be palpable through the abdominal wall. Approximately two thirds of gastrinomas are malignant; although they are usually slow growing, they can metastasize and can be a cause of death.

TABLE 43.4 Postgastrectomy Syndromes

Syndrome Clinical Features Pathophysiology Diagnosis and Treatment Early Stomal dysfunction Vomiting, gastric retention Edema, inflammation, hypokalemia Electrolyte repletion, time, no suction Duodenal stump dehiscence Pain, fever, signs of abscess, sepsis, death Billroth II anastomosis: tension and poor closure, adjacent pancreatitis, excessive inflammation in area of surgery Reoperation Afferent loop syndrome Pain, vomiting bile without food, may occur acutely or chronically Billroth II anastomosis: afferent loop too long, kinked, twisted, herniated etc.; loop fills, then empties Reoperation Vagotomy complications Transient dysphagia Dysphagia Lower esophageal sphincter dysfunction Usually transient, disappears in 1–2 wk Diarrhea Diarrhea transient or slight, 20%–40% of patients; troublesome, 5%; occurs mainly with selective vagotomy and drainage procedure Most common after truncal vagotomy; appears to be related to increased output of dihydroxy bile salts, the cause of which is uncertain Cholestyramine, Amphojel Late or Persistent Dumping syndrome Early phase: with or shortly after meals–nausea, abdominal fullness or pain, cramping, palpations, dizziness, sweating Distension of gastric pouch and upper jejunum from rapid emptying; peripheral intravascular volume depletion from rapid entry of fluid into jejunum due to osmotic changes in jejunum; vasomotor symptoms related to release of vasoactive substances into circulation, such as serotonin and bradykinin Small frequent meals, high protein, low carbohydrate, small volume of liquids only Late phase: symptoms of hypoglycemia Early hyperglycemia → insulin production → late hypoglycemia Gastric cancer Increased incidence of 3%–5% in gastric stump 15–20 yr after surgery Possibly related to chronic gastritis developing after gastrectomy Endoscopy for diagnosis, surgical resection Diarrhea Chronic diarrhea Rapid gastric emptying, lactose intolerance unmasked by vagotomy, malabsorption, Z-E syndrome, bile acid output increased, bacterial overgrowth Lactose-free diet; if no response, malabsorption workup (Chapter 45) Stomal or recurrent ulcer Recurrent ulcer symptoms; hemorrhage in approximately 50% Hyperacidity caused by inadequate resection, incomplete vagotomy, retained antrum, unrecognized Z-E syndrome (gastrinoma) Endoscopy; H2antagonists, (successful in 80%), reoperation Anemia Iron deficiency Chronic blood loss; impaired iron absorption Repletion of deficient nutrient Nutritional anemia Defective vitamin B12 absorption because of decreased intrinsic factor production (resection and gastritis); possible blind loop bacterial overgrowth; folate deficiency Osteomalacia Bone pain Diminished calcium intake, poor vitamin D absorption; duodenal bypass Z-E, Zollinger-Ellison.

P.668

The appreciation of the clinical features of Zollinger-Ellison syndrome changed when measurement of serum gastrin levels became available. The original description of the syndrome focused on the virulent nature of the ulcer diathesis and on the atypical location for the ulcers. It is now recognized, however, that 75% of ulcers in patients with Zollinger-Ellison syndrome occur in the duodenal bulb and appear as routine single lesions. However, the finding of postbulbar and jejunal ulcerations should still alert the clinician to the possibility of the syndrome. More than one fourth of patients undergo ulcer surgery before the diagnosis of the syndrome, which is usually made only when anastomotic ulcers develop.

Diarrhea occurs in approximately one third of patients, and may precede the formation of ulcers by several years. Some patients have diarrhea and never develop an ulcer. Diarrhea is caused principally by the increased secretion of gastric acid that lowers the pH of the normally alkaline duodenal fluid and interferes with absorption of water and electrolytes.

Diagnosis

The diagnosis of Zollinger-Ellison syndrome should be considered when duodenal ulcers do not heal with medical therapy, or when there are giant ulcers multiple ulcers, postbulbar or jejunal ulcers, or anastomotic ulcers. Ulcer disease associated with diarrhea or with radiographic evidence of gastric hypersecretion also suggests Zollinger-Ellison syndrome.

The diagnosis of Zollinger-Ellison syndrome is usually based on the fasting serum gastrin concentration, normally less than 150 mg/mL. Elevations greater than 1,000 mg/mL in association with the typical clinical picture are virtually diagnostic of a gastrinoma. However, in patients with mild elevations of the serum gastrin concentration (between 150 and 300 mg/mL) and in postoperative patients, differentiation between Zollinger-Ellison syndrome and other causes of hypergastrinemia is important. Other conditions that may lead to hypergastrinemia include retained antrum, G-cell hyperplasia, postvagotomy with pyloroplasty, and pernicious anemia. Also, H₂ receptor blockers and PPIs cause a slight increase in the serum gastrin concentration, and therefore should be stopped for at least several weeks before gastrin levels are measured.

