PARTO Procedure for Gastric Varices and Encephalopathy

Plug-assisted retrograde transvenous obliteration, known as PARTO, is a minimally invasive procedure used to treat dangerous enlarged veins (varices) in the stomach and, in some cases, a confusion-causing condition called hepatic encephalopathy. It works by threading a small vascular plug through the venous system to block the abnormal blood vessel that feeds the varix, then filling the varix with an embolic material so it clots off permanently. Developed as a faster, simpler alternative to an older balloon-based technique, PARTO has become a go-to option for interventional radiologists treating patients with portal hypertension, the elevated blood pressure in the liver’s venous system that drives these complications.

Why Gastric Varices Need a Dedicated Procedure

When the liver becomes scarred from cirrhosis or other chronic disease, blood has trouble flowing through it. Pressure builds in the portal vein, the major vessel that carries blood from the gut to the liver, and the body compensates by rerouting blood through alternative pathways. Some of those detour routes pass through the wall of the stomach, where the veins balloon outward into gastric varices. These swollen veins sit just beneath the stomach lining and can rupture without warning, causing life-threatening bleeding.

Gastric varices are trickier to manage than the more common esophageal varices. Endoscopic band ligation, which works well in the esophagus, is less reliable in the stomach because of the anatomy. Injecting a tissue glue through an endoscope can stop active bleeding, but the varices often come back. That gap in treatment is where procedures like PARTO fit in: rather than treating the varix from inside the stomach, they approach it from outside, through the veins themselves.

How the Procedure Works

The anatomy behind PARTO starts with the natural drainage routes that gastric varices use. Roughly 80 to 85 percent of fundal gastric varices drain into the left renal vein through a channel called the gastrorenal shunt. Another 10 to 15 percent drain directly into the inferior vena cava through a gastrocaval shunt.¹ Less commonly, drainage flows through the pericardial vein into the left brachiocephalic vein.² PARTO takes advantage of these outflow routes by entering them “retrograde,” meaning the radiologist goes against the direction of blood flow to reach and seal off the varix.

In a typical case, the radiologist accesses the venous system through a catheter inserted in the neck or groin. The catheter is navigated under fluoroscopic (live X-ray) guidance into the draining vein, usually the left renal vein and then into the gastrorenal shunt. Once the catheter tip is positioned in the outflow channel near the varix, a vascular plug, most often an Amplatzer vascular plug, is deployed. This small, expandable metal device lodges in the vein and acts as a dam.³

With the plug blocking the outflow, the radiologist injects a gelfoam slurry, a mixture of absorbable gelatin sponge particles and contrast dye, back through the catheter. The slurry fills the shunt and the varix itself, packing the entire abnormal channel with material that promotes clotting. Over days and weeks, the body forms scar tissue around the gelfoam, permanently sealing the varix shut.

What Makes PARTO Different from BRTO

PARTO is best understood as an evolution of balloon-occluded retrograde transvenous obliteration, or BRTO, which was developed in Japan in the 1990s. BRTO uses an inflated balloon catheter to block the outflow vein, then fills the varix with a liquid sclerosant, a chemical that destroys the vein lining. The balloon must stay inflated for hours, sometimes overnight, while the sclerosant dwells in the varix. That means the patient is tethered to the catheter in a hospital bed, often for six hours or more, and the sclerosant itself can leak past the balloon and cause damage to surrounding tissue.

PARTO swaps the balloon for a vascular plug and replaces the liquid sclerosant with a gelfoam slurry. Those two changes simplify the procedure considerably. The plug provides a more reliable seal than an inflated balloon, so there is less risk of the embolic material escaping where it should not go. And because gelfoam does not need a prolonged dwell time the way sclerosants do, the procedure can be completed faster. A comparative study found that PARTO procedure time was significantly shorter than conventional BRTO with either ethanolamine oleate or sodium tetradecyl sulfate foam.⁴ Beyond speed, PARTO and its plug-based relatives avoid the complications linked to sclerosant chemicals, including allergic reactions, hemolysis (destruction of red blood cells), and kidney injury.⁵

In head-to-head comparisons, gastric variceal rebleeding rates after BRTO and PARTO have not shown a statistically significant difference, suggesting similar long-term effectiveness.⁶ The practical advantages in speed and complication avoidance have led many interventional radiology teams to favor PARTO, and some experts now consider it a more efficient first-line choice.⁷

Success Rates and Long-Term Durability

Technical success in PARTO, meaning the plug fully blocks the outflow vein and the gelfoam completely fills the varix, is high. In one series of 19 patients, complete occlusion was achieved in about 95 percent of cases.⁸ A Gastroenterology review cited an obliteration rate of roughly 99 percent across accumulated data.⁹ These numbers reflect immediate procedural success, but the more important question for patients is whether the varices stay gone.

