Safety and efficacy of low-cost, off-label balloons during balloon-occluded retrograde transvenous obliteration procedure

Abstract

Objectives:

The objective of the study is to assess the technical feasibility, safety, and clinical efficacy of balloon-occluded retrograde transvenous obliteration (BRTO) using low-cost, off-label balloon catheters.

Material and Methods:

A retrospective study was conducted on 31 patients (20 males, 11 females; mean age 56.6 years; range 47–75 years) who underwent BRTO between January 2009 and November 2023. All patients had gastric varices with a demonstrable gastrorenal shunt on pre-procedural computed tomography; six of them presented with intractable hepatic encephalopathy. Off-label balloon catheters (Swan-Ganz, Fogarty, or stone extractor) were selected based on shunt anatomy. Sodium tetradecyl sulfate foam, with or without Lipiodol, was used as the sclerosing agent. Technical success was defined as sustained sclerosant stasis during balloon inflation. Clinical success was defined as cessation of variceal bleeding or resolution of encephalopathy.

Results:

Technical success was achieved in 29 of 31 patients (93.5%). Balloon rupture or deflation occurred in two patients; however, hemostasis was successfully maintained without any rebleeding. Of the 25 patients with follow-up data (mean 953.6 days; range 1–4015 days), clinical success was observed in 23 (92%). One patient required surgical shunt creation for immediate post-procedure bleeding, and another experienced rebleeding after 1 year, which was successfully managed with a transjugular intrahepatic portosystemic shunt. Among 17 patients with follow-up imaging or endoscopy, complete obliteration of gastric varices was observed in 70.5% and partial obliteration in 29.5%. Balloon-related complications were rare and not associated with adverse clinical outcomes.

Conclusion:

BRTO using inexpensive off-label balloon catheters is a technically feasible, safe, and clinically effective treatment for gastric varices and hepatic encephalopathy. In combination with a modified Lipiodol-free sclerosant protocol, this technique offers a cost-effective alternative to conventional BRTO in resource-constrained environments. Prospective studies are needed to confirm these findings and to develop standardized clinical guidelines.

INTRODUCTION

Gastric varices occur in about 20% of individuals with portal hypertension. While gastric varices bleed less often than esophageal varices, they tend to bleed massively and carry a higher risk of rebleeding and mortality.^[1] In Western countries, the transjugular intrahepatic portosystemic shunt (TIPS) is often used to treat refractory gastric variceal bleeding. However, the TIPS procedure may not completely resolve the varices even when combined with gastric variceal embolization. In addition, TIPS is sometimes not preferred as the primary treatment because of the risk of worsening hepatic encephalopathy.^[2]

The majority of the gastric fundal varices typically drain into the left renal vein through a gastro-renal shunt.^[3] Based on this understanding, a technique has been developed that involves injecting sclerosing agents retrogradely into the gastric varices. This endovascular procedure is called balloon-occluded retrograde transvenous obliteration (BRTO). It is also an effective treatment option for sclerosis of spontaneous portosystemic shunts complicated by hepatic encephalopathy.^[4]

The idea of using transvenous obliteration to address varices resulting from portal hypertension was conceived in the 1970s. The concept of retrograde obliteration via a gastro-renal shunt through the left renal vein was initially proposed by Olson et al.^[2,5] However, it was the Japanese researchers, Kanagawa and others, who significantly refined the clinical use of BRTO as a therapeutic adjunct or alternative to TIPS in the management of the gastric varices.^[2]