Differentiation of the Zollinger-Ellison syndrome from other causes of hypergastrinemia often requires the use of provocative tests, i.e., the secretin and calcium infusion tests, generally performed by a gastroenterologist. The secretin test is preferred because it is more reliable and is safer. In both tests, the response of the serum gastrin level to the infusion of a stimulating substance is monitored. In the secretin test, the serum gastrin level rises, usually within the first half hour after the injection of secretin, in patients with Zollinger-Ellison syndrome, whereas in all other disorders the gastrin level falls or is unchanged.

Gastric analysis may provide further supportive data. Marked hypersecretion is found in both the basal state and after pentagastrin stimulation. Because the stomach is being influenced by an autonomous tumor, further stimulation with exogenous pentagastrin provides little additional stimulation to secretion. Thus, the basal acid output to maximal acid output ratio is 0.6 or greater in this syndrome. However, there is considerable overlap with normal values, so gastric secretory data alone cannot be used to make the diagnosis.

Therapy

Because the tumor mass is rarely localized and therefore rarely curable by local resection, therapy is directed at the end organ. Although total gastrectomy historically was the procedure of choice, in recent years it has become clear that medical therapy is successful in controlling both the ulcer disease and the diarrhea, making surgery often unnecessary. The PPIs (Table 43.3) are also extremely successful in controlling acid secretion in these patients, although high doses are often required. Long-term treatment with these agents is safe. Although gastrectomy prevents the consequences of hypersecretion of acid it does not alter the biologic behavior of the gastrinoma. In the past, patients died from this condition, most often because of the virulent nature of the ulcer disease (with frequent recurrences), because of severe diarrhea (with malabsorption), and as a result of multiple operations and their complications.

Dyspepsia

The term dyspepsia is often used to describe epigastric discomfort, occasionally related to meals and sometimes associated with nausea, belching, or bloating. It is estimated to be present in 7% of the U.S. population, most of whom rarely seek medical attention. When patients with dyspepsia are evaluated with endoscopy, only 20% to 25% have peptic ulcer disease or gastric cancer. Non-ulcer dyspepsia presents a diagnostic and therapeutic dilemma.

Four major disorders are associated with nonulcer dyspepsia: irritable bowel syndrome, cholelithiasis, gastroesophageal reflux, and chronic pancreatic disease. In irritable bowel syndrome (see Chapter 44) the dyspepsia is associated with diffuse abdominal pain and altered bowel habits. Patients with gastroesophageal reflux disease have associated heartburn (see Chapter 42). Chronic pancreatic disease is less common but is usually associated with more severe pain and steatorrhea. The most difficult diagnostic dilemma, because of the high prevalence of gallstones in the general population, is distinguishing patients with

P.669

symptomatic gallstones from patients with dyspepsia and asymptomatic (incidental) gallstones (see Chapter 96). It is now well established that patients with dyspepsia do not respond to cholecystectomy unless they have had an identifiable attack of acute cholecystitis or a history of biliary colic. The absence of either should suggest that gallstones are not the cause of dyspepsia.

Patients without one of these identifiable conditions are said to have essential dyspepsia. The etiology of this condition is unknown. A small subset of patients has been described with delayed gastric emptying of solid foods; however, correlation of improvement of symptoms with treatment has been poor. Approximately 50% of patients have H. pylori gastritis, but symptomatic response to treatment for this infection has been inconsistent. Patients with essential dyspepsia do not have increased basal acid secretion nor has a definite association with emotional stress been documented.

The patient with essential dyspepsia can be difficult to manage. The yield of radiographic and endoscopic procedures is low and may be confusing. Response to empiric therapy with H₂ blockers or PPIs has been disappointing and may be misleading.

Patients younger than 40 years with a short history and no evidence of organic disease by physical examination and appropriate laboratory tests should be treated with reassurance, some modification of their diet, and avoidance of caffeine, alcohol, and tobacco. A 4-week trial of H₂ blockers or PPIs may be used (Table 43.3). If no response occurs after 4 weeks and no evidence of reflux disease or irritable bowel syndrome is present, endoscopy should be performed. If endoscopy is negative, the other diagnoses should be pursued. Patients older than age 50 are candidates for early investigation (within 1 to 2 weeks), particularly if symptoms are severe and have occurred for the first time. In this group the diagnostic yield of endoscopy is 60% (43). Patients who are not responsive to these measures are often treated for H. pylori infection if these organisms are demonstrated on gastric biopsy; however, data from placebo-controlled trials suggest that treatment of H. pylori does not result in long-term relief of essential or nonulcer dyspepsia (44). An individualized approach is necessary using a combination of antisecretory agents, reassurance, and supportive care. The long-term prognosis is good with or without treatment in patients in whom endoscopy is negative.

