Peptic ulcer disease (PUD) results from an imbalance between mucosal defense and aggressive factors such as gastric acid, pepsin, and exogenous injury.^[1]

The principal causes are:

• Helicobacter pylori infection, about 42% of cases.
• Aspirin or NSAID use, about 36% of cases.
• Idiopathic ulcers, unrelated to either, about 22% of cases.^[2]

H. pylori and NSAIDs independently increase ulcer risk and act synergistically. PUD is rare when neither is present.^[3]

Less common causes include:

• Stress-related mucosal disease in critical illness.^[4]
• Zollinger-Ellison syndrome.
• Malignancy.^[5]
• Rare inflammatory, infectious, and anatomic causes.^[6]

Helicobacter pylori is the most common identifiable cause of PUD.^[2]

Global burden. About 57% (95% CI 44–68%) of PUD worldwide is attributable to H. pylori. This corresponds to about 3.5 million cases that could potentially be prevented by eradication.^[12]

Infection is necessary but not sufficient. An infected person has an estimated 10–20% lifetime risk of PUD, at least 3–4 times higher than an uninfected person.^[8] In a Danish population-based cohort, H. pylori infection was a major independent risk factor for PUD (OR 4.3, 95% CI 2.2–8.3).^[13]

Strain virulence. Infection with cagA-positive strains raises ulcer risk above that of cagA-negative strains. In a nested case-control study, people seropositive for both H. pylori and CagA had odds ratios of:

• 5.8 (95% CI 1.1–30.0) for duodenal ulcer.
• 4.4 (95% CI 1.8–10.5) for gastric ulcer.^[14]

The gastritis pattern determines ulcer location.

• Antral-predominant gastritis suppresses somatostatin and increases gastrin release. This causes acid hypersecretion and predisposes to duodenal ulcer.^[15]
• Corpus-predominant gastritis or pangastritis leads to gastric atrophy and low acid output. It predisposes to gastric ulcer and gastric cancer.^[15]
• In a prospective cohort of 1,526 Japanese patients, gastric cancer developed in patients with gastric ulcer but in none of those with duodenal ulcer.^[16]

NSAIDs, including low-dose aspirin, are an etiologic factor in about 36% of PUD.^[2]

Mechanisms of injury. NSAIDs injure the mucosa in two ways:^[9][17]

• Local (topical) toxicity to the epithelium.
• Systemic inhibition of cyclooxygenase-1, which depletes gastroprotective prostaglandins.

Interaction with H. pylori. A meta-analysis of 16 studies of 1,625 NSAID users found:^[3]

• Uncomplicated PUD was more common in H. pylori-positive than H. pylori-negative users (41.7% vs 25.9%; OR 2.12, 95% CI 1.68–2.67).
• Compared with H. pylori-negative non-users, H. pylori-infected NSAID users had an ulcer OR of 61.1.
• H. pylori infection increased the risk of ulcer bleeding 1.79-fold, and NSAID use increased it 4.85-fold.

Clinical presentation. Most NSAID-induced ulcers are asymptomatic, so the first presentation may be a complication.^[9]

Idiopathic ulcers are H. pylori-negative and NSAID-negative.

Frequency. They account for about 22% of PUD.^[2] Their share has risen as H. pylori prevalence has fallen:

• In Hong Kong, idiopathic ulcers rose from 4.2% to 18.8% of bleeding ulcers between 1997–1998 and 2000.^[10]
• In Asia, the proportion reached 10–30% in studies from the 2000s.^[7]

Distinguishing true from pseudo-idiopathic ulcers. Apparent idiopathic ulcers must be separated from:

• False-negative H. pylori testing. Using two or more diagnostic methods increases sensitivity. Serology is useful in patients taking proton pump inhibitors.^[5]
• Surreptitious NSAID or aspirin use. Platelet cyclooxygenase activity testing can confirm it.^[5]

Proposed mechanisms. The pathogenesis of true idiopathic ulcers is not established.^[7]

Prognosis. H. pylori-negative idiopathic bleeding ulcers carry a high incidence of recurrent bleeding and mortality.^[11]

• Stress-related mucosal disease.^[4]
  • Upper gastrointestinal mucosal lesions are present in 75–100% of ICU patients, but only 1–6% develop clinically important bleeding.
  • Bleeding risk is predicted by mechanical ventilation for more than 48 hours and by coagulopathy.
  • Stress-ulcer bleeding carries an estimated mortality of 40–50%, mostly from the underlying illness.
• Acid hypersecretory states. Zollinger-Ellison syndrome is a rare but important cause of refractory gastroduodenal ulcers.^[5]
• Malignancy. Malignancy is a recognized cause of refractory ulcers. Gastric cancer and gastric lymphoma must be excluded in non-healing gastric ulcers.^[5]
• Rare causes. These include:^[6]
  • Crohn’s disease.
  • Eosinophilic gastroduodenitis.
  • Systemic mastocytosis.
  • Radiation injury.
  • Cytomegalovirus or herpes simplex virus infection.
  • Helicobacter heilmannii colonization.
  • Cameron ulcers with a large hiatal hernia.
  • Anastomotic (marginal) ulcers after gastric resection.

These factors increase ulcer risk but are not sole causes. Details belong in the risk factors microchapter.

• Cigarette smoking
  • Independent risk factor for PUD (OR 3.8, 95% CI 1.7–9.8).^[13]
  • Delays ulcer healing.^[5]
• Psychosocial and environmental factors. Risk is increased by psychological stress, family history of PUD, and adverse childhood circumstances such as crowded living conditions and poor water quality.^[2]
• Disaster-related stress. Evidence comes from the 2011 Great East Japan Earthquake:
  • PUD incidence rose 1.5-fold and hemorrhagic ulcers 2.2-fold.
  • Non-H. pylori, non-NSAID ulcers rose from 13% to 24% of cases.^[18]
  • Living in a refugee shelter independently predicted ulcer bleeding (OR 4.4, 95% CI 2.1–9.6).^[19]

• Establish H. pylori status in every patient with PUD. Take a detailed, repeated NSAID and aspirin history that includes over-the-counter products.^[5][20]
• Do not accept a single negative H. pylori test while the patient is on a PPI. Use two or more methods, or serology.^[5]
• Before labeling an ulcer idiopathic, exclude:^[5][6]
  • False-negative H. pylori testing.
  • Occult NSAID or aspirin use.
  • Malignancy.
  • Zollinger-Ellison syndrome.
• A silent ulcer, especially one presenting with bleeding or perforation, should raise suspicion for NSAID use.^[9]