Liquid-Embolic Assisted Retrograde Transvenous Obliteration of Bleeding Gastric Varices: The LARTO Technique

Abstract

Balloon-occluded retrograde transvenous obliteration (BRTO) has been an established technique for treating bleeding gastric varices. However, the traditional BRTO technique requires prolonged balloon inflation and can be resource-intensive. Advanced variations of classic BRTO technique using coils (CARTO) or plugs (PARTO) have been described, but can be cost prohibitive. The authors report a novel variation – liquid-embolic assisted retrograde transvenous obliteration (LARTO) – utilizing the LAVA liquid embolic system to occlude the outflow vein during variceal obliteration, instead of coils or plugs.

Keywords

BRTO - LARTO - liquid-embolic assisted retrograde transvenous obliteration - LAVA - liquid embolic

The balloon-occluded retrograde transvenous obliteration (BRTO) technique has been used for over 25 years for the management of gastric varices. Recent advancements have introduced variations such as coil-assisted retrograde transvenous obliteration (CARTO) and plug-assisted retrograde transvenous obliteration (PARTO) to reduce procedure time and improve efficacy. The authors present a novel variation, liquid-embolic assisted retrograde transvenous obliteration (LARTO), which uses a liquid embolic system (LES) in place of coils or vascular plugs to achieve durable occlusion during variceal obliteration.

A 76-year-old male with hypertension, hyperlipidemia, and diabetes, but no history of liver disease, presented to the emergency department with hematemesis and several days of melena. Contrast-enhanced computed tomography (CT) of the abdomen revealed cirrhosis with gastric varices without paraesophageal varices, splenomegaly, or ascites ([Fig. 1]). Upper gastrointestinal (GI) endoscopy confirmed the imaging findings of gastric varices without active hemorrhage, but clot was seen over the varices indicating recent bleeding. The patient was referred to interventional radiology (IR) for further management of bleeding gastric varices, per our institutional protocol. Pertinent laboratories included hemoglobin of 7.7 g/dL, platelet count of 126 × 10^9/L, prothrombin time/international normalized ratio 11.9/1.14, and Model for End-Stage Liver Disease score of 11. Closer review of the CT showed a type A shunt (based on venous drainage), with a splenorenal shunt draining through an enlarged inferior phrenic vein into the left renal vein and inferior vena cava. The BRTO procedure was performed using a right internal jugular vein approach and 6 French Ansel 45 cm sheath was placed into the left renal vein. The outflow vein (draining into the left renal vein) was catheterized using 5 French C2 catheter and a J-Rosen wire. A balloon-occlusion catheter (Fogarty 5.5 French, McKesson Corporation, New York City, New York, United States) was deployed occluding the outflow varix, and venography was performed demonstrating opacification of the gastric varices ([Fig. 2]). After confirming complete outflow occlusion with balloon-occlusion catheter (evidenced by no leakage of contrast into the left renal vein), a 2.4-French Progreat microcatheter (Terumo Corporation, Tokyo, Japan) was advanced distally into the varix. The sclerosant foam was made using the widely used Tessari method, using 15 mL air, 10 mL of 3% sodium tetradecyl sulfate, and 5 mL Omnipaque 350 contrast in a 3:2:1 ratio. To provide additional embolic effect to the sclerosant, the foam was then mixed with Embocube (uniformly cut gelatin foam, 2.5 mm × 50 mg, Merit Medical, Jordan, Utah, United States). The sclerosant was injected through the microcatheter into the varices, while constantly mixing the foam to prevent settling. Intraprocedural cone-beam CT confirmed good sclerosant distribution within the gastric varices ([Fig. 3]). After the entire sclerosant mix was injected, the microcatheter was removed and a new 2.4 French Progreat microcatheter was placed into the outflow vein, and primed with 0.6 mL dimethyl sulfoxide (DMSO). The LAVA liquid embolic (Liquid Embolic for Vascular Applications; SIR Tex, Woburn, Massachusetts, United States) was slowly injected to occlude the outflow vein while the microcatheter was retracted to the tip of the balloon catheter. The LAVA 36 viscosity variant was utilized and a total of 1.1 mL embolic volume was used. Complete occlusion of the outflow vein was confirmed by slowly deflating the balloon and absence of contrast leakage into the renal vein. Gentle contrast injection was performed showing effective occlusion of the outflow vein ([Fig. 4]), and the catheter was removed. The patient was discharged the following day without recurrent upper GI bleeding. Most recent 6-month follow-up with the gastroenterology outpatient service showed no recurrent episode of upper GI bleeding. The patient was ultimately diagnosed with metabolic dysfunction-associated steatotic liver disease.

