ACP medicine, 3rd Edition


Peptic Ulcer Diseases

Mark Feldman MD, MACP1

Michael S. Le MD2

1William O. Tschumy, Jr., M.D., Chair of Internal Medicine and Clinical Professor of Internal Medicine, University of Texas Southwestern Medical School at Dallas, Director, Internal Medicine Residency Program, Presbyterian Hospital of Dallas

2Resident of Medicine, University of Mississippi Medical Center

The authors have no commercial relationships with manufacturers of products or providers of services discussed in this chapter.

May 2007


Peptic ulcers are holes in the inner lining of the gastrointestinal tract that are attributed to exposure of the mucosa to gastric acid and pepsin. Peptic ulcers extend through the mucosa and the muscularis mucosae, a thin layer of smooth muscle separating the mucosa from the deeper submucosa, muscularis propria, and serosa. Most peptic ulcers are round or oval, but some are linear, triangular, or irregular in shape. Ulcers have depth when viewed through an endoscope. Typically, only a single ulcer is present. An erosion is a focal loss of superficial epithelial cells and glands, without extension through the muscularis mucosae. On endoscopy, erosions appear as breaks in the mucosal lining without depth. At the other extreme, a peptic ulcer may burr itself entirely through the wall of the GI tract, thus connecting the GI lumen with the peritoneal cavity (perforated ulcer), a solid organ such as the pancreas (penetrating ulcer), or another hollow organ such as the intestine or bile duct (fistulizing ulcer).


In the United States, peptic ulcer disease affects up to 10% of men and 4% of women at some time in their lives. The incidence is influenced by age (older persons are more susceptible than younger persons) and gender (males are more susceptible than females). Because ulcer disease is often recurrent, its prevalence exceeds its incidence. Eradication of Helicobacter pylori from the stomach markedly reduces recurrence of ulcer disease. With the widespread use of treatment regimens for H. pylori, the prevalence of peptic ulcer is decreasing in the United States. Reinfection with H. pylori remains an uncommon event in the United States, with an incidence of approximately one reinfection per 100 patients a year.

Pathogenesis and Etiologic Factors

The normal stomach and duodenum are able to resist autodigestion by acid-pepsin. However, high rates of acid-pepsin secretion or impaired mucosal resistance factors, such as prostaglandin deficiency, can predispose to duodenal ulcer formation (typically in the most proximal part of the duodenum, the bulb) or to gastric ulcer formation (typically in the most distal part of the stomach, the antrum).

On rare occasions, peptic ulcers occur in the second, third, or fourth portion of the duodenum (postbulbar ulcer) or even in the proximal jejunum. Ordinarily, alkaline secretions from the duodenum, biliary tract, and pancreas neutralize gastric acid in the duodenum, but high rates of gastric acid secretion (e.g., in Zollinger-Ellison syndrome) can overwhelm these endogenous alkaline secretions and lead to postbulbar or jejunal ulcerations.

In patients with pathologic amounts of gastroesophageal reflux of acid-pepsin, and in many patients with gastric acid hypersecretion, erosions and ulcers may develop in the lower esophagus [see 4:I Esophageal Disorders]. Peptic ulcers may also occur where acid and pepsin are secreted heterotopically, such as in a congenital ileal (Meckel) diverticulum.

Regardless of location and etiology, chronic peptic ulcers are similar pathologically. In addition to the focal loss of mucosal epithelial cells, these ulcers have four characteristic layers at their base: fibrinoid necrosis, exudate, granulation tissue, and a fibrous scar (the deepest layer). A layer of granulation tissue and fibrosis may be absent in acute ulcers that occur in settings of serious trauma or severe surgical or medical illnesses [see Acute Stress Ulcers, below].

Why a peptic ulcer is such a focal lesion is unclear. Although peptic ulcers require the presence of acid-pepsin, acid-pepsin alone is only rarely sufficient to produce an ulcer, unless marked gastric hypersecretion is present, such as in Zollinger-Ellison syndrome. In the majority of patients, there must be another predisposing factor, such as H. pylori infection of the stomach,1 use of nonsteroidal anti-inflammatory drugs (NSAIDs), or stress [see Figure 1].


Figure 1. Etiopathogenesis of peptic ulcers. (a) Helicobacter pylori induces a diffuse, chronic, active superficial gastritis, usually throughout the stomach. Exactly how this infectious gastritis results in peptic ulcer disease is unknown. (b) Nonsteroidal anti-inflammatory drugs (NSAIDs) block cyclooxygenase-1 (COX-1) to reduce the amount of gastroduodenal prostaglandins (PGs) synthesized from their precursor, arachidonic acid. COX-2 selective inhibitors do not reduce prostaglandins and are associated with much fewer peptic ulcers than nonselective COX-1/COX-2—inhibiting NSAIDs. (c) Gastrinoma in the pancreas or duodenum secretes large amounts of gastrin into the circulation. Elevated serum gastrin levels promote the release of histamine by acting on receptors for cholecystokininB (CCKB) and gastrin, which are located on gastric enterochromaffin-like (ECL) cells. Histamine acts on H2 receptors on parietal and chief cells to augment hydrochloric acid (HCl) and pepsin secretion.

  1. Pylori ulcers

The prevalence of H. pylori infection of the stomach is much higher in duodenal ulcer patients and, to a somewhat lesser extent, in gastric ulcer patients than in age-matched control subjects [see Figure 2].2 In addition, cure of H. pylori infection with antimicrobial therapy markedly reduces recurrences of duodenal and gastric ulcers.3,4 The correlation of H. pylori infection with peptic ulcers is not consistent, however. Duodenal ulcers develop in some infected persons; in others, gastric ulcers develop. Most infected persons, however, experience no ulcers at all.


Figure 2. Point prevalence of H. pylori infection assessed by endoscopic biopsy and mucosal histology. The dark-blue bars represent patients with active ulcer (13 duodenal and 13 gastric). The gray bars indicate patients whose ulcers are inactive (20 duodenal and 23 gastric). The light-blue bars represent 26 age- and sex-matched control subjects. None of these ulcer patients were receiving aspirin or NSAIDs.3

The etiologic mechanism linking H. pylori infection and ulcerogenesis is not yet absolutely established, for the following reasons: (1) voluntary ingestion of H. pylori leads to gastric H. pylori infection and to gastritis but not to ulcers; (2) duodenal or gastric ulcers develop in only 10% to 20% of individuals with H. pylori gastritis, implying that only certain persons with additional genetic, anatomic, physiologic, or environmental risk factors are predisposed to ulcers or that only certain H. pylori strains are ulcerogenic; (3) H. pylori induces a diffuse inflammation in the stomach, yet the strongest link between H. pylori and peptic ulcer is with focal duodenal bulbar ulcer; and (4) gastric H. pylori infection is as common in women as in men, yet duodenal ulcer is two to three times less common in women. Currently, H. pylori can be considered the most important risk factor for duodenal and gastric ulcers, but it is clear that the mere presence of H. pylori in the stomach is not sufficient to cause peptic ulcers [see Figure 1, part a]. Patients with duodenal ulcers tend more often to be infected with cagA-positive strains than do H. pylori-infected patients without ulcer.5 Genetic studies of H. pylori strains in East Asia and South America have disclosed a duodenal ulcer-promoting (dupA) gene of H. pylori. This gene was associated with an increased risk of duodenal ulcers and a reduced risk for gastric atrophy, intestinal metaplasia, and gastric cancer.6

NSAID Ulcers

The ulcerogenicity of NSAIDs has been established experimentally by exposing animals, human volunteers, and patients to these drugs. Experimental studies have been corroborated by numerous case-control studies and autopsy studies. Unlike H. pylori-related peptic ulcers, which more often occur in the duodenal bulb, NSAID ulcers typically occur in the stomach. A gastric or duodenal ulcer associated with NSAID use is classified as a peptic ulcer, and it usually heals with potent acid antisecretory therapy, even if NSAID use is continued. NSAIDs can also cause ulcers in the jejunum, ileum, or colon, where little or no acid-pepsin secretion occurs. These ulcers are not actually peptic ulcers.

Although the pathogenesis of NSAID ulcers is multifactorial, by far the most important mechanism appears to be inhibition of cyclooxygenase-1 (COX-1), the rate-limiting enzyme in GI prostaglandin synthesis [see Figure 1, part b]. Prostaglandins normally protect the GI mucosa from damage by maintaining mucosal blood flow and increasing mucosal secretion of mucus and bicarbonate. Blockade of COX-1 activity by NSAIDs reduces prostaglandin synthesis and thus lowers GI mucosal blood flow and secretion of mucus and bicarbonate. Evidence continues to accumulate indicating that H. pylori gastritis increases the risk of gastroduodenal ulcer formation in NSAID users around twofold.7 Moreover, clinically symptomatic peptic ulcers occur much less commonly if H. pylori gastritis is treated before starting long-term NSAID therapy.8 Clinically diagnosed peptic ulcers will develop in approximately 2% to 4% of persons taking NSAIDs per year of exposure. The extent to which the damaging effects of NSAIDs on the stomach are topical rather than systemic is unclear. Many NSAIDs, such as aspirin, are acidic and thus are nonionized in the acidic stomach, where they can be absorbed and initiate gastric mucosal damage. However, NSAIDs (e.g., ketorolac) given by parenteral injection and aspirin given transdermally are ulcerogenic; so too are so-called NSAID prodrugs, such as sulindac and nabumetone, which do not inhibit gastric prostaglandins until they are metabolized to their active form after GI absorption. Evidence suggests that acute mucosal damage by NSAIDs (i.e., hemorrhages and erosions but seldom ulcers) is mainly caused by the topical damaging effects of NSAIDs. In contrast, chronic ulcer formation, often with complications such as bleeding and perforation, is mainly the result of the systemic effect of NSAIDs on prostaglandin synthesis by the GI mucosa.