The procedure details and outcomes of BRTO are well documented in the literature.^[4,6-8] The approach and complexity of the procedure are primarily influenced by the anatomic classification of the inflow and outflow veins of the varix. A key aspect of the procedure is ensuring that the sclerosant remains within the varix or shunt without backflow into the portal circulation or antegrade flow into the systemic circulation, which can potentially lead to serious complications. To avoid these complications, occlusion balloons are strategically placed to control the flow within the varix or shunt.^[6] Several occlusion balloon catheters and combinations of sclerosants have been documented in the literature.^[4] In Asia, reverse-shaped balloon catheters are preferred because they provide easier, more stable access to the shunt; however, these are not widely available in the United States. In the U.S., occlusion balloons, including the Python occlusion balloon (Applied Medical, Rancho Santa Margarita, California), the Equalizer (Boston Scientific, Natick, Massachusetts), standard occlusion balloon catheters (Boston Scientific, Natick, Massachusetts), flow-directed occlusion balloon catheters (Cook Inc., Bloomington, Indiana), and the Coda (Cook Inc., Bloomington, Indiana), are utilized. Dedicated BRTO balloons are often expensive and not readily available in many low and middle-income countries. In resource-constrained settings, this may limit access to BRTO despite proven efficacy. The use of a repurposed low-cost balloon catheter, therefore, represents a practical and potentially scalable alternative. In our institute, we have utilized low-cost, off-label balloons such as Swan-Ganz catheter (Edwards Lifesciences, Puerto Rico, United States of America), Fogarty thru-lumen embolectomy catheter (Edwards Lifesciences, Puerto Rico, United States of America), and stone extractor balloon catheters (Olympus Hong Kong China, Ltd., New Territories, Hong Kong) in BRTO for gastric varices. The primary purpose of these balloons is pulmonary arterial pressure measurement, arterial mechanical thrombectomy, and biliary stone removal through an endoscope, respectively. However, published evidence on the safety, efficacy, and outcomes of these alternatives is scarce. This study was therefore undertaken to evaluate the feasibility of using low-cost balloons for BRTO, to establish practical, affordable options without compromising clinical outcomes.

MATERIAL AND METHODS

The study was approved by our institutional review board. It is a retrospective single-center study to assess the safety and efficacy of off-label, low-cost balloons to perform BRTO. Between January 2009 and November 2023, 31 out of 35 patients with refractory gastric variceal bleeding or hepatic encephalopathy having a gastro-renal venous shunt on computed tomography (CT) angiography (SOMATOM Drive, Siemens Healthcare GmbH, Germany) underwent BRTO, 3 patients underwent plug-assisted BRTO. One patient did not undergo any endovascular procedure because of numerous efferent collateral veins. All 31 patients had clinical and endoscopic evidence of gastric varices and had a patent gastro-renal shunt on preprocedural contrast-enhanced CT (CECT) [Figure 1]. Electronic medical records were evaluated for patient demographics, endoscopic findings, details regarding the endovascular procedure, and clinical and imaging follow-up. Procedural complications were also recorded.

Figure 1:

57-year-old lady with decompensated chronic liver disease who presented with recurrent upper gastrointestinal bleeding. (a) Endoscopy image showing isolated gastric varices (arrow). (b) Contrast-enhanced computed tomography (CECT) axial image showing gastric varices (arrow). (c) CECT coronal image showing a gastro-renal shunt (arrow). (d) Balloon-occluded retrograde transvenous obliteration image showing an inflated balloon and sclerosant-contrast stasis in gastric varices (arrow). (e) Post-procedure dynamic computed tomography axial image showing good dispersion of contrast-sclerosant mixture in gastric varices (arrow). (f) Follow-up CECT axial image showing no residual varices (arrow).

Occlusion balloon characteristics

The Swan-Ganz catheter is a flow-directed balloon catheter initially designed to catheterize the pulmonary artery to monitor the hemodynamics of critically ill patients. It is made of latex material. The catheter is available in lengths ranging from 60 cm to 110 cm and calibers from 4 to 8 French (F). The diameter of the fully inflated balloon for the 7F size measures 13 mm. It necessitates an introducer sheath of at least 8F, and the average cost is around 5,000 Indian rupees (INR). The Fogarty thru-lumen embolectomy catheter was specifically designed for the removal of fresh, soft emboli and thrombi from vessels within the arterial system. Constructed from latex material, this catheter is available in lengths ranging from 40 to 80 cm and calibers of 3–7 F. The diameter of the fully inflated balloon (4F) measures 9 mm, and it requires an introducer sheath of at least 5F. The estimated cost of the catheter is around 5,000 INR. The stone extractor balloon is specifically designed for the endoscopic removal of stones from the bile duct. It was made of latex.

