Gastrointestinal Imaging

Why Is CT Enterography the Next Step for GI Bleeding After a Negative Endoscopy?

·
Why Is CT Enterography the Next Step for GI Bleeding After a Negative Endoscopy?

A 62-year-old patient with a history of iron-deficiency anemia and recent melena returns from the gastroenterology suite. The upper endoscopy was unremarkable, with visualization to the distal duodenum. Despite this, the patient’s hematocrit continues to drop, and the clinical picture points to an ongoing, nonvariceal upper gastrointestinal (GI) bleed. You are now faced with a common and challenging clinical question: with the most likely sources ruled out by direct visualization, what is the appropriate next step in imaging to locate this obscure bleed? This article provides a detailed workflow for this specific scenario, grounded in the American College of Radiology (ACR) Appropriateness Criteria. For an adult with nonvariceal upper GI bleeding and a negative endoscopy, the ACR rates CT enterography as Usually Appropriate for initial imaging.

Who Fits This Clinical Scenario for Obscure GI Bleeding?

This guidance is tailored for a specific patient population: adults presenting with evidence of nonvariceal upper GI bleeding who have already undergone a technically adequate and unrevealing upper endoscopy (esophagogastroduodenoscopy, or EGD). The key inclusion criteria are:

  • Clinical Evidence of Bleeding: This can be overt (melena, hematemesis) or occult (iron-deficiency anemia, positive fecal occult blood test) with a high suspicion for an upper GI source.
  • Negative Upper Endoscopy: A recent EGD has been performed and failed to identify a bleeding source. This implies the source is likely distal to the ligament of Treitz or was missed during the initial examination.
  • Nonvariceal Etiology: The patient has no history or stigmata of portal hypertension to suggest esophageal or gastric varices are the cause.

It is critical to distinguish this situation from similar, but distinct, clinical scenarios. This workflow does not apply if:

  • No Endoscopy Has Been Performed: For patients with suspected nonvariceal upper GI bleeding who have not yet had an EGD, endoscopy remains the primary diagnostic and therapeutic tool.
  • Endoscopy Confirms a Source: If the EGD identifies a bleeding source (e.g., a peptic ulcer) but therapeutic intervention is not possible or fails, the next step is typically interventional radiology, not diagnostic imaging.
  • Bleeding is Postsurgical or Traumatic: These cases have a different differential diagnosis (e.g., anastomotic leak, fistula) and may require different imaging protocols.

What Diagnoses Are You Working Up in This Scenario?

When a standard EGD is negative, the diagnostic focus shifts from the esophagus, stomach, and duodenum to the small bowel. The differential diagnosis for this “obscure” GI bleeding is broad, and the goal of imaging is to identify pathologies that are difficult to visualize endoscopically.

Small Bowel Angioectasia (AVM): Arteriovenous malformations are the most common cause of obscure GI bleeding, particularly in patients over 50. These are small, fragile vascular lesions in the submucosa that can bleed intermittently. They are often multiple and can be challenging to spot on endoscopy if not actively bleeding.

Small Bowel Tumors: While less common, neoplasms of the small intestine are a critical consideration. These can include benign tumors like adenomas or malignant ones such as gastrointestinal stromal tumors (GISTs), adenocarcinoma, lymphoma, or carcinoid tumors. These masses may bleed due to ulceration of the overlying mucosa.

Dieulafoy’s Lesion: This is a large, tortuous, and persistent submucosal artery that erodes the overlying epithelium without a primary ulcer. Because the mucosal defect can be tiny, these lesions are notoriously difficult to identify on endoscopy unless they are actively bleeding at the time of the procedure. They can occur anywhere in the GI tract but are often found in the stomach, just beyond the reach of a standard EGD.

Aortoenteric Fistula: A rare but life-threatening cause of massive GI bleeding. This diagnosis should be strongly considered in any patient with a history of aortic aneurysm repair or other aortic surgery. It represents a direct communication between the aorta and the bowel, most commonly the duodenum.