Epidemiologic studies suggest that NSAIDs vary in their ability to cause ulcers,9 but this issue is complicated by the difficulty of comparing equipotent doses of NSAIDs. All prescription or over-the-counter NSAIDs should be considered ulcerogenic, with the risk of ulcer related to dosages and other patient-related factors, particularly advanced age and previous ulcer history. Epidemiologic studies indicate that the greatest risk of NSAID ulcers is early in the course of treatment (between day 7 and day 30 after initiation), with the risk decreasing thereafter. Even low doses of aspirin used for prophylaxis of cardiovascular disease (i.e., 75 to 325 mg/day) are ulcerogenic in humans.10Neither buffering of aspirin nor enteric coating appears to reduce the incidence of clinically detected ulcer formation.11 Nonacetylated salicylates such as salicylic acid (salsalate) do not block COX-1 and are not ulcerogenic.

Most NSAIDs, including aspirin, block both COX-1 and COX-2. Unlike COX-1, COX-2 is induced and expressed at inflammatory sites but not in the normal GI tract.12 Celecoxib is a selective COX-2 inhibitor that is analgesic and anti-inflammatory, but it offers little GI advantage over traditional NSAIDs and also carries a potential for adverse cardiovascular outcomes.13

Corticosteroids, which block COX-2 but not COX-1,12 are not ulcerogenic when used alone, although they can impair healing of preexisting ulcers. When corticosteroids are used in combination with NSAIDs, the risk of ulcer formation is much greater than when NSAIDs are used alone.

Ulcers in Gastrinoma or Other Hypersecretory States

A gastrinoma is an endocrine tumor of the pancreas or duodenum (usually malignant) consisting of gastrin (G) cells. Gastrinoma causes less than 1% of all peptic ulcers. Peptic ulcers develop in 95% of patients with gastrinoma (Zollinger-Ellison syndrome); ulcers occur most commonly in the duodenal bulb but are also seen in the postbulbar duodenum, jejunum, lower esophagus, and stomach. Multiple ulcers are present in up to 25% of cases of Zollinger-Ellison syndrome.

Patients with a gastrinoma have high circulating levels of gastrin [see Figure 1, part c], which acts on receptors for cholecystokininB (CCKB) and gastrin located on enterochromaffin-like (ECL) cells within the mucosa of the gastric body. ECL cells then release histamine, which acts on H2 receptors present on the membrane of neighboring parietal cells to stimulate (via an adenylate cyclase-cyclic adenosine monophosphate [cAMP]-mediated pathway) the secretion of hydrochloric acid by a unique proton pump, the H+, K+-ATPase pump. Of less physiologic importance, gastrin also acts directly on CCKB/gastrin receptors on parietal cells, increasing cytosol calcium levels in the parietal cells.

Hypergastrinemia in Zollinger-Ellison syndrome results in a continuous high rate of secretion of hydrochloric acid and pepsin, even under basal (fasting) conditions. These secretions overwhelm the buffering and neutralizing capacity of food and upper digestive secretions, as well as mucosal defense factors. Peptic ulceration results, and in many cases, diarrhea (with or without malabsorption) occurs.

Approximately 20% to 30% of patients with gastrinomas have features suggesting a multiple endocrine neoplasia type I (MEN I) syndrome, such as hypercalcemia secondary to hyperparathyroidism, a pituitary adenoma, or both. MEN I is inherited as an autosomal dominant disorder.

Some patients with duodenal ulcer have marked acid hypersecretion but normal serum gastrin levels. A few of these patients have hyperhistaminemia caused by systemic mastocytosis or chronic basophilic leukemia. For the majority of patients with acid hypersecretion, there is no known reason for the disorder (such patients are said to have idiopathic basal acid hypersecretion. Some patients, however, are infected with H. pylori; in these patients, eradication of H. pylori may reduce basal acid hypersecretion.

Idiopathic Ulcers

Up to 20% of chronic gastric and duodenal ulcers in the United States occur in patients who have no evidence of H. pylori infection, deny taking NSAIDs, and have normal serum gastrin concentrations. These ulcers are referred to as idiopathic peptic ulcers. Some patients with this disorder may be taking NSAIDs surreptitiously or are unaware that they are taking these drugs. In others, emotional stress and cigarette smoking, perhaps associated with gastric acid hypersecretion, may be contributing factors.14 Cohort studies of patients with bleeding peptic ulcers in Hong Kong have demonstrated that there has been a recent rise in the relative incidence of idiopathic ulcers compared with ulcers from medications such as NSAIDs, H. pylori, or to both of these factors. Idiopathic ulcers represented close to 20% of recent admissions for ulcer bleeding.15 Furthermore, patients with idiopathic ulcers had a relatively high risk of recurrent bleeding over the ensuing 12 months (13%). Interestingly, 30% of the Hong Kong patients with idiopathic ulcers had histologic evidence of intestinal metaplasia in the stomach or of gastric atrophy, suggesting past H. pylori infection.15

Acute Stress Ulcers

Acute gastroduodenal erosions and ulcers are very common in patients with serious medical and surgical conditions.16 Such conditions include severe head injury (Cushing ulcers); burn injury (Curling ulcers); major surgical procedures; and life-threatening illnesses such as septic shock, respiratory failure requiring mechanical ventilation, hepatic failure, renal failure, and multiorgan failure. Unlike peptic ulcers, stress ulcers are typically asymptomatic, rarely causing dyspepsia or epigastric pain. Approximately 10% to 25% of patients with acute stress ulcers experience painless upper GI bleeding of variable severity. Bleeding may manifest itself in the intensive care unit as a dark (so-called coffee-ground) or bloody nasogastric aspirate, as a declining hematocrit, as an increasing transfusion requirement, or as unexplained hypotension.

The pathogenesis of stress ulcers is not well understood. The common denominator seems to be tissue hypoxia and acidosis, precipitated by mucosal vasoconstriction and ischemia. Systemic hypoxia, metabolic acidosis, anemia, and reduced cardiac output often are contributing factors. Once mucosal hypoxia develops, mucosal defense factors are impaired and the cells lining the stomach and duodenum become vulnerable to damage by acid-pepsin. Acute stress ulcers have become less common because of the routine use of effective prophylactic medications in patients at high risk for this condition (see below).

Cameron Ulcers

Linear gastric erosions that occur in a hiatus hernia are known as Cameron ulcers.16 The erosions are thought to be related either to traumatic injury of the stomach by the surrounding diaphragm or to mucosal ischemia at the point where the stomach herniates through the diaphragm. Like acute stress ulcers, Cameron ulcers tend to present as bleeding without dyspepsia. Both acute and chronic GI blood loss are possible outcomes of Cameron ulcers.


Clinical Manifestations

Peptic ulcers produce a variety of symptoms but none specific for the disease. Also, symptoms of duodenal ulcer are indistinguishable from those of gastric ulcer. Patients with uncomplicated ulcers typically experience mild to moderate abdominal pain, usually in the epigastrium. However, the pain may be localized to the left or right upper quadrant of the abdomen, to the lower chest (subxiphoid or substernal), the midabdomen, or the back. The pain is often gnawing or burning. It may occur in the middle of the night; rarely, it occurs upon first awakening in the morning. Discomfort is typically relieved by food or an antacid. Severe pain or a rapid increase in pain suggests an ulcer complication (e.g., perforation or penetration) or another diagnosis (e.g., acute pancreatitis).

Associated dyspeptic symptoms include nausea, bloating, heartburn, and belching. Although vomiting may occur with uncomplicated peptic ulcers and may temporarily relieve pain, repeated vomiting suggests an ulcer complication (e.g., gastric outlet obstruction) or another diagnosis (e.g., intestinal obstruction).

Peptic ulcers are the most common cause of acute upper GI bleeding. Therefore, hematemesis, melena, or both, even in a patient with no history of ulcer and no dyspeptic symptoms, should suggest the possibility of a bleeding peptic ulcer. Patients who develop ulcers while taking prescription or over-the-counter NSAIDs or low (cardiovascular) doses of aspirin often have no history of ulcerlike pain. Other patients with bleeding ulcers will have experienced dyspeptic symptoms for the preceding days or weeks, only to have these symptoms wane when bleeding ensues.

In addition to a review of the patient's symptoms and ulcer risk factors (particularly NSAID use and smoking), a family history should be obtained. A family history of ulcer can usually be attributed to within-family infection by H. pylori, to NSAID use, or to smoking. However, a family history of ulcer, hyperparathyroidism, kidney stones, or endocrine tumor should alert the physician to the possibility of gastrinoma (Zollinger-Ellison syndrome), with or without autosomal dominant MEN I syndrome.