Long-term follow-up data are encouraging. A study tracking patients after PARTO found that about 5.6 percent developed recurrent gastric varices, though these recurrences appeared at a different location than the originally treated varix. The cumulative freedom from variceal recurrence was roughly 97 percent at one year, 94 percent at three years, 92 percent at five years, and 85 percent at ten years.¹⁰ Those are strong numbers for a population dealing with progressive liver disease, where the underlying portal hypertension never fully goes away. A prospective multicenter study concluded that PARTO could be performed rapidly with high technical success and durable clinical results, and suggested it might serve as a first-line treatment in appropriate patients.¹¹

Using PARTO for Hepatic Encephalopathy

Gastric variceal bleeding is the most common reason for PARTO, but it is not the only one. The same portosystemic shunts that feed varices can also cause hepatic encephalopathy, a condition in which toxins like ammonia bypass the liver and reach the brain. Normally the liver filters ammonia out of blood returning from the gut. When large shunts divert that blood past the liver, ammonia builds up and causes symptoms ranging from subtle confusion and personality changes to severe disorientation and coma.

By plugging the shunt, PARTO redirects blood flow back through the liver, restoring the organ’s ability to clear ammonia. This approach has been effective in patients whose encephalopathy does not respond well to standard medical therapy with lactulose and rifaximin. In one reported case involving a right mesorenal shunt, an uncommon location, the patient’s ammonia levels dropped substantially within two weeks of PARTO, and no encephalopathy recurred during follow-up.¹² When patients have more than one shunt contributing to encephalopathy, PARTO can be combined with coil-assisted retrograde transvenous obliteration (CARTO) to close multiple pathways in a single session.

The Trade-Off With Portal Pressure

Closing off a portosystemic shunt is not a free lunch. The shunt exists because the body is trying to decompress an overloaded portal venous system. When you plug the escape route, portal pressure rises. That pressure increase can have real consequences for patients who already have portal hypertension.

The most commonly observed downstream effects are worsening of esophageal varices and the development or worsening of ascites, the accumulation of fluid in the abdomen. In an emergency series of nine patients treated with PARTO for active gastric variceal bleeding, four showed mild worsening of esophageal varices or transient ascites afterward.¹³ A broader review of transvenous obliteration techniques confirms that these procedures are associated with increased portal pressure and related complications, mainly worsening of esophageal varices.¹⁴

The good news is that for many patients, the liver function scores actually improve after PARTO. A study measuring portal pressure changes found that both the Child-Pugh score and MELD score, two widely used measures of liver disease severity, improved significantly by six months after the procedure.¹⁵ The likely explanation is that redirecting blood flow through the liver improves its ability to function, even if portal pressure goes up somewhat. This is a meaningful tradeoff: the varix risk drops, liver function improves, but esophageal varices need closer monitoring afterward.

Complications Beyond Portal Pressure

PARTO is generally well tolerated, but the procedure carries several acute complications worth knowing about. Fever is fairly common in the days after the procedure, likely a reaction to the gelfoam and the body’s inflammatory response to the clotting material. In one series, two out of nineteen patients had fever alone and two had fever combined with low blood pressure, one of whom was diagnosed with adrenal insufficiency, a recognized but uncommon complication that occurs when the plug or embolic material inadvertently affects the adrenal vein, which sits close to the gastrorenal shunt. Transient blood in the urine was also seen in a few patients.¹⁶

These side effects are generally manageable and resolve on their own, which is a notable improvement over BRTO. The sclerosants used in BRTO carry risks of hemolysis, disseminated intravascular coagulation, and renal toxicity that simply do not apply with gelfoam. That said, adrenal vein injury is something radiologists have to be attentive to during the procedure. The left adrenal vein drains into the left renal vein in the same neighborhood where the plug is being placed, so careful catheter positioning is essential.

