Interventional Radiology Imaging

BRTO for Bleeding Gastric Varices with Encephalopathy and MRI-Detected Shunt

·
BRTO for Bleeding Gastric Varices with Encephalopathy and MRI-Detected Shunt

A 62-year-old male with a history of alcohol-related cirrhosis is admitted with hematemesis. He is disoriented and confused, consistent with hepatic encephalopathy, and his Model for End-stage Liver Disease (MELD) score is calculated at 18. Emergent endoscopy identifies large, high-flow gastric varices as the source of bleeding, and temporary hemostasis is achieved. A recent MRI confirms the presence of a large, spontaneous gastrorenal shunt. The primary team and consultants are now faced with a critical decision: what is the most effective and safest definitive treatment to prevent re-bleeding and manage his complex presentation? This article provides a detailed workflow for this specific clinical scenario, guided by the American College of Radiology (ACR) Appropriateness Criteria, which rate Balloon-occluded Retrograde Transvenous Obliteration (BRTO) as Usually appropriate.

Who Fits This Clinical Scenario for Gastric Variceal Bleeding?

This guidance is specifically for patients who meet a distinct set of clinical and anatomic criteria. The recommendations apply to individuals with established cirrhosis who are actively bleeding (or have recently bled) from large, high-flow gastric varices. Key accompanying features are the presence of hepatic encephalopathy and moderately elevated liver dysfunction, reflected by a MELD score in the range of 18.

The most critical inclusion criterion, however, is the anatomic confirmation of a large, spontaneous portosystemic shunt—specifically a gastrorenal shunt—on cross-sectional imaging like CT or MRI. The presence of this shunt is the lynchpin of the entire treatment decision, as it defines the pathophysiology and dictates the optimal interventional approach.

This workflow does not apply to several similar-appearing but distinct clinical situations:

  • Patients without a large, defined shunt: A patient with a similar MELD score but bleeding from small, low-flow gastric varices and no demonstrable gastrorenal shunt would follow a different management pathway.
  • Patients with extremely severe liver failure: For patients with very high MELD scores (e.g., >25) or those actively listed for transplant, the risk-benefit analysis for any intervention changes significantly, often favoring TIPS as a bridge to transplantation.
  • Patients with isolated splenic vein thrombosis: If gastric varices are caused by splenic vein occlusion (often from pancreatitis), the underlying problem is regional, not systemic portal hypertension, and the treatment is different (e.g., splenectomy or splenic artery embolization).

What Diagnoses Are You Working Up in This Scenario?

In this context, the primary diagnosis of bleeding gastric varices is already established. The crucial “workup” is focused on understanding the underlying hemodynamic drivers and associated complications that guide the choice of intervention. The goal is not to find a new diagnosis but to characterize the existing one to select the most effective therapy.

Bleeding Gastric Varices Secondary to a High-Flow Gastrorenal Shunt
This is the central pathophysiologic process. In patients with portal hypertension, the body forms spontaneous shunts to decompress the high-pressure portal system. A gastrorenal shunt diverts a large volume of portal blood directly into the left renal vein, bypassing the liver. While this decompresses the portal system, it preferentially directs high-pressure, high-volume flow into the gastric fundal veins, causing them to dilate into large varices prone to catastrophic bleeding. The intervention must address this specific outflow pathway.

Worsening Hepatic Encephalopathy due to Portosystemic Shunting
The patient’s confusion is a direct consequence of the gastrorenal shunt. This large-caliber “bypass” allows neurotoxic substances from the gut, such as ammonia, to circumvent hepatic metabolism and enter the systemic circulation. This contributes significantly to the patient’s encephalopathy. An ideal intervention should not only stop the bleeding but also aim to improve or at least not worsen this neurological complication.

Underlying Decompensated Cirrhosis and Portal Hypertension
While the shunt is the immediate anatomic target, it is a symptom of the root cause: severe portal hypertension from cirrhosis. The MELD score of 18 indicates significant liver dysfunction. The chosen procedure must be tolerated by a patient with limited physiologic reserve and should not precipitate further hepatic decompensation if possible.

Why Is BRTO Usually Appropriate for This High-Flow Variceal Bleed?

For a cirrhotic patient with bleeding gastric varices, hepatic encephalopathy, and a defined gastrorenal shunt, Balloon-occluded Retrograde Transvenous Obliteration (BRTO) is rated as Usually appropriate by the ACR. This procedure directly addresses both the bleeding and the encephalopathy by targeting the shunt itself.

In BRTO, an interventional radiologist navigates a catheter, typically from a femoral or jugular vein approach, into the left renal vein and subsequently into the gastrorenal shunt. A balloon is inflated within the shunt to occlude outflow, and a sclerosant agent is injected “retrograde” to fill and thrombose the gastric varices and the shunt. The balloon remains inflated for a period to allow the sclerosant to work before being deflated and removed.

The key rationales for this recommendation are:

  • High Efficacy for Bleeding Control: BRTO directly obliterates the varices and the primary outflow vessel. This approach has demonstrated high rates of technical success and long-term prevention of re-bleeding from gastric varices when this specific anatomy is present.
  • Improvement of Hepatic Encephalopathy: By closing the large portosystemic shunt, BRTO redirects portal venous blood back through the liver. This increases hepatic perfusion and allows for the clearance of ammonia and other neurotoxins, often leading to a significant improvement in encephalopathy.
  • Preservation of Hepatic Function: Unlike TIPS, which diverts portal flow away from the liver parenchyma, BRTO restores it. This can help preserve or even improve liver function in some patients.