Physical Examination Findings

In uncomplicated peptic ulcer disease, examination results are generally normal. The presence of epigastric tenderness does not distinguish dyspepsia caused by peptic ulcer from other types of dyspepsia.

Patients who have complicated ulcers often have tachycardia and hypotension, which are exaggerated when the patient assumes an upright position. These findings may indicate a bleeding ulcer, a perforated ulcer with peritonitis, or an obstructing ulcer with protracted vomiting and volume depletion. Pulse and blood pressure measurements may give misleading information about the extent of volume contraction if the patient has preexisting hypertension, has cardiovascular disease, or is taking medication that can affect these parameters (e.g., a beta blocker or a calcium channel blocker). Fever and tachypnea suggest ulcer perforation with peritonitis.

Special attention should be given to the patient's mental status, skin and mucous membranes, heart and lungs, and, of course, abdomen and rectum. Involuntary guarding, rigidity, rebound tenderness, and a paucity or absence of bowel sounds suggests ulcer perforation with peritonitis. These findings may be less prominent or even absent in the very young, the elderly, and patients on corticosteroids or analgesics. Abdominal distention suggests gastric outlet obstruction or ileus. In a patient who has not eaten in the past 6 hours, a splashing sound over the stomach when the body is shaken (succussion) suggests gastric outlet obstruction or delayed gastric emptying caused by ileus. Melena or a positive fecal occult blood test suggests ulcer bleeding. Hematochezia or maroon-colored stool may be present if bleeding is voluminous and intestinal transit is rapid. Detection of melena, hematochezia, or maroon-colored stool should prompt placement of a nasogastric tube to obtain an aspirate of gastric contents. If this aspirate is grossly bloody, the diagnosis of upper GI bleeding is confirmed and the likelihood of a bleeding ulcer is increased.

Laboratory Studies

Laboratory results are normal in most patients with uncomplicated ulcer. A complete blood count should be done if blood loss or ulcer perforation is suspected; and serum electrolytes, blood urea nitrogen (BUN), and serum creatinine should be measured if the patient has poor oral intake, nausea, or vomiting. An elevated serum calcium level suggests the possibility of hyperparathyroidism and MEN I with Zollinger-Ellison syndrome, but the pretest probability of this condition is too low in patients presenting with ulcerlike symptoms to recommend routine measurement of serum calcium. If the patient has a strong family history of ulcer disease or of renal stones or has a personal history of renal stones, measurement of serum calcium is warranted, as is measurement of fasting serum gastrin once ulcer disease is confirmed and other etiologies of ulcer are excluded. If the calcium level is elevated, a serum parathyroid hormone measurement should be ordered.

Patients with complicated ulcers often have significant laboratory abnormalities, but these abnormalities are not specific for ulcer disease. Patients with bleeding ulcers have anemia and may have leukocytosis. The red cell indices (e.g., mean corpuscular volume) are typically normal. In the first several hours after an acute ulcer bleed, the hemoglobin concentration will not completely reflect the severity of the blood loss until compensatory hemodilution occurs or until intravenous fluids such as isotonic saline are administered. Thus, the pulse rate and blood pressure in the supine and upright positions are better initial indicators of extent of blood loss than are red cell counts. Patients with bleeding ulcers typically have azotemia, with ratios of BUN to serum creatinine concentrations (in mg/dl) exceeding 20:1, resulting from digestion and intestinal absorption of nitrogenous blood components in concert with reduced renal perfusion.

In patients with perforated ulcers and peritonitis, exudation of plasma into the peritoneal cavity (the so-called third space) may result in an increased hemoglobin concentration from hemoconcentration. The presence of leukocytosis, elevated band forms, or leukopenia should raise suspicion of intra-abdominal sepsis. Lactic acidosis with an increased anion gap may ensue as a consequence of a sepsis syndrome or hypovolemia.

Patients with gastric outlet obstruction typically exhibit a hypokalemic, hypochloremic metabolic alkalosis. If volume loss is extreme, a coexistent metabolic lactic acidosis with an increased anion gap may be present, which may cause an elevated serum bicarbonate level to drop toward normal or even below normal. Likewise, mild to moderate hyponatremia often develops in patients with vomiting from gastric outlet obstruction. Prerenal azotemia and a BUN-serum creatinine ratio greater than 20:1 are typical.

Imaging Studies

Although ulcer disease can be suggested by history, physical examination, and laboratory studies, none of these have sufficient specificity to confirm the diagnosis. Ulcers are diagnosed endoscopically, radiologically, or surgically. Once an ulcer is diagnosed, additional studies can help in determining the cause of the ulcer (e.g., H. pylori infection, NSAID use, gastrinoma, or cancer masquerading as benign ulcer).


Endoscopy is the most accurate way to diagnose a peptic ulcer [see Figure 3]. Most patients require local anesthesia of the pharynx and conscious sedation with an intravenous agent such as midazolam. The advantages of endoscopy are its nearly 100% specificity (rare false positives), greater than 90% sensitivity, portability (i.e., it can be performed in the intensive care unit, emergency department, or operating room), and ability to obtain tissue samples to help determine the etiology of the ulcer. The disadvantages of endoscopy are its cost and its potential for serious side effects. The most serious complications of endoscopy are respiratory depression and perforation of the GI tract. When a bleeding or obstructing ulcer is suspected, the stomach should be intubated and emptied with a large-bore tube before endoscopy to decrease the possibility of bronchopulmonary aspiration of gastric contents and to facilitate endoscopic visualization of mucosal lesions. Endoscopy is contraindicated in cases of suspected ulcer perforation.


Figure 3. Approach to a patient with new and undiagnosed ulcerlike symptoms refractory to a therapeutic trial of customary prescription doses of an H2 receptor blocker or with recurrent ulcerlike symptoms when the H2 receptor blocker is stopped.


Despite having a lower sensitivity and specificity than endoscopy, an upper GI series using barium and air (double contrast) may be favored by primary care physicians and patients over referral for endoscopy for suspected uncomplicated ulcer. An upper GI series offers lower cost, wider availability, and fewer complications [see Figure 3]. However, for troublesome and undiagnosed dyspepsia, an upper GI series may be superfluous, because a normal result will often necessitate endoscopy (endoscopy is more sensitive than radiography) and because an upper GI series showing a gastric ulcer will also necessitate endoscopy to biopsy the ulcer to exclude gastric malignancy. In many patients, only a finding of a duodenal bulbar ulcer on an upper GI series will preclude endoscopy [see Figures 4a and 4b].


Figure 4a. Upper GI series in which double contrast (barium and air) is used, showing rounded collection of barium in an ulcer (arrow) in the duodenal bulb of a patient presenting with dyspepsia (uncomplicated duodenal ulcer).


Figure 4b. Upright chest x-ray showing air beneath the right hemidiaphragm (pneumoperitoneum) of a patient presenting with an acute abdomen caused by a perforated duodenal ulcer.

Plain films of the abdomen, abdominal sonography, and computed tomographic scans may be helpful in patients presenting with suspected complicated ulcers, particularly perforated or obstructing ulcers. Upright chest x-rays of a patient with a perforated ulcer may show free intraperitoneal air [see Figure 4b], typically beneath the right hemidiaphragm. When plain films are negative or equivocal, pneumoperitoneum may be diagnosed by abdominal sonography or CT scan. Such studies should be performed only if the diagnosis of perforation is unclear; if physical signs of peritonitis are obvious, the patient should be referred to a surgeon. Patients with gastric outlet obstruction may have an enlarged stomach with old food debris visible on plain film of the abdomen, upper GI series, abdominal sonography, or CT scan.

Surgical Diagnosis

Certain patients will not have ulcers diagnosed until surgery is performed. Such patients include those presenting with an acute abdomen, in whom the diagnosis of perforated ulcer is made at exploratory laparotomy; those presenting with copious upper GI bleeding, in whom it is difficult for the endoscopist to visualize and treat the ulcer; and those with an obstructing ulcer who have a pinpoint pylorus or a duodenal stricture that prevents passage of the endoscope beyond the stenosis.

Tests to Establish the Etiology of the Ulcer

Endoscopic Tests

The endoscopist can biopsy the stomach of an ulcer patient to determine whether H. pylori organisms are present [see Figures 5a and 5b].H. pylori organisms contain abundant amounts of urease, which splits urea into carbon dioxide and ammonia. If the biopsy sample is placed on a urea-containing medium that also contains a pH-sensitive dye, a change in color indicates that ammonia is being produced. This so-called rapid urease test has a high sensitivity and specificity (> 90%) for H. pylori. If the rapid urease test is negative, a separate biopsy specimen should be sent to a pathology laboratory in formalin for histology. H. pylori can be detected with routine hematoxylin and eosin stains [see Figures 5a and 5b] or, if necessary, by special stains. Moreover, the presence of diffuse active chronic gastritis is highly suggestive of H. pylori infection, and its absence excludes H. pylori infection.


Figure 5a. Gastric biopsy samples stained with hematoxy lin and eosin demonstrating chronic active gastritis with a few H. pylori organisms faintly seen in the lumen of a gland.


Figure 5b. Chronic active gastritis with H. pylori organisms more abundant.