Who Benefits Most and Who Faces Higher Risk

Not every patient with gastric varices is a good candidate for PARTO. The procedure depends on having a usable draining shunt, which is present in the vast majority of patients but not all. In those rare cases without a gastrorenal or gastrocaval shunt, alternative drainage pathways like pericardial veins can sometimes be used, though this is technically more challenging.¹⁷

Liver disease severity at the time of the procedure matters a great deal for prognosis. A multicenter ten-year retrospective study identified several factors independently associated with survival after PARTO or its coil-based cousin CARTO. Patients with a pre-procedure MELD score above 27 had a dramatically higher mortality rate of roughly 93 percent, compared to those with lower scores. Concurrent liver cancer, a history of esophageal variceal bleeding, and elevated bilirubin and ammonia levels also predicted worse outcomes.¹⁸ These findings underscore that PARTO treats a complication of liver disease, not the disease itself. In patients with very advanced cirrhosis or liver cancer, closing the varix may buy time, but the underlying liver failure drives overall survival.

PARTO in Emergency Bleeding

One question that matters clinically is whether PARTO can be performed safely during active hemorrhage, not just as an elective procedure. Gastric variceal bleeds are dramatic events: patients can lose large volumes of blood quickly, and the mortality rate from a single bleeding episode is high. A study specifically evaluating PARTO in the emergency setting found that technical and clinical success, meaning complete shunt occlusion and immediate bleeding control, was achieved in all nine patients treated. No procedure-related complications occurred, and all patients had successful outcomes over a mean follow-up of 17 months.¹⁹

This is a meaningful advantage. BRTO, with its requirement for prolonged balloon inflation and sclerosant dwell time, is poorly suited to an unstable, actively bleeding patient. PARTO’s shorter procedure time and mechanical rather than chemical approach make it more practical when speed matters. The fact that the plug provides a stable occlusion without requiring a balloon to stay inflated means fewer things can go wrong during a high-stress clinical scenario.

How PARTO Relates to TIPS

PARTO and TIPS (transjugular intrahepatic portosystemic shunt) approach the same problem from opposite directions. TIPS creates a new channel through the liver to decompress the portal system, lowering portal pressure and reducing the driving force behind variceal bleeding. Transvenous obliteration techniques like PARTO close the existing shunt and seal the varix, which actually increases portal pressure. The two strategies are sometimes complementary: transvenous obliteration can serve as an alternative or an addition to TIPS depending on the clinical situation, and it is less invasive.²⁰

The choice between TIPS and PARTO depends on the patient’s overall picture. TIPS is better at reducing portal pressure across the board, which helps prevent bleeding from esophageal varices and reduces ascites. But TIPS carries its own risks: the shunt can become blocked, it can trigger or worsen hepatic encephalopathy by diverting too much blood past the liver, and it requires a more invasive procedure through the liver itself. PARTO is the more targeted option when the primary problem is an isolated gastric varix or a portosystemic shunt causing encephalopathy, without diffuse portal hypertensive complications that need global pressure reduction.

CARTO and Other Variations

PARTO is not the only plug-era modification of BRTO. Coil-assisted retrograde transvenous obliteration, or CARTO, uses detachable metal coils instead of a vascular plug to block the outflow vein. The coils are packed into the shunt to create a scaffold, and gelfoam or other embolic agents are injected to fill the remaining spaces. Compared to PARTO, CARTO has the advantage of being able to close larger-diameter shunts, because multiple coils can be deployed to fill a wider vessel where a single plug might not seat properly.²¹

In practice, the choice between PARTO and CARTO often comes down to shunt anatomy. If the outflow vein is a reasonable diameter and has a landing zone where a plug can expand and anchor, PARTO is typically preferred for its simplicity. If the shunt is wide or has an irregular shape, coils may be a better mechanical fit. The two can also be used together: in patients with multiple portosystemic shunts, one might be closed with a plug and another with coils during the same procedure.²²

Hepatic Encephalopathy Without Cirrhosis

While cirrhosis is by far the most common cause of the portosystemic shunts that PARTO targets, shunts and hepatic encephalopathy can occur without cirrhosis. Congenital portosystemic shunts, portal vein thrombosis, and other forms of non-cirrhotic portal hypertension can all produce large enough shunts to cause clinically significant ammonia diversion. In these patients, standard medical management with lactulose, rifaximin, and dietary changes remains the first step, but shunt embolization using techniques like PARTO is a recognized option when medications are insufficient.

Non-cirrhotic patients may actually be better candidates for shunt closure in some ways, because their liver function is typically better preserved. Without the background of progressive cirrhosis, the portal pressure increase from closing the shunt is less likely to cause decompensation. The evidence base for PARTO in non-cirrhotic portal hypertension is smaller than for cirrhotic patients, consisting mostly of case reports and small series, so treatment decisions require careful multidisciplinary discussion. Still, the procedure’s favorable risk profile makes it an attractive option when the anatomy cooperates.