Accelerated balloon-retrograde transvenous obliteration (BRTO): an effective tool in the Arsenal against isolated gastric varices (IGV)

Abstract

Gastric and oesophageal variceal bleeding poses high morbidity and mortality in cirrhosis. Amongst all types, isolated gastric varices (IGV) carry the highest propensity to bleed. Successful outcomes combine endoscopic and interventional radiology approaches using ligation, coils, glue or sclerosants. Transjugular intrahepatic portosystemic shunt success is only seen in a subset of patients, while balloon-retrograde transvenous obliteration (BRTO) has demonstrated high efficacy in preventing rebleeding and morbidity in patients with a myriad of anatomies and shunts. The American Association for the Study of Liver disease guidelines do not favour any particular modality; however, recent trials and meta-analyses support BRTO as the first-line therapy. Despite promising results, BRTO adoption is limited by procedural time, patient length-of-stay and equipment compatibilities hindering scalability in academic and community settings. To address these concerns, we present a successfully treated case of IGV with a revised technique called accelerated BRTO.

Background

In the USA, over 4.5 million adults have cirrhosis and inpatient trends reveal an increasing burden of chronic liver disease on national hospitalisation costs.¹ A significant cause of cirrhosis-related morbidity and mortality is oesophageal or gastric variceal bleeding, accounting for 70% of all upper gastrointestinal bleeding, with each successive episode carrying a 20% mortality risk and 50% requiring an intervention.^{2 3} Gastric varices (GV) prevalence is noted to affect nearly half of all patients with cirrhosis. GV are differentiated into gastro-oesophageal varices (GOV) or isolated gastric varices (IGV) based on their relationship to oesophageal varices (EV). EV tracking into GV along the lesser and greater gastric curvatures are considered GOV type 1 and type 2. In contrast, IGV in the fundus are considered IGV 1, while their presence anywhere else in the stomach is labelled IGV 2.^{3 4} IGV 1 is rare (1.6%) but carries the highest propensity to exsanguinate, require transfusions and death.⁵

The American Association for the Study of Liver disease (AASLD) previously recommended transjugular intrahepatic portosystemic shunt (TIPS) as an effective first-line modality for acute presentations and its adoption has been popularised in the USA and Europe.^{6 7} However, studies have suggested that the net benefit in TIPS is only seen in a subset of patients with small shunts or an IGV complex fed by left gastric veins, requiring alternative solutions. Over the years, balloon-retrograde transvenous obliteration (BRTO) has gained widespread acceptance, especially in South Korea and Japan. In the setting of rebleeding prevention, the 2016 AASLD guidance statement does not favour any particular modality with updates currently in development.^{6 7} A recent randomised trial and meta-analysis comparing BRTO to endoscopic cyanoacrylate or TIPS found effective and safer outcomes at 1 year, including prevention, rebleeding and postprocedure hepatic encephalopathy, suggesting that BRTO should be considered the first-line treatment of GV.^8–10

However, BRTO does have its limitations that have hindered its scalabilities, such as equipment compatibility, prolonged length of stay, and balloon complications. Many approaches and variations have been tried, although the core feature of the procedure remains balloon occlusion for up to 24 hours. We present a case of a successfully treated IGV with a technical modification to the conventional procedure known as accelerated BRTO.

Case presentation

A 59-year-old woman presented to our emergency department with haematemesis and melenic stools for 4 days with a history significant for cirrhosis from chronic hepatitis C, chronic thrombocytopaenia, gastro-oesophageal reflux disease and hypertension on propranolol and ranitidine. She denied using analgesics, iron supplements or bismuth. Her model for end-stage liver disease (MELD-Na) score was 9, and this was her first episode of GIB. On arrival, she was haemodynamically stable, appearing pale with a haemoglobin of 9.2 g/L (baseline 12.3 g/L), platelets 65 x 10⁹ k/μL and acute renal failure. Intravenous hydration, esomeprazole and octreotide infusions were initiated with ceftriaxone and pre-emptive blood transfusion. She was admitted to the intensive care unit and underwent an urgent esophagogastroduodenoscopy (EGD) that revealed old blood with adherent clots, two large gastric fundal varices and no evidence of EV or ulcers (figure 1). A CT of the abdomen with contrast revealed a liver with cirrhosis, 13.6 cm spleen and prominent gastric fundal varices (IGV type 1), and a gastrorenal shunt (GRS) as the efferent channel to the left upper renal vein (figure 2). A multidisciplinary meeting recommended accelerated balloon-occluded retrograde transvenous obliteration as an outpatient, as evidence of liver disease progression was seen despite completion of direct-acting antivirals and evidence of sustained viral response over the preceding years. She was discharged on high-dose esomeprazole (ranitidine was discontinued) and propranolol.