Balloon selection was primarily guided by the diameter and length of the gastro-renal shunt as assessed on pre-procedural CECT. Smaller shunts (<10 mm) were managed using 4–7F Swan-Ganz or Fogarty catheters, while larger shunts (>12–15 mm) were treated using stone extractor balloons capable of achieving wider occlusion diameters. Catheter length and shaft support were additionally considered to ensure stable positioning within tortuous shunts. This individualized selection strategy allowed effective occlusion while minimizing the risk of balloon migration or rupture.

Modified sclerosant protocol

The most commonly used sclerosant in BRTO is 3% w/v sodium tetradecyl sulfate (STS) (Samarth Life Sciences, Mumbai, India), which is often mixed with Lipiodol (Guerbet, Aulnay-sous-Bois, France) to provide radio-opacity during fluoroscopic guidance. However, an important technical limitation arises because many occlusion balloons are prone to rupture when exposed to Lipiodol, leading to premature deflation, escape of sclerosant into systemic circulation, and consequent risk of therapeutic failure. To overcome this issue in BRTO, we have adopted a modified sclerosant mixture consisting of STS, a water-soluble contrast medium (Iohexol 300 mg I/mL, Omnipaque, GE Healthcare, China), and air, thereby eliminating the need for lipiodol while maintaining adequate visibility under fluoroscopy. A 2:2:1 ratio of STS, air, and contrast was used in our study. A key contribution of this study is the adoption of a modified Lipiodol-free sclerosant protocol. Lipiodol, although traditionally used to improve radiopacity, is known to interact chemically with latex balloons, increasing the risk of balloon weakening and rupture. Because the Swan-Ganz, Fogarty, and Stone extractor balloons used in our study are latex-based, eliminating Lipiodol is likely to have reduced balloon compromise. Importantly, both technical failures in our series occurred in patients who received Lipiodol-containing mixtures. Following the transition to the Lipiodol-free sclerosant protocol, no further balloon ruptures were observed. This modification may therefore represent a safer approach when using latex-based, low-cost balloon catheters.

Technical success was defined as achieving good stasis of the sclerosant mixture in the gastric varices during the period of balloon inflation. Clinical success rate was defined as cessation of upper gastrointestinal bleed or hepatic encephalopathy due to gastric varices. A follow-up evaluation of the gastric varices was obtained with CECT before discharge. Patients were then evaluated clinically and with CECT and / or endoscopy at 3–6 month intervals [Figure 2]. Obliteration of the varices was defined as the absence of opacification of the gastric varices on follow-up CT and was classified as complete, partial, or failed obliteration.

Figure 2:

47-year-old gentleman with chronic liver disease who presented with recurrent upper gastrointestinal bleeding. (a) Endoscopy image showing isolated gastric varices (arrow). (b) Contrast-enhanced computed tomography (CECT) axial image showing gastric varices (arrow). (c) CECT coronal image showing a gastro-renal shunt (arrow). (d) Balloon-occluded retrograde transvenous obliteration image showing an inflated balloon and sclerosant-contrast stasis in gastric varices (arrow). (e) Follow-up CECT axial image showing obliteration of gastric varices (arrow). (f) Follow up endoscopy image showing no residual varices (arrow).

Statistical analysis

Statistical analysis was primarily descriptive in nature. Continuous variables were expressed as mean and range, while categorical variables were expressed as frequency and percentage. No comparative statistical analysis was performed due to the retrospective descriptive design and limited sample size. Descriptive statistics alone were performed.

RESULTS

The patient group included 20 men (64.5%) and 11 women (35.5%) with a mean age of 56.6 years (range 47–75 years). Causes of liver cirrhosis included alcohol (n = 14), hepatitis B (n = 2), non-alcoholic fatty liver disease (n = 6), and cryptogenic (n = 9). According to the modified Child-Pugh classification, 15 patients had class A disease, 9 had class B disease, and 7 had class C disease before the performance of BRTO. 25 patients (80.6%) had a history of gastric variceal hemorrhage within 4 weeks. In the remaining patients (n = 6, 19.4%), BRTO was performed to manage intractable hepatic encephalopathy. Baseline clinical characteristics of the patients were summarized in Table 1.

