33-Year-Old Woman With Epigastric Pain and Hematemesis

A 33-year-old woman with a history of depression was admitted to the Mayo Clinic for evaluation of epigastric pain. The pain, which the patient described as “severe,” began suddenly while sitting at work 1 week before admission. The pain was constant (although it had periods of worsening) and radiated to her back. It did not vary with position or dietary intake and was not responsive to acetaminophen, hydrocodone, or omeprazole, which had been prescribed by her local physician, who evaluated her 3 times for this symptom. On the day before admission, she had 2 episodes of vomiting “dark, maroon blood with clots,” which was not preceded by emesis without blood. She denied the former or current use of alcohol, tobacco, nonsteroidal anti-inflammatory drugs, or herbal supplements and also had experienced no early satiety or weight loss. There was no family history of gastrointestinal malignant tumors or ulcer disease. Results of a review of systems were otherwise negative. The only medication she was taking was citalopram.

On admission, the patient appeared comfortable. She was afebrile, with a blood pressure of 127/64 mm Hg, pulse of 92/min, respiratory rate of 24/min, and a body mass index (calculated as weight in kilograms divided by height in meters squared) of 35.5. There was no orthostatic change in her blood pressure or pulse. Physical examination revealed normal bowel sounds and tenderness in the epigastrium without guarding or rebound. The Murphy sign was absent, and there was no organomegaly. The results of the rest of the physical examination were unremarkable.

Initial laboratory evaluation of this patient revealed the following: hemoglobin, 10.7 g/dL; mean corpuscular volume, 87.9 fL; red cell distribution width, 13.0%; platelets, 325 × 10⁹/L; leukocytes, 8.8 × 10⁹/L; creatinine, 0.7 mg/dL; blood urea nitrogen, 12 mg/dL; lipase, 15 U/L; amylase, 34 U/L; bilirubin, 0.7 mg/dL; aspartate, 15 U/L; and alanine, 22 U/L.

• 1
  What is the most likely cause of hematemesis in this patient?

  • a
    Mallory-Weiss tear
  • b
    Eosinophilic esophagitis
  • c
    Gastric adenocarcinoma
  • d
    Peptic ulcer disease
  • e
    Esophageal varices

Hematemesis is a common symptom in both the outpatient and inpatient settings. The differential diagnosis is broad, and it is essential to identify the exact cause because some causes can be life-threatening. In our patient, the history is important, but her age and physical examination findings should also be considered when creating the differential diagnosis. A Mallory-Weiss tear is a possibility because its most common clinical presentation is hematemesis accompanied by epigastric pain. However, patients with a Mallory-Weiss tear usually have a history of vomiting or wrenching before the bleeding, making this diagnosis less likely. In adults, eosinophilic esophagitis commonly manifests with intermittent food impaction and dysphagia.¹ Our patient had neither, making this an incorrect diagnosis. Gastric adenocarcinoma may manifest with epigastric pain, but the mean age for patients at presentation is 60 years, and this disorder is more common in men.² The patient also lacks several “red flag” features of gastric carcinoma, such as iron deficiency anemia, progressive dysphagia, unintentional weight loss, vomiting persistently, or mass on physical examination. The most likely diagnosis is peptic ulcer disease, which can present with hemorrhage. Although this patient did not have some of the typical symptoms, such as nocturnal pain, she had epigastric pain with radiation to the back consistent with duodenal ulcer. Acute epigastric pain with radiation to the back is also concerning for a posterior penetrating duodenal ulcer, causing pancreatitis. Her pancreatic enzyme levels, however, were normal. Without a history of liver disease and no clinical evidence of cirrhosis, esophageal varices are unlikely, and this is an unlikely diagnosis.

The patient was kept fasting. Normal saline via intravenous infusion was administered. Hemoglobin levels were measured every 8 hours.

• 2
  In regard to the hematemesis management, what would be the best initial approach?

  • a
    Observe
  • b
    Administer intravenous pantoprazole, 80-mg bolus, followed by an 8-mg/h infusion
  • c
    Administer ranitidine, 150 mg orally twice daily
  • d
    Administer omeprazole, 40 mg orally daily
  • e
    Administer octreotide and perform an emergent endoscopy