Another useful endoscopic procedure is to obtain multiple biopsies from the edges and the base of the ulcer to exclude malignancy. This is routinely done in cases of gastric ulcer because one in 25 to one in 50 benign-appearing gastric ulcers is in actuality an ulcer within a malignancy, usually in an adenocarcinoma. Duodenal ulcers need not be biopsied unless the ulcer is located in a mass distal to the duodenal bulb.

Endoscopy may also demonstrate a neuroendocrine tumor compatible with a gastrinoma on special stains. Such a tumor is usually located in the proximal duodenum.

In an ulcer patient with a negative rapid urease test and no H. pylori-related gastritis or gastric malignancy on histology, further history regarding NSAID use should be obtained from the patient or the patient's family. Many patients with NSAID-related ulcers have erosions, subepithelial hemorrhages, or both, which clue the endoscopist to the possibility of occult or surreptitious NSAID use, and these lesions may occur with or without gastric or duodenal ulcers.7

Serologic Tests

A number of serum antibody tests for H. pylori are available that have a greater than 90% sensitivity and specificity if the patient has not yet received therapy for H. pylori.17 In patients in whom active ulcers are diagnosed by radiology or surgery and in whom gastric tissue is not available, H. pylori serology can confirm infection with high accuracy [see Figure 3].

In ulcer patients with no evidence of H. pylori infection or NSAID use, the fasting serum gastrin concentration should be measured [seeFigure 3] to screen for gastrinoma (Zollinger-Ellison syndrome). If the serum gastrin concentration is greater than 1,000 pg/ml in a patient with duodenal ulcer, the diagnosis of gastrinoma is confirmed. A modest elevation in the fasting serum gastrin concentration (i.e., an elevation > 150 pg/ml but < 1,000 pg/ml) is suggestive of gastrinoma, but a provocative test should be performed using intravenous secretin (2 IU/kg as a bolus).18 A rise in serum gastrin concentration of more than 200 pg/ml after secretin administration has a greater than 90% sensitivity and specificity for gastrinoma. Because achlorhydria can produce marked hypergastrinemia as a result of the loss of negative feedback of gastric acid on gastrin release, basal acid output or pH should be measured to confirm that the stomach secretes acid in ulcer patients with fasting hypergastrinemia. The combination of achlorhydria, hypergastrinemia, and duodenal ulcer is exceedingly rare, whereas the combination of achlorhydria, hypergastrinemia, and gastric ulcer is sometimes encountered and should suggest gastric adenocarcinoma or NSAID use. A fasting gastric pH measurement will almost invariably distinguish gastrinoma (pH 1 to 2) from achlorhydria (pH 6 to 8), unless the patient has received a potent acid antisecretory agent before pH measurement.

Measurement of serum thromboxane B2 (platelet COX-1 activity) has been used in research laboratories to demonstrate occult or surreptitious NSAID use.19 However, this assay is not widely available.

Breath Tests

A noninvasive method for detecting H. pylori in the stomach, the urea breath test, begins with oral ingestion of urea that has been labeled with carbon-13 (13C) or carbon-14 (14C). If H. pylori, with its abundant urease, is present in the stomach, the labeled urea will be rapidly converted to 13CO2 or 14CO2, which can be detected in breath samples collected during the first 30 to 60 minutes after urea ingestion. Sensitivity and specificity of the breath test are comparable to those of serology.16 In a patient for whom there is no clinical indication for endoscopy, a urea breath test is an alternative to serology for documenting H. pylori infection [see Figure 3]. However, because proton pump inhibitors can suppress H. pylori without eradicating it, use of these drugs should be avoided for 2 weeks before the urea breath test is administered to minimize false negative results.

Because serology is quicker, it is preferred to breath testing for initial diagnosis. Breath testing is more useful than serology in diagnosing failure of eradication of H. pylori or reinfection in patients who were previously treated for H. pylori infection, because the serology will usually remain positive for several months even after successful treatment.20

Fecal Antigen Test

A stool test for H. pylori antigen compares favorably with urea breath tests.21 The test can also distinguish current infection (antigen present in stool) from past infection (antigen not present in stool).

Differential Diagnosis

The most common disorder confused with uncomplicated peptic ulcer is nonulcer dyspepsia; the most serious GI disorder confused with uncomplicated peptic ulcer is gastric cancer [see 12:VI Pancreatic, Gastric, and Other Gastrointestinal Cancers].

Complicated ulcers may be confused with a variety of disorders. These include both intra-abdominal and musculoskeletal processes [seeTable 1].

Table 1 Differential Diagnosis of Peptic Ulcer Disease



Suspected uncomplicated ulcer

Nonulcer dyspepsia, gastroesophageal reflux, biliary colic, pancreatic disease, angina pectoris, gastric cancer

Bleeding ulcer

Varices, Mallory-Weiss tear, esophagitis, vascular lesion (arteriovenous malformation, Dieulafoy lesion, angiodysplasia)

Perforated ulcer

Appendicitis, pancreatitis, cholecystitis, spontaneous bacterial peritonitis, bowel ischemia or infarction, diverticulitis

Penetrating ulcer

Pancreatitis, muscle strain, herniated vertebral disk, ureteral stone

Fistulizing ulcer

Gallstones, GI malignancy, Crohn disease, intra-abdominal abscess


The goals of ulcer therapy are rapid relief of symptoms; healing the ulcer; preventing ulcer recurrences; and reducing ulcer-related complications, morbidity (including the need for endoscopic therapy or surgery), and mortality. The general strategy in a patient with an ulcer should be to treat complications aggressively if present; to determine the etiology of the ulcer; to discontinue NSAID use if possible; to eradicate H. pylori infection if present or strongly suspected, even if other risk factors (e.g., NSAID use) are also present; and to use acid antisecretory therapy to heal the ulcer if H. pylori infection is not present. Smoking cessation should be encouraged. If duodenal ulcer is diagnosed by endoscopy, rapid urease testing of endoscopically obtained gastric biopsy samples, with or without histologic examination, should reliably establish the presence or absence of H. pylori. If duodenal ulcer is diagnosed by x-ray, then it is recommended that serologic testing, a urea breath test, or a fecal antigen test be performed to diagnose H. pylori infection before treating the patient for H. pylori.

Treatment of Uncomplicated Duodenal Ulcers

  1. pylori-Related Duodenal Ulcer

Duodenal ulcer associated with H. pylori infection should be treated with antimicrobial therapy because successful therapy is associated with markedly reduced ulcer recurrences [see Figure 6].3,4 Antimicrobial therapy is usually empirical rather than based on results of culture and in vitro antimicrobial sensitivity testing. No single antimicrobial agent has an acceptably high success rate against H. pylori. Combinations of antimicrobial agents are required, and some regimens that have been approved by the Food and Drug Administration can be recommended [see Table 2].

Table 2 Selected Clarithromycin-Based Regimens to Eradicate Helicobacter pylori

Esomeprazole, amoxicillin, clarithromycin (EAC)
    Esomeprazole, 40 mg b.i.d. for 10 days; then 40 mg q.d. for 18 days if an active ulcer is present
    Amoxicillin, 1 g b.i.d. for 10 days
    Clarithromycin, 500 mg b.i.d. or t.i.d. for 10 days
Lansoprazole, amoxicillin, clarithromycin (LAC)
    Lansoprazole, 30 mg b.i.d. for 10–14 days; then 15 mg q.d. for 14–18 days if an active ulcer is present
    Amoxicillin, 1 g b.i.d. for 10–14 days
    Clarithromycin, 500 mg b.i.d. for 10–14 days
Omeprazole, amoxicillin, clarithromycin (OAC)
    Omeprazole, 20 mg b.i.d. for 10 days; then 20 mg q.d. for 18 days if an active ulcer is present
    Amoxicillin, 1 g b.i.d. for 10–14 days
    Clarithromycin, 500 mg b.i.d. for 10 days
Rabeprazole, amoxicillin, clarithromycin (RAC)
    Rabeprazole, 20 mg b.i.d. for 7 days; then 20 mg q.d. for 21 days if an active ulcer is present
    Amoxicillin, 1 g b.i.d. for 7 days
    Clarithromycin, 500 mg b.i.d. for 7 days


Figure 6. Approach to treatment and follow-up of a patient with an uncomplicated duodenal or gastric ulcer associated withH. pylori infection of the stomach. Maintenance therapy is long-term nightly administration of an H2 receptor blocker or once-a-day proton pump inhibitor [see Table 3].

  1. pylorihas adapted to the acidic stomach, and potent acid antisecretory agents facilitate eradication of H. pyloriby antimicrobial agents. Bismuth compounds, like proton pump inhibitors, suppress the growth of H. pylori but usually do not by themselves eradicate it from the stomach. For this reason, they are frequently employed together with antibiotics.