Figure 1.

An esophagogastroduodenoscopy revealing two large fundal gastric varices (GV).

Figure 2.

Preprocedural axial CT of the abdomen revealing prominent fundal gastric varices (GV).

Outcome and follow-up

Retrograde access to the gastric variceal complex was performed through the internal jugular vein with the positioning of the catheter at the distal GRS, while anterograde access was achieved through the percutaneous transhepatic route (figure 3). The GRS was embolised using metallic coils, and the GV were obliterated using sclerosant injection via the posterior gastric vein with a foam (mixture of one part lipoidal, two parts sodium tetradecyl sulfate 3% and three parts air). Fluoroscopy and final CT demonstrated adequate deposition of the sclerosant within the variceal complex (figure 4). After assuring signs in the recovery area, she was discharged home. In 2 weeks, repeat EGD confirmed no blood or EV with shrinkage in the size of IGV (figure 5). Repeat MELD-Na score was 7, and tissue examination was negative for Helicobacter pylori. At her 4-month follow-up, she denied any bleeding episodes, and her CT angiography confirmed previous complete IGV obliteration, portal vein patency, coils, no thrombi or evidence of worsening effusions, ascites or varices (figure 6).

Figure 3.

Accelerated balloon-retrograde transvenous obliteration overview—GV, GRS, LPV, PV, SV and LRV. Here a microcatheter is fed into the posterior gastric vein to deliver the sclerosant mixture into the GV complex. GRS, gastrorenal shunt; GV, gastric varices; LPV, left portal vein; LRV, left renal vein; SV, splenic vein; PV, portal vein.

Figure 4.

Postprocedure film shows complete obliteration of the GV with the sclerosant mixture with simultaneous occlusion of the distal GRS by metallic coils (arrowhead). GRS, gastrorenal shunt; GV, gastric varices.

Figure 5.

Follow-up esophagogastroduodenoscopy in 2 weeks demonstrating reduced gastric varices filled with sclerosant (arrowhead).

Figure 6.

Follow-up CT of the abdomen at 4 months demonstrating complete obliteration of the gastric variceal complex (arrowhead).

Discussion

BRTO achieves success in 75%–100% of cases with a low rebleeding rate, and initial concerns of efficacy were readdressed in recent studies demonstrating superior 5-year survival in patients undergoing BRTO for GV management.^8–11 Compared with TIPS, BRTO is technically feasible and less prone to puncturing the portal vein.⁹ It involves obliterating shunts and reversing portal pressures, leading to improved liver functioning; hence, gastric variceal bleeding and refractory encephalopathy remain its primary indication. Despite BRTO becoming the first line for GV management, community adoption has been limited due to technical and logistical challenges. Additionally, BRTO use may have been limited as historical models found no change in Child-Pugh scores (CPS); however, CPS does not include creatinine which is commonly deranged, under-reporting BRTO efficacy. We highlight a technical modification to the conventional BRTO known as ‘accelerated BRTO’ described by Darwish et al.

Significant IGV bleeding predictors include stigmata of bleeding, location, size (>10 mm), the severity of cirrhosis, hepatocellular carcinoma and the presence of red wale streaks.² Management of active GV bleeding includes prompt resuscitation, correcting coagulopathy, antibiotics, vasoactive measures and early endoscopy. Endoscopic band ligation is not recommended in IGV, as failure to capture the affected segments can lead to uninterrupted blood loss.⁴ TIPS was not offered to our patient as she did not have EV or ascites and the feasibility of conventional BRTO was limited due to equipment compatibility at our community hospital. The novelty of the procedure was explained to the patient and all of her questions were answered. She tolerated the procedure well and was discharged the same day. She denied any abdominal pain at her 2-week and 4-month follow-up, and her MELD-Na had lowered by 2 points.

A significant limitation in BRTO data within the existing literature is small sample sizes, heterogeneity in baseline liver function of patients and selection biases among technical and logistical challenges.8 9 12 13 To address known limitations of conventional BRTO, we replicated the technique by Darwish et al, where a modification to ‘accelerated BRTO’ demonstrated similar efficacy to conventional BRTO. The modification involves using vascular embolic coils to occlude the shunt, thereby avoiding balloon occlusion for prolonged periods leading to reduced periprocedural morbidity, fewer parts and mixtures required and quicker recovery time. The contrast/gel foam use via the microcatheter is the same in BRTO and coil-assisted retrograde transvenous obliteration (CARTO).¹² This technique uses principles and components of CARTO and the shorter occlusion times demonstrated by Darwish et al and can be performed in most community hospitals such as ours.¹⁴ In cases where IR-guided procedures are limited, endoscopic ultrasound-guided combination coil and sponge injections have shown success.¹⁵

IGV is associated with a high mortality rate and a multidisciplinary approach is essential for a successful outcome. Careful selection for endoscopic or vascular intervention requires a global assessment of these patients with individualised surveillance schedules. Limitations include patients with significant ascites or EV. Future studies are needed to explore the efficacy of accelerated BRTO against other interventional radiology or endoscopic interventions. We additionally performed a systematic search of prior reports of IGV treated by BRTO—table 1.