The initial therapy and need for intervention in patients with upper gastrointestinal bleeding are based on patients' clinical stability and comorbidities. Observation would be reasonable in a stable patient with abdominal pain, but hematemesis is worrisome and could be indicative of a serious clinical event; therefore, preemptive measures should be taken. According to the international consensus recommendations on the treatment of patients with nonvariceal upper gastrointestinal bleeding and the work of Lau et al,³ preendoscopic intravenous proton pump inhibitor (PPI) therapy had a beneficial effect on the need for intervention, supportive cost-effectiveness analysis, and excellent safety profile, especially in patients suspected of having high-risk stigmata.⁴ The standard approach has been to start intravenous PPI therapy before endoscopy and then to continue for 3 days after endoscopy if there was indeed evidence of high-risk stigmata for rebleeding. In addition, the international consensus recommendation states that preendoscopic treatment with PPI reduced the proportion of patients with high-risk stigmata and the need for endoscopic therapy compared with a control group and histamine₂ receptor antagonist.⁴ Therefore, administering ranitidine, a histamine₂ receptor antagonist, would not be appropriate. In addition, high-dose intravenous PPI therapy reduced rebleeding, surgery, and mortality in patients with high-risk stigmata who have undergone endoscopy with hemostatic therapy. Lower doses of PPI reduced rebleeding but did not reduce mortality in the same subset of patients.⁴ The patient has no history of alcohol use or other reason to think portal hypertension has led to variceal bleeding, so octreotide with emergent endoscopy would not be appropriate.

In this patient, we administered intravenous pantoprazole. Six hours later she continued to have epigastric tenderness. Her hemoglobin level remained stable at 10.3 mg/dL, and she had no further hematemesis.

• 3
  What is the best diagnostic test for this patient?

  • a
    Computed tomography of the abdomen
  • b
    Noninvasive Helicobacter pylori testing
  • c
    Esophagogastroduodenoscopy (EGD)
  • d
    Abdominal ultrasonography
  • e
    Barium esophagram

In patients with hematemesis, the decision regarding best diagnostic test should involve cost, yield, and potential complications. In the setting of a physical examination notable for epigastric pain and no palpable mass and laboratory analysis remarkable only for a normocytic anemia, computed tomography is not the first option to evaluate pain and hematemesis. Noninvasive H pylori testing is reasonable in this clinical setting if there are not features, such as hematemesis and severe pain, that are worrisome for ulcerative bleeding or other conditions. Although H pylori is found to be present in 43% to 56% of patients with peptic ulcer bleeding, upper gastrointestinal bleeding can come from additional sources, including tumors, so testing with direct visualization is the best approach.⁵ The best diagnostic test is EGD because it will allow direct visualization of the gastrointestinal mucosa, biopsy of suspicious lesions, and treatment of any active bleeding sites. Abdominal ultrasonography is the best choice if we are concerned for biliary tract disease, but this patient's clinical picture does not fit this concern. She did not experience right upper quadrant pain, and laboratory evaluation was not consistent with cholestasis. A barium esophagram is a reasonable study to perform if she had a history of dysphagia or reflux.

The EGD revealed no active bleeding and 2 well-circumscribed duodenal ulcers. Gastric biopsy specimens were obtained. The patient tolerated the procedure. Her epigastric pain subsequently improved, and she was started on a diet.

• 4
  What is the next best step in the treatment of this patient's condition?

  • a
    Send home with a citalopram prescription
  • b
    Empirically treat H pylori with 2 weeks of triple therapy and PPI
  • c
    Wait for biopsy results before treating H pylori
  • d
    Obtain a serum gastrin level
  • e
    Schedule a surgical consultation

The patient was taking citalopram, which may place her at an increased risk of gastrointestinal bleeding.⁶ If an antidepressant is still needed, it might be best to choose one from another class. In the 1990s, H pylori was present in 80% to 95% of all duodenal ulcers, and although the rates of H pylori–positive duodenal ulcers are decreasing, there is still a high association.⁷ However, in the United States the pretest probability of H pylori infection among young individuals is low. Therefore, empiric treatment for H pylori is not indicated. It would be appropriate to wait for the biopsy result. Serum gastrin measurement is not an initial test to perform because a gastrin-producing tumor, leading to ulceration, is an uncommon cause. It is reasonable to perform if the patient had multiple or refractory ulcers, ulcer location distal to the duodenum, diarrhea, or a personal or family history of multiple endocrine neoplasia type 1.⁸ Because the ulcers looked clean without visual signs worrisome for tumor or perforation, surgical intervention is not warranted at this time.

Biopsy results 3 days later showed moderate active chronic gastritis and H pylori. There was no intestinal metaplasia. The patient had no recurrence of epigastric pain.

• 5
  Given these biopsy results, after treatment what follow-up is recommended?