Antimicrobial agents with activity against H. pylori include metronidazole, tetracycline, amoxicillin, clarithromycin, and fluoroquinolones such as levofloxacin. Most popular in the United States are 10- to 14-day combination regimens, although 7-day courses may be almost as effective22 and are more popular in Europe. A 2-week course of a three-drug regimen that includes a proton pump inhibitor, clarithromycin, and amoxicillin has a success rate approaching 90%. One recent study from Italy demonstrated close to 90% efficacy with a 7-day levofloxacin-based triple therapy; the efficacy rate exceeded those of standard (amoxicillin-based or metronidazole-based) 7-day triple therapies by 10% to 15%, suggesting that this regimen can be an alternative to standard 10-day and 14-day regimens for first-line therapy.23

The major causes of treatment failure are poor compliance with the regimen and clarithromycin resistance. Clarithromycin resistance occurs in 10% to 25% of current strains and is increasing with more macrolide use in the population. Point mutations in the peptidyltransferase region of the H. pylori 23S ribosomal RNA gene (particularly a A2143G mutation) have been implicated as having a major role in clarithromycin resistance. A 10-day sequential regimen (5 days of rabeprazole and amoxicillin, followed by 5 days of rabeprazole, clarithromycin, and tinidazole) was significantly more effective than standard triple therapy in eradicating clarithromycin-resistant strains with the A2143G mutation (eradication was seen in nine of 13 cases with the sequential regimen, versus two of 10 cases with standard triple therapy).24

Metronidazole resistance occurs in 30% to 40% of strains. However, unlike resistance to clarithromycin, which is usually absolute, resistance to metronidazole is relative and can be overcome in some patients.

If treatment fails in a patient who has carefully complied with one of the clarithromycin-based regimens, clarithromycin resistance is likely. In such cases, the re-treatment regimen should generally not include clarithromycin. Most physicians choose a regimen consisting of metronidazole, tetracycline, bismuth (e.g., Pepto-Bismol), and a proton pump inhibitor or H2 receptor blocker. Because of the frequency of metronidazole resistance, other antimicrobials with activity against H. pylori are being used more often.25 These agents include azithromycin, the quinolones norfloxacin and levofloxacin, and rifabutin [see Table 3]. In one study, a 7-day rescue treatment for persistentH. pylori infection using the proton pump inhibitor rabeprazole plus rifabutin and levofloxacin had a 95% success rate, as did a four-drug regimen consisting of rabeprazole, bismuth subcitrate, metronidazole, and tetracycline.26 Results were somewhat less impressive in another randomized controlled trial in patients in whom one or more therapies had previously failed; this study found that triple therapy with levofloxacin, amoxicillin, and pantoprazole given for 10 days eradicated H. pylori much more effectively than quadruple therapy consisting of omeprazole, tetracycline, metronidzaole, and bismuth subcitrate given for 7 days (70% success rate versus 37%).27 In a third study that evaluated patients with H. pylori that was resistant to both metronidazole and clarithromycin, either triple therapy with rifabutin, amoxicillin, and esomeprazole given for 7 days or high-dose dual therapy with omeprazole and amoxicillin given for 14 days eradicated H. pylori in approximately 75% of cases.28

Table 3 Additional Antimicrobial Agents with Activity against Helicobacter pylori





500 mg q.d.

Combined with either amoxicillin or metronidazole, plus a proton pump inhibitor, for 3–7 days


400 mg b.i.d.

Combined with a proton pump inhibitor for 14 days


500 mg q.d.

Combined with either amoxicillin or metronidazole, plus a proton pump inhibitor, for 7 days


300 mg q.d.

Combined with amoxicillin and a proton pump inhibitor for 7–10 days

*These agents are marketed in the United States, but they are not yet approved for H. pylori therapy.

It may also be possible to predict resistance to clarithromycin or to metronidazole by taking a careful history to look for prior exposure to these drugs. Such a history might help the physician choose a first-line regimen that is more likely to succeed.

Side effects of H. pylori-directed therapy are not uncommon but are generally mild. Antibiotic-associated colitis may occur, however.29Physicians should be aware of potential drug-drug interactions if the patient is receiving other medications. If the patient has an active, symptomatic ulcer, an antisecretory drug should be continued at a reduced (standard) dosage for 2 to 5 weeks after completion of the antimicrobial regimen.

After a patient has completed a course of ulcer therapy for an H. pylori-related uncomplicated duodenal ulcer, it is acceptable to follow the patient clinically without confirming eradication, because most compliant patients will be successfully cured of their H. pylori infection [seeFigure 6]. A patient with an H. pylori-related duodenal ulcer that does not recur symptomatically within 2 years after antimicrobial therapy is probably cured. Serology has often reverted to negative by this time.20

Patients in whom recurrent ulcer symptoms develop during the first 2 years after therapy should be assessed either by endoscopy (for ulcer recurrence and for H. pylori persistence or reinfection) or by a urea breath test or fecal antigen test. The most common cause of recurrent ulceration in patients treated for H. pylori-related duodenal ulcer is failure to eradicate the organism. Re-treatment of these patients is indicated [see Treatment of Intractable Duodenal Ulcers or Gastric Ulcers, below]. Rarer causes of duodenal ulcer recurrence include an acid hypersecretory state (e.g., Zollinger-Ellison syndrome), NSAID use, and reinfection with H. pylori.

  1. pylori-Negative Duodenal Ulcer

In a duodenal ulcer patient who is H. pylori negative, the physician should consider NSAID use and gastrinoma (Zollinger-Ellison syndrome).30 Patients with duodenal ulcer who are taking NSAIDs should discontinue the NSAID, if possible. At the same time, an acid antisecretory drug should be administered for 4 to 8 weeks [see Table 4]. The anticipated healing rate with this regimen is 85% to 95%.

Table 4 FDA-Approved Antisecretory Drugs for Active Peptic Ulcer Disease*




Drug InteractionsâĄ

H2receptor blockersâ 


800 mg h.s. or400 mg b.i.d.

Warfarin, theophylline, phenytoin, benzodiazepines, itraconazole, ketoconazole, atazanavir, cefpodoxime, cefditoren, gefitinib, memantine, metformin, and many others


300 mg h.s. or150 mg b.i.d.

Warfarin, atazanavir, itraconazole, ketoconazole, cefpodoxime, cefditoren, enoxacin, gefitinib, memantine, mefformin, tolazoline


300 mg h.s. or150 mg b.i.d.

Atazanavir, itraconazole, ketoconazole, cefpodoxime, cefditoren


40 mg h.s. or20 mg b.i.d.

Atazanavir, itraconazole, ketoconazole, cefpodoxime, cefditoren

Proton pump inhibitors§


20 mg q.d., a.c

Benzodiazepines, ampicillln, atazanavir, digoxin, iron, itraconazole, ketoconazole, voriconazole, methotrexate, tacrolimus


15 mg q.d., a.c.

Ampicillin, atazanavir, digoxin, iron, itraconazole, ketoconazole


40 mg q.d., a.c.

Atazanavir, digoxin, iron, ketoconazole


20 mg q.d., a.c.

Warfarin, ampicillin, atazanavir, digoxin, iron, itraconazole, ketoconazole


40 mg q.d., a.c

Ampicillin, atazanavir, iron, itraconazole, ketoconazole

* Patients with gastrinoma (Zollinger-Ellison syndrome) will usually require much higher dosages of antisecretory drugs than listed here.
† Use for 4–8 wk in the treatment of duodenal ulcer and 6–12 wk in the treatment of gastric ulcer. Duodenal ulcers that do not heal by 8 wk and gastric ulcers that do not heal by 12 wk are considered intractable. Dosage of H2 receptor blockers should be reduced in patients with renal failure.
‥ Micromedex Health Care Services. Most of these drug interactions are minor and not clinically relevant; nevertheless, caution is advised.
§Omeprazole and lansoprazole are approved for gastric ulcers; omeprazole, lansoprazole, and rabeprazole are approved for duodenal ulcers.

Patients with duodenal ulcer as part of the Zollinger-Ellison syndrome should be managed initially with a high dose of a proton pump inhibitor, followed by a maintenance dose guided by gastric acid measurements. If there is no evidence of hepatic metastasis on abdominal CT scan, then exploratory laparotomy for gastrinoma resection, with or without parietal cell vagotomy, should be considered. Radionuclide scintigraphy with octreotide, an analogue of somatostatin, is a highly sensitive and specific preoperative test for detecting and staging gastrinoma, as is endoscopic ultrasonography.

The vast majority of duodenal ulcers, regardless of cause, heal after 8 weeks of antisecretory therapy with a proton pump inhibitor or an H2receptor blocker. Antacids are often prescribed as needed to relieve ulcer symptoms.

In rare cases of idiopathic duodenal ulcer, it is prudent that, after the ulcer has been healed by an acid antisecretory agent, the patient be placed on a maintenance regimen of an H2 receptor blocker given at bedtime to reduce ulcer recurrences. Proton pump inhibitors are also effective in preventing duodenal ulcer recurrences. It is not necessary to confirm duodenal ulcer healing by endoscopy or x-ray before reducing the antisecretory drug dose to a maintenance level.

Treatment of Uncomplicated Gastric Ulcers

  1. pylori-Related Gastric Ulcer

Gastric ulcer associated with H. pylori should be treated with antibiotics [see Table 2]. Because they are larger than duodenal ulcers, gastric ulcers take longer to heal. Thus, after antibiotic administration, the patient should be treated with an acid antisecretory agent [see Tables 3and 4] for an additional 4 to 8 weeks. Patients with gastric ulcers should be followed endoscopically until complete healing has been achieved so that an ulcerated gastric cancer is not missed. Gastric biopsies should be performed during follow-up endoscopy to determine whether eradication of H. pylori has occurred [see Figure 6]. Patients with a history of an uncomplicated gastric ulcer that is currently quiescent should be screened for H. pylori infection; if the result is positive, they should be treated for H. pylori infection to prevent ulcer recurrences.