In contrast, other therapies are rated lower for this specific scenario. A Transjugular Intrahepatic Portosystemic Shunt (TIPS) is rated as May be appropriate. While TIPS effectively decompresses the portal system, it does so by creating a new shunt, which can worsen hepatic encephalopathy—a major concern in a patient who is already encephalopathic. Endoscopic management (e.g., glue injection) is also Usually appropriate for acute hemostasis but is often less durable than BRTO for long-term prevention of re-bleeding from large, high-flow varices.

What’s Next After BRTO? Downstream Workflow

The post-procedure workflow focuses on confirming technical success, monitoring for complications, and managing the hemodynamic consequences of shunt closure.

If the BRTO is technically successful: The immediate next step is close monitoring in an intensive care or step-down unit. The patient should be observed for signs of re-bleeding, though this is uncommon after a successful procedure. A follow-up CT or MRI is typically performed weeks to months later to confirm complete thrombosis of the varices and the shunt. The patient’s mental status should be closely monitored for expected improvement in encephalopathy. Management of their underlying liver disease with a hepatologist continues to be paramount.

If BRTO is technically unsuccessful or not feasible: In a small number of cases, the anatomy of the shunt may be too tortuous or complex to catheterize safely. If BRTO fails, the next step is to reconsider the alternatives. A TIPS procedure (May be appropriate) becomes a primary consideration, accepting the risk of worsening encephalopathy in exchange for portal decompression and bleeding control. A multidisciplinary discussion involving interventional radiology, hepatology, and potentially transplant surgery is essential to weigh the risks and benefits for the individual patient.

If new or worsening ascites or esophageal varices develop: A known consequence of closing a large decompressive shunt like a gastrorenal shunt is a potential increase in portal pressure. This can lead to the formation or worsening of ascites or the development of esophageal varices. Patients must be monitored for this possibility long-term. If significant, this may require diuretics for ascites or endoscopic surveillance and banding for new esophageal varices.

Pitfalls to Avoid (and When to Get Help)

Navigating this clinical scenario requires careful attention to specific details to avoid common errors.

  • Misattributing all encephalopathy to liver failure: In this scenario, the large shunt is a major, treatable contributor to encephalopathy. Not recognizing this can lead to choosing a therapy like TIPS that could worsen the condition.
  • Delaying definitive treatment after initial endoscopic control: While endoscopy can stop acute bleeding, the risk of re-bleeding from large gastric varices is very high. A definitive plan for obliteration (like BRTO) should be made urgently.
  • Failing to monitor for post-BRTO portal hypertension effects: Do not assume the patient is “cured” after BRTO. Institute surveillance for the development of ascites and esophageal varices, which can result from redirecting flow back into the hypertensive portal system.

If the patient has refractory bleeding despite initial measures or if there is uncertainty about the optimal interventional strategy, immediate consultation with an experienced interventional radiologist and a transplant hepatologist is critical.

Related ACR Topics and Tools

For a comprehensive overview of all clinical variants related to this topic, please consult the parent guide. Additional GigHz resources can help you apply these criteria in your practice, from looking up adjacent scenarios to understanding the technical details and safety considerations of the recommended procedures.

Frequently Asked Questions

Why is BRTO preferred over TIPS when the patient already has hepatic encephalopathy?

BRTO is preferred because it closes the large portosystemic (gastrorenal) shunt, which is a major contributor to hepatic encephalopathy. By forcing portal blood back through the liver, BRTO can actually improve encephalopathy. In contrast, a TIPS procedure creates a new shunt, which can worsen the systemic load of neurotoxins and exacerbate encephalopathy.

What if the patient’s MELD score was much higher, like 25?

A much higher MELD score (e.g., >25) indicates more severe liver dysfunction and a poorer prognosis. In this case, the risks of any procedure are higher. The decision becomes more complex, and a TIPS may be reconsidered as a bridge to liver transplantation, as controlling the bleeding becomes the most urgent priority, even at the risk of worsening encephalopathy.

Can BRTO be performed if the patient has a different type of shunt, not a gastrorenal one?

Yes, variations of the BRTO procedure, sometimes called CARTO (Coil-Assisted Retrograde Transvenous Obliteration) or PARTO (Plug-Assisted…), can be used for other large shunts like splenorenal or gastrocaval shunts. The key requirement is a technically accessible, large-caliber shunt that is responsible for feeding the gastric varices.

What are the main risks associated with the BRTO procedure itself?

The primary risks of BRTO include sclerosant-induced hemoglobinuria and potential renal dysfunction, though this is often transient. There is also a risk of non-target embolization of the sclerosant. A significant long-term consequence is the potential for increased portal pressure after shunt occlusion, which can lead to the development or worsening of ascites or esophageal varices.

If the initial MRI didn’t show a shunt, would BRTO still be an option?

No. The presence of a large, definable portosystemic shunt (like a gastrorenal shunt) is a prerequisite for performing BRTO. If no such shunt exists, the varices are being fed by other pathways, and a retrograde approach is not feasible. Management would then rely on other options like endoscopic therapy or TIPS.

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