Table 1:

Baseline clinical characteristics (n=31).

Variable	Value
Demographics
Total patients	31
Mean age (years)	56.6 (range 47–75)
Male (%)	20 (64.5)
Female (%)	11 (35.5)
Etiology of cirrhosis (%)
Alcohol-related	14 (45.2)
Hepatitis B	2 (6.5)
Non-alcoholic fatty liver disease	6 (19.4)
Cryptogenic	9 (29)
Child-Pugh classification (%)
Class A	15 (48.4)
Class B	9 (29)
Class C	7 (22.6)
Indication for balloon-occluded retrograde transvenous obliteration (%)
Recent gastric variceal bleeding (<4 weeks)	25 (80.6)
Refractory hepatic encephalopathy	6 (19.4)

An occlusion balloon catheter was selected based on the size of the shunt. A 7F Swan-Ganz catheter was used in 27 patients (87%), a 4F Fogarty thru-lumen embolectomy catheter in one patient (3.2%), an 8F stone extractor balloon catheter in two patients (6.4%), and an Advance low-profile percutaneous transluminal angioplasty balloon dilatation catheter (Cook Inc., Bloomington, Indiana) in one patient (3.2%). The balloons were inflated by injection of a mixture of contrast media and saline solution sufficient to induce stasis within the shunt. A retrograde venogram was performed through the balloon catheter to demonstrate the gastric varices and efferent venous channels [Figure 3]. In 6 patients, the efferent veins were embolized with a combination of coils and Gelfoam to prevent leakage of the sclerosing agent into the systemic circulation.

Figure 3:

A 52-year-old lady with decompensated chronic liver disease who presented with recurrent upper gastrointestinal bleeding. (a) Contrast-enhanced computed tomography coronal image showing gastric varices (arrow). (b) Balloon occlusion venogram showing opacification of varices corresponding to the computed tomography image (arrow).

A 3% STS foam mixed with iodized oil and contrast in 17 patients (54.8%), 3% STS foam with contrast medium in 13 patients (41.9%), and 3% STS foam with contrast and gel foam in 1 patient (3.2%) under fluoroscopy.

The inflated balloon was maintained for 1–3 h in 27 patients, 25 h in one case, 16 h in one patient, 4 h in one case, 45 min in one case, 30 min in one case, and 10 min in one patient (balloon ruptured). The duration of balloon inflation depended on the size of the varices. In patients with up to 3 h of balloon inflation, the inflated balloon and the migration of the sclerosing agent into the varices were monitored with the use of fluoroscopy every 30 min throughout balloon inflation. Patients with overnight balloon inflation were monitored for 1 h with the use of fluoroscopy. Patients were then transferred to the ward until the next day, and the integrity of the inflated balloon and the migration of the sclerosing agent into the varices were only checked before balloon deflation. Procedure details and outcomes were summarized in Table 2.

Table 2:

Procedural details and clinical outcomes.

Variable	Value (%)
Balloon type used (n=31)
7F Swan-Ganz catheter	27 (87)
4F Fogarty catheter	1 (3.2)
Stone extractor balloon	2 (6.4)
PTA balloon	1 (3.2)
Sclerosant mixture
STS+Lipiodol+contrast	17 (54.8)
STS+contrast (Lipiodol-free)	13 (41.9)
STS+contrast+gelfoam	1 (3.2)
Balloon occlusion duration
1–3 h	27 (87)
<1 h	3 (9.6)
>12 h (overnight)	2 (6.4)
Technical success	29/31 (93.5)
Clinical success (available follow-up n=25)	23/25 (92)
Follow-up imaging/Endoscopy available among patients with clinical follow-up	17/25 (68)
Complete obliteration	12/17 (70.5)
Partial obliteration	5/17 (29.5)

