When a patient has hematemesis but EGD shows blood with no ulcer or varices, think of an intermittently bleeding lesion like a Dieulafoy lesion that can be hard to visualize when not actively bleeding.
Question - Bleeding source
A 45-year-old man with a history of hypertension, type 2 diabetes mellitus, and long-standing alcohol use presents with sudden-onset nausea and vomiting of bright-red blood.
On physical examination, his pulse in 110 beats/minute and his blood pressure is 140/95 mm Hg. Laboratory studies reveal a glycated hemoglobin level of 8.9 g/dL, platelet count of 210,000/µL, and mean corpuscular volume of 92 fL. Upper endoscopy is performed and demonstrates a normal-appearing esophagus with no hiatal hernia. There is evidence of old blood in the stomach without any erosions, varices, or ulcers, and you note a normal-appearing duodenum.
What is the likely cause of this patient's bleeding?
Answer Options:
A. esophageal varices
B. portal gastropathy
C. small intestinal angiodysplasia
D. Cameron lesion
E. Dieulafoy lesion
Dieulafoy lesions are an important board-relevant cause of significant, often sudden upper GI bleeding where endoscopy may show blood in the stomach without an obvious ulcer crater. They represent a large-caliber submucosal artery that erodes through a small mucosal defect, so the surrounding mucosa can appear normal—and the lesion can be missed if bleeding has stopped by the time of EGD.
Current major guidance on nonvariceal upper GI bleeding (e.g., ACG 2021; ESGE 2021) emphasizes that not all clinically important upper GI hemorrhage is ulcer-related. Dieulafoy lesions are a recognized nonvariceal source that may require repeat endoscopy, careful inspection (often proximal stomach/lesser curvature), and endoscopic therapy when identified.
Why This Question Is Often Missed
-
Test-takers over-anchor on alcohol use → varices, despite an explicitly normal esophagus and no varices on EGD.
-
Many expect “upper GI bleed” to equal peptic ulcer disease, but the stem removes ulcers/erosions.
-
Intermittent bleeds (Dieulafoy) are missed because learners assume “normal EGD” excludes a gastric source.
What the Distractors Indicate
| Option | What It Tests / Implies | Why It’s Wrong Here |
|---|---|---|
| A. esophageal varices | Portal HTN/advanced liver disease causing brisk hematemesis | EGD shows **normal-appearing esophagus** and **no varices**. Platelets are normal (not supportive of portal HTN). |
| B. portal gastropathy | Portal hypertension-related gastric mucosal oozing | Endoscopy typically shows **mosaic/snake-skin** mucosa and friability/oozing; stem reports **no erosions/varices** and otherwise normal stomach aside from old blood. |
| C. small intestinal angiodysplasia | “Obscure GI bleeding,” often occult/low-grade, sometimes melena | Usually causes **occult bleeding/iron deficiency** or melena more than sudden large-volume hematemesis; also EGD localizes blood in stomach. |
| D. Cameron lesion | Linear erosions/ulcers at diaphragmatic pinch in a hiatal hernia | Stem explicitly says **no hiatal hernia**. Classic presentation is chronic blood loss/iron deficiency. |
| E. Dieulafoy lesion | Intermittent arterial bleed with normal surrounding mucosa | Fits: **old blood present** with no ulcer/varix; lesion may not be seen if not actively bleeding. |
Internal Medicine High-Yield Pearl for Exam Prep
Hematemesis + EGD showing blood but no ulcer/varix → suspect an intermittent arterial lesion (Dieulafoy) and consider that it may be missed unless actively bleeding.
Core Learning Objectives
1. Differentiate major causes of upper GI bleeding when EGD does not show ulcers or varices.
2. Recognize clinical/endoscopic clues that point to Dieulafoy lesion versus portal hypertensive or hernia-associated bleeding.
