Twenty-five million Americans have symptomatic peptic ulcer disease (PUD), which is defined as the presence of defects in the gastrointestinal mucosa that extend through the muscularis propria. The most common symptom of PUD is a gnawing or burning epigastric pain which often occurs between meals and in the early morning, but that may occur at any time. It may be relieved by eating food or by taking antacids. Other symptoms may include nausea, vomiting, and decreased appetite. Fatigue, dizziness, hematemesis, melena, or hematochezia may occur if there is bleeding. This chapter will provide a brief approach to the outpatient management of peptic ulcer disease and will not address the problem of acute gastrointestinal bleeding.

Most duodenal and gastric ulcers are caused either by Helicobacter pylori infection or by nonsteroidal anti-inflammatory drugs (NSAIDs). Among patients with ulcers who do not use NSAIDs, the prevalence of H. pylori infection is estimated to be 95 percent; the incidence of ulcers in H. pylori-infected persons is six to ten times higher than that of uninfected people. Prolonged NSAID use (for more than two weeks) is an independent risk factor for PUD.

H. pylori is a gram-negative bacillus that was first described in 1983 by Marshall and Warren. Though H. pylori is present in about two-thirds of the world’s population, most infected individuals are asymptomatic. Moreover, only 15 to 20 percent of people infected with H. pylori develop ulcers. In the last 15 years, H. pylori infection has been conclusively linked to peptic ulcer disease; as noted, the relative risk of PUD among H. pylori-infected persons is 4 to 10. Treatment of H. pylori in persons with PUD speeds healing and dramatically decreases the risk of recurrence. In the United States, H. pylori is more prevalent among older adults, African Americans, Hispanics, and lower socioeconomic groups. Its overall prevalence is falling rapidly in developed nations, including the U.S. In the presence of H. pylori, host factors, such as smoking and blood group (O), may predispose a patient to ulceration, although no consistent mechanism of ulcer pathogenesis has been identified.

H. pylori is also related to two-thirds of gastric adenocarcinoma, although only a small percentage of patients with H pylori infection will develop malignancy. H. pylori is believed to predispose to gastric cancer by inducing gastric atrophy and hypochlorhydria. Relatives of patients with gastric cancer, with H. pylori infection, have been found to have an increased prevalence of precancerous gastric abnormalities (i.e. gastric atrophy and hypochlorhydria). Some experts recommend serologic H. pylori screening of family members of patients with gastric carcinoma, especially intestinal type. Family members with positive serologies should then receive antibiotic treatment aimed at H. pylori eradication. Gastric mucosal-associated-lymphoid type (MALT) lymphoma, an uncommon disease, is also associated with H. pylori. There may be regression in 60 to 70 percent of patients with MALT after eradication of infection.

The gold standard test for diagnosis of H. pylori is endoscopy with biopsy and histology. The urease dye test (CLO test) performed with endoscopy requires a biopsy but may provide the diagnosis more rapidly than histologic examination. Culture is the least sensitive of the direct techniques because H. pylori is extremely difficult to culture. Serology is simple and inexpensive, but is not specific for active infection. The urease breath test is both sensitive and specific for active infection, although antimicrobial drugs can temporarily cause false negative results and the test must be used at least four weeks after treatment is completed. The breath test is the best post-treatment test for eradication and would be recommended for everyone if it were readily available and less expensive. Unfortunately, none of the companies that perform it accept Medicaid, and only patients with Medicare and commercial insurance have access. Other patients require a second endoscopy and biopsy in order to prove H. pylori eradication.

TABLE 1: Sensitivity and specificity of diagnostic tests for H. pylori

Deciding which patients to test for H. pylori should be relatively straightforward. The American College of Gastroenterology (ACG) has recently published guidelines for H. pylori management, which stress the following points:

• There is no indication to test patients if you do not plan to offer treatment.
• There is no indication to test asymptomatic patients (unless they have a past history of PUD).
• There is no indication to test patients with abdominal pain other than that due to suspected or proven PUD.
• There is no indication to test patients with GERD, unless they are also suspected of having PUD.

When to Treat:

While experts agree that all patients with PUD and H. pylori infection should be treated for H. pylori, there is significant controversy over the best diagnostic approach. Most guidelines recommend starting with endoscopy if "alarm features" are present (Table 2).^,, Note that this includes all persons over the age of 45, a stance which is colored by the desire to exclude malignant and pre-malignant conditions (such as Barrett’s esophagus). If no alarm features are present, starting with a therapeutic trial of antisecretory or prokinetic agents is recommended. Trials should be stopped after two to six weeks and further evaluation pursued if symptoms persist or recur.⁴

TABLE 2: "Alarm features" which may prompt endoscopy

Another option, if no alarm features are present, is to check the urease breath testi and treat based on the results – with a H. pylori eradication regimen if positive and antisecretory or prokinetic agents if negative. In addition to the fact that Medicaid patients do not have access to the urease breath test in our setting, the disadvantage of this approach is that not all patients with dyspepsia will have ulcer disease. The 15 to 20 percent of H. pylori-infected patients who do have ulcer disease will likely respond to treatment,⁴ but a large number of patients will be exposed to antibiotic therapy with no benefit. Empiric antibiotic therapy without H. pylori testing is strongly discouraged because its prevalence is probably less than 50 percent and widespread antibiotic use may lead to resistance.