PTA: Percutaneous transluminal angioplasty, STS: Sodium tetradecyl sulfate

Technical success was achieved in 29 of 31 (93.5%) patients. Among the two patients who had a technical failure, patient 1 had an inadvertent rupture of a Swan-Ganz balloon after 10 min of inflation. The sclerosant mixture used in this patient consisted of 12 mL of 3% STS foam and 3 mL of Lipiodol. However, the sclerosant was noted to be in place, probably due to the formation of an organized clot. The balloon was not removed so as not to disturb the soft clot. In patient 2, the Swan-Ganz balloon was found deflated after 16 h of contact with the sclerosant; however, it had not ruptured, which was confirmed after removal of the balloon. The sclerosant mixture used in this patient was 36 mL of 3% STS foam and 4 mL of Lipiodol. Neither of these patients developed further bleeding.

Clinical follow-up was available for 25 of 29 patients (86.2%) who had a technically successful procedure at a mean follow-up of 953.6 days (range, 1–4015 days) after the BRTO. Clinical success rate was achieved in 23 of 25 patients (92%). One patient developed a massive hematemesis in the immediate post-procedure period and was taken for surgical splenorenal shunt creation. Moreover, another patient had a recurrence of bleeding after a year and subsequently underwent a TIPS. None of the other patients had any further episode of variceal bleeding or encephalopathy-related admissions during a mean follow-up period of 678.9 days (range 1–4015 days).

Among these 25 patients with clinical follow-up, imaging or endoscopic follow-up was available in 17 patients (68%), demonstrating complete obliteration in 12 patients (70.5%) and partial obliteration in 5 patients (29.5%). Incomplete imaging or endoscopy follow-up may influence the interpretation of true obliteration rates and potentially underestimate late recurrences. The patient follow-up pathway and study cohorts are summarized in Figure 4.

Figure 4:

Patient selection and Follow-up flow chart.

DISCUSSION

Gastric variceal bleeding is one of the major complications of portal hypertension, and it is associated with a higher mortality rate as compared to esophageal variceal bleeding.^[9-11] Gastric varices are also difficult to treat because of their large size, difficult location, and substantial blood flow. There are diverse treatment modalities, including pharmacotherapy, balloon tamponade, endoscopic procedures, endovascular treatment, and surgery.^[3,6]

TIPS has been widely used to manage refractory variceal bleeding with a high rate of initial success. However, gastric varices tend to have a higher rebleeding rate despite lower portosystemic gradients, likely related to their overall larger diameter. Moreover, the presence of substantial spontaneous gastro-renal shunts reduces the portosystemic gradient by decompressing the portal system.^[12] In contrast, BRTO can directly obliterate these shunts, regardless of the portosystemic shunts. Furthermore, BRTO may be particularly beneficial for patients who are unable to tolerate TIPS due to its low-pressure gradient complications, such as hepatic encephalopathy or liver damage.

Compared with TIPS, BRTO provides distinct hemodynamic advantages in patients with large spontaneous gastro-renal shunts. While TIPS reduces portal pressure, it does not directly obliterate gastric varices and may worsen hepatic encephalopathy by increasing systemic shunting. Several studies have reported higher post-procedure encephalopathy rates following TIPS compared with BRTO, whereas BRTO has been associated with improved encephalopathy outcomes due to shunt occlusion.^[13] In addition, BRTO avoids the need for long-term stent surveillance and may be more cost-effective in resource-limited settings. However, TIPS remains preferable in patients requiring global portal decompression or those with concurrent refractory ascites.

This retrospective study indicates that BRTO using off-label, low-cost balloon catheters is a technically feasible and clinically effective treatment for managing gastric varices and hepatic encephalopathy in patients with portal hypertension. Our technical success rate of 93.5% and clinical success rate of 92% are consistent with those reported in larger series using traditional expensive balloon catheters.^[14-16]

Our study was able to achieve this high rate of technical success, despite the use of repurposed balloon catheters such as the Swan-Ganz and Fogarty catheters. It highlights the potential for cost-effective implementation of BRTO in resource-constrained settings. In contrast to the standard BRTO practice in Western and Japanese centers, where dedicated reverse-curve or high-pressure balloons are utilized,^[4,8] our use of widely available catheters significantly differs from convention. Dedicated BRTO balloons cost approximately INR 50,000, compared with approximately INR 5,000 for Swan-ganz or Fograty catheters, representing an estimated 90 % cost reduction. These low-cost options, costing roughly one-tenth of standard BRTO balloons, allowed for procedural adaptability without compromising outcomes, particularly when appropriate sizing and technique were employed.