NSAID-Related Gastric Ulcer

The therapy for an active NSAID-related gastric ulcer is administration of a proton pump inhibitor [see Tables 4 and 5], as well as discontinuance of the NSAID. Healing rates with H2 receptor blockers are nearly as high as with proton pump inhibitors if the NSAID can be stopped.

Table 5 Treatment and Prevention of Peptic Ulcers

Type of Ulcer




Helicobacter pylori–related ulcers

Antibiotics [seeTables 2 and 3] Âą antisecretory agents [see Table 4]

None needed if H. pylorieradicated

Highly cost-effective; document healing in gastric ulcer; document H. pylori eradication in complicated duodenal or gastric ulcer and in intractable duodenal or gastric ulcer

NSAID-related ulcers

Antisecretory agents (proton pump inhibitors have greater efficacy than H2receptor blockers) [seeTable 4]
Discontinue NSAID use, if possible

Misoprostol (600–800¾g/day) or proton pump inhibitor (e.g., omeprazole, 20–40 mg q.d., or lansoprazole, 15–0 mg q.d.) along with an NSAID

Diarrhea may limit compliance in patients treated with misoprostol; avoid misoprostol during pregnancy (abortifacient); proton pump inhibitors are not yet approved by the FDA for prevention of NSAID-related ulcers

Ulcers associated with Zollinger-Ellison syndrome and other hypersecretory states

High-dose proton pump inhibitor

Proton pump inhibitor, adjusted to keep basal acid output < 5–10 mEq/hr

Consider exploratory laparotomy (guided by abdominal imaging studies) to remove easily resectable gastrinomas, if feasible; consider MEN I syndrome (present in 20%–30% of cases)21

Idiopathic ulcers

H2 receptor blocker or proton pump inhibitor

Nocturnal H2 receptor blocker or A.M. proton pump inhibitor

Parietal cell vagotomy for intractable duodenal ulcer and antrectomy for intractable gastric ulcer

Stress ulcers (ICU)

I.V. H2 receptor blocker or proton pump inhibitor (e.g., pantoprazole)

I.V. H2 receptor blocker or proton pump inhibitor; intragastric sucralfate; or intragastric antacid

Maintain pH above 4 with H2 receptor blocker, proton pump inhibitor or antacid; continuous I.V. infusion is superior to I.V. boluses of antisecretory drug

Cameron ulcers (linear gastric erosions in a hiatal hernia)

Iron salts; packed red cell transfusions; endoscopic hemostasis

Hiatal hernia repair, laparoscopic or open

Roles of H2 receptor blockers and proton pump inhibitors are unproved

MEN I—multiple endcocrine neoptasia type I  NSAID—nonsteroidal anti-inflammatory drug

Treatment of Intractable Duodenal Ulcers or Gastric Ulcers

Ulcers That Fail to Heal

Ulcers refractory to pharmacotherapy are rare, and in most cases, prolonging the course of the gastric antisecretory drug, increasing the dose, or taking both measures will lead to healing. The causes of nonhealing include poor compliance with medications, an acid hypersecretory state requiring higher-than-customary doses of antisecretory drugs, continued NSAID use, and persistent H. pylori infection. Often, combinations of these factors and others (e.g., smoking and stress) are present.

Poor compliance with medications necessitates patient education and consideration of elective ulcer surgery. A fasting serum gastrin concentration can be used to screen for an acid hypersecretory state resulting from Zollinger-Ellison syndrome. Physicians should be aware that antisecretory drugs (especially proton pump inhibitors) can also raise serum gastrin levels modestly (to 150 to 600 pg/ml). Definitive documentation of an acid hypersecretory state requires quantitative gastric acid measurement (gastric analysis). NSAID use should be discontinued if at all possible. Persistent H. pylori infection is the result of poor compliance with medications or is caused by drug-resistant strains.25,26,27,28,29 H. pylori has proved to be resistant to metronidazole in 30% to 40% of cases and to clarithromycin in 10% of cases; resistance to tetracycline or amoxicillin occurs in 1% of strains or less.25 Combined resistance to macrolides (e.g., clarithromycin) and imidazoles (e.g., metronidazole) occurs in approximately 5% of patients, in whom infection may prove difficult to eradicate. Culture of gastric biopsy material for H. pylori, followed by antimicrobial drug-susceptibility testing when available, can guide re-treatment. In the absence of this information, the patient should be re-treated for 2 weeks with a proton pump inhibitor, with amoxicillin or tetracycline, and with either clarithromycin or metronidazole (whichever antimicrobial agent the patient did not receive initially). Some physicians use a bismuth preparation (e.g., colloidal bismuth subcitrate, bismuth subsalicylate, or ranitidine bismuth citrate) in place of a proton pump inhibitor [see Table 2]. Several other antibiotics are active against H. pylori and may prove to be useful in rescue therapy for patients in whom treatments have failed; many such patients harbor antibiotic-resistant strains. Agents that are available in the United States include the macrolide azithromycin, the quinolones norfloxacin and levofloxacin, and rifabutin [see Table 3].

Frequently Recurring Ulcers

Another type of intractability is frequent ulcer recurrences (at least three a year). This type of intractability occurs most often when H. pylori has not been successfully eradicated (necessitating re-treatment) or, less often, when NSAID use is resumed or when an acid hypersecretory state is present. Idiopathic ulcers sometimes recur frequently, and patients experiencing such recurrences require lifelong maintenance with H2 receptor blockers or proton pump inhibitors. Some of these patients may choose ulcer surgery (parietal cell vagotomy for duodenal ulcer or antrectomy for gastric ulcer) over lifelong medication.

Treatment of Complicated Peptic Ulcers

Bleeding Ulcers

The first priority in a patient with a suspected bleeding peptic ulcer is to stabilize the vital signs with volume resuscitation, ideally in an intensive care unit. Such an approach is associated with improved patient outcomes.31 Hemodynamic monitoring may assist in fluid and blood replacement, particularly if the patient has significant (New York Heart Association class III or IV) cardiac disease.

After the patient becomes hemodynamically stable, diagnostic upper GI endoscopy is performed [see Figure 7]. If an actively bleeding ulcer or an ulcer with a visible vessel is found, the lesion is treated endoscopically, usually by injection of epinephrine, by thermal application with a heater probe or a bipolar electrode, or by a combination of these methods. A meta-analysis of 16 studies demonstrated that combination therapy was superior to epinephrine injection alone in reducing bleeding, emergency surgery, and mortality.32 Endoscopic therapy is successful in controlling bleeding in approximately 90% of patients; the other 10% are referred for surgery if major bleeding continues.


Figure 7. Approach to treatment and follow-up of a patient with a complicated duodenal or gastric ulcer associated with H. pylori infection of the stomach.

Random gastric biopsies are performed at the time of endoscopy to detect H. pylori by rapid urease testing and, if necessary, gastric histology.

Once an ulcer is demonstrated, intravenous gastric antisecretory therapy, usually with a proton pump inhibitor (e.g., pantoprazole) can be started. Oral therapy with a proton pump inhibitor is superior to no therapy in reducing early rebleeding if endoscopic therapy is not attempted33; proton pump inhibitors also reduce ulcer rebleeding after endoscopic therapy.34 The combination of endoscopic therapy and intravenous proton pump inhibitor is superior to the proton pump inhibitor alone in patients with visible vessels or with clots adherent to the ulcer.35 However, endoscopic therapy may fail or result in a perforation, necessitating surgery.35 As soon as oral intake is resumed, a patient who has been on intravenous therapy should be switched to oral therapy with a proton pump inhibitor. If the rapid urease test or gastric histology is positive for H. pylori, the patient should receive at least two effective antibiotics (e.g., clarithromycin and amoxicillin) for 14 days, along with a proton pump inhibitor [see Figure 7]. The proton pump inhibitor is continued at the same dosage until week 6 for duodenal ulcer or week 8 for gastric ulcer. If subsequent endoscopy shows complete ulcer healing and disappearance of H. pylori by both rapid urease testing and gastric histology, the risk of rebleeding is low34, 36 and therapy can be stopped. Future use of NSAIDs or aspirin is almost always prohibited. Although the use of celecoxib instead of an NSAID is attractive, freedom from rebleeding is not guaranteed. If an NSAID or low-dose aspirin is absolutely necessary, it should be coprescribed with the prostaglandin E1 analogue misoprostol37 or a proton pump inhibitor.38 If the ulcer heals but H. pylori organisms are still present, the patient should be treated again for H. pylori or left on a maintenance H2 receptor blocker or proton pump inhibitor. If the ulcer is not healed, persistent H. pylori infection is likely. Under these circumstances, gastric tissue can be cultured for H. pylori, if available facilities exist, so that antibiotic sensitivities can be determined before re-treatment. Finally, if the ulcer has not healed even though H. pylori has been eradicated, then NSAID use, an acid hypersecretory state, or cancer should be considered. Biopsy samples of the ulcer should be obtained, especially if the ulcer is in the stomach.