TABLE 3: Indications for treatment of H. pylori

The discovery of H. pylori and the development of effective therapy have revolutionized the treatment of peptic ulcer disease. Eradication of H. pylori dramatically reduces, but does not eliminate, ulcer recurrence. The first randomized trial of H. pylori eradication was in 1987, and the first report of successful therapy (with a two-week bismuth-based triple therapy) was in 1991. Subsequent studies have shown that the recurrence of both duodenal and gastric ulcers is reduced to 15 percent after H. pylori eradication compared with a 60 to 100 percent recurrence rate after antisecretory treatment alone. Antibiotic resistance and patient nonadherence are the two main reasons for treatment failure.

It is not clear which is the best treatment regimen. Most trials of therapeutic regimens had small numbers of patients or were not controlled, randomized, or blinded. At least two antimicrobial agents must be used because of the emergence of antibiotic resistance. Most currently used regimens include two antibiotics and an antisecretory agent, usually a proton pump inhibitor. The use of metronidazole OR clarithromycin (not both) is recommended. In regions where metronidazole resistance is increasing, secondary resistance to clarithromycin can be acquired when the agents are used together, leading to selection of a strain with dual resistance.

TABLE 4: Treatment regimens for eradication of H. pylori

Eradication rates in clinical trials range from 61 to 94 percent. The standard 14 day bismuth, metronidazole, and tetracycline (BMT) regimen results in high eradication rates (~89 percent), but adherence is lower than in other regimens because of side effects and frequent dosing. In some settings, BMT may still be the preferred regimen given its efficacy and lower cost. H. pylori can be eradicated with triple antibiotic therapy alone, but the addition of antisecretory agents, such as omeprazole (Prilosec^TM) or lansoprazole (Prevacid^TM) improves symptom relief. When an ulcer is detected, omeprazole should be continued for two additional weeks after the antibiotic course. In addition, smoking cessation is an important intervention that promotes healing and reduces recurrence.

While several studies have demonstrated that one-week regimens can be effective in the setting of a carefully monitored clinical trial, a 10 or 14 day course of antibiotics is still the standard of care. Table 4 outlines some of the many recommended regimens – it is reasonable for a primary care provider to become familiar with two or three "cocktails." The combination of lansoprazole, clarithromycin, and amoxicillin can be prescribed as a PrevPac^TM, in which pills are packaged together for simpler dosing; while this may be the simplest to take (and to prescribe), eradication rates are only 85 percent. The addition of a proton pump inhibitor to the standard bismuth, metronidazole, and tetracycline (BMT) regimen can be up to 95 percent effective. Bismuth, metronidazole and tetracycline can be prescribed as the combination pill Helidac.^TM

American physicians write more than 90 million prescriptions for NSAIDs each year for more than 20 million Americans. In addition, there are more than 200 NSAID-containing products available over the counter, making this one of the most frequently used classes of drugs. NSAIDs inhibit prostaglandin synthesis, and affect the production of gastric acid, bicarbonate and glutathione as well as the integrity of the gastric mucosal barrier, creating an imbalance that may lead to ulcer formation. Symptomatic ulcers occur in approximately one percent of patients taking nonsalicylate NSAIDs after three to six months of chronic use, and in two to four percent after one year, leading to more than 100,000 hospitalizations and 15,000 deaths annually.^,, Benign gastric ulcers are more frequently NSAID-associated than are duodenal ulcers. There does not seem to be an increased risk of ulceration in H. pylori-infected patients taking NSAIDs, although this has recently been debated.

Dyspepsia is not predictive of NSAID gastropathy among patients taking these medications. Oddly, symptoms and pathology are not correlated. In addition, many patients who present with complicated ulcer disease have no antecedent symptoms. Gastroesophageal damage is seen on endoscopy in 20 to 40 percent of patients taking chronic NSAIDs, and the point prevalence of asymptomatic gastric ulcers is 15 to 30 percent. This is in marked contrast to the two to four percent of NSAID-users with symptomatic ulcer disease and the clinical significance of such "endoscopic ulcers" is unknown.