Fig. 1 (A) Contrast-enhanced axial computed tomography (CT) of the abdomen demonstrates enhancing collateral vessels in the gastric region (arrow) consistent with varices. Note the nodular margins of the liver indicating cirrhotic morphology. No ascites, splenomegaly, or periesophageal varices were seen. (B) Coronal 10 mm maximum intensity projection (MIP) reconstruction from the CT showing the outflow vein (long arrow) and the gastric varices. This was used for procedural planning.

Fig. 2 Right internal jugular vein approach venography with the sheath in the left renal vein (black arrow), the tip of the balloon-occlusion catheter in the outflow vein (small white arrow), microcatheter beyond the base catheter, and contrast within the gastric varices (long white arrows).

Fig. 3 Intraprocedural cone-beam computed tomography (CT) after the injection of sclerosant demonstrates opacification of the varices (arrows) confirming appropriate distribution of the sclerosant.

Fig. 4 Fluoroscopic image after the injection of liquid embolic material into the outflow vein (arrows) and deflation of the occlusion balloon demonstrating static contrast within the gastric varices, confirming complete occlusion.

The classic BRTO technique requires prolonged balloon inflation for 4 to 24 hours after sclerosant injection to maximize sclerosant contact with the variceal endothelium.[1] [2] However, this prolonged duration can be resource-intensive and challenging for busy IR departments. Advanced variations like CARTO and PARTO address these limitations by using coils or vascular plugs to occlude the outflow vein, enabling earlier removal of the balloon catheter, thereby reducing the procedure time, while also preventing the risk of systemic sclerosant leakage. Nonetheless, these methods also have technical challenges, including difficult deployment due to anatomical factors and potential complications related to these devices.[3] [4] N-butyl cyanoacrylate glue, although a widely used liquid embolic in embolization procedures, polymerizes rapidly, carries a higher risk of nontarget embolization and catheter entrapment. LAVA, similar in composition to onyx, is a nonadhesive, radiopaque liquid embolic agent composed of ethylene vinyl alcohol copolymer dissolved in DMSO. The LAVA liquid embolic allows controlled delivery, enabling precise and durable occlusion of the outflow vein while minimizing embolic migration. This makes LAVA a safer and more effective choice in the LARTO technique. The authors utilized the LAVA LES,[5] which provided a durable occlusion of the outflow varix, thereby enabling immediate removal of the balloon-occlusion catheter after sclerosant injection, resulting in reduced procedural time while also allowing sufficient dwell time for sclerosant. The LARTO technique using the LAVA embolic system is potentially cost-effective since plug- or coil-assisted techniques require multiple coils or vascular plugs to completely occlude the outflow varix, which can add up the cost, whereas only small amount of LAVA (one vial of 2 mL LAVA 36) was enough in this case to form an occlusive plug. However, the amount of liquid embolic needed will depend on the size of the outflow varix and therefore careful review of preprocedural imaging is crucial for success with this technique. With the increasing availability of liquid embolic agents such as LAVA, the LARTO technique has the potential to offer another advanced alternative to the classic BRTO for patients with bleeding gastric varices. Further studies are warranted to evaluate the broader applicability and long-term efficacy of LARTO technique as a modification of BRTO.

Conflict of Interest

None declared.

Ethical Approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

• References

• 1 Waguri N, Osaki A, Watanabe Y. Balloon-occluded retrograde transvenous obliteration for treatment of gastric varices. World J Hepatol 2021; 13 (06) 650-661
• 2 Saad WE, Kitanosono T, Koizumi J, Hirota S. The conventional balloon-occluded retrograde transvenous obliteration procedure: indications, contraindications, and technical applications. Tech Vasc Interv Radiol 2013; 16 (02) 101-151
• 3 Mukund A, Anandpara KM, Ramalingam R, Choudhury A, Sarin SK. Plug-assisted retrograde transvenous obliteration (PARTO): anatomical factors determining procedure outcome. Cardiovasc Intervent Radiol 2020; 43 (10) 1548-1556
• 4 Kim DJ, Darcy MD, Mani NB. et al. Modified balloon-occluded retrograde transvenous obliteration (BRTO) techniques for the treatment of gastric varices: vascular plug-assisted retrograde transvenous obliteration (PARTO)/coil-assisted retrograde transvenous obliteration (CARTO)/balloon-occluded antegrade transvenous obliteration (BATO). Cardiovasc Intervent Radiol 2018; 41 (06) 835-847
• 5 Arslan B, Razavi MK, Siskin G. et al; LAVA investigators. The LAVA study: a prospective, multicenter, single-arm study of a liquid embolic system for treatment of peripheral arterial hemorrhage. J Vasc Interv Radiol 2025; 36 (03) 436-445.e2

Publikationsverlauf

Artikel online veröffentlicht:
29. September 2025