Why Is CT Enterography the Recommended Study for This Presentation?

The ACR designates CT enterography (CTE) as a Usually Appropriate initial imaging study in this scenario because it is specifically optimized to evaluate the small bowel, the most likely location of an occult bleeding source.

The rationale for CTE’s high rating is based on its technique and diagnostic capability. The procedure involves the oral administration of a large volume of neutral contrast material (like water or a polyethylene glycol solution) to distend the small bowel loops. This distention is crucial; it effaces the normal mucosal folds, allowing for clear visualization of the bowel wall, subtle mucosal abnormalities, and small endoluminal masses that would otherwise be obscured.

Intravenous contrast is administered and imaging is performed in multiple phases:

  • Non-contrast: To identify high-density material like intramural hematoma or recently ingested radiopaque pills.
  • Arterial Phase: Timed to capture the peak enhancement of arteries, this phase is essential for identifying active arterial extravasation (a direct sign of bleeding) and hypervascular lesions like angioectasias or some tumors.
  • Portal Venous Phase: This phase provides optimal enhancement of the bowel wall, mesentery, and solid organs, making it ideal for detecting most tumors, inflammatory conditions, and venous abnormalities.

Compared to other modalities, CTE provides a superior balance of information. While CTA abdomen and pelvis without and with IV contrast is also rated Usually Appropriate, the dedicated CTE protocol with neutral oral contrast provides better small bowel luminal and mural evaluation. Alternatives rated lower include:

  • RBC scan (May be appropriate): A tagged red blood cell scan is highly sensitive for detecting slow or intermittent bleeding (as low as 0.1 mL/min). However, its spatial resolution is poor, often localizing a bleed only to a general quadrant of the abdomen, which is less helpful for guiding subsequent therapy than the precise anatomical map provided by CTE.
  • Arteriography (Usually not appropriate): Conventional catheter angiography is an invasive procedure. It is an excellent problem-solving and therapeutic tool but is not a first-line diagnostic test. It requires a relatively brisk rate of bleeding (at least 0.5 mL/min) to be positive and carries risks associated with arterial access and contrast. It is typically reserved for cases where a bleeding source has been localized by another study.

The primary trade-off with CTE is the radiation dose, which is significant (ACR RRL® ☢☢☢☢, 10-30 mSv). This dose is justified by the high diagnostic yield in a patient with consequential, unexplained bleeding.

Once you’ve decided on CT enterography, our protocol guide covers the technique, contrast, and reading principles: CT Enterography.

What’s Next After CT Enterography? Downstream Workflow

The results of the CT enterography will dictate the subsequent clinical pathway. The goal is to move from diagnosis to targeted therapy.

  • If the CTE is positive for active bleeding: This is an urgent finding. The next step is immediate consultation with Interventional Radiology for catheter-based angiography with the intent of localization and embolization. The CTE provides a crucial roadmap for the interventionalist.
  • If the CTE identifies a tumor or mass: The patient should be referred to a gastroenterologist for deep enteroscopy (e.g., double-balloon or spiral enteroscopy) or video capsule endoscopy (VCE) to obtain a tissue diagnosis. Surgical consultation for potential resection is also warranted.
  • If the CTE is negative but clinical suspicion remains high: A negative high-quality CTE makes a significant bleed or large mass unlikely. However, small, flat mucosal lesions like angioectasias can be missed. In this case, the next logical step is often VCE, which provides excellent mucosal visualization and is highly sensitive for these lesions. If bleeding is intermittent, a tagged RBC scan may still have a role.
  • If the CTE is indeterminate: An equivocal finding, such as focal bowel wall thickening without a clear mass or active bleed, requires multidisciplinary discussion between the clinical team, gastroenterology, and radiology to determine the best next step, which could include a repeat study, VCE, or deep enteroscopy.

Pitfalls to Avoid (and When to Get Help)

Navigating this clinical problem requires careful planning to maximize diagnostic yield and avoid common errors.