Diagnosis of NSAID-associated ulcers is made by history and by endoscopy. If possible, NSAIDs should be stopped. It is not advisable to switch patients with newly diagnosed ulcers to COX-2 inhibitors, as these may slow angiogenesis and ulcer healing. If H. pylori is present, is should be eradicated. The effect of omeprazole is not compromised by the presence of NSAIDS and proton pump inhibitors are preferred over H2 receptor antagonists for complicated, slowly healing, or large ulcers if NSAIDs must be continued.

Clinical factors associated with an increased risk of complicated NSAID ulcer disease include age older than 75 years, history of PUD, history of gastrointestinal bleeding, corticosteroid use, and cardiac disease. In a large study of patients with rheumatoid arthritis taking chronic NSAIDs, ulcer complications occurred in only 0.4 percent of patients with none of these risk factors, in one percent when one factor was present, and in nine percent when all four factors were present. It is these high-risk patients who should be considered for preventive therapy.

In patients with rheumatoid arthritis taking NSAIDs, misoprostol 200ug QID decreased the incidence of bleeding and ulcer by 40 percent. However, the overall incidence of complications was only 1.5 percent per year. Many patients, therefore, would need to be treated at a high cost ($233,000 per case prevented with misoprostol) to prevent a small number of complications.

Preventive therapy in patients taking NSAIDs is indicated if there is a definite history of ulcer disease, simultaneous steroid or anticoagulant use, or serious comorbid conditions that would compromise tolerance of ulcer complications.⁴ Misoprostol is the only FDA-approved drug for prevention of NSAID-induced ulcers, though it is rarely used. Side effects of misoprostol include severe diarrhea and abdominal cramping. Arthrotec^TM is a combined NSAID-misoprostel tablet that may be easier for patients to tolerate; there are no experimental data regarding its efficacy in preventing NSAID-associated PUD. H2-receptor antagonists can decrease the incidence of NSAID-induced duodenal ulcers, but some studies showed they do not prevent NSAID-induced gastric ulcers. One study did show famotidine at 40 mg BID to be effective in preventing both gastric and duodenal NSAID-induced ulcers. There are no data on the prevention of NSAID-induced ulcer complications with proton pump inhibitors, but these may be reasonable alternatives for patients on NSAIDs who cannot tolerate misoprostol.

Two COX-2 inhibitors, celecoxib (Celebrex^TM) and rofecoxib (Vioxx^TM) have been approved for treatment of osteoarthritis and rheumatoid arthritis. Celecoxib appears to be as effective as older NSAIDs as an analgesic, though it may be less effective in treating acute pain. In short-term studies, celecoxib caused fewer "endoscopic" ulcers than older NSAIDs and did not inhibit platelet aggregation or increase the bleeding time.^, However, both celecoxib and rofecoxib if given with warfarin, can increase the INR and prothrombin values and may increase the risk of bleeding.. In a study of patients with rheumatoid arthritis, treatment with rofecoxib was associated with significantly fewer clinically important upper GI events than treatment with Naproxen. Another trial with six months follow up also showed that patients taking rofecoxib had fewer episodes of upper GI tract bleeding. After one year, the cumulative incidence of upper GI tract perforations, gastroduodenal ulcers, and upper GI tract bleeds was 1.3 percent with rofecoxib compared to 1.8 percent with other NSAIDs (p=.046). Whether serious GI bleeding will actually occur less frequently with these agents remains to be seen.

Nonulcer dyspepsia (NUD) has been defined as chronic or recurrent upper abdominal pain or discomfort for a period of more than three months’ duration, with symptoms present for more than 25 percent of the time, in the absence of another organic cause. A simpler definition is that offered by Locke: "persistent or recurrent upper abdominal pain or discomfort with no structural or biochemical explanation for the patient’s symptoms." NUD may include bloating, nausea, early satiety, eructation and heartburn. It is a symptom complex rather than a specific condition.

An organic cause is found in only 40 percent of patients with dyspepsia. The most common of these are gastroduodenal ulcer, GERD, gastroparesis, and gastric cancer. Other causes include cholelithiasis or choledocolithiasis, pancreatitis, carbohydrate malabsorption, intestinal parasites, NSAID or other medication injury, diabetes, thyroid disorders or connective tissue disorders, ischemic bowel, and abdominal cancer. The 60 percent of patients without an organic cause are considered to have "nonulcer dyspepsia," and fall into a continuum of functional gastrointestinal disorders, including irritable bowel syndrome, functional heartburn, and noncardiac chest pain.