Table 1.

Summary of IGV cases that underwent traditional BRTO in the last 10 years (2011–2021)

Author	Country	Indication (prophylaxis, acute management, or rebleeding)	Substance used	Afferent	Efferent	Approach—jugular (IJ), subclavian, saphenous, or femoral	Baseline MELD-Na/Child-Pugh-Hugh	Shunt	Outcomes (short-term, long-term and mortality)
Prior to <201116–28		N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A
Sonomura et al 201129	Japan	Prophylaxis	5% mixture of ethanolamine oleate iopamidol (EOI)	Left renal vein	N/A	R IJ	Pugh-Child class B	Gastrorenal	Minor ascites, left pleural effusions and high fever, varices were gone
Yoshimatsu et al 201130	Japan	Prophylaxis	5% EOI	Left pericardiophrenic	N/A	Left subclavian	Pugh-Child class A	No gastrorenal shunt	Disappearance 1 week, reduction on scope 3 weeks
Janne d'Othée et al 201231	USA	Prophylaxis	Sodium tetradecyl sulfate 3% + contrast	Left inferior phrenic vein (IPV)	N/A	IJ	MELD-Na 12	Gastrorenal	Focal RUQ pain 1 day, complete thrombosis of GV 1 week, haematemesis from EV 2 months
Miura et al 201232 Case 1	Japan	Prophylaxis	5% EOI, 10 mL 50% glucose, NBCA and lipiodol	Left IPV	Azygous	Femoral	N/A	N/A	GV occluded 1 week
Miura et al 201232 Case 2	Japan	Prophylaxis	5% EOI	Left IPV	N/A	Femoral	N/A	No gastrorenal	GV occluded 1 week
Yamagami et al 201433	Japan	Rebleeding	50% glucose + 5% EOI	Left preaortic renal vein	Two branches of left IPV, left pericardiophrenic vein	Femoral	Pugh-Child class A	Gastrorenal	Complete disappearance 1 month
Basseri and Lightfoot 201634	Canada	Rebleeding	3% sodium tetradecyl sulfate foam	IPV	Posterior gastric vein	Femoral	Pugh-Child class A	Gastrorenal	Uncomplicated recovery 4 days, obliteration 3 months, no GV but EV 10 months
Mine et al 201635	Japan	Prophylaxis	5% EOI + 33% mixture of NBCA (glue)	Left IPV	Anastomotic branches to azygous	Femoral	Pugh-Child class A	Gastrorenal	Reduction 2 months
Bundy et al 201836	USA	Acute management	1:1 Ethiodised oil + sodium tetradecyl sulfate	N/A	N/A	N/A	MELD-Na 30	Gastrorenal	Migrated plug into the left pulmonary artery
Srinivasa et al 201837	USA	Rebleeding	3% sodium tetradecyl sulfate, ethiodised oil and air	Unclear	Unclear (created tract between inferior vena cava and superior mesenteric vein)	Right saphenous vein	N/A	Splenorenal (created mesocaval shunt)	No bleeding or encephalopathy 1 month
Ravi et al 201938	USA	Acute management	N/A	Mid-splenic vein	N/A	N/A	Pugh-Child class A MELD-Na 10	Splenorenal shunt	Completely obliterated (no timeframe)

BRTO, balloon-retrograde transvenous obliteration; EV, oesophageal varices; GV, gastric varices; IPV, inferior phrenic vein; MELD-Na, model for end-stage liver disease; NBCA, N-butyl-2-cyanoacrylate.

Learning points.

• The net benefit in transjugular intrahepatic portosystemic shunt (TIPS) is only seen in a subset of patients with small shunts or an isolated gastric varices (IGV) complex fed by left gastric veins. In contrast, balloon-retrograde transvenous obliteration (BRTO) is slowly gaining widespread acceptance.

• Recent randomised trials and meta-analysis comparing BRTO to endoscopic cyanoacrylate or TIPS found compelling and safer outcomes at 1 year including prevention, rebleeding and postprocedure hepatic encephalopathy.

• BRTO should be considered the first-line treatment of GV. However, equipment compatibility, prolonged length of stay and balloon complications have hindered its capabilities.

• A modified technique, accelerated BRTO, can reduce procedure-related time and morbidity and be easily adaptable in academic and community settings with the existing hardware, allowing seamless adaptability.

Ethics statements

Patient consent for publication

Obtained.