The Exam “Test Trick” at Play
The stem deliberately front-loads alcohol use to bait varices/portal hypertensive bleeding, then neutralizes that path by providing a normal esophagus and no varices or gastropathy. The “old blood in the stomach with otherwise normal mucosa” pattern is the exam cue for a hard-to-spot, intermittently bleeding vascular lesion.
Additional Practice Questions and Remediation for Dieulafoy Lesion / Nonvariceal UGIB
Internal Medicine Practice Question 1 — Intermittent hematemesis
A 52-year-old presents with recurrent brisk hematemesis. Initial EGD shows blood in the stomach but no ulcer; repeat EGD during bleeding reveals a protruding vessel in the proximal stomach with minimal mucosal defect. Best diagnosis?
A. Mallory-Weiss tear
B. Dieulafoy lesion
C. Gastric adenocarcinoma
D. Portal hypertensive gastropathy
E. Cameron lesion
Internal Medicine Practice Question 2 — Board-style exclusion
A patient with hematemesis undergoes EGD: normal esophagus, no varices, no ulcer, no erosions, but fresh blood pooled in the fundus. Which lesion is most likely if bleeding is intermittent?
A. Esophageal varices
B. Gastric ulcer
C. Stress gastritis
D. Dieulafoy lesion
E. Gastric antral vascular ectasia
Internal Medicine Practice Question 3 — Hiatal hernia clue
A 68-year-old with chronic iron deficiency anemia has EGD showing a large sliding hiatal hernia and linear erosions at the diaphragmatic impression. Cause?
A. Dieulafoy lesion
B. Cameron lesion
C. Portal gastropathy
D. Esophageal varices
E. Small bowel angiodysplasia
Internal Medicine Practice Question 4 — Portal HTN pattern recognition
A cirrhotic patient has melena and endoscopy reveals a mosaic “snake-skin” gastric mucosa with diffuse oozing; no active variceal bleeding. Diagnosis?
A. Dieulafoy lesion
B. Portal hypertensive gastropathy
C. Cameron lesion
D. Mallory-Weiss tear
E. Small intestinal angiodysplasia
Internal Medicine Practice Question 5 — Next-step reasoning
A patient has ongoing hematemesis. Initial EGD shows large amounts of blood in the stomach but no source identified. After stabilization, what is the most appropriate next diagnostic/therapeutic step to identify a suspected intermittent nonvariceal lesion?
A. Colonoscopy immediately
B. Repeat endoscopy with careful inspection/therapy (consider during active bleeding)
C. H. pylori breath test
D. Empiric anticoagulation for suspected mesenteric ischemia
E. Capsule endoscopy as first-line
Mini Case Discussion Prompt
How would your differential and next steps change if this same patient had (1) thrombocytopenia and stigmata of cirrhosis, or (2) a large hiatal hernia on EGD, or (3) recurrent melena with negative EGD/colonoscopy?
Mini-FAQ
Q1: Why is Dieulafoy lesion “easy to miss” on endoscopy?A1: The ABIM-style concept is that the surrounding mucosa can look normal and bleeding is intermittent, so if it’s not actively bleeding, EGD may show only pooled/old blood.
Q2: Does alcohol use make Dieulafoy more likely than varices?
A2: Alcohol increases the pre-test probability of portal hypertension, but boards expect you to follow the provided endoscopic findings; a normal esophagus with no varices shifts away from variceal bleeding.
Q3: When should you think of portal hypertensive gastropathy instead?
A3: When there’s portal hypertension plus the characteristic endoscopic “mosaic/snake-skin” gastric pattern with diffuse oozing rather than a focal lesion.
Q4: Where does small bowel angiodysplasia fit on exams?
A4: It’s classically tested in “obscure GI bleeding” (often occult or melena) after negative EGD/colonoscopy, rather than sudden large-volume hematemesis.
This question appears in Med-Challenger Internal Medicine Review with CME
Try for free and save. Ace your exams and meet your CME/MOC requirements for just $35 a month!
No matter your program, no matter the size, Med-Challenger for Groups and Institutions can better prepare your program or group, fulfill industry requirements, and increase test scores.