The pathophysiology of nonulcer dyspepsia is not well understood. It was initially speculated that because H. pylori was the main cause of PUD, perhaps it also played a role in NUD. Trials of H. pylori eradication in patients with NUD, however, have not been encouraging. A single study did find some benefit: McColl et al. found that symptoms resolved in 7 percent of patients treated with omeprazole alone compared with 21 percent of patients treated with omeprazole and antibiotics. In contrast, Blum et al. found that dyspepsia resolved in 21 percent of patients treated with omeprazole alone, a rate only slightly lower than that for those treated with omeprazole and antibiotics. Most recently, Talley et al. randomized patients to placebo vs. omeprazole and antibiotics, demonstrated H. pylori eradication in their treatment group and found no clinical difference at 12 months, supporting their conclusion that curing H. pylori has no effect on symptoms of NUD. While we recognize that there is still some controversy about this topic, we recommend following the current AGA guidelines² which strongly state that patients with NUD do not need to be tested or treated for H. pylori.

The evaluation and treatment of patients with nonulcer dyspepsia remains unclear. Initial treatment algorithms are similar to those for suspected peptic ulcer disease. "Alarm features" such as age greater than 45, weight loss, vomiting, dysphagia, and bleeding (Table 2), should prompt early endoscopy in a patient with dyspepsia. If none of these signs are present, and symptoms suggest peptic ulcer disease then management may follow that outlined in Figure 1. Once peptic ulcer disease has been excluded and a diagnosis of NUD has been made, however, evidence-based treatment options do not exist. Trials of antisecretory or promotility agents are often recommended. Simethicone has been anecdotally reported to help some patients with NUD. Some authors recommend antidepressants or supportive psychotherapy and others suggest subspecialty referral. If a patient’s quality of life is seriously impaired by nonulcer dyspepsia, we recommend subspecialty referral. If not, reassurance and symptomatic treatment is the approach of choice.

Gastroesophageal reflux disease (GERD) is often described as heartburn that may be precipitated by eating or lying down. Heartburn is an extremely prevalent complaint; 20 percent of the adult U.S. population has weekly heartburn, and even more experience monthly symptoms. Other symptoms, such as chronic cough, sore throat, hoarseness, chest pain, and asthma may also be related to GERD.

GERD is related to inappropriate transient lower esophageal sphincter (LES) relaxation, which is thought to be more important than a reduced basal LES pressure in causing GERD. Other factors that may contribute to reflux are impaired esophageal motility and delayed gastric emptying. The presence of a hiatal hernia does not cause reflux unless it is associated with reduced esophageal sphincter tone. GERD may cause esophageal complications by mucosal damage. Erosive disease can lead to Barrett’s esophagus, a premalignant lesion. Barrett’s esophagus is diagnosed on endoscopy by gastric-appearing mucosa in the tubular esophagus that has columnar epithelium with intestinal metaplasia. It is found in 10 percent of patients with chronic GERD, though it may be asymptomatic. Other complications of erosive esophagitis include stricture formation leading to dysphagia and upper gastrointestinal bleeding.

It is reasonable to make a presumptive clinical diagnosis of GERD. Endoscopy is not necessary in young patients or in those whose symptoms are relieved with H2 blockers or prokinetic agents. However if a patient requires long-term proton pump inhibition, or if symptoms are unresponsive to medical therapy, then endoscopy is warranted. Some gastroenterologists recommend that every patient with GERD should have an endoscopy at least once in his or her lifetime to rule out erosive esophagitis, Barrett’s and malignancy. Upper endoscopy may also be performed in evaluating "noncardiac" chest pain to look for erosive esophagitis. Intraesophageal 24-hour ambulatory pH monitoring can demonstrate abnormal esophageal acid reflux and is sometimes necessary for diagnosis if endoscopy is negative.

Therapy for GERD is usually life-long. Patients must be educated about the advantage of making lifestyle changes, such as weight reduction and to avoid fatty foods, chocolate, peppermints, and alcohol. They should avoid lying down for three hours after meals and should elevate the head in bed.

If symptoms persist after lifestyle changes, then H2 blockers or prokinetic agents may be tried. Metoclopramide (Reglan^TM) 10 mg po qac and qhs may be a reasonable substitute. If symptoms remain uncontrolled, then proton-pump inhibitors may be tried. Proton-pump inhibitors have been shown to treat symptoms, heal erosive esophagitis, esophageal strictures with esophagitis, and Barrett’s esophagus. They are superior to H2 blockers in treating erosive esophagitis and Barrett’s.¹⁶ When proton-pump inhibitors are used long term in patients with H. pylori, there is an increased risk of atrophic gastritis. However, eradication of H. pylori in GERD is not currently recommended. Fundoplication is another treatment with good initial results, however, long term efficacy is not clear. More than 33 percent of patients develop recurrent reflux symptoms after surgery. Gastroenterolgoists at CPMC are now evaluating non-operative (endoscopic) methods for treatment of GERD. These include ablative and endoscopic suturing methods.

We thank Dr. Reuben Garcia-Carrasquillo for his helpful comments and suggestions.