Gastritis

Gastritis—inflammation of the stomach mucosa—is a nonspecific diagnosis that is made by endoscopic biopsy. It may be variably associated with dyspepsia, although a cause and effect relationship has not been documented. Several types of gastritis are seen in clinical practice.

Acute erosive, or hemorrhagic, gastritis is seen most commonly in seriously ill hospitalized patients; in patients taking NSAIDs, including aspirin; after heavy alcohol ingestion; and rarely in patients taking potassium chloride or iron supplements. Symptoms are variable but usually include nausea or vomiting and gastrointestinal bleeding.

Nonerosive or chronic antral gastritis is a histologic entity commonly seen in the general population, particularly the elderly and often caused by H. pylori infection. However, it is unclear whether the histologic (or endoscopic) entity of gastritis is associated with symptomatic disease. There is no correlation between eradication of H. pylori, histologic resolution of gastritis, and relief of dyspepsia, bloating, abdominal pain, or gas. In practice, patients with these syndromes should be treated as having non-ulcer dyspepsia. Symptomatic treatment of patients with nonerosive gastritis is the same as it is for patients with non-ulcer dyspepsia.

Specific References*

For annotated General References and resources related to this chapter, visit http://www.hopkinsbayview.org/PAMreferences.

1. Munnangi S, Sonnenberg A. Time trends of physician visits and treatment patterns of peptic ulcer disease in the United States. Arch Intern Med 1997;157:1489.
2. Kurata JH. Epidemiology: peptic ulcer risk factors. Semin Gastrointest Dis 1993;4:2.
3. Sonnenberg A, Everhart JE. Health impact of peptic ulcer in the US. Am J Gastroenterol 1997;92:614.
4. Sonnenberg A. Temporal trends and geographic variations of peptic ulcer disease. Aliment Pharmacol Ther 1995;9[Suppl2]:3.
5. The ACG Committee on FDA-Related Matters. Current status of maintenance therapy in peptic ulcer disease. Am J Gastroenterol 1988;83:607.
6. Fry J. Peptic ulcer disease: a profile. Br Med J 1964;2:809.
7. Greibe J, Bugge P, Gjorup T, et al. Long-term prognosis of duodenal ulcer: follow-up study and survey of doctors’ estimates. Br Med J 1977; 2:1572.
8. Penston JG, Wormsley KG. Review articles: maintenance treatment with H₂receptor antagonists for peptic-ulcer disease. Aliment Pharmacol Ther 1997;6:3.
9. Bardhan KD, Hinchliffe RFC, Bose K. Low dose maintenance treatment with cimetidine in duodenal ulcer: intermediate term results. Postgrad Med J 1986;62:347.
10. Bonnevie O. Survival in peptic ulcer. Gastroenterology 1978;75:1055.
11. Pounder RE, Ng D. The prevalence of Helicobacter pyloriinfection in different countries. Aliment Pharmacol Ther 1995;9[Suppl 2]:33.
12. National Institutes of Health. Helicobacter pyloriin peptic ulcer disease. JAMA 1994;272:65.
13. Graham DY, Lew GM, Evans DG, et al. Effect of triple therapy (antibiotics plus bismuth) on duodenal ulcer healing. A randomized controlled trial. Ann Intern Med 1991;115:266.
14. Hentschel E, Brandstatter G, Dragosics B, et al. Effect of ranitidine and amoxicillin plus metronidazole on the eradication ofHelicobacter pyloriand the recurrence of duodenal ulcer. N Engl J Med 1993;328:308.
15. Laine L, Hopkins RJ, Girardi LS. Has the impact of Helicobacter pyloritherapy on ulcer recurrence in the US been overstated? A meta-analysis of rigorously designed trials. Am J Gastroenterol 1998;93:1409.
16. Gisbert JP, Blanco M, Mateos JM, et al. H. pylorinegative duodenal ulcer. Prevalence and causes in 774 patients. Dig Dis Sci 1999;44:2295.
17. Fries JF, Miller SR, Spitz PW, et al. Toward an epidemiology of gastropathy associated with nonsteroidal antiinflammatory drug use. Gastroenterology 1989;96:647.
18. Lanza FL. A guideline for the treatment and prevention of NSAID-induced ulcers. Am J Gastroenterol 1998;93:2037.