A patient with a bleeding peptic ulcer that is negative for H. pylori by rapid urease testing and gastric histology usually has an NSAID-related ulcer. The NSAID is stopped if possible, and a high-dose proton pump inhibitor (e.g., 40 mg of omeprazole or 30 mg of lansoprazole) is prescribed for 8 weeks for gastric ulcer or 4 to 6 weeks for duodenal ulcer. Repeat endoscopy is usually indicated to assess healing of a gastric ulcer; if the ulcer has not healed, biopsy is performed. Whether to perform a repeat endoscopy to assess healing 4 to 8 weeks after an NSAID-related bleeding duodenal ulcer is controversial. If the ulcer is shown to have healed after the patient is no longer taking the NSAID, no further therapy is required unless the patient resumes taking an NSAID or aspirin, even a low dose of aspirin. NSAIDs should be avoided in patients with ulcers that have bled. When NSAID resumption is unavoidable, misoprostol (200 Âľg q.i.d.) is modestly protective against subsequent bleeding.37 There is also evidence that maintenance therapy with a proton pump inhibitor such as lansoprazole or omeprazole is associated with a low rate of NSAID-related ulcer rebleeding. One study found that, in high-risk arthritis patients with prior ulcer bleeding, celecoxib was as effective as the combination of an NSAID, diclofenac, and the proton pump inhibitor omeprazole.39 With either therapy, however, bleeding resumed in 5% to 6% of such patients over the ensuing 6 months. In patients with cardiovascular disease and an aspirin-induced bleeding ulcer, once ulcer healing had occurred, recurrent ulcer bleeding was reduced more effectively by using aspirin and esomeprazole concomitantly than by using clopidogrel alone. Ulcer rebleeding over the subsequent year occurred in only 1% of patients treated with aspirin and esomeprazole, compared with a 9% rate of ulcer rebleeding in patients switched to clopidogrel once the initial bleeding ulcer had healed.40

Patients with idiopathic bleeding ulcers that heal with the use of an antisecretory drug should continue long-term therapy with an H2receptor blocker41 or a proton pump inhibitor.

Acute Stress Ulcers

Therapy for bleeding acute stress ulcers and erosions involves blood transfusion if necessary and attempts to treat the underlying disease state. The role of intravenous H2 receptor blockers or proton pump inhibitors is unproved. Endoscopic therapy is not usually curative, because multiple bleeding lesions are often present. In rare cases, visceral angiography with embolization of the major bleeding site is attempted. Gastrectomy for continuous bleeding or significant rebleeding is used as a last resort and is associated with a very high mortality.

Because of the dire consequences of stress ulcers and the lack of an effective therapy, high-risk patients in intensive care units should be placed on stress ulcer prophylaxis.42 The patients at highest risk for bleeding are those with multiorgan failure and those who are receiving ventilatory assistance for more than 24 hours. The incidence of significant bleeding in high-risk patients is reduced from a range of about 10% to 25% to a range of about 1% to 5% with the use of prophylactic intragastric or oral antacids or sucralfate or with the use of intravenous H2 receptor blockers or proton pump inhibitors given by continuous infusion. Intravenous antisecretory therapy has the advantages of ease of administration, the ability to monitor gastric pH to assess effectiveness (the goal is a pH > 4), and proven efficacy in clinical trials.

Perforated Ulcers

When a perforated viscus is documented or strongly suspected, the patient is started on broad-spectrum intravenous antibiotics covering gram-negative aerobic bacilli, enterococci, and anaerobes such as Bacteroides species and is then taken to surgery for closure of the perforation with a patch of omentum. If the surgeon does not obtain an intraoperative gastric biopsy sample, the patient should undergo postoperative testing for H. pylori by serology, urea breath test, or fecal antigen test; if infection is present, it should be treated.43 Many perforated ulcers are associated with NSAID use rather than with H. pylori infection.44 Regardless of the patient's H. pylori status, antisecretory drugs should be administered for 6 to 8 weeks postoperatively. Endoscopy is then performed to assess healing. At the time of this endoscopy, success of eradication of H. pylori is ascertained by gastric biopsy with rapid urease testing and histology.

Up to 10% of perforated gastric ulcers are in fact perforated gastric cancers. If no biopsy or resection of the ulcer is done at the time of repair of the perforation, postoperative endoscopy with biopsy is imperative before the patient is discharged from the hospital or soon thereafter.

The mortality in patients with perforated peptic ulcers is 5% to 10%. Factors associated with higher mortality include delayed diagnosis and treatment of perforation; advanced age; comorbid conditions, such as cardiac, pulmonary, or liver disease; immunodeficiency; and advanced malignancy.

A small number of patients with suspected or probable perforated peptic ulcer improve rapidly before surgery is performed. Others refuse surgery or are poor surgical candidates. An alternative to surgery in these situations is nonoperative therapy consisting of nothing by mouth; intravenous fluids and electrolytes; a broad-spectrum antibiotic such as ticarcillin-clavulante or pipericillin-tazobactam; and nasogastric suction. Compared with surgical therapy, nonoperative therapy is associated with more abdominal complications (e.g., abscesses), fewer pulmonary complications (e.g., atelectasis), and similar mortality.45

Obstructing Ulcers

The patient with an obstructing ulcer is initially placed on nasogastric suction, intravenous fluids and electrolytes, and an intravenously administered proton pump inhibitor. If the obstruction is the result of edema associated with an active ulcer, the gastric outlet may open as edema subsides and the ulcer heals over several days to weeks. If, on the other hand, obstruction is the result of scarring from previous ulcers, it will not resolve with these measures. In some patients, it is difficult to determine whether edema or fibrosis is the primary cause of gastric outlet obstruction. Because obstruction may resolve with time, early consideration should be given to parenteral hyperalimentation. This intervention prevents or minimizes tissue catabolism during the waiting period and also induces a positive nitrogen balance, which will be beneficial if the gastric outlet fails to open up and the patient requires surgery.

A saline load test can be used to guide management. Thirty minutes after 750 ml of isotonic saline is infused into the stomach through the nasogastric tube, gastric contents (saline plus secretions) are aspirated. A return of less than 200 ml indicates normal gastric emptying of liquids and is associated with a good prognosis; 200 to 400 ml is indeterminate; and more than 400 ml is suggestive of a high-grade obstruction that will likely require intervention. Repeating the saline load test every day or two may also provide information about whether the obstruction is resolving.

If the obstruction resolves within 3 to 7 days, the nasogastric tube is removed and the patient is fed and observed clinically. An oral proton pump inhibitor or H2 receptor blocker is started as soon as feasible. Gastric prokinetic agents (e.g., metoclopramide) should not be used. At least 50% of patients whose obstruction resolves with conservative medical therapy experience another obstruction within about a year. Whether routine treatment of H. pylori infection will reduce this high recurrence rate is unknown. Unlike bleeding or perforation, obstruction is usually a late complication of ulcer disease. Thus, H. pylori eradication in ulcer patients is more likely to be effective in primary prevention of obstruction than in secondary prevention. NSAIDs may cause gastric outlet obstruction as well and should therefore be avoided. Endoscopic therapy is an option for obstruction that does not resolve with conservative therapy.46 Using inflatable balloons placed over guide wires, the endoscopist can dilate a stenotic pylorus or duodenum under fluoroscopic guidance, although complications such as perforation may occur. Endoscopic balloon dilatation, when feasible, is a temporizing measure and rarely obviates surgery. Thus, obstruction that recurs after medical therapy or after endoscopic therapy is an indication for surgery. Pyloroplasty, gastroenterostomy, and resection plus gastroenterostomy are the most popular operations for an obstructing ulcer. Pyloroplasty and gastroenterostomy are typically combined with a vagotomy to reduce the likelihood of recurrent ulceration.

Fistulizing Ulcers

For gastric or duodenal ulcers associated with fistulas, biopsies must be performed to exclude malignancy. Initially, benign ulcers are treated in the manner of an uncomplicated ulcer [see Figure 6]. An antisecretory agent—ideally, a proton pump inhibitor—should be prescribed, along with antibiotics if H. pylori organisms are present. The ulcer may heal with closure of the fistula. If the fistula persists, surgical resection of the fistula is warranted only if significant symptoms are present (e.g., troublesome diarrhea in a patient with a gastrocolonic fistula or cholangitis in a patient with a duodenocholedochal fistula).

Cameron Ulcers

Although acid secretory inhibitors are often used in the treatment of linear erosions in hiatal hernias (Cameron ulcers), their value is uncertain.16 Standard therapy consists of packed red cell transfusions for acute bleeding, oral iron replacement for chronic bleeding, and laparoscopic or open repair of the hiatal hernia when medical therapy fails. Many of these patients are elderly or are at high risk for surgical complications; for such patients, surgery should be undertaken only when medical therapy fails or becomes cumbersome.


Figures 1, 3 Marcia Kammerer.

Figure 5 Courtesy of Edward Lee, M.D.