  • Inadequate Bowel Preparation: For CTE to be effective, the patient must ingest the full volume of oral contrast to achieve adequate small bowel distention. An incomplete prep can render the study nondiagnostic.
  • Wrong CT Protocol: Ordering a “CT Abdomen/Pelvis with contrast” instead of a dedicated multiphase “CT Enterography” or “CT Angiography for GI Bleed” may miss the crucial arterial phase needed to detect active extravasation. Be specific in your order.
  • Delaying the Scan: In a patient with active, hemodynamically significant bleeding, timing is critical. The chance of seeing contrast extravasation is highest during an active bleed. Coordinate with the radiology department to perform the scan expeditiously.
  • Ignoring the Radiation Dose: While necessary, the radiation dose is not trivial. Ensure the indication is appropriate and avoid repeated, non-diagnostic CT scans.

If a patient becomes hemodynamically unstable despite resuscitation, this constitutes a clinical emergency. Escalate immediately to GI and Interventional Radiology for potential emergent angiography, bypassing further diagnostic imaging steps.

Related ACR Topics and Tools

This article focuses on one specific clinical variant. For a comprehensive overview of imaging for all related presentations, or to explore the technical details of the recommended studies, the following resources are available.

Frequently Asked Questions

Why not just do a video capsule endoscopy (VCE) first after a negative EGD?

While VCE is excellent for detecting mucosal lesions like angioectasias, it has several limitations. It cannot detect actively bleeding sources in real-time, cannot provide precise anatomical localization for intervention, and is contraindicated in patients with suspected strictures or obstruction. CT enterography provides a comprehensive anatomical survey, can detect active bleeding, and can identify contraindications to VCE, making it a more robust initial imaging test in this setting.

What if my patient has renal insufficiency and cannot receive IV contrast?

This complicates the workup significantly. A non-contrast CT is of very limited value. MR enterography, rated as May be appropriate, becomes a strong consideration as it does not use iodinated contrast and avoids radiation. However, it is less sensitive for detecting active bleeding than CTE. A tagged RBC scan is another option that does not require iodinated contrast. The choice depends on the acuity of the bleed and local imaging expertise.

Is CTA rated the same as CTE? Which one should I order?

Both CT enterography and CTA of the abdomen and pelvis are rated Usually Appropriate. The key difference is the oral contrast protocol. CTE uses a large volume of neutral contrast to distend the bowel, which is superior for evaluating the bowel wall for tumors or inflammatory causes. CTA protocols may not use this specific oral prep but are optimized for detecting vascular abnormalities and active bleeding. For a general ‘obscure GI bleed’ workup, CTE is often preferred for its comprehensive nature. If the suspicion for a brisk, active bleed is extremely high, a dedicated multiphase CTA is also an excellent choice.

The patient’s bleeding has stopped. Should I still order a CT enterography?

Yes. Even if the patient is not actively bleeding at the moment of the scan, CTE can still identify the underlying cause, such as a small bowel tumor, an arteriovenous malformation, or inflammatory bowel disease. While the chance of seeing active contrast extravasation is gone, the ability to find the structural cause of the intermittent bleed remains high. The diagnostic yield is highest when performed closer to the bleeding episode.

What is the difference between this scenario and the one where endoscopy confirms bleeding but finds no clear source?

This scenario (‘negative endoscopy’) implies the endoscopist saw no evidence of bleeding (e.g., no blood, no visible lesion) within the examined area. The sibling scenario (‘endoscopy confirms bleeding without a clear source’) means the endoscopist saw active bleeding or stigmata of recent hemorrhage (e.g., fresh blood in the duodenum) but could not pinpoint the exact origin. The latter implies a more active, ongoing process, and the imaging choice (often CTA) is more focused on finding active extravasation immediately.

Reviewed by Pouyan Golshani, MD, Interventional Radiologist — May 26, 2026