The modified sclerosant protocol may also have been responsible for reducing balloon rupture, which is a possible complication when Lipiodol is used due to its chemical interaction with latex balloons.^[17] In our study, only two cases of balloon compromise occurred, and neither led to clinical failure or rebleeding. This suggests that even in cases of balloon failure, if the sclerosant achieves variceal stasis, therapeutic benefit may still be achieved, possibly due to rapid clot formation or mechanical occlusion of the varices. Both cases of balloon compromise occurred in patients receiving Lipiodol-containing sclerosant mixtures. No technical failures occurred after adopting the Lipiodol-free protocol, suggesting that chemical interaction rather than operator technique was the primary contributing factor.

Balloon occlusion times varied considerably in this study (10 min–25 h), reflecting procedural evolution over the study period and operator-dependent decision-making. Longer occlusion times were initially adopted to ensure complete thrombosis of large varices. However, subsequent experience demonstrated that effective sclerosant stasis could be achieved within 1–3 h in most cases. The variability was not systematically influenced by balloon type but was occasionally related to variceal size and sclerosant volume. Importantly, no clear association was observed between prolonged occlusion and improved obliteration rates.

The durability of BRTO was confirmed through long-term follow-up, with a mean of 953.6 days. Only two patients experienced rebleeding, one in the immediate post-procedure period, requiring surgical shunting, and another after a year, who subsequently underwent a TIPS procedure. Notably, none of the other patients developed recurrent gastric variceal bleeding, affirming the long-term efficacy of BRTO, as also reported in other studies.^[7,18-20] In addition, the inclusion of six patients undergoing BRTO for refractory hepatic encephalopathy demonstrates the versatility of this technique beyond hemorrhagic indications. Previous studies have shown that BRTO can significantly reduce portosystemic shunting and ammonia levels, leading to clinical improvement in encephalopathy symptoms.^[20,21] Although ammonia levels were not routinely monitored in this study, the absence of further encephalopathy-related admissions in these patients supports the therapeutic efficacy of BRTO in this context.

There were several limitations to the study. First, our study was a retrospective single-center study, which may introduce selection and information bias due to reliance on previously collected data. In addition, the small sample size limits the generalizability of our findings and may affect the statistical power of the results. In addition, the absence of a comparator or control group using dedicated BRTO balloons limits direct comparison of outcomes. Operator-dependent decision-making and increasing procedural experience over time could additionally have contributed to variability in technique and follow-up practices. Finally follow up imaging or endoscopy was available in only 68 %. of patients, potentially underestimating the true obliteration rate or missing late recurrences. Procedural heterogeneity in balloon type, sclerosant composition, and occlusion times introduces procedural variability that may influence outcomes. Yet, these modifications are necessary in resource-limited settings and do not compromise efficacy.

CONCLUSION

Our experience supports the safe and effective use of off-label, low-cost balloon catheters for BRTO in the management of gastric varices and hepatic encephalopathy. When combined with a modified, Lipiodol-free sclerosant protocol, these approaches offer a cost-effective alternative without compromising outcomes. More prospective studies are needed to validate these findings and standardize protocols for the generalization of these techniques.

Funding Statement

Financial support and sponsorship: Nil.

Ethical approval:

The research/study was approved by the Institutional Review Board at Christian Medical College, Vellore, with the number IRB Min. No. 16203, dated March 20, 2024.

Declaration of patient consent:

Patient’s consent not required as the patient’s identity is not disclosed or compromised.

Conflict of interest:

There are no conflicts of interest.

Use of artificial intelligence (AI)-assisted technology for manuscript preparation:

The authors confirm that there was no use of artificial intelligence (AI)-assisted technology for assisting in the writing or editing of the manuscript, and no images were manipulated using AI.