  1. Helicobacter pyloriin peptic ulcer disease. NIH Consensus Development Panel on Helicobacter pylori in Peptic Ulcer Disease. JAMA 272:65, 1994
  2. Cryer B, Faust TW, Goldschmiedt M et al: Gastric and duodenal mucosal prostaglandin concentrations in gastric or duodenal ulcer disease: relationships with demographics, environmental, and histological factors, including Helicobacter pylori. Am J Gastroenterol 87:1747, 1992
  3. Forbes GM, Glaser ME, Cullen DJ et al: Duodenal ulcer treated with Helicobacter pylorieradication: seven-year follow-up. Lancet 343:258, 1994
  4. Sung JJ, Chung SC, Ling TK et al: Antibacterial treatment of gastric ulcers associated with Helicobacter pylori. N Engl J Med 332:139, 1995
  5. Spechler SJ, Fischbach L, Feldman M: Clinical aspects of genetic variability in Helicobacter pylori. JAMA 283:1264, 2000
  6. Lu H, PI Hsu, Graham D et al: Duodenal ulcer promoting gene of Helicobacter pylori. Gastroenterology 128:833, 2005
  7. Papatheodoridis GV, Papadelli D, Cholongitas E et al: Effect of Helicobacter pyloriinfection on the risk of upper gastrointestinal bleeding in users of nonsteroidal anti-inflammatory drugs. Am J Med 116:601, 2004
  8. Chan FK, To KF, Wu JC et al: Eradication of Helicobacter pyloriand risk of peptic ulcers in patients starting long-term treatment with non-steroidal anti-inflammatory drugs: a randomised trial. Lancet 359:9, 2002
  9. Henry D, Lim LL, Rodriguez LA et al: Variability in risk of gastrointestinal complications with individual non-steroidal anti-inflammatory drugs: results of a collaborative meta-analysis. BMJ 312:1563, 1996
  10. Cryer B, Feldman M: Effects of very low dose daily, long-term aspirin therapy on gastric, duodenal, and rectal prostaglandin levels and on mucosal injury in healthy humans. Gastroenterology 117:17, 1999
  11. Kelly JP, Kaufman DW, Jurgelson JM et al: Risk of aspirin-associated major upper-gastrointestinal bleeding with enteric-coated or buffered product. Lancet 348:1413, 1996
  12. Cryer B, Feldman M: Cyclooxygenase-1 and cyclooxygenase-2 selectivity of widely used NSAIDs and other anti-inflammatory or analgesic drugs: studies in whole blood and gastric mucosa of healthy humans. Am J Med 104:413, 1998
  13. Solomon SD, Pfeffer MA, McMurray JJ et al: Effect of celecoxib on cardiovascular events and blood pressure in two trials for the prevention of colorectal polyps. Circulation 114:1028, 2006
  14. Aoyama N, Kinoshita Y, Fujimoto S et al: Peptic ulcers after the Hanshin-Awaji earthquake: increased incidence of bleeding gastric ulcer. Am J Gastroenterol 93:311, 1998
  15. Hung LC, Ching JY, Sung JJ et al: Long-term outcome of Helicobacter pylori—negative idiopathic bleeding ulcers: a prospective cohort study. Gastroenterology 128:1845, 2005
  16. Cryer B, Spechler SJ: Peptic ulcer disease. Gastrointestinal and Liver Disease, 8th ed. Feldman M, Friedman LSF, Brandt LJ, Eds. Saunders Elsevier, Philadelphia, 2006, p 1089
  17. Cutler AF, Havstad S, Ma CK et al: Accuracy of invasive and noninvasive tests to diagnose Helicobacter pyloriinfection. Gastroenterology 109:136, 1995
  18. Metz DC, Buchanan M, Purich E et al: A randomized controlled crossover study comparing synthetic porcine and human secretins with biologically derived porcine secretin to diagnose Zollinger-Ellison syndrome. Aliment Pharm Ther 15:669, 2001
  19. Lanas A, Sekar MC, Hirschowitz BI: Objective evidence of aspirin use in both ulcer and nonulcer upper and lower gastrointestinal bleeding. Gastroenterology 103:862, 1992
  20. Feldman M, Cryer B, Lee E et al: Role of seroconversion in confirming cure of Helicobacter pyloriinfection. JAMA 280:363, 1998
  21. Vaira D, Nimish V, Menegatti M et al: The stool test for detection of Helicobacter pyloriafter eradication therapy. Ann Intern Med 136:280, 2002
  22. Calvet X, Ducons J, Bujanda L et al: Seven versus ten days of rabeprazole triple therapy for Helicobacter pylorieradication: a multicenter randomized trial. Am J Gastroenterol 100:1696, 2005
  23. Nista EC, Candelli M, Zocco MA et al: Levofloxacin-based triple therapy in first-line treatment for Helicobacter pylorieradication. Am J Gastroenterol 101:1985, 2006
  24. De Francesco V, Margiotta M, Zullo A et al: Clarithromycin-resistant genotypes and eradication of Helicobacter pylori. Ann Intern Med 144:94, 2006
  25. Gisbert JP, Pajares JM: Helicobacter pylori“rescue” regimen when proton pump inhibitor—based therapies fail. Aliment Pharmacol Ther 16:1047, 2002
  26. Wong WM, Gu Q, Lam SK et al: Randomized controlled study of rabeprazole, levofloxacin and rifabutin triple therapy vs. quadruple therapy as a second-line treatment for Helicobacter pyloriinfection. Aliment Pharm Ther 17:553, 2003
  27. Bilardi C, Dulbecco P, Zentilin P et al: A 10-day levofloxacin-based therapy in patients with resistant Helicobacter pyloriinfection: a controlled trial. Clin Gastroenterol Hepatol 2:997, 2004
  28. Miehlke S, Hansky K, Schneider-Brachert W et al: Randomized trial of rifabutin-based triple therapy and high-dose dual therapy for salvage treatment of Helicobacter pyloriresistant to both metronidazole and clarithromycin. Aliment Pharmacol Ther 24:395, 2006
  29. Hogenauer C, Langner C, Beubler E et al: Klebsiella oxytoca as a causative organism of antibiotic-associated hemorrhagic colitis. N Engl J Med 355: 2418, 2006
  30. McColl KE, El-Nujumi AM, Chittajallu RS et al: A study of pathogenesis of Helicobacter pylorinegative chronic duodenal ulceration. Gut 34:762, 1993
  31. Baradarian R, Ramdhaney S, Chapalamadugu R et al: Early intensive resuscitation of patients with upper gastrointestinal bleeding decreases mortality. Am J Gastroenterol 99:619, 2004
  32. Calvet X, Vergara M, Brullet E et al: Addition of a second endoscopic treatment following epinephrine injection improves outcome in high-risk bleeding ulcers. Gastroenterology 126:441, 2004
  33. Khuroo M, Yattoo GN, Javid G et al: A comparison of omeprazole and placebo for bleeding peptic ulcer. N Engl J Med 336:1054, 1997
  34. Lau JW, Sung JJ, Lee KK et al: Effect of intravenous omeprazole on recurrent bleeding after endoscopic treatment of bleeding peptic ulcers. N Engl J Med 343:310, 2000
  35. Sung JJ, Chan FK, Lau JY et al: The effect of endoscopic therapy in patients receiving omeprazole for bleeding ulcers with nonbleeding visible vessels or adherent clots. Ann Intern Med 139:237, 2003
  36. Rokkas T, Karameris A, Mavrogeorgis A et al: Eradication of Helicobacter pylorireduces the possibility of rebleeding in peptic ulcer disease. Gastrointest Endosc 41:1, 1995
  37. Silverstein FE, Graham DY, Senior JR et al: Misoprostol reduces serious gastrointestinal complications in patients with rheumatoid arthritis receiving nonsteroidal anti-inflammatory drugs: a randomized, double-blind, placebo-controlled trial. Ann Intern Med 123:241, 1995
  38. Lai K, Lam SK, Chu KM et al: Lansoprazole for the prevention of recurrences of ulcer complications from long-term low-dose aspirin use. N Engl J Med 346:2033, 2002
  39. 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 347:2104, 2002
  40. Chan FK, Ching JY, Hung LC et al: Clopidogrel versus aspirin and esomeprazole to prevent recurrent ulcer bleeding. N Engl J Med 352:238, 2005
  41. Jensen DM, Cheng S, Kovacs TO et al: A controlled study of ranitidine for the prevention of recurrent hemorrhage from duodenal ulcer. N Engl J Med 330:382, 1994
  42. Cook D, Guyatt G, Marshall J et al: A comparison of sucralfate and ranitidine for prevention of upper gastrointestinal bleeding in patients requiring mechanical ventilation. N Engl J Med 338:791, 1998
  43. Ng EK, Lam YH, Sung JJ et al: Eradication of Helicobacter pyloriprevents recurrence of ulcer after simple closure of duodenal ulcer perforation: randomized controlled trial. Ann Surg 231:153, 2000
  44. Reinbach DH, Cruickshank G, McColl KE: Acute perforated duodenal ulcer is not associated with Helicobacter pyloriinfection. Gut 34:1344, 1993
  45. Crofts TJ, Park KG, Steele RJ et al: A randomized trial of nonoperative treatment for perforated peptic ulcer. N Engl J Med 320:970, 1989
  46. DiSario JA, Fennerty MB, Tietze CC et al: Endoscopic balloon dilation for ulcer-induced gastric outlet obstruction. Am J Gastroenterol 89:868, 1994

Editors: Dale, David C.; Federman, Daniel D.