  • a
    No follow-up is necessary
  • b
    Repeat EGD with biopsy in 4 weeks
  • c
    Urea breath test 4 weeks after completion of H pylori therapy
  • d
    H pylori polyclonal stool testing in 4 weeks after completion of therapy
  • e
    Serologic testing for H pylori in 4 weeks

No follow-up would be inappropriate because the American College of Gastroenterology recommends confirming eradication of H pylori in patients with an associated ulcer.⁹ Noninvasive testing is recommended over invasive testing unless there is a clinical indication that endoscopy is necessary, as in a gastric ulcer or antibiotic resistance.⁹ Therefore, repeat EGD in 4 weeks is incorrect. Urea breath testing is the best option. Urea breath testing and stool antigen testing using monoclonal antibodies performed at least 4 weeks after treatment are tests that can be used to confirm eradication of infection, whereas H pylori polyclonal stool testing is less accurate.⁹ Therefore, H pylori polyclonal stool testing is not appropriate in this patient. Serologic testing is not recommended because patients could continue to have positive antibodies for weeks to months after completion of treatment.⁹

The patient was discharged from the hospital after she had no further bleeding, reported decreased epigastric pain, and tolerated a diet. Her hemoglobin level was 11.7 mg/dL on dismissal. She was scheduled for follow-up with her primary care physician in 1 month.

Discussion

Peptic ulcers are defects in the gastrointestinal mucosa that extend beyond the muscularis mucosa and cause considerable morbidity and health care cost annually. A recent study reported annual incidence rates of peptic ulcer disease ranging from 0.10% to 0.19% for physician-diagnosed peptic ulcer disease and 0.03% to 0.17% when based on hospitalization data.¹⁰ Some common clinical presentations include burning or gnawing pain in the epigastrium, although patients can be completely asymptomatic or have only vague pain until a complication, such as bleeding or perforation, occurs. H pylori is noted to be a pathogen in 73% of duodenal ulcers and 95% of duodenal ulcers if nonsteroidal anti-inflammatory drugs were also being used.⁷ It is important to treat patients with H pylori-positive ulcers because the ulcer recurrence rate after treatment is low compared with not treating infection (6% vs 67%).¹¹

The diagnostic approach to suspected ulcerative disease should include an evaluation for H pylori. The sensitivity of all testing for this bacteria is reduced if the patient is undergoing active therapy with a PPI, bismuth, or antibiotics.⁹ In patients requiring endoscopy who have not been taking a PPI within 1 to 2 weeks or an antibiotic or bismuth within 4 weeks of endoscopy, the rapid urease test provides an accurate, inexpensive means of identifying H pylori.⁹ In patients who have been taking a PPI, antibiotics, or bismuth, endoscopic testing should include biopsies with or without the rapid urease test.⁹ In noninvasive testing, the diagnostic accuracy of the urea breath test is 95% in studies and the most reliable test to document eradication.⁹ Antibody testing is inexpensive and widely available but has poor positive predictive value in populations with a low prevalence of H pylori infection, therefore limiting its usefulness in clinical practice. The fecal antigen test provides a reliable means of identifying active H pylori infection before antibiotic therapy. The sensitivity and specificity are 91% and 93%, respectively, for the polyclonal antibody and 96% and 97%, respectively, for the monoclonal antibody.⁹ Posttreatment confirmation of eradication is best done using the monoclonal antibody as opposed to the polyclonal antibody.⁹ The urea breath test tends to be less widely available and reimbursement has been inconsistent, whereas the fecal antigen test comes with the unpleasantness associated with collecting stool.⁹ Either way, the pretest and posttest positive and negative predictive values are similar enough that it is the clinician's preference regarding which test to choose.

Because H pylori eradication significantly decreases recurrence of duodenal ulcers, it is important to initiate treatment when infection is noted. For PPI (standard dose twice daily), clarithromycin (500 mg twice daily), amoxicillin (1000 mg twice daily), or metronidazole (400 or 500 mg twice daily), 14-day treatment is more effective than 7 days. The first-choice treatment in populations with less than 15% to 20% clarithromycin resistance is PPI, clarithromycin, and amoxicillin or metronidazole. In populations with less than 40% metronidazole resistance, PPI, clarithromycin, and metronidazole treatment is preferable.¹² Bismuth-containing quadruple treatments remain the best second-choice treatment, if available.¹² If bismuth is not available, PPI and amoxicillin or tetracycline and metronidazole are recommended.¹²

Because of the association of H pylori and peptic ulcer disease, it is important for clinicians to perform an appropriate diagnostic work-up, prescribe appropriate therapy, and perform timely follow-up. Treatment of H pylori decreases ulcer recurrence. Therefore, clinicians should be aware of diagnostic strategies and current treatment guidelines.