P.670

19. Conn HO, Poynard T. Corticosteroid therapy does not induce peptic ulcer. J Intern Med 1994;236:619.
20. Chan, FK, Hung, LC, Suen BY, et al. Celecoxib versus diclofenac and omeprazole in reducing the risk of recurrent ulcer bleeding in patients with arthritis. N Engl J Med 2002;347: 2104.
21. Silverstein FE, Faich G, Goldstein JL, et al. Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: A randomized controlled trial. JAMA 2000;284:1247.
22. Del Valle J, Cohen H, Laine L, et al. Acid peptic disorders. In: Textbook of Gastroenterology. Vol 1. Philadelphia: Lippincott, Williams & Wilkins, 1999:1370.
23. Piper DW, McIntosh JH, Ariotti DE, et al. Life events and chronic duodenal ulcer: a case control study. Gut 1981;22:1011.
24. McCarthy DM. Smoking and ulcers: time to quit [Editorial]. N Engl J Med 1984;311:726.
25. Kurata JH, Nogawa AW. Meta-analysis of risk factors for peptic ulcer. Nonsteroidal anti-inflammatory drugs, Helicobacter pylori, and smoking. J Clin Gastroenterol 1997;24:2.
26. Sontag S, Graham DY, Belsito A, et al. Cimetidine, cigarette smoking, and recurrence of duodenal ulcer. N Engl J Med 1984;311:689.
27. Stemmermann GN, Marcus EB, Buist AS, et al. Relative impact of smoking and reduced pulmonary function in peptic ulcer risk. A prospective study of Japanese men in Hawaii. Gastroenterology 1989;96:1419.
28. Kirk AP, Dooley JS, Hunt RH. Peptic ulceration in patients with chronic liver disease. Dig Dis Sci 1980;25:756.
29. Kang JY, Wu AY, Sutherland IH, et al. Prevalence of peptic ulcer in patients undergoing maintenance dialysis. Dig Dis Sci 1988;33:774.
30. Aspirin Myocardial Infarction Study Research Group. A randomized, controlled trial of aspirin in persons recovered from myocardial infarction. JAMA 1980;243:661.
31. Dooley CP, Larson AW, Stace NH, et al. Double contrast barium meal and upper gastrointestinal endoscopy. A comparative study. Ann Intern Med 1984;101:538.
32. Committee on Endoscopic Utilization. Appropriate use of gastrointestinal endoscopy. Manchester, MA: American Society for Gastrointestinal Endoscopy, June 6, 1986.
33. McCarthy D. The place of surgery in the Zollinger-Ellison syndrome. N Engl J Med 1980;302:1344.
34. Cutler AF, Havstad S, Ma CK, et al. Accuracy of invasive and noninvasive tests to diagnose Helicobacter pyloriinfection. Gastroenterology 1995;109:136.
35. Cutler AF, Prasad VM. Long-term follow-up of Helicobacter pyloriserology after successful eradication. Am J Gastroenterol 1996;91:85.
36. Fallone CA, Mitchell A, Paterson WG. Determination of the test performance of less costly methods of Helicobacter pyloriinfection. Clin Invest Med 1995;18:177.
37. Faigel DO, Furth EE, Childs M, et al. Histological predictors of active Helicobacter pyloriinfection. Dig Dis Sci 1996;41:937.
38. Graham DY, Evans DJ, Peacock J, et al. Comparison of rapid serological tests (FlexSure HP and QuickVue) with conventional ELISA for detection of Helicobacter pyloriinfection. Am J Gastroenterol 1996;91:942.
39. Chan F, Sung J, Lee YT, et al. Does smoking predispose to peptic ulcer relapse after eradication of Helicobacter pylori? Am J Gastroenterol 1997;92:442.
40. Gisbert, JP, Khorrami, S, Carballo, F, et al. Meta-analysis: Helicobacter pylorieradication therapy vs. antisecretory non-eradication therapy for the prevention of recurrent bleeding from peptic ulcer. Aliment Pharmacol Ther 2004;19:617.
41. Lanzon-Miller S, Pounder RE, Hamilton MR, et al. Twenty four hour intragastric acidity and plasma gastrin concentration before and during treatment with either ranitidine or omeprazole. Aliment Pharmacol Therap 1987;1:239.
42. Jordan PH Jr., Thornby J. Should it be parietal cell vagotomy or selective vagotomy-antrectomy for treatment of duodenal ulcer? A progress report. Ann Surg 1987;205:572.
43. Talley NJ, Phillips SF. Nonulcer dyspepsia: potential causes and pathophysiology. Ann Intern Med 1988;108:865.
44. Talley NJ, Hunt RH. What role does Helicobacter pyloriplay in dyspepsia and nonulcer dyspepsia? Arguments for and against. H. pylori being associated with dyspeptic syndrome. Gastroenterology 1997;113:S67.